I. Honneth’s recognition theoretical approach to justice

Based on the descriptive value and the normative pull Axel Honneth ascribes to ‘recognition’, he proposes a ‘recognition theoretical approach to justice’, where justice is understood in terms of recognizing individuals’ needs and legal and social status in order to allow individuals to form identities and be able to achieve self-realization. Moral progress consists in increasing ‘the quality of social integration’ by widening the scope of recognition in line with normative demands of legitimate recognition claims (Fraser and Honneth 2003: 175).

On the descriptive level Honneth holds that social and political movements and conflicts can best be explained in terms of an experience of ‘moral injury’ caused by withheld (perceived-to-be legitimate) recognition. Individuals are motivated to act by the experience of harm and the goal of social and political action is to gain recognition (of needs, rights or achievements).Footnote ³ Social and political dissatisfaction and conflict arise when one party feels that their needs, legal status or achievements are not recognised appropriately. Social conflicts themselves are really conflicts either about the standards of recognition, which have developed within certain contexts and histories (and hence are the outcome of cultural historical development) or about the application of the accepted grounds for recognition.

Honneth proposes a complex system of recognition. Reminiscent of Hegel’s distinction between family, state and civil society, he distinguishes three different spheres of post-traditional ethical life with their own respective forms of recognition: the sphere of personal relations, the sphere of legal relations and the sphere of social relations. Corresponding to these spheres are different expressions of the principle of recognition: in the sphere of personal relations, which is constituted by our intimate relationships, familial relations and friendships, recognition consists in love, expressed in care for our needs; in the legal sphere recognition consists in respect for our legal rights and duties and the principle of equality; the form of recognition in the social sphere is esteem for our achievements. The three spheres, while distinct, interact in complex ways, e.g., misrecognition on the level of achievement can lead to legal misrecognition (and vice versa), misrecognition in the sphere of love can – at least indirectly – affect recognition relations in the two other spheres, etc.Footnote ⁴

The experience of harm itself depends on the break-down of relations of mutual recognition. These relations are important for our development and maintenance of a sense of self and our ability to achieve self-realization. Honneth draws on a variety of sources to make his point here. Starting with the German Idealist view of subject-formation, where the subject only becomes aware of him- or herself as subject by being recognized as such by other subjects, Honneth finds empirical support for this position in developmental psychology and object-relations psychoanalytic theories. With reference to object relations psychoanalytic approaches, Honneth distinguishes three forms of ‘self-relation’ which are – to some degree – determined by relations of mutual recognition and which correspond to the different spheres of interaction: in his or her relation to primary caregivers and early peers, the child learns to develop self-confidence (‘Selbstvertrauen’), understood as the ability ‘to learn to regard one’s internally experienced needs and desires as a part of one’s person that can be articulated’ (Honneth Reference Honneth2010: 265).Footnote ⁵ Here learning takes the form of experiencing the positive response to one’s needs by one’s caregivers. Next, the growing child, in regular interaction with peers slowly also develops self-respect (‘Selbstachtung’), i.e., the child learns that he or she is capable of forming opinions and judgements and that these are regarded as valuable and reliable by others, so the child experiences him or herself as accountable (‘zurechnungsfähig’). The third kind of self-relation can develop concurrently with the second and is a sense of self-worth (‘Selbstwertgefühl’), where the child learns that he or she has original and valuable talents. Once developed, however, these attitudes towards self are not stable and require constant confirmation through the positive attitudes of others, i.e., recognition. If the development or maintenance of those self-relations fails, the individual is diminished in his or her ability to achieve self-realization. If, on the other hand, those attitudes towards oneself are developed. As a result, the individual is able to autonomously and authentically participate in the social world and thus create and apply the conditions necessary for self-realization.

