1Reference to the past stands out in the landscape of causes of communitarian conflict in Belgium. In order to develop frameworks explaining the roots of communitarian discord, studies often examine the temporal depth of relations between linguistic communities, usually identifying phases of evolution. Alternatively, such studies recount the episodes endured by the Dutch-speaking population under the domination of the francophone ruling classes. The latter subjected their Dutch-speaking counterparts to numerous linguistic annoyances and harassment, which would later appear as the drivers of demands for autonomy.  This approach from the angle of wrongs suffered seeks to shed light on the political instability and linguistic conflicts via the Flemish collective memory – a hotbed for ethno-nationalist movements – which is understood to contain the causal framework explaining these phenomena. The memory dynamic, an ever-vivid representation of the past which is understood to explain the persistence of Flemish demands, typically operates through “constant causes”. Coined by Arthur Stinchcombe, this expression refers to social phenomena that are reproduced over time through the very causes that generated them initially. 
2Therefore, through a process of symbolic memory reactivation by political elites, the set of causes that gave rise to the initial demands for independence continues to fuel identical demands in similar fashion. This type of explanation nevertheless does not exclude a different understanding of the contentious phenomenon: the “historical causes” hypothesis. Identified by Stinchcombe as the alternative path to “constant causes”, it allows for the reproduction of a sociopolitical phenomenon, even in the absence of the causes initially responsible for its emergence. Although this phenomenon originally came into being via a system of causes that is no longer in existence, autonomous processes continue to ensure its sustainability despite the absence of the circumstances at the root of its development.
3The “historical causes” analysis remains clearly distinct from works emphasizing distant causes of Belgian communitarian dispute. While these works have contextualized the genesis of relations between French and Flemish speakers and analyzed the successive historical phases of their development, they have neither theorized the course taken nor specified the processes through which the Belgian past influences its contentious present. This article proposes to identify these processes with reference to an analytical framework, path dependence, which is based upon a historical causality approach.  Path dependence theory posits that the choices made in the past place historical development on a set path towards a given outcome, with little possibility of reversal or deviation. This theoretical approach is adopted here in order to explain the recurring conflicts and the model of increasing autonomy which have characterized Belgian political life since the end of the nineteenth century.
4In 1830, French was the only official language in force in Belgium, used throughout the entire central administration, legal system, secondary and higher education system, and the military. Despite never having been the native language of the majority of the population, French experienced unprecedented dissemination from the eighteenth century, leading to the francization of the upper classes in several regions of Europe, most particularly in the southern Netherlands. When Belgium was created, the political and economic elite, the embodiment of the grande bourgeoisie, was French-speaking, including in Flanders. Against this background a movement comprised of proponents of the Flemish language very quickly formed, seeking to build upon the respect for traditions to push for the use of idiomatic expressions and encouraging the development of a national Flemish literature.  The mobilization of the Flemish movement against the French-speaking elite who held political power via censitary suffrage thus became early on one of the structuring cleavages of national political life, or, in Lipset and Rokkan’s terms, the cleavage between the centre and the periphery. Another cleavage took shape in parallel: one opposing the secular and clerical, which would find its partisan expression in the liberal and socialist parties on one hand, and the Catholic party on the other.
5The demands that arose in the late-nineteenth-century landscape of political and social mobilization were not solely linguistic. Industrialization brought with it a new working class, whose working conditions in the coal mines and metallurgical industry were particularly precarious. From the 1870s, socialist unions attempted to mobilize workers of different trades around the goal of improving standards of living for the working class. Their Catholic counterparts subsequently began to create workers’ organizations, after perceiving a “socialist threat”.
6The central argument here is that the choices operated by socialist and Catholic actors between 1885 and 1910 – the period during which workers’ movements were formed – would play a determining role with regard to how the structure of political relations within Belgium evolved. Before this period, these relations were characterized by the lack of a major linguistic disagreement which would have permitted clear identification of two opposing social aggregates. Broadly speaking, towards the end of the nineteenth century, a break-point occurred within this relationship: against expectations within a society organized primarily around differences of class or political ideology, it was the linguistic marker which became the distinctive element defining new actors, and the focal point for their conflicts and demands. As a result of the heightened politicization of linguistic identities, the idea of separate communities became established as an unavoidable political order. This led both to greater awareness of a common identity, and to the transformation of the linguistic community’s role from a cultural one to a political one. This article contends that the choice made by the community elites concerning the restructuring of the political space, which was made necessary by the politicization of identities, set Belgium on a path which would lead to the appearance of a new institutional system (federalism) coupled with a dynamic of continuous change (increasing autonomy) in the second half of the twentieth century. These models impacted the formulation of actors’ preferences, in the sense that conflicts continued and grew, while the consociational system, to a certain extent, failed.
7More specifically, the socialists – the pioneers of initiatives to organize the working class – did not consider or seize the opportunity for an alliance with the Flemish movement, which was centering its struggle on the improvement of the status of the Flemish language. The lack of attention given to the linguistic issue impeded the expansion of socialism among Flemish workers, while the adoption of these demands by the Catholics translated into a stronger political and social foothold in Flanders. The real alternative – which might have opened the door to other configurations of alliances – for the socialist and Flemish movements to create an alliance, would have tempered the reach of the Catholic church within the Flemish-speaking communities, and of the socialists within the French-speaking community. Yet the choices made in the end had the opposite effect, of associating a specific community with both the Catholic and socialist forces.
8The formation of the socialist and Catholic workers’ movements resulted in a decisive structuring of Belgian political life: the establishment of two sociopolitical categories in the Flemish and Walloon provinces that would constitute the majority for many years to come.
9One was socialist and francophone, the other Catholic and Flemish. From the beginning, their establishment was accompanied by the beginnings of a process of “social closure”, fixing the boundaries of inclusion and exclusion  and thereby placing Belgium on a trajectory towards community differentiation. This gradually drove the country towards a model of increasing autonomy in which communitarian conflicts appear as the inevitable phenomena of a historical process following its own logic.
10I outline below a path dependence model to explore the long-term effects of this relatively contingent identity configuration. The objective of this model is to shed light on the role of the past in communitarian conflicts. The subsequent section will directly address the application of this explanatory framework to a case study that indicates that the model of increasing autonomy characterizing Belgian political life is the result of a paradoxical historical progression. Its roots lie in the emergence of two sociopolitical ensembles referred to here as “categories”. This choice of this label, rather than that of “group”, rests upon an understanding of the identity phenomenon as a contingent event that, according to Brubaker, “may not happen”, revealing “the possibility that groupness may fail to crystallize, despite the group-making efforts of ethnopolitical entrepreneurs, and even in situations of intense elite-level ethnopolitical conflict”.  The emergence of the two said categories in a context of contingency permits the rejection of the culturalist approach to communitarian conflict, by indicating that their appearance is not the inevitable result of an initial cultural order. Rather, it shows that the very real linguistic segmentation in place since Belgium’s birth in 1830 became more complex with the intervention of non-cultural factors which lay within the expression of political, socialist, Flemish, and Catholic actors’ ideologies at the turn of the twentieth century. The entanglement of these ideologies with linguistic segmentation during this period generated a significant structural order which was not only the expression of cultural difference. It is paradoxical, then, that this historical progression, which emerged from a contingent dynamic of identity creation, proceeded with clear determinism towards a model of increasing autonomy.
The path dependence model
11The concept of path dependence provides a means of analytically approaching the phenomena of irreversibility and permanence of certain political choices. “Path dependence” suggests that, “once a country or region has started down a track, the costs of reversal are very high. There will be other choice points, but the entrenchments of certain institutional arrangements obstruct an easy reversal of the initial choice”. 
12This remark suggests that the choices made by actors during certain periods in the past play a decisive role, because they create changes in direction which make future alternatives less likely, thus setting long-term trajectories of development. Without limiting ourselves to the vague idea that history is integral to understanding sociopolitical realities, path dependence offers a specific explanation, formalized in political science in the area of public policy and historical sociology. The accepted research technique uses diachronic causality to retrace a series of interconnected sequences which lead from an interesting result, the causes of which we hope to identify, back to the sometimes unsuspected historical crossroads that generated the process that gave rise to the phenomenon. Certain political phenomena are only explained by the prior appearance of an event that is sometimes far in the past, but whose still-active effect bears witness to the “causal significance of temporally remote events or processes” .
