Which Will Governs? Competing Wills, Codicils, Drafts, and Electronic Documents

Revision note: Substantially revised July 21, 2026. The argument has been reordered according to the classical rhetorical movement, compressed where necessary, and given one consolidated Notes and Further Reading section.

A necessary legal caution

This article uses American probate law as an illustration of documentary identity. It is not legal advice. The law of wills is jurisdiction-specific, fact-sensitive, and subject to change. Anyone facing an actual estate dispute should consult a qualified lawyer in the governing jurisdiction.

That caution is also a theological safeguard. We are not going to derive the doctrine of providential preservation from a Michigan statute or a uniform act. Human laws can be wise or foolish, and legislatures can revise them. Scripture alone supplies the church’s doctrine of Scripture.

Probate law is useful here for a more modest reason. It has spent centuries confronting concrete versions of the question that drives this series: When several documents make overlapping or contradictory claims to be a person’s final testamentary act, what makes any one of them operative?

The law’s answer is never, “They are all the will because they all exist.”

A will is an act, not merely a sheet of paper

In ordinary speech, “the will” can mean the physical document found in a desk drawer. Legally, however, the paper matters because it records a particular act: a competent testator’s final, legally expressed intention for disposition at death.

Consider Michigan’s Estates and Protected Individuals Code as one contemporary example. It treats an individual as having sufficient capacity to make a will if the person meets stated requirements concerning age, mental capacity, property, natural objects of bounty, disposition, and the ability to relate these elements in an orderly plan.¹ The ordinary execution statute requires a will to be in writing, signed by the testator or in the testator’s conscious presence and at the testator’s direction, and signed by at least two witnesses within the prescribed time. It also recognizes a holographic will when the material portions and signature are in the testator’s handwriting, subject to statutory conditions.²

These formalities do not imply that ink, witnesses, and signatures magically create intention. They supply public evidence, solemnity, caution, and protection against fraud. A grocery list signed by two people is not a will. A beautifully drafted estate plan never executed may remain only a draft. A forged document does not become the decedent’s act because it resembles one.

Modern statutes also recognize that rigid formalism can sometimes defeat a genuine testamentary act. Michigan’s “harmless error” provision permits a noncompliant document or writing to be treated as compliant if clear and convincing evidence shows that the decedent intended it to constitute the decedent’s will, a partial or complete revocation, an addition or alteration, or a revival of a formerly revoked will or portion.³

That rule relaxes formal execution in defined circumstances; it does not abolish identity. The proponent still must prove, at a heightened evidentiary level, that this writing was intended to perform a particular testamentary function. The issue shifts from mechanical compliance to demonstrated final intent. It does not become an open license to probate every expression of preference.

Two documents may both have been valid—but not both govern a contradiction

Suppose Martha signs a valid will in 2018 leaving her house to her nephew. In 2024, while competent, she signs a new valid will that expressly revokes all prior wills and leaves the house to her daughter. Both documents were genuine testamentary acts when executed. The first is not a forgery or a false historical artifact. But after Martha’s death, the later governing will controls the house because the earlier disposition was revoked.

So the sentence “both wills cannot both be right” needs refinement.

They can both be authentic documents. They can both have been validly executed at different times. They can both provide accurate historical evidence of the testator’s intentions when made. What they cannot do is impose mutually exclusive dispositions of the same property under the same governing estate plan at death.

That logical point is inescapable. If one operative clause gives the same indivisible house wholly to Anna and another gives it wholly to Benjamin, both cannot be performed in the same respect. A legal system must determine which clause governs, whether one modifies the other, or whether neither is effective.

Michigan law, like that of many jurisdictions, recognizes revocation by executing a subsequent will that revokes the previous instrument expressly or by inconsistency. It also addresses revocation by a physical act performed with the required intent.⁴ The details matter. A later document might revoke the earlier will entirely, alter only part of it, or fail to revoke it at all.

Chronology is important but not sufficient. The newest dated page wins only if it is the testator’s legally effective act. A purported later will might be a working draft, the product of incapacity or undue influence, improperly executed, fabricated after death, or revoked in turn. “Later” is evidence, not self-authentication.

A codicil can make two instruments operate together

Not every difference creates a winner-take-all contest.

A codicil is a testamentary instrument that modifies, explains, adds to, subtracts from, or republishes an existing will without necessarily replacing the whole. Suppose Martha’s will gives her estate equally to three children. A later codicil replaces the named personal representative but leaves the gifts untouched. The will and codicil operate together as one estate plan.

