Bibliotheca Sacra 150: 599 (1993): 273-284.
[Reproduced by permission]

[Bruce M. Metzger is Professor of New Testament Language and Literature, Emeritus, Princeton Theological Seminary, Princeton, New Jersey.]

[This is article three in the four-part series, "Translating the Bible: An Ongoing Task," delivered by the author as the W. H. Griffith Thomas Lectures at Dallas Theological Seminary, February 4-7, 1992.]

The work involved in making a translation of the Bible is both exhilarating and exhausting. It is exhilarating when translators consider the benefits, both spiritual and literary, that the rendering will provide to their readers; it is exhausting when they confront various problems, some of them beyond the possibility of solution. Problems involved in translating the Scriptures are many. Some result from the presence of variant readings among the manuscripts of the Old and New Testaments. Others have to do with the meaning of rare words as well as the uncertainty of punctuation of the Hebrew and the Greek texts. Still others relate to the appropriate renderings in English or any other receptor language and bear on the choice of the literary level and style of phraseology. This article considers examples of these kinds of problems.

Variant Readings among the Manuscripts

The first problem facing Bible translators is the differences in wording among manuscripts of the Scriptures. These differences have arisen because, even with the strongest determination to copy a text without error, a scribe copying a text of considerable length will almost inevitably introduce changes in the wording. It is understandable that mistakes can arise from inattentiveness brought on by weariness. For example instead of the correct reading, "Is a lamp brought in to be put under a bushel, or under a

[p.274]

bed, and not on a stand?" (Mark 4:21, RSV), several important manuscripts read "under the stand." This is obviously a scribal error in repeating the preposition "under" in the third phrase.

Sometimes a scribe's error of judgment works havoc with the text. One of the most atrocious blunders of this kind is in the minuscule Greek manuscript no. 109, dated to the 14th century. This manuscript of the four Gospels was transcribed from a copy that must have had Luke's genealogy of Jesus (3:23-38 ) in two columns of 28 lines in each. Instead of transcribing the text by following the columns in succession, the scribe of MS 109 copied the genealogy by following the lines across the two columns.

In addition to such transcriptional blunders, which can usually be detected and corrected, occasionally a scribe deliberately introduced into the copy a change that seems to clarify the sense or eliminate a difficulty. For example the older manuscripts of Mark 1:2-3 attribute to the Prophet Isaiah the evangelist's composite quotation from both Malachi and Isaiah, whereas later manuscripts (followed by the King James translators of 1611) read, "As it is written in the prophets," an obvious amelioration of the earlier text.

By comparing the surviving manuscript copies, scholars seek to determine what should be regarded as the original wording, and which reading or readings are secondary. Two kinds of considerations are taken into account. One concerns external evidence; this has to do with the age of the manuscripts that present the several different readings, as well as the geographical spread of the witnesses (and these include early versions in other languages) that support each reading. In general, the older manuscripts are, in the nature of the case, separated from the original text by fewer stages of copying and recopying than more recently copied manuscripts. Likewise, the more widespread the witnesses for a given reading, the more impressive is their testimony.

From considerations such as these one can appreciate why the discoveries in the 20th century of much earlier copies than those previously available are so important. Scrolls and fragments of each book of the Hebrew Bible, except Esther, hidden for centuries in caves by the Dead Sea, have brought to light manuscripts that are at least 900 years older than previously known copies. Similarly the acquisition of Greek papyrus manuscripts of various books of the New Testament now provides evidence for the wording of these texts that antedates what was previously available.

Besides external evidence, scholars also take into account what is called internal evidence of the variant readings. This is

[p.275]

of two kinds, involving transcriptional probability and intrinsic probability. Transcriptional considerations have to do with the habits of scribes. When a scribe was confronted with divergent wordings in two or more manuscripts, it was likely that, rather than choosing one and discarding the other, he would sometimes produce a composite reading that embodied both. In such cases the longer reading may be suspected as secondary. For example in the account concerning Stephen in Acts 6:8 some manuscripts describe him as "full of grace" and others as "full of faith." The sixth-century Greek and Latin manuscript of Acts known as Codex Laudianus (E) conflates the two and says that Stephen was "full of grace and faith."

