The roots of Reformed covenant theology are as deep as the Christian revelation and tradition is old. Its importance to the Reformed faith cannot be overstated. The great Princeton theologian, B. B. Warfield called federal (covenant) theology, “architectonic principle” of the Westminster Confession of Faith
Early Fathers: Present But Undeveloped
Until the Pelagian controversy (late 300’s) and the semi-Pelagian (early 400’s) controversies following that, the church did not have a highly developed doctrine of salvation. The early church also had a theology of the covenant which is best described as latent, but undeveloped. The early fathers used the doctrine of the covenant in five ways:
- To stress the moral obligations of Christianity;
- To show God’s grace in including the Gentiles in the Abrahamic blessings;
- To deny that Israelites received the promises simply because they were physical descendants of Abraham;
- To demonstrate the unity of the divine economy of salvation;
- To explain the discontinuity between the old and new covenants in Scripture.
The greatest of all the early fathers, however, was Augustine of Hippo (354–430 AD), the giant upon whose shoulders the rest of the church has stood. In his greatest work, The City of God (16.27), he clearly taught the outlines of what would become central elements in classic Reformed theology, the covenant of works and the covenant of grace.
But even the infants, not personally in their own life, but according to the common origin of the human race, have all broken God’s covenant in that one in whom all have sinned. Now there are many things called God’s covenants besides those two great ones, the old and the new, which any one who pleases may read and know. For the first covenant, which was made with the first man, is just this: “In the day ye eat thereof, ye shall surely die.” Whence it is written in the book called Ecclesiasticus, “All flesh waxeth old as doth a garment. For the covenant from the beginning is, Thou shall die the death.”
There are two Adams in the history of salvation. As the Puritans had it, “In Adam’s fall, sinned we all.” He was our federal representative. Christ, of course, is the Second Adam, the federal representative of all the elect. Augustine considers that God made a legal covenant with Adam, that he was under the Law and he distinguishes here and later in the chapter, between the Law and the Gospel. This distinction was largely lost in the Medieval church, but it was one of the great recoveries of the Reformation.
Considering the history of salvation, the old and new covenants are both expressions of the Gospel. Most importantly we must note that Augustine turned to covenant theology against the Pelagians (who denied original sin) and against the semi-Pelagians, who affirmed original sin, but who argued that we could cooperate with divine grace for our righteousness before God.
For most of the Medieval period, the Western (Latin) church and the major theologians agreed that God says what he says about us, because we are what we are. That is, God can only call people righteous, if they truly are righteous, inside and out. This, they thought, will happen when sinners are infused with grace so that they become saints. Justification was a matter of cooperation with divine grace, faith is obedience and doubt is of the essence of faith.
The major development in medieval covenant theology was the proposition by great Franciscan theologian, William of Ockham (1285–1347) and later by Gabriel Biel (1420–95) that God does not say what he says (e.g., “you are just”) because we really are just, but rather, because we have met the terms of the covenant to cooperate with God. This is known as the Franciscan Pactum theology. Their slogan was, “To the one who does what he can, God will not deny grace.” You know this teaching as, “God helps those who help themselves.”
Ockham and Biel were teaching that God rewards sinners with a kind of merit when they do their best. He overlooks their sins and treats them as if they had fulfilled the terms of the covenant, i.e., as if they had kept the Law. It was against this very teaching that Martin rebelled in the Protestant Reformation.
Covenant Theology in the Reformation
Though Martin Luther (1483–1546) came to hate the covenant theology of the Franciscans, he did not abandon every part of it. Though he did not work out a complete covenant theology, as he became a Protestant (1513–21) Luther taught Paul’s doctrine of original sin, absolute divine sovereignty in salvation (double predestination), the imputation of our sin to Christ and his righteousness to us and faith as the alone instrument of justification. According to Luther, we are not justified because we are sanctified. He, with Calvin and all the Protestants, did not reject the idea of merit, but he learned that it is not our merits produced by grace which satisfies God, it is Christ who merited our justice and his merits are imputed to sinners.
