Editors’ Note: As noted in our recent interview with authors Peter Gentry and Stephen Wellum, Kingdom through Covenant: A Biblical-Theological Understanding of the Covenants (Crossway, 2012) is a groundbreaking contribution to any discussion about the intersection of exegesis, biblical theology, and systematic theology. In 848 pages, Gentry and Wellum have made a substantial case for an independent middle path between dispensational and covenant theology—a case that demands a response.
We’ve invited three noted scholars to evaluate Gentry and Wellum’s proposal: Darrell Bock, research professor of New Testament studies at Dallas Theological Seminary; Douglas Moo, professor of New Testament at Wheaton College outside Chicago; and Michael Horton, professor of systematic theology and apologetics at Westminster Theological Seminary (California). Today we hear from Horton.
For the seriousness with which it handles the issues, its depth of research and analysis, its approach on many issues, and the respectful description of alternative positions, Kingdom through Covenant strikes me as a model for the deeper and richer conversations that we need in our circles. However, since I’m offering a review from a traditional “covenant theology” perspective, I will skip over a host of edifying discoveries and get right to the point.
If I understand it correctly, the main argument of the authors is that dispensationalism and covenant theology both fail to read the Bible in a sufficiently typological way (pointing to Christ), though at different points: an unconditional and inviolable promise of either an ethnic people and geo-political land or of a “genealogical principle” that underwrites the baptism of covenant children and a “mixed body” ecclesiology. Consequently, covenant theology results in a one-to-one correspondence between circumcision and baptism, Israel and the church.
Abrahamic and New Covenants, Israel and the Church: Too Much Continuity?
I’ll grant that especially in anti-Anabaptist polemics, Calvin and his heirs have sometimes so stressed continuity within the one Abrahamic covenant of grace that the newness of the new covenant is insufficiently appreciated. Long ago, Voetius and Cocceius represented the wideness of the spectrum in covenant theology on that question, and more recent Reformed scholars (e.g., Vos, Ridderbos, Murray, Kline, Gaffin, et al.) have explored the qualitatively new blessings in the new covenant. So while I definitely think this criticism keeps us on our toes, there’s enough out there to qualify the charge that we see the Spirit’s work as “basically the same across redemptive history.”
What does hold across the various administrations of the covenant of grace, however, is God’s unilateral promise to provide a Savior in whom the families of the earth will be blessed. That’s just the rub, though, according to the authors. The classification of “unconditional” and “conditional” covenants isn’t helpful, they argue, because there are elements of each in every biblical covenant.
However, their argument assumes that the mere presence of commands indicates a mixture of unconditional-conditional aspects in the basis of the covenant itself. At this point, Reformed theology has traditionally appealed to a distinction between basis and administration. The mere presence of commands says nothing about the basis of a covenant itself. Circumcision (like baptism) identifies the members of the covenant, so if one is not circumcised, he is “cut off.” Nevertheless, one is not justified because he is circumcised, as Paul indicates in Romans 4:11. That would turn conditions into the basis rather than the administration of the covenant. Commands function in a law-covenant as the basis for blessing or curse: the swearer’s perfect, personal, perpetual obedience is the ground, ratified by a public assumption of the covenant obligations on one’s own head. In the covenant of grace, however, commands function as the “reasonable service” that we offer “in view of God’s mercies.”
The Abrahamic promise was a unilateral gift. In Genesis 15, God alone swears the oath and then walks through the pieces assuming its solemn threats. The gracious promise includes an earthly land and seed as well as a heavenly land and seed. This grace is the basis for Israel’s inheritance of the land in the first place, as Deuteronomy 8 and 9 point out so clearly, along with the prologue to the Ten Commandments in Exodus 20:2. The earthly promise of a “holy nation” generates within history a temporary kingdom that is typological of the heavenly promise of an everlasting kingdom with global scope: Christ with his worldwide body.
