Professor Theodore Scaltsas holds the Chair of Ancient Philosophy at the University of Edinburgh. He started his professional career at Oxford University, after receiving his degrees at Oxford, Brandeis and Duke universities. He specialises in Ancient Greek philosophy, and in innovative applications of Information Technology to abstract thought. He has held Research Fellowships at Harvard, Princeton, and Sydney universities. His publications focus on Aristotle, and extend to Plato, Socrates, Zeno and the Stoics.
In his Ethics Aristotle says:
“Of particular justice, one kind is that which is manifested in distributions of honour or money or the other things that fall to be divided among those who have a share in the constitution, and one is that which plays a rectifying part in transactions between man and man”.
Aristotle describes the range of application of the second kind — rectificatory justice — as extending to transactions of sale, purchase, loan for consumption, pledging, loan for use, depositing, and letting, as well as to such actions as theft, adultery, poisoning, procuring, enticement of slaves, assassination, false witness, assault, imprisonment, murder, robbery with violence, mutilation, insult. The list is substantial, and between them distributive and rectificatory justice seem to provide a full spectrum for covering all cases of distribution of goods, exchange of goods and rectification for harm done. My concern in this paper is to provide new arguments in support of the position that there is a domain for the application of justice which is not covered either by distributive or by rectificatory justice. This is the domain of reciprocal justice, which is introduced by Aristotle to complement the other two. One sector of this domain consists of acts of injustice, and the other, of exchanges of goods in society. I shall first be concerned with the meaning of ‘reciprocity’ for Aristotle. Then I shall argue that, according to Aristotle, reciprocal justice fills a gap left by distributive and rectificatory justice. I shall finally offer a utilitarian argument, not offered by Aristotle but needed for his position, to distinguish between reciprocal and rectificatory justice.
It will be useful, for the purposes of characterising reciprocal justice, to describe very briefly the main features of the other two kinds of justice. Distributive justice is practiced before any injustice is done, and the principle regulating the distribution of the goods is geometrical proportion. The coefficient determining the proportion is the merit of the parties involved, so that in distributing the goods, the ratio of the merits of the persons will determine the ratio of the value of the goods they receive: as person A is to person B, so is A’s portion to B’s portion of the goods. On the other hand, rectificatory justice is practiced after injustice is done, and the principle regulating the return to the injured party is arithmetical equality: the injured party will receive a portion from the injuring party that will equalise the difference between them caused by the unjust act. What is significant in rectificatory justice is that all parties are seen as having equal merit. Differences in merit do not enter into the calculations of retribution.
To determine what Aristotle understands by reciprocity we have to look, first, at the example of reciprocity with which he introduces this notion, and then, at the mathematical definition the principle involves.
2. Pythagorean Reciprocity.
Aristotle opens chapter 5 with an ambiguous statement which has been translated variously: T. Irwin: ‘It seems to some people, however, that reciprocity is also unconditionally just’. D. Ross: ‘Some think that reciprocity is without qualification just, as the Pythagoreans said.’ R. Browne: ‘Some people think that retaliation is absolutely just, as the Pythagoreans said’. J. Thomson: ‘There are some who even think that what is just is simple reciprocity, as the Pythagoreans maintained’. H. Rackham: ‘The view is also held by some that simple Reciprocity is Justice’.
The possible readings of this passage are the following. (1) If reciprocal, then just. (2) If just, then reciprocal. (3) Reciprocal if and only if just. For each of these readings we can distinguish between: (a) reciprocity understood in the simple, lex talionis (or, an eye for an eye), sense; and (b) reciprocity understood any way it may be defined. I will argue that the point Aristotle is making in this sentence, to which he will object, is given by (3b) (which could also be made by (2b)).
