The Upper Register Blog

Here’s a summary of Kline’s argument for infant baptism:

(1) Kline’s argument for infant baptism avoids the notion of presumptive regeneration and relies on the principle of parental authority.

(2) Kline appeals to the principle of vassal authority, that is, the principle that when a vassal takes an oath of loyalty to his suzerain, he brings not only himself but also his family under the covenant lordship of the suzerain.

(3) Kline sees this principle of vassal authority in the Abrahamic Covenant.

(4) Kline sees this principle of vassal authority continued from the Abrahamic Covenant (the pre-Messianic administration of the covenant of grace) into the New Covenant (the fulfilled, Messianic administration of the covenant of grace). Although Kline sees discontinuity between the Mosaic/Old Covenant and the New Covenant, he sees fundamental continuity along the underlying substratum of the one covenant of grace as it moves from promise (Abrahamic Covenant) to fulfillment (New Covenant).

(5) The case for continuity in the administration of the sign of the covenant on the basis of parental authority rests on the following pieces of evidence:

a. Paul’s teaching that our children are “holy” (1 Cor 7:14)
b. The parents who brought their children to be blessed by Jesus
c. Paul’s instructions to parents and children (Eph 6:1-4)
d. The oikos formula

With this post I’m done blogging through By Oath Consigned.

The fourth piece of evidence that baptism is administered on the principle of vassal or parental authority: 

(d) The oikos (household) formula

Finally, Kline appeals to the New Testament accounts of household baptisms where the household is baptized or brought into the church on the basis of the confession of faith of the parent (Acts 2:38-39; 10:2, 47-48; 11:14; 16:15, 33-34; 18:8; 1 Cor. 1:16; 2 Tim. 1:16; 4:19). The “oikos formula” refers to the recurring statement that “he/she and his/her (whole) household” believed and were baptized. This paedobaptist argument was famously articulated by Joachim Jeremias in the 1960s.

Kline thinks the oikos formula supports the paedobaptist view that the principle of parental authority carries forward into the New Covenant. Of course, there are difficulties with using these texts:  (a) we don’t know whether small children were actually present in any of these cases, and (b) since slaves were also members of the Greco-Roman oikos (household), the paedobaptist appeal to the oikos formula may lead to the unwanted conclusion that slaves should also be baptized. Kline recognizes these difficulties, but wishes to makes the following point (BOC, pp. 96-98):

We would simply observe that for the purpose of substantiating the authority principle of covenant administration the precise constituency of the households involved would not need to be determined. Whether or not there were infant children in one case or the other, or slaves in this or that household, households are mentioned along with the central authority figures in these instances, and these households had to consist of somebody in the category of household subordinates. Even with respect to the narrower question of whether parental authority is honored in the administration of the New Covenant, it would not matter whether conclusive evidence could be adduced proving that there were no children in any of these households; for if there were no children, then surely the households consisted of servants; and if it could be shown that servants were received into the church on the basis of the authority principle, it would follow a fortiori that the continuity with Old Testament practice included infants too …

The recurring mention of the household along with the central figure, whether in description of an existing God-fearing community, or in an invitation to salvation, or in an account of the acknowledgment of faith, or in a record of the administration of baptism, can very naturally be interpreted as the terminological reflex of a standard missions policy according to which the covenant community would regularly be enlarged through the accretion of household authority units. Indeed, it seems easier, particularly in the cases of prospective announcements of salvation and evangelistic proclamation (Acts 11:14; 16:31), to account for the recurrence of the appended reference to the household as a statement reflecting administrative policy rather than as a prediction based on a possible general rule that the sovereign soteric operations of the Spirit of God permeate intimate groupings of men. To explain the language of these declarations as meaning that the invitation with its terms was not confined to the householder but was extended to the members of his household, they, too, being invited to salvation on the same condition of faith, seems somewhat artificial; moreover, it would not explain the phenomenon of recurrence.

[Note: I have some additional information on the oikos formula based on Jeremias’s work that I’m trying to upload, but Adobe says the file is corrupted. Hopefully I can resolve the problem by tomorrow.]

