Part 3 of a 3-part series. You can read Part 1 here, and Part 2 here. This post includes my theological analysis of the passage as well as the implications I see in it, followed by the conclusion and an addendum. Note: the footnotes restarted at 1 owing to how the post is divided. [UPDATED with working link for the White-Brown debate in footnote 15.]

Theological Analysis and Implications

John Owen,[1] John Gill,[2] John Frame,[3] Michael Horton,[4] Tom Nettles,[5] and Jonathan Gibson[6] are representative of Calvinists who assert Rom 8:32–34 supports limited atonement. I find it interesting that some of the systematic theologians tend to follow this line of thinking, while most of the biblical theologians tend not to do so, even when they agree in principle with the systematic theologians that limited atonement is biblical.

Consider Jonathan Gibson’s argument that Rom 8:32–34 supports limited atonement:

Romans 8:32 provides yet another new insight into Paul’s soteriological framework: redemption accomplished on its own secures redemption consummated, without any reference to redemption applied. What Paul presents here is the efficacy of Christ’s atoning work (without reference to its application); it cannot but produce its intended effect. Put in another way, all those for whom Christ died cannot but be given all things in order to reach final glorification.

As we saw earlier, the demonstrative pronoun toutous shows that the moments of redemption predestined, applied, and consummated all carry the same extent. In verse 32, Paul now shows that redemption accomplished also carries the same extent as the other moments of salvation. Paul presents redemption accomplished and redemption consummated as coextensive: If Christ was given up “for us all” (huper hēmōn pantōn), how will God not also, freely, along with Christ (sun autō), give “us” (hēmen) all things in order to be glorified. This means that unless one wishes to affirm universal salvation, the word “all” must be limited in some way. The context provides the correct referent for “us all” (hēmon pantōn): the “us” of verse 32 is the same as the “us” of verse 31 and those referred to in the preceding verses: those whom God foreknew, predestined, called, justified, and will one day glorify (vv. 29–30). The subsequent verses also support an intended and definite reference: the “all of us” are God’s elect (eklektōn theou; v. 33) and those for whom Christ intercedes…

The mere proposition that the text does not contain the word “only” cannot be used to counter the case for definite atonement in Romans 8, since the text has its own inherent logic, one which demonstrates clearly that (1) those for whom Christ died are the elect, and that (2) Christ’s death is an efficacious substitutionary atonement that cannot but produce its intended effect. The nature of the atonement is in Paul’s purview here, and its nature is one of ultimate efficacy: those for whom Christ died will make it to glory. Paul’s argument is therefore tendential to definite atonement and can point in no other direction.[7]

Gibson’s analysis founders on several points. He has made the same logical mistakes that Owen made. First, nothing in the atonement secures its own application. Where is this stated in the text or any text? Application of the atonement is the purview of the Holy Spirit (even from a Calvinistic perspective when He sovereignly regenerates an individual and then gives the gift of faith such that the individual believes and is justified). Whether faith can be described as a purchased gift only for the elect or not changes nothing. The atonement unapplied saves no one, as Charles Hodge rightly noted.[8] The only application of the atonement according to Paul in Romans is in response to faith. Second, Gibson employs the negative inference fallacy: the proof of a proposition does not disprove its converse. Where does Paul say that redemption accomplished is equivalent in extent as the other moments of salvation in the golden chain? He does not. Third, as demonstrated in the exegesis of the text above, Gibson is equivocating on the term “us” when he attempts to make it apply to all the elect as an abstract class, a hermeneutical and exegetical false move. Fourth, the nature of the atonement is not in focus in this text, nor can Gibson sustain the claim that its nature is one of “ultimate efficacy.” He has imported some aspects of Reformed theology into the text which the text itself does not state.