13If we follow the models developed by Ruth and David Collier, and James Mahoney, in particular, a series of initial conditions define the available options (specific public policy, specific coalition or type of government, etc.), one of which is endorsed by actors during foundational episodes, called critical junctures. Once they have been selected, in a context which is characterised by contingency, these options give rise to a structural or institutional framework that becomes stable and demonstrates great resistance to time. Indeed, the critical juncture possesses the unique characteristic of making any chosen direction sustainable, insofar as the said structural framework is reinforced by an autonomous series of mechanisms that ensure its continuing existence. 
14Following the choices made at the critical juncture, the birth of such a model and its survival trigger “reactive sequences”, a series of reactions and counter-reactions initiated by the positions of actors in relation to the ongoing institutional configuration. These reactions channel historical development up to its final outcome, which signals the resolution of the conflicts which characterize the reactive sequences. This endpoint creates a new structural configuration, such as the formation of a new partisan system, political regime, or electoral system. 
The critical juncture or the plurality of possibilities
15The critical juncture is the starting point of path dependence. It consists of an accelerated opening sequence immediately presenting several possible paths for change; once over, only one path can be pursued, selected from the previously available but now impossible options. The unique characteristic of the critical juncture is that it represents a moment of “structural indetermination”, in which political entrepreneurs manage to forge certain institutional models in a more pronounced manner than they might under normal circumstances.  As a central element of the explanatory framework of change within historical institutionalism, numerous theorists within this current often refer to the critical juncture to “distinguish periods of continuity and ‘critical situations’ in the stream of historical events, that is to say, moments when important institutional changes occur, thereby creating ‘bifurcations’ that engage historical development on a new path”.  Institutional innovation depends, according to this line of thought, on the appearance of a highly fluid context that allows a heavily influential social and political transaction to take place among actors.
16In this context of fast-moving historical events, we are essentially dealing with relatively short periods compared to the length of the causal process whose outcome is supposed to be affected by the critical juncture. These periods define an environment in which the preferences which have emerged become influential over the long term. By generating a configuration with a critical downstream effect, they significantly constrain future developments.  The range of previously envisioned alternatives becomes irreversibly limited after the period of change closes. The setting in motion of a distinct trajectory indicates the abandonment of alternative paths. After this period of openness and change comes a longer period of stability.
17The critical juncture, whether considered as the occasion of “adoption of a particular institutional arrangement from among two or more alternatives”  or the expression of “relatively short periods of time during which there is a substantially heightened probability that agents’ choices will affect the outcome of interest”,  or even qualified as a “turning point”,  attributes causal force to actors’ actions, with regard to historical processes that may be temporally distant from the initially endorsed choice, but are nevertheless linked to it by a series of causal sequences.
18The concept originates in the work of historical sociologists Seymour Martin Lipset and Stein Rokkan, and Barrington Moore. The former identified a series of cleavages that have shaped the party system throughout Western Europe over the long term. These cleavages were themselves the outcome of historical crossroads such as the French and industrial revolutions.  Moore focused on the type and nature of the alliances between the bourgeoisie, peasantry, and landowners during the pre-industrial period, and their impact on the form of political regimes that appeared during the twentieth century.  Although Moore gave greater attention to the role of actors in his analysis, the second iteration of the notion of critical juncture launched by the work of Ruth and David Collier – on the incorporation of workers’ movements in Latin America and their temporally distant effect on a multitude of institutional variables – primarily accentuated the relatively autonomous capacity of actors to forge institutional models,  thereby reducing the impact of structurally defined historical configurations.
19Moreover, the specificity of the critical juncture is characterized by the contingency of the relevant historical sequence. The choices made at this point are sometimes discretionary or are sometimes rooted, to a limited extent, in prior conditions. Jack A. Goldstone notes that the structural framework “is not determined by any particular set of initial conditions”. It is linked “stochastically to initial conditions, and the particular outcome that obtains in any given ‘run’ of the system depends on the choices or outcomes of intermediate events between the initial conditions and the outcome”.  In reality, change happens through the possibility of increased concretization of actors’ choices brought about by the release of structural constraints belonging to the period of stability preceding the juncture.  In other words, the loosening of prior constraints gives free rein to a form of contingency that opens up a space for action.
20By definition, contingency discourages the selection of an option through attempts to predict its outcome. The adoption of one path over another is unexpected and contradicts the prior understanding of the functioning of the historical process in question.  Thus, in the words of Raymond Aron, “the intelligible order, that the historian observes at the macroscopic level, was not determined ahead of time, it is what actually occurred, not necessarily what was possible in advance”. 
21However, the schema emerging from the juncture is not derived, ready formed, from actors’ preferences; it is not sufficient for them to express their intentions at the right time to sustainably determine a given trajectory. While it is an accelerator of change, according to a relatively random method, the juncture nevertheless remains influenced by prior conditions. The selection process for one of the alternatives is, according to Ruth and David Collier, an outcome of “generative cleavages”, which are active when the initial conditions give way to the juncture. By defining the terms, the actors, and the issues at the juncture, these cleavages specify the framework in which the process of selection will take place.
“[t]he activation or exacerbation of the cleavage creates new actors or groups and the critical juncture consists of their emergence […] the cleavage may be important not because it leads to the emergence of new organized actors, but because it raises political issues so compelling as to trigger some kind of larger reorganization of political relationships”. 
The sustainability of the structural framework
23If at the critical juncture, change allows the society to embark on a new institutional trajectory, the following period will maintain and inhibit change. An essential element of path dependence is the permanence of choices adopted. The direction taken will give rise to “increasing returns” phenomena that freeze change, or at least strongly reduce the likelihood of its future occurrence. Through these types of phenomena, “the probability of further steps along the same path increases with each move down that path […] because […] the costs of exit – of switching to some previously plausible alternative – rise”. 
24In addition to “increasing returns” mechanisms developed from rational calculations of costs and benefits, other mechanisms operate to maintain the structural framework generated by the critical juncture. Ensuring institutional reproduction, these mechanisms are distinct from the processes that first enabled the establishment of the institution. The institutions marked by path dependence phenomena perpetuate themselves in the absence of the original causes of their existence. This independent series of mechanisms can be of a functional nature, to the extent that the institution reproduces itself to fulfill a role in the system; it can also arise from a power relation in which a sufficiently powerful elite succeeds in preventing any modification of the framework; and finally, it can derive from the worldviews of actors, who will seek to reproduce the institution because it is perceived as legitimate and in accordance with their beliefs. 
The channeling of historical development: reactive sequences and the final outcome
25The continuing existence of an institutional form triggers a process of positioning by actors who will seek, according to their preferences, either to accommodate it, or, conversely, reject it. The first reactions provoke opposite reactions, then a series of reactions that drive historical development to its final outcome, the legacy of the critical juncture. Mahoney calls this causal string “reactive sequences”. Indeed, there often exists an interval between the critical juncture and the “period of continuity […] To the extent that the critical juncture is a polarizing event that produces intense political reactions […] the crystallization of the legacy does not necessarily occur immediately, but rather may consist of a sequence of intervening steps that respond to these reactions and counterreactions”.  The historical development triggered by a foundational episode evolves through these successive sequences until the final result is attained, which puts an end to the evolution initiated by the juncture. This result, which resolves conflicts through compromise, involves the creation of a new institutional or structural configuration such as a party system, political regime or union organization,  etc.
An explanation of political development in Belgium
Initial conditions: the interaction of cleavages
26During the second half of the nineteenth century the twin-issue conflict, in which the working class and the defenders of the Dutch language opposed the dominant francophone grande bourgeoisie, brought about a particularly fruitful interaction – in terms of political change – between the principal structural cleavages that become “generative cleavages”. These generative cleavages fit the description of initial conditions and define the terms and actors within the juncture which was to come between 1885 and 1910: taking as its starting points first the opposition between owners and workers and the growing protest by the Flemish periphery against the francophone center,  but also the evolution of the confrontation between the clerical and the secular, the question that arises is that of alliances on the national level between these different emerging actors – socialists, Flemish-speakers, and Catholics.
The critical juncture: Catholic and socialist orientations
27The simultaneous appearance of the Flemish movement and an organized working class, and the intensification of the conflict between Catholics and the anticlerical movement accelerated the advent, between 1885 and 1910, of the critical juncture. This juncture evolved in two stages in a context of contingency, without being predetermined by prior conditions, and is identified with the socialist option of ignoring Flemish claims and the Catholic option of endorsing them (Figure 1).