The third piece of evidence: 

(c) Paul’s instructions concerning parents and children

Continuing in the same paragraph, Kline writes (BOC, p. 94):

Another significant fact is that Paul instructed the children of various congregations to obey their parents in the Lord, and in support of his charge cited the pertinent stipulation of the Sinaitic Covenant together with its accompanying covenantal sanction (Eph. 6:1-3; Col. 3:20; cf. Ex. 20:12). Clear confirmation is also found in Paul’s directive to covenant parents to bring their children under the nurturing and admonishing authority of the Lord (Eph. 6:4). In this exhortation the apostle takes for granted that it is the very authority of Christ as covenant Lord that reaches and claims children through the authority of their parents.*

It is therefore a matter of express scriptural teaching that the disciple of Christ is bound to bring those who are under his parental authority along with himself when he comes by oath under the higher authority of his covenant Suzerain. From this it follows that the Scriptures provide ample warrant for the administration of baptism to the children of confessing Christians, for baptism is the New Covenant rite whose precise significance is that of committal to Christ’s authority and of incorporation within the domain of Christ’s covenant lordship.

*The text reads:  “Fathers, do not provoke your children to anger, but bring them up in the discipline and instruction of the Lord [ἐν παιδείᾳ καὶ νουθεσίᾳ κυρίου]” (Eph 6:4). Kline appears to take κυρίου as a subjective genitive. This is the view taken by Ernest Best in his commentary on Ephesians:  ”The Lord [is] seen as the ultimate instructor who works through the father … The father mediates the Lord’s instruction” (Ernest Best, A Critical and Exegetical Commentary on Ephesians [ICC; Edinburgh: T&T Clark, 1998], 569-70). 

The second piece of evidence that the principle of parental authority continues in the New Covenant: 

(b) The parents who brought their children to Jesus to be blessed

Kline’s comments here are brief but telling (BOC, pp. 93-94):

In the discussion of infant baptism the episode of the bringing of the children to Jesus (Matt. 19:13-15; Mk. 10:13-16; Lk. 18:15-17) has been the source of considerable contention. But in support of the point we would make we need gather no more from that episode than that our Lord heartily approved when those with parental authority over these children exercised it to bring them to him and place them under the authority of his ministry. And that much at least would seem to be beyond debate.

Kline’s thesis is that the sign of entrance into the New Covenant (baptism) is administered not only to adult confessors, but also to their children on the principle of vassal or parental authority. The first piece of evidence that Kline appeals to is:

(a) Our children are “holy” (1 Cor 7:14)

Dealing with the problem of (religiously) mixed marriages, Paul says:  “For the unbelieving husband is sanctified through his wife, and the unbelieving wife is sanctified through her believing husband; for otherwise your children are unclean, but now they are holy.”

The problem that must first be addressed is what kind of “holiness” is in view here? Kline argues that it is cultic/covenantal holiness, not internal sanctification of heart that applies only to the elect (BOC, pp. 91-93):

Treating first the case of the children of the mixed marriages under discussion by the apostle, there does not seem to be any way to construe the holiness ascribed to them other than as a holiness of status … In what, then, does the holiness of the children’s status consist? In accordance with the biblical concept of holiness it will have to involve some sort of dedicatory separation unto the name of God, a consecration to his service and glory. Clearly it is not the holy consecration of subjective-spiritual condition … Since I Corinthians 7:14 provides evidence that the cultic corollary of the authority principle was operative in the apostolic church, this passage may be cited in support of the thesis that the authority principle is still in effect in the administration of the New Covenant, at least in the form of parental authority.

In other words, the holiness of covenant children here must be understood in a sense that could apply to both regenerate and unregenerate. It must be understood in a cultic sense as a holiness of status and separation unto God, not in the sense of internal holiness via regeneration. The cultic concept of holiness in the Old Testament applied even to the furniture of the tabernacle and the priests, without indicating salvific status. Therefore, holiness here is tantamount to covenant membership. And Paul clearly says that the children of at least one believing parent are “holy” in that sense.