David Ponter offered a salient critique of the attempt to interpret Rom 8:32–34 as teaching that salvation is infallibly applied only to those for whom it is purchased, such that all for whom Christ died will receive the purchased benefit—salvation. In addition to demonstrating the logical error involved in the argument, Ponter also demonstrated from Calvin’s own writings that 1) Calvin himself never made this argument from Rom 8:32–34, and 2) that Calvin himself repeatedly coupled Rom 8:32 (“us all”) with John 3:16 (“world”), and that Calvin clearly interpreted “world” in John 3:16 to refer to all humanity.[9] Calvin taught unlimited atonement and unconditional election.[10]

The great 19^th century Scottish Calvinist, Thomas Chalmers, commented on the phrase “for us all” in v. 32:

“For us all.” The apostle may perhaps be confining his regards in this clause to himself and to his converts, to those of whom he had this evidence that they were the elect of God—even that the gospel had come to them with power and with the Holy Ghost and with much assurance. But, notwithstanding this, we have the authority of other passages for the comfortable truth, that Christ tasted death for every man—and so every man, who hears of the expiation rendered by this death, hath a warrant to rejoice therein; and that He is set forth a propitiation for the sins of the world—and so it is competent for every one in the world, to look unto this propitiation and be at peace; and that He gave Himself a ransom for all to be testified in due time—and so might each of you who hears this testimony, embrace it for himself, and feel the whole charm of his deliverance from guilt and from all its consequences. Christ did not so die for all, as that all do actually receive the gift of salvation. But He so died for all, as that all to whom He is preached have the real and honest offer of salvation…^[11]

Chalmers affirms an unlimited atonement and does not deduce limited atonement from this text.

Charles Hodge, the esteemed 19^th century American Calvinist theologian, likewise does not deduce limited atonement from this text:

There is no restriction or limitation to be put on the word all in this verse, other than that which the context and the analogy of Scripture imposes. God, says Paul, gave up his Son for us all; whether he means all rational creatures, or all men, all those whom he determined thereby to redeem, and whom he had foreknown and predestinated to eternal life, depends on what the Scripture elsewhere teaches on the subject.^[12]

Conclusion

Theologically, limited atonement may be summed up in three propositions:

1. Christ suffered and died only for the sins of the elect.
2. Only the sins of the elect were imputed to Christ.
3. Christ only laid down a redemptive price (ransom) for the elect alone.

Exegetically, it cannot be demonstrated that Paul asserts or implies limited atonement from Rom 8:31–34. Rather, the exegetical evidence indicates that when Paul uses the term “us” in Rom 8:31–34, he is referencing believers and not the concept of the elect as an abstract class including all the elect: unborn, unbelieving, believing, and glorified.

The crux of the issue is this: to squeeze limited atonement from this text one must somehow interpret the “us all” in Rom 8:32 to mean all the elect qua elect. Reformed systematic theologians often speak of “the elect” as the abstract class of all the elect; Scripture never does. Therefore, it is begging the question when one takes the two uses of “us” in Rom 8:32 and “elect” in Rom 8:33 as meaning the abstract class of all the elect qua elect. It is only believers Paul is speaking about. He is giving special assurance to all the elect who are in a believing state.

The limitarian argument proceeds along theses logical lines:

All those “died for” (the elect qua elect) receive all things.

Some (the non-elect) do not receive all things (the lesser gifts).

Therefore, they (the non-elect) are not “died for” (the greatest gift).

As demonstrated above, here is the fallacy: the first “us” in Rom 8:32 is being converted into “all for whom Christ died,” when contextually the “us” refers to believers, not all for whom Christ died.

Those who cite Rom 8:28–34 as supporting limited atonement are equivocating on the meaning of the word “us” in v. 32 and “elect” in v. 33. There is no place in Scripture where “elect” refers to the abstract class of all the elect qua elect (unborn elect, unbelieving elect, believing elect, glorified elect). Every time the word “elect” occurs in the Old Testament, it always refers to Israel as a nation (except when the reference is to the Jesus as the Suffering Servant in Isaiah). Every time the term occurs in the New Testament, it refers only to believers, either individually or as a group. There are no exceptions. (As an aside, the same is true of the use of the phrase “his people,” common in Calvinistic literature, and often used to refer to the elect qua elect. Never in the NT is the term used for the abstract class of the elect,[13] as most contemporary Calvinists tend to use it. “His people” is always a reference to the nation of Israel, as in Matt 1:21.)

