Palinurus: Engaging Political Philosophy

Journal Home
About Palinurus
Editorial Collective
On The Political
Issue Number

Jean Cohen: A Review of Agnes Heller, Beyond Justice



Republishing a review of Agnes Heller’s Beyond Justice--a work that first came out twenty years ago--it might risk being outdated, but it is not. Heller’s book provides a rich and comprehensive analysis of the concept of justice and an invaluable historical perspective that gives us a clarification of the philosophical issues at stake—through an original and relevant discussion. Beyond Justice is an important book in keeping with current debates in political philosophy.



The contemporary debate in political philosophy between neo-communitarians and rights-oriented liberals revolves around the question of whether it is possible to articulate a formal, universalistic (deontological) concept of justice without presupposing a substantive (historically and culturally specific) concept of the good. Seen from the perspective of the neo-communitarians, rights, be they “human”, civil or political, articulate no more than the particular discourses and traditions of western civilization; they can be defended or challenged only with respect to our way of life. For the political liberal, however, rights constitute the heart of a conception of justice that makes the claims to legitimacy of any contemporary polity plausible. As such, rights constitute universal moral demands, upon all states to enact laws and enforce policies that are tolerant of diversity and neutral with respect to any particular conception of the good. While the first approach seems to be at a loss when it comes to articulating a principle that can serve to ground the choice of those aspects of “our tradition” that one selects as worthy of preserving; the second is perpetually embarrassed when it comes to defending its universalistic claims on grounds other than dogmatic appeals to natural rights. In short, the contenders in the debate seem to have reached a stand-off.


That is why Beyond Justice could not have appeared at a more opportune moment. By providing an extremely rich, learned and comprehensive analysis of the concept of justice and its various shapes over time, the book offers an invaluable historical perspective sorely missing in the current debate. It also gives us, if not the solution, at least a clarification of the philosophical issues at stake, thereby opening a conceptual way out.


For us, the most relevant and original discussion occurs in chapter 5 of Beyond Justice, entitled, “Towards an Incomplete Ethico-Political Concept of Justice”. For it is here that Heller throws down the gauntlet to neo-communitarians, by insisting that a comprehensive, ethical-political concept of justice is neither possible nor desireable today. The latter, of course, refers to a concept of justice that fuses moral and legal norms, “righteousness” or goodness and political justice. It rests on the assumption that morality can be defined by the observance of socio-political norms, while the notion of a just polity can be based upon the foundation of a pre-existing moral order. In short, a complete ethico-political concept of justice is a concept of collective morality or Sittlichkeit. The norms and rules of the (ideal) city are thought of as the precondition for the education (paidea) of the virtuous (just) man and as the end result of virtuous activity itself [1].


The modern condition, of course, undermines this position from three sides: First, the plurality of conceptions of the good life and the absence of any over arching collective definition of the good precludes construing the morality of the individual as the following of social and political norms. Second, cultural modernity, or the differentiation and autonomization of the spheres of morality, science and art each organized around their own distinct validity claims, heralds the dissolution of totalizing world views (Weber) and the end of the substantive unity of reason on which all complete ethico-political conceptions of justice are based. Finally, the very project of a modern, post-conventional, enlightened and free morality opens all norms in principle to critical reflection. It thereby undermines the self-evidence that conceptions of Sittlichkeit rely on. As Heller aptly puts it, if totalizing conceptions of justice reemerge in modern times, they are no longer representative of these times. Thus, Hegel’s Philosophy of Right is indeed the swansong of the complete ethico-political conception of justice.


With respect to Hegel and to the modern condition, Heller states, “The moment that civil society was conceptualized as the source of the legitimation of the state, and the state was conceived as the trinity of the constitution, the sum total of the laws and the lawful government, the complete ethico-political concept of justice had to fall to pieces...” [2]. The state as such can no longer be seen as the supreme good but becomes only an institution which exacts retribution for crime and guarantees negative freedom. Accordingly, the ethico-political concept of justice dissolves into its component parts--social and political philosophy take up the dimensions of justice that refer to issues of distribution and retribution, while moral philosophy pursues the ethical dimension, and seeks to retain its objectivity through becoming formal and procedural.


