HistoricalChristian.Faith

The Seven Sacraments (how the number got fixed)

How did the church come to say there are seven sacraments — no more, no fewer? Not by counting from the beginning. For the first thousand years there was a rich doctrine of the sacrament but no fixed list: the Latin word sacramentum began as the soldier's oath, Greek writers spoke of the mysteria, and the fathers applied the term broadly to many holy signs. This page traces the enumeration's slow crystallization, earliest witness first — from that uncounted breadth (Tertullian, Ambrose, Augustine), through a sixth-century catalogue that arrives at a different set of six (Pseudo-Dionysius) and the twelfth-century schoolmen who first close the list (Hugh of St Victor, then decisively Peter Lombard), to its ecumenical decree at Florence (1439), the Protestant reduction to two (Luther, 1520), and its dogmatic definition "neither more nor less than seven" at the Council of Trent (1547).

Related doctrines: The sacraments — see also penance & confession · confirmation & chrismation · anointing of the sick

Timeline

Tertullian (c. 203)

Tertullian, the first Latin theologian, uses sacramentum in its ordinary sense of the soldier's oath of allegiance: the Christian's "divine sacrament" stands against the soldier's "human sacrament," and baptism is the swearing-in to "the warfare of the living God." The word names a bond of loyalty, not a rite in a numbered list.¹

Ambrose of Milan (c. 390)

Ambrose's catecheses for the newly baptized expound "the sacraments" as a small, concrete cluster of initiation rites — the font, the "spiritual seal" (the anointing), and the eucharist — each made effective by the word of Christ. He has a developed sacramental theology but counts nothing; the "sacraments" are simply the holy rites at hand.²

Augustine of Hippo (c. 410)

Augustine gives the West its enduring definition — a sacrament is a "sacred sign" of an invisible reality, the word joined to the element making a "visible word." The definition is deliberately broad, so many holy signs can be called sacramenta; yet where he numbers them he says the sacraments of the new dispensation are "in number very few," naming only baptism and the eucharist.³

Pseudo-Dionysius the Areopagite (c. 500)

Pseudo-Dionysius wrote the first systematic treatise on the church's rites, and it enumerates six sacred mysteries — baptism, the eucharist, the consecration of the ointment (chrism), the ordination of clergy, the consecration of monks, and the rites over the dead. That a careful catalogue arrives at a different six — including monastic consecration and funerals, omitting penance and marriage — is the plainest proof that no fixed number yet existed.⁴

Hugh of St Victor (c. 1134)

Hugh of St Victor, in the first great scholastic synthesis on the subject, sharpens the definition — a sacrament must not merely signify but contain grace — and rejects the loosest sign-definition (a picture or a written text is a sacred sign, but no sacrament). Yet his own working set remains a large, tiered array sorted into ranks, not a list closed at any number.⁵

Peter Lombard (c. 1150)

Peter Lombard, in Book IV of the Sentences — the theology textbook of the medieval West — did what no one had before: joining the tightened definition (a sign that causes the grace it signifies) to a named enumeration, he fixed the sacraments at seven: "baptism, confirmation, the bread of blessing (that is, the Eucharist), penance, extreme unction, orders, and marriage." Because every theologian lectured on the Sentences, this list became the standard.⁶ Marriage's place leans on the Vulgate of Ephesians 5:32, which renders Paul's Greek mystērion ("mystery") by the Latin sacramentum, so that a Latin reader met a verse apparently calling marriage "a great sacrament."⁷

Thomas Aquinas (c. 1274)

Thomas Aquinas supplies the mature scholastic defence of the number — not a fresh count, but a rationale for the list already "commonly adopted by all." In the Summa Theologiae (III, q. 65) he argues that exactly seven is fitting: the sacraments answer, one to one, the stages of the life of the body — baptism (birth), confirmation (growth), eucharist (nourishment), penance (healing), extreme unction (convalescence), orders (governance), matrimony (propagation) — a scheme he reinforces by a second derivation (remedies for the defects of sin) and a third (a virtue for each). He orders the seven and ranks the Eucharist first, "the greatest of all the sacraments," to which the others are ordained as to their end.⁸

