Today we
conclude our interview with Dr. Maurice Robinson, one of the compilers of the
Byzantine Textform.
Q: Dr.
Robinson, what would you say to someone who
said, “I want to follow the Majority Text, but I want to follow the majority
text in manuscripts up to the year 900, and set aside the late manuscripts,
most of which are Byzantine”?
Robinson: Should others desire to adopt a different type of Byzantine
preference theory, they certainly are welcome to do so. Clearly, a theory for
the Gospels based on a Byzantine uncial consensus (e.g., that of A E F G H K M
S U V Ω, and parts of W) would approximate closely the results obtained when
the wider minuscule consensus is included.
The problem would be more severe,
however, in the Acts and Epistles, where the available uncial
manuscripts
are fewer, particularly those of Byzantine type that
predate the 9th century. The consensus base for those
New
Testament
books would seem to require inclusion of later minuscule
testimony, else the resultant text of those
New Testament
books will be
less Byzantine than what the similar process might produce among the Gospels. Further, the situation in Revelation would be far worse, since the competing
Byzantine groups there primarily depend upon
minuscules made after the 800’s
(generally representing
the
Majority-Andreas and Majority-Koine forms of the text) — leave these out of consideration, and it would be
difficult to say what the resultant text might become (most likely quite
non-Byzantine in nature).
For the Gospels, a pre-9th century
Byzantine consensus would be as strong and even more viable than portions of
the Hodges-Farstad theory where they appealed to less-than-Byzantine minority
groups to settle instances of textual division; certainly such a method also
would be far more reasonable than adopting a recensional form of text found
only among late manuscripts (such as the Family 35 subgroup). Other possible approaches that would result
in a basically Byzantine form of text could include following the archetype of
Family Π/Ka or (perhaps with less likelihood of success) von Soden’s K1 group.
From my perspective, however, none of these alternatives appear superior to the
present Byzantine-priority hypothesis, methodology, and obtainable results.
Q: Any comments on Nestle-Aland 28?
Robinson: Particularly I am disappointed with one aspect of the new
format, namely the editors’ decision to eliminate mention of fluctuating
degrees of minority Greek manuscript support for variant readings (the “
pc”
and “
al” designations). This move leaves users of the
NA28 apparatus
unclear as to the relative amount of support a given variant reading might
have, and
– even worse
– readers might presume that
only the manuscripts cited for a particular
NA28
reading
actually support such
– and this even though the editors
explicitly claim the new format supposedly should
prevent such.
Although
the
editors claim
that “
pc and
al cannot be used in a precisely defined
way, because full collation of all the manuscripts would yield more witnesses
for known variants,” such special pleading appears peculiar, particularly when
the full collation data of
Text
und Textwert are compared
against readings designated
pc or
al in the former NA27 apparatus.
The
Text
und Textwert data regularly
validate the propriety of the
pc and
al designations within a concise, more
limited apparatus. From my perspective, those designations ought to be
reinstated in future NA editions, along with re-inclusion of at least some of
the previous consistently cited witnesses from NA26/27 that no longer appear in
NA28.
On a positive note, the new
typeface is nice, and the NA28 regularization of some orthographic forms was long overdue.
Similarly, I consider the elimination of conjectural suggestions from the
apparatus beneficial, although an appendix listing the more important of these could be informative in terms of
understanding scholarly views on the matter.
Q: In the approach you describe in “The
Case for Byzantine Priority,” a prohibition on conjectural emendation is
Rule #1. What do you think of NA28’s introduction of a conjectural emendation
into the text of Second Peter 3:10?
Robinson: Given that the
UBS/NA editions long have had a conjecture at
Acts
16:12, the inclusion of a new conjecture at
Second
Peter
3:10 (dating from at least the time of Tischendorf — see his
8th
edition’s
apparatus) is unsurprising, particularly since the basic
Alexandrian reading in that location — found in the main text of previous
critical editions dating back to Tregelles and W-H — simply makes no good sense
(
kai ta en auth erga eureqhsetai, literally “and the works in her shall
be found”). The point is well illustrated in the translational circumlocutions
that appear among those English versions based on the critical text.
Consider the following, grouped according to how they render
eureqhsetai:
Lexham: “and the deeds done on it will be disclosed.”
HCSB: “and the works on it will be disclosed.”
NRSV: “and everything that is done on it will be disclosed.”
NIV: “and everything done in it will be laid bare.”
NET: “and every deed done on it will be laid bare.”
CEB: “and all the works done on it will be exposed.”
ESV: “and the works that are done on it will be exposed.”
GW: “and everything that people have done
on it will be exposed.”
ISV: “and everything done on it will be exposed.”
NCV: “and everything in it will be exposed.”
TEV: “with everything in it will vanish.”
NIrV: “God will judge the earth and everything done in it.”
Mess: “and all its works exposed to the scrutiny of Judgment.”
NLT: “everything on it will be found to deserve judgment.”
Voice: “and all the works done on it will be seen as they truly are.”
NAB: “and everything done on it will be found out.”
