Return to Index

The Alexandrian Text Type

The Question of Alexandrian Text Types

"In the course of this inquiry I was driven to suspect that the apparent consensus on the existence of a late Alexandrian text is far from resting on a solid documentary [manuscript] basis. This consensus is more the result of a [recent historical] process, which I shall try to describe briefly, and which starts from the beginning of modern textual criticism.

I. History of an Idea

John Mill, Richard Bentley (1707-1727) and Vaticanus

In the Prolegomena to his great edition of the Greek NT (1707) John Mill had stated that Codex B had been 'in Occidentalium gratiam a Latino scriba exaratum', "written by a scribe for the Western world". He did not think therefore that it was important to collate this MS. ¹ And although, 20 years later, Richard Bentley made some effort to get a full collation of this MS, ² his cards were not made available to J.J. Wettstein for his edition of 1751.

J.J. Wettstein (1751)

Wettstein did not complain about that. He would have liked to know the readings of B not because he thought that they could have been of any help to him for textual decisions, but in order to prove that this MS had no authority whatsoever: 'sed ut vel hoc constaret, Codicem nullius esse auctoritatis'. ³

Return to Index

J.S. Selmer (1767) and the 'Alexandrian Recension'

The first author to mention and define clearly an Alexandrian text of the Gospels was probably J. S. Selmer. In his Apparatus ad liberalem N.T. interpretationem of 1767 he speaks of an Eastern (Antioch and Constantinople), a Western and an Alexandrian Recension of the NT text. The name 'Alexandrian Recension' for him designates the manuscripts whose readings are also to be found in Fathers of the Egyptian Church. This text then was for Selmer the local text of Egypt. ⁴

J.J. Griesbach (1777-1796)

Ten years later J.J. Griesbach, following the path opened by Selmer, produced a list of 9 MSS which were to be assigned to the Alexandrian recension: C L K 1 13* 33 69 106 118. ⁵ We may wonder today why MS B (Vaticanus 1209) was not in this list. In fact B was at that time still under suspicion of having been heavily interpolated by the Latin tradition.

It was therefore something new in the history of textual criticism when Griesbach, in the 2nd edition of his Greek NT (1796) added to the MSS quoted above also Codex Vaticanus (B) as witness to the Alexandrian recension, at least for Mark, Luke and John. He still thought that in the first part of Matthew, B had to remain assigned to the Western 'recension'.

J. L. Hug (1808-1847)

At that beginning of the [19th] century the problem of the Alexandrian recension was subjected again to scrutiny by J.L. Hug. He suggested that the 'Alexandrian recension' was to be dated about the middle of the 3rd century (A.D.), and that it represented the purification of a wild text, which was similar to the text of Codex Cantabrigiensis (D). Out of a similar text the Alexandrian was formed thorugh the removal of interpolations, refinements of grammar and so on. The result was the text of B C L, which was also the text of Athanasius and of Cyril of Alexandria ⁶ It seem thus that in the [19th] century the notion of an Alexandrian text and its presence in MSS like B C L had found a kind of consensus among the critics. ⁷

Return to Index

Westcott & Hort (1881)

This situation remaind almost unchanged until the publication of the Introduction of Westcott and Hort in 1881. ⁸ What was new was the view that was brought in by their inquiry about the Alexandrian text? I shall try to summarize it in these items:

(a) The separation between an Alexandrian and a Neutral text

(b) a careful description of documentary and internal characteristics of the Alexandrian text and of the method for its detection in NT MSS.

(c) an assessment of the secondary nature of the distinctive Alexandrian readings [as against the Neutral].

Return to Index

Hort's 'Neutral text' Concept

On the first point little needs to be said, because the conception of a Neutral text, as distinct from an Alexandrian text type, is widely known as a typical Hortian conception. ⁹ It may be however of some importance to insist on a point that has not won, it seems to me, due recognition. 'Neutral' in Hort's view does not designate another text type on the same level as the others.

'Neutral' is not the denomination for another local text. This text is not related, as such, to certain MSS. Its first characteristics are not positive, but negative. ¹⁰ The term 'neutral' applies to those variants which cannot be characterized either as Syrian or as Alexandrian or as Western, and which can, at least theoretically, be found everywhere, in the text of many different local churches. It is only by an historical chance, according to Hort, that de facto a majority of neutral readings is found mainly in old Egyptian MSS. But we should remember that the only MSS surviving from the 1st centuries are in fact all Egyptian.