Even where a later will contains inconsistent provisions, the revocation may be only partial. Courts read the instruments according to the governing law and demonstrated intent. The later valid disposition controls to the extent of conflict; compatible provisions can remain effective.

This supplies another needed qualification. Two texts with differences can both contribute to one legally operative whole if the testator authorized them to do so. The contradiction principle is narrower and stronger: incompatible provisions cannot both govern the same matter in the same respect.

That principle should sound familiar. The four Gospels can all be inspired without being verbally identical because they are four authorized compositions. The Old and New Testaments can differ because God authorized successive covenant administrations. A will and codicil can differ because the testator intended them to operate together. None of these examples establishes that two contradictory readings at one textual location are both the authorial wording of that location.

Difference is not the problem. Unauthorized or unresolved contradiction is.

Both claimed wills can be wrong

Now suppose two documents are presented after Martha’s death.

The first is an unsigned draft prepared by a lawyer but never approved or executed. The second is a forged document manufactured by an interested heir. They disagree about the house. The court is not compelled to select one merely because there are two candidates. Both can fail.

Or suppose a valid will once existed but was later revoked. A photocopy of the revoked instrument and a typed transcription containing errors might survive. Both papers could preserve much of the former will, yet neither would be the governing testament. Or two dependent copies could repeat the same mistake inherited from an earlier defective copy.

This possibility matters for textual reasoning. When two manuscript readings differ, it is a false dilemma to assume that one of the two forms before us must be original. Both may be secondary. A scribe may replace a difficult wording with one paraphrase; another may replace it with a different harmonization. Textual criticism rightly considers whether the original reading is preserved in a third witness, an early translation, a quotation, or another part of the tradition.

The proposition “both cannot be original” does not entail “one of these two must be original.” Contradictories cannot both be true in the same respect, but alternatives can both be false.

This also shows why generic preservation is too weak when left undefined. It is not enough to say that a substantially correct testament survives across a mass of documents if the governing wording at disputed points remains indefinitely unidentified. The archive may contain both truth and error. Preservation worthy of a normative covenant document must ultimately concern the identity and availability of the words that belong to it, not merely the survival of many witnesses that contain most of them.

Can a revoked will return?

Revival adds another layer. Suppose Martha executes Will A, later executes Will B revoking A, and then revokes B. Does A automatically revive?

Not necessarily. Jurisdictions use different rules, often focusing on the testator’s intent. Michigan’s statute distinguishes circumstances and directs attention to whether the testator intended the previous will to take effect as executed. Evidence may include the circumstances of the revocation and the testator’s declarations.⁵

The point for our illustration is that destroying a later instrument does not always answer which earlier text governs. The legal system asks what act of revocation or revival the testator actually intended and lawfully performed.

Here the analogy with Scripture becomes especially limited. God does not forget which edition expresses his will, lose capacity, or leave courts to infer a change in his eternal purpose from ambiguous behavior. We must not picture providence as a cosmic probate proceeding after the divine Testator has become unavailable.

Yet the legal category still clarifies something. Documentary authority depends upon a history of authorized acts, not upon the bare survival of artifacts. An earlier page, a later page, and a destroyed page do not interpret themselves. Their relation must be established.

The phone note that counted—and the transcript that did not

Electronic documents have forced courts to distinguish medium from testamentary function.

In In re Estate of Horton, a Michigan court considered an unsigned electronic note left in a decedent’s cell phone. The statutory harmless-error rule allowed the court to recognize the note because clear and convincing evidence supported the conclusion that the decedent intended it to be his will. The relevant question was not whether pixels are less real than ink. It was whether the electronic writing represented a final testamentary act intended to operate at death.⁶

In In re Collier Estate, by contrast, the Michigan Court of Appeals confronted a typed transcript of a recorded phone conversation. The decedent discussed wishes concerning property, but the court held that the transcript did not demonstrate the necessary final testamentary intent. The opinion emphasized that no particular form of words is indispensable, yet the document must evince an intention to dispose of property at death and must be final rather than merely deliberative, casual, or precatory.⁷

The comparison is instructive.

• An unsigned phone note can, in exceptional circumstances, qualify.
• A formal-looking transcript can fail.
• The material medium is not decisive.
• Expressed wishes are not automatically a will.
• Finality and intention must attach to the document offered for probate.

