The Sign and The Covenant
Part 3: The Real Baptism Debate
Two installments have been given to the covenants. A third is now given to the sign.
This is not a detour—it is the destination the series has been building toward. The covenants were never the point in themselves. They were always moving somewhere. And where they were moving was here: to the question of who belongs to the covenant community Christ has purchased, and what sign marks their membership in it.
How God constitutes His covenant people determines how His covenant signs are administered. Get the covenants right, and the sign follows. Get the covenants wrong, and the sign will be given to the wrong people, for the wrong reasons, with the wrong theology beneath it.
The debate over circumcision and baptism is not, at its root, a debate about infants. It is a debate about the architecture of redemptive history—about what kind of covenant God is now administering, what kind of people He is now forming, and whether the genealogical principle of a former covenant still governs the new covenant community.
The covenants have been examined. Each has been allowed to define its own ranking, its own rules, its own rewards, and its own ratification. What has emerged is not a single covenant in varied dress, but a series of distinct divine arrangements—each serving its appointed purpose, each pointing forward to the covenant that would finally and fully secure what the others could only promise and picture.
That covenant has now been cut. Its sign has been given.
If you belong to Adam, you bear his covenant curse.
If you belong to Abraham, you receive Abraham’s sign.
If you belong to Israel, you receive Israel’s sign.
If you belong to Christ, you receive His—and with Him, life.
Only one question remains:
Who belongs to Christ?
Circumcision, Baptism, and Covenant: The Heart of the Disagreement
Every major theological debate eventually circles back to the same question: How does God relate to His people across time?
The answer is covenant.
This is the question that hovered over Eden’s groves, where the first covenant shattered like brittle clay. It echoed in Abraham’s starry promise, thundered amid Sinai’s smoke, and found its quiet fulfillment in a blood-sealed upper room. It pulses through the prophets’ visions of a law etched not on stone but on beating hearts.
If covenant is the framework of Scripture’s story, then the signs attached to those covenants are never arbitrary. They reveal what kind of people God is forming, what kind of kingdom He is building, and how He brings His redemptive purposes to completion.
Few debates expose this more clearly than the disagreement over circumcision and baptism. On the surface, the dispute appears to concern the proper subjects of a ritual. In reality, it concerns the architecture of redemptive history—how God structures His covenants, fulfills His promises, and brings the Old Testament types to their appointed telos in Christ.1
What follows is a covenantal, redemptive-historical case demonstrating that circumcision cannot function as a model for Christian baptism. If, as established in parts one and two, the New Covenant alone is the Covenant of Grace—composed exclusively of regenerate members—then its sign cannot be administered on the genealogical principle of a different covenant.
We begin not with water or blade, but with the eternal counsel where Father, Son, and Spirit decreed the salvation of the elect.
We begin not with signs, but with the reality to which they point.
We begin where God begins: with Himself.
A Covenantal Overview: A Brief but Necessary Foundation
Before we ever get to the disagreement about who should be baptized, we have to settle a far more foundational question: What exactly is the New Covenant, and how does it relate to everything that came before it?
Scripture unfolds redemptive history through a series of divinely imposed, oath-bound arrangements—covenants. Each one has its own ranking, rules, rewards/retribution, and ratification. None of them are arbitrary, and none of them can be flattened into the others without doing violence to the text.
Yet all paths trace back to eternity past, where the members of the Trinity covenanted among themselves—the Covenant of Redemption. In this eternal agreement, the Father elects and gives a people to the Son; the Son redeems them, satisfying divine wrath; the Spirit applies that work, sealing and sanctifying the elect. This intra-Trinitarian pact, rooted in divine unity, sets the stage for all that follows, ensuring that every historical covenant serves this redemptive purpose (Ephesians 1:3–14; Psalm 110:1–4; John 10:27–30).
1. Creation & Adam
God orders the cosmos as His cosmic temple and places Adam in a garden-temple as vice-regent. Adam is given a covenant of works: perfect, personal, perpetual obedience would bring him and everything he represented into eternal sabbath-rest enthronement. He fails. The world falls with its federal head. Yet in the very curse God promises a coming Seed who will crush the serpent (Gen 3:15) —not the Covenant of Grace ratified, but the promise of the One who would cut that covenant in history with His own blood— the protoevangelium, the first glimmer of the Covenant of Grace.
2. Noah after the flood
God does not cut (kārat) a brand-new saving covenant. Because it is universal and preservative rather than narrowly redemptive, the Noahic covenant is never said to be ‘cut’ (kārat) anew but ‘upheld’ (hēqîm) toward Noah (Gen 6:18; 9:9). He affirms (hēqîm) with Noah the original Adamic arrangement, now graciously adjusted to a cursed world. While hēqîm can be used in various covenantal contexts, the point here is that the Noahic arrangement is creational and preservative in scope, not redemptive in substance.
