Re-inking the
Pen:
Evidence from P. Oxy. 657 (P13) concerning
unintentional
scribal errors
P.M. Head & M. Warren
Published in New
Testament Studies 43 (1997) 466-473.
This article uses SPIonic Greek Font (© James R. Adair, 2000); to download: SP Fonts
Abstract: This article argues that
the constant necessity to re-ink one's pen provided the opportunity for scribal
distraction at the level of eye, memory, judgement and pen, and would thus have been an occasion for the introduction of
unintentional copying errors. This conclusion is suggested by writing practices
in antiquity and confirmed by several striking examples in P. Oxy. 657 (=P13)
where observable re-inking practices correlate with the introduction of new
singular readings.
I. Introduction
New
Testament textual critics customarily discuss scribal errors under two
headings: intentional errors and unintentional, accidental, or involuntary
errors. Intentional errors are those made deliberately in order to improve or
correct the text and involve corrections made to spelling or grammar,
harmonisation to other passages (whether in the immediate context, in a
parallel passage, or in an Old Testament source), and liturgical or theological
enhancement. These might all be traced to the conscious activity of the scribe.[1] Unintentional
errors, on the other hand, can be distinguished (in theory at least) as arising
not from conscious scribal activity but from involuntary mistakes of eye, ear,
memory, judgment, pen or speech.[2] These errors
involve accidental additions, especially dittography (writing twice something
that is only once in the exemplar); omissions, either by haplography (writing
once something that is twice in the exemplar) or through homoioteleuton (a
confusion due to similar endings of words or lines in the exemplar); or other
sorts of confusions of letters or words.[3]
In
considering the causes whereby unintentional or accidental errors were
introduced by scribes, scholars have often attributed them to factors such as
tiredness or distraction, following Jerome who referred to mistakes made by
'copyists more asleep than awake'.[4] As the Alands
wrote, 'It is certainly no secret that a scribe engaged in copying a manuscript
is susceptible to fatigue, especially when copying continuous script.'[5] Kenyon deduced
from the extremely varied quality of scribal workmanship the truism 'that
scribes are capable of anything' and suggested that some mistakes 'are mere
incalculable blunders of a wandering eye, of inattention, or of misplaced
invention.'[6] Our aim in this
paper is to document and support the suggestion that, in addition to the
general issue of scribal weariness, another factor, namely the constant
necessity to re-ink one's pen, provided the occasion for distraction at the
level of, at the very least, eye, memory, judgment and pen. We shall argue that
such a suggestion is clearly plausible in general and that specific examples of
errors introduced in the re-inking process can be detected in P. Oxy. 657, an
early manuscript of Hebrews.
II. Pens and Writing in
Antiquity
Throughout
the Greco-Roman era it seems that scribes writing on papyrus or parchment
characteristically used the hard reed pen (Greek: ka/lamov; Latin: calamus), with the point split into two
equal parts to form a nib and constantly sharpened.[7] Unless one
needed the facility of easy erasure, for example in composition or
school-exercises where a wax tablet and stylus would be used,[8] the reed pen was
widely used.[9] Greenlee asserted
that 'it is safe to assume that virtually all extant papyrus mss., both of the
N.T. and of other documents, were written with a reed pen.'[10] This assumption
is supported by several accounts involving the procurement of writing
materials. P. Oxy. 326, a letter (of cca AD 45) about the procurement of
writing materials, provides a list of the basic requirements, including 'the ink pot and the reeds and the
pen-knife so that he can split the reeds for writing and the rubber'.[11] The accounts of
the record office at Tebtunis (AD 45-57) show numerous purchases of sheets or
packets of papyrus (at an average of 4 drachma); as well as regular purchases
of ink (P. Mich. 123 Verso III.11, 24, 25; V.25; IX.7; X.8), and single
purchases of a ka/lamov (Verso X.8; cf. Rect XII.39) and a sponge
(Verso II.30).[12]
The
main drawback in the use of such a pen was perceived by Quintillian to be the
problem of constant re-inking:
'It
is best to write on wax owing to the facility which it offers for erasure, though
weak sight may make it desirable to employ parchment (membranarum) by preference. The latter, however, although of
assistance to the eye, delays the hand and interrupts the stream of thought
owing to the frequency with which the pen (calamus)
has to be supplied with ink.'[13]
Other factors were also involved, Persius Flaccus
(AD 34-62), in a satirical illustration of the wasted opportunities of youth,
but which must presumably resemble reality, wrote:
We
now take up our book (liber), and the
two-coloured parchment (membrana),
well cleansed of hair; some paper (chartae)
too, and the knotty reed-pen (calamo).
