Spatialising Peace and Conflict

Mapping the Production of Places, Sites and Scales of Violence

Annika Bjorkdahl, Susanne Buckley-Zistel, Annika Bjorkdahl, Susanne Buckley-Zistel

• Publisher: Palgrave Macmillan
• Year: 2016
• ISBN:
  9781137550484
• Topic: Politik & Internationale Beziehungen
• Subtopic: Vergleichende Politikwissenschaften

This volume brings to the fore the spatial dimension of specific places and sites, and assesses how they condition – and are conditioned by – conflict and peace processes. By marrying spatial theories with theories of peace and conflict, the contributors propose a new research agenda to investigate where peace and conflict take place.

Part I

Territorialities and Scales

1

Overcoming the Material/Social Divide: Conflict Studies from the Perspective of Spatial Theory

Sven Chojnacki and Bettina Engels

Introduction

In conflict studies it has become a trend to link space and conflict theoretically as well as empirically (see e.g. Raleigh et al. 2010; Stephenne et al. 2009; Wucherpfennig et al. 2011). So far, however, few substantial arguments have been made that consider how the relationship between space and conflict can enhance our understanding of conflicts. One challenge is that the landscape of theoretical narratives and empirical studies is rather fragmented along the line of different disciplines. Social geographers, political scientists, conflict researchers and development scholars alike suggest that we broaden our understanding of space in order to reflect on the conditions, dynamics and effects of conflict more precisely, without much engagement between these disciplines. One central question discussed in these different bodies of literature is whether space is an external, material condition that influences human action, or whether a different understanding of space is required.

In this chapter we critically review the corresponding debates in conflict studies and examine how they conceptualize spatial categories, both theoretically and empirically. We begin with an overview of how spatial references are made in the mainstream debates in conflict studies, then introduce theoretical reflections of the concept of space in social geography. Next we link the two debates in order to outline how spatial categories can be applied in the analysis of violent conflict in a more productive way. Although the main objective of the chapter is to introduce the subject and to present the state of the art, it concludes by offering a perspective on conflict as social action that is constitutively related to space, and suggesting that space does not determine conflict but that space and conflict mutually produce and reproduce each other.

In order to unfold this argument the chapter proceeds as follows. First, we review central debates within conflict studies with regard to how they refer to spatial concepts. We begin with studies on international and civil wars that, since the 1970s, have used geographic categories – such as topography, infrastructure, borders or proximity – as explanatory factors for the occurrence of armed conflict. In the study of civil war, natural resources and territorial control are particularly considered. Next the debate on environmental conflicts – that is, research on the relationship between ecological change and violent conflict – is sketched out as another influential strand within conflict studies, which refers strongly to spatial aspects. Following this we show that most studies of international and civil war and environmental conflict reproduce the container concept of the territorially bounded nation-state.

The main problem we identify in conflict studies literature is thus that most studies do not connect physical and material with sociopolitical aspects of collective conflict, therefore perpetuating what has been referred to as methodological nationalism. Wimmer and Glick Schiller (2002) define methodological nationalism as ‘the assumption that the nation/state/society is the natural social and political form of the modern world’ (Wimmer and Glick Schiller 2002: 302). While their concern is with outlining how the epistemic programmes and structures of mainstream social science research are interwoven with modern (capitalist) nation-state formation, in this chapter we confine our critique to one of their three variants of methodological nationalism: the focus on the nation-state as the key, and often only, unit of analysis.

After the literature review, we show how concepts of spatial theory developed in radical geography help to redress these problems. The notions of territorialization and scale are taken as examples to elucidate this in more detail. We conclude by arguing that spatial theory provides helpful analytical categories for conflict studies, but that space does not simply present an additional variable to be added to the analysis of violent conflict. The point is, rather, to examine how conflict can be analysed from the perspective of a spatial theory, and how spatial and conflict theory can enter into a productive dialogue.

