Whereas the idea of negative freedom has hardly any precursors in antiquity or the Middle Ages, the notion of reflexive freedom reaches all the way back to the intellectual prehistory of modernity. Ever since Aristotle, a number of thinkers and philosophers have claimed that in order for individuals to be free, they must be able to arrive at their own decisions and influence their own will.¹ The historical asymmetry between these two concepts of freedom demonstrates that the idea of reflexive freedom cannot be viewed merely as an expansion or more profound version of the ideal of negative freedom. It would be careless of us to regard the notion of an externally secured free-space as a merely preliminary stage of a model of freedom which then resolutely focuses on the internal. Negative freedom is an original and indispensable element of modernity's moral self-understanding; it conveys the demand that all individuals be entitled to act in accordance with their own preferences, without external restrictions and without having to submit their motives to rational judgement, provided they do not violate the right of their fellow citizens to do the same.² By contrast, the idea of reflexive freedom focuses solely on the subject's relationship-to-self; according to this notion, individuals are free if their actions are solely guided by their own intentions.

But even this general determination shows that we can attach very different conceptions to the idea of reflexive freedom. After all, what is meant by an individual's ‘own’ intentions? What does it mean to be ‘guided’ by them? We can think of a number of different answers to these questions and a variety of combined meanings. Isaiah Berlin, who spoke of ‘positive’ rather than ‘reflexive’ freedom, distinguishes between two versions of this type of ‘internally’ directed freedom. The notion that subjects are free only to the extent that they are able to determine their own actions has developed in two different directions: the idea of ‘autonomy’ and ‘self-realization’.³ Picking up on the work of Berlin, Raymond Geuss has even proposed that we distinguish between five variations on the concept of ‘positive’ or ‘reflexive’ freedom; he claims this one idea can be divided into various sets of meanings, each taking into account all the different aspects or modes of what it means to act according to one's own will.⁴

Historically, the idea of reflexive freedom centres on the proposal that we distinguish between autonomous and heteronymous acts – a distinction that goes back to Rousseau and completely realigned the meaning of individual freedom in one stroke: In order to count as free, it is no longer enough for an act to be carried out in the external world without resistance, rather the act must also be traceable to the will of the agent. The modification of human nature required to justify such a distinction can be found in Rousseau's Émile; the chapter entitled ‘Confessions of a Savoyard Vicar’ proposes ideas about human will that anticipate much of what Kant would have to say about moral autonomy three decades later.⁵

Already in The Social Contract, published only a few months prior to Émile, Rousseau noted that humans cannot be free as long as they remain enslaved by the ‘impulse of mere desire’; they are only free once they practice ‘obedience to self-imposed law’.⁶ Rousseau does not further pursue this divide in human nature, in which ‘ethical freedom’ conflicts with ‘desire’, and it is not until Émile that he asks how children are to become capable of self-determination. The considerations that Rousseau puts in the mouth of the vicar begin with a claim that appears to be a critique of the idea of merely negative freedom: ‘When I give myself up to my passions, I act according to the impulse of external objects … I am a slave through my vices.’⁷ An action that occurs as a reaction to sense impulses cannot be described as ‘free’, because this is merely the uninterrupted continuation of the ‘law of bodies’, i.e. natural causality, in human activity. Unlike heteronymous actions that the subject feels compelled to perform, it ‘feels’ that true acts of its free will have come about in a different way, and it feels that it has managed to realize exactly what it has set out to do. In the first instance Rousseau grasps the distinction between heteronymous and autonomous actions as a difference in terms of the self-perception of the acting subject: ‘I consent or I resist; I am vanquished or I am conqueror, and I feel perfectly within myself, when I do that which I would do, or when I only give way to my passions.’⁸ Once humans succeed in realizing what their will imposes, rather than their desires, they are capable of feeling free. Humans thereby interrupt the natural laws that dictate their impulses by obeying not an external demand, but the imperative of a previous decision. Nevertheless, Rousseau has trouble explaining the features of this puzzling entity called the ‘will’. Drawing on Leibniz, he understands it as an ‘immaterial substance’,⁹ which enables subjects to turn rational insights or the stirrings of their conscience into effective motives for action. At the same time, he assumes that such a rational or moral will cannot automatically enable acting subjects to prevail over the onslaught of their natural inclinations. On the one hand, and almost by definition, wherever a ‘free will’ exists, it should always be able to cause the action it performs; but on the other hand, only subjects seem to have the power to obey either their will or their own passions. Rousseau does not yet possess the conceptual tools to resolve these contradictions. Neither is it clear what exactly he means by ‘will’, nor is he capable of understanding what this ‘weakness of the will’ really means. However, his exploratory considerations on autonomy or self-legislation (Selbstgesetzgebung), as well as his definition of free action, were groundbreaking and fruitful enough to lay the foundation for two versions of the modern idea of reflexive freedom.

