Skip to main content

Punishment and Exchange

November 2nd, 2021 | 19 min read

By Oliver O'Donovan

It seems to some the very epitome of a “mere” orthodoxy in the worst sense — too juridical for God’s mercy and too arbitrary for God’s justice — even to those who are not in the habit of yielding too quickly to theological fashions, it is a puzzling thought that God could accept the sufferings of the innocent as an atonement for the offences of the guilty. But it is worth thinking further about how far such puzzled and outraged reactions can take us.

We are puzzled whenever reason confronts a difficulty. It presents us with a threat; we fear the loss of contact with reality. For it is the function of reason in every aspect of its operations – intuitive, discursive, logical, imaginative, prescriptive, etc. — to keep us in touch with reality. If reason sometimes dares to loosen its ties with reality, it is only in order to tie them again more tightly. Reason can explore counter-factual hypotheses, can imagine alternative worlds, can frame purposes and intentions that are to change the world it knows; but all these explorations, imaginations and purposes have to be “realistic,” bound by the conditions of logical consistency and conceivable possibility. And yet reason is not immune to the threat of a breach with reality. It can find itself challenged, astonished, bewildered, even outraged by reality, for reality is always much wider and deeper than reason can accommodate. Kierkegaard described this dramatic affront of reason as “the paradox.”

We experience “paradoxical passion” when reason has run to its limits, and is stopped dead in its tracks. And we can, he believed, respond in one of two ways to this. We can recoil from the paradoxical reality in “offence,” or we can embrace the paradox in a “leap of faith,” suspending the assumptions of reason that stand in its way. Reason then steps back, and looks for a new and more adequate approach to reality. There are some quite ordinary rational processes that illustrate this kind of rational re-calibration. The defeat and revision of scientific theories is one commonplace example, when prevailing theoretical understandings have to give way to new observations, and an imaginative leap is needed to frame new theoretical perspectives. But Kierkegaard’s own interest in the phenomenon was theological. The term “paradox” was suggested to him by Saint Paul’s reference to the “scandal” of the cross.

We are bound to wonder: can moral paradox ever be met by such a leap of faith? Reason, as we have been taught since Aristotle, proceeds on two fronts, theoretical and practical; it considers not only what is the case, but what is to be done. And while theoretical reason follows reality, describing it in the light of observations, practical reason leads reality, forming prescriptions for action out of moral first principles. So it may seem that practical reason, at least in its higher form as moral reason, cannot respond to perplexity simply by surrendering its ground. That would be morally inconsistent, and so irrational. Moral reason is sometimes said to be “unconditional,” by which is meant that its intuitions of good and evil, right and wrong, come first. They set the conditions for whatever it may subsequently learn from engagement with empirical reality. Right is right and wrong is wrong; it makes no difference if we learn that very many people in fact do wrong and only a few do right. Moral reason must stick to its principles and defy the world, as the proverb says, “even though the sky falls!”

That objection would be decisive, if it were in fact true that the sole function of moral reason was to prescribe. Our moral discourse is not confined to saying “Do this!” and “Don’t do that!” It describes the moral quality of actions, events and states of affairs, and it can do it more attentively or less attentively, more selectively, more comprehensively, and so on. It is a mistake to think of moral thinking as a kind of peremptory pronouncement, a rhetorically defiant first and last word. There are, in fact, two operations involved in moral reason.

One, which we call “reflection,” does things like “praising,” “blaming,” “evaluating,” “recognising,” “admiring,” “deploring,” and so on, attending to attending to morally significant realities, to good and evil, beauty and ugliness, importance and triviality, etc., as they show up in the situations around us.

The other, which we call “deliberation,” brings these reflective appreciations to the point of action: it “seeks” goals, “avoids” mistakes, forms purposes, “plans” courses of action, “decides” between alternatives. We ask first whether a situation is good or evil, and only then what is to be done about it. “He has shown you, o man, what is good…” says the prophet (Micah 6:8), “now, what does the Lord require of you…?”

There are two questions, “what is good” and “what does the Lord require.” Two answers are required, connected by a “now,” or a “therefore,” which makes the second answer coherent with the first. Moral reason is a process of inference, not simply an intuition. Its descriptions and its prescriptions correct one another. And that is how moral reason, too, may confront “paradoxes,” realities that it does not know how to describe adequately and so cannot prescribe for. Moral paradoxes arise in the most ordinary day-to-day experience of facing a dilemma.

