Moral Faith and Atonement

Yesterday I talked to you about why the moral agent needs to have moral faith. I distinguished two types of moral faith, faith in the possibility of virtue, and faith in Providence. Yesterday was mostly about faith in Providence, especially what I called the strong belief that the actual world is so ordered that a person's own virtue is consistent with her own happiness, whether other people are virtuous or not. Today I am going to talk about faith in the possibility of virtue. This faith is required, I suggested yesterday, because of the structure of the moral gap. In this structure there is a gap between the moral demand and our initial capacities, which are not adequate to the demand. What is necessary is some kind of transformation of these capacities. Kant talks about a 'revolution of the will', which itself cannot be accomplished by humans on their own but needs God's assistance. I am going to talk today about some Christian doctrines which are tied to the picture of the moral gap; they suppose that the third part of the moral gap structure, namely the at least possible holy being, intervenes so as to change human moral capacity so that it becomes adequate to the demand. God, in other words, calls us to obey his law, by which we stand condemned, and then enables us to live by this law.

The talk today is about how this assistance by God is supposed to work. I realize that I am talking about a mystery of the faith, and that it is presumptuous to think we will reach perfect understanding here on earth, let alone in the next fifty minutes. But I do think I have made some progress in understanding this mystery, and I want to try to share this with you.

I will start with the fact that the atonement has dropped out of the contemporary professional literature on ethical theory. Part of the reason is the general decline of Christianity amongst the academic elite over the last hundred and fifty years, and part of it is unique to the doctrine of the atonement itself, which has seemed particularly scandalous. It is true that especially in America, there have been signs of the beginning of a turn-around in the academy, at least within the discipline of philosophy. It is possible now to hear good papers about philosophy and Christianity at most meetings of the American Philosophical Association. But my own view of this is not altogether encouraging. I think the old-fashioned contempt for Christianity was based on a sense that convincing reasons could be given for rejecting the traditional faith. My father, for example, said twenty five years ago (1970) that it was impossible to hold together three things; belief in a supernatural God, science and philosophy (or logic). Any two of the three can be held, he said, and have been. But not all three together, at least not by a reasonable person. Traditional Christianity was seen as an important enemy that had been defeated. Over the last twenty five years has come a gradual decline in the prestige of science and in the sense that compelling reasons for rejecting Christianity can be given, that any reasonable person ought to accept. But what has taken the place of the enmity is not friendship but a kind of amused tolerance, which in a way takes Christianity less seriously than before. Bits and pieces of Christian doctrine creep into the professional literature, for metaphor or adornment; but except in isolated cases I have not seen any significant movement towards intellectual repentance or conversion to the faith as a whole system of doctrine and practice. It is as if Christianity were a great aunt that you did not like very much, but after she has died and a decent interval has passed, you can start dressing up for fun in some of her clothes.

These are very general remarks, and my main purpose is to focus on the reasons for rejecting the doctrine of the atonement in particular, rather than traditional Christianity in general. I am going to try to defend the doctrine against one main objection to it. It is an objection to the doctrine interpreted in one particular way, namely the atonement interpreted as penal substitution. To see Christ's death in terms of penal substitution is to see it as a punishment undertaken by Christ on our behalf, as our substitute. Kant had an objection. He thought that the traditional doctrine of the atonement, understood in terms of penal substitution, could not be held by a responsible thinker within the limits of reason alone. Again, this is not a talk about Kant. But I will be retracing his steps, and this time I want to disagree with him fundamentally. Kant supposed that morality does not allow transferred liability for wrong-doing. Guilt is not, he said, like a financial indebtedness, which can be transferred from one person to another. With a financial debt, we can see how one person could take over the debt from another. But if I have been guilty of some wrong-doing, then it is my responsibility to make up for it; and this responsibility he thought could not be transferred to another person. In the traditional doctrine of penal substitution, however, it is held that Christ takes over our sins, and pays the penalty for them. This doctrine seemed to Kant to be at odds with morality, as he understood it. That is the central Kantian objection. There have been other moral objections to the atonement, and I have put some of them on the handout. The utilitarians, for example my father, hold that the retributive notion of punishment has to be rejected on moral grounds. To return harm for harm done is, they think, merely to increase the amount of suffering in the world, not to set anything right. Since the traditional doctrine of the atonement relies on the notion that Christ paid a death that we deserved, they reject the atonement as well. There is also a moral objection from recent feminist ethics, to the idea that a caring God would perform a child sacrifice, allowing his own son to die a brutal death in order to satisfy the demands of abstract justice. It is the Kantian objection that will be the focus of this talk. But you will see that there are resources in what I say for beginning a reply to these other two objections as well.