Only a quarter-century later, Kant would build on Rousseau's analysis in order to complete his own account of self-determination; what is significant about Rousseau's work for Kant is primarily that part which presents freedom as the result of autonomy.¹⁰ During this same period, Rousseau's concept of freedom shows up in a second intellectual current, which is less concerned with reason than with self-determination; what primarily interests the latter school of thought, mostly made up of early Romanticists and marginal figures within German Idealism, are those passages in Rousseau's writings in which he demonstrates that freedom is dependent on the articulation of real or authentic desires.¹¹ The intellectual effect of Rousseau's ingenious, though not always coherent, analyses of the distinction between autonomous and heteronymous actions unfolds in two different directions. Although in both cases the aim is to disclose the reflexive structure of individual freedom, the content of that reflexivity, its particularity, is answered in virtually opposite ways, though with recourse to one and the same author.

As mentioned above, Kant builds on those parts of Rousseau's theory that interpret freedom in terms of autonomy. Subjects are ‘free’ if and because they have the capacity to self-legislate and act in accordance with these self-imposed laws. While Rousseau is vague about whether these laws are merely empirical intentions or rational principles, Kant makes a resolute turn toward the transcendental. He is utterly convinced that such self-legislated laws can only bring about freedom if they derive from an insight into correct, that is, rational reasons.¹² Kant arrives at this claim by taking three bold steps, thus removing the ambiguities that afflict Rousseau's concept of the will. First, he points out that when it comes to rational beings, desiring something does not merely mean following one's inclinations. To have and formulate an intention implies resisting the laws through which nature influences our intentions. For Kant, therefore, the mere fact of human will is enough to prove that humans are capable of freedom. But it is only by taking the next step that Kant arrives at what he wants to prove with regard to Rousseau: In order to prove that humans cannot help but obey rational laws, he argues that as soon as individuals formulate an intention and thus inquire into what guides their actions, they only have the criterion of potential universality: Subjects can only obey principles they can also want everybody else to obey. ‘Since I have robbed the will of every inducement that might arise for it as a consequence of obeying any particular law, nothing is left but the conformity of actions to universal law as such, and this alone must serve the will as its principle. That is to say, I ought never to act except in such a way that I can also will that my maxim should become a universal law. Here bare conformity to universal law as such (without having as its base any law prescribing particular actions) is what serves the will as its principle, and must so serve it. … The ordinary reason of mankind also agrees with this completely in its practical judgements and always has the aforesaid principle before its eyes.’¹³

In the final step of his argumentation, Kant claims that this principle of universality also conveys an attitude of universal respect, for as soon as I ask whether all other subjects could agree to the maxim I obey in my own actions, I respect their rationality and treat them as ends in themselves. It is in his formulation of the categorical imperative that Kant captures most coherently the moral yield of his argumentation, which demands that each rational being ‘treat himself and all others never merely as a means, but always at the same time as an end in himself’.¹⁴ Therefore, humans are free precisely because they can obey the moral laws they have imposed upon themselves. Kant concludes by remarking that individual self-determination is identical to the fulfilment of the rational moral principle: ‘As a rational being, and consequently as belonging to the intelligible world, man can never conceive the causality of his own will except under the Idea of freedom; for to be independent of determination by causes in the sensible world … is to be free. To the Idea of freedom there is inseparably attached the concept of autonomy, and to this in turn the universal principle of morality – a principle which in Idea forms the ground for all the actions of rational beings, just as the law of nature does for all appearances.’¹⁵ The reflexive freedom Kant has in mind consists in the insight that we have the moral duty to treat all other subjects as autonomous beings, just as we would expect them to treat us.