The importance of Heller‘s project is that she nonetheless refuses to restrict matters of justice to questions of distribution and or retribution, while recognizing the impossibility and undesirability (pace the neo-communitarians) of returning to a complete ethico-political model. Instead, she draws out the theoretical implications of the “moment” referred to above, if civil society is indeed the source of the legitimacy of the modern state, then matters of justice are matters of political legitimacy and not simply questions of distributive norms. Moreover, insofar as modern states claim not only to secure negative freedoms and distributive justice, but also to be democratic, that is, to draw their legitimacy from civil society, then a modern theory of justice must be a theory of democratic legitimacy. It is thus no accident that chapter 5 of her book relies so heavily on the work of Jürgen Habermas, since his theory of the discourse ethics is by far the best theory of democratic legitimacy available today. In effect, Heller’s “incomplete ethico-political concept of justice” is explicitly a Habermas interpretation, and I believe that the strengths and weaknesses of Heller‘s position can be clarified through an assessment of this interpretation.


Heller maintains that she accepts the discourse ethic as the basic normative foundation for an incomplete ethico-political concept of justice. The latter seeks to establish a common normative foundation for different ways of life without simultaneously purporting to design the best possible way of life. It seeks to answer the question of how a pluralist universe in which each different culture is bound to every other by bonds of symmetric reciprocity is possible. In short, an incomplete ethico-political concept of justice is self-limiting: it focuses on procedures through which socio-political norms and rules can be legitimately established and revised but is does not pose as a general moral theory. The normative foundations of such a conception of justice are not the normative foundations of morals in general, but of social-political norms acceptable to a plurality of forms of life in a given polity and between polities. The task is thus to provide a formal, procedural, universalistic conception of the right, of justice, which is neither empty nor reducible to a particular concept of the good.


To be sure, Habermas views his discourse ethics both as a general moral theory able to replace the Kantian categorical imperative and as a theory of democratic legitimacy that replaces social contract models. It is thus open to the charge that despite its abstract and procedural character, it does tend to collapse the right and the good, to become monistic and overly totalizing. Accordingly, Heller’s restriction of the theory to questions of democratic legitimacy or political justice alone is crucial. Since this restriction is extremely important, let me briefly state the two core tenets of the discourse ethics so that the argument is clear. The first specifies the conditions of possibility of coming to a legitimate rational agreement, the second, articulates the possible content, on a formal level of such an agreement. A norm of action is considered legitimate only if all those possibly affected by it would, as participants in a practical discourse, arrive at an agreement such that the norm should come into or remain in force. What qualifies as a rationally motivated agreement has rather demanding preconditions which Habermas articulates as a set of metanorms implicit in the procedural principles of discursive argumentation itself. All those affected must have an effective equality to assume dialogue roles, the dialogue must be fully public and unconstrained, it must be possible to alter the level of discussion and radicalize the argument at all levels, there can be no taboos. In short, the procedural principles underlying the possibility of a rational consensus involve symmetry, reciprocity and reflexivity. Unlike Rawls, however Habermas insists on an actual rather than a virtual dialogue, for only an actually carried out discourse allows the exchange of roles of each with every actor and hence a genuine universalization of perspective. The emphasis has shifted away from what each individual can will without contradiction to be a general law (Kant) to what each and every one will recognize in consensus to be a universal norm.


With respect to the second dimension of the discourse ethics, Habermas maintains that the norm of action upon which we agree must articulate generalizable interests. Of course, this principle requires in turn an actual dialogue [3].


By restricting the relevance of the discourse ethic to socio-political norms, Heller is able to remain true to one of the great achievements of modernity: the differentiation between morality and legality. At the same time, she can examine on its own terrain, as it were, the ethical principles underlying and legitimizing law. However, Heller accepts the discourse ethic as the basis of an incomplete ethico-political concept of justice only with certain additional and far more problematic revisions which shall spell out. Indeed, I want to argue that (a) Heller’s reasons for the necessity of revision rest on a one-sided interpretation of the discourse ethic, and that (b) this leads her to make claims that are far more sweeping and indefensible than those flowing from the original theory, and (c) that her alternative tends to refuse morality and justice which she so carefully and rightly differentiated.


(a) As Heller points out, Habermas has long been a principle exponent of the view that the validity of norms can be tested in a cognitive procedure [4]. He has insisted on this thesis because without it, one would be unable to distinguish between the de facto social acceptance of a norm and its validity, in other words, between a true and false consensus. Without the possibility of judging the rationality of a consensus, acts of collective choice would simply be acts of will; we could describe them but we could not justify them. Thus, Habermas explicitly rejects the suggestion that one simply equate argumentation with processes of collective will formation and thereby excise the cognitive dimension from the theory altogether. For this would leave one wide open to the classic objection to theories of democratic will formation and majority rule: a mere empirical consensus cannot yield legitimate obligations. Here Heller agrees with Habermas: the viability of democratic will formation depends on the cognitive-rational dimensions of the discourse ethic.


