“Glory to God in the highest, and on earth peace, good will
toward men.” The words of Luke 2:14
have encircled the world as an echo of the angelic chorus proclaimed to the
shepherds on the night of Christ’s birth in
Bethlehem. Few passages of Scripture have had wider
circulation in Western culture, traveling from the Bible to the majestic compositions of
Handel and
Vivaldi,
“Glory to God,
” to Christmas cards, to stained-glass
windows.
Even these clear, plain words, however, have been changed in modern Bible versions. Most of the major new versions do more than change “on earth
peace” into “peace on earth.” There is a
different
meaning in their versions’
rendering of the final phrase. Unlike the
KJV, NKJV, and MEV, which all retain the phrase, “good will toward men,” the
New Living Translation reads, “and peace on earth
to those with whom God is
pleased.” The
New International Version
reads, “and on earth peace
to those on whom his favor rests.” The
Holman Christian Standard Bible reads, “and
peace on earth
to people He favors!”
Thus we have two sets of Bibles which mean two different things in Luke
2:14: in one set, the angels proclaim goodwill
toward people in general, without expressing any particular conditions or
parameters or limitations. In the other
set, the angels proclaim God’s peace specifically, as the
English StandardVersion says, “
among those with whom he is pleased.”
According to a note in the
MacArthur Study Bible, the angelic declaration in Luke 2:14 “is not
to be taken as a universal declaration of peace toward all humanity.” This may seem difficult for some readers to
swallow, especially since the angel declares in verse 10 that his promise of good news is for everybody. But the translators of
the
ESV, NLT, NIV, and HCSB have not taken
it upon themselves to rewrite the angelic song of goodwill toward men. These versions render the last phrase of Luke
2:14 differently because in Luke 2:14,
the
Byzantine Text (and the
Textus
Receptus, the compilation on which the KJV and NKJV are based) disagrees
with the Alexandrian Text, which is the main base-text of the
ESV,
NLT, NIV, and HCSB.
The difference between the Byzantine Text (eudokia, eudokia) and
the Alexandrian text (eudokiaV, eudokias) consists of only one letter, but the difference in
meaning is drastic. Consider the
renderings of Luke 2:14b in The VoiceTM (based on the Alexandrian Text) and the World
English Bible (based on the Byzantine Text):
The VoiceTM: “
And on earth, peace among all people who bring pleasure to
God!”
WEB: “On earth peace, good will toward men.”
One letter in Greek has produced an extra phase in
English.
Now we arrive at the key question: which reading is original: eudokia or eudokiaV? Researchers
have offered different solutions to this question.
In the late 1800s, John Burgon proposed that an early
copyist inadvertently omitted the word
en in the phrase
kai epi ghV eirhnh en anqrwpoiV eudokia,
resulting in
kai epi ghV eirhnh anqrwpoiV eudokia. (A similar error, he pointed out, occurred in
Acts 4:12 in
Codex Bezae, one of the few manuscripts that support
eudokiaV. In addition, according to James Brooks, in
manuscripts of works of the fourth-century writer Gregory of Nyssa, Luke 2:14
is utilized with
eudokiaV and
without
en. According to Wieland Willker’s
Textual Commentary on the Gospels, a few minuscules also omit
en.) From this shortened text, the early Latin
version of the phrase is accounted for, which instead of meaning “And on earth
peace, good will toward men,” means, “And on earth peace to men of good will” –
which does not necessarily mean that divine favor has been dispensed to them,
but can also mean that they are favorably disposed toward God:
in
terra pax hominibus bonae voluntatis.
In an appendix of his 1871 book
The Last Twelve Verses of Mark, Burgon used this variant-unit in
Luke 2:14 to illustrate the importance of patristic testimony. Confronted with the testimony of Codex
Vaticanus, Codex Sinaiticus, Codex Bezae, and Codex Alexandrinus in favor of
eudokiaV,
Burgon cited 17 early patristic writers whose manuscripts of Luke 2:14
supported
eudokia. The supporters of the reading
eudokia include:
● Eusebius (c. 320, in Caesarea),
● Aphrahat (330’s, in Syria),
● Titus of Bostra (c. 350, in Syria),
● Gregory of Nazianzus (c. 360, in central Turkey
and then Constantinople),
● Epiphanius of Salamis
(375, on Cyprus),
● Philo of Carpasia (late 300s, on Cyprus),
● Didymus (380, in Egypt),
●
Apostolic
Constitutions (380),
● Chrysostom (c. 400, in
Antioch
and then
Constantinople),
● Cyril of Alexandria (c. 420, in northern Egypt),
● Theodotus of Ancyra (c. 430, in
central Turkey), and
● Marcus Eremita (435, in Israel).