The cases do not say form is irrelevant. They show that form serves the identification and authentication of a testamentary act. Harmless-error doctrine asks whether compelling evidence overcomes a defect in conventional form. It does not turn every electronic trace into a coequal testament.

Electronic wills change the medium, not the need for identity

The Uniform Law Commission approved the Uniform Electronic Wills Act in 2019 to provide states with a legal framework for wills maintained as electronic records. The act permits electronic execution while retaining safeguards aimed at authentication, witnessing or notarization, and reliable custody. It also offers states choices concerning remote witnessing.⁸

The innovation is substantial, but its logic is conservative. An electronic will must still be identifiable as the testator’s operative instrument. The legal system still cares who signed, who witnessed, whether the act was contemporaneous, whether the record was altered, and which version is final. Tamper-evident storage matters precisely because a digital file can be copied perfectly or altered invisibly.

Digital replication illustrates both the strength and weakness of talking about “copies.” Ten bit-for-bit identical files do not create ten different wills. They are ten instances of the same informational content. Conversely, two files with the same filename can contain contradictory clauses. Their labels and substantial agreement do not erase the difference.

The identity of a text is not identical to one physical substrate. A biblical book can be written on parchment, printed on paper, stored as Unicode, read aloud, or faithfully translated. But freedom from one substrate is not freedom from wording. The proposition “the text is not the paper” does not entail “any wording on any paper is the text.”

Copies, witnesses, and the governing text

At this point we must state the analogy’s most important technical limitation.

Biblical manuscripts are ordinarily not competing wills in the full legal sense. They are copies and witnesses to prophetic and apostolic writings. A medieval copyist was not attempting to execute a new Gospel of John as his own testamentary act. He was attempting to reproduce an antecedent text.

The more exact probate comparison is therefore this: the original apostolic document is like the authoritative instrument; manuscript copies, quotations, translations, and editions are like evidentiary witnesses to its wording. Some are closer, some more remote. Some preserve an error; some correct one; many agree. Their collective testimony is indispensable, but it must not be confused with the object about which they testify.

That distinction allows us to say all of the following without contradiction:

• A manuscript containing an error can still be a valuable witness.
• Two differing manuscripts may both preserve nearly the whole text faithfully.
• At a genuine contradiction, both readings cannot be the one original wording.
• The two readings under immediate comparison may both be secondary.
• The authentic reading can be preserved elsewhere in the tradition.
• The number, age, and genealogy of witnesses matter, but none is inspired merely by surviving.

This is ordinary reasoning about evidence. Confessional bibliology adds a theological premise: the covenant Lord has not allowed his authentic Word to disappear from the church’s possession. His singular providence governs the history in which the evidence survives, is received, and is used.

What probate law gives us—and what only Scripture can give

Probate law gives us categories:

• Capacity: Was this truly the act of the purported maker?
• Execution: Was intention given an authoritative documentary form?
• Finality: Was this a completed disposition or merely a draft?
• Authentication: Is the document what its proponent claims?
• Revocation and succession: Did a later authorized act replace or modify an earlier one?
• Integration: Do several instruments lawfully operate together?
• Contradiction: Which provision governs when terms cannot both be performed?

Those categories expose conceptual imprecision. A pile of documents is not self-authenticating. Substantial similarity is not identity. Later is not necessarily authoritative. Older is not necessarily operative. Two competing claims can both fail. Multiple authorized instruments can operate together, while contradictory provisions cannot both govern the same matter.

Only Scripture, however, gives the premise on which the preservation argument depends: God inspired covenant documents and, by his special providence, keeps his Word for the faith and obedience of his people. A human testator can fail to communicate, a clerk can lose the file, and a court can reach the wrong result. The living God cannot lie, forget, die into permanent absence, or be defeated by the ordinary means he sovereignly governs.

This first half of the series has therefore established the object before addressing the controversy. God binds his people by written words. Covenant and testament overlap without becoming simplistic synonyms. Christ’s death secures the new-covenant inheritance. Authorized canonical succession differs from corruption. And the law of wills confirms the rational necessity of identifying the governing document.

Part Two will turn directly to preservation. We will ask whether Scripture teaches merely that the original wording survives somewhere among all extant variants, or that God has kept his words publicly available to his church. We will examine the claim that all the readings are “there,” the role of ecclesiastical reception, the difference between providential evidence and autonomous reconstruction, and the specific models advanced by contemporary critical-text advocates.

The controlling question will remain simple: What kind of preservation is adequate for words that function as the church’s final covenantal authority?

Notes and Further Reading