Last Updated:

Oct 21, 2010

Fee: the Majority Text

Excerpt for Review: G. Fee, Studies in the Theory and Method of NT TC, "The Majority Text..." (a rewrite of Fee, 1978-9), p.183 fwd, (Eerdmans, 1993)

Page Index

The Majority Text: - G.D. Fee

Hodges' Support of the Majority Text:
    The Byzantine Text: Hodges' arguments

Fee's Refutation of Hodges:
    1. Transmission Theory - rejected by Fee
    2. Transmission Data - facts and conclusions
    3. Maj. Text Dominance - accounting for it

More Articles:

Other - blurb

Hodge's Argument for Majority Text

Quantity, Uniformity, Dominance

"The seminal arguments in this modern advocacy of the Majority Text were offered by Zane Hodges... Besides the obvious theological drive behind his presentation 1 ...his basic arguments are three...:

(1) the inability of scholarship to explain the rise, the comparative uniformity, and the dominance of the Byzantine text;

(2) that the Majority text is the result of "normal" transmission (= uninterrupted by deliberate operations) of the NT text, because "the copies nearest the autograph will have the largest number of descendants" (1970:21); ... I respond only to the 1st two arguments, which for Hodges are basically two parts of one argument, based on both the quantity and general uniformity of the extant MSS whiic support the Byzantine text-type.

On the one hand, he argues that:

... modern textual criticism, while "denying to the Majority text any claim to represent the actual form of the original nevertheless unable to explain its rise, its comparative uniformity, and its dominance in any satisfactory manner" (1970:18).

On the other hand, he argues that:

...this uniformity and dominance "can be rationally accounted for, .... if the Majority text represents simply the continous transmission of the original textfrom the very first" (1970:18)

He is faced with the problem, however, that this text-form is completely unknown by any of the evidence up to 350 A.D., the earliest evidence being in some 4th century Church Fathers, then later in the 5th century in portions of Codices W, A. 2

He counteracts this problem in a two-fold way:

(1) by arguing that "all of our most ancient manuscripts derive basically from Egypt," thus suggesting that they represent "a local text of Egypt" (1970:12-13) and are "merely divergent offshoots of the broad stream of transmission [= the Majority text] whose source is the autographs themselves" (1970:18); 3 and,

(2) by affirming as a "truism" that "the manuscript tradition of an ancient book will, under any but the most exceptional conditions, multiply in a reasonably regular fashion with the result that the copies nearest the autograph will normally have the largest number of descendants" (1970:27).

Original Footnotes:

1. ...The theology in this case, is an urgency about Scripture, that the providence of God would not have allowed Protestant Christianity for so much of its existence to have been wrong about the text of the Bible, and that modern textual criticism stems from Westcott & Hort, whose orthodoxy is suspect on other grounds.

2. It must be emphasized that even though quite a few Byzantine readings existed earlier than this, the text-type itself did not exist. The question here is not a matter of readings, but of these readings all existing together in collocation in the same piece(s) of evidence. The failure to recognize this crucial point is the Achilles' heel of Sturz's study as well; it does no good to argue that some Byzantine readings can be shown to have existed early. What must be shown is that they existed together in this form that early.

3. Hodges never really addresses the question. How is it that only "offshoots" have been found in the first 300 years, if the Majority text represents the "broad stream" that issues from the autographs? Why are there no MSS even partly representing the Majority text until the 5th century and no full-scale representatives until the 8th?

Earlier answers to this, which apparently Hodges endorses, were given by Burgon (1883:319), and by Hills (1870:42), to the effect that "they were read so constantly and copied so frequently that they finally wore out and perished"; and conversely that the "offshoots" survived "because they were rejected by the Greek Churches as faulty and so not used." It is hard to imagine a less historical answer to a historical inquiry than this one.

1. Transmission Theory

Probabilities and Arguments

1. It should be noted first that Hodges' "truism" is simply not true - either theoretically or actually. As a matter of fact, Hodges rests his case on no historical evidence at all, but on the mere theoretical probability of this "truism".

But even theoretically there is no good reason to believe that it is true. It does no good to say that Abraham will have more descendants than Isaac or Jacob, not to mention David or Hezekiah. The question is whether the majority of Abraham's descendants, through either Isaac or Esau, continued through the years to look more like their original ancestor than like their more immediate ancestors through the phenomenon of mixed parentage. 4

The point is that there is not a reason in the world to believe that copies "nearest the autograph" will normally have the largest number of descendants - even if these could be allegedly assumed to bear precise family resemblance.

