There is nothing inherently violent in nationalist ideology. As the Catalan and Quebecois cases on the one hand and the Basque and Chechen on the other clearly demonstrate, there are “nationalisms that bark and nationalisms that bite” (Beissinger Reference Beissinger and Hall1998). Nevertheless it seems that in some parts of the world nationalism is exclusively identified with bellicosity. The popular perceptions of the BalkansFootnote ¹ are of a region prone to rampant nationalism and incessant violence. Such a stance was already present in the early 19^th century when the image of the Balkans gradually replaced that of the Ottoman Empire as the chaotic and uncivilised Other. While the wars of independence against the crumbling “sick man of Europe” were generally greeted with enthusiasm in the West, the attempts by the newly independent Balkan states to establish genuine autonomy from the Great Powers at the end of the 19th and early 20th century provoked profound animosity in the major European capitals. This view is best illustrated by the Evening Post’s depiction of Serbia following the assassination of King Aleksandar Obrenović in 1903:

Interestingly enough such perceptions of the region were often shared by the inhabitants of the Balkan Peninsula, and in particular the intellectuals, who were just as much prone to auto-stereotyping and perceiving their region as intrinsically violent, ethnocentric, barbaric and characterised by a distinct “Balkan mentality” (Cvijić, Reference Cvijić1922; Dvorniković Reference Dvorniković1939).

This alleged ubiquitous violence and entrenched nationalism are often singled out as the principal causes for the region’s economic and political backwardness (Berend Reference Berend2003; Gerolymatos Reference Gerolymatos2002). For example, in his 1993 introduction to the reprinted Carnegie Endowment report on the Balkan Wars of 1912-13 George Kennan directly links the violent conflicts in the region at the beginning and at the end of 20^th century arguing that the lack of development in the Balkans was not rooted in religious differences but in “aggressive nationalism […] [which] manifested itself on the field of battle, drew on deeper traits of character inherited, presumably, from the distant tribal past […] and so it remains today” (Kennan Reference Kennan1993, pp. 4-6).

This paper challenges both of these popular perceptions. Firstly I argue that when compared to the rest of the European continent, South East Europe has experienced less organised violence and significantly less nationalism for much of its history. Secondly and more importantly it was the absence of coherent, popular nationalist ideologies and protracted inter-state warfare that have often proved to be a crucial hindrance for intensive social development in this part of Europe. The core argument combines theoretical and empirical aspects. The first section engages with the bellicose tradition in historical sociology that sees the state as a direct product of warfare and it questions the blanket proposition that ‘wars make states’ in the context of South East European history. I argue that the conventional bellicose argument requires a serious qualification as it seems unable to explain persistent weakness of the state in the 19^th and early 20^th century Balkans. The second section focuses on the relationship between organised violence, state formation and nationalism. It attempts to rehabilitate the classical modernist theories of nationalism, and in particular Gellner’s approach, which are often criticised for not being able to account for the alleged emergence of nationalist aspirations in the early 19^th century Balkans. In contrast to the mainstream Balkan historiography, I argue that there was very little if any nationalism in South East Europe for much of the 19^th century and for large sections of population nationalism was a marginal ideology well into the 20^th century.

South East Europe and the Bellicose Historical Sociology

The idea that wars make states has a long tradition in social sciences, expressed by numerous theorists: from Gumplowitz (Reference Gumplowicz1899), Ratzenhofer (Reference Ratzenhofer1881), Ward (Reference Ward1913), Oppenheimer (1926), Hintze (Reference Hintze1975 [1908]) and Rüstow (Reference Rüstow1980 [1950]) to the more recent theories of Carneiro (Reference Carneiro1977), Downing (Reference Downing1992) Ertman (Reference Ertman1997) and Gat (Reference Gat2006). This bellicose historical sociology gained momentum with Tilly’s (Reference Tilly1975, Reference Tilly, Evans, Rueschemeyer and Skocpol1985, Reference Tilly1992) sophisticated reformulation which emphasised the broader geo-political context in the changing character of imperial inter-state warfare in early modern Europe. For Tilly protracted wars were the principal catalyst of state transformation as they ultimately enabled the development of the omnipotent bureaucratic apparatuses, effective revenue systems, state wide juridical control, integrated regional administration and substantially better financial infrastructure. Furthermore, the constant war-making increased demand for greater military and economic mobilisation of ordinary people whose material resources and military participation became decisive for the long term survival of warring states. The unintended consequence of these changes were gradual but steady payoffs that the state rulers had to make to their subjects including wider citizenship rights, expanding parliamentary systems, religious freedoms, and social protection, all of which also encouraged the development of civil societies.

The west European multi-polar order prevented the emergence of a single empire, thus creating a semi-anarchic environment where rulers were gradually forced to rely on the broader sectors of the population under their control. Hence to effectively fight external threat rulers often embarked on pre-emptive wars while simultaneously trying to neutralise domestic forms of violence. To finance ever-more-costly warfare in the context of rapid military transformations (with the invention and mass manufacturing of more destructive weaponry and dramatic improvements in transport and communication systems) rulers had to constantly increase resource extraction and introduce universal conscription in the territories under their control. As Tilly (Reference Tilly, Evans, Rueschemeyer and Skocpol1985, p. 172) emphasises, this process was highly contingent with states emerging as a corollary of war making, extraction of resources and capital accumulation. In other words to extract money, people and material it was necessary to subdue and disarm internal rivals and defeat external foes. In this process state rulers operated in a similar way to gangs who offer security in exchange for regular financial payment. That is, the state developed as an institutionalised and legitimised large scale protection racket:

Therefore, the bellicose tradition argues, the modern centralised political entities that successfully claim monopoly on the use of violence over their territory, which we call nation-states, were a direct by-product of the intensification in west European war-making from the 16^th century onwards.

This bellicose interpretation has provoked a great deal of attention: it has been applied to various parts of the world with a focus on the structural differences between Western Europe and the rest of the world. For example, Lustick (Reference Lustick1997) interprets state instability in the Middle East through the prism of earlier state development in Europe whereby better organised and more powerful European states colonised the region and in this way averted the emergence of ‘state-building wars’ in the Middle East. Barnett (Reference Barnett1992) provides a comparative analysis of Israel and Egypt arguing that the relationship between war and state-making is highly dependent on the different strategies of state building pursued by the rulers in the two countries. In contrast, Tin-Bor Hui (Reference Tin-Bor Hui2005) pinpoints striking similarities between China in the Spring and Autumn and Warring States periods (656-221 BCE) with early modern Europe: whereas both regions experienced the proliferation of warfare, development of centralised bureaucracy and expansion of trade they nevertheless eventually experienced different outcomes. Taylor and Botea (Reference Taylor and Botea2008) contrast the impact of war on state development in Vietnam and Afghanistan arguing that the relative ethnic homogeneity was a decisive factor contributing to a state making war in Vietnam and a state destroying war in Afghanistan. Centeno (Reference Centeno2002) and Herbst (Reference Herbst2000) have compared the experience of state formation in early modern Western Europe with that of Latin America and Africa respectively, pointing to different organisational trajectories in these two continents where, in contrast to West European protracted wars and strong states, sporadic and limited warfare generated weak and internally polarised states.

