**Carmelo Moreno**

Department of Political Science and Administration, University of Basque Country, Barrio de Sarriena s/n, 48940 Leioa, Bizkaia, Spain; carmelo.moreno@ehu.eus

Received: 7 November 2019; Accepted: 24 December 2019; Published: 31 December 2019

**Abstract:** To analyse the Spanish national question requires considering the relationship between the idea of the nation and the phenomenon of nationalism on one side, and the question of political plurality on the other. The approval of the Constitutional text 40 years ago was achieved thanks to a delicate semantic balancing act concerning the concept of nation, whose interpretation remains open. Academic studies of public opinion, such as the famous Linz-Moreno Question—also known as Moreno Question—that measures the possible mixture of Spanish subjective national identity, are equally the object of wide controversy. The extent to which political plurinationality is a suitable concept for defining the country is not clear because, amongs<sup>t</sup> other reasons, the political consequences that might derive from adopting the concept are unknown. This article sets out the thesis that Spain is a plurinational labyrinth since there is neither consensus nor are there discursive strategies that might help in forming an image of the country in national terms. The paradox of this labyrinth is that, since the approval of the Constitution in 1978, the political actors have accepted that nationality in Spain is insoluble without taking the plurinational idea into account. But, at the same time, it is not easy to assume such plurinationality in practical terms because the political cost to those actors that openly defend national plurality is very high. For this reason, political discourses in Spain on the national question o ffer a highly ambiguous scenario, where the actors seek windows of opportunity and are reluctant to take risks in order to solve this puzzle situation. The aim of this paper is to analyse which indicators are most e fficient for testing how the di fferent actors position themselves facing the phenomenon of the Spanish plurinational labyrinth. The clearest examples are what we refer to here as the concepts of (i) intersubjective national identity and (ii) plurinational governments.

**Keywords:** plurinationality; spain; nationalism; autonomy; intersubjective national identity

## **1. Introduction. Is a Labyrinth Neutral?**

Historically, the national question in Spain has been a major headache, an issue that has often been seen as posing a real existential problem (Laín Entralgo 2010). In general, the national problem has been explained as the result of a *defect*; that is, an imperfect process of unfinished nation-building. It is thus usually said that the "process of nation-building" in Spain was frustrated at some point in history due to the nonexistence of a truly uniform political project, the absence of a social class that could provide cohesion, and the lack of a strategy aimed at creating a market of common national a ffects (Ortega y Gasset 2006; Álvarez Junco 2001). Surprisingly, this idea can still be found in numerous works (Bernal 2005; Sotelo 2006; Béjar 2008; Colomer 2018) that, consciously or unconsciously, continue to accept as an almost inevitable phenomenon that modern states are only successful if they end up

constructing, in one way or another, a certain monist idea of a political nation. The nation-building processes in France, Germany, or United States are extolled as "successful" without consideration given to the enormous human costs and the deep ethical-political dilemmas involved in these processes of reducing identitarian diversity.

In reality, instead of being considered a *defect*, the national problem in Spain could be explained as resulting from an *excess*; that is, an imperfect process of national deconstruction that on many occasions has been viewed as an endless phenomenon (Romero 2012). As some historians point out, the idea of the Spanish nation did not emerge in order to configure a liberal state of free and equal citizens, but instead to provide symbolic cover in identitarian terms to a broad political empire that extended over several continents, a fact that makes a concept like "imperial nation" analytically appropriate (Fradera 2016). It is no accident that the first article of the first Spanish Constitution, approved in Cadiz in 1812, should have defined the Spanish nation as "the coming together of the Spaniards of both hemispheres". In this affirmation, one can observe how at the origins of Spanish constitutionalism, an attempt was made to homologise—in terms of collective identity as members of the same national community—the inhabitants of a series of territories that, under the dominion of the kingdom of Spain, shared the same language and certain religious values. The fact is that this idea of an imperial nation (Fradera 2016), in spite of the attempts to achieve communitarian homogeneity, from the outset established an important asymmetry in the citizenship rights of its inhabitants, according to different criteria of ethnic, group, and territorial ascription. Little by little, this idea of an imperial nation was affected by the independence processes of numerous territories in Latin America and Asia over the whole of the nineteenth century, making it necessary to revise the very concept of the Spanish nation, which was reduced to the borders of today's state following the independence of Cuba and the Philippines in 1898.

At the same time that this idea of the Spanish nation encountered difficulties in adjusting to a waning imperial political structure, Spain as a country also found it necessary to rethink itself on discovering that, within the incipient modern liberal-democratic state, political plurality on the national question was beginning to emerge. This situation even led over the course of the XIX century to the emergence of a debate on the need to have a political party clearly labelled as a "national party" (Fernández Torres 2018, pp. 185–95). In short, if the imperial nation had proved excessive when it came to materialising itself within the boundaries of a dwindling imperial power structure, the Spanish nation now also started to be seen as excessive when it came to materialising itself, even within the boundaries of the peninsular state alone. The appearance of the text by Pi i Margall titled *Las nacionalidades* (*The nationalities*) in 1876 was perhaps the first attempt in the modern Spanish state to think the Spanish national question in plural terms. Since then, the idea of the Spanish nation has coexisted with this concept of nationality, which refers to the same semantic field and poses the challenge of the point to which the Spanish political community can or cannot be viewed in terms of plurinationality.

The transition from a singular imperial nation to an idea of a nation that coexists with other national realities, of all of them having a political character, within the same state is not a simple undertaking. This transition is especially difficult in convulsive political contexts, such as occurred in Spain at the start of the twentieth century, or in periods that were directly undemocratic, during the Francoist dictatorship, for example. In the last four decades, however, the Spanish democratic system has been accumulating political capital that has enabled it to provide some answers to this challenge concerning its supposedly plurinational character. Two types of strategies have been developed to address this challenge. On one side, setting underway the debate on the difference between the *civic* character, as opposed to the *ethnic* character, of the different nationalist projects (De Blas Guerrero 1984; Jáuregu<sup>i</sup> 1999; Grad and Rojo 2003; Ruiz Jiménez 2005; Muro and Quiroga 2005). On the other, putting into circulation a strange phenomenon that we will term acceptance of plurinationality without naming it (Nuñez Seixas 2010). This double strategy has been possible thanks to four pairs of especially hazy concepts that organise Spanish political language: (i) The distinction between nations that are

supposedly based on nationalist citizens and nations that appear to exist without nationalists; (ii) the distinction between plurinational territories and non-plurinational territories; (iii) the distinction between the concepts of identity and national identification; and finally, (iv) the distinction between the concept of nested national identity and the concept of overlapping national identity.

It is worth asking about the extent to which these four pairs of concepts serve in explaining the Spanish plurinational labyrinth, in which paradoxically all the political actors employ the same strategy: Criticising the supposed "nationalist biases" in the positions of their political rivals while at the same time savouring the "nationalist bias" in their own position. This strategy is based on two arguments: On one side, criticising the existence of a ethnical nationalism in the eye of the other, without explicitly recognising the existence of a similar ethnic nationalism in one's own eye; on the other side, demanding recognition for plurinationality in the eye of the other, without explicitly recognising the existence of plurinationalism in one's own eye. This strange narrative of accepting plurinationality without naming it has had three considerable effects: At the juridical level, a high degree of legal-constitutional ambiguity when it comes to defining the national question; at the academic level, an enormous difficulty in measuring the national identity of the citizens; at the political level, an apparent absence of political and electoral incentives for reaching consensus on some idea of national pluralism in Spain. The politically relevant question lies in knowing whether, after 40 years of democratic coexistence, this dynamic of monist nationalisms that are prisoners to the dilemma of accepting plurinationality without naming it is the best framework of relationship for the actors of the Spanish political system. The hypothesis of this paper is that some political actors have benefitted more than others from this labyrinth, with the result that the incentives for maintaining it are strongly asymmetrical. In sum, it is worth enquiring into the possible neutrality of this labyrinth on the national question. These questions are especially pertinent, as we will see later, to try to understand the political dynamic that is currently being lived in the Spanish region of Catalonia, popularly known as "el procés" (Cuadras-Morató 2016; Kraus and Vergés 2017): A sequence of political events conducing to a situation of strong challenge to the Spanish political order. This process, to a large extent, might be interpreted as a result of a problem to accommodate the national plurality within the Spanish institutional system.