These witnesses, and others, demonstrate that
ευδοκία was such a widespread
reading – not just in the Middle Ages, but already in the 300s and early 400s – that even with some patristic writers on the other side of the
equation (including Gregory of Nyssa, who is not listed in the UBS apparatus), it is not easy to explain how, if it is not original, it managed to come from
behind, overtake the original text, and leave it far behind. The
evidence from the Armenian version, the Bohairic version, the
Syriac Peshitta version and the Harklean Syriac version add their weight to
this consideration; they support
eudokia.
F. H. A. Scrivener, another text-critical scholar of the
1800s, concurred with Burgon’s conclusion, but argued that
eudokiaV originated by a simple “transcriptional blunder,” which was subsequently meticulously
perpetuated. Scrivener also emphasized
that the internal evidence is against
eudokiaV on the grounds that the
same thought could be expressed in much clearer terms.
Bruce Metzger, in his widely circulated and influential Textual Commentary on the Greek New
Testament, favored the reading in Vaticanus (what a surprise!), and offered
the explanation that a copyist might have considered eudokiaV too difficult, or may
have misread his exemplar – which would be particularly easy to do if, in the
exemplar, the word eudokiaV was
written at the end of a line, with the final letter superscripted instead of at
its normal size. However, Metzger’s imaginative proposal can easily be reversed
if one imagines, instead, an exemplar in which a punctuation-mark was mistaken
for a sigma.
 |
| James Hardy Ropes |
James Hardy Ropes, a minister and Harvard professor in the
early 1900s, favored the reading
eudokia. Ropes usually
advocated Alexandrian readings in his textual research – as he spectacularly
demonstrated in
his commentary on the epistle of James – but he concluded, in
an article which appeared in Harvard Theological Review in 1917, that
eudokiaV originated when an early
copyist read the angelic hymn as a distich, that is, as two poetic lines, the
first of which was, “Glory to God in the highest and on earth.” The copyist, having gotten that far, would
come to
eudokia and
sense a disruption in the poetic balance of the two-line proclamation; thinking
that this was due to an error in his exemplar, he changed
eudokia to
eudokiaV.
Thirty-six years earlier, Hort had approved eudokiaV, the
reading which was in Vaticanus and Sinaiticus when they were made – even though
in both of those fourth-century manuscripts, the final sigma was erased so as to bring their text into conformity to the
usual reading. (It is difficult to tell
at what point these erasures were carried out.
For all we know, it might have been when the manuscripts were proof-read
in the scriptorium, before their pages were bound together.) Hort argued in his 1881 Introduction that the angelic hymn was indeed meant to consist of
two parallel statements: (1) Glory to God in heaven and on earth,
and (2) Peace to men of His favor. Hort accounted for the grammatical harshness
of this rendering as an effect of its nature as a Hebraism.
This line of
reasoning, however, did not satisfy Ropes, who countered that Hort’s proposal
turns the angelic proclamation into an irregular
distich, but “With eudokia, the verse is a tristich, and is easily translatable into
three lines of formal poetry in either Hebrew or Aramaic.” Matthew Black confirmed, regarding the text of
Lk. 2:14 with eudokia,
“When rendered into Aramaic it falls naturally into rhythmic structure, and,
with the earlier words for poimhn, poimnh, gh, preceding ευδοκία, there is an example of paronomasia.” [p. 125, An
Aramaic Approach to the Gospels and Acts.]
Ropes and I disagree with Hort’s conclusion, but Hort did
get one thing right: he stated that if
eudokiaV is
not original, “it must be Western.” The
reason is that
eudokiaV is read by Codex Bezae, and seems to be the basis of the
Old Latin rendering:
hominibus bonae
voluntatis is accounted for by
en anqrwpois eudokiaV more naturally than by
en anqrwpois eudokia (unless, as Burgon
thought,
en was
omitted). (The author of
the NET’s note at Luke 2:14 seems to take for granted that Burgon’s guess is correct,
stating that the Old Latin witnesses “reflect a Greek text which has the
genitive
eudokiaV but
drops the preposition
en.”)
But are there any other reasons why – other than by sheer accident (as Burgon thought),
or due to a misunderstanding of the phrase-division of the sentence (as Ropes
thought) – an early copyist whose exemplar read
eudokia might change it to
eudokiaV?
To
answer this question it maybe helpful to take a closer look at something
written by Origen (a prolific writer in the first half of the
200s). Origen quoted Luke 2:14 with
eudokia in
Against Celsus 1:60, and in Book One of his
Commentary on John (according to the 1989 edition by Ronald E. Heine, page 49) but he is thought
to have used a different form of the text when he wrote
Homily #13 on Luke in the 230’s.
