1. Introduction
[T]he words “subject” and “sovereign” are identical correlatives, the meaning of which is brought together in the single word “citizen”. (Rousseau Reference Rousseau and Cranston[1762] 1968, 138).
Public reason, according to John Rawls, ‘is the form of reasoning appropriate to equal citizens who as a corporate body impose rules on one another backed by sanctions of state power’ (Rawls Reference Rawls2001, 92 [my emphasis]). In this paper I flesh out this conception of citizens’ shared political autonomy by formulating what I call the ‘civic people’ account of public reason. On the basis of a distinctly political conception of mutual respect that I call ‘civic respect’, citizens can constitute a kind of corporate moral agent – a self-governing civic people – despite their adherence to different religious, moral, and philosophical views. As members of a civic people, citizens decide fundamental political questions by means of shareable public reasons.
Since the conception of civic respect and the civic people account of corporate moral agency that I outline in this paper are advanced as contributions to the theory of political liberalism, Footnote 1 I summarize the relevant elements of that theory in Section 2. Drawing upon these elements, I outline in Section 3 the main features of civic respect. Civic respect is a form of mutual ‘recognition respect’ that should govern free and equal citizens’ shared exercise of political power within pluralist societies. I then show in Section 4 how a robust form of shared political autonomy can be realized within contemporary pluralist societies based upon citizens’ mutual civic respect. Drawing upon Michael Bratman’s theory of shared agency (Bratman Reference Bratman1999, Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, Reference Bratman2014), I formulate the idea of a civic people. This idea, I propose, vindicates Rawls’s claim that legitimate political power is ‘the power of free and equal citizens as a collective body’ (Rawls Reference Rawls2001, 40 [my emphasis]).Footnote 2 A civic people is constituted by citizens’ ‘shared policy’ to decide fundamental political questions by means of shareable public reasons. In Section 5 I address an objection to my account of a civic people, namely, that Bratman’s theory of shared agency cannot be applied to groups as large as political societies. I overcome this objection by modifying Bratman’s theory so that it can apply to certain kinds of large-scale groups.
In Section 6 I compare the civic people account of shared political autonomy to what I call the ‘constrained proceduralist account’. According to the constrained proceduralist account, citizens share a policy to decide fundamental political questions through fair democratic procedures, but such decisions are ‘constrained’ by the equal protection of the basic liberal democratic rights of all. Political decisions within the constrained proceduralist account need not be made via shareable public reasons; instead, citizens can justify their political decisions with reasons drawn exclusively from their respective comprehensive doctrines. I reject the constrained proceduralist account for two reasons. First, the constrained proceduralist account does not realize citizens’ political autonomy as fully as does the civic people account. Second, a civic people likely will enjoy greater stability over time than will a society characterized by the constrained proceduralist account.
I then consider in Section 7 the ‘convergence’ account of public justification advanced in recent years by Gerald Gaus and Kevin Vallier (Gaus and Vallier Reference Gaus and Vallier2009; Gaus Reference Gaus2010, Reference Gaus2011; Vallier Reference Vallier2014, Reference Vallier2015). According to this account, non-shareable (‘non-public’) reasons may be employed when justifying political decisions – so long as those reasons all ‘converge’ in supporting those decisions (in contrast, the constrained proceduralist account does not require such convergence). The convergence account of public justification, like the constrained proceduralist account, permits a greater range of reasons to be employed by citizens when deciding fundamental political matters than does the civic people account. Insofar as citizens are committed to realizing the ideal of shared political autonomy, though, they should reject the convergence account. This is because the kind of shared policy that can constitute a civic people is not possible within the convergence account of public justification. Citizens committed to the ideal of shared political autonomy, I conclude, should endorse the civic people account of public reason.
2. Political liberalism
According to the theory of political liberalism, liberal societies are characterized by what Rawls calls the ‘fact of reasonable pluralism’: the fact that persons living in such societies, as a consequence of the free exercise of their reason, invariably will subscribe to a variety of different, typically incompatible, philosophical, moral, and religious ‘comprehensive doctrines’ (Rawls Reference Rawls2005, 36–37, 440–441). In order to accommodate the fact of reasonable pluralism, Rawls holds that the main political and economic institutions of a liberal society should be governed by what he calls a ‘political conception of justice’. A political conception of justice satisfies (what I will refer to as) the ‘basic structure restriction’ and the ‘freestanding condition’. According to the basic structure restriction, a political conception of justice should apply only to the basic structure of society – its main political and economic institutions, taken together as an overall system of cooperation – and not to social, philosophical, or moral concerns that lie beyond this domain.Footnote 3 A political conception of justice satisfies the freestanding condition by being formulated in terms of distinctly ‘political’ ideas (including ideals, values, principles, and so forth). Such political ideas, like that of society as a ‘fair system of social cooperation’, do not presuppose or depend upon the truth of any particular comprehensive doctrine (for instance, Buddhism or utilitarianism). These ideas instead are construed as implicit within the public political culture of democratic society. Political conceptions of justice are compatible with the (‘reasonable’) comprehensive doctrines endorsed by that society’s citizens (Rawls Reference Rawls2005, 11–15, 374–376, 453).Footnote 4
The ‘intrinsic (moral) political ideal’ of political liberalism, according to Rawls, is the ‘criterion of reciprocity’ (Rawls Reference Rawls2005, xlv). This criterion requires that citizens offer terms of social cooperation, terms drawn from political conceptions of justice, that they think other citizens will find acceptable. The relevant constituency of citizens are those whom Rawls calls ‘reasonable persons’. Such persons acknowledge the fact of reasonable pluralism and are committed to satisfying the criterion of reciprocity when justifying fundamental political decisions (Rawls Reference Rawls2005, xliv, 16, 49–50, 54).Footnote 5 Any political conception of justice capable of satisfying the criterion of reciprocity is what Rawls calls a ‘reasonable’ conception.Footnote 6 A reasonable political conception of justice can secure the acceptance, if not the endorsement, Footnote 7 of all reasonable persons in a pluralist society, irrespective of their broader religious or philosophical convictions. And a reasonable political conception of justice, by being acceptable to all reasonable persons, can satisfy the ‘liberal principle of legitimacy’.Footnote 8
Finally, a society with a basic structure that is organized by a reasonable political conception of justice is capable of being (or becoming) a ‘well-ordered society’.Footnote 9 A well-ordered society, roughly, is a society with a basic structure that is supported freely over time by the reasonable persons within it. In a well-ordered society, reasonable persons maintain and participate in their society’s basic structure not because the state threatens them with coercive sanctions, but because they judge the political conception of justice that organizes their basic structure to be acceptable to them, despite their adherence to different comprehensive doctrines. The basic structure of a well-ordered society thus can be understood as a ‘fair system of social cooperation between free and equal persons’ (Rawls Reference Rawls2005, 9). It is a fair system of social cooperation because the terms of social cooperation, the principles of a political conception of justice, are acceptable to the reasonable persons subject to them.