The Noahic covenant is with Noah, his sons, every living creature, and the earth itself. Its purpose is preservation, not redemption. God promises seasonal and creational stability (never again a global flood, seedtime and harvest continue) so that the stage remains set for the promised Seed. This is common grace in service of redemptive grace—the lapdog, not the master.
An unavoidable reality
Before turning directly to the question of circumcision and baptism, one more step is necessary. The issue before us is not merely what the covenants say, but how they are being read.
At this point, the question of method can no longer be avoided. Both sides in this discussion bear a burden of proof. The issue is not whether a system can be constructed that accounts for the data, but whether that system arises from the data itself2. In Parts 1 and 2, I have attempted to do precisely that—identifying the pattern provided by scripture and examining each covenant in turn according to its own terms: its ranking, its rules, its rewards and retribution, and its ratification. The argument has not been that the covenants cannot be related, but that they must first be allowed to speak in their own categories before any synthesis is imposed upon them3.
The Presbyterian approach, however, begins at a different point. It states that the covenants share one underlying substance while differing only in administration. But this is not a category supplied by the text. Presbyterian theologians are not thereby abandoning Scripture as their sole infallible rule of faith and life. The question is not whether they appeal to scripture, but whether the categories by which they organize the covenants arise from Scripture itself or are brought to it from without. Historically, this framework has been expressed in terms that are indebted to Aristotelian distinctions between substance and accidents45. That does not make it false; the use of categories is unavoidable. But it does mean it must be justified. The burden lies with those who employ such categories to demonstrate that Scripture itself warrants them, rather than assuming them as a starting point and reading them back into the covenants. The question then is whether these categories are derived from the covenant texts themselves or imposed upon them?
Until that justification is done, the Baptist is not rejecting a biblical conclusion, but declining to grant an unproven premise. The covenants must first be read as they are given. Only then can we ask how they relate.
The question is not whether the covenants can be made to fit a system. The question is whether the system has first been drawn from the covenants themselves.
3. Abraham
Here God does cut (kārat) a new covenant (Gen 15). It is one covenant with two intertwined (but not identical) modes of inclusion: by flesh (preserving the genealogical line) and by promise (anticipating spiritual blessing through the coming Seed).
• a physical seed, a land, a nation, kings—all typological and earthly;
• a promised singular Seed (Christ, Gal 3:16) through whom all nations will be blessed.
Circumcision is given as the sign of this covenant. It marks entrance into the fleshly, genealogically defined community from which Messiah will come—not the spiritual community that belongs to Messiah now that He has come. Its typological significance is real (see Section 4 below), but its administrative function was genealogical. The sign itself makes the point. Ishmael was circumcised (Genesis 17:23–27). Ishmael was explicitly excluded from the covenant promise — 'through Isaac shall your offspring be named' (Genesis 21:12; Romans 9:7). The sign did not make him a member of the promissory strand. It marked his inclusion in the genealogical community, the physical line through which the Messiah would come. Paul draws this conclusion explicitly in Romans 9:6–8: 'not all who are descended from Israel are Israel.' The sign and the spiritual reality were separable from the beginning — it was a feature, not a bug—which is precisely what the New Covenant eliminates.
Yes, there is one people of God united in Christ across time. But the covenantal form of that people pre-Christ is typologically entangled with a national, genealogical structure that is not itself the Covenant of Grace in substance. Presbyterian theology maintains that the Abrahamic Covenant is the Covenant of Grace, differently administered across redemptive history, therefore grounding the continuity between circumcision and baptism as visible signs of the same Covenant of Grace under different administrations. This is of course the central point of the first two installments of this series.6 The New Covenant differs from every prior covenant not merely in administration, but in substance—its ranking, rules, rewards, and ratification are all distinct. Most decisively, it is the only covenant in Scripture that secures the full forgiveness of sins for all of its members (Jer. 31:34; Heb. 8:12). Therefore, it alone is the Covenant of Grace. Unlike the sacrifices of the Aaronic priests, when the Great High Priest intercedes for his covenant people with his own blood, that intercession actually forgives their sins (Heb 10:11-14). Where forgiveness is not secured, covenant membership in Christ’s blood is not present.
4. Moses
Another new covenant is cut (kārat) at Sinai—national, conditional in its retention of Canaan, typological, and intentionally mixed. Obedience brings temporal blessing in the land; disobedience brings exile. It is a covenant of works at the national level, superimposed on the prior gracious promise, designed to expose sin and function as a guardian until Christ (Gal 3:19–24).
5. Phinehas
One might object: “But the Phinehas covenant is merely a subset of the Mosaic covenant—an internal administration within the Sinai framework. It does not directly address the Abrahamic covenant or its sign. Therefore, its termination proves nothing about whether circumcision continues as baptism.”
At first glance, this objection seems plausible. But it ultimately exposes a deeper issue—one that lies beneath the entire paedobaptist argument.
The question is not merely how one covenant relates to another. The question is: What governs our interpretation of covenant continuity?
