Discursive control,non-domination and Hegelian recognition theory

Introduction

In recent years, both Pettit’s neo-republican writings on freedom and Hegel’s work on recognition as constitutive of free agency have received increasing amounts of attention, especially as possible forms of extension and improvement of canonical Kantian-liberal views on the conditions for free and autonomous social agency. ¹ While Pettit has forcefully defended the idea that freedom as non-domination offers significant advantages over (Berlin’s) standard liberal view of freedom as non-interference, theorists like Pippin, Brandom, Wood and Pinkard have shown that (at least part of) Hegel’s accounts of rational agency and recognition can be read as useful additions and complements to Kant’s practical philosophy towards a more social-relational reading of freedom and autonomy. ² In fact, republican non-domination and Hegelian recognition theory are two of the most widely and vividly discussed concepts in contemporary political theory. ³

However, both projects and fields also face some problems and open questions. With regard to Pettit’s work it is striking to observe that Pettit’s political conception of freedom as non-domination has got vastly more attention than his equally, if not even more important, work on the normative conditions of freedom, and his notion of discursive control. While non-domination is certainly an important concept, it seems that in order to fully understand its roots and socio-political scope one would have to engage more deeply with Pettit’s views on agency, reasoning and also recognition. Unfortunately, though, despite its impressive scope even Pettit’s own work leaves much to be desired when it comes to spelling out the normative conditions for discursive control, the exact role of recognition in his account, and the connection between non-domination as a political ideal and freedom as a normative concept. ⁴

Similarly, current reactualizations of Hegel’s practical philosophy in general, and of his recognition theory in particular, face criticisms from two sides: either that Hegelian rational agency is too thin a concept to ground a full-blown theory of socio-political freedom and justice, in part because recognition as a concept is unsuitable for carrying significant normative weight; or that the work done on Hegel by scholars like Pippin and Honneth expects too much of recognition and overburdens the notion in such a way that it either obscures the real socio-political problems, or fails to safeguard the freedom of the individual adequately against the perils of modern mass societies. ⁵ The problems for existing theories of Hegelian practical thought are thus twofold: first, Hegel’s account of rational agency must be proven to be substantial enough to ground a moral-political theory; second, the exact role of recognition within such a theory must be determined and shown what recognition theory adds to existing accounts, without overburdening the concept.

While this article can obviously not attempt to address all the issues mentioned above, its aim is to combine Pettit’s account(s) of freedom, both his work on discursive control and on non-domination, with Pippin’s and Brandom’s reinterpretation of Hegelian rational agency and the role of recognition theory within it. The benefits of combining these two theories lie, as I hope to show in the course of the following sections, in three findings: first, re-examining Hegelian agency in the spirit of Brandom and Pippin in combination with Pettit’s views on freedom shows clearly why and in which way a Hegelian account of rational agency can ground an attractive socio-political account of freedom; second, the reconciling of discursive control and non-domination with Hegelian agency shows how the force and scope of recognition become finally tangible, without either falling into the trap of overburdening the concept, or merely rephrasing existing equal respect views; ⁶ third, the arguments from this article also highlight the importance of freedom as non-domination (if based on Hegelian readings of agency) and how this notion is, indeed, as Pettit himself claims, an agency-freedom which aims at successfully securing the social, political, economic and even (some) psychological conditions for free and autonomous agency. ⁷

The article is divided into three sections. Section I gives an overview on Pettit’s accounts of freedom as non-domination and freedom as discursive control, while focusing on the latter, trying to carve out more clearly at which points the concept of discursive control seems compatible with Hegelian ideas. Section II, then, uses a Hegelian account of agency, inspired by the work of Brandom and Pippin, to flesh out the normative conditions of freedom and to explain the exact role recognition is to play within such an account. In section III the normative notion of freedom developed in section II will be brought to the political level, so as to show how rational agency, recognition, freedom and non-domination all hang together. Moreover, section III will argue that the account of freedom developed in this article is uniquely able to secure the conditions for free and autonomous agency under the circumstances of a complex and globalized world.