The text is organised into four parts. The first part is a theoretical consideration of the concept of plurinationality, based on the model for the analysis of political concepts proposed by Reinhart Koselleck. In this part, we will try to detect three types of situations: (a) The difference between the concept of plurinationality and the terms employed in practice to speak about this concept; (b) the importance of the *spaces of experience* and *horizons of expectation* contained in a concept like plurinationality in Spanish political life; and finally, (c) the question of the extralinguistic character of the concept of plurinationality, which enables its politicisation, ideologisation, temporalisation, and democratisation. In the second part of the text, we will briefly review the form in which the concept of plurinationality has been included in the Spanish juridical-constitutional framework, especially the inflated wording on the national phenomenon, which has enabled ambiguous and even conflicting interpretations to be made, while at the same time facilitating the prevalence of some interpretations and hindering the emergence of others. In the third part, we will address the epistemological dilemmas of the famous *Moreno Question* that tries to analytically measure the phenomenon of identity and plurinational identification in Spain, the debates on this question, and the academic shortcomings found in this type of measurements. Finally, in the fourth part, we will deal with the question of the Spanish party system and the logic of the electoral competition based on the nationalist issue. In this section, we will see that national plurality in Spain has been treated as a question of recognition and representation (Alonso 2011, pp. 186–87), but not so much as a problem of governability and, quite specifically, as a question of political alternation (Caron and Laforest 2009). This alternation holds both at the state-wide level and in the regional level, where there are different options for crystallising the national question in their respective governments.

## **2. Plurinationality: A Slippery Concept**

Plurinationality is a term that has had a very brief career in both historical and academic terms in Spain. It is very much used as an adjective to speak, for example, of the di fficulty involved in understanding Spain as a "plurinational state", when in any case, it is not specifically recognised as such (Fossas and Requejo 1999; Rodríguez Bereijo 2018, p. 840); also, for speaking of the virtues or di fficulties in articulating that nebulous and little defined syntagm "plurinational federalism" (Caminal 2002; Requejo 2003, 2004; Caminal and Requejo 2009). At times, the adjective plurinational has been changed for other apparently proximal terms like "multinational state" (Norman 2006; Jewkes 2015) or "multinational democracy" (Gagnon et al. 2003; Zuber 2011). Nonetheless, a certain vertigo still subsists when it comes to giving substance to the phenomenon. In a certain way, the evolution of the term plurinationality has been contrary to what happened with other terms like Liberalism, for example. Liberalism is viewed as a substantive term that has evolved in such a way that today we only know it through its di fferent "liberal" variants, where the di fferent meanings of the adjective bear a certain family resemblance (Rawls 1996), while in the case of plurinationality, the term has become an adjective without having been properly constituted as a noun; that is, as an object that defines a historical phenomenon. However, and this is a thesis that we shall be setting out here, it is possible to a ffirm that the phenomenon of plurinationality exists as a concept in spite of the generalised absence of this term as something substantive and singular. The historian Reinhart Koselleck made an important distinction between concepts and words. In his opinion, a linguistic meaning can adhere to a word, but a concept makes it possible to define an historical phenomenon and, at the same time, offers di fferent political-social meanings that are instruments and, simultaneously, vectors for political action and conflict between di fferent options (Koselleck 2012, pp. 27–43). In the case of the term plurinationality that concerns us here, it is legitimate to think that this concept, in a country like Spain, has undergone a process of unstable, precarious, and troubled crystallisation due to the existence of di fferent discursive strategies aimed at hindering its generalised social and political use as a noun. It is worth tracing how this strange process has occurred and, above all, identifying the reasons why such a situation subsists.

It is possible find as many as four di fferent and complementary strategies in this process of indirect conceptual elaboration of plurinationality in the Spanish case. The first strategy was based on the creation of the concept of "nation of nations" (Leguina 1980; Maldonado 1995; Solozábal 2018; Cruz Villalón 2018). This term nests national plurinationality as a type of hierarchical structure of layers, in which each big nation exists insofar as it includes other smaller nations within itself. From the juridical point of view, it is not clear how a single term can signify a whole and a part at the same time, and perhaps for this reason, the di fficulties inherent in this strategy mean that it has been employed more in the metaphorical field of political action than in the strictly juridical-constitutional field, where the term is not well received. As the Ruling of the Constitutional Court 21/2010, June 28, notes, "the term nation is extraordinarily protean due to the many contexts in which it is usually involved ( ... ) E ffectively, one can speak of the nation as a cultural, historical, linguistic, sociological, and even religious reality. However, the nation that is of concern here is exclusively the nation in the juridical-constitutional sense. In this specific sense, the Constitution does not know of any other than the Spanish nation ( ... )". The term *nation of nations* is an updating of the old XIX century ideal, inspired by a famous Spanish writer, Miguel de Unamuno, who tries to reconcile in a single ontological unit an idea of a country that has one history and multiple intrahistories, but this attempt to hierarchise a single concept on two semantic levels has proved highly unsatisfactory.

The second strategy consisted of modelling two distinct theoretical types of nation that can coexist in the same state: The idea of a civic nation facing the idea of an ethnic nation. Based on the classification used to explain the nation-building processes in France (civic nation) and Germany (ethnic nation) in the XIX century, an abundant academic literature has justified this classification to distinguish between political processes that build their idea of the nation on civic elements, such as individual rights or the idea of equality before the law, as against those that build the nation on ethnic

and cultural elements, such as the existence of a language, a religion, or certain values (Ignatie ff 1993; Ipperciel 2007). This distinction between civic and ethnic dimensions has, moreover, been accompanied by an evaluative asymmetry according to which the concept of civic nation enjoys more democratic recognition than the idea of ethnic nation. This can reach the point where the struggle between di fferent nations that coexist in a political space, as happens in Spain, is usually and to a grea<sup>t</sup> extent an attempt to criminalise the ethnic conception of the opposing nation and, at the same time, an attempt to emphasise the civic vision of one's own nation. Thus, in Spain, the idea of the Spanish nation is usually identified in general with a civic nation, except in the case of the peripheral nationalists that consider it to be an ethnic nation. When it comes to judging the civic or ethnic character of the Basque and Catalan nations, the mainstream opinion in Spain has varied in a way that is certainly significant: Depending on the better or worse relation of the Basque and Catalan nationalists with the Spanish political system, between 1978 and 2019, the Basque nation has gone from being an ethnic nation to being a civic one, while the Catalan nation has gone from being seen as a civic nation to being treated as an ethnic one (Miley 2007; Serrano 1998, 2001; Rubio Caballero 2015). As Ramón Maiz observes, empirical experience shows that the distinction between an idea of a civic nation facing an ethnic nation is an untenable dichotomy. The clearest example is provided by France due to the di fficulty in reconciling its self-conception as a civic nation with the Muslim population that live in that country (Maiz 2018, pp. 123–35). Just as an ethnic nation without a civic dimension is not democratically acceptable, a civic nation without an ethnic or cultural dimension is not politically operative (Maiz 2018, pp. 77–142). All nations have civic and ethnic elements and, consequently, although the relationship between these elements is variable, the strategy aimed at distinguishing the two ideal types of nation is of no help in understanding the concept of plurinationality.