The descriptive value of Honneth’s approach in the context of subject formation and maintenance is controversial. Lois McNay (Reference McNay2008) accuses recognition approaches in general of conflating the normative and descriptive aspects of recognition and thereby giving a warped, oversimplified descriptive account of subject formation, leading to an equally warped view of subjects as autonomous, which in turn is desirable for the normative dimension of recognition theories. In other words, the normative role recognition is thought to play directs the descriptive account provided. She identifies recognition approaches’ emphasis (sometimes even exclusive focus) on face to face interaction as a key problem. In contrast, McNay holds that subject formation, development and maintenance of a sense of self, self-respect, self-confidence and self-esteem take place against the backdrop of impersonal and institutionalised power structures and hierarchies (structures of oppression), the force and pervasiveness of which recognition approaches do not take into account.Footnote ⁶ McNay’s concerns constitute, in large part, the descriptive challenge and so will be discussed in more detailed below. At this point though, it should be noted that Honneth does allow for the impact of impersonal structures of oppression: while he does not explicitly address the impersonal mechanisms involved in subject formation in the detail McNay offers, as shown below he acknowledges these factors regularly and his discussion of ‘decentred autonomy’ only makes sense in the light of a conception of subjects that does appreciate the effects of pervasive and opaque power structures on subject formation.

As indicated above, Honneth also sees a normative value in recognition. Partly, this value rests on the contribution recognition makes to the sense of self and the severe moral injuries experienced by individuals who do not receive (legitimate) recognition. However, severe injury is not automatically a ‘moral’ injury, and so Honneth requires something more than a mere descriptive account of the importance of recognition in order to make the normative case. Consequently, Honneth explicitly commits to an Aristotelian ethics, where self-realization takes centre stage (Fraser and Honneth 2003: 177).Footnote ⁷ Self-realization (like Aristotle’s eudaimonia) is, on the one hand, to be regarded as the fundamental moral value and, on the other hand, Honneth believes that the desire of individuals to self-realize functions as an explanation of ‘moral’ motivation, or it explains the motivation of social movement or social struggles in general (and in this way, self-realization also possesses normative and descriptive value). Within this normative system built on the value of self-realization, recognition is only morally relevant in as far as it pertains to self-realization.Footnote ⁸

II. Decentred autonomy and ‘authenticity’

In light of the normative system described above, it seems that ‘authenticity’ and ‘autonomy’ must play important roles in Honneth’s theory of recognition. Recognition claims derive their normative value from their relation to individuals’ ability to self-realize. In order for recognition claims to protect and promote self-realization, these recognition claims should pertain to issues relevant to self-realization, i.e., to our needs, desires, life-plans.Footnote ⁹ Usually we hold autonomous and authentic agents to be capable of knowing and realising their own needs, desires and life plans. In other words, recognition claims can only have normative value if they are authentic or autonomous or promote or protect authenticity and autonomy.Footnote ¹⁰ This does not mean that all authentic or autonomous recognition claims have (equal) normative value; it is possible for recognition claims to be authentic but morally obnoxiousFootnote ¹¹ so other moral principles have to be employed to settle the normative value of recognition claims and adjudicate between conflicting claims; rather, due to the relation between autonomy and authenticity to self-realization, recognition claims must be authentic or autonomous or promote or protect authenticity and autonomy as a prerequisite for them having normative value at all.Footnote ¹²

Honneth identifies three criteria of classical autonomy, which need to be revised because empirical evidence and new theoretical developments suggest them as unattainable: transparency of self, control over self and linguistic independence. Honneth argues that, especially in light of psychoanalytic theories (and evidence), the self can no longer be plausibly regarded as self-transparent; rather, motives for action are often hidden from subjects’ consciousness. This opacity of motives reduces the control subjects have over themselves. Further, subjects depend for the articulation and exploration of motives, beliefs, etc., on a language system with pregiven meanings, thus diminishing how occurrences (in the broadest sense) can be expressed and interpreted. In this way, Honneth argues, the self-’authorship’ of individuals is affected (Honneth 2007a).