Those theorists who do not primarily regard Rousseau as a theorist of autonomy, but as an advocate of integrity, offer an entirely different interpretation, according to which the reflexivity of individual freedom consists in the fact that individuals only truly become individuals once they have acquired and articulated their own, authentic will through a lengthy process of reflection. This second current in the tradition of Rousseau also justifiably invokes elements of Rousseau's notion of freedom. Already in Émile, but above all in Confessions and Julie, or the New Heloise, Rousseau emphasizes that we are only free once we have the ‘feeling’ of having achieved precisely the desires and intentions that we truly have within ourselves.¹⁶ This ideal of self-realization, which opposes Kant's idea of moral autonomy by placing the individual good before the general good,¹⁷ finds a direct continuation in the writings of Johann Gottfried Herder. In his work On the Cognition and the Sensation of the Human Soul,¹⁸ Herder outlines the reflexive process in which individuals learn to realize their ‘inner I’¹⁹ in the ‘medium’ of ‘language’.²⁰ Herder believes that nature grants each individual a unique soul which, like a seed, only requires proper care in order to grow and prosper while unfolding its potential. Analogous to living organisms, individuals achieve perfection only once they have brought to bear all of their inner powers and sensations to the extent that they can experience their own action as the execution of authentic freedom: ‘The more deeply someone has climbed down into himself, into the structure and origin of his noblest thoughts, then the more he will cover his eyes and feet, and say “What I am, I have become.”’²¹ The reflexive freedom that Herder has in mind consists in an act of acquisition, in the course of which I learn, by learning the general laws of language, to articulate the authentic core of my personality.

These two models of freedom, both of which were developed at the close of the eighteenth century and were heavily influenced by Rousseau, represent two versions of the idea that individual freedom can only be the product of a reflexive act. Both Kant and Herder are convinced that merely negative determinations of freedom are insufficient because they do not penetrate the space of reasons, thus regarding subjects as free in a merely external sense without taking account of whether their realized intentions themselves meet the conditions of freedom. In order to correct this grave omission, both thinkers adopt Rousseau's idea that individual freedom rests on free will. Subjects are only truly free if they restrict their actions to intentions or aims that are free of any trace of compulsion. But when it comes to how subjects carry out such a purification process, the two thinkers part ways: Whereas Kant proposes that we interpret the free will as the product of rational autonomy, Herder assumes that the purification of the will is a matter of discovering one's own, authentic desires. This opposition between self-determination and self-realization, between autonomy and authenticity, laid down the path followed by the idea of reflexive freedom throughout the philosophical discourse of modernity. The reflexive acts that we must always keep in mind when we speak of individual freedom are to be understood, according to Kant and Herder as well, either as a kind of rational self-restriction or as a diachronic process of self-discovery. However, as the discourse would continue to develop, both of these conceptual models would get pruned back to a more modest scope than these authors had originally intended.

Kant's transcendental concept of autonomy would later be weakened either by empirical reinterpretation or by a intersubjectivist correction of its reflexive achievements. In the first case, what Kant viewed as the rational capacity of noumenal subjects would be interpreted as a bundle of empirical skills; the reflexive acts required for the exercise of individual freedom would be described as the outcome of a process of socialization in which all subjects learn to regard themselves as co-authors of morally valid laws. Such empirically reductionist interpretations of moral autonomy can be found today across an entire spectrum of competing positions; both Freud's moral-psychological speculations²² and Piaget's theoretical investigations on child development²³ are seen as empirically demonstrating how children gradually arrive at an understanding of themselves as morally responsible actors. Furthermore, such reinterpretations of what were once transcendental acts can now be found in moral-philosophical analyses; here we find attempts to prove that nearly existential constraints compel subjects to spontaneously adopt the perspective of moral autonomy.²⁴