Most Calvinists admit there is no single text of Scripture that asserts limited atonement. Limited atonement is a doctrine in search of a text. One simply will not find anywhere in Scripture that states explicitly Christ died only for the sins of some people. Yet there are easily a dozen clear NT texts that affirm and assert unlimited atonement: Mark 10:45; John 1:29; 3:14–16; Rom 5:18–19; 1 Cor 15:3–11; 2 Cor 5:14–21; 1 Tim 2:4–6; 4:10; Titus 2:11–14; Heb 2:9; 9:28; 2 Pet 2:1; and 1 John 2:1–2. There are other texts that implicitly affirm unlimited atonement: Luke 22:20–23; John 17:21, 23; Acts 3:26; 10:34; Rom 1:16; 2:11; 3:21–26; 5:15; 11:32; 14:15; 1 Cor 8:11–12; 2 Pet 3:9; Jude 4; and Rev 22:17.

There is no valid micro-contextual or macro-contextual exegesis connecting texts to demonstrate limited atonement. The attempt to deduce limited atonement from disparate texts combined with theological deductions fails from the start, and especially so in light of the many clear texts that assert unlimited atonement. The burden of proof is on those who assert that Rom 8:32–34 supports limited atonement. Their case is not based on solid exegesis but on invalid logical deductions which hark back mostly to John Owen. Nothing in this text or the surrounding context warrants the interpretation of the “us all” as all the elect as such. This is the only relevant issue at hand.

Moderate Calvinists, along with Arminians and other non-Calvinists, have often demonstrated the flaws in the following presuppositions undergirding the limited atonement argument, many of which are smuggled into discussions of Rom 8:31–34:

1. A limited intent to save only the elect entails a limited atonement.
2. Election entails limited atonement.
3. Penal substitution entails limited atonement.
4. Unlimited atonement entails universalism.
5. Unlimited atonement entails Trinitarian disharmony.
6. Christ’s intercessory ministry for the elect entails limited atonement.
7. The so-called “Double Payment” argument that sins cannot be paid for twice (once by Christ on the cross and again by unbelievers in hell) entails limited atonement.
8. John Owen’s “Triple Choice” (Trilemma) argument entails limited atonement.
9. The “effectual” nature of the atonement entails those for whom Christ died must be saved.[14]

Of these nine, only one is an actual attempt to make a biblical argument (#6). The rest are attempted logical deductions nowhere made in Scripture.

Addendum: James White and Romans 8:31–34

By now it should be clear that for Dr. White to sustain his argument, he must make the “us all” of Rom 8:32 all the elect qua elect in order to get his limited atonement argument to work in the context. As is often his habit, he is equivocating on the term “elect” and contrary to biblical usage is assigning the meaning “all those appointed to eternal life” as the referent.[15] White seems to think that if Paul speaks of all the predestined as such in Rom 8:29–30, then it must follow that “us all” in 8:32 is all the elect qua elect. This is his major hermeneutical, exegetical, logical,[16] and theological mistake.

White’s exegesis of Romans 8:31–34 is thin, lacks nuance, is contextually flawed, and is system-driven. White should ask himself how many modern exegetical commentaries on Romans interpret Rom 8:31–34 as he does? How many modern Calvinist authors of exegetical commentaries treat Rom 8:31–34 as he does? Charles Hodge does not. Moo does not. Schreiner does not. Nor do most others.

Finally, it is unfortunate that Dr. White stooped to uncharitable, demeaning remarks about my motives, scholarly interaction, and/or exegetical competency in his “Dividing Line” broadcasts.[17] I have made no effort to respond to them. White’s blood pressure should not be elevated when, in response to his own criticism of my interpretation of Rom 8:31–34, I dare to comment on his. I have sought to engage the text exegetically, theologically, and logically in an attempt to let the text speak for itself and to demonstrate the validity of my points. Others can judge for themselves whether I am successful or not.

Though said in a derogatory way about me, I can at least concur with Dr. White’s sentiments when he stated: “Sometimes we use our scholarship in a way that is displeasing to the Lord. We have our traditions and we defend our traditions. That’s what you have here. You have to learn to critically analyze what someone says.”[18] I trust my readers will critically analyze my exegesis as well as Dr. White’s to determine who most likely is depending upon tradition, and who best represents the meaning of Scripture on the points at issue in Rom 8:31–34.

---

[1]J. Owen, “The Death of Death in the Death of Christ,” in The Works of John Owen, 16 vols., ed. W. H. Goold (New York: Robert Carter and Brothers, 1852), 10:293. Owen’s logical argument is as follows: 1. Christ’s atonement was for an equivalent number of people as is his intercession. 2. Christ’s intercedes only for the elect alone. 3. Therefore, Christ’s atonement was made for the elect alone.