The general tendency among scholars is either to endorse or disprove the “war makes states” thesis. Some authors such as Thies (Reference Thies2007), Ayoob (Reference Ayoob1995) and Herbst (Reference Herbst2000) side with Tilly in the argument that the proliferation of protracted inter-state warfare is conducive to the development of pervasive state apparatuses in other parts of the world such as the Middle East, Africa, or South Asia. Others (Leander Reference Leander, Guzzini and Jung2004, Reno, Reference Reno, Davis and Pereira2003, Kaldor Reference Kaldor2001) have tended to be much more critical of this assumption arguing that the historical context has significantly changed: in the post WWII international system all attempts at unilateral border change are rapidly delegitimised which creates a situation where inter-state warfare is on the wane while intra-state (civil) wars, conflicts that destroy state capacity, proliferate.

The Balkan case remains something of a puzzle as it does not easily fit into either of the two competing perspectives. At first glance it might seem that the South East European experience quickly refutes the bellicose thesis as the common perception of the region is one of relentless civil warfare and weak states. However careful historical sociological analysis shows otherwise: for much of its history the Balkans have not experienced more violence than other parts of Europe. The rulers of the medieval kingdoms of Serbia, Croatia, Bosnia, and Bulgaria waged frequent wars with the Byzantine Empire and between each other but in terms of intensity, human casualties and their organisational features, these wars were no different from other medieval conflicts fought in the same period throughout the European continent (Nicholson Reference Nicholson2004; Keen Reference Keen1999: Fine Reference Fine1994). Just as all wars in the feudal age, these conflicts were essentially small scale, low casualty ritualistic affairs with very few proper and lasting battlefields. “As the knights represented the core of all mediaeval forces, armies were quite small and expensive while direct battles were generally avoided whenever possible making regular warfare no more than plundering expeditions” (Malešević Reference Malešević2010, p. 107). Once the Ottoman Empire conquered much of South East Europe the entire region was successfully pacified for centuries. While some areas in the Balkans were sites of inter-imperial wars between Ottomans, Habsburgs, Venetians and the Russian EmpireFootnote ² for much of the three centuries (16^th-18^th) this region has nevertheless experienced substantially less organised violence than other parts of Europe. With the gradual weakening of the Ottoman Empire in the early 19^th century the Balkans became a zone of interest for the so called Great Powers whose political and cultural elite encouraged violent resistance to Ottoman rule. From Chateaubriand and Flaubert to Lord Byron, Shelley and Goethe South East Europe and Greece in particular were suddenly rediscovered as the exotic cradle of European civilization (Todorova Reference Todorova1997, pp. 89-115).Footnote ³ It is only in this period that one can witness the acceleration of organised violence in the region. Nevertheless, this is hardly unique for the Balkans as scientific, technological and organisational developments made 19^th century warfare much more deadly all over the world: for example, the global war casualties for the combined 16^th and 17^th centuries amounted to less than 8 million whereas this figure of war deaths jumps to over 19 million for the 19^th century alone (Eckhardt Reference Eckhardt1992, pp. 272-273).

In this context the organised violence in the Balkans appears miniscule when compared to the intensity and scale of destruction and human casualties resulting from wars, revolutions, uprisings and industrial conflicts in the large and powerful European states. For example while the French and British polities were fighting wars, colonising the globe and dealing violently with various revolutions and numerous uprisings throughout the 19^th century – with Britain involved in more than sixty major wars – during the same period South East Europe experienced only six violent conflicts and five major (peasant) rebellions (Clodfelter Reference Clodfelter1992) (see table at the end of the text). Furthermore, while at the beginning of the 19^th century much of Europe was engulfed in the extremely destructive Napoleonic Wars with casualties totalling no less than six million, the Balkan uprisings of the same period amounted to a few thousand casualties (Clodfelter Reference Clodfelter1992, p. 322, Biondich Reference Biondich2011). With the clear exception of the Greek War of Independence (1821-9) all the 19^th century Balkan wars and uprisings were small scale conflicts involving several thousand casualties.

Nonetheless the Greek War of Independence was far from being a domestic Balkan affair. It was fostered and supported by the Great Powers and it was their military, economic and political involvement that sealed “Greek” victory as not a single Greek soldier took part in the battles (such as that of Navarino, 1827) that decided the fate of the war. The large casualty rate was primarily a product of highly disorganised infighting among the supposedly unified and “nationally conscious” Greeks and the civilian massacres by various brigand forces (Roudometof Reference Roudometof2001; Mazower Reference Mazower2000).

Early 20^th century warfare was significantly more destructive with two Balkan wars (1912-13 and 1913), World War I and the Greco-Turkish war of 1919-22 defining the character of organised violence in South East Europe for this period. The 1912-13 Balkan wars were particularly important as their scale, intensity and outcome had a profound impact on the perception of the region in the West. Although the massacres of the Greek War of independence dented the until then popular view of the Balkans as a land of “noble savage Christians” who needed to be saved from the “Turkish yoke” and brought back to their ancient Greek heritage, it was really the unexpected context of the Balkan wars that completely changed popular perceptions of the region in the West. As Todorova (Reference Todorova1997, pp. 122-139) convincingly argues, and demonstrates, “violence as the leitmotiv of the Balkans was strictly speaking, a post-Balkan wars phenomenon”. Whereas the start of the First Balkan War was still largely interpreted in Western media through the prism of liberation from the Ottomans (as it involved organised alliance of the small Balkan states against the Ottoman Empire) the swift and comprehensive military victory of the Balkan League changed this perception.Footnote ⁴

The outbreak of the Second Balkan War with the former allies now fighting each other over the former Ottoman possessions provoked outrage in the capitals of the Great Powers. The fact that for the first time in modern history the political elites of the small Balkan polities were in a position to pursue their own geo-political interests and in this process largely ignore the wishes of the Great powers infuriated the political establishment in the major European capitals. In addition, the unprecedented scale of the conflict came as a shock to most external observers. The armies of the Balkan states were well equipped with modern weaponry, well trained in the most recent military doctrines and strategies, relying heavily “on the ideas of the French Colonel Louis de Grandmaison to carry out the attack quickly and in force”, and were able to mobilise hundreds of thousands of soldiers (Hall Reference Hall2000, pp. 15-18). The direct consequence of this speedy, comprehensively organised, well equipped and thoroughly armed mass mobilisation was mass destruction on a scale not seen before in this part of the world. The total casualties of the two Balkan wars amounted to over 150,000 people (Singer Reference Singer1972; Eckhardt Reference Eckhardt and Sivarded Leger1988), the losing sides (Ottoman in the first and Bulgarian in the second) had substantial territorial losses while the winners (Serbia, Greece, Montenegro and Romania) had, in most instances, more than doubled their pre-war territories – i.e. Serbia by 81 %, Greece by 64 %, Montenegro by 62 % (Biondich Reference Biondich2011, p. 78).

