Beyond Justice by Agnes Heller (1987)

The Formal Concept of Justice

On the Formal Concept of Justice In General

The formal (or static) concept of justice is formal (or static) in that (a) it is so defined as to encompass the common properties of all types of justice, and, as a result, (b) it is abstracted not only from all normative content, criteria and procedures of justice, but from the (finite) ideal types as well, which are reasonably construed via different combinations of content, criteria and procedures. Chaim Perelman, in his book The Idea of Justice, on which I rely heavily at this initial stage of my analysis, uses the term ‘formal justice’ in what I should call a misleading way. Weber introduced a particular definition of ‘formal’ justice In his theoretical discourse, in contradistinction to substantive concepts. He meant thereby the formalization of just procedure, a relatively modern type of justice emerging with the rationalization of legal authority, and I see no compelling reason to dismiss the Weberian definition or replace it with any other. Formal justice is thus one kind (or type) of justice. However, the formal concept of justice contains the common features of all kinds of justice, be it ‘formal’ or ‘substantive’. It is not the kind of justice it refers to that is formal, but the concept. The formal concept of justice implies, then, a far higher level of abstraction than the concept of ‘formal justice’.

Being born into any society and any social group means that members of this particular society and social group must learn about and practise the norms and rules of that particular society and social group. This is the process called ‘socialization’. Normally, different sets of norms and rules apply to different groups of people (one fundamental division being that between men and women). Accordingly, learning about and practising norms and rules is tantamount to two distinct sets of behaviour. On the one hand it means knowing, practising and expecting a certain set of norms and rules; on the other it means knowing and expecting, but not practising the same set of norms and rules – in other words, ingroup and outgroup behaviour. The norms and rules regulating the actions of outgroup members have to be known, and the observance of these norms and rules is expected. However, the practice oriented towards the outgroup is regulated not by the norms and rules of this group, but by those of the ingroup. Within an ingroup, expectation and action (speech acts included) are symmetrical: I do in relation to you exactly what I expect you to do in relation to me, for the simple reason that we are both supposed to observe exactly the same norms and rules. By contrast, in relations between ingroup and outgroup, both expectation and action (speech acts included) are asymmetrical: I expect you to do something different from what I do, for the simple reason that we are supposed to observe different norms and rules, or, at least, different ones in addition to common ones. This is obviously a simplified model. Social relations are far more complex than this even when they occur in a so-called ‘primitive’ milieu. But this simple schema is an ideal point of departure for the establishment of a preliminary, formal concept of justice.

Theories of justice conditioned by the view that justice must be explained in terms of the ‘equality-inequality’ dichotomy depart, as a rule, from the Aristotelian dictum that being just means treating equals equally and unequals unequally. However, this dictum (or reference) is normally analysed out of its original context, dislocated from the complete Aristotelian conception, where it did make sense . Aristotle never believed that two persons could be each other’s equals. Rather, in criticizing Plato’s Republic (in the Politics) he emphatically argued the contrary. In contrast to Aristotle, those who have established a theory of justice based on the principle that equality means treating equals equally and unequals unequally really do intend that people who are in fact equal should be treated equally, and that those who are in fact unequal should be treated unequally. But what does it really mean that two persons are equals, or, for that matter, unequals? People obviously differ from one another and each and every person is unique. On what ground, then, can we compare them? Human beings cannot be quantified (at least, not without a significant residue considerably impairing the value of such quantification), so it would be more than ridiculous for one to establish the equation A = B = C, etc., or similarly to establish that A > B > C. An obvious objection to this is that the statement ‘equals should be treated equally and unequals unequally’ refers not to A and B as wholes but to one aspect of A and one aspect of B – for instance, stating that A, B and C are girls or are persons all eighteen years of age. But the mere fact that A, B and C are equally girls or equally persons eighteen years old does not make them equals as girls or as eighteen-year-olds, unless there are certain norms and rules which apply to all girls or to all eighteen-year-olds. To be ‘equally such and such’ means that we share certain characteristics. Indeed, even if we all are singular human beings we do share certain properties with other human beings, and, again, we share certain essential properties with each and every human being. However, Rousseau was clearly aware of how little these shared properties have to do with social equality or inequality, unless they are created, or at least reinforced, by social norms and rules. Everyday users of language may identify the adverb ‘equally’ with the noun ‘one’s equal’ if a social order is taken for granted as quasi-natural (where only those who are one’s, equals are treated equally). This not being the case, even everyday language-users tend to arrive at the standpoint from which I began this discussion – that equality and inequality are normatively constituted. In other words, if the same norms and rules apply to a cluster of people, we refer to the members of this cluster as equals. If different norms and rules apply to two clusters of people and the asymmetry of the pertinent behaviour of the members of the two clusters towards each other is constant, we refer to the relation of the members of the two clusters as unequal and to the members themselves as (each other’s) unequals. If A, B, C are free citizens and D, E, F are not, the norms and rules of free citizenry apply to A, B, C and do not apply to D, E, F. As far as citizenship is concerned, A, B and C are equal with one another (are each other’s equals), and D, E, F are, again, equal with one another, but they are unequal with A, B, C (members of the different groups). Social equality and inequality is not an ontological given; it is constituted by the application of distinct sets of norms and rules. Conversely, distinct sets of norms and rules constitute social inequality only if the following conditions prevail: (a) a strict division obtains between the sameness of knowledge, action and expectation within the ingroup, and the difference between knowledge and expectation on the one hand, and action on the other hand, in relation to the outgroup; (b) the asymmetry is constant.

Let me disregard for the time being the relation between various social clusters, and talk only about a single social cluster, where the same norms and rules apply to every member of this cluster. But even where every member must observe the norms and follow the rules, they can observe and follow them with different degrees of intensity and consistency. The infringement of norms and rules can be regarded as a sin, crime, offence or blunder according to the gravity of transgression of failure to comply. Norms and rules themselves provide the standards for the establishment of merits and demerits (in moral or other forms of excellence). Comparing persons according to their merits and demerits – which is, after all, a most common form of doing justice and injustice – is a procedure that can hardly be understood in terms of the dichotomy ‘equality-inequality’, not even within the framework of a modern, highly technical and formalized legal system. If both A and B are sentenced to six years imprisonment, there is no reason to assume either that A and B are alike, or that their respective, crimes are alike. Speaking of merit, if A, B, and C are awarded the Nobel Prize, it is presumed that all three excel in something, but it is not presumed that they excel in the same thing or that they are ‘alike’ in their excellence.

Perelman offers three definitions of the formal concept of justice, or, to use his phrase, of ‘formal or abstract’ justice. This concept is supposed to be ‘a principle of action in accordance with which beings of one and the same essential category must be treated in the same way’; and, further, it ‘means ‘applying rule to all members of an essential category’ and ‘observing a rule which lays down the obligation to treat in a certain way all persons who belong to a given category’.’ These definitions serve as my own starting-points as well, with certain modifications. Perelman’s definitions might create the impression that there are certain pre-existing essential categories (or clusters) to which the same rules should then apply. But in my view it is precisely the application of the same norms and rules to a certain cluster of people that constitutes that very cluster, and not the other way round.

Further, I shall not use the term ‘rule’ (or ‘rules’) alone, but shall speak of ‘norms and rules’. To my mind it is wrong to assert that all social regulations assume a rule character, even if this is widely held in sociology or modern philosophies. Rules can only be followed in a single and definite way. If rule Y applies to case X, there can be doubt whether X is the case: there can be no doubt, however, concerning the ways and modes of following rule Y if X is the case. If X is the case, rule Y has to be followed without de liberation and reflection, and there is only one way to follow this rule. Norms, on the other hand, are different in nature. There are concrete and abstract norms. Concrete norms are rule-like, abstract norms are not: for example, the virtue of ‘acting with civility’ is an abstract norm, whilst norms which require that we treat a visitor in a particular manner are concrete. Though related, they are clearly different. Abstract norms enjoin us to act always in a certain manner; concrete norms tell us how to act in particular situations. However, while concrete norms are rule-like in nature, they are not rules. How strongly they enjoin compliance depends on their proximity to abstract norms. The observance of norms is not exhausted by following rules – at least, not in a complex society or ‘essential category’. This holds true for the simplest game situation as well. During a soccer match, the referee is only concerned with the rules; there can be no doubt whatsoever in his mind whether or not – in the case of fault – the rules prescribe a free kick against the offending party. On the other hand it is not his business whether a player is showing dedication or just idling around during the match. However, the latter Is one of the main concerns for both coach and fellow players, utmost personal dedication being one of the norms (and not a rule) of the game. And, given that social life is not a game situation and cannot be adequately characterized in any game theory, the use of the term ‘rule’ instead of ‘norms and rules’ inhibits the understanding of its complexity.

My redefinition of the formal concept of justice is as follows: the formal concept of justice means the consistent and continuous application of the same norms and rules to each and every member of the social cluster to which the norms and rules apply.

This definition sounds tautological, but it is not (nor is Perelman’s). The formal concept of justice is the maxim of justice; the consequence of non-observance is self-contradiction. But it is far from easy to act, judge and distribute without manoeuvring ourselves, time and time again, into such a self-contradiction.

There are two distinct ways of deviating from acting, judging and distributing under the guidance of the maxim of justice. This happens, first, if the norms and rules are not applied to the members of the particular social cluster to which they ought to be applied (if other norms and rules are applied), and secondly, if the application of the norms and rules which should be – and indeed are – applied is inconsistent or discontinuous (discontinuity being a special case of inconsistency). Observing the maxim in the first way poses more complex theoretical problems than observing it in the second way, particularly when it comes to applying norms. Even if we normally know which norms apply to which cluster, deciding whether a person should be treated according to one norm or another can present a serious problem if the person belongs in at least two different clusters. At this juncture moral conflicts may arise; for instance, as to whether the person who committed a particular crime is mentally ill or sane. Obviously, the decision made on this question entails different forms of reaction to the crime, and the risk of coming to a wrong decision, of passing a wrong judgement, is quite obvious. Normally, more than one norm applies to members of social clusters, which raises the question of which (for instance, which merit or demerit) should be given priority. Further, one can be lenient or strict in a greater or lesser degree in the application of the same norms to each and every member of the cluster to which they apply. What are the limits of rigorousness and liberalism if the proper norms and rules are applied? There are no general answers to these problems, but they indicate in their being that the formal concept of justice is indeed a maxim.

The maxim that the same norms and rules should apply to each and every member of a social cluster constituting that cluster regulates two, relatively distinct, kinds of action, speech acts included. The individual A who belongs to cluster X must himself observe the norms and follow the rules, and his interaction with all other members of cluster X should depend on whether they have observed the same norms or followed the same rules, and if so, to what extent. Even though these two factors are regulatively interwoven, inconsistency may appear in applying norms and rules if one factor is more constitutive than the other. People tend to judge others harshly for deeds they themselves have committed, and this is certainly unjust, even if they make these judgements about others consistently. Eventually, the obverse can be the case as well: severity in self-judgement, lenience in the judgement of others. If this inconsistent judge judges others consistently, we might say that he is unjust to himself, but normally we do not call him an unjust person.

Let me demonstrate the formal concept of justice by a very simple example. A rule exists that every student who achieves 340 marks in his or her High School Certificate (HSC) should be admitted to the law school of any university. This rule establishes the cluster of students who have indeed achieved at least 340 marks. Yet such a rule does not make the students ‘alike’, not even in respect of their mental excellence. But, irrespective of their remaining personal differences, all are equalized by the rule, because it applies to all of them. Therefore, should any one student with 340 points or more not be admitted to the law school of a university, we have a case of injustice. There cannot be the slightest doubt about this conclusion, for the cluster in question was constituted by a rule (not a norm). As mentioned, if we apply a rule, we merely have to determine whether X is the case. If it is, then Y should be done, and it can only be done in one way (in our case, if at least 340 points ‘Is the case’, the student gains admittance to any law school).

