Ed.’s note: All comments will be pre-moderated, meaning that each comment will post after moderator approval.
by Dr. David L. Allen
Dean of the School of Theology
Professor of Preaching
Director of the Center for Expository Preaching
George W. Truett Chair of Pastoral Ministry
Southwestern Baptist Theological Seminary
“Review of From Heaven He Came and Sought Her: Definite Atonement in Historical, Biblical, Theological, and Pastoral Perspective, David Gibson & Jonathan Gibson, eds. (Wheaton: Crossway, 2013) – Part 2
In this final summary review, I will cover the section on definite atonement in theological perspective. Since I am well over the limit for this post, I will have to address the final pastoral section of the book, along with the foreword, preface, and introduction, at a later time.
A more detailed analysis of many of the chapters in the volume will appear later this month on my new website www.DrDavidLAllen.com.
Definite Atonement: Theological Perspective
There is much helpful material in the theological perspective chapters. Probably the most significant is the attempted interaction with the Amyraldian and Hypothetical Universalism views within the orbit of Reformed theology. This is commendable and opens the door for further dialogue to occur. However, this interaction is not without its problems, as there appear to be a fair amount of misunderstanding and mis-characterization of these positions.
When one examines the arguments made by the authors in the theological section, certain common key presuppositions become evident. At the risk of oversimplification, I see these to be:
1. The notion that God’s intention in the atonement (to save the elect) entails a limitation in Christ’s sin-bearing such that there is an imputation of the sins of the elect only to Christ at the cross. Limited intent entails limited extent.
2. God’s salvific intention is expressed in the divine decree of election, and election entails the notion that atonement is only made for the sins of the elect. Election entails limited extent.
3. Penal substitution entails limited extent.
4. Hypothetical Universalism in all forms entails Trinitarian discord.
5. Old Testament priestly typology entails New Testament definite atonement.
6. There are only two options: either definite atonement or universalism.
Macleod’s chapter addresses definite atonement and the divine decree. After dealing with Supralapsarianism and Infralapsarianism as versions of predestinarianism, he turns to a discussion of Hypothetical Universalism and election (422-434).
One troubling feature of this chapter is what appears to be Macleod’s misunderstanding of HU. In a discussion on the importance of not trying to pry into the imponderables of God’s secret will but rather engage in indiscriminate preaching, he notes that Hypothetical Universalists fare no better than their high-Calvinist counterparts: “Hypothetical Universalism provided no solution. How could they put their trust in a hypothetical redemption? How could they believe at all unless they were elected to faith?” (431). Macleod’s first question appears to misunderstand just what it is that is “hypothetical” in Hypothetical Universalism. For all HUs, the atonement is not hypothetical for the non-elect, it is actual. What is hypothetical is the conditionality of faith: if anyone believes, he shall be saved based on the fact that there is an atonement for sin. Conditionality is operative in all orthodox Christian approaches to salvation: faith is a necessity. With respect to the second question, as Calvinists, all HUs believe in election and the necessity of effectual calling.
Macleod asks: “Is it not fatally incoherent that God should simultaneously decree that the cross of Christ should redeem all the non-elect and provide him with grounds for their greater condemnation?” (434). It did not seem so to Calvin, who states this very point.
Robert Letham contributes a chapter entitled “The Triune God, Incarnation, and Definite Atonement.” The chief burden of his chapter is to demonstrate that all forms of universal atonement create Trinitarian disharmony (439). Letham lumps Amyraldians with Arminians when he states that “the atoning death of Christ does not of itself secure the salvation of anyone in particular, since it is contingent on the human response in the case of Arminianism or on the particular work of the Spirit in terms of Amyraldianism. Moreover, since the atonement is not intrinsically efficacious, it cannot yield a doctrine of penal substitution” (440).
This statement is problematic on two counts. First, note the operative phrase “of itself.” Interestingly, I recall Charles Hodge and W. G. T. Shedd stating that the atonement “of itself” secures the salvation of no one. No one is saved until, through the work of the Holy Spirit, they are brought to faith in Christ. This again confuses extent with application. Second, it is incorrect to argue that lack of intrinsic efficacy negates penal substitution. Letham may be slipping into the web of a commercialistic view of the atonement here.
