Theories of the Right of Secession: A Republican Analysis

Republican theorists have paid little attention to the normative problems of secession conflicts. So far, there is no such thing as a democratic republican theory of the right of secession (TRS), nor any comprehensive analysis of current TRS has ever been undertaken from a democratic republican point of view. This article tries to fill this second gap as a previous step in order to approach the first one. In doing so, it is shown how secession conflicts pose threats for two core democratic republican values: freedom and inclusion. The threats are, concretely, those of exclusion, blackmailing minorities, arbitrary permanent majorities, and instability. The article also shows how, due to their respective pro-unionist or pro-secessionist biases, no current TRS seems to be able to handle those threats, and briefly outlines how a democratic republican TRS, based on a non-unilateralist logic, could be developed.

This article aims to undertake a normative analysis of current theories of the right of secession (TRS) from the point of view of democratic republicanism. There has hardly been any analysis of this kind concerning these two bodies of normative literature, which I regard as a necessary step in order to fill another major gap: the lack of a democratic republican TRS. The few scholars who have worked on the issue have done so in a rather exploratory way (McGarry and Moore, 2011), usually as a secondary issue within broader works on nationalism (Ovejero, 2006, pp. 81-104), international law (Sellers, 2006, pp. 158-166) or self-determination considered more broadly (Klabbers, 2006). 1 And even scholars with an affinity to republicanism, such as Weinstock (with Nadeau, 2004) or Miller (2008), have not used republican concepts and principles in their works on secession (Weinstock, 2000(Weinstock, , 2001Miller, 2003). Only Catala (2017) has outlined some ethico-political duties of a potentially secessionist group, concerning non-domination, in one particular area (distributive justice).
This article tries to fill this gap, being a first step in order to develop a democratic republican TRS. Indeed, in order to develop such a theory, we need to find out in the first place whether it must be based on one current TRS, on a synthesis of them or on a completely new basis. As we will see at the end of the article, I am rather inclined to this second path. The article is structured in five sections. In the first two, I summarize the main traits both of democratic republicanism and current TRS, respectively. The third section shows the threats that secession conflicts imply for two core democratic republican values: freedom and inclusion. This third section shows as well how and why each TRS fails to provide a framework capable of protecting the two main factions of secession conflicts (i.e. unionists and secessionists) from domination and exclusion, since all them tend to be based on a pro-unionist or pro-secessionist bias. In the fourth, using these findings, I briefly outline how an alternative democratic republican approach to secession could be developed through further research. The fifth section summarizes my findings and conclusions.

Democratic Republicanism: A Definition
Republicanism, as any tradition of political thought, can be defined in different ways. This article is based in what nowadays is probably the main rational reconstruction of the republican tradition: that of Philip Pettit (1997Pettit ( , 2012. For Pettit, republicanism: (1) stands for freedom as non-domination, with domination meaning the arbitrary power of one individual or group over other people; 2 (2) argues that, in order to promote republican freedom, private sources of power (e.g. wealth) must be checked, controlled and dispersed by the state and by a civically virtuous citizenry; (3) argues that, in order to prevent the state itself from becoming dominating, it must be organized as a constitutional republic, 3 with its own powers being checked, dispersed, prevented from being monopolized by any faction, and kept under the rule of law and the vigilance of citizenry; and (4) argues that civic virtue is unattainable without freedom, and vice versa. Besides, democratic republicanism, as opposed to more oligarchic brands of the republican tradition, (5) seeks republican citizenship, based on republican freedom and civic virtue, to include as many people as possible. Thus, domination and exclusion, so defined, are the main concerns of democratic republicanism.
There is a point implicit in Pettit's work that I think he does not properly highlight: the key importance that, for republicanism, is held by the conflict between factions, which we can generally define as groups of people with some common perceived interest translated into a common political purpose. Almost all historical republican thinkers (including Aristotle, Machiavelli or Madison, among others) understood that every political community was divided between factions contending for political power. In this sense, they regarded constitutional checks and balances as needed, among other reasons, in order to prevent a factional takeover, which would end the definition of the political community as a res publica (public matter) by transforming it into a res privata (private matter) of the ruling faction. This feature of the republican tradition will be, as we will later see, central to my republican analysis of TRS.
