I. Introduction
Law brings people together. It protects, and it enforces relations between individuals. At times, it imposes on individuals requirements that are unwelcome. The precise detail of the impositions conveyed within legal relations lies at the heart of a technical mastery of the law. Legal disputes are fought and resolved over what one party can legally require of another. The exact basis for the relationship the law upholds between the parties goes to the justification of the requirement that is made on one party for the benefit of the other. When that requirement is regarded as an unwelcome imposition, the call for a justificatory basis becomes more acute.
Legal relations accordingly attract both analytical-technical and normative-justificatory attention, and in both cases correlativity features as an explanatory term. An analytical-technical role for correlativity was famously employed by Hohfeld, pairing off his fundamental legal conceptions in correlative relations.Footnote 1 Contemporary scholars such as Ernest Weinrib, working within a Kantian tradition, have invoked a normative aspect of correlativity with its connotation of mutuality between the parties in a legal relation: “Correlativity locks the plaintiff and defendant into a reciprocal normative embrace … The only pertinent justificatory considerations are those that articulate the correlational nature of right and duty.”Footnote 2 Here, the justificatory impetus adopted for correlativity is aimed towards a defence of social or legal relations between those enjoying an equal status as autonomous agents. This reflects Kant’s principle of right with its grand precondition for the use of force in civil society established in respecting the equality of the wills of its members, in a “reciprocal relation of choice”.Footnote 3
Weinrib’s influential theory of private law, narrowly conceived as corrective justice, promotes Kant’s principle of right through an appeal to correlativity and (as the above excerpts reveal) a close association with reciprocity. There is a tendency for reciprocity to tag along with correlativity, being suggestive of the beneficial aspect of the mutual relationship. Indeed, reciprocity is sometimes treated as little more than a synonym for correlativity. Even Peter Cane and Hillel Steiner, two of Weinrib’s critics who find fault with his neglect of distributive justice in private law, seem content to follow the use of correlativity and reciprocity as synonyms.Footnote 4 In Ariel Zylberman’s recent work on a Kantian, non-instrumentalist approach to human rights the two become fused in a notion of reciprocal correlativity.Footnote 5
As well as being employed to represent analytical or normative aspects of legal relations, correlativity is also used within argument to support a particular understanding of legal relations, and through that an understanding of private law, human rights, or the law more broadly. For Hohfeld, the argumentative support from correlativity for his analytical framework is basically intuitive: correlativity simply lends an air of mutual entailment to the fundamental legal conceptions. They can be presented as requiring no further explanation beyond their inter-relationship within “a scheme of ‘opposites’ and ‘correlatives’”, where the distinguishing characteristics of one conception are to be drawn from its “correlative (and equivalent)”.Footnote 6 Zylberman too seems content to rely on an intuitive grasp of the significance of correlativity, making no effort to provide a precise understanding of it. Weinrib, however, as a key part of his argument refers to a logic of correlativity, which he takes from Aristotle, and then depends on in reaching the Kantian ideal of an equality of autonomous wills.Footnote 7
Despite a widespread interest in using correlativity to represent and argue over legal relations, there has been little progress in delivering a general understanding of correlativity, together with a rigorous assessment of its implications. This situation prevails even though there is some consensus over its use. The paradigm correlation between a claim-right and a directed duty occurs in both Hohfeldian and Kantian frameworks. However, efforts that have been made to provide greater understanding of the idea are suggestive of contestability, with different concepts being advanced to promote different normative agendas and competing analytical perspectives. In 1998, Matthew Kramer attempted an extensive and definitive account of Hohfeldian correlativity only for it to be immediately rejected by Nigel Simmonds.Footnote 8 David Frydrych’s contribution to a celebration of Hohfeld’s centennial has provided a critical and informative survey of the different perspectives on normative correlativity. Frydrych concludes his survey by describing it as a “starting point” for exploring “the correlativity of normative positions”, and concludes that this is “an area where more research is required.”Footnote 9
The danger is that further research will simply expand the opposing perspectives. It is not my aim to dispel this contestability by proposing an authoritative understanding of correlativity and demonstrating its superiority over all rivals. The more modest objective of this article is to provide a scheme of intelligibility for correlativity so as to clarify different uses of correlativity—different understandings, perhaps, though part of the burden of this project is to show that correlativity has frequently been invoked without any clear understanding. In distinguishing these different uses, I argue it is possible to see precisely what assumptions accompany and what implications follow from each use. In particular, the precise scope of a “logic of correlativity” can be revealed: its attachment to a particular use of correlativity and the wholly unwarranted inferences drawn from linking it to an alternative use. An additional advantage I claim for clearing up the uses of correlativity, is that it sheds light on the altogether more complicated relationship between correlativity and reciprocity.
Work on the scheme of intelligibility is undertaken in the following extensive section, comprising a sequence of observations on different uses of correlativity. This encompasses a survey of factual correlativity and a variety of types of normative correlativity, an investigation of the logic of correlativity, and recognition of three distinct forms of rights in legal relations involving correlativity. Subsequent sections draw on these observations to provide critiques of Weinrib’s use of correlativity and Zylberman’s amalgam of reciprocal correlativity. For Weinrib, the critique takes in interrogation of his established outlook on corrective justice; for Zylberman, the critique questions his view of human rights as ensuring equal dignity founded on “the same basic right to independence and the same basic duty of respect.”Footnote 10
The discussion involves careful re-examination of Aristotelian texts to reveal insights that, I argue, have been neglected or even misrepresented. Certain Hohfeldian insights enter the discussion but ramifications for Hohfeldian scholarship are not explored here. Although the scheme of intelligibility does have implications for Hohfeld’s analytical scheme, that topic merits separate attention. The current article does not encourage a rigid division between analytical and normative aspects of correlativity, but its emphasis is firmly on the normative side. A brief concluding section accordingly draws together some general lessons emerging from the discussion of correlativity on the importance of the deep asymmetry of law; suggests an understanding of corrective justice based on asymmetry rather than equality; and, more speculatively, raises doubts about the core conviction of Kantian thinking on legal and social relationships.
II. Making Correlativity Intelligible
A. Factual Correlativity
Although this article is concerned with normative correlativity, it is worth stepping back to acknowledge correlativity in non-normative contexts, or, more specifically, situations where the correlative relationship is not between normative positions—even though normative evaluations may be regularly attached to the positions involved. This may help to provide a sense of an underlying structure of correlation before working through its implications for a normative setting. The paradigm of normative correlativity has already been mentioned, correlation between a claim-right and a directed duty, linking the two normative positions of holding a claim-right and holding a duty. The very relationship between these two positions has normative significance: it informs us what is required to be given by the one party and what is required to be received by the other party. It tells us what ought to be the case.