Is it the exegesis of each covenant in its own context, or a prior commitment to a theological system that determines, in advance, how those covenants must relate?
The Phinehas covenant functions as a test case.
If the Mosaic covenant is, as Westminster theology asserts, an administration of the Covenant of Grace, then what occurs within that administration cannot be dismissed as irrelevant. It becomes a controlling example of how covenantal language—particularly terms like “everlasting”—actually functions within redemptive history.
And here the dilemma emerges:
Either the Mosaic covenant is an administration of the Covenant of Grace, or it is not.
If it is not, then the Westminster framework collapses at a foundational point, since it explicitly teaches that the covenant after the fall “was differently administered in the time of the law” (WCF 7.5).
But if it is, then the Phinehas covenant—embedded within that administration—must be allowed to inform how covenantal language, covenant continuity, and covenant signs are interpreted across redemptive history.
You cannot appeal to the Mosaic covenant as an administration of grace when it supports your system, and then dismiss elements within it when they threaten it.
The Phinehas covenant forces the issue. It does not introduce a new principle—it exposes whether a consistent one is actually being applied
The Substance of the Priesthood
The priesthood given to Phinehas was not merely a different administration of the same realities fulfilled in Christ. Its substance was fundamentally distinct.
Phinehas and his sons:
• mediated ceremonial cleanness,
• offered sacrifices for ritual defilement,
• guarded the tabernacle from profanation,
• preserved Israel’s standing before God within the land.
This was their covenantal function: maintaining ceremonial holiness sufficient for dwelling in God’s earthly sanctuary (Lev. 10:3; Num. 18:1–7).
Christ mediates a different order entirely:
• not cleansing of the flesh, but of the conscience (Heb. 9:14),
• not repeated sacrifices, but a once-for-all offering (Heb. 10:10),
• not access to an earthly tent, but entrance into the heavenly sanctuary (Heb. 10:19–22).
The Aaronic priesthood, including the line of Phinehas, was typological. It pointed forward to Christ. But its primary function was not the administration of New Covenant realities in earthly form. It maintained Israel’s ceremonial standing under the Mosaic covenant.
And here a basic principle must be stated:
A type, by necessity, cannot be the antitype it signifies7.
A type is not the reality itself. It is a God-ordained correspondence that anticipates, prefigures, and points beyond itself to something greater. To treat the type as though it were the same in substance as the antitype—merely administered differently—is to collapse the very distinction that makes typology intelligible.
And if that principle holds everywhere else in redemptive history, it must hold here as well.
And yet—this typological priesthood, with its distinct and earthly substance, was called “everlasting.”
What “Everlasting” Does—and Does Not—Mean
This is the critical point.
The language of “everlasting” covenant, as applied to Phinehas, does not guarantee:
• identical substance continuing unchanged,
• administrative forms persisting in the same mode,
• genealogical succession as a permanent principle,
• or covenantal structures outlasting their appointed purpose.
The priesthood endured—not by continuation in its original form, but by fulfillment in its antitype.
Therefore:
If “everlasting” covenant language can apply to a typological priesthood that truly gives way8 to its antitype in Christ9, then such language cannot, by itself, establish the perpetuity of genealogical covenant membership under Abraham.
The Pressure on the Presbyterian System
This creates a direct challenge.
The Presbyterian argument depends on reading Genesis 17’s “everlasting covenant” as requiring the continuation of its administrative (accidental) form—namely, genealogical inclusion under the covenant sign.
But Phinehas demonstrates that “everlasting” covenant language can function within a typological structure that is:
• real in its time,
• necessary for its purpose,
• and yet temporary in its form.
The dilemma sharpens:
Why is typology allowed to terminate the priesthood, but not circumcision?
Why is it accepted that a perpetual priesthood, which is explicitly typological, gives way to its antitype in Christ,
yet denied that a covenant structured around genealogical descent could do the same?
A Consistent Hermeneutic
The Reformed Baptist position is not introducing a new principle. It is applying one consistently across all of covenant history:
• Aaronic priesthood → its antitype in Christ (Heb. 7–10)
• Physical circumcision → heart circumcision (Rom. 2:28–29; Col. 2:11)
• Genealogical Israel → spiritual Israel (Rom. 9:6–8)
• Mosaic covenant → New Covenant (Jer. 31:31–34)
In each case:
• the type is real,
• divinely instituted,
• and necessary,
• yet gives way when the antitype arrives.
Phinehas is not an exception. He is a pattern.
What this means.
Phinehas had a real priesthood with real substance serving real purposes. But it was typological. It was not the final reality. It endured only as long as its role as a type required.
So too with circumcision.
It marked the genealogically governed line through which the Messiah would come. It distinguished the covenant people. It carried real meaning and real purpose.
But it was a type.
And a type, by necessity, cannot remain once the antitype has come.
God did not fail to keep His promise to Phinehas. He fulfilled it—in Christ.
God has not failed to keep His covenant with Abraham. He has fulfilled it—in Christ.
