Realms of Faith


 

Comparing Bible Translations: Analysis

Issue #2: Textual Basis

NOTE: For the sake of consistency, from here on translations will be listed in chronological order of publication, according to the latest edition reviewed. They proceed in this order: Tyndale, GEN, KJV, RV, Young, Darby, ASV, Weymouth, Moffatt, Montgomery, AAT, BV, William, Lamsa, PME, Norlie, NWT, AMP, JB, MLB, CPV, RSV, LB, NEB, TBV, NKJV, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, INC, KJ21, LITV, NASB, GW, NLT1, NIrV, ISV, NET, ESV, TNIV, MSG, CW, HCSB, NLT2.

While there are over five thousand manuscripts of the Greek New Testament, no two are identical. Those who copied them by hand sometimes made mistakes, and sometimes made changes. One of the tasks of the biblical scholar is to determine what the original reading was wherever there are variant readings. This task is called textual criticism. It is a necessary first step before translation, because someone must decide which reading is to be translated.

Toward which manuscript tradition does the translation lean?

When someone has compared the manuscripts so as to compile a Greek New Testament that resembles the original text as closely as possible, this is called a critical edition. The first printed critical edition was done by Erasmus, and it and its revisions came to be known as the Textus Receptus, or Received Text. Erasmus' work was later revised several times by Stephanus and again by Theodore Beza. As more manuscripts were discovered and techniques were refined, other editions followed. The one in use today is available in two identical forms (differing only in format)–the UBS Greek New Testament (now in its fourth edition) and the Nestle-Aland Novum Testamentum Graece (now in its 27th edition). These editions are known as the Critical Text, since they represent the present state of textual criticism.

Erasmus did most of his work from a few later Greek manuscripts. They come from what we call the Byzantine tradition, a fairly standardized text that came to dominate the majority of manuscripts, since Greek was native to Byzantine Europe. For this reason, the Byzantine tradition is also called the Majority Text. After the time of Erasmus, however, other traditions were uncovered as earlier manuscripts were discovered. The "Western" tradition, influenced by Latin though not limited to Western Europe, tended to embellish the stories of the Bible and is generally longer. The Alexandrian tradition, associated with Egypt and Palestine, tends to be shorter and more careful to avoid confusing similar passages. Some textual critics propose there is a fourth tradition, the Caesarean, associated with Origen and marked by a tendency to combine the other traditions.

When the KJV was translated, various revisions of the Textus Receptus (TR) were virtually the only viable option for a New Testament text base. Today, since Alexandrian readings are supported by the most ancient papyri and other evidence, most modern translations use the Critical Text. Still, the KJV has established the TR firmly as the traditional text, and the English verse divisions are based on it, and so it and the Majority Text still have many advocates.

(Evangelical scholar Daniel Wallace has an excellent analysis of the main theological arguments for TR/Majority Text advocacy.)

 

Summary: Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV work strictly from the Textus Receptus. The CW is about 3/4 Byzantine, and the TBV and Darby's version have an even mix of Alexandrian and Byzantine readings. Since Lamsa works from the Syriac (Western) tradition, his readings match the Byzantine text most of the time, but he has several unique readings. Moffatt, Norlie, LB, ISV, and MSG are also fairly eclectic. The other translations surveyed almost always go by the Critical Text. The consistently Alexandrian versions are AAT, NWT, JB, NIV, TNIV, NJB, NAB, NRSV, INC, NASB, CEV, ESV, and HCSB. The Amplified Bible usually places Byzantine readings in italics, and BV, MLB, NASB, and HCSB use parentheses or brackets when an entire sentence or verse is in question.

Presence or absence of a divine name

One feature of the Byzantine text is the tendency to make the name of God explicit where it was not expressed originally. This was often done out of reverence or for clarification.

Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV have the Byzantine reading in all five instances. Norlie has the longer reading all five times, but in Romans 1:16 has the gospel of Jesus Christ, an expansion not supported by any Greek text. The LB and TBV choose the Byzantine three times, and the Darby and CW twice. The RV, ASV, Moffatt, Lamsa, RSV, NLT1, NLT2 and ISV each have only one Byzantine reading of the five. The AMP chooses one Byzantine reading and includes the other four in italics. Weymouth, Montgomery, AAT, BV, William, PME, NWT, JB, MLB, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NIrV, NET, ESV, TNIV, MSG, and HCSB take the Alexandrian reading all five times.