However, Heller accuses Habermas of winding up with a position, malgré lui, very close to the one just described. She does so by locating the cognitive character of the discourse ethic exclusively in the principle of universalization and by arguing that this principle cannot deliver on the cognitive claim. If the sole criteria of the rightness of norms is whether everyone accepts its consequences and side effects, then the discussion will focus on these and not on determining via a cognitive procedure whether the norm is right or just. Moreover, Heller insists that interests or needs cannot be subjected to discussion oriented to consensus at all since the discussion of needs can never be conclusive. If as Habermas himself maintains, need interpretations are informed by values and norms, then we are led into a vicious circle: “…where needs themselves are informed by values and norms but where the will to the acceptance of a norm is informed by needs” [5].


Heller‘s solution to this apparent dilemma is twofold. First, she argues that the discourse should be conducted among values. It is here that the cognitive process sets in, namely in testing the truth, falsity, rightness or wrongness of values. Second, she maintains that such a discourse is possible (i.e. can be settled via free consent) only if there is a “higher order” consensus about the unconditional and absolute validity of at least one value prior to the discourse. This higher order consensus has a normative power beyond reasoning. And yet Heller insists that “if we do not accept any substantive value as taken for granted, as not open to testing,consensus can be based on will-formation alone, thus it will not be rational” [6].


(b) It is here that I must raise certain grave doubts. First, does not the specter of decisionism which both Heller and Habermas seek to avoid by insisting on the cognitive character of ethics return with this revision? Who is to decide which substantive value is supreme and unquestionable? To insist on the unconditional, absolute validity of one supreme value prior to discourse, to refuse to open it to testing, is either to dogmatically assert the value without giving grounds or to claim that such a value is already an empirical universal, that it is already recognized by everyone. Heller’s revision in fact does both. Decisionism with respect to the ultimate supreme values (for even if everyone now shares them, someone could change their mind or new arrivals might disagree) is coupled with the astonishing claim that two and only two values fit the bill: freedom and life, and that these values “…are factually universally valid (freedom) or in the process of factual universalization (life)” [7]. They are, according to Heller, transculturally universal.

By this she means that the opposite of the value can’t be chosen by anyone: “Since Hitler, not even the worst kind of tyrant can now publicly confess that he prefers unfreedom to freedom” [8]. One is really at a loss to figure out what such a statement can mean in the face of Khomeini, anti-modern religious fundamentalisms at home and abroad, although not so Stalinist systems alive and well in many parts of the globe.


(c) My second objection is that the insistence that political discourse on norms requires a prior consensus on at least one supreme value undermines the very distinction between morality and justice that Heller started out with. For it implies that the precondition of political justice is after all a shared substantive morality. To claim that the discourse has to be on the truth or falsity of values, and that it can be settled only if all contesting values are related to the supreme one or by showing that they are only different interpretations of the same value, is to reassert the primacy of morality over justice and to return to a rather more emphatic conception of Sittlichkeit than seems plausible.


It seems to me that Heller encounters these difficulties because she has partly misinterpreted the proper locus of the cognitive claim in Habermas’ theory and consequently has chosen the wrong solution to a nevertheless genuine problem. Habermas does not locate the cognitive dimension of his theory solely in the principle of universalization. Rather, he maintains that the cognitive component of norms is not limited to the propositional content of normed behavioral expectations. The normative validity claim itself has a cognitive thrust in the sense of the supposition that it can be discursively redeemed  grounded in the consensus of the participants through, argumentation and the giving of reasons. In this sense, rational grounds can be given not for the truth of values per se, but for their incorporation into socio-political norms. The principles of argumentation can provide a meta-norm (symmetric reciprocity) to which participants in a dialogue can appeal in testing the outcomes (norms) of an empirical consensus. Although it is true that every consensus is empirical, we are not left only with an arbitrary collective will. The rationality of the consensus can be tested by referring validity claims back to the procedural conditions or meta-norms that alone can make it valid and obligatory. One can challenge the validity of a political norm by claiming that the consensus was unfree, that all arguments were not heard, that some individuals were excluded from the discussion, etc., and by proposing counter arguments. In this sense, the objectivity of the judgement can be rooted in the structure of argumentation itself- it is not brought in from the out side in the shape of an absolute value that we happen to choose.