Dilemmas have not been well treated by recent moral philosophy. It has preferred to focus our thoughts on the high ground of virtues and ideals, and has encouraged us to relegate dilemmas to the margins as a kind of anomaly, a disturbance on the calm surface of virtuous practice. But this is a dangerous temptation, for dilemmas are part and parcel of our real moral condition. Conducting ourselves well in the real world is often a puzzling business. Authentic moral conviction is rarely given us in a flash, but usually has to be searched for, sometimes with great anxiety. There are dilemmas that are purely deliberative, concerned with finding means to achieve ends: I know what I want to achieve, but I must look hard for the most suitable ways to achieve it. Others arise on the borderline between reflection and deliberation, when an unusual situation demands something that goes against the grain of an ordinarily virtuous habit of mind:

I am used to thinking carefully about decisions, but must now be decisive and resolute; I am used to speaking frankly, but must now be guarded and evasive. Others again arise wholly within the reflective sphere. Others again are true “paradoxes,” challenges to settled moral belief, which create the suspicion that the world is morally at odds with itself, inhospitable to virtue: I may believe that force is opposed to peace, and yet realise that peace may require force to defend it; I may believe in candid speech, and yet recognise situations in which only lying can protect someone who depends on me, and so on. The paradoxes associated with the death of Christ belong to this third type: we believe in the inviolability of innocence, but cannot deny the truth of what Saint John called the “prophecy” of Caiaphas: “It is expedient that one man should die for the people.” In these conflicting pairs of principles, each appears valid on its own terms, and neither can be simply refused. Moral reason seems to contradict itself, and so we may come to doubt the ground of moral reasoning itself, the faith that good undergirds the world of action. This is the moment of “offence,” which will leave us either sceptical or desperate.

There is, however, a theological strategy on which we may pin hopes of escaping from the perplexity. There is one absolute limit on the competence of moral reflection: it cannot tell us about the being and action of God. Moral reason thinks by analogy, judges every particular situation and action in the light of features it shares with other situations or actions of a similar kind. God’s sovereign action, however, allows no comparisons. Can we not resolve the paradox, then, simply by removing God from the scene in which it arises, insisting that he is, as some medieval theologians liked to say, exlex, “exempt from all laws?”

We should not expect to trace the outskirts of God’s ways. “Let earth adore!” we may conclude. “Let angel minds enquire no more!” But then we face a problem: if earth can and must adore, can it also praise? To “praise” is to declare that what the sovereign God does is supremely good, but if moral reason is dumb before the mystery of God’s will, it is clearly impertinent to praise God for his goodness. Of course, there are moral questions about God’s acts that are so badly conceived that they should never be asked. “Would God not have done better by creating the world earlier than he did?” was one such question that used to annoy the church fathers. But not all moral questions about God’s deeds are like that. If the praise of God’s goodness is to have any place in creaturely worship, there must be ground for the moral faith that God upholds and vindicates the moral order. Which is not to deny, of course, that we must learn about the order God upholds and vindicates from what he actually does. Yet the reason we can learn is that we can recognise him as the sovereign good. “Good art thou, and doest good,” said the Psalmist, and continued, “teach me thy statutes!” (Psa. 119:68). The good that God is, leads to the good God does, and then to the good God teaches us. Moral reason is not left dumb in the face of his works and commands.

So much in general about moral reason and its paradoxes. Now we must focus on the particular paradox presented by the theologians’ theory of “penal substitution,” that on the cross Christ suffered punishment for the world’s offences, the innocent for the guilty. Not everything that is said and thought about Christ’s death need, or should, be a “theory” in this sense. Catechetical teaching is often formed non-theoretically, as a simple string of declarations, and the language of praise often revels in paradoxes that it makes no effort to resolve. The famous hymn of Venantius Fortunatus, “Sing, my tongue, the glorious battle” parades a string of unresolved paradoxes about battles won by dying on a cross; it would be foolish to criticise it for that. It is precisely the rhetorical genius of praise to evoke a deeper sense of the mystery by heightening paradox. But it would be foolish, too, to think of these paradoxes as a kind of theory.