In other cases scribes amplified and rounded off phrases by the addition of natural complements and similar adjuncts. A good example of a "growing" text is Galatians 6:17, where the earliest form of the text reads, "I carry the marks of Jesus branded on my body." In later centuries scribes expanded the simple and unadorned mention of "Jesus" with various additions, producing "the Lord Jesus," or "the Lord Jesus Christ," or "our Lord Jesus Christ."

Intrinsic probability has to do with considerations of what the author is likely to have written. Naturally attention should be given to such considerations only after all other kinds of evidence have been canvassed and evaluated. At that stage, one is in a position to test the validity of tentative conclusions as to the original reading. If a reading is contrary to the immediate context and/or is out of harmony with the usage of the author elsewhere, serious doubt is cast on the originality of that reading, despite the weight of the external evidence. In some cases, therefore, opinions will differ on the original wording.

Obviously all such decisions as to textual variants have been made by editors of the original texts, and translators generally depend on the expertise of those who have produced the printed editions of the Hebrew Scriptures and the Greek New Testament. The most widely used printed editions at the end of the 20th century are the Biblia Hebraica Stuttgartensia (1977; ed. sec. emendata, 1983) and the United Bible Societies' Greek New Testament, prepared by an interconfessional and international committee (1966; third edition corrected 1983; a fourth edition is shortly to appear). Translators, however, may give further consideration to the evaluation of textual evidence and occasionally will adopt a reading different from that in the printed text. In 1 Thessalonians 2:7 Greek manuscripts are divided; some read "gentle," some "infants." The difference in Greek is only one letter,

[p.276]

and . The last letter of the previous word in the Greek is "," and so no one can say whether scribes wrote "n" twice instead of once, or once instead of twice. "Infants" is found in many good manuscripts (and is printed in the United Bible Societies' Greek New Testament), but "gentle" makes better sense,[1] and is adopted by most translators (RSV, NIV, REB, NRSV) who otherwise follow the UBS text.

The Meaning of Hebrew and Greek Words

After translators have decided which wording of the Hebrew or Greek text should be taken as the basis of the English rendering, the next problem has to do with ascertaining the precise meaning of the words. Lexicographers are constantly attempting to learn more exactly the meaning of ancient Hebrew and Greek terms and expressions. The Hebrew Bible contains about 300,000 words, comprising 8,674 different words, of which, according to one method of calculation, about 1,500 occur only once.[2] In many cases a similar word occurs in the literature of other Semitic peoples, notably in Arabic, Assyrian, Eblaic, and Ugaritic. By comparative linguistics and archaeological finds scholars are able in some cases to define more precisely the meaning of rare Hebrew words. One such Hebrew word, which has never been found in other Semitic literature, is (1 Sam 13:21). Because of the context the King James translators took this word to mean "a file," used by blacksmiths to sharpen hoes and other agricultural tools. In the first part of the 20th century, however, archaeologists discovered at various places in Palestine ancient sets of weights used for business transactions, each bearing a Hebrew word. One of these, weighing almost two and two-thirds ounces, is marked and so translators now know this was the amount that the blacksmiths charged for sharpening various tools.

Even when the meaning of individual Hebrew words can be determined with a degree of certainty, there is sometimes also the problem as to how they are to be understood in relation to each other in the sentence. What has been called the most obscure verse in the Book of Proverbs (26:10 ) involves a whole nest of problems. Many combinations of the words have been made, along with

[p.277]

attempts at emendation. Because the meaning of more than one word in the verse is subject to various interpretations, at least 10 different translations of it have been proposed.[3]

