The Problems of Penal Substitution
For Christians, the atonement of Jesus Christ is indisputably the central event of salvation history. However, there is no clear consensus as to its precise nature or rationale. Over the centuries a number of atonement theories have been formulated, with a theory known as penal substitution predominating. In this paper, I will explore the penal substitution theory. My task will be primarily negative–I will argue that penal substitution is an enormously problematic theory of the atonement.
Penal Substitution has antecedents in the 12th century satisfaction theory of St. Anselm, but did not crystallize until the Reformation. According to the satisfaction theory, God’s honor is vitiated by human sin and can only be restored by some form of remuneration. Penal substitution substitutes justice for honor. There is an eternal law of justice that God cannot abrogate, either because justice is an essential part of God’s nature or, in some Mormon adaptations of the theory, because there is a metaphysical principle of justice in the universe which God is duty-bound to uphold. God cannot simply forgive us when we sin, for justice demands punishment. However, because God is also loving and merciful, He accepts Christ as a substitute for sinful humanity. Repentant human beings are empowered to escape punishment for their sins because Christ receives the punishment in their place, thereby satisfying the demands of justice.
An implicit penal substitutionary understanding of the atonement, likely unconsciously imbibed from mainstream Christianity, is common among Mormons. This is perhaps unsurprising given the relative youth of our religious tradition and the evocative power of some of the surface features of a penal substitutionary framework. Viewed from a certain angle, the penal substitutionary narrative can be profoundly moving, even transformative; the idea of a morally perfect individual willingly submitting to excruciating pain on our behalf is unquestionably awe inspiring. Even so, I believe penal substitution to be a deeply flawed theory. God is no doubt more concerned with the condition of our hearts than the acuity of our theology–far be it from me to decree that God can only speak to those in possession of a sufficiently analytical atonement theory. But this needn’t entail that reason has no part to play in our wrestle with God. In my view, careful analysis shows the penal substitution theory to be both morally and logically deficient.
My primary objection to penal substitution is the injustice objection. This objection is grounded in the self-evident principle that it is wrong to punish the innocent. Following Dennis Potter, I will call this the innocence principle (73). The very foundations of our judicial system rest on such a principle. It is manifested in our visceral reaction whenever we learn of persons who have been wrongly convicted of crimes–we feel in our bones that there is something seriously wrong with punishing the innocent in the place of the guilty. The problem for penal substitution should be obvious: penal substitution runs afoul of the innocence principle. On the penal substitutionary view, we who deserve to be punished escape punishment while the only person in the history of the world who emphatically does not deserve punishment is punished in our place. The deep irony at the heart of penal substitution is that a theory that begins by valorizing God’s justice ends up subverting it.
One might respond that it is not unjust to punish Christ in our place because he willingly takes our punishment upon himself. However, if we examine our intuitions I believe that we will see that the fact that the innocent third party’s consent has been secured does not alter the essential injustice of punishing the innocent in place of the guilty. Consider the OJ Simpson murder trial. Imagine a counterfactual in which the jury recognizes that the evidence against OJ is overwhelming and finds him guilty of murder. He is sentenced to life in prison without the possibility of parole. OJ’s devoted fans are devastated by this outcome. The president of OJ’s fan club tells Judge Ito that she would like to be sentenced in OJ’s place. Judge Ito reasons that, so long as some punitive response to OJ’s actions is inflicted, it doesn’t really matter whether it falls on OJ himself or a consenting third party. So Judge Ito allows the president of OJ’s fan club to take his place. What would we think of this? It seems to me that stipulating that the penal substitute is a volunteer does nothing to dislodge the innocence principle.
Indeed, one could argue that inflicting punishment on the innocent is not merely unjust, but conceptually incoherent. What is punishment? Primarily, a particular act whereby one inflicts some harm or deprivation on an offender in response to her offense. Following Brent G. Kyle, we can imagine a father who comes home from a bad day at work and vents his anger and frustration on his child by spanking her, even though she has done nothing wrong (208). Clearly, the father inflicts a form of punishment that harms his child, but it seems wrong to say that he has punished her. It seems reasonable to think that for an act to count as an instance of punishment, the punisher must believe that the recipient is responsible for an offense of some kind. But God, being omniscient, could not have erroneously believed that Christ was responsible for human sin. So the notion that God punished the innocent Christ in the place of sinners may not even be logically coherent.
