What It Means to ‘Build’ a Theology from a Text

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.

Nearly every Christian theologian can be said to use the Bible. He may quote it, arrange chapters under biblical subjects, or place a column of proof texts after a doctrinal sentence. None of those acts, considered alone, proves that his theology has been built from the text cited.

To build is a stronger claim than to use. It asserts dependence. A decorative quotation may be removed while the doctrine remains untouched. A load-bearing premise cannot be removed without rebuilding the argument—or admitting that it no longer stands.

That distinction governs this series. The question is not whether modern critical editions can be used by Reformed theologians. They plainly can. The question is whether the exact propositions of the Reformed confession can be reconstructed from their operative readings without quietly borrowing a premise from another text, a translation, or the inherited system.

One safeguard must come first. Before a confessional sentence is divided into logical parts, its historical meaning must be established. Westminster is not ours to modernize and then test. We must hear its words in their seventeenth-century setting, identify the doctrine they affirm and the errors they exclude, and only then ask whether every material part can be derived from the designated text.

The governing definition is therefore this:

To build a doctrine from a text is to display a public and reproducible chain of warrant from identified readings, through grammatically warranted local propositions and valid canonical inferences, to the full doctrinal conclusion—without counting a material premise supplied only by another reading, a translation, or an inherited system as though it arose from the designated text.

Every part of that definition matters.

Theology Is Received, Not Invented

The language of construction may suggest that the theologian manufactures doctrine as an architect invents a house. Reformed theology denies it. God knows himself perfectly. He speaks truly. Theology begins with the reception of that divine speech, not the invention of religious ideas. The Shorter Catechism accordingly calls the Word of God the Christian’s “only rule” and says that Scripture principally teaches what we are to believe concerning God and what duty God requires of us.¹

The theologian receives, interprets, distinguishes, relates, and orders. Construction belongs to the public demonstration and arrangement of our theology; it does not belong to the creation of divine truth.

This rebukes both parties to the present dispute. The advocate of the received text may not declare a doctrine biblical merely because it belongs to the inherited Reformed system. He must show its textual warrant. The advocate of the critical text may not retain the old conclusion, exchange one of its premises, and assume that the new premise performs the same work. He too must show the warrant.

The test is not whether a proposition is old or new. It is whether the revealed premises establish everything the proposition asserts.

Scripture Is the Source; Editions Are Controls

Christians do not derive theology from NA28, UBS6, BHS, BHQ, Stephanus, Beza, or Scrivener as though editors were inspired authorities. Theology is derived from Holy Scripture. An edition is an editorial presentation of what its producers judge the biblical wording to be, sometimes accompanied by evidence for alternatives.

Nevertheless, an experiment about textual differences must fix its controls. “Scripture teaches D” depends upon which words are being treated as Scripture at the disputed place. An undefined appeal to “the Bible” cannot decide whether a proposition follows from one wording, another wording, or both.

Five layers must remain distinct:

An apparatus works chiefly at the first layer. A translation already embodies judgments about the second. A confession operates at the fifth. Confusion begins when evidence from one layer is silently made to do the work of another.

The editions named in this project are therefore controls: they fix the textual input being tested. They do not replace the Holy Spirit speaking in Scripture as the final authority of Christian theology.

Westminster’s Two Ways of Deriving Doctrine

The Westminster Confession gives the proper rule. The whole counsel of God is “either expressly set down in Scripture, or by good and necessary consequence may be deduced from Scripture.”²

That statement rejects two errors.

The first is a narrow biblicism that recognizes only doctrines stated in one verse with later theological vocabulary already present. On that rule, the church could not confess the Trinity or the two natures of Christ unless a single verse supplied the completed formula.

The second is uncontrolled system-building. Westminster does not authorize any consequence that is possible, suggestive, or attractive to the interpreter. The consequence must be both good and necessary.

A doctrine is expressly set down when a passage directly asserts its material content. This is a semantic relation, not a verbal trick. John 1:1 expressly teaches the deity of the Word although the later sentence “the Word is divine” is not a letter-for-letter reproduction of the Greek clause. Direct teaching requires us to preserve the subject, predicate, relations, modifiers, and negations of the text in its discourse. A shared word does not guarantee a shared proposition, and different words may communicate the same truth.