The Baptist objection to the paedobaptist use of 1 Cor 7:14 is that if one wants to use this passage to prove that the child of at least one professing believer is to be regarded as a member of the church, then to be consistent one must include the unbelieving spouse as well. Here is Kline’s counter to the Baptist objection (BOC, p. 93):

While the similarity of terminology in verse 14a and 14b of I Corinthians 7 requires that the sanctification be of the same kind (i.e., cultic) in the case of the unbelieving parents and of the children, it is an unjustifiably wooden approach to the apostle’s words that insists that this cultic sanctification must apply in exactly the same manner in the two cases.  Rather than think of sanctification of status in the case of the unbelieving parents it is possible and, it seems, preferable to understand that their holiness, which Paul describes as possessed in the believing spouse, is a sanctification of these unbelievers in the functioning of the marriage relationship and particularly in that role which fulfills the central and distinctive purpose of marriage. In effect, the force of the language is then that the marriage relationship itself was sanctified by virtue of the presence of the believer unto the service of the holy covenant of God and specifically unto the securing of a holy seed.

(2) The principle of vassal authority in the New Covenant

The next step in Kline’s argument for infant baptism is to show that the principle of vassal authority is also present in the administration of baptism under the New Covenant. Specifically, the principle of vassal authority is expressed in the form of parental authority. Kline rejects the theory of presumptive regeneration as the basis for infant baptism (BOC, p. 90), and argues for infant baptism by appealing to the principle of parental authority, a principle of covenant administraion which he sees as continuing from the Abrahamic Covenant into its fulfilled form, the New Covenant.

BOC, p. 91:

For us the pertinent question is whether the covenant for which baptism serves as oath sign of incorporation is, like the divine covenants of the Old Testament and the parallel vassal covenants of the ancient world, a relationship of authority spheres rather than simply of individuals … The pattern of authority is not peripheral but central in the vassal covenant form, and therefore the whole weight of the historical case for identifying the New Covenant as a continuation of the earlier Suzerain vassal covenants presses for the conclusion that this New Covenant is administered to confessors not just as individuals but as heads of authority units.

Kline appeals to four pieces of NT evidence in support of his thesis that the sign of entrance into the New Covenant (baptism) is administered on the principle of vassal or parental authority. I’ll explain that evidence over the next four posts.

Chapter six is the final chapter of By Oath Consigned. In it, Kline ties everything together and makes his case for infant baptism. Aside from a few phrases related to issues that I have already covered (e.g., identifying the New Covenant as a law covenant), there is little in BOC ch. 6 that Kline would have wanted to revise. For this reason I will only be examining the continuities and will not have any posts on discontinuities.

(1) The principle of vassal authority in the Abrahamic Covenant

Kline’s argument for infant baptism begins by appealing to the ANE suzerainty treaties. He quotes a number of these treaties and shows that they were covenants that were made not merely between individuals but between a great king (or suzerain) and a servant king (or vassal), together with his subjects and descendants. For example, one treaty states that Esarhaddon “has made you take an oath that you will relate [the treaty provisions] to your sons and to your grandsons, to your seed, to your seed’s seed which shall be (born) in the future” (BOC, p. 85).

On the basis of such statements in the ANE treaties Kline makes the following observation (BOC, p. 85):

It is of course obvious from the whole purpose of these treaties that the vassal king in taking the ratificatory oath did so in his capacity as king and thus brought his kingdom with him into the relationship of allegiance to the suzerain. Moreover, from express statements in the treaties we know that the vassal king assumed responsibility for his sons and more remote descendants, consigning them with himself in his covenant oath. Consequently, these descendants are mentioned in the curses as objects of divine vengeance if the covenant sworn by the vassal king should be broken.

Kline then shows the principle of vassal authority was also at work in the administration of circumcision under the Abrahamic Covenant (BOC, p. 89):

We conclude, then, that the principle of vassal authority was integral to the administration of circumcision as sign of entrance into God’s redemptive covenant. Confession of Yahweh’s lordship as a matter of personal faith constituted the necessary nucleus and historical beginning for the administration of the rite, and thus for the formal establishment of the covenant community for which circumcision was (paradoxically) the sign of inclusion. There had to be an Abraham. But Abraham could not enter into this oath and covenant simply as an individual. It was Abraham the parent householder, Abraham the patriarch, to whom God gave the covenant of circumcision. In keeping with the nature of the covenant as that may be discerned in the light of the most relevant biblical and extra biblical data, covenantal incorporation into the kingdom of God did not proceed exclusively in terms of individual confession. The formation of the ancient covenant community was rather a process of incorporating households which were under the authority of a confessing servant of the Lord.