It appears Letham assumes that, for Amyraut, Christ died equally for all in terms of intent as well as extent (438-440). This is true for Arminians but not true for Amyraut and his followers. When Amyraut spoke of Christ dying “equally’ for all, it is clear he meant that his death was equally sufficient for all, not in the sense of intention. Amyraut believed Christ died with the special intent of saving only the elect.[3
Letham’s critique of T. F. Torrance scores some points, but some things should be noted. First, Letham may disagree that the incarnation, for Torrance, entails an atonement for all humans (447), but let’s not forget that this was the general position of the Church Fathers and the medieval church. Second, Letham is correct to note Torrance’s error in claiming Calvin rejected the Lombardian formula, but then errs himself when he concludes from this that Calvin held to definite atonement. This fails to take into account the revision of the formula after Calvin’s death by supporters of definite atonement (see previous reviews).
Finally, Letham’s assessment of Davenant’s position (444) appears to be incorrect. Davenant taught, on the basis of 2 Cor 5:18-20, that God was reconciled to the world objectively, but not all people in the world are subjectively reconciled to God. Davenant’s understanding of this universal provision did not “overshadow” God’s decree to save the elect. Letham is arguing that Davenant’s construct entails conflict and incoherence, but Davenant is arguing that the Trinity acts in unity to accomplish a dual intent in atonement and redemption.
Garry Williams writes the next two chapters on penal substitution and the double payment argument. The upshot of these two chapters is an attempt to argue the point that penal substitution entails definite atonement and that the double payment argument (that God cannot demand payment for sin twice) is a valid construct and supports definite atonement.
First rattle out of the box, Williams comes across as a bit condescending when he states his chapter is “. . . simply intended to show brothers that at this point they are wrong, . . .” Perhaps it would have been better to word this along the lines of showing brothers “reasons why perhaps they may be wrong in their assessment.”
In his critique of Broughton Knox, Williams discusses distinctions between intent, extent, and application (471). The non-sequitur is his statement “The separation is unsustainable: with an ontology that gives proper place to the constitutive role of God’s will, the prior, determinative divine intention for the sufferings of Christ makes them what they are and thus makes them definite in nature.” This is an assertion with no support.
Williams attempts to show that in the HU approach, the suffering of Christ is not “identifiable as punishment for the sin and sins of particular persons . . . (471). The NT use of “sin” singular (generic) and “sins” plural (specific) indicates that the use of one form does not exclude the other. Rightly so. His conclusion, however, does not follow: “Although none of these NT writers were self-consciously addressing our question, they evidently held that Jesus died bearing specific sins committed by particular people” (474). We will consider this issue in Williams’ next chapter.
Williams concludes his first chapter with a false either/or, statement: “An indefinite atonement must either embrace universalism or it must contradict the biblical doctrine of penal substitution” (481). This conclusion does not seem to be warranted given the evidence in the chapter.
Williams’ next chapter addresses the double payment argument. Classically formulated by John Owen, the double payment argument asserts that God’s justice does not allow the same sin to be punished twice, first in Christ and then in the sinner. Owen’s defense of this argument employs a commercialistic understanding of the atonement, for which he has been rightly criticized by Calvinists and non-Calvinists alike. The gist for Owen is that “debt” language in Scripture moves beyond the metaphor and actually describes the mechanism for the payment of sin. Owen and Williams assume that since the satisfaction is for “sins” plural, and not for “sin” in abstraction, that it therefore must be definite (limited to only the elect). The transaction is commercial: so much is owed and so much is paid. If Christ paid for all sins, then God cannot demand a second payment from any sinner. Seems like an open and shut case.
Here are the problems. The metaphor is pushed beyond its legitimate point of analogy and becomes, for Owen and Williams’, the actual mechanism whereby sin is paid for. Williams’ dependence upon Owen’s treatment of the parable of the Unforgiving Servant in Matt 18 leads him to misinterpret the point of the parable. The context of the parable is not atonement but forgiveness between brothers by way of a commercial debt metaphor. The point of the parable is the mechanism for forgiveness, not the mechanism for satisfaction of sins.
Williams concludes that Christ’s satisfaction is a “repayment.” The mistake is viewing God as a creditor from the fact that sin is metaphorically described as a debt. Sin as debt is about obligation, not about the death of Christ being a payment to a creditor (God).
As part of his argument, Williams posits that it is “both possible and necessary to hold together the idea of an unquantifiable punishment and an inherently definite atonement” (499). Recall Williams’ conclusion in his previous chapter that only “sins” were laid on Christ at the cross, not “sin” generically. This is fundamentally an unnecessary and even flawed bifurcation. No one claims that Christ dies for “sin” without dying for “sins.” Of course Jesus did not die for some abstract notion of sin. He died for real people; all of them. He accomplished this by becoming “sin” for us (2 Cor 5:21). Christ died for “the one and the many,” for “sins” and for “sin.”