The socialist choice to ignore linguistic demands
28Although the cradle of the socialist movement – for which Brussels but especially Ghent provided a structure, a political party and an ideology – was Flemish, the movement was to develop later and less widely in Flanders than in Wallonia. Several reasons, including the absence of an industrial tradition (except in the textile sector) and the extension of liberal progressivism to the detriment of its socialist cousin, explain the feeble growth of socialism in Flanders. But the main reason for socialism’s failure to develop is due more to the Church’s containment efforts. In Flanders, the Church reacted in time to stop the socialist recruitment process, whereas in Wallonia the Church’s strategy was implemented too late.  The effectiveness of this move remains surprising to the extent that, at the end of the nineteenth century, the Church did not really appear to be well-placed to respond to the expectations raised by industrialization and democratic advances, since it sought rather to slow down the expansion of the workers’ movement. This reluctance is evident in the repressive reaction to the movement to fight inequalities led by Adolphe Daens, a priest, which reveals that in their early understanding of social issues, the leaders of the Church were troubled by the demands to challenge the established order and rapidly adopted a solidly conservative position.
29Given this starting point, how did the Catholic world of the 1890s succeed in turning those social groups who enjoyed social and political power, especially in Flanders, away from the socialist appeal? The real question is rather, given the anteriority of their political and organizational structure, to understand “why the socialists failed to mobilize all segments of the Belgian working class, and why Catholicism emerged as socialism’s chief competitor”. 
30In reality, although the interruption of the progression of secular thought in Flanders was the work of the Church, this is nonetheless a cause that fed off failure, procrastination, and insufficient investment by the socialists in Flanders. The answer to the question is thus perhaps partly to be found in the strategy deployed by the socialist movement to recruit the greatest number of workers. Looking first at the industrialized zones, the leaders of the workers’ movement concentrated their efforts on Wallonia. The recruited workers were coal miners and metallurgists, of which there were many in the south of the country. Hence, in part, the use of French among the national leaders of the socialist workers’ movement to target the francophone constituency. The use of Dutch was also low, even non-existent, among the national authorities of the Parti ouvrier belge (POB) [Belgian Workers Party], as evidenced by its exclusively French-language publications.  Flemish workers were therefore less exposed to socialist discourse for linguistic reasons.
31Nevertheless, other elements underline the socialists’ efforts to include the greatest number of workers in their organizations, including those from the north of the country. The socialists’ declared anticlericalism did not constitute an insurmountable obstacle to the rallying of the working masses. Certain key figures, such as Emile Vandervelde, president of the POB, encouraged the recognition of free education and religious associations during the adoption of the Charter of Quaregnon in 1894.  This moderate approach, reflected in the dropping of the term socialist in the title of the new party in 1885, was conducive to bringing together all workers, in particular those in Flanders, most of whom had a religious affiliation. Nor did the difference with the Catholic approach in relation to the inclusion of Flemish workers in a dynamic of social and political mobilization lie in recruiting workers from a wider range of socio-economic levels: both the socialist unions and the Catholic unions later in the 1890s, enlisted workers from the same trades: weavers, mechanics, cotton workers. 
32However, in important contrast to the Catholic approach, was the socialist ignorance of the Flemish movement’s linguistic demands. This ignorance had long-term consequences in the sense that, ultimately, the socialists’ primary election base was Wallonia, whereas the social and political expression of Catholicism was strongly and above all rooted in Flanders. This situation was to be equally critical for the Flemish movement, as M. De Vroede notes.
“The fact that the Flemish movement missed its opportunity with the workers’ movement was of capital importance for how it subsequently developed. The workers’ movement first took shape within the socialist movement. Socialism was not, in principle, anti-Flemish, but it viewed social issues in a context broader than that of an exclusively cultural perspective, and gave priority to the improvement of workers’ material conditions. The socialists […] misunderstood the possibilities that the combination of language and class oppositions contained, when in fact they could be cumulated.” 
34There was certainly no antinomy between the socialist class struggle and the linguistic battle of the Flemish movement. The two even joined forces to challenge certain privileges of the dominant classes, which meant the bourgeoisie for the socialists, and the francophones for the Flemish-speakers. From 1830, those who held economic power and those who did not speak the people’s language were one and the same. They were the two faces of the same actor, the francophone bourgeoisie of Flanders, Brussels, and Wallonia, who could be blamed both for the precariousness of workers’ rights and the refusal to improve the status of Dutch. The socialists were even sensitive to the balanced use of the languages in public administration. There were few things to discourage them from supporting the demands of the Flemish movement, which would have added a cultural dimension to the struggle for material rights; and the Flemish movement could have benefited from the broad political base it lacked.
35However the socialists made the decision to center their fight primarily on securing a change in the social situation of workers from an essentially material point of view. It also declared an anti-religious stance and an uncompromising vision of class struggle, which may have put off the Flemish petite bourgeoisie, which favored a more conciliatory vision of relations with the upper social classes. This may partly explain the lack of liaison between socialism and Flemish nationalism. But none of the above predetermined the missed opportunity. All throughout the nineteenth century, the Flemish liberals who had the most significant presence in the Flemish movement were the most visible proponents for the “restitution of [linguistic] rights”.  Anticlericalism was not an element unknown to the Flemish movement, which readily accommodated its anti-religious members so long as they contributed to the linguistic struggle.
36To a certain extent, the outcome of these dilemmas contradicts the prediction that the socialists would have found a natural ally in the Flemish movement to overturn the economic and linguistic privileges of the bourgeoisie. In doing so, it highlights the environment of contingency surrounding this choice. After the missed opportunity for the socialist workers movement and the Flemish movement came the successful blend of linguistic demands with the aspirations of the Christian workers’ movement.
The Catholics’ choice to support Flemish linguistic grievances
37In the race against the socialists to recruit the greatest number of workers, the Catholic unions, on the other hand, took up Flemish linguistic grievances. Unlike the socialist leaders who never, either in Ghent or, particularly, in Brussels turned to the issue of language equality – despite its popularity among certain workers – the Catholics managed to get hold of the issue and include it in their framework of demands. Even if, “in many ways it is paradoxical that the Catholics should outflank the socialists in this respect. The upper levels of the Catholic Church in Flemish areas were heavily Francophone, and the Church strongly identified with the Francophone national government that had given it both financial support and the freedom to run its own affairs”. 
38Faced with the transformation of the working-class landscape, the Church reacted strongly, attempting to hamper socialist progression within the working class. Putting an end to the ban on political activity in 1891 by issuing the Rerum Novarum encyclical, the Vatican simultaneously encouraged Catholics to form unions with the specific objective of undermining the socialist hold. In was in this context that the first – conservative and paternalist – Catholic unions took shape. Around 1891 in Ghent, Arthur Verhaegen, a renowned Catholic, took it upon himself to bring together certain neutral – that is to say non-socialist – unions, with groups of Catholic workers. This coming together absorbed the independent unions into the realm of the majority religion’s political activities. They became Catholic unions encompassed in the “Antisocialistische Werkliedenbond” (Antisocialist Workers’ League) created in December 1890. From the beginning, they positioned themselves in direct opposition to the socialists, in line with the Church’s policy of containing the expansion of anticlerical thought and protest against the established order.  They also worked to form unions constituted of workers whom the socialists had ignored: the workers in the suburbs, women, commuters from country villages, unqualified workers, and of particular interest to us, Flemish workers.
39It was on the language issue, which had been largely neglected by the socialists, that Christian unionism was able to set itself apart. Supported in their efforts by the lower-ranking clergy, which was closer to the people, the Catholics moderately but determinedly endorsed the linguistic struggle. The duality of the National Federation of Catholic Unions, comprised of two organizations, one Flemish and the other francophone, allowed for the regular organization of Flemish social weeks, to discuss not only social matters, but also the failures in terms of language rights.  The exclusively national structure of the Belgian Workers Party did not allow, on the other hand, for debates more specific to Flanders to blossom. In parliament, it was the Catholic deputies who most often introduced and voted on laws which brought linguistic advances, such as the law recognizing Dutch as the second national language in 1898. Using the linguistic issue as a medium, was also a way for the Catholic Party to become predominant. For instance, it opposed military service spread over several years, which, they claimed, as well as contributing to the catastrophic depopulation of the countryside by young people, posed the additional difficulty of professional development in a military body in which French, which most young people did not know, was the official language. 