Before I start with the actual argument, though, it is important to stress that it is Pettitian and Hegelian thought the article is concerned with. That is to say, that the article aims to be neither an exercise in ‘correct’ Hegel exegesis, ⁸ nor an exact reconstruction of Pettit’s ideas. Instead, the article uses Hegelian and Pettitian ideas in order to argue for an agency-centred notion of freedom, which pays special attention to both recognition and non-domination. In presenting an account of its own this article is thus different from other highly interesting work on the connection between Hegelian recognition theory and neo-republican non-domination, such as the question of whether Hegel was a republican. ⁹

I: Pettit on discursive control and non-domination

The classical conception of republican freedom is freedom from domination, meaning that no person should stand to another person in the position of a master, that is, be able to interfere arbitrarily. ¹⁰ Freedom as non-domination is thus worried about a person’s ability to arbitrarily interfere with another person’s life and not about interference as such (like standard liberal views of freedom), even if this ability is never exercised, as already the mere existence of this interfering potential is identified as a principal threat to freedom.

As Pettit points out in response to Carter and Kramer, both of whom defend non-interference views of freedom, republicans believe that ‘there may be freedom in the presence or absence of interference, and there may be unfreedom in its presence or absence’, since the threat to freedom lies for republicans in the possibility of arbitrary interference, or alien control, that is, domination. ¹¹ According to freedom as non-domination, for an agent to act freely not only does alien control need to be absent, but that agent – with her or his relevant choices – also needs to be generally well protected against arbitrary interference. ¹² Republicans like Pettit, therefore, stress the importance of effective protection against both domination by private individuals and domination by the institutions of the state. The ideal of non-domination, thus, predominantly cares about a person’s free agency, which must be protected through constitutional authority and mechanisms which themselves are not allowed to become dominating factors. ¹³ Moreover, as Bohman rightly points out, republican freedom is egalitarian in nature, but non-domination in itself does not require equality as a distributive standard. ¹⁴

Phrased differently, freedom as non-domination cares about the nature of certain social relationships, since alien control requires interpersonal relationships and the power to arbitrarily interfere with somebody’s actions is a positional good. ¹⁵ Freedom as non-domination is egalitarian insofar as it aims to secure a state of affairs in which all members of society are in discursive control, that is, people ‘have the ability to discourse and … have access to discourse’. ¹⁶ In order to be in discursive control, people have to stand in discourse-friendly relationships to each other, as only in such relationships can agents exercise their agency freely. As Pettit puts it, free persons are persons who can exercise their agency freely, agency which is ‘allowed to them by their standing relative to others’. ¹⁷ In other words, ‘[w]e stand in a variety of relationships to others, each of them characterized by its own distinctive pattern of power and vulnerability, authority and liability’, and only in certain forms of mutually discourse-friendly relationships can we rightly be considered to be free. ¹⁸ Pettit’s account of freedom is, thus, indeed very sensitive to the relationships we stand in with regard to others, recognizing the freedom-influencing effects of power structures and their associated vulnerabilities. Only if certain forms of social relationships persist can we enjoy freedom as discursive control. So what exactly does freedom as discursive control require?

A free person in discursive control must have, according to Pettit, ‘the ratiocinative capacity to take part in discourse, and the relational capacity that goes with enjoying relationships that are discourse-friendly’. ¹⁹ That is to say, the person in question has to be reason-responsive, able to reflect on reasons, as well as actually able to participate in discourse and be recognized as a discursive authority. There are two very important points here which seem to get overlooked in the reception of Pettit’s account of freedom: first, Pettit stipulates rationality criteria which he takes to be necessary aspects of a free agent, since for Pettit only free agents are fit to be held properly responsible; second, Pettit explicitly connects the idea of discourse-friendly relationships with the concept of mutual intersubjective recognition. ²⁰ Both observations are of crucial importance for the discussion in this article, since only an in-depth analysis of the nature of discourse-friendly relationships will further elucidate the normative conditions of Pettit’s conception of freedom, and show how Pettit’s ideas offer various connection points to Brandom’s and Pippin’s work on Hegelian agency.

If one takes seriously Pettit’s remarks on the intricate connection between free agency and responsibility, and on the importance of recognition, discourse-friendly relationships can be redescribed as relationships of recognition, in which persons are socially recognized as reason-responsive and reflecting discourse participants. Within these relationships, agents give, either implicitly or explicitly, (and sometimes discuss) the reasons for their actions and stand by their reasons as responsible rational agents. However, while Pettit explicitly refers to the need for social recognition of an agent’s status as a legitimate discourse participant, he does not engage in depth with the actual role recognition plays in creating and sustaining social relationships within which free agency is possible. Moreover, Pettit’s notion of discursive control still seems to lack a more detailed account of responsible rational agency and its social conditions. While it is certainly true that we commonly assume that a free person is also a person we can hold responsible for his or her actions, it is yet unclear how potential arbitrary interference, which never actually occurred, should influence the way we assess an agent’s responsibility for a specific harmful action A.

