Abstract
This article deals with the management of public order during a crucial period of Spain's twentieth-century democratization process. A regional case-study is employed in order to analyse the detail of political violence and gain a clearer understanding of both the participants and the role of the police and the civil governor. It is based on an exhaustive database containing all the episodes of violence that led to deaths and serious injuries. Quantitative research, combined with adequate contextualization, is vital to unravel the evolution of this violence and identify the main actors. This article pinpoints important differences in the case of Asturias – compared to the nationwide data available – and contributes to the academic debate on the authorities’ responsibility for the rising toll of victims during the spring of 1936. It shows that police participation in the generation of victims was of minor relevance, partly due to a deliberate policy pursued by Asturias’ civil governor, which ruled out the preventative and recurrent use of the police to control conflictive situations.
On 16 February 1936, the third general election was held of the short-lived Spanish Second Republic – born in April 1931 and considered by specialists as ‘the first Spanish democracy’. 1 The general framework of this research is the government's management of public order and political violence during the spring of 1936. Close attention is paid to the figure of the civil governor, the main authority in charge of public order in the province – the most relevant unit of the country's territorial division and under the authority of Ministry of the Interior. While the focus is on a relatively short time period – from the formation of a new national government on 16 February 1936 to the 17 July coup d’etat which precipitated the civil war – these months are particularly relevant to understanding the difficulties democratization generated in twentieth-century Spain. 2
The civil governor was the government's delegate in the province – a political appointment designated by the national government in Madrid. In addition to being responsible for the police, the civil governors were the national government's eyes and ears in each province. Although they received and carried out orders from the ministry of the interior, the civil governors enjoyed relatively broad freedom to make decisions on issues of public order. For instance, they exercised authority over a range of matters that affected individual rights and freedoms: the authorization of political rallies and protests; the holding of public meetings; the authorization and supervision of strikes; police recordkeeping and police checkpoints on the streets; and the issuing of weapons licences. 3
According to the public order law of 28 July 1933, there were three levels of state of emergency: prevention, alarm and war. The intermediate level was in place throughout the spring of 1936 – from 17 February to 17 July. 4 This state of alarm could be decreed by the government when ‘state security’ required it ‘in the event of evident and imminent danger’, but the government then had to justify this action to parliament. 5 While in force, civil governors possessed the power to apply a priori censorship and to disperse protests and gatherings using force if required. This was in addition to two other powers that are vital to the functioning of a democracy: firstly, they could order gubernatorial detentions, which deprived citizens of their freedom without the need to bring them before a judge; 6 and secondly, they could keep individuals in prison who had been freed by a court of law. 7
The position of civil governor had traditionally served to reward the elites of the parties that formed the national government. However, despite their party affiliation and their hierarchical subordination to the ministry, the governors were challenged to develop autonomous provincial leadership. The Spanish spring of 1936 was a particularly complex period when it came to party mobilization, democratic coexistence and the exercise of public freedoms. For this reason, it provides a fitting context for an analysis of how the new governors made use of their authority. 8
Asturias was selected as a case study due to its great relevance to republican politics between 1934 and 1936. In October 1934, the Spanish Socialist Workers’ Party (PSOE) responded to the ascent to government of the Confederation of Autonomous Right-Wing Groups (CEDA) with a revolutionary movement. They considered that this right-wing Catholic coalition was an enemy of the republic. In Asturias the ‘October rebellion take the form of a prolonged assault on state power and an experiment in revolution’. 9 Over 80 per cent of the total deaths recorded in Spain (1051 civilians, 111 civil guards and 173 members of other government forces according to official statistics) occurred in that province. 10 It was ‘the most important episode of revolutionary upheaval in Europe between the early 1920s and the Spanish Civil War’, according to Kerry's recent research (2020). He also points out that there were other episodes of mass left-wing mobilization in Europe during that decade, such as Austrian socialists’ rebellion against the authoritarian drift of Chancellor Dollfuss, but ‘these defensive reactions did not match the scale and revolutionary pretensions of the Asturian insurrection’. The Socialist movement in Asturias was ‘the last attempt at the seizure of state power via a mass armed insurrection by the working class in Europe’. 11
The legacy of conflict that stemmed from the failed revolution of October 1934 and the subsequent repression was decisive for the evolution of Spanish politics in the year and a half that followed. The repression expelled the Socialists from institutions and public life, deeply weakening their trade unions. Another consequence was a greater radicalization of right-wing narratives and the weakening of liberal and moderate republican leaders. At the beginning of January 1936, a general election was called. During the campaign, the Asturian revolution was the main topic of debate. The Popular Front, the coalition of almost all left-wing groups – including the PSOE, the Communist Party of Spain (PCE) and the two main parties of the republican left, Republican Left (IR) and Republican Union (UR), defended the legacy of the 1934 revolution and demanded a general amnesty. The opposing front of the right-wing parties aggressively denounced the threat of a revolution in the making as ‘anti-Spain’. 12
The relevance of Asturias as a case study has also other facets. In first place, the legacy of October 1934 ‘and the long 1935 conditioned the nature of politics in the province’. 13 An important consequence was the problematic position of the police in general, and the Civil Guard in particular. They were strongly brought into question after the Popular Front's victory because of their role in the repression. In second place, socialist trade unionism had a long tradition in the province and was particularly well-established in coal-mining, the main sector of this industrial region. Socialist mineworkers had created SOMA (Mineworkers Union of Asturias) in 1910, which became ‘a powerful voice for the coal miners’ in that decade. Their membership doubled in the first two years of the Second Republic: in 1932 they claimed ‘69 per cent of the workforce’. Without a doubt, they were the ‘main union force in the coal mines’. 14 In third and final place, the Asturian economy was experiencing a serious structural crisis, which had its roots in the past and impacted upon the mining areas, bringing with it major social consequences. 15 Thus, the economic crisis was added to the social crisis that derived from the repression of the October movement.