By contrast, we can recognize a non-normative, or simply factual, correlativity connecting two positions where the significance of the relationship is to inform us what is the case for the one party in its relation to the other party and what is the case for the other party in its relation to the first party. An obvious example is the correlation between a mother and a child.Footnote 11 Another example of factual correlativity is the correlation between an employer and employee. Although in both of these cases we could argue that there are normative connotations attached to being an employer (and employee) or mother (and child), the relationship itself informs us what exists as a matter of fact between the two parties. We may want to suggest that because you are an employer (mother), this is the way you should behave towards your employee (child), but the relationship itself rests on what has happened between the parties, not on what is required to happen between them.
Perhaps the simplest way of capturing factual correlativity is in the form of the active and passive aspects of a single occurrence of conduct. (1) There has been a punch. A punched B is the correlative of B was punched by A. (2) There has been a gift. A gave the gift to B is the correlative of B was given the gift by A. Note this active-passive structure holds even if we opt for a different term to describe the passive encounter of the conduct. (2A) A gave the gift to B is the correlative of B received the gift from A. (3) There has been a birth. A gave birth to B is the correlative of B was given birth to by A. Note we can still discern an active-passive structure even if we find it more fluent to refer to the correlative positions in terms of nouns rather than verbs. (3A) A is the mother of B is the correlative of B is the child of A.Footnote 12 Similarly, (4) There has been an employment. A employed B is the correlative of B was employed by A; (4A) A is the employer of B is the correlative of B is the employee of A.
What is common to all of these instances of factual correlativity is that the correlative positions are linked within a single occurrence of conduct affecting the two parties, while affecting them differently in that one experiences the conduct actively while the other experiences it passively. This gives rise to distinct experiences not just different perceptions of the same experience. The experience of sore knuckles is quite distinct from the experience of a bloody nose at the active and passive ends of a punch.Footnote 13 Nevertheless, it is impossible for one party to experience the conduct alone without the experience of the other party. One cannot punch a person, give a gift to someone, give birth, or employ someone, without the other party being involved; and equally for the person on the receiving end. In sum, if the conduct happens at all it happens for both correlative positions; and, each correlative position experiences the same conduct, albeit in different ways (the active-passive distinction). We have here the stirrings of a “logic” of correlativity. It follows from what has just been said that it is possible to infer one correlative position from the other. Quite how this logic operates, on both a factual and normative plane, will be examined in more detail shortly.
B. Normative Correlativity
We can now revisit our paradigm case of normative correlativity, the correlation between claim-right and directed duty, and trace a similar active-passive structure to the correlativity here. Take as a standard example, a contractual arrangement whereby A is to pay $200 to B.Footnote 14 Once we place the subject matter of this arrangement within a normative requirement, which we can indicate by R, then that content follows the structure of factual correlativity, as follows. (5) R (there is a payment of $200). R (A pays B $200) is the correlative of R (B is paid $200 by A). And in a similar move to that found in (3A) and (4A), we can restate this as (5A) A’s duty to pay $200 to B is correlative to B’s claim-right to be paid $200 by A.
The move is similar but not precisely the same in that this time we do not merely have a switch from verb to noun to convey what is the case for a correlative position. Now the noun governs the factual content of what is required and conveys the nature of the normative requirement upon the party. Nevertheless, if we accept that it is accurate to portray {A’s duty to pay $200 to B} as {R (A pays B $200)}, and {B’s claim-right to be paid $200 by A} as {R (B is paid $200 by A)}, then this exercise is capable of demonstrating a parallel between factual and normative correlativity over the presence of the active-passive structure.
Two amplifications need to be made at this stage, regarding the active-passive structure of normative correlativity. The illustration used above is not only a standard illustration of a claim-right and directed duty. It is also a straightforward and convenient illustration, in that the content of the duty and right amounts to conduct that is (actively and passively) directly experienced by the two parties. Suppose the contract between A and B specifies that A is to pay $200 to C. Cases such as this can still be understood as fitting the active-passive structure if we take the contract at a more abstract level to be concerned with the giving and receiving of a benefit. A by paying C is giving a benefit to B Footnote 15 and on payment being made B receives that benefit. Similarly, the more abstract giving and receiving of benefit can be applied to undertakings that do not involve third party beneficiaries but where one party does not personally experience the conduct performed. A is a gardener employed by B to work in his garden. B does not himself directly experience the work done by A in the garden but does thereby receive the benefit given by A.
The second amplification relates to those claim-rights and directed duties governing omissions. A particularly important group, including your duty not to punch me on the nose.Footnote 16 At first sight, it appears contrived to suggest that the duty not to do something can be located at the active end of an active-passive structure, and equally that the claim-right not to have something done to one can be located at the passive end of the same structure. Yet, once we examine what it is that is required not to happen the familiar active-passive structure is evident. (6) R ¬ (there is a punch on another person). R ¬ (A punches B) is the correlative of R ¬ (B is punched by A). And in a similar move to that found in (5A), we can restate this as (6A) A’s duty not to punch B is correlative to B’s claim-right not to be punched by A. This case of conduct by omission shares all the characteristics of factual correlativity noted above. It is impossible for A not to punch B to occur unless it is also the case that B is not punched by A; and the presence of the one can be inferred from the other.
It is obviously possible to say more about the normative positions of right and duty. We shall consider below why some may be motivated to say more about the normative position of a right than that it expresses the passive reception of conduct from a duty holder; or, correspondingly, to say more about the normative position of a duty than that it expresses the active undertaking of conduct given to a right holder. The point for the moment is to stress that whatever else a claim-right may be taken to involve normatively, the basic correlativity between claim-right and directed duty can be expressed at a very elementary level in terms of the correlation found in an active-passive structure attached to the conduct that is required.
We shall also consider below why some are motivated to insist on acknowledging correlation with rights other than claim-rights (in a narrow sense), and it is appropriate to concede now that the basic notion (active-passive structure) of normative correlativity applied here to the paradigm of claim-right and directed duty depends entirely on there being a single occurrence of normatively required conduct linking the two parties in this way. And with that, to acknowledge other occurrences of normative requirements affecting two parties in different ways may not satisfy these limitations.
Two potential advantages follow from starting with this limited basic case of normative correlativity. For one thing, it enables us to test a simple case of normative correlativity in order to explore how correlativity functions without being distracted by complications that may beset other types of normative correlativity. For another thing, if we obtain some valuable results in this simple case, it may assist us in determining whether the same results apply to other more advanced or more complex cases.
