Introduction
In the study of ethnicity and nationalism, understanding the link between cultural consolidation and state-building has been a central agenda. A usual starting point, for both theoretical and empirical works, is the analytical distinction between nations as culturally consolidated groups and states as organizations that safeguard them. Seminal works hold that sovereign states come before nations emerge (Anderson Reference Anderson2006; Breuilly Reference Breuilly1993; Gellner Reference Gellner2006; Hobsbawm Reference Hobsbawm1990; Tilly Reference Tilly1994). Some point to the existence, besides, states, of an “ancient root” or a “symbol” as a precondition for the rise of nations that helps generate and sustain the belief that members of the community have a common descent (Smith Reference Smith1986). Recent research also shows that for groups that achieve political independence in the late nineteenth century on, the construction of a coherent and discrete community is key before obtaining independent statehood (Aktürk Reference Aktürk2012; Mylonas Reference Mylonas2012; Roeder Reference Roeder2007; Wimmer Reference Wimmer2002; Wimmer and Feinstein Reference Wimmer and Yuval2010). In practice, however, cultural consolidation and state-building are often seen as mutually reinforcing. One reason is that the nation, which is a population that comprises a set of well-established cultural practices, is often deemed as possessing a political right to its own state—hence a right to national self-determination.Footnote 1
The stylized fact in extant research is that cultural consolidation and state-building occur in tandem closely if not mechanically. If one accepts the notion of national self-determination, it is reasonable to assume that once a community claims as a nation, political mobilization toward autonomy or independence is expected to follow. Thus, those nations that fail to do so, or “fail to bark” (Gellner Reference Gellner2006: 42–43), seem counterintuitive. One explanation in the scholarship is that many nations that seek sovereignty remain dormant until the breakdown of institutions occurs, which heretofore put constraint on them (Beissinger Reference Beissinger2002; Brubaker Reference Brubaker1996). A close connection between cultural consolidation and state-building is conceptually and empirically appealing because the sovereign territorial state can function as the ultimate guarantor of the culture of a population. In “ideal types” of nation-states, it is required that a group comprise a state that possesses the monopoly of the legitimate use of violence over its own territory (Weber Reference Weber1978). Building on this model, scholars seek to leverage their empirical analysis either by varying the enforcement capacity of the state or by varying analytical contexts—based on temporality, geography, or levels of analysis (i.e., international or domestic). Cases in developing countries fit well with the former because they tend to deviate from the Weberian definition of the state. States in developing societies do not typically have as much the capacity to enforce rules or raise revenue as wealthy states do. Society in the developing world also tends to be ethnically more heterogeneous. In these cases, analysts examine the occurrence of violent conflicts or the provision of public goods (Alesina et al. Reference Alesina, Baqir and Easterly1999; Bates Reference Bates, Rothchild and Olorunsola1983; Habyarimana et al. Reference Habyarimana, Humphreys, Posner and Weinstein2007, Reference Habyarimana2009; Miguel Reference Miguel2004; Miguel and Gugerty Reference Miguel and Gugerty2005). In the latter category on the analytical contexts, empirical studies focus on the varying incentive structure that key political actors have in terms of the benefit of rallying their community as a nation or mobilizing toward independence (Alesina and Spolaore Reference Alesina and Spolaore2003; Lawrence Reference Lawrence2013).
Few studies empirically investigate the tightness of the link between cultural consolidation and state-building. More specifically, variation in the degree of cultural consolidation among nations and its effects on political mobilization have yet to be examined. Extant research addresses the process of cultural consolidation by exploring how elites of a nation impose their cultural practices as standard through the state apparatus. Another approach focuses on how elites of a minority nation command loyalty within their population to stir nationalist mobilization. These works typically accept the notion of the “nation” as a cohesive, well-established group and operationalize it either as an outcome variable or a unitary actor seeking a state. Yet these approaches leave out the possibilities in which ethnic groups fail to consolidate their culture or, following consolidation, fail to mobilize toward self-rule.
In this article, I examine whether elites of stateless ethnic groups establish institutions that facilitate political mobilization after they consolidate culture. To address this question, I use quantitative and qualitative methods to operationalize cultural consolidation and select cases. Evidence draws primarily from the case of failed mobilization by the Occitans, a minority ethnic group in France, but I also briefly consider the Bulgarians as a successful case for comparison. I first summarize major conceptual frameworks in the literature that connect the link between cultural consolidation and state-building. This exercise demonstrates that, among the ethnic minorities in late-nineteenth-century Europe, extant research has yet to address the mechanism of political mobilization after cultural consolidation. To fill this gap, I construct a simple hypothesis that not all ethnic groups invest in institutions even when given an opportunity. I then use propensity-score matching methods to select a pool of potential groups for empirical analysis. My choice of the Occitans is motivated by the “deviant case” method in case-selection strategy, in which the researcher purposefully investigates theoretically counterintuitive, “underperforming” observations to gauge the explanatory power of a hypothesis. The Bulgarian case serves as an expected outcome, and I use it as my reference category to explore the causes of the failed institution-building among the Occitans. My main finding is that although the Occitans achieved cultural consolidation in the late nineteenth century by standardizing their language early and establishing an active, growing literary society with successful publications, the society’s senior leadership had little motive to pursue the politically more ambitious goal of federalism or self-determination. They remained unable to resist the effort of the French counterpart on language and cultural unification through free, compulsory primary education across the country. Occitans confined their organization’s activity largely to the less ambitious goal of cultural preservation. This inaction proved fateful in the long run, as the movement limited to language revival never regained initial popularity. Today, the political status of the distinct vernacular for the Occitans is uncertain and the vitality of language use in the community is under threat.
This article makes two contributions. First, it adds a new hypothesis to the existing distinction between cultural consolidation and state-building, where I identify “institution-building” as an intermediate category that can be used as a distinct outcome. My goal is more to widen the scope of empirical analysis in the existing mechanism than to construct a new theory. Introducing this category as an outcome of interest yields a handful of new empirical questions that have not been previously explored: Does cultural consolidation lead to institution-building? Does institution-building lead to state-building? How tight is each link? What explains its strength? Second, I address one of these empirical questions, namely the connection between cultural consolidation and institution-building. I measure cultural consolidation by focusing on language standardization as one of the most important dimensions of culture and operationalize it by using the first publication date of vernacular dictionaries for 104 ethnic groups in Europe. Then I investigate whether ethnic elites build institutions as my outcome variable in a detailed case study of the Occitans. Its main findings is that not all groups that achieve cultural consolidation engage in institution-building. Occitan elites had the capacity and resources to build institutions by starting schooling in the vernacular and expanding their literary society, but they opted for a more modest goal and not to compete openly with the French for self-rule. My process-tracing analysis shows that the decision of minority elites of a leading cultural organization plays a crucial role in understanding whether institution-building occurs. It documents evidence that the linkage between cultural consolidation and state-building is not as tight as typically assumed. My analysis suggests that there are likely unobserved factors that make the connection and need to be identified for a fuller understanding of why some ethnic groups succeed in winning their own state.