I want to start by tying atonement to the notion of forgiveness, because this will help explain what version of penal substitution I am trying to defend. How does forgiveness work? Suppose we think of two persons, a victim and an offender. The victim forgives the offender, we can say, when she stops holding the offence against him. But what does this mean? In my view, following Richard Swinburne on this (though I have some objections to his view of forgiveness as well), the victim can expect the offender to perform various tasks, such as repentance, apology, and reparation. When she forgives him, she accepts these tasks as completed. But, it has been objected, surely to forgive is to let the offender off at least some of these tasks. The prodigal son returns from abroad, and his father rushes out to meet him on the way. The father forgives, even though his son is not able to make any reparation for the harm he has done by wasting his share of the estate. But I think we have to be careful here, and it is easy to make a mistake. Forgiving a debt is indeed to release the debtor from payment. If the debtor has paid the debt, then forgiveness cannot take place; there is nothing any longer to forgive. But forgiving a person is not like forgiving a debt in this way. When we are forgiving a person, accepting that the tasks have been performed can just be forgiving the person, or ceasing to hold his offence against him. Why does it matter to the doctrine of the atonement whether or not we make this distinction between forgiving debts and forgiving persons? It matters because the failure to make the distinction has given rise to an objection to penal substitution, which was made, for example, by Faustus Socinus in the sixteenth centruy, but has often been repeated. The objection goes like this. God forgives us, and forgiveness means, according to the objection, that the person offended lets the offender off at least some of his tasks. The atonement, then, could not be understood as Christ completing the offender's tasks on our behalf. For if Christ did so, there would be nothing left for God to forgive. But I do not think this objection is true to the nature of forgiveness between persons. Sometimes, forgiving means that the person offended accepts the offender's tasks as completed.

What counts as completing the tasks? When, for example, has the offender made enough reparation? In outline, the tasks would be completed if the offender could make things as good for the victim as they would have been if the offence had not been committed. Sometimes, in fact often, the offender will not be able to do enough to bring this about. For rape, how could there be enough reparation? What happens with forgiveness when this is so? In particular, can the victim simply forgive the offender without reparation? If so, would not this be admirable, and a mark of magnanimity? If it is admirable, then here is another objection to penal substitution. It is number five on your sheet. This forgiveness without requiring reparation would be the mark of a good person. And if so, it would be a defect in God to require penal substitution, and thus the full performance of the offender's tasks.

I want to suggest a partial answer to this very difficult question. The answer will come from reflection about the justification of punishment. Why do we punish people? There have been several answers in the tradition of reflection about this. We punish in order to deter, or in order to rehabilitate, or in order to prevent the offender from further offenses by putting him in custody. These three answers all look for the justification of punishment towards the future. The offender (or others like the offender) are to be deterred in the future. Or offenders are to be changed into better people by putting them in a healthful environment, like a prison, in which they can be morally trained. Or they are to be prevented physically from future offences, by being put behind bars. There is, however, another rationale, different from any of these, namely that we punish because the offender deserves it. This justification looks not to the future but to the past. An eye for an eye sets the balance of justice right. It remedies the past imbalance created by the offence. A utilitarian, as I said before, will usually find this retributive theory of punishment morally objectionable. She will evaluate actions morally by looking at their future consequences, and by trying to maximize happiness impartially. To her, retributive punishment (the eye for an eye) merely adds to the sum of misery in the world, and is not in itself good. But there is a variant of the retributive theory which meets the utilitarian's central point. I will call this variant 'the expressive theory of punishment', taking the term from Jean Hampton, who died this year, alas, before she could finish her project on the foundations of morals.

