Abstract The three most important ruptures in Spanish anarchism in the 20thcentury (1931, 1945 and 1979) occurred during political transitions towards more open and democratic systems. How did the political context influence these processes of factionalism and rupture? This article attempts to answer this question by means of a qualitative comparison of the three episodes. The results highlight the role played by two environmental factors: the isolation imposed by the previousrepression, which encouraged the growth of divergent groups; and the growth in resources associated with the start of the political transition, which encouraged the transformation of symbolic conflict into political competition for control of the movement. Moreover, the results show how certain organizational procedures and features of the movement accelerated the internal conflict. They show in particular how the calling of congresses in the new context of visibility served to catalyze political competition; and how the symbolic capital of the movement became reduced to two antagonistic and mutually destructive strategic positions. At the theoretical level this case study examines the scope of two different models. On the one hand ,it extends the viability of the alternating phases of latency and visibility model by applying it outside the context of new social movements. On the other, it questions two biases of the protest cycle model by dissociating innovation and internal competition in social movements from the initial and final phases of the cycle, respectively.

‘Liberty is to faction what air is to fire, an aliment without which it instantly expires.’ James Madison, 1787

Anarchism found fertile ground in Spain in which to put down roots for a powerful social movement. Much has been written about this movement, especially with respect to its participation in the contentious 1930s. However, there is still a lack of studies which, with a broad historical perspective, compare the Spanish experience with that of anarchist movements in other countries, or compare the distinct phases of the Spanish movement itself. This article consists of the second type of comparison and is centered on processes of rupture. Factionalism, that dynamic in which informal groups fight for control of the mother organization, is a phenomenon common to all kinds of organizations (Hirschman 1970). Numerous scholars have devoted research to it, particularly in the area of political parties but also in that of other, less institutionalized forms of collective action such as social and political movements, and also in areas that are not strictly speaking political, such as religion or the world of business (see Boucek 2009). All forms of social action with a minimum of continuity experience clashes between different internal groups and Spanish anarchism is no exception, on the contrary, its history is strewn with struggles which have sometimes led to splits.

In line with the thinking of Madison quoted above, one might think of anarchism as fertile ground for factionalism. However, this argument lacks theoretical support because while it is indeed possible to perceive a tension between the anarchist commitment to individual freedom and the effectiveness of political action (Arvon 1979), the organizational flexibility and democratic deliberative processes defended by anarchism in the heart of its orga- nizations ought to facilitate the resolution of conflicts (see Romanos forthcoming). In any case, there is a need for studies which empirically examine the phenomenon. This article aims to contribute to filling this gap through a comparison of the three most important ruptures suffered by the Spanish anarchist move- ment in the 20th century, those of 1931, 1945 and 1979. The first occurred a few months after the foundation of the Second Spanish Republic, the second took place in France after its liberation among Spanish anarchists in exile and the third, the effects of which are still in place, occurred during “the Spanish transition to democracy” that followed the death of the dictator Francisco Franco in 1975.

The ruptures in Spanish anarchist movement were centered on one organization, though the involvement of other organizations or groups and their rapid diffusion throughout the anarchist movement in general made them ruptures in a social movement and not just one social movement organization (SMO). The organization in question was the Confederación Nacional del Trabajo (CNT), the anarchosyndicalist trade union created in 1910, which was, indeed, the most powerful and influential organization within Spanish anarchism.1 Certain characteristics of the CNT appear to have acted as risk factors for the appearance and non-resolution of conflicts. These included, for example, its emphasis on doctrinal purity and a system of authority that was not very centralized; there was a national committee but the degree of autonomy of local and regional federations was high. However, a relatively homogenous social base worked in its favor, (though there were always differences between occupational, geographical and age groups), as well as an inclusive membership and a largely formalized, even bureaucratized structure (Zald and Ash 1966; Gamson 1990; Balser 1997).

Did doctrinal purity and decentralized authority influence the processes of rupture? They probably did but these organizational features remained relatively constant down through the years and so do not explain why the ruptures occurred in certain historic contexts and not others. The political context in which the three episodes under review took place was one of political transition towards more open and democratic systems: from the dictatorship of Primo de Rivera to the Second Republic in Spain, from the German occupation to the Fourth Republic in France and from Franco’s dictatorship to the constitutional monarchy in Spain. Thus I believe it is important to maintain a focus on exogenous factors related to the political environment and which configure themselves in the relationship which the organization established with other actors such as the state, the general public and other social movements and organizations (Balser 1997). This is the line of enquiry I will pursue in this study of ruptures which, although they occurred in similar political contexts, have important distinctive features.

What relationship was there between the political context and the ruptures? What combination of exogenous and endogenous factors influenced the dynamics of the ruptures? This article aims to provide material to help answer these questions through a comparison over time. The study of contentious politics is no stranger to the question of time but the kind of comparison that has most commonly been carried out has been between spaces, among cases generally from different countries, in search of the distinctive features of each one. The events and processes under review in this article, however, are situated in a number of historical events undergone by the same movement, and this with the intention of analyzing continuities and change. The comparison carried out of these events is qualitative in nature as it involves only a few cases (three) with a common base (factionalism and rupture) in order to maximize comparability (McAdam et al. 2001). The data analyzed, which deals with the organizational and symbolic resources of the anarchist movement, comes from the relevant literature and, on occasion, from archive material and the testimony of participants, especially in the case of the second rupture, that of 1945.

The article opens with a brief setting out of the theoretical framework in which the relevant models and hypotheses for this study are presented. There follows a descriptive section which presents the ruptures in chronological order after which they are compared in order to seek differences and similarities, to identify factors of rupture and to see to what extent this broader approach confirms or contradicts the specialized literature. The article closes with a conclusion in the form of a summary of its main findings.