The challenges thus presented to the notion of self-determinacy in the classical notion of ‘autonomy’ can be met in one of three ways: by abandoning ‘autonomy’ as attainable at all, by taking a neo-idealist position and clinging to the classical ideal of autonomy in the face of evidence to the contrary, or, lastly, by reconceiving the self and autonomy in light of these challenges. In ‘Decentred Autonomy’ Honneth proposes a new conception of autonomy. Rather than self-transparency, decentred autonomy requires increasing (and never-ending) self-disclosure through the articulation ‘without fear’ of unconscious impulses, needs and desires (Honneth 2007a: 189). This in turn renders the autonomous subject dependent on others in two ways: first, there needs to be a growing linguistic scope to enable articulation; and second, based on evidence from developmental psychology, Honneth holds that the required articulation and self-discovery can only take place within a context of ‘permanence of care shown…by concrete other…’(ibid.). Because transparency has given way to disclosure and articulation without fear, it is no longer possible to exert self-control by hierarchically ordering one’s motives and desires along a chosen principle (as this may preclude the possibility of new disclosure); rather than a hierarchical organisation of one’s needs, desires and motives, decentred autonomy merely requires that one can construct a coherent narrative, able to integrate ‘disparate parts’(ibid.: 190), where this narrative is ethically reflective, i.e., it is coherent from the position of meta-wishes, or second order desires the agent reflectively embraces (where these values cannot be reduced to one single value). Lastly, Honneth argues that the classical aspect of moral autonomy, the application of universalizable moral principles, has to be modified in light of the constant influence of social pressure on subjects: the decentred-autonomous agent is always accompanied by feelings of guilt, arising from an inability to fulfil all the various social and moral demands and will thus become sensitive to ‘the concrete predicaments and neediness which other subjects have to struggle with in their lives.’(ibid.: 191). As a result, instead of universalizability, the autonomous agent will be ‘context sensitive’ in the application of moral principles.

So, the decentred-autonomous agent, that is the agent who is autonomous ‘under conditions in which the psyche is in principle not wholly under our control’ (ibid.) is ‘in a position to disclose needs creatively, to present his or her entire life in an ethically reflected way, and to apply universalist norms in a context-sensitive manner…’ (ibid.).

Especially from the perspective of moral autonomy, however, Honneth’s proposal seems peculiar. The ethically self-reflective narrative subject, combined with one or other variation on the ‘superego’ seem to suggest a new notion of autonomy that has lost the very core of the value of classical autonomy.Footnote ¹³ But, here I am interested in autonomy only in as far as it relates to self-realization. For this issue, what seems to protect the capacity to self-realize (i.e., that aspect of autonomy concerned with authenticity) is the notion of the disclosure of unconscious aspects of self, through articulation without fear (and the notion of reflective ethical narrative).

Before turning to a discussion of Honneth’s view of decentred autonomy, I want to give a brief outline of Honneth’s position on ‘authenticity’ or ‘self-relation’.Footnote ¹⁴ In Reification (Honneth Reference Honneth2005) he proposes a view of authenticity as ‘correct self-relation’. The term ‘reification’ is borrowed from Lukács to describe one of the ‘pathologies’ of modernity, a practical attitude we have towards the world and ourselves that deviates from correct praxis.Footnote ¹⁵ But, whereas for Lukács reification consists in every form of ‘objectification’ or ‘de-personalization’,Footnote ¹⁶ Honneth argues that some forms of objectification are necessary. To him only those instances of ‘objectification’ or ‘de-personalization’ which involve a ‘forgetting of [due] recognition’ are properly regarded as instances of ‘reification’. Following Lukács, Honneth identifies three spheres of reification: the objective world, the intersubjective world and the intrasubjective realm. For our purposes only the last is relevant, because it is in contrast to the reification of self that we find the notion of proper self-relation or ‘authenticity’.Footnote ¹⁷ Self-reification is also an objectification combined with forgetting (i.e., inability to realise the objectification). The self can be objectified in two ways: first, one can take a view of oneself as the object of one’s knowledge, passive, already given, something to be discovered or detected; second, one can take the opposite view and regard oneself as the agent of all of one’s character. In the first case the self becomes a passive object of detection, in the second case the self becomes the object of creation. Honneth contrasts self-reification with proper self-relation, what he calls ‘expressionism’:

