Good Fences: The Importance of Setting Boundaries for Peaceful Coexistence

We consider the conditions of peace and violence among ethnic groups, testing a theory designed to predict the locations of violence and interventions that can promote peace. Characterizing the model’s success in predicting peace requires examples where peace prevails despite diversity. Switzerland is recognized as a country of peace, stability and prosperity. This is surprising because of its linguistic and religious diversity that in other parts of the world lead to conflict and violence. Here we analyze how peaceful stability is maintained. Our analysis shows that peace does not depend on integrated coexistence, but rather on well defined topographical and political boundaries separating groups, allowing for partial autonomy within a single country. In Switzerland, mountains and lakes are an important part of the boundaries between sharply defined linguistic areas. Political canton and circle (sub-canton) boundaries often separate religious groups. Where such boundaries do not appear to be sufficient, we find that specific aspects of the population distribution guarantee either sufficient separation or sufficient mixing to inhibit intergroup violence according to the quantitative theory of conflict. In exactly one region, a porous mountain range does not adequately separate linguistic groups and that region has experienced [amazon asin=0765808684&template=*lrc ad (left)]significant violent conflict, leading to the recent creation of the canton of Jura. Our analysis supports the hypothesis that violence between groups can be inhibited by physical and political boundaries. A similar analysis of the area of the former Yugoslavia shows that during widespread ethnic violence existing political boundaries did not coincide with the boundaries of distinct groups, but peace prevailed in specific areas where they did coincide. The success of peace in Switzerland may serve as a model to resolve conflict in other ethnically diverse countries and regions of the world.

Introduction

Efforts to resolve conflicts and achieve sustained peace are guided by perspectives about how conflict and peace are based in interpersonal and intergroup relationships, as well as historical, social, economic and political contexts. Recently, we introduced a complex systems theory of ethnic conflict that describes conflicts in areas of the former Yugoslavia and India with high accuracy [1]. In this theory, details of history, and social and economic conditions are not the primary determinants of peace or conflict. Instead the geographic arrangement of populations is key. Significantly, our theory points to two distinct conditions that are conducive to peace – [amazon asin=1478344687&template=*lrc ad (right)]well mixed and well separated. The first corresponds to the most commonly striven for peaceful framework: a well integrated society [2]. The second corresponds to spatial separation, partition and self determination – a historically used but often reviled approach [3]. Here we consider a more subtle third approach, that of within-state boundaries in which intergroup cooperation and autonomy are both present. The success of this approach is of particular importance as the world becomes more connected through international cooperation. As illustrated by the European Union, the role of borders as boundaries is changing.

In order to evaluate the role of within-state boundaries in peace, we considered the coexistence of groups in Switzerland. Switzerland is known as a country of great stability, without major internal conflict despite multiple languages and religions [4][5]. Switzerland is not a well-mixed society, it is heterogeneous geographically in both language and religion (Fig. 1). The alpine topography and the federal system of strong cantons have been noted as relevant to coexistence; their importance can be seen in Napoleon’s statement, after the failure of his centralized Helvetic Republic, that “nature” had made Switzerland a federation [6][8]. But the existence of both alpine and non-alpine boundaries between groups and the presence of multiple languages and religions within individual cantons suggest partition is not essential for peaceful coexistence in Switzerland. In identifying the causes of peace, the literature has focused on socio-economic and political conditions. These include: a long tradition of mediation and accommodation; social cleavages that “cross-cut” the population rather than coincide with each other; unwritten and written rights of proportionality (fairness) and cultural protectionism; a federal system with strong sub-national units; a civil society that fosters unity; direct democracy through frequent referenda; small size; historical time difference between cleavages in language and religion; neutrality in international warfare; and economic prosperity [4][6],[9][13]. Geography plays an unclear, presumably supporting, role in these frameworks. The analysis of coexistence in Switzerland is also part of a broader debate about whether social and geographical aspects of federalism promote peace or conflict [15].

thumbnail

Figure 1. Maps of Switzerland showing the 2000 census proportion of (A) linguistic groups, (B) Catholics and Protestants (Mercator projection).

