No individual or body can count as the representative of another unless selected or authorized to act in that role. The representative speaks and acts in the name of another individual or group, and this, unlike speaking or acting on another's behalf – say, speaking or acting as a self-appointed advocate – requires authorization in the role. This authorization raises two questions. First, is it legitimate? Is the representative selected by a suitable agent or agency and under suitable rules? Second, is it motivated? Is the selection made on the grounds that the candidate is distinctively eligible or qualified for a representative role?

Under electoral arrangements, authorization comes via the selection of the representative, directly or indirectly, on the basis of a popular vote. And it is the disposition to be responsive to the attitudes of electors, which that very mode of selection is designed to encourage, that qualifies the candidate to serve in a representative role. The candidate or deputy may be responsive at only a general level to the attitudes of electors – say, to their values or interests – or responsive to detailed wishes and instructions; in the first case, such a deputy will count as a trustee, in the second as a delegate.

Recent exercises in the renewal of republican theory have focused on the Italian–Atlantic tradition of republicanism and on its demise with the rise of classical liberalism in the early nineteenth century. The focus on this tradition has been theoretically fruitful in generating a novel way of thinking about freedom and government in the contemporary world. But there is a distinct Franco–German tradition of republicanism, developed from the time of Rousseau and Kant, and it is important not to confuse this with the older way of thinking on which neo-republicans focus. My aim here is to reduce the risk of confusion by laying out the key differences between the two traditions and putting them in historical context. Inevitably I have to stylize the traditions I discuss but I hope that I do no serious injustice to the figures I address.

While Rousseau and Kant remained faithful to some core ideas in the Italian-Atlantic tradition of thinking – in particular, as we shall see, to the idea of freedom as nondomination – the way of thinking about citizens and the state that they ushered in was as inimical to the tradition as classical liberalism. Indeed, as liberalism came to displace traditional republican doctrine as the main ideology of the English-speaking world, the name “republicanism” came to designate the new Franco–German doctrine. It is primarily with this doctrine, rather than the Italian-Atlantic tradition, that critics of liberalism like Hannah Arendt (1958; 1973) and Michael Sandel (1996) identify, for example.

In order to gain a good understanding of the concept of freedom as non-domination it will be useful to focus first on what is required for freedom in one or another choice. As we saw, republicans traditionally concentrated on the freedom of the person, period – the free status of the liber, or ‘free-man’ or citizen – rather than on the freedom of a person’s particular choices (Pettit 2007e; Skinner 2008a). But once we know what freedom of choice requires, we can represent people’s status freedom as a function of their freedom over a common range of choices, secured on the basis of common norms and laws. We will return to that issue in the next chapter.

A choice is identified by a set of mutually exclusive, jointly exhaustive options, as in the choice you may have between doing X, doing Y and doing Z. The options are available insofar as two conditions hold, one objective, the other cognitive. Objectively, it is true that you can do X, or you can do Y, or you can do Z, and that’s it: there is nothing else you can do instead. And this truth registers cognitively: it holds according to your own perceptions of the scenario. Thus if you have a choice between X, Y and Z, then you must understand each of those options under its relevant aspect, as a case of X-ing or Y-ing or Z-ing; realize that you can choose any one of them, taking it under that aspect; and recognize that there is no further alternative. Whether you do X or Y or Z is up to you and you only; you can think truly ‘I can do X’, ‘I can do Y’ and ‘I can do Z’. It may not be the case that any option chosen is logically guaranteed to materialize as a result of your decision – the letter you mail may not reach its destination because of problems at the post office – but it is enough that as a matter of contingent fact there are no obstacles in the way. Sending a letter to your correspondent is an option for you just to the extent that the world happens to be compliant – there are no postal problems – and success is in your hands.

THE PROJECT

Every philosophy of the good society starts with an account of the canonical complaint that the state should help to put right: the evil that the society should drive out by means of political organization and initiative. The complaints targeted for political rectification come in two broad families. On the one side, personal afflictions like misery or poverty or inequality; on the other, social failures like division or disorder or perhaps an excess of customary restriction.

The more personal complaints generate a powerfully motivating agenda, since most of us would rejoice in a state that silenced them. But these complaints are liable to seem politically over-demanding. While it would be good to be rid of misery or poverty or inequality, not everyone will agree that the state could, or should, be given the job of dealing with them. The removal of the less personal evils is not politically over-demanding in the same way, for most people will think that the state is able to remedy such failures. But these complaints may fail to motivate appropriately: their rectification falls short of what many of us feel that we in a politically organized society can and should collectively provide for our members.

According to the argument developed in Chapter 3, a state will be legitimate to the extent that the order it imposes on its citizens is imposed under popular control. But, so the argument continued, people will control the state just insofar as they have an individualized, unconditioned and efficacious influence that pushes it in a direction that they find acceptable. First, the people must have an influence on government that is individualized, unconditioned and efficacious. And, second, they must exercise this influence to a purpose or direction that individuals – or at least those who are prepared to live on equal terms with one another – find acceptable. People must have such a power over government that the regime can be described, in a rich, egalitarian sense of the term, as democratic: a regime that establishes the kratos, or ‘control’, of the demos, or ‘people’.

