Carson on The Atonement

Don Carson has written a great chapter on the love of God and the intent of the cross. I’ve blogged on this before, but have had a number of conversations recently about this topic. So here’s the whole extract. If you’re not a Christian, go watch this video.

The Love of God and the Intent of the Atonement

by D. A. Carson

Here I wish to see if the approaches we have been following with respect to the love of God may shed some light on another area connected with the sovereignty of God – the purpose of the Atonement.

The label “limited atonement” is singularly unfortunate for two reasons. First, it is a defensive, restrictive expression: here is atonement, and then someone wants to limit it. The notion of limiting something as glorious as the Atonement is intrinsically offensive. Second, even when inspected more coolly, “limited atonement” is objectively misleading. Every view of the Atonement “limits” it in some way, save for the view of the unqualified universalist. For example, the Arminian limits the Atonement by regarding it as merely potential for everyone; the Calvinist regards the Atonement as definite and effective (i.e., those for whom Christ died will certainly be saved), but limits this effectiveness to the elect; the Amyraldian limits the Atonement in much the same way as they Arminian, even though the undergirding structures are different.

It may be less prejudicial, therefore, to distinguish general atonement and definite atonement, rather than unlimited atonement and limited atonement. The Arminian (and the Amyraldian, whom I shall lump together for the sake of this discussion) holds that the Atonement is general, i.e., sufficient for all, available to all, on condition of faith; the Calvinist holds that the Atonement is definite, i.e., intended by God to be effective for the elect.

At least part of the argument in favor of definite atonement runs as follows. Let us grant, for the sake of argument, the truth of election. [Footnote 1: If someone denies unconditional election, as an informed Arminian (but not an Amyraldian) would, most Calvinists would want to start further back.] That is one point where this discussion intersects with what was said in the third chapter about God’s sovereignty and his electing love. In that case the question may be framed in this way: When God sent his Son to the cross, did he think of the effect of the cross with respect to his elect differently from the way he thought of the effect of the cross with respect to all others? If one answers negatively, it is very difficult to see that one is really holding to a doctrine of election at all; if one answers positively, then one has veered toward some notion of definite atonement. The definiteness of the Atonement turns rather more on God’s intent in Christ’s cross work than in the mere extent of its significance.

But the issue is not merely one of logic dependent on election. Those who defend definite atonement cite texts. Jesus will save his people from their sins (Matt. 1:21) – not everyone. Christ gave himself “for us,” i.e., for us the people of the new covenant (Tit. 2:14), “to redeem us from all wickedness and to purify for himself a people that are his very own, eager to do what is good.” Moreover, in his death Christ did not merely make adequate provision for the elect, but he actually achieved the desired result (Rom. 5:6-10; Eph. 2:15-16). The Son of Man came to give his life a ransom “for many” (Matt. 20:28; Mark 10:45; cf. Isa. 53:10-12). Christ “loved the church and gave himself up for her” (Eph. 5:25).

The Arminian, however, responds that there are simply too many texts on the other side of the issue. God so loved the world that he gave his Son (John 3:16). Clever exegetical devices that make “the world” a label for referring to the elect are not very convincing. Christ Jesus is the propitiation “for our sins, and not only for ours but also for the sins of the whole world” (1 John 2:2). And much more of the same.

So how shall we forge ahead? The arguments marshaled on both sides are of course more numerous and more sophisticated than I have indicated in this thumbnail sketch. But recall for a moment the outline I provided in the first chapter on the various ways the Bible speaks about the love of God: (1) God’s intra-Trinitarian love, (2) God’s love displayed in his providential care, (3) God’s yearning warning and invitation to all human beings as he invites and commands them to repent and believe, (4) God’s special love towards the elect, and (5) God’s conditional love toward his covenant people as he speaks in the language of discipline. I indicated that if you absolutize any one of these ways in which the Bible speaks of the love of God, you will generate a false system that squeezes out other important things the Bible says, thus finally distorting your vision of God.

In this case, if we adopt the fourth of these ways of talking about God’s love (viz. God’s particular and effective love toward the elect), and insist that this is the only way the Bible speaks of the love of God, then definite atonement is exonerated, but at the cost of other texts that do not easily fit into this mold and at the expense of being unable to say that there is any sense in which God displays a loving, yearning, salvific stance toward the whole world. Further, there could then be no sense in which the Atonement is sufficient for all without exception. Alternatively, if you put all your theological eggs into the third basket and think of God’s love exclusively in terms of open invitation to all human beings, one has excluded not only definite atonement as a theological construct, but also a string of passages that, read most naturally, mean that Jesus Christ did die in some special way for his own people and that God with perfect knowledge of the elect saw Christ’s death with respect to the elect in a different way then he saw Christ’s death with respect to everyone else.