The third strategy for tackling the phenomenon of plurinationality without mentioning it as such consisted of saying that there are nations founded on nationalist ideologies and nations whose foundations do not include a nationalist discourse. In this case, the strategy was based on a constant process of demonizing nationalism as an ideological phenomenon, which is supposedly used by some nations but not by all (De Schutter 2007). In few countries can one find as much literature as in Spain that, over the course of time, has criticised political nationalism so reiteratively and so severely, especially when referring to the so-called regional nationalisms. However, the consequences of this process in which many people have only been able to build their national identity through a process of "Spanish nationalist denial" are, on one side, relegating their own national identities to marginal areas (Fusi 2006) or, on the other, adopting attitudes of desertion ( Béjar 2008). If we lose sight of the presence of this kind of nationalist self-denial, it will be di fficult to understand the nature of the Spanish nationalism and its presence, especially in regions like Basque Country and Catalonia where it competes with other nationalisms.

Finally, the fourth strategy for addressing the phenomenon of plurinationality consisted of an exercise of dividing the noun of the national question by using two terms, nation and nationality, to talk of two experiences with sociological similarities but that in principle have di fferent juridical levels of political articulation. As we will see in the following section, this was the practical option adopted during the juridical-political debate that took place when drafting Article 2 of the Spanish Constitution in 1978 (Solozábal 1980; Corcuera 1993). In Koselleck's terms, the wording of Article 2 of the Spanish Constitution in 1978 can be analysed as a point of encounter between a space of past experience and a horizon of future expectations (Koselleck 1993, pp. 333–57). On the one hand, the space of experience consists in the mutual recognition by all the political actors of the existence in the past of situations of misunderstanding involving the di fferent national realities that must be superseded. On the other hand, the horizon of expectations consists of leaving the door open to a possible coexistence amongs<sup>t</sup> the national political experiences by means of ambiguous and unspecific political formulas, constitutional "silences" (Rodríguez Bereijo 2018, pp. 837–39), "apocryphal compromises" as they were called by Carl Schmitt, that serve for postponing a definitive decision because, in fact, it is thought that only future historical time can give shape to that decision.

This strange process of crystallisation of the concept of plurinationality in Spain, which avoids directly relating the ideas of nation and political plurality by means of this network of elusive terms, complicated strategies, and semantic indeterminacy, has hindered the development of the four dimensions inherent in every political concept described by Koselleck, namely its politicisation, its ideologisation, its temporalisation, and its democratisation (Villacañas and Oncina 1997, p. 23). As we will now proceed to examine, the politicisation of the concept of plurinationality implies the possibility of including an increasingly large number of political situations in Spain that could come under this umbrella term; the ideologisation of the concept implies that plurinationality is not a natural phenomenon that must be accepted without public deliberation, but on the contrary, it can and must be oriented towards disparate ideological positions; the temporalisation of the concept implies that plurinationality is not a static situation, a fixed photo to which a representative channel must be given, but is instead a dynamic experience involving political changes in the exercise of power over time and that it must be analysed through the existence of real political alternatives; the democratisation of the concept implies that plurinationality is not an attribute that can be demanded by one concrete actor alone (in this case, the state or the institutions involved in the territorial organisation of power), but instead it is a quality that can be demanded at all political levels and in all spheres of political action. In practice, it has not been possible to develop any of these four dimensions in Spanish political life.

## **3. Nation, Nationality, and the Spanish Constitution**

It is well-known that the current Spanish Carta Magna, adopted in 1978, makes no specific mention of the idea of plurinationality. However, the fact that it contains a plurality of terms associated with the national question implies that, in practice, an attempt was made to address this phenomenon in some way. If the Spanish constitutional text, as happens in a large part of the fundamental rules in the European sphere, had decided to dispense with any mention of the national question in substantive terms, the question of plurinationality as a political phenomenon would perhaps have ceased to be an option. The problem is that that option became inviable because during the democratic Transition the political actors understood that no Constitution would be legitimate without providing at least a minimal and explicit answer to the question of national plurality. It is well-known that during the debates addressing this issue, the idea of considering the plurinational character of Spain was in circulation, especially amongs<sup>t</sup> the leaders of the Socialist Party (henceforth, PSOE) although it is true that this term was never seriously proposed for inclusion in the constitutional text (San Antón 2018, p. 200). Finally, during the constituent process, an overblown formula concerning the national question was chosen: The use of two terms, *nation* and *nationality* (both contained in Article 2 of the Constitution), as well as a mention of the *Spanish nation* as the repository of sovereignty (Preface) and an allusion to the idea that *national* sovereignty resides in the Spanish people, from whom the powers of the state emanate (Article 1). This conceptual architecture concerning the national question was intended to simultaneously guarantee both the unitary character of the Spanish nation (in Article 2 three di fferent expressions are used to reiterate the same idea that the Spanish nation is "indissoluble, common and indivisible") and the right to an autonomous political power of the so-called nationalities, which are generically named in a di fferent way from the rest of the regions that make up Spain (De Blas Guerrero 2018).

It is important to underline that the chosen option involved a definition of national plurality that was closer to the model that Miller (2001) called "nested nationalities" than to the model we could call "overlapping nationalities". According to the nested nationalities model, di fferent national forms can coexist if a relationship of parallelism is established amongs<sup>t</sup> them, in which each one in fact acts at a di fferent political level. That is, these forms of national expression relate to each other in a model of coexistence with a minimum level of substantive contact, to thus avoid a greater degree of conflict. Hence, in the Spanish case, while there is recognition that the Spanish nation exists politically, so to guarantee the unity of the country in a civic way and without nationalist apriorisms, the existence of the di fferent nationalities, which are not mentioned expressly in the Constitution and whose territorial

delimitations are not established, is accepted in order to develop the maximum possible level of political autonomy in certain territories of the country, with the sole restriction that they should not jeopardise Spanish unity. Similarly, it is spelt out that the degree of political autonomy of the di fferent nationalities and regions is to be an open and asymmetrical process, based on the so-called principe dispositif, since the level of competences is directly chosen by each of the territories. In sum, it is accepted that coexistence between the two levels of the national reality is possible because, in reality, both levels operate on parallel political levels. Therefore, there is no overlap or intersection between the di fferent national realities. Any speculation about what the degree of substantive relationship between the di fferent national realities might be is discarded. Instead, the option chosen is that this relationship between di fferent national realities should be as small as possible, based on a spatial division of slightly interconnected political systems. As we will see, this model of nested nationalisms, which is more defensive than interactive, more formal than substantive, complicates any measurement and interpretation of overlapping national identities, which is what the Moreno Question hypothetically attempts to do.

