A Primer on Constructing a Penal Substitutionary Atonement Model by Jonathan C. Rutledge

Growing up in the Bible belt as I did brings you into contact with a wide range of ideas of what constitutes the Christian Gospel. Strange as it might sound, one of the worst accusations someone could have said of me in my high school days was that I held a belief that was inconsistent with the Gospel. Truly, that would have worried me deeply. As a result and ever since, when someone claims that doctrine-x is an essential part of the Gospel, my critical self takes over and I want nothing more than to clarify precisely what they think doctrine-x actually is.

Recently, my critical self awoke again when I came across a resolution by one major Protestant denomination that said something very close to “doctrine-x is an essential part of the Gospel”. The following line is what caught my eye:

[We] reaffirm the truthfulness, efficacy, and beauty of the biblical doctrine of penal substitutionary atonement as the burning core of the Gospel message and the only hope of a fallen race” (italics mine).[1]

Now, penal substitutionary atonement (henceforth, ‘PSA’) is a family of models of the atonement that claims that salvation was achieved by Christ’s being punished in our place for sin. Or at least, that is a typical definition assigned to PSA. The problem is this: PSA is a family of models, which means that there is more than one version of PSA that someone might affirm. And not only are there multiple versions of PSA, but some versions of PSA are inconsistent with each other. As a result, it’s possible for two different individuals to both affirm a version of PSA while one of the individuals believes something false. In other words, on the assumption that PSA is in fact a part of the Gospel message, one of the individuals involved would fail to believe the Gospel despite affirming a version of PSA. So perhaps one could be forgiven for asking: Which version of PSA do the drafters of this resolution have in mind?

Well clearly, even if we were to allow that PSA might be a core constituent of the Gospel (i.e. a claim on which I remain neutral here), it is incumbent upon those who advance PSA—like the drafters of the above resolution—to speak clearly about what they mean. And this is true if for no other reason than to ensure that we really understand what believing the Gospel, on this assumption, really requires. For instance, did Christ bear the punishment for our sins or merely suffer what would have been punishment if applied to us?[2] In what did Jesus’s punishment or suffering consist—e.g. being the object of God’s wrath or suffering death and eternal torment?[3] What was the justification for Christ’s (or humankind’s) punishment? Different answers to these questions deliver different versions of PSA, and if PSA is a core component of the Gospel, then different answers to these questions deliver different (and possibly incompatible) articulations of the Gospel.

I cannot hope in such a brief post to map all of the logical terrain on which different forms of PSA might be located. I do seek, however, to offer a beginners guide of sorts for someone to clarify what they mean when they affirm that Jesus died to serve as my penal substitute.

Penal Substitution

What does it mean to say that Christ is our penal substitute? It is at least to say that he (in some sense) substitutes for our punishment. So understanding the concept of punishment is critical for clarifying what one means in endorsing PSA. So, what is punishment, exactly?

As a working definition, we can say that punishment is at least…

(i) harsh treatment, applied to an alleged offender for

(ii) breaking some sort of norm (i.e., a rule)—moral or social, let’s say—

(iii) by someone with the authority to apply such harsh treatment.[4]

Accordingly on this definition, vigilante activity doesn’t count as punishment (i.e. because of condition (iii)). Furthermore, due to condition (i), punishment cannot occur without an accompanying form of suffering of some sort, such as physical or psychological suffering (i.e., pain or shame). Thus, solitary confinement counts as punishment given that it constitutes a form of social suffering, but presenting an offender with, for instance, a key to the city would not, in any normal case, count as punishment.

Now that we have a general understanding of what punishment is, we must ask what in particular serves as the punishment for sin according to PSA. One traditional answer is to say that the punishment is death. But what else must hold in order for death to generally count as a punishment as defined above? A proponent of PSA adopting such a definition of punishment would also need to say that God was the authority figure—satisfying condition (iii)—who imposed the death as harsh treatment for the offense of sin—satisfying condition (ii).

One complication accompanying such a view is that there is a distinction between punishment as defined above and the natural consequences of an action. If death is a punishment for sin and not a natural consequence, then it would be possible for sin to enter the world without death. And this is because, at least in normal cases, what constitutes the punishment of an offense is not required by the offense. For example, if my son disobeys me when I tell him to clean his room, there’s nothing in particular about failing to clean his room that requires that I punish him by putting him in timeout. I could just as easily choose a different punishment, such as no longer taking him to the movies. Thus, if death is merely a punishment for sin in this sense, then death seems a bit arbitrary. However, if death is a natural consequence of sin, then the arbitrariness disappears. The difficulty in that case is that it’s no longer clear why death would count as punishment—given condition (ii)—any more than the consequence of getting sick from eating too much candy should count as a punishment. So how should we understand death (or whatever other particular punishment for sin you have in mind)? Is it a punishment, natural consequence, or somehow both?

Next, we must consider whether in the act of punishing one is condemning the one punished. If so, then Christ can only be punished on our behalf, strictly speaking, if Christ is also condemned. Indeed, some versions of PSA affirm this very claim while others resist it, sometimes with fervor.[5] Obviously though, adopting either definition of punishment results in a very different looking version of PSA. Which version(s) is (are) constitutive of the Gospel? It’s up to the particular proponent of PSA to let us know.

