Foundations: No.70 Spring 2016

The Covenant of Redemption according to John Owen and Patrick Gillespie

John Owen and Patrick Gillespie made profound contributions to the Reformed understanding of the “Covenant of Redemption”, or pactum salutis. Owen discusses it in at least sixteen of his works from 1645 onwards. Gillespie’s work, The Ark of the Covenant Opened, 1677, has been described as “the most elaborate work in the English language” on the subject. The importance of the doctrine, in Owen’s view, is apparent in the Preface that he wrote to Gillespie’s work. He says “the truth herein is the very centre wherein all the lines concerning the grace of God and our own duty do meet, wherein the whole of religion doth consist.” Both authors regard the doctrine as the intra-Trinitarian foundation of the Covenant of Grace apart from which no man is saved. Both explain that God’s salvation plan was the result of the eternal counsels of the persons of the Trinity, such counsels having the features of a covenant and including Christ’s distinct personal concurrence. Other theologians have rejected the notion of an intra-Trinitarian covenant, on the basis that it contradicts the undisputed truth that God has one indivisible will, and have sought instead to explain the plan of salvation simply in terms of divine decree. Yet Owen and Gillespie regard the “pure decree” explanation as an inadequate account of the Scriptural data – and hence an inadequate account of the whole foundation of God’s covenantal dealings with his people which underpin the whole of theology. In their view, the Covenant of Redemption provides a more compelling and faithful account, and does so without dividing the indivisible Trinity. This essay explores the alignment of their thinking on this vital issue.


John Owen (1616-1683) and Patrick Gillespie (1617-1675) were contemporaries and friends.[1] Both wrote extensively on the Covenant of Redemption, meaning the ad intra[2] Trinitarian pact in eternity, from which in time flowed ad extra the Covenant of Grace and the salvation of the elect.[3] Owen tends to refer to the Covenant of Redemption as the eternal covenant, compact or counsel between the Father and the Son; but also as the “covenant of the mediator” or “covenant of the redeemer”. He discusses it in at least sixteen of his works between 1645 and his death.[4] The first express reference to it is in his Greater Catechism (1645).[5] Thereafter his most extensive treatments are in The Death of Death (1648),[6] Vindiciae Evang-elicae (1655),[7] Hebrews Exercitations 27 and 28 (1674),[8] and The Doctrine of Justification by Faith (1677).[9]

Gillespie discusses it principally in The Ark of the Covenant Opened (1677).[10] He also calls it the “covenant of suretiship”. Trueman describes this as probably the most elaborate work in the English language on the Covenant of Redemption”.[11] It is said to be the second of five works setting out his covenant theology. The first, The Ark of the Testament Opened (1661), discusses the Covenant of Grace.[12] The third, fourth and fifth volumes are said to have been completed but lost.[13] I note, however, that this second volume as posthumously published appears to encompass much more than is advertised on its own contents page.[14] Given other internal evidence that this second volume was meant to finish with chapter 6 of the published work,[15] it seems to me possible that chapters 7 to 23 might belong to one or more of the “lost” volumes. These later chapters are principally concerned with Christ’s temporal role as mediator of the New Covenant (chapters 7-17); and with the various “relations that Christ sustains in the Covenant of Grace” (chapters 18-23). In any event, they do not add a great deal concerning the Covenant of Redemption that has not been expressed in the first six chapters.[16]

Owen’s and Gillespie’s theologies of the Covenant of Redemption are similar in many respects. The most direct evidence of this is in the preface that Owen wrote for The Ark of the Covenant Opened around two years after Gillespie’s death. He says that “for order, method, perspicuity… and solidity of argument”, Gillespie’s work is better than any he has read; it is “entirely compliant with the doctrine of the gospel”; and it lacks nothing “unto what is practical.”[17]

We therefore approach Gillespie’s work expecting a large amount of common ground with Owen’s writings; and this we find, but some points of difference also. I will first discuss the common ground and then move on to three issues wherein Owen and Gillespie differ, or are said to differ, or appear to differ. My aim has been to survey and briefly describe the positions of Owen and Gillespie, before interacting with some contemporary analysis of their positions. I will focus on points where it seems to me that their positions have not hitherto been adequately analysed, in particular:

  • why both Owen and Gillespie regarded the Covenants of Redemption and of Grace to have been separate and distinct;
  • whether they under- or over-stated the role of the Holy Spirit in the Covenant of Redemption;
  • whether they were agreed on the capacity in which Christ contracts in the Covenant of Redemption;
  • whether they were agreed that the atonement, pursuant to the Covenant of Redemption, was absolutely necessary if God was to save sinners without prejudice to his justice; and
  • whether they regarded the Covenant of Grace, unlike the Covenant of Redemption, as “embracing a broader category than only the elect”.

Common ground

1. Synopsis of the elements and purpose of the Covenant of Redemption

Owen and Gillespie agree on the essential elements of any covenant: it must have parties; a prescription of works; consent by one party to undertake the works; and a promise of reward upon completion. When these are present, a covenant is present, whether or not it is expressly so called.