‘We neither perceive our mental states simply as objects, nor do we constitute them through our articulation, but we articulate them in accordance with that with which we are internally already familiar…’ (Honneth Reference Honneth2005: 88).Footnote ¹⁸ We are already familiar with our feelings and desires because our mental states are given (not created through our articulation or consciousness of them) and we have learned to regard them – during the process of socialisation – as part of who we are and, moreover, as a part of who we are that we communicate to and share with others. Mental states may strike us as unfamiliar, if they are genuinely new states or if they go beyond our linguistic scope. Proper self-relation means that we relate these unfamiliar states back to familiar states and thus expand the horizons of familiarity (ibid.). So, against constructivist theories of self, Honneth’s expressionism holds that desires and feelings are given. We are familiar with them. But familiarity is not the same as knowledge. We may be familiar with mental states that are diffuse and not a ‘proper object of knowledge’ (ibid.: 86). Moreover, we do not just have a neutral epistemological interest in them. In order to ‘articulate’ our mental states, Honneth maintains, we must first have a positive attitude towards these states. An attitude of ‘care’ or ‘recognition’.Footnote ¹⁹ We must regard our mental states as worthy of articulation. Authenticity thus consists of a recognition of one’s own feelings, needs and desires where recognition is an attitude of affirmation of value and care which prompts articulation. The authentic agent, in Honneth, is one who recognizes him or herself, who practices self-recognition.Footnote ²⁰

III. Authenticity, autonomy and the value of recognition

With Honneth’s account of authenticity and inauthenticity and his notion of decentred autonomy in mind, I want to return to the roles those concepts must play within his recognition theoretical approach to justice.

Autonomy and authenticity are (assumed to be) bearers of normative value because of their (putative) relation to self-realization. Recognition claims can only be considered to have moral value if they are related to autonomy and/or authenticity in specific ways: recognition claims must be expressions of autonomous and authentic agents and/or they must promote or protect authenticity and/or autonomy. So, autonomy and authenticity seem to be conditions of self-realization (it is only through autonomous and authentic recognition claims that self-realization and the circumstances of self-realization can be achieved). However, at times, it seems that self-realization is constituted by autonomy and authenticity (see esp. Honneth Reference Honneth1994 and 2011), so only those subjects who have achieved autonomy and authenticity are self-realized, and once they have achieved autonomy and authenticity they (where achieving autonomy and authenticity are on-going processes). While there is a difference between the conditional and the constitutive accounts, on both accounts autonomy and authenticity carry the normative weight. While autonomy and authenticity play an essential normative role, they must also be descriptively accurate or attainable. If we recall, one of the problems with Honneth’s recognition approach is that it supposedly misrepresents subject-formation in such a way as to give rise to the belief that individuals can attain full autonomy and authenticity, such that their recognition claims are normatively reliable (or such that they can fully participate in the social world as to achieve self-fulfilment), but that in reality, due to the impact of oppressive power structures, individuals cannot attain full autonomy or authenticity (hence there can be no recognition claims issued by autonomous or authentic agents, though there might still be recognition claims that promote autonomy accidentally).Footnote ²¹

It seems that what I call a ‘weak’ reading of decentred autonomy and expressionist self-relation can successfully meet the descriptive challenge. On the weak reading individuals are autonomous if they are capable of self-discovery, articulation and evaluative reconciliation, irrespective of the source of aspects about themselves they discover or the impact social structures have on the way individuals reconcile unconscious aspects and social expectations. They are authentic if they recognize and positively affirm their needs and desires, irrespective of the source of these needs and desires. In this sense autonomy and authenticity are achievable even in the context of permanent and pervasive influence of impersonal structures and dynamics of oppression on subject formation. Decentred autonomy and Honneth’s conception of authenticity are cut to fit a view of the self, which is opaque to itself, fragmented and unable to exert complete self-control.

However, in so far as Honneth succeeds on the descriptive level, his decentred autonomy and expressive self-relation fail on the normative level. Autonomy and authenticity have been transformed into concepts that no longer obviously carry the same value as classical autonomy and authenticity, and they no longer relate to self-realization in the required way (except if self-realization is to be re-conceived in light of a ‘damaged’ sense of self, in which case self-realization would lose at least some of its normative value).