The relevance of Asturias as a case study also lies in the high number of episodes of political violence that occurred in the five months before the civil war. This research, as explained in detail below, has confirmed 125 victims between 20 February and 17 July 1936. The intensity of political violence in the province is striking: a rate of 0.16 victims per thousand inhabitants. Compared to the nationwide average (0.09), this is a very high figure. In absolute terms, only Madrid, a very conflictive province and the location of the country's main city, had a greater number of episodes and victims than Asturias. 16
Finally, to broaden the focus of a specific case allows us to better contextualize the decisions taken by the authorities and to unravel the circumstantial elements that surrounded police action. In the case of Asturias, the leadership of a new governor, Rafael Bosque, who arrived in the province in March 1936, proved to be deeply controversial and resulted in his resignation in early July, after a public confrontation with the main leader of the anti-republican right-wing, the monarchist José Calvo Sotelo. This article examines the following hypothesis: unlike other provinces, the gubernatorial policy in Asturias is crucial to explaining the reasons for the relative lack of police participation – in the generation of victims of political violence – compared to nationwide data currently available. The purpose is also to contribute to the academic debate on the provincial authorities’ responsibility for the rising death toll during the spring of 1936. 17
Between 17 February and 17 July 1936, political violence reached extremely high levels in many areas of Spain. During these months, the right reported a death toll of more than 200 among its ranks, a figure that would be subsequently reproduced in the Franco dictatorship's propaganda. Nevertheless, there was no published academic study of these deaths until 1978; there the author identified 273 victims between 31 January and 17 July 1936. 18 For many years there was no further research on these figures 19 , until Cruz (2006) estimated a similar death toll of 262. Shortly afterwards, Blázquez Miguel (2009) surprisingly increased the figure to 454 deaths and 1686 injured, based on a painstaking but methodologically flawed study of the press. Overall, the study that carries the greatest academic weight at the time of writing is that of González Calleja (2015), which identified 384 deaths, although it did not include an estimate for the number of injured. 20
Moreover, the latter work reflects a recent tendency to combine rigorous quantitative data with a qualitative analysis that includes information such as the party allegiances of victims, cause of death, the geographical and chronological distribution of deaths and the role of police intervention. Regarding the latter, both Cruz (2006) and González Calleja (2015) have maintained that the violence was exacerbated by mismanagement on the part of the forces of law and order, and therefore the authorities: there was ‘a disproportionate deployment of force, rather than of authority’. 21 This is a thesis that furthers the notion of state brutality as a legacy of the monarchy and the failure to modernize the police. In part, it also follows the argument of Ballbé (1983): the republican state maintained repressive practices and a militarized understanding of public order, an approach that proved to be incompatible with the demands of a more active citizenry. 22 Del Rey (2007) added nuance to this claim: one thing is that the police ‘found themselves embroiled in the majority of “the confrontations” …, but it is quite another to conclude that these were provoked by the police’. 23 Furthermore, Blaney (2007) brought into question the theory of the brutalized state, instead attributing the death toll to the conduct of two actors: the Falangist – members of Falange Española (FE), the fascist party – and the revolutionary wing of the working-class left – socialists, communists and anarchists. 24
As a contribution to this interesting debate, this article demonstrates that case studies based on archival and primary sources can provide us with more detail on concrete incidents, thereby improving our understanding of ‘the exact circumstances that surrounded the violence’. 25
The Popular Front coalition achieved a narrow victory in the February 16 election. However, in Asturias the left's victory was clear: its candidates obtained 20,000 more votes than their opponents – 53 versus 47 per cent for the right-wing coalition. Thanks to the majoritarian electoral system, the Popular Front won 13 of the 17 seats in the province. The distribution reveals the preeminence of the socialist organization in Asturias: PSOE (6), IR (3), PCE (2), Left Independents (2). The right-wing coalition won the four minority seats: CEDA (3) and the Liberal-Democratic Republican Party (PRLD) (1). In this case, the strongest party was the possibilist Catholic right. The PRLD, a historic party of republicanism during the Restoration monarchy that had moved to the right and was clearly anti-socialist, maintained notable influence in several cities in the province. 26
The first public-order problems arose in Asturias in a context similar to that experienced in other parts of the country: in the aftermath of an ‘impassioned’ electoral campaign and troubled count, a new government was formed on the afternoon of 19 February, led by Manuel Azaña, leader of the Republican Left (IR) and one of the founding fathers of the Second Republic. 27 Almost at the same time, there was a broad mobilization of the supporters of the victorious parties, who took to the streets not only to celebrate their victory, but also to demand rapid measures and political changes in local councils. Conscious of these tensions, Azaña prepared a message of pacification to be broadcast on the radio to the whole country. 28
At the same time, the new government extended the state of alarm, thereby placing exceptional measures at its disposal, but at the same time rendering unlikely a victimless police response to the left-wing mobilization. Article 38 of the public order law empowered the ‘authorities’ to ‘prohibit the gathering of all classes of groups on the streets’, adding that ‘the use of force to restore order’ was permitted where there was ‘a failure to obey’ police orders after ‘three warnings’. Moreover, the law made clear that ‘prior warning would not be necessary if the forces of law and order found themselves under attack’. 29
On 20 February, a few hours after the new government had assumed power, some 5000 people from Oviedo's working-class neighbourhoods gathered outside the prison amid rumours that the doors were to be opened. Under pressure, the governor ordered the release of around 400 prisoners. The situation was even tenser in the province's main port city, Gijón. Here crowds also amassed in the streets, laying down the gauntlet to the prison governor, who eventually allowed the doors to be opened and ordered the soldiers who guarded the facility not to use their weapons. While a riot was averted, along with a police charge that could have produced numerous victims, dozens of common prisoners took advantage of the situation to escape, something which alarmed even the prime minister. 30