Like Owen, Williams appears to be operating from a sort of transference view of imputation: specific guilt for specific sins is laid on Christ. But this is problematic. While our sins are imputed to Christ, before our conversion we remain under the wrath of God as Paul states in Eph 2:1-3. As Dabney says, God holds the unbelieving elect subject to wrath until they believe. It would appear that Williams fails to address this objection by Dabney that the living unbelieving elect are under the wrath of God.
Second, would Owen consider the imputation of Christ’s righteousness to believers as the transference of so many acts of law-keeping? It would seem not. Are believers credited with specific acts of righteousness on Christ’s part? No, we are credited with a quality of righteousness. All of Christ’s acts of obedience fall under the somewhat abstract class of “righteousness.” Just as believers are not imputed with something like so many bits of righteousness but rather with righteousness, so also Christ was not imputed with “sin-bits” but rather with sin in a comprehensive way. He was treated as though he were sinful. Owen, and it would seem Williams as well, has a faulty notion of imputation. Christ died one death that all sinners deserve under the law. In paying the penalty of what one sinner deserves, he paid the penalty of what every sinner deserves. He suffered the curse of the law as defined by the law. Owen’s double payment and trilemma arguments undermine the true meaning of imputation and operate on the assumption of the transference of specific sins.
Charles Hodge, in contrast, has retained the proper understanding of imputation:
What was suitable for one was suitable for all. The righteousness of Christ, the merit of his obedience and death, is needed for justification by each individual of our race, and therefore is needed by all. It is no more appropriate to one man than to another. Christ fulfilled the conditions of the covenant under which all men were placed. He rendered the obedience required of all, and suffered the penalty which all had incurred; and therefore his work is equally suited to all.
Williams is at odds with Hodge.
Finally, Williams trades on the false dilemma fallacy when he asserts: “If God punishes all sin, then Christ must have died for the sin of unbelief, and if he did that for all without exception, then all without exception must be saved.”
Stephen Wellum addresses the subject of priesthood, atonement, and intercession in an effort to demonstrate that only definite atonement takes account of Christ’s unified priestly work: Christ only dies for the sins of those for whom he intercedes. For Wellum, all general atonement views fragment Christ’s priestly work of offering and intercession (530).
Wellum’s appeal to the typology of Christ as our High Priest simply cannot carry the freight he wishes to place on it. Caution should be exercised so as not to read categories of “elect” and “non-elect” from the New Testament back into the OT. Neither should we read OT categories of the sacrificial system into the NT unless we have specific biblical justification to do so.
While Wellum considers objections from contemporary scholars, I do not see where he addresses the critique of his position made long ago by the likes of Baxter and Bunyan.
Space considerations limit me to speaking specifically to one of the arguments: the supposition that if Jesus’ intercession (John 17) is limited to the elect, then he only dies for the sins of his elect. But nowhere in the text itself does it state that Jesus dies only for those for whom He prays, let alone the fact that John 17 makes no mention of the death of Christ at all. Laying aside for the moment the possibility that in context this is most likely a reference to the disciples, and even taking it as extending to the believing elect at the time, even then one is not warranted to draw the conclusion that the text means that Jesus did not die for the sins of all people, elect and non-elect. Wellum falls prey to generalizing that election entails limited atonement. The mistake here is a collapsing of the intercession of Christ into His expiation for sins. This merely begs the question.
Henri Blocher’s chapter on constructing a systematic theology of definite atonement rounds out the theological section, and appropriately so since he attempts to develop something of a systematic theology of definite atonement.
I am especially grateful that Blocher engages my critique of definite atonement at numerous points in his chapter. Though there is much with which to engage here, I shall only select two or three points.
Blocher asks where the decisive difference lies between HU and DA:
In the relationship with election. Is the purpose of the atonement identical for all, elect and reprobate? Hypothetical Universalism answers yes; definite atonement answers no. Or, in the transaction that took place on the cross, which is described by such phrases as “bearing sins,” “satisfying divine justice,” “paying the ransom-price,” are the reprobate included as well as the elect? Hypothetical Universalism: Yes; definite atonement: no. Or, did atonement secure eternal life in such a way that those for whom it was accomplished according to its main purpose and operation shall infallibly receive it at the end? Definite atonement, yes; Hypothetical Universalism: no (548-49).