Kant's original conception has since been stripped of its transcendental features, not only by having been translated into empirical claims, but also by having been reformulated in intersubjectivist terms. This detranscendentalization undertaken by Karl-Otto Apel and Jürgen Habermas, drawing on Peirce and Mead, locates the moral subject within a communicative community;²⁵ what was previously the act of a solitary, self-referential subject thus came to be interpreted, through a speech-theoretical turn, as the communicative product of the members of a speech community. The argument goes that the normative presuppositions of speech compel individuals to view themselves as participants in a discourse in which each person must respect the autonomy of the others. The idea of reflexive freedom, still an entirely methodological matter for Kant, thus took on an intersubjective significance, which in turn gave this notion of freedom a much stronger foothold in the social structures of the lifeworld. On this view, individuals gain autonomy by being socialized into a communicative community in which they learn to regard themselves as addressees of the universal norms they bring about in cooperation with others. But as we will see later, the expansion of the ‘I’ into the ‘We’ of autonomy is not sufficient for fully grasping the import of the idea of intersubjective freedom, for it entirely ignores the fact that both the ‘I’ and the ‘We’ can only achieve self-determination once institutional relations within social reality offer opportunities to achieve these aims.

The same fate that befell Kant's concept of self-determination by the twentieth century at the latest would also befall Herder's notion of self-realization soon after his death. Step by step, Herder's theory would also become detached from its metaphysical premises and adapted to the intellectual conditions that gradually began to prevail within a more sobered modernity. After Nietzsche and Freud, it would become increasingly difficult to conceive of the process of self-realization as the reflexive liberation of a primitive and – furthermore – natural core of one's personality. In modernity, a person's ‘self’ is assumed to be something that is formed socially; we can withstand these formational processes to a certain extent, but not as a core that already contains the entirety of our individual character. Along with the abandonment of the notion that we have a certain personality core, the idea would soon fade that self-realization is a process of discovery, of finding out the truth about oneself. If there is no primal, ‘true’ self, then self-realization cannot be understood as a process of self-discovery, but only as an essentially constructive process that demands standards other than those of reflection or identity with oneself. The outcome of all of these theoretical restrictions is that the discovery of one's own authentic desires has increasingly been set in opposition to the process of self-realization. The internal link that Herder simply assumed to exist between these two processes was in danger of being torn asunder, because the loss of the premise of an antecedent personality core also entails the loss of any possibility of unifying these two processes. The consequence is that even today, the ideas of authenticity and self-realization remain alienated from each other. Whereas the freedom that consists in acting solely according to one's own true desires is largely interpreted as a one-time act of identification or articulation, the freedom of self-realization is viewed diachronically as the capacity for creating the story of a unified self.

Harry Frankfurt has certainly ventured the furthest when it comes to defining authenticity by assuming a stage-like hierarchy of the human will. Unlike animals, humans are able to view first-order desires from the perspective of a higher-order desire, accepting, rejecting or affirming the former.²⁶ For Frankfurt, even if we act on a desire we judge to be acceptable or worthy from a higher-order perspective, our actions are not yet completely free. What is needed is a separate act of identification, of emotional agreement, in order to turn a desire into a motive for action that can truly be felt as ‘free’.²⁷ The distinction between this idea of authentic freedom and all models of self-realization becomes apparent once we recognize that for Frankfurt, the possibility of completely identifying with a given desire does not depend on biographical continuity. I do not need to be able to grasp a desire which fulfils me entirely as a new stage or component in the process of my personal development in order to act authentically. By contrast, the ideal of self-realization necessarily assumes biographical continuity. As much as the fictional character of such continuity is now emphasized, freedom in the sense of self-realization still must be grasped as the outcome of a reflection on the diachronic whole of our biography.²⁸ What Herder regarded as a unity between authenticity and self-realization has now been torn in two: The acts demanded by reflexive freedom are entirely different from those presupposed in the formation of an authentic will.

