Horton on the Atonement

Michael Horton’s presentation

In chapter 4, Michael Horton addresses the third point of  “TULIP,” which he prefers to call “particular redemption”` rather than “limited atonement,” arguing that it is “specific or definite in its intention and scope” (80.) He begins with a discussion of “the nature and effects of Christ’s work on the cross,” positing that “penal substitution has always been at the heart of Reformed (as other) accounts of Christ’s redemptive work” (81). But Horton describes and assesses other understandings of the atonement – recapitulation, Christ’s victory over the powers, satisfaction of divine honor, moral influence and the governmental theories. He reports on a common contemporary preference for subjective atonement theories, which focus on the change that the cross effects in us, rather than  on the satisfaction of God’s justice and righteousness. Instead, Horton emphasizes the objective work of Christ, while affirming the truth he sees in the various theories, each of which has some biblical ground, but needs to be carefully formulated. What Calvinism points out, however, is that “none of these other aspects can actually be realized unless Christ’s work is first of all a vicarious substitution addressing the objective problem of guilt before a holy God” (90).

Horton identifies three options in regard to the extent of Christ’s work: (1) Christ’s death redeemed every person, (2) Christ’s death made the salvation of every person possible, and (3) Christ redeemed all the elect. As representatives of the second option, Horton identifies the Dutch Remonstrants (Arminians) and what he calls “a mediating position between the orthodox Calvinism defined by the Synod of Dort in 1618-19 and Arminianism,” usually referred to as “hypothetical universalism” or “Amyraldianism.” As examples of evangelical Protestants who fit in this category, Lewis Sperry Chafer and Robert Lightner are cited (92-93).

Horton views the third alternative, “that Christ died for all the sins of the elect, thereby redeeming them at the cross,” to be the position affirmed by Dort. But he acknowledges that Dort followed the common formula, “sufficient for the whole world but efficient for the elect alone,” in chap. II, art. 3. There the Reformed Church stated that Christ’s death is “of infinite worth and value, abundantly sufficient to expiate the sins of the whole world,” but Horton hears the synod to be asserting that “Christ objectively and effectively bore the sins of the elect alone” (92).

Naturally, Horton defends what he considers to be the position affirmed by Dort, and he makes a case for “particular redemption” (92-96). His arguments are that:

(1) “this view maintains that Christ’s death actually saves” (92-93),

(2) “this view emphasizes the relationship between the Trinity and redemption” (93-95),

(3) “this view places the focus entirely on Christ rather than on the believer” (95-96).

Finally, Horton responds to 3 objections. To the objection that “the NT teaches clearly that Christ died for the world” (96), he proposes that “the question is never the sufficiency of Christ’s work but the purpose of the triune God” (97). To those who object that “unbelievers are condemned not for their sins but for their unbelief in Christ,” Horton proposes that “unbelief is simply one of the sins for which people will be condemned on the last day” (97). Finally, Horton deals with the concern that it would be unjust for God to condemn people if Christ didn’t die for them. He asserts “God’s sovereign freedom to choose whom he will out of a mass of condemned humanity,” but he also observes that “the death of Christ is sufficient for everyone,” so that “no one is left out except those who refuse this gift” (98).

My Reflections

I commend Michael Horton’s description of what Christ accomplished in his atoning work. He is aware of the numerous contemporary protests regarding penal substitution, but he wisely defends its truth and fundamental importance, while refusing to consider it the only or complete truth about the atonement, as taught in Scripture. There is also much about Horton`s case for particular redemption which I affirm. When he gets to his discussion of the “extent of Christ’s work,” however, I think that his proposal is much less helpful and, indeed, is confusing and misleading in a number of significant ways.

Extent and effect: Calvinism and Arminianism

The “L” of TULIP is the point most disputed among Calvinists these days, and I did not find Horton’s chapter as helpful as I had hoped it would be in addressing the in-house differences. On the first page of the chapter, he states that “Calvinists believe that [the atonement] is limited (or definite) in its extent, but unlimited in its nature and efficacy: Christ’s death actually saved the elect.” By contrast, he purports, “Arminians believe it is unlimited in its extent, but limited in its nature or efficacy: Christ’s death makes possible the salvation of everyone, but does not actually save any” (80).