Perhaps Owen’s most succinct expression of the necessary elements is:

An absolutely complete covenant is a voluntary convention, pact or agreement, between distinct persons, about the ordering and disposal of things in their power, unto their mutual concern and advantage.[18]

He finds all of these things present in “that compact, covenant, convention or agreement… between the Father and the Son, for the accomplishment of the work of our redemption by the mediation of Christ.”[19] In this covenant were fixed all the terms of the Covenant of Grace.[20] The whole purpose of this work is to manifest “the glorious properties of the divine nature” especially his wisdom, justice, grace, mercy, goodness and love.[21] By virtue of their “distinct personal actings”, involving the Son’s voluntary undertaking of suffering (something which is not natural to God), and the willingness of the Father to accept his obedience and suretiship on behalf of the elect, Owen says that the arrangement “differeth from a pure decree”.[22] It is “more than a decree, and hath the proper nature of a covenant or compact” because, with the parties’ consent, it eternally brings into being “a new habitude of will in the Father and Son towards each other that is not in them essentially… [or] naturally”.[23] He draws these conclusions from numerous Scriptures including Hebrews 2:9-16 and 10:5-9, John 17, Zechariah 6:13, Isaiah 49:5-9, Proverbs 8:22-31 and Psalm 40:7-8.[24]

With similarly painstaking exegesis of many of the same passages of Scripture, Gillespie agrees with Owen that these elements were present in the eternal counsels of God. He identifies the elements in a daisy-chain sequence, constituting a pact between “Jehovah and Christ” (a) wherein God holds forth “commands with promises”; (b) these being “promises with conditions”; (c) which are received as “conditions with consent”; (d) followed by “consenting with performing”; with thereafter the “asking and giving” of the promised reward.[25] Substantively, Christ is thereby established as mediator of the New Covenant (which he will fulfil as prophet, priest and king)[26] and as its surety (perfectly performing every stipulation of the Covenant of Works in the stead of his people, who would fail to keep it).[27] All of these necessary elements work together to become “the foundation of the covenant [of grace] made with us” whereby God can deal with the elect voluntarily, graciously and justly; otherwise he might simply “in justice have prosecuted the covenant of works”, to every man’s eternal condemnation.[28]

Owen and Gillespie are therefore agreed that the Covenant of Redemption is the eternal foundation and cause of the temporal Covenant of Grace, finally expressed in the New Covenant. In eternity, Christ graciously accepted the obligations which he executed in time. As Gillespie puts it, “nothing is here transacted in time which was not from eternity concluded in the counsel of God’s will.”[29] These obligations included the requirement that Christ: assume the human nature of those he was to save; submit to the general law applying to all mankind; submit to the especial law of the church; and by keeping those laws where mankind failed to do so, on behalf of his elect, and by suffering the punishment due to them for their failure, save them by his perfect obedience. The Father, having assisted the Son in the performance of the work, graciously promises to accept the imputation of Christ’s righteousness to the elect, such that they may justly be accounted righteous; and to reward the Son with glory.[30]

2. Two covenants or one?

Owen and Gillespie are agreed that the Covenants of Redemption and of Grace are two distinct covenants. Gillespie says we must avoid “two extreams, [such] that we neither confound nor divide these two covenants.”[31] In this they differ from a number of Puritans, including Thomas Boston, John Brown, Edmund Calamy, John Gill, Samuel Petto and also the “Antinomian wing of Puritanism”, who for different reasons regarded the atemporal “establishing” covenant and the temporal “effecting” covenant as one and the same.[32]

In The Death of Death, Owen refers to the Covenant of Redemption as God’s “eternal counsel for the setting apart of his Son incarnate” for the office of mediator, noting that “this is an act eternally established in the mind and will of God, and so not to be ranged in order with the others, which are all temporary and had their beginning in the fullness of time.”[33] Complexity alone does not prevent one agreement from dealing with different issues at different times with different parties. But in Owen’s view, we are not to consolidate these two pacts, being essentially different in kind, into one. The atemporal cannot be “ranged in order with” the temporal, though it is its “spring and fountain.”[34] This, for Owen, is the principal reason why the two covenants must be regarded as distinct.

Gillespie identifies eight similarities between the two covenants, nine ways in which they differ, and then a “five-fold connexion” between them.[35] Of the differences, Gillespie agrees with Owen that the “eternal pact” versus “temporal pact” distinction means that they differ in an essential “property”; and hence they must be separate covenants.[36] Summarising his other points of difference, Gillespie also says that (a) the parties to the Covenant of Redemption – Father and Son – are mutual sources of grace, whereas God alone is the source of grace shown to sinners in and through the Covenant of Grace;[37] (b) the Covenant of Redemption is made between equals and requires no mediator;[38] (c) the specific promises, commands and conditions are different because one is the establishing covenant, the other is the effecting covenant.[39] One is the “fountain”, the other is the “stream”; one is the “root”, the other is the “branch”; one is “a deed… drawn for children and heirs not yet born”, the other is the “conveyance” which in time gives effect to it.[40] All of these differences argue that the two covenants must be regarded as distinct.