A Conceptual Framework on Political Mobilization among Stateless Nations
Among numerous hypotheses in the literature on ethnicity and nationalism about the linkage between cultural consolidation and state-building, a common framework seems to exist that applies to the stateless ethnic groups. It abstracts a causal path from cultural consolidation to state-building. The path refers to the stylized fact about how ethnic groups consolidate their culture by establishing a set of rules that define their cultural practices before mobilizing their community for political ends—typically autonomy or independence. This generalization emerges from the experience of ethnic minorities in the late nineteenth century through the early twentieth century, in which they highlight cultural differences between them and their imperial masters to stir a nationalist movement and press for independence in the name of anticolonialism. The mobilization took place across the globe as much of the world at the time was ruled by European empires. Some definitions are in order before discussing this mechanism in detail. In this article, I follow Weber (Reference Weber1978: 389) to define ethnic groups as those that “entertain a subjective belief in their common descent” based on what they perceive as a unique set of shared attributes that delimit membership. Building on this concept, I define nations as those ethnic groups that successfully standardize their cultural attributes. These definitions that center on cultural dimensions are useful for distinguishing cultural standardization from institution-building for political purposes, the latter of which I examine in terms of “success” and “failure.”
My analytical approach is distinct from that of classical works whose frameworks rest on the assumption that nation-states are essentially a function of having a sovereign territorial state. These works uncover patterns or important factors that explain how ethnic groups consolidate their culture (Anderson Reference Anderson2006; Breuilly Reference Breuilly1993; Gellner Reference Gellner2006; Hobsbawm Reference Hobsbawm1990; Tilly Reference Tilly1994). One of the contributions of this research is to produce numerous forms of “ideal types” on nations and nationalism that inform empirical research in subsequent generations (Brubaker Reference Brubaker1992; Kohn Reference Kohn2005).
Recent works address the variation of outcomes in nationalist and secessionist movements by constructing testable hypotheses based typically on a set of detailed case studies. For instance, Brubaker (Reference Brubaker1996) examines the impacts of major institutional breakdowns, such as the collapse of the Soviet Union, on different forms of secessionism. Similarly, Mylonas (Reference Mylonas2012) considers how international and geopolitical factors shape the outcomes of nationalist mobilization by drawing on minority groups in the Balkans. These studies are important in that they specify conditions under which nationalist or separatist movements achieve their ultimate goal of acquiring a sovereign territorial state. Their hypotheses allow for distinguishing between groups that are more or less likely to succeed when certain conditions are met.Footnote 2
I build on these classical and recent contributions to explore whether ethnic minorities foster institutions toward greater political autonomy after standardizing their cultural practices. Cultural consolidation among ethnic minorities in the late nineteenth century, particularly those in Europe, happened in reaction to nationalist movements in earlier decades (Anderson Reference Anderson2006: ch. 3). Many early-starters, such as the Greeks and the Serbs, successfully achieved autonomy or independence and then began to rationalize their culture through national citizenship and universal education (Hobsbawm Reference Hobsbawm1990; Tilly Reference Tilly1994). By contrast, latecomers, such as the Slovenes and the Bretons, typically took the opposite path by consolidating their culture first before seeking political freedom. Ethnic elites in the latter pattern would typically organize a literate society and issue journals, history, and novels in their vernaculars. These publications are intended not only to make minority ethnicity salient across their community but also to help bring about language standardization, because in many empires their tongues were used in everyday communications but rarely in official channels. Intellectuals of minority groups may also start schooling taught in a vernacular that would have the same effect. In this conceptual framework, once linguistic unity is achieved, ethnic elites are expected to follow their earlier examples to mobilize toward territorial sovereignty. A similar pattern is observed in contemporary cases of anticolonial and secessionist movements beyond Europe (cf. Cunningham Reference Cunningham2014; Manela Reference Manela2009).
The mechanisms of political mobilization following cultural consolidation among stateless ethnic groups have not been widely investigated. Recent quantitative works provide evidence regarding which minority communities achieve “nation-building” or the status of “nation-states” (Roeder Reference Roeder2007; Wimmer Reference Wimmer2018; Wimmer and Feinstein 2010). Yet few consider the linkage between cultural consolidation and institution-building toward political mobilization. I distinguish between institution-building and state-building as analytically different concepts. The former is a necessary step to the latter; investing in political institutions would foster state capacity in terms of rule enforcement, taxation, and other administrative transactions. It is crucial to explore this mechanism because, as I show later, the connection between the two is not automatic. Doing so requires one to take a close look at the process of whether ethnic elites strengthen their own institutional capacity. Detailed case studies are appropriate for this type of hypothesis testing because it is not readily captured in quantitative analysis.
My outcome variable in this article is whether elites of ethnic minorities invest in institutions for political mobilization. Territorial sovereignty would be the ultimate goal for them, and the movement typically takes a nationalist character in that ethnic elites seek to generate mass support by making their national identity politically salient and framing the action in the name of their nation. As extant research suggests, the success of nationalist mobilization depends on the strength of political institutions such as parties, religious centers, unions, and other types of associations. They play a crucial role as an organizational hub that enables ethnic elites to promote their causes and mobilize their community on a large scale. I argue that the development of these institutions, in turn, rests on prior cultural consolidation. My main explanatory variable is cultural consolidation focusing on language rationalization among ethnic minorities. It matters to institution-building not only because it facilitates within-group communication but also because, once certain rules about cultural practices are established, elites may use them as political instruments to command mass loyalty to the group and build organizational capacity.
My analytical approach has two advantages. First, it is consistent with the distinction that classical works on nationalism makes between state-building and cultural consolidation. Second, it adds an “intermediate” category in this distinction by exploring the mechanism of political mobilization among ethnic minorities. The new category, which is my outcome variable, is also analytically separate from either sovereign territorial states or cultural consolidation. The rest of my article investigates this mechanism in a detailed study primarily on a failed case of institution-building. It not only demonstrates the conceptual significance of the mechanism but also provides evidence through a systematic measurement of cultural consolidation and case selection.
Empirical Strategy
Process-Tracing Methods
I use process-tracing methods to investigate the mechanism linking cultural standardization and institution-building toward political mobilization. My goal is to uncover a causal process of whether ethnic elites who achieve early language standardization invest in institutions. Process tracing is a method to identify causal processes between an explanatory variable and an outcome variable (George and Bennett Reference George and Bennett2005: 206). The main rationale is to establish what scholars of this methodological tradition call “causal-process observations,” which refer to “pieces of data that provide information about context, process, or mechanism,” the kind of observations that enables analysts to draw causal inference (Brady and Collier Reference Brady and Collier2010: 318). This is an appropriate method for this article because few investigations address the mechanism in question as an outcome of interest and the mechanism as a causal process has not been firmly established. It also fits the purpose of my article because ethnic elites’ decision about whether to foster institutions is not easily quantifiable (or has yet to be done on a large scale) (Gerring Reference Gerring2007). Thus, a key advantage of process tracing is its malleability in application: the researcher can use it to both develop conceptual frameworks and test hypotheses (Bennett and Checkel Reference Bennett, Checkel, Bennett and Checkel2014; Mahoney Reference Mahoney, Mahoney and Rueschemeyer2003, Reference Mahoney2010).