I have said that Kline in his mature thought did not believe that the New Covenant threatened any curses. In response someone may object, “But isn’t excommunication a covenant curse?”  The mature Kline would say, “No, because the New Covenant is not a covenant of works in which blessings and curses are conditioned upon obedience or disobedience.” The church’s act of putting someone outside of the visible church is not itself a punitive act, or a covenant curse, or the exercise of the wrath of God. It is merely a fallible judgment that a particular community of believers does not regard this person as a fellow believer. If the church’s judgment is correct, the person in question will indeed face covenant curse and divine wrath at the day of judgment — but not from the New Covenant per se.

I think it is helpful to remember that the New Covenant in Kline’s thinking is tied very closely to the pactum salutis (the Father’s covenant of works with the Second Adam). The New Covenant is an earthly, visible administration of the pactum salutis. Now the pactum salutis clearly threatens no curses. The curse has been borne by Jesus, 100%, with no remainder. All who are in view in the pactum salutis (i.e., the elect) are guaranteed to be eternally saved and glorified. The New Covenant, of course, differs from the pactum salutis in that, prior to the consummation, it embraces both elect and non-elect. This is so because it is an earthly, visible community and we cannot read people’s hearts. The membership roll of the New Covenant, in this present age, does not correspond perfectly to the roster of the pactum salutis. The church sometimes makes mistakes in who it lets in and who it puts out. The members themselves sometimes fake it. Only God knows their hearts.

But in spite of these administrative problems, the New Covenant is the principal means that God uses to gather and perfect the elect who are envisioned in the pactum salutis. The proper purpose of the New Covenant is the salvation of the elect through the preaching of the gospel, the administration of the sacraments, and church discipline. So, given the close connection between the pactum salutis and the New Covenant, it doesn’t make sense to view the latter as a conditional covenant with blessings offered for obedience and curses threatened for disobedience. The New Covenant is merely the earthly, temporal, visible outworking of the pactum salutis, and since Christ has fulfilled the pactum salutis by his perfect obedience, no curse remains to be administered by the New Covenant. The meritorious obedience of Christ as the Second Adam under the terms of the Father’s covenant of works with him (= the pactum salutis) is the ground of the blessings that are administered, applied, and sealed to the elect members of the New Covenant. True, there may be non-elect members in the New Covenant prior to the eschatological consummation when “they shall all know the Lord” (Jer 31:34), but they are temporary branches that do not properly belong and they will be excised. The New Covenant itself is the administration of the eternal soteric blessings of justification, sanctification, and glorification. It is an administration of blessing, grace, and life, not an administration of curse, judgment, and death. 

The non-elect who are temporarily members of the New Covenant fail to believe in Christ and thus fail to receive the blessings purchased by Christ under the pactum salutis. But just because they fail to receive the blessings of the New Covenant, that does not mean they are cursed by the New Covenant. When they are finally removed from the New Covenant by excommunication, they are taken out to be judged by God. But then the curse comes from God himself according to the terms of the broken Adamic covenant of works, not from the New Covenant per se.

So excommunication from the church of the New Covenant is not a covenant curse. It is merely an administrative act of being removed from the New Covenant by the officers of the visible church. Barring repentance and restoration, such apostates will indeed suffer an eschatological curse, but the curse comes from a separate covenant, the Adamic covenant of works. The New Covenant has Christ as its mediator and surety (Heb 7:22; 8:6); therefore, properly speaking, it threatens no curses, but offers nothing but blessings. Even an excommunicated person may repent and return to the covenant fold, lay hold of Christ and his righteousness, and receive the blessings. In a covenant of works, by contrast, restoration is impossible once the covenant has been violated. In the New Covenant, the message is grace, grace, infinite grace. It is ever and always a message of blessing offered freely to all who will believe — even to the poor, the wretched, the repeat offender, and yes, even to the apostate: 

“The one who comes to me I will certainly not cast out” (John 6:37).

“We beg you on behalf of Christ, be reconciled to God!” (2 Cor 5:20).

“Let the one who is thirsty come; let the one who wishes take the water of life without cost” (Rev 22:17).

Indeed, so far from being a covenant curse, excommunication may even be the means that God uses to reclaim the offender and bring them back to the fold.

(2) Description of the New Covenant

If the Abrahamic Covenant is a promissory anticipation of the New Covenant, and if Kline revised his description of the Abrahamic Covenant, then it should come as no surprise that he also revised his description of the New Covenant.

