Sinaiticus

If the testimony of Codex Sinaiticus is to be considered accurately, then
three things about its testimony must be carefully considered: (1) The four
pages upon which Mark ends and Luke begins were not produced by the
copyist who made the surrounding pages; they are replacement-pages. (2)
On these replacement-pages, the rate of letters per column varies in an
extraordinary way. (3) A decorative design after Mark 16:8 indicates that
its maker was aware of verses 9-20.

Why were the four pages produced by the main copyist removed and
replaced? By analyzing the rate of letters per column in the replacement-
pages that contain Mark 14:54-16:8 and Luke 1:1-56 (consisting of a single
parchment-sheet, folded in half, with four columns on each page), we can
provisionally answer that question.

The copyist who produced the surrounding pages wrote about 635 letters per
column of text, averaging slightly more than 13 letters per line. In the four
replacement-pages at the end of Mark and the beginning of Luke, though,
the rate of letters per column varies significantly: these 16 columns contain
the following amounts of letters: in Mark, columns 1-10 contain 635, 650,
639, 707, 592, 593, 604, 605, 552, and 37 letters. In Luke, columns 11-16
contain 681, 672, 702, 687, 725, and 679 letters. The low number in column
10 is accounted for because that is where the text of Mark ends. The
variation in the rate of letters per column in the remaining columns is
striking.

The rate of letters-per-column in the first three columns is not appreciably
different from the main copyist's rate earlier in Mark 14. But in the fourth
column, the rate drastically increases to 707. In columns 5-7 it drops
consistently lower than the main copyist's rate, by about 30 letters. Then in
column 9, the rate of letters-per-column drops drastically to 552. In column
10 there are only 37 letters, because here the Gospel of Mark ends. The
remaining six columns average 691 letters per column, remarkably higher
than normal. The explanation of this phenomenon may help us discern what
elicited the replacement of the pages that had been produced by the main
copyist.

What apparently happened is that the main copyist accidentally omitted part
of the text of Luke, perhaps carelessly skipping from the Greek word
"EGENETO" at the beginning of 1:5 to the same word at the beginning of
1:8, thus leaving out 319 letters. If the main copyist had written Luke 1:1-56
in these six columns, but left out 319 letters, he would have written 3,827
letters in six columns, thus averaging 337 letters per column, pretty close to
his usual rate.

The proof-reader, discovering the main copyist's mistake in the text of the
first chapter of Luke, decided that it would be better to replace all four pages
than to attempt to add the missing words in the margin. So after removing
the pages with the error, he made new pages to replace them. Probably the
proof-reader realized that the text of Luke 1:56a at the end of the sixteenth
column would need to match up exactly with the text of 1:56b on the following
page; realizing also that failure to accomplish this would require another
attempt on another parchment-sheet, he probably lessened his risk of wasted
effort by beginning to write on the replacement-pages at the beginning of
Luke, in column 11.

After successfully rewriting Luke 1:1-56, he began to rewrite the text of
Mark at the normal rate of about 635 letters per column in columns 1-3.
Then he compressed his handwriting in column four, with the result that this
column contains 707 letters. Upon reaching 15:19 in column 5, he stopped
compressing his lettering and began stretching it, filling the column with less
than 600 letters. He continued to stretch his lettering in columns 6, 7, and 8,
clearly attempting to fill space so that the text of Mark would extend into
column 10. However, he accidentally skipped 76 letters in Mark 15:47-16:1,
and 12 letters in 16:6. So in column 9, he had to drastically stretch his
lettering, filling the column with only 552 letters, in order to have some text
to place in column 10. (He also wrote the name "Jesus" in full in 16:6
instead of using its normal contracted form, to take up space.)

All this has some interesting implications.

First, we can deduce that the four pages initially made by the main copyist
did not contain Mark 16:9-20. The first ten columns of text on the
replacement-pages (containing Mark 14:54-16:8) contain 5,614 letters. Add
to that the 88 letters that the proof-reader skipped, subtract the extra four
letters that he added in the name "Jesus" in 16:6, and we arrive at a total of
5,698 letters. Add the 971 letters in Mark 16:9-20, and we reach a total of
6,669 letters. So if the first ten columns in the replaced pages included
Mark 16:9-20, they would have had to average 667 letters per column,
significantly higher than the main copyist's normal rate of 635.