Theological textbooks have sometimes claimed to discover in such texts a “theory” of the atonement as a victory. Whatever this language does, it does not theorise. Theory is not all-important to theology, but it does have its own limited importance. It is responsible for the rational consistency of faith, resolving paradoxes that can be resolved, and accounting for those that cannot be resolved in terms of the basic mysteries of faith itself. It does this in order to help us think, and if our praise is to amount to more than exuberant rhetoric, we are wise not to ignore the help it offers. Where Scripture uses many terms to describe the death of Christ — “sacrifice,” “judgment,” “ransom,” “remission,” “reconciliation,” “victory,” “payment,” “humiliation,” etc. — a hymn, a prayer or a sermon may select, or may cheerfully mix them up together, while a theoretical account seeks to set them in order, deciding which have exegetical and logical priority, suggesting how those that appear to be in tension can shed light on one another, and so on. And though an account of Christ’s death is not primarily a task of moral theory, if it is to speak of God’s overcoming of the sin and meaninglessness of human action, it must at least satisfy moral reason.

When we speak of the redemption wrought by the cross of Christ, we speak of a wholly unique act, one for which there are no general types, no regular explanatory patterns, no set of premises on which it might have been predicted. Here God acted once and decisively, to determine the ultimate fate of the created universe. That is why this act creates a paradox for a reason accustomed to look for regularities and patterns. But there are other explanatory resources than regularities. There is a logic proper to narrative, which treats every event as unique, yet displays the consistency and intelligibility in a sequence of events. This is the logic we recognise in God’s faithfulness to his own covenanted acts: “he cannot deny himself” (2 Tim. 2:13). The unique can be made intelligible by analogies. The language used by Scripture to speak of Christ’s death draws on a range of analogies. These are real likenesses, not fanciful metaphors, and each identify different aspects of what the atoning event is, while none of them identifies everything that the atoning event is. In justice to the uniqueness of God’s act one analogy is qualified by others. Valid analogies may be many, and in revealing the truth of such an event as that many are needed. The appearance of this language as conceptually rather over-furnished is only an appearance; the reconciliation of all things is not treated extravagantly by being viewed from many angles. Theological theory, which seeks to organise this multiplicity of analogies, does not have the decisiveness of either Scriptural or credal formulations, and different ways of weighing the different elements, alternative theories of equal validity can often co-exist.

The textbooks and encyclopedias are fond of saying that theories of the Atonement were a late arrival in theology, and a speciality of the West. Two twelfth-century examples provided by Anselm and Abelard, one founded on the idea of exchange, the other on the idea of Christ as archetype, were the source of all subsequent atonement theory. I would hesitate to put it quite like that, and cannot imagine how any account of Christ’s death could have proceeded without the presence in the tradition of a great essay from the fourth-century Eastern church, Athanasius’ On the Incarnation. However, the twelfth century was a moment when new questions were prompted by the rediscovery of law, especially questions about the relationship of law to morality. These elaborations of the going accounts of Christ’s death respond especially, and in very different ways, to the demand for a juridical account of it.

Out of the centuries of subsequent discussion, I would like to draw attention to one text in particular. It is the work of Hugo Grotius, On the Satisfaction of Christ, written in 1617 to refute a sceptical attack on the soteriology of the cross by the anti-trinitarian Faustus Socinus, and it is notable for the emphasis it lays on the analogy of punishment. One of the interesting aspects of Grotius’ essay is that the author introduced it by disavowing any pretension to be a theologian. To his contemporaries he was known internationally as a classical scholar of distinction, while in his native Netherlands he was known also as a controversial lawyer and politician.[1] That he was a distinguished philologist affects every page of this frankly rather dry work, which pursues Socinus exhaustively — to the reader, exhaustingly — through the Greek New Testament text by text and word by word, starting with a review of the uses of the Greek prepositions dia, huper and anti, “on account of,” “for the sake of,” and “instead of.” What makes it dry to us made it the gold standard for preachers of the next two centuries with any pretensions to scholarship. Yet the main thrust of the essay lies not in its contribution to New Testament philology, but in the legal point of view it takes: Grotius reminds theologians that when speaking of the justice of God they need to think in terms of public law, not private law. Socinus had seen the relation of the human race to God as that of a debtor to a creditor, which was why he could make no moral sense of it. A creditor who enforces his debt at all costs (including costs to the innocent) is not a morally admirable figure. But God is not a private creditor who could perfectly well write off his debt, Grotius insists, but a ruler and a judge, upholding the moral order of the universe.