On balance, this plurinational model of nested and not overlapping national identities has generated three types of problems during the 40 years in which the Constitution has been in e ffect. The first is related to the comparative and permanent grievance amongs<sup>t</sup> the territories of Spain based on the evidence that the introduction of an asymmetrical autonomous model was open to exploitation by the so-called "di fferential national realities", which has led to an inflationary argumen<sup>t</sup> about the national question in many territories. While during the Spanish democratic Transition, the term nationality was understood in a more restrictive way to refer to territories like Catalonia, the Basque Country, or Galicia, over time this term was adopted in the di fferent Autonomy Statutes of numerous territories, starting with Andalusia (1981) and followed by the Canary Islands (1996), Aragon (2007), Balearic Islands (2007), and the Valencian Community (2019). The project to reform the Autonomy Statute (Estatut) of Catalonia, which included the option of considering that territory as a nation and that was annulled by the sentence of the Constitutional Court STC 31/2010, was inscribed in that same inflationary logic (Rodríguez Bereijo 2018, p. 841). On the other hand, attempts to harmonise the autonomous system as occurred with the approval of the Organic Law on Harmonisation of the Autonomous Process (LOAPA), which was declared to be partially unconstitutional by the STC 76/1983 following an appeal lodged by Basque and Catalan nationalists, demonstrate that with the present constitutional design of nested plurinationality it will be di fficult to reform the autonomous system, which was designed in an asymmetrical way that di ffers greatly from the aims of a federal model (Lluch 2011).

The second problem with this Spanish model of nested and not overlapping national identities is related to the progressive reduction of the perception of national plurality in those territories conceived as *nationalities*, especially the Basque Country and Catalonia. The rejection of the idea that these national realities overlap at the Spanish level has resulted in the non-acceptance of plurinational realities at the regional level. It is no accident that those regions where there are politically strong nationalist parties are the ones where political alternation between parties with di fferent national sensibilities has been less feasible and, above all, more controversial. The examples of the Basque Country in 1986, Catalonia in 2003, and Navarre in 2019 are the best examples of this situation.

In 1986, the exceptional situation in the Basque Country arose in which the two leading parties, the PNV and the PSE-PSOE, formed the first coalition governmen<sup>t</sup> in Spain, giving shape to the first governmen<sup>t</sup> that included di fferent national sensibilities in the entire country (Llera 1988); a governmen<sup>t</sup> that laid the institutional, economic, and political foundations of the Basque Country for the following decades, depriving ETA's terrorist violence of any hint of legitimacy. It was no accident that ETA, by means of a massive terrorist onslought, made 1987 the bloodiest year of both the Transition and its own history: It was without doubt the main actor that wanted to contest the supposedly plurinational character of the new Basque governmen<sup>t</sup> and its policies. This was especially true of the educational system, which was introduced by politicians chosen by the Socialist Party of the Basque Country (PSE-PSOE), José Ramón Recalde and Fernando Buesa, both of whom were the target of attacks by this terrorist gang; a coalition government, in short, designed precisely to achieve coexistence between di fferent national sensibilities. Decades later, the reality is that there has been no political alternation in the governmen<sup>t</sup> of the Basque Country, if we ignore the episode of alternation led by the socialist Patxi <sup>L</sup>ópez in 2009, made possible by the non-participation of the radical left-wing regional nationalists in that year's election to the regional parliament.

In the case of Catalonia, the election of 2003 in which the Catalan socialist Pasqual Maragall (PSC) was elected president of the regional governmen<sup>t</sup> (*Generalitat*) was possibly the first opportunity to make a real alternation between the di fferent national sensibilities in this autonomous community. In light of reactions to the speeches and political strategy of Pasqual Maragall, with his aspiration to initiate a reform of the Autonomy Statute (*Estatut*) and even bring about more far reaching change in the way the political relation is established amongs<sup>t</sup> the territories of Spain (Maragall 2005, 2008), it is doubtful that his initiative was applauded as an opportunity for alternation in terms of plurinationality. Quite the opposite, in general Maragall, initiatives were interpreted critically as the victory of the nationalist discourse within the political imaginary of Catalan socialism (De Blas Guerrero 2003; Elorza 2003; Vallés 2005; Peregrín Gutiérrez 2004; Blanco Valdés 2005, pp. 180–81; Madrid Delgado 2005; Espelosín Atienza 2005; Quiroga Fernández de Soto 2008). Perhaps the *Maragall symptom* can be seen as the political experience that has taken questioning of the constitutional design of nested plurinationality to the limit, forcing the grea<sup>t</sup> majority of political actors, both Spanish and Catalan, to take up the challenge of plurinationality. In this sense, Maragall's political disavowal following the approval of the reform of the Statute agreed between the Spanish prime minister Zapatero and the leader of CIU, Artur Mas, in January 2006, behind the back of the Catalan plurinational governmen<sup>t</sup> of the PSC and the left-wing regional nationalists of ERC, can be seen as an alliance of those critical of the idea of *overlapping plurinationality.* That is, as an alliance of those who consider that in the final instance the optimum relation between the "nations" should be one in which the parties that "represent" them do not interfere in their respective spaces of power; i.e., *it is inconceivable for there to be a plurinational government in Spain*. At that time, the idea of overlapping plurinationality underwent one of its most disappointing episodes.

In this sense, the tensions and controversy in the Spanish political debate during 2019 over the formation of a plurinational governmen<sup>t</sup> in the region of Navarre between the PSOE and some nationalist parties from this territory have followed the same logic: The idea that institutionalised plurinationality in governments can encounter resistance and that opposing it politically can easily bring electoral benefits. As we stated above, our starting hypothesis is that these plurinational governmen<sup>t</sup> formulas receive asymmetrical answers from the di fferent political actors, but one can wonder whether or not opposition to this governmen<sup>t</sup> formula has a promising future.

In relation to the above, there is a third problem in this Spanish model of nested and not overlapping nationalisms: The di fficult fit involved in accepting possible political interdependencies and, more specifically, the possible support that regional nationalist parties can provide for Spanish governability. In 1993, the PSOE lost its absolute majority for the first time since the restoration of democracy and was forced to seek stable parliamentary support in order to govern. Since then, centre-right nationalist parties like CIU and PNV have been the main options chosen to assist this new multiparty governability, but subsequently, the most recurrent argumen<sup>t</sup> in the Spanish public opinion since then has consisted of criticising the excessive influence that these regional nationalist parties have in the general power system in Spain. The underlying idea is that *national* policy undergoes deterioration due to the presence of actors that do not include the search for issues conceived as having a Spanish "national" character amongs<sup>t</sup> their priorities. It is significant that in 40 years, the option of forming a coalition governmen<sup>t</sup> in Spain from a plurinational perspective has not been seen as desirable, or even possible, although this has happened, albeit with many problems, in the Basque Country, Catalonia, and now in Navarre. Perhaps the negative interpretation of the plurinational experiences in regional governments in Spain, highly visible in the case of Catalonia above all, have for the time being closed a horizon of expectations favourable to this possibility. While a plural governmen<sup>t</sup> formula between parties that represent the Spanish nation is not possible in the future, the most usual way of evaluating the policy of alliances of the parties forming the central governmen<sup>t</sup> has been by measuring their relation of dependence, and therefore possible weakness, facing the regional nationalist parties. It is important to highlight that this argumen<sup>t</sup> had always been employed from a single ideological direction, namely from the right-wing party Partido Popular (PP) to criticise the left-wing party (PSOE), and this argumen<sup>t</sup> is paradoxically now being used against the right-wing governmen<sup>t</sup> (PP) by other right-wing parties that have not so far had executive responsibilities, like Ciudadanos and the extreme right-wing party Vox. This situation has taken the argumen<sup>t</sup> to the point where, at present, the only available alternatives to the "constitutionalist parties" (with this expression taking for granted that there is only one basic way of interpreting the Constitution amongs<sup>t</sup> "Spanish national" parties) are three in number: (i) Waiting for the formation of a governmen<sup>t</sup> only by these parties, (ii) fighting against the appearance that governability does not involve making concessions to the regional nationalists parties, or (iii) refusing to make any concessions at the cost of blocking all governmen<sup>t</sup> activity. This trilemma is the result of a model that holds that Spanish national realities must resist a process of political intersection. The question, as we will see, is knowing what incentives, benefits, and costs the di fferent actors are capable of accepting to maintain or, in its case, change this framework of (badly) nested plurinationality.