Furthermore, unless one supposes that subsequent copies were regularly checked against known earlier copies, one must also reasonably assume that whatever errors were made in any of the copies would also be transmitted to their offspring as copies became exemplars.

What one may reasonably assume therefore, is the precise opposite of Hodges' presumption. More copies mean more errors, unless there were to be a systematic attempt to correct subsequent copies against earlier ones. But this is precisely not what one would expect in the earliest period, when

(a) copies would not have been made by trained scribes in scriptoria,

(b) copies were being made for pragmatic reasons, not necessarily with a sense of copying Scripture, and

(c) the earliest copies were probably very early carried away from their place of origin (or first destination).

Therefore, the proliferation of copies with numerous differences from the autographs would continue until certain factors converged to stop the process of proliferation and diversity. And when such a check occurred, it would freeze the form of the text then current - but a text that would most likely be far removed from the original.

When one turns to a variety of historical evidence, including the NT, one finds this to be exactly the case. In fact, what Hodges calls a "truism" turns out to be based on some form of theoretical "logic" not on hard data, since the "truism" does not exist anywhere in antiquity in the data themselves.

He uses the Latin Vulgate, for example, as an illustration to support his view of transmission. The enormous diversity and cross-contamination of the more than 8,000 Vulgate MSS amply demonstrates, Hodges argues, the inability of an "official" edition to arise out of diversity and uniformly claim the field (1970:18).

But this is a poor choice of illustration, for what the MSS of the Vulgate doo demonstrate is that Hodges' view of transmission does not work out. If it is true that the later hundreds of medieval copies of the Vulgate lacked the uniformity one finds in the Greek MSS, it is also true that they are far more like one another than they are like Jerome's original. 5

This is precisely as with the Greek NT, except for greater uniformity of the latter, which has another explanation. 6

Original Footnotes:

4. For some reason Hodges simply cannot see the rather total illogic of his theory at this point, and offered a considerable rebuttal in his response. But the fact is that in both human and manuscript lineage the further away one gets from the original "parent(s)" the more the lineage picks up cross-fertilization, especially when one throws in the clear geographic changes brought about by easy mobility.

Thus in response (1978:158 n.2) I suggested the following analogy:

"Let us assume that two very Swedish parents (tall, blonde, blue eyes, and so forth), Olaf and Helga Olson, have two boys, Karl and Sven. These two sons also marry Swedish young women, and each has two children. One of Karl's sons emigrates the USA and marries a very Italian young woman (short, black hair, dark eyes, and so on), and this son happens to be prolific and has 12 children. Of these children seven marry Italians, one marries a Swede, and 4 marry American "mixtures". Meanwhile in Sweden, catastrophe has struck thehouse of Olson and the two children of Sven are killed in an accident. The other son of Karl had one child, a daughter, who turned out to be barren. Thus the only descendants of Olaf Olson are in America, and the vast majority are now assuming very Italian features."

That something very much like this happened to the NT MS tradition is precisely why the argument from numbers is totally irrelevant.

5. Hodges also responded that this statement is "wildly untrue". But here he does not seem to have checked the actual data for himself. One might try collating any 10 late-medieval Vulgate MSS with Wordsworth & White's edition and see the results. After all, the Sixtene & Clementine editions, for all their differences, agree more often with one another than either does with Wordsworth & White.

6. In the original article I also noted that the same holds true with all critical editions of the words of the Church Fathers. A good critical edition of a Father's text, or the discovery of early MSS, always moves the Father's text of the NT away from the TR and closer to the text of our modern critical editions.

2. Transmission Data

Facts and Conclusions

2. In contrast to this theoretical (and unrealistic) view of the transmission of the NT proposed by Hodges, the actual historical data show an enormous fluidity in the earliest period, which disappears in later decades. 7

Hodges' contention that all early evidence derives basically from Egypt is patently false. What we theorized above about the earliest copies (not made by trained scribes, made for practical purposes, and each book transmitted independently over a widely scattered geography) seems in fact to have been the case.