Next we complain that the ink is thick and clots upon the pen; that when water
is poured in, the blackness disappears, and that the tube sprinkles the diluted
stuff in blots upon the paper. ... How can I work with a pen like this?[14]
The
constant need for re-inking is also suggested in two wall paintings from Pompei
which depict various types of writing materials. In one of these a reed pen is
balanced on a hexagonal ink pot next to a half unround scroll.[15] In another
(Naples Museum Inv. 4676) a reed pen is shown resting against a double ink-well
next to a partially written scroll.[16] The close
association in the literature of the reed pen and the necessary ink, as
attested in the phrase "pen and ink" (3 John 13) is significant in
this regard.[17]
The
literary evidence suggests, following Quintillian in particular, that the
re-inking of the reed-pen could easily have served as an opportunity for
distraction from the task at hand. In the next section we shall argue that
several examples of this generally plausible suggestion can be seen in the
re-inking practices of a particular scribe.
III. Evidence for Re-inking
in P. Oxy. 657
The
best scribes clearly took care that the process of re-inking their reed pen did
not adversely effect the presentation of their document. Among less competent
scribes, however, the gradual fading of the letters is sometimes followed by a
dramatically bolder word or two.[18] In the early
period very clear examples of this can be found in P. Oxy. 657, a manuscript
containing portions of Hebrews and dated to around AD 300.[19] Various lines of
evidence suggest that the scribe responsible for this manuscript should be
classified as 'non-professional'. In the first place, an analysis of the
manuscript reveals a distinct lack of discipline in column width and a wide
variation in the number of lines per column, neither of which would
characterise a professional scribe paid by the line. The available evidence is
summarised in the table below (in most cases the column widths have to be
estimated by extrapolation from the extant width).
Column column width (mm. est.) lines of text height (mm)
mz (47 = 1 in GH) 135-140 27 200
mh (48 = 2 in GH) 135-140 26 205
mq (49 = 3 in GH) 145-150 25 195
n (50 = 4 in GH) insufficient evidence 25 192
cb (63 = 5 in GH) 150-155 24 205
cg (64 = 6 in GH) insufficient evidence
cd (65 = 7 in GH) 170-175 23 195
[ce] (66 = 8 in GH) 170-175 23 195
cv (67 = 9 in GH) insufficient evidence 26 200
ch (68 = 10 in GH) 170-175 24 205
[cq] (69 = 11 in GH) insufficient evidence
A further indication of the scribes
'non-professional' behaviour is the marked deterioration in the writing in the
course of the extant London fragment: letter formation becomes increasingly
erratic and the individual letters increase in size, while lines begin to slope
up at the end and the lower lines become longer. The scribe takes some care
with the content of the text, as is clear from the twelve corrections he makes
to single letters or diphthongs by overwriting or superscripting the
correction, but he takes less care with the general appearance of his
manuscript.[20]
A
careful survey suggests that our scribe needed to re-ink his pen at least once
every line and possibly more often (the edges of the manuscript have been
damaged and so beginnings and ends of lines are often either not extant or
rather damaged, hence most of our examples come from the around the middle of
the line). A number of these occur at natural divisions in the text, often
corresponding with punctuation marks of one form or another. The gradual fading
of the lettering is followed by a very dark serious of letters which eventually
tails off into faint lettering again. Such an appearance cannot be attributed
to wear of the manuscript or preservation problems as there is no pattern to
their location on the page. As our earlier discussion suggested the scribe must
have very often had to stop writing in order to dip his pen into the ink and,
having re-inked the pen, resumed writing. Many dozens of clear examples can be
observed in this manuscript.