Consider the bully on the school playground, tormenting his victim. He is giving outward expression here to a low view of the victim, putting him at the bottom of the playground hierarchy. The playground, especially if it is relatively unsupervised, is a good place to observe what I said yesterday. People who know each other well, and know how much what they do hurts, can go ahead despite this knowledge with relish. They establish a pecking order, like chickens in a coop, and the people at the bottom get wounds from which they sometimes never recover. But the victim's moral value as a human being with dignity is higher than this, and is in fact equal to the bully's. The function of punishing the bully is then to give an outward expression to the lowering of the bully from his assumed status, 'putting him', as we say 'in his place', and raising the demeaned value of his victim. The suffering of the bully is not something good in itself; here the utilitarians are right. But it is good because it expresses the right relative value of himself and his victim. On the expressive theory of punishment, punishment is good because it takes the victim seriously.

I do not claim that I have given here more than a suggestion of a theory of punishment. But it enables us to see what might be good about God's accepting reparation. For we most often wrong God by wronging each other, not by wronging God directly (though we can do that too). The punishment of the offender takes seriously the value of the victim, and this is not given expression if the offender is let off. If it is the victim who lets the offender off, this might express his own value (if the situation is not inherently coercive). But on the expressive theory of punishment, God's punishing us expresses not so much his own value as the value of the people we have wronged. Or we could put this distinction in a different way. The punishment expresses God's value in his image-bearers who have been abused, rather than his value in his own person. To go back to the example of the bully. If the victim says to the school that he does not want the bully punished, this might be an expression of his magnanimity and a restoration of his value; though he might also be afraid of the bully's implied threat in the future, and just be continuing his role as victim in an inherently coercive situation. But if the school on its own account lets the bully off his punishment, it fails to acknowledge the value of the bully's victim.

It might be objected that this expressive account of punishment cannot fit the atonement, because Christ agreed to die for us voluntarily. Surely the bully on the playground is not put back in his proper place, if his punishment is contingent upon his voluntary agreement to it; that is, if he will not be punished unless he agrees to it first. Punishment, if it is to have this expressive function, needs to be forcible. But I think this is not quite right. Think of the example of Moses, who offered to take death on behalf of the Israelites, when they had worshipped the golden calf. It was their punishment that he offered to take on voluntarily, though God did not allow the transfer in this case. There is nothing here that is inconsistent with the nature of punishment. Once the originally innocent party has agreed to be punished on behalf of the originally guilty, the punishment can be exacted in the normal way. This is true also with the human punishment which Christ suffered under Pontius Pilate. He could have prevented this punishment taking place at any time, with legions of angels or his own divine power. But he undertook voluntarily to be constrained, and to come under human power. And so the nature of the punishment as exacted was preserved.

I want to go back now to the central Kantian objection I mentioned before. This is the objection to transferred liability. It is the first objection on your sheet. What sense can we make of the notion that Christ took on our sins? Is there, as Kant thought, something about the idea of this transfer that is repugnant to morality? The topic here is the 'substitution' part of penal substitution, rather than the 'penal' part which I was discussing just now. I will try to respond to the Kantian difficulty by talking about what I will call 'a partial merger of identity', which I think is a familiar part of our experience. Indeed these partial mergers of identity account for a large part of what makes a human life a good life. My thinking about this has been influenced especially by recent writing in feminist ethics, which has objected to the picture of moral praise and blame and moral responsibility as strictly individual. My strategy will be to start with inter-human experience, and then move to the notion that Christians are 'in Christ'. The idea I shall try to work out is that our union with Christ effects one of these partial mergers, and that in the context of these partial mergers the transfer of liability makes sense.