Currently, those polities closer to democratic republican values are modern democracies, 4 in which citizenship rights (whether civil, political and socioeconomic), separation of powers and constitutionalism protect citizens against arbitrary power to degrees of depth and extension unparalleled in history. Pettit acknowledges this, while pointing out serious weaknesses of these polities in democratic republican terms (e.g. lack of civic virtue or unchecked economic power) (2012, p. 23). Republicanism, so defined, has been developed to deal with different topics of political theory: the definition of concepts like freedom or civic virtue, the proper institutional design to promote republican values, the tense relationship between civic virtue and commerce or the place of republicanism in national or international law.
However, historically, republicanism as a tradition of political thought has not produced much theory about the problems of nation-building, from language regulations to immigration policies or cultural diversity. An oblivion shared, we must notice, by almost all theories of democracy until recent times. As Requejo and Caminal have said, What seems increasingly untenable is not what traditional democratic liberalism and other ideologies say, but what they do not say because they take it for granted: a series of theoretical assumptions and common places of a 'statist' nature that characterize the nation-building processes. (2011, p. 2).
Secession is one of those problems that republican thinkers have barely theorized about. And what TRS do, precisely, is to question the statist assumptions pointed out by Requejo and Caminal, whether to criticize them or to find out their best normative foundations.

Current Theories of the Right of Secession: Main Features and Criticisms
Current TRS tend to focus, implicitly or explicitly, on the unilateral right of secession, which is "the principal focus of interest for theorists of secession" (Pavkovic and Radan, 2007, pp. 200-201). They try to establish who has a moral (i.e. not necessarily a legal) right to secede, and why. Though discussions on institutionalization are not excluded from this field, the main concerns of TRS are the two just mentioned. They are usually classified in three groups: plebiscitarianism, adscriptivism and remedialism. The first two conceive secession as a primary right, i.e. a right to which some groups of people are entitled a priori, with no need to justify their decision; they only differ on the definition of which people(s) are entitled to this primary right.
For plebiscitarianism, the subject of the right of secession is simply any territorially concentrated group containing a secessionist majority. Wellman (1995), Beran (1984) and Philpott (1998) are some well-known examples of these theories. Despite their differences, they always seem to appeal to the same idea: in liberal-democratic terms, the state is the servant of the people, not vice versa; therefore, a state's legitimacy is based on consent. In second place, adscriptivism restricts that primary right to certain groups which are linked by objective features such as language, history, self-government institutions, traditions, world-views and so on. Miller (1995Miller ( , pp. 84-85, 2003, Nielsen (1998) and Margalit and Raz (1990) are typical representatives of adscriptive theories. Broadly speaking, the idea is that some objective features create a common identity (usually labeled as national) among the people who share them; this common identity provides some important goods to these people; and this fact makes it justifiable for such people to build their own state, if that is their will.
These primary right theories have received different criticisms, two of the most common ones being: (1) secessio ad infinitum, i.e. the risk that a primary right of secession could lead to recursive secessions, and hence to instability; and (2) the blackmail threat, that is, the risk of giving privileged minorities the power to threaten the whole polity, and even to undermine citizenship rights and social justice (e.g. the Slave States' secessionism in the antebellum US). In addition, adscriptivism has been further criticized for: (3) weak operationalization: it is difficult to give an empirically operational definition of the adscriptive features that a human group must share in order to be considered a nation or an encompassing group; (4) unclear normative logic: it is not clear why an encompassing group should have the right to its own state, since the reasons usually pointed out in this sense (e.g. common interests, importance for individual identity and values) are easily applicable to other kinds of groups, such as social classes; and (5) the threat of exclusion on ethnic-cultural grounds: those who live in the territory of the adscriptive group entitled to secession, but who do not share its adscriptive (normally meaning cultural) features, might be seen as second-class citizens, be excluded from the decision on secession, or even be excluded of citizenship altogether.