In fact, Pettit explicitly refers to the work of Honneth and Darwall when talking about recognition, but doing so leaves still many questions unanswered, as Honneth famously differentiates three spheres of recognition (love, respect and esteem) using a Hegelian framework, while Darwall extends Kantian respect-recognition for his construction of the second-person standpoint and its rules of address. ²¹ It hence remains open to discussion what kind of recognition we actually need in order to create discourse-friendly relationships and what this recognition enables us to do. It is here that Brandom’s and Pippin’s readings of Hegelian agency can help, since they offer an account of the role of recognition and responsibility, which sits nicely with some of the things Pettit writes on these issues. As my arguments in the following section will suggest, the key to solving the puzzle lies in grasping what kind of social relationships discourse-friendly relationships actually are; that is, to carve out more clearly the socio-normative conditions of discursive, or free, agency.

II: Hegelian agency, recognition and freedom

As we saw in the preceding section, Pettit’s view of freedom as discursive control establishes two important criteria which must be fulfilled so that a person can be considered to be free: first, the agent in question must be capable of participating in discourse as a critical, reflective reason-responsive person; second, the agent in question must be a socially recognized reason-giver who stands in discourse-friendly relationships with others. Only if both these conditions are satisfied is a person free, and therefore fit to be held responsible. Pettit’s view of freedom and free agency is thus distinctly social, a feature it shares with Hegel’s post-Kantian account of free agency.

Hegel extended Kant’s principal insight that humans are normative beings. For Kant, our capacity to reason, that is, our ability to reflect on reasons, pass judgement and use concepts, forces us to commit ourselves normatively, to wit, to identify and embrace rules and principles. This normative capacity of ours makes us free and responsible beings, as in order to act according to reasons and to endorse norms legislated by our own will, we have to presuppose practical freedom. As Kant puts it, ‘the will of a rational being, [must] be regarded as free’. ²² Hegel takes this Kantian insight of the normative nature of freedom and extends it, by adding a distinctly social dimension. Hegelian agency starts from the insight that we are always already part of a social community, and thus subject to a range of discursive, social and cultural norms and practices. Within this social world we exercise our agency by reflecting on reasons, committing to norms and participating in the social and discursive practices of society. As Brandom observes, as ‘concept-mongering’ ‘normative creatures’ we ‘live and move … in a normative space’. ²³ This normative space is a shared social space of reasons, as every subject, surrounded by conceptual norms, has to apply rules in order to apply concepts, and justify its actions on the basis of reasons. ²⁴ That is to say, the subject’s social and discursive practices express the rules and norms the subject endorses, and they function as a test for the subject’s commitment. In other words, it is through social practices of reason-giving and reason-taking that a subject’s actions and normative commitments are judged and interpreted.

As Pippin points out, being in the space of reasons is a necessary aspect of being a rational agent, as being socially recognized as a legitimate and competent reason-giver is a precondition for successfully exercising one’s agency. As shown in Hegel’s master–slave dialectic, mutual recognition is necessary for establishing social relationships in which two agents acknowledge each other and establish themselves as authoritative reason-givers, because huge power imbalances such as in the case of master and slave make freely given mutual recognition impossible. A rational agent is an agent who critically reflects on reasons and acts accordingly, who recognizes other members of society as competent judges, and who is likewise recognized by others as a legitimate rational agent. Put crudely, free rational agency is a social-recognitive and normative state. In many ways this social-recognitive account Brandom and Pippin give seems to fit with the more cryptic remarks Pettit made in his explication of freedom as discursive control. Precisely because every human being is caught in a complex web of social relations and intersubjective relationships, freedom and the exercise of rational agency require mutual recognition.