Hort would certainly have objected to the division, which has now become common, of the Alexandrian text into proto- and later Alexandrian, where by 'proto-Alexandrian' the 'Neutral' text is meant. For him the Alexandrian text, as distinct from the Neutral, had its origin very early in Egypt, perhaps already in the 2nd century. It was not due to a recension by one hand, but to a trend towards a better Greek language in the transcription of MSS. The distinctively Alexandrian variants are the 'work of careful and leisurely hands, and not seldom display a delicate philological tact.' ¹¹

As for the method of detection of these variants, Hort was well aware of the fact that all extant MSS are mixed MSS. ¹² Therefore the Alexandrian text cannot be found as such in any of the existing documents. Instead of trying to detect which are the MSS that are nearer to a pure Alexandrian text (as his predecessors did, and as is common also today through different methods of quantitative analysis) Hort introduced the method of the observation and comparison of contrasted groupings in successive variations.

This means that, given a group of witnesses that are liable to have preserved Alexandrian readings (like ALeph, C L X 33, Coptic version ,Alexandrian Fathers), one should carefully observe how they combine together against other distinctive (Syrian and Western) readings in successive variations. It is then not enough to have the support of all of these witnesses for a variant, or even of a majority of them, in order to define it as a distinctive Alexandrian reading. Two other conditions are necessary: contrast to other (distinctive) readings, and some sign of internal secondary character.

This method requires the use of both the external and internal evidence, and it may be successful, according to Hort himself, only when the textual history of a given variant does not present too many complications. ¹³

Return to Index

'Text-type' as Abstract Collection of Readings

THis is not the place to judge the validity or the practical utility of this method. What I would like to emphasize here is that in this way a new concept of a text-type is brought forward, which presents the following characteristics:

(a) A distinctive text-type is primarily not a group of MSS, but a set of readings.

(b) This set is limited:; it does not cover all the readings of the NT. This view is entirely different from the silent presupposition which seems to be common today in textual treatments of the NT.

(c) This list of readings cannot be established only by external attestation. Internal examination is required. A statistical approach is not sufficient.

(d) Readings of different text-types do not overlap, except when a certain genealogical relation is proved. For instance, Hort strictly maintained that a reading can never be Western and Alexandrian together, or Neutral and Alexandrian. There is mutual exclusion among text-types. ¹⁴

Later Developments

In the earlier view of Griesbach which considered only local texts, a view which has been subsequently revived in Germany by von Soden and in England by B.H. Streeter and is common today, a reading can be assigned to different text-types.

This happens when we find that a variant was current, for instance, in Alexandria and also in the Western Church. The geographical diffusion of a variant is taken today as an important factor in determining its originality. Of course, Westcott and Hort thought the same, but they carefully tried to avoid confusion between geographical consideration and the assessment of text-types. They maintained that a distinctive variant cannot be at the same time Alexandrian and Western, because this would entail the independant rising of the same corruption in different places ¹⁵ For them text-types are always to be considered not as simple historical entities, but as variations from the original texts.

Return to Index

The Secondary Character of Alexandrian Readings

The last point worth mentioning in this brief recollection of the contribution of Westcott and Hort is their assessment of the secondary character of the Alexandrian readings. The Alexandrian Text is, like the Western, an aberrant text.

It tries to smooth difficult passages.

"The changes made have usually more to do with language than matter, and are marked by an effort after correctness of phrase." [i.e. Atticisms] ¹⁶

Wherever we can find a distinctive Alexandrian reading, it can be contrasted with a variant which is less elegant, more rude, less 'correct': it is the Neutral reading.

From all this it follows that ther is no place, in the view of Westcott/Hort for a 'late Alexandrian text', as a distinctive type of text. THere is however in their system a possible acceptance of the denomination 'late Alexandrian', provided that it receives a very clear delimitation, which should sound like this:

...an eclectic text of the 3rd or 4th century, which would have accepted, together with distinctively Alexandrian readings, also a share of Western readings which had obtained a wide popularity in Egypt since the most ancient times. ¹⁷

Hort would have envisioned in a way like this the work of Hesychius, which could then be considered as a late development of the text in Alexandria.

It is clear how in this description of a late Alexandrian text the term 'text' is taken in a looser way than was the case of the distinctively Alexandrian. The work of Hesychius would thus have been, according to this suggestion, the putting together of variants derived from other existing texts and so repeating their readings, only in a new combination: a step towards the Byzantine Text, if not already a first form of it.

Return to Index

Newer Viewpoints

Let us now consider how subsequent theories of the history of NT text have dealt with the Gospel text of Alexandria.

Common to almost all of them is the rejection of the idea of a Neutral text, and the transformation of the readings which Hort had assigned to it into readings of the local text of Alexandria.