In order for authenticity or expressive self-relation to be relevant for a morally significant notion of self-realization, we would have to give what I call a ‘strong’ reading of Honneth’s account: authenticity consists of our recognition (including our affirmation) of our true needs, feelings and desires, where ‘true’ here emphasizes needs, desires and feelings that are not the result of external structures of oppression, which are all pervasive in our lives. However, this interpretation is not equipped to respond to the descriptive challenge. In other words: if we read Honneth’s account in a way that meets the descriptive challenge (weak reading), the authentic agent recognizes (and values) his or her needs and desires, even though at least some of these desires will inevitably be the result of pervasive, impersonal, institutionalised powers of oppression and are part of the reproduction and maintenance of these power structures (which are at their very core obstacles to self-realization). A strong reading of Honneth’s conception of authenticity, in turn, would exclude needs and desires that are the result of oppressive power structures, but would fail to be practically attainable in the light of the pervasiveness of these power structures. In so far as authenticity is practically unattainable, recognition claims do not originate with authentic agents, and if they promote or protect potential authenticity then they will do so accidentally.Footnote ²²

The notion of decentred autonomy does not fare any better, and for similar reasons. As Honneth admits, the notion of disclosure of unconscious motivation through articulation without fear depends on others, not merely for their continued care and support (in the sphere of love) but also for the availability of a language in which these needs, motivation and desires can be articulated. But language is a prime example of oppressive power structures at work (see McNay Reference McNay2008: 69-74). Decentred autonomy relates to expressive self-relation in a similar way as autonomy relates to authenticity in Taylor: decentred autonomy is the potential we have to discover our needs, desires and motivations within the constrictions of oppressive structures, which are what classically autonomous agents seek to overcome. At first we might think that a strong reading of decentred autonomy is possible, after all, Honneth’s account of decentred autonomy contains a notion of reconciliation between the unconscious and conscious aspects of self. The decentred autonomous agent not only self-discloses unconscious aspects, but makes ‘reflective decisions’ (Honneth 2007a: 189) about them. This seems to allow some space for at least the possibility to overcome the effects of power structures embedded in language and similar external, oppressive influences on self-disclosure. The ideally autonomous agent is free from unconscious motives which he or she has rendered conscious and rationally judged. However, Honneth himself insists on the fact, that self-disclosure is an ‘unending’ process (ibid.), as such decentred autonomy is never fully achieved. Moreover, given our dependence on language and others in the process of self-disclosure, it seems that there is no guarantee that those motives we take ourselves to self-disclose are actually accurately identified. If we misidentify our unconscious motives (or mis-ascribe), our rational judgements of them are possibly irrelevant (as they do not relate to our ‘true motives’) and do not help to overcome oppressive structures. Honneth himself seems to suggest that ‘due to the impact of poetic innovations’ (ibid.), our language has sufficient scope to express previously inexpressible unconscious impulses. This however, is not satisfactory. It may well be the case that as language develops the expressive scope widens (though some would probably insist that the opposite is the case), but that is not to say that we can now articulate and differentiate between different unconscious motives reliably, i.e., that the scope is such that all our unconscious features can be expressed or detected or that we (and the others we rely on) approach self-disclosure in a sufficiently unbiased manner. Lastly, if Honneth takes McNay’s worries seriously (and hence answers to her descriptive challenge) he possibly would have to take into account the effect of pervasive impersonal power structures on our rational judgements themselves. Overall, however, the fact that full self-disclosure is regarded as unachievable, combined with the dual dependence on others suffice to rule out a strong reading of decentred autonomy.Footnote ²³

Above arguments leave us with two possibilities regarding the relation between autonomy, authenticity and self-realization. On the one hand, if we conceive of self-realization in terms of decentred autonomy and proper self-relation, thus assuming a strong link between the three, self-realization itself seems to lose moral value. Here, self-realization would consist in the articulation and affirmation of needs, desires and motives, which may be the result of oppressive power structures (and perpetuate these structures). That is, self-realization here is not in any obvious way opposed to oppressive structures. Alternatively, we could maintain a view of self-realization which relates to the actualization of individuals’ true desires, choices, motives, etc. These true choices would be threatened by oppressive structures and hence self-realization is opposed to such structures and retains moral value. But this conception of self-realization is no longer intimately linked to notions of decentred autonomy and proper self-realization, i.e., it is not clear why we should assume that recognition claims issued by agents who approximate decentred autonomy and have proper self-relation promote or protect this strong sense of self-realization. This is problematic because on this account we can ignore recognition claims by subjects and just refer to our theory of objective true needs and desires.