On a Hegelian reading, then, rational agency is seen as a normative social state in which the individual subject is able to act freely and autonomously for reasons which are its own, embedded in a social space of mutually reciprocal relationships of recognition. As Pippin points out, ‘freedom is understood by Hegel to involve a certain sort of self-relation and a certain sort of relation to others’. ²⁵ Hegelian freedom is social-relational, acknowledging every subject’s indeterminacy and constitutive dependency, that is, the independence of the free and autonomous subject is at the same time a kind of social dependence, namely mutual constraint by norms. ²⁶ In short, being a free agent depends ‘on being recognized as, taken to be such, a free agent within a community of mutually recognizing agents’. ²⁷

There are two aspects to achieving free rational agency, then, namely an ‘outer’ state of standing in social relationships of mutual recognition, and an ‘inner’ state of seeing oneself as the legitimate source of reasons. The ‘inner’ condition refers to a person’s ability to reflect critically, to think independently and to stand socially by her or his reasoning, which means that a person is willing to take responsibility for her or his reasons and actions. The ‘outer’ condition meanwhile refers to a set of reciprocally recognitive social relationships. Using the language of freedom and autonomy we can say that a Hegelian reading of free agency is based on freedom as recognition and autonomy as responsible endorsement. What does autonomy as responsible endorsement mean?

While autonomy is often misportrayed as a claim for perfect self-sufficiency, the idea of autonomy as responsible endorsement is actually a social-relational concept, which grasps the ‘inner’ condition of Hegelian agency; namely, to stand behind one’s reasons. Only if a person is able to see herself as a legitimate source of claims and to endorse norms and reasons freely in a reason-responsive manner, willing to take responsibility for these reasons and claims, can she be taken to be autonomous according to the idea of autonomy as responsible endorsement. However, as Anderson and Honneth observe, a person’s autonomy, that is, a person’s self-belief and rational self-confidence, is extremely sensitive to the social relations he stands in, meaning that freedom and autonomy often go hand in hand, since both depend to a significant degree on social relationships of mutual recognition. In other words, there is a range of alienating and dominating conditions, such as certain social practices or discriminating institutions, which threaten both the autonomy and the freedom of an agent.

This dual structure of free rational agency thus nicely mirrors Pettit’s description of discursive control as being subject to the agent’s ratiocinative capacity and the existence of discourse-friendly relationships. The relationship between these two aspects of freedom is mutually supportive and dependent, as one is obviously only going to feel like a legitimate source of reasons if one receives social recognition, while one will not be considered by others to be a reason-responsive discourse participant unless one displays the ratiocinative capacities required. ²⁸ However, it is crucial to highlight the fact that one’s reason-responsiveness and ability to take responsibility for one’s reasons and actions, as well as the social recognition one gets as being a participating member of social discourse worthy of address, are matters of degree, as the exercise of one’s rational agency will get better by practising it, which in turn affects one’s reputation as a socially recognized reason-giver.

From what was said so far, we can see how closely freedom as discursive control and recognition seem to be related. In fact, applying our Hegelian reading of rational agency as freedom as recognition and autonomy as responsible endorsement allows us to establish a direct link between the recognitive relationships which exist in a given society and an agent’s access to discursive control. However, we still have to define more clearly what this kind of mutual intersubjective recognition actually entails and how the normative dimension of freedom and agency is connected to the political dimension. In the remainder of this section, I will thus focus on the kind of recognition rational agency and freedom as discursive control require.

Within contemporary political philosophy there exists a myriad of different theories of recognition, many of which claim to be in one way or another Hegelian in nature. Instead of engaging with all these varying conceptions, which would be an article-length endeavour of its own, I will try to carve out the role of recognition for our account by reconstructing the clues we already have from what was said thus far, as especially Pippin gives us plenty to work with.

In the account of agency we presented above reciprocal intersubjective recognition was defined as a complex social-relational state in which two important conditions are fulfilled: first, every agent recognizes the other members of society as competent judges and legitimate sources of claims and reasons, while enjoying reciprocally this very same recognition as a legitimate reason-giver; second, every agent sees himself or herself as a legitimate source of reasons, meaning that an agent is able to stand by his or her judgment and take responsibility for it. These two features of reciprocal recognition we summarized in the idea of free agency as a person enjoying freedom as recognition and autonomy as responsible endorsement. In this account of free agency recognition thus is a complex social-relational state, which can be achieved neither by instituting a range of legal guarantees, nor by simply distributing the good of esteem recognition as a means to securing what Rawls calls the social bases of self-respect. ²⁹ In other words, neither is recognition a tangible and quantifiable good we can dole out so as to ensure that every person feels treated fairly, nor can recognition be reduced to a set of formal legal guarantees. As Pippin points out, relationships of mutual recognition are actualized, that is, lived in and through social practices and their associated norms.