Luther expressed these truths in his distinction, in justification, between Law and Gospel. The latter is the good news about what Christ has done for sinners. The former is bad news for sinners. Any time Scripture says, “Do this and live” (Luke 10:28) it is speaking Law. Whenever it says, “I have done that you might live” it is speaking Gospel. Though some Reformed theologians have suggested that we disagree with Luther on this principle, B. B. Warfield reminded us that it is “misleading to find the formative principle of either type of Protestantism in its difference from the other; they have infinitely more in common than in distinction.” Our doctrine of justification is one of those things we have in common.
One reason why Luther did not speak much about covenant in his later writings was that the idea had come to be associated with the Franciscan theologians whom he had publicly repudiated. Another possible reason is that Huldrych Zwingi (1484-1531) spoke more about the covenant. Zwingli, however, also taught a covenant of works before the fall and a covenant of grace after the fall. He especially described the sacraments in terms of the covenant, and our response to grace. His emphasis on our responsibility in the covenant made it sounds to Luther as if he agreed with Ockham.
One of the lesser-known Protestants between Luther and Calvin was Johannes Oecolampadius (†1531). For the time, Oecolampadius taught a remarkably mature covenant theology including the doctrine of the covenant of redemption, the covenant of works and the covenant of grace. Indeed, the great Reformed theologian Amandus Polanus considered Oecolampadius the first Reformed covenant theologian.
For example, the covenant of redemption, though a relatively new doctrine and though not worked out in detail, is found more fully in Oecolampadius’ than in other theologians of the period. In 1521 He spoke of the Father’s covenant with the Son, and taught that the covenant of grace is an outworking of this covenant.
As it came to be expressed in 17th century Reformed theology, the covenant of redemption (pactum salutis or consilium pacis, the counsel of peace) taught that the Father required that the Son should obey in the place of the elect, that he should be their surety, i.e., he would meet the legal obligations of the elect, to atone for their sins, to bear the punishment for their sins and to meet the demands of the covenant of works (Law) and to merit the forgiveness of sins and positive righteousness (imputed) to his people.
The Son, as the second party to this covenant, graciously, freely, willingly accepted the terms of this covenant. The Father promised several things, among them a sinless humanity, the Holy Spirit without measure, cooperation in the Son’s work, the authority to dispense the Holy Spirit and all authority on heaven and earth, numerous rewards for completing the probation as the 2nd Adam. Should the Son meet the terms of this covenant, he would merit the justification of his people and be vindicated by his resurrection.
His most important work on covenant theology were his lectures on Isaiah delivered in 1523-24. In those lectures he described the covenant of grace as one-sided in origin and two-sided in administration. Therefore, the covenant of grace, considered as God’s Gospel offer to sinners, must be said to be unconditional in the sense that we do not prepare for it, nor do we cooperate with it. We simply believe the Gospel promise. The covenant of grace can be said to be conditional when we consider the administration of the covenant in the life of the church. Christians are obligated, as a response to grace to attend to the preaching of the Gospel and the administration of the sacraments. These are the basic lines of all Reformed covenant theology through the 19th century.
Like Luther, John Calvin (1509–64) taught the substance of the more highly developed federal theology. Like Bullinger, most of his discussion of the covenant concerned the history of redemption from Adam to Christ and the continuity of the covenant of grace. Nevertheless, he taught the substance of what became classic Reformed federal theology: the covenant of redemption in eternity (pactum salutis), the covenant of works before the fall and the covenant of grace after the fall.
Some scholars deny that Calvin taught the same covenant theology as the later Reformed theologians since he did not use the same vocabulary as they did. This is ironic since Calvin himself complained about the Romanists who would not allow him to use the expression “faith alone” (Institutes, 3.11.19) since the word “alone” (sola) is not used expressly in Scripture. For Calvin the Law (covenant of works) kills sinners and the Gospel (covenant of grace) justifies and sanctifies them through faith alone, in Christ alone. He used the covenant to express those fundamental truths.