Although the Mosaic covenant is certainly in service to the covenant of grace in various ways, in both form and content it is quite different from the one-sided promise to Adam and Eve after the Fall, Abraham and Sarah, David, and the new covenant. At Sinai (as in Eden), the servant-people swear the oath and Moses is the mediator, splashing blood on them “in accordance with all the words you have spoken, saying, ‘All this we will do'” (Exod. 24:8). As the Last Adam and True Israel, Jesus fulfills this law-covenant, confirming his oath with his own “blood of the covenant.” So now we inherit in a covenant of grace what our mediator has merited for us by fulfilling the covenant of works. Gentry and Wellum offer a tremendous defense of Christ’s active obedience, but this requires the sharp distinction between unconditional and conditional covenants that the original context of each covenant supports. So Israel inherits the earthly land by (Abrahamic) promise, but remains in the land by (Sinai-treaty) obedience. With respect to this covenant, E. P. Sanders is exactly right with his formula: “Get in by grace, stay in by obedience.” However, confusing these two covenants is precisely what provoked Paul’s argument in Galatians.
In the meantime, the promise of the Seed in whom all the families of the earth would be blessed continued—as it does today (Gen. 3:15; 28:14; Acts 3:25; Gal. 3:16). Paul doesn’t treat the Abrahamic covenant merely as typological of Christ, but sees the new covenant as the only possible fulfillment of the worldwide promise to Abraham. The Mosaic covenant was essential in the historical unfolding of the covenant of grace leading to Christ, but it was strictly temporary, typological, conditional, and limited to one geo-political nation.
The new covenant is indeed new—not like the Sinai covenant, which had “planned obsolescence” built into it (Jer. 31:32; cf. Heb. 8:9). However, it is the realization of the Abrahamic covenant (Acts 3:25). So despite repeated discontinuity between old and new covenants (Sinai and Zion, “two covenants”: law and promise, etc.), it is striking that the apostolic message explicitly treats the new covenant as the consummation of the Abrahamic covenant rather than its abrogation (for example, Acts 10:9-43, 15:7-17; Romans 4 and Galatians 3-4). Therefore, in spite of the obvious differences between the redemptive-historical quality of blessings enjoyed by Old and New Testament saints, respectively, believers and their children are “heirs according to the promise” just as Abraham and his heirs were. In both cases, there are those who embrace the promise and those who don’t. That’s why Romans 9 hits its mark: God’s promise hasn’t failed; even in the OT, God has always reserved his freedom in election even among those who are outwardly identified with the covenant. The apostles address their Jewish-Gentile congregations just as the prophets addressed Israel, only with respect to the everlasting-spiritual promises (which are not abrogated) rather than the geo-political theocracy (which is).
Gentry and Wellum seem to agree that the “geneological principle operative in the Abrahamic covenant” is indicated by the phrase, “you and your seed” from Genesis 17:7 (633). I agree, but if that’s true, then what do we say of Peter’s repetition of this formula in Acts 2:39, specifically in connection with baptism? When we add to this the instances of household baptisms (with only one believing parent mentioned) in Acts (16:15, 32-33; 18:18) and 1 Corinthians 1:16, and the mention of children being sanctified by one believing parent in 1 Corinthians 7:14, the cumulative case seems to place the burden of proof on the Baptist position. The new covenant is certainly greater. For one thing, it’s more inclusive (not less!): not only Jews, but Gentiles; not only males, but females also receive the seal of the covenant and are indwelled by the Spirit. Where are the explicit passages to indicate that with such a profound expansion of blessings to “all the families of the earth”—indeed, to believers and their seed (Acts 2:39)—children are excluded in an ostensibly better covenant?
The authors contend, however, that the new covenant is distinguished from the Abrahamic by defining the church (or covenant community) as identical with its elect and regenerate members. I’m not at all persuaded that the often-cited source for this position—Jeremiah 31:34—entails (much less requires) that conclusion. Does “all will know me” in Jeremiah 31:34 mean each and every member without exception, or is it, like “all the families of the earth,” an expression of the pervasiveness of blessing in the new covenant? It will reach from the greatest to the least, women as well as men, children as well as elders, Gentiles as well as Jews, and so forth. Isn’t it an argument for the priesthood of all believers, not that each and every professing member will believe?