This particular outcome shocked the political and cultural elites of major European states. The highly violent struggle for territory which was now justified in explicit nationalist terms reinforced the perception that the Balkan region is an eternal epicentre of aggressive nationalism. This attitude was already present in the first academic analysis of the two Balkan wars with American scholar Jacob Schurman (Reference Schurman1914, p. 47) projecting war aims into the past: “For ages the fatal vice of the Balkan nations has been the immoderate and intolerant assertion by each of its own claim [for territory] coupled with contemptuous disregard of the right of others”. A very similar attitude followed the just as devastating Greco-Turkish War (1919-22) that ultimately helped entrench the perception of the region as being characterised by incessant violence and xenophobic nationalist aspirations.Footnote ⁵

Nevertheless rather than being an omnipresent feature of Balkan history this sudden eruption of organised violence was a completely novel product of intensive 19^th century state building, largely based on Western European models. Unlike other parts of Europe before the mid-19^th century the Balkan region did not have a history of large scale protracted violent conflicts as its polities did not possess even rudimentary organisational means to initiate and wage such conflicts. The direct legacy of the Ottoman Empire was an almost non-existent civil service, poor transport and communication networks, no significant urban centres, dispersed and haphazard power and military structures, undeveloped banking and commerce, no proper legal system or any significant industry.

At the beginning of the 19^th century the new Balkan polities had virtually no bureaucratic apparatus: in 1813 Serbia had only 24 civil servants, the entire Dunabian Principalities (present day Romania) less than 1,000 office-holding boyars, whereas even as late as 1878 the whole of Bosnia and Herzegovina was administered by only 120 civil servants (Pavlowitch Reference Pavlowitch1999, p. 31; Glenny Reference Glenny1999:268; Stokes Reference Stokes1976, p. 4). In addition, no South East European polity had a national bank, factory, railroad or town with more than 30,000 inhabitants until well into the second half of the 19^th century. The mountainous terrain and tiny network of paved roads (by the mid-19^th century Greece had only 168 and Serbia just under 800 km) made transport extremely cumbersome and slow (Roudometof Reference Roudometof2001; Stoianovich Reference Stoianovich1994).

The lack of state development made violent conflict difficult; the Greek War of Independence (1821-9) and the two Serbian uprisings (1804-13; 1815-7) were largely chaotic, highly contingent events fought by disorganised and poorly armed units consisting of local notables, foreign trained volunteers and banditry with no proper military instruction (Glenny Reference Glenny1999; Meriage Reference Meriage1977). Such weak polities could not fight large scale protracted and destructive wars. Hence the sporadic and disordered violence that characterised the conflicts of the early 19^th century Balkans could not, and did not, create strong states. Even in the case of the region’s most intense and lengthy war, the Greek War of Independence, the war experience did not result in substantially enhanced state capacity of the new Greek polity. Thus the idea that warfare by itself can automatically create potent states has to be questioned. Nevertheless although this particular outcome goes against the general premise of the bellicose approach – as developed by its forbearers such as Gumplowitz (Reference Gumplowicz1899) or Oppenheimer (Reference Oppenheimer2007 [1926]) or some recent articulations (Gat Reference Gat2006, Carneiro Reference Carneiro1977), it does not falsify Tilly’s thesis completely as his argument emphasises that low intensity warfare is less likely to increase the organisational capacity of the state. The fact that the Greek War of Independence was militarily and politically decided by the Great Powers, and that the conflict also had all the hallmarks of a civil war, complicates this issue further.

A much better testing ground for the bellicose argument in general and Tilly’s thesis in particular is the end of the 19^th and beginning of the 20^th century when one can observe a dramatic increase in the development of the organisational and infrastructural powers of states in the Balkans. Modelling their state apparatuses on the French, Belgian and Prussian example, political elites in South East Europe managed, in a very short time, to build potent state and war machines. The size of the civil service changed beyond recognition: what started off as a handful of administrators in the early 1800s grew to hundreds of thousands by the end of that century. For example, even in 1837 the Serbian civil service consisted of less than 500 administrators while by 1902 over 22 % of all Belgrade households were inhabited by civil servants and their families. The administrative apparatus in Greece and Bulgaria underwent event greater transformation so that in less than 20 years the Greek state administration grew by 43 % and by the 1930s civil servants and their families constituted more than 650,000 in each case which amounted to between a third and fourth of the country’s entire urban population (Roudometof Reference Roudometof2001, pp. 156; Stoianovitch Reference Stoianovich1994). In a similar vein transport and communication networks expanded dramatically with much better roads, railway systems, and commercial shipping (i.e. Greece): between 1885 and 1912 the total length of railway lines increased by 841 % in Bulgaria, 613 % in Greece and 285 % in Serbia (Lampe and Jackson Reference Lampe and Jackson1982, p. 211).

By the end of the 19^th century all governments in the region were preoccupied with the greater centralisation of state power. To achieve this objective their constitutional arrangements were modelled on the highly centralised Prussian (1850), Belgian (1831) and French (1830) constitutions (Pippidi Reference Pippidi, Van Meurs and Mungiu-Pippidi2010, p. 125). Furthermore, most states adopted the hierarchical top down models of internal organisation often imitating French-style district prefects and substantially enhancing the position of the top administrators. The dominant view among the political and cultural elites was that successful state building requires excessive centralisation. As the one time Bulgarian minister of education and Czech intellectual, Konstantin Josef Jireček, put it: “Bulgarian politicians wished to arrange an omnipotent state machine on the French model, ruled by a centre with thousands of officials paid by the state and depending on it and the political parties” (Bechev Reference Bechev, Van Meurs and Mungiu-Pippidi2010, p. 142).

The administrative expansion was paralleled with the intensive development of the military and police. Military budgets increased substantially, together with the size of the armies and officer corps. For example, at the beginning of the 20^th century one third of Bulgaria’s annual budget was allocated to the military, and other states in the region had a very similar situation (Pelt Reference Pelt, Van Meurs and Mungiu-Pippidi2010, p. 240). The introduction of universal conscription meant that within two or three decades the militaries of the Balkan states expanded considerably: in 1903 the Serbian army was four times larger than it was ten years earlier whereas the Greek officer corps grew between 1872 and 1895 by a staggering 240 % (Roudometof Reference Roudometof2001). The military build-up was also accompanied by increasing investments in technology and industry linked to weaponry and military logistics. Hence Romania’s petrol production rose so much that by the late 1920s the country became one of the leading petrol producers in the world and Serbia’s industrial infrastructure at the same time was enlarged threefold compared to what it was twenty years earlier (Pearton Reference Pearton1971; Vucinich Reference Vucinich1968).

The ultimate result of this military, bureaucratic and state expansion was the capability of mobilising large sectors of the population and field enormous armies in the two Balkan wars and World War 1. It is these violent conflicts that brought about huge human casualties and material destruction never seen before in this part of the world. Hence, rather than representing an alleged continuity of violence supposedly inherent in the region’s past, the proliferation of organised violence was a completely novel phenomenon emerging as a direct consequence of intensive modernisation and state building on Western European models. In this context the Balkan wars of 1912-3 were not a throwback to the past but a distinctly modern phenomenon. The scale of violence and destruction witnessed here was a sign of things to come, something that the rest of Europe was to experience just one year later.

The fleeting look at the two Balkan wars would suggest that their experience is fully congruent with the bellicose argument and particularly with Tilly’s point that mobilisation for war and protracted warfare are likely to enhance the capacity of states: indeed what one can observe here is the parallel development of state power, war preparations and gradual pacification of domestic resistance. There is no disputing the fact that late 19^th century Balkan polities were much more robust state and war machines able to control resources, people and materiel, than their early 19^th century predecessors. However, the complexity of the Balkan case challenges the simple formula that war makes states and states make war.

