Shepherd of Hermas

The Shepherd of Hermas is a Christian text written in Rome around the middle of the second century, cast as a series of visions given to a Roman Christian named Hermas, first by an aged woman who represents the church and then by an angelic messenger who appears dressed as a shepherd.¹ Widely read and valued in the early church, it was nonetheless excluded from the New Testament canon: the Muratorian fragment rejected it by name as written too recently to carry apostolic authority, and it appears on no fourth-century canonical list.¹

Content

The work falls into three parts: Visions, Commands (or Mandates), and Similitudes, the source of its full title.² The four Visions come through the ministry of “an aged woman, who is the church of Christ”; the twelve Commands and ten Similitudes come instead from a figure who appears to Hermas “in the habit of a shepherd, clothed with a white cloak, having his bag upon his back, and his staff in his hand.”² Both sections work to “unfold the truths of Christianity with its doctrines and duties.”² Barrows judged the writing itself the product of “a most luxuriant imagination,” admitting that a reader of the Visions and Similitudes in particular can find himself “bewildered in a thicket of images and similitudes, some of them grotesque and not altogether congruous,” while still allowing that the work “throws much light on the religious ideas and tendencies of its age.”²

Confessional Definition

The 1689 Second London Baptist Confession, compiled more than fifteen centuries after the Shepherd was written, has no chapter devoted to it and never names it. Its chapter “Of the Holy Scriptures” states the general principle under which the work was excluded from the canon: only the sixty-six books of the Old and New Testaments make up the church’s rule of faith, and any other writing, “not being of divine inspiration,… [is] no otherwise approved or made use of than other human writings.”³ That describes the Shepherd’s own reception: read approvingly by many early Christians, yet never received as Scripture.

Historical Development

Authorship and date

The Shepherd makes no claim to apostolic authorship.¹ Origen connected its author to the Hermas whom Paul greets in Romans 16:14, a connection Kruger judges “erroneous”¹ and Barrows doubts too, since it rests on nothing more than a shared name.² The Muratorian fragment instead names the author as the brother of Pope Pius, and gives the clearest datable claim about the work’s composition: Hermas, it says, “wrote very recently in our times, while the bishop Pius, his brother, occupied the chair of the church at Rome.”⁴ Muratori discovered the fragment bearing his own name in the Ambrosian Library at Milan; the later of the two dates usually given for Pius’s death is AD 157, which places the Shepherd’s composition around the middle of the second century.⁴

Reception and exclusion from the canon

The Muratorian fragment itself, usually dated to around AD 180,⁵ rejects the Shepherd by name as written “very recently, in our times” and not during the time of the apostles.¹ Irenaeus, writing around the same period, uses the Greek term for “Scripture” (γραφή) of the Shepherd on one occasion but never cites it again afterward, leaving scholars such as Grant and Campenhausen uncertain whether he meant the word in its technical sense or merely as “writing.”¹ Clement of Alexandria cites the Shepherd more often than either of them but never affirms its canonical status.¹ Origen, writing somewhat later, cites it alongside scriptural books yet explicitly leaves it out of his own canonical lists.¹ Tertullian rejects the book outright as false, adding that “every council of Churches” had judged it apocryphal.¹

By the fourth century the judgment had hardened further. Eusebius does not even place the Shepherd among the “disputed” books, ranking it instead as outright “spurious” (νόθα).¹ Athanasius’ Thirty-Ninth Festal Letter (AD 367)⁵ denies it canonical status while still affirming its usefulness for catechetical instruction, and the work appears in none of the fourth-century canonical lists.¹ Its continued popularity survived that exclusion: the Shepherd is bound at the end of Codex Sinaiticus alongside the Epistle of Barnabas, a placement Kruger attributes to the common scribal practice of appending disputed or generally useful books to a manuscript rather than to any claim of canonical status, the same practice that placed 1 and 2 Clement at the end of Codex Alexandrinus.¹

Teaching on Repentance

The Shepherd’s own content bears on a live question in the post-apostolic church: whether, and how often, a Christian who sinned after baptism could be restored. Against a more rigorist view that allowed no further repentance after baptism at all, Bavinck locates in the Shepherd’s Visions, Commands, and Similitudes the judgment “that at least one additional conversion was possible” for the lapsed.⁶ Bavinck’s own note on the point adds a caveat: “but elsewhere he expresses himself less generously,” since the work is not perfectly consistent on how far that allowance extends.⁶ This one-time allowance fed into the wider penitential system the early and medieval church developed for postbaptismal sin, a system Bavinck elsewhere judges a departure from the New Testament’s own teaching that repentance is not “a second plank after shipwreck” but a continual return to the grace already given in baptism.⁶

Objections and Responses

Doesn’t Irenaeus’s use of the word “Scripture” for the Shepherd show it was once considered canonical? Irenaeus uses that term for the work only once and never cites it again, and Grant and Campenhausen read this as ambiguous evidence for how technically he meant the word.¹ The Shepherd’s wider reception did not share even that ambiguity: the Muratorian fragment named it too recent for apostolic authority, Tertullian called it false, and Eusebius classed it as spurious rather than merely disputed.¹

If the Shepherd was excluded from the canon, why does Codex Sinaiticus, one of the oldest and most important biblical manuscripts, include it? Appending a respected but non-canonical work to the end of a manuscript was ordinary scribal practice for books judged useful without being treated as Scripture, the same practice that placed the Epistle of Barnabas alongside the Shepherd in Sinaiticus and 1 and 2 Clement at the end of Codex Alexandrinus.¹

Related Doctrines

References

Michael J. Kruger, Canon Revisited: Establishing the Origins and Authority of the New Testament Books (Wheaton, IL: Crossway, 2012), chap. 8, “Problem Books and Canonical Boundaries”: “In the end, there is little here to suggest that the Shepherd ought to have a place among the canonical Scriptures.” ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵
Elijah Porter Barrows, Companion to the Bible (New York: American Tract Society, 1867), chap. xxxii. ↩ ↩² ↩³ ↩⁴ ↩⁵
The Second London Baptist Confession of Faith (1677/1689), chap. 1, “Of the Holy Scriptures,” para. 3. ↩
Barrows, Companion to the Bible, chap. ii. ↩ ↩²
Kruger, Canon Revisited, chap. 6, “The Corporate Reception of the Canon”: “Our earliest canonical list comes from the Muratorian fragment toward the end of the second century (c. 180).” ↩ ↩²
Herman Bavinck, Reformed Dogmatics: Abridged in One Volume, ed. John Bolt (Grand Rapids: Baker Academic, 2011), chap. 19, “Faith and Conversion.” ↩ ↩² ↩³