p. 138
The Byzantine text is attested by about 95 % of the total number of surviving manuscripts of the Greek New Testament.33 In the Gospels, depending on what kind of variants we include, the differences between the main form of the Byzantine text and the text of the modern critical editions affect about 7 % of the wording.34 During the past 200 years there has sometimes been a tendency to assume, as if it were an incontrovertible truth, that most Greek manuscripts of the Byzantine period contain an inferior text of the New Testament. According to that point of view, the form of text which was prevalent throughout the Greek-speaking churches for more than 1,000 years was a relatively corrupted text, which modern scholarship has successfully corrected and replaced by making use of recently discovered manuscript evidence.
If that approach to the New Testament text is correct, it requires the conclusion that Erasmus’ dependence upon Byzantine manuscripts caused him to produce an unreliable text, now of little practical use and of interest mainly to historians. However, it is necessary to question the assumptions on which that point of view is based. Here we will just briefly consider, in outline, a few limitations affecting the evidence and criteria which are central to current text-critical theory.
The “Lateness” of Erasmus’ Byzantine Manuscripts
The natural expectation is that the earlier manuscripts of the Greek New Testament would be more reliable, because they would theoretically reflect fewer stages of copying by scribes, and hence would provide fewer opportunities for those scribes to introduce errors into the text. Erasmus did not formulate the principle in these terms, but it is implied by his emphasis on the superior authority conferred by antiquity. His esteem for ancient sources also led him to make exaggerated claims about the age of the manuscripts which he used.
Yet the comparatively late date of Erasmus’ manuscripts is of less significance than is commonly stated. His Byzantine manuscripts, mainly dating from the 11th century onwards, reproduced a form of text which had a much older history behind it, stretching back at least as far as the 4th century, and possibly earlier.
33 Calculations of the number of “Byzantine” manuscripts can produce differing totals, depending on what criteria are used for excluding manuscripts which have some degree of divergence from the norm.
34 Editions adopting a mainly Alexandrian form of text include those of K. Lachmann (1842-50), S.P. Tregelles (1857-72), C. Tischendorf (1869-72), B.F. Westcott / F.J.A. Hort (1881), Eberhard and Erwin Nestle (1901 and later editions), K. Aland et al. (known as the Nestle-Aland edition; latest revision 2012), and the United Bible Societies (latest revision 2014).
p. 139
For the most part, therefore, Erasmus was transmitting an ancient text. If he had made use of earlier Byzantine manuscripts the resulting text might not have been much different. Many of the errors in his Greek text could have been avoided by making comparisons with a slightly larger number of manuscripts, rather than by using manuscripts that happened to be earlier in date.
One of the chief arguments advanced in favour of modern critical editions
of the Greek New Testament is that they are based on “earlier” manuscripts,
drawing upon evidence from the 2nd, 3rd and 4th centuries. In general, the crit-
ical texts agree more frequently with the earliest available uncials and papyri,
the early Church Fathers, and ancient versions of the New Testament in other
languages. By contrast, there is not yet any direct evidence that the distinctive
pattern of readings displayed by the Byzantine text was in existence during the
2nd or 3rd centuries. It nevertheless seems highly probable that all or nearly all
of the individual Byzantine readings did already exist at that early date, as more
and more of such readings have found support in recently-discovered papyri
from that period. The possibility cannot be excluded, that the main form of
the Byzantine text was derived from earlier manuscripts which have not been
preserved.
It is widely accepted that the surviving early manuscript copies of the Greek
New Testament probably constitute only a small fraction of all the copies that
were formerly in existence, and that most of the early copies perished by repeated
use or through accidental or deliberate destruction. There is hence no reason to
think that the few surviving early copies provide a balanced representation of all
forms of text that used to exist in the earliest period. Nor is there any basis for
confidence that the earliest extant manuscripts supply the most accurate attaina-
ble wording of the original text. The existence of numerous discrepancies among
the early papyri, and between the earliest manuscripts and the scripture citations
by patristic writers, demonstrates that textual variation, scribal errors and de-
liberate alterations had already become widespread during the first centuries of
manuscript copying. Under these circumstances, there is an increased possibility
that later manuscripts could preserve a superior form of text, drawn from earlier
ancestors that are no longer extant. A comparable phenomenon is observed in
the manuscript transmission of other ancient texts, where textual critics accept
that a later manuscript or group of later manuscripts can in principle preserve
genuine readings that are not attested by the earliest surviving witnesses.