One distraction thus avoided by considering this simple case is the burden to show that all instances of rights are correlated with all instances of duties, in order to reach an understanding of normative correlativity. We are only examining normative correlativity for those claim-rights and directed duties whose subject matter falls under the active-passive structure. For this subset the correlation is based not on a general understanding of right and duty but emerges from the structure of the single “factual conduct” that links the claim-right and duty, even though the factual conduct here is subject to a normative requirement rather than simply being the case.
As for seeking a valuable result for this simple case of normative correlativity, one obvious issue to examine is whether a logic of correlativity can be explained here along the lines of the inference we noted could be drawn in a case of factual correlativity, given the common active-passive structure we have identified in these two cases. We look into this issue next, before widening our observations to take in more advanced and more complex cases of normative correlativity.
C. The Logic of Correlativity
When Weinrib introduces the logic of correlativity,Footnote 17 he draws on Aristotle’s discussion of correlativity in the Rhetoric. We shall consider the exact way Weinrib makes use of this source in section III, but one point needs to be addressed here. This is the suggestion Weinrib makes that “Aristotle provides no answer” to the question of how normative implications arise out of the recognition of correlativity.Footnote 18 In a work that is supposed to offer sound types of proof to win over an audience, it would be strange for Aristotle to assert that certain normative implications are borne by correlativity and yet to offer no explanation of how this occurs.Footnote 19 I shall argue here that on this point, Weinrib is clearly wrong,Footnote 20 and that the explanation Aristotle does provide for the normative implications of correlativity helps us to make further progress in linking together factual and normative correlativity, as well as providing a platform for a more refined analysis of the normative implications of different uses of correlativity.
To see how Aristotle does explain how normative implications arise out of correlativity, it is worth quoting in full the passage from the Rhetoric that Weinrib himself quotes:
Another instance arises out of things that are mutually related (correlativesFootnote 21). If it is properly said of the one party that it did well or justly then it is also proper to say of the other party that what happened to it was good or just. As where one commanded and the other was prevailed on to carry out the command, or as Diomedon the tax collector pointed out about the business of tax collecting, “If it is not shameful for you to sell the right to collect taxes, it is not shameful for us to buy it.” Just as where it is proper to say of the person to whom something happened that it was good or just then it can also be said of the person who did it that he acted well or justly, so too if it is proper to say that of the person who did it then it is also proper to say it of the person to whom it happened.Footnote 22
The structure of this passage is fairly clear: (i) introduction of another type of proof, from correlatives; (ii) a statement of the general way it operates, allowing an evaluation of one party to the correlative relation to be transferred to the other party; (iii) an illustration, regarding commanding and being commanded;Footnote 23 (iv) another illustration, regarding the selling and buyingFootnote 24 of the right to collect taxes; (v) a fuller explanation of how this proof operates, based on the general mutual transferability of evaluation from the position of a party to whom something happened to the position of the party who did it and vice versa.Footnote 25
Two straightforward points can be made. First, Aristotle does explain how normative implications arise out of the recognition of correlativity. They arise due to the correlative positions being linked as the party who did something and the party to which that thing happened—the active and passive aspects of a single occurrence of conduct. If it can be said of the conduct that it is just for the person doing it then the conduct must be just for the person receiving it, and likewise the other way round. Secondly, the normative implications of correlativity do not amount to an evaluation of correlativity itself. It is not that the correlative positions are good or just because they are correlatives. What correlativity does is to allow the transfer of the evaluation made of one correlative position to the other, whether as good or bad, just or unjust. So, to return to Diomedon the tax collector, if it had been accepted that selling the right to collect taxes was shameful, he would have had to admit that his buying of the right was equally shameful.
These points reinforce the connection suggested above between factual correlativity and basic normative correlativity through the common active-passive structure. In the same way that the normative requirement, R, stood outside the correlative relationship based on the single factual conduct linking the parties, so too with Aristotle’s account of correlativity explored here. For Aristotle, the correlativity is found within the single occurrence of conduct, and the evaluative element attaches to the position of one party to that conduct from an external assessment of that position.
Although Aristotle’s focus is on conduct being good or just and the previous focus is on conduct being required as duty or right, this different focus does not alter the basic normative correlative structure. Admittedly, Aristotle’s concerns are far wider in encompassing active and passive aspects of a single occurrence of conduct that links parties in ways where correlative rights and duties are not part of the picture.Footnote 26 However, Aristotle’s broader understanding of normative correlativity is capable of being applied to those cases where the single occurrence of conduct does link parties where correlative rights and duties are involved.Footnote 27
Once it is accepted that a normative scheme is capable of requiring conduct as well as merely expressing an evaluation of conduct,Footnote 28 then for such conductFootnote 29 the shift in focus from what is regarded as good or just to what can be required as a duty, or right, is simply a matter of perspective.Footnote 30 So those cases of claim-right and directed duty correlativity we have identified as falling under basic normative correlativity can be viewed as coming under Aristotle’s broader principle of normative correlativity, and with that as following its logic of correlativity. This raises the question whether it is possible to move beyond a case of basic normative correlativity to find other cases that satisfy this logic.
D. From Basic to Deeper Normative Correlativity
The active-passive structure of basic normative correlativity is fairly shallow. It can be detected on the surface of the relationship without examining deeper issues that might be relevant to the correlative positions. One such deeper issue is the justificatory basis for the evaluation made of the correlative positions.Footnote 31 We learned from Aristotle that whichever side of the relationship found in a single occurrence of conduct the evaluation is first attached to, it can be transferred by the logic of correlativity to the other side. For the logic to work, there is no need for any reason to be given as to why the one side was selected first, or why exactly the evaluation or requirement holds.
Others have been more concerned with these more advanced matters, specifically, wanting to provide a justificatory rationale for the required conduct. This justificatory element is frequently taken to form part of the understanding of correlativity, so offering a deeper understanding of correlativity.Footnote 32 However, it is not obvious why the justificatory element should be regarded as integral to the correlativity.
Recall that previously we have found an understanding of correlativity dependent on the active-passive structure that applies equally to factual and basic normative correlativity. In the case of normative correlativity, there is an additional element of evaluation or normative requirement, but this operates outside of the factual conduct that forms the basis of the correlation. This understanding of normative correlativity is indifferent to the justificatory basis for the evaluation—so long as some evaluation has been made. So, if it is right for A (a citizen) to pay taxes to B (a tax collector) then it is right for B to receive payment of taxes from A, and we can speak of the correlativity of A’s duty and B’s claim-right, irrespective of whether this requirement is justified on any of the following grounds: (i) A’s responsibility as a citizen to contribute to the general welfare places him under a duty to pay taxes to the appointed collector, B; (ii) B’s right to collect taxes has been acquired by a legitimate purchase from the authorities and gives him sufficient interest to impose the duty on A to pay him the taxes; (iii) the King has decreed that one of his subjects A should pay taxes to another of his subjects B.