Scholars have developed numerous process-tracing strategies and standards to establish causality, but implementation remains a challenge (Dunning Reference Dunning2012, Reference Dunning, Bennett and Checkel2014). Case-study approaches are most useful when validating the statistical correlation between two variables or when identifying alternative hypotheses or omitted variables. Yet it remains unclear which cases would be appropriate to make causal inferences and how they can be drawn because of the need to control for background conditions before proceeding to comparison. In traditional strategies, for example, the methods of agreement and of difference, the researcher conditions on the value of a main independent variable or the dependent variable. But the probabilities that these values occur or the conditioning of control variables attract less attention (Sekhon Reference Sekhon2004).Footnote 3 To mitigate this problem, I use a strategy based on matching to select cases systematically.
Descriptive Data on the Link between Cultural Consolidation and Institution-Building
Before proceeding to my strategy of case selection, I provide descriptive data on the connection between cultural consolidation and institution-building. Of myriad cultural attributes that ethnic groups can standardize, I focus on language. Once the vernacular is codified, an ethnic group has an easier time communicating within group, which makes the standardization of other cultural dimensions more effective. Ethnic elites may use it as a political instrument to begin an effort to mobilize their community and build institutional capacity based on political parties and religious centers. Language standardization is crucial for setting this process in motion. However, it is one of the most difficult dimensions to standardize because it requires literacy when compared to other cultural attributes such as flags and anthems, which do not. In addition, codification requires high fixed costs. Languages evolves through importing and exporting words from one another, and its codification typically entails the “purification” process by studying the etymology of words and removing foreign ones to demonstrate that language serves as an untainted symbol of political identity for an ethnic group (Burke Reference Burke2004).
Figure 1 documents the cumulative output of language standardization for 104 European ethnic groups from 1750 to 2000. I proxy language standardization by taking the first publication year of vernacular dictionaries for each ethnic group. This operationalization demonstrates the difficulty and high fixed costs of vernacular codification as captured in Figure 1. Specifically, it offers three insights. First, language rationalization is difficult and entails high fixed costs. Of the 104 groups, only 27 communities (or 26.5 percent) did so by 1900, and most instances take place in the twentieth century. Some ethnic groups that comprise a state, such as the Italians, the Swedes, the French, and the Russians, cover these costs by founding a specialized, government-sponsored academy. Those that did not comprise an independent state may launch a literacy society or a cultural organization to promote their vernacular and teach it at local schools.Footnote 4 Second, Figure 1 provides evidence that cultural consolidation, through language standardization, and state-building are distinct processes as postulated in the literature. In particular, it begs a question as to why not all ethnic groups that have codified vernaculars comprise sovereign states. Third, there is little difference in the instances of language standardization between those groups that comprise a state and those that do not by the turn of the twentieth century. When the year 1900 is used as the cutoff for “early standardization” (the vertical bar in Figure 1), there are 27 cases in total, in which 16 had an independent state while 11 did not. This evidence is consistent with my hypothesis that among ethnic minorities, cultural consolidation is not necessarily connected to institution-building toward autonomy or independence.
FIGURE 1. Cumulative number of vernacular dictionary publication among 104 European ethnic groups. Source: Burke (Reference Burke2004), Dalby (Reference Dalby1998), Price (Reference Price1998), among others.
Descriptive evidence drawn from Figure 1 motivates my case-selection strategy for empirical analysis. Specifically, I use the “deviant case” method to investigate the link between cultural consolidation and institution-building (Gerring Reference Gerring2007: ch. 5). It allows the researcher to choose cases that overperform or underperform relative to an expected value on one’s prediction (ibid.: 107). In my context, deviant cases are the ethnic groups whose leaders fail to build institutions toward political mobilization. They “underperform” in that they are expected to invest in their institutional capacity not just to augment the chances of autonomy but also to ensure survival as a distinct community. If elites do not make such a move, their group is likely to remain as a minority—or, worse, may face the prospects of cultural assimilation or integration by their majority ethnic group in the context of waves of nationalist movements in turn-of-the-century Europe. Choosing a deviant case generates evidence for these groups that make this counterintuitive decision. As Gerring (Reference Gerring2007: 106) points out, the purpose of this method is “to probe for new—but as yet unspecified—explanations.” In the nationalism literature, King (Reference King2010) makes among the first attempts to construct a framework on “failures” of nation-building as the main outcome of interest. The deviant-case method is the initial step in my strategy of case selection to shed new light on the linkage between cultural consolidation and state-building.
Matching Methods for Case Selection
In the next step, I use matching methods for case selection and causal inference. Matching is a nonparametric preprocessing approach to adjust observational data before the parametric analysis (Ho et al. Reference Ho, Imai, King and Stuart2007; Stuart Reference Stuart2010). Social science research often relies on observational data when experiments are infeasible. Unlike experimental data, observational data almost always lack balance or are not randomly distributed across the sample. Imbalance creates difficulty for causal inference because it violates the assumption of ignorability or the absence of selection on observables (Ho et al. Reference Ho, Imai, King and Stuart2007: 206–7). If this violation is suspect, one needs to address the concern that unobserved factors might determine the outcome variable and/or explanatory variables.
The primary goal of matching is to improve balance in the distribution of covariates (Ho et al. 2007: 215; Stuart Reference Stuart2010: 1). This strategy offers two advantages. First, preprocessing data reduces dependence on one’s choices of statistical modeling and specifications (Ho et al. 2007). By taking an extra step to achieve balance in the distribution of covariate data between the treatment group and the control group, the link between the treatment variable and the control variables is eliminated or diminished (ibid.: 212). Subsequent statistical analysis, if the researcher chooses to do so, is less dependent on the assumptions of their proposed models and specifications. Second, matching methods focus on the distribution of covariates where there is not much overlap between the treatment and control groups when causal effects are estimated. Statistical models tend to perform poorly when overlap is insufficient, but researchers become more aware of data limitations in their claims about causal estimates when using smaller data through matching methods (Stuart Reference Stuart2010: 2).
Propensity Score Matching
In this articleFootnote 1 , I use propensity score matching. It is a popular strategy for case selection and causal inference and is widely used not just in social sciences but also in medicine and health (ibid.: 1). The propensity score is an index of the likelihood that a key explanatory (or “treatment”) variable occurs based on observational data. Studies that employ this method typically proceed in two steps. First, the researcher estimates a model, where the matching variable is the outcome, to calculate propensity scores of the matching variable. These scores are used to assign observations into the treatment group (the matching variable that occurred) or the control group (the matching variable that did not). In the second step, the analyst estimates the effect of the propensity as the approximate average treatment effect on the outcome of interest.