Transitions and Cycles

I have identified two theoretical models which explain the general implications of the changing political environment in the movement and intra-movement dynamics and whose arguments and hypotheses appear to be of relevance for this study. They are the alternating phases of latency and visibility model and the protest cycle model. The first was originally proposed by Melucci, who saw in “latency and visibility . . . the conditions of the contemporary movements, which ceaselessly oscillate between them” (1996: 174). The latency phase is characterized by the underground activity of networks distant from official institutions and with actors who are strongly committed to and involved in cultural production through the elaboration of alternative codes and meanings different to those that predominate in society (Melucci 1994, 1989; see also Taylor 1989). In general, periods of latency coincide with political contexts that are not favorable to the demands of the movement; during these periods the group manages to keep itself alive due to its underground structure until the development of a context more favorable for action, when such a context develops the movement enters into a period of visibility. This involves intense public political activity, open confrontation with the authorities and the putting into action of the mobilizing frames prepared during the period of latency. Recruitment processes are activated or intensified on the basis of pre-existing solidarity networks. The phases of latency and visibility described in the model correspond to those experienced by the anarchist movement before and during the political transition, respectively. In the former the anarchists were obliged to carry out their activities underground to escape the repression organized by the state, in the latter, on the fall of the repressive regime, they were able to emerge from the underground and carry out their activities and demonstrations in public.2

This model of alternating phases, though often cited has been applied in a partial manner. A number of studies have looked at the importance of latency phases in the continuity of social movements (e.g., Staggenborg 1998; Mooney and Hunt 1996; Taylor 1989; Lindgren 1987), but insufficient attention has been paid to the transition from these phases to those of visibility (della Porta and Mosca 2007), and even less to the relationship between this transition and what occurs outside the movement, at the level of the political context. This present study attempts to contribute to a better understanding of the relationship between the political transition and the transition as it was experienced in the heart of the movement through an examination of the effects of these processes on factionalism and the ruptures. Furthermore, while the model of alternation between latency and visibility has usually been applied to new social movements, this article analyzes it in what might be described as a movement of “the old sort”.

The second theoretical model is that of cycles of protest. The cycles and waves of protest are metaphors which try to capture the rhythm of contention. The idea is that mobilization does distribute itself at random over time but rather in phases of relative calm and others of great intensity (Koopmans 2004). Tarrow defines a cycle of protest as “an increasing and then decreasing wave of interrelated collective actions and reactions to them whose aggregate frequency, intensity, and forms increase and then decline in rough chronological proximity” (1993: 287). The cycles are partly the result of changes in political opportunities (aspects of the political environment which, on being perceived by the actors, encourage or discourage collective action), and partly the result of the action of “early risers”, powerful movements which show the vulnerability of the authorities and reduce the cost of political action for other actors (Whittier 2004: 533). The disruptive frames, forms of organization and tactics of the “early risers” influence and serve as an example for other movements, “spin offs” which support, imitate or oppose the demands of their predecessors. In the cases under review here, the phase of the beginning of the cycle corresponds to the contentious sequence which follows the beginning of the political transitions, in which various social and political groups (the anarchists among them) perceived new opportunities for collective action and, as a result, intensified their mobilization, expanded their demands, and recruited more participants for their campaigns and protests.

With regard to the ruptures, the protest cycle model distinguishes two kinds of fragmentation: one is non-traumatic, or at least less so, and has to do with the expansion, dissemination and diversification of demands within the social movement sector through the “spin offs”, while the other is a traumatic fragmentation such as the ones the anarchist movement underwent, which is the result of strategic competition. Research has shown the importance of this second type of competition in social movements, sometimes carried out by way of the SMOs (e.g., King 2008; Benford 1993; Barkan 1986). The inclusion of exogenous factors in this conflict has centered on the intervention of the state through the channeling of the moderates and the repression of the radicals (Gupta 2007; Koopmans 1995, 1993; Tarrow 1989; Haines 1984; cfr. Jung 2010). Observing their evolution over time, some scholars have located strategic conflict and factional competition in general in phases of reduction in mobilization potential (the final phase of the cycles of protest), with internal ruptures themselves accelerating the decline (della Porta 1990; Tarrow 1989; Miller 1983; Gamson 1990). Normally they are associated with the reduction in resources that occurs “during the declining phases of mobilization - when a part of the movement became institutionalized and the movement organizations more exclusive” (della Porta 1995: 110). The comparison of the three ruptures will serve to confirm, refine or refute this observation.

Finally this article aims to shed light on an important issue in the process of factionalism in Spanish anarchism. This issue is innovation and it is one regarding which the models set out above have opposing views. The working hypothesis in this paper is that the production of new sensibilities, alternative tactics and ideological positions led to internal conflicts. The alternating phases model situates innovation (understood as cultural production) in the latency phase prior to intense mobilization. The protest cycle model, by contrast, situates it at the starting phase of the cycles (which would correspond with the visibility phase in Melucci’s two-poles model), when the contacts between actors, participation in protests, interaction with the authorities, and the production of new frameworks of mobilization accelerate significantly (Tarrow 1993; Snow and Benford 1992; Kriesi et al. 1995; cfr. Tilly and Tarrow 2007: 30).3

To sum up, the two theoretical models used in this study are partially complementary as they focus on different periods of the life of the movement. They are also partially contradictory in that they disagree on the question of innovation. Their application to the case of Spanish anarchism will serve to contribute to resolving this disagreement about innovation and establish the scope of their presuppositions in the context of the political transition in which the ruptures occurred.

