In the past, a proliferation of translations of Scripture, especially of the New Testament, has usually resulted in issuance of an “authorized” translation. Bishop Damascus of Rome (ca. 382) asked Jerome to settle the problem of differing Latin translations by a single “authorized” version, which became the Latin Vulgate. The Syriac Peshitta perhaps originated as an authorized attempt to supplant differing translations. And, of course, this is also the history of our English Authorized Version (AV), more commonly known as the King James Version (KJV).

We are once again in a period of “competing” translations. To some people this seems desirable: a comparison of translations may give the lay reader a broader perspective on the possible meaning of the Hebrew or Greek text. But for others it results in hopeless confusion. Compare, for example, three strikingly different alternatives given by the Revised Standard Version (RSV), the New American Standard Bible (NASB), and the New English Bible (NEB) to the literal but ambiguous text of the KJV in First Corinthians 7:36, “But if any man think that he behaveth himself uncomely toward his virgin.…” (1) “If any one thinks that he is not behaving properly toward his betrothed …” (RSV). (2) “But if a man thinks that he is acting unbecomingly toward his virgin daughter …” (NASB). (3) “But if a man has a partner in celibacy and feels that he is not behaving properly towards her …” (NEB).

Those who see value in having a variety of translations either buy several, or use a compilation in parallel columns (such as CHRISTIANITY TODAY’s The New Testament in Four Versions), or resort to something like The New Testament From 26 Translations (Zondervan).

For the rest there remains the strong desire for an “authorized” version. However, the possibilities of this in our day seem very slender. To be sure, the National Council of Churches placed its imprimatur on the RSV, but that very fact made it suspect to thousands of Christians who oppose the NCC. The best of British scholarship offered us the NEB, but for many an imprimatur from “liberal” scholarship was as bad as one from the NCC.

The final alternatives are either to cling to the old (the KJV) as a kind of nostalgic refuge in the midst of uncertainty or to ask some “authority” his opinion. But alas, if one asks more than one “authority,” he is likely to end up in the same confusion that prompted the question.

My purpose in this essay is not to set myself up as another “authority,” though I do have some strong preferences, but rather to suggest some bases for choice that even the “authorities” must use, and to give reasons for confidence in any one of several contemporary translations as generally adequate English representations of the original Greek text.

Article continues below

The most common criterion for choosing a translation is, for want of a better term, taste. By “taste” I mean all those factors involved in an adequate communication of words and ideas from one language to another. It will include questions of (1) the propriety of the English idiom (people raised on the classical idiom of the KJV may think the Cottonpatch paraphrase borders on blasphemy), (2) the ability of that idiom to communicate to the reader (one of the reasons why such paraphrases as those by J. B. Phillips and Kenneth Taylor are so popular), and (3) the suitability of the new idiom for expressing the intent of the original author (in this area the science of linguistics plays a vital role).

In this regard, such contemporary translations as the Jerusalem Bible (JB), Today’s English Version (TEV), the RSV, and the NEB succeeded with varying degrees of excellence. (Many, including myself, feel that the NASB fails on point three: it tends to be a bit too wooden, too literal a rendering of the Greek idiom.)


However, the basic criterion is not taste but text. That is, the basic question is whether the Greek text used by the translator adequately represents what the biblical author originally wrote. At this point, the contemporary translations as a group have one thing in common: they tend to agree against the KJV and its eccentric offspring, King James II, in omitting hundreds of words, phrases, and verses. The reason is that, although these translations have usually been based on various Greek texts, those differing texts have a common base. This means that although the problems of finding the “original” New Testament text is not fully solved, there is a consensus among scholars—and that consensus is in favor of a text differing noticeably from the Greek text from which the KJV was translated.

But who is to say what the “original” text looked like? What is the layman to think when he hears conflicting “authorities” speak with confidence about the “original Greek”? Although these questions will not be answered to everyone’s satisfaction, at least it may be helpful to know what textual scholars have been, and are, up to.


The problem faced by the textual critic is a combination of three factors: (1) The originals, probably written on papyrus scrolls, have perished. (2) For more than 1,400 years everything was copied by hand, and the scribes made every conceivable kind of copying error as well as deliberate changes to the text. (3) There are now, in whole or in part, 5,338 known Greek handwritten copies (manuscripts), plus hundreds of copies of early translations, plus the evidence from the biblical quotations in the writings of the early Church Fathers.

