Although there are some other variant-units in this verse,
let’s focus today on this one: “had
brought forth a son,” or “had brought forth her firstborn son.” With some data derived from Jonathan Clark
Borland’s Textual Commentary on the Greek New Testament blog, we can obtain some hard figures about the quantities
involved in the support for each variant.
These numbers are slightly obsolete but nevertheless they indicate the
proportions involved: τον
υιον αυτης τον πρωτοτοκον (her firstborn son) is supported exactly by 1,446 MSS , and inexactly by 13 MSS; υιον (a son) is supported exactly by 7 MSS , and inexactly by 1 MS.
The seven manuscripts which
support υιον include Sinaiticus (À) and Vaticanus (B). Also listed in UBS 4 is Z (035), that is, Codex Dublinensis, a palimpsest from the mid/late 500’s. According to Swanson, 1, 1582*, 33, and 788 (a member of f13) also
support υιον. Borland describes this
slightly differently, including them all, along with 071vid (400’s
or 500’s, discovered at Oxyrhynchus) and 1192 (a member of f1), but
qualifying Z as Zvid.
The
testimony of 071 merits closer investigation.
This was the first item presented in 1910 in Volume 3 of Grenfell & Hunt’s series on the Oxyrhynchus Papyri and therein we find this acknowledgement: “The vestiges are
indecisive between υιον (ÀBZ, W-H.) and τον υιον αυτης
τον πρωτοτοκον (CDEKLM, T-R.), since with either reading the letters αυ would
come where they appear to do in l. 14, and there is not enough at the beginning
of l. 15 to show whether the word to which ν belongs was abbreviated or not.” Thus 071 cannot legitimately be regarded as a
witness for either reading. (UBS 2 listed 071vid as a
witness for υιον but UBS 4 does not.)
Mt. 1:25 in Codex Bezae (D) |
Willker
provides data about the versional evidence (see variant-unit #10 in his Textual Commentary on the Greek Gospels);
the Old Latin and Palestinian Aramaic are split; the Peshitta and the Vulgate
and the Harklean Syriac favor τον υιον αυτης τον πρωτοτοκον. (The Vulgate reads: “Et non cognoscebat eam donec peperit filium suum primogenitum: et vocavit
nomen ejus Jesum.”) The Nubian version, of which
only scant remains are extant, favors τον υιον αυτης τον πρωτοτοκον. The Armenian and Ethiopic versions also favor
τον υιον αυτης τον πρωτοτοκον, although the Old Georgian supports υιον. The Sahidic and Bohairic versions favor υιον,
the Curetonian Syriac favors υιον, and the Sinaitic Syriac wanders off on its
own with a reading that means “to him a son,” which is an aspect of the
thorough corruption in the Sinaitic Syriac (shared, to an extent, by Codex
Bobbiensis) in Matthean passages pertaining to the relationship between Joseph
and Jesus. The Gothic version is a
non-witness here because Codex Argenteus is non-extant in Matthew 1:1-5:14. The Middle Egyptian manuscript (Schoyen 2650) supports υιον.
Mt. 1:25 in Codex Regius (L) |
The
testimony of the Latin-writing author of Opus
Imperfectum in Matthaeum regarding Matthew 1:25 has been contested. This work, from the early 400s, (it is worth
mentioning that this composition was edited by Erasmus in 1530) quotes Matthew 1:25 as “Et non cognovit eam, donec
peperit filium suum primogenitum” according to Migne’s P.G. vol. 56, col. 635,
on lines 37-38 – supporting the Byzantine reading. However, in the recent edition of Opus Imperfectum in Matthaeum translated
by James A. Kellerman (in the Ancient
Christian Texts series), the quotation of Mt. 1:25 is presented as if it agrees
with the Alexandrian reading. However
the content of what immediately follows indicates that the author read τον υιον
αυτης τον πρωτοτοκον: the author
mentions the view of followers of Eunomius and states that “he calls Christ the
firstborn because we call him firstborn whom other siblings follow.”
Mt. 1:25 in Codex Sangallensis (Delta) |
The
reference in UBS 4 to Didymusdub refers to De
Trinitate, 3:4, where the author (either Didymus, or someone else in Egypt
in the late 300s) states: “It helps us
to understand the terms ‘firstborn’ and ‘only-begotten’ when the Evangelist
states that Mary remained a virgin ‘until she brought forth her first-born
Son;’ for neither did Mary, who is to be honored and praised above all others,
marry anyone else, nor did she ever become the mother of anyone else, but even
after childbirth she remained always and forever an immaculate virgin.” Also,
in the Ancient Christian Commentary on Scripture series, in the volume on Matthew, Chromatius is presented as
quoting Matthew 1:25 in support of τον υιον αυτης τον πρωτοτοκον in his Tractate on Matthew 3:1. In the same volume on the same page,
Chrysostom quotes Matthew 1:25 with υιον.