Surely it is best not to introduce disjunctions where God himself has not introduced them. Of one holds that the Atonement is sufficient for all and effective for the elect, then both sets of texts and concerns are accommodated. As far as I can see, a text such as 1 John 2:2 states something about the potential breadth of the Atonement. As I understand the historical context, the proto-gnostic opponents John was facing though of themselves as an ontological elite who enjoyed the inside track with God because of the special insight they had received. [Footnote 2: I have defended this as the background, at some length, in my forthcoming commentary on the Johannine Epistles in the New International Greek Testament Commentary (NIGTC).] But when Jesus Christ died, John rejoins, it was not for the sake of, say, the Jews only or, now, of some group, gnostic or otherwise, that sets itself up as intrinsically superior. Far from it. It was not for our sins only, but also for the sins of the whole world. The context, then, understands this to mean something like “potentially for all without distinction” rather than “effectively for all without exception” – for in the latter case all without exception must surely be saved, and John does not suppose that that will take place. This is in line, then, with passages that speak of God’s love in the third sense listed above. But it is difficult to see why that should rule out the fourth sense in the other passages.

In recent years I have tried to read both primary and secondary sources on the doctrine of the Atonement from Calvin on. [Footnote 3: One of the latest treatments is G. Michael Thomas, The extent of the Atonement: A Dilemma for Reformed Theology from Calvin to the Consensus (1536-1675), Paternoster Biblical and Theological Monographs (Carlisle: Paternoster, 1997).] One of my most forceful impressions is that the categories of the debate gradually shift with time so as to force disjunction where a slightly different bit of question-framing would allow synthesis. Correcting this, I suggest, is one of the useful things we may accomplish from an adequate study of the love of God in holy Scripture. For God is a person. Surely it is unsurprising if the love that characterizes him as a person is manifest in a variety of ways toward other persons. But it is always love, for all that.

I argue, then, that both Arminians and Calvinists should rightly affirm that Christ died for all, in the sense that Christ’s death was sufficient for all and that Scripture portrays God as inviting, commanding, and desiring the salvation of all, out of love (in the third sense developed in the first chapter). Further, all Christians ought also to confess that, in a slightly different sense, Christ Jesus, in the intent of God, died effectively for the elect alone, in line with the way the Bible speaks of God’s special selecting love for the elect (in the fourth sense developed in the first chapter).

Pastorally, there are many important implications. I mention only two.

(1) This approach, I content, must surely come as a relief to young preachers in the Reformed tradition who hunger to preach the Gospel effectively but who do not know how far they can go in saying things such as “God loves you” to unbelievers. When I have preached or lectured in Reformed circles, I have often been asked the question, “Do you feel free to tell unbelievers that God loves them?” No doubt the question is put to me because I still do a fair bit of evangelism, and people want models. Historically, Reformed theology at its best has never been slow in evangelism. Ask George Whitefield, for instance, or virtually all the main lights in the Southern Baptist Convention until the end of the last century. From what I have already said, it is obvious that I have no hesitation in answering this question from young Reformed preachers affirmatively: Of course I tell the unconverted that God loves them.

Not for a moment am I suggesting that when one preaches evangelistically, one ought to retreat to passages of the third type (above), holding back on the fourth type until after a person is converted. There is something sleazy about that sort of approach. Certainly it is possible to preach evangelistically when dealing with a passage that explicitly teaches election. Spurgeon did this sort of thing regularly. But I am saying that, provided there is an honest commitment to preaching the whole counsel of God, preachers in the Reformed tradition should not hesitate for an instant to declare the love of God for a lost world, for lost individuals. The Bible’s ways of speaking about the love of God are comprehensive enough not only to permit this but to mandate it. [Footnote 4: Cf. somewhat similar reflections by Hywel R. Jones, “Is God Love?” in Banner of Truth Magazine 412 (January 1998), 10-16.]

(2) At the same time, to preserve the notion of particular redemption proves pastorally important for many reasons. If Christ died for all people with exactly the same intent, as measured on any axis, then it is surely impossible to avoid the conclusion that the ultimate distinguishing mark between those who are saved and those who are not is their own will. That is surely ground for boasting. This argument does not charge the Arminian with no understanding of grace. After all, the Arminian believes that the cross is the ground of the Christian’s acceptance before God; the choice to believe is not in any sense the ground. Still, this view of grace surely requires the conclusion that the ultimate distinction between the believer and the unbeliever lies, finally, in the human beings themselves. That entails an understanding of grace quite different, and in my view far more limited, than the view that traces the ultimate distinction back to the purposes of God, including his purposes in the cross. The pastoral implications are many and obvious.

D. A. Carson, The Difficult Doctrine of the Love of God (IVP, 2000), 84-90.