The cluster of students constituted by the rule in question is a result cluster, in that it results from a previous procedure. We can term a procedure just or unjust (from the perspective of the formal concept of justice) if the norms and rules of a procedure constitute a social cluster. In our example, the cluster previous to the result cluster consists of all students who sit their HSC, and sitting their HSC means following rules and observing norms related to ‘sitting for the HSC’. (In this respect, the minimum number of subjects a student must take is a rule, and so it can only be observed in one way.) On the other hand, preparing for the HSC is a norm, not a rule: different students do it in different ways and with different degrees of commitment. However, the procedure for HSC constitutes not one but two clusters (ingroup-outgroup): one cluster which meets the standards set externally, the other verifying whether the members of the ingroup have indeed met the externally set standards (the marking teachers).

Under further scrutiny, three conclusions can be drawn from our simple model.

  1. The formal concept of justice applies to procedure as well. Or, more precisely, a procedure can only be termed just or unjust if the formal concept of justice applies. There is no specific procedural justice as contrasted to the formal concept of justice. However, one specification has to be made. If rule or norm Y applies to case X, we must first ascertain whether X is indeed the case before applying the rule (norm). Prior to the emergence of modern formal justice (which is now termed ‘formal’ in the narrower sense of the word), procedures were admitted or eventually invented to establish the facts (for instance, innocence or guilt); procedures to which the formal concept of justice was not applied. In this vein, the procedure invented by Solomon to determine the real mother of a child cannot be brought under the heading of the formal concept of justice, whilst his final decision, after having ascertained the truth, obviously can. This and similar idiosyncratic procedures are still with us, even if they do not operate in the legal domain.
  2. The application of the same norms and rules to each and every member of a cluster to which the norms and rules in question apply is, together with consistency in application, a moral imperative, even the norms and rules themselves are not o la moral nature. Conversely, making exceptions in the application of these norms and rules is a moral offence, even if the subject matter is not of moral provenance. So an unjust act is morally wrong in itself, irrespective of whether marking, judging, comparing, or distributing has anything to do with the morals of those marking, judging and distributing, or with the morals of those towards whom we act or to whom we distribute.
  3. It was not just for simplicity’s sake that I took the norms and rules for granted. Clearly, they can be questioned. The norms and rules constituting a social cluster can be declared unjust, or the procedure itself may be declared unjust, even if the norms and rules in question are consistently applied to each and every member of the cluster. How and when they can and should be declared unjust will be discussed in chapter 3. Yet, by declaring either the norms and rules or the procedure unjust, we already form in our mind (in communication with certain others) the idea of different norms and rules which have the telos, or at least the potential, to constitute social clusters different from existing ones. But even in this ‘Ideally existing’ cluster the same norms and rules should apply to each and every member of the cluster. One cannot criticize existing norms and rules rationally without proposing alternative ones, or at least an alternative procedure for establishing social clusters. Thus the formal concept of justice itself is not tested and queried, even if norms and rules are.

The Virtue of Justice and the Formal Concept of Justice

Viewed from the standpoint of static justice, judgement is passed on merit and excellence and their respective opposites. Except when we keep our judgement to ourselves, all judgements are actions. They are speech acts, and often include acts other than speech acts. Scolding and praising are speech acts; punishing and rewarding include acts other than mere speech acts. The speech act of judging may have a perlocutionary and an illocutionary power, or it may have only the latter. If one turns to the person one scolds or praises, the perlocutionary power of judgement is intended; the case of public scolding and praising is similar, even if the person in question is absent. Privately undertaken acts of judgement in the absence of the person on whom judgement is passed can only have an illocutionary power. But, whatever power these speech acts have, all judgements passed on merit and excellence imply either scolding or praising or both, if only implicitly. Further, all judgements imply ranking and/or comparing from the standpoint of norms and rules the persons upon whom judgement is passed. Not even judgements about another person’s taste or mode of dress are exceptions. If there are no norms and rules concerning proper taste or modes of dress for members of a social cluster, judgement is neither just nor unjust (it is precisely in this vacuum that we may say that personal taste is not open to discussion). Obviously, not all forms of ranking and comparing are related to justice, only those passed on persons according to social norms and rules.

Judgements passed on merit and excellence can be viewed as acts of distribution. Praising or scolding mean distributing honour and shame If such speech acts are followed by direct actions, it is punishment or reward that Is being distributed. This usually happens via the distribution or redistribution of ‘social positions’ or material goods. Whether kings distribute lands according to the loyalty of their vassals, or authorities of a liberal state distribute promotion on the basis of excellence, being lust means observing the norms and rules and applying them consistently. Consequently, one can be unjust in two ways. First, one can apply to the members of a social cluster norms and rules other than those that should be applied to them. If this happens, one will normally be regarded as unjust by the members of the cluster, even if these norms have been applied consistently to each and every member of the cluster. Secondly, one can inconsistently and discontinuously apply the norms which do apply.

Moreover, there are two kinds of inconsistency. First, there is inconsistency in comparing and ranking persons, applying norms and rules to some but not to others (the exception could be one person), being more severe to some, more lenient to others. Inconsistency of this kind is normally attributed to personal motivations, such as liking or disliking, vested interest, passions aid the like. Secondly, one can be inconsistent in comparing groups. The distinction between a group and a social cluster is constituted by the norms and rules themselves. In so far as the same norms and rules apply to different groups of persons, those groups belong to the same cluster and the members of them must be judged by common norms and rules. If one judges one severely and another leniently, when the same X (action, form of behaviour or forbearance) applies to both, this is unjust. This inconsistency is called the application of double standards. It is attributed to social (and occasionally ideological) bias. Of course, if two groups do not belong to the same social cluster (being constituted by different norms and rules), then, if we apply different standards to the members of the respective groups, we are not guilty of applying double standards because we have applied the proper (because different) rules to these clusters, and we have done so consistently in accordance with the formal concept of justice. Distinguishing between different and double standards is of the highest importance. If, for example, excellence is the only accepted norm and rule in selecting people for certain positions (without provisos and further conditions), anyone who consistently prefers males to females applies a double standard, However, there can be provisos and conditions built into the norms and rules themselves, such as preventing criminals from exercising their civil rights. This is a very simple case because it concerns a rule and not a norm (or norms). However, if at least two different norms can apply to a situation, one must have priority. In the United States the debate is still continuing on how to apply the norm of excellence in conjunction with another norm (that of historical equity) in order to rectify previous discriminations against disadvantaged groups (blacks, chicanos, women); there are several different general guidelines and opinions as to how to make a just decision and how to apply the second norm without infringing the first. Even so, irrespective of these guidelines, people making the concrete decisions have been forced to decide, at least in the more complex cases, in favour of one norm as against another. In this and all similar cases each and every concrete decision has to be justified. To follow the argument of Marcus Singer, [2] we can say that the generalization principle falls in such a situation (we cannot act as everyone should act in a similar situation), and consequently the justification principle has to be applied. We have to justify why we have acted (decided, judged, distributed) in a particular way in a particular case. Although we have dealt only with the problem of inconsistency in the treatment of members of different groups, if this inconsistency is built into the norms themselves then the same can be said of the inconsistent treatment of persons under the same conditions. The only exception to this is judgement as a mere illocutionary act, where different norms can be applied simultaneously without contradiction. But, if we act towards persons (punishing, rewarding, distributing, or anything else) in a context where two norms apply, and we prefer to apply one norm rather than the other (which is inevitable), we are again obliged to justify our action through having recourse to a norm which applies to all members of the social cluster to which the person so treated belongs.

The allegory of justice is the blindfolded woman holding up the scales of balance. Eventually, she may also hold a sword. This is the typical graphic representation of the formal concept of justice. The allegory of justice as depicted by Giotto in the chapel of Arena is, however, a representation of the dynamic concept of justice, not of static justice. Justice appears here as a queen holding a statue in both hands, the angel of war and the angel of peace, the latter being heavier than the former. And the figure is not blindfolded; her eyes look forward, into the future.

It is the allegory of the formal concept of justice to which we restrict our discussion here. The deity is blindfolded for very good reasons: she must not see whose acts are going to be weighed. Justice must be impersonal and impartial. it must not be influenced by liking or disliking, passion or interest, nor should it be influenced by charity, pity or goodness of heart. These latter stipulations are just as important as the former. If ten people are sentenced to death in accordance with the requirements of certain norms and rules, those who are moved by pity and rescue even one of them are unjust. This is so whether those who make the exception do so because the person is a friend or relative or because they are moved by the natural dignity of his behaviour in the face of impending execution. Clemency and mercy are only just if practised according to norms and rules. Deeds motivated by the most sublime feelings can be unjust. Being just is a cold virtue, sometimes even a cruel one. How mild or how cruel it actually is depends on the norms and rules themselves, but this problem cannot be addressed in a discussion of the formal concept of justice. [3] The Romans, both with their maxim of fiat justitia pereat mundus (‘let justice be done though the world perish’), and with their heroes Brutus, who killed his son, and the second Brutus, who killed Julius Caesar, his stepfather, represent the epitome of unflinching impartiality. But even in other times, such as ours, when mercy and forgiveness are part of the normative system of personal interaction (if not in the legal system), justice still preserves its chillness, this quality which sometimes borders on cruelty. Every teacher who ever failed a student, every parent who ever punished a child, every person who ever ranked, graded, distributed and judged (and we all have), has felt the coldness and even the cruelty of justice. Fiat justitia is an unconditional imperative of the formal concept of justice, even if the qualification pereat mundus need not be appended to it.

Still speaking from the standpoint of the formal concept of justice, persons can be called just if they consistently and continuously apply norms and rules to each and every member of any social cluster to which those norms and rules apply. ‘Any social cluster’ is an important qualification. A just father can yet be an unjust friend. Anyone who applies double standards in any respect is an unjust person, even if he so acts consistently and continuously. A ‘just person must be just overall; it is not acceptable to be just in one respect and unjust in another. Cold as this virtue might be, it is a major virtue all the same. The legitimacy, the validity of any system of norms and rules depends on whether the members of a given society practise this virtue and to what extent. Generally speaking, as long as the system of norms and rules is legitimized, the members of this society will on average be more just than unjust. Injustice is the exception and justice is the rule. Were it otherwise, no system of norms and rules could operate, for they simply do not exist outside the sphere of their application. As a rule, people are not ‘just overall’ (they are not just persons), but they are basically just in all areas in which their personal interest, passion, liking, disliking and bias do not make them act otherwise.

I have mentioned several human qualities that make a person just. All are encompassed in the notion of impartiality. Impartiality is the precondition of objectivity in human affairs. We must be detached from our personal likings or dislikings and our vested interests to establish in a given situation ‘what the case is’. I emphasize here that impartiality in our judgements presupposes partiality for our norms and rules. Fanaticism and justice do not exclude one another, unless norms exclude fanaticism. Members of religious or political groups can be extremely fanatic. Partiality for their own norms and rules can reach the level where no minor deviation from any norm is tolerated. Yet they can be just in that they censure or punish equally or proportionally and with the utmost consistency. However, partiality for our own norms and rules distorts our judgements of people who do not share them, for it excludes even relative impartiality in our judgements of norms and rules. People tend to judge members of another culture by their own standards, which is the extreme opposite of the application of double standards, but does not occur less frequently. Religiously, politically or ideologically underpinned or over-determined norms are usually not open to any kind of relativization. If such norms happen to shape social clusters within a common social milieu, and the members of these clusters I judge members of other clusters by their own norms, the outcome may be blatant injustice, even if those norms are applied consistently and continuously. How those who are ‘Just overall’ (within their ingroup) can be ‘unjust overall’ (in judging members of a group with different norms and of a different persuasion) is a problem I can only outline here, not discuss.