In my case, the outcome is whether an ethnic group builds institutions toward political mobilization after early language standardization. I define early as language standardization by 1900 because only 16 percent (27 out of 171 ethnic groups in my data set) had published a vernacular dictionary by the turn of the twentieth century (26.5 percent when NAs are omitted).Footnote 5 I use this indicator Ti as the treatment variable to assign ethnic groups in my data set:
$$\eqalign{\it {T_i}\, = \Bigg\left\{\hskip-2.3pc\matrix{{\rm 1}{\rm{\,\,if\,\,ethnic\,\,group\,\,standardized\,\,language\,\,by\,\,1900}} \cr \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,{\rm 0}{\rm{\,\,if\,\,ethnic\,\,group\,\,did\,\,not\,\,standardize\,\,language\,\,by\,\,1900}} \cr}} $$
Ti = 1 refers to a subset of my data set where ethnic group i codified its vernacular by 1900 and Ti = 0 indicates the rest of the ethnic groups that did not do so, given the pretreatment characteristics of the homeland city for each ethnic group. More formally, this case selection strategy aims to capture the effect of Ti on the outcome, Yi , which is whether ethnic groups invest in institutions:
$$\it{ {Y_i \equiv \,Y_i(T_i)\, = \,(T_i)Y_i\left( {\rm 1} \right)\, + \,({\rm 1} - {\rm{ }}T_i)Y_i\left( {\rm 0} \right)$$
(1)
To assign ethnic groups in the treatment and control groups, I first compute propensity scores that denote the probability of achieving vernacular codification by 1900, conditional on a vector of ethnic group attributes Xi :
$${\it{ {P(X_i)\, = \,{\rm{Pr}}(T_i = \,{\rm 1}\,|\,X\, = \,X_i)\, = \,E[T_i|\,X\, = \,X_i]$$
(2)
This data preprocessing allows me to capture the effect of early language standardization on whether ethnic groups start institution-building by conditioning on potential confounding variables for language standardization.
It is important to note that matching based on propensity scores is not without problems. One is that causal estimates from matching models can suffer from model dependence. It is the problem in which the researcher makes inferences from fitted models but extrapolations are not supported by the existing data or these extrapolations are outside the bounds of underlying data (King and Zeng Reference King and Zeng2007). Propensity-score matching does not, in theory, reduce model dependence (Iacus et al. Reference Iacus, King and Porro2012). But the rationale of using this matching strategy in my article is to select cases for hypothesis generation rather than to estimate causal effects of the models. More specifically, my goal is not to extrapolate the impact of early language standardization on state-building but to find cases that bear empirical similarities and differences as a preparatory step for in-depth investigation, rather than to estimate the causal effects of my models, the latter to which these limitations apply.
I estimate propensity scores using a logistic regression model, where the outcome variable is a binary indicator of whether 171 European ethnic groups standardized their language by 1900. Following Sasaki (Reference Sasaki2017), I proxy language standardization by taking the first publication date of vernacular dictionaries. I include all covariates used in the estimation in this article, but I made two major changes. One is that I truncate the data to pre-1900 to capture the pretreatment effects of the variation in language standardization. Second, I transform the original time-series and cross-sectional data set to enable cross-sectional estimation because the original data set contained information from 1400 through 2000.Footnote 6 More specifically, I recode the following time-varying covariates: First, I obtain the total time (in years) that ethnic groups have the printing press by calculating the time elapsed since print adoption. I also create the total time for the university and bishoprics for each group using the same method. Second, I calculate the mean for the war frequency and urban potential variables between 1400 and 1900. War frequency represents the average occurrence of war during this period. Urban potential is a measure of economic growth used in economic history by capturing the extent to which population growth of a given homeland city is affected by cities nearby. Higher values indicate that cities are surrounded by other competing (i.e., growing) cities and lower values mean that growth occurs in a sparsely populated area. Third, I recode the country fixed effects for the Ottoman, Russian, and Habsburg empires by calculating time under each of these polities. Some stateless ethnic groups, especially those in Eastern and Central Europe, underwent multiple state “owners” over the course of the five centuries. Fixed effects do not represent this change, so I calculate total years of each of these imperial rulers. The rest of the variables is time-invariant and needs no transformation.
Table 1 documents logistic and OLS estimators on language standardization by 1900 (complete results are available in the Supplementary appendix). Consistent with the results in Sasaki (Reference Sasaki2017), the variable on the printing press is significantly positively associated with language standardization. More concretely, elapsed time since the adoption of print technology increases the chances of vernacular codification by 0.6 percentage points. Similar results obtain when the OLS is used. Compared to the estimates in Sasaki (Reference Sasaki2017), the magnitude of the printing press variable is small because only 27 of 171 ethnic groups achieved language standardization by 1900. The vast majority of cases occurred in the twentieth century.
TABLE 1. Regression outputs of language standardization for ethnic groups by 1900
Source: See Clair (1976) and Febvre and Martin (1976) primarily for the sources of the printing press. For other variables and the coding strategy, see the Supplementary appendix.
Note: Western Europe is used as the reference category for region fixed effects and thus omitted. Catholic is used as the reference category for religion fixed effects and thus omitted. The Wald Chi Square is the effect of time since print adoption. R2 for logistic regression is pseudo-R2 . The complete results are available in the Supplementary appendix.
*p < 0.1. **p < 0.05 ***p < 0.01.
I use the parameter estimates of the logistic model to calculate propensity scores. Figure 2 displays the densities of the scores for two groups: those that standardized language by 1900 and those that did not. It shows that ethnic groups with standardization have largely higher scores, while those without it are typically concentrated in low propensity scores, especially in the scores between 0 and 0.25. Still, the distribution of both subsets is similar in that the scores are largely less than 0.5.
FIGURE 2. Distribution of propensity scores for two sets of ethnic groups on language standardization by 1900.
Table 2 further breaks down the distribution for 16 groups with the propensity score of 0.5 or above. It adds two columns of information: dictionary publication by 1900 and territorial sovereignty today. Two points merit attention. First, high propensity scores reflect the high likelihood of early vernacular codification. Of the 16 groups, 10 achieved codification by 1900. Second, nine groups acquire or maintain an independent state. Observations are more equally distributed between those with a state and stateless units.
TABLE 2. Ethnic groups with the propensity score above 0.5
Source: Author calculations.
Note: Theoretically counterintuitive outcomes are highlighted in bold.
This variation provides purchase for case selection with a focus on deviant cases. It raises an empirical puzzle as to why early language standardization does not always lead to state-building. This is at odds with the hypothesis discussed earlier, in which ethnic groups, once standardization is achieved, would have an incentive to build institutions and mobilize toward self-rule to increase the chances of survival. The decision not to build on this momentum appears counterintuitive. In the European context, it is a highly risky choice to the extent that it is near self-destruction because of recurring warfare and territorial conquest in the early modern period through modern times (McNeill Reference McNeill1982; Tilly Reference Tilly and Tilly1975, Reference Tilly1992; Wimmer Reference Wimmer2013). The inability to raise revenue and manpower for defense would leave ethnic groups vulnerable to political incorporation by others that comprise an independent state. However, descriptive evidence from Table 2 shows that this empirical puzzle offers a case-selection strategy based on the counterfactual framework.