The Three Ruptures

The Rupture during the Second Republic

The dictatorship which followed the coup by General Primo de Rivera in 1923 applied a policy of closure and detentions to the CNT (González Calleja 2005). The trade union suffered a number of strong internal conflicts during this period and these became more severe from 1927 when its legal status was made conditional on the explicit acceptance of the corporative organization of the state and management-worker committees to settle disputes. Two groups emerged during these disputes: the more trade union orientated vs. the defenders of the trabazón (“connection” or “link”), the name given to the organic relationship between the union and the anarchist groups that were members of the Federación Anarquista Ibérica (FAI), the “specific” anarchist organization created in 1927 to safeguard the essence of anarchism. Primo de Rivera resigned in January 1930 and the CNT was finally legalized on April 30th that year. However, it continued to hold underground meetings in some regions for some time.

After the foundation of the Second Republic, on April 14th, 1931, the CNT grew significantly in terms of members and resources, especially in the first months. According to the noted activist Federica Montseny, this tendency had begun in 1930, the year in which the CNT “had flourished with exceptional vigor” (1987: 60).4 By the time it held its Third Congress in Madrid (June 11th to June 16th, 1931) it had 535.566 members,5 a number that was to grow to some 800.000 by the autumn of the same year, mainly concentrated in Catalonia and Andalusia. This high level of participation was maintained well into 1932. After that it began to decline and did not recover until the early months of 1936. At the tactical level the Third Congress approved important initiatives which saw in the advent of the new republic the necessary conditions for recovering from the damage done by the dictatorship of Primo de Rivera and convert the CNT into an organization able to take on the growing power of capitalist concentration (Casanova 2004; Lorenzo 1972). These included a plan for a series of minimum demands to be put to the Cortes Constituyentes (Constitutional Assembly) in areas such as education, individual and collective liberty and the orga- nization of work, and the creation of the Federaciones Nacionales de Industria. This latter implied an end to the old craft union system and its replacement by an entity that would bring together all the workers in one industry. The approval of the two resolutions was not without controversy. Both received strong criticism, especially the plan of demands, with some thirty unions describing it as collaborationist with the institutions of the state, something which contradicted the objectives and principles of anarchism.

On the 30th of August, 1931, just two months after the Third Congress, thirty prominent activists of the CNT published in a Barcelona newspaper the so-called Manifiesto de los treinta (for the number of its signatories), criticizing what they described as the “demagogic” concept of revolution defended by the FAI, which it described as a vanguard organization that encouraged destructive courses of action among the CNT constituency. After its publication, the signatories of the manifesto were stigmatized and those in favor of union action, who at first had emerged strengthened from the Third Congress, were progressively marginalized and removed from representative posts. At the same time the FAI gained decisive influence in the CNT through the monopolization of the defense committees created by both organizations and the control it exercised over prisoner support committees, whose solidarity campaigns were the platform on which an ever more active campaign of agitation against the government of the republic was mounted (Casanova 2004).6 After various expulsions the treintistas confirmed the split by creating a parallel organization, the so-called Sindicatos de Oposición, which had its own media outlets (among them Sindicalismo), and which grouped around the Federación Sindicalista Libertaria, founded in January 1933 and which held its First Congress in July 1934. Meanwhile the FAI had carried out a period of insurrection which between 1932 and 1933 decimated its ranks, along with the resources of the CNT.

The Sindicatos de Oposición that led the split were strongest in Catalonia where in the summer of 1933 they had around 26.000 members, and the Levante where in 1934 they organized a meeting attended by 23.258 members. They also had some influence in Madrid and Huelva and some sympathizers in Asturias and Galicia (Íñiguez 2001; Vega 1987, 1980). The “treintistas” were finally admitted to the CNT in its Fourth Congress (held in Zaragoza from May 1st to May 15th 1936, with 559.294 members of whom 60.621 were from the Sindicatos de Oposición) during which both the treintista tactic and that defended by the FAI were recognized as failures. With the two sides somehow reconciled they promoted confrontation with the state based on close cooperation with the socialist UGT federation, of which it demanded that it abandon its tactic of parliamentary and political collaboration.

The Rupture in Exile

Just two months after the Fourth Congress, the anarchists found themselves in control of different cities and regions after having defeated the military uprising that sparked the Spanish Civil War (July 17th, 1936–April 1st, 1939). With the collapse of the state, they succeeded in enacting a programme of libertarian communism in the form of industrial and agrarian collectives in the rearguard, especially in Catalonia, eastern Aragon and Valencia. The Civil War represented the apogee of the anarchist movement, but ended in severe and cruel repression. The postwar Francoist regime brutally repressed the activists, rendering their organizations illegal. Many flee into exile. Most concentrated in France, its first and most important destination. The regrouping began in concentration camps with hardly any international and institutional help. In February 1939 the Consejo General del Movimiento Libertario (CGML) was formed by some leaders who had crossed the border after the fall of Catalonia but it barely functioned and after the outbreak of the Second World War in September of that year it practically disappeared. Some anarchists were able to leave the camps to work in war industries. A large and active group in the department of Cantal participated in the construction of the l’Aigle dam, where a comisión organizadora (later comité de relaciones) of the Movimiento Libertario was formed.

After the liberation of France there was a clash between the Cantal group, which by then had become the National Committee of the CNT in France and the former members of the CGML. This was not the first clash that had occurred in exile,7 but was important for its virulence and because it anticipated the two groups and two positions which would confront each other in the next rupture. In October 1944, Juan Manuel Molina, secretary general of the National Committee (NC) criticized Germinal Esgleas (secretary general of the CGML since June 1939) for his silence during the occupation, for excesses committed in the appropriation of respon- sibilities and the mismanagement of funds. Esgleas denied these allegations while requested that the NC reaffirm its “uncompromis- ing position against the state and all forms of power” against “the suicidal goal” of collaboration with other anti-Francoist political forces, which was then the chosen tactic of the clandestine activists in Spain.8 The position attacked by Esgleas had been advanced in a document produced by the NC in June 1943 which, subordinating itself to the view of the activists in Spain, defended the legality of the Republic, with an eye on the position of the eventually victorious Allies.