Because of such considerations, it would be unwise to assume that the mere
antiquity of the Alexandrian manuscripts guarantees the genuineness of the
readings which they contain, even when supported by ancient patristic or ver-
sional evidence. In the same way, the comparative “lateness” of manuscripts
which contain the Byzantine text (or which were used by Erasmus) does not
justify the conclusion that the Byzantine readings are individually or collectively
false.
p. 140 ndrew ]. Brown
The Criterion of the “Shorter Reading”
Another fundamental concept which underlies current textual theory is that the
intrinsic superiority of the text of the most ancient manuscripts can be demon-
strated by the application of commonly accepted procedures and criteria. These
criteria or guiding principles, which form part of the armoury of every textual
critic, can be very useful when considering the merits of the competing textual
variants at individual passages. Unfortunately, in practice, the application of
these criteria often leads to contradictory results.
A case in point is the criterion of lectio breuior potior: that the shorter reading
is to be preferred. This principle arises from the observation that scribes might
add words and phrases to the text, perhaps to provide an explanation of a difficult
word, or to emphasize a particular doctrine, or to produce a more elegant turn
of phrase, or accidentally importing into the text a comment which had previ-
ously stood in the margin. On the other hand, recent analysis of scribal habits
has shown that certain types of longer variant are more likely to be genuine, by
demonstrating that errors by scribes have a greater tendency to result in omis-
sions rather than additions of words.35 With regard to deliberate scribal changes,
while it is undoubtedly true that some scribes made pious or explanatory expan-
sions to the text, it can equally be argued that other scribes chose to omit words
and phrases which did not conform with their own theological views, or which
they considered to be superfluous.
/----------------—------r---------
The possibility of scribal additions to the text was well known to Erasmus,
who regularly refers to them in his Annotations. There are also instances where
he expresses doubts over some of the longer variants attested by his Greek manu-
scripts. Usually he adopts the longer Byzantine form of wording for his printed
text. Besides this, there are places where his adoption of a longer variant is the
result of his retranslating from the Latin Vulgate into Greek, at passages where
his Greek manuscripts attest to a shorter wording. In some of these passages, the
additional words seem to have originated as insertions by the medieval copyists
of the Latin text.
When the Byzantine and Alexandrian texts are compared, it can be seen that
the former contains a greater number of passages that have a longer form of
wording. At each of these passages the same two questions can be asked: is the
longer wording an erroneous addition, or is the shorter wording an erroneous
omission? There has been a tendency to insist that the longer forms of wording
are typical embellishments by later scribes, and to assume that the later date of
the Byzantine manuscripts proves that this is the case. On this point, however,
35 Cf., for example, J. R. Royse, Scribal Habits in Early Greek New Testament Papyri (Leiden
2008); A. Wilson, ‘Scribal Habits in Greek New Testament Manuscripts’, Filologia Neotestamen-
taria 24 (2011), 95-126.
p. 141
there is a danger of circular reasoning, by assuming that the shorter readings of
the Alexandrian text are superior because of the earlier date of the manuscripts
which contain them, while simultaneously claiming that the same manuscripts
are intrinsically superior because of the shorter readings which they contain. In
fact there are no grounds for certainty that the Alexandrian text had an earlier
origin or that its shorter wording is correct.
The Criterion of the “Harder Reading”
Another important textual criterion is the principle of lectio difficilior potior: that
preference should be given to the “harder reading”. The idea behind this is that a
scribe is more likely to change a word or phrase which is harder to understand,
or less elegantly expressed, or more problematic in its theological implications.
