HistoricalChristian.Faith

Epiphanius of Salamis on the Old Testament Canon

Dates: c. 385 AD (his canon catalogues span the Panarion, c. 377, and On Weights and Measures, c. 392; Epiphanius c. 310/320–403) · Wikipedia
Verdict on the deuterocanon: NO by the list — but his own usage cites it as Scripture. By the catalogue: Wisdom and Sirach kept as "useful," Baruch folded into Jeremiah, Tobit/Judith/Maccabees absent. By the practice: he quotes Sirach as "Scripture," argues the Incarnation from Baruch as "the sacred scripture," and even tails his summary of "all the sacred scriptures" with "the Wisdoms, I mean Solomon's and Sirach's." The great heresy-hunter of Salamis gives the same Hebrew-sized canon as the Eastern Fathers before him: twenty-two books by the count of the Hebrew letters (twenty-seven when the doubled books are separated), with Tobit, Judith, and the Maccabees simply absent. But like Origen, Cyril, and Athanasius, his "No" is not a bare thirty-nine: Baruch and the Epistle of Jeremiah ride along inside the canonical Jeremiah, and the two Wisdom books — the Wisdom of Solomon and the Wisdom of Sirach (Ecclesiasticus) — are named, judged "useful," and explicitly set outside the canon, "of disputed canonicity."

Claims & evidence

The canon: "twenty-seven recognized canonical books, but twenty-two when counted according to the letters"

In On Weights and Measures Epiphanius embeds his Old-Testament catalogue inside the Septuagint translation-legend: what the Seventy-(two) rendered, he says, was the recognized canon — counted two ways.¹

…and thus were translated the twenty-seven recognized canonical books, but twenty-two when counted according to the letters of the alphabet of the Hebrews. For the names of the letters are twenty-two. But there are five of them that have a double form… Therefore in this manner the books also are counted as twenty-two; but there are twenty-seven, because five of them are double.

He then runs the list out in four "pentateuchs" plus two leftover books: the Law (Genesis–Deuteronomy); five "in verse" (Job, Psalms, Proverbs, Koheleth/Ecclesiastes, Song of Songs); the Writings (Joshua, Judges with Ruth, 1–2 Paraleipomena/Chronicles, 1–4 Kingdoms); the Prophets (the Twelve as one, Isaiah, Jeremiah, Ezekiel, Daniel); and the two remaining — Ezra (the two counted as one) and Esther.¹ Measured against the disputed books the verdict on the canon proper is plain: no Tobit, no Judith, no Maccabees. This is the same Hebrew-letter reckoning Origen, Cyril, and Athanasius gave — the standard fourth-century Eastern canon. (The parallel list in the Panarion agrees: "twenty-seven books of the Old Testament, which are (also) numbered as twenty-two."²)

The Wisdom books: "helpful and useful, but… not… of the recognized" — "of disputed canonicity"

Epiphanius does not pass the two great Wisdom books over in silence; he names them, praises them, and rules them out — the clearest statement on this page of a graded "No":³

For there are two (other) poetical books, that by Solomon called "Most Excellent," and that by Jesus the son of Sirach and grandson of Jesus… which also are helpful and useful, but are not included in the number of the recognized; and therefore they were not kept in the chest, that is, in the ark of the covenant.

The Panarion says the same thing in a word: the Jews "have two more books of disputed canonicity, the Wisdom of Sirach and the Wisdom of Solomon, apart from certain other apocrypha."² This is precisely the middle posture the Eastern Fathers had been refining: not canon, not heretical forgery, but useful — the same rank in which Athanasius had placed the deuterocanon ("appointed by the Fathers to be read") a generation earlier. The vivid image is Epiphanius's own: helpful books, but not in "the chest… the ark of the covenant."

The bundled exception: "the epistles of Baruch and of Jeremiah" inside Jeremiah

As with every Eastern list on this timeline, the Hebrew count is not perfectly clean of the Greek extras. Closing the catalogue, Epiphanius notes that the twenty-two are reckoned "even with the book of Psalms and those by Jeremiah," and spells out what "those by Jeremiah" means:⁴

…those by Jeremiah—I mean Lamentations and the epistles of Baruch and of Jeremiah, although the epistles are not in use among the Hebrews, but only Lamentations, which is joined to Jeremiah.

So Baruch and the Epistle of Jeremiah are folded into the canonical Jeremiah, exactly as the Greek manuscripts carried them — the recurring Eastern signature seen in Origen, Cyril, and Athanasius. Epiphanius is even franker than the rest about the seam: he grants that "the epistles are not in use among the Hebrews" — an honest admission that what he keeps inside Jeremiah is a Greek addition the Hebrew text does not have. The Panarion list likewise gives "the Prophet Jeremiah, with the Lamentations and the Epistles of Jeremiah and Baruch."²

Yet in his own writing the disputed books do the work of Scripture

Set against the catalogue, Epiphanius's running practice across the Panarion is the same list-versus-usage strain seen in Cyril, Hilary, and Athanasius. The Father who lists Sirach "outside… the recognized" and keeps Baruch only as an appendix to Jeremiah elsewhere reaches for both as Scripture in the heat of argument — and once even lets them ride at the tail of "all the sacred scriptures."