Trueman says that Gillespie’s “central Biblical justification for arguing that the covenant of redemption exists as a separate covenant… is rooted in Biblical texts which… make no reference to [Christ’s] Seed.”[41] This, argues Trueman, is because the Covenant of Redemption is a “personal covenant” “between two parties which is focused on their mutual obligations and benefits”; hence it is distinct from the Covenant of Grace, which is “an arrangement which has an impact upon the progeny of the parties.”[42]

However, we should note that Gillespie’s distinction between “personal covenants” (concerning just the direct parties) and “real covenants” (concerning corporate or federal obligations) is just one of his nine arguments for the distinction between the two covenants.[43] It is debatable whether it is his “central” justification. That the Covenant of Redemption should be a “personal” covenant, as defined by Gillespie, does not itself mean that it could not also have included “real”, corporate or federal obligations. At least in the context of covenants between men, there is nothing unusual about a party covenanting in several capacities. In the same agreement a party may take on “personal” obligations in one clause and corporate or fiduciary obligations (for example, regarding his sub-contractors or beneficiaries) in the next. The most fundamental of Gillespie’s nine points of difference, which most compellingly necessitates two distinct covenants, is the “eternal pact” versus “temporal pact” distinction,[44] wherein he agrees with Owen.

3. Is the Holy Spirit party to the covenant?

Owen and Gillespie are agreed that Christ in particular must be party to the Covenant of Redemption, in a way that is not so obviously the case with the Holy Spirit, because Christ was therein appointed to roles that involved economic subordination, condescension and humiliation, notwithstanding the ontological equality of the persons of the Trinity.

Owen says that the personal nature of the works that Christ must undertake for the Father “indispensably introduceth an inequality and subordination in the covenanters… however on other accounts they be equal.”[45] That follows simply and unavoidably from the Father being the prescriber of duties and the Son being the undertaker. Those passages in Scripture that speak of the subordination of the Son to the Father, such as Psalm 16:2, are “expressions [which] argue both a covenant and a subordination therein.”[46] But the status of subordination is all the clearer given the nature of the duties that he undertook. At least in human terms, no-one is appointed a legal surety except with his consent.[47] Furthermore, consent to the subordination was essential for the efficacy of Christ’s work. A priest’s offering is valueless if it is given grudgingly. “His death could not have been an oblation and offering had not his will concurred.”[48] Thus the economic subordination of the Son, notwithstanding his essential equality with the Father, not only results from the Covenant of Redemption but is evidence for the necessity of a covenantal arrangement rather than a non-covenantal “pure decree”.[49]

Similarly, Gillespie says that it is because Christ’s work required him “to leave the throne of glory, and come down to his footstool, there to be in disgrace” that his consent was required.[50] There was a necessity of nature for the Spirit to support the work of the Father and Son, but there was “no necessity of nature that [Christ] should make himself a sacrifice for our sins.”[51] It is because the Covenant of Redemption is essentially concerned with the appointing of a divine person to the roles of mediator, surety and messenger, all of which involve unnatural condescension and economic subordination – “and this person was the Son only, not the Father, nor the Spirit” – that Gillespie considers it right to describe it as an agreement “betwixt Jehovah and the Mediator Christ.”[52]

By focussing on the Father and the Son as parties, both Owen and Gillespie are potentially open to the criticism that they are downplaying the role of the Spirit and diminishing his involvement in a work of the Trinity ad extra.[53] So Letham says that Owen’s treatment of the Covenant of Redemption in Exercitation 28 is essentially “binitarian”, and that “Amazingly the Holy Spirit receives no mention!”[54] He argues that Owen is part of a Western Church tradition that sees the Holy Spirit as “subordinated and depersonalized” to the point of being “merely the bond of love between the Father and the Son.”[55] Trueman, by contrast, commends Owen for making,

a distinctly Trinitarian advance on the works of Fisher and Bulkeley who, with their exclusive attention to the Father-Son relationship were arguably vulnerable to the accusation of developing a sub-Trinitarian foundation of the economy of salvation.[56]

Trueman’s is the fair assessment. Although Owen frequently refers to the covenant as being between Father and Son, he does not omit mention of the involvement of the Holy Spirit, and certainly does not reduce him to being a “mere bond of love” between Father and Son. The Holy Spirit is not prominent or named as a contracting party in his treatment of the Covenant of Redemption; but that is fitting, since the covenant is primarily concerned with the appointment of the Son to his “new habitude or relation” to the Father, involving humiliation, “which is not natural or necessary unto them, but freely taken on them.”[57] Furthermore, in two works – including the Exercitation immediately preceding the one that Letham criticises – Owen expressly refers to the concurrence of the Holy Spirit in the eternal counsel of the Trinity.[58]

Trueman says that Owen avoids sub-Trinitarian thinking “by describing the various roles played in the covenant of redemption by Father, Son and Holy Spirit.” Of the Spirit’s role, Trueman says that “the Holy Spirit is engaged in the work of the incarnation and of Christ’s earthly ministry, his oblation and his resurrection.”[59] O’Donnell rightly notes that Trueman is here in danger of conflating the distinction between the Spirit’s role in the Covenant of Redemption and his temporal role.[60] The Spirit’s work during the incarnation does not per se prove that the Spirit was party to the eternal transaction. Hypothetically his work could be procured pursuant to a purely bilateral eternal compact between Father and Son.[61] By way of human analogy, a prime contractor can agree to procure the work of a sub-contractor without the sub-contractor being made party to the principal-level agreement. O’Donnell therefore says that Trueman’s praise of Owen is “too strong”. His conclusion is that “The most that can be said is that [Owen] neither ignores completely nor develops satisfyingly the Spirit’s role in the pactum.”[62]

