Rethinking Castoriadis’s Concept of Self-Limitation: The Case of Swiss Women’s Suffrage

Abstract

For there to be democracy, there must be self-limitation. This is one of the key ideas in the thought of Cornelius Castoriadis. Yet he says little about what exactly self-limitation means. The examples he gives by way of illustration – Sophocles’ Antigone and the graphe paranomon – are only of limited help. His interpretation of the former is highly idiosyncratic; his account of the latter relies on considerable simplification. This article makes up for this lack of clarity. We show that there are two kinds of self-limitation. One is about matters within the demos; the other concerns who should belong to it. Our interest is in shedding light on the latter. We do this by drawing on the history of women’s suffrage in Switzerland and by reading Castoriadis against Castoriadis. Our argument is that self-limitation is best understood as a reciprocal process, as something required of both those who already belong to the body politic and those who wish to join it.

Introduction

In 1848, in the heart of a continent largely under the rule of monarchs, a unique federal state sprang into existence: democratic Switzerland. Before long, the country put into place a series of direct democratic procedures – most notably the right to hold referenda in 1874 and the right of popular initiative in 1891 – thus securing its image as one of the ‘oldest democracies in the world’ (Studer, 2021: 26, our translation). A Willensnation, a nation by will, rather than a community bound together by a common language, religion, or origin, Switzerland was and still is characterized by its strong federalism, with a high degree of autonomy at both the local and the regional level. At first blush, its record on women’s rights seems equally impressive. It was a pioneer among German-speaking countries in opening universities to women and introduced women’s suffrage not in the wake of a regime change or political upheaval but by popular vote. And yet, from today’s perspective, there is something rotten in the history of the state of Switzerland: it was not until 1971, long after its neighbouring countries had granted women equal rights, that Swiss men voted in favour of women’s enfranchisement, leaving a permanent blot on the escutcheon of Swiss democracy.

Nevertheless, the case of women’s suffrage in Switzerland is not just yet another reminder of the power of patriarchy (as if such a reminder were needed). There is more to it than that, especially when looked at from the vantage point of democratic theory. In what follows, we argue that the introduction of Swiss women’s right to vote points to the pivotal role of self-limitation in democratic societies. The idea of self-limitation comes from Cornelius Castoriadis and lies at the heart of his conception of democracy. And yet, as we are about to show, he leaves it strangely undertheorized, the consequence being that its potential is still not fully understood. The purpose of this article is to address this problem by offering a fresh take on what democratic self-limitation is.

This article has three sections. The first section offers an analysis of the two main examples Castoriadis gives to get his idea of self-limitation across: Sophocles’ play Antigone and the graphe paranomon, a legal device in Athenian democracy. Both are only of limited help. Castoriadis’s interpretation of the former is highly eccentric; his account of the latter is oversimplified. As a result, the concept of self-limitation is very much in need of clarification. We attempt to meet this need by turning to the example of Swiss women’s suffrage. Prior to that, however, we make a distinction between two kinds of self-limitation. The first refers to affairs internal to the demos; the second concerns who should belong to it. Our focus is exclusively on the latter. In the second section, we give a brief overview of the history of women’s suffrage in Switzerland, concentrating in particular on women’s activism. The third section uses the example of Swiss women’s enfranchisement to show that Castoriadis’s poorly developed theorization of self-limitation allows for two contradictory interpretations. This leads us to try to clarify the concept of self-limitation by thinking with Castoriadis against Castoriadis. The position we arrive at is that self-limitation as regards the question of membership depends on reciprocity.

