Home | Downloads | About CWI | Donate | Site Map | Contact
Covenant Worldview Institute Home


The Covenant of Works:
A Litmus Test for Reformed Theology?

by Rev. Ralph Allan Smith

There are Reformed writers today who virtually insist that the Covenant of Works should be a litmus test doctrine. It is implied that those who deny it repudiate what is essential to the Reformed faith and its confession of God's saving grace. In particular, it is suggested that the Covenant of Works is indispensable to a genuinely Reformed view of justification by faith.[1] This is simply wrong.

First, though the doctrine of the Covenant of Works is included in the Westminster Standards, it is not even mentioned in any of the other Reformed Confessions. This is not to say or imply that only the Presbyterians believed in a Covenant of Works. On the contrary, the doctrine of a Covenant of Works became standard reformed theology in the 17th century. In the 16th century, however, it was not included in the Thirty Nine Articles — originally a Calvinistic and Reformed statement of faith — the Heidelberg Catechism or the Belgic Confession. The early 17th century Reformed statement, The Canons of the Synod of Dort, in rejecting Arminian theology and defining a Reformed doctrine of salvation makes no mention of a Covenant of Works, though it was no doubt widely believed. Concerning these and other Reformed creeds, John Murray wrote the following.

Towards the end of the 16th century the administration dispensed to Adam in Eden, focused in the prohibition to eat of the tree of the knowledge of good and evil, had come to be interpreted as a covenant, frequently called the Covenant of Works, sometimes a covenant of life, or the Legal Covenant. It is, however, significant that the early covenant theologians did not construe this Adamic administration as a covenant, far less as a covenant of works. Reformed creeds of the 16th century such as the French Confession (1559), the Scottish Confession (1560), the Belgic Confession (1561), the Thirty-Nine Articles (1562), the Heidelberg Catechism (1563), and the Second Helvetic (1566) do not exhibit any such construction of the Edenic institution. After the pattern of the theological thought prevailing at the time of their preparation, the term 'covenant,' insofar as it pertained to God's relations with men, was interpreted as designating the relation constituted by redemptive provisions and as belonging, therefore, to the sphere of saving grace.[2]

Why the doctrine of the Covenant of Works did not attain confessional status is an interesting historical question. But the fact that it did not is an indisputable truth. That being the case, it is either ignorant or perverse to raise the doctrine of the Covenant of Works to the status of a litmus test in our day. Denial of the Covenant of Works is not tantamount to a denial of Reformed theology.

Second, just as the doctrine of the Covenant of Works is not indispensable to the Reformed Confessions and Catechisms, so also it has not been considered indispensable by all Reformed theologians. The United Reformed Churches of North America published an unofficial statement of the view of the covenant "generally held in the United Reformed Churches." Concerning the Covenant of Works, the statement reads:

Frequently the Reformed churches have used the wording "the covenant of works" as applying before man's fall into sin, and "the covenant of grace" referring to God's gracious and just deeds and promises after and in response to the fall. This bi-focused view of God's relationship with His creature man is questionable. The use of the former in particular has limitations as to its usefulness, since the Bible does not suggest nor employ the wording. We believe that the simple designation "God's covenant" is preferable.[3]

Thus, Reformed churches following the continental Reformed confessions rather openly dismiss the doctrine of the Covenant of Works. Again, this is not to say that no theologians or ministers among the Dutch Reformed believe in a Covenant of Works. Many do. Some of them with all the enthusiasm of the Westminster Presbyterians. That does not, however, change the fact that among the continental Reformed churches, the doctrine of a Covenant of Works is not considered essential to Reformed orthodoxy.

One of the most famous Dutch Reformed theologians of the 20th century, G. C. Berkouwer, cautioned his readers about misconstruing the doctrine of a Covenant of Works. Almost prophetically, he rejected an approach to the Covenant of Works nearly identical to the one later to be set forth by Meredith Kline as the touchstone of Reformed orthodoxy.

Can we speak of a prelapsarian state in which there was a "law"? Was there a "nomological" existence of man apart from and even prior to the distinction of the law and the Gospel? If there was, can we search there, perhaps, for the fundamental structure of what it means to be a "man"? R. Schippers, in weighing all of these question, has affirmed that there was a law in man's "prolapsarian state," and that that law was there apart from the Gospel. At the same time, we no longer may speak of this law in abstraction. Schipper's statement has reference to the creation of God which must certainly be distinguished from man's guilt and fallenness and therefore from the Gospel of God's grace which saves.

It stands to reason that we may not draw conclusions concerning the relation of the law and the Gospel or build a case for the "priority" of the law on such a basis as this. Man's original life under God's rule cannot be regarded, for even a moment, apart from God's love and communion. Within that communion man was subjected to God's holy and good command; furthermore, because of that communion the commandment was never an impersonal or statutory rule. God's commandment expresses his lordship over life. Therefore, any discussion of the usus legis, in its various dimensions, is only conceivable in terms of this absolute goodness of God's commandment for creaturely man. The fact that this accent was sounded so frequently in Reformation and post-Reformation times is no evidence of the darkening of the Gospel, and is no recognition of a "legal order" above or before the "order of grace." What we see in this accent is only the enigmatic nature of guilt in the face of God's loving communion or the goodness of his rule.

