This year marks the 400th anniversary of the Synod of Dort. Thus, I was not surprised to come across another article on its Canons. I am always one who is encouraged to find popular Reformed authors defending the Reformed confessions of the 16th and 17th centuries. In fact, if there is one document which defines the boundaries of what should and should not pass for Reformed orthodoxy – at least on the famous “Five Points of Calvinism” – surely that would be this international Synod and its Canons, which were approved not only by the Dutch Reformed, but also by the various German, Swiss, and British delegations attending the Synod.
But I was concerned, as I read, at how the main points the article asserted departed from a good reading of its subject matter; and I was surprised to find, as I got to the end, that its author was Dr. R. Scott Clark, Professor of Church History and Historical Theology at Westminster Seminary California. The article, which gives a summary of hypothetical universalism and rules it incompatible with the theology of the Canons, contains historical inaccuracies, clear misinterpretation, and a heavy reliance on an encyclopedia article rather than primary sources.
Before examining the article, an important caveat is worth making. I readily admit at the outset that I am not an expert on all forms of early modern hypothetical universalism. My nearly completed dissertation is on John Davenant’s hypothetical universalism; and so, because I have not read them carefully, I will generally avoid making any specific claims as to the theology of Moises Amyraut or the other so-called French Amyraldians. My knowledge about what they taught is largely based on their critics, such as Francis Turretin. As a historian, I have always been distrustful of such sources. If I want to know what, say, the Roman Catholic Robert Bellarmine taught, good historiography demands that I read him carefully – even if that means reading him in Latin – rather than accept his opponents’ depictions of him. With that caveat in mind, I now turn to the article itself.
Dr. Clark begins his article with a brief history of Franciscus Gomarus leading up to his time at the French Reformed academy at Saumur, where Gomarus was succeeded by John Cameron, whom Clark asserts made a “sharp departure” from Gomarus’s theology. By “sharp departure,” Clark has in mind two points. First, he identifies Cameron’s doctrine of regeneration, which taught that because the human will always and necessarily follows the last judgment of the practical intellect (in accordance with general faculty psychology of the day), spiritual regeneration does not directly work upon the will, but on the intellect. Seeing that “the will depends upon the intellect, with the renovation of the intellect, it [i.e., the renovation of the intellect] produces the renovation of the will.” 1 Hence, God’s action of regeneration “does not attach itself only to the human intellect, but reaches (pertingit) to the will itself,” yet mediately so, as the intellect is the instrument by which that action of regenerating the will takes place. 2 Whether such an opinion was a departure from “the way that the Augustinian and Reformed theologians and churches had hitherto taught,” as Clark claims, is perhaps subject to further inquiry; that would demand a fairly broad survey of the topic amongst a whole host of earlier Reformed and Augustinian theologians (the latter, including perhaps, early modern Roman Catholic theologians). 3
Dr. Clark’s real concern with Cameron’s “sharp departure” from Gomarus, however, is Cameron’s hypothetical universalism. It is at this point in Clark’s essay where things get especially messy. Clark cites an encyclopedia article – admittedly written by the very-learned B. B. Warfield – on Calvinism. 4 Using this encyclopedia article, Clark summarizes the teachings of the hypothetical universalists. Interestingly, Dr. Clark does not emphasize that Warfield contextualizes this doctrinal system as that which “retained a position within its [i.e., “the Calvinistic system”] limits” and takes its “start from a fundamental agreement in the principles which govern the [Calvinistic] system.” 5 We will return to this point towards the end of this critique. For now, let us notice what Clark does quote from Warfield. Warfield summarizes hypothetical universalism in terms of a particular ordo decretorum, that is, an order of God’s decrees, where God is said (a priori) to have ordained the Fall and the redemption purchased by Christ previous to his decree of election. This particular ordering of God’s decree, according to Warfield, entails the belief that “Christ died to make satisfaction for the sins of all men without exception if—if, that is, they believe: but that, foreseeing that none would believe, God elected some to be granted faith through the effectual operation of the Holy Spirit.” 6