Castoriadis’s Concept of Self-Limitation

One, if not the, basic premise of Castoriadis’s (1991b: 162) thought is that society must do without ‘an extrasocial source of nomos’. Those who look for a norm outside the body politic to guide them as to its proper institution are on a fool’s errand. For there is no such norm. Neither nature nor reason nor history contains anything that would pass for a firm foundation. It follows that it is entirely up to the demos to decide on the kind of society it wishes to create. What marks out a democratic society is its awareness of this fact. The people know that they are the source of the law they give themselves (Castoriadis, 1991a: 114–15). In this sense, democracy is ‘the regime of autonomy’ (Castoriadis, 2005b: 203). ¹ Yet whatever law the demos comes up with, Castoriadis (1993: 329) argues, it ‘will never exhaust the question of justice’, ever-changing circumstances and intractable individual cases routinely causing it to fall short. This, then, is where the rub lies: on the one hand, there is no society without law; on the other, no law can ever presume to be fully just. The conclusion Castoriadis (1993: 329) draws from this is that a ‘just society is not a society that has adopted, once and for all, just laws. A just society is a society in which the question of justice remains constantly open’. ² Together with the inevitable clash of different opinions and the unpredictable consequences of human action, this lack of closure necessitates the constant self-institution of society (Castoriadis, 1993: 329).

Now, if all power is vested in the people and the belief in ultimate foundations has been dispensed with, there is, at least in principle, nothing the people cannot do, no external boundary to tame them. As Castoriadis (2005b: 202) puts it, ‘as soon as society no longer accepts any transcendent or merely inherited norm, there is nothing that might, intrinsically, be able to set the limits where this power has to stop’. In other words, there are no safeguards outside society against the excesses of hubris – that is, against grave abuses of power, including the abolition of democracy by (seemingly) democratic means. The corollary is that the demos itself must somehow bridle any harmful or self-destructive impulses: ‘In a democracy, people can do anything and must know that they ought not do just anything’ (Castoriadis, 1991a: 115). What is needed, then, is self-limitation, which turns out to be just as essential to democracy as its capacity for self-rule. Democracy is not only the regime of autonomy but also the ‘regime of self-limitation’ (Castoriadis, 1991a: 115).

What are we to make of this idea of self-limitation? Although he puts great stress on its centrality to democracy, Castoriadis says relatively little about what exactly he has in mind. Indeed, instead of providing us with a clear definition, he seeks to outline self-limitation by way of illustration, focusing in particular on two examples from ancient Greece – tragedy and the graphe paranomon. Let us consider them in turn.

Tragedy, according to Castoriadis (1991a: 117–18), is predicated on the notion that ‘Being is Chaos’, meaning that the protagonists of the classical Greek, or Athenian, ³ plays have to deal with an ‘absence of order for man’, have no control over the consequences of their actions or what they come to signify, and tend to fall prey to their own hubris. Under these conditions, self-limitation is key. Think, Castoriadis (1991a: 120) suggests, of Sophocles’ Antigone: The reason the play ends in carnage – that is, with the deaths of Antigone, Haemon, and Eurydice – is that ‘both Creon and Antigone insist on their own reasons, without listening to the reasons of the other’, a folly captured by the Greek monos phronein. While Creon pig-headedly takes the position of the law, Antigone declares it null and void. The former does not tolerate dissenting opinions and disregards that one cannot control the consequences of one’s actions; the latter brushes aside the simple truth that ‘no city can exist [. . .] without nomoi’ (Castoriadis, 1991a: 120). Predictably, catastrophe ensues, the moral being that self-limitation is paramount if democracy is to survive.

Antigone is [. . .] a high point of thought, of the democratic political attitude. That attitude bars and condemns monos phronein, recognizes the intrinsic hubris of men, responds to it with phronēsis, and faces square-on the ultimate problem of autonomous man: the self-limitation of the individual and of the political community. (Castoriadis, 2005a: 27)

This is the genius of Sophocles’ Antigone. It brings to the stage one of the key principles of democracy – namely, that ‘even when we are right, it is possible that we might be wrong’ (Castoriadis, 2005a: 26). In this sense, it serves as a ‘constant reminder of self-limitation’ (Castoriadis, 1997c: 93), urging us to accept that no one can ever claim to have succeeded in instituting a just society.

Note, however, that this is a rather idiosyncratic interpretation of Antigone. The whole structure of the play seems to support the view that the brunt of the blame lies with Creon, not with Creon and Antigone. ⁴ Also, scolding Antigone for refusing to accept that no city can do without laws is highly dubious. For she opposes not so much the law as such as a particular tyrannical decree that is at odds with Greek custom. ⁵ In short, the case for Antigone as a play about self-limitation is not as straightforward as Castoriadis makes it out to be.