Overall, in order for self-realization and notions deriving value from the ideal of self-realization to have normative value they need to refer to some sense of self that is opposed to that self whose needs, motivations and desires are the outcome and reproduction of those influences thought to hinder individual freedom and development. What is required is either a notion of a ‘true self’, or a more substantive notion of human well-being (which is to be realized and the realization of which requires our standing in relations of mutual recognition). The notion of a ‘true self’ – on the practical level – seems to fall victim to the descriptive challenge. Worse than that, however, it is implausible in the context of a recognition approach. Here the self is the outcome of interpersonal relations, which is what makes recognition so valuable and the lack thereof so harmful. It would appear that the only option left for Honneth is to propose an account of human well-being that does not rely on notions of self-realization and recognition but itself provides a normative grounding for the value of both.

At this point one might object: it does not follow from the unattainability of strong notions of decentred autonomy and proper self-relation that Honneth’s normative framework is in trouble. Moral theories can appeal to an ideal as normative standard which can never be actually obtained. The reply here basically accuses above criticism of a conflation of the normative and the empirical. After all, Honneth has in place a theoretical framework of judging recognition claims by asking a) whether the claim promotes emancipatory self-realization (even if such self-realisation can never be complete) and b) whether the claim increases or at least does not diminish the scope of recognition (i.e., the number of people who receive legitimate recognition). So, it offers mechanisms to distinguish between genuine, normatively valuable recognition claims and morally obnoxious, oppressive recognition claims. The question of whether it is practically possible to move towards a society in which autonomy and authenticity can be fully realised is a separate question and not one Honneth has to answer.

However, if this response is aimed to defend the Honnethian account that relies on decentred autonomy and authenticity and resulting (weak) notion of self-realization (short: the weak reading), rather than an account supplemented by an objective account of genuine self-realization or human well-being, it misses two important points:

First, on the weak reading it is not even obvious that we theoretically ought to value decentred autonomy and proper self-relation and the resulting form of self-realization. So that the weak reading does not even theoretically serve as a standard of evaluating recognition claims.

Second, the weak reading does not even theoretically serve as a means to distinguish legitimate from illegitimate claims. Let me illustrate this with an example: consider the relatively new phenomenon sociologists have called the ‘precariat’. The precariat is a social group consisting of people from various socio-economic and educational backgrounds who all find themselves in a (long-term) situation of insecure working conditions (temporary contracts, often low pay, etc.). We can imagine two members of the precariat belonging to two different social movements, Karl and Robert. Karl’s movement aims at changing employment policies which have led to the creation and spread of the precariat: they want to abolish the precariat (or rather the insecure work conditions). Karl can make his argument along Honnethian lines: members of the precariat are withheld their due recognition on the level of social relations because their contribution to the market is not adequately valued, moreover they are prevented from participating in certain social practices (they cannot get mortgages, become house owners, plan a holiday, contribute to private pension funds, etc.). The specifics of the work situation also diminishes the opportunity of members of the precariat to engage in self-sustaining relations of mutual recognition in the private sphere as they have no sufficient stability (e.g., they may have to relocate often due to changing work contracts) or leisure time to build and maintain intimate relations or families.