From 150-225 A.D. we have firm data from all over the ancient world that a variety of text forms were in use, but in all these materials there is not a single illustration of the later Majority (-Byzantine) text as a text form.

The evidence from Egypt is basicaly singular. The earliest Greek MSS (P66 P75 P46 P72 ca. 175-250), the citations of Clement (ca. 190-215) and Origen (215-245), and the earliest translations (Sahidic & Bohairic Coptic) all bear witness to a single text-type.

There are indeed some scattered readings in Clement and P66 from the so-called Western tradition and a few readings in P66 where it now has the earliest evidence for Byzantine readings, but these are so few as to alter the text of these witnesses only slightly (see Fee 1968b).

The point is that the Byzantine text simply did not exist in Egypt in the early period. P75 is therefore not a reject; it represents the only "broad stream" that existed there. This is further evidenced by Origen who apparently used several different Bibles, and P66, which was corrected against a 2nd MS. In none of these does one find evidence for the existance of the Majority Text.

The "Western" Text-type

The same is true elsewhere in the Christian world in the 2nd and 3rd centuries. The other type of text that existed in the 2nd century is commonly called "Western" because variants peculiar to it are firmly established in texts found in North Africa (Tertullian, Cyprian, some OL), Italy (Novatian, some OL), and southern France (Irenaeus). "Western" however, is something of a misnomer, for many of its peculiar variants are also found in the East (Tatian and the OS) and occasionally in Egypt (some quotations in Clement, John 6-7 in P66).

But despite this early and widespread attestation, these various witnesses lack the homogeneity found in Egypt and in the later Byzantine text. The textual relationships are not consistently sustained over large portions of text; rather, "Western" describes a group of witnesses, obviously related by hundreds of unusual readings, sometimes found in one or several, sometimes in others, but apparently reflecting an uncontrolled, sometimes "wild" tradition of copying and translating.

Again however, in none of these areas does one find a single witness to the Majority text as a text form, but only sporadic attestation to the existence of some of the Byzantine readings.

One might argue, of course, that all the early translations (Latin, Syriac, Coptic) and early Fathers (Justin, Irenaeus, Tatian, Clement, Tertullian, Origen etc.) had the misfortune to use only the "rejected offshoot" MSS. But if so, who represents the "broad stream" that "wore out" the copies more like the autographs? The obvious answer is that the Byzantine text form simply did not exist in the 2nd and 3rd centuries, although many of the variants that were to be found in it had already come into existance.

The Majority text as a full-fledged form of text, distinguishable from the Egyptian and "Western", does not appear in history until about 350 A.D. NT citations that are closer to the TR than to the Egyptian and "Western" texts first appear in a group of writers associated with the church of Antioch: Asterius the Sophist, the Cappadocians, Chrysostom, Theodoret of Cyprus.

But even so, these Fathers had a NT only about 90% along the way to the full Byzantine text of the later Middle Ages. The earliest Greek MS to reflect this text is from Alexandria (Codex W, ca 400 A.D. - Luke 8:14-14:53 only) and is only about 85% Byzantine, while the earliest full witnesses to it are uncials from the 8th and 9th centuries (Codices E F G H M Ω) - and even these reflect a slightly earlier stage of the text finally found in the TR. 8

The fact is that even this text, as generally homogeneous as it is from about 400 - 1500 A.D., has clearly evolved from an earlier form, where the kinds of readings peculiar to it become more thoroughgoing at a later stage.

These are the historical data. They are "objectively verifiable" and incontrovertable. It is true that the actual origins of the Byzantine text as a text-type are shrouded in mystery, but that is scarcely an argument in its favor.

If it were indeed the closer to the autographs, the same "mystery" would prevail for the origin of the Egyptian text. In either case one has to argue for recensional activity at its beginning. But that has been demonstrated not to be true of the Egyptian text (see Fee 1974[ch 13]), whereas the Byzantine text has all the earmarks of a recension - of a kind for which there is firm evidence of its existence (1974:30-31).

The idea that the Majority text of the Middle Ages reflected the "broad stream" of the transmission of the text going back to the autographs is simply a myth.

Original Footnotes:

7. For the best recent attempt to write a history of the text, see Birdsall, 1970.

8. For the full display of the data demonstrating these judgments, see Fee 1974 and 1971a [now Chapters 13, 15].

3. The Dominance of the Maj. Text

But the question still must be answered: How does one account for [The Majority Text]'s dominance and general uniformity?