This tension and disparity of views served as a prelude to what occurred at the congress of the Movimiento Libertario Español (MLE)-CNT9 in France, held in Paris from the 1st to the 12th of May, 1945. This congress marked the culmination of the process of reunification of the Spanish anarchists dispersed by the conse- quences of the Second World War. It was prepared in haste with different factions passing around “guides and suggestions . . . by various means” among the activists, which obliged the NC to make a call for order, apparently without result.10 Victory went to those in favor of “apoliticism”, who had made much effort in proselytizing to discredit the pro-alliance position of the activists in Spain and they emerged strengthened from the congress. There was little delay before “fratricidal struggle” (Peirats 1990: 94) broke out. In September the new NC elected by the congress refused to recognize the nomination of two anarchists as ministers of Public Works and Agriculture of the government of the Republic in exile. Their nomi- nation had been an initiative of activists in Spain who favored alliances with other opposition forces and meant a return to government after the participation in it during the Civil War. The decision of the NC led a group of prominent exiles to sign a critical manifesto published on the 27th of October with the title Con España o Contra España (With Spain or Against Spain). Its publication confirmed a split which from then on confronted two groups of activists gathered together in parallel organizations; there were those who supported the pro-alliance position of the activists in Spain and created a National Sub-Committee (NS-C) in France, leaving the general representation of the movement to the organization in Spain, and those who opposed it, who defined the pro-alliance position as alien to the methods and goals of anarchism. This latter group adhered to the NC elected by the Paris Congress. With the split confirmed the groups dispersed in exile aligned themselves either with the NC or NS-C with accompanying rounds of expulsions from both sides.

The split tore apart the movement in exile but also distanced the official CNT in France (led by the NC) from the clandestine organization in Spain, which found an interlocutor and support in the dissident exiled sector (led by the NS-C). With regard to the number of members that supported each faction in exile, it has been estimated that of the 30.000 registered at the 1945 Paris Congress, about 73% (22.000) stayed with the official CNT, while 15% (4.500) aligned themselves with the dissidents and 12% (3.500) resigned their membership. Membership declined over the years on both sides. In 1947, the official CNT had lost 20% of its members and was left with around 18.000. By 1948 only 13.000 remained. The decline was more pronounced in the dissident sector. By 1950 it had lost 75% of its members if the membership calculated at the moment of the split is taken as correct. Once this rapid loss of members had occurred, the remaining thousand or so stayed affiliated for a decade (Herrerín 2004; Borrás 1998).

In the summer of 1961, the official CNT and the dissidents reunified themselves in a congress held at Limoges, France. The reunified body had a membership of 7,135. By then the long exile and the negative consequences of the split had bled the movement white with the grassroots distancing itself from a more active and cliquish minority.

The Rupture during the Transition to Democracy

After three decades of repression, the anarchist movement in Spain at the start of the 1970s found itself dispersed among a series of committees which claimed to be representing the CNT, and a number of new groups made up of young anarchists who were partly influenced by the renewal of anarchism that occurred in the 1968 anti-authoritarian struggles. Some of these young people took part in a conference in the spring of 1973 which was aimed at being the first step towards an anarchist congress which would permit a clarification of ideology and the development of a consensus on a common strategy. The attempt, however, failed (Romanos 2007).

Most of the groups of young people decided to dissolve themselves, some doing so more quickly than others, during the process of reconstruction of the CNT which followed the death of the dictator Franco, which occurred on November 20th, 1975. This reconstruction started at regional level, firstly in the Asturias region, then Madrid and at the start of 1976, in Catalonia (Gómez Casas 1984). It held its first national meeting in July 1976, by which time it had some 130.000 militants and activists. Of these 70.000 were located in Catalonia, 20.000 in Andalusia and 15.000 in the Levante. During 1976, the anarchist movement grew progressively in a context of expectations of change and social effervescence in which it tried to make itself visible and occupy public space in demonstrations to which the authorities more or less consented (Zambrana 1999). Until it was legalized in May 1977 the situation was one of a certain degree of tolerance or permissiveness with demonstrations being allowed or banned depending on the circumstances. Many young people (from the groups that appeared towards the end of Franco’s rule, or new sympathizers with anar- chism) came to these demonstrations but their participation tailed off due to a great extent to problems generated by the generation gap which separated them from many older members of the CNT, for example, regarding the question of whether to reconstruct a trade union federation or try to form a new anarchist movement with room for new sensibilities. Among these young people there existed a yearning to leave behind the old economicist vision of anarcho-syndicalism and instead commit to an integral revolutionary project which would deal with all aspects of capitalist exploitation, among others those in the “personal” sphere (Torres Ryan 1993).

The anarchist movement reached its zenith in 1977 with the holding of mass meetings in San Sebastián de los Reyes on the 27th of March, which over 30.000 people attended, and one in Barcelona on the 2nd of July attended by between 100.000 and 200.000 people. It also organized the Jornadas Libertarias Internacionales (International Libertarian Conference) in Barcelona from the 22nd to the 25th of July, these were attended by some 600.000 people. The Jornadas were the most visible sign of the cultural effervescence within the anarchist movement, an effervescence that covered areas like worker management, anti-militarism, ecology, sexology and women’s issues and made efforts to recover a largely mythologized historical memory (Zambrana 1999). This was a period of significant strikes, pro-amnesty demonstrations, assemblies and attempts to protect historic buildings through occupations. Anarchist thought circulated in new magazines (e.g., Ajoblanco, El Viejo Topo, Bicicleta) and in books published by Campo Abierto, La Piqueta, and Tusquets’ “Acracia” series, among others.