Robert belongs to a social movement that wants to abolish rigid work conditions. For him the precariat is not insecure but flexible and his movement aims at changing the policies so as to increase the number of people who are thus flexible. Robert also argues his point along Honnethian lines: increased flexibility to him is an increased freedom; people are forced out of their comfort zones and must thus discover new aspects about themselves, new confidence, new strengths, and new desires. Precarious employment contracts also remove traditional class boundaries (as the precariat consists of people from various backgrounds) and increase social solidarity, thus broadening the scope of relations of genuine mutual recognition amongst employees. Moreover, such contracts contribute to market developments that allow everybody to satisfy various needs and desires quickly and cheaply. Robert and his fellow fighters also experience suffering, they suffer from the prejudice of those who ascribe to the dogma of ‘secure work’ and ‘planning’, which leads to misrecognition of their freedom (as insecurity); traditional, rigid views on work life threaten new found solidarity, undermine new relations of mutual recognition and endanger the kind of market progress that allows fulfilment of needs and desires.Footnote ²⁴ Importantly we can easily imagine Karl and Robert (and the fellow members of the respective social movements) to satisfy the criteria of decentred autonomy and proper self-relation (weak reading).

Honneth has two ways of responding. He can try to adjudicate between the competing claims (i) with reference to the standards mentioned above, i.e., his notion of ‘moral progress’, emancipation and genuine appeal to recognition principles or (ii) he can adapt a version of his ‘recognition as ideology’.

(i) As mentioned earlier, Honneth has a framework that allows him to evaluate recognition claims by examining, whether they overall aim to widen the scope of those who can be legitimately recognized (moral progress) and whether they promote emancipation and refer to one of the principles of recognition. Applied to Karl and Robert, it seems both of them aim at widening the ‘recognition scope’ (as opposed to racist or other obviously morally obnoxious claims, who aim to exclude people from relations of mutual recognition) according to their respective views on self-realization; equally, according to their views their claims promote emancipation and refer to the recognition principle.

(ii) In ‘Recognition as Ideology’ and Das Ich im Wir (Honneth 2007b, 2010)Footnote ²⁵ Honneth considers a slightly different problem: the possibility of acts of recognition as ideology, i.e., as a means to encourage individuals to internalize, submit to and perpetuate oppressive power structures. He puts forward criteria by which we distinguish genuine (institutional) acts of recognition from acts of recognition as ideology. Genuine recognition consists of acknowledging and positively appreciating the value of newly discovered aspects of individuals. This positive appreciation is not merely cognitive but should also bring about a modification in behaviour. The newly valued aspect, should be aspects the recognized individuals themselves can evaluatively affirm (i.e., aspects that the individuals themselves judge to be positive and do not experience as restrictive). The evaluative aspect of recognition must involve structural changes in society so as to ‘ensure the material prerequisites for realizing [these] new evaluative qualities…’.Footnote ²⁶ Recognition as ideology mimics the evaluative aspect but lacks this material component. For various reasons this material criterion is problematic as it standsFootnote ²⁷ , but we can modify it slightly to include a preparedness to change material structures. Applied to the problem at hand we could say that those recognition claims that merely express an internalization of recognition as ideology are illegitimate. So, we would now try to claim that Robert’s (or Karl’s) recognition claims are actually merely internalized forms of recognition as ideology, e.g., Robert has fallen for the neo-liberal rhetoric, which passes off a socially disadvantageous positions as ‘new opportunity’, but which actually only aims to encourage individuals to submit to and help to reproduce oppressive power structures rather than change structures so as to materially manifest the value of these ‘new opportunities’. We are allowed to make this judgment because genuine recognition of the ‘precariat’ as a valuable social group would come with a change to the current (institutional) structures, so as to match the new value with material resources (for example). But as it is, the ‘recognition’ of the precariat lacks any impetus for structural change in that way. This seems an odd reply to Robert, who makes no claims about the material structures in such a way: what he values about the precariat is (structurally) intrinsic to the precariat itself, e.g., flexibility. Further, his claims grow out of a willingness to change material structures (e.g., loosen rigid labour regulations). So it seems the internalization of ideology does not offer a solution, except if we have a more objective account of the structural changes that we ought to aim at.Footnote ²⁸