Basic normative correlativity holds irrespective of whether the justification for the required conduct is found in A’s position as in (i), B’s position as in (ii), or their joint positions as in (iii). A further point to note is that in any of the three cases the justificatory position of any party differs from that party’s correlative position. In (i) the responsibility of A differs from the active participation in the required conduct; in (ii) the sufficient interest of B differs from the passive reception of the required conduct. And in (iii) their common status as subjects bound by a decree of the King differs from their respective active and passive engagements in the conduct. Stated generally, we could say that what justifies the conduct differs from what conduct that justification requires, even if both can be connected to the same party.Footnote 33
None of these three justificatory bases we have just considered adds anything to our understanding of basic normative correlativity and the logic of correlativity. All any one of them does is to provide something in addition, a justificatory explanation of why the conduct is required, surplus to our understanding of correlativity and its logic. To integrate a justificatory basis within a deeper understanding of correlativity, there must be an expanded understanding of correlativity in which mention of the justificatory basis is not redundant.
The successful accomplishment of this is hard to envisage, given that in justificatory bases (i) and (ii) the justification falls entirely on one party without another (correlative) party being involved; and in (iii) the justification falls on status held in common by both parties rather than producing distinct correlatives for them. In the absence of an expanded understanding of correlativity, there is no opportunity for an associated new logic of correlativity to emerge. It remains possible for correlativity to be used in a way which insists on an added justificatory layer, but this usage does not produce a deeper understanding of correlativity, nor yield a different logic of correlativity.
E. Correlations with Different Types of Rights
Before considering an alternative way in which the basic model of normative correlativity might be modified by taking on a more complex character, it would be helpful to undertake a slight digression into Hohfeldian territory to clarify the different types of rights that may be found in correlative relationships. For present purposes we can restrict ourselves to the first two of the four different correlative relations Hohfeld identified.Footnote 34 The distinction between a claim-right and a liberty, with their respective correlations, is key to the Hohfeldian scheme of analysis and is one of the most commonly confused aspects of that scheme.Footnote 35 The fundamental difference lies in a claim-right always covering conduct undertaken by another party (the duty holder), whereas a liberty always covers conduct undertaken by the liberty holder (the other party having no-right that the conduct should not be performed). Loosely speaking, it is the difference between my right that you do (or omit) something and my right that I do (or omit) something.
So far, so clear. The first problem arises when strict Hohfeldian vocabulary is enforced, and the shorter “right” is preferred exclusively for “claim-right”. Now, I can have a right that you do something but I cannot have a right to do something myself. Contrasted with colloquial usage, it appears counter-intuitive to ban the possibility of speaking of my right to do something, even if a “liberty” to do something is an option.
The problem of awkward vocabulary is compounded by a second problem. Things get worse in that the correlative of a Hohfeldian liberty is a no-right.Footnote 36 So the liberty as a right to engage in conduct has no protection from other parties who might interfere with it. All that the Hohfeldian liberty secures is that the liberty holder does not owe a duty to the other party not to perform that conduct. The failure to secure non-interference has often been regarded as a defective representation of my rights to engage in conduct—sometimes referred to as “active rights”, though that can be misleading since such rights need not relate to active conduct on the part of the right holder, being capable of covering omissions (a right against self-incrimination, rights of conscientious objection, a right not to turn up for work when sick, etc.). However, a fuller appreciation of Hohfeld’s scheme indicates this concern is unfounded.
The right comprising a protected liberty can be recognized. Hohfeld simply insists it has to be broken down into its components, as Ted Sichelman has recently clarified.Footnote 37 If I have a liberty as against you, then any prohibition you are under not to interfere with that liberty is a separate protective duty—in fact, a potential series of duties corresponding to all the different forms of interference that the law prohibits. Each of these is correlative to a claim-right I hold that you do not engage in that interfering conduct. So, if I enjoy a right of way over your land, my positive right to walk on your land amounts to my having a liberty to walk on the land (no duty to stay off) together with any protective rights against your interfering with my so walking on your land.
The advantages or otherwise of adhering to Hohfeld’s analytical and terminological strictures form a subject that we need not pursue here. What is relevant to note are the three types of rights emerging from the above discussion:
(a) a claim-right over your conduct;
(b) a bare liberty, acknowledging I may engage in conduct without breach of duty to you;
(c) a protected liberty amounting to a positively protected right to engage in conduct, which we can refer to as a positive right. Footnote 38
There is the possibility of finding each of these in correlative relationships with other normative positions. We have clear correlatives for the first two in orthodox Hohfeldian terms: (a) (claim-right)-duty correlation; (b) liberty-(no-right) correlation. However, it is worth pausing to recognize that although (a) is familiar as the paradigm case of basic normative correlativity, (b) displays some significant differences. Most significantly, there is no active-passive structure directly linking the two correlative positions. A’s liberty to walk on B’s land exhibits an active aspect of conduct which does not feed into a passive aspect of that conduct exhibited in B’s no-right. Even if we reach for an abstract formulation of the conduct, as we did earlier in speaking of the giving and receiving of a benefit under a contract, this does not help. There is no reception of anything by B amounting to a passive aspect of the conduct found in A’s liberty.
The difficulty here can be addressed by looking more closely at the negation process involved between type-(a) correlation and type-(b) correlation. The move from a duty in (a) to a liberty in (b) is not a straightforward negation.Footnote 39 The straightforward negation of A’s duty to stay off the land in (a) would be A’s no-duty to stay off the land. Instead we find in (b) A’s liberty not to stay off the land, or more fluently, A’s liberty to enter the land. This amounts to a negation in two stages:
(I) duty to do TO no-duty to do;
(II) no-duty to do TO liberty not to do.
The first stage is a simple negation, whereas the second stage involves a deontic implication.Footnote 40
If we kept to the simple negation at (I), then the liberty in (b) would be recast as a no-duty, giving us: A’s no-duty to do something for B correlates with B’s no-right that A do that thing. We still do not have an active-passive structure involving a single occurrence of required conduct, as found in the basic case of normative correlativity in (a). But what we do have is a simple negation of that. The basic normative correlativity and the logic of correlativity from (a) is then followed in (b)—by way of negation. The additional stage (II), although perfectly appropriate, serves to conceal this point, and we need to trace the formation of type (b) correlation back to stage (I) in order to appreciate it. This diversion to establish that liberty-(no-right) correlation can be assimilated by way of negation under basic normative correlativity serves as a helpful introduction to the discussion of correlation involving the third of the three types of right we have encountered, (c) a full positive right.
