Family 1 and Family 13

A Professor Who Died Before He Finished

W. H. Ferrar, a fellow of Trinity College in Dublin, noticed that four gospel manuscripts kept agreeing with one another against everything else, and set out to collate them and prove they came from a common ancestor. He died in 1871 before he could publish. His colleague T. K. Abbott finished the work and brought it out in 1877 under both their names, and the title page says what happened plainly: the collation is "by the late William Hugh Ferrar", edited by Abbott, who wrote that the credit for "the detailed and laborious comparison necessary in order to prove that they actually represent a single archetype, is due to the late Prof. Ferrar". The group has been called the Ferrar Group ever since.

The second cluster was established by K. Lake in 1902, in a Cambridge monograph on Codex 1 and its allies. Lake demonstrated that a set of minuscules headed by manuscript 1 descended from a lost archetype of their own, and while he was at it he invented the notation everybody still uses. His own explanation of why is disarming: "It is very cumbrous to have to write 13 — 69 — 124 — 346 — 543 — 788 — 826 — 828 every time that one wishes to quote the consensus of all the Ferrar mss." So he wrote fam¹³ for Ferrar's group and fam¹ for his own, and those two numerals have named the two families ever since.

The members have kept being added to. Lake's Family 1 was a handful; A. Anderson collated thirteen for Matthew; A. Welsby seventeen for the Gospel of John; the collation of that gospel published by D. Parker for the Birmingham institute in 2024 lists twenty-one. For Family 13, Lake's eight have become ten in the current lists, with a lectionary sometimes added. These are not fixed clubs, and a reader who meets fam¹ in an apparatus is reading a shorthand for whatever the editor of that edition thinks belongs in it.

What a Family Is For

A family is a claim about descent. If a dozen late manuscripts agree in a peculiar reading, either they all inherited it or they all invented it, and inheritance is very much likelier. So the family stands in for a book that no longer exists, and the value of a fourteenth-century copy is not its own age but the age of the ancestor it lets you reconstruct. Family 1's archetype is generally reckoned to be considerably older than any surviving member.

The clearest single piece of evidence for that is a note rather than a reading. In Codex 1, the account of the woman caught in adultery is not printed at John 7:53 where a reader expects it; it is added at the end of the gospel, after a note saying that the passage does not appear in many manuscripts and was not commented on by John Chrysostom, Cyril of Alexandria or Theodore of Mopsuestia. Welsby's study observes that the note stands in Codex 1 and in Codex 1582 "without any significant variation", that manuscript 565 has an abridged version, and that "as this note is not known to exist in any other manuscripts it provides compelling evidence of an extremely close link between 1, 565 and 1582". Somebody in the ancestry of these books had checked, and wrote down what he found.

Codex 1582 has since displaced Codex 1 at the head of the family, and it is worth knowing why: it is dated, by a colophon on folio 287r, to the year 948, and that colophon names the monk Ephraim as the scribe. Where the book was made is a separate question from where it sits now, and the two are easily run together, because a shelfmark records where a manuscript is and never where it was written. Codex 1582 is Vatopedi 949 on Mount Athos today; A. Welsby's summary of the scholarship is that "most scholars agree, based on 1582's quality and fine script, that it was a product of Constantinople", and that "little is known of 1582's previous owners before it came to the Vatopedi monastery". The colophon survives only in a later hand's transcription of it, but the date it gives is generally accepted and the script agrees with it. Lake thought Codex 1 the best witness to the archetype; A. Anderson, after collating both in full, "nominated 1582 as the leading Family 1 manuscript" and drew attention to inaccuracies in Lake's edition, and the Birmingham collation of the Gospel of John now states flatly that "1582 is the most reliable witness to the text of the archetype". Lake's central finding stands and his ranking does not, which is a normal and healthy thing for a hundred-and-twenty-year-old conclusion to have done.

The Rule Everybody Repeats About Family 13

Family 13's famous feature is that its manuscripts move the woman caught in adultery out of the Gospel of John altogether and put it after Luke 21:38. C. R. Gregory stated it in 1907 in the form still quoted — though what he was describing was one manuscript and not a family. Writing of minuscule 13 itself, immediately before noting that Ferrar had seen its likeness to 69, 124 and 346, he says that "the interpolation about the adulteress is not at John 7:53 to 8:11, but is placed directly after Luke 21:38." The sentence has been generalised to the whole group ever since. It is a genuinely striking thing to find in a manuscript, and it is used everywhere as the test of whether a manuscript belongs to the family.

It does not work as a test, and a specialist who has studied the family at book length says so in as many words. J. Perrin's study of the family in the Gospel of John states its purpose against exactly this practice: "Every Family 13 study to date has assumed that the location (or relocation) of the Pericope Adulterae from John 7:53-8:12 was a secure rubric for designating Family 13 constituency. This study rejects the validity of this criterion." His evidence is concrete in both directions. Manuscript 1689 is a member of the family on every other measure and has the passage in its ordinary place at John 7:53. Manuscripts 1141 and 2900 put the passage after John 21:25 and are not members. His own summary is that the placement after Luke 21:38 is usual in the family but that the rule is unreliable. An article that stated the Luke 21:38 fact flatly would be repeating a shibboleth a specialist has published against.