Dr. Clark, on the basis of this encyclopedia article, then gathers from Warfield’s summary that “the Hypothetical Universalists removed the limit from the atonement and pushed it back a step to the application of redemption.” 7 This is far too general of a statement. Does Dr. Clark have in mind all hypothetical universalists? Some? Or only the ones mentioned in the encyclopedia article? 8 He does not tell us. He does add, however, that because the hypothetical universalists removed the limit from the atonement, this allowed them to say that Christ died for each and every human being. Clark correctly notes that this language is used by the Remonstrants, but he fails to mention the more salient point that the Contra-Remonstrants themselves, at the Hague Conference in 1611, explicitly admitted that Christ died for all human beings (understood individually, pro singulis hominibus), as regards the sufficient power and merit of Christ’s death to take away the sins of all human beings. 9
This observation highlights the fact that there is more at play than just that language of “Christ dying for all.” The debates in the early modern period cannot be simply boiled down to mere linguistic shibboleths. In fact, one of the most prominent shibboleths in the early modern period, the Lombardian formula (“Christ died for all, sufficiently; Christ died for the elect, effectually”) engendered many different reactions from the Reformed: some defended it, some denied it, and some modified it. 10 More on that to come. Dr. Clark proposes two motivations for wanting to affirm that Christ died for all: (1) “to appease” the Lutherans and (2) to make better sense of the Scriptures. However, there is no hint at all of the first motivation in either Bishop James Ussher’s famous two letters where he lays out and defends his version of hypothetical universalism, nor in the whole of Davenant’s treatise De Morte Christi. 11
Dr. Clark next turns his attention to Davenant’s hypothetical universalism. Clark writes:
Well before the Amyraldian Controversy broke out in Saumur, however, there were Reformed theologians, including at least two delegates to the Synod of Dort, who were toying with alternative language concerning the atonement. In his 1627 Dissertation on the Death of Christ John Davenant (1572–1641) argued for a kind of hypothetical universalism.
There are some historical infelicities here. Dr. Clark wrongly attaches the date 1627 to Davenant’s De Morte Christ. It was first published posthumously in 1650, and was subsequently republished in 1683. 12 Later, in the 19th century, the treatise was then translated into English and appended to the English translation of Davenant’s commentary on Colossians. 13 Indeed, in a November 1628 letter to Samuel Ward, Davenant does say that De Morte Christi is finished (terminus ad quem), but that is not in 1627. 14 As best I can tell, perhaps Dr. Clark was following the chapter title of Davenant’s 19th century biographer, Morris Fuller. 15
To more substantial concerns, we might ask what it is that previous to Dort constitutes “normal Reformed language concerning the atonement”? Does the language of Article 31 of the 39 Articles count? 16 Q&A 37 of the Heidelberg Catechism, where it says that Christ sustained the wrath of God against the sin of the whole human race? 17 Is it the Lombardian formula? 18 If the latter, then it was very many of the non-hypothetical universalists among the Reformed churches who were “toying with alternative language concerning the atonement.” Theodore Bèza, Johann Piscator, Heinrich Alting, and Gomarus, just to name some, all self-consciously began not merely “toying with” but in fact denying the ancient distinction 19 of Christ dying for all sufficiently, and for the elect effectually. 20 For his part, Davenant spends a significant amount of time explaining and defending the classic distinction over and against some of his fellow Reformed. 21 To be sure, Lutherans, Remonstrants, and others were also defending the Lombardian formula as well. 22 Notably, however, their interpretation of the distinction is different than Davenant’s. 23 Regardless, Davenant was not, at least as he saw it, “toying” (Clark’s word) with theological language. He was committed to defending classical theological language, not inventing new theology. Indeed, John Owen himself realized Davenant’s commitment to Augustinian modes of speech when it came to the atonement:
The whole of those persuasions [found in Davenant’s De Morte Christi], I confess, which he endeavoureth in them to maintain, is suited to the expressions of sundry learned men, as Austin [i.e., Augustine], Hilary, Fulgentius, Prosper, who in their generations deserved exceeding well of the church of God … 24
Does Owen represent Davenant as someone toying with alternative language regarding the atonement? Maybe Clark thinks that the early Reformers themselves toyed with alternative language (distinct from Augustine, Prosper, Hilary, and Fulgentius!) regarding the atonement. Ironically, it is Davenant who argues against those toying with the received theological language, i.e., those Reformed theologians who were denying, either explicitly or implicitly, that Christ died for all sufficiently.