The question, then, is unavoidable:
If we recognize this pattern everywhere else in covenant history,
on what exegetical grounds is it applied everywhere else, but not here?
6. David
Again God cuts (kārat) a new covenant, promising an everlasting throne to David’s greater Son. It too is typological and finds its substance in Christ, requiring an obedient son to claim its eternal rewards (2 Sam 7:12–16; 1 Kgs 6:11–13).
7. The New Covenant
Jeremiah 31:31–34 and Hebrews 8–10 announce something beautifully different: a covenant that God will cut (kārat) with the house of Israel and Judah that is “not like” the one He made when He brought them out of Egypt.
• All members know the Lord (no evangelism of covenant members needed).
• All have the law written on their hearts (regeneration).
• All enjoy true forgiveness (“I will remember their sins no more”).
• There is no possibility of being in this covenant and perishing (Heb 10:14–18).
This is the first covenant in history whose native membership is exclusively regenerate10. Every previous covenant—Adamic, Noahic, Abrahamic, Mosaic, Phinehas, Davidic—contained both elect and non-elect, believer and unbeliever, wheat and tares, by divine design. The New Covenant does not. It is the first purely spiritual, inwardly efficacious, unbreakable covenant.
Here is the decisive point:
The New Covenant is the Covenant of Grace in its fulfilled, unveiled form. Every prior covenant was gracious in its own way, but none of them was the Covenant of Grace in substance. The Covenant of Grace is not Genesis 3:15 in embryonic form, nor Genesis 6, 9, 12, 15, nor 17. The Covenant of Grace is Jeremiah 31, inaugurated by the blood of Christ (Luke 22:20), mediated by our great High Priest at God’s right hand right now (Heb 8:6).
Why does this matter for baptism?
Because the sign must fit the covenant to which it belongs. The Presbyterian argument depends heavily on reading circumcision and baptism as successive signs of the same covenantal principle. Appeal is often made to Colossians 2:11–12 as decisive proof that baptism has taken the place of circumcision in the New Covenant. But this reading assumes precisely what must be proven11. Circumcision fit a covenant that included unbelievers by birth, pointed to physical descent, and served a typological purpose. Baptism belongs to a covenant whose native constituency is regenerate, forgiven, indwelt by the Spirit, and destined never to fall away. You do not administer the sign of a regenerate-membership covenant to those who have given no credible profession of the realities that define that covenant.
Everything hangs on getting the covenants right. If you use Aristotelian categories to flatten them into “one covenant of grace under different administrations” then paedobaptism seems natural. However, when you let each covenant define its own terms, its own membership, and its own trajectory toward Christ, then the credobaptist conclusion follows naturally. The New Covenant is new. Its reality is new. Its sign must be new as well.
This essay is drawn from a larger work currently in development. For clarity and readability in an online format, AI tools were used in an editorial capacity only. All theological content, arguments, and conclusions are the author’s own.
Works Consulted for Part 3: The Real Baptism Debate
Adams, Brandon. Calvin vs. 1689 Federalism on Old vs. New (2016), available at: https://contrast2.wordpress.com/wp-content/uploads/2016/05/calvin-vs-1689f-on-old-vs-new_051616.pdf.
Barcellos, Richard C. “An Exegetical Appraisal of Colossians 2:11–12.” In Recovering a Covenantal Heritage. Palmdale, CA: Reformed Baptist Academic Press, 2014.
Barcellos, Richard C. The Covenant of Works: Its Confessional and Scriptural Basis. Palmdale, CA: Reformed Baptist Academic Press, 2017.
Beale, G. K. A New Testament Biblical Theology. Grand Rapids: Baker, 2011.
Calvin, John. Institutes of the Christian Religion. 4.16–4.17.
Coxe, Nehemiah. “A Discourse of the Covenants.” In Covenant Theology from Adam to Christ. Palmdale, CA: Reformed Baptist Academic Press, 2005.
Denault, Pascal. The Distinctiveness of Baptist Covenant Theology. 2nd ed. Birmingham, AL: Solid Ground Christian Books, 2017.
Fowler, Stanley K. More Than a Symbol: The British Baptist Recovery of Sacramentalism. Cape Coral, FL: Founders Press, 2015.
Fowler, Stanley K. Rethinking Baptism: Some Baptist Reflections. Eugene, OR: Wipf & Stock, 2015.
Gentry, Peter J., and Stephen J. Wellum. Kingdom through Covenant: A Biblical-Theological Understanding of the Covenants. 2nd ed. Wheaton, IL: Crossway, 2018.
Jeremias, Joachim. The Origins of Infant Baptism: A Further Study in Reply to Kurt Aland. Naperville, IL: A.R. Allenson, 1963. https://archive.org/details/originsofinfantb0000jere/page/14/mode/1up
Jewett, Paul K. Infant Baptism and the Covenant of Grace. Grand Rapids, MI: Eerdmans, 1978.