In order to overcome these problems, different primary right theorists propose placing restrictions on this right. Secession would then be a primary right of certain groups, if certain conditions are met. Each primary right theorist has his own set of limits to unilateral secession (Wellman, 2005, p. 36;Beran, 1984, pp. 30-31;Philpott, 1998, p. 80;or Margalit and Raz, 1990, pp. 459-460), but normally all of them refer to the foreseeable capability (and will) of both the seceding state and the former host state to keep fulfilling the obligations of a functional modern democratic state. However, the more limits we put on a primary right of secession, the more it needs an arbiter to interpret when those limits have been passed (which, of course, raises the problem of who should be the arbiter), and the more it loses one of its strong points: the open questioning of the arbitrariness of most boundaries as they have been developed throughout history (since we will be assuming that those boundaries can only be modified by those secessionist groups which happen to satisfy a series of conditions that we are not demanding of currently existing states in order to continue their existence).
Lastly, we find remedialism, that is, those theories which define the right of secession not as a primary right but as a remedial one. For remedialists, there are no groups that have this right for their own sake. Secession must be seen as a last resort to be used by those groups that are the victims of serious and constant injustices and/or grievances. What these injustices and/ or grievances are that legitimize a group to exercise secession is a matter of discussion within this theoretical family, but they generally agree on some basic ones, such as massive and constant violations of basic human rights, or unjust annexation. Buchanan (1991Buchanan ( , 2007 is probably the best-known representative of this group, within which we can also find scholars like Birch (1984), Norman (1998Norman ( , 2003Norman ( , 2006 or Seymour (2007). The idea is that a state's legitimacy is not based on consent, but on the state's capability and willingness to keep basic standards of justice among its citizens. Hence, in a modern democracy, unilateral secession would be forbidden a priori.
These theories have the appeal of avoiding the weak points of primary right theories. However, remedial theories have also been criticized as biased towards the statu quo, assuming the legitimacy of current boundaries while putting the burden of proof on secessionists. This is quite problematic, since most boundaries are the result of historical processes (e.g. wars) rather far from complying with the normative values that remedial theories rest upon. Remedialism seems to imply an answer to this objection: as long as states are reasonably just, the shape of their borders is not relevant. However, between the massively oppressed minority and the privileged, blackmailing one, there are a lot of intermediate stages in which there is an absence of intolerable injustice; there are genuine discussions on certain issues related to state or nationbuilding (e.g. language regulations, decentralization or territorial allocation of economic resources) in which territorialized, permanent minorities (e.g. Quebecers within Canada, Basques or Catalans within Spain) are usually the weak part under state-majority rule.
Remedialists tend to employ two main strategies in order to avoid these problems: (1) to argue for intrastate reasonable degrees of autonomy for minorities (Buchanan, 2007, pp. 401-424); but this can hardly be seen as a solution since the definition of reasonable degree of autonomy would be differently interpreted by majorities and minorities and, again, here the weak part would generally be minorities; and (2) to extend the catalog of "just causes" for secession in order to include, for instance, the failure of a state to adequately recognize the national identity of the secessionist target group (Bauböck, 2000;Patten, 2002;Seymour, 2007). However, the more we extend the catalog, the more remedialism loses one of its main appeals: clarity in the delimitation of a demos with a reasonable claim against staying within its host state. It is relatively easy to determine if a group is a victim of massive violations of human rights, but it will usually be controversial whether a group has a national identity, in the first place, and whether it has been, or not, adequately recognized by the state.

Democratic Republicanism And Current TRS
As we have seen, all TRS are in some way criticized for being unfairly or dangerously biased either towards the state or towards secessionists. This is not an ultimate reason to reject them all, but if we explore the reasons behind this common weakness, we will find some useful insights in order to understand how democratic republicanism must look at secession conflicts. In this sense, I would say that the main reason for that common weakness is that all current TRS choose an a priori winner in secession conflicts, an actor who is not asked to prove its legitimacy to unilaterally define the boundaries of the state. This poses an evident problem in terms of building a consensus between the two main actors in any secessionist controversy, i.e. unionists and secessionists. And, as we have seen, there do not seem to exist clear ultimate reasons to impose the burden of the proof on either of them. In my view, this shared trait is the consequence of a common search: the search for the demos of democracy. All current TRS implicitly assume that, once you find this demos, the main normative problem of secession conflicts is basically solved.