Owen errs because he is attempting to apply the passage to all the elect as an abstract class. Paul is referring only to believers. Romans 8:32–34 does not address those who have not been justified. Therefore, while the text does prove that Christ died for everyone he intercedes for, it does not prove that he died for no one else. This is Owen’s logical mistake.

[2]J. Gill, A Complete Body of Doctrinal and Practical Divinity (1839 repr. Paris, AR: The Baptist Standard Bearer, 1987), 461; John Gill, The Cause of God and Truth (Grand Rapids: Baker, 1980), 101–102.

[3]J. M. Frame, Systematic Theology: An Introduction to Christian Belief (Phillipsburg, NJ: P&R, 2013), 906.

[4]M. Horton, The Christian Faith: A Systematic Theology for Pilgrims on the Way (Grand Rapids: Zondervan, 2011),518.

[5]T. Nettles, “John Calvin’s Understanding of the Death of Christ,” in Whomever He Wills: A Surprising Display of Sovereign Mercy, ed. Matthew M. Barrett and Thomas J. Nettles (Cape Coral, FL: Founders Press, 2012), 293–315.

[6]Jonathan Gibson, “The Glorious, Indivisible, Trinitarian Work of God in Christ,” in From Heaven He Came and Sought Her, eds. David Gibson and Jonathan Gibson (Wheaton: Crossway, 2013), 340–42.

[7]Ibid., 341–42.

[8]C. Hodge, Systematic Theology, 2:471–72; 555; 557–58. See also W. G. T. Shedd, Dogmatic Theology (Phillipsburg, NJ: Presbyterian & Reformed, 2003), 726.

[9]  David Ponter, “Review Essay (Part Two): John Calvin on the Death of Christ and the Reformation’s Forgotten Doctrine of Universal Vicarious Satisfaction: A Review and Critique of Tom Nettles’ Chapter in Whomever He Wills,” Southwestern Journal of Theology 55.2 (Spring 2013), 263–64. For the Calvin references, see Calvin, Sermons on Psalm 119 (Audubon, NJ: Old Paths Publications, 1996), 133–34; Calvin, Sermons on Melchizedek and Abraham (Willow Street, PA: Old Paths Publications, 2000), 50–51; Calvin, Sermons on Deuteronomy (Edinburgh: Banner of Truth, 1987), 764. In Calvin’s Sermons on Timothy, he speaks of those who “cast away the grace which was purchased for all the world by Jesus Christ” (John Calvin, Sermons on Timothy, trans. Robert White, Sermon 15, 1 Tim 2:5–6 [Edinburgh: Banner of Truth, 2018], 177–78). Calvin also stated concerning those who reject Christ that “Jesus Christ is their redeemer,…“And thus we see now, how men are not partakers of this benefit, which was purchased them by our Lord Jesus Christ,”…(see full citation in Ponter, “Review Essay (Part Two),” 263–64).

[10]For a full listing of Calvin quotes demonstrating conclusively Calvin held to unlimited atonement, see Allen, Extent of the Atonement, 48–96.

[11]Thomas Chalmers, Lectures on the Epistle of Paul to the Romans, Lecture LXII—Romans 8:31, 32 (New York: Robert Carter & Brothers, 1863), 319. For other examples of Chalmers’ assertion of unlimited atonement, see Allen, Extent of the Atonement, 318–22.

[12]C. Hodge, Romans, 288.See also Hodge, Systematic Theology, 2:555 where he affirms Christ’s unlimited atonement. For other examples, see Allen, Extent, 330–32. Notice also how R. Dabney, Hodge’s contemporary, cites Hodge as affirming universal atonement. Dabney, Systematic Theology (1878; repr. Edinburgh: Banner of Truth, 2002), 527.

[13]Titus 2:14 is no exception, since the phrase “a people for his own possession” refers to the church and is patterned on the Old Testament Hebrew phrase so translated in the LXX in Ex 19:5; Deut 7:6; 14:2; and 26:18 in reference to the people of Israel. See T. Desmond Alexander, Exodus, in Apollos Old Testament Commentary, vol. 2, eds. David W. Baker and Gordon J. Wenham (London: Apollos/Downers Grove, IL: InterVarsity, 2017), 367–68.