F. From Basic to Complex Normative Correlativity
By the expression a full positive right we have been referring to a right to do something, a liberty, which also includes protection from acts of interference that would frustrate the exercise of that liberty. Bringing the duty not to interfere within the ambit of the positive right makes this is a far more meaningful position for the other party in its relationship with the right holder than the empty no-right correlated with the bare Hohfeldian liberty. It comes as no surprise when the suggestion is made that this duty not to interfere should be regarded as the correlative position of the liberty, producing a liberty-duty correlation.
Negatively, this move does two things. It banishes the previously recognized no-right as a correlative to the liberty, supplanting liberty-(no-right) with liberty-duty as the recognized correlation. Also, it ignores any claim-right correlatives that would otherwise attach to the duty or duties not to interfere, those claim-rights held by the liberty holder correlative to the duties not to engage in conduct interfering with that liberty—recognized by Hohfeld, as we noted above. These negative implications are not insignificant but will not be explored here. It is the positive outcome of the move that is of interest here, found in the distinctive correlation between a liberty and a duty not to interfere.
A terminological clarification is crucial at this juncture. The basic Hohfeldian distinction between those rights that I have over your conduct and those rights I have over my own conduct holds true, whether or not we adopt his terms and whether or not we use his correlatives. Frequently, the single word “right” is employed to cover both types of rights. Although accurate as a report of colloquial usage, this is misleading in a discussion of correlativity. And it remains misleading even if we are discussing not merely (b) bare Hohfeldian liberties but (c) full positive rights, the liberty plus the protection found in a duty not to interfere.
A matter not immediately obvious, but of fundamental importance is that the protection from a duty not to interfere incorporated within a full positive right can only attach to a liberty, understood as a right to my conduct. It cannot attach to a claim-right, understood as a right to your conduct. For it is nonsensical to speak of your being under a duty not to interfere with your own conduct. Unfortunately, both Weinrib and Zylberman fail to grasp this key point, as we shall see below, through blurring the liberty/claim-right distinction. For the moment, it is important to signal that claim-right and liberty will be used strictly here to distinguish between rights to your conduct and rights to my conduct, irrespective of whether we are otherwise adopting the Hohfeldian scheme at the time. A supplementary point on this terminological convention, is that liberty (which is more widespread) will be used consistently in preference to privilege (the synonym that Hohfeld himself and others have preferred) for a right to my conduct.
The task of considering correlation involving (c) a full positive right as a liberty-duty correlation is made easier by referring initially to Heidi Hurd’s proposal to recognize it. Hurd explicitly constructs her version of this correlation on a liberty and so avoids any confusion between liberty and claim-right within her proposal. She does, however, suggest modifying the terminology when advancing her analysis of a full positive right, which is referred to not as a protected liberty but as a permission.Footnote 41 Along the way, she dispenses with liberty as a term by imposing an unjustified restriction on a bare Hohfeldian liberty to a setting that is normatively inert (“a moral state of nature”).Footnote 42 Hurd’s route to her liberty as permission (i) commences with a liberty to do an act, (ii) takes the liberty to be a right, and (iii) concludes it must be afforded the protection from a duty not to interfere.Footnote 43 Since (i) and (ii) employ straightforward features of a Hohfeldian liberty, anything added by (iii) may strengthen the liberty but does not alter the basic character of the right-holder’s normative position as a liberty.Footnote 44
In her 1999 book, Hurd does not speak of the connection between liberty and duty not to interfere as correlation (although she has discussed it in such terms subsequentlyFootnote 45). Instead, she promotes it through what she calls the “correspondence thesis”. This regards an action and another action which would permit or prevent the first action as “codependent actions”, and holds that the justifiability of the former determines the justifiability of the latter.Footnote 46 Hurd explains her thesis in the following way:
The correspondence thesis rests on the intuition that, since an action cannot be simultaneously right and wrong, it cannot be the case that one actor may be justified in performing an act while another may be simultaneously justified in preventing that act.Footnote 47
An outworking of the correspondence thesis is found in the “correspondence”, or correlation, between the liberty of one party and a duty not to interfere of another party. A corollary follows which fortifies Hurd’s major campaign to avoid moral combat, or normative conflict. Being under a duty not to interfere, the other party cannot assert a liberty to do anything that amounts to an interference, and so cannot maintain a conflicting liberty with the other party’s original liberty. Hence no possibility of a conflict between opposing liberties.Footnote 48
It is striking that Hurd’s codependence and correspondence between the liberty and the duty not to interfere suggest the same kind of transfer of normative evaluation as we found above in basic correlativity: because the liberty is justified so too is the duty not to interfere with it. However, this type of correlation is altogether more complex, and does not follow the simple logic of correlativity we found in basic normative correlativity. The reason is that the liberty-duty correlation is concerned with connecting two separate actions as codependent.
Hurd’s exposition of the correspondence thesis excerpted above does not make this point totally clear.Footnote 49 In speaking of the underlying intuition as applying to a single “action”, and then describing the positions of the correlative parties as amounting to performing and preventing a common “act”, a strong sense of connection through a single occurrence of conduct is produced. The reality is that there are two different actions, or occurrences of conduct, in play here. The conduct of the liberty holder in exercising the liberty, and conduct of the duty holder amounting to an interference with that conduct.
It is possible to find correlatives for each of these cases. Take as an illustration:
(1) A’s liberty to walk on B’s land
(2) B’s duty not to interfere with A’s walking on the land
Then the correlative of A’s liberty in (1) is B’s no-right, and the correlative of B’s duty in (2) is A’s claim-right.Footnote 50 What Hurd’s correspondence-correlativity aims to do is to link A’s liberty in (1) to B’s duty in (2). This cannot be achieved through basic normative correlativity and cannot then rely on the associated logic underwritten by Aristotle. Hurd herself admits that there is no necessary logical connection between the two.Footnote 51 So this more complex, type-(c), correlation requires the introduction of a further premise into a particular normative scheme—seen in the very proposal of Hurd’s correspondence thesis.
However, even if we adopt the use of this more complex type of correspondence-correlativity, there are doubts about the soundness of reasoning with it. These doubts arise from the same feature that took it outside basic normative correlativity, the presence of two occurrences of conduct. Now the problem is that the description and justification of the one conduct does not in itself authoritatively determine the description and justification of the other. With one occurrence of conduct, there may, of course, be disputes over how that conduct is to be understood, but once that has been determined, the determination holds good for both the active and passive aspects of it. With an attempt to relate together two forms of conduct (the content of the liberty and what counts as interference with it) not only are the opportunities for indeterminacy doubled, but there is an additional complexity in that the determination of the one may then create uncertainty over the determination of the other—producing a dynamic interplay of indeterminacy between the two corresponding actions.