40In the wake of the successful union between the antisocialist workers movement and the Flemish movement many Flemish nationalists were drawn towards Christian democracy, which brought together Catholic workers’ organizations with progressive tendencies. Christian democracy thus became a pillar of the Flemish movement and its political expression, and definitively reinforced its Catholic component to the detriment of its liberal counterpart, despite the latter’s hold on power throughout the nineteenth century.  The deliberate inclusion of the linguistic dimension and the initiatives that accompanied this decision made Christian unionism “quantitatively more widespread in Flanders than in Wallonia […] better matching Flemish cultural identity: that will be its strength”. Better adapted to the economic, social, and cultural fabric of the north of the country, Flemish leadership was clear about this, while there simultaneously existed a ‘certain overlap between the Christian workers’ movement and the Flemish movement”. 
The birth of a structural framework: the formation of communitarian and ideological categories
41The multiplicity of relationships forged between 1885 and 1910 draw attention to the construction of the two groups that would confront one another throughout the linguistic cleavage. With the closing of the critical juncture, a new structural order was established. Together, the confirmed and ignored choices led to the emergence of two dominant sociopolitical categories: one Catholic and Flemish (occupying a majority in the Dutch cultural space); the other socialist and francophone (the majority in the francophone cultural space). Although the possibility for the socialists to effect a link with the Flemish movement could have tempered the concentration of these categories within communities, the socialists’ lack of interest in this possibility enabled the Catholics to integrate the Flemish movement into their political and social organizations. These choices had multiple consequences.
421/ The votes from the Flemish movement allowed Catholics to fully benefit from the demographic superiority of Flanders. From 1884, they held power alone for thirty years.
432/ The socialists’ lack of understanding of the linguistic issue limited the expansion of socialism in Flanders. More importantly, it pushed the linguistic issue into the heart of the secular/clerical cleavage. It was within the religious (Catholic) pillar that this matter would henceforth be addressed in de facto opposition to the secular pillar. From this point on, the Catholics would consider the linguistic issue through the prism of the philosophical cleavage.
443/ The center/periphery cleavage (Flemish periphery versus francophone center) was thus eventually absorbed by the secular/clerical cleavage. According to Jan Erk, it was during this period that “Flemish nationalism was internalized into the Catholic/secular cleavage of Belgian politics, while the secular left increasingly became associated with French-speakers”. Taking the period in which workers’ movements were formed, and studying them in connection with the establishment of universal suffrage in 1893, Erk equates the Catholics’ choice to create an alliance with the Flemish movement with the political complexion of the right as Flemish sub-nationalism, and of the left as Walloon sub-nationalism. 
454/ Although the intersection of the center/periphery and secular/clerical cleavages is an undeniable reality, the configuration of alliances which came about as a result of the establishment of workers’ movements prevented this intersection from achieving greater balance, allowing a tangible “cross-adding” effect. Undeniably, the Walloon side – separated from the Flemish by the linguistic cleavage – was also divided by the philosophical cleavage, which also separated the two groups in Flanders. The intersection of cleavages operates through the alliance of a majority actor in one of the two linguistic regions with another minority actor in the other region. The intersection of cleavages results in the attenuation of conflicts, since each side is held back in its mobilization by the link that connects him to a part of the opposing camp. 
46The sociopolitical categories that emerged, Catholic and Flemish on one hand, francophone and secular on the other, are both made up of a majority that emerged from each pillar within the two cleavages: for the first category, it is a double majority, made up of clerical and periphery actors, whereas for the second, it is a majority emanating from the secular and francophones who supported, or partly supported, the center. As a result, the Catholics were able to secure over 60% of the vote in Flanders, while the predominant socialist group enjoyed a comparable political base in Wallonia. Yet, this supremacy never cut both ways, as the Catholics consistently remained a political minority in Wallonia, and the socialists were always the underdogs in Flanders. This situation created imbalance in the dynamic of the intersection of cleavages and limited its pacifying impact.
47The very real intersection of cleavages nevertheless allowed another reality to persist, attributable to the manner in which sociopolitical groups were formed, and with additional attributes which tended to galvanize linguistic conflict. The superimposition of the communitarian and philosophical cleavages effectively directs the intensity of the first cleavage towards communitarian conflict and the intensity of the second towards philosophical conflict. This surplus of conflict also appears when actors’ membership of a social class coincides with their community origins. Here we observe an alignment between class, philosophical choice, and language, apparent when the Flemish provinces were subjected to francophone cultural and economic ascendancy, until the 1930s.
48This accumulation of cleavages, sometimes ignored by analysts focusing primarily on the phenomenon of the intersection of cleavages, only plays out fully through the deeply rooted perceptions of actors, who, in a context of conflict, are inclined to relegate reality to the back burner. In this regard, Lode Wils evokes the “polarizing effect of mutual representations that is undoubtedly more important than the different political realities: of a clerical and fascist Flanders and of a Marxist Wallonia. Furthermore, it is difficult to argue that these representations have no connection with reality”. 
Structural persistence: reproduction mechanisms
49Communitarian categories did not decline in the first decades of the twentieth century. Quite the contrary: they were considerably reinforced by sustainably defining the response of political actors to community conflict. The durability of the structural framework is ensured by the “pillarization” system which characterizes Belgian political life. Indeed, each of these categories is consolidated and differentiated from inside the Catholic or socialist pillar. As the point of aggregation for the interests of the pillar from which they originate, political parties were themselves consolidated by the establishment in 1899 of proportional representation, which preserved multi-partisanship, and reinforced their role as privileged representatives of a segment of the population. The implementation of universal suffrage in 1893, despite being tempered by the plural vote of the wealthy classes, had already permitted – thanks to the electoral weight of the population exposed to the pro-Flemish discourse – the Flemish issue to acquire a new acuity. 
50However, the longevity of communitarian categories primarily depends on a power mechanism, a direct consequence of the arising configuration of relations.
The power mechanism
51Explanations for institutional persistence based on power relations posit that the newly created structural framework inequitably distributes costs and benefits. Preserving institutional continuity is thus eminently contentious, as it is based upon the diverging interests of actors who do not identically profit from the new order. Certain actors gain strength from this change at the expense of other actors. The group with newfound advantage may previously have been less powerful, perhaps even dominated. But it will use its power gain to consolidate the institution and further destabilize the power relations with the disadvantaged group. Thus, the appearance of a new structural framework “alter[s] the power structure within society by strengthening previously subordinate actors at the expense of previously dominant ones”. 
52The advent of a series of new alliances, which allowed for the formation of a majority sociopolitical category in the two regions, led to a real change in the power structure in the sense described above. The formerly marginalized Flemish were able to make gains thanks to their recent connection with the Catholics, who tightened their grip on power, whereas previously they had alternated power with the liberals. From the last decade of the nineteenth century, the process of adoption of linguistic legislation consolidated and broadened its areas of intervention.  This period corresponds in its entirety to the Catholic Party’s accession to power in 1884, which it would hold onto, and would govern the country without sharing that power for thirty years.
53The dominant force – the francophone bourgeoisie – lost some of its power. In addition to the marginalization of the liberal party, which included a good number of leading figures of the traditional political elite, as a result of the establishment of universal suffrage in 1893, the majority of representatives of this francophone bourgeoisie were sidelined from the top levels of the Catholic Party after the First World War and the establishment of a uninominal voting system based on universal suffrage in 1919. Broadly speaking, the Flemish wing of the party was strengthened by the votes of supporters of the Flemish movement. Thus, in the 1921 elections, 46 of the 80 seats secured by the Catholics were attributable to the Flemish section.  The universalization of electoral suffrage magnified the effects of the structural framework. It boosted the Flemish issue and the popularity of actors who supported it, while penalizing those who countered it.
54In parallel, the emerging forces – namely the socialists – had to wait for decades before acceding to power. In the 1912 elections, liberals and socialists came together in the hope of putting an end to Catholic domination. But the cartel of the left, the majority in Wallonia, was defeated at the national Belgian level by a gap of only 75,456 votes, which was sufficient, however, to confer an absolute majority status on the Catholics once again. 