Second, after deducing that the text of Mark 14:54-16:8 in the replaced
pages contained 5,698 letters, we can get a good idea of what these columns
looked like in the replaced pages: written at a rate of 635 letters per
column, the text of Mark 14:54-16:8 would end near the bottom of column 9
and column 10 would be blank.

So there are at least two feasible reconstructions of the contents of Mark

in the replaced pages:

(1) Writing at his normal rate of 635 letters per column, the main copyist
ended the Gospel of Mark at the end of 16:8 in column 9, left column

10 blank, and began Luke at the top of column 11.

(2) Writing at slightly less than his normal rate, the main copyist ended the
Gospel of Mark in column 10.

(A third possibility also exists: that the main copyist

wrote at his normal rate and included the Shorter Ending

after 16:8, with the result that the text extended into

column 10. However, this is unlikely, because Codex

Sinaiticus was almost certainly made at Caesarea in the

mid-300's, and Eusebius of Caesarea, who described how

his copies of Mark ended around 330, did not mention

any copies with the Shorter Ending.)

Now let's consider the decorative lines that appear in

Codex Sinaiticus' replacement-page after Mark 16:8.

These decorative lines are similar to this copyist's usual

decorative lines (which appear in Codex Sinaiticus at the

ends of the Apocrypha-books of Judith and Tobit, and at

the end of First Thessalonians), but they are uniquely

emphatic here after Mark 16:8. The significance of this

feature was noticed by George Salmon in the 1800's;

Salmon stated that although the decorative line's basic

design occurs three or four times in Codex Vaticanus,

"The prolongation of the arabesque has no parallel in

either MS. We see that the scribe who recopied the leaf

betrays that he had his mind full of the thought that the

Gospel must be made to end with EFOBOUNTO GAR [the

last words of Mark 16:8 in Greek], and took pains that

no one should add more." Another scholar, John Gwynn,

wrote in 1883 that it is not correct to assert that Codex

Sinaiticus betrays no sign of consciousness of the existence

of verses 9-20, "For the last line of ver. 8, containing only

the letters TOGAR, has the rest of the space (more than

half the width of the column) filled up with a minute and

elaborate "arabesque" executed with the pen in ink and

vermilion, nothing like which occurs anywhere else in the

whole MS. (O.T. or N.T.)." The idea is that this uniquely

emphatic decorative line was intended as a sort of fence

to prevent any addition, thus implying an awareness of

such an addition on the part of the person who made the

decorative line.

AN ADDITIONAL QUESTION: DID EUSEBIUS

OF CAESAREA SUPERVISE THE PRODUCTION

OF CODEX SINAITICUS AND CODEX VATICANUS?

Eusebius of Caesarea, a prominent bishop and historian

(d. 340), mentioned in his Life of Constantine that Emperor

Constantine instructed him to produce 50 Bibles to be used

in the new capital, Constantinople. Some researchers

have suggested that Vaticanus and/or Sinaiticus may be

two of those 50 copies. However, while Vaticanus and

Sinaiticus were certainly stored together at Caesarea at

some point in their history (when a series of liturgical marks

were added to them both in the first half of the book of Acts),

it is unlikely that Eusebius was directly involved in

the production of either copy. Eusebius was the

creator of the Eusebian Canons, a cross-reference

system for the Gospels, in which sections of text

were numbered and the section-numbers were

arranged in ten groups in a chart prefacing the

Gospels. Each group had a different kind of

parallel, from the first group, containing material

found in all four Gospels, to the tenth group,

containing material unique to a single Gospel.

Codex Vaticanus and Codex Sinaiticus both are

missing some verses to which Eusebius assigned

section-numbers. In addition, in Codex Sinaiticus,

the Eusebian section-numbers, which were added

while the manuscript was still in production, are

written with many mistakes. It is unlikely that

Eusebius would oversee such an erroneous

presentation of his own cross-reference system.

It is more likely that Codex Sinaiticus was

produced under the supervision of Eusebius'

successor, Acacius the One-eyed, who, according

to Jerome, oversaw the production of parchment

copies to replace papyri in the library at Caesarea

that were wearing out.