Covenant Membership: How New Is the New Covenant?
I disagree with the authors when they suggest ambiguity in Reformed thinking regarding the members of the covenant of grace. True, the Westminster Confession says that the elect are the proper subjects, but the Confession goes on to say that “the visible church . . . consists of all those throughout the world that profess the true religion, and of their children, and is the kingdom of the Lord Jesus Christ, the house and family of God, out of which there is no ordinary possibility of salvation” (25.2). This is the standard Reformed view, drawing on the invisible-visible church distinction, which the authors seem to reject.
I wonder how the authors can affirm “the running tension between the ‘already-not yet'” (690) if the visible church is already identical with the invisible church. This has enormous consequences for pastoral ministry. In traditional Reformed treatments it is frequently warned that pastors and elders have authority only to determine credible professions of faith, not to actually determine whether people are elect or regenerated. While correcting or excluding members exhibiting open unbelief and non-repentance, a long-standing Reformed criticism of Anabaptist ecclesiologies is that they assume a separation of sheep and goats prior to Christ’s final judgment. I think this is dangerous on a number of practical levels in dealing with people under our charge.
Especially given the robust exegesis and theological argumentation elsewhere, I was expecting to find more rigorous engagement with the “apostasy” passages. It almost seemed like these passages (especially in Hebrews) were irrelevant because of the a priori commitment to discontinuity between Israel and the church (692).
Hebrews 6 assumes a category of covenant members (Jewish Christians) who are in some sense beneficiaries of the Spirit’s common work in the church through the means of grace, but revert to Judaism. Hardly an empty set, they are covenant members “who have once been enlightened” (ancient church documents use “baptized” and “enlightened” interchangeably), “who have tasted the heavenly gift [the Supper], and have shared in the Holy Spirit, and have tasted the goodness of the word of God [preaching] and the powers of the age to come, and then have fallen away” (vv. 4-6a). Specifically, they have fallen away from the new covenant, reverting to the old. However, there is no forgiveness anymore if they go back to the sacrificial system of the temple (v. 6b). “For land that has drunk the rain that often falls on it, and produces a crop useful to those for whose sake it is cultivated, receives a blessing from God. But if it bears thorns and thistles, it is worthless and near to being cursed, and its end is to be burned” (vv. 7-8). Through their covenant membership they have shared in God’s blessings, and now, if they respond in unbelief, they will bear the covenant curses.
Concerning covenant theologians, the authors suggest, “Ironically, however, they agree with the Arminian exegesis and conclusion as applied to full covenant members who are not elect” (75). This isn’t quite accurate. We hasten to add the qualification in verse 9: “Though we speak in this way, yet in your case, beloved, we feel sure of better things—things that belong to salvation” (v. 9). The writer does not know for certain that each and every reader is regenerate, but exercises charity since they are not among the open apostates.
A Baptist interpretation cannot account for this category of covenant beneficiaries who spurn the objective blessings delivered to them and fall away, while an Arminian interpretation cannot account for the distinction of this group from those who were in fact united to Christ. If our theological system cannot account for this group—neither non-members nor regenerate—then we need a different theological paradigm. It’s covenant theology that accounts for this tertium quid between “foreigners to the covenant” and “regenerate members.” In fact, the warning is emphatic in chapter 10 against “the one who has spurned the Son of God, and has profaned the blood of the covenant by which he was sanctified, and has outraged the Spirit of grace” (v. 29). The writer warns his readers not to follow the example of Esau “who sold his birthright for a single meal” (12:16).
All of this fits well with Jesus’s distinction in his parables between a seed that at first begins to grow but is choked by weeds, or weeds sown among the wheat, or fish caught in the net and sorted out (Matt. 13). It isn’t just a distinction between the world and the church, since the fish are in fact caught in the covenantal net of the kingdom and then sorted on the last day.