Some have tried to forestall the injustice objection by arguing that Christ took upon himself not merely our punishment but our guilt as well. The innocence principle was not violated because an innocent person was not in fact punished. However, this maneuver has bizarre implications. Accepting it commits us to the position that it is possible for a person to be guilty without actually having done anything wrong. Yet a guilty person, by definition, is someone who did something wrong. If Judge Ito had told the fan club president that it would be unjust for him to punish an innocent person in OJ’s place, but he could get around this constraint by transferring OJ’s guilt to her prior to imprisoning her, he would have simply compounded the absurdity of the situation.
It should be easy for Mormons to detect the sleight of hand at work here. Mormons rightly reject the traditional Christian understanding of original sin because it fundamentally misunderstands the nature of guilt (cf. Article of Faith #3). The same kind of semantic distortion is inherent in any attempt to impute our guilt to Christ. Moral guilt just isn’t the kind of thing that can be passed from one person to another, like currency or colorblindness. Ergo, this claim is conceptually muddled, false by the very meaning of the terms employed.
Moreover, according to the penal substitution theory, Christ’s innocence was a necessary condition for his being eligible to perform the atonement. It follows that Christ’s becoming guilty so that He could be justly punished for our sins would have rendered him incapable of carrying out an atonement. Thus, even if it were logically possible for someone who has done nothing wrong to somehow acquire the guilt of others, Christ could not have availed himself of this loophole without losing his capacity to atone.
Perhaps the atonement should be understood as analogous to a third party’s stepping in to pay a fine on one’s behalf. There is clearly an injustice done if OJ murders two people and we send his fan club president to prison in his place; however, no injustice is done if OJ, having squandered his fortune, incurs an exorbitant parking ticket and his fan club raises money to pay the fine on his behalf. I believe that we can make sense of our differing intuitions with respect to these two scenarios if we draw a distinction between actions that are legally prohibited but not morally wrong and those that are legally prohibited at least in part because they are morally wrong (Ostler 271-272). When an action is contrary to a statute, the state may affix a penalty to such acts both to deter them and as a means of raising money. But crimes like fraud, abuse, rape, and murder are not merely illegal, but morally wrong as well. We respond to them by imposing punishment.
There is an element of censure and reprobation built into the concept of punishment, an expressive function, that is not present in the concept of a penalty (Feinberg 400). Punishment consists of some form of hard treatment or deprivation understood as, among other things, the instantiation of the community’s moral disapprobation of the criminal’s act. Punishment is thus, by its very nature, not the sort of thing that admits of substitution. If we think of sins as moral wrongs analogous to crimes–and what is a sin if not a moral wrong?–we should reject a penal substitutionary understanding of the atonement.
Now, there may be cases in which a fine is intended not only to penalize but to punish. Nonetheless, if the offender in one of these cases is fortunate enough to have an affluent benefactor willing to come to his assistance, the state would typically not step in to prevent this third-party from paying the fine. Might the innocence principle be less absolute than it first appeared? I don’t think so. It seems to me that the fact that we permit punitive fines to be paid by third parties is less a function of what we think is just than it is a function of expediency. It would be overly burdensome and intrusive for the state to try to prevent third-party intervention in such cases. Consequently, we allow considerations of expediency to take precedence over considerations of justice.
Invoking this aspect of our criminal justice system is of little help to the penal substitution theory for two reasons. The first is that our tolerance for these deviations from justice is strictly circumscribed–if our hypothetical offender, emboldened by his ability to avoid the financial consequences of his actions, comes to believe that he can commit a serious crime and escape punishment by sending his indulgent patron to prison in his place, he is in for a rude awakening. The second is that one of the distinguishing characteristics of the penal substitution theory is its inflexible conception of justice. The penal substitutionist believes that justice must receive its due, so it should be immaterial to him that our criminal justice system is willing to forego justice in the interest of expediency in certain limited circumstances.
The injustice objection is sufficient by itself to discredit penal substitution, but it is not the only objection that can be raised. I believe that the assumption that a punitive response to wrongdoing is mandatory, that the universe will be out of joint unless every offense is counterbalanced by the requisite quantity of pain, is misplaced. Contra the penal substitutionist, I submit that there is no reason to think that it is impossible to forgive without punishment. It is wrong to punish the innocent; it is not necessarily wrong to refrain from punishing the guilty. I call this the forgiveness objection.
