Internal Evidence

Mussies: Linguistics
and Methodology

Excerpt for review from: G Mussies,
The morphology of Koine Greek, as used in the Apoc. of St.John,
(Brill, 1971)

Page Index

Last Updated: Feb 23, 2010

Review: - T.C. and Linguistics: G. Mussies
    MS Analysis - evaluating MSS individually to avoid circularity
    MS Selection - genealogy and mixture
    Linguistic Phenomena - in the transmission stream

Return to Index

Mussies on
Linguistics and Method


G. Mussies brings concise clarity to the basic questions of methodology in the use of linguistic information for evaluating variant readings. He especially highlights the danger of circularity in selecting readings as original or 'ancient', before evaluating the age and nature of sources of linguistic data, such as manuscripts on an individual basis.

Excerpted for Review from: G Mussies,
The morphology of Koine Greek, as used in the Apoc. of St.John,
(Brill, 1971), ch 2

Headings have been added for clarity and navigation purposes.


2.1 Introduction

The Complexity of and Sources for Linguistic Data

Anyone who wants to describe the (use of) language of a literary work or other document that has been handed down in several manuscripts, meets with the problem of textual criticism. Often that text edition which is known to be the best is chosen to serve as the basis for linguistic description. If the number of manuscripts is not too large the variant readings can be included in description, but in the case of the Apocalypse [Revelation] this will be absolutely impossible: it is simply impracticable to make a description that accounts for some 230 manuscripts at the same time. Therefore a choice cannot be avoided.

Moreover, there is a second problem. The science of textual criticism itself used more and more linguistic data in its method of determining which are the oldest variant readings. These data, however, cannot and may not be furnished by descriptions of the language as found in eclectic editions.1

Avoiding Circularity in Method

If we want to avoid circular reasoning we shall have to start from the use of language of the separate manuscripts [abbreviated: MSS] themselves, these being the only facts at our disposal. 2 Studies of this kind have been undertaken by P.W. Hoogterp and J.D. Yoder respectively with regard to the Codex Bobiensis 3 and the Codex Bezae 4, but their works covered several N.T. writings at the time.

In such cases, we think, one would have to go still further in order to achieve methodological consistency and keep the N.T. writings apart in the description, not only because each author has stylistic peculiarities of his own, but also because the works of one and the same man have not been copied together right from the beginning.

They circulated separately and had as it were a textual life of their own. That is why one and the same codex may be trustworthy with regard to, say the Gospel of St. Luke, but not as far as Acts is concerned.

Even when the books had been combined into larger collections they were often corrected each from a different manuscript [singular: MS], sometimes the model was even changed during the collation of one and the same book. 5

Classifying Individual MSS and Scribes

Only when linguistic descriptions have been made of the MS or MSS that are considered to be the more important witnesses (for instance on the ground of the highest number of lectiones difficiliores which they contain) one can proceed to make a guess at the use of language in the original. The result may be that many variant readings are to be discarded on the ground of this reconstruction, not only in the inferior MSS, but also in those serving as the basis of our description. It stands to reason that absolute certainty can never be attained.

Conclusion. Instead of describing the use of language of a text edition which is inevitably eclectid the linguist will have to choose the best MS available and describe in the first place the idiom of that one alone.

2.2. Choice of the Manuscript

...As is well known, the oldest manuscript is not necessarily the most trustworthy. The number of intermediate copies between a 10th cent. cursive MS and the autograph need not be larger than between the latter and a 4th century Uncial. 6 ...each copy increases the number of variants, omissions, improvements, errors, etc. is one of the difficult tasks of T.C. to shed as much light as possible on their interrelationships. However, the MS reality is much more complex than we may have suggested by our comparison of lineal descent, precisely because "lineal descent" is rare: [a scribe] copied and corrected at the same time, using several models [i.e., exemplars, master-copies] simultaneously.

Such a conflation of texts arose also when marginal notes were here and there inserted by a copyist because he thought them better than the wording of the text itself. But the most frequent cause of conflation was undoubtedly the copying of [exemplars] that had already been corrected by their successive possessors.

2.3. Manuscripts and Language

A second question that may be raised is, whether the older MSS are closer to the autograph than the younger MSS, as far as the language is concerned. A priori it may seem that the answer will be positive, as younger MSS are likely to contain more modern linguistic phenomena. Nevertheless some reserve is in place since a twofold evolution can be discerned in Greek MSS dating from the Christian era. 7

2.3.1 Phonology and Morphology.

First there is the gradual and very slow process of modernizing, perceptable for instance from the increasing number of phonetic spellings, especially during the first period of text transmission. 8

Occasionally, there are also morphological innovations, for instance at Rev.21:17 in cursive MS 2050 (c.1107 A.D.) the numerals (cf. Mod.Gk ) and (Mod.Gk ). This process follows the changes of spoken language and will mostly have worked unintentionally, through slips of the pen, and incidentally.

Second, there is the much stronger archaizing influence of Atticism. It was carried through on purpose but never consistently, and moreover, by different copyists to different degrees. From observation we learn that the older MSS have both types of aorist side by side, and that the one in -α tended to be replaced in the younger MSS 025 and 046 by the Attistic -ον etc. This is quite contrary to the evolution of the spoken language: Modern Greek has only imperfects and aorists ending in -α.

That Atticisms worked incidentally in the MS tradition and did not expel all the newer forms at once, appears from the mixture of new and old found.

Original Footnotes:

1. Especially not when the frequency of of current words like the articles, και,the personal pronouns,etc. has to be determined for a specific book. Their presence or absence in a given passage is very often the point of difference between the manuscripts.

2. Cf. A. Thumb, Die griechische Sprache im Zeitalter des Hellenismus, Strasbourg, 1901 p.12.

3. Etude sur le Latin du Codex Bobiensis (k) des Evangiles, Wageningen, 1930.

4. The Language of the Greek Variants of Cod.Bezae Canta. Princeton, 1958; cf. Novum Testamentum 3, 1959, p. 241-248.

5. H.J. Vogels, Handbuch der neutestamentlichen Textkritik, Bonn, 1955, 2nd ed. p. 20-21. L. Vaganay, Initiation a la critique textuelle neotestamentaire, Lyon, 1934, p.60.

6. Vogels, pl 63, 153, Vaganay, p.25.

7. cf. W. Michaelis, Einleitung in das Neue Testament, Berne, 1946, p. 381 and F. Blass-A. Debrunner, Grammatik des neutestatmentlichen Griechisch, Gott. 1959 10th ed. par 3.

8. cf. W. Michaelis, Der Attizismus und das Neue Testament, Z.N.W. 22, 1923, p.121.

Return to Index

Return to Index