Let us now turn to the graphe paranomon. This device of Athenian democracy made it possible to prosecute a citizen for proposing to the demos a motion that ‘was contrary to the laws in force, hostile to the interests of the people, or procedurally invalid’ (Schwartzberg, 2013: 1049). Once a graphe paranomon had been brought, it fell to the People’s Court, whose members were chosen by lot, to adjudicate on the case in question, possibly resulting in the rejection of the motion and severe sanctions against its proposer (Hansen, 1991: 205–8). Because the People’s Court did not consist of legal experts but was a selection of members of the demos, Castoriadis (1991a: 117) celebrates the graphe paranomon as a ‘magnificent example of an effective institution of self-limitation’.

Since the source of the law is the people, ‘control of constitutionality’ could not be entrusted to ‘professionals’ – in any case, the idea would have sounded ridiculous to a Greek – but only to the people themselves acting in a different guise. The people say what the law is; the people can err; the people can correct themselves. (Castoriadis, 1991a: 117)

The graphe paranomon is certainly a more clear-cut case of self-limitation than Antigone. On closer inspection, however, one cannot but notice that Castoriadis simplifies things considerably in order to make his point. It is true that in fifth-century BCE Athenian democracy ‘any citizen could bring a graphe paranomon against any decision of the Assembly or the Council’ (Hansen, 1991: 175), with the first documented use dating back to 415 BCE (Atkinson, 2015: 2; Hansen, 1991: 205). But when democracy was restored in 403 BCE after being crushed by the Spartans, it saw some profound changes, not least to the role and function of the graphe paranomon (Cartledge, 2016: 105–6; Hansen, 1991: 151). And because ‘the overwhelming bulk of evidence for “the” Athenian democracy is concentrated in the thirty years [. . .] between about 350 and 322 [BCE]’ (Cartledge, 2016: 105), it is precisely this latter, revised version of the graphe paranomon that historians and classicists usually focus on. How does it differ from its earlier counterpart? From 403 BCE onwards, the Athenians made a distinction between laws (nomoi) and decrees (psephismata) – that is, between general norms with no time limit and individual norms with a clearly defined validity period (Hansen, 1991: 161–2). Whereas the former contained the basic rules of the polis, the latter dealt with citizenship grants and honorary decrees, issues of foreign policy, questions of cult and religion, and, to a very small extent, financial and legislative matters (Hansen, 1991: 156–8). Although Castoriadis (1991a: 116) is aware of this contrast between laws and decrees, he fails to mention that ‘the graphe paranomon could be used only against decrees’ (Hansen, 1991: 212); in order to challenge laws, one had to bring a graphe nomon me epitedeion theinai, a mechanism that was employed much less frequently (Hansen, 1991: 212).

Granted, one might brush aside our criticism of Castoriadis’s lack of nuance as excessive pedantry. But even then, the graphe paranomon is not as obvious a case of self-limitation as he claims. True, some scholars’ understanding of the graphe paranomon does indeed seem to be in agreement with, or at least not incompatible with, Castoriadis’s interpretation. Melissa Schwartzberg (2013: 1051), for instance, describes the graphe paranomon as a device that affirms ‘the ongoing popular capacity for reflective judgment’. Others, however, consider the graphe paranomon not so much a means of self-limitation as a precursor of what is now dubbed judicial review. The most prominent voice in this camp is the Danish classical philologist Mogens Herman Hansen (1991: 211), who calls the graphe paranomon a ‘bulwark of the constitution and the only sure defence of the laws’. According to Schwartzberg (2013: 1050), this is the ‘dominant explanation’. Moreover, Castoriadis fails to note that not every member of the demos could become part of the People’s Council judging a graphe paranomon; only those who were at least 30 years old were considered fit for the task. ⁶ This ‘thirty-years-of-age rule’, Hansen (1991: 181) argues, ‘was a quite considerable limitation on the recruitment of jurors, seeing that out of the 30,000 adult male citizens there would not be more than 20,000 eligible for jury service’. The rationale for this age limit is not clear. Hansen (1991: 181) suspects that the Athenians wanted to ‘put the power of judging into the hands of the eldest and most experienced’, thus making the People’s Court a court of experts; Schwartzberg (2013: 1052) rejects this argument. Finally, it should not go unmentioned that the graphe paranomon introduced ‘a new kind of corruption’ (Hansen, 1991: 208) into Athenian politics. Because controversial decrees carried the risk of severe sanctions, ⁷ some members of the demos tried to get someone else to put forward a proposal on their behalf, so that they could then support it in the Assembly without risking punishment (Hansen, 1991: 207–8). The upshot is that while it would be wrong to say that Castoriadis’s (1991a: 117) depiction of the graphe paranomon as an instance of ‘the demos [. . .] appealing against itself in front of itself’ is misguided, it would be equally wrong to ignore that it comes at the price of considerable simplification.