In that light, Grotius understands the self-imposed moral necessity under which God acts. For if he is not a private creditor, neither is he an “absolute” ruler, who may impose his will arbitrarily. It is the covenant he has made with creation that extends the two limits of justice and mercy that determine his action. We cannot help noticing here that Grotius has something to say by implication about the justice not only of God, but of human governments. Constituted order, as opposed to arbitrariness, is essential; so is the prudent and compassionate elaboration of constituted order to meet emerging needs. Ordinary public justice knows of a “merciful judgment,” as it knows of a “remission” or “mitigation” of criminal liability. Ordinary public justice knows, too, of representative status, an exchange of positions in which one person fulfils the office of acting and suffering on behalf of many persons. This is not a fiction, but the basis of all political order. It is what constitutes public action and public events.

Public justice has as its further goal not simply the enforcement of legal liabilities but evoking and sustaining a general will for civil obedience and good order. Only God, Grotius holds, could punish without a prospective view to enabling a social life, and God has covenanted that he will not do so. This implies something he is prepared, in defiance of the going Calvinist orthodoxy of his day, to call the “conditionality” of atonement, i.e. the demand it lays on human faith in appropriating it. The point of this language, anticipating the theological path later to be followed by the Wesleys, is not to set a limit on the redemptive power of God, but to safeguard the resolution of the paradox of the cross in the restoration of human fellowship with God. Like many other Western reflections on the redemptive work of Christ, Grotius’ account focuses too sharply on the cross at the expense of the resurrection. Nevertheless, he is prepared to say that the exchange of the cross, with its “imputing” of righteousness, is not the last word in God’s dealings with man, but the next-to-last word; the last word rests with the transformation of moral life by the “imparting” of the Spirit of righteousness.

Yet if we take the decisive character of God’s act with full seriousness, the “next-to-last” is really the presenting aspect of the ultimate. There are not two redemptive acts, but one. Once again the analogy of political life comes to our aid. To be a member of a community is to be represented; it is to be subject to the representative roles in which actions and sufferings are undertaken by one “on behalf of” everyone. Yet to be represented in this way is not to be left out of the action and suffering, but to be really involved in it. The many are identified with the sufferings and actions of the one. When soldiers fall in battle, we speak, quite properly, both of their suffering on behalf of those whose safety they defend, and of the people’s suffering in the loss of their soldiers. Through their suffering, the suffering of the community becomes a real fact of history.

To achieve a view of what was accomplished by this approach to the question, we need to ask two questions of it. Why should a political analogy of God as a ruler be more illuminating of what he has done than a private-law analogy of God as a creditor? And why should punishment be taken as the paradigm case of political judgment?

The answer to the first question is that the political analogy directs us to a sphere of quite ordinary experience in which the very tensions that trouble us in relation to the paradox of the cross are typically experienced as a matter of course: tensions between the demands of justice and the demands of prudence. Neither justice nor prudence can be dispensed with in political life; it is a condition of political authority that they should be safeguarded together.

The answer to the second question is that punishment is the model of justice in its most purely retrospective aspect, at its farthest remove from prudence. There are kinds of justice that can be assimilated to prudence — “distributive” or “social” justice especially so — but there is no just punishment except in relation to something that has been done. Though human justice cannot, and divine justice will not, punish simply for punishment’s sake with no proportionate good to be achieved, punishment as such enacts the logic of what is due to past deeds. Without that retrospective reference it is not punishment at all, merely persecution.

The concept of representation belongs primarily to the prudential aspect of political authority. That selected representatives are authorised to take initiatives on behalf of the many is the way the action of a whole people takes form. By such representative action great historical innovations are launched. But the bond of political representation is not an occasional bond, assumed to get a project started and then put aside. Continuity is of its essence. And a bond formed to allow collective action is also a bond that allows collective suffering, too. The mishaps of the representative agent become the mishaps of the people, and especially those mishaps that occur precisely through the failures of the representative action.

One such failure, and perhaps the most fundamental one, is when it incurs liability for blame and censure. We are not used to associating the idea of representation with that of blame and punishment. We act together to positive ends, we assume, but incur blame separately by our individual faults. But that is an illusion, and it is one virtue of the idea of a penal substitution that it calls our bluff on that illusion. Guilt is woven into our social communications. Wrongdoing is done by community and suffered by community more fundamentally, in fact, than it is done and suffered by individuals. In all moral transactions we represent one another. Grotius believed that he could produce examples of representative punishment from the political life of his age; our own, with its fondness for holding that somebody must be to blame for every accident, certainly offers many more.