In comparison with the difficulties of ascertaining the meaning of words in the Hebrew Bible, the Greek New Testament is much easier, at least with regard to the number of lexical problems. According to statistics collected by an assiduous researcher, the Greek New Testament contains 137,328 words, comprising a total of 5,436 different words, of which 1,934 occur only once.[4] The great majority of these hapax legomena occur also in other Greek sources,[5] and so the meaning of most of them is not often in dispute. The meaning, however, of a word in the Lord's Prayer as recorded in Matthew 6:11 and Luke 11:3 has often been debated. Does "Give us this day our bread" mean "daily bread" or "bread for tomorrow"? Except in subsequent quotations of the prayer, no other piece of Greek literature is known to contain this word. The only time it seems to have turned up was in 1889 when A. H. Sayce edited a fragmentary Greek papyrus containing a householder's account-book listing the purchase of provisions. Here, according to Sayce, in one of the broken lines of the list was - , with the end of the word defaced. It is most unfortunate, however, that scholars who wish to double-check this information are unable to do so, for the papyrus fragment has disappeared and cannot be found. Furthermore its loss is particularly distressing because Sayce (whose shortcomings as a decipherer of Greek papyri were generally recognized) may have misread the householder's list.[6] And in any case, even if Sayce did correctly read the word, lexicographers do not know much more about its meaning than was known before, namely, that the expression signifies either "daily bread" or "bread for tomorrow." In such cases when a word is susceptible of two equally

[p.278]

legitimate renderings, translators have no choice except to place one in the text and the other in a footnote.

The Punctuation of the Text

Once translators have decided which form of text to translate and what the Hebrew and Greek words mean, the problem of punctuation arises. In antiquity it was customary to write Hebrew and Greek manuscripts with few, if any, marks of punctuation. The beginning of a sentence was not identified by a capital letter. Not until the eighth or ninth century A.D. did Greek scribes begin to be more or less systematic in the use of punctuation marks.[7] Though exegetes can learn something concerning the history of the interpretation of a passage by considering the punctuation in the manuscripts, translators need not feel bound to adopt the punctuation preferred by either the scribe or the editor of the printed text. Furthermore, since there are no quotation marks in any of the manuscripts, the decision of where to insert these in the translation is totally in the hands of the translators.

Naturally the opinions of translators as to appropriate punctuation will sometimes differ. There is no infallible rule to follow; judgments must be based on what seems to provide the fullest and most appropriate sense in the context. The beginning of a direct quotation can usually be determined without any trouble when it is indicated by a verb such as "said," "asked," "replied," or the like. But problems can arise concerning the close of a quotation, especially when it is the final sentence of a series of comments of a conversation. It is uncertain, for example, whether the last statement made by Jesus to Nicodemus is intended to end at John 3:15 (so the RSV) or at 3:21 (so the NIV and NRSV).

The position of a comma within a sentence can totally alter the sense. In Revelation 5:1 the traditional punctuation describes the scroll held in the right hand of God as "written on the inside and on the back, sealed with seven seals." The Greek text, however, may also be understood in a different way, resulting in the translation given in the NRSV footnote on this verse, "written on the inside, and sealed on the back with seven seals."

Changing the position of a comma can sometimes expand the sense. The third petition in the Lord's Prayer in the King James Version reads, "Thy will be done in earth, as it is in heaven" (Matt 6:10), whereas most modern versions punctuate it differently, "Thy will be done, on earth as in heaven." The principle translators follow is to use the punctuation that provides the best and fullest sense. In this case the second way of punctuating is better, for it permits the phrase "on earth as in heaven" to be taken with all three preceding petitions, thus enlarging the scope and meaning of the prayer.

A theological point is involved in the placing of a comma in Luke 23:43. According to the traditional way of understanding the passage, the repentant robber asked Jesus on the cross to remember him when Jesus entered His kingdom. To this request Jesus responded, "Truly I say to you, today you will be with Me in paradise." In the interest of supporting the doctrine of "soul sleep" held by Jehovah's Witnesses, the translators of the New World Translation of the Christian Greek Scriptures have moved the comma so that the verse reads, "Truly I tell you today, You will be with me in Paradise." But surely here the robber knew that Jesus was speaking to him that day, and so the correct punctuation is that of traditional translations.