I think this way of looking at secessionist politics is, to a great extent, an inheritance of the three basic modern conceptions of nation, which in the end is one of the most common ways to name the demos of modern democracy: (1) a voluntaristic conception which Renan (1996) famously summarized in the metaphor of the nation as a daily plebiscite; (2) an organic conception of nation as a group of people linked by some common objective traits (e.g. language, culture or even race), as we can read in Fichte (1997); and (3) a statist conception of nation as the citizenship represented in the legislature of the state, which can be found in Sieyès (1982). It is easy to see links between these conceptions and plebiscitary, adscriptive and remedial TRS, respectively, and it is important to recall that all three were designed in order to argue for concrete state and nationbuilding projects. 5 This means that to handle secession conflicts from one of these three conceptions is the equivalent of handling them from a point of view raised by some of the conflicting factions in order to win those conflicts. 6 That is not an ultimate reason to reject current TRS, but it is a way of understanding the roots of their difficulties in reaching consensus.
So rather than asking who is right and who is wrong in secession conflicts, we must conceive and develop rules and institutions able to channel these conflicts in civilized ways. 7 I think that democratic republicanism, with its republican conception of factional conflicts as a source of domination, and its democratic concern regarding inclusion, is well suited to face this challenge; it will only need to update its catalog of factional conflicts, as it has done other times in the past. 8 A democratic republican approach to secession must therefore look at contenders of secession conflicts as factions of the ultimate expression of center-periphery conflicts, usually confronting permanent majorities and permanent minorities. Thus, the normative analysis of secession conflicts must be concerned on minimizing the risks of factional domination and/or exclusion. 5 Thus, Renan, in 1882, wanted to argue for the continuity of Alsace-Lorraine within the French Republic, which was the will of the majority of their inhabitants despite their German ethnolinguistic heritage. Fichte, in 1808, was concerned with the unification of the Germanspeaking states. And Sieyès, in the wake of the French Revolution, wanted to identify the French state with its citizenship, thus replacing the old absolutist identification between the state and the monarch. And, in different times, different movements have assumed one of these concepts of nation in order to legitimate their aspiration to create a state or to keep one united.
6 Some remedial TRS are linked to some notion of plurinational federalism (Bauböck, 2000), and therefore it could be argued that they do not share this link with these three brands of nationalism. However, I think this would miss one important point: plurinational federalism recognizes, indeed, the existence of internal nations, in a certain sense, within a shared federal state; but nevertheless it does not question the very fact of the sovereignty of that federal state (unless the federal state infringes what plurinational federalism considers to be the rights of national minorities). And what those classical theorists of the nation had in mind when they formulated their theories was the justification of certain claims of sovereignty. In this very crucial sense, a remedial TRS grounded on plurinational federalism retains a statist background à la Sieyès.
7 In a way, this is what Sanjaume-Calvet (2020) attempts in his realist approach to secession. 8 In Greek-Roman ancient times the main concern of democratic republicans was to prevent exclusion and domination based on social class or political party; early modern republicanism incorporated a concern about exclusion and domination based on religion; and further developments also included concerns about exclusion and domination based on gender or race, as well as regarding domination between states or different levels of government between multilevel states. However, there is a key point in which secession conflicts are different from any of the factional conflicts that republicanism has ever been concerned about: what is at stake is not how the state should be governed, who should govern it, or even how power must be distributed within the borders of the state; instead, the objects of the conflict are precisely those borders. A problem that traditional republican solutions are ill-prepared to handle, for they are designed to work within the state (or, in republican International Relations theories, between states). However, the idea of preventing the arbitrary rule of one faction over the other, I think, still makes full sense in secession conflicts. Therefore, the principles inherited from the republican view of factional conflicts are still valid, although we would probably need to translate them into rules and institutions which are likely to be sensibly different from others previously envisioned by republican thinkers.
In my view, if we look at secession conflicts from this democratic republican point of view, we can identify four main normative threats linked to them: (1) exclusion: some people who would be directly affected by secession may be excluded from deciding on the issue and/or from full citizenship in the new state (e.g. an ethnic minority within the seceding territory); (2) domination by blackmailing minorities: in case of being entitled to secession, powerful minorities might be able to blackmail the rest of the polity; (3) domination by arbitrary permanent majorities: secession conflicts are usually an ultimate expression of particularly deep center-periphery conflicts between a central permanent majority and a peripheral permanent minority; 9 both of them defined along constant disagreements on how the state should be conceived and organized in terms of economy (i.e. territorial organization of it), territory (i.e. territorial distribution of political power) and identity, 10 so that without a feasible exit option, permanent minorities would be at mercy of arbitrary permanent majorities; and (4) instability, i.e. the risk that a wrong management of secession conflicts is likely to promote instability, 11 eventually triggering exclusion and/ or domination. Thus, democratic republicanism must evaluate current TRS in terms of their capability to confront these four threats.