With respect to the first question, it is not accurate to say that HU asserts the purpose of the atonement is identical for all. The extent of the atonement is identical for all, but the intent (purpose) according to HU is the same as with DA: to secure the salvation of the elect. With respect to the third question it is also inaccurate to state that HU does not affirm that the atonement secures eternal life “according to its main purpose” since supporters of HU affirm that this is accomplished in the application on the grounds of both God’s intent and the atonement’s extent.
Blocher also appears to misread both Andrew Fuller and Charles Hodge, whom he assigns to the DA camp, in disagreement with my placing them in the HU camp. Blocher thinks many of my quotations “fail to convince because the flexibility of the language used by definite atonement supporters is not recognized…” (551).
Actually it is the other way around. Blocher fails to recognize flexibility of the dualism of “intent” and “extent” in these names I provided. He confuses Hodge’s clear statement about Christ’s sin-bearing for the world with the issue of the universal offer (554). For Hodge, it is the universal nature of the atonement that grounds the universal offer, which is Hodge’s point in the very quotation Blocher cites (555). Hodge clearly states that though Christ did not die “equally” for all men (Arminianism), yet he died for his “sheep” and “Church,” (Calvinism) and “He did all that was necessary, so far as a satisfaction to justice is concerned, all that is required for the salvation of all men.” (Hypothetical Universalism). If Christ’s death accomplished “all that is required for the salvation of all men,” then it cannot be a limited substitution as in the definite atonement scheme. Hodge affirmed Christ’s universal sin-bearing.
Blocher offers an inadequate response to Gary Schultz’s point about the gospel content of Paul’s kerygma to the unconverted Corinthians: “Christ died for our sins” in 1 Cor 15:1-3. He attempts two defeaters: 1) Paul does not reproduce the wording verbatim; 2) “for us” may have meant Paul’s team and any who would join them. This statement by Paul clearly implies he preached Christ died for the sins of all.
While the book will likely be too much for some laypeople to digest, I would encourage all theological students, pastors, and scholars to take the time to read it and digest it. It is probably the most comprehensive defense of definite atonement available. On the surface, it looks formidable, but it has a soft underbelly and is vulnerable to a number of criticisms.
It only takes one clear statement in Scripture that Christ died for the sins of all people to confirm unlimited atonement no matter how many statements indicate Christ died for a specific group of people. Likewise, it would only take one clear statement in Scripture that Christ died only for the sins of the elect to confirm definite atonement. There is not one single statement in Scripture that overtly states Christ died only for the sins of the elect. There are easily a dozen New Testament Scriptures overtly stating Christ died for all people.
The burden is on the authors of this book to prove that a simple positive statement can entail a universal negation. This is the book’s claim. The hill which the authors must climb is to prove, exegetically from Scripture, that Christ died only for some people’s sins (a limited imputation of sin). If exegetically, DA fails, then no amount of theological flying buttresses will support it.
 Calvin, Sermons on Isaiah, 141.
 Charles Hodge, Systematic Theology, 2:557-58. W. G. T. Shedd, Dogmatic Theology, 3:418.
 See http://calvinandcalvinism.com/?p=2293.
 Dabney, Lectures in Systematic Theology, 521.
 Charles Hodge, Systematic Theology (Grand Rapids: Eerdmans, 1993), 2:545.
 For the exegetical argument that Jesus prayed both for the elect and for the world, interested readers might consult Harold Dekker’s analysis of John 17 in, “God’s Love to Sinners — One or Two?,” The Reformed Journal 13 (March 1963), 14-15. Dekker was formerly professor and Academic Dean at Calvin Theological Seminary. See also Leon Morris, The Gospel According to John (Grand Rapids: Eerdmans, 1971), 725, who argues the same point.
 For the evidence that the later Fuller came to reject limited atonement, see Peter Morden, Offering Christ to the World: Andrew Fuller and the Revival of Eighteenth Century Particular Baptist Life, vol. 8, Studies in Baptist History and Thought (Waynesboro, GA: Paternoster, 2003), 26-27. See also my “Preaching for a Great Commission Resurgence,” 292-94, in Great Commission Resurgence, eds. Lawless & Greenway (Nashville: B&H Academic, 2010).
 Notice also how Dabney, Hodge’s contemporary, cites Hodge as affirming universal expiation. Robert Dabney, Lectures in Systematic Theology, 527.