That the intended meaning is not (1) (whether (a) or (b)) is seen from the fact that, in what follows the passage above, Aristotle says that the Pythagoareans ‘defined justice unconditionally as reciprocity’. This shows that it is (3) which was held by the Pythagoreans: i.e. just if and only if reciprocal. It could be thought that what Aristotle is trying to stress here is that some people, like the Pythagoreans, believe in lex talionis. But I think that the reason why Aristotle introduced the Pythagorean definition is to show that there are some people who believe that justice is exhausted by reciprocity. What Aristotle wants to emphasise is that for the Pythagoreans, justice just is reciprocity. The sentence that follows in the text shows clearly that this is Aristotle’s concern. He says that reciprocity does not fit either distributive or rectificatory justice, which he then proceeds to demonstrate. It follows that, given his concern to show that reciprocity is only a species of justice, he will disagree with the Pythagoreans insofar as they claim that reciprocity covers all cases of justice, no matter how the Pythagoreans define reciprocity. Therefore, the message that I take Aristotle to be conveying in the initial sentence of this chapter is that some people think that justice is reciprocity however reciprocity may be defined.
The translation that would give this meaning is: ‘Some even think that reciprocity is simply the just, as the Pythagoreans said’. This, taking into account the definition that immediately follows, is the reading in (3b) — just if and only if reciprocal, however reciprocity is defined – although it is also compatible with (2b).
What follows from the above reading of the passage is that there is no compelling reason to take the term ‘reciprocity’ in its third occurrence in the chapter to have the same meaning as in its previous two occurrences (reciprocity in the Pythagorean sense). Whatever the meaning of the term in its first two occurrences, the point Aristotle is making in its third occurrence still remains, namely that justice is not coextensive with reciprocity. Thus, we can take Aristotle in the third sentence (where he rejects the coextensiveness of justice and reciprocity) to be introducing reciprocity in the sense that he is concerned to define it, rather than assuming that he is still talking of reciprocity in the Pythagorean sense of lex talionis.
3. Rhadamanthean Justice.
Before showing why reciprocity differs from distributive and rectificatory justice, Aristotle anticipates an objection to this claim. He says:
yet people want this to mean even the justice of Rhadamanthus:
Should a man suffer what he did, right justice would be done.
This is a crucial passage, as Aristotle gives us the first indication of what he understands by reciprocity, by telling us what he does not mean by it. Taking the Rhadamanthean example to be an example of reciprocity (in the common, but still not the Aristotelian sense of the term) is crucial in giving us a first indication of Aristotle’s understanding of reciprocity: he does not mean talionis retribution, since that would collapse reciprocity into rectificatory justice.
4. The Official’s Example as an Instance of Reciprocal Justice.
Aristotle shows that reciprocity does not fit either distribution or rectification by the example of the official, which is the first positive indication he offers of what he understands by reciprocity:
If an official has inflicted a wound, he should not be wounded in return, and if some one has wounded an official, he ought not to be wounded only but punished in addition.
Aristotle’s concern in the example of the official is to stress that the just return for the injury inflicted will not be on the principle of lex talionis. (Yet, such a return would fit rectificatory justice). What the layman who injured an official will get in return is not only what he has inflicted on the official, but an additional punishment as well. Moreover, the official who injures a layman should not suffer the same injury in return. This does not preclude a lesser punishment of the official, such as a reprimand. What is clearly excluded is the Rhadamanthean-Pythagorean conception of reciprocity: if reciprocity were talionis return of the harm inflicted, then the official should be wounded in return and the layman not punished in addition.
Aristotle offers the example of the official in order to show (albeit by exclusion) what he means by reciprocity and thereby justify the claim he has just made that reciprocity is different from rectification and distribution. Otherwise the reader would remain in the dark with respect to what it is that Aristotle is distinguishing from rectification and distribution, while being offered an example of the already familiar rectificatory justice.