The term “extent” is ambiguous and hence problematic. Classic monergist soteriology spoke of Christ’s atoning work as sufficient for all but efficient only for the elect. Clearly, the “efficacy,” which Horton states to be “unlimited” in Calvinism, is not the same thing as that which monergists had classically called “efficient,” for Horton’s own reading of Dort was that it affirmed the atonement to be efficient only for the elect. If we are to speak of “extent,” therefore, we do best to distinguish between the extent of its sufficiency (for all) and the extent of its efficiency, that is, efficacy (for the elect). Ironically, from the classic Arminian perspective of the election of individuals to salvation, based upon foreseen faith, even Arminians can affirm this classic formula. The critical difference between Calvinists and Arminians is therefore regarding the conditionality/unconditionality of election, not regarding the “extent” of the atonement.

Consequently, if we use the term “extent” less ambiguously, Horton’s statement that “Arminians believe it is unlimited in its extent, but limited in its nature or efficacy,” is inaccurate. Nor is it correct to say that Arminians believe that: “Christ’s death makes possible the salvation of everyone, but does not actually save any.” In the second of the “Five Arminian Articles” (1610), they asserted that Christ “died for all men and for every man, so that he has obtained for them all, by his death on the cross, redemption and the forgiveness of sins” (Art. II). In other words, Christ’s death was sufficient for the redemption of all people, just as Dort and the classical tradition had asserted. Had everyone been finally saved, no further atonement would have been necessary than was accomplished by Christ. Dort concurred: “The death of the Son of God is the only and most perfect sacrifice and satisfaction for sin; is of infinite worth and value, abundantly sufficient to expiate the sins of the whole world” (Chap. II, Art. 3, as Horton himself pointed out [92]). Here, both the Arminians and the orthodox of Dort appear to have agreed.

Having cited Dort in regard to the sufficiency of Christ’s death for all, Horton believes Dort to have asserted that “Christ objectively and effectively bore the sins of the elect alone” (92). At that point, I think confusion sets in, for Dort did not speak in that way. In fact, when Dort supports its affirmation of the sufficiency of Christ’s death “to expiate the sins of the whole world” (II, 4), it appeals to the “infinite value and dignity” of the person of Christ, and to the fact that his death “was attended with a sense of the wrath and curse of God due to us for sin.” It is precisely this universal sufficiency that grounds “the promise of the gospel,” namely, “that whosoever believeth in Christ crucified shall not perish, but have everlasting life.” Consequently, “this promise, together with the command to repent and believe, ought to be declared and published to all nations, and to all persons promiscuously and without distinction” (II, 5).

Thus far, I see the Arminians and the orthodox at Dort to be agreed. Christ’s death made expiation for the sins of the whole world, but only those who believe have everlasting life (Dort, II, 5, which concords with Arminian Article II, where Jn 3:16 was also cited). If we speak in the same manner as both the Arminian Remonstrants and the orthodox at Dort did, Arminian Article II is not problematic, and it is clearly supported by Dort, II, 3-6.

Increasingly, I see the “L” of TULIP to be detrimental. It has confused and misled both Arminians and Calvinists in their important conversation about the atonement. At Dort, they agreed, but TULIP has made them appear to be in disagreement. The “L” is also unhelpful to the conversation between Calvinists. Much has been made about a difference between 4 and 5 point Calvinists, the former purportedly rejecting the “L” of TULIP. I do not deny that there are now differences among us Calvinists in regard to how we should speak about the atonement, but these derive from TULIP not from Dort! Much progress can be made if we start with that recognition.

I am arguing that the dispute between Arminians and Calvinists is not about the extent of the atonement. Both the Remonstrants and the orthodox affirmed that the death of Christ was unlimited in its sufficiency but limited in its efficiency/efficacy. Both were also agreed that faith is necessary for the death of Christ to bring salvation to an individual. Furthermore, both were agreed that salvation, including the faith that is necessary for its appropriation, is of God’s grace, not of human merit. That point is extensively made in Articles III and IV of the Remonstrance. Article III stated that because humans are in “the state of apostasy and sin,” they are unable to think, will, or do “any thing that is truly good (such as saving Faith eminently is).” [All my citations of the Remonstrant articles and the Canons of Dort are from Schaff, Vol. III.] They must be “born again of God in Christ, through his Holy Spirit, and renewed in understanding, inclination, or will, and all [their] powers, in order that [they] may rightly understand, think, will, and effect what is truly good” (citing Jn 15:5). Article IV goes on to emphasize the comprehensiveness of grace in human salvation, it is “prevenient or assisting, awakening, following and co-operative,” so that “all good deeds or movements, that can be conceived, must be ascribed to the grace of God in Christ.” Thus far, as a Calvinist, I must say a hearty “Amen” to fellow Christians who are orthodox in their Arminianism.