Being just (in the sense of the formal concept of justice) is the result of practice. One has to learn the habit of being just. ‘Impartiality’ does not imply that one cannot pursue interests, cannot like some and dislike others, cannot be passionate, jealous and envious, cannot have a ‘good heart’. It simply means applying the same norms and rules consistently irrespective of personal interest and emotional involvement (warm and positive or otherwise). It is generally presumed that highly emotional people are less capable of practising justice. So most societies provide the opportunity for such people to ‘put things right’ before they consider them liable to punishment; this may take the form of self-humiliation, asking for forgiveness and the like. If such a gesture is accepted then the ‘balance of justice’ is restored. These and similar habits and incidents ‘teach’ people to be just rather than unjust, in the first instance simply to avoid further humiliations. Self-humiliation, asking for pardon or forgiveness, can restore a social hierarchy and authority relationships, but only if the person asking forgiveness belongs to a hierarchically lower social cluster than the person able to grant it.

Apart from the virtue of (qualified) impartiality, just practice requires practical wisdom, a kind of phronesis, to use Aristotle’s terminology. One has to know the facts, the circumstances, the persons, to make a proper judgement, to act properly or distribute justly. If, during a famine, foodstuffs in short supply are to be distributed, one must know if hidden reserves exist, and who has them, in order to distribute justly. A judge must establish the facts before passing verdict. In most cases not all the relevant facts are known, but there is a threshold to knowledge below which just action or judgement can only occur through luck or intuition. Knowledge is a prerequisite for the application of both norms and rules, but phronesis (practical wisdom) is normally activated only in the former case, for here the question of the method of application arises. Incidentally, however, phronesis might serve as the vehicle for finding out what the case really is. Stories about wise rulers meeting their people in disguise are illustrations of practical wisdom as the ability to unveil the facts on the grounds of which just judgement can then be passed.

No person can be just in all of his or her actions, judgements, acts of distribution, even if this person is basically just. At times the available knowledge is below the threshold level, action is still necessary, and intuition falls. Further, if norms are to be applied, and different norms can be applied to a given situation, there is no single just action that can be called upon, but only several approximately just ones. Being just means covering a certain area, not illuminating a single point. Genuinely just persons are nearly always aware of the approximative character of justice. Should they accidentally commit an injustice because of ignorance, incorrect assessment, the false character of available information (if they are deceived), they will rectify it by passing just judgement on themselves and asking for forgiveness from the offended.[4]

Social Conflicts Viewed from the Perspective of Static Justice

From the perspective of static justice, three major sources of social conflict may arise:

  1. the inconsistent application of norms and rules pertaining to a social position;
  2. an attempt to alter the width of the social cluster to which certain norms and rules apply;
  3. asymmetry of ingroups and outgroups.

At this stage the process by which norms and rules themselves are tested cannot be discussed, and neither can the social conflicts arising therefrom.

(1) We have already seen that norms and rules constitute social clusters (in Perelman’s terminology, ‘essential categories’), and that they have to be applied consistently and continuously to every member of the social cluster to which they relate. This fact stabilizes expectations, as well as establishing a relative causal nexus between an action and its consequences. By ‘relative causal nexus’ I mean a strong probability. In our actions we can usually rely upon a strong probability. If I interact with people who are obliged, according to the norms and rules, to keep their promises, I can expect them to keep their promises. In the same vein, if I interact with an outgroup wherein ‘keeping a promise’ does not belong to the norms and rules – where, for example, cheating, deceiving and lying are accepted behavioural forms – I cannot expect that any promises will be kept. Of course, we can be disappointed in our expectations even in the former situation, for injustice can never be excluded as a possibility. The relation between the probability of expectation fulfilled and the possibility of expectation disappointed is variable, and depends on several factors. As a rough estimate one could say that the probability is very high in respect of the application of rules, less so in respect of the application of norms. But since people are just rather than unjust if their interests, passions and biases do not dictate otherwise, the relation between expectation fulfilled and expectation disappointed is always that of probability-possibility.

One situation exists in which this network of expectation breaks down. This happens when there is someone (or more than one person) to whom no norm or rule applies. Persons with no norms and rules applied to them can choose (though they do not necessarily do so) not to apply any norms and rules to members of any social cluster, but to act according to the maxim sic volo, sic jubeo (‘as I wish, so I command’). These people are tyrants. A tyrant not only excludes himself from the application of certain norms and rules; he regards himself as a person to whom they do not apply at all. Tyranny is the state of absolute injustice. In this situation all expectations break down. The relative causal nexus between acting in a certain manner and the consequences of this action is replaced by mere possibility and conjecture. One has no idea whether any promise will be kept; one cannot know whether any rule will be applied. Observing the law does not make it more probable that one will not be executed or imprisoned; at best, there is only a possibility of this. In the state of absolute injustice, ‘reading the tyrant’s mind’ is usually substituted for observing norms and rules.

Absolute injustice can be total or partial. It is total if the tyrant controls the whole of society, and partial if he controls a distinct social cluster (in which case we must speak of tyrants in the plural). Stalin’s Russia is a good example of the first. In Stalin’s time all expectations broke down. The relative causal nexus between acts and the consequences of acts disappeared. Slavery is a good example of the second, if – and only if – slaveholders are not obliged to observe certain norms and rules in the treatment of their slaves (as they indeed were in the ancient Greek city states). In the case of slavery (except if implemented in the state of total injustice), norms and rules apply to society, to all ‘free’ clusters of society, but not to slaves. It is always indicative of the nature of a society whether it tolerates or even encourages absolute injustice in the form of partial injustice. (Total injustice can be endured, but never encouraged, by any society.) There were, and still are, societies encouraging absolute injustice within the family. The tyranny of the paterfamilias is a typical form of partial absolute injustice.

The state of absolute injustice (total or partial) is the hotbed of hate, resentment and, occasionally, social conflict. Such conflict can assume violent forms. With total absolute injustice it often ends with the gesture of tyrannicide (if the tyrant does not die naturally). But the end result of this conflict, whatever form it takes, is the reinstatement (or creation) of the validity of norms and rules that were not applied in the state of absolute injustice. This ensures again the relative predictability of human action.

(2) Social conflicts may emerge because of an attempt to widen or narrow down social clusters to which certain norms and rules apply. Again, the validity of the norms and rules is not queried, What is questioned is the exclusion of certain groups from, or the inclusion of certain groups in, a particular social cluster. I am of course alluding here to conflicts fought to the slogan of equality or inequality. I stated at the outset that both equality and inequality are created by norms and rules. People are made equal by creating a cluster through the application of a certain set of norms and rules: alternatively, people are made unequal by applying different norms and rules to the members in another cluster. If the same norms and rules apply to each and every member of an entire society (regardless of the social groups to which these members belong), these norms and rules make everyone equal from the standpoint of the norms and rules in question. ‘Narrowing down’ or ‘widening’ the clusters to which the same norms and rules apply are only two aspects of the same procedure. ‘Equalization’ (in other words, the widening of the application of the same norms and rules to social clusters) can take three distinct forms. First, norms and rules applied to an ‘upper cluster’ may be applied to the ‘middle’ and eventually also to the ‘lower’ cluster (aristocracy, oligarchy and democracy in ancient Athens). Secondly, norms and rules hitherto applied only to the middle, and then to the lower cluster, may be applied to the upper cluster (the paying of taxes extended to the nobility, a form of financial equalization). Thirdly, norms and rules that hitherto have been applied only to the lowest cluster may be extended to the whole of society (for example, that no one should have property). Social conflicts related to these three clusters usually go hand in hand with testing and changing the norms and rules themselves; but this problem cannot be addressed from the viewpoint of the formal concept of justice.

The social struggle for the right to be different is another form in which the issue of equality is raised. Let me refer back to the problem briefly mentioned a short while back: if norms are religiously, politically or ideologically underpinned or overdetermined, the ingroup judges the members of another group with its own, and not with their own, norms and rules. The struggle for the recognition of the difference of a normative system is a struggle for equality, for what is being sought is equal recognition. The contest concerns the pluralism of value systems and the recognition of this pluralism, restricted (for instance, to two value systems) or general. The struggle for the general recognition of pluralism appears only at the dawn of modernity, and cannot be dealt with (at least in toto) in the context of social conflicts viewed from the perspective of the formal concept of justice. But the struggle for the recognition of a restricted pluralism can. If two systems of norms and rules are equally taken for granted by the members of respective clusters, one does not need to test the validity of either by granting at least limited recognition to both. Even though before the emergence of modernity this and similar forms of mutual recognition were usually only temporal, and non-recognition could be either reinvoked or reversed (previous outgroups imposing their norms on previous ingroups), social conflicts occurred repeatedly over this issue.

(3) If different sets of norms and rules apply to two clusters of people within the same milieu or life world, and the asymmetry of the members of the two clusters is constant, the relation of the members of the two clusters is socially unequal. This is the social constellation that comes first to mind in connection with the dictum that unequals should be treated unequally. Of course, there are other kinds of inequality; for instance, inequality in merit or excellence. The Aristotelian idea of proportionality refers precisely to this kind of inequality. Proportionality in action, judgement and distribution presupposes the existence of a common yardstick (a common set of norms and rules) for measuring merit or excellence. The imperative of proportionality (another name for the principle of consistency) is valid in the treatment of people who are each other’s social equals. Proportionality should be effective among equals. This is the idea underlying Aristotle’s discussion of commodity exchange. In his view if two kinds of commodities are exchanged, there must be one common (equal) standard with which to establish the proper proportion. And Aristotle is only highly consistent when he comes to the conclusion that it is convention, certain norms and rules of economy, that applies to each and every person who performs any kind of labour to produce commodities. Equality is to be understood here, just as everywhere else, in the sense that the same norms and rules apply to each and every member of the cluster, and proportionality is established by the yardstick of this norm by which one measures different forms and acts of labour.

What happens if no common norm or rule applies to two social clusters? Here the imperative of proportionality becomes absolutely meaningless, because respective members cannot be compared. In Aristotle, one in fact cannot compare ‘proportionally’ the action of a free man with the action of a slave, the merit of a free man with the merit of a woman, the merit of a father with that of his son, a minor. One can compare the merits of free men only with those of other free men, the merits of women only with those of other women, the excellence of slaves only with that of other slaves. The moment we compare all of them, we declare thereby that men and women, masters and slaves, adults and minors in fact belong to the same cluster; or at least we do so in one important respect, in that there are certain norms and rules which apply to all of them.

Asymmetry, or social hierarchy among various clusters or essential categories’ of people, implies, however, something above and beyond the social division of norms and rules. We should remember the ancient dictum that to be just means to treat equals equally, and unequals unequally. The question implicitly, raised here is this: who can treat whom equally or unequally? Equals can treat their equals, those in higher social positions can treat those in lower social positions, but those in lower social positions can never treat people in higher social positions, either equally or unequally or proportionately. Under the conditions of an all-embracing social hierarchy, both being just and being unjust are prerogatives allotted to the repositories of social authority. The ‘God-anointed king’ can be just or unjust with his subjects, but his subjects cannot be either lust or unjust with him (except when he is a mere primus inter pares), only loyal or disloyal, faithful or treacherous. Similarly, a paterfamilias can be just or unjust with his wife, but the converse cannot hold: his wife can only be obedient or disobedient, good or bad, faithful or unfaithful, and the like. The master refers to faithful servants, obedient slaves, never just ones; or, if he does, he has in mind servants and slaves who are just to their fellow servants and fellow slaves. Gods can be just and unjust; the ‘living God’ can only be just. Creatures, on the other hand, can transgress or sin against gods or God, but they cannot be unjust either with them or Him. Not only is vengeance mine, but judgement is as well these are the words of a consistent God.