A second reason why the official’s example cannot be seen as an example of rectification has to do with the rationale offered for the punishments of the official and the layman. What has to be explained is why Aristotle says that the official should not be struck back while the layman should be struck back and punished in addition, when the principles of rectification require equal return. The usual line of defense for the compatibility between differential treatment and the principles of equal return is that the layman does not only harm the person of the official but also the state by offending a public officer. In other words he is committing a double offence, one against the person of the official and one against the status of the official. So his punishment should be severe. To see the shortcomings of this explanation, let us consider a familiar example, the notorious incident of General Patton striking down a soldier because he thought it was cowardly of the soldier to be in the hospital recuperating from shell shock. Patton acted unjustly in striking the soldier. But his punishment, a public apology, was very different from the punishment that should have been inflicted on a soldier if the soldier had struck a general because he did not think the general’s plan courageous enough. In such a case the soldier would be severely punished. Following the rationale mentioned above, one would justify the differential treatment of Patton and the soldier by saying that the soldier committed an offence both against the person of Patton and against the state, given the status of Patton as a general. But let us follow up this kind of rationale. When Patton unjustly strikes the soldier, it could equally be said, he is not only committing an offence against the person of the soldier, but also against the state, given the status of the individual as a soldier. Patton has the obligation to respect the rank of the soldier even when he has a legitimate complaint against the soldier (which Patton would have to process through the judicial procedure of the army), let alone when he is acting unjustly towards the soldier. Therefore, the reasons for which it was claimed above that the soldier should be punished in addition to being struck back are also applicable to Patton: he is committing an offence against the person of the soldier and one against the public status of the soldier. Hence, on this reasoning (according to rectificatory justice) the official should not only be struck back, but punished in addition. Yet Aristotle says that the official should not be struck back. It follows that on the proposed rationale, the principles of rectification could not justify Aristotle’s example of the official (or, for that matter, the way the army treated Patton — merely a reprimand).
If it were claimed that the layman’s striking the official is more harmful to society than the official’s striking the layman, this would explain why the layman should be struck back and punished in addition; but there would be no explanation, on the principles of rectificatory justice, of why the official should not even be struck back.
A different rationale for the justification of the differential treatment of the official and the layman on the principles or rectificatory justice is to assume an asymmetry of justification between the official’s case and the layman’s case. Thus, Burnet says:
if an officer strikes a private, that is not merely a blow, but may also be an act of discipline; if a private strikes an officer, that is not merely a blow, but also an act of mutiny.
But why does Burnet assume that, whereas the officer had some just reason for striking the layman, the layman had no just reason for striking the officer? If we assume partial justification of the action in the one case, we should do so in the other, too. The layman could have been trying to prevent the officer from inflicting a great, possibly fatal, injury on someone else. Mitigating factors can be taken into consideration on both occasions, in which case it is far from true that the layman who strikes an officer should be struck back and punished in addition. What Burnet has done (which is the usual line of defence) is, in order to show that in both cases (when the official is the offender, when the layman is the offender) the punishment is in accord with the principle of equal return (i.e. rectification), to assume an asymmetry between the two cases: the official is partially justified in striking the layman (hence, discipline), while the layman is not justified in striking the official (hence, mutiny). Unless this asymmetry of justification is assumed, both the layman and the official are guilty of unjustly striking another person. But we have no indication in the text that the mitigating circumstances are different in the two cases. Hence, things being equal, rectificatory justice cannot explain why the two should be treated differently.
One further line of thinking purporting to show that inflicting different punishment to the official and the layman is compatible with the principle of equal return–rectification — could be the following: because of their difference in merit, a mild retributive act against the official will cause as much harm as a severe one against the layman. Thus, a reprimand of a high official may be as harmful to the official as flogging would be to the layman. (It is compatible with the text to assume that the official might be punished, only not as severely as equal return would require.) But calculating punishment (in the cases of injustice done) on the basis of the difference in merit leads to unreasonable consequences: if a rich person steals from the poor, should the rich person pay back the stolen amount or an amount that represents for the rich person what the amount stolen represents for the poor person? (I owe this point to Vinit Haksar.) Conversely, if the poor steal from the rich, should the offence be treated as negligible if the loss the rich suffered is negligible to them? There is no indication that Aristotle is proposing anything as radical as that.
The explanation that I shall supply in section (7) of this paper for the justification of the differential treatment of the official and the layman introduces a further factor for consideration in determining their punishment: the utility and disutility of punishing the official and the layman for society. It is true that both the official and the layman commit offences both against a person and against the state in their respective acts; but it is not true that punishing them equally will be equally beneficial for the state. Hence, for utilitarian reasons relating to their punishment, given the position (merit) of the two parties, rectificatory justice should not be applied in their case; rather, the principles of proportionate return of reciprocity should decide their respective punishment. This utilitarian consideration can offer us the grounds for the asymmetry between the two cases. There is no scope for the introduction of such utilitarian factors in rectificatory justice, which is why reciprocal justice is marked off as a different type of justice from rectification.