Article continues below

The vast quantity of biblical material far surpasses that of all other documents of equal antiquity combined, and is for the New Testament textual critic both his fortune and his problem. It is his fortune in that with such abundance, he can be reasonably sure that the original text is to be found somewhere in the available material. Unlike those searching for other original texts (including the Old Testament), he scarcely ever needs to resort to emendation, or correcting the text because he cannot make sense of what is found in all the manuscripts.

However, no two copies anywhere are exactly alike, and the greater the number of copies, the greater the number of differences (variants, readings) among them.

His problem, then, is to sift through all the material, detect the errors, and try to determine what the inspired biblical author himself actually wrote.

Although this seems to be a formidable task and one on which 100 per cent certainty is never obtainable, careful study has led scholars to approximately a 95 per cent certainty about the original text. Where uncertainty remains, it is due partly to the emphasis on different principles among scholars and partly to the very difficulty of choice in some cases. It is in this small percentage of uncertainty that one will find most of the differences among contemporary translations. A closer look at each of factors two and three—scribal errors and changes, and the vast number of copies available—should serve to reveal the textual scholar at his craft as well as to indicate where we are in the quest for the original text.


To begin with factor three: the problem of sifting through the manuscripts is not as formidable as it might at first appear. Although no two copies are exactly alike, many are enough alike that the copies tend to group themselves into three (some scholars think four) major families of texts (text-types).

There is a group that derives basically from Alexandria in Egypt, headed by the recently discovered Bodmer papyri (P75 and P66) in the Gospels (ca. A.D. 200), the Chester Beatty papyrus (P46) in Paul (ca. 225), the Bodmer papyrus (P72) in Peter and Jude (ca. 275?), the great Codex Vaticanus (B, ca. 325), and the quotations in Origen (225–250). It is also supported to a lesser degree by several other manscripts (e.g., Aleph, C, L, W, 33) and by the later Alexandrian Fathers (Didymus, Athanasius, Cyril).

Article continues below

A second group is equally as early but shows nothing of the homogeneity of the first. It is often called the “western” text because it is headed by the earlier of the Old Latin versions, the great bilingual Codex Bezae (D, ca. 600), and by the Latin Fathers (e.g. Tertullian, Cyprian, Ambrose). The heterogeneity of this group has led some of us to refer to it as the western “textual tradition,” which describes a group of manuscripts obviously related by some peculiar readings but apparently reflecting an uncontrolled, sometimes “wild” tradition of copying and translating.

The final group, the “Byzantine” or “majority” text, is made up of over 80 per cent of all the manuscripts. The earliest full witnesses to this text are manuscripts from the eighth century (E and Omega), though it is also represented for the most part in the Gospels text of Codex Alexandrinus (A, ca. 475) and in the biblical quotations by several Church Fathers from Antioch and Constantinople in the latter half of the fourth century.

It has been argued recently that the peculiar variants of this text-type may be found earlier in such a manuscript as P66. However, this is a half-truth. Such readings are occasionally supported by an early witness; but P66, rather than supporting specific variants of the Byzantine text-type, tends to vary from its Alexandrian cousins with a rash of patently non-original readings of the same kind as appear solidly and specifically entrenched in the Byzantine text-type. (See my Papyrus Bodmer II (P66): Its Textual Relationships and Scribal Characteristics, University of Utah Press, 1968.)

It is from manuscripts of this third text-type that the Greek text lying behind the KJV has come. The printed edition of this Greek text goes back to Erasmus, who in 1516 published the first Greek text of the New Testament chiefly on the basis of two inferior twelfth-century manuscripts. As more manuscripts came to light in the following years, the vast majority supported this kind of text; those that did not were generally considered “eccentric.” This became known as the “received text” (textus receptus=TR) and was universally accepted on the basis of “majority rule.”

Article continues below

The problem with “majority rule,” of course, is that the majority may be wrong. Indeed, it eventually came to be recognized that quality is far more important than quantity in making textual choices.

In the case of the New Testament text, the problem with the majority was twofold: (1) As more and more older manuscripts came to light, it became readily apparent that the older the manuscript, the less it looked like the majority text. (2) An analysis of scribal errors showed that the later manuscripts had tended to accumulate all the kinds of errors scribes are known to have made, as well as to accumulate thousands of deliberate additions and alterations to the text from earlier centuries.

