Re: Copying parameters for early NT MSS

Finney@csuvax1.murdoch.edu.au asked several questions evidently
intended to help create a probabilistic model for NT MS propagation.
I don't think hard answers are available for them, mainly for two
reasons, (1) they depend on knowledge of lost MSS, and (2) the
answers just DEPEND.  Some thoughts/suggestions/ruminations
with varying degrees of reliability follow.

> How often were NT MSS copied (per year)?
This depends on the MS, how old and venerable it was, how good a copy
it seemed to be, whether it was situated in an important church, etc.
Because there are so very few extant MSS known to be copied from another
extant MS, it has been suggested (S Lake?) that exemplars were often
destroyed after they were copied once.  On the other hand, MSS like Pi
seemed to have been highly popular as exemplars, on account of the number
of known MS relatives and the number of correctors writing on the original.

> What was the expected life of a MS (years)?
A well used copy could wear out in a few decades, an accident like
fire could happen any time, while a MS lost in the desert might last millenia.
Results like "the probability of surviving 1000 years given that the MS
already survived 500 years" might be estimated by looking at Aland's
"Text of the NT" and counting extant MSS from each century, adjusting
in some heuristic fashion for the MSS catalogued in modern times but now lost.

> What was the probability that a MS would be destroyed (per year)?
> What was the probability that a MS would be discarded (per year)?
These probably could be given Gamma distributions, but the parameters
would depend on knowing about lost MSS.

> What was the probability of an error (per word)?
This depends strongly on the scribe and on the copying circumstances.
Professional scribes in later centuries could attain quite high accuracy,
better than OCR software I would say, while others could be abysmal,
either through inattention, not knowing Greek, copying from dictation,
feeling free to improve the text, or just having a bad day (or a bad decade).
The earliest MSS seem to vary the most in quality.
Critical studies generally examine singular or sub-singular readings of a MS,
which readings can be assumed to be innovative errors,
such errors are characterized and counted, and conclusions about the
copyist's methods and reliability are drawn.
Colwell did this for three papyri (p45, p66, p75?), Kubo for p75 (I think),
Hurtado for W in Mark, others for Sinaiticus, Vaticanus, Bezae,
and for precious few, if any, other MSS.
Unfortunately, apparatuses, even the IGNTP Luke, omit most singular readings,
so you have to go back to the MSS and collations themselves to count.

> What was the probability of a correction (per word)?
This depends strongly on how well the first hand copyist performed,
also on whether the corrector used the same exemplar as the first hand used.
Wisse said that correctors tended to be very diligent at the beginning
of a book, but slack off a lot later in the book.  Many old MSS had
several distinguishable correctors over the centuries.
A rough range of correction rates might be measured from the IGNTP Luke
apparatus for the cases of Aleph, W, and L, say, by just counting samples.

> What was the probability that a number of MSS would be used as exemplars?
It seemed to be very common for the text of the tangible exemplar to be
mixed with the text(s) existing in the mind of the copyist.  Also,
the text of the corrector's exemplar would get mixed in with the text
of the first hand's exemplar when the MS became an exemplar itself for copying.
Block mixture of texts occurred when exemplars were fragmentary
or different exemplars were used on different days in a scriptorium, say.
Rough statistics on block mixture might be obtained from Wisse's profiles
of MSS of Luke for the IGNTP.

> How many copies of the NT MSS were made altogether?
A lot more than survived to the present.
Wild guesses might be made from population estimates
and ranges like 1-5/church/century.
Changes in the distribution of Greek speakers during the early centuries
make things difficult, but you might get solid numbers from the late middle
ages by comparing populations with counts of extant MSS.

Vincent Broman,  code 572 Bayside                        Phone: +1 619 553 1641
Naval Command Control and Ocean Surveillance Center, RDT&E Div.
San Diego, CA  92152-6147,  USA                          Email: broman@nosc.mil