In my view, O’Donnell’s criticism of both Trueman and Owen is too strong. Trueman is aware that the Covenant of Redemption is “the nexus between eternity and time with respect to salvation”, thus he is willing to link closely the Spirit’s temporal work with his eternal concurrence.[63] He is not saying that Owen’s argument consists simplistically of: “the Spirit does these things in time, therefore he must be a party to the Covenant of Redemption in the same way that the Son is”.[64] Owen’s position is more subtle than that. This is apparent from Exercitation 27, where he refers to the “peculiar, internal, personal transactions between the Father, Son and Spirit… [involving] mutual distinct actings and concurrence of the several persons in the Trinity… expressed by way of deliberation”;[65] and to the eternal compact as “a personal transaction, before the creation of the world, between the Father and the Son, acting mutually by their one Spirit, concerning the state and condition of mankind.”[66] This is consistent with it being the Father’s role to prescribe duties; the Son’s to consent to undertake them (those two being the principal parties); and the Spirit’s role to concur in eternity and to operate instrumentally in time in Christ’s conception, oblation, resurrection and exaltation and in applying his work to his people. Thus Owen is content frequently to refer to the eternal transaction as being “between the Father and Son”. In doing so he is neither overlooking the Spirit’s role, nor failing to recognise the essential feature of it that required it to be transactional rather than merely decretal: the willing condescension of the Son. As Tay puts it, “the Spirit’s role in the pactum salutis is instrumental and thus not directly pactional.”[67] Tay regards this careful expression of the Spirit’s tacit role as being consistent with “Owen’s understanding of the filioque-based order of divine operations”, in which the “order of subsistence” of the persons of the Trinity flows through into the “order of operations”.[68] Thus the Spirit does not act as originator or primary undertaker of the eternal compact, but as implementer and perfecter, wherein “the Son’s incarnation, mediatorial sufferings and resurrection were all made possible and actual by the work of the Spirit.”[69] Owen is seeking to go only as far as Scripture permits, not to satisfy every curiosity.[70] If O’Donnell finds Owen’s explanation of the Spirit’s role “unsatisfying”, I suspect that Owen would say that he could with propriety go no further.[71]

Gillespie is somewhat ambiguous about the distinction between the Spirit’s role in the eternal covenant on the one hand, and in the temporal covenant on the other. Having consistently referred to the eternal covenant as having been made between Father and Son, he says nonetheless that the Spirit “undertakes to unite the humane nature to Christ, by miraculous conception… to be a Spirit of unction… to shed abroad the love of God in our hearts,” and to do other applicatory works.[72] Did he only undertake these things in time? Apparently not: it does seem that Gillespie is referring to the Covenant of Redemption at this point.[73] If the Holy Spirit made these undertakings in eternity, it presumably follows that Gillespie regards him as at least in some sense a party to the eternal covenant. Gillespie does not specify his position more clearly, but he appears to be a little more willing than Owen to see the Spirit’s involvement as “pactional” in eternity, as opposed to being merely concurring. We can at least say that Owen’s view is the more internally consistent.

Before leaving this point, I note that Beeke and Jones introduce an overly anthropomorphic twist by asking whether the Spirit “was involved directly in the covenant of redemption as a negotiating partner.”[74] If we try to picture the negotiation of the Covenant of Redemption as the persons of the Trinity gathered around a table, discussing who must do what to save sinners, wondering if perhaps the Holy Spirit might have been “out of the room at the time?”,[75] we are overlooking the eternal and immutable simplicity of God. The Covenant of Redemption did not involve negotiation, but rather the eternal concurrence of the single divine will from the perspective of each person of the Trinity.

4. The Savoy formulation

It follows from what has been said so far that Owen and Gillespie would both concur with the description of the Covenant of Redemption in the Savoy Declaration. Chapter 8.1 of this says:

It pleased God, in his eternal purpose, to choose and ordain the Lord Jesus his only begotten Son, according to a covenant made between them both, to be the Mediator between God and man; the Prophet, Priest, and King, the Head and Saviour of his Church, the Heir of all things and Judge of the world; unto whom he did from all eternity give a people to be his seed, and to be by him in time redeemed, called, justified, sanctified, and glorified.[76]

Owen can be taken to have approved this wording, having been involved in the drafting of the Savoy Declaration: indeed, along with Thomas Goodwin, [being] one of [its] principal architects.”[77] Trueman considers that the express reference to the Covenant of Redemption is “almost certainly” the result of Owen’s role in writing the Savoy document.[78] As for Gillespie, all of the elements of chapter 8.1 may be found in the first chapter of The Ark of the Covenant.

There is no reason to think that they would not also have concurred with the corresponding article (Chapter 8.1) in the earlier Westminster Confession. This differs only in that it lacks the wording underlined above. This is not to say that those who formulated the Westminster Confession in the mid-1640s had in mind the Covenant of Redemption, a term which was not in common use by that time. But it is to say that neither Owen nor Gillespie would have had significant objection to what they found in Chapter 8 of the Confession, which Owen was happy to adopt and expand upon in the Savoy Declaration.

Points of difference

Given Owen’s almost unreserved endorsement of Gillespie’s work in his Preface, it is perhaps surprising that his view of the Covenant of Redemption differs in a number of respects. Most of the differences are those of expression, such as Owen in his Preface tells us to expect. But one difference which I will examine is substantial. Another is said to be substantial, but in my view the difference has been over-stated. Another is real, and worthy of comment, but not a significant point of disagreement between Owen and Gillespie.