It is clear from Honneth’s writings on the importance of employment and the role it plays in social esteem (Honneth 2007a: 75-77) and his discussion of the employment market (Honneth Reference Honneth2011: 457-458), whose claims he would judge to be genuine recognition claims. Such judgements, interestingly, refer implicitly to objective views of the human good. What Honneth ultimately judges is the ‘truth tracking’ of Karl’s and Robert’s claims regarding true self-realization. Honneth claims that self-realisation requires stability in the form of reliable work conditions, including clear separation of work-time and leisure time, etc., (see Honneth Reference Honneth2011: 457): he does not refer to the demands made by decentredly autonomous agents or requirements of decentred autonomy. Once Honneth has in place an account of genuine self-realization (or ‘true needs’ or human well-being) he can claim that Robert acts from ‘false consciousness’, he is hindered in his emancipation without even realising it. Honneth can now also make claims about Robert’s proper self-relation (which he has achieved to a somewhat lesser degree than Karl). Honneth could possibly now adopt a strong reading of decentred autonomy and conclude that Robert and his friends do not really engage in rational reflection on desires, but are swayed by advertisements and industry rhetoric. But all of these judgements stem from objective, substantial conceptions of the good and they are made irrespective of any specific recognition claims that are made by actual subjects.

In some sense, Honneth is comfortable with this position. He regards it as one of the strengths of his account that he does not depend on the articulation of the experience of injustice or recognition claims, that his account can also deal with ‘silent injustices’, those violations that people suffer from either unknown to them or (at this point) nexpressible for them. Moreover, some some recognition claims are illegitimate they have to be overridden. However, the account of ‘true needs’ or genuine self-realization (or emancipation) that Honneth has to provide (and implicitly assumes in much of his discussions) is more detailed, further removed from the realm of experiences and more objectivist than the one he takes himself to be committed to.

This should trouble Honneth slightly. First, there is a certain tension between fleshing out a more detailed account of true needs and human well-being and the place he gives to discursive practices in discovering needs. When he discusses possible remedies to his analysis of contemporary consumerism (and the powerlessness of consumers), a ‘discursive mechanism’ is not only necessary, it seems, to mobilise people to act (and become an effective social movement) but also to establish (discover through discourse) exactly what needs ought to be satisfied (and can be satisfied) by what form of consumerism (see, for example, Honneth Reference Honneth2011: 400-410). But if he provides a conception of human well-being that is detailed enough to distinguish legitimate from illegitimate claims, it seems he must already have decided which needs must be fulfilled, and which should be ignored (because their fulfilment perpetuates oppression).Footnote ²⁹

Moreover, any account he gives will be provisional. Being authentic and autonomous in the weak sense himself – like all of us – and thus having a diminished capacity for distinguishing between true and false needs or true emancipation and oppression perpetrating self-realization. Again, this is a price that he can pay, but it would put him in the very position he claims Fraser is in with regards to the norms that govern participatory parity (Fraser and Honneth 2003).

Lastly, the normative aspect of Honneth’s account will lose its descriptive pull and, as such, he again faces a similar problem to Fraser: he has difficulty in explaining the motivation for social struggles, at least if he wants motivation to be strongly linked to moral legitimacy. And of course, the further he abstracts from the empirically given and experienced, the more he is in danger of imposing a theory (possibly of proposing a new ‘ideology’) and thus leaving the path of Critical Theory moving towards traditional theory.Footnote ³⁰

To sum up then, in this article I have shown that Honneth’s recognition theoretical approach to justice and especially his proposal of a revised account of decentred autonomy and authenticity as proper self-relation manages to meet the descriptive challenge presented to recognition theories, but only at the cost of losing the normative pull. If Honneth wants to propose a normative framework within which social critique is possible he will have to give up some of the descriptive pull of his theory and provide a fairly detailed account of genuine self-realization, true needs or objective human well-being. However, given the pervasiveness of oppressive power structures, the facts of opacity and dependence, such an account itself will always be provisional and open to revision. There is of course a tension between the project of Critical Theory itself (and its task of ideology critique) and the need to give an objective account of well-being which will have to be somewhat removed from the actual experiences of injustice and oppression, which Honneth would need to address once more, were he to supplement his recognition theoretical approach.