Dr. Clark proffers his strongest criticism for the perceived ambiguity in Davenant’s treatise. Clark writes: “I say ‘some version’ [of hypothetical universalism] because even in Davenant’s Dissertation it [sic] thesis is not always entirely clear. As a teacher, should a student have submitted this as a Master’s Thesis I should have returned it with a demand that he state unequivocally his position and clearly contrast it with other views.”
This rhetorical framing is meant to suggest a weakness in Davenant’s scholarship, that his work would not pass muster in Dr. Clark’s classes. But it may tell us more about Dr. Clark and contemporary academic expectations than about Dr. Davenant, former Lady Margaret Professor of Theology at Cambridge University – the same post held by Erasmus when he was at Cambridge. Let us put this in the most charitable way possible. Davenant’s treatise was written for scholars, not amateurs. It was a scholastic treatise originally written in Latin, which had as its intended audience those who were fully aware of and had carefully read the most important documents surrounding the Remonstrant controversy. I would submit then that any person not instructed under an educational system which demanded that college students in their first semester (in Davenant’s case, as a 15-year-old) master Franco Burgersdijk’s system of Logic, Institutionum Logicarum, will have a significant learning curve in mastering the contents of Davenant’s work.
Dr. Clark’s first mistake in trying to interpret Davenant’s De Morte Christi is relying on the English translation, of which he cites two pages. Thus some of the problems in “clarity” may be found in this preliminary step. The translation is filled with errors and should not be depended upon by someone who claims to read Latin. A second error is claiming that Davenant only has a singular “thesis,” when, in fact, he defends five. Davenant’s last thesis ,where he defends the notion that Christ died for the elect alone effectually, actually undermines a significant part of Dr. Clark’s argument regarding Davenant’s hypothetical universalism. Clark’s third error is his unwarranted claim that Davenant does not “clearly contrast” his view “with other views.” Not only does Davenant name interlocutors throughout the work, but at nearly every significant point, he draws up – usually in syllogistic form – possible/actual objections to his position and then gives a response to each. In so doing, Davenant presumes a lot of background information, but he also gives a lot of information as well, even if that necessitates tracking down his citations and reading the quotes in their contexts.
Oddly, even after admitting that he had a hard time understanding Davenant’s hypothetical universalism, Dr. Clark does not shy away from attempting to explain and eventually criticize Davenant’s argument. Clark writes: “There is a sense, [Davenant] argued, in which Christ may be said to have died for all men but that he did not actually obtain salvation for all. The limit then is not in the atonement but in the will of God to apply salvation to all. God, he argued, has not determined to apply the work of Christ to all.”
The first and third sentences of this section are accurate, but the second sentence is wildly misleading. It is inaccurate to claim that Davenant does not hold to any limit in what Christ accomplished (in distinction from redemption applied) on the cross. It is true that Davenant affirms that Christ’s death obtained a universal remedy applicable for the sins of all human beings. 25 Yet, and equally important to Davenant, the death of Christ also merited or obtained all the to-be-applied saving benefits (such as faith and repentance) for the elect and the elect alone. 26 This latter point – the point which fundamentally separated Davenant’s position from the Remonstrants – is almost never acknowledged in expositions of hypothetical universalism, even though Davenant spends a whole (and large) chapter defending it in De Morte Christi. In brief, in redemption accomplished (not applied) there is something accomplished for all, and there is something accomplished for the elect alone. It is not just that Davenant believes the saving benefits of Christ’s death will be applied to the elect alone; he also says that Christ actually died, and purposed to die, for the elect alone in a certain sense.
