
The Shepherd of Hermas was read with serious attention in early Christianity, yet the Muratorian Fragment says it could not be read publicly to the people in church.
That formula records a distinction that modern talk of “included” and “excluded” can easily flatten. A work might be worthy of reading and instruction, yet not authorized for public proclamation among the writings assigned to prophets or apostles. The Shepherd’s long afterlife, visible in major manuscripts and in later catalogues, makes the distinction more important.
A Text That Travelled Widely
The Shepherd is a large early Christian work presented through visions, commands (often called Mandates), and parables or Similitudes. Its voice is practical but also urgent: repentance, moral repair, communal discipline, and the Church imagined as a structure still being built. Whatever one concludes about its precise date or authorship, the work’s survival and circulation show that it was important to readers across several centuries.
One striking witness is Codex Sinaiticus, the fourth-century Greek Bible manuscript. The Shepherd appears after the Epistle of Barnabas in the codex’s later leaves. Its presence does not, by itself, settle a canon list; it does, however, make visible a world in which a book’s devotional, instructional, and manuscript life could be substantial without fitting every later boundary in the same way.
The Sentence That Governs the Debate
The most famous ancient ruling on the Shepherd occurs in the Muratorian Fragment, a damaged Latin list of books and authors. The following Latin text preserves the lineation and odd spellings in Henry M. Gwatkin’s printed transcription: spaces within words such as “nuperrim e,” “cathe / dra,” and “se pu / blicare” must not all be treated alike. The first and last examples reproduce breaks in the print; the middle of a word in “nuperrim e” is itself printed with a space.
pastorem vero
nuperrim e temporibus nostris in urbe
roma herma conscripsit sedente cathe
dra urbis romae aecclesiae pio eps fratre
eius et ideo legi eum quide oportet se pu
blicare vero in eclesia populo neque inter
profetas completum numero neque inter
apostolos in fine temporum potest
Lines 73–80. Spelling and division are retained; this is a transcriptional witness, not a normalized critical text.
“And therefore it ought indeed to be read; but it cannot be read publicly to the people in church either among the Prophets, whose number is complete, or among the Apostles, for it is after their time.”
English translation of the Fragment, lines 77–80.
The crucial contrast is built into the sentence itself. Legi… oportet allows reading; publicare… in eclesia populo denies a public ecclesial use. The ruling then gives two limits: the Shepherd is not to be placed among the Prophets, “whose number is complete,” or among the Apostles, because it comes after their time. In other words, the text is treated as useful, but its usefulness is not made identical with apostolic or prophetic status.
Read, but Not Publicly Proclaimed
Later fourth-century witnesses preserve similarly graded categories, though they are not simply paraphrases of the Muratorian sentence. Eusebius says the Shepherd had been disputed by some, was considered useful for elementary instruction by others, and had been publicly read in churches; in his summary classification, he places it among writings he calls “rejected,” while distinguishing them from works he dismisses as heretical inventions.
Athanasius’s Thirty-Ninth Festal Letter (367) makes the same sort of distinction in a different register. After listing books “included in the Canon,” he names the Shepherd among works “appointed by the Fathers to be read” by those newly joining and seeking instruction. The point is not that every community used identical labels. It is that “to be read” could be a positive category without being the same as “canonical.”
This is why the phrase “a book read and a reading refused” is more precise than saying the Shepherd was simply rejected. The surviving language speaks of a boundary around one form of use, public reading to the congregation, not a blanket instruction to avoid the work.
The Manuscript Is Not the List
Another distinction is essential. The surviving Latin copy is not automatically the date of the list it transmits. The Biblioteca Ambrosiana catalogue identifies the manuscript as I 101 sup., a parchment codex of IV + 77 + I leaves, and gives an uncertain catalogue date of 676–750. The shelfmark is sometimes printed with J rather than I in older references; the Ambrosiana record itself uses I 101 sup.

Samuel Prideaux Tregelles described the fragment’s extent as the two sides of folio 10 and the first twenty-three lines of folio 11 recto. That is a nineteenth-century report about the location of the fragment in the codex. It should not be confused with a date for the underlying list.