So, for example, A insists on a liberty as in (1) above and claims B is in breach of a duty as in (2) when B forcibly removes A from his land. It turns out that A has been walking up and down the small section of a public right of way adjoining B’s residence for more than an hour at 3.00 am, in a drunken state, singing raucously. B argues that his action amounts to abating the nuisance caused by A, rather than interfering with A’s walking on the land under (2); and, moreover, argues that A’s liberty to walk on the land in (1) is restricted to using the right of way for passage and does not extend to continuous occupation of a small section of the right of way, nor to engaging in a nuisance. This demonstrates how one cannot simply transfer a normative evaluation of one correlative position to the other within Hurd’s correspondence relationship, as was possible under the logic of correlativity applicable to basic normative correlativity. Now we have the possibility of balancing countervailing normative considerations: here is a rationale for justifying the liberty; there is a rationale for holding the alleged interference as justifiable, which, if accepted, would oppose the rationale for the liberty and ultimately restrict the liberty.
The possibility of competing normative evaluations emerging from each correlative position, instead of a common evaluation being transferable in either direction between them, leads to another significant difference. The correlation between liberty and duty not to interfere is not a stable relationship. It is possible that the evaluation of the liberty side—say, a positive assessment of the benefits of engaging in free speech—will compete with a competing evaluation on the interference side—say, a recognition of the value of individual reputation—producing different outcomes on different occasions when the balancing of the countervailing considerations is undertaken. In one jurisdiction, the outcome is to provide strong support for reputation in general, so restricting the liberty in a uniform way. In another jurisdiction, a distinction is made for political reputation, with less restriction on the liberty in cases of political speech. There is, then, the possibility of differentiated protections accompanying the liberty, as a result of how the actual balancing works out. This set of differentiated correlative positions for the liberty is a logical impossibility in a case of basic normative correlativity where the positions are tied together as the active and passive aspects of a single occurrence of conduct.
This leaves us with two different kinds of correlativities being used for different types of rights: the type-(a) claim-right correlationFootnote 52 amounts to a case of basic normative correlativity and follows the logic of correlativity; the type-(c) full positive right (protected liberty) correlation involves the more complex correspondence-correlativity which fails to adhere to the logic of correlativity. In the light of that finding, abiding by a consistent terminology for different types of rights is of paramount importance in discussions of normative correlativity.
III. Weinrib’s Use of Correlativity
We noted in the Introduction Ernest Weinrib’s association of correlativity with reciprocity. Interestingly, Weinrib builds his idea of private law as corrective justice by first taking the correlativity of corrective justice as being highlighted by Aristotle before supplementing the correlative structure with reciprocity derived from Kant.Footnote 53 Had Weinrib sought immediate endorsement from Aristotle for his close connection of the two within corrective justice, he would have been disappointed. An explicit statement from Aristotle notes that (despite the efforts of the Pythagoreans to connect them) “Reciprocity however does not coincide … with Corrective Justice”.Footnote 54 Enough for the moment to raise doubts over synonymity between correlativity and reciprocity. A full challenge is delayed until section IV, where Zylberman’s effort to fuse together correlativity and reciprocity is considered.
In this section I shall concentrate on four flaws in Weinrib’s approach to correlativity. The flaws are intertwined in much the same way as the elements of his idea of private law, described as “mutually supporting ideas”.Footnote 55 They can be conveniently addressed in the following related topics: Weinrib’s identification of corrective justice with correlativity; his associated view of a logic of correlativity; his location of justificatory considerations within correlativity; and, his related inference of equality for the correlative parties from those justificatory considerations (paving the way for reciprocity).
A fifth, technical flaw found in Weinrib’s failure to recognize the distinct types of claim-right correlation and protected liberty correlation can be regarded as exacerbating each of the other four,Footnote 56 but we shall not expand the discussion to include this here. Again, this topic is postponed to the next section, since Zylberman deals more obviously with liberties alongside claim-rights in his portrayal of human rights.
The truth of the matter is that Aristotle does not highlight correlativity in his discussion of corrective justice. It is fair to say that the passage Weinrib refers to contains examples of correlativity,Footnote 57 but Aristotle does not highlight this feature. He does not even mention correlativity.Footnote 58 Nor are there other grounds for identifying correlativity with corrective justice. As was pointed out above, dishonourable or unjust relations can exhibit correlativity as much as just ones. And also, relations exhibiting the active-passive structure for which no question of corrective justice arises, as with the relation between commander and commanded. Even in the case of (claim-right)-duty correlation, it is perfectly intelligible to express the view that the legal duty of an employee to work oppressively long hours within a particular legal system is unjust, without denying the correlativity between the employee’s duty and the employer’s right.
Weinrib’s view of a logic of correlativity builds upon his flawed identification of corrective justice with correlativity, in taking the logic to be displayed in terms of the specific operation of corrective justice rather than to be exhibited in the general features of correlativity. To maintain this logic, Weinrib has to do two things. Negatively, he has to ignore the logic that is apparent in the features of correlativity itself, based upon its active-passive structure.Footnote 59 Positively, Weinrib then has to provide a logic implicated in what he sees as the values of corrective justice.
The positive stage is more sophisticated than the stark denial found in the negative stage.Footnote 60 In order to convert the logic of correlativity into a logic of corrective justice, Weinrib takes the logical issue to be found in a corrective justice context and to be concerned with the process of reasoning that Aristotle indicates is peculiar to corrective justice. Weinrib selects a key sentence from Aristotle to advance his cause:
For it makes no difference whether a good man has defrauded a bad man or a bad one a good one, nor whether it is a good or a bad man that has committed adultery; the law looks only at the nature of the damage, treating the parties as equal, and merely asking whether one has done and the other suffered injustice, whether one inflicted and the other has sustained damage.Footnote 61
Weinrib’s argument turns on his division of this sentence around the semicolon, asserting that the first part indicates what does not matter and the second part what does matter for corrective justice.Footnote 62 In this way, the phrase “treating the parties as equal” becomes a key part of what does matter for corrective justice. However, a better reading of the sentence (brought out in the translation provided) is that this phrase after the semicolon simply summarizes what has been illustrated in the examples before the semicolon. That is to say, it repeats what does not matter for corrective justice—no reference to the status of the parties, they are treated the sameFootnote 63—alongside what does matter.
In support of the view taken here of what amounts to a better reading, three points can be briefly made. First, Aristotle’s contrast with distributive justice falters if we understand corrective justice to be concerned with the status of the parties in a positive way. The contrast is diluted as holding only between the possibility of a differentiated status for distributive and a uniform status of equality for corrective justice—equality having already been discussed as one possible status for distributive justice in Nicomachean Ethics, V, 3. Secondly, if “treating the parties as equals” were to be regarded as a positive requirement for corrective justice, one would expect further discussion of it—as appeared with the positive requirement of status for distributive justice. Yet if the phrase is taken to indicate that status is not a consideration, the absence of further discussion we actually find is perfectly natural. Thirdly, there is further discussion in this passage of what Aristotle does consider significant for corrective justice. Aristotle continues after the quoted sentence, until the end of chapter 4, with a discussion of corrective justice for the parties as amounting to seeking equality as a mean between gain and loss.