This case against the majority text was set forth in its classic form by the British scholars B. F. Westcott and F. J. A. Hort, in their (Hort was chiefly responsible for it) famous introduction to the Greek text that served as the basis for the Revised Version of 1881. Besides the factors noted above, the cornerstone to their case was the fact that no Church Father earlier than 350 had New Testament quotations supporting the peculiar readings of the Byzantine text-type.

But if the majority of manuscripts are late and do not reflect the original text, what then? Hort was faced with the fact that the other two earlier text-types had equal antiquity (they could be traced back into the second century). So on the basis of internal evidence (the tendencies of scribal changes and judgments as to what the author probably wrote), he chose in favor of the Alexandrian text-type, which he dubbed “neutral,” and which he believed really was.

Westcott and Hort did their work so thoroughly and with such exceptional skill that textual work since then has been either in reaction to or in implementation of theirs. Some indeed were uneasy with their term “neutral”—after all, no manuscripts have totally escaped scribal errors. What is significant is that even those who tended to disagree with Westcott and Hort’s obvious preference for Codex B published Greek texts that differed very little from theirs—but different in many places from the “received text.”

It should be noted, however, that in recent years some attempts have been made to give the majority text a new hearing. One such case is by the British scholar G. D. Kilpatrick and his students. They have developed a method of choosing among variants strictly on the basis of internal evidence, with total disregard for the age, quantity, or quality of the witnesses to a given variant. Although this falls short of wholesale acceptance of the majority text—in fact, their results are far from it—their work does have the effect of giving the majority text equal hearing with the earlier witnesses. To a degree this seems to be a welcome corrective, for age alone should not determine originality. The biggest fault Kilpatrick’s critics find with his method concerns his choice of “internal” criteria on which to make decisions (it is felt that he places too much emphasis on an author’s style). Although he has not intended it to be so, his method appears to be the ultimate in subjectivity.

Article continues below

Less significant attempts to give the text of the KJV a new hearing have also been made by men such as Jay Green of the Religious Book Discount House (King James II) and David Otis Fuller (editor of Which Bible?). Their basic arguments, however, are theological (and sometimes emotional) and have little to do with the actual data. Fuller, for example, argues: “If we believe the original writings of the Scriptures were verbally inspired of God, then of necessity, they must have been providentially preserved through the ages.” One might superficially answer that they have been preserved, but in the earlier, not the later, manuscripts. But the point is that desiring to have exact providential preservation of the original New Testament by the medieval church does not make it so. (It seems strange that one in the Baptist tradition, who rejects so much from medieval Christianity, should argue so strongly for this point.) In any case, no amount of wanting can make the account of the angel at the pool (John 5:3, 4), for example, a part of John’s Gospel if John indeed did not write it—and its absence from all early manuscripts as well as its non-Johannine language does not seem to allow the remotest possibility that he did.

A more significant attempt to give the majority text a new hearing has come from Zane C. Hodges of Dallas Seminary, whose article is reprinted in Fuller’s Which Bible? Hodges has revived the argument from quantity: so many manuscripts cannot be wrong. His basic argument is 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.” Although this may be hypothetically true, the question for the New Testament is whether it is really true. All the actual “hard” evidence speaks to the contrary.

Article continues below

During the second to the fourth centuries, the readings peculiar to this text are generally unknown (I say “generally” because an occasional variant of this type may be found here and there) in Syria (the Old Syriac), Egypt, Africa (the Old Latin, Tertullian, Cyprian), Italy (the Old Latin, Novatian), and southern France (Irenaeus). The first witness to this text (as an entity) is St. John Chrysostom, who apparently carried such a text from Antioch to Constantinople. Its wide medieval acceptance was probably the result of the immense influence of Chrysostom, the drying-up of copies of the Alexandrian text owing to the demise of Christianity in Africa, and the exclusive use of Latin in the West, where few Greek manuscripts were transcribed.

In any case, the argument is irrelevant. Even if the parent of the majority text could be shown to be early, the ultimate question is not which parent (Alexandrian, Byzantine, Western) produced the most offspring (the Latin Vulgate wins this one) but rather which parent is more likely the original.