3. Civic respect
According to Rawls, citizens in a well-ordered society cooperate with one another in maintaining their society’s basic structure on the basis of mutual respect (Rawls Reference Rawls2001, 28, 91). In this section I outline the main features of a conception of mutual respect – civic respect – appropriate for political liberalism.Footnote 10 It is a conception that can serve as the basis for social cooperation between free and equal citizens in their collective exercise of political power despite their endorsement of different comprehensive doctrines.Footnote 11
Building upon political liberalism’s idea of reasonable persons, the conception of civic respect is characterized by five features.Footnote 12 The first feature is that civic respect requires that citizens acknowledge the fact of reasonable pluralism. Acknowledging that this pluralism is ‘reasonable’ means, inter alia, that citizens understand and accept that decisions regarding the basic structure of their society – decisions that involve the exercise of public political authority, and hence, when necessary, the power of the stateFootnote 13 – should satisfy the freestanding condition.Footnote 14
Second, civic respect is a kind of respect that Stephen Darwall calls ‘recognition respect’ (Darwall Reference Darwall2006). Recognition respect is a form of respect that involves acknowledging and according to persons a certain status and authority in one’s deliberations and actions towards them, or in one’s deliberations and actions that concern them in some normatively significant way, in virtue of those persons’ capacities to be reasonable.Footnote 15 In other words, recognition respect involves respecting others as having a kind of ‘dignity’, and hence possessing an authority to hold others accountable for their actions and decisions with respect to them. Such recognition respect, moreover, is owed equally to all agents who can occupy the relevant second-person standpoint (roughly, the standpoint of mutual accountability).Footnote 16 Civic respect is a form of recognition respect in that it is owed to persons in virtue of their standing and authority as free and equal citizens. One practices such respect by taking this standing and authority into account when deciding fundamental political questions in concert with one’s fellow citizens. Equal ‘civic’ recognition respect is owed to all adequately reasonable persons in one’s political society.
Third, because civic respect is owed to persons qua citizens, it is limited in its scope of application to persons’ relations within the basic structure of society. According to political liberalism, roughly, persons’ lives qua citizens are delimited by their participation within their society’s basic structure.Footnote 17 Civic respect, then, can be distinguished from other, more ‘comprehensive’ forms of recognition respect, including those that apply to all aspects of persons’ lives, such as that required by Kant’s ‘Formula of Humanity’ (Kant Reference Kant and Gregor1998, 36–43).Footnote 18
Fourth, civic respect requires that citizens recognize the criterion of reciprocity as a prescriptive norm for governing their public political relations with each other, as manifested in the various institutions that make up the basic structure. (The criterion of reciprocity, recall, holds that citizens must offer terms of social cooperation, and in particular political principles of justice, that they think other citizens will find acceptable.)
The fifth feature of civic respect – which I suggest in the next section should be understood as a ‘shared policy’ among reasonable citizens based upon the first four features of civic respect – is that citizens are committed to the idea of public reason. Public reasoning, according to political liberalism, is the shared form of reasoning that the citizens of a pluralist democratic society should use when deciding ‘constitutional essentials’ and ‘matters of basic justice’ (Rawls Reference Rawls2005, 214, 227–230). Public reasoning makes the realization of the ideal of fair social cooperation amongst free and equal citizens possible in pluralist societies; accordingly, Rawls holds that it should be understood as ‘part of the idea of democracy itself’ (Rawls Reference Rawls2005, 441). By employing public reasons when deciding fundamental political questions, citizens relate to one another as equal co-sovereigns and ensure the legitimacy of their shared exercise of political power.