One could object to all this that, if the members of a hierarchically higher cluster can be just or unjust to everyone belonging to a lower cluster, the members of this lower cluster can pass at least one judgement on the members of this higher cluster. They can say the following; that X (a member of a higher cluster) treated us justly (or unjustly). This is clearly possible. But the gist of the matter is that members of the hierarchically higher cluster cannot be treated by the members of the lower cluster either justly or unjustly. Members of the lower cluster cannot confer anything upon the members of a higher cluster, neither wealth, nor position, nor honour. Even, if they pass judgements on the members of a higher cluster, this judgement is meaningless to the latter because it bears no authority. They do not feel honoured by a complimentary opinion, and they are not ashamed if blamed. Whether right or wrong in these judgements of their hierarchical superiors, hierarchical inferiors cannot do them either justice or injustice, at least not in any direct way. But even the indirect way is seeking Justice rather than doing justice, in that it is supplication submitted to an authority superior, or at least equal in standing, to the member of the hierarchically higher cluster. This authority can, then, lend an ear to the grievance of the members of the lower cluster (if he wishes to do so) and rectify injustice.

Hierarchical relations thus imply two kinds of inequalities:

i inequalities which are due to the mere existence of ‘lower’ and ‘higher’ social clusters to which different sets of norms and rules apply;

ii inequalities due to the fact that the members of a lower cluster cannot treat the members of a higher one either justly or unjustly.

If (ii) is the case, (i) Is necessarily the case. If (i) is the case, (ii) is not necessarily the case.

If there are no common norms and rules at all for two clusters of people in a hierarchical interactive situation, (i) and (ii) both hold. If there exist some common norms and rules, but the set of norms and. rules differs (and it must, otherwise we could not speak of two social clusters), then (i) is certainly the case, but (ii) is not, as there exists’, or at least virtually exists, a common standard for measuring the actions and behaviour of people belonging to two hierarchically ordered clusters. One could roughly say that, if the hierarchical relation in question is a relation of personal dependence, (ii) is normally the case, even if there are some shared (common) norms and rules. If, however, the hierarchical relation is not a relation of personal dependence, and if (but only if) certain norms and rules apply to both clusters in question, then, even though (i) is still the case, (ii) is usually not the case.

Several social conflicts can emerge, in connection with the formal concept of justice, from the hierarchical (asymmetrical) relation of social clusters. In static societies, where norms and rules themselves are normally taken for granted, and the question of their justness or unjustness is not yet raised, almost all social conflict can be intelligible from the perspective of the formal concept of justice, and almost all springs from, or is interrelated with, the hierarchical (asymmetrical) relation of social clusters.

Those who cannot commit either justice or injustice in relation to the members of a higher social cluster must take the norms and rules for granted (otherwise they could commit both justice and injustice). Social conflicts can often become personalized, and here norms and rules remain unaffected; the tax-collector, not the tax system, is the target of hostility. In such cases it is anger and not indignation that fuels a conflict that may swell into rebellion. Anger and supplication are not mutually exclusive. Justice is done if the particular offender is dismissed or disciplined by the higher authority, and the members of the higher cluster properly observe their norms in interactions with members of the lower cluster.

Conflicts may occur around the transformation of personal dependence into a relationship of social hierarchy. At this juncture, a most important social category, that of the rights of the members of a lower cluster, makes its appearance. There is still an enormous difference between the situation where the members of the higher cluster must observe norms in their behaviour towards the members of the lower, and the members of the lower cluster can expect them to do so, and the situation where the member of the lower cluster have established rights, perhaps different from those of the higher cluster, but rights none the less. The difference can be summed up as follows. First, references to rights are substituted for supplication. Secondly, where rights exist, obligations exist as well, and it is a duty to describe what the obligations consist of. Accordingly, where members of a lower cluster ‘have’ certain rights, the enactment of justice or injustice can be relatively reciprocal. Sometimes the gaining of certain ‘rights’ does not change the norms and rules at all. But it does change the relationship between the members of the higher and lower clusters, for the members of the lower cluster can to some extent commit justice or injustice in relation to the members of the higher cluster.

Conflicts can also aim at the transformation of personal dependence into social hierarchy in such a primary relationship as marriage. Even if a woman is ‘treated well’ by her husband, the phrase infers that she has no rights: she is being acted upon, she cannot act upon her husband. She has duties to perform, and failure to carry them out is not injustice but disobedience. If the husband treats her badly and infringes his own norm of ‘treating a wife well’, the only channel of redress available to the wife is her father or brother (an equal authority), who may achieve justice for her where she cannot achieve justice for herself. If, however, she has certain rights (and obligations), justice and/or injustice can be relatively reciprocal, as under Roman law. But the complete transformation of personal dependence into a hierarchical relation is never possible in marriage, because the relationship is irrevocably personal. This is why social conflict cannot stop short of lifting the ‘cluster barriers’ between the Sexes; that is, stop short of applying exactly the same norms and rules to both men and women.

The Rules of Static Justice

Studies of justice often refer to the ‘golden rule’, the shorthand of formal justice. ‘I do unto you what I expect you to do unto me’ is indeed a rule, and a formal one. It tells me that I should do unto you what I expect you to do unto me, but it leaves open the question of what exactly it is that I should do.

The golden rule can be interpreted in the following way: ‘I do unto you the same as I expect you to do unto me.’ Although this is not the only interpretation and the only way to observe this rule, it can be stated that the golden rule applies only to those interactions where the rule can be precisely so interpreted. In short, the golden rule only applies to symmetrical relations, never to asymmetrical ones. Let me recall the preliminary definition of the formal concept Of justice: the norms and rules applying to a social cluster should be applied to each and every member of this cluster. The golden rule advises us how to act in order to observe static justice. If the norms and rules should apply to every member of the social cluster, they apply to any A, or any B of that cluster. Thus A should act in the same way unto B as B should act unto A: I (A) expect B to observe the norms and rules in acting unto me, therefore I (A) should observe the norms and rules in acting unto him.

The presumption that the golden rule has the status of an unrestricted generality Is to be attributed to the Jewish-Christian tradition, under which at least some norms, and fairly abstract norms at that – to begin with, the Ten Commandments – were to be applied to everyone. In other words, there were (abstract) norms valid for everyone, regardless of the social cluster to which they belonged. In principle every A had to observe the commandment, of the Lord in interaction with every B, and could expect every B to do exactly the same. Yet it is common knowledge that even abstract norms were concretized and interpreted in different ways within the sets of norms and rules pertaining to different social clusters Observing the presumed generality of the golden rule was the case of supererogation. The universalization of certain values and the subsequent dismantling of the social division in moral norms along the lines of stratification had to occur before the golden rule could become a universal regulative idea, no longer restricted to the observance of a few common norms. As such, it functions not as a rule but as the idea of a rule.

To sum up: the golden rule can orient us in all our actions only if human relations are symmetrical – that is, if our interactions are those of social equals, though not equals in merit or excellence. As we already know, if the same norms and rules apply to each, any member must be treated proportionately (according to merit and excellence), where proportionality presupposes social equality (the common yardstick). Viewed from this angle, various formulae involving the complete or restricted application of the golden rule have to be distinguished.

(1) I do unto you the same as I expect you to do unto me.
This is the formula of absolute reciprocity. Equality is presupposed, but not proportionality. (‘I help you in your need: I expect to be helped by you in my need’.)

(2) I do unto you what you deserve; I expect you to do the same thing unto me if I deserve it.
This formula does not imply that I expect you to do the same thing unto me as I do unto you. When I express my gratitude for something you have done I do not expect you to express gratitude to me, unless I have done something of equal merit. Proportionality is taken into consideration.

(3) I do unto you what you deserve; you should do the same thing unto me if I deserve it.
Again, this formula does not imply that I expect you to do the same thing unto me that I do unto you. But it differs from formula 2, for I do not expect you to do the same thing unto me. This is the formula applied in the case of just punishment. I do not expect to do something that deserves punishment. But I can punish only to the extent that others should punish me if I deserved it. Proportionality is taken into consideration.

But if we turn to the discussion of asymmetrical human relations, we shall immediately see that none of these formulae for the golden rule applies to them. The master cannot tell the slave or the servant, the lord the serf, the husband the wife, ‘I feed you well and I expect you to feed me well.’ Or, conversely, the slave, the servant, the wife cannot say to the master, the lord or the husband, ‘I serve you faithfully and I expect you to do the same.’ The master cannot tell the slave, ‘I do unto you what you deserve. I expect you to do the same thing unto me if I deserve it.’ In fact, the golden rule turns out to be completely nonsensical if we apply it to asymmetrical human relations.

This, however, does not mean that there are no abstract rules of justice to be applied to asymmetrical social-personal relations. Normally, the set of norms and rules of higher clusters includes norms for action and behaviour in relation to the members of lower clusters, and vice versa. Of course, there are exceptions, and we have referred to them in our brief discussion of tyranny. In tyranny, there is no justice at all, and there are no rules of abstract justice either.

The rules of abstract justice in asymmetrical relations (which are far from golden) are the following.

(i) I do X unto you; I expect you to do Y unto me.
It is in this sense that Confucius’ definition of reciprocity, ‘What you do not want done to yourself, do not do to others’ is interpreted in The Doctrine of the Mean (XIII 3f.): ‘To serve my father, as I would require my son to serve me ... ; to serve my prince, as I would require my minister to serve me ... ; to serve my elder brother, as I would require my younger brother to serve me ... ; to set the’ example in behaving to a friend, as I would require him to. behave to me.’ The doctrine makes a strong distinction between asymmetrical and symmetrical reciprocity. Asymmetrical reciprocity is emphasized by the use of the verb ‘I serve’, symmetrical by the use of the verb ‘I behave’.

This general formula implies that I do such because I am what I Am, and I expect you to do such because you are what you are. (Supreme authority, God, says, ‘I am what I am.’) Thus, ‘I Command, you obey’ (‘I do X unto you; I expect you to do Y unto me’) implies that I am a commanding officer, you are a private; I am your husband, you are my wife, and so on.

This is the formula of asymmetric reciprocity. It is asymmetrical because X and Y are not interchangeable. But it must include reciprocity, provided that the members of the higher cluster can be both just and unjust to the members of the lower cluster, as we have assumed. As a result, the formula ‘I do unto you; I expect you to do unto me’ can be read as follows: ‘I have spent my whole life in your service; I expect you not to let me starve in my old age.’ This is a matter of justice, because the servant can expect that in all probability he will be fed in his old age if he spends his working life in his master’s service.

(ii) I do X unto you as you deserve. You should do Y unto me, for I deserve it (because I did X unto you).
This can only be applied by members of the higher cluster to the members of the lower. ‘I liberated you from slavery, for you were faithful, and now you should be faithful to me because I liberated you.'

(iii) I do X unto you, which is due to you; I expect you to do Y unto me because I have done what was due.
For instance: ‘I followed you through all adversities; I expect you to be kind to me because I ... .'

(iv) I do unto you something you deserve; you should accept it.
The formula applied in respect of just punishment is not reciprocal; not even analogic reciprocity can be fathomed in it. at least not from the perspective of the formal concept of justice. And, given that the rules of justice, all formulae included, are but the rules of static Justice (those of the formal concept of justice), we must terminate our discussion at this point.

The Ideas of Justice from the Perspective of the Formal Concept of Justice

Perelman enumerates the following ideas of justice:

1 To each the same thing.
2 To each according to his merits.
3 To each according to his works.
4 To each according to his needs.
5 To each according to his rank.
6 To each according to his legal entitlement.