The forgiveness objection is informed by the observation that humans are capable of forgiving others without requiring that punishment first occur. Suppose my son becomes angry with me and attacks me, causing me severe pain and injury. I have the right to bring an action for assault and battery against him. But what if he comes to me and sincerely apologizes and asks for my forgiveness? It doesn’t seem like it would be morally wrong for me to forgive without pressing charges. In fact, assuming that my son’s repentance is genuine, choosing to forgive may be the more loving course of action. I certainly don’t commit some offense against justice if I choose to forgive without requiring punishment first. That said, I clearly would offend justice if I were to tell my son that I am willing to forgive him, but only if he can find someone else to be punished in his place. To make my forgiveness contingent on such a demand would be neither just nor forgiving.
We institute systems of punishment in our societies in the service of certain extrinsic ends–deterrence, incapacitation, rehabilitation, and the previously mentioned expressive reasons. The presumption of guilt on the part of the offender is an important constraint on our punishing practices–only the guilty are proper candidates for punishment. But if no purpose apart from sheer retribution would be served by a particular instance of punishment, if we somehow came to possess certain knowledge that an offender’s character has been transformed, that she is fully repentant for her wrongful action, and that pardoning her would not undermine the deterrence of other wrongdoers or the moral ecology of the community, then it seems implausible to say that we would be committing an offense against justice by choosing to forego punishment. Such a fortuitous set of conditions may rarely, if ever, obtain this side of the eschaton, and it might be objected that, in any event, we could never know for certain either that the offender is truly penitent or that pardoning her would have no deleterious effect on the social fabric. Yet surely God, who knows our hearts and whose kingdom rests on firmer foundations than ours, is able to make assessments of this kind.
As Jacob Morgan points out (69), we often quote Alma’s teaching that mercy cannot rob justice (Alma 42:25), but we have largely ignored Amulek’s teaching that mercy can “overpower” justice: “And thus he shall bring salvation to all those who shall believe on his name; this being the intent of this last sacrifice, to bring about the bowels of mercy, which overpowereth justice, and bringeth about means unto men that they may have faith unto repentance” (Alma 34:15). This scripture states that mercy can overpower justice on condition of repentance, which seems to imply that punishment is not required where there has been true repentance. It hardly makes sense to say that mercy has overpowered justice if justice had to be placated via punishment before forgiveness could take place. There is no need for suffering (vicarious or otherwise) once we have mended our sinful ways.
At this point, I wish to reiterate that it is not my intent to cast aspersions on those who understand the atonement through the lens of penal substitution. But for those who become aware of its problems, a penal substitutionary understanding of atonement can become more of a stumbling block than a springboard. Fortunately, the process of putting aside our paradigms and revising our assumptions is constructive, not merely destructive. There is a rich history of Christian reflection on the atonement that eschews the broken categories of penal substitution. In particular, the moral influence theory presents a framework for thinking about the atonement that, while certainly not answering every query one might have, provides a fruitful and spiritually nourishing alternative to penal substitution. Regrettably, an exposition of the moral influence theory is beyond the scope of this paper.
In conclusion, the penal substitution theory, for all its persistence and power, cannot withstand logical analysis. It clashes with the innocence principle and undercuts God’s capacity to forgive. Paradox and mystery are the inevitable companions of anyone who wrestles with the divine, but they should not be conflated with outright contradiction and confusion. Accordingly, the penal substitution theory must be abandoned.
Feinberg, Joel. “The Expressive Function of Punishment.” The Monist, vol. 49, no. 3 (July
1965): 397-423. Print.
Kyle, Brent G. “Punishing and Atoning: A New Critique of Penal Substitution.” International Journal for Philosophy of Religion, vol. 74, no. 2 (2013): 201-217. Print.
Morgan, Jacob. “The Divine Infusion Theory: Rethinking the Atonement.” Dialogue: A Journal of Mormon Thought, vol. 39, no. 1 (Spring 2006): 57-81. Print.
Ostler, Blake. “The Compassion Theory of Atonement.” Exploring Mormon Thought: The Problems of Theism and the Love of God, Greg Kofford Books, 2006, 235-281. Print.
Potter, Dennis. “Did Christ Pay for Our Sins?” Dialogue: A Journal of Mormon Thought, vol. 32, no.4 (Winter 1999): 73-86. Print.