Of these,
only the Roman Catholic New American Bible comes close to the base meaning of the problematic construction. So certainly, the NA28 conjectural inclusion of
ouk before
eureqhsetai makes
far better sense without requiring alteration of the proper meaning of the word in the process (thus NA28 in conjecture: “shall [not] be found”). The proposed conjecture, therefore, is
quite good, and similar in quality to what
Rendel Harris suggested for First Peter 3:19, where the main text
en w kai should be supplemented by the conjectural addition of
Enwc (thereby reading, “in which also Enoch”) — a
brilliant conjecture; yet equally without manuscript evidence, and equally recognized by most scholars as
non-original, just as they ought to regard the current NA28 conjecture at Second Peter 3:10.
Put simply, researchers —
particularly those involved in the study and use of actual manuscript testimony — should not invent or prefer readings that have no known existence among the Greek
manuscript base merely because such might “make better sense” than an otherwise
problematic preferred reading (not that I consider eureqhsetai to be original over against the Byzantine katakahsetai,
but obviously the critical text editors do so presume). As I recently commented
on the Evangelical Textual Criticism blog,
The problem I have with conjectural
readings is not restricted to a
priori concerns related to a
Byzantine Priority or majority text
position, but rather as ultimately involving transmissional considerations;
i.e., any conjectured reading — assuming such
supposedly to be more reasonable than what appears among the existing witnesses
— would have to explain transmissionally how and why such would utterly disappear from our
known transmissional history. Were such conjectures actually superior to all extant alternatives, I would
consider their lack of perpetuation to be inexplicable.
Q:
Could
you briefly explain how the NA28 has many conjectural emendations if one
considers short series of variant-units instead of just single variant-units?
Robinson: I have already written extensively on the so-called
“zero-support”
verses in the Nestle-Aland editions, both in a published
essay in Translating
the New Testament: Text, Translation, Theology and in a subsequent
ETS presentation in 2012.
To summarize: if individual variant support in NA27 is considered in a
linear manner (i.e., the stated documentary support for an entire verse
containing at least two variant units, when reduced to its combined joint
agreement), at least 105 whole verses exist in the critical text that
apparently lack any actual existence in the form published, whether from any
Greek manuscript, ancient version, or patristic source. In those instances, the
result obviously becomes de facto conjecture.
In my follow-up paper, I examined two-verse
segments in NA27 using the same criteria, and found an additional 210
similar portions of text that again as published lack attestation
from any known witness (suffice it to say that among
the manuscripts comprising the Byzantine Textform, such never occurs
in relation to passages of similar length). Note that the same findings apply
to the NA28 edition as well, since its main text and apparatus support
basically remain the same.
Q:
Could you explain again why
Revelation is so different in the Hodges-Farstad compilation?
Robinson: Although more
differences appear between H-F and RP in Revelation than elsewhere in the
New
Testament, they are not that extensive as when
either text is compared against the Old Uncial form of the critical editions.
Rather, the primary differences between the Byzantine form of text in H-F and
RP mainly involve H-F utilizing a particular stemmatic approach that prefers a
minority Byzantine subgroup that they considered original – a group that at
times represents less than 30% of the Byzantine
manuscripts of that book. In contrast,
RP2005 presents a non-stemmatic model
representing a general consensus among
the two primary Byzantine groups
within that book.
Where these
two groups divide, RP generally follow the
Majority-Koine group except where
a significant number of its manuscripts align with the Majority-Andreas group – this because the Majority-Andreas group appears
to reflect a single archetype derived from the Andreas commentary that usually
accompanies those particular manuscripts.
Robinson: What Dallas critic might have suggested such (I speak as a fool)? I
would prefer to say that many have a preference for a text similar to the
Byzantine who might claim to be majority or Byzantine supporters, but who
speedily dissent from such whenever the Byzantine Textform departs from their
favored Textus Receptus/KJV
type of reading.
Beyond Pierpont and myself, among
those who are not TR/KJV partisans but who favor some form of the Byzantine
text (not necessarily agreeing with our specific theory or methodology nor
resultant form of the text) would include Hodges and Farstad, John Wenham,
Jakob van Bruggen, Peter Johnston, Harry Sturz, Wilbur Pickering, Paul
Anderson, Thomas Edgar, James Davis, Donald Brake, Timothy Friberg (not all
still living) and others, including several more Europeans along with many of
my own students. Not all of
these have published in relation to textual matters, and thus some names may be
unfamiliar; yet in general they remain pro-Byzantine to some degree. There also
are numerous laypeople that have communicated with me or these others over the
years who hold to some sort of Byzantine or majority text position, but I only
mention here a few who have published within academia.
Q: Finally: in NA28, in
Second Peter 2:18, the editors rejected oligws and adopted ontws, even though the adoption
of oligws had previously been given an “A”
ranking (as if the editors were certain that it was correct). Any idea how that
happened?
Robinson: The answer
apparently is the “wag the dog” influence of CBGM and little else; this
particularly in view of Metzger’s previous strong defense of oligws in his Textual Commentary. I also note
that in the process UBS5 lowered the rating from “A” to “C” – again without
providing any particular reason or justification for such. Nice that they here adopted a
Byzantine reading, but clearly not for the same reason that I would do so.
Thank you for sharing your
thoughts.
— end of interview —