Space in the analysis of international and civil wars

As early as the 1970s, research on international war conceptualized geographic conditions (e.g. direct neighbourhood and spatial distance) as explanatory factors for the occurrence of armed conflict and war (see Diehl 1991; Starr 1991). Ever since, physical features such as borders or contiguity have been identified as enabling interaction between territorial entities, or as providing opportunity structures under which actors formulate preferences and make decisions (Siverson and Starr 1991). Quantitative studies on the causes of interstate wars emphasize the importance of territorial aspects (notably boundary disputes) as the most conflicting among all contentious issues (Vasquez 2009). Here, both the understanding of geography and the twin concept of opportunity and willingness (Most and Starr 1980) are oriented towards the relatively static notion of boundaries as central to methodological nationalism.

At present, most quantitative empirical studies of the field operate in a territorial container, though, when focusing on correlations between physical variables such as topography, infrastructure, borders or proximity and conflict. In other words, most studies are based on an understanding of geography that restricts space to allegedly fixed material factors, such as the availability of resources or physical demarcation (bordering) in order to explain and predict armed conflict. One central argument holds that the dynamics of armed conflict are conditioned by the location of and distance to political, natural or other resources – for instance, when arguing that in strong regimes, civil wars are located further away from the capital (Buhaug 2010), or that conflicts last longer if they are located along remote international borders, in regions with valuable minerals or at a distance from the main government (Buhaug et al. 2009).

Other studies show that the existence, concentration and type of natural resources impacts on the occurrence of civil war. In this reading, diamonds and oil have highly significant effects, while agricultural goods are hardly significant at all (Fearon 2005; Lujala et al. 2005; Ross 2004). Moreover, centralized resources, such as petroleum or easily accessible mines, are considerably easier to monitor and protect than geographically dispersed resources, such as opium plantations, alluvial diamonds or tropical forests (Le Billon 2001). One critical aspect is also the proximity of key resources to a fighting faction’s headquarters or the capital. Buhaug and Rød (2006) demonstrate furthermore that armed conflict correlates with the spatial distribution of features such as relative road density, while O’Loughlin et al. (2011) found evidence that the proximity to strategic locations (military installations, administrative institutions) affects the incidences and diffusion patterns of violence over time.

Most of these studies have in common that they rarely question the relationship between conflict and external resources as a material condition of conflict. But material features only become resources through social relations, and it is these social relations that explain why a certain resource enables or disables certain conflict action. Even though places and territories have locatable and measurable properties, the meaning and impact such conditions have for individuals and social groups are contingent on time and space. The relationship of space and conflict hence revolves around the intangible and dynamic qualities that are attributed to these conditions by social groups or individuals (e.g. notions of ownership, ideas of cultural identity¹ that are connected to places and territories). These social and political meanings also have to be reflected in approaches to conflict studies, which draw on topography, borders and resources.

Studies on civil war link conflict to the importance of territory for maintaining order, emphasizing the significance of territorial control for the political and strategic relations between violent actors and the civilian population. Using a microlevel approach, Kalyvas (2006, 2012) concludes that the level of territorial control by armed actors, as well as their desire to minimize information asymmetries, allows for predicting the spatial variation of violence against civilians in civil wars. Depending on the degree of control (reaching from areas fully controlled by one actor to areas contested by armed groups), violence becomes a function of different zones of territorial control. Kalyvas thus offers a pioneering perspective on the interconnection of information, territorial control and types of violent action. Still, his model remains within the logic of territory as a physical condition and fails to consider its cultural meanings or its political and social power relations.

Control over territory is a precondition for the extraction of natural resources. Access to these resources is directly related to political power, as well as to the opportunity and incentive to wage war. The value of territorial control, however, varies with the economic and strategic importance of specific areas. For state actors and armed groups alike, both territories with valuable resources such as diamonds, gold or oil, and those which possess an inherent value for strategic action, such as capitals, harbours and transport routes, are more important than a piece of land in the periphery without major resources.