#### **4. The Moreno Question: To Be** *and* **Not to Be**

Together with political-constitutional analysis, a productive course for exploring the scope of plurinationality in Spain has been to study opinions on the phenomenon of subjective national identity through questions in surveys. For many decades, in a pioneering way that was later exported to other countries and international contexts (Cussó et al. 2018), a question known as the *Moreno Question* (Moreno 1988, 1995, 2006) was used in Spain. This tries to analyse the degree of ethno-territorial identification of the country's inhabitants in relation to their identification with Spain. It consists of five options that range from 1 to 5 according to whether people consider themselves *only (Spanish identity); more (Spanish identity) than (regional identity); as (Spanish identity) as (regional identity); more (regional identity) than (Spanish Identity); or only (regional identity).* Over time, numerous critical commentaries have pointed out the problems generated by this question: problems deriving from the mere wording of the question to adapt it to di fferent contexts (Ruiz Jiménez 2007); problems about its static character when it comes to capturing the dynamic construction of identities according to di fferent political contexts (Lecours 2004); problems in measuring the intensity and supposed linearity of di fferent identities when these are compared within a single question (Guinjoan and Rodón 2015); and problems on how to measure so-called dual identity in countries like Spain, which contains regions that di ffer greatly when this question is addressed (Martínez-Herrera 2002; De Nieves and Diz 2019; Molina and Quiroga 2019). Apart from these substantial limitations, perhaps the prior question that must be posed is knowing whether this question really serves for measuring the national question and, once this doubt is resolved, seeing whether the alternatives that have been proposed for solving the *Moreno Question*'s problems contribute to measuring Spanish plurinationality in a more reliable way or, on the contrary, hinder such knowledge.

The most usual interpretations of the *Moreno Question*—not only those that extol its validity but, strangely enough, also those that criticise it—assume that it is possible to establish a certain analogy between the ethno-territorial dimension and the national dimension of political identity. It is well-known that the two dimensions, the ethnic and the national, refer to political phenomena that are close in theory while they can be very di fferent in practice. However, the reason for making this analogy is that the national dimension, as mentioned above, is highly conceptually ambiguous and full of vague theoretical constructions, as occurs in the confused argumen<sup>t</sup> over the concepts of civic and ethnic national identity, that not only contaminate the political debate, but academic analysis as well. It is no accident that the *Moreno Question*, which supposedly measures subjective national

identity, has never utilised the term "nation", "national" or "nationalism" in its wording or in the answers o ffered to the interviewee. Given that the national question can arouse certain suspicions when it comes to recognising it, suspicions that are especially visible in the case of the *Spanish* national question, the ethno-territorial question has been accepted on numerous occasions as an epistemological shortcut for trying to approach an understanding of the national question. As Ramón Maiz points out, every national phenomenon always refers to some ethno-cultural component (Maiz 2018), but the paradoxical thing is that not everyone accepts this proximity between the national and ethno-cultural dimensions in the same way and with the same intensity. In fact, in Spain, there are many people who implicitly accept that the so-called Spanish national question can be understood in civic terms without any need to link it with any ethnic component. This asymmetrically distorts their predispositions on the national question, especially when they perceive the existence of national identities that are in dispute within the same territory. Without any doubt, this situation poses a serious problem when the ethno-territorial dimension is used as a means for measuring national identification approximately, amongs<sup>t</sup> other reasons because ethno-territorial sentiment holds no negative bias (identifying with the territory as such is generally socially accepted). However, that negative bias can exist regarding the nationalist question because identifying with nationalist content might be problematic when territory and national identity do not fully coincide.

The basic reason that the ethno-territorial and the national dimensions are usually seen as analytically distinct is because the ethno-territorial dimension is usually perceived in a positive way inasmuch as it is accepted as a non-political dimension that, in theory, does not necessarily translate into an expression of national or nationalist character. However, the argumen<sup>t</sup> that we will be defending here is a very di fferent one: We believe that the entire ethnic dimension is usually translated politically into some type of national expression, although this is not always done in the same way. Or, stated di fferently, both the Spanish ethno-cultural dimension and the ethno-cultural dimension of each Spanish region are *always* translated into national expressions, but they do not manifest themselves in the same form and they are not perceived as being national in the same way. In the Spanish case, there are regions where the ethno-territorial identifications of some of the inhabitants end up creating a national identity that crystallises in regional nationalist parties of di ffering political strength (as occurs, in the Spanish case, in territories like Aragon, the Balearic Islands, the Canary Islands, Catalonia, Galicia, Navarre, or the Basque Country); just as ethno-territorial identification is generally low in other regions (as occurs, for example, in Madrid, Valencia, or the two Castiles); or the ethno-territorial identification can be strong or very strong in other regions, but does not crystallise into durable regional nationalist parties (as occurs, for example, in Andalusia, Asturias, Cantabria, or La Rioja). Assuming that in the former cases we are dealing with territories with a single national identity of a regional type, and in the latter cases we are facing regions that lack national identity, or that are "weakly nationalised" (De Nieves and Diz 2019), is tantamount to a ffirming the absence of a Spanish ethno-cultural national identity in all these territories. This a ffirmation is the logical consequence of accepting that Spanish national identity is chiefly civic and that only in extreme cases, such as that of a violent Spanish nationalism, does it have an ethnic character. It is perhaps worth imagining a very di fferent a ffirmation: It is possible that in territories like Catalonia, the Basque Country, Navarre, or Galicia, what we are facing is plurinational societies in which, nonetheless, a defective method has been employed for conceptualising and measuring them as such; while in territories like Andalusia or Asturias, we are dealing with communities that, far from being weakly nationalised, are simply less plurinational; that is, these are communities where a dual ethno-cultural identification exists but, nevertheless, this duality is convergen<sup>t</sup> in the national dimension.

In this sense, the use of the so-called *Moreno Question* for measuring the subjective national identity of citizens in plurinational political settings generates many explanatory problems, above all in those answers situated in intermediate mixed categories, where dual identification with Spain and the specific region of each interviewee is possible. In those regions where regional national identity weighs very heavily, as in Catalonia or the Basque Country, the ethno-cultural questions

associated with Spanish national identity are usually very hazy, either for reasons of stigmatisation (Dinas 2012) or due to cognitive weakness when it comes to indicating how those mixed sentiments can be articulated. As Marc Guinjoan and Toni Rodón observe, many citizens of communities with regional national identities who situate themselves in intermediate positions of the *Moreno Question* do so for very di fferent reasons, without it being possible to establish a specific profile in terms of dual identity (Guinjoan and Rodón 2015, p. 137). On occasion, the Spanish ethno-cultural sentiments of these interviewees are *nested* in the sentiment of regional ethno-territorial belonging, in such a way that one sentiment can be articulated in national terms and the other sentiment can be articulated in more administrative terms; on other occasions, on the contrary, the two ethno-cultural sentiments, both Spanish and regional, *overlap* in national-identitarian terms. The problem is that, to date, the most usual alternative that has been suggested for analysing the di fferences amongs<sup>t</sup> interviewees holding these intermediate positions, between those that give nationally nested answers and those that give nationally overlapping answers, is to compare the position of these interviewees on the two nationalist scales separately (Guinjoan and Rodón 2015, p. 135; De Nieves and Diz 2019, p. 32). That is, measuring state and regional nationalist sentiment using two independent metric variables on the assumption that these two variables are comparable when in fact they measure very di fferent things. The regional nationalist scale usually does fuse the ethnic, territorial, civic, and national dimensions in a fairly balanced way, so that the answers of interviewees can be analysed lineally in identitarian terms. However, on the Spanish nationalist scale, the relationship between the ethnic, civic, and national dimensions has a very high degree of heterogeneity, amongs<sup>t</sup> other reasons because many interviewees simply define themselves as "not nationalist" and this attribution prevents the expression of strong positive sentiments on the Spanish national question. The confusing relation between national question and nationalism when one thinks of Spain, as opposed to the relation between national question and nationalism in regional nationalisms that are less problematic, makes comparison between the two scales very di fficult to observe because the linearity of each is sociologically di fferent. Instead of helping to understand the results of the *Moreno Question*, it is possible that they contribute to increasing the complexity of the plurinational labyrinth.