Johnson, Jeffrey D. The Fatal Flaw of the Theology Behind Infant Baptism. Free Grace Press, 2010. https://www.amazon.com/Fatal-Flaw-Theology-Behind-Baptism-ebook/dp/B071JQ6J68
Kline, Meredith G. By Oath Consigned: A Reinterpretation of the Covenant Signs of Circumcision and Baptism. Grand Rapids: Eerdmans, 1968.
Kline, Meredith G. Images of the Spirit. Eugene, OR: Wipf & Stock, 1999
Kline, Meredith G. Kingdom Prologue. Eugene, OR: Wipf & Stock, 2006.
Kline, Meredith G. Treaty of the Great King: The Covenant Structure of Deuteronomy. Grand Rapids: Eerdmans, 1963.
Kline, Meredith G. “Oath and Ordeal Signs—I.” Westminster Theological Journal 27 (1964–65): 115–139. https://meredithkline.com/klines-works/articles-and-essays/oath-and-ordeal-signs-part-1/
Kline, Meredith G. “Oath and Ordeal Signs—II.” Westminster Theological Journal 28 (1965–66): 1–37. https://meredithkline.com/klines-works/articles-and-essays/oath-and-ordeal-signs-part-2/
Louis Berkhof, Principles of Biblical Interpretation (Sacred Hermeneutics) (Grand Rapids, MI: Baker Book House, 1950), [147], https://dn721201.ca.archive.org/0/items/principlesofbibl00berk/principlesofbibl00berk.pdf.
Malone, Fred A. The Baptism of Disciples Alone. Cape Coral, FL: Founders Press, 2003.
Marcel, Pierre Ch. The Biblical Doctrine of Infant Baptism. 1959. https://archive.org/details/biblicaldoctrine0000unse_a4y8/page/n1/mode/1up
Mendenhall, George E. Law and Covenant in Israel and the Ancient Near East. Pittsburgh: Biblical Colloquium, 1955. https://archive.org/details/lawcovenantinisr0000mend/page/3/mode/1up
Micah Renihan and Samuel Renihan, “Reformed Baptist Covenant Theology and Biblical Theology” (presentation to students at Westminster Seminary California, n.d.), https://thelogcollege.wordpress.com/wp-content/uploads/2012/11/rb-cov-theo-renihans.pdf.
Moo, Douglas J. The Letters to the Colossians and to Philemon. 2nd ed. Grand Rapids, MI: Eerdmans, 2024.
Murray, John. The Covenant of Grace. London: Tyndale Press, 1954.
Murray, John. Christian Baptism. Phillipsburg, NJ: P&R Publishing, 1980. https://gereformeerd.org/wp-content/uploads/2023/03/Murray-J-Christian-Baptism.pdf
Owen, John. Communion with God. E4 Group, 2017.
Owen, John. An Exposition of the Epistle to the Hebrews. Vols. 5–7. Edinburgh: Banner of Truth.
Renihan, Samuel D. The Mystery of Christ, His Covenant, and His Kingdom. Cape Coral, FL: Founders Press, 2019.
Robertson, O. Palmer. The Christ of the Covenants. Phillipsburg, NJ: P&R, 1980.
Schreiner, Thomas R., and Shawn D. Wright, eds. Believer’s Baptism: Sign of the New Covenant in Christ. Nashville, TN: B&H Academic, 2006.
Van Dorn, Douglas. Waters of Creation: A Biblical-Theological Study of Baptism. 2012.
Vos, Geerhardus. The Teaching of the Epistle to the Hebrews. Phillipsburg, NJ: P&R, 1956.
Walker, Austin. Hot Water: A Baptismal Controversy from the 1690s and Its Relevance for Today. Broken Wharfe, 2024.
White, James R. “The Newness of the New Covenant (Part I).” Reformed Baptist Theological Review 1, no. 2 (July 2004): 144–168.
White, James R. “The Newness of the New Covenant (Part II).” Reformed Baptist Theological Review 2, no. 1 (January 2005): 83–104.
Witsius, Herman. The Economy of the Covenants Between God and Man. Translated by William Crookshank. Reprint, CrossReach Publications, 2017.
Scripture employs both shadow/substance language (e.g., Heb. 8:5; 10:1; Col. 2:17) and typological categories (e.g., Rom. 5:14; 1 Cor. 10:6, 11; cf. Heb. 9:24). While related, these are not strictly identical. In this essay, type/antitype language is used where precision is required, since a type is not the same in substance as the antitype it signifies, but gives way to it in fulfillment. This accords with Vos’s reading of Hebrews, in which earthly copies yield to heavenly realities, and with the broader argument that the New Covenant is not merely the old order in another administration, but its fulfillment in Christ. See Vos, The Teaching of the Epistle to the Hebrews; Redemptive History and Biblical Interpretation; Kline, Structure of Biblical Authority; and Micah and Samuel Renihan, “Reformed Baptist Covenant Theology and Biblical Theology.