[14]For detailed analysis of these and other arguments for limited atonement and refutation by moderate Calvinists, Arminians, and other non-Calvinists, see Allen, Extent of the Atonement; and Allen, The Atonement.

[15]In White’s recent second response to Mike Winger, he again used the “union with Christ” idea to refer to all the elect as such being united to Christ in His death. This amounts to decretalizing “union with Christ,” when the biblical phrase is never used of anyone other than those who are in vital, living union with Christ by virtue of salvation. In other words, it is only believers who are properly in union with Christ. The unbelieving elect are never described in this way in Scripture.

Moreover, White interprets the “children of God who are scattered abroad” in John 11:52 as all of the elect as an abstract class, when it is absurd and patently unbiblical to call non-existent people, or unbelievers, (except in a creational sense [Acts 17:28–29]), “children” of God.

White does the same thing with the “you” in John 5:34 (“I don’t receive human testimony, but I say these things so that you may be saved.”) in a 2010 podcast discussion with Michael Brown at minute 25:54–26:11 of “The Calvinism Debate with Michael Brown Pt 2,” (The Dividing Line, 1/28/2010). White takes the “you” as meant for the elect qua elect, totally contrary to Sam Waldron’s solid exegesis of the same text as found in his Modern Exposition of the 1689 Baptist Confession of Faith (Durham, England: Evangelical Press, 1989), 121–22, and more thoroughly and recently in his The Crux of the Free Offer of the Gospel (Greenbriar, AR: Free Grace Press, 2019), 17–24.

When White speaks of all of the elect as such being in “union with Christ,” “saved at the cross,” and are the “children of God” before they believe, he does not seem to realize, and certainly never acknowledges, just how close he is to opening the door to eternal justification. Decretalizing the sense of those words was the pattern of the antinomians and hyper-Calvinists in the 1800’s, as demonstrated and documented in Curt Daniel’s excellent work “Hyper-Calvinism and John Gill” (PhD diss., University of Edinburgh, 1983).

It was also clear in White’s response to Winger that he doesn’t seem to understand the issues of the free offer of the gospel that Winger was raising. I wonder whether he even understands the state of the question with respect to the issue of the free offer of the gospel’s relation to the atonement. One wonders whether he has read much of the literature on the subject, given his responses. The free offer problem for Calvinists who affirm limited atonement does not deal with the limited noetic information of human preachers. That is a given. The free offer problem lies with the fact that the omniscient God Himself is making offers (i.e., proffers) to those whom He knows He has not appointed or effectually purposed to eternal life, and even more problematic, those whom God knows there is no atonement for their sins. In the high (not hyper) Calvinist limited atonement system, the non-elect have no atonement for their sins and are therefore not saveable, and yet the omniscient God Himself (not merely the human preacher) is proffering eternal life to them.” If they are not saveable, it is obvious they are not “offerable” either, as even God cannot genuinely offer something to someone that does not exist for them. Salvation is an impossibility for any person for whom there is no blood atonement (Heb 9:22), and the doctrine of limited atonement, by definition, means a limited satisfaction for sins, such that Christ only died for the sins of the elect.

[16]In a clever but abortive debate tactic to disparage my use of logic, White does not seem cognizant of my point that I am merely demonstrating the logical steps taken by John Owen and all who follow him, including White, with respect to Rom 8:32–34. Unsound logic (but with a valid modus tollens form) is what is behind the limitarian claim that Rom 8:32 supports limited atonement.

[17]As, for example, when he stated: “There is no concern for context. The only concern here is we have to make sure that Southern Baptists don’t believe in particular redemption. That’s the ultimate drive of these pages. Not what is in Scripture” (“The Dividing Line,” Oct. 29, 2019). Throughout my many years of engaging Calvinists who affirm limited atonement, I am occasionally taken aback at how some of them react and respond with borderline, if not outright, hostility. I have come to believe this is sometimes due to insecurity in this sense: Some Calvinists are so wedded to the TULIP system that they view any challenge to it as a direct challenge to Scripture, even God himself. If any part of the TULIP goes, it’s like the blue screen of death for these Calvinists. System crash. You’re no longer an interlocutor or debate partner—you’re the enemy.

[18]See: “The Dividing Line,” Oct. 29, 2019.