1. The capacity in which Christ contracts within the Covenant of Redemption

There is a significant difference between the opinions of Owen and Gillespie on the covenanting capacity of Christ.

I will consider Gillespie’s argument first, since it is peculiar. Essentially he says that Christ could not have been party to the Covenant of Redemption in his divine nature, as the pre-incarnate Son. God has one will and therefore “Christ God, the second person, could not constitute a party covenanting distinct from God considered essentially… Father, Son and Spirit.”[79] There must be another distinct party for there to be two wills in agreement. So Gillespie proposes that God contracted in eternity with Christ already as God-man. “Christ had a will distinct from Jehovah’s will only as he was God-man: for as God, his will is one and the same with his Father’s will and undistinguished from it.”[80] He develops this argument by saying that “all satisfaction is taken away” if the same party proposes the work, does it and rewards it.[81] Furthermore, if the work must be done on earth by the God-man, it must be the God-man who contracts with Jehovah: “In what consideration [i.e. capacity] Christ did perform the Covenant of Redemption, in the same respect he is to be considered as a party-undertaker.”[82]

Gillespie appears to have taken a seriously wrong turning at this point. If he had been speaking of the Covenant of Grace, which was made between God and Christ incarnate in time not eternity, neither Owen nor the Westminster Confession would disagree with him. But he was speaking of the Covenant of Redemption, which – as Gillespie himself says – was not entered into when Christ was man, but “long before his coming in the flesh, [because it] must needs relate to such transactions as were betwixt God and Christ in the counsel of his will from all eternity”.[83] Christ was not the God-man until the incarnation.[84] Per Owen: “The same person – who before was… not man – was made flesh as man.”[85] He had no pre-incarnate human soul, nor human will.[86] He covenanted to become the God-man, and to do the appointed work; it was not necessary that he should covenant qua God-man in order to do that. There also appears to be confusion in his thinking as to the distinct personhood of Father and Son. The Son was just as much a distinct person before the incarnation as afterwards – so Gillespie’s “all satisfaction is taken away” argument is groundless.[87]

I have not found any contemporary comment on Gillespie’s opinion on this point. Beeke and Jones, Trueman, Muller and Carol Williams in her PhD thesis all give some space to his views on the parties to the Covenant of Redemption and the unity of the divine will, but none of them discusses Gillespie’s opinion that Christ contracted as God-man.[88]

Owen’s solution to the challenge of finding two willing covenanters in a one-willed God is more compelling. As early as 1655 he explained that “the will of God the Father, Son, and Holy Ghost is but one: but in respect of their distinct personal actings, this will is appropriated to them respectively, so that the will of the Father and the will of the Son may be considered [distinctly] in this business”[89] In other words, the one will of God has distinct applications to the distinct acts of each person of the Trinity ad extra. His position is the same, but more developed, in 1674. The clarity of his argument warrants an extended quotation:

Father, Son and Spirit have not distinct wills. They are one God, and God’s will is one, as being an essential property of his nature… How, then, can it be said that the will of the Father and the will of the Son did concur distinctly in the making of this covenant? This difficulty may be solved [by having regard to] the distinction of the persons in the unity of the divine essence, as… they act in natural and essential acts reciprocally one towards another… And as they subsist distinctly, so they also act distinctly in those works which are of external operation. …The will of God as to the peculiar actings of the Father is the will of the Father, and the will of God with regard unto the peculiar actings of the Son is the will of the Son; not by a distinction of sundry wills, but by the distinct application of the same will unto its distinct acts in the persons of the Father and the Son.[90]

Writing with reference to Zechariah 6:13 just a couple of years before he wrote his Preface to Gillespie’s work, in direct contradiction of Gillespie’s opinion, Owen says: “God takes counsel with [the Son] as he was his eternal Wisdom, only with respect unto his future incarnation”; he did not do so “absolutely as he was a man, or was to be a man, for so there was not… ‘counsel’ between God and [Christ incarnate].”[91] It was in the accom-plishment of the covenant that he took on humanity, not in the formation of it.[92]

Thus the solution to Gillespie’s difficulty is not to be found in the two wills (or the human will) of the incarnate Christ, for his human will was not extant or operative pre-incarnation. Nor is it to be found by drawing back into the Covenant of Redemption the God-man as he will become pursuant to the Covenant of Redemption. Instead, Owen’s solution upholds the unity and simplicity of God, while also maintaining the possibility of the divine persons acting willingly and distinctly in relationship with each other – for example, in freely and willingly loving each other – and concurring in the proposed acts of redemption ad extra.[93] The oneness of God cannot rule out the possibility of concurrence between the three persons; nor does the certain inevitability of their concurrence, or the impossibility of disagreement, contradict the possibility or reality of that concurrence.