The movement lost much of its influence after 1978. The reasons for this are various. Among them were the lack of consensus on central issues, the dissatisfaction of some groups with the development of the reconstruction project and the repression by the authorities of a movement that had declared itself to be opposed to the Pactos de la Moncloa.11 In April, after the transfer of the CNT secretariat from Madrid to Barcelona, the internal conflicts intensified and discontent continued to grow. By the time the CNT held its Fifth Congress in Madrid (December 8th to December 16th 1979) it had lost many of the members with whom it had started its reconstruction. Its membership had then declined from 116.900 in September 1977 to just 30.288. The congress also failed to reach agreement on any of the most controversial topics; it served rather to catalyze the differences which were to break the CNT, with many groups which had at first joined the reconstruction project already outside it. The atmosphere was very tense right from the outset, with few mechanisms of control over those attending and the suspicions that this generated. The lack of time and the excessively long agenda which prevented points from being discussed in detail only made matters worse. From the motions passed it can be seen that the congress reaffirmed old. Anarcho-syndicalism as a general orientation was reduced to its economic aspects with the focus on the world of work as the central space of emancipation. The most controversial point was the strategy for action. In 1978 the CNT had boycotted the union elections but some of its members had run on a personal basis. From then on the question was a controversial one. However, the congress was unable to generate a consensus on the matter. The resolution that was passed reaffirmed support for the traditional methods of paralyzing the production process by boycotts, strikes and sabotage (Carmona 2004). The impossibility of using legal methods by way of union elections led 53 unions to abandon the congress, accusing the organization of being ineffective for blocking the exposition of arguments, debate and the search for solutions.

The closure of the congress was quickly followed by a legal challenge to its validity which was supported by 8.000 members, more than a third of those registered to attend it. The split took shape in the Comisión Técnica Impugnadora (the technical commission which dealt with the legal challenge) which was created at the start of 1980. It organized a new congress in Valencia from the 25th and 27th of July and among the resolutions it approved was one that authorized participation in union elections. As had happened in previous ruptures, the CNT started to expel the dissidents, whom it defined as enemies of trade unionism. The dissident faction wanted to continue to use the classic acronym “CNT” but a judge’s decision prevented it from doing so and from 1989 it became known as the Confederación General del Trabajo (CGT).

Comparisons

The ruptures at the heart of the Spanish anarchist movement occurred in the context of a general opening of political opportunities coming after a period of repression. The foundation of the Second Republic, the liberation of France and the starting pistol shot that signaled the start of democratic transition on the death of the dictator Francisco Franco opened new, previously unknown opportunities for political and social mobilization in their respective contexts (see, for example, Cruz 2006; Buton 2004; Sánchez- Cuenca 2010). The three splits were “transformative events” in so far as they represented turning points in the structure of political opportunities which produced a dramatic increment in the level of mobilization (Hess and Martin 2006; Flacks 2004). They were also “moments of madness”, to use Zolberg’s phrase, in that the mobilization produced intense confrontation between many people who shared great hopes of drastic social change and many others who at the same time were very much afraid of suffering reprisals (Zolberg 1972).12

These tensions, characteristic of the initial phase of a cycle of protest, outlined the political context in which anarchists mobilized themselves. In 1931 they did so in the hope of giving greater social content to the nascent Second Republic, for example, by pressing the new regime for a “radical solution to the agrarian problem” (la tierra para quien la trabaja, the land for those who work it), a demand the CNT had been making since its foundation. Euphoria was also felt by anarchists exiled in France in 1944 not just because of the end of the Second World War but because they thought the victorious Allies would not stop at the Pyrenees and would go on to overthrow the government of Franco, a regime constituted with the help of the Axis, as it had been characterized by the Potsdam Conference.13 Many exiled anarchists mobilized to encourage the Allies down this path. Finally, the death of Franco in 1975 not only meant the disappearance of the person who had run the country with an iron fist for forty years but also that of the man who had controlled the power struggles between elites, the Francoist “families”, especially fierce in the last years of the dictator’s life. With the equilibrium broken, the elites strove to acquire a better position in the struggle for power and finally accepted a move towards democracy, even if in a weak and uncertain manner in a context in which political forces that had had to work underground for decades mobilized outside the framework of the institutions, alongside new social movements that had sprung up at the end of the dictatorship.

In the three political transitions visibility was a synonym for strength and the disintegration of the former mechanisms of repression allowed the anarchists to obtain new and more visible spaces for mobilization. However, due to the loss of influence it had suffered while operating underground the anarchist movement cannot be considered to be what McAdam (1995) has called an “initiator” of the cycles. The anarchists did not have a decisive influence on the opening up of political opportunities, a process advanced more by other actors and factors, or at least by a heterogeneous coalition of these. Once the new opportunities had opened up the anarchist movement actively participated in the sequence of mobilization which, propelled by euphoria, sustained itself through an overall growth in resources. Though the clandestine nature of the movement in the preceding period means that there is a lack of credible data relating to the resources then available to it everything points to a considerable growth in the number of members and sympathizers following the regime change. In 1931, the focus of recruitment was mainly landless peasantry and the industrial proletariat; in 1975, young people attracted to anarchist ideals. In the case of postwar France, however there were no new recruits; the revival was manned by old activists, many of whom had been operating underground in territory occupied by the enemy or who had served in combat against it.