Sometimes a sentence in the Greek New Testament can be construed as either a statement or as a question. This ambiguity accounts for the change at Romans 8:33 between the RSV, "Is it Christ Jesus, who died…" and the NRSV, "It is Christ Jesus, who died…" (the latter returns to the interpretation of the King James translators). At Mark 15:2, in response to Pilate's question, "Are you the King of the Jews?" Jesus answered with a statement, "You say so." It is possible, however, to understand the Greek here as a question, "Do you say so?"[8]

Modern translators occasionally find that an exclamation mark brings out most appropriately the force of the original. The awe and wonder of the scene described in Revelation 4:1-2 is then disclosed in the RSV: "After this I looked, and lo, in heaven an open door! …At once I was in the Spirit, and lo, a throne stood in heaven, with one seated on the throne!" In the Old Testament, particularly in the Psalms, the RSV translators were overzealous in their use of exclamation marks, and in the NRSV many of them have been replaced with a period (as in the King James Version).

To Translate or to Transliterate?

Yet another problem confronting translators arises when a Hebrew or a Greek word can be either translated or transliterated. What should be done with proper names that can also be used as common nouns? For example, is both a common noun

[p.280]

meaning "a man" and the proper name "Adam." In translating Genesis the question soon arises at what point in the narrative one should begin to use "Adam" rather than "man." On this matter there has been wide disagreement among translators. Some versions make the change at Genesis 2:7 (Targum), others at 2:16 (Septuagint), and still others at 2:19 (KJV), 2:20 (NIV), 3:17 (RSV), 3:21 (NEB), or 4:25 (NAB, REB, NRSV).

The traditional rendering of Psalm 84:6 in the King James Version is, "Who passing through the valley of Baca make it a well." Since, however, means "a balsam tree," the New Jerusalem Bible translates it, "As they pass through the Valley of the Balsam," and the New American Bible has, "When they pass through the valley of mastic trees." Since, however, the similarly pronounced word means "to weep," the American Standard Version reads, "Passing through the valley of Weeping."

A similar problem in the New Testament concerns the Greek word , which can be transliterated "Christ" or translated "anointed one," or "Messiah." Several modern translations (e.g., NEB, NAB, NRSV) have replaced most occurrences of "Christ" in earlier renderings of the Gospels (e.g., KJV, ASV, NASV) with "Messiah." The reason for making the change arises from the recognition that it was only after the message of the early followers of Jesus was addressed to Gentiles that the word as a title (Jesus the Messiah) would come to be understood as a proper name (Jesus Christ). The transfer of understanding was total when, still later, the expression Christ Jesus is sometimes used in the Epistles and in the Book of Revelation.

The word "Hades" in Greek () was originally a proper noun, the name of the god of the underworld. In time the word came to denote a place or state, and in the King James Version it is usually rendered "hell," and once "grave" (1 Cor 15:55). In the RSV the word is usually transliterated, but in Matthew 16:18 it is rendered "[powers of] death," where the NRSV transliterates.

Besides proper names other words are sometimes translated and sometimes transliterated. The Greek verb has traditionally been transliterated "baptize." About 1885 the American Baptist Publication Society of Philadelphia issued the New Testament in two forms, one that used the traditional rendering, "baptize" and the other that translated the verb with the word "immerse." "John the Baptist" became "John the immerser."

The Greek words and , usually translated "elder" and "council of elders," can also be transliterated "presbyter" and "presbytery." The Greek word means "overseer" but is often transliterated (through the

[p.281]

Old English "bisceop") as "bishop." Also , which means "servant," is transliterated "deacon."

Stylistic Problems

Since the Bible is a source of both information and inspiration, translations must be both accurate and esthetically felicitous. They should be suitable for rapid reading and for detailed study, as well as suitable for reading aloud to large and small groups. Ideally they should be intelligible and even inviting to readers of all ages, of all degrees of education, and of almost all levels of intelligence - all without sacrificing accuracy, in either matter or manner. Besides the several problems already considered as to text, meaning of words, punctuation, and the like, the following are illustrations of some of the more delicate stylistic problems that confront Bible translators.

1. Not only the choice of English words but also the order in which they are arranged often makes a difference in meaning. In the words of the institution of the Lord's Supper, the rendering in the King James Version, "Drink ye all of it" (Matt 26:27), leaves it uncertain whether Jesus meant all who drink or all of the contents of the cup. Since the Greek text here uses the plural form of the word "all," the English translation should be something like, "Drink from it, all of you."