Both Owen and Gillespie held to the orthodox view that Christ, from the creation of man, was mediator according to both his divine and human natures.[94] Owen says plainly that prior to the incarnation, pursuant to the Covenant of Redemption, Christ was mediator only in his divine nature; whereas following the incarnation he was mediator according to both natures. Gillespie seems to agree, though there is some ambiguity in his expression: “before Christ’s incarnation he was a mediator virtually and undiscernedly… (though the way of his acting in that office… transcend our understanding) but after his incarnation… then he was manifestly an actual mediator… [in that] he did visibly act that part upon earth.”[95] We see from this that the point of difference between Owen and Gillespie lies in Christ’s covenanting capacity in eternity, rather than in the temporal outworking of the Covenant of Redemption.

2. Whether the Covenant of Redemption was an absolute necessity

A number of contemporary authors argue there is a substantial difference between the views of Owen and Gillespie on whether, if God was minded to save sinners, he must do so by means of the Covenant of Redemption and Christ’s atoning work on the cross; or whether in his sovereign freedom he might do so in some other way.

By 1652, Owen was firmly settled in his view that the atonement, by means of the incarnation, death and resurrection of Christ and in no other way, was a conditional necessity: that is, conditional upon God’s free decision to ordain the creation of all things including rational creatures, who are morally dependent on God and are capable of responding in obedience; conditional upon the giving of law to these rational creatures; and conditional upon the actual sin of these creatures. “God… necessarily punishes sin; not… from an absolute necessity of nature, as the Father begets the Son, but upon the suppositions before mentioned.”[96] At no time did Owen espouse a voluntarist position, as if God could do whatever he willed, unconstrained by his nature; but in 1648 he had argued that God was free (one might equally say constrained) to do what was most conducive or fitting to his glory.[97]

To put it another way, using the language of Anthony Burgess,[98] one might say that in 1648 Owen regarded the necessity as constrained by “the final cause” (God’s glory) rather than the “efficient cause” (God’s natural justice), whereas from 1652 he regarded the necessity as constrained by both.[99]

Gillespie distinguishes between three types of necessity. Like Owen, he denies that the atonement, pursuant to the Covenant of Redemption, was an “absolute necessity”. God was free to redeem sinners or not to redeem them: “God… might have not entered that Covenant with his Son; for it was not absolutely necessary that man should be redeemed.”[100] Nor was it a “natural necessity”, proceeding “as the fire burns, necessarily and naturally”, otherwise God must “either have punished all that sinned to the utmost as soon as they had sinned, or he should have shewed mercy upon all as soon as there were qualified objects for mercy, because all natural agents work to the utmost they can.”[101] Rather, the Covenant of Redemption was a “hypothetical” or “consequential” necessity: consequential upon God determining to make known both his justice and his mercy to men, having “creat[ed them] in a blessed but mutable state” and upon their fall into sin.[102] God did not then “acquit [man] without a satisfaction to justice” because that did not accord with his “infinite wisdom”; instead he determined to redeem lost men through Christ “because [he] thought fit that it should be so.”[103] Thus far, Owen would have been in agreement, both in 1648 and subsequently.

Gillespie then asks the Burgess question: whether that consequential necessity was dictated by “the final cause” (God’s glory) or the “efficient cause” (God’s natural justice).[104] Certainly it was dictated by the former, he says; but was it also dictated by the latter? Gillespie is reluctant to be dogmatic on this, noting that there is a difference of opinion among the orthodox and regarding it as not “useful to dispute about the possibility of another way of taking away sin… [since God hath]… plainly pitched upon this only way: that he will not pardon sin without a price and satisfaction.”[105] He goes on to reject the “extream and… dangerous” and “wholly arbitrary” voluntarism of the Socinians: “[God’s] punishing sin is not merely from his will.”[106] But he also declines to endorse the “other extream”, which he attributes to Burgess, by which he means the view “that because justice is in some sense a natural property in God, and his punishing of sin is not meerly from his will; therefore God punisheth sin by necessity of nature, and cannot but punish it, or require satisfaction, more than he can deny his own nature, or cease to be God.”[107]

At first sight, this second “extream” view which he is rejecting seems close to Owen’s post-1652 position. Certainly Owen, in his Dissertation on Divine Justice, insists that justice is a “perfection of the divine nature”, and that God punishes sin necessarily and not merely voluntarily.[108] Trueman goes so far as to say that Gillespie “repudiates… the position (which looks very like that of the later Owen) which regards punishment as necessary on the basis of God’s being, if sin is to be forgiven.”[109] Similarly, Beeke and Jones say that Gillespie “reject[ed] Owen’s position on the absolute necessity of the atonement.”[110] But are Owen and Gillespie as far apart as Beeke, Jones and Trueman assert? There are certainly sentences in Gillespie’s work that could be quoted to support the claim; but a more detailed consideration of his argument suggests a large measure of agreement.