To begin with, I see the threat of exclusion as particularly present in adscriptive TRS. If, for instance, we define Quebecois people in linguistic (French-speaking) terms, and we give this people the right to secede, then English-speaking Quebecers might be excluded from a referendum on the secession of Quebec. Besides, some scholars have argued that this threat is also present in plebiscitary theories (Ovejero, 2011, pp. 155-201). The argument goes like this: republican freedom requires, as we have seen, the citizens' right to participate in politics, as well as civic virtue. On the other hand, democracy requires a maximally inclusive citizenship. This ideal is realized, however imperfectly, in modern democracies. To give some people the unilateral right to secede from a democratic political community would therefore mean: (1) to let some people decide over a matter that affects the whole political community; and (2) to declare all the rest of the political community as strangers; that is, as non-citizens.
In my view, these are misleading objections. Concerning the first one, we can only consider that secession affects the whole political community if we take for granted that the political community (the nation, one could say) is the host state. If not, secession may "affect" the host state very much like many decisions taken by the host state (e.g. concerning tariffs or immigration policies) may affect its neighbors. If this is an argument against plebiscitarian secession, it should equally be an argument against the independence of the host state. 12 Concerning the second objection, we must realize that secession does not exclude the people at the other side of the new border in the same way in which, for instance, African-Americans were excluded from full citizenship in the Southern U.S. before the 1960s. Taking again the example of Quebec: if Quebec seceded unilaterally after a referendum in which all Quebec residents had been able to participate, and if all Quebec residents were entitled to the new Quebecer citizenship, then the people of the rest of Canada would not have been more excluded from Quebec than they currently are from the U.S. They would continue to be full members of a democratic community. The territorial borders of their democracy would be smaller; their inclusion within that democracy would not.
Nevertheless, a threat that indeed affects both adscriptive and (particularly) plebiscitary TRS is the threat of blackmailing minorities. As we have seen in the previous section, primary right theorists propose overcoming this risk, as well as the risk of exclusion, by putting restrictions on such primary right, normally linked to the foreseeable capability (and will) of both the seceding state and the remainder one to keep fulfilling the obligations of a functional modern democratic state. However, we have also seen the weaknesses of this solution: (1) the need of an arbiter; and (2) the weakening of the open questioning of the arbitrariness of most states' boundaries.
These two threats of exclusion and blackmail, linked as they are with primary right TRS, have led republican scholars like Ovejero (2006, p. 81) or Sellers (2006, pp. 158-166) to embrace remedialism. I regard remedialism, however, as ill-prepared to handle the threat of arbitrary permanent majorities: by putting the burden of the proof on the secessionists' shoulders, remedialism gives the high ground to state-level permanent majorities to arbitrarily decide what degree of autonomy, recognition or economic promotion they will give to permanent minorities. Certainly, remedialism places a limit on this arbitrariness: the state, and therefore a ruling permanent majority, is not allowed to commit evident injustices against minorities. But a permanent majority can be highly dismissive towards permanent minorities' identitary, territorial or economic aspirations while essentially respecting the limits imposed by basic human rights and modern democratic institutions. For instance, it can decide, through strictly democratic procedures, to remove the teaching of the indigenous language of a permanent minority (if it happened to have one) from public education across the state.
From a democratic republican point of view, this criticism of remedialism can be answered by arguing that, in order to protect republican freedom and civic virtue, democracy should be understood in a deliberative way. Citizens of a democratic republic have to be open to argue their views and, in the process, to convince and to be convinced by other people by means of public reasons, i.e. reasons that appeal to the common good of all citizenship, and not just to the particularistic interest of one individual or faction. This way, factional domination would be expelled from political conflicts, including center-periphery ones. Starting from this point, it is easy to follow: if the permanent minority's claims can be sustained on reasons of common interest, they will prevail; if not, they will fail. And in no way could this be attributed to their permanent minority status, but only to the fact that their claims would have a factional nature. To give them a primary, unilateral right of secession would undermine all this deliberative architecture.