Apart from showing that reciprocity is not talionis return, the example of the official shows at the same time that reciprocity is not distributive or rectificatory justice. It is not distributive justice since such cases as that of the official concern retribution, whereas, as we saw, distributive justice is practiced in cases where no injustice has occurred. It is not rectificatory justice since reciprocity takes the differences of merit between the two parties into consideration, while rectificatory justice treats everyone as having equal merit.
5. Reciprocal Proportion in Exchanges of Goods.
From the example of the official, which is an application of reciprocity after injustice has been done, Aristotle moves to the application of reciprocity in exchanges in society, giving reciprocity a fundamental role in the very constitution of the fabric that holds society together: ‘in associations for exchange this sort of justice does hold men together — reciprocity in accordance with a proportion and not on the basis of precisely equal return’. Here we have a clear statement of the connection between reciprocity on the one hand, and proportional, as opposed to equal return, on the other. This comes both as an explanation of Aristotle’s rejection of the Pythagorean-Rahdamanthean definition of reciprocity as lex talionis, and as an explicit formulation of the principle that he already introduced through the example of the official.
Associating reciprocity with exchanges of goods brings Aristotle to a detailed description of the nature of the proportion that is employed in reciprocal exchanges. He states that: “as farmer is to shoemaker, the shoemaker’s work is to the farmer’s work”. One would have expected Aristotle to have said: as farmer is to shoemaker, the farmer’s work is to the shoemaker’s work. I shall first show that the passage is, in fact, consistent with the prior specifications of geometrical proportion, and then turn to some problems that will be raised concerning the relation between reciprocity and the other two kinds of justice — distributive and rectificatory.
First, we must establish that Aristotle is not introducing here a new kind of proportion. We saw above that he has distinguished between geometrical and arithmetical proportions. The geometrical one (which is required of reciprocity) was determined in the cases of distribution of goods as follows: as the merit of A is to the merit of B, so the portion of A will be to the portion of B. It is the attempt to follow along the principles of this proportion that leads the reader to expect in the case of reciprocity, too, that as farmer is to shoemaker, so will food be to shoes. Why then does Aristotle say that “as farmer is to shoemaker, the shoemaker’s work is to the farmer’s work”? The key towards understanding Aristotle’s statement is given in another example of reciprocity: “the number of shoes exchanged for a house must correspond to the ratio of builder and shoemaker”. What we see here is that Aristotle is comparing the value of the work that is exchanged, not the value of work per unit of production. Thus, because the value of the farmer’s work is greater than the value of the shoemaker’s work, the amounts of goods exchanged should be inversely analogous to the value of their producer’s work. Hence, as farmer is to shoemaker, the number of shoes (the farmer will receive) are to the amount of food (the shoemaker will receive).
This way of reading the rationale behind the proportion in the case of reciprocity shows that Aristotle is retaining one and the same kind of proportion in the cases of distributive justice and of reciprocal justice. But whereas in the case of distribution the operative value is the merit of each party (which will determine the value of goods they each deserve), in the case of reciprocal exchange the operative value is the value of the products (which will determine the value of the craftsmen). In the case of distribution the merit of each party is determined by such values as the status of free person, the degree of wealth of the individual, the nobility of the individual, the excellence of the individual, etc. Whereas in the case of reciprocal exchange the value of the product is determined by the degree of need for the individual’s products: “All goods must be measured by some one thing this unit is, in truth, need”. The difference, therefore, between distribution and reciprocal exchange is that in the first, the regulative value is the merit of the person, while in the second, the regulative value is the value of the product exchanged.
Joachim complains that Aristotle is being mysterious in his comparison of the producer’s work: ‘How exactly the values of the producers are to be determined, and what the ratio between them can mean, is, I must confess, in the end unintelligible to me. Apparently Aristotle is thinking of their values to one another in the particular exchange in question: i.e. their demand respectively for one another’s services. But this does not take us much further’. It is not clear to me what sort of algorithm Joachim is requiring, over and above the principle that Aristotle did give us in the passage quoted above, namely, that the value of the goods will be determined by the respective need for them in society. One might disagree with Aristotle as to whether it is need, or demand, or some other principle that determines the value of the goods. But having such a principle goes most of the way towards explaining the determination of the value of the goods.