By the evening of 20 February, the civil governor claimed that calm reigned throughout the province. 31 This was not entirely accurate. In Oviedo there were a number of incidents. A young right-winger was assaulted and there were various attempted attacks on the premises of conservative parties, but the most significant violence occurred as a result of clashes with the assault guards – the police force created by the republic with the aim of modernizing the control of public order, particularly in urban settings. Although some assault guards were injured, these incidents were not as serious as those seen in other Spanish cities. There was also an attempted attack on a provincial newspaper, which was averted thanks to the intervention of local socialist leaders. 32
These events establish a preliminary fact that may be surprising, given the recent history of Asturias and the aftermath of October 1934 revolt: the opening hours of the post-electoral mobilization failed to produce either deaths or serious injuries. In fact, clashes with the police were neither generalized nor led to fatal consequences.
Nevertheless, political violence erupted in Asturias with great force during the days that followed. There was a first phase, which lasted until the third week of March, in which acts of aggression or attacks within the trade union sphere predominated: basically these consisted of reprisals against workers deemed to be sympathetic to employers who were considered to be ‘scabs’ by socialists, communists or anarchists. Apart from the powerful socialist miners’ union (SOMA), anarchists were well-established in the port city of Gijón and in the iron and steel industries in the town of La Felguera. However, the anarchists of the CNT (National Confederation of Labour) were weak in the mining sector and their union – the Mineworkers’ Single Union (SUM) – had very few members. The communists (PCE) made some progress among the coal miners during the spring of 1936, although they were a minority compared to the socialists. 33
The first phase of violence was clearly connected to preexisting, entrenched labour disputes. Significant tension had resulted from the application of the emergency degree passed by the national government, which mandated the rehiring of workers that had been laid off in the aftermath of the October 1934 general strike. In turn, some employers refused to dismiss the workers they had hired to replace those on strike; however, for the left, those workers were strikebreakers or ‘scabs’ who deserved to be punished. At the same time, some of those who lost their jobs, falling victim to the political changes, joined the ranks of reaction and readied themselves to inflict violence on their adversaries.
On the other hand, the months of March through May saw extremely high levels of labour conflict in the province. There were numerous strikes, including the prolonged stoppage of the Gijón dockers, and these contributed to both fuelling tensions and increasing the likelihood of violence. 34 Therefore, the labour market and the trade unions’ struggle to control it, compounded by ideological factors, was one of the fundamental sources of the violence seen during the first phase. Moreover, this phase included one of the attacks that had the greatest impact in Gijón. It took place on 29 February, on a tram full to the brim with passengers; two armed left-wingers shot a foreman from the employers’ loading and unloading team at the El Musel port at point-blank range. The victim, Eduardo López Peña, was aged 51 and had previously been on the receiving end of threats. His murder was an extreme manifestation of the violence that affected a number of Gijón's industrial workers. On the one hand, it was an act of revenge, but it was also a warning shot in the context of the exclusion of those workers who had replaced the strikers of 1934. 35
Along similar lines, there were some violent incidents in the weeks that followed the election whose purpose was clearly evident: to undermine the working-class organizations linked to the Catholic church, which had been the main beneficiaries of the socialists’ exclusion after the revolution of 1934. On March 7, with the cry of ‘At them, for they are fascists’, a group of young leftists attacked the offices of a union affiliated with the Catholic Confederation of Spanish Workers’ Unions (CESO) in Avilés. This was followed by various acts of aggression against the secretary of the Port Board of Works, two industrialists and at least two young right-wing workers. 36
As a result, the tumultuous violence characteristic of the weeks immediately following the election was not of great significance in Asturias – for instance, the violence frequently seen in other provinces which stemmed from unauthorized gatherings or from attacks on property or the premises of political parties, and which led to clashes with the police. In Asturias the settling of scores among trade unionists and acts of aggression rooted in partisan differences predominated, as the following examples demonstrate. On 2 March, a young catholic was beaten up in Oviedo, having fended off an attack carried out by two other young men. This was a clear act of revenge, given that both of the attackers had been victims of the post-October reprisals, with one imprisoned while the other fled. 37 On 8 March, the body of a railway worker associated with the right was found in Llanes; he had been attacked the night before after an argument with political opponents. 38 That same day also saw a farmer from Soto del Barco, who was also a rightist, seriously injured. 39 Only hours later, a young rightist suffered serious stab wounds at the hands of a socialist neighbour in Sotondrio. 40 10 March saw two residents of Ciaño, on this occasion socialists, suffer firearms injuries at the hands of local political adversaries. 41 In the midst of these intense weeks, on 13 March the President of the Cudillero Fishermen's Guild died as a result of stab wounds inflicted in a clear act of revenge that was driven by a mix of political and labour-related motives. 42
The number of victims rose further from 14–20 March, all of these subject to acts of aggression and attacks, such as that carried out against a Falangist leafleteer in Gijón. 43 Others were the result of politically motivated arguments or premeditated attacks that descended into brawls, such as the 14 March murder of a leftist in Sama de Langreo at the hands of an individual whom the victim had previously rebuked with accusations of being a fascist. 44 Only a few hours later, an attack on a socialist worker in La Foz de Morcín after an argument with a worker affiliated with the Catholic miners’ union also proved to be fatal. 45 Bringing to a close three weeks marked by numerous acts of violence, 15 and 16 March saw two further victims: firstly, a right-wing resident of Mieres died after receiving various shots in what was a clear act of revenge; 46 and subsequently the school teacher in Carabanzo was shot by two local right-wing brothers who forced their way into his house at night and murdered him in cold blood in front of his family. 47
On 20 March, an important change took place which would impact the development of public-order policy in Asturias. This was the arrival of a new civil governor, Rafael Bosque, a member of Azaña's Republican Left (IR). While he had prior experience in the civil governments, he was unfamiliar with the region. The previous governor, José Friera, had remaining as acting governor for over a month, a situation that was exceptional when compared to the rest of Spain.