If the protest cycle model allows for greater understanding of the social movement sector once the political transition had begun, Melucci’s two-poles model outlines the change that occurred inside the movement with respect to the previous period. As we have seen, the new political opportunities allowed the anarchists to merge from the catacombs and find a new visibility in the general wave of mobilization. According to the model, alternation between phases of latency and visibility is inherent in new social movements. Their continued existence and effectiveness depend on a series of conditions. These include diversity in the environment of the movement, elasticity in the political system and the existence of transitory organizations (Melucci 1994; see also Klandermans 1992). In the Spanish anarchist case, retraction and latency were not a choice but rather a situation imposed by a context of repression: perse- cution and illegality during the dictatorship of Primo de Rivera, the hounded members of the resistance during the Second World War and attempts at annihilation during the Francoist dictatorship. However, and although each case had its own special features, this was overcome by i) the strong commitment of a highly politicized minority, ii) a clandestine organizational structure which, though defective, ensured a minimum of internal contact, iii) permeability to a range of stimuli in the environment, among them the develop- ment of other political forces, public discourse and the influence of libertarian intellectuals and pro-anarchist international protests, and finally iv) the material conditions of possibility offered by the political processes of democratization, including the variant of this which followed the Allied victory in France.

With regard to the question of innovation, a comparison of the three ruptures shows how it took the form of a change of values, interests and/or sensibilities in the heart of the movement during the latency phase prior to the political transitions. The case of Spanish anarchism corroborates the hypothesis of the two-poles model and at the same time criticizes that of the protest cycle model. The emergence of symbolic innovation can be explained on the basis of the experience of a crisis inside the movement and the creation of the means to overcome it.14 In the first episode, innovation took place at the heart of the CNT, whose secretaries general, as well as other trade unionists, struggled to strengthen and transform it into an organization of influence which would lead the way out of the crisis then being experienced by the anarchist movement. This is the “trade union alternative” which Elorza (1973) examined in his work on the Second Republic, but which began to appear clearly during the dictatorship of Primo de Rivera when its representatives sought to distance or even completely remove libertarian communism from the CNT’s programme in order to avoid having the organization banned. The groups that were members of the FAI critically confronted this move. The foundation of the “specific” anarchist organization can be indeed interpreted as a “spin off” of this change of direction (see Whittier 2004). Those who supported the FAI demonstrated their dissatisfaction with the direction being taken by the CNT and sought to change it. Matters were more complex in the second episode. The Spanish Civil War had been a profoundly traumatic experience not only for its violence and the conflicts within the movement but also for the bloody repression and difficult exile that the end of the war ushered in. In conditions that were very unfavorable for communication different groups arrived at individual conclusions. The successive underground national committees inside Spain appealed to the “exceptionalness” of the situation to subordinate the final revolutionary goal of anarchism to the more immediate one of national liberation from the cruel regime of Franco. In this context they opted for the anti-Francoist alliance as a means of action and also participated in the government of the republic in exile. One sector of those in exile supported this position but another rejected it because it saw the renouncing of the principle of the movement as being one of the causes of the recent disaster in the war. Finally, innovation developed towards the end of Franco’s rule with the arrival on the scene of new groups of young anarchists influenced, though in a scat- tered and weak manner, by the visibility during the 1968 international protests of an anarchism that was to a considerable degree renewed. These groups represented a symbolic alternative to the original movement, which in those years was practically inoperative and dispersed across a constellation of parallel committees struggling to control it. In the three episodes examined here, the cultural change (in the form of movement innovation) came before the social change (in the form of political transition), the expectation and promotion of which had strengthened an adaptive evolution in the movement.

Innovation led to internal conflicts. Between 1927 and 1931 this took the form of a clash between those who favored strengthening the trade unions and those who supported the anarchist trabazón; between 1943 and 1945 the anarchist underground in Spain favored an alliance with other opposition forces, later this was supported by some in exile and opposed by others who favored the classic isolationist position; and from 1968 new anarchist groups sprung up that were the carriers of new sensibilities that exceeded the bounds of the economic perspective of anarcho-syndicalism.15 It might not be going too far to say that the non-resolution of the internal conflict was partly an unintended consequence of the repression. This kind of outcome has been studied in the social movement sector, for example with regard to the facilitating of coalitions between diverse organizations (Chang 2008). The case of Spanish anarchism suggests an examination of the internal situation of the movement, where the repression may have impeded a long peaceful and open debate which would have produced a unified position. No such debate occurred, in part because of the repression and the elimination of spaces for both confrontation and dialogue, all of which kept the internal confrontation alive until the initial phase of visibility, when it grew more radical.16

In the three cases analyzed the transition to greater visibility occurred during the start of the waves of mobilization, times of great activity in the social movement sector, in terms of relations with other movements, but also in terms of processes activated inside the movements themselves. In this context the first two ruptures produced similar patterns of events, which refutes the observation by the protest cycle model that associates strategic competition with a reduction in mobilization potential during the final phase of the cycle. Both ruptures occurred in the midst of a contrary dynamic, one of general growth in the number of members and the quantity of resources; an ascending curve that the ruptures contributed to pulling down in the Spanish anarchist case.

What factors were involved in the process? In the first place, state repression, to which reference has already been made. This compounded the difficulties arising from the lack of agreement among positions taken on the symbolic change related to innovation. In the second place, the substantial increase in material resources available at the start of the visibility phase that encouraged the rival groups to fight for control of the movement, which was growing in influence and importance. This dynamic is similar to that pointed to by Zald and Ash (1966: 337), who saw how “[a]s movement organizations approach gaining power, latent conflicts over means, ends, and the future distribution of power, which have been suppressed in the general battle, rise to the fore.” The case of Spanish anarchism shows how the activation of internal conflicts, or the transition of these conflicts from latency to visibility, which amounts to the same thing, does not require the success of the movement but simply a similar transition within the movement itself with the dynamic of growth that goes with it. The visibility of internal conflicts and the sustained inability of the movement to cope with this, led to an increase in public awareness and finally to many activists and potential sympathizers becoming disillusioned and drifting away from the movement.