Because of that fact we can never construe an antithesis between the covenants of "works" and "grace." We err if we interpret this distinction as though God's original covenant had to do with our work or our achievement or our fulfillment of his law, while the later covenant of grace has reference to the pure gift of his mercy apart from all our works. If we assume this we are compelled to say that God's original relation to man was strictly "legal," or that the structure of that relation was determined by man's merit. In that case, we lose sight of the fact that man's obedience to God's command can never be different from a thankful response to God's own fellowship. Therefore S. G. DeGraaf has rightly said that the concept which sees God's favor only at the end of man's way of obedience is open to serious dispute. Man participates in God's favor, communion, and love already at the very beginning. In that fact we see the awful reality of his guilt and apostasy.[4]

Though Berkouwer does not himself reject the language of the Covenant of Works, he questions its usefulness.[5] Berkouwer's testimony may be considered suspect by some because of the influence of Karl Barth on his thinking, but there is nothing in the quotation above to justify dismissing his views out of hand, nor will there be many who would wish to deny that he is a Reformed theologian, however inadequate his views on certain issues.

More respected among conservative Reformed people, though less well-known, is the theologian mentioned in the quotation above, S. G. DeGraaf, who altogether repudiated the notion of a Covenant of Works because he believed it created more problems than it could solve.[6] If Berkouwer is less than thoroughly Reformed in the eyes of some, the same cannot be said of DeGraaf. His eminence as a Reformed theologian is only reduced by the fact that so much of what he wrote is only available in Dutch.

The most famous example of a Reformed theologian denying the covenant of works is that of John Murray. Here we have a theologian whose credentials as a guardian of Reformed orthodoxy are unquestioned. Also, in distinction from the groups and men mentioned above, John Murray was a Presbyterian, holding to the Westminster Confession of Faith. He did not consider taking exception to the doctrine of a Covenant of Works a denial of theology of the Westminster Confession, though in general his approach to the Confession is strict.

Rather than extending the list of Reformed theologians who deny or question a Covenant of Works or getting into the subject of interpreting the Covenant of Works — for even among those who regard the doctrine as important, the interpretation of the Covenant of Works varies significantly[7] — suffice it to say that we can only insist that the doctrine of the Covenant of Works function as a test of Reformed orthodoxy if we are willing to create a new standard, one which adds not only to the Three Forms of Unity a doctrine which they do not profess, but which also condemns Reformed denominations and theologians who have hitherto been respected leaders. Is this really what we want to do? Rather than protecting Reformed orthodoxy, would we not, rather, be establishing a Reformed fundamentalism?

One final point, those who imply that the doctrine of justification by faith stands or falls with the Covenant of Works have proved too much. Conservative Lutheran churches, following Luther himself, clearly confess the doctrine of justification by faith. But neither Luther nor his followers ever connected the doctrine of justification by faith with the notion of a Covenant of Works. Moreover, the Lutheran doctrine of the law and the Gospel is not exactly parallel to the Reformed doctrine of the Covenant of Works and the Covenant of Grace. Beyond the Lutherans, Evangelical Protestants in various denominations confess the doctrine of justification by faith without holding to a doctrine of the Covenant of Works. Do we really want to maintain or imply that Lutherans, and Evangelical Christians, in addition to covenant theologians like John Murray, all implicitly deny or fundamentally compromise the doctrine of justification by faith?


The doctrine of the Covenant of Works is controversial among Reformed theologians and thinkers. The debate should go on and participants can be expected to be enthusiastically committed to their views. This is a sign of health. At some point, by the grace of God, debate should bring consensus. The idea of an endless debate is a denial of the possibility of progress in theology beyond the Reformation. But a consensus has certainly not been reached yet. What we need now is serious discussion among Reformed pastors and teachers, seeking to come to a better understanding of the Scriptures.

In the same way that no single eschatological position has yet come to be recognized as the Reformed position and gained virtual confessional status, the doctrine of the Covenant in Reformed theology has been open to various interpretations. There is no one doctrine of the Covenant that can lay claim to a position of authority. Until Reformed churches do attain a united confession in this area, it does not promote the Reformed faith to cast aspersion on those who disagree with us, denying their right to be regarded as Reformed or implying that they are undermining the Reformed faith. The Westminster Standards do not have "papal" status among the Reformed creeds, nor do Reformed Christians regard the Reformed creeds as the ultimate criterion for truth in theology. If we believe in Sola Scriptura, we need to practice it, while working together in love to promote the growth of the Truth of Christ and His Gospel.


[1] Meredith Kline, in particular, clearly implies that without the doctrine of the Covenant of Works, the Reformed Gospel is compromised. See for example, "Covenant Theology Under Attack." Joseph Morecraft, III, also seems to be of this opinion, beginning his discussion of the Covenant of Works with a quotation from Wilhelmus a Brakel that avers those who deny a covenant of works will not understand the covenant of grace of the mediatorial work of Christ. In, "The Covenant of Works" in The New Southern Presbyterian Review, vol. 1, no. 2, Fall, 2002, p. 112.

[2] John Murray, Collected Writings of John Murray: 4, Studies in Theology, "Covenant Theology," (Edinburgh: Banner of Truth, 1982), pp. 217-18.

[3] "Contours of God's Covenant," A Unity Committee Report, Clarion, vol. 48, no. 16, Aug. 6, 1999.

[4] G. C. Berkouwer, Studies in Dogmatics: Sin (Grand Rapids: Eerdmans, 1971), p. 206-207.

[5] Ibid., p. 207.

[6] Ibid.

[7] One of the ironies of this whole debate is that Joseph Morecraft, apparently one of those who wishes to insist that the Covenant of Works is a test of Reformed orthodoxy, interprets the Covenant of Works in a manner that Meridith Kline claims to anti-Reformed. If Kline's interpretation of the Covenant of Works is correct, Joseph Morecraft's affirmation of that covenant is so compromised by his interpretation that he, too, is guilty of undermining the Reformed faith! See, Joseph Morecraft, III, "The Covenant of Works" in The New Southern Presbyterian Review, vol. 1, no. 2, Fall, 2002.

 site design and maintenance
BERITH.ORG  —  Copyright © 2002 by Ralph Allan Smith.  All rights reserved.