The critical difference between Arminians and Calvinists arises when we explain why it is that some from among the “all” for whose sins Christ made expiation believe to salvation, and others do not. The Remonstrants put that issue on the table, later in Article IV, in their denial that the operation of God’s grace is “irresistible.” The orthodox at Dort asserted that the unbelief of many who hear the gospel “is not owing to any defect or inusfficiency in the sacrifice offered by Christ upon the cross, but is wholly to be imputed to themselves” ((II, 6), and I can hear a loud “Amen” from Arminians. Article 7 of the second chapter made the same point that the Remonstrants had made, though in many fewer words, that those who “truly believe and are delivered and saved from sin and destruction through the death of Christ, are indebted for this benefit solely to the grace of God given them in Christ from everlasting, and not to any merit of their own.” It is in Article 8 of that second chapter that Dort took issue with the point made only briefly by the Remonstrants, in their Article IV. The Remonstrance had insisted that God’s grace was not “irresistible,” but Dort asserted that it “was the sovereign counsel and most gracious will and purpose of God the Father, that the quickening and saving efficacy of the most precious death of his Son should extend to all the elect, for bestowing upon them alone the gift of justifying faith, thereby to bring them infallibly to salvation: that is, it was the will of God, that Christ by the blood of the cross, whereby he confirmed the new covenant, should effectually redeem out of every people, tribe, nation, and language, all those, and those only, who were from eternity chosen to salvation, and given to him by the Father; that he should confer upon them the faith, which, together with all the other saving gifts of the Holy Spirit, he purchased for them by his death.”

Here we can see clearly, I suggest: (1) that the “L” of TULIP is a troublemaker that should be laid aside, and (2) that in regard to the atonement, the Arminian Remonstrants and orthodox Calvinists have no dispute. We both affirm heartily the universal sufficiency of Christ’s atoning death, and its efficacy only for those who believe. Where we diverge sharply is in regard to the role of God’s grace in that necessary faith. Thus, the dispute arises at the “I” of TULIP, but we can shake hands in regard to the “L” which is seen to serve no helpful purpose.

Added on June 12: On further thought, I would rather say that a point concerning the intent of Christ in the atonement does deserve to be made, but this is not well done by the “L” of  TULIP. The significant truth worth stating is in regard to the limited intention of Christ, that the saving benefits of his universally sufficient sacrifice should be applied to those whom the Father chose in him, by the Holy Spirit’s effectual calling of these people. The high Calvinism of what has been called “5 point” Calvinism, and classic Arminianism, have in common their belief in a single intent for the atonement. Classic-moderate Calvinists, on the other hand, recognize that Scripture teaches a broader and narrower intent in God’s gracious work, thus indicating a double intent. 

 “Four-point Calvinism”

What I have said in the point above has spelled out further my previous (http://thoughtstheological.com/for-whom-did-christ-die/) assertion, that Dort affirmed two divine intentions in regard to Christ’s atoning work, (1) an intent that he should make satisfaction for sin that is universally sufficient (II, 3), and (2) that he should effectively save all those whom the Father gives to him, and to whom he gives justifying faith (II, 8).

I am now convinced that the purported distinction between 4 and 5 point Calvinism is murky and that its fuzziness derives from the unhelpfulness of the “L” in the formula. It begins to look to me like a distinction between those who affirm a double intent and those who affirm only a single intent, for Christ’s atoning work. What particularly muddies the water, I suggest, is that an affirmation of single intent, identified with affirmation  of the “L,” is presented as acceptance of the position of Dort concerning the atonement. Those who affirm a double intent, are generally prepared to state that they reject the “L,” holding only to the other four points, and to grant that, on this point, they depart from the position of Dort. As my work in this post and the earlier one (linked above) makes clear, I hope, a double intent best represents the position put forward in the second chapter (or head) of the Canons, and the “L” of TULIP has laid down a very unhelpful smoke screen.

Amyraldianism

A major problem arises in Horton’s 4^th chapter when he lays out three options concerning “the extent of Christ’s work,” and lumps together Arminianism and Amyraldianism, which he identifies as “a mediating position” between the “orthodox Calvinism defined by” Dort and Arminianism (91). Horton defined this second option as the belief “that Christ died to make salvation of every person possible” (91). As I have demonstrated above at some length, this is a conviction affirmed by both the Arminians and the orthodox Calvinists, so it is highly confusing to state the option in this way. Amyraut saw a double intent in the atonement, and I have argued that this is correct. Where he went astray, I think, is in his decretal formulation of this double intent, as a distinction between two logical stages in the divine decree. That speculation is neither helpful nor necessary, I propose. What should be obvious by now, however, is that I think the double intent which Amyraut discerned in God’s will for Christ’s atoning work was precisely what Dort defined, in its reaffirmation of the death of Christ as universally sufficient but efficient only for the elect.