Public reasons are freestanding with respect to particular comprehensive doctrines as they are drawn from the reasonable political conception of justice (or family of conceptions) endorsed by citizens (Rawls Reference Rawls2005, 450f), Footnote 19 as well as more general civic virtues (such as transparency and toleration) and modes of inquiry (such as rules concerning evidence, logic, and the like). By deciding fundamental political questions with public reasons, citizens express their (recognition) respect for each other as equal members of political society, despite their adherence to different comprehensive doctrines. Citizens thereby fulfill what Rawls calls their ‘duty of civility’ (Rawls Reference Rawls2005, 444–445).Footnote 20
To summarize, civic respect:
Requires that citizens acknowledge the fact of reasonable pluralism;
Is a form of recognition respect;
Is limited in its scope to citizens’ public political relations with one another, that is, their relations within the basic structure;
Requires that citizens comply with the criterion of reciprocity;
And on the basis of (1)–(4):
Requires that citizens employ public reasons when deciding fundamental political questions.
When citizens are committed to interacting with one another on the basis of civic respect, it is possible for them to enjoy and exercise what Rawls calls ‘full political autonomy’. There are two elements to citizens’ full political autonomy. The first is an ‘institutional’ element: to be politically autonomous, citizens must enjoy the rights and resources that enable them to take part as (roughly) equal contributors to their society’s main political decision-making processes.Footnote 21 The second element of full political autonomy is ‘justificatory’ in nature: citizens are politically autonomous when fundamental political decisions – decisions to which they are subject – are made via reasons that they find acceptable.Footnote 22
Citizens exercise political autonomy, institutionally, ‘by participating in society’s public affairs and sharing in its collective self-determination over time’ (Rawls Reference Rawls2005, 78). The conception of civic respect helps explain how institutionally autonomous citizens can share in society’s collective self-determination over time in a way that enables all citizens to enjoy justificatory political autonomy as well – namely, by being parties to a shared policy to decide fundamental political questions via public reasons and by participating in the formation of such decisions in accordance with this policy. Consequently, my aim in the next section is to explain why a shared policy to decide fundamental political questions by means of public reasons, civic respect’s fifth feature, would be adopted on the basis of the first four features of civic respect, and how citizens thereby would constitute a corporate moral agent, a self-governing civic people.
4. A civic peopleFootnote 23
Recall that ‘public reason is the form of reasoning appropriate to equal citizens who as a corporate body impose rules on one another by sanctions of state power’ (Rawls Reference Rawls2001, 92 [my emphasis]). In this section I explain that citizens attempting to realize the ideal of shared political autonomy in a pluralist society would commit themselves to a ‘shared policy’ to give weight only to public reasons in their decisions concerning fundamental political matters.Footnote 24 Such citizens, through this shared policy, would constitute a ‘corporate body’, what I call a civic people.
Michael Bratman’s account of shared intentions extends his planning theory of intentions from individual agents to groups (Bratman Reference Bratman1999, Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, Reference Bratman2014).Footnote 25 According to Bratman, when individuals do something together as a group they have a ‘shared intention’. By a shared intention Bratman does not mean some kind of ‘super-intention’ of a ‘collective mind’. His account of shared intentions and group agency is thoroughly reductionist in nature. An intention is shared, roughly, when each member of a group has an intention of the following form: I intend that we do x, where my ‘intention to do x’ is (1) contingent on my belief that the others among us also intend to do x, (2) these intentions to do x are mutually responsive to each others’ intentions to do x, and (3) we intend that our shared activity of x-ing go by way of our respective intentions and ‘meshing sub-plans’. The matching individual intentions, together with the members’ common knowledge that they hold matching intentions, their mutual responsiveness with respect to them, and their desire to realize the shared activity in question via their respective intentions and meshing sub-plans, constitute the group’s shared intention to do x.
To illustrate, three people – Gordon, Jean, and Jo-Anne – might decide to host a party together. Their individual intentions to host the party are dependent on, and responsive to, the corresponding intentions of the others. Hence if Gordon determines that he no longer can take part, Jean and Jo-Anne would have to revise or abandon their intentions with respect to the party. Gordon, Jean, and Jo-Anne all intend that their shared activity of ‘hosting the party’ go by way of their respective intentions to host the party together. Finally, they intend that the various sub-plans that they adopt to realize this common goal are mutually compatible or ‘mesh’. This does not require (at least not normally) that each participant individually be aware of all of the details of the others’ relevant sub-plans, only that those sub-plans be compatible. For instance, Gordon may be in charge of invitations, Jean preparing the food, and Jo-Anne the wine and music, without each necessarily knowing the specifics of the others’ relevant sub-plans (Jean need not know the specifics of Jo-Anne’s wine purchases and music selection, and so forth). Of course, their shared intention to host a certain kind of party, among other considerations, may place certain constraints on each person’s relevant sub-plans. The party is a product of shared agency because the planning and activity of each participant aims at the same activity, and is mutually responsive to, and dependent on, the planning and activities of the others.
Bratman also extends his account of individual self-governing policies to groups. Self-governing policies are special kinds of intentions: they are intentions to assign certain kinds of justifying significance or ‘weights’ to certain considerations or reasons in one’s deliberations regarding particular subjects, under appropriate circumstances. A self-governing policy can reflect, among other things, an individual’s moral commitments or value judgements. For instance, an individual may adopt a policy of giving considerations of revenge ‘no weight’ in her reasoning about how to interact with other persons (despite, perhaps, her frequent desire to do so).