Reactions and counter-reactions: the channeling of historical progression towards the final outcome
55The persistence, or perhaps even the reinforcement, of communitarian categories set into motion a series of reactions and counter-reactions that can be characterized as a dynamic of community differentiation. The wronged actor is the first to trigger this causal sequence of interconnected positioning by opposing certain consequences of the structural framework, namely the unfavorable power relations that ensue and successively affect each community. Once established, this dynamic of community differentiation progressively influences historical development so as to increase the discrimination between the two newly created entities. The organization and nature of alliances construct a rationale in the political system which exclude it from the processes, observed elsewhere, of building a Belgian state. Pressure is then exerted on the unitary state structures leading to the multiplication of institutions and of boundaries which are intended to correspond as closely as possible to these mobilized sociopolitical categories.
56The structuring of the Catholic workers movement ensured broad circulation and visibility for the demands of the Flemish movement, transforming it into a mass movement on the eve of the First World War. In response, a nascent Walloon regionalism rejected the bilingualism that was called for at one time by the Flemish movement, thus contributing to the assertion of unilingualism in both of the two large regions. Consequently, the linguistic conflict shifted from a social conflict between the francophone elites of Flanders and the actors of the Flemish movement, to a territorial conflict opposing north and south. This shift, without which the conflict could have remained as solely a class struggle, was the result of the widening perception in Wallonia of a Catholic Flanders with many cultural demands but backwards economically. The opposite perception of a socialist and anticlerical Wallonia, in collusion with the francophone center on linguistic matters, also took shape in Flanders. All of these elements – the links between them and their joint attribution to the same actor – would not have been evident without the relatively random formation of two sociopolitical categories as a result of the choices of the workers’ movements, who sometimes combined, and sometimes ignored, the socioeconomic, ideological, and cultural claims at their heart.
Shifting the axis of conflict: from class struggle to the territorialization of conflict
57In 1870 in Flanders and Brussels the francophone side took a defensive stance on the first laws introducing Dutch into the administration. These beginnings of the Walloon movement were instigated by francophone civil servants who had emigrated from Wallonia to Flanders. In Brussels in 1888, the Walloon Propaganda Society was founded, pursuing the objective of maintaining the predominance of French culture as the cement of national unity, and of defending the positions of Walloon civil servants which were threatened. The first public demonstrations by the Walloon movement pushed for a unilingual Belgium united around francophone culture.  The opposition subsequently won over the liberal socialist groups in Wallonia as a result of the Catholic dominance of the government from 1884. 
58It was the realization that “the Flemish national movement and Catholicism were, after all, indivisibly linked in the reality of political power and in the collective imagination”  which altered the support of the Belgian Workers Party (POB) for linguistic equality. In 1898, its deputies had voted overwhelmingly for the Coremans-Devriendt law recognizing language equality. The Catholics’ continued hold on power and close connection with the Flemish national movement convinced the POB, whose deputies were primarily from the Walloon constituencies, to omit language equality from its 1909 program. The legislative and provincial elections of June 1912 which saw the defeat of the POB-liberal cartel provoked an exasperated reaction in the face of the repeated marginalization of the smaller Wallonia. Strikes exploded in Hainaut. In Liege, incidents degenerated and four demonstrators were killed. In August 1912, Jules Destrée, socialist deputy for Charleroi, published his letter to the King expressing what is described as a dynamic of community differentiation in these terms: “In Belgium there are Walloons and Flemings, there are no Belgians.” 
59The linguistic law of 1932 was a vehicle for these reactions. It was inspired by the principle of regional homogeneity and the desire of the Walloons to sanction the principle of language territoriality out of fear of a generalized bilingualism. The reaction of the Walloon movement thus had several effects.
601/ The axis of conflict shifted. The principle battle fought by the Flemish movement was oriented towards the francophone Flemish elite, the “fransquillons”, essentially a class struggle within Flanders. The socialists were initially favorable to linguistic advances. But the concretization of the alliance between Catholics and the Flemish nationalist movement during the critical juncture and its control over the exercise of power changed the situation, and provoked a Walloon and socialist reaction, resulting in an association with the francophone center against the extension of Dutch. The Walloon rejection triggered by the structural framework and its consequences evolved from integral unilingualism to a defense of bilingualism in Flanders, and finally to the recognition of the unilingualism of both regions.
612/ Simultaneously, the conflict became territorialized. The linguistic boundary was enacted in 1932, while the Brussels district became bilingual. From then on, Flemish and Walloon nationalists sought to defend the cultural and linguistic integrity of each region. “This choice for territorial unilingualism set the country on course for territorial rather than group conflict”,  announcing from the outset the shape of future reforms, and signposting a federal state.
On the road to federalism and a model of increasing autonomy
62The Walloon reactions were such that they radicalized the Flemish movement with regard to identity, and it is clear that “its Catholic character came out of it stronger”.  This was also true of the movement’s demands, since as a result of the brake on the adoption of measures improving the status of Dutch (the minimalist program for which had been formulated during the First World War by Flemish moderates), Flemish nationalists demanded cultural autonomy as early as the 1930s. The idea of a federalist state was part of Catholic Party thinking too. On the Walloon side, in order to counter the industrial decline and escape the domination of Flanders, the autonomist solution gradually established itself: structural reforms were viewed by both sides of the linguistic cleavage as a solution  to the conflict marking the sequence of reactions initiated by the critical juncture.
63In the post-war years, two major episodes were part of this sequence of positionings and drove the process marked by reactive sequences to its resolution: the Royal Question and the winter strike of 1960-1961.
The Royal Question
64The Royal Question refers to the political crisis provoked by the nagging issue during the post-war years of the return of King Leopold III to Belgium at the end of his captivity in Germany. The referendum on the return of the king organized in March 1950 led to the emergence of a double majority, but with opposing results: 72.2% of Flanders voted in favor of the king’s return, whereas 58% of Wallonia and 52% of Brussels voters rejected his re-accession to the throne. But the overall outcome of the referendum was favorable, on the national level, to the lifting of the impossibility of sovereign reign and to the return of the king. Riots immediately broke out in Wallonia, with incidents leaving several dead. When the king returned to the country, he finally renounced the throne and abdicated in favor of his son.
65Although the Royal Question unfolded at the intersection of cleavages, its conclusion – which sparked a rebellion in Wallonia and trauma in Flanders – had to do with the cumulative effect of the superimposition of two cleavages: the philosophical and the communitarian. During the crisis, in effect, “little by little, the traditional cleavages of Belgian politics and those that were revealed by the royal question started to coincide, such that the conflict began to resemble a confrontation between a Catholic and conservative Flanders and a secular and leftist Wallonia”. 
66Sometimes described as a typical case of cleavage intersection,  the Royal Question was concluded in reality on the basis of a cumulative effect that rendered ineffective this very real intersection’s power to limit tensions. Prior to the referendum, the Catholics favored the return of the king while the socialists opposed it. Here we have a Catholic/secular cleavage dividing the two communitarian sides (intersection of cleavages). The conflict was still contained, among the elite, by the left- and right-wing parties (conflict regulation). After the referendum and the clarity of the results, the cumulative effect of the two cleavages, activated both in terms of political domination by a simultaneously communitarian and philosophical majority and in terms of collective representation, played out to its full potential (conflict intensification). This then became a Flemish/Walloon cleavage on which the philosophical cleavage was traced.
67From the outset, the outcome of this historical episode begged the question of greater regional autonomy. In effect, “the royal question seems to end with a Walloon left victory which succeeds in preventing the return of Leopold III as desired by the Flemish. For the latter, getting rid of the king, experienced as a humiliation, revived their claims in favor of structural reforms”. 
The general strike of winter of 1960-1961
68The general strike followed the Eyskens administration’s decision, against a background of decolonization, to implement a significant austerity program which was introduced in a legislative text called the Loi unique (Single Law). This particularly uncompromising strike was undertaken primarily in Wallonia by the General Belgian Workers Federation (FGBT), under the iron rule of union activist André Renard. Although this social uprising was a failure in relative terms, its impact on community relations is notable.
69Following this social struggle, Flemish politicians came out “frightened by the perspective of a ‘collectivist’ Wallonia”.  The impression left by the isolation of the Walloon mobilization was accentuated by Cardinal Van Roey’s message to the workers of the Confederation of Christian Unions/Algemeen Christelijk Vakverbond, dissuading them from joining the strike. The leaders of the Christian unions were also closely linked to certain socio-Christian ministers. These ministers were about to succeed in eliminating the linguistic section of the census, hence the willingness of the Catholic world, which strongly supported this Flemish demand, to avoid destabilizing the government by supporting the strike. 