Just like the idea of negative freedom, the various concepts of reflexive freedom have led to various specific conceptions of how we must approach the issue of social justice. While the first idea of freedom seemed to present a clear connection between the concepts of freedom and justice, it becomes utterly opaque once we turn to reflexive freedom. The concepts of autonomy and self-realization represent two ideals at odds with each other, whose implicit conceptions of justice can hardly be boiled down to a common denominator. When it comes to moral autonomy, the corresponding relations are relatively apparent: Because individual freedom is interpreted in the Kantian tradition as a kind of self-determination guided by the principle of universal respect, the principles of social justice must be conceivable as the outcome of cooperation among the totality of individual free actors. In essence, the idea of moral autonomy amounts to a procedural conception of justice. The procedure of individual self-determination is transferred to a higher stage of the social order once it is viewed as a shared process of will-formation in which equal citizens deliberate and decide on the principles of what they consider to be a ‘just’ social order. The ‘substantive’ content of such a conception of justice is thus not offered by the theory itself, as the latter restricts itself to determining the procedure of collective will-formation, perhaps along with a few principles that, for reasons of fairness or equal opportunity, precede these procedures.²⁹ It might also name a ‘system’ of individual rights that give shape to the procedures for forming a constitution;³⁰ but otherwise, the concrete definition of justice is the outcome of the procedure of collective self-determination. Just as the idea of negative freedom ultimately leads to a conception of justice that promotes a social system based on individual egotism, the idea of moral autonomy necessitates a procedural conception of justice that serves a social system based on cooperation or democratic deliberation. However, in the second case, the substance of this system is not determined in advance, because for conceptual reasons the theory cannot anticipate decisions that autonomous subjects must make on their own.

As clear as the methodological connections between the idea of freedom and the conception of social justice might be when it comes to self-determination, they become all the more ambiguous once we interpret reflexive freedom in terms of ‘self-realization’ of ‘authenticity’. As we saw above, the modern understanding of justice depends almost exclusively on an idea of individual freedom. If we think of freedom as a reflexive act, and if we interpret this act as a life-long process of self-articulation, the resulting conception of justice will necessarily be a social system in which each subject can pursue self-realization without harming others. In this case, the cooperating subjects play a much smaller role when it comes to determining the actual substance of that just order than in the case of the ideal of autonomy. After all, the theorist has at least a rough idea of the social conditions under which subjects strive for self-realization. Unlike the pluralism that is characteristic of conceptions of justice that define freedom as self-determination, conceptions of justice guided by the ideal of self-realization usually have some substantive content. Though they must not anticipate the aims or the direction of individuals’ self-articulation, they can present external knowledge of the social conditions that individuals require for pursuing their aims.³¹

Yet the conceptions of justice stemming from the ideal of self-realization can be further divided into two sub-classes, as it is possible to grasp the idea that individuals can only attain freedom by articulating their ‘true’ selves in both an individualistic and collectivist manner. In the first case, in which the reflexivity of self-realization is interpreted as the exclusive act of individual subjects, the corresponding conception of justice must have an individualist character: Here a just order is generally understood as a sum of social resources and cultural conditions that allow individual subjects to freely articulate their authentic selves over the course of their lives. The best example of such a conception of justice are without a doubt those parts of John Stuart Mill's writings that do not present a merely negative idea of freedom, but are instead guided by an ideal of self-realization.³² Drawing on Wilhelm von Humboldt, Mill argues that government has the duty to create a social ‘atmosphere of freedom’³³ through appropriate educational measures and by reliably ensuring the pluralism of public opinion – an atmosphere in which the members of society can ‘unfold’ their individual ‘faculties, capacities and susceptibilities’ to the fullest possible extent.³⁴ For Mill the freedom of subjective self-realization – which the state must secure by means of basic educational measures, diversity of opinion and cultural life – is restricted solely by the famous ‘harm principle’.³⁵ Within the limits that are to prevent subjects from violating the equal rights of other subjects, all individuals enjoy what ultimately amounts to a claim, vouched for by the state, to discover their own ‘originality’ and realize it in their own lives.³⁶