Like an individual’s self-governing policy, a shared self-governing policy is a special kind of intention. Bratman explains: ‘S1 and S2 have a shared policy concerning the justifying significance or weight to be given to consideration C in relevant contexts of shared deliberation’ (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 4). Moreover, insofar as self-governing policies for individuals sometimes can be understood as a kind of valuing, so too shared policies for groups sometimes can be understood as a kind of shared valuing (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 15). So, for example, a group’s shared policy to give considerations of revenge ‘no weight’ in its deliberations concerning how to interact with other groups can be understood as a form of ‘shared valuing’ among the members of that group.
In ‘core cases’ of self-governing shared policies, according to Bratman, the following two conditions obtain, and it is common knowledge that they obtain, among the relevant parties:
We each intend
that we give weight to R [a particular consideration or reason] in relevant shared deliberation, and
that (i) proceed by way of each of our (a) (i) intentions and their meshing sub-plans.
There is mutual interdependence between each of our (a) intentions. (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 21–22.)
Shared policies, then, ‘constitute a commitment, on the part of the individuals in the group, to structuring their shared deliberation and planning in a certain way’ (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 19; see also Reference Bratman2014, ch. 7). For example, a hiring committee at a university department may adopt as a policy to govern its deliberations about whom to hire the policy that the ‘personal affability’ of the various job candidates should not be a factor in their deliberations. That is, the department in question is committed to giving personal affability ‘no weight’ in their collective evaluations of the job candidates’ applications.
Finally, Bratman suggests that ‘particular shared policies concerning what to treat as a reason in certain contexts of shared activity are more or less definitive of the group whose policies they are’ (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 19; see also Reference Bratman2014, ch. 7). To return to the departmental hiring committee example, the various policies that the committee adopts to govern its deliberations about whom to hire will reflect the overall purpose of the group, as it is a group with a specific mandate, and one that will adopt policies pursuant to that mandate, namely, policies about what weights to assign to different hiring criteria (such as candidates’ teaching records, research, and so forth).
A commitment on the part of reasonable citizens to employ public reasons when deciding fundamental political matters, I propose, can be understood as an instance of what Bratman calls a shared policy. This shared policy defines the group in question as a self-governing civic people. Thus Bratman's account of shared agency can render plausible Rawls’s references to free and equal citizens as making up a collective or corporate body, that is, it can show why thinking of citizens as constituting a collective or corporate body need not commit us to anything metaphysically or normatively controversial.Footnote 26
Turning back to the conception of civic respect advanced in Section 3, recognition of the fact of reasonable pluralism (the first feature of civic respect) means that citizens are aware that only certain kinds of reasons in public political deliberation are going to be mutually acceptable. However, the deliberation in question is concerned only with matters that have to do with the basic structure (the third feature of civic respect), specifically constitutional essentials and matters of basic justice, and thus has a limited scope. Furthermore, the equal recognition respect that citizens have for each other (the second feature of civic respect), and the commitment to the criterion of reciprocity that they have in their relations and deliberations with each other (the fourth feature of civic respect), mean that citizens recognize that they all have an equal standing in the determination of fundamental political matters. Together, these features of civic respect give citizens sufficient reason to commit themselves to a shared policy to give weight only to public reasons when deciding fundamental political questions. In other words, those citizens who would agree to conduct their public political relations with each other on the basis of the first four features of civic respect also would agree to form or join a shared policy to decide fundamental political matters by means of public reasons.Footnote 27
This shared policy is compatible with accommodating the fact of reasonable pluralism, as citizens may justify their commitment to it to themselves in terms of their respective comprehensive doctrines. Bratman notes that while a ‘shared policy […] will normally be to some extent responsive to the relevant judgments of value on the part of the various participants [, ] […] [n]evertheless, the shared policy […] requires neither agreement nor disagreement in underlying value judgments’ (Bratman Reference Bratman, Wallace, Pettit, Scheffler and Smith2004, 23). Likewise, the fact that the ideas of public reason – such as the conception of citizens as free and equal, and the idea of society as a fair system of social cooperation – can be supported by, or at least rendered compatible with, the various beliefs and values of the comprehensive doctrines endorsed by different citizens does not mean that citizens will agree on the underlying grounds or justifications for the political ideas in question.Footnote 28
By being parties to a shared policy to decide fundamental political questions via public reasons, and thereby constituting a civic people, citizens can enjoy a form of shared political autonomy. Insofar as citizens wish to exercise their political autonomy and take part in their society’s collective self-determination over time, then, they will take themselves to owe it to other citizens to offer justifications for their positions concerning fundamental political matters in terms of public reasons. Citizens, that is, will take themselves to be subject to the duty of civility.Footnote 29
5. On the size of a civic people
An objection may be raised here regarding the size of the group agent outlined in the previous section. Although he does not rule out the possibility of extending his account of shared agency to large groups, and notes some recent work by others along these lines, Footnote 30 Bratman’s discussions of shared intentions and policies focus on small groups, such as a few people painting a house together.Footnote 31 Perhaps, then, the idea of a civic people stretches Bratman’s account of shared policies too far?Footnote 32 After all, a central feature of Bratman's account of shared policies is the interdependent nature of the participants’ intentions. Within even a well-ordered society, though, it is implausible to think that all citizens’ intentions could be directly interdependent in the way apparently required by Bratman’s account.