Gillespie goes on to clarify what it is that he is rejecting at this point, as he seeks a middle course between “these extreams”.[111] What he is particularly rejecting is the second of the three types of necessity that he previously identified and discussed: that is, “natural necessity”, as fire burns.[112] He says that “God hath plainly revealed that he will not pardon sin, without a satisfaction and an atonement made, and this decree of his doth infer an hypothetical necessity of Christ’s dying by way of satisfaction.”[113] Owen agrees.[114] Gillespie next says that “supposing that… punitive justice be natural to God, and among his properties”, it does not follow that he must then necessarily act upon his natural justice to “punish sin as soon as ever it is committed [and] punish sin to the utmost degree of punishment… without mercy shewed to any.” That would be “natural necessity” at work. Owen says much the same.[115] But God has a discretion: just “because ’tis natural to man to speak… he might notwithstanding never… speak, as pleaseth him.”[116] Again, Owen agrees, using very similar language.[117] Gillespie then distinguishes between, on the one hand, the natural properties of God that operate with “absolute necessity of nature” because they require no freely-created object to operate upon – such as his wisdom, holiness, goodness and essential justice; and, on the other, those properties that require a freely-created “object ad extra” in order for them to operate. In the operation of these latter properties, which include punitive justice, God has freedom. He “doth not punish sin by necessity of nature, as the fire burns; since the exercise of justice, yea, the choice of objects upon which he will exercise it, are subject to his free will and sovereignty as is manifest from Rom. 9:18, ‘Therefore hath he mercy on whom he will have mercy; and whom he will, he hardeneth’.”[118] What Gillespie is eschewing is the “natural necessity [which] excludes freedom both in the principle of action and in the act itself.”[119] Gillespie, like Owen, believes that “in the act itself” God has a degree of freedom that the second “extream view” denies.[120] Gillespie, like Owen, denies that God is bound to punish sin absolutely, in the sense that he has no discretion in the creation of the objects to be punished, or in the election of some to be punished and some to receive mercy, or in the mode and manner of punishment of those he wills to punish.

The difference between Gillespie and Owen on this question is, then, more subtle than Trueman, Beeke and Jones have suggested. Returning to the Burgess question, Gillespie’s conclusion is that the constraint upon God – that is, the consequential necessity – “does mainly respect the final cause and ends which God had purposed.”[121] His position is that the necessity is “mainly” constrained by the requirement that God’s glory be maximised, rather than by God’s natural justice necessarily consuming everything in its path. He leans that way because he sees a degree of freedom in the exercise of God’s natural justice that he cannot deny. Owen does not deny that degree of freedom either. Trueman overstates the difference between them in his summary of Gillespie’s conclusion by substituting the word “simply” in place of “mainly”: “for Gillespie, it is simply in terms of final causality that one can understand this necessity to operate.”[122]

Thus is seems that both Owen and Gillespie believe that the Covenant of Redemption, and the merciful provision it makes against God’s vindicative justice, are conditional necessities, not absolute or natural necessities. They agree on what those conditions are. The difference between them is more a difference of expression than a difference of substance. Thus it is perhaps less “surprising” than Trueman suggests,[123] on this point, that Owen should feel able to endorse the Ark of the Covenant without significant reservation.

3. A covenant just for the elect?

The last point of difference upon which I will comment arises as a further facet of the “two or one covenant” question. Van Asselt distinguishes between those who regarded the Covenant of Redemption as different from the Covenant of Grace, including Cocceius, Dickson and Rutherford, and those such as Boston who regarded them as “one and the same covenant.”[124] Van Asselt goes on to say that the former (“two covenants”) group regarded the Covenant of Redemption as “the eternal pact which concerns the elect” and the Covenant of Grace as “embracing a broader category than the elect only”. Van Asselt puts Owen in this group, along with Cocceius.[125] As we have seen, he is right to say that Owen (like Gillespie) regarded the two covenants as distinct. But he is probably wrong to say that Owen regarded the beneficiaries of the temporal covenant as being a “broader category” than those benefited by the eternal covenant. It seems that Owen regarded the true beneficiaries of both covenants as being the elect and only the elect. In this, there is some difference of expression, at least, between Owen and Gillespie.

Owen explains in The Death of Death that there is exact co-extensivity as between those who were promised to Christ as the reward for his work on earth; those for whom he died; and those for whom he intercedes. It is on the basis of “the compact and agreement that was between the Father and the Son” that he insists that “the oblation and intercession of Christ are of equal compass and extent in respect of… the persons for whom he once offered himself.”[126] On the basis of the reward promised to him in eternity, there is then a “strict connexion [between] the oblation and intercession of Jesus Christ.”[127] With reference to Jeremiah 31, he explains that there is no-one taken into the New Covenant who does not receive the full salvific effects of it.[128] He holds to the same position in his later works. The “actual efficacy of his oblation… unto the church” is established by the Covenant of Redemption.[129] His death and intercession were on behalf of the elect according to that covenant. The purpose of his intercession is to present “his offering and sacrifice for the procuring of the actual communication of the fruits thereof unto them for whom he so offered himself.”[130] Explicitly: “The covenant of grace in Christ is made only with the Israel of God, the church of the elect.”[131] Just as the beneficiaries of the Covenant of Redemption were specifically and only the elect, so are the beneficiaries of the Covenant of Grace. The alternative, in Trueman’s words, is to “indulge in the speculative and futile move of trying to assign a meaning to [Christ’s death] outside of the determination of the covenant structure.”[132]

Owen is aware that some theologians argue that unregenerate members of the visible church are true members nonetheless: “Divines of all sorts do dispute [whether] hypocrites and persons unregenerate may be true members of visible churches.”[133] No doubt an unregenerate person may enjoy some of the benefits enjoyed by members of the Covenant of Grace – in particular “church-communion”.[134] This is enjoyed not least by children of believers, whose baptism as infants “[gives] them thereby an admission into the visible catholic church”.[135] But, contrary to Van Asselt’s suggestion, it does not appear that Owen finds any use for a “broader category” of covenant membership in the temporal covenant as opposed to the eternal.[136] The unregenerate within the visible church need not be regarded as true members of the Covenant of Grace.