One larger claim about both families should be handled with the same care. For most of the twentieth century these manuscripts were classed together with Codex Koridethi and some of Origen's quotations as witnesses to a "Caesarean" text — a text-type supposedly used at Caesarea in Palestine. That hypothesis has been under attack for decades. D. Lafleur, revisiting Lake's work in 2014, writes that "since then, the question of the «Caesarean» text-type has been a very disputed matter. It still remains an important textual issue", and points out that Lake himself hedged in 1900 — the Caesarean origin "cannot yet be regarded as certain" — and then removed the sentence from later editions. It is worth being honest about the limits of this article's own reading: no published statement was found declaring the Caesarean hypothesis dead, and the two doctoral studies of these families cited below mention the term only in their bibliographies. That silence is suggestive and it is not a citation. What can be said is that a family is a demonstrable thing, a text-type is a much larger and shakier one, and the second should not be smuggled in on the strength of the first.

Scripture references

• John 7:53 — Where the account of the woman caught in adultery usually stands, and where Family 13 does not put it
• Luke 21:38 — Where most Family 13 manuscripts place the passage instead
• John 21:25 — Where two manuscripts that are not family members place it
• John 8:11 — The end of the disputed passage

Sources

• Primary textW. H. Ferrar and T. K. Abbott, A Collation of Four Important Manuscripts of the Gospels — William Hugh Ferrar and Thomas Kingsmill Abbott · Hodges, Foster, and Figgis, Dublin; Macmillan, London; 1877. Public domain, read at the Internet Archive. Source of the title page's own description of Ferrar as "the late", of Abbott's statement of what Ferrar had done, and of Abbott's judgement that the position of the passage marks the independence of the group. A second scan of the same book on the Archive is OCR'd unusably; this is the readable oneSupports: Ferrar's death before publication, Abbott's role, and the original argument
• Primary textK. Lake, "Codex 1 of the Gospels and its Allies", Texts and Studies VII.3 — Kirsopp Lake · Cambridge University Press, 1902, printed as the third part of Texts and Studies volume 7, which is the item digitised — there is no standalone scan. Public domain, read at the Internet Archive. Source of Lake's own explanation of the fam¹ and fam¹³ notation, quoted verbatim including his list of the eight Ferrar manuscripts, and of his account of the note at the end of John in Codex 1. It is also the source of his own membership list, which runs to four manuscripts: his second chapter is headed "The relations subsisting between 1 118 131 209", and page xxii sets 205 aside in as many words — "I was convinced when I studied the question at Venice that 205 was a copy of 209... It is for this reason that no further notice has been taken of 205"Supports: The origin of the family notation and Lake's 1902 findings
• Holding institutionITSEE, University of Birmingham: ECM John, Collation of Family 1 — D. C. Parker, Institute for Textual Scholarship and Electronic Editing, University of Birmingham · January 2024, built on A. Welsby's doctoral work. The current membership list of twenty-one witnesses is taken from here, as is the statement that "1582 is the most reliable witness to the text of the archetype"Supports: The current membership of Family 1 and the standing of Codex 1582
• ScholarshipJ. Perrin, "Family 13 in Saint John's Gospel" (doctoral thesis, University of Birmingham, 2013) — Jac Perrin, University of Birmingham · Open access, 497 pages. Source of the quoted rejection of the Luke 21:38 placement as a membership criterion, of manuscript 1689 as a member with the passage in situ, and of 1141 and 2900 as non-members that relocate it after John 21:25Supports: That the famous Family 13 test does not work as a test
• ScholarshipA. Welsby, "A Textual Study of Family 1 in the Gospel of John" (doctoral thesis, University of Birmingham, 2012) — Alison Welsby, University of Birmingham · Open access, 336 pages; the basis of her De Gruyter volume, which is not. Source of the account of the note about the woman caught in adultery in 1, 565 and 1582 and of the quoted conclusion drawn from it, of what Lake established in 1902, and of A. Anderson's case for 1582 over Codex 1. Section 2.1.4, pages 19 and 21, carries the colophon of 948 naming the monk Ephraim, the note that it survives in a later transcription, and both quoted sentences on provenance. Worth keeping straight while reading it: Vatopedi 949 is a shelfmark and answers only where the book is kept — where it was made is Welsby's argument from its script and quality, not the monastery's recordSupports: The scribal note, the link between three manuscripts, and the reordering of the family's leading witness
• ScholarshipD. Lafleur, "The «Caesarean» Text of the Gospel of Mark: Lake Revisited" — Didier Lafleur, Institut de recherche et d'histoire des textes · BABELAO 3 (2014), 145-169 with an appendix to 289; open access. Source of the quoted statement that the Caesarean question remains disputed, of the observation that Lake removed his own hedged sentence from later editions, and of Lafleur's dates for LakeSupports: That the Caesarean text-type is contested rather than settled
• Primary textC. R. Gregory, Canon and Text of the New Testament — Caspar René Gregory · 1907; public domain, read at the Internet Archive. Source of the quoted statement about the placement after Luke 21:38 and of the account of Ferrar's and Abbott's work. That sentence stands on page 372 inside Gregory's description of the single manuscript 13, so it is a statement about one book rather than the family-wide rule it is usually quoted as. Note for anyone following the trail: this book contains no description of minuscule 33 and no mention of the "queen of the cursives" epithet, which are in his Textkritik of 1900 insteadSupports: The Luke 21:38 placement as the standard description, and that Gregory said it of manuscript 13
• EncyclopediaWikipedia: Family 13

Further reading

• W. H. Ferrar and T. K. Abbott, A Collation of Four Important Manuscripts of the Gospels (1877)
• K. Lake, Codex 1 of the Gospels and its Allies, in Texts and Studies VII (1902)
• ITSEE, University of Birmingham: the ECM collation of Family 1 in John

Reflect

A scribe copying John stopped, wrote a note saying that a famous story was missing from many of his sources, and copied it anyway at the end where you could see the join. What would it cost you to show your working like that?