Firstly, it is not so apparent that warfare itself had so much impact on the state building in the region as the period of most intensive state expansion was generally characterised by prolonged peace. In fact, in the second half of the 19^th century, a period when the Balkan countries made most progress in enhancing the infrastructural powers of their polities, the region was involved in only three small scale wars of very short duration: the Serbo-Ottoman War (1876), the Serbo-Bulgarian War (1885) and the Greco-Ottoman war (1897). It is difficult to see how these rather insignificant conflicts could have had a direct impact on the dramatic intensification of state building that took place in this period. Moreover, large scale protracted conflicts such as the 1912-3 Balkan wars, World War 1 and the 1919-22 Greco-Turkish war all came after the period of concentrated state development.

Secondly, and more importantly, the outcome of these major wars provides a direct challenge to Tilly’s thesis. If the central argument is that prolonged and successful war making leads not only to the creation of strong and stable states but also developed civil societies, parliamentarism and economic prosperity then the Balkan case shows otherwise. The aftermath of all major wars fought in South East Europe at the beginning of the 20^th century shows that regardless of whether a particular state found itself on the winning or losing side this made little or no difference to its post-war development. More specifically although Greek, Montenegrin, Romanian and Serbian states were clear winners of the 1912-3 Balkan wars as well as wwiFootnote ⁶ as they acquired large new territories, population and resources, their post-wwi state development was almost identical to that of the states that found themselves on the losing side: Bulgaria and Turkey. In other words instead of further enhancing their state capacities, expanding civil societies, parliamentarism and economic growth, the 1920s and 30s were periods of economic stagnation, weakening state power, curtailing of civil liberties and stifling of parliamentary institutions which eventually ended in rigid authoritarianism. The Balkan states became heavily indebted and reliant on foreign capital and in this resembled more the colonies of Imperial powers than the stable and strong sovereign states (Biondich Reference Biondich2011; Mungiu-Pippidi Reference Mungiu-Pippidi, Van Meurs and Mungiu-Pippidi2010; Mann Reference Mann2004).

Therefore the conventional bellicose approach seems unable to explain these different trajectories in the relationship between war and state-making in the Balkans. The key question here is: why have intensive inter-state wars not created strong and vibrant polities in South East Europe? To properly answer this question it is paramount to explore the internal configuration of societies in the Balkans and especially the relationship between social stratification, ideology and warfare. As the conventional bellicose approach tends to overemphasise external, geo-political and economic factors at the expense of internal and specifically ideological sources of conflicts there is a need to go beyond Tilly’s analysis. In particular the focus needs to move towards the role ideologies, especially popular ideological doctrines such as nationalism, and internal social divisions play in mediating the relationship between state and organised violence. Much of the bellicose tradition of analysis, and Tilly in particular, downplay the role of ideological power in modernity (Malešević Reference Malešević2010, p. 79-84). Nevertheless to account for the Balkan case it is crucial to take ideology and especially nationalism much more seriously than the conventional bellicose tradition does.

Nationalism and State Formation in the Balkans: Rehabilitating Modernism

Sharing the general view that human beings are essentially interest driven, materialist creatures much of the conventional bellicose historical sociology devotes little attention to the ideas, values and corresponding practices espoused by the agents involved in various social conflicts. Charles Tilly is no exception here as his approach emphasises the role of material interests and political institutions at the expense of collective meanings and individual perceptions. This is particularly visible in his treatment of nationalist ideology which is never a sui generis phenomenon but rather a weak, parasitic force dependent on the actions of states and their rulers. As Brubaker (Reference Brubaker2010, p. 380) rightly argues, Tilly’s understanding of nationalism is overly state-centred, materialist and instrumentalist: “the theory addresses the political form of nationalist claims-making while ignoring the cultural content of nationalist sense-making”. Hence, to explain persistent state weakness in the wake of intensive inter-state wars in the Balkans, my attention will shift to the role of ideology, and nationalism in particular, and its link with social stratification.

If we understand ideologies not as closed and inflexible belief systems but as dynamic, changing, processes through which human beings make sense of their everyday experiences, then the analysis of ideology is a precondition for understanding large scale social processes such as war and state-building. As political facts and social events cannot speak for themselves but entail particular interpretation we inevitably rely on different interpretative maps to understand social reality. In this sense all human beings are ideological creatures (Malešević, Reference Malešević2010; Reference Malešević2006). More specifically, as we live in a world where the principal unit of mass social organisation is the nation-state, with much of everyday experience articulated in nation-centric terms, it is crucial to understand how nationalism came to be and was established as the dominant source of state legitimacy in the modern world.

Seminal scholars of nationalism such as Gellner (Reference Gellner1964; 1983; 1997), Hobsbawmn (Reference Hobsbawmn1990), Breuilly (Reference Breuilly1993) and Anderson (Reference Anderson1983) among others have emphasised the inherent contingency, relative historical novelty and revolutionary character of nationalist ideology. Rather than being a natural, normal, primeval and permanent form of collective identity, national attachments are understood to be historically specific and atypical, generated by the actions of distinct social organisations, malleable and heavily dependent on on-going structural transformations such as industrialisation, urbanisation, or secularisation. For Gellner nationalism could not emerge before modernity as pre-industrial societies were rigidly hierarchical in an economic, political and cultural sense. Instead of unifying rulers and their subjects in the pre-industrial world, culture was used to reinforce social distinctions between the aristocracy and top clergy on the one hand and the peasantry and urban dwellers on the other. Hence for much of history the dominant socio-political units were either much smaller (i.e. city-states, principalities, chiefdoms, etc.) or larger than the nation-state (i.e. empires). Consequently, instead of nationhood, one’s sense of identity tended to be expressed in terms of locality, religion, kinship or status.

Gellner (Reference Gellner1997, Reference Gellner1988) makes a sharp distinction between the agrarian and industrial age: the former is characterised by stringent hierarchies, stagnation, sluggishness and stability where ‘people starve according to rank’ and the latter is defined by its vibrancy, dynamism, social mobility, innovation and the striving towards continuous economic and scientific growth. The system’s ingrained changeability, the clearly articulated division of labour and the demand for constant growth enhances the role of expert knowledge. Hence unlike Agraria where work is essentially manual and discursive knowledge is ancillary if not completely irrelevant, Industria is dominated by semantic labour where universal, context-free literacy becomes a norm. Moreover, as the industrial world is dependent on both social and geographical mobility, with large sectors of the population moving from the countryside to the cities, there is an organisational need for greater linguistic uniformity. Consequently, the proliferation of industrial development goes hand in hand with the expansion of large scale educational systems which ultimately turns a culturally diverse peasantry into a homogenous nation. In other words, mass education does not just provide a skilled workforce necessary for the perpetuation of its industrial base but more significantly it also forges strong national identities that could not exist beforehand. In Gellner’s (Reference Gellner1983, p. 36) own words:

Therefore, rather than preceding modernity nationalism is a by-product of industrialisation as it is only in the modern era that the trans-class cultural homogeneity makes sociological sense. In the pre-industrial era neither the peasantry nor the aristocracy could conceptualise the world in national terms: while the social universe of illiterate peasants rarely expanded beyond one’s village, close kinship or religious affiliation, the aristocrats and top clergy relied on culture to reinforce the internal status divide. In contrast, modernity entails a substantial degree of egalitarianism where commonly shared “high” national culture, inculcated through the educational system and other state institutions, replaces the sea of vernacular, oral “low” cultures and establishes itself as a principal source of political legitimacy.