In short, the *Moreno Question* attempted to tackle the issue of "to be *and* not to be" of the Spanish plurinationality in a fairly elaborate way. On one side, it accepts the existence of plurality in terms of ethno-territorial identification but, on the other, it refuses to articulate that plurality in strictly national terms. In this sense, many of the attempts to improve this shortcoming, such as using metric variables that measure the terms "national" or "nationalist" separately to explain subjective national identity at the Spanish level and the regional level, as if they were parallel worlds, are exercises that involve taking a step backwards instead of forwards, insofar as they refuse to directly ask, in a combined manner, about the question of plurinationality. With respect to this resistance, it is very significant to observe how the Spanish people, in di fferent surveys conducted over many years, have been asked to respond to a question about what Spain means to them, by choosing from amongs<sup>t</sup> the following five options: *1* = *It's my country; 2* = *It's the state of which I am a citizen; 3* = *It's a nation of which I feel a member; 4* = *Something special, resulting from history, which unites those who live here, and that I wouldn't know how to define; 5* = *It's a state formed of several nationalities* (Martínez-Herrera and Miley 2016, p. 200). A second alternative to the same question was asked in some regions, whose inhabitants could choose from amongs<sup>t</sup> the following five options to refer to Spain: *1* = *It's my country; 2* = *It's a nation of which I feel a member; 3* = *It's the state of which I am a citizen 4* = *It's a state formed of several nationalities and regions; 5* = *It's an alien state of which my country does not form part* (Fernández Manjón 2010, p. 124). It is possible that these two questions have come closest to asking the citizens their direct opinion on plurinationality in Spain. However, the problem with these two questions is that they assume that plurinationality (i) solely refers to the sum of the regional nationalities, separated from each other in di fferent territories, but without including Spanish nationality in a clear form; and (ii) refers solely to the Spanish territory as a whole, but not to each of its regions, where it seems it does not even consider that there might be some type of plurinationality. It is not surprising (but, at the same time, it is revealing about the

predominant idea of plurinationality in Spain) that a question like this has never been posed to refer to the autonomous level, not even in those territories like the Basque Country, Navarre, Catalonia, and Galicia, where the plurinational character of their population is quite clear.

## **5. National Intersubjectivity and Plurinational Governments**

Analysis of the plurinational question from the juridical-constitutional perspective or from that of opinion surveys runs the risk of placing excessive emphasis on the "created" (Lecours 2004, p. 82), or better put, "static" elements (Ruiz Jiménez 2007, p. 163) of the di fferent national realities. That is, overvaluing the nested character, at the expense of the overlapping character, of the national identities. These approaches avoid analysing controversial situations deriving from the interrelation between national identities in plural societies, as occurs to a large extent in Spain, at the cost of completely ignoring the essential aspect that defines the plurinational question as such (Stojanovic 2011). Facing this, another perspective is possible that observes the ever-changing dynamics in which the di fferent political actors move and the national preferences of both the political parties and the institutions, or of the citizens themselves, who tend to develop an attitude towards the national question that is more strategic and less essentialist in character. According to this perspective, the national identity of individuals, or better put, the political appearance of that identity, is not so much an objective or subjective question, but a relational one; or as Habermas would put it, an "intersubjective" identity (Habermas 1994, p. 113) that takes shape depending on how the di fferent national identities deliberate with each other in the public sphere. When a political actor in Spain holds a certain position on a national question, their position is not generally related so much to the essentialist content of their identity seen in isolation (subjective national identity) as to the advantages that position brings them in relation to the others (intersubjective national identity). In each political circumstance, a more ethnic or a more civic nationalist position opens up certain opportunities for expression and closes others. Therefore, study of the national identity of the actors, especially the political parties, should be done according to the calculation of the advantages and risks their position involves facing di fferent national offers. In sum, political actors are not so subject to national political identity but are instead agents with strategic and adaptive capacity that use the national question as a tool (Humlebaek and Jiménez 2018) that is articulated in an unstable framework of political negotiation.

In the debate over applying the concepts of recognition and redistribution to plurinational political systems, recognition has been the term most used for explaining how relations amongs<sup>t</sup> the di fferent national realities of a country ought to be (Gagnon and Grégoire 2015, p. 82). Nonetheless, recognition tends to limit the intersubjective character of the national question as it proposes that nations should above all strengthen themselves *inwards* within a certain territorial framework. On the contrary, the concept of redistribution tends to widen the focus *outwards*, towards the space of political action (Shorten 2015) insofar as the actors accept the divisibility and negotiation of so-called national goods. As Christina Zuber notes, the plurinational phenomenon in decentralised political structures with a certain degree of asymmetry is a dynamic game of actors that moves simultaneously in two directions: On one hand, it is a vertical movement that is understood as a political conflict in national terms between the centre and the di fferent peripheries; on the other, it is a horizontal movement that is understood as an intergovernmental political conflict, also in national terms, amongs<sup>t</sup> *all* the di fferent territories that make up a country, given that any territory, and not only those that have regional nationalist parties, end up using the national question to their own benefit (Zuber 2011). The result of this process is a dynamic readjustment, both symbolic and material, of the plurinational question; that is, a permanent de-alignment and realignment of the ethnic and national boundaries of the people and the strategies of the political parties (Serrano and Bonillo 2017; Corujo et al. 2019). From an institutionalist perspective of rational choice, plurinationalism defines an intersubjective network of actors in which the latter relate to each other in a complex manner according to their power of negotiation and their forms of negottiation, on both the vertical and horizontal planes (Petersohn et al. 2015). As di fferent authors observe, the relation among actors within a plurinational state culminates in one of three scenarios: Secession, autonomy, or federalism (Mcewen and Lecours 2008; Lluch 2014; Basta 2018).

In the Spanish case, lack of recognition of the concept of *plurinationality*—a useful tool in political debate that could favour an intersubjective redistribution of the national question amongs<sup>t</sup> di fferent political parties, governments, and public opinion in general—has provoked two very specific situations. The first is a distortion in the political evaluation of the really existing model of territorial organisation in Spain (the autonomous community model), in comparison with the apparently less plausible model (the secessionist model) and with what appears to be a more favourable alternative (the federal model). As Bossacoma and Sanjaume-Calvet suggest, it is not clear if the Spanish Constitution defended asymmetry in the past but now it seems that "it was a transitory and potential asymmetry more than a permanent and actual asymmetry" (Bossacoma Busquets and Sanjaume-Calvet 2019, p. 457). Confusion in evaluating these three models has meant that in Spain a series of political actions are carried out that, in practice, go against the very idea of political decentralisation of power (Linz 1999; Maiz et al. 2010). Related to this, the second consequence of this lack of acceptance of plurinationality in Spain is the existence of a highly centripetal dynamic in the political relation amongs<sup>t</sup> the political actors, which prefer to maximise the adversarial component ("seeing the speck of sawdust in your brother's eye and paying no attention to the plank in your own eye") instead of exploring the possibilities, which are more costly and have a less certain outcome, of using a political language, a strategy and a framework of alliances that is more consociational and inclusive in national terms.