Accordingly, in Davenant’s interpretation of the Lombardian formula (Christ died sufficiently for all, efficiently for the elect), the two adverbs “sufficiently” and “effectually” modify the mode of Christ’s oblation, “Christ died for all” and “Christ died for the elect alone” respectively. 27 For Davenant, and every other Reformed theologian of the period, whatever Christ actually accomplished on the cross, whether it be for all or for the elect alone, it is grounded upon God’s absolute will. 28 What Clark misstated was that the limitation was simply in God’s will to apply the death of Christ to the elect alone, when, in fact, Davenant goes much farther, saying that it was God’s will for Christ himself to pray (in John 17, which Davenant understands as a prayer for God’s elect) and die for the elect alone, in order to merit for them the gifts of saving faith, repentance, etc.. Still, Davenant believed it was incoherent to say that Christ intended to die only for the elect while at the same time maintaining the original language of the Lombardian formula. 29 (This is where he and those like Piscator, who simply rejected the Lombardian formula, agreed.)
Davenant concluded that a proper affirmation of the Lombardian formula, according to the meaning he thought it was held by the medieval scholastics and the majority of the Christian church before his own time, required the Reformed (in distinction from the Lutheran and Remonstrant) understanding of the latter part of the Lombardian formula, meaning that God ordained Christ’s death to merit certain saving benefits for the elect alone! Still, Davenant maintained that a proper interpretation of the former part of the Lombardian formula entailed the belief that the Father intended Christ to make a satisfaction for the sins of all (not conditionally, however, as Warfield says; remission is conditional, but the act of satisfaction is not 30).
We see, then, that the conversation is more complicated than previously explained, with later critics not always attending to important qualifications and distinctions. And this is not the end of Dr. Clark’s misinterpretation of Davenant’s work. Dr. Clark goes on: “[Davenant and Amyraut] applied a twofold distinction in the divine decree (the absolute will of God and the antecedent will of God) to the atonement and they made the atonement conditional. Christ died for all to make salvation possible but he died ‘for you’ if you believe.” Stated in this way, this is incorrect.
Perhaps it is because Davenant’s scholasticism obscures the clarity of his thesis (or maybe because Clark presumes that Davenant and Amyraut teach the same thing), but Clark poorly describes Davenant’s twofold divine decree as a distinction between God’s antecedent will and his absolute will. 31 The distinction Davenant consistently uses, culled straight from Aquinas, is the distinction between God’s providential will and his will of predestination. 32 In both of these wills, God gets what he intends. 33 According to the former will, God intended that Christ’s death be a universal cause of salvation. According to the latter will, God intended Christ’s death to infallibly bring about the salvation of the elect alone. Dr. Clark should have described Davenant’s hypothetical universalism as: Christ died for you – whether you believe or not – insofar as he intended his oblation to provide a universal remedy for your sins; but he has not died for you – if you are non-elect – insofar as Christ did not intend to merit for you the infallible application of the saving benefits offered in the gospel. As briefly noted earlier, Davenant interprets all of the normal “Christ died for the elect alone” Scripture texts (e.g., John 17 and John 10) as referring to this latter divine intention to die for the elect alone. 34
It is worth emphasizing, again, that this latter divine decree does not simply entail that only the elect will be saved (something any Remonstrant would affirm); but it includes a divine will in the work of Christ as accomplished. Christ offered himself for Peter in a way that he did not for Judas. God the Father ordained the death of Christ, and Christ offered himself on the cross in order to procure for the elect alone an infallible application destined. 35 There is a limited aspect to Christ’s death in its accomplishment.