Other leading scholars share many aspects of Gellner’s modernist explanation while downplaying his economistic account. Instead they argue that nationalism was an invention of political elites in times of dramatic political and social changes (Hobsbawm Reference Hobsbawmn1990), or a by-product of the development of the modern bureaucratic state (Mann Reference Mann1993; Breuilly Reference Breuilly1993), or a new form of collective imagination resulting from the expansion of print-capitalism (Anderson Reference Anderson1983). This modernist paradigm, and especially Gellner’s theory, has provoked much criticism. Some have singled out the rampant functionalism and historical determinism that underpin the paradigm (Malešević Reference Malešević, Malešević and Haugaard2007; Mouzelis Reference Mouzelis, Malešević and Haugaard2007; 1998; O’Leary Reference O’Leary and Hall1998); others have condemned cultural essentialism, Eurocentric assumptions and a rather nostalgic view of the European empires (Eriksen Reference Eriksen, Malešević and Haugaard2007; Hann Reference Hann2001; 1998). Nevertheless the most critical are the ethno-symbolists, perennialists and “early modernists” who insist on the substantial degree of continuity between the pre-modern ethnies and modern nations and who also emphasise the role emotions play in generating nationalist action (Smith Reference Smith2009, 1986; Hutchinson, Reference Hutchinson, Leoussi and Grosby2007; Roshwald Reference Roshwald2006; Gorski Reference Gorski2000; Taylor Reference Taylor and Hall1998). In the eyes of these critics South East Europe stands out as a clear historical case that refutes the key tenants of the modernist paradigm and in particular Gellner’s theory. Hence Drakulić (Reference Drakulić2008); Canefe (Reference Canefe2002); Minogue (Reference Minogue, Hall and Jarvie1996); Kedourie (Reference Kedourie1985), Orridge (Reference Orridge1981) and Wilson (Reference Wilson1970) argue that nationalism emerged in the Balkans long before any visible signs of modernity or industrialisation. For example Minogue (Reference Minogue, Hall and Jarvie1996, p. 120) insists that “nationalism long precedes the coming of industrialism, as in the case of Greek nationalism” whereas Hupchick (Reference Hupchick2002, p. 187, p. 212) writes about the “national revolutionary activity […] among the Serbs and the Greeks during the opening decades of the nineteenth century” and a “sense of ethnic group awareness, based on recognition of a common language and shared history, [that] grew and spread among the various Balkan populations (a process termed “national revival” […]) so too did the idea of group self-governance”. In a similar way Wilson (Reference Wilson1970, p. 28) describes the First Serbian Uprising of 1804 as “the first of the great nationalistic movements of the nineteenth century”.

The modernist response to these criticisms was either to soften their concepts and explanatory claims, to look for exceptional historical conditions for the Balkans or to emphasise the indirect influence of modernisation on the region. Thus Mouzelis (Reference Mouzelis, Malešević and Haugaard2007; 1998) and Hall (Reference Hall2010) reformulate Gellner’s concept of industrialism as “modernity” which would then be able to encompass the advent of nationalism in regions such as the Balkans, Latin America or Ireland where industrialisation came much later. Gellner’s own defence of his theory combined the idea of unique circumstances and indirect influence. He saw the Balkan merchants and bandits as key generators of nationalist doctrine: whereas the merchants were depicted as being affected by western ideas through international trade, the religious difference of the mostly Orthodox banditry vis-à-vis their Muslim Ottoman rulers led to their gradual transformation into nationalist rebels. As Gellner (Reference Gellner1997, p. 42) puts it:

As Hall (Reference Hall2010) and O’Leary (Reference O’Leary and Hall1998) rightly point out Gellner’s defence of his argument is not only unconvincing but it is also based on highly speculative assumptions which are extremely difficult to prove. This attempt at an ad hoc justification weakens the central premise of the theory. However, expanding Gellner’s concept of “industry” to “modernity” and searching for alternative signs of modernisation, other than industrialism, in the Balkans has not proved to be a particularly fruitful strategy. In this sense Gellner’s approach, as is the case for other modernist theories of nationalism, seems unable to explain the Balkan case. In other words, just as the conventional bellicose approaches could not make clear why intensive inter-state wars did not produce strong states in the region so the modernist theories of nationalism seem resigned to the view that the Balkans are a blind spot for their approach. Nevertheless, similar to the ‘war makes states’ paradigm both the criticisms as well as the defences of the approach have been focused on the wrong target. Rather than attempting to prove the impossible – that industrialisation or modernisation were in some form present in the early 19^th century Balkans – the emphasis should move towards the question of whether the Balkan uprisings of the early 19^th century and later had anything to do with nationalist ideology.

My argument is that even more so than organised violence, nationalism was a latecomer to the region. More specifically in what follows I demonstrate that nationalist goals and principles were a largely insignificant source for social mobilisation not only among the majority of the population but also among the political elites that were at the helm of these early uprisings. Furthermore, I argue that even when the Balkan states embarked on large scale protracted wars at the beginning of the 20^th century nationalism still remained an ideology that was not shared by the majority of the Balkan populations. In other words, contrary to popular perceptions the modernist theories can account for the Balkan case quite well. However, as classical modernism, and Gellner’s model in particular, do not devote enough attention to the complex relationship between nationalism and social stratification they need some revising with a view to explaining the peculiarity of the Balkan case.

If one follows Gellner’s (Reference Gellner1983, p. 1) well known definition of nationalism as a theory of political legitimacy built on “a political principle, which holds that the political and national unit should be congruent” then the early 19^th century uprisings in South East Europe could not be described as nationalist even in a minimalist sense. One of the important legacies of life under the Ottoman Empire was the dominance of religion, kinship and status ranks over “ethnic” attachments. The millet system fortified religious divisions and in this process subdued any sense of articulated cultural difference. For example the Rum millet included all Orthodox Christian populations regardless of their “ethnic” origin and the vernacular languages spoken. Since Greek became a lingua franca of the Rum millet, mastering the language well and moving to the city usually meant becoming a “Greek”. In other words being a “Greek” was a status category, a mechanism for social mobility, that had neither ’ethnic’ nor political meaning as an overwhelming majority of middle class Orthodox Christians (“Greeks”) had no inclination towards forming an independent Greek nation-state (Kitromilides 2010, Roudometoff Reference Roudometof2001).

In this context the various Balkan uprisings of the early 19^th century such as the First (1804-13) and Second Serbian Uprising (1815-17), the Wallachian and Cretan insurrections of 1821 and the Greek War of Independence were the result of internal turmoil within the Ottoman empire, and to a lesser extent the geo-political pressures of the Great Powers, rather than what they were seen to be in the West – “the revolutions for national liberation”. Not only did mass participants of these uprisings have no sense of what the sovereign nation means but even the political leadership of these movements had no ambition to establish independent nation-states. All these uprisings, including the Greek War of Independence, were chaotic, highly contingent events comprising elements of social discontent, fear, opportunism and necessity where nationalist principles were virtually nonexistent.

