Criteria for Evaluating Variant Readings
Adapted and Expanded from Metzger, pp. 209-210.
- External Evidence
- Types of Witnesses
- Greek Manuscripts
- Versions (esp. Latin, Syriac, and Coptic)
- Scripture Quotations in Early Writers
- the date of the witnesses
- the date of the families or versions
- Geographical Distribution
- this information is not readily available
- a manuscript may not have been written where it
- Genealogical Relationships of texts and families
- Internal Evidence
- Transcriptional Probabilities
- A reading which is hard to understand is more likely
to be original that one easier to understand
UNLESS it is so hard to understand that it makes
no or almost no sense.
- All other things being equal, a shorter reading is
more likely to be original than a longer one
- the ending of the omitted text is the same as
the text just before the omission.
- the beginning of the omitted text is the same
as the text just after the omission.
- the omitted text is the length of a column of
text in a manuscript.
- the omitted text may have seemed to be harsh,
superfluous, or contrary to belief, liturgy,
- A reading which is different from a parallel passage
or Old Testament quotation is more likely to be
original than one which is exactly the same.
- A reading which contains a little used word is more
likely to be original than one with a more
- A reading which is less refined is more likely to be
original than one which fits Attic standards of
grammar and word usage.
- A reading which makes sense but merely implies small
words such as pronouns, conjunctions, and
expletives is more likely to be original than a
smoother text which includes them.
- A reading which contains only one word is more
likely to be original than one which contains two
synonyms (conflation of readings).
- A reading which is unique to a book is more likely
to be original than one which uses the same
language as other places in the book.
- Intrinsic Probabilities
- A reading is more likely to be original if it fits
- the style and vocabulary of the author,
- the immediate context,
- the usage of the author elsewhere,
- grammatical patterns used by the author,
- cyclical or chiastic patterns in the context.
- Note that these probabilities often conflict with #8 above.
The overriding principle is "choose the reading which best
explains the origin of the others" (Metzger 207).
Last Updated March 25, 2002
Page maintained by
Copyright © 2002 Bruce Terry