One familiar part of human experience is the union between mother and infant child. Suppose I offer to hold a baby, and its mother hands him over. I take him on my shoulder and make a few soothing noises, and the child is then sick all over my shirt. What does the mother do? She apologizes, takes her baby back, tries to wipe off the mess, and offers to wash the shirt. She behaves, in other words, just as if she had been sick all over my shirt herself. She performs the offender's tasks of repentance, apology and reparation. This is because of a kind of union between her and her child. This was even closer in the womb, when the foetus was literally 'in' its mother, and took its nourishment through the umbilical chord. But if she is nursing, there is still the physical link between what she consumes and what he gets to drink. Think of the times when you invite the mother to some gathering, and you assume that it is mother-plus-baby that you are inviting. The two of them form a unit; they come and go together. It is not just mothers, but what Virginia Held calls 'mothering persons' who can form units of this kind. I remember carrying my infant child in a pouch tied around my stomach, and feeling as close to pregnant as male nature allows. Was I proud of my baby? Or was I proud of myself for having this baby? Or is this a bad distinction? What I want to suggest is that the mothering person and child form a kind of unit for evaluative purposes, and the sense of pride or (when the child spits up) the sense of shame carries across or transfers so that one can be proud or ashamed of what the other is or does.

How does the Kantian objection look in this context? Is there something about this transfer that is repugnant to morality? I think the Kantian will be indignant at this question. This is not the sort of thing he was talking about. He was talking about adults, not babies. And he was talking about guilt, not shame. I want to take these responses one at a time. First, it is just the feminist point that our account of morality has ignored the mothering relation, as though what is morally important is confined to the relations between adults. But even if we do put the mothering relation aside, as somehow a special case or not properly paradigmatic, we are still faced with many other partial mergers of identity in which this kind of transfer takes place. Consider, first, the case of an adolescent with a younger sibling. The adolescent has a friend home for a meal, and his younger sister does something unspeakable. Perhaps, to keep on with the sequence of liquid examples I have been using, she is sitting next to the friend, and she clumsily knocks over her mug of soup, so that it drenches the front of his pants. Her brother is mortified. He feels he has to apologize, and he offers a pair of his own trousers as a temporary substitute. He may not offer to clean the drenched ones, but after all he is a teenager. Part of what is going on in this example, I think, is that the brother feels part of a unit, the family, which has together violated the duties of host to guest. This is why he is ashamed of what his sister has done, and why he feels he has to make it good, as far as he can and within his conception of his limits.

Now consider marriage. How is it that I can be ashamed or proud of what my wife does, and she can be ashamed or proud of what I do? When she does well at an interview and gets the job she wants, I am pleased and proud for her. But the truth is also expressed by saying that she and I together form a unit, and I am pleased and proud for the both of us. It is true that I may also feel envy and resentment, and gird myself for new battles about the division of responsibilities. But if the marriage is functioning well, you might see a new spring in my stride, a greater sense of self-confidence, that would be altogether mysterious if we could not understand the transfer or transmission of pride within the unit. Similarly if I fail in some disastrous way at work, especially if it seems that the whole world knows about it, my wife and I will both be dragged down by it. It is not just that she feels sympathy for me in my predicament. It is that she has been humiliated together with me, and we will both feel the urge to hide from public view.