The
following examples are of particular interest as they suggest a correlation
between the re-inking procedure and the introduction of new variants into the
text. The four examples are found in columns two and three where the re-inking
procedure is most clearly observed, and are each extremely obvious, practically
undeniable, examples of the use of a freshly re-inked pen. In the list below we
note the beginning of the re-inking in bold type:
Example
1. Col 2. line 4 (Heb 3.10b): aei [pla]n[w]ntai en th
kardia autwn dio ouk egnw[san
Example
2. Col 2. line 6 (Heb 3.11-12): ... thn katapausin mou : blepetai
a[del
Example
3. Col 2. line 9 (Heb 3.13): para]kalesate eautouv kaq ekasthn hme[ran
Example
4. Col. 3. line 8 (Heb 4.3-4): kos]mou
genhqentwn eirhken pou peri
thv
These
four passages are of particular interest due to the fact that in these places
evidence of re-inking coincides with singular readings (readings attested in no
other Greek manuscript) in P. Oxy. 657. The presence of unique or very unusual
readings at the same point in the text as a renewed bold text clearly suggests
that, at least for this scribe, the physical process of re-inking his pen
provided the opportunity for inattentiveness when writing resumed. We shall
discuss these in reverse order, as examples 2, 3, and 4 are fairly
straightforward.
Example
4. Col. 3. line 8 (Heb 4.3-4): kos]mou
genhqentwn eirhken pou peri
thv
This variant involves the omission of the connective
ga/r from Heb 4.4 at the very point where the boldness
of the ink indicates the scribe has re-inked his pen (cf. ei1rhken ga/r pou peri\ th=j e9bdo/mhj ou3twj, Kai\
kate/pausen o9 qeo\j e0n th=| h9me/ra| th=| e9bdo/mh| a0po\ pa/ntwn tw=n e1rgwn
au0tou= ). Although a similar omission is found in some
non-Greek witnesses (some vulgate manuscripts and the Syriac Peshitta), the
correlation of re-inking and apparent omission should be noted.
Example
3. Col 2. line 9 (Heb 3.13): para]kalesate eautouv kaq ekasthn hme[ran
This variant offers an even more clear cut example,
one of comparatively few which occur in the middle of a word. In this case there
is also an unusually long space between the eta and the sigma, consonant with
the probability that the scribe has lifted his pen from the page. Here P. Oxy.
657 offers a clear example of a singular reading, para]kalesate (cf. NA27:
a0lla\ parakalei=te e9autou\j kaq 9 e9ka/sthn
h9me/ran)) at the very point of re-inking. The aorist
indicative of P. Oxy. 657 hardly makes sense in the context and loses the
assonance with kalei=tai in the second half of the verse. Once again
we have a notable correlation between re-inking and the introduction of a
singular reading.
Example
2. Col 2. line 6 (Heb 3.11-12): ... thn katapausin mou : blepetai
a[del
This example shows that at a natural place to pause,
as marked by a punctuation sign ( : ); after the conclusion of the citation
from Psalm 95 in verses 7 to 11, our scribe has dipped his pen in the ink-well
and produced a unique reading, blepetai (cf. NA27: Ble/pete, a0delfoi/, mh/pote e1stai e1n tini u9mw=n
kardi/a ponhra\ a0pisti/aj e0n tw=| a0posth=nai a0po\ qeou= zw=ntoj). This is hardly
of great textual significance, as it corresponds to a standard phonological
variation: 'There is a very frequent interchange of ai and
e in
all phonetic environments from the beginning of the Roman period on, indicating
the identification of the short diphthong /ai/ with the simple
vowel /e/.'[21]
Although
this last variant may be a simple phonetic error, independent of the re-inking
process, the correlation between the bold ink and a singular variation in these
three cases supports the suggestion that re-inking one's reed-pen might lead to
confusion or distraction at some level. Our final example is more complicated,
but once again the re-inking practice of our scribe is an important factor
which must be taken into account.