Origen’s
Homily #13 on
Luke, written in the 230’s, is extant only in a Latin translation prepared
by Jerome, and so it is not perfectly clear what text Origen cited. Commenting on the angels’ statement from Luke
2:14, Origen said that a careful reader of Scripture might wonder how, inasmuch
as Jesus said (in Matthew
10:34)
that He did not come to bring peace on earth, the angels could chant “Peace on
earth.” Origen’s solution to this
problem runs as follows (based on
the English translation made by Joseph Lienhard, pages 53-54):
“If Scripture said, ‘Peace on earth,’ and the sentence ended
there, then the objection would be valid.
But something is added. After
‘peace,” Scripture says, ‘among men of good will.’ [Latin:
in hominibus bonae voluntatis] This
answers the objection. The peace on
earth that the Lord does not give is not the peace ‘of good will.’ [Without the
double-negative, that’s “The peace on earth that the Lord gives is the peace of
good will.”] For he did not say simply
[in Mt.
10:34], “I have not come to
spread peace.” He added [still in Mt.
10:34], ‘Upon earth.’”
Hort argued that Origen “manifestly reads
ευδοκίας,
combining it in construction with
eirhnh, not with
anqrwpoiV.” It seems to me
that Origen’s resolution still works with
eudokia, but in the interest of
brevity I won’t argue that point here.
The thing to see is that Origen was concerned that readers might sense an
inconsistency between Mt. 10:34 and Luke 2:14.
And here we may have an explanation for both the creation of
eudokiaV in
the Western text-stream and for its preference by a copyist or copyists in the
ancestry of Vaticanus, Sinaiticus, and Alexandrinus: by specifying that peace is extended to a
particular group of people, an objection about the appearance of
inconsistency between Matthew 10:34 and Luke 2:14 is nullified.
For more than 100 years, textual critics have allowed an apologetically motivated Western reading which contaminated part of the Alexandrian Text’s transmission-stream to stand in the text of Luke 2:14. This should have changed a long time ago. Not only was Ropes’ appeal to internal evidence sufficiently strong, but we have external evidence of which Hort knew nothing. The minuscule 892 (a manuscript of the Gospels from the 800s, with strong Alexandrian affinities – more than any other minuscule) supports
eudokia; had
Hort known this, he might have given further consideration to the idea that
eudokiaV is a
Western attempt to prevent the perception of an inconsistency.
In addition, Ephraem Syrus’ commentary on the
Diatessaron is much better-known than it was in 1881. In Ephraem’s
Commentary on Tatian’s Diatessaron, Ephraem cites Luke 2:14 as
follows: “Glory to God in the highest
and peace on earth” – not to the animals or beasts, but, “Good hope for human
beings.” [p. 67, Carmel McCarthy,
Saint Ephrem’s Commentary on Tatian’s Diatessaron: An English Translation of Chester Beatty Syriac MS 709 with Introduction and Notes, Copyright Oxford Univ. Press 1993] On the premise that Ephraem was quoting from
the same text upon which he was commenting – and this seems a very reasonable
premise – we have here an echo of Tatian’s
Diatessaron, that is, a composition
made around 172.
Besides the
testimony of Tatian and minuscule 892, the Sinaitic Syriac manuscript – discovered at St.
Catherine’s monastery in 1892 – also provides early support for
en anqrwpoiV eudokia. After Agnes Smith Lewis and her sister
Margaret Gibson obtained access to this manuscript (which was produced around
400),
she translated it into English in 1894, and in that translation she
rendered the angelic hymn as, “Glory to God in the highest, and peace upon
earth, and good-will to men.” According
to
Matthew Black, the Sinaitic Syriac has the Syriac word
‘ar‘utha, which is different from the Peshitta’s term,
sabhra tabha; both are paraphrastic but
clearly support
eudokia.
To sum up: ever since
1881, most new English translations of Luke 2:14 have been based on eudokiaV, even though the
external evidence in favor of its rival eudokia has grown stronger and
stronger, and the rationale for ευδοκίας based on internal considerations has been effectively
countered. If the textual critics who
are currently preparing base-texts for translators of the New Testament do not feel
compelled to adopt eudokia –
the reading with more widespread patristic support, the reading with overwhelming
more abundant attestation, the reading that is more stylistically appropriate
to the author, the reading that fits the context better (considering 2:10b), and the reading that is more difficult (from an apologetic
perspective) – then this confirms what many readers have long suspected: despite the text-critically significant
discoveries that have been made in the past 135 years, the pro-Alexandrian compilers
of the Nestle-Aland and United Bible Societies’ texts have not progressed very much beyond Westcott and Hort, and have no intention of doing so.
