HistoricalChristian.Faith

Does the exception permit remarriage, or only separation?

Type: supporting argument (off-timeline) · Assessment: contested — grant that Jesus permits putting away a wife for porneia; does the exception dissolve the bond so the innocent party may marry again (the Erasmian/Protestant reading, with the lone Latin father Ambrosiaster before it), or license only separation from bed and board with the bond intact and no remarriage (the patristic and Roman Catholic reading — Hermas, Clement, Chrysostom, Jerome, Augustine)? The Greek syntax of Matthew 19:9 is the crux — and it will not settle the question by itself. The field is genuinely divided · this is the central disputed question of the whole doctrine

Related arguments: whether the bond dissolves — see also is the marriage bond indissoluble while both spouses live? · the Pauline privilege — does desertion free the believer to remarry?

The data every reading must fit

Matthew 19:9 reads, in the Greek Erasmus first printed, “hos an apolysē tēn gynaika autou mē epi porneia kai gamēsē allēn moichatai” — “whosoever shall put away his wife, except it be for fornication, and shall marry another, committeth adultery” (Matthew 19:9). The whole dispute turns on the position of the three-word exception, mē epi porneia (“except for fornication”), which sits between the two verbs it might qualify: the putting-away (apolysē) that comes just before it, and the marrying-another (gamēsē allēn) that comes just after. Does “except for fornication” carve its exception out of the divorce alone — so that a man may dismiss an unfaithful wife but still not remarry — or out of the whole conditional, divorce-and-remarriage together, so that for porneia a man may both dismiss her and marry again?¹

Two further data bound the question. First, the Sermon on the Mount form is built differently: “whosoever shall put away his wife, saving for the cause of fornication, causeth her to commit adultery” (Matthew 5:32) — there the exception plainly attaches to the putting-away, and remarriage is not in that clause at all, so 5:32 leans, if anything, toward the narrower scope. Second, the three parallel witnesses carry no exception and each condemns the remarriage as such: “whosoever shall put away his wife, and marry another, committeth adultery” (Mark 10:11–12; Luke 16:18), and Paul’s direct report of the Lord’s charge tells the separated wife to “remain unmarried” (1 Corinthians 7:10–11). Any reading that lets Matthew’s exception reach the remarriage must square that with three parallels in which the remarriage is the very thing forbidden.

The two readings

(A) The exception reaches the remarriage — the bond is dissolved, and the innocent may marry again. This is the Erasmian and majority-Protestant reading. Its grammatical champion is D. A. Carson, who argues that the exceptive clause must “be understood to govern the entire protasis,” so the sense is: “anyone who divorces his wife and marries another commits adultery — but this does not hold in the case of porneia.” His decisive point is that the narrow reading renders the sentence nonsensical: “anyone who divorces his wife, except for porneia, commits adultery” is simply untrue as it stands (a bare divorce is not adultery), so the adultery being excepted must be the adultery of the remarriage — which means the exception is releasing the innocent party to remarry.² The reading is as old, in the Latin West, as Ambrosiaster (c. 375), who let an innocent husband “marry another woman, if he divorces a wife who sins”; it was recovered by Erasmus in 1516–19, hatched by Luther and given careful exegesis by Calvin, and passed through the Reformers into the Westminster Confession — the settled majority position of Protestantism.³

(B) The exception reaches only the divorce — the bond survives, and no remarriage is licensed. This is the reading of the early church and, later, of Rome. Its modern grammatical case was pressed by William Heth and Gordon Wenham: the exceptive phrase is elliptical — it has no verb of its own — so the verb to be supplied is the one it stands immediately after, “put away,” not the “marry another” that follows. Being adverbial, mē epi porneia modifies the nearer verb (“divorces…except for fornication”) and leaves the remarriage clause under the unqualified judgment “commits adultery.” On this reading Jesus says: it is always wrong to divorce, except for unchastity; and remarriage after divorce is always adultery, exception or no.⁴ This is not a modern invention read back onto the fathers: it is the oldest attested reading, held with near-unanimity from Hermas (c. 140) — “if he put his wife away and marry another, he also commits adultery” — through Clement, Tertullian, Basil, Chrysostom, and Jerome, grounded by Augustine in the marriage sacramentum, enforced by the Council of Elvira, and fixed as Roman Catholic dogma by the Council of Trent, whose canon 7 anathematizes the claim that “the innocent one who gave not occasion to the adultery” may “contract another marriage, during the life-time of the other.”⁵

Weighing the readings

Where reading (A) is strong. Carson’s “nonsensical” objection has real bite: taken at face value, the bare protasis “whoever divorces his wife…commits adultery” is false, because divorcing is not itself the act of adultery — the adultery enters with a new union, so the exception looks, on the plainest reading of the whole sentence, to be qualifying the composite act (divorce-and-remarry) rather than the divorce alone. Word order helps him too: in 19:9 the exception is embedded in a single conditional whose apodosis (“commits adultery”) is triggered by the remarriage, and a first-century reader tracking the sentence as a unit would naturally let the one exception cover the one condition. That the Reformers, reading the Greek fresh, arrived here in numbers — and that the reading has a genuine, if lonely, patristic precedent in Ambrosiaster — shows it is not a Protestant novelty invented to license divorce.⁶

