Covenant or Testament? Berith, Diathēkē, and the Limits of the Legal Analogy

Revision note: Substantially revised July 20, 2026. The argument has been reordered according to the classical rhetorical movement, compressed where necessary, and given one consolidated Notes and Further Reading section.

One English Bible, two familiar titles

Christians open one Bible and find it divided into an Old Testament and a New Testament. Then they read the words of institution and hear Christ speak of “my blood of the new testament” (Matthew 26:28) or “the new testament in my blood” (Luke 22:20; 1 Corinthians 11:25). In many modern translations, the same passages read “covenant.” Hebrews speaks of Christ as “the mediator of the new testament” and then, in the King James Version, of the death and force of a “testament” and a “testator” (Hebrews 9:15–17). A few verses later it returns to the first testament and the blood sprinkled by Moses.

Is a covenant a contract? Is the New Covenant Christ’s last will? Does testament mean a collection of canonical books, an arrangement of salvation, a legal disposition effective at death, or all three? The answer is not obtained by choosing one English gloss and forcing it into every verse. Scripture uses words in contexts; translation must attend to those contexts; and theology must respect both the unity and the variety of biblical usage. The legal analogy will prove illuminating, especially in Hebrews 9. But if it is made absolute, it will obscure the very covenant theology it was meant to explain.

Berith: the bond God establishes

The principal Hebrew word translated “covenant” is בְּרִית (berith). It appears in relationships between God and human beings, among human parties, and in political and personal arrangements. Scripture speaks of God’s covenant with Noah and every living creature (Genesis 9:8–17), his covenant with Abraham (Genesis 15; 17), the covenant at Sinai (Exodus 19–24), the covenant with David (2 Samuel 7; Psalm 89), and the promised new covenant (Jeremiah 31:31–34). It also uses the word for arrangements between Abraham and Abimelech (Genesis 21:27, 32), Isaac and Abimelech (26:28), and David and Jonathan (1 Samuel 18:3; 23:18).

The word’s etymology is disputed. Proposals connect it with eating, binding, fettering, selecting, or an Akkadian term, but no proposed root should be made to carry a theology the contexts themselves do not establish.¹ Words do not arrive in a passage accompanied by the entire history of their possible roots. Their meaning is governed by actual usage.

Across that usage, a berith is a solemnly established relationship or arrangement with stated commitments. It may include promises, obligations, signs, oaths, witnesses, sacrifice, blessings, and curses. The exact configuration differs. God’s promise to Noah is not identical in administration to Sinai; David and Jonathan do not stand in the same relation as the Lord and Abraham; Jeremiah’s new covenant is not merely Sinai repeated. Yet the word ordinarily marks something firmer and more public than a private intention. A covenant establishes and defines a bond.

“Cutting” a covenant is a frequent Hebrew idiom (for example, Genesis 15:18; Exodus 24:8; Deuteronomy 5:2–3). Sacrificial death and self-maledictory oath can form the solemn background: may the covenant breaker become like the divided animals. Genesis 15 dramatizes that sanction when the smoking furnace and burning lamp pass between the pieces while Abram is in a deep sleep. Jeremiah 34:18–20 later makes the judgment explicit against those who transgressed the covenant after passing between the parts of a calf.

But berith is not simply Hebrew for a modern negotiated contract. A contract normally presupposes parties whose consent produces reciprocal enforceable duties within an existing legal order. God’s covenants are acts of the sovereign Lord who creates, rules, judges, and graciously binds himself. No law above God compels him. No creature bargains with him as an equal. Even where a covenant contains mutual obligations, the relationship is sovereignly constituted.

That is why Reformed theology has often spoken of divine covenants as sovereign administrations. God condescends to reveal his will, confirms his promise, and orders communion with his people. Human faith and obedience are real covenant responses, but they do not place God and man on one legal plane.

Diathēkē: covenant, disposition, testament

The Greek word διαθήκη (diathēkē) has a semantic range that includes a disposition or arrangement and, in legal settings, a will or testament. Greek translators of the Old Testament used diathēkē with remarkable regularity for Hebrew berith. Consequently, when New Testament writers used the Greek word, they inherited a rich scriptural history. Noah, Abraham, Sinai, David, and Jeremiah were already associated with diathēkē in the Greek Bible.²

This fact prevents a common mistake. We cannot begin with the ordinary meaning of “last will and testament” in a contemporary probate code and pour that meaning into every New Testament occurrence. When Paul speaks of “the covenants of promise” (Ephesians 2:12), when Luke records God remembering his holy covenant and the oath sworn to Abraham (Luke 1:72–73), or when Hebrews cites Jeremiah’s promise of a new covenant (Hebrews 8:8–12), the Old Testament covenantal background is controlling.

