Simon Gathercole defends substitution

Read time: 9 minutes

Simon Gathercole is worried that the doctrine of substitutionary atonement is going out of fashion so he sets out to defend it in this brief book Defending Substitution: An Essay on Atonement in Paul. It’s a very limited argument: in two main exegetical chapters he considers two statements that Paul makes: Christ died “for our sins” (1 Cor. 15:3; and Christ died “for the ungodly… for us” (Rom. 5:6-8). I ended up unconvinced by his defence, but not quite for the reasons I expected. 

In the Introduction, Gathercole explains that he thinks substitution is important for both doctrinal and pastoral reasons. He provides a straightforward definition: “I am defining substitutionary atonement for the present purposes as Christ’s death in our place, instead of us” (15). He distinguishes between substitution and other atonement ideas: penalty, representation, propitiation, and satisfaction. Finally, he addresses a number of theological, philosophical and logical criticisms of the “doctrine of substitutionary atonement”, including Steve Chalke’s notorious claim that substitutionary atonement amounts to “cosmic child abuse”, which he dismisses as “extremely shallow”, and Christopher Hitchens’ fierce objection to vicarious redemption: “I cannot absolve you of your responsibilities. It would be immoral of me to offer, and immoral of you to accept” (27). But this is just the introduction to a small book whose focus is on exegesis, so don’t expect anything more than a passing appraisal.

Chapter 1 examines three exegetical competitors to the substitutionary account of atonement: i) Otfried Hofius’ argument that in the atonement Christ includes the sinner in his own death in order to take him or her through to new life; ii) Morna Hooker’s “interchange” theory, whereby Jesus is not substituted for people on the cross but (in Gathercole’s words) “goes to the place where they are and takes them from there to salvation” (39); and iii) the “apocalyptic deliverance” view of the recently deceased J. Louis Martyn, which is that Jesus’ death liberates humanity from enslavement either to Sin or to the Law. These competing theories have their merits, but they also present problems, notably in that they tend to treat “sin” in abstract existential or corporate terms; they do not take into account concrete sins in the plural. Gathercole maintains that “Sins, transgressions, individual infractions of the divine will are… integral to Paul’s account of humanity’s plight” (54). For this reason—the logic is left unclear—he thinks that substitution has to be retained as a legitimate aspect of Paul’s understanding of the atonement.

Christ died for our sins according to the Scriptures (1 Cor. 15:3)

For I delivered to you as of first importance what I also received: that Christ died for our sins in accordance with the Scriptures…. (1 Cor. 15:3)

For the sake of argument, let’s assume that Gathercole is right—that when Paul says that “Christ died for our sins in accordance with the Scriptures”, he has in mind generally the Old Testament idea that people should normally die because of their own sins and not because of sins committed by another person (cf. Deut. 24:16 LXX; 1 Kgs. 16:18-19), and in particular the exceptional story of Isaiah’s servant, who suffered because of the sins of others. So we have a two-part logic: “(1) Jesus died because of the sins of others, and (2) because he has done that, he thereby saves others from the consequences of sin” (72). Whether this logic amounts to substitution I’ll consider at the end.

It is the next part of the argument that I have always found problematic. Gathercole holds that because Paul includes Gentile Corinthian believers in “our”, he must be thinking theologically rather than historically:

Jesus’s death is for Paul a theological consequence of sins rather than a straightforwardly historical one. By “theological” here I mean something that cannot be explained merely in terms of historical causation, for instance, Jesus dying as the result of the judicial verdicts of Herod, Pilate, and others. It is caused by others in another sense, however. Christ’s death for sins is theological because the divinely ordained consequence of sin is always death. (72-73)

I think the theological reframing of the logic of atonement is questionable. What exegetical reason is there for supposing that in 1 Corinthians 15:3 the “universal scope of the sins” is in view? Jesus “died for our sins in accordance with the Scriptures”. Why go to the trouble of establishing the biblical basis for the language and idea and then simply drop the biblical framework? There is no Old Testament basis for the idea of a vicarious redemptive death for universal sin. The servant suffers because of Israel’s sin, and it is Israel which benefits from his suffering. Likewise, it is Israel which will rise up on the third day and live in the presence of God (Hos. 6:2 LXX).