3. It was suggested above that one would expect a proliferation and diversity among copied texts until certain factors would combine to stop that process. But that would not at the same time guarantee that one of these texts should emerge as dominant and thereby become the uniform text of all Greek Christendom.

Such in fact did nevertheless happen - but only with regard to the Greek text. The Latin, Armenian, and Syrian churches, among others, developed their own dominant and generally uniform text, which did not coincide with the Greek text.

But our interest is with the Greek. How did the Byzantine text become dominant? The answer lies in a combination of several factors that converge between the 4th and 7th centuries.

(a) By the 4th century all of the factors that led to diversity had been superceded by their opposites.

First, instead of untrained scribes who copied parts of the Bible for pragmatic purposes, there had emerged the trained Christian scribe, whose work was being produced in scriptoria. This began early in Alexandria, as the Egyptian MSS bear abundant witness, and probably was thoroughgoing after Constantine (c. 310 A.D.).

Second, the concept of canon brought in an ecclesiastical concern over the wording per se, which did not exist among the copyists of the 2nd century, as the NT citations in all the Church Fathers of this century bear witness. The origins of the Latin Vulgate are to be explained precisely for this reason. 9 This will not guarantee uniformity, of course, but it will surely lessen the amount of "new variation" and, conversely, will add to the process of cross-checking and "correcting" existing MSS (note the several times this happened to Codex Sinaiticus, always away from its Egyptian standard and in greater conformity to the Byzantine [see. Fee 1968a: 43-4]).

Third, instead of copies being made to be carried off to some other center, copies were now being made to remain where they were - for study purposes. Herein lies one of the most significant factors both for "dominance" and uniformity. After all, it is not by accident that the vast majority of extant Greek MSS were found in large quantities in monastery and university libraries!

(b) One can scarcely underestimate the influence of Chrysostom in the history of the Greek church.

As Quasten notes,

"None of the Eastern writers has won the admiration and love of posterity to such a degree as he" (1960:3.429).

Prime evidence of this influence is both the abundant quantity of extant MSS of his own writings (by far greater than for any other Greek Father) and the great number of extant spurious writings attributed to him, whose authors sought immortality for their writings under the prestige of Chrysostom's name. Its is almost inevitable that the text-form Chrysostom used first at Antioch and then later carried to Constantinople should become the predominant text of the Greek church. 10

(c) The most important factor for the dominance and general uniformity of the Byzantine text is directly related to (b) above.

By the end of the 7th century, the Greek NT was being transmitted in a very narrow sector of the Church - namely, the Greek Orthodox Church with its dominant patriarchate in Constantinople. By the time of Chalcedon, Greek was almost unknown in the West, 11 and after Chalcedon the decline of Alexandria and the subsequent rise of Islam narrowed Greek-speaking Christendom still further.

All of these factors together ensure both the dominance and general uniformity of a text-form properly called Byzantine. 12

9. Damasus of Rome asked Jerome to make a new translation because of the great diversity that existed in the Latin Bibles.

10. For reasons not easy to decipher, Hodges chose to refute this reality, arguing that studies showed that Chrysostom did not in fact witness to a Byzantine text. But his argument only underscored a point made above, that Chrysostom simply had an early expression of the Byzantine text that was about 90% along the way to its full medieval expression.

Furthermore, Hodges chose to cite an earlier study by Geerlings and New (1931) that seemed to dispute this reality about Chrysostom's text. But my own sampling of work on Chrysostom's text made me realize that something was desperately wrong with that study. The result of this interchange with Hodges caused me to go back over the work of Geerlings and New. It turned out that the failure was methodological. The results of that inquiry were published as Fee 1979 and demonstrated that Chrysostom's text of the NT was consistent throughout.

11. Another point Hodges took exception to. But to say that someone like Augustine knew Greek is a far cry from saying that he used a Greek NT, which he did not.

12. The rest of the original article responded to Hodges' final argument as to the "subjectivity" and "circularity" of the reasoningn of modern textual criticism. Since most of that response was a summary of some arguments presented in Fee 1974 [ch. 13], and the rest of it can be found in the critique of Pickering that follows, I chose to omit it from the present presentation.

Return to Top