Table 3 displays the relationship between language standardization and institution-building, where the ethnic groups in the table are examples drawn from Table 2. The top left cell represents the expected outcome, where ethnic groups, such as the Bulgarians, the French, and the Dutch, would invest in state institutions following early language standardization. The top-right cell denotes the counterintuitive outcome, in which ethnic minorities such as the Occitans and the Alsatians failed to build institutions for political mobilization despite language standardization.Footnote 7 Tables 2 and 3 inform case selection. The outcome in the first group may seem unsurprising given that groups like the French already comprised their own state and had resources for language rationalization and means for enforcing the vernacular. Yet they also include groups like the Bulgarians that were long under the yoke of imperial powers. They were able to codify their vernacular without first having an independent state, and only in the twentieth century did they achieve sovereignty. These are “near miss” observations, as they could have belonged to the bottom row of Table 3 if they did not achieve language standardization early. Similarly, those in the top-right cell constitute near misses for a different reason in that they missed the opportunity to acquire greater political autonomy. These borderline cases provide a rationale for detailed investigation because they could shed light on the causal mechanism in question (Lieberman Reference Lieberman2005), which in my article is the link between cultural consolidation and state-building.
TABLE 3. Two-by-two table on the relationship between language standardization and state-capacity building
Source: Table 2. The blank cells in the “Country” column indicates no sovereignty for the given ethnic group.
I first offer a brief discussion of the Bulgarian case as one that codified the vernacular by 1900 and obtained territorial sovereignty in the twentieth century. Bulgarians came late in terms of both cultural consolidation and political mobilization for independence. They did not have a strong tradition of vernacular use until the nineteenth century but managed to codify their vernacular in 1895 while under Ottoman rule (Dalby Reference Dalby1998). In a separate effort, Bulgarians achieved “religious independence” by erecting their own church in the mid-1870s. For them, the independent church served as a key institutional basis from which their elites launched the independence movement later. This example constitutes the reference category in my empirical analysis. By contrast, the Occitans serve as a deviant case that did not experience institution-building. They were among the early language standardizers by publishing the first monoglot dictionary in 1875 (ibid.). By the mid-nineteenth century, the Occitans had long lost their sovereignty, but there was a growing window of opportunity for political autonomy amid the debate over the possibility of a federalism in France. Unlike the Bulgarian counterpart, few Occitan elites were interested in building institutions and pressing for federalism—or political freedom more generally. I show in the next section that this inaction proved critical in explaining the divergent outcome between the Bulgarians and the Occitans.
Investigating the Link between Language Standardization and Institution-Building
The Case of the Bulgarians
The Bulgarians were the last of the Balkan groups under Ottoman rule to start political mobilization toward autonomy or independence. Geographical proximity to Constantinople likely played a major role for this laggedness (Jelavich and Jelavich Reference Jelavich and Jelavich1977: 128). Like many minority vernaculars, Bulgarian was a local language spoken primarily in day-to-day communications; until the late 1870s, Turkish was the administrative language and Greek was the language of service in the Bulgarian Orthodox Church (Price Reference Price1998: 42). Linguistic development initially took place outside of what is today’s Bulgarian territory. It was in Venice in 1566 that the first book in Bulgarian—a collection of prayers—was printed. Before the mid-nineteenth century, Bulgarian speakers had to rely on non-Bulgarian towns such as Belgrade, Budapest, and Bucharest for the production of printed books in their tongue (Macdermott Reference Macdermott1962: 136). Similarly, in 1838 when the officially-recognized Bulgarian printing press was set up, the print shop was built in Salonika, today’s Greece, and not in the area of Bulgarian residence (ibid.: 136). Bulgarians had to wait another eight years until the press was set up in Samokov, a town near Sofia. The absence of vernacular print in the area of Bulgarian residence was indicative of the extent to which the linguistic norm of Bulgarian, of its modern form, existed before modern times (Price Reference Price1998: 42).
Two relevant developments also promoted the standardization of Bulgarian. The first is schooling. The first Bulgarian-speaking school opened in 1835 in Gabrovo. The founder, Vasil Evstatiev Aprilov, emulated the European style that was secular, free, and open to all (Macdermott Reference Macdermott1962: 126). Literate Bulgarians were able to amass support for vernacular education among the peasantry. Despite the Greek opposition, they established primary schools across approximately 1,479 towns and villages (ibid.: 300). The second is the founding of literate societies. Among the first was the Bulgarian Letters (Bulgarski Knizhnina) formed in Constantinople in 1852. It published the Bulgarian Papers (Bulgarski Knizhitsi), which took up secular themes such as science, commerce, and agriculture (ibid.: 141). In a similar vein, Marin Drinov, a historian and philologist, helped found the Bulgarian Literary Society (Knizhovno Druzhestvo) in 1869, which later became the Bulgarian Academy of Sciences upon independence. Drinov was a major advocate of the “revival” of the Bulgarian people by writing extensively about the “historical roots” of the population through the society’s publication, Periodical Journal (Daskalov Reference Daskalov2004: 2). Despite the relatively late start, Bulgarian elites effectively used schools, literary organizations, and their publications to codify their vernacular by the turn of the twentieth century.
The critical factor that allowed the Bulgarians to move from the phase of cultural consolidation to that of state-building was the establishment of an independent Bulgarian church: the Bulgarian Exarchate of 1870. It constitutes a critical institution-building episode for the Bulgarians as a nation. Previously, ecclesiastical affairs on Bulgarian land were under Greek control. Granted by the Ottoman decree, Bulgarians in 1870 took control over their own religious affairs. The independent church gave the Bulgarians the freedom to appoint bishops to the dioceses within the newly constituted province (i.e., the exarchate). It also came with the authority to appropriate revenue from the schools, churches, and other ecclesiastical properties (Mylonas Reference Mylonas2012: 60). Furthermore, the church offered extensive organizational networks that were deeply rooted in each locale. Its hierarchical structure made policy implementation effective. Thus, the church provided a strong platform equipped with the capacity not only to organize collective action for political ends but also transmit political identities (cf. Kalyvas Reference Kalyvas1996).Footnote 8 Although religious independence was not intended as part of the movement for political independence (Jelavich and Jelavich Reference Jelavich and Jelavich1977: 134–35; Macdermott Reference Macdermott1962: 167), the church served as a crucial organizational basis through which Bulgarian elites mobilized their community toward sovereignty in the first decade of the twentieth century.