The system of territorial organisation of political power in Spain shows that it is one of the countries with the highest levels of political decentralisation and intergovernmental asymmetry in the world (Hooghe et al. 2008a, p. 221; 2008b, pp. 137–39). Nonetheless, the idea subsists that its functioning is atypical when compared with other countries where federal principles are operative (Aja 2014; Caamaño 2014; Maiz et al. 2018) and, to a certain point, it is also viewed as incomplete when compared with the asymmetrical federal agreements that characterise a model of *plural federalism* suitable for plurinational societies (Requejo 1999, p. 334). However, it is also true that, to date, no convincing explanation has been given of what a "plurinational federation" would actually consist of, beyond the conceptual tools of territorial federalism (Resnick 1994, 2004; Romero 2013). As Enric Fossas observes, one of the reasons for this distorted view of the decentralising model found in Spain lies in comparing it with egalitarian federalist proposals. In his opinion, symmetrical formulas for the decentralisation of territorial power that organise the power-sharing in terms that are more horizontal than vertical, in a multilateral way amongs<sup>t</sup> the di fferent territories and not in a bilateral way between some territories and the central power, are not the most suitable for the characteristics of the Spanish state (Fossas 1999, p. 279). In fact, in countries where there are sub-state communities with a strong di fferentiated national identity, reactions against the federal model are very large and make its viability very di fficult (Guibernau 2003; Lluch 2011, p. 134). In plurinational societies, the classical federal model does not have suitable tools for accommodating diversity and political asymmetry amongs<sup>t</sup> regions, unlike the autonomous model (Suksi 1998). The latter could, in principle, guarantee the original diversity and political asymmetry amongs<sup>t</sup> the territories according to their di fferent national composition, exactly as occurs in the model of decentralisation created in the Spanish Constitution of 1978. As argued in some works comparing di fferent federalising processes, the final failure of the Autonomous Agreement of 1992 between the two main Spanish parties, PSOE and the Popular Party (PP), which attempted to establish a process of harmonisation of the Spanish territorial system in terms of federal symmetry, explains the di fficulties in understanding the dynamic of the Spanish model from a strictly federal logic (Petersohn et al. 2015, p. 640). Perhaps the explanation for the problems of the Spanish territorial model does not lie in an apparent federal shortcoming that must be corrected, given that the initial intention of the Spanish democratic system never was federalism, nor perhaps is it the most appropriate model for the country's characteristics. On the contrary, the explanation lies in the dispute over how to understand the autonomous model of territorial power. In fact, if the concept of plurinational federalism has any practical value, it perhaps lies in associating it not with the theoretical language of federalism, but instead with the characteristics of the autonomous model, which has its own foundations and its own problems.

The Spanish autonomous model was designed in 1978 to combine the principles of the explicit will of regional self-government, political diversity, territorial asymmetry, and accommodation of differential national realities within the same common institutional political system (Fossas 1999, p. 291). Unlike the federal model, basically conceived for societies with national plurality where the function of coordination amongs<sup>t</sup> territories is directed more towards horizontal than vertical mechanisms in their relation to the state, the autonomous model combines multilateral and bilateral elements in the relation of the territories with the state. Obviously, the autonomous model has its benefit, sbut it is also more problematic, due to its asymmetrical character in trying to respond to heterogeneous national plurality. In the Spanish case, the autonomous model presents four characteristics that make it a peculiar case.

In the first place, it was initially conceived as an open model, through the so-called *principe dispositif*, which has enabled the territories to take the initiative in articulating their framework of competences and elaborating their legal status (Guénette 2016). This open model created the idea that the territories could have *sine die* a type of non-conditional will of self-government in order to redefine their degree of autonomy over time, a conception that is currently under discussion (Fossas 2008, p. 168). In the second place, the Spanish autonomous model is applicable to the whole territory with the result that the entire country, excepting the enclaves of Ceuta and Melilla, is organised into regions with a certain degree of autonomy. At the bottom, this fact introduces the idea that it is possible to articulate homogenising dynamics or, at least, it accepts that each territory will over time create an autonomous dynamic that will demand respect and a minimum degree of adaptation from the rest. In the third place, the Spanish autonomous model is a deeply asymmetrical one, both at the level of competences assumed by the territories themselves, and, above all, due to the juridical-constitutional articulation of two territories: The Basque Country and Navarre. These have a special status and an especially autonomous economic regime protected by the First Additional Provision and the Second Derogation Provision of the Constitution, while a third territory, the Canary Islands, also has special economic protection contained in the Third Additional Provision. This asymmetry has served as a referent for every type of political discussion on the model for financing the Spanish autonomous system (León 2015) and, above all, it opens up discussion on the advantages of this asymmetry and, where appropriate, the possibility of extending this status to other territories. Finally, the fourth element of the Spanish autonomous model proposes decentralisation with the aim of guaranteeing "the autonomy of the nationalities and regions"; that is, to enable the political development of "those collectives that have a differentiated political identity" (Fossas 1999, p. 289). This final criterion accepts that there are territories that have certain different identitarian features that deserve to be guaranteed because they incorporate national elements that do not appear in other territories. The problem here is that the decentralisation of power in Spain has not been carried out in terms of national groups, but in territorial terms instead. If a national minority coincides with a concrete territory to the point where there is a full symbiosis between national identity and territory, there will certainly be no room for differentiating between territorial and national decentralisation. However, if in those territories where there is a national minority no such fusion exists because that "collective endowed with a differentiated political identity" is in fact a plurinational political community, as indeed occurs in many regions in Spain, then a problem arises when it comes to attributing the concept of autonomy to the national idea instead of associating the idea of autonomy with a territory. Calling a territory a "nationality" or a "nation" is confusing, if in that territory there is, in the final instance, a plurinational society. It is not clear if autonomy is conceived for a part of that collective or for the territory that contains people with different national feelings. On this point, only two solutions are possible: (i) To aim to reduce over time the political distance between the plurinational collective and the territory to the point where the territory becomes a national collective; or (ii) to define the territory from a plurinational perspective as one that in fact exercises its autonomy through a plural and intersubjective national negotiation amongs<sup>t</sup> its inhabitants.

Having reached this point, the relevant question is to identify the dynamic developed by the political actors, especially the political parties through their involvement in the multilevel governance over 40 years of democracy (Fernández-Albertos and Lago 2015; Field and Hamann 2015; Field 2016) to tackle the phenomenon of plurinationality. Or, stated differently, to ask who has benefitted from, or been negatively affected by, denying the existence of plurinationality, both in the country as a whole and in the different territories where there are particular national sentiments. One indicator for testing this question is comparing the amenability of different political parties in Spain to forming plurinational governments, that is, governments formed by parties that respond to different national identities. Tables 1 and 2 refer to this indicator: In the first case, with reference to Spanish governments; in the second case, with reference to autonomous governments.

**Table 1.** Support received from regional nationalist parties in the investiture processes of governments in Spain, 1993–2018.