This was done in many different ways. W. Bousset endeavoured to prove that the characteristic text of Aleph/B is the fruit of the Hesychian Recension. ¹⁸

Independantly, von Soden tried to recover the three main forms of text existing in the Church at the beginning of the 4th century. One of them was the Hesychian Form, represented chiefly by Aleph/B etc. ¹⁹

In Britain B.H. Streeter insisted that the reconstruction of the text of local churches and the comparision between these texts was the only possible way for textual criticism. One of the chief local texts was that of Alexandria, represented by Aleph/B L etc. ²⁰

It can be said in general that in the first decades of our century textual critics practically identified the 'Neutral' and Alexandrian text. They took for granted the existence of a revised text of the Alexandrian Church, due to a recensional activity which could be dated between the 3rd and 4th century, which was present in a dozen well-known MSS, especially in Vaticanus (B) and Sinaiticus (Aleph).

Return to Index

The Discovery of the Papyri

A new period was inaugurated by the discovery of the Chester Beatty and Bodmer papyri. At first they were studied in the light of earlier classifications of text-types. But eventually it became clear that they were imposing a reconsideration of textual history.

As for the Alexandrian text, it is now more and more recognized that the text of B, together with the text of Bodmer Pap XIV-XV (P⁷⁵), cannot be considered as a part of an Alexandrian recension, to be dated towards the beginning of the 4th century.

They represent a much more ancient text which is, in all probability, an unrevised text. ²¹ What then is to be derived from all this for the so-called Alexandrian text?

The solution which is now commonly proposed, as we said at the beginning of this article, is to divide it into two parts: a proto-Alexandrian text, which could be dated in the 2nd century A.D., and which is preserved especially in P⁷⁵/B; ²² and a late Alexandrian text, which represents a later stage of the same text.

But, as we can see from the historical consideration, if there are good reasons to postulate an old Alexandrian form of the text of the Gospels, the notion of a late Alexandrian form rests on a very limited basis of inquiry.

II

The work which should be done to answer the question whether the notion of a late Alexandrian text, as contrasted with the early Alexandrian, corresponds to an historical reality and is to be retained at all in our view of the history of the NT text, is a major one.

...

---

Original Footnotes:

1. J. Mill, H KAINH ΔΙΑΘΗΚΗ, (Oxonii, 1707) p. clxiii. He states expressyly about the Cod. Vaticanus: 'nec dignum esse, cuius variationes multa cum difficultate comparandae essent'.

2. In the year 1720 a ferst collation of Cod. Vaticanus was prepared for R. Bentley by A. Mico, scriptor of the Vatican Library. A 2nd collation was made after 1726 by Rulotta and sent to Bentley.

3. J.J. Wettstein, Novum Testamentum Graecum, Tomus I (Amstel., 1751), p.24.

4. J.S. Selmer, Apparatus ad liberalem NT interp. (Halle, 1767), pp. 45 ff.

5. J.J. Griesbach, Novum Testamentum Graecum, 2nd ed. (Halae, 1796), prlegomena, p. lxxxi.

6. J. L. Hug, Eins. in die Schriften des NT 1st ed. 1808. I followed here 4th ed. vol. 1(Stutt., 1847), 168 ff.

7. At least among those who accepted the existence of recensions. In fact there were critics, among them C. von Tischendorf, S.P. Tregelles, who did not care too much about classification of MSS. They preferred to consider each MS in itself and to work in the field of collation to record exactly its readings.

8. Westcott/Hort, The NT in its Original Greek, Introduction (1881, 2nd ed. 1896)

9. See for instance B. Metzger, The Text of the NT, (1968), pp. 133-4

10. This corresponds to the 1st principle of the two great critics, which is proposed at the beginning of their Introduction:

'The office of TC, it cannot be too clearly understood at the outset, is always secondary and always negative' (Intro p.1; emphasis mine)

It would be interesting to study how much this principle affects the whole of the conception of W/H.

11. W/H, Introduction, p. 131.

12. THis is especially true for W/H, of the Alexandrian text:

'Since it has so happened that every MS containing an approximately unmixed Alexandrian text has perished, the Alexandrian readings can have no strictly primary attestation among extant documents, and are therefore known only through documents containing large other elements.' (Intro, p. 166)

13. Introduction, pp. 130-2, 166-7.

14. For instance, see Introduction, pp. 208-9.

15. Introduction, pp. 209

16. Introduction, pp. 131

17. Introduction, pp. 208-9

18. W. Bousset, Textkrititsche Studien zum NT (Text und Unter., 11) (Leip. 1894), m. Die Rec.des Hesychius, pp. 74-110

19. H. von Soden Die Schriften des NT I. (Berlin 1902) II. (1907)

20. B. Streeter, The Four Gospels (Lond., 1924)

21. See my study Il problema della recens. cod. B alla..Bod. XIV (Roma, 1966), and G.D. Fee P75/76 and Origen ...(Grand.Rap. 1974) pp.19-45

22. See above, p. 285,n. 3