Nicholas Rescher [6] lists the following possible ideas of distributive justice:

1 As equals (except possibly in the case of certain ‘negative’ distributions such as punishment).
2 According to their needs.
3 According to their ability.
4 According to their efforts and sacrifices.
5 According to their actual productive contribution.
6 According to the requirements of common good or public interest, or the welfare of mankind, or the greater good of the greater number.
7 According to a valuation of their socially useful services in terms of their scarcity in the essentially economic terms of supply and demand.

Ideas (alias principles) of justice can be discussed at two levels: as constitutive or as regulative ideas. In discussing the constitutive use of ideas, we have in mind a use in which the ideas actually prescribe the norms of proportionality. For example, the ideas involved in saying ‘to each according to his merits, works, rank’ prescribe the Ways of establishing proportionality, whereas the idea ‘to each the same thing’ disallows proportionality of any kind. In discussing regulative ideas, one has in mind philosophies, social theories or ideologies which suggest the acceptance of one or two all-embracing ideas that would prescribe methods of establishing proportionality in, order to make society just. As yet, such ideas do not concern us, because they raise the issue of the ‘just society’ (an Issue I shall address in chapter 5). Here, I restrict my discussion to the ideas of justice as constitutive ideas.

‘The ideas of justice mediate between the substance of justice – that is to say, the concrete norms and rules applying to each and every member of a social cluster – and the criteria of justice: that is, the values on which justice is based, which are always values other than justice. For example, the idea ‘to each according to his merits’ leaves wide open the question of what merit we have mind, what kinds of actions are considered meritorious, and what kinds are not. All sorts of values, such as the good of the state, the salvation of the soul, fraternity and efficiency, can be mediated by the idea ‘to each according to his merits’. ‘What merit really Is’ will be different in all cases.

Perelman starts his discussions with the remark that the ideas of ce he lists are irreconcilable. He later solves the problem of irreconcilability by the introduction of the formal concept of justice, which encompasses all ideas of justice and to which I too subscribe. Indeed, in introducing the formal concept of justice he discovered maxim of justice, reconsidering all ideas from the viewpoint of maxim. In the main, I shall follow this example. But I wish to emphasize that the ideas are irreconcilable only if they are supposed to regulate or constitute the same social or human relationship, for they can be easily reconciled if they are supposed to regulate or constitute different social or human relations (or the same relations from a different viewpoint). As constitutive ideas, all of them have actually regulated human intercourse simultaneously. Should we even have in mind the regulative use of ideas (whereby one idea is supposed to be all-encompassing in order to establish the ‘just’ or ‘the good’ society), the simultaneous application of different ideas is not excluded. Moreover, I have never heard it suggested that any of these ideas should apply exclusively to any society, real or imaginary. For example, the ideas ‘to each the same thing’ and ‘to each according to needs’ were (and are) suggested as ideas of distributive justice, and this alone. The first concerns an egalitarian society, the second the ideal of a society beyond justice. Whatever our opinion of these ideas might be, they restrict the application of the idea of justice to the distribution of material wealth (and occasionally of other kinds of wealth). When, for example, Marx distinguished between the so-called ‘first’ and ‘second’ stages of communism in his Critique of the Gotha Programme, seeking the acceptance in the first stage of the principle ‘to each according to his work’, and in the second that of ‘to each according to his needs’, he was clearly referring to the distribution of material wealth. Perelman struggles with self-created ghosts when he writes as follows about the idea ‘to each the same thing':

According to this conception, all the people taken into account must be treated in the same ways, without regard to any of their distinguishing particularities. Young or old, well or sick, rich or poor, virtuous or criminal, aristocrat or boor, white or black, guilty or innocent – it is just that all should be treated in the same way, without any discrimination or differentiation. in popular imagery the perfectly just being is death, which touches every man on the shoulder regardless of any of his privileges.[7]

But is there such a thing as an all-encompassing idea of justice regulating all social relations? To my knowledge there is not. At the same time, it can be safely stated that the idea (or principle) to leach the same thing’ appears as one regulative idea of justice (among many) in several different social relations, legal justice being perhaps the most important. Of course, it has its obvious limitations. For example, Rescher realized (even if he mentions it only in an aside) that the idea ‘to each the same thing’ cannot be rationally applied at all in the situation he calls ‘negative distribution’ (punishment).

Put properly, the problem we are facing is as follows. Although different ideas of justice apply to all (real or imaginary) societies, there are always dominating ideas of justice, as well as restricted ones (of greater or lesser degree). Indeed, certain dominating ideas may exclude the validity and application of certain other ideas operative in a society as a whole. Other dominating ideas can relegate lesser ideas to a subordinate position, to particular realms, or may allocate only a corrective function to them, though without eliminating them.

Even though I believe that Perelman fought against logical possibilities and not genuine conceptions, I shall follow in his footsteps by =her analysing the idea ‘to each the same thing’. Let us imagine that the same norms and rules apply to all humankind, and thus must be applied to every human being. Here, all human beings would have exactly the same rights and obligations. The principle would then be ‘to each the same’. Application of the idea ‘to each the same’ would not exclude application of the idea ‘to each according to his merits’ or ‘to each according to his excellence’, because different people could live up to more or less the same norms, and could be judged accordingly (proportionately); but it would certainly exclude application of the principle ‘to each according to his rank’, which implies the exclusivity of certain rights and obligations. There are numerous possible and actual interpretations of the idea ‘to each the same thing’, but even the most egalitarian of them (which recommends equality in the distribution of material goods) is far from nonsensical, as much as I disagree with it.

In fact, we always operate with the principle ‘to each the same thing’, even if it is far from being the dominating idea. At this point we may recall the example of the High School Certificate. According to its charter, everyone obtaining more than 340 marks is eligible to enter the law school of any university. The principle ‘to each the same thing’ is applied, then, in this case. If women struggle for equal wages for equal work, their struggle Is guided not by the Idea ‘to each according to her work’, but by the idea ‘to each the same thing’. They believe that exactly the same norms and rules should apply to their work as to the work of men; that the yardstick by which the idea ‘to each according to his/her work’ is measured should be exactly the same for both sexes. Moreover – this already follows from the Aristotelian interpretation of equal proportionality – the idea ‘to each according to his merit’ (or demerit) cannot be applied except in conjunction with the idea ‘to each the same thing’. If people should be punished in proportion to the crimes they commit, two persons who commit exactly the same crime should suffer the same type and amount of punishment, whether black or white, rich or poor, sick or well, aristocrat or boor, young or old, to repeat Perelman’s examples. In other words, the idea ‘to each the same thing’ applies.

This brief discussion of ideas illustrates by way of example my initial thesis that no idea of justice can be exclusive, that the various ideas are in fact not all irreconcilable, whilst at the same time certain ideas are indeed irreconcilable with cert am others. If we analysed the other ideas listed by Perelman we should reach exactly the same conclusions. I should mention here that I am only discussing Perelman’s list because Rescher’s list is more haphazard and also fairly inconsistent. Rescher confuses the criteria of justice (values) with the ideas of justice. In point 6 he lists the following ideas: ‘According to the requirements of common good or public interest, or the welfare of mankind, or the greater good of the greater number’. However, these are (mostly utilitarian) criteria of justice, not ideas of justice. In point 7 he simply defines the content of an idea of justice: ‘According to a valuation of their socially useful services’, where ‘socially useful services’ stands for ‘merits’ or ‘works’. It is only in points 1, 3, 4 and 5 that he enumerates ideas proper, if in a fairly confused way. In point 3 he mentions ‘According to their merit’, and in point 4 ‘According to their efforts and sacrifices’, and does so as if these were two distinct ideas, although ‘effort’ and ‘sacrifice’ are really only two sub-cases of merit. These inconsistencies make it better to work from Perelman’s list.

It should be noted that in my analysis of the formal concept of j ustice it sufficed to operate with Perelman’s ideas 1, 2, 3 and 5. In discussing asymmetrical relations, I referred to the idea ‘to each according to his rank’ (5); in discussing symmetrical relations I referred to the ideas ‘to each according to his merits’ (2) and ‘to each according to his excellence’ (3), [8] and finally to the idea ‘to each the same thing’ (1). For very good reasons, I did not discuss either the idea ‘to each according to his needs’ (4), which also figures in Rescher’s list, or the idea ‘to each according to his legal entitlement’ (6).

The idea ‘to each according to his rank’ is operative in every hierarchical society. The ideas ‘to each according to his merit’, ‘to each according to his excellence’ and ‘to each the same thing’ can be operative in any society, real or imaginary. If norms and rules apply at all (any norms and rules), these three ideas apply as well. Given that we cannot imagine the existence of any society that does not recognize the validity of at least a few norms and rules, neither can we imagine a society in which these three ideas of Justice do not apply, regardless of which is dominant. Nevertheless, should we imagine such a society, it will be beyond justice, a society having either justice nor injustice. I shall return to this problem in chapter 5. For the time being it will suffice to bear in mind that, if we mention justice in universal terms, we speak of the application of exactly these three ideas of justice.

Thus the ideas ‘to each the same thing’, ‘to each according to his merits’, ‘to each according to his excellence’ are the universal ideas of justice. ‘To each according to his rank’ is a general idea of justice, operative in every hierarchical society.

‘To each according to his legal entitlement’ is neither a universal nor a general idea. To ‘have’ a ‘legal entitlement’ there must exist, apart from an ethical-normative structure, a legal system, which is not always the case. Also, in almost all pre-modern societies, ‘legal entitlement’ was identical with rank, the reward for excellence. In modern societies, at least in democratic states (and in all democratic theories), the ‘rights of man and citizen’ have been established, and so ‘to each according to legal entitlement’ is identical with ‘to each the same thing’. As a result, the idea ‘to each according to his legal entitlement’ can only be conceived of as a restricted idea applied to distributive justice, and based on property ownership. The acceptance of a restricted idea of justice while advocating universal ideas of .justice caused a lot of theoretical trouble to those philosophers aware of the inconsistency this involved. This is why Locke made excessive efforts to prove that property results from work (excellence), and that the idea ‘to each according to his legal entitlement’ is only an example (sub-case) of the application of the idea ‘to each according to his excellence (works)’. All theorists who have challenged the right of inheritance have wanted to eliminate the idea ‘to each according to his legal entitlement’ as applied to property by suggesting that such legal entitlement (to property) should be the reward for excellence and merit alone. For this to occur, ‘property’ would have to be shifted from the context according to’ to the ‘to each

Before discussing the idea ‘to each according to his needs’, which In my view is not an idea of justice, I should like to mention an idea ‘Of justice which has not appeared in either list scrutinized so far. It is easy to grasp the reasons for its absence: first, it can never become the dominating idea of justice; and, second, in Perelman’s view it is identical with the idea ‘to each according to his needs’, which he did list. This idea is ‘to each what is due to him by virtue of being a member of a social cluster or essential category’. If charity is a social norm, those belonging to the cluster ‘starving people’ should be provided with food Irrespective of their merit and excellence.

However, this does not mean or imply that we provide them with food ‘according to their needs’, only that we feed them because th ey are in need’. (I analyse this example because Perelman used it to illustrate the relevance of the idea ‘to each according to his needs,.) It depends on the norms and rules of a society whether life itself has a claim (Irrespective of social rank, merit or excellence), though of course it often has. Certain cultures have developed norms to be applied to foreigners and members of an outgroup, and these define precisely what is due to such people. At this point the Old Testament comes to mind: if you do not gather the crops from the margin of your field, but leave something for the alien as well, You simply ‘give them a chance’. Once again, if the enemy’s children are spared, it would be strange to claim that this is due to the idea ‘to each according to his needs’. In other words, the relevant idea of justice here is not ‘to each according to his needs’, but ‘to each what is due to him’ (by virtue of being the member of a particular cluster, because of a socially acknowledged norm). If proper education is due to the children of a particular stratum, and a child from this milieu is brought up in ignorance, he can justifiably accuse his parents of injustice; this is part of a well-known tale from Shakespeare. It could be stated that this idea of justice is only a subcase of the idea ‘to each the same thing’, but this is not generally true, even if the two principles can occasionally overlap in practice. The norm that all sick people should be treated does not imply that all must get the same treatment (‘the same thing’). This would be ludicrous. Nor does educating all the children of a family mean education in the same thing; similarly, charity does not entail the distribution of equal amounts of food of identical quality. Of course, Perelman was not misguided when he applied the idea ‘to each according to his needs’ to at least one of the above-mentioned cases: food, life, good education, medical treatment – all of these are indeed needs. All the examples of the application of the idea ‘to each what is due to him by virtue of being a member of a social cluster or essential category’ have something to do with needs, whereas the examples of the application of the principle ‘to each the same thing’ have not, with the exception of the example concerning human rights, where the application of the principles overlap.