I think that this worry can be overcome.Footnote 33 To see how, it is important to appreciate that the responsibility to decide fundamental political questions falls most directly upon public officials within what Rawls calls the ‘public political forum’. This forum is where fundamental political issues are debated and authoritative decisions regarding them are made. It consists of three parts: ‘the discourse of judges in their decisions, especially of the judges of a supreme court; the discourse of government officials, especially chief executives and legislators; and finally, the discourse of candidates for public office and their campaign managers’ (Rawls Reference Rawls2005, 443). Such public officials help determine (or, in the case of candidates, aspire to help determine) the ways in which political power is exercised, through their shaping, implementation, and interpretation of laws. Knowledge of whether such officials comply with the shared policy to decide fundamental political questions via public reasons is, or would be within an adequately well-ordered society, generally accessible – that is, it is (or generally can become) public knowledge. If a public official violates her duty of civility, say, by deciding a matter of basic justice on grounds that presuppose the truth of her comprehensive doctrine, and for which there is no sufficient public reason justification, Footnote 34 this fact readily can become public knowledge, given the open democratic procedures and free press of an adequately well-ordered liberal society. Thus whether public officials adhere to the shared policy to decide fundamental political questions by means of public reasons is (adequately) knowable, both by other public officials and citizens in general.Footnote 35
But what about the justifications for political decisions made by citizens who are not public officials? How can citizens in general be parties to the shared policy in question, and thus be members of the civic people? Citizens, Rawls claims, fulfill their duty of civility by holding public officials to the idea of public reason when evaluating their performance within the public political forum, especially (though not exclusively) when voting (Rawls Reference Rawls2005, 444–445). Of course, knowledge about the specific voting intentions of other citizens – in particular, whether those intentions involve or incorporate the shared policy in question – is not available to citizens. In the small-scale endeavours discussed by Bratman, like a few people painting a house together, the shared intentions of the relevant agents are responsive to each other and thus directly interdependent. In contrast, within a political society it would be absurd to think that citizens could be knowledgeable and responsive to all other citizens’ intentions in a similar way.
Nonetheless, I propose that citizens within a well-ordered society can be construed as parties to the shared policy that constitutes the civic people if they know (or have good grounds for believing) that an adequate number of their fellow citizens also are reasonable persons, share the policy in question, and thus reliably act in compliance with that policy. Knowledge of other citizens ongoing compliance with the shared policy would be available through the outcomes of elections – say, in light of whether citizens in sufficient numbers vote against public officials who violate the duty of civility – as well as within the background culture and public political discourse of an adequately well-ordered society.Footnote 36 So long as this threshold of participation in the shared policy is satisfied, I propose that it is plausible to think of the reasonable citizens of a well-ordered society as parties to the shared policy to decide fundamental political questions via public reasons, and thus sufficient for the realization and persistence over time of a civic people.
6. Shared political autonomy and the constrained proceduralist account
One might argue that the ideal of shared political autonomy by free and equal citizens can be realized adequately through a less demanding shared policy, say, one to abide by what Simon May calls the ‘democratic principle of legitimacy’ (May Reference May2009).Footnote 37 According to this principle, ‘political power must be exercised within a constitutional order that respects the equal status of all citizens and that effectively guarantees each citizen the basic liberal rights and entitlements necessary to participate in political processes on equal terms’ (May Reference May2009, 136). I refer to this alternative as the ‘constrained proceduralist account’. According to this account, citizens share a policy to: (a) decide fundamental political questions via fair democratic procedures, and (b) ensure that such procedures and decisions respect the equal liberal democratic rights of all citizens. (Commitment (b) ‘constrains’ commitment (a).) On the constrained proceduralist account, then, citizens do not commit themselves to a shared policy to decide fundamental political questions by means of public reasons. In committing themselves to a shared policy to comply with the democratic principle of legitimacy, citizens constitute a collective agent, but a different kind of agent from a civic people, given the different content of the shared policy in question.Footnote 38
I think that the civic people account has (at least) two advantages over the constrained proceduralist account of shared political autonomy.
The first advantage is that the civic people account realizes more fully citizens’ political autonomy than does the constrained proceduralist account. Recall (from Section 3) that the idea of full political autonomy has two elements: an institutional element (concerning citizens’ rights and resources vis-à-vis their society’s main political decision-making institutions) and a justificatory element (concerning the reasons underpinning their society’s main political institutions and laws). The constrained proceduralist account may realize the institutional element of citizens’ political autonomy. Citizens are fully politically autonomous, though, only when the decisions made by their society concerning fundamental political questions are made for reasons that they themselves endorse or at least judge to be acceptable.Footnote 39
Fundamental political questions made by a civic people are made with reasons that all reasonable citizens affirm as legitimate. According to the constrained proceduralist account, in contrast, decisions concerning fundamental political questions can be made on grounds that significant numbers of citizens reject, that is, grounds that are incompatible with (at least) some of their deepest beliefs and values.Footnote 40 For instance, on the constrained proceduralist view, a political society could legitimately pass legislation prohibiting abortion under any circumstances, or, alternatively, requiring abortion under certain circumstances, and justify this legislation exclusively on religious or utilitarian grounds. Such a society could do this via democratic procedures despite the presence of substantial minorities who reject the religious or utilitarian doctrine in question. In contrast, such a law could not be justified on the civic people account, as it fails to rely upon public reasons.Footnote 41 So while all citizens within a society governed by the constrained proceduralist account of legitimate decision-making might enjoy institutional political autonomy, they all cannot enjoy full political autonomy.Footnote 42
The second advantage of the civic people account over the constrained proceduralist account has to do with stability. A society that decides fundamental political questions by means of shareable public reasons is more likely to be capable of securing the free support of its citizens for those decisions over time. If fundamental political questions are decided on the basis of reasons that all citizens affirm as legitimate, even if they do not always agree with the particular decisions themselves, then citizens will have reason to support freely those decisions. Citizens, that is, will understand that the political decisions in question are made for reasons that they themselves judge acceptable, ones that do not deny the truth of (any core element of) their comprehensive doctrines. Such a society likely will be, as Rawls puts it, ‘stable for the right reasons’ (Rawls Reference Rawls2001, 202; Reference Rawls2005, 390, 459).