6. Reciprocal Justice Differs from Distributive and Rectificatory Justice.
The position we have reached so far is the following. Distributive justice is practiced before injustice is done, on the basis of a principle of geometrical proportion, regulated by the relative merit of the parties involved. Rectificatory justice is practiced after injustice is done, on the basis of a principle of equal return, all parties being credited with equal merit. Reciprocal justice branches out into two sections: the first kind is concerned with exchanges of goods and is practiced before injustice has been done, on the basis of geometrical proportion, regulated by the value of the goods exchanged; the second kind is concerned with retribution for injustice done and is practiced on the basis of geometrical proportion, regulated by the merit of the parties involved.
Both kinds of reciprocal justice employ the principle of geometrical proportion according to value: the value of the products for the one, and the merit of the parties for the other. This distinguishes reciprocal from rectificatory justice, which operates on the principle of equality, disregarding differences of merit.
Reciprocal justice is also different from distributive justice. In the case of exchange of goods, reciprocity (in reciprocal justice) is determined on the basis of the value of the goods, while distribution (in distributive justice) is determined on the basis of the merit of the person that receives them. In the case of injustice done, reciprocity is different from distribution since distribution does not apply to it; distribution applies before injustice is done, not after. Thus the interpretation of reciprocity proposed in this paper pays justice to Aristotle’s claim that reciprocal justice is different from both, distributive and rectificatory justice.
Yet, a problem has now emerged; namely, it appears that the principles of rectificatory justice contradict the principles of reciprocal justice. We saw above that Aristotle said rectificatory justice applies to such cases as abuse, assault, mutilation, murder, false witness, imprisonment, etc. It would follow that in the case of the injury inflicted by the official, the two parties should be treated as equal under the law. But on the present interpretation (which treats the example of the official as an example of Aristotle’s conception of reciprocal justice) the official and the layman will be judged on the basis of their difference in merit, as reciprocity requires. This seems to create a direct conflict in the treatment of one and the same case within Aristotle’s theory of justice. If the claim of the present interpretation about the role of the example of the official in Aristotle’s theory of justice is correct, this apparent conflict must be explained away. This can be done if we show that there is reason to treat the case of the official, and other such cases, as falling only under the domain of reciprocal justice, and not under the domain of rectificatory justice.
7. A Utilitarian Justification of Reciprocity.
Aristotle does not provide criteria for distinguishing between the cases that fall under rectification and the ones that fall under reciprocity, nor do the commentators offer a justification on his behalf. Only Michael Ephesius provides a brief statement which establishes their difference in merit as the reason for the asymmetrical return, but does not explain why this should be so. Furthermore, Michael Ephesius interprets the Aristotelian notion of reciprocity so broadly, it is not clear at all how he could draw, and defend, Aristotle’s distinction between rectification and reciprocity. He spreads a reciprocity umbrella that is broad enough to cover all moral acts in the social context, leaving no distinct domain for rectification.
In order to provide a set of criteria in support of Aristotle’s distinction between rectification and reciprocity we shall turn to the example of the official. We are looking for a reason why the difference in merit between the layman and the official should be taken into account in restoring the injustice done. Only then could the case be treated as that of reciprocity rather than rectification. That the parties have different merit is not itself sufficient to require different punishment, since their difference in merit does not result in different punishment in the cases of rectificatory justice.
The argument I will offer in support of introducing the difference in merit in the case of the example of the official (and other cases like it) is utilitarian in nature. The question we need to answer is: what is special about the layman striking the official that requires merit to play a role, as opposed to e.g. the case of a layman committing adultery with the official’s wife, where merit makes no difference? Why should the one follow the principles of reciprocal justice, while the second, of rectificatory justice? What is special about the first case is that the layman’s action goes against the very essence of the difference in rank between him and the official. Their difference in rank consists in the layman obeying the official. Being so obeyed is constitutive of the official’s role as an official. But treating the two offences as being of the same kind would undermine the official’s authority in issuing commands. The preservation of the authority of the official is of such value to the state that it justifies an additional punishment in the case of the layman, while it counterbalances the requirement for the punishment of the official (which rectification would have demanded). The introduction of such a utilitarian principle is over and above the considerations that are operative in rectificatory justice. This extra coefficient distinguishes reciprocity from rectification.