This kind of example can be extended, we can now see, into relations of close friendship, and beyond that into relations of collegiality at work, and so on. I will not labour the point. I think there is a general principle. There is a kind of partial merger of identity which comes in degrees, and the limits of the transfer of moral evaluation depend on the extent of the merger. It is tempting to think, like the Stoics, of what they called appropriation (in Greek, oikeiosis, which means making something my own). Hierocles the Stoic, I quote, 'drew a picture of a man at the centre of a number of concentric circles. In the innermost he stands himself, with his body, and the satisfaction of his physical needs, in the next are his parents, brothers, wife and children, then more distant relatives, then members of his deme (ward or village), of his city, of neighbouring cities, of his country, of the human race. Hierocles suggests that we should try to contract the circles, treating e.g. uncles like parents: the ultimate aim would be to treat all men as our brothers.' In this picture the degrees of merger go out from the centre to encompass ideally the whole of humanity. But there are several things wrong with this picture. First, it is almost all about males. Second it starts from the centre, as though the boundaries of the self were firmly established in opposition or distinction from anyone else. The feminist literature I have been reading disputes this both as a chronological account of development and as a normative account of the proper structure of moral life. Third, the proper ordering of moral relations does not end with collapsing special relations into the love of humanity in general. A proper place of preference for special relations within limits has to be preserved. But what is right about the Stoic picture is the idea of a continuum of relations within which identity can be merged, so that the other can be seen, in Aristotle's term, as 'another self'. When this merger occurs, I can, to different degrees, identify with the prescriptions of other people (my children, for example), so that the moral evaluation of myself extends to different degrees outside my own skin. The 'I' in moral evaluation becomes elastic.

John Donne has a typically metaphysical way of putting this phenomenon, and I have put the quotation on the handout. 'When love with one another so/ interinanimates two souls,/ that abler soul which thence doth flow/ defects of loneliness controls.' He has invented here a new latinate word, namely 'interinanimate'. His idea is that a soul (anima) is inserted into (in) the space between (inter) two people. Inter-in-anima-tion. The love between them creates a new life shared between them. This new life, which both parties live, has the power (which neither individual life had) to control the isolation of the two parties on their own.

I want to return briefly to the distinction between shame and guilt. Often the distinction between so-called 'shame cultures' and 'guilt cultures' is drawn in such a way as to suggest that the shame cultures are primitive, that we have grown out of them, and that the Western world has matured into replacing shame with guilt. It does indeed seem that whereas all cultures have something like shame, guilt is a specialized concept and is culturally specific. Homeric society is often given as an example of a shame culture, which has no recognizable concept of guilt. What is the difference supposed to be? To complicate the question, we need to distinguish also between what I will call 'objective' and 'subjective' guilt, and 'objective' and 'subjective' shame. Subjective guilt is when I feel that I have done something morally wrong, whether or not I have in fact done so. Objective guilt is when I have done something morally wrong, whether or not I feel that I have done so. Similarly subjective shame is when I feel that what I am or have done is defective by public standards. Objective shame is when it is defective, whether I feel it to be so or not. This is not as familiar to us as it was for example to the ancient Greeks, but the concept of disgrace is more familiar in its objective sense. 'When in disgrace with fortune and men's eyes', says Shakespeare, 'I all alone beweep my outcast state.' But I can be in disgrace whether I acknowledge that or not. One thing that distinguishes guilt from shame, on most accounts, is that they can have different objects. Subjective and objective guilt are both for something I have done. Shame may be for something I have done, but it may also be for what I am, for example if I am naked in public or physically deformed from birth. Suppose I have been sunbathing naked in what I thought was a private place, and I am suddenly surrounded by a group of curious tourists with cameras. I may or may not feel guilty for having sunbathed naked. But that is something different from the sense of shame in my nakedness. I can be ashamed of something, whether or not it was the product of my deliberation and rational choice. In this way, shame is more inclusive than guilt. A second difference between shame and guilt lies in their circumstances. Shame seems to carry with it the sense of being public, or at least being in front of or being looked at by some person or group of persons. In the example I gave earlier about my disaster at work, my wife and I will be ashamed especially if the whole world seems to know about it. In this way guilt is more inclusive than shame, since it does not necessarily carry with it the sense of an audience or spectator.