Bosque was well-received by the two main parties of the working-class left (the PSOE and the PCE), and during his first weeks in office he displayed empathy with socialist trade-unions demands in a number of key labour disputes. 48 Nevertheless, his tenure proved to be problematic, lasting until the first week of July. The right accused him of bias, in particular for his appointment of a number of gubernatorial delegates, who supplanted the local mayors in their role as ancillaries to the implementation of public-order policy. 49
As regards the trajectory of political violence, a change was evident shortly after Bosque assumed office. From April onwards, the vast majority of victims were the result of aggression that was driven by strictly partisan considerations. There were still some victims of violence stemming from labour disputes, such as a building contractor who was shot in Muros de Pravia in early May, and an employer who suffered the same fate in Gijón in the second week of July. 50 However, from April onwards this type of incident became less common, in stark contrast to the previous month. On the other hand, there was a continued absence of large-scale confrontations with the police that produced seriously injured victims.
Overall, taking into account the changing number of victims that political violence produced, the spring of 1936 in Asturias can be divided into four clear phases.
1- The first of these phases coincided with the month during which Friera was acting governor, and includes the first days of Bosque's tenure, with 45 victims (12 deaths and 33 injured). The critical moment of this phase occurred on 22 March, with the well-known attack on the former Liberal-Democratic (PRLD) deputy Alfredo Martínez, who had been minister of labour in one of the governments of the conservative biennium. Martínez was shot dead by gunmen outside his home in Oviedo, in an act of reprisal for the death of a young leftist.
51
2- The second phase basically encompassed the month of April. Here the situation stabilized somewhat, with the data trending downwards, although there were still numerous incidents, with 24 victims in total (five deaths and 19 injured). 3- The third phase included the month of May and the early days of June; during this period the violence decreased notably. Up until 10 June there were 40 victims (three deaths and 35 injured), but 21 of the injured resulted from a single confrontation between assault guards and armed leftists in Oviedo on 24 May. For this reason, the total number of victims is misleading, given that the number of violent episodes fell. 4- The fourth phase began in the second week of June and ended on 17 July. It was characterized by an upturn in both violent incidents and in the number of deaths: 18 victims (six deaths and 12 injured) (Figures 1–3)

Total number of victims by month in the province of Oviedo.

Total number of victims by month in the province of Oviedo (excluding 21 injured on 24 May).

Deaths by month in the province of Oviedo.
This research has confirmed that between 20 February and 17 July 1936, there was a total of 76 episodes of political violence in Asturias. These produced 125 victims: 26 deaths and 99 injured with varying degrees of severity. Minor injuries have not been included here. 52 Until now there has been no specific research study of the whole political violence in Asturias during spring 1936. Kerry (2020) has a valuable, if brief, epigraph on the political violence during the spring of 1936 in Asturias. However, its main objective is not the analysis of global figures, but rather a limited qualitative approach. The figures arrived at here (26 deaths and 99 injured) are slightly higher than those published in the only nationwide studies that include data broken down by province: González Calleja (2015) and Blázquez Miguel (2009) computed 22 deaths and 18 deaths respectively (plus 98 injured in the case of the latter, who also counted these). 53 The difference between these totals and those arrived at here is in all likelihood due to the limited geographical scope of this study, which has allowed for the comparison of a greater number of primary sources. Moreover, this study's originality also stems from its inclusion of a source that has, until now, been widely overlooked in the study of political violence: the proceedings of criminal cases seen by the emergency courts in the province 54 (Figure 4).

Total number of victims by date in the province of Oviedo.
These 76 incidents with victims can be classified using a typology of generic actions as follows 55 (Table 1).
Typology of generic actions.
Practically all of these incidents (72 out of 76) took place as a consequence of partisan differences and, to a lesser degree, trade-union related conflicts. In very few instances were the police present, among other reasons because almost none of these 76 episodes took place in the context of mob violence, as was the case with attacks on party premises and religious buildings or illegal gatherings.