In my opinion, this view confuses wishes with reality. This deliberative ideal has indeed a good republican pedigree (Pettit, 1997, pp. 187-190;Sunstein, 1988;Bello Hutt, 2018), and from a republican point of view it is clearly desirable for democracies to work, as far as possible, in a deliberative way, and therefore to design their institutions to promote this deliberative dynamic. The fact is, however, that more often than not, our modern democracies do not work that way. That is not to say that rational discussion and common interest have no place in our modern democracies, but they do not have a clear, ultimate place which is strong enough as to discredit any secessionist claim as anti-deliberative. Factional conflict does not disappear, nor it is well handled, just by saying that "we should not be factional" and that "we should think about the well-being of the whole political community"; it is rather the other way around: factions will find themselves forced to discuss each other through public reasons if we minimize, through proper institutional designs, their chances of plainly and arbitrarily impose their will onto other factions.
Indeed, McGarry and Moore (2011) provide, from a republican and remedialist point of view, a brief and interesting exploration of how, within democracy and the rule of law, the majority group in a state can "ensure its domination over state institutions and through this, over the society as a whole" (2011, p. 436). However, I find some weak points in their analysis: (1) they do not explore the threats of domination that other TRS pose for republican freedom; (2) they describe the ways in which a state-level majority can cheat within the democratic game in order to ensure its permanent hegemony (e.g. gerrymandering or a self-favoring design of the electoral system), but they do not clearly consider the fact that, even with reasonably fair democratic rules, a permanent majority still retains an a priori power to arbitrarily ignore consistently predominant demands and views of permanent minorities, unless they become entrenched with the means to counter such arbitrary power; and (3) they conclude that remedialism should include a republican perspective on domination and apply it to this sort of democratic domination of majorities over minorities, but they neither explore the possibility of just choosing another TRS instead of remedialism, nor realize that there are reasons that remedialism as such can be unsuitable for a republican reformulation.
As we have already seen, there are two remedialist strategies in order to overcome this bias against permanent minorities. The first one was the defense of reasonably high degrees of intrastate autonomy. Unfortunately, we have also seen a critical weakness in this strategy: it is rather hard to establish objective criteria on what should be seen as a reasonably high level of autonomy. The second strategy was, on the other hand, the inclusion of insufficient self-government, discriminatory redistribution and/or failure of recognition within the catalog of just causes for unilateral secession; Bauböck (2000), Patten (2002) and Seymour (2007) are some proponents of this version of remedialism. As we have seen, however, this strategy weakens the most interesting feature of remedialism: its clarity in defining who has a right to secede without allowing exclusion or minority blackmailing.
The last of the four threats we saw at the beginning of this section was instability. It is the only one that, in my view, is common to all three TRS and, more importantly, also to the option of rejecting all three without an alternative. Under rules which could be reasonably presented as biased towards one faction of secession conflicts, the disadvantaged faction (whether secessionists or unionists) is likely to reject such rules, which in turn will lead to instability and to a might makes right logic, the nemesis of republican freedom. In the absence of any rules at all, unless both factions are spontaneously reasonable and open to dialogue, such instability and logic will arise even more quickly. So it seems that all current TRS, when reviewed from the point of view of democratic republicanism, share the same mistake: they all favor one faction of the conflict at stake, instead of looking for a way to balance the power between them so that no-one can dominate the other.
In the end, we seem to be stuck in a cul-de-sac: on the one hand, secession conflicts pose serious threats for democratic republicanism; on the other hand, no current TRS seems to be able to overcome them all. However, it is unlikely that neither democratic republicanism, nor current TRS, would be better suited to handle secession by simply ignoring this problem. So the exploration of the best way to reconcile both theoretical fields seems to be a relevant research issue. But then, new questions arise: is this reconciliation possible? That is: can we develop a democratic republican TRS? I think we can, and before finishing this article I would like to point out the basic tasks that must be fulfilled in order to develop such theory.