Source: Elaborated by the author. The table reflects only those investitures in which the winning party did not obtain an absolute majority and need to negotiate support. In the rest of the elections where the wining party ge<sup>t</sup> the absolute majority in the Parliament, the support of the regional nationalist parties to the winning party (as CIU and CC towards PP in 2000) did not have grea<sup>t</sup> political effects. \* In 2018, the investiture resulted from a motion of censorship.

**Table 2.** Participation of Spanish state-wide parties in the plurinational governance of the Autonomous Communities, 1986–2019.


Source: Elaborated by the author. The table reflects the participation of the state-wide parties (PSOE, PP) in plurinational autonomous governments at different levels: The column marked *Gov* refers to years of participation as leaders of a regional government; the column marked *Partner* refers to years of participation as partners in a government; finally, the column marked *Invest* refers to the number of times the party voted favourably in the investiture of an autonomous governmen<sup>t</sup> of a regional nationalist party. The two columns on the right (*Plurinational Vote*) refers to the percentage of votes for regionalist nationalist parties and state-wide nationalist parties in the regional elections (*RegElect*).

In light of the absence of coalition governments in the modern democratic history of Spain, in Table 1 we show one of the highest levels of participation by different regional nationalist parties in Spanish governance: Their vote in the investiture processes in those cases where the party aspiring to govern did not have an absolute majority. It can be seen that there have been significant differences between the PSOE and the PP. In the case of the PSOE all the regional nationalist parties have on some occasion shown their support for the investiture of its three candidates, especially the centre-right nationalist parties (CIU, PNV, and Coalición Canaria) but also the left-wing nationalist parties (ERC, EA, and BNG, and even the left-wing Basque nationalist party in the motion of censorship in 2018). However, in the case of the PP, it was only in 1996 when it received support from the centre-right nationalist parties, while on the two remaining occasions, it did not receive their support. A small di fferential strategy can be observed: In the case of the PSOE, its governance has sought to accommodate regional nationalist claims to some extent, while in the case of the PP, there has been much greater resistance to organising its governmental action through agreements with regional nationalist parties. In the exceptional case of the year 1996, the support between the PP and CIU consisted in a reciprocal agreement, the so-called *Majestic Pact*, to enable the formation single-party governments in both Spain and Catalonia.

If we analyse the participation of the PSOE and PP in plurinational governance at the autonomous level, which can be seen in Table 2, the di fferences between the two parties are even more striking. The PSOE has formed plurinational governments in all the autonomous communities where there are parties with a regional nationalist identity: In eight communities (Catalonia, the Canary Islands, Navarre, Aragon, the Balearic Islands, Galicia, the Valencian Community, and Andalusia), it has formed plurinational governments led by a socialist leader, while in the other two (Basque Country and Cantabria), it has formed plurinational governments as a partner in government. Furthermore, by abstaining in the investiture process, it has made possible the formation of some other regional nationalist governments in Catalonia and Navarre. In the case of the PP, its participation scores much lower in all the indicators: it has only participated in plurinational governments in three communities where the regional nationalist governments have little political weight (Cantabria, Aragon, and the Balearic Islands), and it has only enabled the formation of regional nationalist governments in two other communities (the Canary Islands and Catalonia; in the latter case, through the reciprocal agreemen<sup>t</sup> mentioned above). The strategy of the two parties has thus been very di fferent. In the case of the PSOE, its degree of involvement in plurinational governments at the autonomous level has been extensive and covers all types of autonomous community, not only where regional nationalist identity has no dispute with Spanish national identity, but also in those communities where regional and Spanish identities have a higher level of confrontation (the Basque Country, Catalonia, and Navarre). The involvement of the PSOE has been so extensive that it includes the anomalous situation of its having formed plurinational governments with a centre-right nationalist party, the PNV, in the Basque Country for 14 years in di fferent waves, in spite of the existence of an important left-wing nationalist party in that community. The link between this left-wing Basque nationalist party to the terrorist organisation ETA in the recent past makes reaching any type of plurinational agreemen<sup>t</sup> with it very di fficult. The PP has employed a very di fferent strategy: Its involvement in plurinational governments at the autonomous level has been quantitatively much less than that of the PSOE and has been concentrated in those regions where regional national identity has no quarrel with Spanish national identity. In those communities where regional national identity is in dispute with Spanish national identity, even where there are ideologically similar nationalist parties with which some type of understanding could be established (especially the Basque Country, Catalonia, and Navarre), the strategy of the PP has been to avoid such agreements. The exception to this was the case of Catalonia in 1996, mentioned above, which was justified on the grounds of reciprocity and had the aim of supporting two non-plurinational governments, without involving any type of coalition between the parties.

When analysing the position of the regional nationalist parties with respect to the Spanish state-wide parties, it is important to note their di fferent strategies when focusing on the plurinational political question. In the case of the left-wing regional nationalist parties, insofar as their political position is not hegemonic in their respective communities, their strategy has generally oscillated between reproducing the conservative regional nationalism strategy or being open to negotiation with Spanish state-wide parties if this decision gave them an opportunity to enter into regional governments: This was the strategy of the BNG in Galicia (2005), Compromís in the Valencian Community (2015), and above all the strategy of ERC in Catalonia (2004 and 2006) in order to compete with the nationalism of CIU as an alternative in governance. In the case of right-wing regional nationalist parties, insofar as they have been hegemonic in their respective communities (this is especially clear in the case of the PNV in the Basque Country and CIU in Catalonia), their strategy has in general been aimed at seeking any type of agreemen<sup>t</sup> to maintain that dominant position. In the case of the PNV, due to the ostracism of left-wing regional nationalism because of its link to violence, its strategy has consisted of guaranteeing its prevalent position (Zuber and Szocsik 2015) through a plurinational agreemen<sup>t</sup> with a left-wing state-wide party, enabling it to consolidate its hegemonic position with reasonable ease (Moreno 2000). In the case of CIU, the existence of a real alternative of a left-wing plurinational governmen<sup>t</sup> in Catalonia, as occurred between 2004 and 2010, obliged it to reconsider its strategy: To seek a non-plurinational Catalan governmen<sup>t</sup> or to seek a plurinational governmen<sup>t</sup> with the PP. Since 2010 CIU opted for the first strategy to recover power: It demanded a greater level of autonomy in Catalonia, as proposed in the reform of the Catalan statute initiated by the left-wing plurinational government, which was rejected by the PP and subsequently by the Constitutional Court in Ruling 31/2010 as well. Since then, reaching any plurinational agreemen<sup>t</sup> in Catalonia has become a very difficult undertaking.

The question remains of whether or not the political problem that has opened up with the so-called *procés* in Catalonia marks the high point of a model based on the explicit negation of plurinationality. This is clear in the strategies of the main centre-right nationalist parties in Spain: Above all, the PP and CIU; less so in the case of the PNV. To date, the asymmetrical character of the Spanish autonomous system has enabled some centre-right parties in Spain to maintain a political dialectic involving zero sum nationalist arguments, which has hindered any strategy of plurinationality. In the case of the centre-left Spanish parties, their doubts have lain in knowing which of the following three options is best. The first is a strategy in defence of plurinationality as an alternative model to that of the conservative parties: It is difficult to implement this model and it involves a high probability of electoral defeat. The second option is to succumb to the logic of nationalist confrontation that provides the right-wing parties with such substantial electoral benefits: Pursuing this option has been quite clear in the case of the Basque nationalist left, a little less obvious in the case of the Catalan nationalist left ERC, and much less so in the case of the PSOE. Finally, the third option is to wait until the clash between nationalisms converts the defence of low-profile plurinationality into a winning position: This is the option used by the PSOE, and also by the ERC in Catalonia at certain very specific moments in its recent history, with uneven results and a limited horizon of expectations.