The kind of recognition an agent requires for being free and autonomous is the social-relational standing of being a legitimate reason-giver and source of reasons, which is expressed in the social and discursive practices of society. Freedom is thus not a state which can be formally achieved but it must be actualized in a society’s shared social space of reasons, a fact which Pettit also highlights in his account of freedom as discursive control. ³⁰ It is this particular conceptualization of recognition which explains how intricately freedom and intersubjective recognition are linked, and it also explains why we should interpret freedom as an agency-centred idea, which focuses on the nature of certain social relationships. However, construing recognition in this way is not only crucial for understanding the conditions and workings of rational agency and freedom, but also – as paradoxically as that might sound – for realizing the limitations of the concept of recognition. As Bohman rightly observes, recognitive statuses in the way we described them ‘are not sufficient to avoid domination’. ³¹ Recognition is a normative condition of freedom, but on the socio-political level the safeguarding of every agent’s freedom requires further protective measures.

Only if we understand, first, how vulnerable our rational agency is, and second, what recognition can do and what it cannot do, can we see the clear connection between recognition, free agency and the republican ideal of non-domination. Thus, the following section will deal with this intricate relationship and highlight how our normative account of freedom ties in with the political idea of non-domination.

III: Freedom, agency and non-domination

At the beginning of this article I claimed to show that Hegelian agency is ‘thick’ enough to be the basis of a theory of social and political freedom, and that combining Hegelian recognition theory with Petit's account of freedom allows us to see the way in which non-domination is indeed an agency-centred concept. After what was said thus far, critics might still wonder how the Hegelian account set out above can be translated to the political level, as we focused on the normative workings of recognition and its limits. Thus, the aim of this last section is to show that the protection of free agency requires non-domination, and that there is a direct link between the idea of freedom as recognition and freedom as non-domination.

As pointed out earlier, in order to be a free agent a person requires two things, namely freedom as recognition, which assures the person’s standing as a free and equal legitimate reason-giver, and autonomy as responsible endorsement, which is necessary for being a reason-responsive and responsible agent. However, neither recognition nor responsible endorsement is an effective safeguard against all the possible threats to free agency such as arbitrary interference, domination (or alien control), and alienation.

Rahel Jaeggi, for instance, offers an excellent analysis of the different faces of alienation and how within our advanced capitalist production processes many forms of lifestyles, certain groups of society (e.g. a large number of women, single parents, people in cleaning jobs, most immigrants, etc.) and a range of belief-systems are systematically subject to alienating economic, political and cultural practices and structures. ³² Instances of alienation do not lead to a situation in which a person’s freedom is dependent on the goodwill of a benevolent dictator, or subject to potential arbitrary interference, but they create conditions in which an agent is unable to reason autonomously, that is, to responsibly endorse her or his own reasons and judgments. Alienation threatens the ‘inner’ conditions of free agency we discussed earlier, while domination, be it by private persons or through the institutions of the state, undermines the ‘outer’ conditions of free agency.

Hence, while Pettit’s account of freedom as discursive control is clearly sensible to the problems triggered by alienation, non-domination might be taken to struggle with the problems posed by alienation. However, as should be clear from what we said above, freedom as non-domination properly conceived, that is, as an account of freedom which tries to safeguard every person’s rational agency through providing the material, social and recognitive condition for exercising this capacity, goes hand in hand with what we can call autonomy as non-alienation. Let me explain what this means, and how non-domination and non-alienation are key aspects of social freedom.

Non-domination is, according to Pettit, an agency-centred conception of freedom because it cares about the actual freedom of an individual (1) to make decisions freely, and (2) to feel for the right reasons as an equal citizen, while also providing through constitutional guarantees and institutional mechanisms ‘a systematic sort of protection and empowerment against alien control’. ³³ As we saw earlier, mutual recognition is a necessary requirement for points (1) and (2), as free agency can only be actualized in discourse-friendly relationships, but only the safeguards of non-alienation (especially for point 2) and non-domination can effectively secure a person’s socio-political freedom. That is to say that an agent can only considered to be free, if she or he is properly protected against and free from alien control, as well as not subject to alienating social conditions, which means the ability to see eye to eye with fellow citizens. ³⁴ Protecting an agent’s freedom means to provide her or him with the social conditions in which choices, judgments and actions are both content-independent and context-independent.