In sum, the Bulgarians in the nineteenth century achieved both cultural consolidation through language standardization and institution-building through religious independence while under Ottoman rule. These were two distinct processes and not based on a grand, coherent vision, but they nicely converged upon the decision of Bulgarians to pursue sovereignty. The independence of the church proved significant for their subsequent political movement, which, as hypothesized in the nationalism literature, made a bridge between cultural consolidation and state-building. By contrast, the Occitans followed a different path. As I show next, they did not build on their successful linguistic codification to invest in institution-building. I demonstrate that this inaction plays a key role in explaining the divergent outcome on political independence between the Bulgarians and the Occitans.
The Case of the Occitans
The geographical area of Occitania, the territory that Occitans claim as their homeland, broadly refers to the southern half of today’s France (the “Midi”). The hilly region that cuts across the middle of France historically served as the natural borders for demarcation. The south is climatically and culturally different from the north. Prior to French incorporation in the early thirteenth century, those who lived in Occitania had closer cultural ties with Spanish and Italian speakers than with French speakers (Paterson Reference Paterson1993). The most important region within Occitania is Languedoc, the southernmost region of France that faces the Mediterranean, whose political and economic centers comprise the cities of Toulouse, Carcassonne, and Narbonne (see Figure 3). This area began to be the target of political incorporation by the Capetian monarch in 1209 through his participation in what is later known as the Albigensian Crusade (1209–29).Footnote 9 The Capetian assertion of rule continued through much of the thirteenth century (Given Reference Given1990: 39). Although the struggle for control over this territory with English monarchs continued intermittently through the mid-fifteenth century, Occitania remained primarily under French rule (Beik Reference Beik1985: 37). For the Occitans, the Albigensian Crusade and the subsequent French conquest constituted the watershed because any prospects for independence began to disappear with the incursion (Marvin Reference Marvin2008: 306). For the Capetians, annexing the large territory of Occitania provided a new major source of income (Strayer Reference Strayer1971: 46–47).
FIGURE 3. Region of Occitan speakers at the time of the French Revolution. Note: The Occitan-speaking region is denoted by gray and Languedoc, a province within the Occitan region, is denoted by black. The white dot within Languedoc represents Toulouse.
In the wake of the Crusade, the Capetians installed what may be described as “indirect rule,” in which local institutions were retained in preference to the transplantation of royal institutions and local aristocrats managed to preserve and exercise limited authority over legal and economic affairs (Given Reference Given1990: 57–63). One rationale is that this is a typical governance strategy in Europe from the Middle Ages until around 1750 because the political centers like Paris were too weak to project power over their nominal domain (Elliott Reference Elliott1992; Koenigsberger Reference Koenigsberger1987). It is an inexpensive yet practical compromise that ruling monarchs were willing to make so long as the autonomous local elites pledged loyalty in terms of taxation.Footnote 10 Another reason that indirect rule was preferred is because of the differences in law, society, and culture between northern France and Occitania, which made direct rule too costly at the time (Strayer Reference Strayer1971: 51).
Beginning in the early fifteenth century, a new institution of indirect rule was established. Languedoc, along with other departments in the countryside, was accorded with the status of a pays d’états, one of the two types of administrative jurisdiction (provincial estates) under the Ancien Régime.Footnote 11 It is a department-level administrative category for the purposes of taxation. Residents of this status possessed the right to hold representative assemblies, where elites could negotiate tax burdens with the monarch and collect taxes on their own (Kwass Reference Kwass2000: 70). This institution thus allowed the residents to retain a degree of self-government in terms of rights and privileges. Languedoc, whose provincial capital was located in Toulouse, convened the assemblies roughly once a year from 1418 until 1789 when all the Ancien Régime institutions were abolished.Footnote 12 In addition to the matters of taxation, attendees introduced legislation and debated budgetary allocation on public goods (Swann Reference Swann and Doyle2012). The assemblies of Languedoc enjoyed greater discretion over the regional affairs than those of other regions, such as Brittany, Burgundy, and Artois, and as a result, the residents were able to preserve strong regional identities (Beik Reference Beik1985: 38).Footnote 13
A degree of self-rule implies that the Occitans would enjoy the freedom to use and develop the vernacular as the primary means of communication. It seems that the vast majority of them knew little French and, especially in Languedoc, spoke virtually no French throughout the early modern period. Although direct evidence is scant, this linguistic situation can be inferred from the 1863 survey, which shows that few Occitans throughout the region spoke French words well into the late nineteenth century (Weber Reference Weber1976: 68). By contrast, a small group of literate Occitans learned French (Cohen Reference Cohen2000). They had a stronger incentive to master French rather than developing the Occitan language because the ability to use the language of the political center allowed them to negotiate directly with the monarch over taxation at the provincial assembly. These literates could take advantage of the language market to gain access to the information that was available only in French and advocate the tax regime to their own benefit (Laitin Reference Laitin1988). Moreover, from the sixteenth century on these local elites formed a marriage of convenience with Parisian bureaucrats so that local elites could retain their privilege and the central government could raise revenue at the expense of the peasantry (Bates Reference Bates1988; Root Reference Root1987).
The nineteenth century saw sort of a Renaissance in intellectuals’ attitudes toward local history and culture. Influenced by German Romanticism, scholarly interest in the Occitan language and the history of southern France grew. Both French and Occitan authors took up this effort. For instance, Jean-Bernard Mary-Lafon, a historian, dramatist, and linguist, published a series of influential works on the history of the Occitan region that recounts not just politics but also religion and literature (Roza Reference Roza2003: 59). In addition, there were French-language works on the Troubadours, the poets and singers during the “golden age” of the Occitan culture developed in the thirteenth century, and on Occitan grammar books. Occitan literates also composed poems in the vernacular. These poems described the lives of the working-class people in the region and gained popularity among regular Occitans, as Occitan poets, who also belonged to the working class as bakers, hairdressers, and wig makers, traveled across the region (ibid.: 61). This growing intellectual effort to understand the history, language, and culture of the Occitans culminated in 1875 in the publication of the first comprehensive vernacular dictionary, Dictionnaire analogique et étymologique des idiomes méridionaux (Explanatory and etymological dictionary of the southern dialects). This 1,344-page volume was edited by Louis Boucoiran, a little-known historian originally from Nîme. To be clear, the publication was not tantamount to the unification of various Occitan dialects. Although the growing investment in and the popularity of the Occitan language was a reaction to the parallel—and larger—effort on the standardization of French by the Third Republic, it seems that the Occitans had no political entrepreneur who took the effort to unify the Occitan tongues (Smith Reference Smith1973). Vernacular codification took place, but there was little institution-building to launch a political movement.