Example
1. Col 2. line 4 (Heb 3.10b): aei [pla]n[w]ntai en th
kardia autwn dio ouk egnw[san
Here P. Oxy. 657 (P13) offers a text differing from
others at a number of points. This can be observed most simply in the following
table:
NA27 P. Oxy 657 (P13)
Kai\ ei]pon, 0Aei\ planw=ntai th=| kardi/a| Kai\ ei]pon,
0Aei\ planw=ntai e0n th=| kardi/a| au0tw=n
Au0toi\ de\ ou0k e1gnwsan
ta\j o9dou/j mou dio/ ou0k e1gnwsan ta\j o9dou/j mou
In his recent commentary H. W. Attridge has argued,
albeit tentatively, for the originality of the text given by P. Oxy 657 here:
'Although P13 may display a simple idiosyncratic corruption, it is likely that
the process of making the text of the psalm conform to the LXX was operative
and that the unusual wording is original.'[22] In support of
Attridge it is undoubtedly true that scribes often conformed their NT text to
readings they knew from the LXX.[23] Nevertheless,
the opposite case, that the readings of P. Oxy 657 result from 'a simple
idiosyncratic corruption', might be supported by the observation that the
relevant line of script in P. Oxy 657 contains two clear cases of re-inking
which relate to the singular readings already noted: in the first, having
writen planwntai our scribe adds en boldly; in the
second case the boldly inked dio stands in contrast to the
otherwise universally attested de.
Consideration
of this problem suggests that even if one took these two cases as involving
accidental errors correlative with the distraction caused by re-inking, the
implied exemplar for P. Oxy 657 would not necessarily have reflected the text
represented by NA27. P. Oxy. 657's en th kardia autwn (cf. NA27: th=| kardi/a| . Au0toi\) is not a simple alteration
but a re-organisation of the thought of the verse, so that the pause for
re-inking corresponds precisely with the end of the clause (for P. Oxy 657
although not for NA27). One might argue that, having made that alteration, de would no longer
make sense and so dio is substituted by the scribe. But such a view would
necessarily attribute the text of P. Oxy 657 to the conscious activity of the
scribe; if the variants reflect deliberate alteration this would strengthen the
argument that conscious assimilation towards the text of the LXX is more likely
that otherwise inexplicable conscious departure from it, and thus strengthen
Attridge's case.
For
the purposes of our argument it is sufficient to note that the presence of
indications of scribal re-inking in P. Oxy 657 must be taken into account in
the evaluation of the readings of this manuscript, that several cases exist
where clear indications of such re-inking correlate with singular readings in
such a way as to suggest that the re-inking process served as an opportunity
for distraction, and that the peculiar character of this particular manuscript
demands, in the well-worn but still fitting words of F.J.A. Hort, that
'knowledge of documents should precede final judgement upon readings'.[24] We are not by
any means proposing that distraction caused by re-inking provides a
revolutionary key to textual criticism by which every slip can be attributed to
the individual scribe. Our brief discussion of example 1 shows that no easy
answer appears available at the moment.
IV. Conclusion
The specific
examples of scribal re-inking of the reed-pen coincident with singular and
non-original readings observed in P. Oxy 657, combine with the mechanics of the
scribal setting and the comments of writers cited above to support the
contention we began with, that the
constant necessity to re-ink one's pen provided the opportunity for scribal
distraction at the level of eye, memory, judgment and pen.
Just a note that after presenting a preliminary version of the findings of this paper to Keith Elliott’s Textual Criticism Seminar in Leeds, Lionel North mentioned a recollection of a discussion between Tregelles and Tischendorf on the subject of pen-dipping. On 26.10.98 he kindly sent me a copy of the page in John Eadie, The English Bible: An External and Critical History of the Various English Translations of Scripture, with remarks on the need of revising the English New Testament (London: Macmillan & Co., 1876; 2 vols.); vol. 2, p. 364. This recollects that Tischendorf excluded John 21.25 on the grounds that in Sinaiticus it appeared to be written with fresher or darker ink. Eadie reports: ‘when he showed the page to Tregelles, the English scholar at once exclaimed, “O yes, I see; the scribe took a new dip of ink after writing verse 24th.”