Hodges has also raised the cry of subjectivity—that choices based on internal criteria are inevitably subjective. In fact, textual work may be made to appear circular: one chooses the reading supported by the best manscripts, which are determined to be the best because they have the best readings.

As a theoretical argument this criticism looks convincing, but when one examines the data, in this instance by comparing the readings of specific manuscripts, the accusation of circularity seems to be only theoretical. All this leads us to look at the second factor: scribal errors and alterations.


Scribal changes in any text fall into two large categories: intentional and unintentional. The latter are the more numerous and the less significant. They are generally easy to discern (usually added or dropped letters or words because of slips of the eye, ear, or mind), and they have affected the New Testament text much less than the intentional changes.

Most of the problems, then, lie with those intentional changes. However, lest one think of these early scribes as vicious fellows, let it quickly be pointed out that one may overdo the concept of deliberate. The majority of these changes were attempts to help out the text. The early scribes appear to have been far more concerned with the inspired message of the text than with the very words, for they frequently made that message clearer by smoothing out grammar, adding nouns or pronouns where there might be ambiguity, substituting common synonyms for uncommon words, clarifying difficult phrases, and confirming one passage to another (especially in the Gospels). There are some instances, of course—and these are usually very important ones—where whole sentences or narratives have been added (or subtracted) in the interest of either doctrine or completeness (such as the catechetical addition before the baptism in Acts 8:37, which is supported by only a handful of witnesses, yet found its way into the KJV).

Article continues below

Examples of this process may help. Let us begin with some from John’s Gospel, which are untranslatable, and yet which clearly show scribal tendencies.

1. A common feature of Greek prose is the abundant use of conjunctions and particles to add nuances between sentences and phrases. John’s Gospel, however, has many sentences that lack these conjunctions (the technical term for this lack is asyndeton). Over hundreds of years, scribes in various ways tended to correct asyndeton, i.e., they tended to conform John to the more standard style. (There are often as many as four independent attempts to do this in the same sentence!) Elimination of conjunctions for reasons of style is uncommon, so that a text marked by asyndeton is not likely to have become so by editing. What is significant here is that this clearly secondary scribal correction, the addition of conjunctions, is found predominantly in the majority text, not the earlier, especially Alexandrian, manuscripts.

2. Something similar is true with the use of the definite article with personal names. The usual practice in Greek was to include the article with names (=“the aforementioned …”). John is the great exception to this usage in the New Testament; and again, the tendency of later scribes was to add the definite article in conformity to ordinary usage. Again, the TR has picked up the great majority of these additions, so that it has the article eighty-four times with the name Jesus where Codex B does not, and B never has it where TR does not.

As a further example (now of translatable variants), take the tendency of scribes to conform one biblical passage to another. In the earliest manuscripts and Church Fathers, both east and west, the Lord’s Prayer in Luke’s Gospel lacks the phrases “who art in heaven,” “thy will be done on earth as it is in heaven,” and “but deliver us from evil.” On the other hand, these phrases are not lacking from any known copy of Matthew’s Gospel. It seems evident, therefore, that the later scribes of Luke’s Gospel eventually conformed his version of the prayer to the longer and more often quoted version of Matthew’s Gospel. (For further information see J. Jeremias, The Lord’s Prayer, Fortress, 1964.)

Article continues below

This harmonizing tendency appears early in the western texts and has become thoroughgoing in the majority text. Such harmonizations appear on nearly every page of the Synoptic Gospels (especially Mark), as well as in parallel accounts in Acts (see how Acts 9:4 was conformed to 26:14) and parallel passages in Paul (compare Ephesians 1:7 with Colossians 1:14). It can always be argued, of course, that some scribe later deliberately omitted or changed texts for the sake of variety, and thus that the harmonized texts were earlier, but there is no evidence that happened and it seems to be contrary to all reason for him to have done so.

The point is that such examples may be multiplied hundreds of times over. Therefore, the more significant readings peculiar to the later manuscripts are already suspect by the company they keep. Moreover, when one applies such basic criteria as “the more difficult reading is probably original” (on the grounds that scribes tend to clear up difficulties, not create them) or “the variant which best explains how the others came to be is probably original,” the majority text comes out second (or third) best in most cases.