Here are some relevant quotes from BOC where he comments on the New Covenant. In BOC, he treats the New Covenant, like the Abrahamic, as a law covenant with dual sanctions (blessing and curses) conditioned on faithfulness.

BOC, p. 75:

For all its difference, the New Covenant of Jeremiah 31 is still patterned after the Sinaitic Covenant. In fact, Jeremiah’s concept of the New Covenant was a development of that already presented by Moses in the sanctions section of the Deuteronomic renewal of the Sinaitic Covenant (Deut. 30:1 10). According to Jeremiah, the New Covenant is a writing of the law on the heart rather than on tables of stone (v. 33; cf. II Cor. 3:3), but it is another writing of the law. It is a new law covenant. Hence, for Jeremiah, the New Covenant, though it could be sharply contrasted with the Old (v. 32), was nevertheless a renewal of the Mosaic Covenant.

BOC, pp. 77-78

Both blessing and curse are included in the administration of the true New Covenant … We are bound to conclude, therefore, that the newness of the New Covenant cannot involve the elimination of the curse sanction as a component of the covenant … So also the semi-eschatological phase of the New Covenant moves on towards a messianic ordeal which will bring for the justified meek the inheritance of the earth, but judicial exposure and the curse-sentence of excision for the apostates.

Unsurprisingly, Kline’s revised description of the New Covenant in KP corresponds to his revised description of the Abrahamic Covenant. In other words, in KP he no longer defines the New Covenant as a renewal of the Old/Mosaic Covenant (i.e., as a law covenant) and instead stresses the contrast between the Old and the New Covenants. The Mosaic Covenant was a covenant of works and was breakable. The New Covenant is a covenant of grace and is fundamentally unbreakable (although the sense in which it is unbreakable must be carefully defined).

KP, p. 322:

Paul was resuming Jeremiah’s classic analysis of the covenants when he contrasted the new covenant to the old (the old viewed in the restricted but distinctive terms of its typological dimension).  In contrast to the new covenant which could not be broken, founded as it was on God’s sovereign, forgiving grace in Christ, the old covenant, according to Jeremiah, was breakable (Jer 31:32).  Individual members of the new covenant community might prove false and be broken off as branches from a tree while the covenant tree remained intact, pruned and flourishing.  But the old covenant’s typological kingdom order as such could be and was terminated.  The axe of God’s judgment was ultimately laid unto the roots of the tree and the tree itself was felled. Jeremiah’s identification of the old covenant as breakable was the equivalent of an assertion that it lacked the guarantee afforded by the grace principle and was instead based on the principle of works.

Note that Kline defines “breakability” as “the tree itself was felled,” as opposed to individual branches being broken off while the covenant tree remains intact. The New Covenant is unbreakable only in the sense that the covenant itself cannot be terminated, not in the sense that apostasy of individuals within the covenant is impossible.

Kline further wants to affirm that the discontinuity between the Old/Mosaic Covenant and the New Covenant is not total, since there is an underlying continuity between the Abrahamic substratum that underlay the Mosaic Covenant and the New Covenant.

KP, p. 345:

The new covenant is not a renewal of an older covenant in the sense of confirming the continuing validity of the old.  If we speak of the new covenant as a renewal of the old it must be to express their continuity as two administrations of the Covenant of Grace or, more specifically, the continuity of the new covenant with the underlying, foundational stratum of the old covenant, the substratum of gospel-grace as the way to the ultimate heavenly hope in Christ.  But with respect to the old covenant as a typological realization of the promised kingdom realm, the new covenant does not confirm the continuing validity of the old but rather announces its obsolescence and end.

Necessarily so.  For, as the Jeremiah 31:31-34 prophecy indicated, the old covenant in its typological kingdom aspect was not a permanent order of the grace-guarantee kind but a probationary arrangement informed by the works principle, hence breakable.  And having been broken, it was perforce terminated.