70At the very moment when Catholic workers seemed convinced by the action taken by their socialist colleagues, the nature of the link between the Christian unions and the Flemish political world played a role that would impede the social mobilization of Catholics even around causes they did support, leaving the Walloon part of the FGTB to face the government alone. It was in this context that André Renard judged that “the Loi unique is yet another law imposed by one part of the country on the other part of the country”. 
71In the aftermath of the social uprising of the winter of 1960-1961, André Renard created the Walloon Popular Movement aiming to “deliver Wallonia from the yoke of clerical Flemish conservatism. It seemed to him that it would not be possible to realize salutary economic and political structural reforms for Wallonia in a state dominated by a Flemish majority who self-identified as conservative. Federalist tendencies would come to express themselves increasingly vigorously in Wallonia”.  Under the effects of the constraints imposed by the nature of the alliances between the Christian workers’ movement and the Flemish movement, a growing awareness of the need for structural reform developed among the Walloons.
Inheritance of the critical juncture: federalism and the increasing autonomy model
72The conflict marking the reactive sequences resulted in the adoption of federalism. In these sequences, the actors sought to escape, via community differentiation, from the structural constraints that had appeared with the establishment of sociopolitical categories. These constraints were apparent in the experiences of each category of attempts by the other side to exercise dominance. Both Flemish and Walloon actors found themselves urged to get involved in parallel processes of disengagement of unitary structures. This was the response that they formulated when faced simultaneously by the permanence of these categories and of the prior positions of actors on the other side. The structural framework that arose from the critical juncture pushes actors to move in a separatist and conflictual direction (Figure 2).
Reciprocal fashioning of actors’ choices and structural/institutional frameworks
Reciprocal fashioning of actors’ choices and structural/institutional frameworks
73In less than a decade, between 1972 and 1978, the party system that was constructed on a national base split into two. From then on, liberal, socialist, and Catholic families were represented by two parties situated on either side of the linguistic boundary. The solution to the conflict also appeared as the introduction of the early elements of federalism in 1970. Flemish and French cultural communities were instituted as well as the three regions: Flanders, Wallonia, and later Brussels. The regional bodies would only be created in 1980 and 1989, accompanied by a net increase in their autonomy. Federalism has continued to expand from this point, to the detriment of central power.
The increasing autonomy model
74The gradual completion of the process of federalization begun in 1970 established a system of progressive autonomy which is still in force in Belgium. Multiple factors play a role in the dynamic, synonymous with continual change, which underlies increasing autonomy. The formation of the political elite in particular, and its election at regional level, means that the legitimization of the decision-making process operates within the public opinion of each community. Given that “Belgian general interest” has no representatives in the absence of federal constituencies, centrifugal tendencies are part of the rationale of the electoral system. In certain respects, there is reason to think that institutional reforms will continue in the absence of communitarian mobilization, given that the success of regionalization has institutionalized the dynamic of power transfer towards the federated entities. 
75As the result of actors’ response to opt for community differentiation when faced with the critical juncture, these different elements, which regularly challenge the prerogatives of the federal state, constitute recurring sources of blockages in the formation of governments or in the resolution of certain current issues. Thus, it took over a year and a half of negotiations following the June 2010 elections for the federal government to finally be constituted. The dispute that shook Belgian political life for decades was primarily related to the elimination, in Flemish territory, of an electoral and judicial district, Bruxelles-Halle-Vilvorde, which comprised a majority of francophones. Elected on a community basis, like the francophone parties, the Flemish parties pushed for the principle of territoriality to be respected, and in the end triumphed over the former, who accepted the compromise in order to preserve the federal state’s powers to reallocate wealth. Other contentious issues, such as economic reforms, have seen a majority right-wing Flanders oppose a majority left-wing Wallonia. In the background, we can clearly identify the cause of current conflicts. These are always linked back to the trajectory of Belgian political development, which pursued territorial differentiation based on divergent ideological constructions.
76In sum, this progressive autonomy model follows its course. Set in motion by the resolution of conflicts in the post-critical juncture period, situated between the beginning of the twentieth century and the 1960s, this course was in turn subjected, in many ways, to processes of self-reinforcement that with each step brought about further advances down the same path. By creating situations in which incompatible interests are likely to be expressed, increasing autonomy fuels current communitarian conflicts – which, analytically, appear as the epiphenomena of this model.
79The approach taken throughout this analysis has been to apply the concept of path dependence in order to evaluate the role of the past in Belgian community relations. By reconsidering the Belgian case in light of the historicity of the communitarian conflict between Flemish-speakers and francophones, I have demonstrated the usefulness of this approach as a relevant analytical angle for understanding the emergence of communitarian groupings and the institutional trajectories that ensue. More specifically, the identification of a critical juncture, crucial for the future development of community relations, helps us to grasp that the conflict between francophones and Flemish-speakers was not inevitable in terms of historical development.
80This assessment is clearly distinct from the most often-repeated reading of the communitarian conflict, which finds a reflection not just in everyday understanding of the situation but also in numerous political speeches.  This reading suggests that the opposition between Flemish-speakers and francophones concerning the incessant redefinition of the perimeter of the common state structure that constitutes the federal state is the outcome of the cohabitation of two distinct cultures within a shared political space. The cultural disparities, which imply deep divisions, represent the primary explanatory factor for the crises experienced by Belgium. The emphasis placed on these tensions echoes the vision that plural societies are primarily “defined by dissensus and pregnant with conflict”,  the inevitable emergence of which arises from the diversity of “values, […] a profound source of instability”.  This perspective, essentialist in overtone, holds the original cultural differentiation of a political collectivity responsible for the appearance of incompatible objectives among its components, laying the foundations for a conflict that would unfold, in the Belgian context, “between natural objective entities” shaped by an implacable deep-seated opposition. 
81However, the quantitative studies that examined the correlation between social heterogeneity and conflict highlight in certain cases the absence of a relationship between the two phenomena.  Yet, they do emphasize the role of external factors that seem to further increase the probability of conflicts in societies split by communitarian cleavages, than in less fragmented societal configurations.  The analysis developed here proceeds in the same direction. It indicates that initial linguistic diversity only becomes a decisive political issue when cultural markers are politicized by factors not inherent to pluralism. The changing perception of these markers may be due to the intervention of exogenous historical factors, such as ideology, or even the contingency of events, which makes the cultural scope of diversity more complicated. In other words, the presence of cultural pluralism, in itself, has no contentious effect. The two identity and politico-geographical referents that are Flanders and Wallonia did not pre-date the political entity which oversaw their emergence: they are the fruit of a specific interaction that only happened after the creation of Belgium. Following this line of thought, this article highlights the fact that linguistic groups were crystallized at the end of the nineteenth century only through a process marked by relatively contingent choices regarding how to integrate the different emerging actors of the period into the political sphere. The boundaries of linguistic groups, their symbolic systems, their repertoires of collective action, were thus not fixed from the beginning by the existence of two cultural spheres, of Latin or Germanic origins, within Belgium.
82Thus the perspective adopted restores the influential role of collective orientations against the vision that perceives the conflict as a consequence or extension of linguistic difference. Furthermore, it re-establishes the contingency of identity formulations against “primordialist” visions which lend credence to the idea of the natural character of community groupings; and argues for the importance of actors’ collective choice over the course of history, as opposed to progression supposedly historically predetermined by cultural differences.  In the words of Flemish historian L. Wils, it should be noted that “the ethnic opposition between Walloons and the Flemish arises from ideological opposition”.  After untangling the contingency of choices from the intricate web of relations built around workers’ issues, the resulting order is immediately subjected to reinforcement mechanisms. These mechanisms re-establish a determinism that is visible when we attribute to the course of history an almost irreversible objective of increasing autonomy.
83By also articulating the varied dynamics of political development, path dependence theory reveals itself to be particularly fruitful for explaining the evolution of plural societies. When we account effectively for contingency, this allows us, for instance, to recognize the phases of “identity uncertainty” that these societies have experienced. Without this account, the history of political communities would lose sight of the unnecessary and constructivist nature of the genesis of sociopolitical actors, and consequently of the particular historical process through which pluralism is organized within state structures. Moreover, in the context of plural societies, this process appears strongly to depend on the consequences of periods of politicization of identities during the early phases of state consolidation. This moment, which witnesses the development of the shape of national groups, is unique in that it produces effects that have a pronounced impact on a community’s later development.