A third factor has to do with internal procedures, more specifically with congresses. The ruptures all occurred just after the holding of a congress. My hypothesis is that the calling of meetings of members at the start of the new period of visibility helped transform the symbolic differences that had arisen from cultural change in the period of latency into open competition for control of the movement.17 The congresses can be seen as spaces that were designed to integrate diverging positions into agreed messages for mobilization but this is not how they turned out. Repression by the state created enormous difficulties in the preparing of the ground for agreement and meetings, hurriedly called and poorly organized, which either radicalized existing internal conflict (1931) or erroneously closed it off (1945). In these meetings the existing disagreement between conflicting positions developed during the period of latency presented themselves. They were arenas for the presentation and representation of the internal conflict. During the discussions, factions radicalized their positions and many defended their visions by declaring their rivals to be enemies. The final result, with one side of the struggle splitting off from the main body of the movement showed what little flexibility there was in the movement for dealing with internal dissent (see Valocchi 2001), in spite of the supposedly traditional anarchist mechanisms for deliberation.

The fourth and final factor has to do with the symbolic resources of the movement, more specifically their scarcity and the consequences of this for the dynamic of confrontation. The internal struggle always centered around two dominant conflicting positions. The anarchist movement was not able to gather together enough symbolic resources to have more than two intellectual positions in debate at a time. The absence of a plausible third position generated destructive mutual antagonism.18 The adding of symbolic resources is exactly what distinguishes the third episode from the two previous ones. As in 1931 and 1945, the 1979 rupture was preceded by a cultural change that occurred in the period of latency. In this case it had developed towards the end of Franco’s rule with the arrival on the scene of new groups of young anarchists. As in the previous cases repression placed difficulties in the way of an internal debate aimed at unifying points of view between the younger and older generations of the movement, but it is important not to overlook the role of the change in generations. In any case, with the demands of the young people the anarchist movement was able to incorporate sufficient new symbolic resources to prevent the internal struggle turning into a clash at the strategic level. This interlude, however, was brief. The young people were an important force in the reconstruction of the CNT during the transition, even though they decided to dissolve their own groups in the process. This process deprived them of a platform from which to set out their position in a unified way, the much longed for “ideological transformation”, and this in turn deprived them of the possibility of formalizing the cultural change that had arisen in the latency period into a plausible alternative. Once the disillusioned young people withdrew from the debate, the door was opened for the reproduction of the strategic confrontation, in this case between those who were for and those who were against participation in union elections. Their positions became visible and grew more radical during the 1979 congress, which, just like in the previous two episodes of rupture, served as an arena for the representation and formalization of the rupture.

Conclusions

This article shows the degree to which environmental factors were important in the dynamic of factionalism and rupture within the Spanish anarchist movement; a dynamic which began under repressive regimes and finished during the political transitions to more open and tolerant regimes. In the first stage the repression favored both the formation of groups and their isolation. Furthermore, being obliged to operate underground brought with it an internal crisis, a search for solutions to it and hence the elaboration of distinct positions. The repression impeded internal debate, made communication between groups difficult and it was not possible to accommodate the new positions. Later the beginning of the political transitions, with their consequent opening of the structure of political opportunities, caused a sudden and accelerating increase in the material resources available to the movement. This converted what had until then been a symbolic conflict around divergent strategic positions into a power struggle between factions intent on leading the movement in the new and at the outset promising era. Returning to Madison’s metaphor, it might be said that if repression made it hard to put out a fire lit in the heart of the Spanish anarchist movement under undemocratic regimes, the winds of liberty produced by political transitions stoked it and eventually produced an explosion which destroyed the movement.

As well as these exogenous factors certain features and procedures of the movement aggravated the dynamic of rupture. The mass conferences convened at the start of the political transitions served as catalysts for the power struggle between the factions. The closing of the congresses represented the beginning of the process with one of the parts splitting off, or trying to take control by unorthodox means after the success or failure of which the losers would split off. A second endogenous factor had to do with the symbolic capital of the movement being limited to two antagonistic strategic alternatives, which channeled the conflict down an alley from which it was difficult to find a way out. In this regard, the 1979 episode had special feature: the intervention of an alternative third position, which, while it lasted, inhibited conflict between radicals and moderates. When this position dissolved itself the path was smoothed for antagonistic political conflict the outcome of which, as in the two previous episodes, was rupture.

To what degree does the Spanish anarchist case expand or refute the theoretical models used in this study? In the first place the comparison of the various episodes of rupture corroborates the validity of the alternating phases of latency and visibility model beyond the usual borders of its application (the new social movements). The old Spanish anarchist movement was moved by the alternation. The leap to the phase of latency, here identified as the move underground, was not a decision taken freely by the movement but rather a sine qua non condition for its continued existence. The Spanish anarchist case shows how the leap in the opposite direction, towards visibility, was a complex process and crucial for the final rupture. Furthermore, the case of Spanish anarchism is not adequately explained by the protest cycle model which, though it has been able to explain what occurred at the level of the social movement sector at the start of political transitions, has two biases relating to intra-movement dynamics. The first has to do with innovation, which the model associates with the initial phases of the cycle. The Spanish anarchist case shows that innovation, in the form of internal tendency which thinks and tries to come up with new strategies to resolve the internal crisis, took place at a largely different period, the latency phase, thus corroborating that observed by the two-poles model in this regard. The second bias has to do with strategic competition, which the protest cycle model associates with the reduction in mobilization potential in the final phase of decline. To the contrary, in two of the episodes reviewed here, those of 1931 and 1945, this completion was radi- calized by the rapid growth in material resources and intensification of mobilizations produced by the start of the transitional processes.