Additional examples: Matt. 12:25; 13:51; Mark 10:52; Luke 22:31; 24:36; Acts 2:30; 7:30; 15:17-18; 20:25; 22:16; 1 Cor. 15:47; Gal. 3:17; 4:7; Phil. 1:14; 4:13; 1 Tim. 2:7; Heb. 1:3; 10:30; 1 Pet. 1:22; Rev. 20:9.

Differences in length of a divine name

Related to the above issue is the lengthening of a name. Christ becomes Jesus Christ, which becomes the Lord Jesus Christ, which becomes our Lord Jesus Christ. This tendency is easy to trace through the manuscripts and is not limited to biblical texts.

Once again, Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV have the Byzantine reading all five times. Lamsa, Norlie, and CW have three Byzantine readings. (CW also rewrites Matt. 16:20.) Darby, Moffatt, LB, TBV, and AMP each have one Byzantine and four Alexandrian readings (the AMP again using italics for rejected readings). The others consistently use the Critical Text–RV, ASV, Weymouth, Montgomery, AAT, William, PME, NWT, JB, MLB, RSV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, NASB, GW, NLT1, NIrV, ISV, NET, ESV, TNIV, MSG, HCSB, and NLT2. INC is Alexandrian but replaces Lord with Savior in Acts 16:31. (MSG uses Master instead of Lord.)

Additional examples: Matt. 8:29; 16:13; Luke 4:41; John 4:42; 9:35; Acts 15:11; 19:4, 10; 1 Cor. 5:4; 9:1; 16:23; 2 Cor. 5:18; 11:31; Col. 1:28; 1 Thes. 3:11; 2 Thes. 1:8, 12; 1 Tim. 1:17; 5:21; Phe. 6; Heb. 3:1; 1 John 1:7; 2 John 3; Jude 4; Rev. 1:9; 11:17; 12:17

Other pious expansions

A striking example of the growth in manuscripts is in the title of Revelation. From The Revelation to John it grows in ms. 1775 to The Revelation of the all-glorious evangelist, bosom-friend [of Jesus], virgin, beloved to Christ, John the theologian, son of Salome and Zebedee, but adopted son of Mary the Mother of God, and Son of Thunder. Most changes are not quite as striking, but liturgy and asceticism have led many a scribe to "heighten" the text at certain points. Often this is easy to detect, as it shows up in only one or two manuscripts. But some of these caught on and are included in the Byzantine tradition:

Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV continue to stand together with all Byzantine readings here. Lamsa has four readings, Norlie and CW three, and RV, Darby, ASV, Weymouth, and MSG have two Byzantine readings apiece. (The MSG paraphrases the Byzantine coda of the Lord's Prayer as You're in charge! You can do anything you want! You're ablaze in beauty! Yes. Yes. Yes.) The PME, NEB, TBV, GNC, REB, and ISV each take one Byzantine reading. The Montgomery, NASB, and HCSB place the doxology of the Lord's Prayer in brackets but otherwise have Alexandrian readings. AMP is Alexandrian here but puts three of the Byzantine readings in italics. MLB is Alexandrian with two of the Byzantine expansions in brackets. The others–Moffatt, AAT, William, NWT, JB, RSV, LB, NIV, NJB, NAB, NRSV, CEV, NCV, GNT, GW, NLT1, NIrV, NET, TNIV, ESV, and NLT2–take the shorter, Alexandrian readings each time. INC is Alexandrian but reads Chosen One in John 9:35.

Theological clarification

On a few occasions, scribes made slight alterations to the text for theological reasons. Perhaps they felt free to add precision to the author's point, or they may have suspected that a text had been changed already and that they were restoring it to its original form. In any case, the tendency in such changes is to make difficult texts harmonize with orthodox theology.

Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV, which consistently have the Byzantine reading, are joined by the Lamsa in four cases and the TBV in three. Norlie has two Byzantine readings and two Alexandrian, but paraphrases the crucial phrase out of Luke 2:43. CW is likewise split and rewrites Mark 10:24 as, It isn't easy for those who make a lot of money to change, even if they've earned it legitimately. Several translations have one Byzantine reading (usually in Mark 10:24): RV, Darby, Moffatt, ASV, Weymouth, Montgomery, PME, LB, ISV, and MSG. The AMP is Alexandrian but includes two of the Byzantine readings in italics, and Moffatt and MLB bracket the reading in Mark 10:24. The fully Alexandrian versions here are AAT, William, NWT, JB, RSV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, NASB, GW, NLT1, NIrV, NET, TNIV, ESV, HCSB, and NLT2. The Cotton Patch Version chooses the Alexandrian reading for both verses in Acts, and the INC is Alexandrian except its unattested reading Mary and Joseph in Luke 2:43.

Additional examples: Luke 2:33; 23:42; John 1:18; 6:47; 1 Cor. 5:7; Col. 2:11; 1 John 4:3

Possible marginal glosses

When scribes made mistakes, they or others often placed corrections in the margins. Unfortunately, the margin was also a place where scribes would write explanatory comments (glosses) and other remarks about the text. Rather than risk omitting part of the text, a later scribe might incorporate such a note if he thought it might be part of the text itself. A number of longer variants with very weak textual support are suspected of having arisen in this manner.

Nine translations choose the longer reading all four times: Tyndale, GEN, KJV, Young, Darby, TBV, NKJV, KJ21, and LITV. Three times the Lamsa, Norlie, GNT, ISV, and CW have the longer reading. (CW adds a full sentence to Luke 9:54 referring to Mount Carmel and asking that the towns be wiped off the map). The ASV, Moffatt, Montgomery, PME, MLB, LB, GNC, NASB, and HCSB each include one of the longer readings in brackets. The AMP uses italics for the longer reading three times, and the other versions omit the longer readings: RV, Weymouth, AAT, William, NWT, JB, CPV, RSV, NEB, NIV, NJB, NAB, NRSV, REB, CEV, NCV, INC, GW, NLT1, NIrV, NET, ESV, TNIV, MSG, and NLT2.

Story embellishments

As a narrative is transmitted over the centuries, whether or not it is part of the Bible, little details and snippets of dialogue tend to find their way into the text. This is most prominent in the book of Acts, where the Western text (followed in English only by the Lamsa Bible) is ten percent longer than the Alexandrian or Byzantine versions of the book. Since these are usually innocuous details, there is no reason they would have been left out had they been original.

Darby joins Tyndale, KJV, Young, NKJV, KJ21, and LITV as fully Byzantine here, but the GEN has the Alexandrian reading at John 8:59. CW has four Byzantine renderings, Lamsa and TBV have three, Norlie and LB two, and the NIrV and AMP each have one. (The AMP also includes three Byzantine readings in italics, and MLB brackets the addition to Acts 28:16.) The majority of translations stay with the older, shorter readings: RV, ASV, Moffatt, Weymouth, Montgomery, AT, NWT, JB, CPV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NLT1, ISV, NET, ESV, TNIV, MSG, HCSB, and NLT2.

Stylistic changes

Some words or phrases are especially common in the Bible, or in the writings of a particular author. When the author then departs from that characteristic style, scribes may "correct" the text to bring it in line with other verses. This may also happen unintentionally, as certain ways of expressing things tend to become habitual.

As usual, Tyndale, GEN, KJV, Young, Lamsa, NKJV, KJ21, and LITV follow the Byzantine readings, but a surprising number of others join them. The TBV has four of the five Byzantine readings, Darby, Norlie, LB, and CW have three, Montgomery has two, and Moffatt, AMP, PME, NCV, and NLT1 have one apiece. That leaves the RV, ASV, Weymouth, AAT, William, NWT, JB, MLB, RSV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, GNT, INC, GW, NIrV, ISV, NET, ESV, TNIV, MSG, HCSB, and NLT2 with all Alexandrian readings. (Free translations reading dear friends in 1 John 2:7 are working from the Alexandrian text.)