In contrast, a society that permits fundamental political questions to be decided on the basis of reasons that not all citizens accept, reasons that presuppose the falseness of (at least) some core elements of the comprehensive doctrines endorsed by (at least) some reasonable citizens, risks alienating those citizens. Adherents of minority comprehensive doctrines in particular (for instance, Muslims or atheists in a predominantly Christian society) may come to feel that they regularly ‘lose out’ in the political decision-making processes of their society if it decides political questions in compliance with the democratic principle of legitimacy (even if their basic rights remain secure). Consequently, such citizens may become alienated from their society’s basic structure, and find it increasingly difficult to support freely its institutions and laws.Footnote 43 Such a society would less likely be ‘stable for the right reasons’ over time.Footnote 44
7. Shared political autonomy and convergence public justification
In recent years, Gerald Gaus and Kevin Vallier have advanced a ‘convergence’ account of public justification as an alternative to Rawlsian public reason (see: Gaus Reference Gaus2010, Reference Gaus2011; Gaus and Vallier Reference Gaus and Vallier2009; Vallier Reference Vallier2014, Reference Vallier2015). The convergence account (like the civic people and constrained proceduralist accounts) is an account of how to decide political questions fairly within pluralist societies. Central to the convergence account is what Gaus and Vallier call the ‘Public Justification Principle’ (hereinafter ‘PJP’). The PJP states: ‘L is a justified coercive law only if each and every member of the public P has conclusive reason(s) R to accept L as a requirement’ (Gaus and Vallier, 53). According to the PJP, political decisions need not be made via shareable public reasons. Rather, so long all reasonable citizens, whom Gaus and Vallier call ‘members of the public’ (Gaus and Vallier, 53–54), Footnote 45 have their own sufficient (public or non-public) reasons that ‘converge’ in supporting the political decisions in question, those decisions are legitimate. Political decisions thus can be justified by a range of incompatible reasons drawn from the disparate comprehensive doctrines of the members of the public.Footnote 46
Could citizens comprise a kind of self-governing body by committing themselves to a shared policy to satisfy the PJP when making political decisions? Such a ‘convergence’ conception of self-government would allow citizens to appeal to reasons that presuppose the truth of their respective comprehensive doctrines when participating in political decision-making. It thus would avoid the restriction to shareable public reasons that is part of the civic people account. According to the proposed self-governing shared policy:
All (reasonable) citizens would intend to support or find acceptable only those political decisions that satisfy the PJP (that is, political decisions that can be supported with adequate reasons drawn from all citizens’ systems of beliefs and values, including their various comprehensive doctrines); and
There would be mutual interdependence amongst citizens with respect to (a).
Condition (b) means that the intentions of citizens to satisfy the PJP (condition (a)) when making political decisions is contingent upon (an adequate number of) other citizens’ intentions to satisfy PJP (condition (a)) over time.
Such a shared policy is not feasible. By requiring that legitimate laws be justifiable to all reasonable citizens in terms of their overall systems of beliefs and values, including their respective comprehensive doctrines, the convergence conception imposes considerable informational burdens on citizens and their political representatives. It invariably will be quite difficult, if not impossible, for most citizens to ascertain whether there exist justifications for any proposed law with respect to all other reasonable citizens’ views.Footnote 47 Because of this, it seems unfeasible that such a shared policy could be endorsed and maintained over time by citizens.Footnote 48
But perhaps we could construe the shared policy in question as not requiring of citizens that they themselves ensure that the PJP be satisfied when engaging in political decision-making? Instead, citizens need only commit themselves to a policy of granting to one another ‘veto power’ over any proposed law. Such ‘vetoes’ could be based upon reasons drawn from citizens’ respective comprehensive doctrines.Footnote 49 While such a policy would be somewhat less epistemically demanding than the one under consideration, it would still be unfeasible. If citizens lack knowledge of one another’s comprehensive doctrines, there is no way for them to know whether objections to proposed laws are normatively justified or not – that is, whether citizens’ vetoes are based on reasons drawn from the values and beliefs of sincerely held religious and philosophical views – rather than strategic considerations of self-interest, motives of domination, or other considerations that violate reciprocity and equal respect.Footnote 50
In contrast, because the civic people account requires that shareable public reasons be used when deciding fundamental political questions, it is comparatively easy for citizens to evaluate directly those reasons.Footnote 51 Public reasons are inherently transparent; reasons drawn from comprehensive doctrines, however, often can be quite opaque, at least to citizens who do not endorse those doctrines. The informational demands imposed on citizens within the civic people account of democratic self-government are comparatively modest. Citizens merely need to be able to determine whether public officials (or themselves in certain cases, such as when participating in referenda) decide fundamental political questions on the basis of shareable public reasons. Knowledge of other citizens’ comprehensive doctrines is not required by political liberalism’s duty of civility.