Perelman was aware that the idea ‘to each according to his needs as it stands, is not an idea of justice. Instead, he suggested a restricted version of the idea: ‘to each according to his essential needs’. However, the crux of the matter is that we simply cannot divide needs into ‘essential’ and ‘Inessential’ ones, as Perelman himself admits elsewhere. Apart from mere drive reduction, there is nothing naturally ‘essential’ in needs, and even the objects of drive reduction are socially patterned. Food is essential for human living, but what kind of food is considered essential is an entirely different matter. Systems of needs are shaped by the norms and rules (the values) of each and every society, and thus different systems of needs are allocated to different social clusters. The systems of needs allocated to a social cluster are recognized needs. It is generally presupposed that members of a social cluster ‘have’ needs which are allocated to them by the norms and rules. Thus the needs recognized as ‘essential’ needs of each and every member of a social cluster are precisely those that have been allocated to the very cluster to which they belong. It depends on the norms and rules as to which needs are considered ‘essential’ and which not. If, for example, ‘good education’ is allocated by the norms and rules to the members of the social cluster ‘nobility’, the need for good education is allocated also to the sons of noble families, and so any father who keeps his sons in ignorance is unjust, for he withholds the satisfaction of a recognized need. If the normative system of a society recognizes life as a value (which is far from always being the case), then the need for mere survival is recognized as well, and the allocation of food to those in need is meant as the satisfaction of an ‘essential need’, and is therefore a just act.

This why I believe that even the restricted version of ‘to each according to his needs’ – that is, ‘to each according to his essential needs’ – is inconclusive, for it involves a naturalistic fallacy. And so I would replace it with the idea ‘to each what is due to him by virtue of being a member of a social cluster or essential category’, which in my view formulates clearly the crux of the matter. The norms and rules themselves define here what is due to someone by virtue of being a member of a social cluster or essential category, and, if someone does not receive the material and/or spiritual goods or the proper means due to him, injustice has taken place. Finally, it should be clear why I have augmented ‘social cluster’ with ‘essential category’. There are essential categories which are not connected with any social clusters. An example of this is that ‘medical treatment is due to sick persons’. Further, ‘essential categories’ can be distinguished within the same social cluster; for example, the sons of a nobleman all belong to the same social cluster, but the first, second and third son, because they may belong to entirely different ‘essential categories’, may receive very different forms of education, as history attests.

However, individual systems of needs are very rarely, if at all, completely identical with the system of social needs allocated to each and every member of a social cluster or essential category. Usually, socially allocated needs define the framework in which personal systems of needs may evolve. That which is due to me or those like me regulates my aspirations, excites my fantasy or keeps it under tight control. Yet the particular system of individual needs varies from person to person. No norm or rule can completely shape the system of needs; in other words, no norms and rules can apply to the ever-existing systems of individual needs. Moreover, norms and rules have a levelling function. Their character includes abstraction from the concrete system of needs which we can call individual as against social needs. Since individual and social needs almost never completely coincide, and no norm or rule applies to the former, justice stops short of the singularity of the person. Individual structures of needs are the limits to justice. Of course, an idea pertaining to individual needs can, in principle, be formulated and brought close to the form of the ideas of justice. For instance, it can be said, ‘to each according to his singularity’. But this idea would no longer be an idea of justice. Singularity is by definition beyond comparison and ranking, and without the latter there is no justice. Grasping singularity is an interminable task. One can state that in general judgement is passed on the finite, not the infinite, a constellation clearly grasped by all religious concepts of justice. The French proverb tout comprendre, c'est tout pardonner is perhaps morally suspect, but it certainly conveys an important message. To comprehend everything is an infinite task, and a full understanding of even a single action of a single person may involve an infinite number of interpretations and reinterpretations. As a result, if one intends to understand everything, one can neither forgive nor condemn. Still, judgement is passed via the application of norms and rules, and one should understand only so much of a person’s action as to allow the application of the proper norms in this process of judgement (by comparing the action in question to the actions of other people).

So far I have briefly argued that the idea ‘to each according to his needs’ is not an idea of justice. The reader must bear in mind that here we are dealing with ideas of justice in their capacity of constitutive and not regulative ideas. Though this is only a preliminary discussion, some further qualifications should be made. Even if a typology of singular needs is not possible (for they are singular, and so cannot be typified), a typology of relations between singular systems of needs and various sets of norms and rules can be constructed. The attempt I shall make at this is nothing more or less than that: it is an attempt, and does not lay claim to being a complete explanation.

(1) At least in cultures such as our own, there are moral norms which are not norms of justice because they do not apply to every member of a cluster. Apart from certain others (which do not need to be discussed here), these are the norms that apply to personal relations, qua personal relations. The norm of satisfying the needs of others stands out among them as the most important: in other words, goodness (or ‘goodness of heart’). If a friend or someone we love says, ‘I need you’, the answer is, ‘Here I am.’ If that person says, ‘I need this’, the answer is, ‘Here you are.’ The person who says, ‘I need you’ or ‘I need this and this’, is not obliged to argue on behalf of his need or for this ‘something’. Anyway, needs cannot be argued for. Rather, the converse is true: it is the one who has been addressed but has rejected the appeal who has to argue on behalf of his non-action. The justification principle elaborated by Marcus Singer applies also in this case. [9] Justification must take recourse to norms other than the norms we fall to observe. Thus I can justify my refusal to satisfy the needs of others by pointing out that the satisfaction of the need in question would infringe certain norms or values, including the value of justice. If the norm is that sick persons should be treated, then every sick person should be treated. This is a matter of justice from the viewpoint of the idea ‘to each what is due to him by virtue of belonging to a social cluster or an essential category’. But a person watching over a sick friend day and night does not practise ‘Justice'; he does not satisfy a social, but rather an individual, need. While he is indeed acting to satisfy the need of his friend (‘I need you’ – ‘Here I am’), this and similar acts cannot be subsumed under the idea ‘to each according to his needs’, for no society can ‘provide’ people with good friends. Not even the following formulation is possible: each friend of a sick person must attend the latter day and night. This formulation is not possible because, for a start, some sick people often prefer to be left alone. The satisfaction of personal-singular needs is one of the most sublime virtues, but because of this it is beyond justice.

(2) By allocating certain social needs, at the same time normative systems ‘outlaw’ some others. They can do this in a relative and in an absolute sense.

Needs are ‘outlawed’ relatively if social clusters are hierarchical and the relation between their members is asymmetrical. Consequently, needs that are socially allocated to the members of a higher cluster can only become individual in respect of the members of the lower clusters. Education is not ‘due’ to a serf, and so his claim to education is not considered just.

Needs are ‘ outlawed’ absolutely, and must remain purely individual, if they are unrecognized by norms and rules whatever social cluster their ‘bearers’ might belong to. What I have in mind is not ‘repression’ in general, which is too broad a problem to be discussed here, but a phenomenon which can be illustrated by the following example. Politicians normally feel the need for power; that is why they choose politics as a profession. However, in a democratic body politic they have only limited power, and their individual need for unlimited power is outlawed. This need cannot even be expressed in the form of a just claim, so long as the prevailing social conditions (political democracy) remain.

(3) Even if the idea ‘to each according to his needs’ is not an idea of Justice, it can still be applied as a corrective idea of Justice, as an idea of equity. Whichever idea of justice should be applied in a given action, those who are ranking and comparing, acting, judging and distributing can take into consideration the singularity of each and every person, and first of all the singular system of needs of the persons in question, though this can occur only to a degree. This degree is determined and prescribed by the norms themselves, whereas, in the case of the application of rules, equity, is out of the question. Equity, as corrective justice, can modify judgement, action and distribution and decrease the rigidity of norm application. None the less, the same norms and rules as otherwise establish the particular idea of justice prevailing here have to be applied without extreme rigidity. Corrective justice is thus a sub-case of justice, as Aristotle correctly claimed. If one considers the singularity of persons in grading, ranking, acting, judging and distributing through violating the norms of justice, one cannot be either just or equitable (though one can still be good). However, equity is usually conditional, whereas justice is always unconditional.

It should also be remembered that the idea ‘to each according to his needs’, as an idea of equity (corrective justice), does not entail the application of this idea in full. Only those types of needs are considered which are absolutely relevant from the perspective of the application of a norm (normally there is only one such need). Further, ‘to each according to his needs’, as t e principle of equity, very rarely includes the satisfaction of needs; rather it urges the provision of means for the satisfaction of certain needs. Also, there are needs requiring satisfaction to which neither justice nor equity applies, such as the need for love.

The Formal Concept of Justice and ‘Humankind’

Contemporary political action is not confined to one country or one culture, nor even to several countries or cultures. ‘World history’, once a construct of philosophies of history and applied ‘to all times’, has become the reality of our time. Different cultures (thus different histories) not only share the same planet; they also share the fate of this globe. When we mention ‘humankind’ we have this undeniable fact in mind. ‘Humankind’ is defined as the ‘sum total of human beings who inhabit our globe’.

Since the dawn of modernity, philosophies have taken excessive pains to discover the common elements of all existing the possible human cultures. The notion ‘culture’ itself is a product of this inquiry. ‘Common elements’ must not only be factually and verifiably common; they must indicate the ‘essence’ of the human species as well. They must include propensities other than the merely visible. These essential features, which are attributed to all human beings (in all human cultures), are called human universals. When referring to humankind, we may have these human universals in mind, or at least we have one of them in mind. We can refer to humankind as ‘Homo sapiens, by virtue of X’, where X stands for any human universal. (The scientific category Homo sapiens was also coined in modern times, and it has no meaning at all without combining this seemingly purely ‘anatomical’ notion with at least one human universal.)

From the viewpoint of values, human universals are construed as interpretations of certain empirical facts. True enough, a form of evaluation is also involved in the definition of humankind as ‘the sum total of human beings who inhabit our globe’. However, to operate with a slightly modified Kantian categorical framework, ‘humankind’ defined as ‘the sum total of human beings who inhabit our globe’ is a merely empirical concept, whereas defined in terms of ‘human universals’ it is not an empirical concept, even if the definition involves recourse to interpreted empirical facts. ‘Humankind’ understood in this way is rather a theoretical idea in which values are embedded. The merely empirical definition of humankind is not binding, for the imperative ‘As a member of humankind, you should act such’ does not follow from the assertion that you share the globe with members of the same species. Yet not even the definition of humankind through human universals is morally binding. Whether ‘work’ or ‘language’ or ‘division of labour’ figures as such a universal, it falls to provide us with any moral imperative.