84Therefore, the type of power relation established as a consequence of identity politicization can determine the trajectories of political change in the long run, giving rise, depending on the configuration of relations which develop between dominants and subordinates, to unstable and contentious institutional models or practices. In particular, as A. R. Zolberg remarks, “Where there has existed a long history of cultural inequality, its consequences are so deeply imbedded in the general culture of the relevant society […] that the process of rectification of wrongs is almost endless”.  In other words, when conflicts start from a situation of power imbalance which is, moreover, established simultaneously with the development of communitarian categories, the resulting struggle between the relevant actors persists beyond its social or economic causes; it becomes part of the rationale of perpetuation of symbolic systems and internal networks of solidarity. To this perceptual origin of conflicts is added the role of formal institutions that plural societies often inherit from their past identity politicization. Institutional characteristics (the distribution of power among groups, political economy, etc.), by creating a unique sociopolitical environment which contributes to the definition of actors’ preferences, shape the collective orientations that may then develop in a direction which fosters processes of instability or, on the contrary, political immobility.
85The “conflictual” trajectory that may emerge is not, however, inevitable, as other paths leading to lower-intensity conflict are generally possible alternatives. The trajectory is not definitive either, because although it is maintained by deterministic development, the advent of other historical conditioning can put an end to it, leading to a re-composition of the identity landscape and the reactivation of a new institutional sequence.
Aristide R. Zolberg, “The making of Flemings and Walloons. Belgium: 1830-1914”, Journal of Interdisciplinary History, 5(2), 1974, 179-235; John Fitzmaurice, The Politics of Belgium. Crisis and Compromise in a Plural Society (New York: St. Martin’s Press, 1983); Anthony Murphy, “Belgium: regional divergence along the road to federalism”, in Graham Smith (ed.), Federalism. The Multi-Ethnic Challenge (London: Longman, 1995); Xavier Mabille, Histoire politique de la Belgique, facteurs et acteurs de changement (Brussels: Éditions du CRISP, 2000); Kaz Deprez, Louis Voz (eds), Nationalism in Belgium. Shifting Identities, 1780-1995 (Basingstoke: Palgrave, 2001); Liesbet Hooghe, “Belgium: hollowing the center”, in Ugo M. Amoretti, Nancy Bermeo (eds), Federalism and Territorial Cleavages (Baltimore: The Johns Hopkins University Press, 2004), 55-92.
Arthur L. Stinchcombe, Constructing Social Theories (Chicago: University of Chicago Press, 1968, 101-29).
The reproduction of a structural consequence in line with the “historical causes” model first appeared in Lipset and Rokkan’s cleavage theory (Seymour M. Lipset, Stein Rokkan, “Cleavage structures, party systems and voter alignments: an introduction”, in Seymour M. Lipset, Stein Rokkan (eds), Party Systems and Voter Alignments. Cross National Perspectives (New York: The Free Press, 1967), 1-64), and subsequently in a Belgian version (André-Paul Frognier, “Application du modèle de Lipset et Rokkan à la Belgique”, Revue internationale de politique comparée, 14(2), 2007, 281-302). Cleavage theory, a variant of path dependence, offers a “freezing hypothesis” of partisan choice in Western Europe. The “institutionalization” of cleavages until the 1960s suggests that the party system reflects the structure of cleavages established in the 1920s. To the extent that cleavages, and the party system that embodies them, are reproduced over time in the absence of the recurring national and industrial revolutions at their origin, we are dealing with a “historical causes” framework. This article distinguishes itself from cleavage theory in that it fits into a perspective focused on actors’ choices rather than structurally determined configurations. It also considers the result of the activation and interaction of these cleavages at one – as opposed to several – historical crossroads, namely the formation of workers’ movements at the end of the nineteenth century.
Lode Wils, Histoire des nations belges. Belgique, Flandre, Wallonie: quinze siècles de passé commun (Ottignies LLN: Qorum, 1996), 157.
Max Weber, Économie et société. Tome I, Les catégories de la sociologie (Paris: Pocket, 2003).
Rogers Brubaker, Ethnicity without Groups (Cambridge: Harvard University Press, 2004), 12. According to Brubaker, mentioning the group right from the beginning suggests that analyzing ethno-national conflicts involves well-defined entities that have always existed. Yet it is important not to lose sight of the “phases of extraordinary cohesion and moments of intensely felt collective solidarity” that do not imply “implicitly treating high levels of groupness as constant” (12). The use of alternative notions of “groupness” and “category” permit an understanding of the contingency dimension of identity. A category is the basis for a potential group in the process of being formed, in which the elements of communication, interaction, and solidarity have not yet reached full maturity. It is in these terms that we can analyze the formation of the two sociopolitical categories that arose in each of the Belgian cultural spaces at the end of the nineteenth century.
Margaret Levi, “A model, a method, and a map: rational choice in comparative and historical analysis”, in Mark Irving Lichbach (ed.), Comparative Politics: Rationality, Culture, and Structure (Cambridge: Cambridge University Press, 1997), 28.
Paul Pierson, Politics in Time. History, Institutions, and Social Analysis (Princeton: Princeton University Press, 2004), 11.
Paul Pierson, “Increasing returns, path dependence, and the study of politics”, American Political Review, 94(2), 2000, 251-67.
James Mahoney, “Path dependence in historical sociology”, Theory and Society, 29(4), 2000, 507-48. For critiques of the concept of path dependence, see especially Adrian Kay, “A critique of the use of path dependency in policy studies”, Public Administration, 83(3), 2005, 553-71 (562). For a variation on the theory in relation to change, see Guy Peters, Jon Pierre, Desmond S. King, “The politics of path dependency: political conflict in historical institutionalism”, The Journal of Politics, 67(4), 2005, 1275-300.Online
James Mahoney, The Legacies of Liberalism. Path Dependence and Political Regimes in Central America (Baltimore: The Johns Hopkins University Press, 2001), 7.
Peter A. Hall, Rosemary C. R. Taylor, “La science politique et les trois néo-institutionnalismes”, Revue française de science politique, 47(3), 1997, 469-96 (476). For a critical evaluation of this notion and of the methodological pitfalls of the associated research technique, see Michel Dobry, “Les voies incertaines de la transitologie: choix stratégiques, séquences historiques, bifurcations et processus de path dependence”, Revue française de science politique, 50(4), 2000, 585-614 (604-06).Online
P. A. Hall, R. C. R. Taylor, “La science politique et les trois néo-institutionnalismes”, 475.
J. Mahoney, “Path dependence in historical sociology”, 513.
Giovanni Capoccia, R. Daniel Kelemen, “The study of critical junctures. Theory, narrative, and counterfactuals in historical institutionalism”, World Politics, 59 (3), 2007, 341-369 (quotation on 348; emphasis in the original).Online
Andrew Abbott, Time Matters. On Theory and Method (Chicago: University of Chicago Press, 2001), 243.
S. M. Lipset, S. Rokkan, “Cleavage structures…”.
Barrington Jr. Moore, Social Origins of Dictatorship and Democracy. Lord and Peasant in the Making of the Modern World (Boston: Beacon Press, 1966).
Ruth Berins Collier, David Collier, Shaping the Political Arena. Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America (Princeton: Princeton University Press, 1992).
Jack A. Goldstone, “Initial conditions, general laws, path dependence, and explanation in historical sociology”, The American Journal of Sociology, 104(3), 1998, 829-45 (quotation from 834, emphasis in original).Online
G. Capoccia, D. Kelemen, “The study of critical junctures…”, 343.
J. Mahoney, “Path dependence in historical sociology”, 514.
Raymond Aron, Leçons sur l’histoire (Paris: Éditions de Fallois, 1989), 401. Taking contingency into consideration spurs thinking on the different alternatives available to actors. Evaluating a critical juncture includes appreciating the scope of the selected path relative to those that were not pursued. The counterfactual method is used to precisely judge the causal weight of certain factors in the outcome of a given historical process (see G. Capoccia and D. Kelemen, “The study of critical junctures…”). With this technique, the researcher constructs a different scenario, with the assumption that an alternative option to the one actually chosen was selected, and reconsiders history accordingly. By demonstrating that the choice of a plausible alternative option would have produced an entirely different final result, this method corroborates the centrality of the critical juncture relative to the explained phenomenon.