In this opinion, Owen differed at least to some degree from many of his Reformed contemporaries and predecessors. But the difference can be over-stated. The Westminster Directory, for example, states that the infants of believers, whether those infants are elect or not,

have, by their birth, interest in the covenant, and right to the seal of it, and to the outward privileges of the church, under the gospel, no less than the children of Abraham in the time of the Old Testament; the covenant of grace, for substance, being the same.[137]

Owen would not disagree that such infants have some kind of interest in the Covenant of Grace, albeit outward or external. Likewise, the Directory notes that the children of believers, by baptism, are only received into the visible church; as such they are to be regarded as Christians, but it is not presumed from this that they enjoy inward grace and regeneration.

Gillespie says,

it was agreed betwixt Jehovah and Christ… who should be the redeemed people… he was not Surety and undertaker for all mankind… for his undertaking is no wider nor larger than his dying, purchasing redemption, bearing iniquity, praying etc, these being of equal extent and efficacy.[138]

Christ’s work “is to bring the elect within the bond of the covenant [of grace]”.[139] Thus far he agrees with Owen. He goes on to say that reprobates within the visible church may be regarded as being “within the covenant externally.”[140] To the extent that he recognises this category of covenanted-reprobates we may say that he aligns with Cocceius a little more closely than does Owen. However, this is probably more a difference of expression than of substance.


Given Owen’s Preface to Gillespie’s work, we are not surprised to find a considerable measure of agreement between them – even on the question of the necessity of the Covenant of Redemption, where at first sight they appear to be quite a distance apart.

Van Asselt’s explanation for the similarity of opinion is that Owen was “influenced by… Gillespie’s The Ark of the Covenant.”[141] This seems unlikely, given the precedence of Owen’s writings. It is more likely that Gillespie would have read his friend’s earlier works on the subject,[142] and discussed their respective views face to face.

But what Van Asselt goes on to say about Owen’s view of the Covenant of Redemption is apposite and equally true of Gillespie’s: neither saw it as “a reworking of, or an alternative to, the doctrine of the divine decrees”; rather,

the decree was taken up in a Trinitarian and covenantal language which… underscored the sincerity and faithfulness of God in his covenantal dealings with men. [It] articulated that the entire work of salvation has a Trinitarian-covenantal form at its very roots.[143]

Both Owen and Gillespie give clear reasons why purely decretal language is inadequate in seeking to explain the eternal counsel of God. To speak of God’s decrees is, rightly and appropriately, to speak of the one and indivisible will of God. There is one God and he has one will. But to speak only of decrees risks failing to speak of the personal, relational, intra-Trinitarian counsels that give rise and effect to them.[144] As soon as we admit of eternal concurrence and counsel within the Trinity, and contemplate the abasement that the incarnation required, the language of covenant (or compact, convention or counsel, if these other terms used by Owen and Gillespie are preferred) becomes appropriate if not unavoidable. Properly expressed in these terms, the notion of the Covenant of Redemption faithfully expresses the Scriptural data concerning the Trinitarian plan of salvation, and does so without implying any disunity or disharmony within the indivisible Trinity. On this fundamental point, and on much of its outworking, Owen and Gillespie are agreed.

The most important difference between them concerns Gillespie’s ill-considered or at least ill-expressed view of the covenanting capacity of Christ in the Covenant of Redemption. Owen’s view is plainly to be preferred, and so far as I am aware no-one has suggested a more compelling alternative. It is perhaps curious that Owen did not distance himself from Gillespie on this point of difference, particularly given that Owen says in his Preface that it was with “some diligence and great satisfaction” that he had perused Gillespie’s work.[145] Perhaps he overlooked it; or perhaps he regarded it as falling within the range of “ways of explanation” that Owen describes as “not unuseful”, when the

same truth, especially that which is of so great importance as what concerneth the Covenant [of Redemption], be variously handled [by different writers] according unto the measure of the gift of Christ which they have received.[146]

According to Owen, we must expect and charitably tolerate some such differences on the basis that “perfect harmony and universal agreement in all things is the priviledg only of the sacred writers who were divinely inspired.”[147] Certainly the difference between Owen and Gillespie on the necessity of the Covenant of Redemption and the resultant atonement falls within that “not unuseful” range; when scrutinised, it is not as substantial as Beeke, Jones and Trueman have suggested.

The interest of both Owen and Gillespie in these matters was not driven by a love of theological or philosophical conjecture, but by pastoral concerns. This is apparent from all of their writings on the subject. In Owen’s words, their common aim was to bless those “who desire to be edified in the truth that is after Godliness”, by contemplating as far as Scripture permits “the truth [that is] the very center wherein all the lines concerning the grace of God and our own duty do meet, wherein the whole of religion doth consist.”[148] The Covenant of Redemption is then the “well-head or the fountain of salvation, … the immediate sacred spring and fountain of the priesthood of Christ.”[149] By drawing back into eternity the eternal destiny of every man, no room is left for vain notions of sin catching God unawares[150] and necessitating a recovery plan by way of after-thought; nor of man contributing anything to his own salvation. Thus here we find the basis for an assurance, for all who are in Christ, that was founded and guaranteed in eternity.