The photograph reproduced here must therefore be named carefully: it is a photograph of a medieval copy containing a disputed ancient list, not a photograph of the original moment when that list was first composed. The difference is basic textual criticism, but it is frequently obscured in popular retellings.
Why It Does Not Settle the Date?
The Shepherd notice is often enlisted in arguments for a late-second-century date for the Muratorian Fragment. Its wording says that Hermas wrote the work “very recently, in our times” in Rome while Pius, identified as his brother, occupied the episcopal chair. If that notice is taken at face value as a near-contemporary recollection, it has obvious chronological force.
But the language is not a numbered date, and the Latin is transmitted in a notoriously difficult form. Geoffrey Mark Hahneman’s 1987 Oxford D.Phil. thesis presses this problem hard. On p. 296 he argues that the cumulative evidence makes it more reasonable to redate the list as a fourth-century Eastern document, possibly from Palestine or western Syria; on p. 300 he proposes that it most probably derives from sometime after Eusebius.
Those claims should be reported as an argument, not smuggled back into the transcription as though the Latin itself said “fourth century,” “Palestine,” “western Syria,” or “after Eusebius.” The connective et ideo leads from the notice about Hermas to the rule about reading; it does not turn the sentence into a precise year. The conventional early date and revisionist proposals must remain distinct from the date of the surviving copy.
What the Shepherd Still Reveals
The Shepherd’s reception is evidence for the process by which Christian communities weighed authority, usefulness, and public reading. A book could be cherished for moral formation, copied beside highly valued texts, and still encounter limits in liturgical or public use. That layered reception is the historical fact worth recovering.
The Muratorian Fragment makes the tension unusually explicit: “it ought indeed to be read,” yet it cannot be read publicly among the prophets or apostles. The result is a reading permitted, preserved, and debated, precisely the kind of reading that keeps the Shepherd important for the history of early Christianity.
By Aris Manus
References
Bible Researcher. The Muratorian Fragment. Includes Bruce M. Metzger’s English translation and the Latin transcription from Henry M. Gwatkin, ed., Selections from Early Writers Illustrative of Church History to the Time of Constantine (London: Macmillan, 1937), pp. 82–88. Accessed 4 October 2026. https://www.bible-researcher.com/muratorian.html
Codex Sinaiticus Project. See the Manuscript | Shepherd of Hermas. Digital manuscript page, including BL f.341r. Accessed 4 October 2026. https://codexsinaiticus.org/en/manuscript.aspx?__VIEWSTATEGENERATOR=01FB804F&book=61&lid=en&side=r&zoomSlider=0
Biblioteca Ambrosiana. Manoscritti antichi: I 101 sup. Catalogue record: IV + 77 + I leaves; parchment; uncertain date 676–750. Accessed 4 October 2026. https://ambrosiana.comperio.it/opac/detail/view/ambro:catalog:76502
Tregelles, Samuel Prideaux, ed. Canon Muratorianus: The Earliest Catalogue of the Books of the New Testament. Oxford: Clarendon Press, 1867. See the description of the fragment’s placement on folio 10 and folio 11 recto.
Eusebius. Ecclesiastical History 3.3.6 and 3.25. Trans. Arthur Cushman McGiffert, in Nicene and Post-Nicene Fathers, second series, vol. 1. Digital text accessed 4 October 2026. https://www.newadvent.org/fathers/250103.htm
Athanasius of Alexandria. Festal Letter 39 (367), §§5–7. Trans. R. Payne-Smith, in Nicene and Post-Nicene Fathers, second series, vol. 4. Digital text accessed 4 October 2026. https://www.newadvent.org/fathers/2806039.htm
Hahneman, Geoffrey Mark. The Muratorian Fragment and the Development of the Canon. D.Phil. thesis, Christ Church, University of Oxford, Trinity Term 1987, pp. 296, 300. Digitized record in Oxford University Research Archive; the record carries a 1989 publication date. Accessed 4 October 2026. https://ora.ox.ac.uk/objects/uuid:cdc17076-98b3-4085-9c49-022041e9bd29
Metzger, Bruce M. The Canon of the New Testament: Its Origin, Development, and Significance. Oxford: Clarendon Press, 1987, pp. 191–201, 305–307.