The evidence of Aristotle’s text points to the conclusion that Aristotle is not concerned with the equality of the parties, and that the only equality he is concerned with in his discussion of corrective justice is an equality between gain and loss. And correlativity plays no role in this discussion. The correlativity illustrated in passing between the doing and suffering of harm, as setting the scene for an inquiry into corrective justice, cannot connect the gain and loss since they are not connected as active and passive aspects of conduct. That analysis could work for the loss, in isolation: the causing and suffering of loss. It does not work at all for the gain, since the party gaining from the involuntary transaction of causing loss to the other produces the gain for himself. Aristotle remarks that to even speak of gain and loss, and equalizing them, is problematic in these circumstances. He attempts to resolve this conundrum, not with the resources of correlativity but by drawing an analogy with exchange in voluntary transactions.Footnote 64
Weinrib is, accordingly, wrong in taking the logic of correlativity to be displayed here in Aristotle’s understanding of the operation of corrective justice. However, once correlativity and its logic have been identified with corrective justice by Weinrib, two further errors are committed with little extra effort. The justificatory considerations required for corrective justice are located within “the correlativity of doing and suffering”;Footnote 65 and also the equality of the parties that has been interpolated by Weinrib into Aristotle’s understanding of corrective justice is implicated in these justificatory considerations. Weinrib proclaims “normative equality” as “implicit in the correlativity of doing and suffering”.Footnote 66
These last two steps in his argument are needed to prepare Weinrib’s representation of corrective justice as linked to his view of Aristotelian correlativity for completion in a full account of equality furnished by Kantian right.Footnote 67 The implications of a corrected understanding of Aristotelian correlativity for Weinrib’s enterprise should not be underestimated. The misrepresentation of Aristotle is not simply found in a secondary argument used by Weinrib to strengthen the case for his understanding of correlativity within private law, so that disagreement over the understanding of Aristotle does not impair Weinrib’s central case for his view of the structure of private law. Weinrib’s case depends upon a logic of correlativity and without the support of Aristotle there is no logic to draw upon. Moreover, it is not simply a matter of seeking alternative grounds for a logic of correlativity, that are not attributable to Aristotle, which can offer support for correlativity which does deliver an equality between the parties. The burden on Weinrib is also to deal with the detailed argument Aristotle has provided for a logic of correlativity that, as we saw in section II.C, permits a transfer of evaluation between correlative parties without having anything to say about an equality between the parties. Indeed, as the examples Aristotle provides illustrate, there may well be asymmetry between the parties (commander and commanded, seller and buyer).
The further implications of a corrected understanding of Aristotelian correlativity for the general viability of an equality derived from Kantian right will be considered in section V. Meanwhile, one final point can be made on Weinrib’s mischaracterisation of correlativity. Apart from a detailed consideration of his arguments and an intricate examination of Aristotelian texts, there is a simple way of capturing Weinrib’s error in attributing normative equality to correlativity. Weinrib notes that in a case of correlativity, what is “predicated of the doer can equally be predicated of the sufferer.”Footnote 68 What is predicated of the doer and sufferer in this passage is an evaluation (of injustice), and it is sound to say that the same evaluation can be applied to both doer and sufferer through the recognition of correlativity. So we could say, using Weinrib’s formulation, that the same evaluation can equally be applied to doer and sufferer. Still sound. What changes three lines down is Weinrib associating “equality between the parties” with correlativity. This is something completely different. An evaluation equally applied to both does not amount to an evaluation of both as equals.Footnote 69
IV. Zylberman’s Reciprocal Correlativity
Equality of persons also figures prominently in Ariel Zylberman’s vision for human rights, but his route to attaining it differs from Weinrib’s. Zylberman establishes equality (and with that non-instrumental human rights) from the outset in “a master norm of Reciprocity”,Footnote 70 whose status as a “relational, deontic norm” is tied to correlativity, in that “it represents the correlation of directed norms”.Footnote 71
This last phrase is rather puzzling. Is it intended to convey norms involving correlation between the parties, or something concerned with the relationship between the norms? Zylberman amplifies the phrase immediately as: “the rights of one person against another or, equivalently, the directed duty of respect any person owes another.”Footnote 72 This is not particularly helpful. It seems to be alluding to basic (claim-right)-duty correlation, but in a rather oblique way. More is revealed in Zylberman’s definitive statement of Reciprocity:
Reciprocity: A has a basic claim right to independence against B; or, equivalently, A has a basic duty to respect B’s independence.Footnote 73
The equivalence between right and duty is suggestive of correlation, yet the switch between A being owed the duty from B to A owing the duty to B is indicative of reciprocity. Is Zylberman’s statement of his master principle a coherent fusion of these two qualities or just a muddle?
We need to examine precisely how these two qualities can combine. The two may coincide, but any such coincidence is far from being a simple matter. Take the giving of a gift. This provides a case of correlativity: X being the donor of the gift to Y is the correlative of Y being the donee of the gift from X. The relationship between them is correlative but not symmetrical: one gives, the other receives. Neither can it be described as reciprocal, since that requires a response in kind: one gives, the other gives. For that we need another gift coming in the other direction. X gives a birthday present to Y and on X’s birthday Y reciprocates by giving a birthday present to X. Now we have correlativity and reciprocity, but in order to achieve this, two occasions of correlation are needed. And, the correlation in the one case is the inverse of the correlation in the other. When reciprocity between X and Y does occur it does not coincide with any single case of basic correlativity but across two relationships of correlativity, where one expresses the inverse of the other.
The response on behalf of Zylberman might be, the master principle of Reciprocity is relying on just such a case of inverted correlativities in order to establish reciprocity. However, Zylberman’s principle does not simply stipulate an occurrence of inverted correlativities for reciprocity. It requires them to be equivalents. But any normative equivalence based on a mutually transferable evaluation within a single occurrence of basic correlativity does not travel across inverted correlativities.
Recalling the example of the reciprocal gifts, the initial position of donor cannot be related as equivalent to the subsequent inverse position of donee in the same party, as Zylberman’s statement has it. It is not the same gift. The particular birthday present that you give me in return will not be the identical present I gave you.Footnote 74 The reciprocal receipt of a gift neither coexists with, nor are its practical details inferable from, the initial giving of a gift. There is no logic of correlativity that can tie these together.