Between the narrowly causative account (Jesus died because of decisions taken by Herod, Pilate, and others) and the universal theological conception is a distinct third possibility: that Paul speaks on behalf of a historically sinful people that has been saved from the consequences of its sins by the suffering of Jesus. The Corinthians are “being saved” not because Jesus died for the sins of all humanity—that is outside the purview of Paul’s Jewish narrative—but because they have been incorporated into that community whose sins led to Jesus’ death.

The vicarious death of Christ and classical parallels (Rom. 5:6-8)

For while we were still weak, at the right time Christ died for (hyper) the ungodly. For one will scarcely die for (hyper) a righteous person—though perhaps for (hyper) a good person one would dare even to die—but God shows his love for us in that while we were still sinners, Christ died for (hyper) us. (Rom. 5:6–8)

The approach taken in this section comes as a bit of surprise. In 1 Corinthians 15:3 Christ died for “our sins”; in Romans 5:6-8 he dies for people. Gathercole appears to attribute the distinction—he doesn’t quite make it this explicit—to the literary background: the “died for our sins” formula reflects Old Testament language and ideas, but in the Romans passage the background is Greek-Roman: “Paul links the death of Christ with other heroic deaths from his cultural environment—what he refers to as the rare examples of deaths for good or righteous individuals” (86).

Gathercole makes little attempt to justify the contention. It is suggested that Jewish tradition “does not have much to say about individuals dying for other individuals”—the Maccabean martyrs tended to die “for piety” or “for the covenant of our fathers”. Much is made of the supposedly well known story of Alcestis, who died “on behalf of” her husband. But at risk of damaging my egalitarian credentials, I find it hard to believe that Paul had such an example of conjugal love in mind as a model for the substitutionary death of Christ.

It is hardly surprising that classical literature affords us instances of one person “dying for” another person—even daring to die for another person—but this does not prove that Paul meant to invoke such stories as models for Christ’s death for others. It’s not out of the question, of course, that his language echoes—for secondary rhetorical reasons—the motif of the noble death for others, but the Jewish conceptual background may have been too readily rejected. 

In the first place, it looks to me as though in verse 7 Paul has parenthetically reworked Proverbs 11:31 LXX: “If the righteous is scarcely saved (ei ho men dikaios molis sōizetai), where will the impious (asebēs) and the sinner appear?” The righteous person is scarcely saved by a person dying for them; a person might dare to die for the “good” person who is not impious and a sinner; but Christ “died for the impious (asebōn)”. So I’m not persuaded that “in talking of one person dying for another good or righteous person, the most natural link in Romans 5 is with examples of vicarious death in classical texts” (90). Proverbs 11:31 is quoted in the course of a similar argument in 1 Peter 4:12-19.

I also don’t think the relevance of the Maccabean martyrs can be ruled out. The last of the seven brothers speaks of giving up “body and life for our ancestral laws”. But he also prays that their deaths will “bring to an end the wrath of the Almighty that has justly fallen on our whole nation” (2 Macc. 7:37–38). They are suffering because of the sins of Israel; but the prospect is raised that their deaths will bring to an end the punishment of Israel. We have the same thought in 4 Maccabees 17:22: “through the blood of those pious people and the propitiatory of their death, divine Providence preserved Israel, though before it had been afflicted”. Paul goes on to say that having been justified by Jesus’ blood, they may now hope to be “saved by him from the wrath of God”. This reinforces my basic point: Jesus dies for the sins of God’s people in order to preserve them as a people from the coming wrath. Gentile believers in Corinth have simply been assimilated into this narrative.

Has a case been made for substitution?

My standard objection to substitutionary models of atonement has been that they are constructed theologically rather than historically. I haven’t had a problem with the idea in principle, just with how it’s been used. But this book has left me wondering whether substitution is required at all to explain Jesus’ death for sins or for others.