The first two significant rebellions of the early 19^th century, the two Serbian uprisings of 1804 and 1815, were profoundly contingent historical events which were neither inspired nor undertaken in the name of sovereign nationhood. The principal leaders of the rebellions, Đorđe Petrović-Karađorđe and Miloš Obrenović, were opportune traders who quickly realized that the social frustrations of local peasantry could be channeled in a direction that would benefit their personal influence and ultimately help their ambition to establish a monopoly on pork trade with the Habsburg Empire (Meriage Reference Meriage1977; Paxton Reference Paxton1972).Footnote ⁷ Both leaders were illiterate, their lifestyle and system of rule was modeled on the Ottoman example and instead of demanding independence for Serbia they were engaged in internal Ottoman conflict vouching ‘to restore the order on behalf of the Sultan’ and remove the disloyal and undisciplined ayans and janissaries (Roudometoff Reference Roudometof2001, p. 231; Djordjević and Fischer-Galati Reference Djordjević and Fisher-Galati1981; Pavlowitch Reference Pavlowitch and Clogg1981). As there were no discernible intellectuals on the territories of Ottoman SerbiaFootnote ⁸ at this time the support for the uprisings came from the cultural and religious elites based in the Habsburg Empire. However even these individuals, such as the leader of the Habsburg Serbian Orthodox population, Stevan Stratimirović of Karlovac, did not envisage the formation of an independent Serbian state for the Serbian nation but instead advocated the establishment of a Slavic Orthodox Empire “ruled by a Russian grand duke” (Meriage Reference Meriage1977, p. 189). Hence, neither political, cultural nor economic elites had any inclination towards national self-determination whereas the majority of, the essentially peasant, population had no understanding of what a nation was (Stokes Reference Stokes1976). While there is no doubt that in “the Greek War of Independence” the ideas and principles of national sovereignty were more present than in the Serbian (or later Bulgarian) case, the driving force of the conflict had very little to do with nationalism. Despite the later, nationalist, re-interpretation of the events leading to this war and the war itself as being motivated by the clearly defined goals of “national liberation’ this is far from the truth.

Not only were most senior Orthodox clergy, wealthy merchant families and the Ottoman Christian administrators not particularly interested in the demise of the Ottoman system under which they largely prospered as the leaders of churches, trade, banking, administration and foreign policy, but the Phanariot and Boyar (Christian) families were often the pillars of this very system enjoying various privileges and large scale estates. Hence the outbreak of conflict, later dubbed the “Greek War of Independence”, was essentially a power struggle between the two camps of Christian elites. The uprising started not in Greece but in the Dunabian Principalities (later day Romania) and it consisted of a chaotic and messy series of events involving prolonged internal rivalry initially between Boyars and Phanariots and later between different sections of very diverse “Greek society” with the final outcome decided exclusively by the direct involvement of Britain, France and Russia (Glenny Reference Glenny1999; Mazower Reference Mazower2000). The attempt to trace a direct link between these unpredictable and chaotic events with the activities of small secret societies, such as Filiki Eteria, the Philorthodox Organisation, or the Big Brotherhood, based outside of the Greek territories which allegedly had devised plans for the establishment of Greek nation-state, are completely unfounded. The prominent members of these societies advocated different and often mutually exclusive visions of cultural, religious or social renewal with the focus on the restoration of the Byzantine Empire rather than pursuing the goal of an independent Greek polity (Kitromilides 1994; Roudometoff Reference Roudometof2001).

Traditional historiography has made much of the role played by the social bandits (hajduks, hajduts, uskoks, khlepts, and kaçaks) in the early 19^th century Balkan uprisings, depicting them as guerillas fighting for the national cause. Nevertheless most were completely ignorant of the nationalist aspirations, were often simple opportunists willing to switch sides and prey on both the Muslim and Christian peasantry (Pelt Reference Pelt, Van Meurs and Mungiu-Pippidi2010, p. 224; Pavlowitch Reference Pavlowitch1999, Glenny Reference Glenny1999). The idea of shared national heritage and history meant next to nothing for most khlepts, hajduks and kaçaks. A flattering comparison made by a foreign visitor comparing a klepht leader Nikotsaras and the ancient Greek hero Achilles provoked an outburst of anger in Nikotsaras: “What rubbish are you talking about? Who is this Achilles?” (Kakridis, Reference Kakridis1963, p. 252).

Since the overwhelming majority of the population in South East Europe at that time were peasants whose sense of belonging oscillated between local (kinship, village) and religious attachments (Orthodox Christianity) there was no room for comprehending the world in national terms. To enhance their support in the uprisings the leaders had to rely on clan and family name recognition, religiously inspired prophecies that linked the collapse of the Ottoman Empire to the Second Coming of Christ and “the authority of the Orthodox [Russian] tsar […] and on loyalty towards the sultan”. However, in most instances, coercion was the decisive source for mobilization. For example Karađorđe “threatened to burn villages of those who did not appear [on the battlefield] […] or to decapitate Serbs [Christians] who helped Turks [Muslims]” (Stokes Reference Stokes1976, p. 83). Therefore the Balkan case does not falsify the modernist accounts of nationalism as there was no nationalism in the Balkans before industrialization or modernity. Not only were the majority of the population oblivious to nationalist ideology but so were, for the most part, the leaders of the early 19^th century uprisings in South East Europe. Furthermore Gellner’s emphasis on literacy and the role of educational institutions in forging viable national identities finds much support in the post-independence period.

Although the second half of the 19^th century saw the new Balkan states invest heavily in building large scale nation-state centric state apparatuses it took a long time to create a literate and schooled citizenry. For example as late as 1864 the literacy rate in Serbia was only 4.2 % while by 1884 there were just 4.5 % literate men and 1.5 % women in Bulgaria. Greece was in a slightly better position but the population was still overwhelmingly illiterate: in 1840 only 12.5 % men and 6.3 % women had basic literacy skills (Ekmečić Reference Ekmečić, Vucinich and Emmert1991; Roudometoff Reference Roudometof2001). The first primary Bulgarian and Montenegrin schools were opened in 1835 and 1934 respectively whereas the Albanian speaking population did not have a single primary school until 1887 (Biondich Reference Biondich2011, Lederer Reference Lederer, Sugar and Lederer1969).

As, before independence, the Orthodox Church was a most significant institution of what Gellner (Reference Gellner1983) would call “exo-socialisation”, it is interesting to briefly explore its relationship with nationalist ideology. Although after independence the autocephalous Orthodox Churches had become the beacons of ethno-nationalisms during Ottoman rule they were generally opposed to ideas and principles of national sovereignty. The Orthodox patriarchs, bishops and other senior clergymen enjoyed a privileged position within the Ottoman’s millet structure and were certain that independence would open the door for the potential proselytism of the Roman Catholic and Protestant churches. Furthermore the church establishment was particularly hostile to the Enlightenment and Romanticism inspired secularism, liberalism and republicanism that underpinned the nascent nationalist movements, and sternly resisted any initiatives that promoted the ideas and practices associated with the pre-Byzantine (i.e. pagan) traditions. Hence the Greek Orthodox Church opposed the standardization of Greek vernacular on its ancient model. The Patriarch Gregory V’s encyclical on education (1819) strongly condemned the practice of naming children after ancient Greek heroes and denounced even rudimentary attempts to propagate republican and nationalist ideas describing them as “the plots of the devil that often masquerade behind the clamoring for liberty and equality’ (Kitromilides 2010, p. 38).