Conversely, a scribe is less likely to introduce a more difficult form of wording
deliberately. At individual passages this concept is useful to the textual critic,
helping to diagnose the cause of a scribal alteration. While Erasmus does not use
quite the same terms to describe this textual criterion, he clearly made use of the
underlying principle. He recognized, for example, the possible genuineness of
textual variants which might at first sight appear “absurd” to a scribe.36
It is noticeable that the ancient manuscript witnesses of the Alexandrian text,
generally followed by the modern critical editions, have a higher frequency of
“harder readings” than the manuscripts which transmit the Byzantine text. This
gave rise to the theory that the Byzantine text, not just at isolated passages but
as a constantly recurring feature, was characterized by a tendency to modify,
36 In epistolam ad Corinthios priorem annotationes (Basel 1519), on 1 Cor. 15,51: “quoties
ueteres fatentur lectionem esse diuersam, semper mihi suspectior esse solet ea quae prima spe-
cie uidetur absurdior, ut consentaneum sit, lectorem uel parum eruditum, uel parum attentum,
offensum absurditatis imagine, mutasse scripturam” (ASD VI-8,310,11. 774-777). He expresses
himself a little obscurely here but the sense can be paraphrased as: “Whenever the ancients ac-
knowledge that there is a different reading, the one which seems quite ‘absurd’ at first sight [but
which can be shown to make good sense when examined more closely] always tends to arouse
my suspicion that a reader, who was either poorly educated or insufficiently attentive, and who
was offended by the imagined ‘absurdity’, probably altered the scripture text [by substituting
a variant in which the apparent ‘absurdity’ did not occur].” Erasmus is not speaking of every
kind of “harder reading” at this point, but appears to refer solely to passages where the early
fathers (ueteres) acknowledge the existence of divergent readings, and where one of the variants
has a meaning which could initially seem “quite absurd” or “more absurd” (absurdior). In the
context of his statement, absurdior refers to a form of wording which, at first sight, could appear
to be contrary to orthodox doctrine or inconsistent with what is taught by other passages of
scripture, and which therefore does not make good sense to a superficial reader. J.H. Bentley,
‘Erasmus, Jean Le Clerc, and the Principle of the Harder Reading’, Renaissance Quarterly 31
(1978), 309-321, at 318 fi, relying upon a rare connotation of suspectior, makes the ingenious
suggestion that this word should mean “more esteemed”. However, “more suspect” or “causing
more suspicion” is the usual sense.
142
Andrew J. Brown
simplify and smooth out any difficulties that were present in the original word-
ing.37 Although this theory has been frequently repeated in the literature of
textual criticism, it is necessary to mention that here again there is a risk of
circular reasoning, in assuming that the harder readings of the Alexandrian text
are superior because of the earlier date of the manuscripts which contain them,
while simultaneously thinking that those manuscripts are intrinsically superior
because of the harder readings which they contain.
In essence, the principle of lectio difficilior is based on a paradox. The difficulty
of a particular phrase, that might cause an ordinary reader to suspect that there
is a mistake in his copy, is the very feature that the textual critic uses as evidence
that the phrase is genuine. Because of its paradoxical aspect, this textual crite-
rion needs to be used with caution, and its usefulness in assessing the overall
character of a manuscript or text-type can be questioned. There are many kinds
of accidental scribal errors which are likely to disrupt rather than to improve
the theological, contextual or grammatical cohesion of a passage. By their na-
ture, errors of that kind produce a “more difficult” reading.38 The presence of a
greater quantity of “difficult” readings does not by itself establish the superiority
of the manuscript or text-type which contains them, but might merely reflect an
accumulation of ancient errors within that particular branch of the manuscript
tradition. Like the other commonly-used textual criteria, the concept of lectio
difficilior potior fails to supply proof that the Byzantine manuscripts contain an
intrinsically inferior text.
Taking such considerations into account, there is little justification for con-
demning Erasmus for his use of Byzantine manuscripts. Although the “standard”
modern editions of the Greek New Testament provide many valid corrections of
Erasmus’ text, some of the textual theories and assumptions underlying those
editions have not been conclusively established. It remains possible that the text
which Erasmus produced 500 years ago is, at numerous passages, better than
some of the more recent editions which have attempted to replace it.
37 See especially the comments by F. J. A. Hort, in id. / B. F. Westcott (eds.), The New Testament
in the Original Greeks 2 vols. (Cambridge 1881), vol. 2, part 1 (= Introduction), 115f., 134 f.
38 Cf. Wilson, ‘Scribal Habits (op. cit. n. 35), 105-126.