A consequence is good when its premises have been responsibly established, its terms retain their proper senses, and its form of reasoning is valid. Logic cannot rescue a mistranslation; exegesis cannot rescue an invalid inference.

A consequence is necessary when the conclusion follows from the premises and cannot be denied without surrendering or contradicting at least one of them. The argument need not fit into three lines of an elementary syllogism. It may unite many passages and distinctions. Yet a merely possible harmony is not a necessary deduction.

Finally, the consequence must be visible. The theologian must identify his premises and show what they entail. A row of references is not an argument, and traditional terminology cannot be allowed to hide an inferential gap.

Scripture Reasons from Scripture

The legitimacy of theological inference is not a concession imposed upon the Bible by later systematicians. Scripture itself reasons from revealed premises.³

In Matthew 22:31–32, our Lord draws a conclusion about resurrection from the covenantal words, “I am the God of Abraham, and the God of Isaac, and the God of Jacob.” In verses 41–46, he reasons from David’s calling the Messiah “Lord.” Hebrews 7 relates tribe, priesthood, law, promise, and Christ to establish the change in priestly administration. First Corinthians 15 traces the denial of resurrection through preaching, faith, apostolic witness, the state of the dead, and Christian hope.

These arguments are cumulative, but they are not careless. They may turn upon a title, a tense, a relation, a negation, or a conditional. Canonical reasoning does not make exact words less important. It shows how exact words become premises in a larger demonstration.

The biblical model also forbids atomism. Jesus reads Exodus within the covenant. Hebrews relates narrative and law to promise and fulfillment. Paul follows one denied proposition through the gospel. Building theology is neither verse collecting nor free construction; it is disciplined reasoning within one divine revelation.

The Construction Chain

The normal movement from text to confession may be stated simply:

identified reading → contextual meaning → local proposition → canonical relations → doctrinal synthesis → confessional formulation

Each arrow represents work that may succeed or fail.

1. Identified reading: Establish the wording and its status in the operative edition.
2. Contextual meaning: Examine grammar, syntax, discourse, genre, speaker, audience, and historical setting.
3. Local proposition: State what this passage teaches without importing the completed system.
4. Canonical relations: Bring other passages as confirmation, qualification, distinction, or additional premise.
5. Doctrinal synthesis: Relate the collected propositions coherently.
6. Confessional formulation: Compare that synthesis with the exact affirmation and denials of the confession.

Consider the Shorter Catechism’s confession that Christ is God and man “in two distinct natures, and one person, forever.”⁴ No single verse contains that finished sentence. Yet its parts are scripturally warranted: John 1:1 identifies the Word as God; John 1:14 says that the same Word became flesh; Hebrews 2:14–17 teaches his true participation in human nature; the Gospels attribute divine and human acts to the one Christ; and the resurrection and continuing priesthood establish that the union does not dissolve. The church’s technical terms guard relations already given in revelation.

That is construction. The finished formula contains more organization than any one proof text, but it contains no foreign theological material.

Keep an Open Premise Ledger

Grammar, lexicons, historical evidence, creeds, confessions, and earlier theologians are indispensable servants. They help us discover the sense of words, recover the setting of an argument, name a distinction, or expose an inconsistency. But a servant must not be promoted secretly into a source of revelation.

Every material claim should therefore be entered in an open premise ledger:

Prior theology may frame our questions and provide language. No reader approaches Scripture with an empty mind, nor should the church forget the teachers Christ has given it. The test is not mental amnesia. It is public accountability. Can the argument stand when every indispensable premise is traced to an admitted source?

This is also why discovery and justification must be distinguished. Augustine or Owen may teach me where to look. Their insight may be the occasion of discovery. The doctrine is justified as divine doctrine only when its warrant is shown from Scripture. Ecclesiastical testimony is real and weighty, but ministerial.⁵

Compatibility Is Not Derivation

The familiar assertion that “the doctrine is still there” often conceals very different claims:

Only the final two establish the doctrine under Westminster 1.6. Compatibility may show that a doctrine has not been disproved; it does not show that it has been built. A best-explanation argument may become a necessary consequence when rival constructions are excluded by further premises. Until then, its degree of warrant must not be exaggerated.

Proof texts also perform different work. One may state a principal proposition; another may connect two propositions; a third may establish a boundary that excludes a rival reading; a fourth may merely confirm what is already established. Losing a boundary text may be more serious than losing several illustrations. Counting references cannot tell us what doctrinal weight each one bears.