Where reading (B) is strong. Three things pull the other way. Grammatically, Heth and Wenham are right that the exception is verbless and adverbial, so its default anchor is the verb it abuts — and in 5:32, where remarriage is not even in the clause, the exception unmistakably governs only the putting-away, which is at least a Matthean pattern. Canonically, the three exceptionless parallels do not merely omit an exception; they positively brand the remarriage as adultery (Mark, Luke) or forbid it outright (Paul, “let her remain unmarried”), and reading (A) has to explain why Matthew would silently reverse, for the innocent party, what Paul states as the Lord’s own charge. Historically, the separation-only reading is not one voice among many but the overwhelming consensus of the first fifteen centuries — a datum that tells heavily on how the clause was in fact heard by readers far closer to its idiom than we are.⁷

Why neither wins on the grammar. The honest verdict of the specialist literature is that the syntax of 19:9, taken by itself, is ambiguous — both scopes are grammatically available, and the sentence can be construed either way without violence. The exception’s position between the two verbs is exactly what makes it undecidable: it is near enough the remarriage to be swept into the whole conditional (A), and near enough the divorce, as its immediately preceding verb, to attach there alone (B). So the decision is forced back onto context, canon, and history — and there the two readings weight the evidence differently rather than one refuting the other: (A) privileges the internal logic of the single Matthean sentence; (B) privileges the three parallels and the reception. A telling marker of the deadlock is that William Heth, who co-wrote the fullest modern defense of (B), later changed his mind and came to permit the remarriage of the innocent party — while his co-author Gordon Wenham did not; the same evidence moved two of its ablest weighers apart.⁸

Adversarial assessment

What holds. This is a real, unresolved crux, not a manufactured one. The two readings are each independently advocated by serious grammarians and rest on genuine features of the text: (A) on the unitary logic of the 19:9 conditional and the awkwardness of an exception that excepts nothing actionable; (B) on the verbless, adverbial character of the exception, the 5:32 pattern, and the three exceptionless parallels that make remarriage the thing condemned. Each is also carried by a long and living tradition — (A) by the whole Protestant mainstream (and, before it, by Ambrosiaster), (B) by the patristic consensus and Roman and Eastern practice.

What is overstated on each side. Three cautions. (1) It is an overreach, common in the Reformed literature, to present (A) as the plain and obvious sense that only special pleading resists: the grammar does not compel it, and the three exceptionless parallels are a standing difficulty it must argue past, not a detail. (2) It is equally an overreach, common in the strict literature, to claim — as Heth and Wenham did — that “no one until Erasmus” permitted the innocent party to remarry: the Latin Ambrosiaster did so two centuries into the tradition, the Eastern churches came to permit remarriage after divorce by pastoral “economy” (a strand Origen already noticed), and a near-unanimous tradition is not a strictly unanimous one. (3) The syntactic argument is not decisive for either camp, and a page that leans its whole weight on the grammar — in either direction — claims more than the Greek delivers. The reading a person reaches is largely set by what they let govern the ambiguity: the internal sentence, or the canon and the fathers.

Assessment

Contested. Whether Matthew’s exception licenses remarriage or only separation is the central disputed question of the doctrine, and it is genuinely undecided: the Greek of Matthew 19:9 is ambiguous as to whether “except for fornication” governs the divorce alone or the whole divorce-and-remarry conditional, and each scope is defended by capable grammarians (Carson for remarriage; Heth and Wenham for separation-only) and carried by a major tradition (the Protestant mainstream and Ambrosiaster on one side; the patristic consensus, Trent, and Eastern discipline on the other). The syntax cannot settle it, so the verdict turns on whether one weights the internal logic of Matthew’s single sentence or the three exceptionless parallels and the church’s reception — strong by one measure, weak by the other, with no clear winner. This argument adjudicates the crux between the two camps that the timeline documents; it presupposes, but does not re-decide, that the exception means a wife’s marital unfaithfulness (weighed under the exception-clause readings) and that the clause is authentic to the tradition however it originated (weighed under why Matthew alone carries it). It is kept distinct from the separate question of whether the Eastern church may dissolve a marriage by pastoral “economy” (oikonomia) — a mechanism of church practice, not a claim about what the text itself licenses. A narrower proof-text bearing on the same crux — that Matthew 5:32a “makes her an adulteress” presupposes the wife’s remarriage (assessment: sound) — shows a causeless divorce leaves the bond intact, but by that very restriction does not settle the innocent party’s case after a divorce for unfaithfulness, which is the question weighed here. Underneath this exegetical crux lies the definitional one — whether marriage is a covenant breakable by treachery or an indissoluble one-flesh bond — weighed on the covenant-versus-ontological-bond page.

Sources