As a matter of fact, a crucial aspect of the maintenance of public order concerns questions about the role of the police: the conduct of the police; whether the use of force was proportional; whether they were guided by outdated, repressive criteria; or whether police methods produced avoidable casualties. In this context, the mobilization of the Popular Front's supporters, in particular of socialists and communists, and the proactive violence of the Falangists rendered the spring of 1936 a complex period. The tragic events of these months prompted public debate over whether police use of force had been disproportionate. This was particularly evident in cases involving the civil guard, such as those that occurred in Pechina (Almería) on 23 February, in Escalona (Toledo) on 8 March, in Cartaya (Huelva) on 23 April, and the much better-known events in Yeste (Albacete) on 29 May. These were all situations of extreme tension in the countryside, in which the civil guard confronted the people and used their firearms. There were deaths in all cases, but in Yeste it was very serious: a group of peasants brutally attacked a patrol of civil guards who were guarding some detainees and there were several casualties, including some civil guards. Subsequently, the civil guard raided the town and committed serious abuses. There were 18 deaths and 29 seriously injured – 15 of them civil guards, including one death. 56 It was the bloodiest clash during the spring of 1936 and the most important in which the civil guard was involved.
In the case of Asturias, the situation was compounded by the fact that the civil guard carried the stigma of its role in the repression that followed the revolution of 1934 – and the fact that ‘an entire region had risen up against [the civil guards] meant that for the first time they carried out a systematic program of instilling fear in that entire region’. 57 Here both socialists and communists demanded a thoroughgoing purge of its command, which they considered to be hostile to the republic. A communiqué issued by the Asturian Popular Front deputies demanded the ‘immediate removal’ of those elements of the civil and assault guard command deemed to be involved in the ‘repression’. In addition, the association of civil-guard officers and local ‘fascism’ was habitual in communist discourse. 58
In order to analyse the data on police involvement in the 76 cases that produced casualties, it is first necessary to identify those who participated in these incidents. Of the total of 125 victims, it has been possible to determine the party affiliation of 61: 28 belonged to the right and 33 to the left, with the following distribution (Tables 2 and 3).
Party affiliation of right-wing victims.
aThe Free Trade Unions encompassed trade unions allied with the employers.
Party affiliation of left-wing victims.
If the analysis is limited to deaths only, the research data is more complete as the identities of 25 of the 26 victims have been determined: 14 sympathized with or were members of left-wing parties, nine sympathized with or were members of right-wing parties and two were civil guards. Therefore, as regards the distribution of deaths, the data on political affiliation tends towards the left, with one-third more victims than their adversaries. Nevertheless, if the affiliation of the perpetrators is included, the distribution immediately becomes more balanced: both left and right produced 12 victims each, excluding one death at the hands of a civil guard and another at the hands of a priest.
Following other recent studies, the number of victims and perpetrators can be equated with responsibility for carrying out this violence. 59 Thus, the data above indicates that while the left accumulated around 50 per cent more deaths than their adversaries, they were the perpetrators in a similar proportion to the right. From this it could be concluded that the right was more proactive and as a consequence, held greater responsibility for triggering the violence.
However, this derives from a flawed methodology. By assuming a purported, but unproven, equivalence between the perpetrators and those responsible for the violence, it suggests a misleading causal relationship. It is evident that whoever uses a weapon against another person, including to the point of killing them, incurs both moral and legal responsibility, even when acting in self-defence. Nonetheless, this neither exhausts the analysis of political violence nor allows for a consideration of whether the perpetrators were those who had initiated the incident. For this reason, it is also vital to consider the detail of how each episode unfurled and who bore primary responsibility for the violence.
In the case of Asturias, primary sources enable us to identify who was responsible for initiating 62 out of 72 recorded episodes of violence, thereby allowing us to arrive at fairly reliable conclusions. The aggressor was also the instigator in 50 of these 62 cases, while the victim was the instigator in the remaining 12. The distribution by political affiliation is as follows (Table 4).
Party affiliation of instigators (total episodes).
aRadical Republican Party (PRR), the main republican group from 1931 to 1935, was a centrist party that practically disappeared after the 1936 national election.
If we add together the aggressors and victims whose affiliation has been determined, the affiliates or sympathizers of the left were responsible for instigating 41 incidents, which led to 19 deaths and 40 injured. In turn, the right initiated 11 incidents, which produced three deaths and 11 injured.
If we consider only episodes which resulted in deaths, the balance is similar in terms of who was responsible for instigating the events: the left provoked incidents that produced 19 deaths and the right were responsible for incidents that left four dead, excluding three instances in which the affiliation of the instigators is unknown (Table 5).
Party affiliation of instigators (episodes which results in deaths).
One pattern clearly stands out: the police were by no means major instigators of the violence. In the whole of the spring of 1936, the police intervened in only nine of the 76 episodes with recorded victims: this comprises four acts of aggression, three shoot-outs and two clashes with the police. Representing less than 12 per cent of the total number of incidents, these nine cases produced three deaths and 34 injured. It must also be recalled that the figure of 34 injured distorts the perspective, since as seen above, the shoot-out that took place in Oviedo on 24 May between leftists and the assault guards alone produced 21 of those injured.
During the spring of 1936, the civil guard incurred two fatalities in Asturias: one of these deaths resulted from shots fired by the occupants of a vehicle that had been ordered to stop, and the other was the result of a deliberate attack whose motivation was clearly political. To these we may add some cases of minor injuries sustained during the course of dispersing protests or breaking up clashes between rival groups. Furthermore, in terms of the perpetrators, the actions of civil guards produced one death and one seriously injured. While the assault guards were the protagonists of the aforementioned May shoot-out in Oviedo, they neither caused any deaths nor incurred any fatalities among their ranks.
