The Oxford divine relies too much on a “logic of inevitability.”
The extent of the atonement has a been a topic of study for me for years. On and off I have read various defenses of the definite atonement position, as well as those for unlimited atonement. While the 2013 publication of From Heaven He Came and Sought Her may challenge this, for a very long time many people pointed to John Owen’s Death of Death in the Death of Christ as the unanswerable defense of definite atonement. However, I think that assessment is exaggerated. I’ve read Owen, and marked up my copy pretty well, but the most powerful arguments against Owen may have come from an unlikely source: A master’s degree thesis at Reformed Theological Seminary in Jackson, Mississippi.
Neil Andrew Chambers wrote his thesis back in 1998, and while it has never been published, I did acquire a copy and read through it carefully. As an aside, Chamber’s work highlights the fact that there has always been a diversity of views within the Reformed tradition. One need not affirm limited atonement to be Reformed. I’m aware that some will disagree with this, or say that one is not truly Reformed without upholding all five points of Calvinism. But the number of proponents of unlimited atonement who are still within the broadly Reformed tradition argues against this. For an exhaustive survey of these theologians, I recommend David L. Allen’s The Extent of the Atonement: A Historical and Critical Review.
What does Chambers say that undercuts Owen’s position? As I’ve written a few blog posts about this, this will be a summary with links to some of those posts.
To begin, much has been made about the “double payment” argument. If Jesus’ death pays for the sins of all, and some of these are not saved, then we cannot say their sins have been paid for. An offshoot of this is the question, “is unbelief a sin? If it is, and Christ paid for the sins of all, why aren’t all saved?” Chambers asks, “If unbelief is a sin atoned for, then why are the elect, those for whom Christ died, not saved at the cross regardless of their belief or unbelief? If unbelief is just like other sins, what is its relationship to the unforgivable sin, and are there ‘types’ of unbelief which require a more nuanced presentation of the significance of unbelief in relation to the atonement?” (More on this topic below.)
The double payment argument is not as it seems. It is instead a claim that extent and atonement cannot be separated. If the double payment argument is valid, the elect would be saved at the cross, with no application of faith necessary. The elect are among the sons of disobedience, under God’s wrath until they believe. If their sins are paid for, how is this possible? The answer is that there is a distinction between extent and application—a distinction even advocates of definite atonement admit. The double payment argument ends up as a rhetorical device more than a real proof.
Secondly, the definition of “world.” Owen insists that the word “world” must be read contextually and that in John 3:16, to cite the most famous example, the word indicates the elect. Chambers shows that this cannot be exegetically demonstrated. The world, particularly in John’s theology is a quality of rebellion and opposition to God, it is not a numerical quantity. By viewing it as a collection of people (the elect) Owen makes a category mistake, due to his commitment that the extent of the atonement and the application of the atonement can never vary.
Thirdly, Owen bases much of his argument for definite atonement on the covenant of redemption, which for him is a pact that the Father and the Son entered into prior to time. The Father would give the Son and the Son would agree to die in the stead of the elect only. Yet such a pact is by no means clear in Scripture. It can only be arrived at by inference. Chambers and others have pointed out some implications of a pre-temporal covenant such as Owen relies on. A covenant implies a change from what previously prevailed to a new, agreed upon state. Does this imply change in the Godhead? Moreover, there are no examples in Scripture of a covenant within the Godhead. As presented by Owen, the covenant of redemption introduces obligation. That is, the elect are in fact owed salvation due to the covenant, and the element of grace is minimized if not excised from redemption.
Fourthly, and as part of the covenant of redemption, Owen believes that faith is purchased for the elect, and that this faith will infallibly be transferred to them. “For Owen Christ’s death must be directly causally responsible for the faith of those being saved.” Chambers shows that this relies on the logic Owen employs rather than any passages of Scripture. No verse of Scripture—even those that speak of a gift—can support a view that says faith is purchased at the cross. Owen conceives of faith as a “thing,” a transferable object, but this mischaracterizes faith in the New Testament as something other than a response to the gospel. In saying it is a response to the gospel, some fear that this somehow puts man in charge. Scripture does not infer nor state this.
This is also where the view that unbelief is a sin atoned for at the cross runs into challenges. It exposes Owen’s commercialistic and quantitative mindset. In his view, sins are a set number of acts that are atoned for, and only for the elect. The transactional nature of Owen’s view misrepresents guilt in the language of debt. The bookkeeping metaphor is prominent in Owen, but Chambers notes that debt language is used infrequently in the New Testament, and where used, that debt is canceled rather than paid. “It is τοῖς δόγμασιν [the ordinances, Col 2:14], our failure to keep the law, which is unquantifiable due to the nature of the demand. Further, this ‘certificate of indebtedness’ is not paid but canceled, by being destroyed, with no suggestion of a payment of a debt to anyone.” Over-reliance on commercial language distorts Owen’s claims.
Finally, Chambers questions Owen’s whole methodology of beginning with logic, and then moving to exegesis of the passage to demonstrate the truth of his logic. In the first two books of The Death of Death, Owen presents the case portraying definite atonement as unavoidable, i.e., if not all are saved, then it must mean that provision was not made for all, for had provision been made, they would be saved. “Between both these, end and means, there is this relation, that (though in sundry kinds) they are mutually causes one of another. The end is the first, principal, moving cause of the whole.” If one accepts the logic of means and ends that Owen stipulates, it lends this air of inevitability to his conclusion. As Chambers notes, “The argument is really won or lost in Books One and Two, for acceptance of Owen’s presuppositions here makes his conclusions and exegesis inevitable, as Owen himself suggested at the conclusion of Book Two.”
Chambers observes that Owen affirms “with God outcome and intention are one. This is a crucial observation for it allows him to arrive at God’s intention for the cross directly from its outcome, and every statement in scriptures concerning the result of Christ’s work becomes a statement of the intention of that work, an argument employed repeatedly in Book III where Owen will move from Christ’s death being spoken of as e.g. reconciling or sanctifying to the conclusion that it is only ever intended for those who are actually reconciled or sanctified.” Yet in passage after passage, Chambers shows it to be other than Owen affirms.
Whether one thinks definite atonement is a true depiction of what the Scriptures teach on the extent of the atonement, it is not the case that John Owen has provided an airtight argument of this. The only way to affirm this is to accept Owen’s presuppositions.
 Neil Chambers, “A Critical Examination of John Owen’s Argument for Limited Atonement in ‘The Death of Death in the Death of Christ,’” (ThM thesis, Reformed Theological Seminary, 1998), 40.
 Chambers, 35.
 Chambers, 253-254.
 John Owen, The Death of Death in the Death of Christ (Edinburgh/Carlisle, Banner of Truth, 1967), 48-49.
 Chambers, 31.
 Chambers, 36.