R. B. Collier, D. Collier, Shaping the Political Arena…, 33.
P. Pierson, “Increasing returns…”, 252.
Reinforcement mechanisms produce a cyclical effect: an institution created in a relatively contingent manner – whether it be maintained by an elite enjoying a favorable power relation or who have developed a perception of its legitimacy – will, through persistence, accentuate the magnitude of legitimate belief and extend the reach of the elite’s power, such that they will further consolidate the institution. Here we are dealing with selfreinforcing sequences.
R. B. Collier, D. Collier, Shaping the Political Arena…, 37.
J. Mahoney, “Path dependence in historical sociology”, 527. Although those marked by processes of institutional reproduction tend to be self-reinforcing, the reactive sequences describe a historical trajectory that is not deployed via the reproduction of a prior event, but rather by many new developments attributable to the successive positioning of actors in relation to this event.
For the majority of the nineteenth century, the center was occupied by the francophone elite, while the Flemish movement was active in the periphery. In subjective terms, Flanders was culturally and linguistically dominated by the francophones for a much longer period, even when the latter considered themselves to be a political and economic minority relative to an increasingly dominant Flanders. Conceptions of center and periphery thus vary depending on the subjectivity of actors and whether we attempt, independently from actors’ perceptions, to objectivize the relations which characterize a historical situation (on this topic, see Vincent de Coorebyter, “Clivages et partis en Belgique”, Courrier hebdomadaire du Centre de recherche et d’information sociopolitiques, no 2000, 2008, 49).
Els Witte, “Une Flandre appauvrie”, in Els Witte (ed.), Histoire de Flandre des origines à nos jours (Brussels: La Renaissance du Livre, 1983), 215-18.
Carl Strikwerda, “The divided class: Catholics vs. socialists in Belgium, 1880-1914”, Comparative Studies in Society and History, 30(2), 1988, 333-59 (342).Online
C. Strikwerda, “The divided class…”, 335.
V. de Coorebyter, “Clivages et partis en Belgique”, 27.
C. Strikwerda, “The divided class…”, 350. The only difference lies in the location affiliation, urban for the socialists, suburban and rural for the Catholics.
Maurits de Vroede, Le mouvement flamand en Belgique (Anvers: Conseil culturel flamand, Institut d’information, 1975), 39.
M. de Vroede, Le mouvement flamand en Belgique, 36.
C. Strikwerda, “The divided class…”, 355.
Paul Gérin, “Catholicisme social et démocratie chrétienne (1884-1904)”, in Emmanuel Gérard, Paul Wynants (eds), Histoire du mouvement ouvrier chrétien en Belgique (Louvain: Leuven University Press, 1994), 58-113 (90).
C. Strikwerda, “The divided class…”, 357.
L. Wils, Histoire des nations belges…, 162.
M. de Vroede, Le mouvement flamand en Belgique, 40.
P. Gérin, “Catholicisme social et démocratie chrétienne (1884-1904)”, 113.
Jan Erk, “Sub-state nationalism and the left-right divide: critical junctures in the formation of nationalist labour movements in Belgium”, Nations and Nationalism, 11(4), 2005, 551-70 (554). The analysis here is similar to this approach, but strives to go further by demonstrating the structural consequences of the Catholics’ choice on a set of variables; a choice which can be explained by the socialists’ early decision to rule out certain determinant options. The perspective adopted in this article goes beyond the deduction of Belgian sub-nationalisms’ political orientation by retracing, from a methodological perspective, the causal sequences linking the critical juncture to the appearance of an increasing autonomy model which impacts current conflicts.Online
Seymour M. Lipset, Political Man. The Social Bases of Politics (Baltimore: The Johns Hopkins University Press, 1960), 88-9.
L. Wils, Histoire des nations belges…, 332.
Gérin, “Catholicisme social et démocratie chrétienne (1884-1904)”, 84.
J. Mahoney, “Path dependence in historical sociology”, 521-2.
In 1898, Dutch was recognized as an official language; in 1906, it was introduced to the Cour d’assise de Brabant [criminal court] and in 1909 to the Conseils des prud’hommes [labor tribunals]; in 1914, a law was passed regarding primary education, which definitively established the lingua franca as being the mother tongue.
Jan Craeybeckx, “L’ère des mutations accélérées”, in E. Witte (ed.), Histoire de Flandre des origines à nos jours, 250-9.
Michel Dumoulin, Emmanuel Gérard, Mark Van den Wijngaert, Vincent Dujardin, Nouvelle histoire de Belgique. Vol. 2: 1905-1950 (Brussels: Complexe, 2006), 72.
Marie-Thérèse Bitsch, Histoire de la Belgique. De l’Antiquité à nos jours (Brussels: Complexe, 2004), 123.
The Flemish Catholics’ hold on power was very clear. It often exceeded their demographic weight. The distribution of ministerial positions was the following: Walloons held 26% of positions from 1884 to 1894, 16% from 1884 to 1896, 13% from 1896 to 1899, 24% from 1899 to 1907, 22% in 1907, 21% from 1908 to 1911, and from 1911 to 1914, no more than 38% (Yves Quairiaux, L’image du flamand en Wallonie. Essai d’analyse sociale et politique (1830-1914) (Brussels: Labor, 2006), 30).
Sophie de Schaepdrijver, La Belgique et la première guerre mondiale (Brussels: Peter Lang, 2006), 30.
Jules Destrée, La lettre au Roi (Gembloux: Éditions de la Wallonie libre, 1968), 16.
L. Hooghe, “Belgium: hollowing the center”, 58.
L. Wils, Histoire des nations belges…, 214.
Denise Van Dam, “Histoire du mouvement Wallon”, in Marco Martiniello, Marc Swyngedouw (eds), Où va la Belgique? Les soubresauts d’une petite démocratie européenne (Paris: L’Harmattan, 1998), 73-84.
Marc Reynebeau, Histoire belge. 1830-2005 (Brussels: Racine, 2005), 225.
Nathalie Schiffino, Crises politiques et démocratie en Belgique (Paris: L’Harmattan, 2003), 32.
M.-T. Bitsch, Histoire de la Belgique. De l’Antiquité à nos jours, 199.
J. Craeybeckx, “L’ère des mutations accélérées”, 291.
Pascal Delwit, La vie politique en Belgique de 1830 à nos jours (Brussels: Éditions de l’Université de Bruxelles, 2009), 128.
Jean Neuville, Jacques Yerna, Le choc de l’hiver 60-61. Les grèves contre la loi unique (Brussels: Pol-His, 1990), 52.
On this topic, see Britt Cartrite, “Contemporary ethnopolitical identity and the future of the Belgian state”, Nationalism and Ethnic Politics, 8(3), 2002, 43-71.
M. de Vroede, Le mouvement flamand en Belgique, 84.Online
Take for example the attitude of separatist Flemish leader Bart de Wever, the president of the most significant party in the kingdom following the 2010 elections, who regularly highlighted cultural difference as a factor in the understanding of the increasing detachment and tensions between the north and south of the country. This vision is conveyed through this type of speech: “I note that, in Latin cultures, the voter does not react negatively when authorities grant themselves privileges whereas in Flanders, we are attacked for far less. Look at the electoral results in Hainaut: nothing has changed.” (“De Wever oppose la Flandre à une ‘culture latine’ des prébendes”, La Libre, 22 December 2010).
Michael G. Smith, The Plural Society in the British West Indies (Los Angeles: University of California Press, 1965), xiii.
M. G. Smith, The Plural Society in the British West Indies, 90-1.
Marco Martiniello, “Culturalisation des différences, différenciation des cultures dans la politique belge”, Les Cahiers du CERI, 20, 1998, 3-41 (19).
Paul Collier, Anke Hoeffler, “On economic causes of civil war”, Oxford Economic Papers, 50, 1998, 563-73.Online
Randall J. Blimes, “The indirect effect of ethnic heterogeneity on the likelihood of civil war onset”, Journal of Conflict Resolution, 50(4), 2006, 536-47.Online
A. R. Zolberg, “The making of Flemings and Walloons. Belgium: 1830-1914”, 180.
L. Wils, Histoire des nations belges…, 207.
A. R. Zolberg, “The making of Flemings and Walloons. Belgium: 1830-1914, 233.