Attempts to institutionalize self-limitation are not confined to ancient Athens, however. They can also be found in modern times, with constitutions and the separation of powers being the most prominent examples, according to Castoriadis. With regard to the former, he tells us that he supports their existence for both pragmatic and pedagogical reasons (Castoriadis, 1997b: 409), noting, though, that ‘perfectly “democratic” constitutions have served to mask the bloodiest tyrannies and continue to do so’ (Castoriadis, 1997b: 408) and that ‘the oldest “democracy” in the liberal West, Britain, has no “Constitution” at all’ (Castoriadis, 1991a: 115). A constitution alone will not do to protect democracy (Castoriadis, 1997a: 316); to think otherwise is to subscribe to the ‘illusion of constitutionalism’ (Castoriadis, 1997c: 94). As to the separation of powers, Castoriadis (1997b: 409) concedes that the idea ‘has been much developed in theory’, but he also points out that in practice legislative power and executive power (and sometimes even judicial power) are often ‘in the hands of the same effective instance of authority: the majority party’. In addition to constitutions and the separation of powers, he cursorily mentions human rights, the rule of law, and the plea of illegality or unconstitutionality as further efforts to institutionalize self-limitation (Castoriadis, 1997b: 408).

To sum up, Castoriadis provides us with a diverse assortment of instances of self-limitation. In one way or another, they all seek to protect against abuses of power and to stave off the self-destruction of democracy. Other than that, however, they do not appear to have much in common. What is more, Castoriadis’s portrayal of them is often questionable. The puzzle remains: What is democratic self-limitation?

In the light of Castoriadis’s understanding of democracy, there seem to be two kinds of self-limitation. We have already noted that democracy is the regime where the people themselves function as legislators and where, by implication, there is ‘equality of the law for everyone’ (Castoriadis, 1993: 316). But we have not yet discussed what is meant by law in this context. Castoriadis (1993: 316) has two things in mind, ‘the “formal” law, which in modern societies is written, law in the narrow sense’ and ‘the institution of society’. The first meaning of law refers to the regulation of matters internal to the demos; the second, to who should belong to it. If this is so, self-limitation, too, can be exercised either within the demos or as concerns the question of membership. Our interest is primarily in properly theorizing the latter. This we try to do by looking at what we think should be interpreted as an example of self-limitation with regard to who should belong to the body politic: the case of women’s suffrage in Switzerland.