Even so, those who promoted the ideas of ‘freedom’ or ‘reason’ to the rank of human universals were not simply offering a theory; they were proposing a moral commitment as well. The rediscovery and reinterpretation of ‘natural law’ bears witness to this commitment. The recognition of a certain kind of ‘natural law’ (the secularized divine law) which must not be infringed by any positive law includes a practical commitment, at least In a negative sense. No other human universals offer this much. But even ‘natural law’ has serious flaws, the most severe being that (as Croce pointed out) it does not exist. In order to be committed to ‘freedom’ and ‘reason’, if only in a negative sense, one must first choose the idea of natural law, which by definition means that one can reject it as well (it is precisely in this sense that natural law does not ‘exist’). Secondly, even if there are many who interpret ‘natural law’ as a moral commitment (for instance, in the following form: ‘As a member of’ humankind you should respect the freedom and the reason of all human beings and should act such that they can make proper use of them’), this commitment does not directly follow from the theory of natural law (as a human universal). What does directly follow from the theoretical statement is the claim, ‘As a member of humankind I am born free and endowed with reason; I have the right to make use of my freedom and reason.'

The notion ‘humankind’ can also be defined by substantive goals. Evolutionist philosophies of history normally establish such substantive goals. Thus ‘humankind’ is not the sum total of persons who inhabit our globe, even when defined as all born free, equally endowed with reason and conscience, but is history itself, progressing towards the state of freedom and/or reason. ‘The greatest happiness of the greatest number’ is also a substantive goal, though this utilitarian creed is based on the assumption that ‘every human being strives for happiness’. Understanding humankind as history, and from the perspective of its end result, implies a commitment to act in harmony with the substantive goal. However, what a happy, free or rational humankind looks like is a matter of decision. Thus different interpretations of this end result carry different commitments, and often conflicting ones. Further, if the essence of humankind is an end state, the problem of the means to this end state must be raised. Action, judgement and distribution in the present are mere means to this end, and must be assessed through the substantive result they are supposed to promote. The universal claim of any action is therefore only indirect, never direct, and as such is not practical but merely pragmatic. Almost all kinds of actions can be rationalized in their capacity as means to the substantive goal.

Today, all political actors pay lip service to humankind. We all act ‘on behalf of humankind’, and we undertake many different things in this service. We may have entirely different interpretations of ‘humankind’ in mind: eventually, even the three interpretations of ‘humankind’ can merge. It is generally held that acting ‘on behalf of humankind’ is just. Hence, from the viewpoint of the actors, every action Is considered just, for all actors rationalize their actions as ‘on behalf of humankind’. But the crux of the matter is that ‘acting on behalf’ is neither a definition nor an idea of justice; in fact is has nothing to do with justice. ‘To act on behalf of X’ can be both just and unjust.

Of course, very different things are done in the name of humankind. In the last half century, certain acts have met with the full approval of our moral sense, our sense of justice, but certain other acts have met with our full disapproval as immoral and unjust. But, even In relation to the former, it would only cause confusion if we tried to apply the formal concept of justice to the actions in question.

I feel free to disregard here those actions committed in the name of ‘humankind’ and ‘Justice’ which meet with the disapproval of our moral sense and sense of justice. They are so numerous and widely known that they need no further comment. I intend rather to address an action which has met with the full approval of a general sense of justice. The Nuremberg Trials is an appropriate example for this purpose.

Clearly, the Nuremberg Trials were conducted on the basis of natural law; if they had not been, the notion of ‘crimes against humankind’ would make no sense whatsoever. Humankind has no common positive legal system, so one can infringe international legal agreements, but it is not possible to infringe the (positive) law of humankind. Such a notion presupposes that all members of the human species are born with certain inalienable rights, the most fundamental of which is the right to life. The violation of this ‘right of nature’ is the infringement of the supreme law, and those who have violated it have committed a crime against humankind, even if they act under the command and the authority of national and positive systems of law. Even those who do not generally subscribe to the theory of natural law feel that the verdict handed down in this trial was just in the most sublime meaning of the word, and did not represent merely the revenge of the victors over the vanquished (as Goring suggested in a remark on the moral status of the trial).

We feel it; we are aware of it; we are committed to it. But we cannot explain it.

Let us stipulate for a moment that ‘the right to life’ and ‘the right to liberty’ are in fact the rights of every human being. This would mean that we are all subject to the authority of the same norm. if the right to life and to liberty constitute the social cluster ‘humankind’, then every member of this cluster should be treated in accordance with these rights, and everyone should respect the life and liberty of all others (’to each the same’). Everyone who violates this norm must be punished, and in proportion to the severity of his offence (’to each according to his offence’).

This norm exists as a regulative, but unfortunately not as a constitutive idea. There is in fact no such norm. One judge presiding at the Nuremberg Trials was guilty of crimes similar to those he was sitting in judgement upon (crimes against life, against liberty).[10] Similar crimes have been committed a hundred times since and have remained unpunished. Through the Nuremberg Trials, humankind established a representative exception, but not a rule. We cannot explain why the judgement was just, even if we feel and know that it was just. In the verdict, the claim had been raised that humankind should become a social cluster, that it should become something which it is not. And it is still not a social cluster.

As has been mentioned, all representative political actors, whatever interests they may pursue, pay lip service to ‘humankind’. Everything is done ‘for the best of humankind’, ‘on behalf of humankind’, ‘In the interest of humankind’. This reference to humankind is almost empty, just as the reference to God was almost empty once, for it bestows a blessing upon good and evil deeds alike. In all ways, the simile is but approximative: the laws of God are valid for the members of the religious group which seeks His blessing, whereas one can refer to humankind in any of its abovementioned interpretations, among which only the theory of ‘natural law’ can constitute certain (if only virtual) binding norms. Thus the reference to humankind is emptier than the reference to God. It sounds sacrilegious to say that we act ‘for the best of God’, ‘on behalf of God’, and in particular ‘In the interest of God’. However, one can still act ‘for God’s sake’, for so to act means, first and foremost, to observe His commandments. One can claim to act ‘for the sake of humankind’ only if one observes ‘commandments’ that are binding for the human species. But no such ‘commandments’ (norms) are recognized by empirical humankind, by the ‘sum total of human beings who inhabit our globe’. And so after this detour we still come back to Kant: there is an unbridgeable gap between empirical humankind and the idea of humankind.

We reproach one another for applying double standards in our judgements In general, and in our judgements of political actions grid actors in particular. If identical acts of oppression and violence are committed in countries with different political institutions or cultural traditions, and X denounces these acts as ‘horrible’ in one case and justifies them as ‘appropriate measures’ in another, we accuse X of having and practising double standards.

The notion ‘double standard’ carries the meaning ‘unjust’, but why? According to the formal concept of justice, which is the all-encompassing concept of justice, if the same norms and rules apply to a social cluster, the same norms should be applied to every member of the cluster. ‘Double standards’ can be defined as follows: certain common norms apply to two groups of people that make them, even though they remain members of distinct groups, members of the same social cluster: nevertheless, we apply different norms to these two groups. The application of double standards is unjust because it contradicts the formal concept of justice. But to apply different standards to entirely different social clusters is by no means unjust. This is expressed in the Latin dictum quod licet Iovi, non licet bovi (What is permitted to Jove, is not permitted to an Ox’). Thus if we indignantly reproach X for applying double standards, we must first inquire whether the notion ‘double standard’ makes sense at all in the case under scrutiny. And we come to the extremely disquieting conclusion that what we so indignantly reject in certain actions and patterns of behaviour is the application not of double standards but of different standards, for the simple reason that humankind is not a social cluster. There are no common norms and rules that apply to all human beings irrespective of their culture, and so there is no common yardstick for comparing and ranking human actions, not even a single yardstick recognized as unconditionally valid by each and every member of humankind.

At this point, our sense of justice will protest that we do have such a yardstick, namely humankind. But which humankind? History as telos can certainly not serve in this capacity; neither can the empirical notion of humankind or human universals. The only human universals capable of providing us with an answer (life, freedom, reason) can be chosen only within the framework of a theory of natural law, a theory we can accept as well as reject, and which, once accepted, is only conditionally binding. But our sense of justice continues to protest. This protest cannot be accounted for in terms of the formal concept of justice, for it transcends the latter’s sphere of competence. In chapter 3 I shall return to this problem of the ‘sense of justice’, but it must be stated here that, whenever this ‘sense of justice’ makes its voice heard, it heralds the claim for new norms and thus new social clusters, and this is why it does not contradict the formal concept of justice, even though it transcends its sphere of competence. The sense of justice protests against the application of ‘double standards’, although it is indeed confronted with the application of different standards. It is exactly this sense of justice which transforms the different standards into ‘double standards’. In so doing, our sense of justice raises the claim that humankind should become a social cluster, that humankind should be the essential cluster, that at least a few common norms and rules should apply to all cultures and all members of Homo sapiens, that there should be a common yardstick for comparison and ranking. The sense of justice which transforms different standards into ‘double standards’ (in that it perceives double standards where ostensibly, there are only different ones), implies a new notion of humankind, not identical with the empirical or the teleological-substantive concept of humankind, or with the definition of humankind through ‘universals’ alone. ‘Humankind’ as the essential human cluster constituted by certain common norms and rules is a concept of humankind which can only be conceived of as binding. For, if certain common norms constitute humankind, these very norms are valid for all human beings and should be applied to all members of humankind, and all members of humankind should apply them equally. And, further, to apply these norms to cases to which they, should be applied is just, whilst not to apply them is unjust; to apply them consistently is just, whilst to be inconsistent in this application is unjust. ‘Humankind’ as the essential social cluster is the unity of empirical humankind and the idea of humankind. It is empirical humankind because it encompasses the sum total of human beings who inhabit our globe. It is also the idea of humankind because membership of the essential social cluster is constituted by universal and binding norms. The norms which constitute the essential cluster cannot be both chosen and rejected; they can only be observed or infringed. Everyone is obliged to observe the common norm and to apply it to everyone else (to every human being) continuously and consistently. There is a common yardstick for comparison and grading; the formal concept of justice is validated.

The sense of justice which induces us to denounce the application of different standards as ‘double standards’, and therefore as unjust, cannot be accounted for but as the expression of the will to construe humankind as the essential social cluster.

Logical as it may sound, this solution leaves several doubts in mind. And these doubts stem from the same quarter as gave birth to the most reasonable, even though still unsatisfactory, idea of humankind – the area of human universals. What we have in mind here is the human universal of ‘culture’, which can be interpreted in the spirit of extreme cultural relativism.

In what follows, I shall argue that extreme cultural relativism contradicts the claim of the sense of justice. If we accepted extreme cultural relativism, we could no longer denounce the application of double standards. Extreme cultural relativism can make its appearance equally in Geistesgeschichte, in cultural hermeneutics and in straightforward positivism, though I cannot deal here with their respective specific features. It suffices to enumerate the basic tenets of such relativism.

Let us assume that all values, opinions, world views, patterns of actions and beliefs are so many prejudices, all relative to the time and space in which they are held. In any given period, including the present, there are several distinct coexisting cultures, each with its own values, opinions, world views and patterns of actions, all taken for granted as ‘beliefs’. Each is different from every other; there is no way to compare them. Yet it is not only cultures as wholes that cannot be ranked. There is not a single action pattern or custom of a particular culture that can be compared with any single and functionally equivalent action pattern of another culture. This is for the simple reason that all action patterns are embedded in their respective cultures as wholes. The same belief which is true and the same action which is right within the tradition of one particular culture may be false and wrong within the tradition of another. We cannot even raise the question of whether one action pattern is more correct (or better) than another, because there is no common standard by which an evaluation can be made. Whenever we carry out a ranking of any kind, whenever we apply common standards in judging the same action performed by people of different cultural backgrounds, we do so from the position of our own particular culture, from the standpoint of values and standards of measurement that we take for granted. Thus we superimpose our prejudices on other cultures. The application of double or even multiple standards is more just than the application of common standards to all cultures. Those operating with double or multiple standards Simply place themselves within the patterns of cultural traditions different from their own. They identify with the different cultural traditions and understand (interpret) their own heritage as an alien system of prejudices.

