Re: tc-list text criticism of Mark
- On Mon, 12 Jan 1998, "Vinton A. Dearing" <dearing@...> wrote:
>The November issue of Literary and Linguistic Computing (vol. 12, no.I can't say anything about the articles cited, since I haven't seen
>4) has two items of interest.
> 1. "Measuring Genre Differences in Mark with Correspondence
>Analysis," by David Mealand. This seems mostly an exellent
>example of statistical analysis--all the data are given and the
>results are fully diagrammed--but I wonder whether it is enough to
>decide that a sample size which has "worked" for others is
>satisfactory for one's own work. Also, the statistical methods are
>simply named, never explained.
> 2. A review of Richard J. Finneran, ed., The Literary Text in the
>Digital Age, University of Michigan Press, 1996.
> If any listers or lurkers have explored the question of sample size
>I should be interested to know their opinions.
them and don't recall ever seeing the journal.
But I have noticed a terrible tendency among TC professionals to use
inadequate sample sizes. (E.g. the Alands' "Thousand Readings" use
only three or four readings in some of the shorter Pauline Epistles.)
The sample size one needs depends on the nature of the data one
is trying to extract. But the Central Limit Theorem -- one of
the basic rules of statistics -- only applies when one uses "large"
numbers, which in practice means a minimum of thirty to fifty
meaningful points of data.
What this means for TCers is that if they want to determine the
affinities of two mixed manuscripts, they need to find at least
30 to 50 readings for which both manuscripts are unmixed (non-Byzantine).
Even for overall affinities, they need 30-50 readings. This
can render many results (e.g. the Alands' for the shorter books,
but also some methods such as the Claremont Profile Method)
invalid if they don't use enough text.
I hope I will be able to offer more information on this subject
at a future time.
Robert B. Waltz
Want more loudmouthed opinions about textual criticism?
Try my web page: http://www.skypoint.com/~waltzmn
(A site inspired by the Encyclopedia of NT Textual Criticism)