• 1The birth of Spanish anarchism as an organized movement was, however, not with the CNT. It is usually credited to the visit to Madrid in November of 1868 of Giuseppe Fanelli, a delegate of Bakunin’s International Alliance of Socialist Democracy and the subsequent creation under his influence of the Spanish Regional Federation, the Spanish section of the International Workingmen’s Association (IWA). See Lorenzo (1974). By “Spanish anarchist movement” I will refer to the conglomerate of organizations, associations and publications grouped around the CNT. The other major organizations were the Federación Anarquista Ibérica (FAI), the Federación Ibérica de Juventudes Libertaria (FIJL), and the Federación Nacional de Mujeres Libres.
• 2It should also be said that while the repression declined it did not disappear. Anarchists continued to have problems with the authorities, they were still  watched and sometimes certain groups and organizations were banned.
• 3McAdam (1995) asociates innovation with the action of “early risers”. In this regard, Kim (2010) suggests that the cycles of protest model suffers from a bias in that it ignores innovations that might occur in phases other than the initial one (see also della Porta and Diani 2006; Whittier 2004).
• 4This and further translations from Spanish are made by the author.
• 5The First and Second CNT Congresses were held in September 1911 (in Barcelona) and December 1919 (in Madrid) respectively. For the Third Congress, see Casanova (2004).
• 6It seems that there was also a difference between moderates and radicals in terms of age, with the latter being the younger (González Calleja and Souto 2007).
• 7Report, CGML branch in England, 20 August 1939 (International Institute of Social History, Amsterdam [henceforth IISH], Fernando Gómez Peláez papers [henceforth FGP], 802).
• 8Report, Comité Nacional, Toulouse, 28 October 1944 (Fundación Salvador Seguí, Madrid, Archive Historia Oral); Letter from Esgleas to Molina and “the other companions of the so-called Comité Nacional” of the MLE-CNT in France, 29 December 1944 (IISH, FGP, 981).
• 9The MLE had its original form as the comité nacional de enlace (national liaison committee) created in the Pleno Nacional Conjunto (joint national meeting) held by the CNT, FAI and FIJL in Barcelona in October 1938. Its intended purpose was to become “an organism of cooperation and political advice for the Libertarian Movement” (IISH, Archive FAI-Comité Peninsular, 59.4; Meeting proceedings at the IISH, Archive FAI-Comité Peninsular, 46C.4–5).
• 10Circular letter number 36, Comité Nacional in France (IISH, FGP, 765). Of the general spirit of the congress, Manuel Buenacasa, one of its organizers said “passion predominated over reason . . . even though motives related to ideals, tactics and doctrine were wildly invoked.” (“El famoso Congreso de París y sus consecuencias”, Hoy, Marseille, 22 June 1946, year 2, issue illegible).
• 11These were agreements signed on October 25th between the government, the main political parties represented in parliament, business asso- ciations and the Comisiones Obreras trade union with the objective of achieving political stability and an end to the economic crisis. One of the most serious episodes of repression is known as the Caso Scala. It involved an attempt to criminalize the anarchist movement by accusing it of being behind a fire in the Scala function room in Barcelona which cost four lives and which occurred after a demonstration called by the CNT against the Pactos on January 15th, 1978.
• 12Sartre, writing about what he had seen on the 22nd of August, 1944 (“L’Insurrection”, published in Combat six days later), used a different simile: “The street has once again become –as it did in 1789 and 1848- the theater for great collective movements and social life [. . .] In this time of drunkeness and joy everybody feels the need to pludge back into the collective life” (quoted in Contat and Rybalka 1974: 101). I am grateful to José Luis Ledesma for the reference. Enrique Marco Nadal, Secretary General of the CNT in Spain between May 1946 and May 1947, illustrates the tension between hope and fear when he writes about what was expe- rienced in 1975 in his memoirs, “When Franco died everyone in Spain let out a sigh; of satisfaction in the case of those who had lost out during the dictator’s rule; of uncertainty in the case of those who had lived well from it” (Marco Nadal 1982: 321).
• 13For the role of emotions in the continuity of anarchist activism in Franco’s Spain, see Romanos (2011).
• 14On crisis and innovation, see Kim (2010).
• 15It must be realized that factionalism did not only arise from ideological and tactical issues. It was also the product of “resentments, ambitions, responsibilities and mental routine”, as was pointed out on the occasion of the second rupture in the report sent by José E. Leiva to the CNT and FAI national committees (Paris, 6 February 1946 [IISH, Ramón Álvarez Palomo papers]), Personalities played an important role but once they had aligned themselves with one side or another (sometimes in a leadership role) personal rivalries were diluted into the ideological and tactical discourse and it was the discourse that became the principal battleground for the disputes.
• 16During the latency period the whole of the movement was the target of state repression. This differs from repression in democratic contexts in which, according to the cycles of protest model, it is concentrated on the most radical sectors and is associated with the final phase of decline.
• 17The process corresponds, to use the language of the literature on social movement outcomes, with the transformation of internal effects of a cultural nature (value change) into internal effects of a political nature (power relations) (see Giugni and Bosi forthcoming). It is worth point out that, being cultural and political effects, the ruptures had a strong impact on the lives of those who participated in them. Some chose to write memoirs to give their side of the story and, sometimes, to settle accounts with the past.
• 18Sometimes the polarization (and vilification) may be the result of strategies which in an intra-movement framing contest, “accentuate differences and draw sharp ingroup/outgroup distinctions” (Wiktorowicz 2004: 165; see also McCaffrey and Keys 2000).