While the senior clergy was actively hostile to nationalism the ordinary priests were largely ignorant of nationalist ideology. As they were hardly distinct from other peasants nearly all regular priests were illiterate, lacked even basic knowledge of the Bible, were not able to deliver sermons and were prone to combine Christian and pagan traditions. For example “Serbian clergymen were almost identical to other peasants. They were farmers and herdsmen. The only difference was their clothes and beards” (Radić Reference Radić and Perović2003, p. 158). In this sense, just as with the population at large, the local priests could not conceptualize the world in nation (ist) terms. It is important to emphasize that despite Orthodox Christianity being an important source of collective identification and probably the only pre-modern mechanism for exo-socialisation, the Church’s institutional influence was very weak. Throughout the Balkans church attendance was low and sporadic, confessions and communions were rare and the religious focus was on “sin and preservation of one’s honor and moral purity” and worshiping individual saints rather than on following Christian teachings and observing Church practices (Folić Reference Folić2001). Hence before independence and the development of state apparatuses no social institution, including the Church, was able to significantly penetrate the microcosm of local, kinship and village based traditions. Simply put, there was no institutional and organizational space for nationalism in the pre-independence and early post-independence Balkans.

It is the intensive state building that took off in the second half of the 19^th and early 20^th century that had direct impact on the presence of nationalist discourses. In other words rather than being a cause of state formation, nationalism was an outcome of state building as all Balkan governments started investing much energy and resources in what Mosse (Reference Mosse1991) would call “nationalization of the masses”. Hence the expansion of education coupled with ever increasing literacy rates became a crucial vehicle for state sponsored inculcation of nationalist ideologies. Thus, for example the newly formed state intellectuals in Serbia were commissioned to write textbooks and other literature which depicted Serbs as the ‘first and oldest people in Europe’ who have “founded Belgrade several thousand years before Christ” (Milojević Reference Milojević1871, p. 74). They also insisted that the Serbian people were very numerous in the past, have ‘inhabited three continents, Asia, Africa and Europe long before Christ’ and that all the peoples in the entire pre-Ottoman Balkans “spoke old Serbian language” (Petković Reference Petković1926, p. 57; Gopčević Reference Gopcevic1889, p. 12). Similarly, the Greek and Bulgarian authorities financed publications of books, plays, paintings and musical creations that either traced the origins of their nations far into the past or simply glorified the “national genius” of Greeks and Bulgarians respectively. So Constantine Paparrigopoulos’s “The History of the Greek Nation” (1865-74), which insists on the uninterrupted continuity between ancient and present-day Greeks, was financed by the Greek government, was quickly instituted as the official account of national origin and is still used in schools and colleges all over Greece (Roudometoff Reference Roudometof2001, pp. 107-10). In addition to the military, bureaucracy and police, national(ist) education became a budgetary priority. Thus Bulgaria’s budget for education expanded from 1.5 % in 1879 to 11.2 % in 1911, and it grew by a staggering 650 % in just over a 20 years period (Pippidi Reference Pippidi, Van Meurs and Mungiu-Pippidi2010, p. 128; Biondich Reference Biondich2011, p. 54). The lavish, state sponsored, institutions of “high” culture such as national academies and learned societies, universities, theaters, opera houses, museums and concert halls were set up in all capitals and some provincial cities throughout Southeast Europe. The Balkan governments provided financial support for newspapers and other mass media so that by the early 20^th century a large number of nationalist newspapers were in circulation. For example, by the late 1920s and early 1930s Yugoslavia had 50 major daily papers (Case Reference Case, Van Meurs and Mungiu-Pippidi2010, p. 294). The Balkan states also supported irredentist movements in the neighboring countries and attempted to project their geo-political ambitions, such as the visions of Greater Greece (Megali Idea), Greater Serbia (Načertanije, Homogena Srbija) and Greater Bulgaria (Velika i obedinena Bulgarija), into educational and artistic institutions.

Nevertheless despite the unprecedented nationalist propaganda penetrating all state apparatuses, and especially education, mass media and even civil society, nationalism remained an ideology of the elite minority, largely not shared by the rural population and urban laborers until well into the 20^th century. As the intensive state building of the second half of the 19^th century clearly privileged civil servants, the police and military establishment, state intellectuals and top business people it is only this social strata that became fully loyal to the new states and espoused fierce nationalist, irredentist and expansionist ideas. On the other side were those that did not benefit from state development: the peasantry, manual workers, and a tiny sector of the lower middle classes. Regardless of the fact that upon independence in both Serbia and Bulgaria most peasants become de jure small land holders while in Greece they found themselves de facto in such a position,Footnote ⁹ this did not substantially improve their economic or political position. However as the administrative, military and other state sector increased dramatically in a very short period of time it was the rural population, and to a lesser extent the small but rising urban proletariat, that had to provide resources to maintain such a gigantic state apparatus. The new states demanded cash taxation and this forced farmers to enter the market and seek credit, usually at quite unfavorable rates. Such a policy provoked deep animosity towards the city and its most recognizable representatives – civil servants. The peasant demonstrations and occasional jacqueries expressed this in slogans such as: “All kaputaši [wearers of city coats; townsfolk] should be killed” (Biondich Reference Biondich2011, p. 60). In other words the swift, state imposed modernization and industrialization created a sharp class polarization with on the one hand an urban, state created strata, favoring further state expansion both externally and internally, and on the other hand, mostly rural (and some urban) producers, favoring state transformation or attenuation of state power. These different structural positions fostered development of different ideological orientations: whereas the state bureaucracy, police, military establishment and the state intellectuals supported étatisme and nationalism, both of which were seen as the principal source of state legitimacy, the peasantry and urban poor were more sympathetic to religious conservatism, peasant populism and anti-statism (Pippidi Reference Pippidi, Van Meurs and Mungiu-Pippidi2010; Stojanović Reference Stojanović and Perović2003; Milosavljević Reference Milosavljević and Perović2003; Roudometof Reference Roudometof2001).

Therefore, despite their palpable visibility Balkan nationalisms of the late 19^th and early 20^th century were neither very deep nor extensively widespread. Most peasants opposed irredentist adventures, wars of territorial expansion and urban uprisings (i.e. the Serbo-Ottoman war of 1875, the April Uprising of 1876, the Serbo-Bulgarian war of 1885) and were often coerced to economically and politically support the war effort (Biondich Reference Biondich2011; Pippidi Reference Pippidi, Van Meurs and Mungiu-Pippidi2010). Furthermore to make the rural population more receptive to nationalism the state authorities relied on anti-Semitism, xenophobia and the scapegoating of religious and ethnic minorities. In this sense, rather than being a spontaneous expression of “age old” animosities and hatred, a great deal of violence was deliberately orchestrated from the top to engineer a sense of inter-ethnic fear and ultimately to provoke internal (national) cohesion. For example, in the run-up to the Balkan wars the Greek, Serbian and Bulgarian officers and soldiers were dispatched to Macedonia to ferment inter-ethnic hostilities, and were dressed up as the indigenous brigands “in order to conceal what was essentially state-sponsored violence” (Biondich Reference Biondich2011, p. 71). Similarly the state intelligentsia, government ministers and state sponsored education and mass media were often the principal vehicles of anti-Semitism, Islamophobia and hatred against other minorities in Romania, Bulgaria, Greece and Serbia (Case Reference Case, Van Meurs and Mungiu-Pippidi2010; Roudometof Reference Roudometof2001; Mazower Reference Mazower2000).