It is possible that the scant police involvement in the episodes of violence with recorded victims in Asturias is a circumstantial exception in the context of a spring that proved to be deeply complex for the civil governors of other provinces. A probable withdrawal on the part of the police, particularly in the case of rural civil-guard posts, is a plausible hypothesis, given that these were aware that they were identified as the repressors of the labour movement after 1934 revolt. Nonetheless, this research has uncovered solid evidence of a relationship between the governor's policies and this limited police involvement.
In the case of Asturias, the first peculiarity was the absence of the vacuum of power that existed in many other provinces during the days immediately after the change of national government on 19 February. Although an appointee of the previous government, Governor José María Friera carried out the orders of the new minister and quickly facilitated the reinstatement of the municipal councillors that had been removed in 1934 and the application of the decree that mandated the reinstatement of workers that had been laid off. The absence of a power vacuum also removed an incentive for the victory celebrations to get out of hand: for example, descending into the attacks on political-party premises that occurred in other provinces, which in turn obliged the police to confront small, violent groups on the streets. 60
In addition, Friera allowed the pressure exerted by the left's supporters – especially in the case of socialists – to force the pace of some political changes, thereby avoiding the need to preventatively deploy the police. This occurred in two situations which could have otherwise provoked serious confrontations with the forces of law and order. Firstly, Friera elected not to order the police to disperse the mass protests outside the prisons in Oviedo and Gijón that occurred shortly after the change of government in Madrid. As mentioned above, in Oviedo hundreds of prisoners were released; in Gijón, a number of common prisoners managed to escape thanks to the speakers’ ability to capitalize on the mass gathering, the presence of explicit communist symbols and the open threat of an assault on the prison if the doors were not opened. The second of these situations occurred during the rapid changes that took place in the city councils of Gijón and Oviedo. On Saturday 22 February there was a demonstration in the Plaza de la República, with various Popular Front leaders in attendance. The governor ordered the mayor to swear in the reinstated councillors, but the protestors gained even more. Taking advantage of the pressure building up on the streets, a commission representing the demonstrators demanded control of the city council. The Liberal-Democrat (PRLD) mayor, who had been in office since 1931, was forced to accede to this demand against his will. While Friera was aware that the left lacked a majority on the city council and that the appointment as mayor of the Socialist Ángel Martín Pérez violated the law, once again he chose not to send in the assault guards to disperse the demonstration. 61
During the electoral campaign Friera had recognized that a victory for the left would render ‘inevitable’ certain ‘expansions’. It was as if he was already preparing himself not to use the police against the demonstrators. 62 In the eyes of the right, this constituted an abdication of responsibility. Yet the governor thereby avoided police involvement in clashes which would have produced victims on both sides. 63 Until he was relieved on 20 March, there were no recorded clashes with the forces of law and order which produced serious injuries. Specific circumstances produced police charges without tragic consequences, such as occurred in Oviedo railway station when a group of leftists attempted to disrupt an act of homage to a military detachment that was returning home. 64 As has already been seen, the majority of political violence was the result of aggression and attacks, reaching its peak in mid-March. Friera elected not to make intensive use of the civil guard to arrest this situation. However, this proved insufficient to gain him the backing of the left. Shortly before his departure, Friera was censured by Oviedo's Popular Front deputies. These remonstrated with the minister of justice that the violence of ‘these days’ in Asturias had been provoked by ‘fascist elements’ who acted with impunity because they enjoyed the ‘shameless protection of the public authorities’. 65
When it came to the selective deployment of the police, the new governor, Rafael Bosque, maintained a similar tone to his predecessor. Both the delicate situation of the Asturian economy and numerous labour conflicts greatly complicated his term in office. A detail description of his governorship is beyond the scope of this article. Bosque's tenure ended abruptly in early July, when the government withdrew its confidence after the controversy generated by a telegram that had supposedly been sent from the Oviedo civil government, and which contained harsh criticisms of the right-wing deputy Calvo Sotelo. 66 Nevertheless, it is worth considering certain data related to Bosque's relationship with the police and his gubernatorial policies.
One aspect that placed Bosque radically at odds with the opposition parties (CEDA and PRLD) and the Catholic media was the appointment of gubernatorial delegates in a number of towns. For several weeks, these carried out Bosque's orders and facilitated a policy of arresting conservatives which intensified from mid-April onwards. Under the state of alarm, gubernatorial detentions were legal and the civil government justified them as a brake on ‘fascist’ subversive activities. 67 However, this policy led to controversy as there were hundreds of arrests in the province, including many non-Falangist rightists and members of the clergy. The conservatives accused Bosque of bias, while the working-class parties, and the communists in particular, applauded the governor's policy. 68 Despite these tensions, there were no clashes with the police when hundreds of family members of the detainees organized a mass visit to the prisons in protest. Neither did discipline break down in the ranks of the civil and assault guards, who carried out the orders of the delegates and the provincial authorities by arresting Falangists and confiscating their arms, as well as searching their political premises.