Beginning with the basic distinction between Law (guilt) and Gospel (grace) he also used the covenant to include a more prominent place for sanctification or gratitude. We know these as the three parts of our catechism.
Heinrich Bullinger (1504–75) published perhaps the first treatise devoted to explaining the covenant in 1534. Like Calvin and the early fathers, he used the covenant to teach the unity of God and his salvation. He contributed to the Reformed tradition of using the covenant of grace as a summary of biblical theology. Caspar Olevian (1536-87) would later do this same thing in three works, chiefly in his book, On the Substance of the Covenant of Grace Between God and the Elect (1585) and Johannes Cocceius (1609–69) and Herman Witsius (1636-1708) write entire systematic theologies structured by the covenants of redemption, works and grace.
Covenant Theology in Reformed Orthodoxy
The two most important Reformed covenant theologians of the late 16th century were the chief authors of our catechism, Caspar Olevianus (1536–87) and Zacharias Ursinus (1534–83). Ursinus lectured on the covenant theology of the catechism in Heidelberg for about fifteen years and later, until his death, at his school in Neustadt. His covenant theology is clear from his lectures and Larger Catechism (1561) which he used in his seminary and university classes.
Ursinus defined covenants in general in terms of the covenant of works, since the Gospel can only be understood against the background of the Law. In the covenant of works, God placed conditions upon Adam, the head of all humanity, which he accepted, to obey his covenant God. The sign of the covenant was the tree of life. If Adam had kept the covenant, he would have entered a state of eternal blessedness. For the same reason, transgression of the Law covenant meant eternal punishment.
According to Ursinus (and all the classic Reformed theologians) Christ, the representative of all the elect, fulfilled this covenant in his active and passive (suffering) obedience. Because Christ obeyed the Law for his people, there is a Gospel covenant. Unlike the covenant of works made with sinless Adam, the covenant of grace is made with sinners, who need a mediator, a covenant keeping Savior, who fulfilled the Law, satisfied God’s just wrath for sinners. This is the difference between Law and Gospel (Larger Catechism, Q. 36).
For Ursinus, as for Olevian and the mainstream of Reformed Orthodoxy, the covenant of works stands for the Law, which is not gracious but relentless in its demand for perfection. The covenant of grace stands for Gospel, which means that Christ our Mediator and substitute has met the terms of the Law for us. It was a covenant of works for Christ and he has made and Gospel covenant for us.
In his On the Substance of the Covenant, Caspar Olevian argued that the covenant can be considered in a broader and narrower sense. In the narrower sense, the covenant can said to have been made only with the elect. It is the elect who are united to Christ by grace alone, through faith alone, in Christ alone, who receive the benefits of the covenant, strictly speaking.
Considered relative to the administration of the covenant of grace, the covenant must be said or thought to be with all the baptized, since only God knows who is elect. Therefore we baptize on the basis of the divine command and promise, and we regard covenant children (before profession of faith) and all who make a credible profession of faith as Christians until they prove otherwise. Those who are in the covenant in the broader sense or externally, do receive some of the benefits of the covenant (Hebrews 6:4-6), but they do not receive what Olevian called the “substance of the covenant,” or the “double benefit of the covenant: justification and sanctification. Only those who are elect actually appropriate, by grace alone, through simple faith alone, the “double benefit” of the covenant. Both Olevian and Ursinus taught the pactum salutis, the covenant of works as a Law covenant and the covenant of grace as a Gospel covenant.
Why is the covenant not more prominent in the Heidelberg Catechism? The answer is in two parts. One of the chief aims of the catechism was to present the Reformed faith to Lutherans in the Palatinate. By 1562, when the work on the catechism was underway, the Lutherans had strongly criticized Reformed covenant theology. Therefore, the committee wanted a more ecumenical tone for the catechism. The second reason is that Ursinus and Olevian were commissioned to explain the catechism in the schools in terms of what we know as the classic Reformed federal theology: covenant of redemption, covenant of works and covenant of grace. Even though the catechism did not use the technical covenant language, the authors of the catechism clearly understood the catechism to teach the substance of covenant theology.