The condition of content-independence refers to a state of affairs in which agents are able to choose and act in the way they want independent of whether they choose X, Y, or Z. In a situation in which an agent A chooses X and gets X, but had the agent chosen Y would not have got Y, content-independence is violated. Similarly, context-independence prescribes that the success of a person’s choices and actions should not be depending on the arbitrary benevolence of another person P, as if that were the case (that A’s successful choice depends on P’s good will) it would present a case of domination. It is crucial to include both content- and context-independence in an account of freedom since if we restrict freedom exclusively to content-independence with respect to choice, we fail to grasp certain aspects of the agency-dimension of freedom we discussed in the preceding section. Moreover, it is necessary to make clear that context-independence and successful rational agency are not only threatened by benevolent dictators, which seems to be the standard case in much of the existing literature; we also need to highlight the different faces of domination, as well as the potentially agency-crippling effects of alienation. It is for this task, namely, to highlight the intricate social-relational condition of free and autonomous rational agency that we need both Hegelian recognition theory and republican non-domination. While the two concepts seem to operate in very different spheres, to wit, the normative and the political, they complement each other within a normatively sound and socio-politically forceful account of freedom.

As Pettit repeatedly stresses, non-domination is supposed to safeguard the actualization of a person’s free agency, a task which can only be achieved if one’s conception of freedom is sensitive to all three issues, namely, the negative effects of relationships of misrecognition, disrespect and inequality, the alienating effects of discriminating and agency-hampering social structures, as well as the possibility of arbitrary interference through other members of society or the institutions of the state. To put it somewhat crudely, free and autonomous agency requires a set of reciprocally recognitive intersubjective relations, which can only be safeguarded and maintained, that is, practically lived and sustained, within a society that is built on the ideas of non-domination and non-alienation.

Using Pettit’s language from his Theory of Freedom, we can put this conclusion also in other terms, namely, that freedom as discursive control requires both mutual recognition as condition for actually achieving discursive control and non-domination in order to maintain and protect every person’s freedom as discursive control. The reason for this is that freedom is a complex normative social state which is vulnerable to a range of threats which go well beyond direct interference, such as structural inequalities, status differences, sexist and racial stereotypes, to name but a few. Non-domination and non-alienation, then, simply secure that every agent enjoys the ‘inner’ and ‘outer’ conditions for freely and autonomously exercising his or her agency.

Conclusion

At the beginning of this article, I started with the observation that despite their immense popularity republican accounts of freedom and Hegelian recognition theory both face a range of criticisms. I claimed that marrying a Hegelian reading of agency, based on the writings of Pippin and Brandom, with Pettit’s ideas on discursive control and non-domination would provide significant benefits with respect to at least three issues: (1) the claim that Hegelian agency is too thin a concept to ground a substantial theory of socio-political freedom; (2) the charge that recognition is either a weak, and thus redundant, concept, or an overburdened notion which thus cannot help us to explore the normative conditions of free agency; and (3) the criticism that non-domination is not as important a concept as republicans want us to believe.

However, as the analysis in this article hopefully clearly showed, that all three issues can be resolved, so that Hegelian agency provides a very suitable normative background to republican non-domination, and that the role recognition plays within this account is a crucial one. Moreover, the central importance of non-domination and non-alienation as agency-centred concepts was clearly highlighted.

Freedom, understood as a normative social state, requires a whole set of recognitive intersubjective relationships, in which an agent can see herself or himself as the legitimate source of claims and reasons while also acknowledging other members of society as competent judges. Only within such a set of recognitive relationships can an agent be said to act freely and autonomously, that is, to take responsibility for her or his own reasons and actions. However, being a rational agent who enjoys freedom as discursive control is a vulnerable social state which calls for protection from a range of threats to agency. This protection is offered by the ideas of non-domination and non-alienation, which means that in this article not only did we flesh out the connection between freedom and recognition but we also clarified the scope and importance of non-domination. Overall, using the insights from our arguments one can see (1) that Hegelian agency is a thick enough concept for a theory of social freedom, (2) which role recognition can and should play within an account of freedom, and (3) how the republican idea of non-domination is indeed, as Pettit claims, an agency-conception of freedom.