Additional examples: Mark 6:16, 33; 7:16; 9:24, 44, 46; Luke 9:55-56; 11:29; John 1:27; 11:41; 16:16; Acts 23:9; 28:29

General deliberate changes

Scribes would also sometimes make additions to the text in order to explain or clarify it, or else to balance out a thought.

Translations fall along the usual lines in this category. Tyndale, GEN, KJV, Young, Lamsa, Norlie, NKJV, KJ21, and LITV are Byzantine four times. RV, Darby, Moffatt, ASV, Weymouth, AAT, NWT, JB, RSV, NEB, NIV, NJB, NAB, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NLT1, NIrV, ISV, NET, ESV, TNIV, MSG, HCSB, and NLT2 are Alexandrian four times. Of the remainder, CW has three Byzantine reading TBV has two, and Montgomery, PME, LB, and GNC are Byzantine at 1 John 4:19. AMP has one Byzantine reading in plain type and one in italics. MLB is Alexandrian but has brackets at 1 Cor. 7:39.

Additional examples: Matt. 12:35; 28:2; Mark 1:2; 13:15; Acts 17:26; 24:15; Romans 15:29; 1 Cor. 6:20; 11:29; Phil. 3:16; 1 Tim. 4:12; Rev. 6:1, 3, 5, 7; 16:17

Similarity to other passages

Those who copied the Scriptures had much of the Bible memorized and were thoroughly familiar with the Gospel accounts. This familiarity sometimes must have worked against them, for it appears that at times they inserted into one passage material that is found in another. This is most common with parallel passages in the Gospels. An example is the Lord's prayer, which is quite different in Luke than in Matthew (for both the Byzantine and Alexandrian manuscripts), but in a few manuscripts they are absolutely identical. Some of this appears to be deliberate harmonization, or else influence from a harmony of the Gospels produced by Tatian in the early centuries of the church. But whether accidental or deliberate, when material from one passage ends up in another, this is called an interpolation.

Most translations are cautious about interpolations. Only Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV have the longer reading all five times, Lamsa and CW have four, Norlie has three, and Darby and TBV have only two. The PME, LB, and ISV have the long version of 1 Cor. 11:24, and Moffatt is Byzantine at Luke 23:38. The rest are Alexandrian–RV, ASV, Weymouth, Montgomery, AAT, William, NWT, AMP, JB, MLB, CPV, RSV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NLT1, NIrV, NET, ESV, TNIV, MSG, HCSB, and NLT2–although AMP includes four of the Byzantine readings in italics.

Additional examples: Matt. 9:13; 17:21; 18:11; 19:9; 20:16, 22; 23:14; 25:13; Mark 6:11; 10:21; 11:26; 15:28; Luke 1:28; 4:8; 6:48; 9:35; 11:4; 22:64; 23:17; John 3:15; Rom. 8:1; 14:10; 16:24; Gal. 3:1; Eph. 5:30; Col. 1:14; 2 Tim. 1:11; Titus 1:4; 2 Pet. 2:17; 1 John 5:13; Rev. 1:8, 11; 2:13

Old Testament citations

Often the biblical writers would direct their readers back to an Old Testament passage by giving only a partial quote. Scribes familiar with the OT source must resist the temptation to finish out the quote automatically when the author leaves it incomplete. They generally did well, and most excesses lie only in a few manuscripts. But a few made it into the Majority Text preserved in the Byzantine tradition.

The TBV joins the usual list of Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV in having all Byzantine readings. (So might CW, except that it paraphrases Matt. 15:8 beyond recognition.) There are three in Lamsa and Norlie, and two each in the LB and ISV. MSG appears Alexandrian, but Rom. 9:28 and 10:15 are paraphrased beyond recognition. MLB is Alexandrian with two of the Byzantine readings in brackets. The rest take the older Alexandrian readings: RV, ASV, Weymouth, Moffatt, Montgomery, AAT, William, PME, NWT, JB, CPV, NEB, NIV, NJB, NAB, GNC, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NLT1, NIrV, NET, ESV, TNIV, HCSB, and NLT2.