For example, the well loved story of the adulterous woman in John 8, which almost certainly actually happened to Jesus, seems just as certainly not authentic to John’s Gospel. The evidence? It is entirely unknown in Greek and Syriac manuscripts before A.D. 600. It is likewise unknown to Church Fathers before the end of the fourth century. The story first appears in John 8 in some, but not all, of the Old Latin translations, and when it finally is found in Greek manuscripts, there are two significantly different accounts, and the story is found in three other places (following John 7:36, John 21:24, and Luke 21:38). Furthermore, there are several examples of non-Johannine words and grammar in the story.

If this story were original to John’s Gospel, the only way one could explain all the data is to assume it was omitted (for what reason?) in a very early copy after the original, that other copies of the original were misplaced for three or four centuries, that in the meantime the only copies of John circulating throughout the Christian world were from the manuscript with the omission, and that eventually the misplaced “original” reappeared and won the day. However, this appears to be an exercise in mental gymnastics. The point is that the one reading that explains all the others happens also to be the best attested reading.

Article continues below

Since one hates to lose an account that has such a “ring of truth” to it, the alternative is to do what most translations have done: acknowledge that it is not original to John, but recognize that it is probably genuinely historical, and therefore include it in brackets, italics, or footnotes. It was simply one of those “many other things Jesus did” (John 21:25); it did not make its way into one of the four Gospels but continued to live in the Church, and was eventually placed in the Gospels (at various points) because it was too good to lose.

But not all textual choices are so easy. And this is why, finally, that although almost all contemporary translations are a vast improvement on the KJV, the Greek texts they use do not always look alike. For example, the JB and NEB translate a Greek text of John 1:34 “Chosen One of God” where the others read “Son of God.” Although the latter has the better of it in quality and quantity of witnesses, the former is found early, and with broad geographical distribution.

The choice here is not easy; however, “Chosen One” seems to be preferable as John’s original. Both titles are messianic, one from Psalm 2:7 and the other from Isaiah 42:1, and both are associated with Jesus’ baptism (Mark 1:11). Since John also calls Jesus the Lamb of God, probably in light of Isaiah’s “suffering servant,” “Chosen One” would be contextually well placed. But if it was original and has such meaningful messianic acclaim, why would a scribe change it? Probably for orthodox theological reasons: “Chosen One of God” would have been well suited to adoptionist Christology, which believed the human Jesus was “adopted” (=“chosen”) to be God’s Son. Of course, on the other hand, it could have been a deliberate adoptionist change to the text. But if so, why only here in the Gospel?

Such are the questions raised and the way textual choices are made. Since there are some obvious variables, what is the layman to do? The following concluding suggestions may be helpful:

1. Textual choices are first of all matters of history, not of faith. Our “manuscripts” were some believers’ Bibles. Our question is, when their “Bibles” differ, which variant is most likely the original? Here we must be historians—and good ones—if we are to know what God’s Word actually said. It should be noted of historical choices that any number of hypothetical possibilities might be raised, but that not all possibilities have equal probability. Ultimately textual critics must choose what seems to them the most probable solution.

Article continues below

2. Textual criticism is not the stronghold of “unbelieving” scholars. What is most probable in textual choices transcends confessional boundaries. Hence, confessional evangelicals are generally at one with other scholars on the principles, if not always on the actual choices, of textual criticism. (See, for example, the excellent Introduction to New Testament Textual Criticism by J. Harold Greenlee [Eerdmans, 1964], which elaborates and generally says far better what is said in this article.) Two newer translations, the NASB and the forthcoming New International Bible, were both done by committees of committed evangelical scholars.

3. For a generally thorough commentary on most of the translatable textual variants in the New Testament, see Bruce M. Metzger, A Textual Commentary on the Greek New Testament (American Bible Society, 1971). Here one can see how textual choices were made by a committee of five well-known experts.

4. The layman finally should learn to use all the available helps. Consult several of the better translations, use them carefully and comparatively, and in places where they differ consult Metzger and the better commentaries for the data and kinds of options that are available.

Such a procedure is surely better than having an “authorized” version, for the possibilities for understanding the Word of God are greatly enhanced.

George M. Marsden is associate professor of history at Calvin College, Grand Rapids, Michigan. He has the Ph.D. (Yale University) and has written “The Evangelical Mind and the New School Presbyterian Experience.”

Have something to add about this? See something we missed? Share your feedback here.

Our digital archives are a work in progress. Let us know if corrections need to be made.