The Council of Florence (1439)

The Council of Florence, in its decree of union with the Armenians (Exsultate Deo), lifted Lombard's list into ecumenical teaching: "There are seven sacraments of the new Law, namely baptism, confirmation, eucharist, penance, extreme unction, orders and matrimony," which — unlike the sacraments of the old law — "both contain grace and bestow it on those who worthily receive them." The decree even explains the count: five for the individual, two for the whole church.⁹

Martin Luther (1520)

In The Babylonian Captivity of the Church, Luther broke the consensus at its root. Applying his own test — a sacrament needs both a sign divinely instituted and a promise of forgiveness — he denied there are seven, opened with three (baptism, penance, the bread), and closed "strictly speaking, but two… baptism and the bread." He noted too that the one ancient witness cited for seven "omits marriage and so has only six."¹⁰

The Council of Trent (1547)

Against the Reformers, the seventh session of the Council of Trent turned the enumeration into defended dogma: whoever says the sacraments "are more, or less, than seven, to wit, Baptism, Confirmation, the Eucharist, Penance, Extreme Unction, Order, and Matrimony," or that any one of them is not truly a sacrament, "let him be anathema." The number Lombard first fixed and Florence first decreed is here made a test of orthodoxy.¹¹

Arguments

Related arguments — the sevenfold count

The timeline's central question — is the number seven an apostolic datum the church always held, or a twelfth-century construction? — divides into two genuinely advocated readings, best read together. The Protestant side presses that no closed list of seven exists before Peter Lombard: the term was used broadly, Augustine counts “very few” (two), and the one systematic early catalogue reaches a different six. The Catholic side grants the late enumeration but denies it constructed anything: the seven were always practised even where they were not yet counted, so the number is a later recognition of an apostolic reality, as the canon of Scripture was fixed late for early books. Each is weighed adversarially on its own page.

Is the sevenfold count a twelfth-century construction rather than an apostolic datum? (the construction reading) — sound

As a claim about the enumeration, the record is unambiguous: no closed list of exactly seven grace-conferring sacraments precedes Peter Lombard (c. 1150); the number is decreed ecumenically only at Florence (1439) and dogmatized at Trent (1547). The first-millennium sources use sacramentum/mysterion broadly, Augustine numbers “very few” (two), and Pseudo-Dionysius reaches a different six — a finding even Catholic reference works concede. It stays sound only kept narrow: it fixes when the count was set, not that the rites were invented (that stronger polemical form overreaches).

Were the seven always practised and only later recognized, not invented? (the discovery/recognition reading) — contested

The Catholic reply: the individual rites are genuinely ancient (baptism, eucharist, ordination, penance, the anointing of James 5, marriage-blessing, chrismation), so the number Lombard fixed recognizes a treasure received from Christ rather than inventing it — as the church fixed the biblical canon late for early books. Strong on the antiquity of the rites and the force of the canon analogy; but weak where the selection of exactly these seven, their closure, and their parity as grace-conferring depend on a definitional test that is itself twelfth-century, and on a prior claim of authority to define — recognition doing the work of construction. Strong by one discipline's lights, weak by another's, with no neutral tiebreaker (contested).

Does Ephesians 5:32 make marriage a sacrament?

The seventh item on Lombard's list is also the one whose scriptural warrant is most contested. Marriage's place leans on the Vulgate of Ephesians 5:32, which renders Paul's Greek mystērion (“this is a great mystery”) by the Latin sacramentum, so that a Latin reader met a verse apparently calling marriage “a great sacrament.” Whether that verse actually supports numbering matrimony among the grace-conferring sacraments is a genuinely disputed interpretive question — affirmed in the Latin tradition (Lombard, Aquinas, Florence, Trent), denied by the Reformers — and is weighed adversarially on its own page.

• Does Ephesians 5:32 make marriage a grace-conferring sacrament? — weak (the verse soundly teaches that marriage signifies the Christ–Church union; the leap to a numbered grace-conferring rite rests on the word sacramentum, which the same Vulgate uses for “the mystery of godliness” with no rite in view — and Augustine's “great sacrament” is the bond's indissolubility, not a numbered sacrament).

Sources