Davenant’s subtlety and careful distinctions, of course, do not prevent Clark from pronouncing a sweeping judgment over hypothetical universalism. He writes: “Both the Remonstrants and the Hypothetical Universalists make the atonement conditional. Davenant, however, still held that the Spirit grants faith to the elect to receive the benefit of the atonement. For the Remonstrants, we have it within ourselves, by common grace, to believe and to appropriate the benefit of the atonement. In neither case, however, is the atonement said to be unconditional. In that neither the hypothetical universal view nor the Remonstrant view agree with the Canons of Dort.”
Throughout his article, Dr. Clark insists that the hypothetical universalists such as Davenant “make the atonement conditional.” He finds this point crucial. But again, Davenant believes (to use George Smeaton’s phrase) that the atonement did contain an element of its own application to the elect. Smeaton, the 19th century Scottish theologian, has a different (and more faithful) interpretation of Davenant’s hypothetical universalism:
They [i.e., the British Delegates at Dort] said that Christ died for the elect according to the love and intention of God the Father and of Christ, that He might actually obtain, and infallibly bestow, remission of sins and eternal salvation; and that faith and perseverance are given to these elect persons out of the same love, by and on account of (per et propter) the merit and intercession of Christ. This draws a wide line of demarcation between the theology of Davenant, or of the Church of England, and that of Amyraldus, which insisted on a view of the atonement which, […] did not contain the element of its own application. 36
Dr. Clark’s final and most outlandish claim is that Davenant’s hypothetical universalism does not agree with the Canons of Dort – outlandish because Davenant was a delegate at Dort and continued to defend and even promote Dort in England. Indeed not even Dr. Clark’s colleagues at Westminster Seminary California agree with him, including Dr. Godfrey, who wrote his dissertation on this precise topic! 37 In fact, I am not aware of a single contemporary historian who argues this. 38 Not even Warfield attempts to say this. While Warfield does think that hypothetical universalism falls outside the boundaries of Westminsterian orthodoxy, Clark neglects to note that Warfield does not think it falls outside the bounds of the Calvinistic system:
That, despite its confessional condemnation [i.e., WCF and Helvetic Formula Consensus], Post-redemptionism has remained a recognized form of Calvinism and has worked out a history for itself in the Calvinistic Churches (especially in America) may be taken as evidence that its advocates, while departing, in some important particulars, from typical Calvinism, have nevertheless remained, in the main, true to the fundamental postulates of the system. 39
Dr. Clark continually says that “[o]n the premise of a universal atonement, whether hypothetical (Davenant) or absolute (Remonstrants), one cannot say that “Christ loved me and gave himself for me,” 40 but as we have seen, Davenant explicitly affirms that. In fact, the British Delegation, in their Suffrage, rejected the Remonstrant position that denied this, even before the Canons were drawn up! 41 In other words, they were not even coerced to reject that position. Davenant writes:
When it is said that Christ died that he might gather together these children, that he might sanctify this church, that he might purchase this church, etc. in Christ’s offering himself, there is not denoted an ineffectual will or some conditional intention, which might fail of the intended effect; but an effectual will and design conjoined with an infallible event … he so loved his sheep, his children, and his church that he decreed to effectually merit for them, by his death, faith and eternal life. 42
Indeed, most of what Dr. Clark says regarding what Dort teaches about the atonement are things that Davenant actually affirms.
John Davenant is the subject of my own dissertation, and as such this essay is something of a “pet topic” for me. But this case reveals more than just careless academics. What concerns me even more is that Dr. Clark’s treatment of Davenant highlights broader problems, namely in how theological gatekeepers wrongfully restrict the bounds of orthodoxy and seek to impugn historical figures through guilt by association and overgeneralized prejudicial summaries. History, then, becomes little more than a battering ram. Instead of treating men on the merits of their own work, they are placed into a larger group of supposedly problematic figures, interpreted in light of related but distinct debates, and then unfairly dismissed. In this case, all hypothetical universalists are treated the same, and they are described as being essentially in agreement with the Remonstrants. A common and overly simplistic answer is given to them alike. This is a problem which is broader than one critical treatment of John Davenant, and as such, is worth refuting.
We can and should do better.