The self-proclaimed ‘red guards’ were another potential source of conflict between April and June. These were socialist and communist patrols that carried out parapolice activities in the villages, engaging in illegal searches and checkpoints. In some places they also implemented boycotts by placing armed guards at the doors of commercial premises. 69 Such activities were the cause of numerous confrontations with young rightists, and Falangists in particular, leading to shoot-outs and acts of aggression that produced various victims. The presence of active ‘red guards’ in some Asturian localities is confirmed by the proceedings of various criminal cases for bodily harm and homicide. 70 Bosque, while aware of this situation, reacted belatedly and there is no evidence that he ordered his delegates to impede these activities. Indeed, it is noteworthy that not a single one of the 76 episodes of political violence that produced victims stemmed from a clash between the ‘red guards’ and the police. 71 It was not until late June and after receiving orders from the government that Bosque would ask the mayors to impede ‘the activities of any persons unable to show their credentials as an agent of the authorities’. 72
The Asturian Popular Front deputies had accused Friera of failing to purge the police command and of allowing the fascists to do as they pleased. Bosque kept these accusations very much in mind. Two episodes in particular illustrate relevant aspects of his public-order policy profile and his relationship with the police.
On 24 April a civil guard named Manuel Vela was killed by gunshots fired from a stationary vehicle on the Gijón coastal road, apparently when he was in the process of stopping and searching said vehicle. Internal civil-guard reports pinpointed the culprits as members of the Casa del Pueblo and concluded that the attack had been premeditated. Vela's death caused great disquiet among his colleagues, who blamed it on the governor's policy of tolerance towards extremists. During the funeral these criticisms became open: some commanders disobeyed the governor's orders regarding the route to be followed. Similarly to what had happened in Madrid at the burial of Lieutenant Anastasio de los Reyes a week previously – very serious clashes with firearms between extremists (leftists and Falangists) and assault guards which resulted in several deaths and serious injuries, the Falangists attempted to capitalize on Vela's death. When Bosque, along with the Socialist mayor, arrived at the funeral, he was greeted with hostility and surrounded by armed supporters of the fascist party, led by their local leader, Enrique Cangas. There were shouts of ‘Long live the fasces’ and the Falangists accused the governor of being a ‘murderer’ and a ‘criminal’. Escorted by army officers who objected to the Falangists’ aggression, Bosque was forced to leave the funeral. All of this took place in the face of passivity on the part of some civil-guard commanders in attendance. A number of Falangists were arrested and tried for illegal possession of arms, disrespect and disorder. 73 It is significant that only a few hours after Vela's funeral, something statistically unusual within the context of political violence in Asturias that spring took place: on 26 April, two civil guards opened fire on a vehicle which failed to stop when ordered to do so, leaving one of its occupants seriously injured. 74
The scandalous incident that took place during Vela's funeral laid bare the open fracture between Bosque and elements of the civil-guard command. At the same time, it revealed the worrying identification of some civil guards with the Falange. Episodes such as this convinced Bosque that police action, under the guise of maintaining public order, could serve as a brake on popular aspirations. Indeed, his notion of public order was mediated by a reading of the situation which he made no attempt to hide. In a ceremony held at the city hall shortly after arriving in Oviedo, when the socialist mayor expressed his hope that ‘the men who seek to disrupt the Republic [would] return to their lairs and remain there’, Bosque claimed that while he would ‘ensure compliance with the law’, he would ‘dispense with it if it [were to] constitute an obstacle to attending to the people's aspirations’. 75
A second moment which nicely illustrates this complex relationship between the governor and the police occurred around Labour Day. Given that the Casas del Pueblo had been shut down since 1934, this celebration held a special significance for the Asturian left. Its purpose was not only to mount a show of force, but also to condemn economic and social conditions in the region. It is significant that in this research's tally of political violence, only a few isolated acts of aggression possibly related to the May-Day mobilization were identified. There is no record of serious confrontations with the police, and this was no mere coincidence. Bosque took the decision not to order a special police operation. The demonstrations were authorized and Bosque entrusted their organizers to maintain order. Hours before the event, he expressed his ‘complete faith’ that there would be a ‘atmosphere of exemplary peace and coexistence’. No special measures were necessary, he added, because the representatives of the ‘working-class masses’ and the Socialists and Communist Youths had assured him that they themselves would immediately extinguish any attempt to disrupt public order. 76
A few days previously, on the occasion of the publication of a manifesto by the socialist miners’ union (SOMA), the governor had already expressed his belief that the workers’ leaders had a ‘clear vision of reality’ and ‘a spirit and sense of sacrifice for which I cannot find the words’. Therefore, he entrusted them not to lead the masses down ‘the path of illegality’ which fuelled the ‘disruptive campaigns of the Republic's enemies’. 77 Thus Bosque made clear prior to 1 May his confidence in the ability of the workers’ leaders to keep public order, while at the same time openly criticizing the conservatives. In a memo, he characterized the latter as ‘naturally fearful’ people who were full of ‘prejudices’ about public order and were therefore sowing the seeds of ‘alarm’. However, in his view it would not be the police, but rather the ‘sensible and admirable citizenry’ of the workers who would maintain order, as ‘they had always done’. 78
Bosque's distrust of certain elements within the police did not only extend to the civil guard, some of whose provincial commanders were transferred after Vela's funeral. 79 There were also difficulties with the Oviedo assault guards, as some of their officers showed signs of a lack of discipline. This was made sharply evident in the central streets of Oviedo during the weekend of 23–24 May. The government and the conservative press provided versions of these events that were directly counterposed. Nevertheless, the information compiled indicates that a group of out-of-uniform assault guards at an open-air dance clashed with various leftists, with the latter being arrested by a lieutenant and taken to the barracks. The following day, disobeying the governors’ orders not to go out on to the streets, at least two vans of assault guards did precisely that. They found themselves engaged in a shoot-out with leftists in the Plaza de Acevedo which resulted in 21 injured, some of whom extremely seriously. 80