It should be noted that the use of scholastic categories was not unique to Presbyterian theology, nor is it inherently problematic. The entire confessional period—Reformed, Lutheran, and Particular Baptist alike—was deeply shaped by the intellectual tools of Reformed scholasticism, itself drawing on Aristotelian distinctions such as substance and accidents, essence and subsistence, and primary and secondary causation. The framers of the 1689 Second London Baptist Confession operated within this same conceptual world, employing these categories as instruments of clarity and theological precision.
The question, therefore, is not whether such categories may be used pedagogically or descriptively. The question is whether they are first derived from the exegesis of Scripture on its own terms, or assumed in advance and then imposed upon the covenantal texts as a controlling framework for their interpretation. The issue at stake is not the use of categories, but the direction of their use—whether they arise from Scripture or govern its reading.
It is often objected that if one rejects the substance/administration distinction as overly philosophical, consistency would require the rejection of Nicene Trinitarianism on the same grounds. This objection fails to recognize a crucial difference in how and why extra-biblical categories are employed.
In the case of the doctrine of the Trinity, the church was compelled by the plain teaching of Scripture to hold together truths that, if left unarticulated, would result in contradiction or heresy. Scripture affirms unequivocally that there is one God (Yahweh) (Deut. 6:4), while also ascribing the divine name, attributes, and works to the Father, the Son, and the Holy Spirit individually (e.g., Isa. 45:23 with Phil. 2:10–11; Acts 5:3–4; John 1:1). These three are not merely modes or manifestations, for they are distinguished from one another in personal relations (Matt. 3:16–17; John 14:16–17), and yet they are not three gods, but one. Further, each person is presented as fully and equally divine, not in a hierarchy of being, but in unity of essence (John 5:18; Heb. 1:3).
This creates an unavoidable theological pressure: one Yahweh, three who are called Yahweh, personally distinct, yet fully equal and not divided in essence. The categories later formalized at the Council of Nicaea and developed in subsequent Trinitarian theology—essence, person, homoousios—were not imposed upon the text as an external framework, but rather arose as necessary linguistic tools to preserve all that Scripture requires without contradiction. They function as guardrails, ensuring that none of the biblical affirmations are denied or collapsed into error (whether tritheism, modalism, or subordinationism). In this sense, the philosophical terminology is subservient and explanatory, not generative; it gives conceptual clarity to truths already demanded by the text itself.
By contrast, the substance/administration distinction in covenant theology does not arise from the same kind of textual necessity. Scripture does not present us with a formal tension such as “one covenant in substance” alongside “multiple covenants in substance” that must be reconciled. Rather, it presents a series of covenants that must be allowed to define themselves according to their own ranking, rules, rewards/retribution, and ratification. When these are examined on their own terms, the covenants display real differences in structure and effect—some explicitly breakable (Jer. 31:32), and one explicitly unbreakable, internal, and effectual for all its members (Jer. 31:33–34; Heb. 8:6–13).
The pressure point, then, is not a metaphysical contradiction requiring resolution, but a question of whether these covenants can be said to share the same substance when their ranking (who belongs), rules (conditions), rewards/retribution (outcomes), and ratification (how they are established) differ in ways that are not merely accidental but constitutive. The substance/administration framework does not arise to resolve an unavoidable tension in the text, but to maintain a particular form of continuity—namely, that all redemptive covenants are one in substance as the covenant of grace, differing only in outward administration.
Unlike Nicene categories, which are required to prevent denial of explicit biblical claims, this distinction is not textually compelled in the same way, but rather represents one theological synthesis among several. The question must therefore be faced: are these philosophical categories being used to clarify what the text demands, or to secure a conclusion about covenantal continuity that the text itself does not explicitly require?
The issue is not whether theological categories may be used—they must be—but whether they arise from the demands of the text or are imposed to organize it according to a prior system. In Trinitarian theology, the categories of essence and person are necessary to say what Scripture says without contradiction. In covenant theology, the categories of substance and administration are used to argue for a level of continuity that must be demonstrated, not assumed, particularly when the covenants themselves, on their own terms, present differing structures along the lines of ranking, rules, rewards/retribution, and ratification.
Thus, the appeal to Nicene Trinitarianism does not justify the substance/administration distinction by analogy. The former represents the church’s compelled articulation of revealed truth under the pressure of the text; the latter represents a systematic proposal for relating that truth across redemptive history—one that must itself be proven from the text rather than secured by philosophical parallel.
Samuel Rutherford, a commissioner to the Westminster Assembly, explicitly speaks in terms of the “substantials” of the covenant over against priesthood, sacrifices, circumcision, baptism, and other outward forms as “accidents” to that substance. See Samuel Rutherford, The Covenant of Life Opened (Edinburgh, 1655), 111.
https://www.digitalpuritan.net/Digital%20Puritan%20Resources/Rutherford,%20Samuel/The%20Covenant%20of%20Life%20Opened.pdf): This does not by itself prove the Presbyterian position false, but it does show that the substance/accidents framework was consciously operative within seventeenth-century Reformed covenant theology.