Contrary to such individualist notions of self-realization, which amount to no less individualist conceptions of social justice, collectivist approaches grasp the achievement of self-realization as an eminently communal, cooperative endeavour.³⁷ According to this view, individuals cannot achieve self-realization on their own, because their authentic self is so much an expression of a social community that it can only be unfolded in collective action. Therefore, the notion of freedom presupposed here is the outcome of a reflexive act that can only be performed by a collective. The conception of justice to which this notion of self-realization leads can take on various forms, but all share the methodological necessity of viewing a desirable social order as one that embodies the actions in which subjects realize the aims they have in common. The democratic version of this conception of justice is represented by liberal republicanism; according to this view, advocated by Hannah Arendt or, to a lesser extent, Michael Sandel,³⁸ members of society come together to discuss and publicly negotiate their common affairs, such that intersubjective debate in the public sphere must be grasped as a collective form of self-realization. The connection between this notion of self-realization and a corresponding conception of justice lies in the fact that given institutional arrangements are examined in terms of whether they can preserve the necessary solidarity among the citizenry. What counts as ‘just’ is ultimately whatever is capable of promoting social attitudes of solidarity, a necessary prerequisite for shared activity in the public sphere. The substance of this abstract idea of justice depends on what we regard as necessary for guaranteeing the social integration of the political community; the spectrum of possibilities runs from social-egalitarian approaches that urge the social inclusion of all citizens to forms of political elitism that can also occasionally be found in Hannah Arendt's work.³⁹

It is extremely difficult to determine whether this version of reflexive freedom that focuses entirely on the periodic, momentary identification with one's own desires can produce an independent idea of justice; after all, there are many indications that the concept of authenticity leads to the same model of a just order that we saw in the individualist notion of self-realization. Even if we interpret self-realization not as a continuous process but as a discontinuous series of acts of identification, the decisive criterion for judging whether a given social order is just or not must be whether the members of society are given sufficient space and resources to perform such acts. On the whole, however, this conception of authenticity is not comprehensive enough to produce an independent idea of justice. Therefore, we are probably justified in stating that this conception of freedom is neutral, or rather, indifferent to matters of justice.⁴⁰

As we have seen in this rough overview, it is not easy to find a common denominator for the conceptions of justice that go together with the idea of reflexive freedom. It is true that they all differ from the conception of justice linked to the idea of negative freedom by focusing on cooperation rather than a social system founded on individual self-interest. The degree to which subjects must cooperate in order to achieve the social conditions needed to realize reflexive freedom is significantly higher than in the case of negative freedom. Beyond this rather formal commonality, however, we find a variety of differences related to the fact that reflexive freedom can be understood in terms of both autonomy and self-realization. And depending on which model we apply, the basic institutions that make up a just order, that is, the institutions intended to guarantee the realization of freedom, are characterized in very different ways. Nevertheless, the methodological procedure for determining the corresponding conceptions of justice remains the same: On the basis of reflexive freedom – whether as self-determination or as self-realization – we deduce ideas about which institutional circumstances are needed to guarantee that all individuals can realize either notion of freedom.

Therefore, neither of these two models of reflexive freedom interpret the social conditions that enable the exercise of freedom as elements of freedom itself. Instead, these conditions do not come into view until the issue of a just order is raised, and thus the social chances for realizing these prerequisites. So in essence, ideas of reflexive freedom stop short of the conditions that enable the exercise of freedom in the first place; they artificially bracket out the institutional circumstances and forms that are crucial for the successful completion of the process of reflection. After all, a key element of self-determination consists in the institutional availability of moral aims; and part of self-realization is the actual availability of the goods required for realizing our desires. In both cases, social circumstances only come into play once the exercise of freedom has already been defined; they are then added externally, as elements of social justice, but not as an inherent aspect of the exercise of freedom. An exception to this ‘after-the-fact’ logic is the discourse theory of reflexive freedom: Because the act of reflection is tied to participation in discourse, the latter cannot be interpreted as a merely external extension of freedom, but must be viewed as one of its core elements. Such an institutional expansion of the concept of freedom serves as the criterion for the third, social concept of freedom, according to which the idea of reflexive freedom cannot be realized without taking into account the institutional forms that enable its realization.