The Struggle for Women’s Suffrage in Switzerland

The history of women’s suffrage in Switzerland constitutes something of a ‘paradox’ (Wecker, 1998). While men were granted the right to vote as early as 1848, women had to wait until 1971 to gain suffrage. The question that most preoccupied both legal experts and the populace at large during the long run-up to the introduction of women’s right to vote was who should decide on their enfranchisement. Was it the male citizens, who would automatically undermine their own legitimacy as the legislative body if they voted in favour of women’s suffrage, or was it enough for the Federal Supreme Court to reinterpret the principle of equality enshrined in the Swiss Constitution in order to grant women the right to vote? Moreover, should the question of women’s suffrage be tackled at the federal level or by each canton separately? And what should be – or, looking at it from a 21st-century perspective, what was – the role of women in all this? Such were the challenges surrounding what the Swiss legal scholar Werner Kägi, in his expert report on the question of women’s enfranchisement, calls ‘the most important constitutional decision of our state since 1848’. In his opinion, the conflict within the Swiss Constitution between the article on the right to vote and the article on the equality of all Swiss citizens should be settled not by simply reinterpreting the constitution and introducing universal adult suffrage but by ‘a formal decision by the constitutional legislature’ – for both ‘legal and political reasons’ (Kägi, 1956: 54, our translation). His compatriots agreed. The consensus eventually arrived at was that whether women should be granted the right to vote should be left to the discretion of the male citizenry, who, it turned out, were anything but eager to enfranchise them. Again, it was not until 1971 that Swiss men came to support women’s suffrage, making Switzerland the odd one out among European states. That it took them so long is a testament to the insidiousness of patriarchy and should, of course, give us pause. And yet, as we see it, properly understood, it is also an outstanding example of self-limitation, of a demos curbing its own power in the name of democracy. To see why this is so, it is necessary to take a closer look at the events preceding the enfranchisement of Swiss women. More specifically, we need to introduce into our discussion their century-long struggle for the right to vote. Given the limited scope of this article, we confine ourselves to outlining its main stages. ⁸

From the 1860s, a growing number of women began to campaign for equal rights and women’s suffrage. This led to the establishment of the Swiss Union for Women’s Suffrage (Schweizer Verband für Frauenstimmrecht) and the Federation of Swiss Women’s Associations (Bund der Schweizerischen Frauenvereine) at the turn of the century, an important step towards more concerted action. As early as 1886, 139 women presented a petition to Parliament demanding the right to vote, arguing that the exclusion of women from elections constituted a violation of their rights as human beings. Seven years later, the Swiss Workers’ Union became the first major organization to call for women’s suffrage, which, together with the First Congress for Women’s Interests (Erster Kongress für Fraueninteressen) in 1896, brought Swiss women’s struggle for the vote to the attention of a wider public. In 1918, it cropped up among the demands of the Swiss general strike. That same year, two motions for women’s suffrage were brought before the National Council (Nationalrat). Once they were passed on to the Federal Council (Bundesrat), however, they were set aside in favour of more pressing concerns, or so the argument went. Around ten years later, the Swiss Union for Women’s Suffrage submitted to Parliament a petition for women’s right to vote signed by almost 250,000 people. This petition, too, fell on deaf ears, with the Federal Council blocking it for more than 20 years. In other European countries, women’s suffrage was introduced in the wake of significant historical changes – think of the transition from monarchy to democracy, the restructuring of the state after the two world wars, or anti-colonial struggles; in Switzerland, there were no similar revolutionary upheavals to fast-track the cause of women’s suffrage.

When in the midst of the Cold War the Federal Council started debating whether civil protection service should be made compulsory for women, the issue of women’s suffrage was put back on the agenda, leading to the first federal referendum on 1 February 1959. The result was dismaying: 66.9 per cent of voters rejected the introduction of the right to vote for women. At the cantonal level, though, things began to change. The canton of Vaud was the first to vote in favour of women’s suffrage, followed by Neuchâtel and Geneva.

In the years that followed, the struggle for the right to vote was helped along by geopolitical changes. In 1963, Switzerland joined the European Council. Five years later, in 1968, it signed the European Convention on Human Rights but decided to leave the question of women’s suffrage to the cantons. Predictably, this caused considerable outrage among many Swiss women, culminating in the 1969 March on Bern, with 5,000 women protesting in front of the Federal Palace (Bundeshaus). That same year, left-wing activists founded the Women’s Liberation Movement (Frauenbefreiungsbewegung), which adopted provocative, attention-grabbing methods to bring the oppression of women to the attention of the Swiss public. With the growing debate about the denial of women’s suffrage as a human rights violation, social-cultural changes, and more radical forms of protest, Switzerland’s image as a progressive democracy began to show some cracks. To address what the rapporteur of the National Council at the time called the country’s ‘anachronistic unicum’ (Studer, 2021: 153, our translation), Switzerland held a second nationwide referendum in 1971. For the first time, all the government parties, as well as the Trade Union and the Farmers’ Association, supported the cause of women’s suffrage. In the end, 65.7 per cent of the electorate opted to grant women the right to vote. ⁹