Advocates of the convergence account of public justification acknowledge the informational burdens of their view; consequently, they reject such political ideas as ‘deliberative democracy’ and shared political autonomy (Gaus and Vallier Reference Gaus and Vallier2009, 65–70; see also Vallier Reference Vallier2015). Vallier, for instance, holds that most citizens have no duty to try to ensure that their society’s laws satisfy the PJP, even when engaging directly in their society’s political life, such as when voting (Vallier Reference Vallier2014, ch. 6). He writes: ‘My view allows citizens to act on whatever reasons they like’ (Vallier Reference Vallier2014, 190). In contrast, legislators are subject to a duty related to the PJP, what Vallier calls the ‘Principle of Convergence Restraint for Legislators’ (PCRL):
A legislator should not vote for law L in order to contribute to M’s becoming or remaining law (where L may be equivalent to M) if he justifiably believes that members of the public lack sufficient reason R n to endorse M. (Vallier Reference Vallier2014, 191; see also Vallier Reference Vallier2015, 154.)
Even with respect to legislators, though, Vallier concedes that the informational demands of this principle are quite strong, and so specifies that the PCRL ‘require[s] that legislators be sensitive to information about public justification when they encounter it’ (Vallier Reference Vallier2015, 155 [my italics]). Hence it is likely that laws will violate the PJP regularly, given how onerous it is for legislators to ensure that proposed laws satisfy the PJP. Since the civic people account relies upon shareable public reasons, though, it is not vulnerable to this problem: the legitimacy of all laws concerning fundamental political issues can be ensured by the public officials and citizens of a civic people.
Perhaps I am overstating the likelihood that the PJP would be violated regularly in a society governed by the convergence account of public justification? In such a society – one that has the kinds of institutions and practices envisioned by Gaus and Vallier, as well as the kind of public political culture that likely would emerge over time – legislators (and other public officials) might acquire the skills and knowledge necessary to determine whether political proposals satisfy the PJP. One reason this may be so is that citizens would be motivated to communicate (what they take to be) ‘defeaters’ drawn from their respective comprehensive doctrines concerning political proposals to their representatives. I concede that this may be the case.Footnote 52 But while this possibility might improve the likelihood that laws will satisfy the PJP in a society governed by the convergence view, the more fundamental difference between it and the civic people account with respect to citizens’ shared political autonomy remains, specifically, citizens themselves cannot act as co-sovereigns by forming a shared policy to ensure that their society’s laws comply with the PJP.
Rather than enabling citizens to deliberate politically together and thereby act as co-sovereigns, then, I think that the PJP most plausibly is interpreted as a criterion for laws to be legitimate with respect to citizens as subjects. Instead of an ideal of shared political autonomy, that is, convergence theorists like Gaus and Vallier advocate relying upon institutional arrangements, such as properly designed constitutional structures, and legislators’ compliance with the PCRL, to ensure that the PJP is satisfied by a society’s laws adequately over time. This strategy, whatever its merits, is an alternative to the ideal of citizens’ shared sovereignty, rather than an alternative version of it.Footnote 53 The convergence account of public justification does not constitute a rival way to realize the ideal of a self-governing people at all.
Now Gaus and Vallier likely would object to this characterization of their view. This is because, drawing upon their reading of Kant’s moral philosophy, they hold that if the PJP is satisfied, then each citizen ‘is both subject and legislator: each is subject to the law, yet each legislates the law, and so all are free and equal under the law’ (Gaus and Vallier Reference Gaus and Vallier2009, 52).Footnote 54 The sense in which citizens ‘legislate’ here, though, is quite passive: the responsibility for ensuring that the PJP is satisfied, as we have seen, does not require anything of most citizens. So the version of ‘self-legislation’ that Gaus and Vallier rely upon differs from the Rawlsian conception of shared political autonomy with which I am concerned in this paper; the latter involves the collective determination by citizens of the laws to which they are subject.Footnote 55
A convergence theorist might object here that the Rawlsian conception of full political autonomy is unfeasible, and so the version of autonomy that Gaus and Vallier outline is the best realizable one. This is because part of Rawls’s conception of full political autonomy involves citizens living within a well-ordered society; such a society is ‘well ordered’ because its basic structure complies with a political conception of justice that citizens themselves all accept (Rawls Reference Rawls2005, 77–78). But in Rawls’s final writings on political liberalism he acknowledges the existence of ‘a family of reasonable political conceptions’ of justice (Rawls Reference Rawls2005, 450). This suggests that the idea of a well-ordered society, a society in which (inter alia) all reasonable citizens endorse the same conception of justice, is no longer plausible. Consequently, Gaus argues that Rawlsian political liberalism must give up on the idea of a well-ordered society (Gaus Reference Gaus2016, esp. 150–154). And because of this, full political autonomy no longer seems achievable for all reasonable citizens, as there invariably will be some reasonable citizens who do not endorse the conception of justice that is realized in their basic structure.Footnote 56
Against this conclusion, I propose that a society can be considered ‘well ordered’ if its basic structure is organized in compliance with (at least) a reasonable political conception of justice, that is, a conception that all reasonable citizens find acceptable because it satisfies the criterion of reciprocity, the freestanding condition, and the basic structure restriction.Footnote 57 And as Paul Weithman (Reference Weithman2017) holds, so long as citizens’ basic structure is organized in compliance with a reasonable political conception of justice, then those citizens can enjoy full political autonomy. This is because: ‘The fundamental terms of citizens’ association are those they would give themselves on the basis of their own freedom and equality’ (Weithman Reference Weithman2017, 102 [my emphasis]).Footnote 58 In other words, if citizens’ basic structure is organized via a conception of justice that they all find acceptable, then their full political autonomy is realizable, even if they do not all endorse that conception.Footnote 59