In the mid-nineteenth century, Occitans did mobilize for cultural revivalism. In 1852 they founded the Félibrige, a literary society, primarily to promote the greater use of the Occitan language throughout the Midi or the Occitan homeland. The most prominent member was Frédéric Mistral, a poet and one of the seven founding members, who wrote not only a series of poetry but also edited two-volume dictionaries in the late nineteenth century. His literary work was recognized with the award of the Nobel Prize in Literature in 1904. The appellation, Félibrige, was derived from felibre, the Occitan word for a doctor which, in turn, denotes “teacher” and “sage” (Roza Reference Roza2003: 74). The motive of the organization is epitomized in Article 1 of the mission statement:
The Félibrige is established to forever preserve for Provence its language, its color, freedom for its grace, its national honor and its great ranks of intelligence, for it is Provence that agrees with us. By Provence, we mean the entire Midi of France. (ibid.: 96)
Joseph Roza (Reference Roza2003: 65) discerns two concrete goals from the record of the organization’s activities. The first is to remove the social stigma attached to using the Occitan tongue; the second is to secure survival as a literary language while Occitan was widely spoken within southern France. Contributing members sought to achieve these goals through publication on a regular basis and participation in social events.
The most important publication that the Félibrige put together was the Armana Prouvençau, an almanac. First released in 1855, this annual publication contained a collection of news, poems, and essays in Occitan. Literates decided on the almanac as an ideal outlet because it would not only convey their thoughts in the vernacular but also chronicle events that were important specifically to the Occitans (ibid.: 76). The almanac’s first edition was printed for around 500 copies, but it quickly gained popularity so that it sold up to 3,000 copies by 1872 and 10,000 copies annually in the decade between the 1880s and the 1890s (ibid.: 83). With the Armana Prouvençau, Mistral sought to encourage lay readers to use Occitan in everyday communication without fear of being chastised not to use French. He argued:
From a habit of imitation and bourgeois prejudice—which unfortunately is spreading more than ever—people had become accustomed to avoiding as “vulgar” the words most ingrained in Provençal speech, and … commonly and quite uncritically used the corrupt, bastardized, Frenchified forms that were spoken in the streets. Once we had made up our minds to write in the language of the people, [the Félibre] had to bring out and set off to advantage its characteristic bluntness, vigor, and richness of expression, and we agreed to write the language purely as it was spoken in places free of outside influence. (Mistral Reference Mistral1986: 81)
When the Félibrige worked to solidify its foothold, the political and social environment in France was an auspicious one. In 1868, Napoleon III lifted the imperial decree of 1852 on the press, which gave an impetus to publish on politically sensitive topics. More important, the mid- to late nineteenth century was a time for great uncertainties in France. There was a constant struggle for the constitutional system between monarchism and republicanism. The uncertainties were especially acute in 1870 following the sudden, unexpected defeat in the Franco-Prussian War. These created room for new ideologies, which some political entrepreneurs exploited to champion their causes. In the new Third Republic that began in 1871, several parties with new ideologies fought for supremacy to fill this ideological space (Hanson Reference Hanson2010: ch. 4). Ambitious leaders of the Savoyards and the Bretons within France stoke nationalist sentiments among their respective populations for self-determination. At the same time, the political uncertainties gave other emerging political actors a pause. For example, Catholic elites were considering forming a Christian party and mobilizing lay followers to secure their influence in the government. They feared that if “secular” Republicans came to power for a prolonged period, this would threaten the political power that the Church enjoyed under centuries of monarchical rule. The Catholics eventually chose not to form a party as they predicted—incorrectly—the republicans in power not to last long (Kalyvas Reference Kalyvas1996: ch. 3). The uncertain environment thus offered a historic opportunity for new ideologies to arise. This period was also prior to the introduction of free, mandatory primary education. Thus, in this context the Félibrige could have taken advantage of the instability in the political center to mobilize lay Occitans, invest in their own institutions, and promote their cause.
The political uncertainties about what type of the constitutional structure would emerge and how long it would last afforded a historic opportunity to demand a change in governance arrangements. More specifically, demographic minorities seeking greater self-rule may pursue autonomy or some form of a federal system that can guarantee political freedom. In France, indirect rule gave way to more direct rule following the Revolution. In the initial stage of the process, Jacobin leaders improvised an institutional arrangement that granted commercial centers like Marseille and Lyon with regional autonomy. However, as the revolutionary government took shape, it instead installed a centralizing system of direct rule on key dimensions of governance like taxation, policing, and public goods across the country (Tilly Reference Tilly1992: 109–10).Footnote 14
The transition period between the Second Empire and the Third Republic in the 1860s through early 1870s provided another opportunity for institutional change. Extant scholarship hypothesizes that a federal system is likely to emerge when the gap in military capacity between the core unit and the peripheral units is small (Riker Reference Riker1964). Alternatively, a country may adopt federation if subordinate units had the institutional capacity to enforce rules and administer policy within them and can raise revenue on behalf of the political center (Ziblatt Reference Ziblatt2004, Reference Ziblatt2006). Given that the political and military turmoil during this period left the state capacity of Paris relatively underdeveloped, it could give minority groups like the Occitans an incentive to build indigenous institutions to strengthen bargaining leverage and demand concessions from the Parisian authorities (Gerring et al. Reference Gerring, Ziblatt, van Gorp and Arévalo2011: 380).Footnote 15
Moreover, a group of French intellectuals supported federation as a constitutional foundation in the event of the collapse of monarchism toward the end of the Second Empire. Federalism emerged as one of the three major ideological strands among the Parisian intelligentsia.Footnote 16 French federalists used the commune as the primary political unit and highlighted unity based on liberty, equality, and loyalty to the state (Hazareesingh Reference Hazareesingh1998: 284). At the same time, this ideology advocated democracy as it espoused consensus and self-governance among the communes. As Auguste Vermorel, a social republican, argues: “[Unity] should result from the free will of groups and their harmonious adhesion to common principles, which are the guarantee of the greatness, independence, and prosperity of the nation. It is the communes … that should direct it” (ibid.). Decentralization was deemed preferable to state centralization because advocates believed that support for republicanism still remained weak in the countryside and that federation would compensate for this reality (ibid.: 285). Although federalists eventually gave way to Jacobinists who promoted centralization under the Third Republic (Hanson Reference Hanson2010: 101–2), federalism was a viable ideological force in mid-nineteenth century France.
In this political climate, the Félibrige continued to expand in the first decade of the Third Republic. It established the new branch in Paris, through which to interact with French intellectuals on a regular basis and expand political and social networks. The new office was also used to raise public awareness about the linguistic situations in the south and garner support for the Occitan language. This strategy initially proved effective in securing the right to use the vernacular in higher education. In 1875, Occitan literates placed the petition in the National Assembly to restore the teaching and study of Occitan at the Universities of Aix, Toulouse, and Montpellier. In three years it was approved for Montpellier (Roza Reference Roza2003: 242). They could have taken advantage of this initial win to reinforce their cause across a wider circle of Occitan speakers. However, this was the apogee of institution-building by the Félibrige, and it seems that the hope of the greater use of the language through university education petered out in the mid-1870s.