This utilitarian principle is different from considerations of the utility and disutility of the offending acts of the layman or the official. What is introduced here is a utilitarian evaluation, not of their acts, but of the punishments of the layman and the official. As we saw in section (4) above, the utilitarian evaluation of the offending acts can explain why the layman should be punished in addition to being struck back, but it will not explain why the official should not be struck back. An objection to the suggested justification for the differential treatment of the official and the layman might be thought to be the counter-suggestion that there may be utilitarian reasons for punishing the official more severely than the layman. Namely, unjust behaviour on the part of the official towards the layman undermines the trust in the judgement of the official, which is requisite for bestowing power and authority to the official. As such, the unjust act of the official is more harmful than that of the layman. I do not think this line of reasoning can be dismissed right off hand. One can very well imagine societies in which unjust conduct of officials is punished more severely than unjust conduct of laymen, the reason being the utility of the preservation of trust in the official. But this would not be an objection to the interpretation offered here for the justification of reciprocal justice. In fact, it would be a further instance of exactly the same utilitarian principle being applied, only with a different utility valuation in play. Whether the official is punished less severely than the layman or more severely, we are deviating from rectificatory justice, applying the principles of reciprocal justice. In both cases, a utilitarian argument is being offered against the principle of equal return. Thus all such cases would be treated on the basis of proportional return, dictated by reciprocal justice, rather than equal return, dictated by rectificatory justice.
It is important to realise that the utilitarian principle is not being introduced to settle a question of rights, but a question of punishment. Whether it is the official striking the layman or vice versa, the acts are assumed from the start to be unjust acts. Thereby, their rights are forfeited. The utilitarian principle comes into the picture only with regard to the severity of punishment that should be inflicted in each case, by considering (in the case of the official’s example) the utility or disutility of such punishment for the preservation of the structure of authority in society.
In the Politics Aristotle says that,
the principle of reciprocity, as I have already remarked in the Ethics, is the salvation of the states. Even among freemen and equals this is a principle which must be maintained.
Aristotle is here treating reciprocity, as compatible with equality between people. Professor Dragona Monachou, who pointed out this passage to me, also referred me to Gregory Vlastos’ classic account of the compatibility of the geometrical proportion according to merit (used by Aristotle in distributive justice) with egalitarianism; ‘Justice and Equality’, in J. Waldron (ed.) Theories of Rights. As Vlastos argues in his paper, the most perfect form of equal distribution is “To each according to his need”. Since, according to my interpretation, both kinds of reciprocity involve geometrical proportion according to merit, I need to show that reciprocity is compatible with egalitarianism. But there is nothing non-egalitarian in determining the value of the farmer (qua farmer) on the basis of the need for his/her products in society. This would be independent of what the farmer would deserve qua member of society, e.g. health-care benefits, social security benefits, etc. Similarly, there is nothing non-egalitarian in determining the punishment of the offending parties on the basis of the utility and the disutility of this punishment for society. If the principle of rectification was applied (= equal return for the injury inflicted to each party) the state would lose out with regard to the preservation of the authority structure within it (in the official’s example). To avoid this, their merit (in this case derived from the authority status of the parties in society) is taken into consideration. The parties are punished unequally because of the harm that their equal punishment would bring to society.
About the Author:
Professor Theodore Scaltsas holds the Chair of Ancient Philosophy at the University of Edinburgh. He started his professional career at Oxford University, after receiving his degrees at Oxford, Brandeis and Duke universities. He specialises in Ancient Greek philosophy, and in innovative applications of Information Technology to abstract thought. He has held Research Fellowships at Harvard, Princeton, and Sydney universities. His publications focus on Aristotle, and extend to Plato, Socrates, Zeno and the Stoics. [http://www.ppls.ed.ac.uk/philosophy/people/theodore-scaltsas]