At the same time, the legal sense “testament” is not imaginary. Diathēkē could denote a disposition of property, including a will. That sense becomes exegetically important in Hebrews 9:16–17, where the writer speaks of death, the one who made the diathēkē, and the instrument’s force in relation to that death. The King James Version renders the cluster “testament” and “testator.” Many modern translations do the same in those two verses, even when they translate diathēkē as “covenant” elsewhere.

The word therefore presents neither a simple equation nor an unbridgeable ambiguity. Covenant and testament overlap in the biblical history of salvation, while their legal features come into sharper focus in particular contexts. Hebrews can exploit that range because Christ’s saving work is both the inauguration of the promised new covenant and the death through which the eternal inheritance is bestowed.

Why not synthēkē?

An often repeated explanation says that the Septuagint translators avoided συνθήκη (synthēkē), a mutual agreement, and chose diathēkē, a unilateral testament, because God sovereignly dictates his covenant. There is a theological truth underneath the observation: God is not one negotiator among equals. But the lexical argument is often pressed too neatly.

Ancient words do not divide themselves according to modern systematic categories. Diathēkē was capable of uses broader than a last will, and synthēkē was not restricted to balanced agreements among equals. The translators’ consistent use of diathēkē created a scriptural convention whose meaning was increasingly shaped by the Hebrew text it represented. One cannot simply open a nonbiblical Greek dictionary, select “will,” and conclude that every biblical covenant is testamentary in precisely the modern probate sense.

The safer procedure is contextual and canonical:

• Ask what relationship the passage describes.
• Attend to the promises, commands, oath, sacrifice, signs, sanctions, and inheritance in that passage.
• Observe how the passage echoes earlier Scripture.
• Then determine whether “covenant,” “testament,” “will,” or another English expression best communicates the emphasis.

Translation is not a contest in which one gloss defeats all others. It is the disciplined rendering of meaning in context.

Three things a biblical covenant is not

Clear definition is helped by removing three misconceptions.

Not a bargain between equals

When God covenants with Abraham, Abraham does not negotiate the divine promise. God calls him, promises a seed and land, passes through the pieces, institutes circumcision, and commands covenant faithfulness. Abraham believes and obeys, but divine grace and sovereignty establish the relation. At Sinai, Israel truly vows, “All that the LORD hath spoken we will do” (Exodus 19:8; 24:3, 7). Yet Israel’s assent does not make God Lord. The preface to the Decalogue begins with accomplished redemption: “I am the LORD thy God, which have brought thee out of the land of Egypt” (Exodus 20:2). Grace grounds obligation.

Not an informal promise

God’s bare word is perfectly trustworthy. He cannot lie. Nevertheless, Scripture records him confirming promises by oath, sign, blood, and written memorial—not because his truthfulness is deficient, but because he condescends to our infirmity and publicly administers the relationship. Hebrews says that God, “willing more abundantly to shew unto the heirs of promise the immutability of his counsel, confirmed it by an oath” (Hebrews 6:17). The oath does not transform falsehood into truth. It gives the heirs strong consolation by placing the immutability of God’s purpose beyond dispute. Covenant form is gracious assurance.

Not merely the physical document

The book of the covenant is genuinely a covenant document, but the covenant cannot be reduced to ink and material. It includes the Lord, his people, the historical acts by which he redeems them, the promises and obligations expressed, and the sanctions administered. A scroll can be copied while the covenant relationship is violated. Josiah can possess the book while previous generations have neglected its terms. That distinction guards us from bibliolatry. Christians do not worship paper, ink, a manuscript, an edition, or a translation. We worship the God whose authoritative speech Scripture inscripturates. Because we worship him, we refuse to treat his words as fungible.

Three things a last will can help us see

A carefully limited comparison with wills and testaments clarifies features of a covenant document.

The maker’s intention governs

A will is not an invitation for beneficiaries to decide what disposition would have been wisest. Courts seek the testator’s legally expressed intent. Evidence, formalities, later instruments, and rules of construction serve that end. Heirs may dispute the words; they do not thereby become authors of them. Likewise, the meaning and authority of Scripture arise from the divine Author. The church receives rather than creates the canon. Copyists witness to rather than constitute the text. Scholars serve rather than rule the Word.

Identity matters at points of difference

Two copies of a will can be substantially identical. A misspelled street name may cause no uncertainty. Yet if one gives the house to Anna and the other gives it to David, the contradiction cannot be dismissed by saying the testament is preserved “in both documents” without deciding which wording belongs to the governing instrument.