The logic of argumentation here is obviously fallacious. And the problem it raises, but does not solve, is undoubtedly real. Let us first analyse the logical fallacy and then proceed to the problem itself. Anyone who states that all cultures are unique, and that we cannot compare or rank various cultures from any common viewpoint moreover, anyone who holds that this is true not only about cultures as wholes, but also about all particular beliefs and patterns of action inherent in each and every culture – raises in uno actu a claim to truth and a claim to right (good) judgement. The claim is that the proposition ‘All cultures are unique and cannot be either compared or ranked by any standards’ is true, and that the norm ‘Do not compare or rank cultures by any common standards because they are unique’ is right. However, in so far as someone claims that this proposition is true and this norm is right, a comparison of various cultures has already taken place. For the following should at least be acknowledged by implication. All cultures being unique and equal, in the sense that every one of them contains true opinions and right norms (in that the norms and opinions prevalent in them are regarded there as true and right), those cultures holding the opinion that all cultures are unique and equal and claiming the norm that, as a result, they should not be compared or ranked are superior to those others not sharing such an opinion, for the simple reason that they claim one more true opinion and right norm than all other cultures. Given that Western culture is one of the very few cultures which at least permit the raising of such claims, it is not equal in value with all other cultures, but is superior, at least as far as one norm and opinion is concerned. The end result, then, is a highly ethnocentric proposition which shows the inner vulnerability of the entire model. However, there is more to be said on the matter than just outlining this (ineliminable) inconsistency.

Let me divide the argumentation of extreme cultural relativism into (1) statements of facts, and (2) imperatives or recommendations.

(1) It is a statement beyond reasonable doubt that human cultures are different and that each of them is unique. And, though not all critics of extreme cultural relativism would go along with me, I also accept the further proposal that unique cultures as wholes cannot be either ranked or compared. Whenever cultures are in fact graded or ranked in this way, ranking is performed by the standards of one culture among many, and the normative system of the assessing culture is taken for granted and recognized as the ‘true’ and the ‘right’ one, or at least the ‘advanced one’, in contrast to the baseness and imperfection, eventually the ‘primitiveness’ of the assessed cultures. What Is now called ‘ethnocentrism is the natural attitude of all cultures towards alien ones. Philosophies of history are only highly sophisticated and streamlined versions of this ‘natural attitude’.

Extreme cultural relativism deduces its imperative from this statement of fact. To deduce an ‘ought’ sentence from an ‘Is’ sentence is, in my view, a legitimate procedure. The gist of the matter in this case, however, is that the imperative deduced from the statement of fact does not follow from the facts which have been established.

(2) It indeed follows from the statement that all cultures are unique that they ought to be understood in their uniqueness. However, the imperative that they ought not to be ranked, compared or graded at all does not follow from the above assertion. Let us briefly analyse an analogous case. All individuals are unique. One can never fully understand individuals. The process of their understanding is in principle Infinite, and the patterns of their interpretations are also in principle infinite. But comparing and ranking individuals has very little, if anything, to do with their uniqueness. One compares individuals by applying norms to particular actions and performances using the norm as a yardstick to grade and compare with. Of course, there is a difference between the uniqueness of individuals and the uniqueness of cultures because cultures are systems of norms and rules. But, if humankind were the overarching (essential) human cluster, there would be at least one norm shared by each and every culture. Cultures would not cease to be unique, but they could and should be compared by the standard of the very norm they share. Thus the recognition of the uniqueness of all culture, on the one hand, avid the possibility of comparing certain patterns of various cultures, on the other hand, do not exclude each other in principle. What is excluded, however, is comparing, ranking and grading cultures as wholes, because the uniqueness of each and every culture consists of the specificity of norms, rules and action patterns unshared with any other culture.

The sense of justice protests against extreme cultural relativism. If different and unique cultures cannot be compared, ranked or graded in any way at all, there is no justice, and accusing anyone of the use of ‘double standards’ is simply meaningless, null and void. But the same sense of justice does not protest against moderate cultural relativism. If I state that animism has its proper place in culture X but is an aberration in culture Y, no one’s sense of justice will protest. The objection could still be made that I am wrong, even that am a poor judge in such matters, but it would not be that I apply double standards and so am unjust. The sense of justice indicates, therefore, that ‘something’ should be compared, but not that everything should be compared. If the sense of justice expresses the will that humankind should become the ‘overarching (essential) cluster’, and if the same sense of justice demands that something, but not everything, should be compared, then the ‘good’ that this will aims at allows for a plurality of ‘goods’ – of normative systems and value systems – within the limits of the shared ‘good’. But what is this ‘shared good’ at which the sense of justice aims?

It is not necessary to take recourse to a hypothetical situation of choosing ‘under the veil of ignorance’ in order to discover which cases we must apply a common standard to if we intend to live up to the normative demands of the sense of justice. Nor is it necessary to make statements of history (of the past, present and future), nor even to identify the so-called ‘universals’ of the human species. it suffices to drop our anchor in the present age and to cast a glance at contemporary historical consciousness, to interpret the empirical evidence disclosed by this historical consciousness.

Let me ask the simplest of questions: when, and in which cases, do people accuse each other of applying ‘double standards'? When, and in which cases, does our sense of justice denounce the application of different standards as unjust? This happens exclusively when we judge or assess acts of domination, coercion, force and violence. If people are imprisoned, tortured, killed, humiliated, even discriminated against, within particular cultures, it is then, and only then, that the sense of justice enjoins us to apply exactly the same standards to all of them. Whether women wear veils or miniskirts is a matter of particular (unique) cultural taste. However, should they be forced to wear veils, it ceases to be a matter of cultural uniqueness’ and becomes a matter of coercion or force.

The claim for a common standard is a claim for grading and comparing. The sense of justice, while transforming different standards into double standards, not only demands the application of the same standards in the assessment of actions and behavioural patterns and systemic constraints which involve domination, coercion, force or violence, but calls for cultures to be compared and ranked from the viewpoint of these standards as well.

The postulate raised by the sense of justice is a decrease in domination, coercion, force and violence. All those protesting against the application of double standards must compare, rank and grade different cultures with the guidance of this postulate.

Obviously, all cultures contain domination or coercion or force or violence in one or another way and to a certain degree. One particular culture will score higher In a certain aspect and lower in another when it is compared to other cultures according to a common standard. And cultures with a lower overall level of domination, coercion, force and violence will score higher than se in which the level of these constituents is higher.

The procedure described above is just and can be accounted for the formal concept of justice. According to the latter, the norms rules which constitute a social cluster have to be applied consistently and continuously to each and every member of the cluster. Members of the cluster have to be ranked and compared by the common norms and rules as standards for ranking and comparing. But on what grounds do we compare various cultures, if humankind is not a social cluster? And indeed it is not.

Here we return to our starting-point: the sense of justice expresses will to constitute humankind as the overarching (essential) cluster. All those not applying double standards in their judgements act as if humankind were the overarching (essential) cluster. Although I have now returned to my point of departure, I have done at a different level. First I came to the conclusion that the sense of justice expresses the will that certain norms, or at least one norm, should be common for all humankind, for all cultures: the norm of constituting humankind as the essential cluster. Now we know above and beyond this: we know which kind of norms (or norm) should constitute humankind as the overarching (essential) cluster. Since our sense of justice warns against the application of double standards only if acts or patterns of domination, coercion, force and violence are to be judged or assessed, the common norms of humankind as an essential cluster should be such that their would minimize domination, coercion, force and violence in every unique culture. There can be different norms of this kind. Kant has already formulated the most sublime and the most abstract of them all: men should not be used as mere means by other men.

Humankind does not consist of single individuals who can enter into a social contract, real or virtual, under the veil of ignorance or otherwise. Humankind consists of various cultures and various histories. Individuals are socialized into different cultures and are part of different histories according to different traditions. There is something precious, as well as something abominable, in all of them. Only cultures, not individuals, can enter into a ‘contract’, symbolically speaking, by accepting a few common norms, in particular the one which enjoins us to respect the life and liberty of all ‘Its members and the members of all ‘outgroups’. If we speak of humankind as ‘the sum total of human beings who inhabit our globe’, we only obscure the fact that we are talking about the sum total of human cultures.

This circumstance does not draw its importance from the speculation about the possibility or desirability of a ‘social contract’ in the version hinted at above. Rather, its importance is that we cannot be just without facing it, here and now. The gist of the matter is that, even if we apply a few common standards to all cultures, then, if an action or systemic pattern includes domination, coercion, the use of force and violence, we cannot compare individual actors of different cultures by this common standard alone. Nor can we rank them according to their merits or shortcomings by applying only these common standards, even if they in fact apply – that is to say, in cases of force, violence and the like. It is a matter of the gravest concern how far the ban on the application of ‘double standards’ should regulate our comparison of individuals and in which cases. Without offering a foolproof recipe, I think that the Weberian ‘ethics of responsibility’ can serve as a guideline for such a decision. One could say that, in political actions proper, a common standard should apply to individual actors as well, and it stands to reason why. World history, once a construct of philosophies of history, has become a reality in our time. Whatever happens in one corner of the globe can have, and more often than not does have, fateful consequences for the lives of the denizens of every culture, every country. Since political action proper influences – in whatever culture it might take place – the lives of the denizens of each and every contemporary culture, the consequences of any political action proper transcend the limits of the culture in the framework of which it has taken place. Political actors cannot be assessed by the standards of their own respective cultures alone, but should also be assessed by the standards of those cultures which bear the consequences of their actions. However, if we assessed a political actor by applying a standard of any other culture but his own, the judgement would be unjust, since we should be applying norms and rules to a member of social cluster in which those norms were not valid. And should we assess a political actor by applying all standards of the empirically coexisting cultures, we could not reach agreement in our assessment. The only way to be just in our judgement is to apply to the actions of the political actors proper the norms which do not yet constitute humankind as the overarching essential) cluster. Respect for life and liberty qualifies for universal normative use. Political actors can and should be compared on the grounds of their respect or disrespect for the lives and liberties of all people, their own Included. In saying this, I only reaffirm the claim to a moderate cultural relativism. Only those actions including domination, coercion, force and violence should be ranked or compared at all. Political actions proper include at least one of these alternatives to a greater or lesser extent; and it is precisely this ‘greater’ or ‘lesser’ extent that provides a firm yardstick to compare and rank individual political actors.

To sum up my argument: extreme cultural relativism contradicts our sense of justice. If we accepted extreme cultural relativism, we could not protest against the use of ‘double standards’. In this case, we either could not raise the claim to just judgement and action at all, or we should have to accept ethnocentric assessments as just. And the sense of justice does exist, for we denounce the application ‘double standards’ as unjust. An obvious objection to this would be that, although we reject the use of ‘double standards’, we certainly apply them all the time. Political adversaries accuse each other of using double standards, but they are not apprehensive about their own use of them. This sad fact is, however, does not refute the assertion, ‘The sense of justice exists.’ The same thing or something similar happens whatever norms ought to be applied to each and every member of a social cluster. It is always easier in every respect to judge others justly than to judge ourselves. The second possible objection is more serious. One can argue that there is a certain concept of humankind that defines the latter through some substantive ends, and that this concept itself excludes justice and encourages the application of different standards from the viewpoint of the substantive end. This is undeniably true. However, it only proves that the sense of justice is not general, not that it does not exist. The will that humankind should be the overarching (essential) cluster to which certain norms should apply, and that these norms should enjoin us to respect the lives and liberties of all human beings, no matter what their culture, does exist, but not every human being shares this will. If all did, humankind would already, in fact, be the overarching (essential) cluster. For humankind to become this, humans of all political affiliations must share this will, and this is clearly a very difficult goal to achieve. But, even so, we can and should act as if they did share this will. Regardless of other people act, we ourselves should not apply double standards.