Sirach, introduced "Scripture says." Refuting the Valentinian Ptolemy, Epiphanius caps his ridicule of the heretic's converts with a proverb he ascribes to Scripture outright — quoting Sirach 13:16:⁵

Scripture says, "Every bird flocketh with its kind, and a man will cleave to his like."

The very book his Old-Testament list rules "not… of the recognized" he here cites with the same formula he would use for Proverbs or the Psalms — "Scripture says."

Baruch, pressed as "the sacred scripture" for the Incarnation. Twice Epiphanius makes the Christological argument turn on Baruch 3:36–38. Against the Arians, expounding why the Son could say "my God and your God" only after the Word took flesh, he folds the Baruch line into a chain of proof-texts running through Genesis, the Psalms, and John — that it was "when he had taken our body,"⁶

…[when he] "appeared on earth and consorted with men," and become one of us, then he said "my God and your God…" to his disciples…

And against Photinus — who denied the Son's pre-existence — Epiphanius reads the same Baruch verse as proof of both the Word's eternity and his coming in the flesh, calling it "the sacred scripture": "He which hath found out every path of knowledge" implies "his preexistence," while the next clause, "He appeared on earth," he says, "indicates his coming incarnation."⁷ The book his list keeps only as an appendix to Jeremiah here carries, in his own hand, a load-bearing proof against two heresies.

The sharpest concession: the two Wisdom books named among "all the sacred scriptures." Confronting the Anomoean over a term "scripture never mentions," Epiphanius bids him search the whole of Scripture — and his sweep of "all the sacred scriptures" runs the Old Testament, the Gospels, the Epistles, Acts, the Revelation of John, and then:⁸

…the Revelation of John, and the Wisdoms, I mean Solomon's and Sirach's—and, in a word, in all the sacred scriptures…

This is the one place where his usage does not merely cite a disputed book but seems to list one: the two Wisdoms appended to a roll he calls "the sacred scriptures." It sits in plain tension with On Weights and Measures, where the same two books are "helpful and useful, but… not… of the recognized" and "not kept in the chest… the ark of the covenant." Modern editors flag the contradiction as unresolved — he ranks the Wisdoms below the canon in the catalogue, yet here lets them stand inside the body of "all the sacred scriptures."

None of this overturns his formal verdict: the canon proper is the Hebrew twenty-two, and Tobit, Judith, and the Maccabees stay out of every list he gives. But it shows what his graded "useful, not recognized" amounted to in practice — quoted as "Scripture," pressed into doctrine, and once shelved among "all the sacred scriptures." The gap between Epiphanius's list and Epiphanius's usage is one more instance of the wider pattern weighed on the "did the early church reject the deuterocanon?" argument page.

Notes

• List and usage do not line up. Epiphanius is often enlisted as a flat witness against the deuterocanon, and his catalogues do exclude Tobit, Judith, and the Maccabees. But two things cut against the flat reading. First, even his lists are not the bare thirty-nine: Baruch and the Epistle of Jeremiah ride inside Jeremiah, and the two Wisdom books are named and kept "useful." Second, and sharper, his practice quotes the disputed books as Scripture — Sirach "Scripture says," Baruch as "the sacred scripture" proving the Incarnation — and his Panarion sweep of "all the sacred scriptures" even tails off with "the Wisdoms, I mean Solomon's and Sirach's." The lead that brought this practice onto his page flagged exactly this list-versus-usage tension.
• NO to the deuterocanon's canonicity, but not to its worth. Epiphanius is a clean fourth-century witness against the canonical standing of Tobit, Judith, Wisdom, Sirach, and the Maccabees — the canon is the Hebrew twenty-two. Yet he singles out the two Wisdom books as "helpful and useful" and "of disputed canonicity," and keeps Baruch and the Epistle inside Jeremiah. The honest summary is the familiar Eastern one: not Scripture, but worth reading.
• The "ark of the covenant" image. Where Athanasius spoke of books "appointed by the Fathers to be read," Epiphanius gives the same middle rank a concrete picture: useful books that "were not kept in the chest, that is, in the ark of the covenant." It is one of the sharpest fourth-century articulations of the canonical-versus-edifying line.
• Two works, one verdict. The fuller list (with the "ark" line) is in On Weights and Measures (c. 392); a parallel catalogue, with the "disputed canonicity" phrase, stands in the earlier Panarion (c. 377). They agree, which is why the witness is dated to the mid-380s between them.
• The numerology, and what it does not dissolve. In On Weights and Measures Epiphanius ties his twenty-two to a full number-scheme — "twenty-two works that God did in the six days … twenty-two generations from Adam to Israel … twenty-two signs of the letters … twenty-two books from Genesis to Esther." Gary Michuta enlists exactly this to argue that the patristic 22/24 counts are mystical numerology, not canon boundaries. But Epiphanius is the witness who most plainly defeats that inference: the same treatise names the two Wisdom books and rules them "not … of the recognized," "not … in the ark" — an explicit canon-verdict standing beside the numerology, not displaced by it. The point is weighed on its own argument page (assessment: weak).
• The canon set inside the Septuagint legend. Epiphanius gives this catalogue while retelling the story of the Seventy-two translators — the same legend weighed on the Origin and Inspiration of the Septuagint page. He thus holds the two together: the Greek Bible is the inspired translation, yet its canon is the Hebrew twenty-two, with the Wisdom books and the extra Jeremianic pieces marked off as not "recognized."

Sources