Towards a democratic republican theory of the right of secession
I have argued that current TRS share a common weakness based on a common search: the search for a sovereign demos. In my opinion, in contrast, the idea of popular (or national) sovereignty is an abstraction with no intrinsic, but only instrumental, value, based on a twofold usefulness: (1) to the extent that it prevents states from arbitrarily interfering in the affairs of other states, unless in case of extraordinary circumstances, hence promoting peace and order in the international arena; and (2) to the extent that it establishes the state as servant, and not as master, of the people, therefore encouraging common people to not tolerate domination by governments and elites. But it is still an abstraction: peoples exist in a certain sense, but they do not exist in the same sense as individuals do. Peoples do not decide anything, nor do they violate any right or become oppressed, literally: individuals do, either alone or in conjunction with other individuals. Therefore, the idea of a people holding sovereignty despite what individuals might think is, in my opinion, a case of reification, in which we are treating an abstraction as if it was a concrete, objective real event.
As long as the individuals who happen to be citizens of one given state recognize each other as members of the same people or nation, the idea of popular or national sovereignty remains useful in the two senses I have pointed out; we can label these cases as uncontested political communities (e.g. Iceland, Switzerland or Japan, to name very different cases of the same concept). But then we can also think of a situation in which a good number of the individuals who happen to live in one given part of one given state happen, on the contrary, to challenge this assumption of being a people together with the rest of the citizens of the state, and more concretely happen to campaign for secession; we could label these cases as contested political communities (e.g. Canada and Quebec, Spain and Catalonia, UK and Scotland, or Belgium and Flanders). To use the concept of popular or national sovereignty as a way of discrediting the aspirations of either unionists or secessionist members of such communities is not likely to make any positive contribution in order to solve this kind of controversies.
Then, from a democratic republican point of view, the question, in contested political communities, should not be "who is the demos?" but rather "how can a contest over who is, or should be, the demos, be handled in a way that maximizes people's protection against exclusion and domination of any sort?". It is clear that, from this point of view, a unilateral right of secession must be a last resort in the face of very concrete grievances committed by the unionists, and in particular by the state. A democratic republican TRS would therefore agree with remedialism in this sense; the difference would be that such theory would also consider the unilateral right of the state to suppress or ignore a demand of secession as a last resort in the face of very concrete grievances by the secessionist group. If I am right, a democratic republican TRS should be based on a sort of double-direction remedialism, in which the last-resort logic is applied to both sides of the conflict, not just to secessionists; and on the other hand, this TRS will also share some tenets of primary right theories, by providing a non-unilateralist path for secessionist demands and, at the same time, denying the host state the right to unilaterally suppress or ignore them.
In other words, in the absence of these very concrete grievances by either faction, neither the right of secession of the regional-level secessionist majority, nor the right to territorial integrity of the state-level unionist majority would be unilateral a priori; instead, some institutional design, based on a non-unilateralist logic, should be in place in order to handle a negotiation process between the two parts. Therefore, the two basic tasks in order to develop a democratic republican TRS would be: (1) to develop this institutional design as the normal framework for handling secession conflicts; and (2) to specify under which concrete conditions one of the two parts would be allowed to unilaterally impose its will. For sure, this will imply the exploration of the possibility of a constitutional right of secession; but we will need to be careful in distinguishing the idea of a non-unilateralist normative framework for secession conflicts, on the one hand, from its legal translation, on the other. This last could be entrenched in constitutional law, in another legal body (e.g. international law) or in no legal body at all, remaining just as a set of criteria to be applied ad hoc by states, secessionists and international organizations on a case by case basis.

Conclusion
Through this article, I have undertaken a normative analysis of current TRS from the point of view of democratic republicanism, finding that: (1) secession conflicts imply threats for democratic republicanism in terms of exclusion, minority blackmailing, arbitrary permanent majorities and instability; (2) democratic republicanism shows that all current TRS, as well as the option of simply rejecting them all, fall into some of these threats; (3) therefore, current TRS and democratic republicanism seem to be in a normative relationship characterized by tension; (4) nevertheless, it is unlikely that neither two will be better suited to handle the normative problems of secession by simply ignoring each other; and (5) a non-unilateralist TRS could be the proper embodiment of democratic republicanism when applied to secession conflicts. Indeed, I hope this article will be seen as a first step in this direction. This is a small part of a longer road that, however, I think is worth taking.