Although E. J. Goodspeed's translation of the New Testament (1923) usually employs American idioms, here and there one finds curious slips in sentence arrangement. Hebrews 10:1 reads, "The same sacrifices…cannot wholly free those who come to worship from their sins." In Hebrews 9, where Goodspeed uses "chest" and "agreement" in place of "ark" and "covenant," verse 4 reads, "the ark that contained the agreement, entirely covered with gold." The ark, not the covenant, was gold-covered.

The New Revised Standard Version corrects several misleading RSV renderings. Instead of Moses leaving "Pharaoh in hot anger" (Exod 11:8), it now reads "in hot anger he left Pharaoh," and instead of "Joshua was standing before the angel, clothed in filthy garments" (Zech 3:3), the NRSV reads, "Joshua was dressed with filthy clothes as he stood before the angel."

2. Translators must pay attention to what can be called the color or tone of their rendering. For example, though the verbs "to dwell" and "to live" are more or less synonymous, translators need to be sensitive to the context in which one word is more appropriate than the other. Translators generally agree that "dwell" is to be preferred in contexts that speak of God in heaven, such as the traditional rendering of Isaiah 57:15 (which is retained in the

[p.282]

NRSV), "I dwell in the high and holy place." On the other hand the word "live" is certainly more appropriate in matter-of-fact statements, such as "Jabal…the ancestor of those who live in tents" (Gen 4:20, NRSV), where earlier versions continued with the King James rendering "dwell."

3. Care must be taken in choosing words that are susceptible of being understood in the wrong way. Modern English versions avoid the King James rendering of Matthew 20:17, which says that Jesus "took the twelve apostles apart in the way." Though James Moffatt struck off many happy phrases in his translation, occasionally one finds an ambiguous rendering. The wording in his 1913 translation spoke of two men in one bed (Luke 17:34), but his 1934 revision reads "two men in bed" (i.e., not a double bed). The RSV in 1 Kings 19:21 says of Elisha, "Then he arose and went after Elijah"; this is modified in the NRSV to read, "Then he set out and followed Elijah." The earlier rendering of Psalm 50:9, "I will accept no bull from your house," is altered to read in the NRSV, "I will not accept a bull from your house."

Also under the category of words that can be misunderstood are homophones, that is, words that have the same sound but differ in spelling and meaning, such as "there" and "their." To prevent possible ambiguity during oral reading, the statement "because there God had revealed himself" (Gen 35:7, RSV) was altered in the NRSV to "Because it was there that God had revealed himself." Another kind of oral ambiguity can arise when one hears Luke 22:35 read aloud: "'Did you lack anything?' They said, "'Nothing.'" The NRSV renders the second sentence, "They said, 'No, not a thing'" to prevent hearers from thinking the sentence read, "They said nothing."

4. The maxim of the committee that produced the New Revised Standard Version is that the version was to be "as literal as possible, as free as necessary." Though, as expected, there would be differences among the members as to when to reject a literal rendering, they agreed that expressions that reflected ancient ideas of psychology should be replaced by modern terms. Both the Old and New Testaments contain references to one's kidneys as the seat of affections and emotions. Because the King James translators used the older English word "reins," which meant kidneys, most readers of that translation today have no idea or, at any rate, a wrong idea of the meaning of such passages as, "My reins also instruct me in the night seasons" (Ps 16:7), or "I am he which searcheth the reins and the hearts" (Rev 2:23). In present-day English the equivalent is "heart" or "mind." The King James literal rendering of Philippians 2:1, "any bowels and

[p.283]

mercies," does not convey the idea intended by the original text. Modern translators employ a variety of equivalent terms, such as "warmth or sympathy" (NJB), "kindness and compassion" (GNB), "warmth of affection or compassion" (REB), "compassion and sympathy" (NRSV), "tenderness and compassion" (NIV).

5. In recent years yet another problem has begun to confront those who translate the Bible into English, namely, the question of the suitability of using masculine-oriented language in passages that obviously apply to men and women alike. The movement for women's "liberation," with its occasional extravagances, has made people conscious as never before of deficiencies in the way humans speak of each other. Many publishers, as well as church educational boards, now issue guidelines as to how best to express oneself in "inclusive" language. No doubt such concerns will not go away, and translators of the Scriptures obviously do not wish to offend and put off readers by using what is increasingly coming to be regarded as unacceptable English.