1 On the continuity of this idea going all the way back to Aristotle, see Ernst Tugendhat, ‘Der Begriff der Willensfreiheit’, in Philosophische Aufsätze (Frankfurt: Suhrkamp, 1992), pp. 334–51.

2 Albrecht Wellmer in particular has emphasized the irreplaceability of negative freedom in ‘Models of Freedom in the Modern World’, pp. 3ff.

4 Raymond Geuss, ‘Auffassungen der Freiheit’, Zeitschrift für philosophische Forschung, 49, 1 (1995): 1–14.

5 Jean-Jacques Rousseau, Émile, or On Education (New York: Basic Books, 1979), pp. 266–313; see Schneewind, The Invention of Autonomy, ch. 21, pp. 474–7.

6 Jean-Jacques Rousseau, The Social Contract and Other Later Political Writings (Cambridge: Cambridge University Press, 1997), pp. 3–150. An extremely convincing interpretation of the role of self-imposed law in ‘The Social Contract’ can be found in Frederick Neuhouser, Rousseau's Theory of Self-Love: Evil, Rationality, and the Drive for Recognition (Oxford: Oxford University Press, 2008), pp. 214–17.

10 On the influence of Rousseau on Kant's notion of moral autonomy, see Schneewind, The Invention of Autonomy, pp. 487–92; Susan Meld Shell, Kant and the Limits of Autonomy (Cambridge, MA: Harvard University Press, 2009), ch. 2.

11 On the literary influence of Rousseau's ideal of authenticity, see Lionel Trilling, Sincerity and Authenticity (Cambridge, MA: Harvard University Press, 1972), pp. 58–67; on its philosophical influence, see Christoph Menke, Tragödie im Sittlichen: Gerechtigkeit und Freiheit nach Hegel (Frankfurt/Main: Suhrkamp, 1996), ch. 4c.

12 Immanuel Kant, Groundwork of the Metaphysic of Morals, trans. H. J. Paton (New York: Harper and Row, 1964).

16 On the ideal of authenticity in Rousseau's Julie, or the New Heloise, see Alessandro Ferrara, Modernity and Authenticity: A Study of the Social and Ethical Thought of Jean-Jacques Rousseau (Albany: SUNY Press, 1993), ch. 5.

17 See Menke, Tragödie im Sittlichen, ch. 4; Taylor, The Ethics of Authenticity, ch. 3.

18 Johann Gottfried Herder, ‘On the Cognition and Sensation of the Human Soul’ in Philosophical Writings (Cambridge: Cambridge University Press, 2002), pp. 187–244; Christoph Menke has published an impressive reinterpretation of this work in Kraft: Ein Grundbegriff ästhetischer Anthropologie (Frankfurt/Main: Suhrkamp, 2008), ch. III; on the relation of Herder's work to Rousseau, the best text remains that by Hermann A. Korff, Geist der Goethezeit: Versuch einer ideellen Entwicklung der klassisch-romantischen Literaturgeschichte (Leipzig: Koehler and Amelang VOB, 1923), vol. 1, part I, ch. 1.2.

22 John Deigh, The Sources of Moral Agency: Essays in Moral Psychology and Freudian Theory (Cambridge: Cambridge University Press, 1996); David Velleman, Self to Self: Selected Essays (Cambridge: Cambridge University Press, 1996), esp. chs 5, 6 and 12.

23 Picking up on Jean Piaget's groundbreaking study The Moral Judgment of the Child (1932) (New York: Free Press, 1997), Lawrence Kohlberg in particular has presented Kantian-inspired empirical studies on moral development: The Psychology of Moral Development: The Nature and Validity of Moral Stages (Essays on Moral Development, Volume 2) (New York: Harper and Row, 1984).

24 Christine M. Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996).

25 Karl-Otto Apel, Towards a Transformation of Philosophy (Marquette: Marquette University Press, 1998); Jürgen Habermas, Moral Consciousness and Communicative Action (Cambridge, MA: MIT Press, 1990), chs 3 and 4.

26 Harry Frankfurt, ‘Freedom of the Will and the Concept of a Person’, in The Importance of What We Care About (Cambridge: Cambridge University Press, 1988), pp. 11–25.