Environmental conflict

Another influential strand in conflict research that refers to spatial categories is the study of environmental conflicts. Since the early 1990s, a debate has emerged about whether and how environmental change influences the occurrence of violent conflict on different levels (local, national and international) (Baechler et al. 2002; Gleditsch 1997; Homer-Dixon 1999). The fourth assessment report of the Intergovernmental Panel on Climate Change, Climate Change 2007 (IPCC 2007) was circulated widely in politics, media and academia. In the following years, scholars focused on the question of how climate change effects, such as rising temperature, increasing rainfall variability and extreme weather, affect violent conflict. By now, most scholars agree that ecological change becomes relevant for collective conflict only through social and political mediation (Gleditsch 1998; Salehyan 2008), although some recent studies still confirm the argument that rainfall patterns and violent conflict correlate (Raleigh and Kniveton 2012; Theisen 2012). Remarkable progress has been made with regard to empirical evidence, as well as to the specification of social and political factors that mediate the relationship between environmental change and violent conflict (Benjaminsen and Boubacar 2009; Benjaminsen et al. 2012). Some weaknesses remain, however, in particular regarding the persisting focus on the nation-state as the most important level of analysis.

Regarding environmental conflict research, few studies explicitly refer to a theoretical concept of space. Changes in the physical-material environment, as well as violent conflict, come across in different ways across time and space. Also, violence is a socially differentiated phenomenon, meaning that at a specific location at a particular moment in time, not all people are affected by collective violence in similar ways. Likewise, the social effects of ecological change vary horizontally and vertically, as numerous studies have shown (see Adger 1999, 2006; Wisner et al. 1994). Where people live (in terms of locations and territories) affects their vulnerability, but it only constitutes one aspect among several and is related to other conditions, such as social structures.

Thus the analysis of the spatial dimension of environmental conflict that exclusively refers to physical and material features and detaches them from the social structures in which they are embedded – by presenting certain areas such as river deltas, coastal zones and savannah regions as being particularly prone to ecological change effects and conflict risks, for instance – fails to deal adequately with the complexity of ecological and social systems. However, an a priori denial of the relevance of any physical materiality for violent conflict falls equally short. Existing models that try to integrate environmental and sociopolitical factors often assume a linear, causal relationship between ecological change and conflict, regarding social and political institutions as intervening variables (Homer-Dixon 1999).

Persisting containers: Analytical levels and scales

Recent research on civil wars and the study of environmental conflict have attempted to overcome the container concept of the nation-state (Buhaug and Gates 2002; Buhaug and Rød 2006). Nevertheless, many studies, in one way or another, still conceptualize the state as a territorial container. Even if complex disaggregated geographical data are used, the state is simply replaced by arbitrary grid cells as the central units of analysis. Such a shift in the level of analysis has yet to fully overcome methodological nationalism without merely replacing it by ‘methodological territorialism’ in the sense of assuming smaller, but still territorially bounded, entities. While the construction of grid cells does not reproduce the arbitrary demarcations of nation-states, it creates new containers and equally arbitrary boundaries, not taking into account any social, political or cultural meaning of territories and places. Furthermore, by deliberately disregarding nation-state borders, grid cells analysis is unable to tackle the historical and political meaning these borders have for local conflicts.

Still, recent studies using disaggregated geographical data shed some light on the occurrence of civil wars: they show, for instance, that topographic variables such as forests and mountains affect the manner in which internal violent conflicts are conducted and how they impact on the prospect of winning a battle or a war (Buhaug and Gates 2002; Buhaug and Rød 2006; Gates 2002). The analytical problems related to nation-state centrism are, however, not resolved solely by simply ‘downscaling’ the level of analysis (Agnew 1994). Rather, the potentials of integrating structural and geometric vector data² methodologically are only fully beneficial if they are systematically linked to theoretical reflections on both space and conflict.

The analytical bias in favour of the nation-state level is reflected in qualitative as well as quantitative studies (Deligiannis 2012). Even though in recent years an increasing number of case studies on civil wars and environmental conflicts have focused on the substate (mainly local) level, they hardly investigate the interscalar relationships between the local, natio...