In his attempt to link as equivalents and as mutually inferable, A’s duty to respect B’s independence and B’s duty to respect A’s independence,Footnote 75 Zylberman is making an unwarranted assimilation of correlativity and reciprocity. That is, if the above observations apply just as much to reciprocity over human rights as they do to reciprocity over birthday gifts. This could be challenged on the ground that, unlike the reciprocal birthday gifts whose specifications lie in the hands of the individual donors, the reciprocal duties have a common origin and specification in a single authoritative source.
This defensive move is unsound. It confuses a common source for the duties of A and B with a common specification of their actual duties (and the respective correlative rights). The move is encouraged by Zylberman portraying the pairs of rights and duties in terms of the general wording of a common source, such as a duty to respect independence, or a duty not to use force. However, the actual duties on A and B as owed to B and A, derived from this common source, will vary considerably due to the different circumstances of both parties in the relationship. A’s actual duty to respect B’s independence will depend upon the actual mode of independence that B is capable of pursuing, and upon the actual opportunity A possesses to interfere with it; and these contingencies will be different when their roles are reversed. And if we extend the analysis to rights of contract and property,Footnote 76 it is evident that A/B’s actual duty to respect B/A’s property will depend upon the actual property that B/A possesses as well as the opportunities afforded to A/B to interfere with it. And similarly, with contractual rights.Footnote 77
The point being made here could be made more eloquently. We are not all equal under the law simply because the same law is applied to our inequalities. Required perhaps to combat Zylberman’s own eloquent peroration for the principle of Reciprocity:
[T]he reciprocal character of this principle means that A and B have the same basic right to independence and the same basic duty of respect. You and I are one in the basic right to independence, equal in our dignity as persons.Footnote 78
More prosaically, once an accurate analysis of the connections between reciprocity and correlativity has been undertaken, we can see that A and B have neither the same right nor the same duty proclaimed by Zylberman.
The equality that Zylberman spuriously attributes to the parties is vulnerable on another front. The logical inference imported from the basic form of (claim-right)-duty correlation is put to work within his statement of Reciprocity to support a protected liberty correlation. The “interdependent inferential relations” are found where: “A has a right to independence against B” and “B has a duty to A to respect A’s independence”.Footnote 79 This amounts to a protected liberty, or full positive right, in a liberty-duty correlation as recognized by Hurd.Footnote 80 This point is obscured by Zylberman occasionally reverting to clear cases of (claim-right)-duty correlation in his subsequent discussion, such as A has a claim-right that B do not enslave him and B has a duty not to enslave A.Footnote 81 However, since the master principle of Reciprocity asserts as foundational A’s right to independence, then any obligation on B not to enslave A has to be taken as secondary, as a means of protecting A’s independence.Footnote 82
Acknowledging that Zylberman’s “right to independence” amounts to a full positive right over the holder’s conduct and not a claim-right to the conduct of another has implications beyond clarifying the type of correlation involved and recognizing the absence of a logic of basic correlativity. We saw in our discussion of Hurd’s correspondence thesis that an attempt to relate together two forms of conduct (the content of the liberty and what counts as interference with it) carries with it a prospect of competing normative evaluations of the two and, accordingly, the possibility of differentiated protections accompanying the liberty. This means that once the actual liberties of different parties have been reached, allowing different degrees of protection to be established, the anticipated equality of independence under the abstract principle of Reciprocity becomes even more remote.Footnote 83
V. Concluding Remarks
The scheme of intelligibility for the uses of correlativity developed in section II has been illustrated and amplified in the preceding two sections through applying it in critical assessments of Weinrib’s idea of private law and Zylberman’s approach to human rights. The work of these two authors is representative of a growing body of scholarship relying on Kantian inspiration to make sense of contemporary demands on law to promote, and be defensible as promoting, mutually respectful relations between members of society.Footnote 84 Since both authors have invested in views of correlativity and reciprocity at a foundational level, any force in the critiques mounted here is likely to have far-reaching implications. These implications start with our understanding of some basic characteristics of law.
Correlativity is clearly present at the point of resolution of a legal dispute when a normative evaluation (a judgment) is expressed that mutually affects both parties.Footnote 85 What has become clear is that despite a mutually transferable evaluation being conveyed by correlativity, the correlative positions of the parties may manifest an asymmetry not an equality. Typically, at the point of judgment the status of one party is dominant and that of the other is subordinate.Footnote 86 Once this asymmetry is generally recognized within a correlative relationship at the point of legal judgment, it can be found in a judgment specifically concerned with matters of corrective justice. Once it is admitted at the point of judgment, there is nothing to stop it being traced all the way down to the underlying social injustice that a case of corrective justice is intended to remedy. Corrective justice to restore a proper asymmetry between employer and employee, between landlord and tenant, etc., becomes plausible; even corrective justice to establish through redistribution a fairer asymmetryFootnote 87 between members of society is now on the table.
The implications for Weinrib’s theory of private law relying on a narrow conception of corrective justice are particularly significant. His reinforcement of the bonds between correlativity, corrective justice and equality is essential for securing a structure for private law that is self-contained all the way down, from its manifestation in positive law to its justificatory basis.Footnote 88 It is the security of this structure that banishes any consideration of other justificatory elements, including distributive concerns.Footnote 89 We noted in the Introduction, the reservations of Cane and Steiner over Weinrib’s banishing of distributive justice from private law. The rejection of Weinrib’s basic tenets in this articleFootnote 90 amounts to a dismantling of his fortified structure for private law and, accordingly, opens up the justificatory basis for private law correlativity, as well as for mechanisms of corrective justice, to distributive and other concerns.Footnote 91
A general problem for Kantian thinking on social and legal relationships is raised by the failure of the efforts of Weinrib and Zylberman to construct a viable understanding of equality between the parties through their understandings of correlativity and reciprocity within private law and human rights. Fairly obviously, if the Kantian approach is going to appear credible it will need to have a demonstrable impact at the level of private law and human rights, the level at which Weinrib and Zylberman have laboured so energetically, but a level with which Kant himself did not engage. Kant’s principle of right intended to secure an equality of willsFootnote 92 has need for a plausible instantiation at the level of legal practice. This need may turn out to be greater than the need an “equality of corrective justice” has for acquiring “normative force from Kantian right” that Weinrib has falsely assumed.Footnote 93 If we conclude that correlativity and corrective justice are not bound up with an equality issue after all, then that has knock-on effects for our views of private law and human rights, and their need to have any recourse to Kantian right. Conversely, Kantian right is deprived of convenient ways of satisfying its need for instantiation. More troubling still, the dismissal of the equality issue as spurious to our understanding of correlativity, and the admission of asymmetry, bring into question the conviction that social relations are based on dealings between those enjoying an equal status as autonomous agents.