If we confine moral condemnation to cases of guilt, we will be impoverishing our moral lives. I do not want to deny that guilt is important, but we have given it too central a place. There is a historical irony here. Authors like Bernard Williams who follow Nietzshe in rejecting guilt in favour of shame think of themselves as going back to recover the ancient sense of nobility and tragedy in human life, before its suppression by Christianity. Thus Nietzsche says 'The advent of the Christian God, as the maximum god attained so far, was therefore accompanied by the maximum feeling of guilty indebtedness on earth. Presuming we have gradually entered upon the reverse course, there is no small probability that with the irresistible decline of faith in the Christian God there is now also a considerable decline in mankind's feeling of guilt.' But if we think about the Christian idea of what we have to be saved from, the two differences I just mentioned between guilt and shame give a better fit with shame. I need to be saved not only from the sins that I have deliberated about and rationally chosen, but from my sin, that I was born with. Even Kant talks about the radical evil in us, which is, he says, both innate and imputable. I need to be sorry, before God, for what I am, not merely for what I have chosen to do. Indeed what I am is the source or ground of what I have chosen to do. Sometimes I can be struck by a sense of acute embarrassment about the sort of person I am; and I think I should feel that way at least some of the time, because it is an acknoweldgement of the truth about myself. Secondly, the Christian sense of sin is always sin before God, in Latin coram Deo, (a phrase which is cognate with os, or face, before the face of God). Kierkegaard makes this the transition from guilt to sin, that sin is before God, or before his face. The idea of sin carries with it the idea of God's look. My claim, then, is that atonement and the transfer of liability belong with shame as well as with guilt. There may be an account of atonement that can manage with guilt alone, but it will not be the full account. How atonement does work with guilt is a question for another talk. My point against Williams is just that he is trying to recover a conceptual resource that Christianity has never lost.

The notion of being 'in Christ' can be understood as a partial merger of identity, or, in Donne's term, as 'interinanimation'. This helps us understand how Christ's death could be substitutionary. The New Testament is full of images for our union with Christ. We are members of his body, the church is his bride, we are grafted into the vine. Jesus compares his relation to us to that between a mother hen and her chicks. He calls us his friends. We are adopted into his family. I will call what I think is central to these images the idea of 'incorporation'. The idea of incorporation is a significant help in understanding the atonement. We can see that many of these images allow extension into cases of transferred liability or transferred imputation. For example, a child is adopted into a family and has some defect of character. Suppose he tends to steal things from shops. By adoption, this becomes a defect of the family, and his parents and siblings become ashamed of it. This is no small matter. In some cases the pain can be severe. But there is also possible a transfer in the reverse direction. The new member of the family can start to live the kind of life that is characteristic of the family. In Donne's terms, there is an 'abler soul' which is now lived between the members of the family, and this soul 'controls the defects' of the new member on his own. The child starts to take on the moral coloration and the pattenrs of the unit to which he now belongs. I will not pause to do this, but similar two-way transfers can occur in the context of many of these biblical images of incorporation. The idea is this: Christ chooses to incorporate us, and in virtue of this incorporation our sin can be shared with him and the quality of his life, the new life, can be shared back with us. The incorporation is not itself the atonement. It belongs with the doctrine of justification, which I would like to discuss further but cannot at this point. But the incorporation, I am suggesting, provides the proper background for understanding the atonement.

What does incorporation have to do with Christ's death? I think an account of the atonement needs to articulate both what it means to share in his death, so that in his death we die, and also what it means that he died instead of us. These are two features of one of the most puzzling passages about the atonement, 2 Cor. 5: 14, which I have put on the handout. 'For the love of Christ constraineth us; because we thus judge, that if one died for all, then were all dead.' What I am concerned about is the 'then'. Paul seems to be saying, first, that Christ died for us, and that therefore, second, we died as well. What sense can be made of this? Some interpreters try to deny that 'for us' means 'instead of us'. I am not a biblical scholar, but I think there is strong biblical and extra-biblical evidence that they are wrong. In any case, we can make sense of this inference, this 'therefore'. Consider the context of representation. If I appoint you as my representative in some negotiation, two things become true. First, you act instead of me. Second, when you act, I act as well. I act through you. To put this in terms of the relevant inference, you act instead of me, and therefore I act as well. In general, the contexts in which this kind of inference makes sense are the contexts in which there has been a partial merger of identity, or what I called 'incorporation'. In an ancient or medieval battle, I am incorporated into a national unit, and we may have a champion who fights for us, say Goliath. He fights for us, or instead of us, but therefore when he is defeated, we are all defeated and find ourselves fleeing for our lives. This double truth, containing both halves of the inference, is what we should mean by 'substitution' in the phrase 'penal substitution'. Because Christ has chosen to incorporate us, two things are true about his death. First, he dies instead of us, and therefore, second, when he dies, we die with him.