How a Variant Reaches a Doctrine

Suppose doctrine D requires P, Q, and R, and a disputed reading in Text A directly supplies P. If the operative text changes, several outcomes are possible.

• Text B may independently and expressly supply P. A proof changes, but the doctrine remains complete.
• Texts B and C may necessarily imply P. The doctrine remains derivable, though less directly.
• The remaining passages may support only a general form of P while failing to establish a required relation, boundary, or exclusion.
• The inherited doctrine may supply P, which is then read back into passages that do not entail it.
• The corpus may permit both P and a materially different P₂ without deciding between them.
• The new reading may assert not-P in the same respect.

The theological effect of a variant therefore cannot be measured by the number of altered words. A one-word negation may reverse a proposition. A long omission may remove only repetition. The relevant question is what the reading asserts and where that assertion stands in the dependency structure.

Later reconstructions will be judged by five measures: directness, completeness, exclusivity, stability, and coherence. How many steps separate the reading from the doctrine? Does every material part follow? Are contrary constructions excluded? Is one declared textual base used consistently? Do the resulting premises agree across the canon?

An indispensable premise will be marked as explicitly stated, necessarily entailed, cumulatively probable, merely compatible, not established, or contradicted. The final judgment may not outrun the weakest indispensable premise.

The Strongest Objections

The defender of modern critical editions may object that this method demands a kind of mathematical certainty foreign to ordinary exegesis. It does not. Many historical and exegetical judgments are made with degrees of probability. But a binding confessional proposition is a stronger claim than an interesting interpretation. If the church says that a doctrine is deduced by good and necessary consequence, the necessity must appear at the doctrinal level even when some supporting historical judgments remain prudential.

He may also object that the method unfairly tests the critical text against a confession written in the received-text era. That historical relation is the very reason for the experiment, but it does not decide the result. No received reading will be treated as authentic merely because it supplies a convenient Westminster premise. Textual authenticity and doctrinal consequence remain distinguishable questions. The first asks which wording should be received; the second asks what follows if that wording is received.

The received-text advocate faces the equal burden. He may not reason from the desired theology to a preferred reading and then cite that reading as independent proof. A more explicit reading is not automatically authentic. Its historical, transmissional, and theological case must be made on its own proper grounds.

The method is thus demanding of both sides. The claimant of derivation must show a sufficient chain. The claimant of failure must examine the strongest responsible alternative proofs rather than declaring defeat after one familiar proof text changes.

The Rule for This Series

The following rule will govern every later article:

A Westminster proposition is built from the designated text when all its material theological content—including necessary subjects, predicates, relations, qualifications, and denials—is either expressly taught by the operative readings or follows from them by good and necessary consequence, under a consistent textual base and a canonically coherent interpretation.

---

1. Westminster Shorter Catechism 2–3, Orthodox Presbyterian Church. ↩
2. Westminster Confession of Faith 1.6, Orthodox Presbyterian Church. ↩
3. See Matthew 22, Hebrews 7, and 1 Corinthians 15. These references concern the form of canonical reasoning; the ESV does not control disputed original-language details. ↩
4. Westminster Shorter Catechism 21, Orthodox Presbyterian Church. ↩
5. Westminster Confession of Faith 1.5, 1.9–1.10, Orthodox Presbyterian Church. On the church’s ministerial reception of Scripture, see John Cosin, A Scholastical History of the Canon of the Holy Scripture, available through the Internet Archive, and Louis Gaussen, The Canon of the Holy Scriptures Examined in the Light of History, available through the Internet Archive. ↩

The construction is not independent when an indispensable element is supplied only by an excluded reading, an unmarked translation departure, an inherited conclusion the remaining passages do not entail, a merely possible inference, an equivocation, or inconsistent editorial rules.

This definition is exacting because Westminster’s claim is exacting. The Standards do not say that the whole counsel of God is merely compatible with Scripture or can be made plausible by Scripture. They say it is set down there or may be deduced from there.

The critical question in each later article will therefore be concrete: when the operative reading changes, which local proposition changes, which doctrinal dependency is affected, what alternative premise remains, and does the complete confessional formulation still follow?

Only after those questions are answered can one responsibly say that a theology has—or has not—been rebuilt.

Notes and Further Reading