This conclusion gave us the task of spelling out how the institutions of a society might meet this design specification and implement a republican conception of democracy. We took up that task in the previous chapter, when we looked at how the people might be given individualized, unconditioned and efficacious influence over government. Without being able to go into detail, we saw that that job specification was likely to require an open electoral system under which individualization would be achieved by the possibility of individualized contestation; unconditioned independence by the resistive character of the citizenry; and efficacy by the insulation of the channels of popular influence against the distorting effects of electoral pressures and private lobbies. Under this image, the political institutions would require many amendments in even the best-practice democracies today; but they would not be so distant from current arrangements as to seem utterly infeasible or utopian.

The argument of this book has taken us over a wide terrain, introducing the republican perspective, traditional and contemporary; presenting the ideal of freedom that lies at its core; sketching a theory and model of the social justice that this ideal would support; defending a matching, republican theory of political legitimacy; and then outlining a model of the democratic institutions that might be thought to satisfy that theory. In conclusion, I think the best thing I can do is to provide a summary of the claims maintained in the development of the argument. While the summary is inevitably sketchy and inexact, I hope that it will help to facilitate readers in finding their way through a book that I wanted to make shorter and simpler than it has turned out to be.

INTRODUCTION. THE REPUBLIC, OLD AND NEW

1. The main ideas in the republican tradition are: freedom as non-domination, the mixed constitution and the contestatory citizenry. Appearing in the Roman republic, in medieval and renaissance Italy, in seventeenth- and eighteenth-century Europe and Britain, and eventually in revolutionary America, they suggest that the state should enable its citizens – however inclusive – to act as free, undominated persons in the sphere of the fundamental liberties, being protected under a mixed, contestatory constitution.

2. This republican tradition came under sustained attack at the hands of Jeremy Bentham and William Paley in later eighteenth-century England, as they introduced a theory of freedom as non-interference. They argued that while the state should cater for the freedom of all citizens – now understood more inclusively – it should do so with only this less demanding ideal in view.

3. Italian–Atlantic republicanism was also challenged in the late eighteenth century by the communitarian republicanism of Jean Jacques Rousseau. While he continued to think of freedom as non-domination, he followed Jean Bodin and Thomas Hobbes in giving up on the mixed constitution and the contestatory citizenry. He argued that there had to be a single sovereign in any well-functioning state and thought that in a republic this should be the assembled, incorporated people.

4. The aim of this book is to build philosophically on the main republican ideas, developing a theory of social justice and, in particular, political legitimacy, where justice governs people’s relations with one another, legitimacy their relations with the state. The republican theory of legitimacy gives the state a democratic job specification, requiring it to operate under equally shared, popular control.

5. …

In the last chapter we focused on issues of social justice: that is, justice in the relations between people within a state, including relations mediated by their belonging to groups and bodies of various kinds. According to the theory developed, justice requires the state to promote the freedom as non-domination of all its citizens – broadly, all adult, able-minded and more or less permanent residents – safeguarding their fundamental liberties on the basis of public laws and norms. This focus on relations amongst citizens leaves out of consideration the relations between citizens as a whole and the state itself. It ignores the question of whether the state operates with political legitimacy in imposing a social order, however just that order might turn out to be. It is one thing to argue that the social order imposed by a state is just, it is quite another to argue that the political imposition of that order is legitimate.

Social justice does not entail political legitimacy, by this account, nor does political legitimacy entail social justice. Thus, to take the second dissociation first, a state might be fully legitimate, by whatever criterion, and yet not succeed in furthering the cause of social justice very well; it might support misconceived, if not ill-willed, policies. It is a failure of this kind that Rousseau had in mind when he acknowledged the possibility that the perfectly legitimate regime – in his terms, the regime that seeks to enact the general will – may still go astray: ‘By itself, the people always wills the good, but by itself it does not always see it’ (1997: II.6.10).

The idea in this chapter and the next is to explore the institutional possibility that the people in a polity might have such control over those who run the state that they are not individually dominated by the interference that the state practises in taxation, coercion and punishment. To the extent that they have a control that makes such interference non-dominating, the citizens will not lose out in freedom just by the fact of living within that state and, by republican criteria, the state will count as politically legitimate. A state that is legitimate in that sense may not achieve a great deal in guarding against private domination and achieving social justice, though it probably has to achieve some minimum threshold if citizens are going to be capable of exercising control over its doings.

Control, as we saw, depends on two distinct elements, influence and direction. Thus the people will achieve control over the state insofar as they attain influence, on the one side, and succeed on the other in using that influence to impose a suitable direction on government. Such popular control will be suited to republican purposes, guarding against domination, to the extent that it gives each citizen an equal share in the control, particularly an equal share in a form of control that is suitably unconditioned and efficacious. People must enjoy an equally accessible form of unconditioned and efficacious influence that imposes an equally acceptable direction on the state. In this chapter I look at how people might enjoy the required influence and I turn in the next chapter to how this influence might impose the required direction. In this chapter’s discussion of the institutional means whereby people might come to enjoy the required influence I shall anticipate the discussion in the next and assume that the influence they enjoy can support a popular direction and not merely be wayward in character.