There are, it is true, differences between our incorporation into Christ and other partial mergers of identity. One major difference is that there is no other relation in which we share in the death of somebody else two thousand years ago. But I think we can understand the notion of retroactive incorporation, even though there is no merely human parallel to the extent of the retroactivity in the atonement. Consider marriage, in which I am incorporated into my wife's family and this includes her family's past. Sensitive areas for her, areas of past hurt, become sensitive no-go areas for me as well. There are certain topics we just do not talk about when we visit her parents. Or consider those educational institutions that require a new member to subscribe to obscure sixteenth and seventeenth century documents as a condition of lonlg-term employment. Why do they do this? They are trying to bring it about as far as possible that the events recorded in these documents become events in his past as well. Surprisingly, this can work. He can become part of a community which is together proud of some of these events and ashamed of others. How can he be proud or ashamed of the past of another cultural grouping quite different than his own original one? By retroactive incorporation. This is an idea which makes sense within our human experience. But our union into Christ's past is also different from any merely human experience. His death two thousand years ago was followed, in Christian doctrine, by his rising and then by his presence with us here on earth through the Holy Spirit. I have a life in union with him which I cannot have with anyone else who has died, because I cannot have a life by interinanimation, a life between him and me to which we both currently contribute. Perhaps the closest parallel is the kind of ancestor worship which recognizes the spirits of my ancestors, even quite remote ancestors, as sticking around to help me after they have died.

To conclude, I am not claiming to have found human analogies for the doctrine of the atonement in all its parts. It remains a mystery of the faith. There are features of the doctrine which my picture of incorporation does not help with, and may even make harder to understand. For example, return to the picture of the mother with her infant child. Does not this picture give comfort to the enemies of Christianity. It is, after all, one of the standard objections to Christianity that it is an infantile religion, one that reduces us to babies in our relation to God. What is left, on my picture, of the call to repent and believe, and the chosen response of faith? I am not going to try to solve this problem. But it is helpful to return to the context in which Jesus uses the analogy of the mother hen and her chickens. He is grieving over Jerusalem, and he says 'How often would I have gathered thy children together, as a hen doth gather her brood under her wings, and ye would not!' The relation of incorporation, in which we are dependent on him, is represented as something he longs for, but not something he forces upon us. We have to choose to be dependent, to be like children. In my particular subculture within Christianity, I was taught that this choice is itself a gift; but it is another long question how that could make sense. Here we are in the difficulty about the extent of the atonement, and how God could, consistently with his love, choose to incorporate some and not others.

Even if we could understand this, there are other grave difficulties. I have concentrated on the link between incorporation and the imputation of sin. But that leaves the question of how God could incorporate human beings into a life together with God. And it leaves the question of how the sin, taken on by Christ, could then be defeated or overcome by him. I think we have to be content with little bits of understanding God, as though we were gathering up the crumbs under his table. For me, this picture of incorporation and its relation to the atonement is one such crumb that has nourished me, even though I know it is not the whole loaf. It helps me to see how Kant's central objection to the atonement fails. This is important in itself. But more than this, it helps me see better how I am reIated to Christ and his work on my behalf. I offer it to you in the hope that it can be helpful for you also.