So Kline affirms that the New Covenant can be broken at the level of individual members who fail to believe in the gospel and who are thus branches to be cut off. The classic Baptist objection must be raised at this point:  How does Kline deal with Jer 31:34 (“They shall all know me”) which seems to imply that all members of the New Covenant are elect and regenerate? Kline would still stand by what he wrote back in BOC to explain this. It is a matter of the prophet’s eschatological perspective. What we see as two mountain ranges (the already and the not-yet), Jeremiah saw as one (BOC, pp. 76-77):

The aspect of New Covenant consummation that Jeremiah does deal with he views from the limited eschatological perspective of an Old Testament prophet. He beheld the messianic accomplishment in that perfection which historically is reached only in the fully eschatological age to come, as the ultimate goal of a process which in the present semi-eschatological age of this world is still marked by tragic imperfection. But the theologian of today ought not impose on himself the visionary limitations of an Old Testament prophet. By virtue of the fuller revelation he enjoys (cf. Lk. 10:24; I Pet. 1:10, 12) he is able to distinguish these two distinct stages in the history of the New Covenant and to observe plainly that the imperfection of the covenant people and program has continued on from the Old Covenant into the present phase of New Covenant history. It is in accordance with this still only semi-eschatological state of affairs that the administration of the New Covenant is presently characterized by dual sanctions, having in particular, anathemas to pronounce and excommunications to execute.

The only thing that Kline would edit in the above paragraph is the statement that the New Covenant itself has “anathemas to pronounce.” The Kline of KP would agree that the New Covenant has “excommunications to execute” on apostates, but that is not quite the same thing as anathemas or curses. The curse of divine judgment comes from the Adamic covenant of works. Any baptized member of the New Covenant “disclaiming the grace of the covenant and thus breaking it, would undergo in himself the judgment due to Adam’s fallen race” (KP, p. 316).

Summary:

(1) The New Covenant is not a renewal of the Mosaic Covenant but the fulfillment of the Abrahamic Covenant.

(2) It is not a new law covenant but a promise/grace covenant.

(3) As a covenant of works at the typological layer of Israel’s retention of the land, the Old/Mosaic Covenant was breakable.

(4) As a covenant of grace founded on the pactum salutis, the New Covenant is unbreakable in the sense that it can never be terminated.

(5) Although the New Covenant itself does not pronounce the Adamic curse, it is possible for individuals to fall away from the New Covenant through unbelief, to be cut off from the covenant tree, and thus to become subject to the Adamic curse.

I’m still working my way through the middle section of BOC (chs. 3-5), the chapters where Kline deals with the topic of baptism and circumcision as judicial ordeals. I’ve explained the three areas of continuity with KP. Now it’s time to turn to the areas of discontinuity, that is, the things that Kline would have wanted to revise in light of his mature thought.

(1) Description of the Abrahamic Covenant

The first thing in this section of BOC that Kline would have wanted to revise is his description of the Abrahamic Covenant as fundamentally or generically a law covenant, with a promise covenant core that applies only to the elect who experience the curse in Christ. It’s important to recall Kline’s earlier rule of thumb that you know a covenant is a law covenant whenever the vassal takes the oath, binding himself to obedience to the suzerain under threat of a curse. Since the Kline of BOC views circumcision as the vassal’s ratification oath, by this rule of thumb (as applied to Gen 17), the Abrahamic Covenant seems to be a law covenant.

BOC, pp. 40-41:

Since in this covenant [Gen 17] the Suzerain is also the divine Witness, the promissory obligations which Yahweh undertakes as Suzerain are also a blessing sanction which he will honor as the divine Witness when he beholds faithfulness in the covenant servant. Another element of the treaty pattern, viz., the sanctions, is thus included here among the stipulations. Curse sanction appears too, appended to the stipulation regarding circumcision (v. 14) … In short, the transaction recorded in Genesis 17 may be identified as a covenant of the vassal type, an administration of the lordship of the covenant Giver, binding his servant to himself in consecrated service under dual sanctions, blessing and curse.

BOC, p. 48:

The broader import of circumcision is determined by the specific nature of that covenant of which it is declared to be a sign, and especially, since circumcision is a sanction sign, by the peculiar nature of the judgment in which that covenant issues. As for the covenant, it was a law covenant, not a simple guarantee of blessing but an administration of the lordship of God, a covenant therefore which confronted the servant with dual sanctions, curse and blessing.

In KP, Kline shifts his position and argues that the ratification ceremony of the Abrahamic Covenant is recorded in Gen 15 (not Gen 17), when God took the self-maledictory oath on himself. Circumcision as revealed in Gen 17 is merely a “supplementary seal” added to the already-ratified Abrahamic Covenant. Therefore, on this view, the Abrahamic Covenant is fundamentally a covenant of grace, not a law covenant.