Probable copyist's errors

In the last category are simple mistakes. Similarly drawn letters may confuse the scribe, or the eye may glance from one string of letters to a similar string, causing him to leave out or duplicate a line. If the text was being read aloud to the scribe, errors of hearing might result. Errors of this sort are usually obvious and therefore isolated, but some make enough sense that they slip in.

Darby, Lamsa, and TBV join Tyndale, GEN, KJV, Young, NKJV, KJ21, and LITV on three of the four readings. Norlie is Byzantine only at 1 Timothy 3:16, and GNC only at 1 John 4:19. CW is Alexandrian once, Byzantine once, and has major rewrites for the other two verses. Aside from the fact that LB's paraphrase of 2 Peter 3:9 removes the pronoun altogether, all the other versions have Alexandrian readings here: RV, ASV, Weymouth, Moffatt, Montgomery, AAT, William, PME, NWT, AMP, JB, MLB, CPV, RSV, LB, NEB, NIV, NJB, NAB, NRSV, REB, CEV, NCV, GNT, INC, NASB, GW, NLT1, NIrV, ISV, NET, ESV, TNIV, MSG, HCSB, and NLT2.

Additional examples: Heb. 10:34 (my bonds vs. those in bonds); 1 Peter 2:2 (unto salvation is missing in Byzantine versions)

How does the translation handle debatable variant readings?

Not all textual decisions are as simple as choosing one tradition or another. Often the textual evidence may be mixed, or internal probability and textual support seem to conflict with each other.

Difficult Old Testament texts

Decisions in the Old Testament can be particularly perplexing, because the manuscripts do not date as close to the time of composition. The Hebrew scribes were much more cautious about preserving the text than those who transmitted the New Testament, but some questions remain. Most versions follow the Masoretic Text, which is the standardized text in Hebrew, though they often consult Greek, Syriac, Latin, and other ancient versions for comparison.

Difficult texts in the Gospels

The "Western non-interpolations"

The Western manuscripts almost always have longer readings than the Alexandrian or Byzantine ones. However, there are several places near the end of Luke in which the Western text has a shorter reading; something is left out. Scholars are divided over whether the longer or shorter reading is original. On one hand, the combination of Majority support and Alexandrian support is almost always enough for certainty, but on the other, some see the material as interpolations from the other post-resurrection accounts that arose in the Alexandrian texts and made their way into the Byzantine without reaching the Western texts.

The verses involved are significant. They include the angel's words He is not here, but has been raised, Peter running to the tomb, Jesus' greeting, Peace to you, Him showing the disciples His wounds, and His ascension. All this material can be found in Matthew, John, and Acts. Having examined 24:3, 6, 12, 36, 40, 51-52, it is apparent that most translators accept the longer readings here. All of them appear in Tyndale, GEN, KJV, RV, Young, Darby, ASV, Weymouth, Lamsa, Norlie, LB, TBV, NKJV, NIV, NAB, CEV, NCV, GNT, LITV, NASB, GW, NLT1, NIrV, ISV, NET, ESV, TNIV, MSG, CW, HCSB, and NLT2. Five of the six are in Montgomery, PME, GNC, and NRSV. Moffatt accepts three and places three in brackets. However, the CPV and REB each contain only one (24:3), and AAT, William, RSV, and NEB choose the shorter reading each time. (The William uses brackets for four of the readings, and the MLB does so only at verse 12.)

Other New Testament texts

How does the translation handle the larger traditional but unlikely readings?

Grades (this category only)

Top 5: NET (highest), NLT2, NAB, MLB, CEV

A: TNIV, ESV, NRSV, HCSB, AAT, NWT, NIV, NIrV, NASB, NLT1, Weymouth, NJB

B: GW, GNT, RV, MSG, NCV, ISV, INC, Montgomery, AMP, ASV, Moffatt, REB, PME, NEB, CPV

C: JB, GNC, LB, William, RSV

D: Darby

F: TBV, Lamsa, Norlie, CW, Young, Tyndale

Bottom 5: GEN, LITV, NKJV, KJ21, KJV (lowest)



Translations Compared


 

For a concise statement of my beliefs about the Bible, see my Declaration of Faith.

 

Return to Realms of Faith.

Return to the MAIN PAGE.