[amazon asin=094546620X&template=*lrc ad (left)]In this paper we analyze the geographical distribution of groups in Switzerland based solely upon the hypothesis that spatial patterns formed by ethnic groups are predictive of unrest and violence among the groups [1]. The model also allows that topographic or political boundaries may serve as separations to promote peace [1][16]. We test the ability of the theory to predict peaceful coexistence in the context of internal country boundaries in Switzerland. Where explicit boundaries do not exist, such as in mixed cantons where alpine boundaries are absent, violence might be expected, and the results of the model in these areas serve as a particularly stringent test of the theory. In most such cases violence is not predicted, consistent with what is found. In one area a significant level of violence is predicted, and in fact violence is actually observed. The analysis sheds light on the example of Switzerland as a model for peaceful coexistence. The former Yugoslavia serves as a contrasting example of widespread violence. The theory also correctly identifies areas of conflict and areas of peace in the former Yugoslavia. The precision of the results provides some assurance of the usefulness of the theory in planning interventions that might promote peace in many areas of the world.

Geographical Distribution Theory[amazon asin=B001D0MPYK&template=*lrc ad (right)]

The geographical distribution theory [1] is independent of the specification of the individual types – consistent with a universality of type behavior. Violence arises due to the structure of boundaries between groups rather than as a result of inherent conflicts between the groups themselves. Even though diverse social and economic factors trigger violence, it occurs when the spatial population structure creates a propensity to conflict, so that spatial heterogeneity itself is predictive of local violence. The local ethnic patch size serves as an “order parameter”, a measure of the degree of order of collective action, to which other aspects of behavior are coupled. The importance of collective behavior implies that ethnic violence can be studied in the universal context of collective dynamics, where models can identify how individual and collective behavior are related.

As we consider it here the analysis is applicable to communal violence and not to criminal activity or international warfare. In highly mixed regions, groups are not large enough to develop strong collective identities, or to identify public spaces as associated with one or another cultural group. They don’t impose upon and are not perceived as a threat to the cultural values or social/political self-determination of other groups. At the other [amazon asin=B00KN0K6EM&template=*lrc ad (left)]extreme, when groups are larger than the critical size, they typically form self-sufficient entities that enjoy local sovereignty. Partial separation with poorly defined boundaries fosters conflict. Violence arises when groups are of a geographical size that they are able to impose cultural norms on public spaces, but where there are still intermittent violations of these rules due to the overlap of cultural domains. Hence, we expect violence to arise when groups of a certain characteristic size exist. The model depends on population geography and not on the mechanisms by which the population structure arose, which may include individual choice as well as internally or externally directed migrations. The use of population geography, determined by census, to predict violence may work well because geography is an important aspect of the dimensions of social space, and other aspects of social behavior (e.g., isolationism, conformity, as well as violence) are correlated to it.

Physical boundaries such as mountain ranges and lakes or national and subnational political boundaries that establish local autonomy may prevent the violations of cultural norms that cause friction between groups and promote self-determination, inhibiting the triggers of violence. By creating autonomous domains of activity and authority, the boundaries shield groups of the characteristic size from each other when they correspond with their geographical domains.