This was the only serious confrontation between the police and left-wing workers to take place in Asturias that spring. Although there were no deaths, its importance was magnified greatly because it highlighted the gulf between the governor and the police command in the capital. The unions called a general strike in response and the city was shut down for more than a day. Once again, the illegal role of socialist and communist activists in controlling public order, supplanting the police, was revealed. ‘Circulation was controlled through safe-conducts, a “red guard” kept order and traffic access into the city was restricted’. 81
As a result of the confrontation, the government removed or transferred various commanders, while Bosque purposefully applied censorship to prevent the press from publishing any version of the events other than that issued by his office. He claimed to have given the ‘categorical order’ that the guards were to remain in their barracks after the first incident, and he openly accused a group of guards of acting arbitrarily and being responsible for the victims. With that in mind, he branded the events as a ‘true and immeasurable aggression’ against the people of Oviedo, who had known how to conduct themselves ‘in all of their popular demonstrations of joy with a truly exemplary rectitude’. This statement brought into sharp relief an understanding of gubernatorial and police management which discredited the right's notion of deteriorating public order and which considered the ‘people's’ violence to be a reaction to the prior provocations of the fascists. 82
This research demonstrates the relevance of case studies to the advancement of the analysis of political violence and public-order management in a complex context of democratic breakdown. 83 A database was developed which goes beyond what has become customary in the specialized historiography, including not only episodes that involved deaths, but also those that produced serious injuries. This significantly increases the quantitative and qualitative value of the sample. In this respect, it confirms the importance of identifying the political affiliation not only of the assailants, but also of those responsible for instigating the violence in a high percentage of cases. This enables us to better contextualize the political conflict that the governor at the time had to manage, as well as the role of the police.
The case of Asturias is of undeniable importance in the circumstances of political violence before the Spanish civil war. The findings presented here will enable a more thorough comparison with those of other provinces once these become available. The data on political violence in Asturias reveals at least three notable singularities. In the first place, once the initial phase had been overcome, labour conflicts gave way to violence that was tied to rivalries and vengeance of a purely partisan nature, with acts of aggression and attacks predominating (two out of every three cases). In second place, the tally of victims on the broad categories of left and the right was very similar, albeit slightly higher in the former. However, if the focus is on those responsible for initiating the incident, then the distribution alters, confirming that left-wing activists were more proactive in this respect, with a ratio of three to one compared to their opponents. This data ought not to be used to draw simplistic and partisan conclusions about who was responsible for the political violence, but rather to interrogate the context in which violence occurred and the complex relationship between intensive popular mobilization and democratic coexistence. In this respect the following characteristics can be highlighted: the exacerbation of ideological differences to the point of demonizing the adversary; local acts of vengeance linked to the complex aftermath of the failed revolution of October 1934 and the subsequent repression; and the desire to exclude the adversary who had been defeated at the polls. In addition, this research has identified another factor that drove the left's proactive role in instigating violence: numerous confrontations with fatal consequences were triggered by small patrols of armed leftists carrying out illegal searches. The third singularity is the fact that episodes of violence with victims which involved the police barely exceed 10 per cent of the total, compared to almost 30 per cent nationwide according to the data currently available. 84
As a result, this research confirms a high level of conflict in the case of Asturias, with a chronological development that is relatively similar to the rest of Spain during the decisive spring of 1936, albeit with one notable peculiarity: the relative lack of police involvement compared to the nationwide data. 85 It has been shown here that this peculiarity was to a considerable extent due to gubernatorial policy. Two arguments in particular should be highlighted. Firstly, there was no vacuum of power in the weeks immediately after the election, and the acting governor appeared able to channel the victors’ mobilization without the hasty deployment of the police. Secondly, the incumbent governor linked his public-order policy to the defence of the regime itself. Aware that the presence of the police on the streets was ill-received by the Popular Front's supporters, Bosque treated the problem of public order as inseparable from the defence of the Republic in the face of its enemies. Thus, his management of public order stemmed from the premise that the mobilized socialists and communists were not the source of conflict, but rather the solution to the threat of fascism. This research indicates that this notion of public order ruled out the preventative and recurrent use of the police to control potentially conflictive situations. Nonetheless, it is also valid to point out that this deliberate withdrawal, along with the appointments of numerous gubernatorial delegates, served to incentivize the paralegal conduct of extremist elements on the left. These may well have judged themselves to enjoy the moral backing of the authorities, gubernatorial delegates included, to engage in a struggle against fascism of their own accord. The national right-wing parties and their press converted the governor into a scapegoat and subjected him to relentless criticism, while the socialists and communists defended him as an example of antifascist fearlessness. His controversial policies reveal the dilemma faced by the provincial authorities in much of Spain during these critical months: while many socialists and communists deemed the mobilization of their supporters vital to the defence of the regime and the defeat of fascism, at the same time distrusting any role whatsoever for the police, the anti-republican right championed the claim that the government had relinquished its responsibilities, accusing the authorities of permitting anarchy to reign and of opening the door to the revolution.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article is part of the research project supported by the Spanish Research Agency (PID2020-113986GB-I00).