The theology of the early 17th century Reformed theologians William Ames (†1633), Johannes Wollebius (†1629) and Amandus Polanus (†1610) was written in the same direction as that of Olevian and Ursinus. The high point of Reformed federal theology was doubtless the work of Johannes Cocceius (†1669), Francis Turretin (†1687), J. H. Heidegger (†1698) and Herman Witsius (†1708).
Cocceius is notable for writing the most comprehensive account of the Biblical covenants, perhaps in Christian history. He was opposed in several respects by Gisbert Voetius (†1676). Cocceius’ chief work was his Summary of the Doctrine Concerning the Covenant and Testament (1648). He is most famous for doctrine of the progressive abrogation of the covenant of works in history. This along with his rather more liberal view of the Sabbath along with his support of the new philosophy of Rene Descartes (†1650), provoked a strong reaction from the Voetians. He was primarily a Biblical theologian interested in the progressive revelation of the accomplishment of salvation. His opponents were more interested, perhaps, in systematic theology and the application of redemption to sinners.
Like most of the earlier federal theologians, he saw the history of salvation as the expression of the eternal covenant of redemption. He distinguished strongly between the covenant of works as Law and the covenant of grace as Gospel. On these main points, he found complete support in Heidegger, Turretin and Witsius.
Francis Turretin is most famous for his Institutes of Elenctic Theology (1679–85) which was the theology text for Princeton Seminary until Charles Hodge wrote his own system in English. Turretin supported the mainlines of Reformed covenant theology and defended them against the Socinians, Arminians and Amyrauldians, all of whom attacked Protestant theology and Reformed federalism because they believed it would give Christians an excuse to sin and because they thought it unreasonable.
J. H. Heidegger and Turretin produced the Helvetic Consensus Formula (1675), a brilliant summary of Reformed covenant theology in the late 17th century. In Canon VII they taught that, “Having created man in this manner, he [God] put him under the Covenant of Works, and in this Covenant freely promised him communion with God, favor and life, if indeed he acted in obedience to his will.” If Adam kept the covenant, he would enter into eternal blessedness, which was signified by the Tree of Life (Canon VIII). What Adam refused to do Christ the Second Adam did for us. They explicitly criticized the Arminians who rejected the covenant of works (Canon IX). Following the Reformed mainstream, they also taught the eternal covenant of redemption (Canon XIII). Against the Remonstrants, they upheld faith (sola fide) as the only condition for entering the covenant. Obedience flows from justification out of gratitude. “In accordance with these two ways of justification the Scripture establishes these two covenants: the Covenant of Works, entered into with Adam and with each one of his descendants in him, but made void by sin; and the Covenant of Grace, made with only the elect in Christ, the second Adam, eternal. [This covenant] cannot be broken while [the Covenant of Works] can be abrogated.”
Herman Witsius attempted to explain, summarize and develop Reformed covenant theology, trying to build bridges between the Cocceians and the Voetians. Like the tradition before him, he identified the covenant of works with the Law and the covenant of grace with the Gospel. The difference between the two covenants is that Christ our Mediator has met the terms of the Law for all elect sinners.
One of the tensions, which remained unresolved in the 16th and 17th centuries, was the matter of the nature of Israel’s relations to the covenant of works. All the classic theologians took some account of the works language, while maintaining the essential unity of the covenant of grace. Some, such as Cocceius and Witsius suggested that Israel was in a sort of probation relative to the land, but not justification.
Since Calvin, Reformed theologians have also spoken of God’s graciousness in entering into covenant relations with Adam. This language has often been misunderstood. John Owen and John Ball (like Herman Bavinck later) called attention to the disproportionality between God and Man and God’s freedom in making the covenant of works. None of these theologians, however, denied that the covenant of works was a legal covenant and none of these theologians said that the covenant of works was a gracious covenant. This important distinction has sometimes been ignored.