With the external evidence
described, we now turn to internal considerations.
Metzger expressed the judgment
of the UBS Committee when he dispatched the Byzantine reading in a
single sentence: “The Textus Receptus,
following C D* K W Δ Π most minuscules al,
inserts τόν before υιον and adds αυτης τόν πρωτότοκον (“her firstborn son”)
from Lk 2.7.” If this appraisal is
correct, the words must have been inserted very
early so as to appear in witnesses as diverse as D, W, the Vulgate, the
Peshitta, the Diatessaron, and 087 (from the 500’s). Against this consideration, however, one may
counter that the reading υιον may be a natural conformation to the wording
of Matthew 1:23 (which itself quotes from Isaiah 7:14 ). A charge of harmonization
can be made against the Byzantine reading, to the effect that a copyist reached
into Luke to find the basis for an expansion, but a charge of harmonization can
also be made against the Alexandrian reading, to the effect that a copyist
reached back two verses to find the basis for an abridgment which yielded a
tighter symmetry between the prophecy (in verse 23) and its fulfillment (in
verse 25).
Mt. 1:25b in MS 490 |
One might say, however, “If this
was such a problem, why was the passage in Luke 2:7 left untouched?”. But if we consider data which was unavailable
to Westcott and Hort, we can see in Codex W that Luke 2:7 was not left
altogether untouched: although Codex W
refers to Christ as “her first-born Son” in Matthew 1:25, in Luke 2:7 (where,
as Willker notes, W’s text is predominantly Alexandrian), τόν πρωτότοκον is
absent. This is a fairly clear symptom
of a theological concern. And if it
could happen in part of the early Alexandrian text-stream in Luke 2 (as seen in
one Greek manuscript), it could happen in another part of the early Alexandrian
text-stream in Matthew 1 (as seen in seven manuscripts).
Mt. 1:25 in MS 72 |
Two authors’ use of the same source can give a false impression that one author is dependent upon the other. Is there an identifiable source which employs the phraseτον υιον αυτης τον πρωτοτοκον? No. However,
it does not seem implausible that two authors could independently use the same common
words to make a connection to to Exodus 4:22 – “Then you shall say to Pharaoh, ‘Thus
says the Lord, ‘Israel is My son, My firstborn.’” (LXX: σὺ δὲ ἐρεῖς τῷ Φαραώ· τάδε λέγει Κύριος· υἱὸς πρωτότοκός μου ᾿Ισραήλ.) An explicit identification of Jesus as
“firstborn” is consistent with Matthew’s treatment of Hosea 11:1: in Mt. 2:15, Matthew rejects the Septuagint’s
rendering and follows instead the Hebrew reading of Hosea 11:1 – “Out of Egypt
have I called My Son” – so as to construct a parallel between Israel, the
anointed people, and Jesus, the anointed Person.
Mt. 1:25 in MS 478 |
When one considers
(1) the second-century support for τον υιον αυτης τον πρωτοτοκον from the
Diatessaron,
(2) the wide-ranging patristic support for τον υιον αυτης τον πρωτοτοκον –
from North Africa (Augustine) to Cyprus (Epiphanius) to Egypt (De Trinitate) to Syria (Peshitta) to
Constantinople (Proclus),
(3) the likelihood that early scribes could regard τον υιον αυτης τον πρωτοτοκον
as potentially scandalous, drawing the doctrine of the perpetual virginity of Mary into question,
(4) the evidence from Codex W that τον πρωτοτοκον was considered objectionable
somewhere in the Alexandrian text-stream,
(5) the
close proximity of Mt. 1:21, compared to the relatively distant proximity of
Luke 2:7, rendering the former more likely to be the basis for a harmonization,
(6) the
relative scope of support for the rival readings: τον υιον αυτης τον πρωτοτοκον has the support
of approximately 99.4% of the Greek manuscripts, drawn from members of every
text-type, whereas the epicenter of the shorter reading appears to be in Egypt , and
(7) the
thematic consistency of a description of Jesus as a firstborn son in 1:25,
echoing Exodus 4:22 (where Israel is the subject) in a way similar to the way
in which Matthew 2:15 treats Hosea 11:1,
the
evidence, on balance, favors τον υιον αυτης τον πρωτοτοκον as the original
reading; the Alexandrian reading is a conformation to the wording in Mt. 1:21
and 1:23.