One important reason is the absence of leadership. There is little evidence that suggests that the Félibrige sought to transform the literary society to launch a political platform. It appears that it was a deliberate decision at the organizational level not to seek regional autonomy or independence. Through the interactions in Paris, Occitan literates concluded that overt political activism for language rights would make too high risks to bear in an industrializing France where proficiency in French was not only a major political goal but also a highly desirable skill for productivity and upward mobility. In this environment that was adverse to cultural diversity, the Félibrige highlighted the importance of using Occitan, or non-French languages more broadly, “to improve students’ knowledge of French” because French was not a native tongue for most pupils in the south (ibid.: 248). This defensive strategy reflected the Félibrige’s belief that French policy makers would never accept bilingual education in Occitan and French, let alone Occitan monolingualism in the Occitan-speaking region.
After the summer of 1886, the Félibrige abandoned its goal to press for language rights and promote the greater use of Occitan in France in favor of a more modest objective of language preservation. In April 1887, Alphonse Roque-Ferrier, a prominent member, contributed an article in the Montpellier daily, Messager du Midi:
The goal of the Félibrige is then the maintenance of this language that we do not have to learn in prep school, and the desire to see it spoken with the same purity as the language of Corneille, the first serving the teaching of the second, and the second serving the teaching of the first. (ibid.: 250)
This trajectory suggests that the lack of institution-building among the Occitan elites lies primarily in the absence of leadership to pursue formal rights to use the Occitan language at school or political activism for greater autonomy.
The Félibrige’s in action can be further attributed to two factors. The first is the generational divide. The Félibrige was never a monolithic group; on the contrary, there were significant differences in political goals across generations. Young members embraced the ambition of a federal France because Paris’s initiative of institutional centralization was perceived as promoting cultural homogeneity and thus a grave threat to the Occitan culture. For instance, Paul Meriéton, a young Félibre and one of the most vocal members calling for federation, argued that centralization would lead to “vaingloriousness” and the “depravation and death” of the Occitan culture that his organization was seeking to revive (Roche Reference Roche1954: 27). In February 1892, he put forward his political goal in the “Déclaration des Félibres Fédéralistes,” addressed to the president of the literary society:
Yes, we want an Assembly at Bordeaux, at Toulouse, at Montpellier; we want one at Marseilles or at Aix. And these Assemblies must control the administration, the courts, the schools, and universities, as well as public works. (ibid.: 64)
The statements by Meriéton and other young members who shared the view were soon printed in newspapers in both Occitan and French and widely circulated. They raised the suspicion from Paris that the Félibrige was harboring the intention to pursue not just federation but ultimately separatism, the charge that older members had carefully avoided (Wright Reference Wright2003: 46–47).
Older generations, however, remained modest about the political ambitions for the literary society. As research on intellectual history finds, Frédéric Mistral, a founding member and the inaugural president, always equivocated the questions about the group’s political goals or the politicization of its activity. While he seemed sympathetic to younger members’ call for political activism, he neither condoned nor reproved such demand and left virtually no writings on politics (ibid.: 20, 47). It seems that the primary goal of the older generations was a renaissance of what they perceived as the distinct civilization that had achieved its apex prior to the Albigensian Crusade. Evidence from older generations’ remarks suggests that the victory of the Crusaders constituted the “sources of spiritual life dried up.” As Mistral noted, “[B]ecause, what really subdued, note it well, was less the South materially speaking, than the spirit of the South” (Roche Reference Roche1954: 52). The Félibrige’s leadership consistently reminded that its purpose was cultural preservation, and Mistral maintained that only through loyalty to la grande patrie (the big, native land of France) could Occitans’ language and culture achieve “communal glory” (ibid.: 65).
The second reason for the lack of political mobilization is the internal discord within the literary society. Julian Wright (Reference Wright2003: 48) points out that the generational divide was found not just over the organization’s political goals but also over the politics within the organization. Elections, titles, and statutes were main sources of the fight. The leadership positions and titles were scarce goods, and officeholders or titleholders could exercise disproportionate influence over the group’s political future. For example, these members can ultimately decide whether to pursue greater language use in school instruction or whether to form a political party. Given that the Félibrige was the leading literary society of the Occitans, access to these goods would be highly desirable, even essential, if some members espoused the ambition for self-rule. The presence of bureaucratic politics suggests that founding members occupied senior positions while denying younger members access to them who were dissatisfied with their leaders’ apolitical attitudes toward group activity. The 1892 declaration of the demand for federalism could be characterized as an outburst of the internal discord.
The generational divide eventually led to the split of the organization. Meriéton and other like-minded young members left the Félibrige to join an existing political party or found a new one. The breakup was the harbinger of the more formal—yet, again, unsuccessful—political movement of regionalism, most notably the Fédération régionaliste française, that arose in the twentieth century. Occitan speakers led the way by inheriting their ideological inspirations from the federalist strand of the constitutional debate that French intellectuals had toward the end of the Second Empire (Beer Reference Beer1980; Wright Reference Wright2003).
For the Félibrige, the turn of the century was the period of sharp decline as a literary society. More and more school-age children went to school and studied only in French in preference of Occitan (or other non-French languages spoken at home) (Weber Reference Weber1976). Vocal young members quit the organization. Senior members, by contrast, achieved a modest success at securing the right to teach Occitan at one university but largely abandoned the strategy of persuading the Parisian counterparts to recognize bilingual education in favor of a more moderate, practical goal of language preservation. As founding members aged and began to die, the organization’s activity shrank in terms of the volume of publications and the number of public events (Roza Reference Roza2003: 274). Although the Félibrige has still been in existence today, their role has remained the same—language and cultural preservation.
Conclusion
This article explores whether elites of ethnic minorities build institutions for political mobilization after cultural consolidation. My innovations are to provide a measurement of cultural consolidation through language standardization and to choose cases among stateless ethnic groups for in-depth investigation based on the degree of vernacular codification by 1900. I use process tracing to compare the successful case of the Bulgarians to the failed case of the Occitans. My empirical analysis indicates that while Bulgarian elites established an independent religious institution through which to win mass support for political autonomy, the Occitan counterparts did not make a similar action and ended up content with cultural preservation. My main finding is that not all ethnic groups invest in institutions following language rationalization. I show that ethnic elites’ indecision or perceptions of taking political risks accounts for this seemingly counterintuitive finding.
One of the broader implications of this study is that the connection between cultural consolidation and state-building may not be as strong as suggested in the literature on nationalism. My article proposed institution-building as one intermediate category within this link. Future research could identify others for a sharper conceptual framework. For instance, one might ask if ethnic elites call a deliberative assembly as part of their effort to foster state capacity and eventually to win sovereignty. Extant scholarship shows that some semblance of state institutions is required for aspiring ethnic groups to be nation-states (Roeder Reference Roeder2007), but what causes the development of state capacity has yet to be examined. This development would be a useful outcome variable because it constitutes a precursor to the development of a sovereign state. This approach opens a new avenue for empirical analysis that leads to the better understanding of why some ethnic groups consolidate culture and win achieve sovereignty more effectively than others.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/ssh.2019.6.