Similarly, what about the warning that Paul gives in Romans 11 against Gentile bragging? He argues that “if the root is holy, so are the branches.” It’s one tree. Jewish branches that didn’t yield faith were broken off to make room for living Gentile branches that share the faith of Abraham in Christ. And yet he adds, “They were broken off because of their unbelief, but you stand fast through faith. So do not become proud, but fear. For if God did not spare the natural branches, neither will he spare you” (vv. 16b-21). The whole tree is holy, but dead branches will be pruned.
The whole church of Corinth is addressed as “the church of God that is in Corinth, to those sanctified in Christ Jesus” (1:2). And yet, among that very number are members he will later upbraid them for not excommunicating! The unrepentant member is “leaven” that will corrupt the whole lump, implying that such a person is in fact part of the lump in some sense (1 Cor. 5:6-7). In fact, Paul clearly says that this is a judgment exercised within the church, toward professing members who live in open rebellion. He says explicitly that we have no business “judging outsiders.” “Is it not those inside the church whom you are to judge? God judges those outside. Purge the evil person from among you” (5:9-13). Nor is there a call to determine who is truly regenerate, but only who is living in flagrant and unrepentant contradiction to his or her public profession. Those who are excommunicated are “cast out” of the covenant community, “removed from among you” (Gal. 4:30; 1 Cor. 5:2).
What accounts for this category: holy by public identification, but not united to Christ through faith? To be claimed as part of God’s holy field comes with threats as well as blessings. Covenant members who do not believe are under the covenant curse. How can they fall under the curses of a covenant to which they didn’t belong? If faith is the only way into membership (693), then why all the warnings to members of the covenant community to exercise faith and persevere in faith to the end?
Which Comes First: God’s Promise or Ours?
Reformation confessions teach that the church is present “wherever the word is rightly preached and the sacraments are rightly administered.” Reformed confessions add church discipline. The accent falls on defining the church as the place where God is at work creating and confirming faith in the hearts of its visible members through preaching, sacrament, and spiritual oversight. By contrast, the Anabaptist-Baptist ecclesiology lodges the church’s visibility in the members. Their response to the covenant, not God’s promise, creates the church. The church comes into being not whenever God speaks, washes, and feeds his flock, but when the hearers trust and obey. Thus, baptism and the Supper become chiefly our means of response rather than God’s means of grace. In my view, this reversal of the priority in the covenantal relationship tends to work against the best monergistic impulses of Calvinistic Baptists.
When the authors summarize the meaning of baptism as signifying the believer’s faith, marking one as God’s child, and add that “baptism always assumes faith for its validity” (700, emphasis added), it’s clear that our differences concern the nature of baptism itself and the relative priority of God’s promise and our faith. Covenant theology doesn’t teach that the covenant of grace itself is “breakable” (67). God promises his saving grace in Christ to each person in baptism, whether they embrace this promise or not. Yet they must embrace the promise in faith. Otherwise, they fall under the covenant curse without Christ as their mediator. The word proclaimed and sealed in the sacraments is valid, regardless of our response, but we don’t enjoy the blessings apart from receiving Christ with all of his benefits. Is baptism the believer’s act of testifying to a personal response, or God’s act of testifying to his everlasting pledge, which itself is the means through which the Spirit creates persevering faith in his elect? How we answer that question has a lot to do with whether the inclusion of children is even conceivable.
Gentry and Wellum do not interpret Colossians 2:11-13 as treating baptism as a replacement for circumcision: “[This] raises an obvious question: If the covenant signs are so similar in meaning then why did circumcision disappear as a covenant sign, especially for the Jewish Christian?” (80) However, the disappearance of ritual circumcision is entirely understandable if baptism replaced it. According to Paul circumcision was—at least for believers like Abraham—the seal of the righteousness Abraham had by faith (Rom. 4:11). Adult converts like Abraham receive the sign and seal of the covenant upon professing faith, while their children receive it unto that profession.
There are many other important issues I wish I could engage and many fruitful insights to explore in Kingdom through Covenant. In short, I look forward to the continuing conversation provoked by this thoughtful work.