Although by the beginning of the Balkan wars of 1912-3 and World War 1, nationalist ideologies had penetrated greater sectors of the population and wars had helped foster a degree of inter-group cohesion even then a majority of the population in South East Europe was not fully ‘nationalized’ and most peasants and urban laborers remained skeptical if not completely opposed to these wars. For example when the Greek activist persistently insisted on finding out whether the local peasantry of Salonika see themselves as Greeks or Bulgarians the peasants did not understand the question: “whenever I asked them what they were – Romaioi [i.e. Greeks] or Voulgaroi [Bulgarians], they stared at me incomprehendingly […] Well, we’re Christians – what do you mean, Romaioi or Vulgaroi?” (Mazower Reference Mazower2000, p. 50).

As Boindich (Reference Biondich2011, p. 43) points out ‘in the period between 1878 and 1923, when the Balkans experienced some of its worst political violence, the bulk of the population, the peasantry, still lacked a strong national consciousness’. Most peasant recruits were unwilling to fight, were inclined to desert or avoid conscription and were not particularly enthusiastic about the territorial expansion of their states. As well documented by Leon Trotsky’s 1912-3 Balkan war correspondence, Serbian peasant conscripts were apathetic and inimical to war efforts and nationalist projects. He characterized them as “depressed and extremely homesick for their villages” (Trotsky Reference Trotsky1980 [1913], p. 121).

Hence South East Europe is not a blind spot for the modernist theories of nationalism. Not only was there no nationalism in the Balkans before modernization and industrialization, but nationalist ideology, even more so than organized violence and warfare, came very late to this part of the world. However, modernism requires some fine tuning to account for the often inversely proportional relationship between social stratification and nationalism. While most modernist theories see nationalism as the decisive social cement that binds diverse citizenry into stable and cohesive societies able to generate economic growth and social development, the Balkan experience indicates an alternative trajectory. The 19^th and early 20^th century history of the region shows how, rather than acting as a cohesive social force, nationalisms in South East Europe were often a source of internal discord: the state supported irredentism and the pursuit of territorial expansion were regularly perceived as the ideological projects of the city elite which could only, and usually did, bring misery to the country side. Hence, instead of generating greater social unity and solidarity and thus potentially stimulating social development the state enforced attempts at nationalist mobilization tended to further polarize already extremely divided societies (Malešević Reference Malešević2006, 2011).

Making Wars, States and Nations in the Balkans

There is a widespread view of the Balkans as the region which for most of its history was brim-full of protracted violence and nationalist euphoria. From the early 19^th century ‘national’ uprisings to the 1990s wars of Yugoslav secession, South East Europe has been seen as a “powder keg of Europe” (Kaplan Reference Kaplan1994). Moreover, both nationalism and violence are regularly singled out as the most important impediments for the social development of the region. Nevertheless, a careful historical sociological analysis shows that neither of these two common perceptions holds up well to empirical scrutiny. Instead of being an inherent feature of the Balkan landscape, both nationalism and organised violence are, historically speaking, fairly recent arrivals to the region. In a similar vein, it is not the abundance of nationalisms and wars than have stifled the development of the region but in fact it was often the lack of their organised prevalence that hampered wider social advance. The fact that much of the warfare in the Balkans was small scale, sporadic, and disorganised meant that such wars could not help enhance the organisational capacities of states in the region. Likewise the uneven, narrow, belated, and rigid, top down spread of the nationalist ideology often mitigated against the development of internal social cohesion, thus preventing the emergence of a degree of social consensus necessary for economic and political development. Nonetheless, this is not to say that war-making by itself inevitably generates strong states and societies, or that the proliferation of nationalism automatically leads to societal wellbeing and economic prosperity.

The view of conventional bellicose historical sociology that protracted warfare is likely to eventually yield infrastructurally strong, centralised states capable of creating political stability, social order, and economic growth, requires major amending. Although the institution of the state might have originated in warfare, its viability and expansion requires much more than sustained violence. Not all wars make states and even those that make states do not necessarily create strong polities. As Tallet (Reference Tallett1992, p. 198) points out, Tilly’s account suffers from the chicken and egg symptom. It is not clear what comes first: “whether an efficient and developed bureaucracy was the precondition for the growth in size and complexity of the armies […] or whether growth in armies stimulated growth in the bureaucratic structures of the state”. Thus warfare is better understood as a test of state strength rather than as an impetus for state formation. Balkan warfare clearly demonstrates the complexity of this relationship. Wars did not create independent polities in the early 19^th century Balkans. Instead independence was a highly contingent event resulting from the combination of the internal organisational structures of the weakened Ottoman Empire and the geo-political interests of the Great Powers.Footnote ¹⁰ Similarly, as they were small scale events in the mid to late 19^th century, wars did not forge strong states either. The beginning of the 20^th century saw South East Europe becoming an epicentre of mass scale warfare but much of the intensive state building took place long before the 1912-13 Balkan wars and World War 1. While the mobilisation for warfare certainly played a part in the centralisation of state power and the expansion of the state’s infrastructural and bureaucratic capacities, the outcomes of these high intensity wars were not strong states and vibrant civil societies. Instead both the victorious and the defeated Balkan states shared almost an identical developmental trajectory in the 1920s and 1930s: authoritarian rule, stifling of parliamentarism and civil society, huge indebtedness to international creditors, state monopolies in the economy, rampant nepotism and corruption, heightened class polarisation and perpetual animosity between the urban and rural population.

To fully understand this particular outcome, which in many respects contradicts the central thesis of the conventional bellicose approach, it is crucial to focus on the role of nationalism and social stratification in the Balkan societies. In contrast to normative views that see rampant nationalist attachments as an obstacle for social progress, historical and sociological reality indicates that the degree of national solidarity is often a precondition for effective political and economic development. The modernist paradigm in the study of nationalism, and Gellner’s work (1997, p. 25; 1964, p. 114) especially, argue that not only are nationalism and economic growth fully compatible but, as the two are the main sources of political legitimacy in the modern/industrial era, they entail each other. Furthermore, and in contrast to the primordialist and perennialist accounts, the modernists rightly insist on the historical novelty of this synergetic relationship between nationalism and socio-economic development. The Balkan experience fully vindicates both of these claims.

Firstly, contrary to the primordial and perennial interpretations, the early 19^th century uprisings in South East Europe had nothing to do with nationalism but were a direct by-product of imperial geo-politics and internal weaknesses of the Ottoman social order. Rather than being a motivational source of state-building, nationalist ideology was a consequence of state formation. Nevertheless, even after decades of intensive, state sponsored “nationalisation of the masses”, this ideology remained a profoundly weak force unable to motivate the majority of Balkan populations until well into the 20^th century.

Secondly the corollary of the state’s inability to swiftly turn peasants into Serbs, Greeks, Bulgarians or Albanians was a lack of internal consensus on developmental goals and ambitions. As nationalism did not penetrate most layers of Balkan societies there was no adequate social glue to provide a shared vision of national solidarity necessary for radical economic, political and social reforms. This Gellnerian account of the relationship between nationalism and social progress requires an analytical extension to capture the internal dynamics of South East European societies where, rather than being a device of social cohesion, nationalism was often a source of deep class friction.

Therefore, the popular image of the Balkan Peninsula as a historical hub of aggressive nationalisms and perpetual violence, both of which allegedly thwarted its progress, is really an inverted, image of the historical reality. It is actually the absence of protracted organised violence and society-wide nationalisms that have heavily contributed to the often sluggish development of the region.

Table List of 19^th and early 20^th century Balkan Wars and Major Uprisings