It is true, however, that Habermas also claims that only if norms embody generalizable interests are they based on a rational consensus. And to do so is indeed problematic. If “generalizable interests” refer to raw need interpretations Heller is right, the discussion must be inconclusive. On the other hand, if it refers to the “objective interests” of a group, it cannot be used as the critierion for the Tightness of a norm without reviving the naturalistic fallacy and/or an authoritarianism made all too familiar in the Jacobin-Leninist tradition. Since the concept of general interest in this sense is by definition ascertainable only from the observer’s point of view, it would tend to replace the opinion of participants with an objectivistic judgement about what is good for them. And yet the concept of interest cannot be dispensed with in political justice because interests, as well as norms comprise the stuff or content of what is discussed in political debates. The way out of this dilemma is not to insist that the discussion on norms presupposes the prior acceptance of an absolute uncontestable value, but rather to see that participation in a discourse allows us to shape, articulate, assess and or discover what if anything we who participate in it have in common.


The discursive process constitutes what is shared, it creates or reaffirms a „we,“ a common identity and a new or reinforced solidarity based on shared norms and opens the way for the recognition of common interests with respect to the preservation of that commonality. (It is in this way that humanity itself could become a cluster and human rights, objectively, but not dogmatically, grounded). Such a discourse does not judge entire forms of life, particular substantive value systems, or, for that matter, need interpretations. Rather, first and foremost, it provides a principled procedure for discovering and selecting what we who come into contact with one another have and wish to preserve in common. In a pluralistic universe comprised of individuals with different forms of life, participation in discourses on norms implicitly commits the participant to the meta-norms of symmetric reciprocity as principles we can come to embrace in and through discussing our differences and discursively settling our conflicts. Understood in this way, it does indeed provide the core of an incomplete ethico-political concept of justice.


The neo-communitarian could of course object that willingness to engage in discursive conflict-resolution is simply a characteristic value of “our” (western) liberal-democratic tradition. Discourses do not create values and solidarities ex nihilo but draw on an already shared commonality and culture, i.e. life world. The norms generated by participation in a discourse would thus not be universal, but specific to those who value this form of interaction. Moreover, it is of course obvious that one can enter into a dialogue for all sorts of empirical reasons and that one could relate strategically to the outcome of a discussion.


What then makes the discursive meta-norms themselves obligatory and universal? Two things: first, there is indeed something that comes in from the outside, as it were. This is not, as Heller argues, a prior consensus on a value, but rather, the experience of normativity itself. Everyone understands ought statements, and everyone has the experience of normatively oriented action. The discursive situation in which we raise and test the validity of norms thus has the prospect of bringing one’s own norm orientation to the post-conventional level through argumentation. It is in such a process that discursive conflict resolution itself is normatized. Second, a life world which has, however imperfectly and partially, institutionalized universalistic principles to guarantee freedom and justice is, to be sure, ours, but this historically contingent fact does not render claims to democratic legitimacy particularistic. On the contrary, claims to democracy and rights that can be discursively redeemed reveal the particularistic and pernicious dimensions of those forms of life, existing within a modernized life world, that are incompatible with a political way of life guided by the principles of justice. Against the liberal claims to absolute neutrality, the neo-communitarian is quite correct  democratic and liberal institutions embody norms and principles that are often substantive and comprise a political way of life. But this is the advantage of an incomplete ethico-political concept of justice: As a political way of life, the principle of democratic legitimacy grounded by the discourse ethic would indeed exclude all forms of oppression, violence, domination, illegitimate inequality and exclusion in both civil society and the polity. In other words, it would not tolerate an absolute right to difference, nor would it be compatible with an infinite plurality of ways of life. But it is the only principled political ethic that is able to provide for justice, autonomy and the widest range of plurality compatible with the principles of freedom and democracy. It is surely the great merit of Beyond Justice that it has pointed us in this direction.




1. Agnes Heller, Beyond Justice (New York: Blackwell, 1987), 93.


2. Heller, Beyond Justice, 91-2.


3. For a fuller discussion see Jean Cohen and Andrew Arato, “The Discourse Ethic and Civil Society,” in Civil Society and Social Theory (M.I.T. Press).


4. Heller, Beyond Justice, 237.


5. Heller, Beyond Justice, 238.


6. Heller, Beyond Justice, 240.


7. Heller, Beyond Justice, 251.


8. Heller, Beyond Justice.



First published by International Praxis, (8:4) January 1989.