For a historical and theological analysis of how the substance/administration distinction developed in response to post-Reformation debates—particularly in relation to infant baptism—and the tensions it introduces in interpreting passages such as Jeremiah 31 and Hebrews 8, see Brandon Adams, Calvin vs. 1689 Federalism on Old vs. New (2016). Adams traces the shift from Augustine’s distinction between the Old and New Covenants as two in substance, to the later Reformed emphasis—especially in Heinrich Bullinger and John Calvin—on a unity of substance with diversity of administration, and highlights John Owen’s return to a two-covenant reading grounded in the exegesis of Hebrews.
Presbyterian covenant theology classically argues that there is one Covenant of Grace running through all of redemptive history, administered differently under the Old and New Covenants. Within this framework, the Abrahamic covenant is understood as a primary administration of the Covenant of Grace, and circumcision as its covenant sign. Accordingly, baptism is viewed as the New Covenant counterpart to circumcision, applied to believers and their children as members of the visible covenant community. See John Murray, Christian Baptism (Phillipsburg, NJ: P&R, 1980); Pierre Marcel, The Biblical Doctrine of Infant Baptism (1959); and Paul K. Jewett, Infant Baptism and the Covenant of Grace (Grand Rapids: Eerdmans, 1978). For modern treatments, see also Michael Horton and others within the Westminster tradition.
Typology in Scripture entails real correspondence between type and antitype, yet not identity of substance or form. A type is preparatory and provisional, pointing beyond itself to a greater fulfillment, and therefore cannot be the antitype it signifies. As Louis Berkhof observes, “it is necessary to have due regard to the essential difference between type and antitype. The one represents truth on a lower, the other, the same truth on a higher stage…from the external to the internal, from the present to the future, from the earthly to the heavenly.”¹ This accords with the broader biblical pattern in which earthly copies and shadows give way to their heavenly realities in Christ (cf. Heb. 8:5; 9:24; 10:1; Col. 2:17).
In saying that the former priesthood gives way to the latter (Heb. 7:12–19), the apostle is not suggesting that God failed to uphold His promise of a perpetual priesthood according to Aaron. Rather, by divine inspiration he demonstrates that the Aaronic priesthood was typological in nature and therefore rightly fulfilled in the antitypical priesthood of Christ according to the order of Melchizedek. The “perpetuity” of the former is thus realized, not through unending continuation in its original form, but through its divinely intended fulfillment in a greater priesthood.
One might object: If a prior “everlasting” institution can give way to another, what assurance do we have that the promises of the New Covenant will not likewise be fulfilled in some future arrangement? Does this not undermine certainty? By no means. The very logic of the Epistle to the Hebrews establishes the opposite. The Levitical system gave way precisely because it could not bring anything to its intended goal (telos): “if perfection (teleiōsis) had been attainable through the Levitical priesthood…” (Heb. 7:11). Its repetition, externality, and inability to cleanse the conscience marked it as inherently provisional (Heb. 9:9; 10:1–4). But Christ, by contrast, “has perfected (teteleiōken) for all time those who are being sanctified” (Heb. 10:14).
This is the decisive distinction. What was formerly administered could not reach the telos; in Christ, that telos has been reached. When Christ declares tetelestai (John 19:30), He announces not merely the cessation of His work, but its completion—its arrival at the divinely appointed end. Indeed, He is Himself the telos (Rom. 10:4), the one in whom all prior institutions find their fulfillment. The former priesthood, therefore, does not continue alongside the latter, nor is it merely adjusted; it is brought to its goal and thus gives way to the reality it anticipated.
The question, then, is unavoidable: can the same covenantal structure govern both pre-telos and post-telos realities? If the former covenant is said to be unable to bring its members to the telos, while the latter actually does so, the question must be faced whether “substance” can meaningfully be predicated of both in the same sense. Can the same covenantal substance both fail to accomplish its end and successfully accomplish it? Hebrews grounds its argument not merely in clarity of administration, but in the incapacity of the former and the efficacy of the latter. The change, therefore, is not incidental but necessary: a covenant that cannot bring its members to the telos must give way to one that does.
For this reason, the language of “everlasting” must be understood in a manner consistent with apostolic usage: not as requiring the unbroken continuation of a typological form, but as securing its fulfillment in the antitype to which it was always ordered. Yet this same logic guarantees the immutability of Christ’s priesthood. Unlike the former, His work is not anticipatory but complete, not repetitive but once-for-all, not provisional but final. Because His priesthood has reached the telos, it does not give way to another. It abides for all ages precisely because nothing remains to be accomplished.