This is a conceptual point, not a shortcut through textual criticism. Many textual variants are trivial; some readings can express the same meaning; some differences are easily explained. But where readings truly contradict, both cannot be the same authorial wording. A collection of witnesses is not identical to the text witnessed.

Authorized revision differs from corruption

A competent testator may execute a later will or codicil according to law. A stranger may not alter a clause and call it succession. The maker’s authorized act and an unauthorized change are categorically different even if both result in later pieces of paper bearing different wording.

Scripture likewise distinguishes progressive revelation from corruption. God may add prophetic words, institute the new covenant promised by Jeremiah, or authorize apostolic writings. A scribe may not add to or remove from God’s words. Later in this series we will show why the Old and New Testaments are not rival recensions of one unstable document.

Four places where the analogy fails

Legal analogies become dangerous when their limits are forgotten.

First, a human will is ordinarily revocable during the testator’s lifetime. God’s counsel is immutable. Scripture can describe changes of administration and the promised obsolescence of the Mosaic covenant without implying that God discovers better plans or corrects earlier mistakes. Second, a human will ordinarily becomes operative because the testator ceases to live and act. Christ truly dies, but he rises, ascends, intercedes, and personally administers the blessings purchased by his blood. The New Testament is not an estate managed in the permanent absence of its deceased maker.

Third, a will chiefly disposes of property to beneficiaries. Biblical covenant embraces lordship, worship, reconciliation, law, communion, peoplehood, mission, and new creation. Inheritance is central, but it is not exhaustive. Fourth, human probate operates under a civil authority above the testator. God swears by himself because he can swear by no greater (Hebrews 6:13). His covenant is not validated by a superior court. His own truth, holiness, and sovereign oath are its ultimate security.

These limits do not make the comparison useless. They teach us to use it analogically. A sound analogy identifies a genuine similarity while preserving greater dissimilarity. The testamentary lens in Hebrews 9 shows something true about death and inheritance; it does not domesticate the new covenant into a modern estate plan.

A lexical bridge, not a lexical trick

Some interpreters have argued that Hebrews 9:16–17 temporarily shifts from “covenant” to “will,” while others contend that diathēkē should remain “covenant” and that the death is the death of a covenant-making sacrifice rather than a testator. Still others see a deliberate interplay: the writer uses a term capacious enough to connect covenant ratification, death, and testamentary inheritance. The first is that the debate cannot be decided by asserting that diathēkē “always means covenant” in biblical Greek. Hebrews 9:16–17 contains language that naturally evokes a testamentary disposition, and historic interpreters recognized it long before modern critical editions.

The second is that “testament” must not be detached from the chapter’s covenantal argument. Verse 15 identifies Christ as mediator of the new diathēkē so that the called may receive “the promise of eternal inheritance.” Verse 18 returns to Moses, blood, the book, and the people. Verse 20 cites the covenant formula of Exodus 24: “This is the blood of the testament which God hath enjoined unto you.” Whatever semantic turn occurs in verses 16–17, it operates within the biblical theology of covenant and sacrifice.

The word is a bridge rather than a trick. It allows Hebrews to show that the promised inheritance comes through death and that the new covenant is inaugurated by blood. The author is not playing with a coincidental homonym. He is drawing together realities that reach their unity in Christ.

What the analogy can—and cannot—prove about preservation

The analogy of a will establishes a rational expectation: an authoritative instrument must have identifiable terms, and contradictory clauses cannot simultaneously express one author’s governing disposition in the same respect. It exposes the weakness in speaking as though every surviving reading simply belongs to “the preserved text.”

But the analogy cannot prove providential preservation. Human wills are lost, forged, destroyed, superseded, and misunderstood. A courthouse archive can burn. Witnesses can lie. If Scripture were only like a human instrument, its loss would be possible. The doctrine of preservation rests instead upon God’s own identity, acts, promises, commands, and covenant purposes. The legal comparison clarifies the kind of object involved; Scripture supplies the divine promise concerning that object. We do not infer God’s singular providence from the efficiency of probate courts. We learn from Scripture that the God who inspired his Word also keeps it for his church.

Nor does the analogy, standing alone, identify the Textus Receptus, the Majority Text, the critical text, one printed edition, or one translation as the authentic text. That conclusion requires a broader doctrine of Scripture and a historical account of reception and transmission. A metaphor cannot do the work of exegesis, confession, and history.

Used within those limits, however, the analogy presses a necessary question: if the Bible is the documentary word of the covenant Lord, what would it mean for that testament to be available only as a set of mutually incompatible possibilities? Before answering, we must look directly at the passage in which covenant, testament, blood, death, and inheritance meet: Hebrews 9.

Notes and Further Reading