Let me now contrast directly the civic people and the convergence views of political association. A civic people is a corporate moral agent: the citizens that comprise it do so through their shared policy to decide fundamental political questions by means of shareable public reasons. This shared policy enables citizens within a pluralist society to realize a robust form of shared political autonomy. Consequently, citizens can be characterized as co-sovereigns. In contrast, because determining whether the PJP is satisfied when advancing a political proposal involves considerable informational demands, citizens cannot form a shared policy to ensure that political proposals satisfy the PJP over time. The PJP (or a duty derived from it) cannot serve as the basis for shared political sovereignty. Instead, the PJP is construed most plausibly as a criterion that must be satisfied in order for laws to be formed and imposed legitimately upon citizens qua subjects. Citizens have no duty of civility to ensure that the laws to which they are subject satisfy the PJP; they need not exercise political sovereignty over themselves. Rather, the PJP is to be satisfied through proper institutional design and the decision-making behaviour of key elites (public officials like legislators) within those institutions.Footnote 60
Convergence theorists such as Gaus and Vallier readily accept that it is impossible for citizens to realize a robust ideal of shared political autonomy within societies governed by their account of public justification. The fact that the PJP is incompatible with this ideal thus is unlikely to trouble them. But this just shows that the convergence account of public justification cannot realize the political ideal that underpins the idea of public reason. Consequently, citizens attracted to that political ideal – the ideal of citizens as co-sovereigns, according to which political power is ultimately their power, the ‘power of free and equal citizens as a collective body’ – have reason to reject the convergence account. Gaus criticizes as insufficiently motivated Rawls’s duty of civility and its requirement that citizens use shareable public reasons when deciding fundamental political questions (Gaus Reference Gaus and Wall2015). Yet because complying with the duty of civility enables citizens to realize the ideal of shared political autonomy under conditions of reasonable pluralism, it follows that Gaus’s criticism misfires. Rawls’s duty of civility is well motivated, namely, by the ideal of shared political autonomy.
8. Conclusion
In this paper I outlined a distinctly political liberal conception of mutual respect – civic respect – and suggested that this conception can provide the basis for social cooperation among citizens in contemporary pluralist liberal societies (Section 3). Building upon this conception of civic respect, I employed a modified version of Bratman’s account of shared agency to show that a form of shared political autonomy, what I call the civic people account, can be realized in contemporary pluralist liberal societies (Sections 4–5). According to the civic people account of self-government, citizens in a liberal well-ordered society can be understood as parties to a shared policy to decide fundamental political questions by means of shareable public reasons. Such citizens are co-sovereigns with respect to the laws of their society. In collectively committing themselves to this shared policy, citizens constitute a kind of corporate moral agent, a civic people.
A shared policy among citizens to decide fundamental political questions in compliance with the democratic principle of legitimacy also can create a corporate moral agent (Section 6). But such a corporate moral agent, one that realizes the constrained proceduralist conception of shared political autonomy, is not as successful as a civic people in realizing citizens’ full political autonomy or in securing stability for the right reasons. Turning to the convergence account of public justification, a shared policy based upon (or derived from) the PJP is not possible (Section 7). So the convergence account of public justification cannot realize the ideal of shared political autonomy. Convergence theorists accept this feature of their view, despite the consequence that it renders a political ideal to which many reasonable citizens are (or may become) committed unrealizable. Citizens who endorse the ideal of shared political autonomy thus should reject the convergence account of public justification and endorse the civic people account.
Acknowledgements
Early versions of this paper (covering, roughly, Sections 2–4) were presented at: the Shared Agency Workshop at Stanford University (2004); the Canadian Philosophical Association annual conference (2006); the Political Studies Association of Ireland annual conference (2006); the Department of Philosophy Colloquium at University College Dublin (2006); and the Royal Irish Academy’s Committee for Philosophy panel on ‘Conflict, Dialogue, and Philosophy’ (2006). I’m grateful to the audiences at those presentations for their helpful questions and comments. Those versions also benefited from feedback from: Elizabeth Anderson, John Callanan, Stephen Darwall, Jürgen De Wispelaere, Louis-Philippe Hodgson, Evan Kirchhoff, Mika Lavaque-Manty, Scott Shapiro, Yonatan Shemmer, David Velleman, and Daniel Watts. I especially am grateful to Michael Bratman for his support and feedback. Later versions (which included Sections 5 and 7) were presented at: the Philosophy Colloquium at Queen’s University (2014); the annual joint meeting of the Montréal and Toronto Centres for Ethics (2015); the Association for Social and Political Philosophy annual conference (2015); the Department of Philosophy Colloquium at Marquette University (2016); and the North American Society for Social Philosophy annual conference (2016). The audiences at those presentations provided me with valuable feedback, which greatly improved the paper. I especially am grateful for comments from: James Boettcher, Simone Chambers, Yoon Choi, Ryoa Chung, Gordon Davis, Gerald Gaus, Will Kymlicka, R.J. Leland, Andrew Lister, Margaret Moore, Julius Sensat, Phil Smolenski, and Margaret Walker. The final versions of the paper received extensive and invaluable comments from two anonymous reviewers and an editor (Section 6 was added in response to some of those comments). I thank them for their considerable time and help. I regret that I (almost certainly) have forgotten some people who have helped me improve this paper over the past fourteen years. Finally, I own all of the flaws that remain.