KP, p. 5:

It must be noted here that not all oaths of covenantal commitment function as ratification oaths. For example, the role played by the oath ritual of circumcision (Gen 17) is that of a supplementary seal added to the Abrahamic Covenant, which had been ratified by God’s oath on an earlier occasion (Gen 15).

In keeping with his later emphasis on Gen 15 (rather than 17) as the ratification ceremony, Kline argues that the Abrahamic Covenant is fundamentally equivalent to the gospel of grace. It is a pre-Messianic version of the New Covenant.

KP, pp. 294-95, 302:

God’s promise arrangement with Abraham is made synonymous with the gospel of grace … By its identification with the gospel of Jesus Christ the Abrahamic Covenant is seen to be a promissory anticipation of the new covenant … Considerable emphasis falls on the divine sovereignty in the revelation of God’s grace in the promises of the Abrahamic Covenant.

However, a couple of qualifications must be added.

First, this emphasis on the grace-character of the Abrahamic Covenant should not be taken as implying that all members of this covenant are elect. Membership in the Abrahamic Covenant is broader than the circle of election which is contained within it:

KP, pp. 306-7:

Not all who were of covenant status were children of the promise, chosen according to the purpose of grace.  The eternal promise commitment to Abraham coincided not with the bounds of the community established by the Abrahamic Covenant but with the bounds of the Father’s commitment to the Son as second Adam in the prior covenant in heaven, the commitment to give him the elect people for whom he should become surety … Promise, in its true and final meaning, coincides with election and election is a narrower circle within the broader circle of the covenant as historically administered.

Second, although it is a covenant of grace, there are stipulations in the Abrahamic Covenant, like the command to “walk before me and be blameless” (Gen 17:1), as well as the necessity of circumcision (Gen 17:14). Kline agrees that obedience is necessary, but not on the basis of the works principle:

KP, pp. 319-20:

Under the Abrahamic Covenant human obedience was indispensable … Such indispensability of obedience did not, however, amount to the works principle.  For in the Abrahamic Covenant, human obedience, though indispensable, did not function as the meritorious ground of blessing.  That ground of the promised blessings was rather the obedience of Christ, in fulfillment of his eternal covenant with the Father.  And man’s appropriation of salvation’s blessing was by faith …

Now, the obedience indispensable to reception of the ultimate blessings of the Abrahamic Covenant is the inevitable accompaniment of the faith through which the righteousness of God is appropriated.  For it is included as a fruit of the same divine work of spiritual renewal from which springs faith … And because of this inevitable connection of obedience with faith, obedience functions with respect to the acquisition of the promises as a criterion of the validity of confessed faith.  It is a confirmatory witness of the presence of the genuine faith which appropriates the promised gift of grace.  Absence of obedience would betray absence of faith.

Thus, the Abrahamic Covenant is not a generic law covenant that happens to become a promise covenant for those who are elect. Rather, it is fundamentally a promise covenant. But this promise covenant is not equivalent to election, even though the salvation of the elect is its main purpose. The Abrahamic Covenant is not unconditional. It is conditional upon faith. And even obedience is necessary, not as the legal ground of receiving the blessings, but as the necessary fruit and evidence of the genuineness of one’s faith. The necessity of obedience is consistent with its character as a promise covenant, since the blessings are received by faith alone. Nevertheless, because the Abrahamic Covenant is conditional, and not all members of this covenant are elect, apostasy is possible.

Here are the summary points from the Kline of KP regarding the Abrahamic Covenant:

(1) The Abrahamic Covenant is not a law covenant but a promise covenant.

(2) Circumcision is a supplementary seal, not the formal ratification oath of the Abrahamic Covenant.

(3) The formal ratification oath of the Abrahamic Covenant is God’s self-maledictory oath in Gen 15.

(4) The Abrahamic Covenant is a promissory anticipation of the New Covenant.

(5) The Abrahamic Covenant is synonymous with the gospel of grace, the blessings of which are received by faith alone.

(6) The Abrahamic Covenant also demands obedience as the fruit of faith.

(7) The Abrahamic Covenant establishes a visible community marked out by circumcision, thus making its membership larger than the circle of election.