Democratic Self-Limitation as a Reciprocal Process

That women were not allowed to vote until 1971 is an indelible blemish on the record of Swiss democracy. And yet, because the question of women’s suffrage was decided by popular vote, many Swiss consider its belated introduction not so much a cause for shame as a testament to the country’s strong commitment to direct democracy. Indeed, the fact that women got the vote not thanks to a reinterpretation of the Swiss Constitution but because the male electorate decided that they should get it fits in well with Switzerland’s image as a paragon of democracy and is often used as an excuse for its curious delay in enfranchising women (Studer, 2021: 156–8). ¹⁰ In a certain way, this take seems to be very much in line with Castoriadis’s position. For he, too, is a champion of direct democracy, calling it the only ‘genuine democracy’ (Castoriadis, 1994: 144). Or as he puts it in one of his essays on socialism, ‘To decide means to decide for oneself. To decide who is to decide already is not quite deciding for oneself. The only total form of democracy is therefore direct democracy’ (Castoriadis, 1988: 98).

In the context of our discussion, the main question is the following. In what sense is the 1971 referendum in favour of Swiss women’s suffrage an example of self-limitation? At first glance, the answer seems obvious. By enfranchising women, men limited their own sovereignty and broadened the demos. On reflection, however, it is not quite so simple. To portray the 1971 referendum as a case of self-limitation simply because of how the male electorate voted is to risk rendering invisible Swiss women’s struggle for suffrage and to conceal Switzerland’s long history of exclusion. As we have shown in the previous section, Swiss men did not vote for women’s enfranchisement in a sudden fit of magnanimity; on the contrary, it was women’s tireless activism that, despite its many setbacks, paved the way for a new understanding of equality and pushed their male counterparts to introduce universal adult suffrage. Contrary to Castoriadis’s (1997d: 5) assertion that ‘the autonomy of the collectivity [. . .] can be achieved only through explicit self-institution and self-governance’, the driving force behind Switzerland’s self-limitation and thus extension of autonomy was not the given demos but those excluded from it – Swiss women.

Regrettably, this is not the story told in present-day Switzerland. According to Katrin Meyer (2022: 89), the majority of the Swiss population and most of the literature on the subject subscribe to what she calls the liberal narrative, which interprets the long-lasting withholding of the right to vote from women as a mere anachronism and considers their final enfranchisement nothing more than a quantitative change. Meyer (2022: 86) describes this position as ‘hegemonic silence’: ‘By silencing the topic of women’s rights, we are told a story about Swiss democracy in which women’s exclusion and inclusion have no weight.’ ¹¹

The point is this. To construe the 1971 referendum as self-limitation on the part of Swiss men, as a benevolent gesture, threatens to sweep under the carpet women’s century-long struggle for the right to vote. But what if, for the sake of argument, we tweak the Swiss example a little? Would it be easier for us to categorize the 1971 referendum as a case of self-limitation if both men and women had been able to vote? The reason we think it worth indulging this counterfactual is that Castoriadis’s notion of direct democracy seems to suggest as much. Writing of the ‘town meetings during the American Revolution, sections during the French Revolution, the Paris Commune, the Workers’ Councils, or the Soviets in their original form’ as instances of a rediscovery of direct democracy along the lines of the Athenian polis, he argues that ‘in all these cases, the sovereign body is the totality of those concerned’ (Castoriadis, 1991a: 107). If direct democracy, properly conceived, requires the inclusion of ‘the totality of those concerned’ in the decision-making process, then this means that Switzerland should have let both men and women decide whether to grant the latter the right to vote. ¹² And in the event of a positive outcome, this could be described as a case of genuine self-limitation.