In the United States the Princeton theologians e.g., Charles Hodge (†1878), B. B. Warfield (†1921) G. Vos (†1949) and J. Gresham Machen (†1936), and in the Netherlands H. Bavinck (†1921) followed the main lines of the classic view, teaching the covenant of redemption, the covenant of works (Law) and the covenant of grace (Gospel).
The single greatest influence on covenant theology in the 20th century has been that of the Swiss theologian Karl Barth (†1968). Having rejected genuine historicity of Scripture in favor of a theology of personal encounter with the Word, Barth rejected much of classic Reformed covenant theology as “scholastic” and unbiblical. He rejected the covenant of redemption and the classic distinction between the covenant of works and the covenant of grace as “legalistic.” In Barth’s theology, grace overwhelmed Law. Many contemporary Reformed theologians, including T. F. Torrence and G. C. Berkouwer followed this critique of the Reformed tradition.
In the Netherlands, with the 1892 merger of the Afscheiding (1834) and the Doleantie, (1886) tensions grew between the followers of Abraham Kuyper (†1920) and some of the Secession theologians, culminating with the deposition of Dr. Klaas Schilder (†1952) near the end of Synod Sneek-Utrecht (1939–43). Like the tradition, Kuyper taught the three-covenant view, distinguishing between those who were in the covenant only outwardly and those who were in the covenant inwardly. He also revised the traditional views in certain places.
Whereas some in the Reformed tradition had discussed the possibility of eternal justification, Kuyper elevated that speculation to a central place in his doctrine of the covenant, identifying it with the covenant of grace, concluding that we baptize on the basis of presumed regeneration rather than on the basis of command and promise. From the perspective of the tradition, this was a move bound to provoke a reaction.
Worried about the possibility of moral laxity among the covenant people, Schilder rejected the Kuyperian distinctives, emphasizing the unity of a gracious covenant before and after the fall, and the responsibility of those within the covenant to appropriate its benefits. In so doing, he also rejected important aspects of the traditional view including the covenant of redemption and the distinction between the covenants of works and grace as well as the distinction between the broader and narrower senses of the covenant of grace.
In the controversy between the Kuyperians and the Schilderites, however, covenant theology turned away from relating covenant to justification in favor of relating covenant to election.
In the first half of the 20th century, in the United States, M. J. Bosma and Louis Berkhof (†1957) upheld the classic view. At Westminster Seminary, however, John Murray (†1975) was also re-formulating covenant theology. He rejected the terms “pactum salutis” and “covenant of works,” though he continued to teach the substance of both. In reaction to fundamentalist dispensationalism which rejected the unity of the covenant of grace, Murray emphasized the continuity of the covenant by defining the covenant primarily in terms of grace. Nevertheless, he taught the Protestant doctrine of justification.
As in the case of Schilder, Murray’s revision of the tradition left a tension between his covenant theology and his doctrine of justification. Professor Norman Shepherd, also of Westminster, resolved the tension by proposing a revision of the doctrine of sola fide, which created a serious controversy culminating in his departure from the seminary in 1981.
In reaction to Murray and Shepherd, Meredith Kline (1922–2007) returned to the classic correlation between the Law and Gospel distinction and the distinction between the covenant of works and grace. To answer the liberals and dispensationalists, he argued that there is one covenant of grace in the history of salvation, but that the Mosaic covenant, though gracious with respect to justification and salvation, had a works element relative to Israel’s tenure in Canaan. In this way, the Mosaic theocracy becomes a re-publication of the covenant of works. Though it appears to some, in our time, to be novel the main features of this view are quite traditional.
It is clear that, through the 20th century, the great consensus which had been sustained since the Reformed covenant theology since the 1520’s has fragmented. The causes seem to be three: The influence of Barth, even among confessional theologians has been greater than many recognize. Second, in reaction to Modernity, many have become practical fundamentalists, little interested in the Reformed tradition and third, we have forgotten our Protestant ABC’s in the doctrine of justification and salvation. Certainly, a first step toward repairing the consensus is to take a serious look back at our greatest covenant theologians.