Yet since the prophet Ezekiel we have been told that each person shall suffer for his or her own sin. There is to be a perfect correspondence between judgment and deed, and between deed and agent. In practice, human justice hardly satisfies this condition, and probably never has satisfied it. We have neither the judges nor the criminals we need in order to make it a reality: judges who understand exactly what has been done, criminals who wholly express themselves in their crime. Pure retribution presupposes pure innocence, pure guilt, and pure discrimination between them, without remainder or qualification, contrary to all our experience. So elusive is the idea of pure retributive justice, in fact, that philosophers often write it off as incomprehensible.

Yet we are not in a position to set the idea aside and do without it. The expectation that each person should be judged by what he or she has done retains its hold upon our consciences. It articulates a principle to which we know ourselves bound, that suffering may be knowingly inflicted only in respect of deserts. The denial of that principle creates insurmountable offence. So we are left with the thought that the ideal of just retribution is one form in which eschatological hope is given to us. In a social world where collusion in guilt is the primary moral reality, hope could be justified only by the demonstration of an innocence that stands outside that circle of collusion and of a retribution adequate to condemn it. Faith in the event of pure innocence and pure retribution is the act that anticipates the era of which Ezekiel spoke, an act in which we are at last set free to answer for ourselves.

What the analogy of public justice contributes to our thought about Christ’s death, then, is to make us aware of a wider range of moral responses than we might at first have suspected we were capable of. Not only are we encouraged to let our experience of public justice shed light on God’s justice, but we are encouraged to let our bewilderment at the paradoxes of public justice shed light on the bewilderment we experience at God’s action. Wider and more self-questioning moral judgments then call our immediate and intuitive ones to account. That alone, even if there were nothing more to say, would be a recommendation for this whole train of thought as an exercise of moral education. It is as though we are taken again through the logic of the parable of the unforgiving servant. Faced initially with the abstract question of an infinitely wealthy sovereign confronted by a debtor who cannot pay, we know the answer at once: the infinitely wealthy sovereign must be infinitely generous and forgiving. But when that situation is extended in time, and we are faced with the concrete problem of a forgiven debtor who does not know how to forgive, we are forced, like the sovereign of the story, to complicate the situation with an act of condemnation.

The penal-substitution theory of the death of Christ is itself a kind of parable, though it is a parable about a real historical action which intends to make us think about its significance in real world-history. It points us to an act that reconciles and fulfils all history, and asks us about the moral conditions on which such an act is conceivable, probing our readiness to understand history as a moral history, a story of the overcoming of good and evil by a final good. It presents us with ourselves in a state of irresolvable tension, constituted by a past of performance and a future of aspiration that are quite incompatible; each of which is, nevertheless, wholly and essentially ourselves. We are asked how we may establish any moral coherence between these two aspects of our existence.

In the end, such a reconciliation does not lie within our powers, but must be effected by one who purposes and directs history, and yet, since it is precisely our moral history that is at stake, it must be possible for us to participate in it willingly and from the heart. It must be accomplished at the heart of history once and for all, but in a form open to imitation and appropriation in every period of history. It must be accomplished through an act in which evil is truthfully and unsparingly judged for what it is, and good is upheld in the face of it. And to help us discern and understand such a unique act, we are told of a punishment willingly accepted on our behalf by an innocent representative. Such a conception inevitably conveys less than the whole truth of that act, but not something other than the truth. It displays together and in one view the scope of evil and the power of generosity in the bonded, social reality of human existence where we constantly represent, and are represented by, each other.

It is a striking feature of the times we are living through that we do not know how to deal with the evils of history, and that we struggle, through frankly symbolic gestures such as demolishing statues, impotently to bring the past under the control of a justice that will somehow set the wrongs of the past right. If any times ought to understand intuitively the moral need that a representative punishment addresses, surely they are ours!

Enjoy the article? Pay the writer.

Personal Info

Donation Total: $0


  1. His international reputation came to rest later on his masterpiece of international law, The Right of War and Peace. A major edition of The Satisfaction of Christ by Edwin Rabbie, with Latin text and English translation, appeared in 1990 as the first volume in a series of Grotius’ Opera Theologica projected by the Royal Netherlands Academy (Assen & Maastricht, Van Gorcum, 1990). For an overview of the man and his work see The Cambridge Companion to Hugo Grotius, edited by Randall Lesaffer and Janne Nijman and due to appear imminently, to which I have contributed a chapter on “The Theological Works”. Grotius’s services to theology were, in fact, considerable. He was the author of the first all-Bible commentary conceived as a philological and historical commentary for the use of preachers. His apologetic work on The Truth of the Christian Religion was no less influential than The Satisfaction of Christ.