Mathematically (see Methods), evaluation of the model begins by mapping census data onto a spatial grid. In this work we included the fraction of every population type on each site. The expected violence is determined by [amazon asin=B005S28ZES&template=*lrc ad (right)]detecting patches consisting of islands or peninsulas of one type surrounded by populations of other types. These features are detected by pattern recognition using the correlation of the population for each population type with a template (filter) that has a positive center and a negative surround. The template used is based on a wavelet filter [1][17][18]. The wavelet filter is a conventional filter for identifying the geographic size of spatial regions. It corresponds also to the on-center off-surround detection elements of spatial size of regions in the human retina. Other methods that identify the size of groups provide the same results. We also find that the results are insensitive to the specific values of the parameters and therefore the analysis is highly robust to variations in the methodology and its parameters. The diameter of the positive region of the wavelet, i.e., the size of the local population patches that are likely to experience violence, is the only essential parameter of the model. The parameter is to be determined by agreement of the model with reports of violence, and results were robust to varying the parameter across a wide range of values. To model the effect of boundaries, we include only the populations within each of the autonomous areas to determine the expected violence. Where boundaries are incomplete, as might be the case for mountains, lakes and convoluted political boundaries, we include only the populations that are in line of sight through gaps or past ends of boundaries to determine the expected violence within a region. For each location, populations past boundaries of the line of sight are replaced by neutral populations. The result of the correlation of population with the wavelet filter is a single value at each location, the theoretical “propensity to violence”, and the locations of expected violence are obtained by applying a threshold to that value. The location of groups of a certain size is indicative of a violence-prone area, but the precise location of violence is not determined. We construct maps of the proximity of every location to the identified violence-prone groups, and the proximity of each location to reported violence. The correlation of these maps tests the ability of the theory to predict violence prone and peaceful regions. The model was validated without boundaries [1] by applying it to the former Yugoslavia, yielding correlations of up to 0.89. The results were robust to varying the characteristic length between 18–60 km, and thresholds in the range 0.2–0.4. Our revised method with fractional population values on every site gave similar results with correlations of up to 0.87.

Results

Switzerland

We now consider the linguistic and religious groups in Switzerland, each in turn. Initial analyses and the historical sequence of boundary formation suggested considering topographical barriers when discussing language groups, and political barriers when considering religious groups. The geography of languages primarily reflects the extent of invasions prior to the existence of current political boundaries and has remained stable in most areas for over a thousand years [5]. The modern state was established afterwards, and religious conflict played a role in establishing the internal political boundaries [5][7]. Census data were obtained for 2634 municipalities (communes) in Switzerland (bfs.admin.ch), yielding a high spatial resolution.

Language and topographical barriers.

We study the three main language groups – German, French and Italian (Fig. 2A) – which together comprise 91% of the total population in the 2000 census (Romansh, the fourth official language, accounts for less than 2%). We considered only the effect of physical boundaries due to lakes and mountain ranges (Fig. 2, B and C). We determined the presence of topographical boundaries using an edge detection algorithm on topographical heights (Fig. 2D). This process identifies where there is a sharp change in height, i.e., a cliff, or steep incline, that runs for a significant distance forming a natural boundary. Elevation data with a spatial resolution of approximately 91 m [19] was coarsened to pixels of size  km. Edges were identified where there was an increase of more than 1.8 km in height over a distance of 9.1 km (11.5°) using a discretized Laplacian differential operator [20] with a mask size of a single pixel. The conclusions are robust to variations in the elevation angle (Methods). Calculations of the propensity to violence are shown in Fig. 2, E and F for the characteristic length of 24 km. Without boundaries, the correlation of the wavelet filter yields a maximum propensity to violence value of 0.48. With topographical boundaries the maximum propensity is reduced to 0.33. Results across filter lengths in the range 24–56 km, shown in Fig. 3 (maps in the methods section), consistently show that the propensity for violence is high for calculations without topographical boundaries and is dramatically reduced by their inclusion. Between the German and French-speaking areas the Jura mountain range and Lake Neuchatel serve as mitigating boundaries in the northwest, and the Bernese Alps are mitigating boundaries in the south. The interface between Lake Neuchatel and the Bernese Alps through the canton of Fribourg has no mitigating boundary, but is almost straight – neither side is surrounded by the other, so the propensity is not high. Between the Italian and German-speaking areas, the Lepontine Alps dramatically reduce the calculated propensity.

Read the rest of the article