Horton states that “many evangelical Protestants hold to either an Arminian or Amyraldian view, in either case agreeing with the position expressed by Lewis Sperry Chafer: ‘Christ’s death does not save either actually or potentially; rather it makes all men savable’” (92). As I read the position of the Arminian Remonstrants, that is decidedly not what Arminianism teaches, nor does it represent Amyraldianism. It looks to me to be an innovation of a different kind. I’ve heard Chafer identified as a “4-point Calvinist,” but if this is what that means, I see it as a departure from Calvinism that is even greater than that found in Arminianism. Horton proceeds to quote Robert Lightner who “says that he rejects the Calvinist view ‘that the work of Christ on the cross was effective in and of itself’” (92). Once again, we have a statement that represents neither Arminianism (as I have demonstrated above) nor Calvinism. Remonstrants, the orthodox at Dort, and Amyraldians, all assert that Christ’s death effectively saves all who believe. If Lightner is rejecting the view that Christ’s death saves people without faith, I don’t know whose view he is rejecting, but it is not the Calvinist view any more than it is the Arminian view. Clearly, there are some very strange ideas wandering around in evangelical Protestantism, and they have not descended from either Arminius or Amyraut!

I see an unresolved tension in Horton’s case for particular redemption, and it is rooted in a failure to consistently follow Dort’s affirmation of the universal sufficiency of Christ’s atoning work. Confusion arises frequently because of the way that sufficiency and efficacy have been described, as I pointed out above. The outcome is that Horton effectively denies the universal sufficiency of Christ’s atonement, despite his robust affirmation of it in a number of clear instances. I’m puzzled, for instance, by a statement like this: “Scripture nowhere teaches that Christ came into the world to make salvation possible, much less that it becomes actual because of faith in Christ. This would make the instrument of receiving salvation (viz., faith) the basis of salvation” (92). I can see how this may be addressed to Chafer and Lightner, as per Horton’s quotes. But I certainly hope that he wasn’t aiming it at Arminianism or Amyraldianism, though I fear such may have been the case. For both those perspectives, as for the orthodox at Dort, it looks clear to me that faith is not the basis of salvation, but it is a necessary instrument, thus faith does actuate the saving benefit of Christ’s death in the life of an individual. Without the Spirit`s work in an individual`s life, the Son`s work does not bring about salvation. The Son`s limited intent is brought to realization by the Spirit, in accordance with the Father`s eternal election.

Later, Horton writes: “All for whom Christ died have been redeemed, reconciled, and saved from the wrath of God`(93). Again there is ambiguity because of the way sufficiency and efficiency have been distinguished on some occasions but confused on others. Not all for whom Christ`s death was sufficient for redemption are actually redeemed, only those to whom it is applied through faith. Defenders of the “L” in TULIP (including me, in Who Can Be Saved?) have often taken great care to explain how passages that speak of Christ’s death for all should be understood in a more limited sense. On occasion, this may be justified exegetically, but a serious affirmation of the universal sufficiency of Christ’s atoning work, in the way Dort affirmed it, makes this sometimes tortuous exegesis unnecessary, and thereby removes a needless offense to Arminians. It is certainly true, as Horton points out, that Christ “’has . . . accomplished redemption for his people’ (Luke 1:68 NASB, emphasis added)” (95), but we must not forget that the redemption he accomplished was sufficient for all, though only efficacious for those to whom God graciously gives justifying faith. Furthermore, it was efficacious in other ways for the others, including common grace and the forgiveness of unintended sin, for instance. The critical thing that all Calvinists (whether dubbed 4 or 5 point) affirm, contrary to Arminianism, is that Christ’s death does effectively bring about the eternal salvation of all those whom the Son intended to save, namely, those who were given to him by the Father. In this particular intent, there was a limitation which is denied by Arminians, but which is essential to preservation of the inner-Trinitarian unity.

I anticipate that what I have laid out in this post will sound strange to many. I would not have spoken this way myself, a year ago. At present, this looks to me both clear and correct, but I remain open to further conversation, both with Arminians who are not used to hearing this from Calvinists, and from traditional 5 point Calvinists who are not used to thinking about Dort in these terms. I hope that the ongoing conversation which I seek will be mutually beneficial, as we work together on ways to represent most accurately the teaching of Scripture concerning the wonder of what Christ did for us as Savior of the world, who eternally saves those who praise and glorify him as such, by his grace.

Earlier posts in this series: Part 1, Part 2.