However, two things do not add up in this fictional scenario. The first is that it does not take into account that every democracy presupposes the institution of a certain demos, whose exclusionary nature can be challenged but not ignored. The second is that the claim that a democratic body politic must welcome all those affected by its decisions in order for it to be truly democratic proves to be untenable. Let us consider these two quibbles in turn. In doing so, we will arrive at a new understanding of what democratic self-limitation is and why the 1971 referendum, properly conceived, can indeed be seen as an example of it.

It is one of the basic tenets of Castoriadis’s (1991b: 144) thought that every society is ‘always already instituted’: ‘The community of citizens – the demos – proclaims that it is absolutely sovereign [. . .]. This is the self-position, self-definition, of the political body, which contains an element of arbitrariness – and always will’ (Castoriadis, 1991a: 106). In other words, as a form of self-government, democracy cannot do without some conception of this very self, of who is part of the demos. This conception is the result of historical power relations and therefore a matter of contingency. Whatever it looks like, though, it will be accompanied by certain exclusions. Think, Castoriadis (1991a: 106) tells us, of the situation of women, slaves, and foreigners in ancient Greece or of chattel slavery in antebellum America. ¹³ What these examples point to is the general truth that there is no such thing as a fully inclusive demos. Every process of institution draws a line between those who belong and those who do not. This, however, is incompatible with Castoriadis’s (1993: 316, emphasis added) notion of direct democracy as a form of self-rule where there is ‘equality of the law for everyone’. ¹⁴ To put it differently, there is a tension between Castoriadis the thinker of institutionalization and Castoriadis the advocate of direct democracy. We side with the former. ¹⁵ Every decision, no matter its subject, is also a decision about who gets to decide and who does not. It follows that it is a blind alley to ponder whether the 1971 referendum could count as an instance of democratic self-limitation if Swiss women had been given a say in it. For it is predicated on the direct democratic fantasy of a demos that includes everyone.

As to the idea of welcoming into the demos whoever wants to become part of it, here, too, it is a matter of reading Castoriadis against Castoriadis. If democracy is the regime of self-limitation, it stands to reason that only those who are themselves committed to the idea of self-limitation should be allowed to join the demos. From this we can infer that self-limitation, at least when the subject in question is the make-up of the body politic, is based on reciprocity. Those who already belong to the demos have to limit their own sovereignty; those who wish to become part of it must limit themselves to demanding an equal say and resist the impulse to retaliate for their former exclusion. This is precisely how we should remember the 1971 referendum if we want to highlight, and keep alive, the democratizing potential of self-limitation. Swiss men acknowledged the contingency of the given demos and agreed to curb their own sovereignty; Swiss women confined themselves to campaigning for equality, rather than advocating the overthrow of men. Indeed, because women constituted the majority of the Swiss population at the time, some activists made their commitment to the idea of self-limitation public. Consider, for example, the following draft letter to members of the male electorate, printed in the women’s magazine annabelle shortly before the referendum.

Dear . . .,

[. . .] This time I would like to ask you [. . .] for a favour. You probably know that on 7 February 1971 there will be a vote on whether we women should have the right to vote. You can decide, I can’t. So I would like to ask you [. . .] to put a ‘yes’ in the ballot box. For me. I promise you [. . .] that I will never try to abolish men’s right to vote. (Haller, 2017, our translation, emphasis added)

Provided that we add the relative clause ‘on the condition that they, too, commit themselves to self-limitation’, the last sentence sums it up perfectly. Democratic self-limitation as regards the question of who should belong to the demos requires reciprocity. It must be exercised both by those who are part of it and by those who are about to be. The scenario suggested by Castoriadis’s understanding of direct democracy, on the other hand, is a dead end. The idea that anyone affected by the decision-making process should be able to participate in it ignores the instituted, and therefore inevitably exclusionary, character of the demos and overlooks that extending its membership risks turning sour if it is not accompanied by the condition that those wishing to join demand nothing more than an equal say. What made the 1971 referendum an instance of self-limitation was that both Swiss men and Swiss women, the existing demos and its aspirant members, were committed to it. And because it was women’s indefatigable activism that put the vote on their suffrage on the political agenda, rather than the goodwill of the male citizenry, they should be the focus of this remarkable, if embarrassingly belated, story of democratization.