Gathercole accepts that in 1 Corinthians 15:3 huper in itself does not have a substitutionary sense: it means only that Christ’s death was a consequence of “our sins”. Rather, “It is when this is set in the framework of one person doing this for the sins of others (and not for one’s own) that the substitutionary sense is achieved” (74). But Paul does not say here that Christ died for the benefit of others. His argument in this chapter, in fact, is that the Corinthians are “being saved” by holding fast to the word preached to them (15:2). Their faith is in the resurrection; and if Christ has not been raised, they are still in their sins. He does not say: if Christ did not die, they are still in their sins (15:17). It is those who are “in Christ” who will be raised from the dead, “made alive” (15:22). This would lead us to conclude that Christ died because of the sins of his people, whereas those who believe that God raised him from the dead will share in his resurrection. Substitution appears to have nothing to do with it.

The Romans passage also looks rather different if we take the wider argument into account—Gathercole is guilty of treating both brief statements in isolation from their context. The argument of Romans 5:6-11 runs as follows: Christ died for the ungodly; therefore, the ungodly are justified, reconciled to God; therefore, they will be saved from the wrath to come.

If “wrath” is simply taken to mean God’s final judgment on the individual, then perhaps substitution has something to do with it—we simply need the sort of qualification of “death” that Gathercole provides in an excursus (“Why, Then, Do Christians Still Die?”). But if, as I maintain, “wrath” signifies historical judgment on a people or nation or empire or civilization, then the point would be that by dying for the ungodly Christ has ensured the future survival and viability of God’s people. This apocalyptic outcome could still be expressed in substitutionary terms, along the lines of both Isaiah 53 and the Maccabean martyr stories, but we are some way from the sort of individualized theologizing that Gathercole has been defending.

One concept I can’t get around is the idea of Jesus’ death being a redemption.

I’m thinking of Exodus 13:11-16.  To commemorate the occasion of God killing the firstborn sons of the Egyptians but redeeming the sons of Israel, God commands that, for the passover, Israel will sacrifice the firstborn males of her livestock but will “redeem” their firstborn sons with a sacrificial lamb.

This isn’t substitionary atonement in the traditional sense of Jesus paying a penalty that was meant for someone else, but at the same time, there is an element of substition involved, isn’t there?  Isn’t there a sense in which the martyrdom of Jesus “buys back” Israel’s life?

@Phil Ledgerwood:

The firstborn of a donkey (for whatever reason) is redeemed with a lamb. We might want to say that the lamb has been substituted for the donkey, but it appears rather that the lamb is a payment for it. The death of the lamb may be implied, but it’s not a factor in the instruction; it’s not part of the logic.

The “firstborn of a man” is likewise redeemed but the price is not specified (nor in 34:20).

It’s only the firstborn son who needs to be redeemed—and so not the rest of the children. This doesn’t work for an atonement model, clearly. Jesus is often himself the firstborn.

There is no sin involved—other than Pharaoh’s stubborn refusal to let the people go.

Does the New Testament make use of this passage to explain the atonement? I don’t think so.

The way I see it, historically speaking, the martyrdom of Jesus changed things for Israel, not in itself perhaps, but because YHWH then raised him from the dead (hence it is the resurrection that is effective, not the death in 1 Cor. 15) and delegated to him authority to judge and rule. In light of this outcome, it was natural to redescribe his death metaphorically in sacrificial or redemptive terms—or in terms of the story told in Isaiah 53. This is a linguistic or rhetorical theory of the atonement!

@Andrew Perriman:

Oh, there’s absolutely no sin involved.  I don’t think there’s anything going on about a substitute payment for sin or penalty paid by sacrifice.

Your points are all true.  I guess I thought the lamb would be killed because, if the donkey’s owner won’t redeem it, he breaks its neck.  But you’re right, it doesn’t mean you kill the lamb, just that the lamb is given to the Lord.  Also, there’s a parallel between what Israel is doing with the livestock and their firstborn sons and what God did to Egypt with their firstborn livestock and sons.

But Jesus is described in the NT as our Passover lamb, which is certainly not a subtitutionary offering for sin, but it seems like there is some aspect of substitution going on in some sense, right?  The lamb is a redemption of the firstborn.  Maybe I’m too entrenched in old categories.  In what sense is Jesus a passover lamb?