Similar “everlasting” (ʿôlām) language appears elsewhere in the Old Testament without requiring literal, unchanged continuation into the New Covenant age. The Sabbath is called a “perpetual covenant” and “sign forever” between God and Israel (Ex. 31:16–17), yet Hebrews 4:9 declares that a greater Sabbath rest remains for the people of God, fulfilled in Christ (Heb. 4:1–11; cf. Col. 2:16–17). The “everlasting doors/gates” of Psalm 24:7, 9 are poetic and typological, pointing to access to God’s presence ultimately realized through Christ, the true temple (John 2:19–21; Heb. 8:5; 9:24; 10:19–22). Though Hebrews does not name Phinehas, it demonstrates that the entire Levitical priesthood—with all its “perpetual” elements—was typological and gives way to Christ’s permanent priesthood after the order of Melchizedek (Heb. 7:11–28; 8:13). In each case, God’s faithfulness is shown not by preserving the shadow indefinitely, but by fulfilling the type in the antitype. These examples reinforce the consistent hermeneutic applied throughout this series: typological realities (including the Phinehasian priesthood and circumcision) serve real, God-ordained purposes in their redemptive-historical context but reach their appointed telos in Christ and the New Covenant, whose substance is exclusively regenerate membership secured by God’s unbreakable promises (Jer. 31:31–34; Heb. 8:8–12).
The exclusive regeneracy of New Covenant membership is established most directly by Jeremiah 31:34 — "they shall all know me, from the least of them to the greatest" — and its citation in Hebrews 8:11. The universal knowledge of God within the covenant is not an eschatological ideal but a defining feature of the covenant itself, distinguishing it from every prior arrangement in which covenant membership and saving knowledge of God were not coextensive. For detailed treatment see Renihan, The Mystery of Christ, His Covenant, and His Kingdom; Denault, The Distinctiveness of Baptist Covenant Theology; cf. Owen, An Exposition of the Epistle to the Hebrews on Hebrews 8. See also Parts 1 and 2 of this series.
The paedobaptist reading of Colossians 2:11–12 typically argues that Paul presents baptism as the direct covenantal successor to circumcision, administered to the same class of recipients under a new form. Presbyterian covenant theology classically grounds this continuity not in an explicit New Testament command, but in the assumption that the same covenantal principle governing the administration of the sign under Abraham remains operative under the New Covenant. As John Murray argues, “the same principle… is embedded and operative in the administration of the covenant of grace under the new,” and therefore the inclusion of infants continues apart from any express statute authorizing the practice. On this reading, the proximity of circumcision and baptism in Colossians 2 confirms that functional equivalence, and the household structure of the Abrahamic covenant carries forward into New Covenant sign administration. See John Murray, Christian Baptism (Phillipsburg, NJ: P&R, 1980), esp. 59–61. https://archive.org/details/christianbaptism00murr/page/61/mode/1up
This reading, however, requires Paul to be doing something the text does not actually do. Richard C. Barcellos, in “An Exegetical Appraisal of Colossians 2:11–12,” demonstrates that the controlling grammatical relationship in verse 11 is not between physical circumcision and baptism, but between circumcision made with hands and circumcision made without hands — the latter being the inward, Christ-wrought reality to which the former pointed. Baptism enters Paul’s argument in verse 12 as a participial clause dependent on the spiritual reality already established in verse 11, not as a covenantal counterpart introduced on its own terms. The movement is from shadow to substance, and then from substance to its visible sign — not from old sign to new sign. Douglas Moo’s recent commentary reaches a comparable conclusion, noting that Paul’s concern throughout the passage is with the sufficiency of what Christ has accomplished, not with establishing sacramental succession. See Barcellos, “An Exegetical Appraisal of Colossians 2:11–12,” in Recovering a Covenantal Heritage (Palmdale: Reformed Baptist Academic Press, 2014); Moo, The Letters to the Colossians and to Philemon, 2nd ed. (Grand Rapids: Eerdmans, 2024).
Note to the reader:
Much of what is assumed in this section depends upon distinctions already established in Parts 1 and 2 and carried into Part 3. For the sake of readers who may not have worked through those earlier installments, those distinctions are briefly restated here. The New Covenant is defined by its membership, not by the church’s ability to identify that membership with perfect accuracy; the sign is therefore administered on the basis of credible profession, not omniscient knowledge, and the presence of false professors reflects the limits of human judgment, not the structure of the covenant itself. Likewise, that a sign does not confer the reality it signifies does not render it theologically indifferent; signs derive their meaning from the covenant to which they belong, and their recipients are determined by that covenant’s terms, not by any inherent efficacy in the sign itself. Further, typology does not operate at the level of symbolism alone. When a typological order gives way to its fulfillment, the structures that governed that order give way as well, so the transition from shadow to substance is covenantal, not merely illustrative. For this reason, the question is not fundamentally about infants, but about the nature of the covenant itself, since the proper recipients of the sign are determined by the covenant’s own membership principle, whatever conclusions that may yield.
I appreciate your clear writing on this topic, as it is very important. How do you interpret passages like Ezekiel 37:24-28 and Jeremiah 32:38-41, which seem to include children of Christians in the new covenant? Thanks!