Penal Substitution

Once one identifies what punishment in general is and what specific punishment is applied to sinners, one must go on to clarify what is meant by ‘substitution’. For instance, we might immediately think that Christ becomes the bearer of punishment in place of humanity. This is complicated, however, for does Christ bear the same exact punishment (i.e. death and eternal separation from God) as humanity? Or does Christ bear an equivalent punishment to our own?

Perhaps punishment includes condemnation, in which case, a PSA theorist might say that Christ does not bear the punishment for sin so much as a penalty that would have been our punishment had it been applied to us.[6] So Christ is the bearer of suffering rather than bearer of punishment on that view. And this is importantly different since it allows those who understand punishment as partly constituted by condemnation to avoid the puzzling feature of God condemning Godself on the cross (i.e., a feature arguably in tension with good Trinitarian theology).

Some proponents of PSA, however, might prefer to incorporate an element of participation into the substitutionary aspect of PSA, especially in light of verses that imply we suffer with Christ (e.g. Col 1:24 – “…I rejoice in my sufferings…I am filling up what is lacking in Christ’s afflictions for the sake of his body” ESV).[7] In such a case, Christ indeed bears our punishment, but he does so with us (perhaps because we are in Christ in some sense). Now, someone might be suspicious that no real substitution is to be found on such a view, but that’s not quite accurate. For whereas humanity initially suffers death for sin according to the biblical narrative, after Christ’s atoning work a new set of persons absorb that punishment; namely, humanity and Christ. One set of persons is substituted for a new set of persons, and so, a form of substitution is undoubtedly present. Such a view does not include the typical characteristic of PSA models—namely, that Christ is punished and we are not—but whether that characteristic is required for a model to count as a version of PSA is an interesting and additional question meriting independent consideration.


As I have continued to delve into the world of theology, I have witnessed a wide range of agreement and disagreement amongst intellectual giants in the history of theology. Within PSA alone, I can think of several different authors from a number of different traditions who arguably could be thought to have affirmed one or another form of PSA (e.g. Hugo Grotius, John Wesley, John Calvin, Jonathan Edwards, NT Wright, etc.). With such figures looming large in my mind, I would think that someone should be very reluctant to deny that PSA has any reasonable biblical merit. As a result, I hope that these reflections will encourage those committed to PSA to clarify which version of it they intend to affirm. Such clarifications can only hope to benefit the church as we corporately aim to better understand the teaching of scripture on the important issue of Christ’s sacrifice.



Jonathan C. Rutledge is a Junior Research Fellow with the Logos Institute for Analytic and Exegetical Theology. He holds a Ph.D. in philosophy from the University of Oklahoma, where he studied under Linda T. Zagzebski, and a Ph.D. in divinity from the University of St Andrews under Alan J. Torrance. His primary academic interests lie in the areas of epistemology, philosophy of religion, and systematic & analytic theology. His current projects include work on the nature of forgiveness, a sacrificial account of atonement, philosophical Arminianism as an account of divine creation, and constructing a new Foley-inspired account of epistemic rationality & defeat.


[1] See the Southern Baptist Convention’s resolution, “On the Necessity of Penal Substitutionary” (2017): http://www.sbc.net/resolutions/2278/on-the-necessity-of-penal-substitutionary-atonement.

[2] William Lane Craig. “Is Penal Substitution Incoherent? An Examination of Mark Murphy’s Criticisms.” Religious Studies (2017): doi:10.1017/S003441251700018X.

[3] Thomas H. McCall. “Is the Wrath of God Really Satisfying?” Christianity Today (29 March 2018): http://www.christianitytoday.com/ct/2018/march-web-only/is-wrath-of-god-satisfying-good-friday-cross.html; DeYoung, Kevin. “In My Place Condemned He Stood” The Gospel Coalition (3 April 2018): https://www.thegospelcoalition.org/blogs/kevin-deyoung/place-condemned-stood/

[4] You get something akin to this definition in Mark C. Murphy. “Not Penal Substitution but Vicarious Punishment.” Faith and Philosophy 26.3 (2009): 253-273.

[5] For a Christus Odium version of PSA, see Joshua R. Farris and S. Mark Hamilton. “Does God the Father Hate His Beloved Son?” The Evangelical Pulpit (12 December 2017): http://www.patheos.com/blogs/evangelicalpulpit/2017/12/god-father-hate-beloved-son/. For staunch resistance to such a version of PSA, see Stott, John R. W. 1986. The Cross of Christ. Downers Grove, IL: InterVarsity Press, 151.

[6] Cf. Craig, “Is Penal Substitution Incoherent?”

[7] See, for instance, Tim Bayne and Greg Restall, “A Participatory Model of the Atonement,” in New Waves in Philosophy of Religion, Eds. Yujin Nagasawa and Erik Wielenberg (New York, NY: Palgrave Macmillan, 2009): 150-166.