The problems that are easiest to correct are, of course, those passages where earlier translators inserted the word "man" or "men" but where the Hebrew or Greek text lacks such a term. The traditional rendering of Jesus' words in John 12:32 is, "And I, when I am lifted up from the earth, will draw all men to myself." Here the King James translators inserted the word "men," italicizing it to indicate (as they were accustomed to do) that it is not in the Greek. In order, however, to show in modern English usage that the passage does not intend to limit the reference to male adults only, translators have rendered the passage either "draw everyone to myself" (NAB, 2d ed., and REB) or "draw all people to myself" (NJB and NRSV).

According to the King James Version, at the wedding feast held at Cana of Galilee, the comment was made, "Every man at the beginning doth set forth good wine; and when men have well drunk, then that which is worse" (John 2:10). Here the words "man" and "men" do not appear in the Greek text, nor are they italicized in the translation. The REB and NRSV have independently of each other avoided the masculine bias of the King James Version by using "everyone" instead of "every man" and "guests" instead of "men."

Somewhat more difficult to assess are the passages that do contain the Hebrew or Greek word for "man" ( or ) but where it would be wrong to understand the passage as restricted to adult males. For example, "Man shall not live by bread alone" (Deut 8:3, quoted in Matt 4:4 and Luke 4:4) is rendered in the NRSV, "One does not live by bread alone."

[p.284]

When the apostle referred to the work of the Holy Spirit in strengthening "the inner man" (Eph 3:16), what should translators do? Should they assume that the expression "inner man" is a stereotyped phrase that would be understood by women? Or is it better to use the expression "your inner being" (NIV and NRSV) or even to replace "strengthen" with the noun "strength" and render the phrase "grant you inward strength"?

Of course in other passages the word "man" or "men" must remain in English. One recognizes that the congregation in most Jewish synagogues in antiquity consisted exclusively of men. Furthermore the presence of the Greek word in Mark's account of the number of those who had eaten at the feeding of the 5,000 (Mark 6:44) must be rendered "men."

A recurring difficulty facing translators is the lack of a common gender third person singular pronoun in English. It is ungrammatical to say, "everyone must bear their own burden," and it is restrictive to say, "everyone must bear his own burden," but it would be cumbersome to say, "everyone must bear his or her own burden." In such cases the NRSV translators considered that the least unsatisfactory solution was to represent the meaning by pluralizing, "All must carry their own burden."

6. Several problems are virtually impossible to resolve. How should poetry be translated? To turn Hebrew poetry into prose has been compared to playing on a violin a score written for the organ.

Plays on words in Hebrew and Greek are especially difficult to handle. Frequently the only solution is to supply explanatory footnotes. At Jeremiah 1:11-12 the NIV adds the note, "The Hebrew for watching sounds like the Hebrew for almond tree," and the RSV provides in notes the transliteration of the two words in question, "Heb shaqed" and "Heb shoqed." The Greek name Onesimus means "useful," to which Paul alluded in Philemon 9-10 . This is handled gracefully in Weymouth's rendering, "Formerly he was useless to you, but now - true to his name - he is of great use to you and to me."

The presence of an acrostic format in such passages as Psalm 119 and Lamentations 1-4 is the despair of many translators. Ronald Knox, however, was no ordinary translator and he managed to present in English the equivalent kind of structure. To take Lamentations 4:1-10 as a specimen, the opening word or words of Knox's rendering are as follows: "All dim…. Bright…. Cub…. Dry throat…. Even they feared…. Faithless Juda…. Gone…. Here…. It were better…. Juda brought low…" and so forth.

References

[1] Two of the five members of the United Bible Societies committee (Allen Wikgren and the present writer) have expressed their preference for the reading "gentle" (Bruce M. Metzger, ed., A Textual Commentary on the Greek New Testament [London: United Bible Societies, 1971], 630).

[2] For a discussion of such words, see Frederick E. Greenspahn, Hapax Legomena in Biblical Hebrew (Chico, CA: Society of Biblical Literature, 1984).

[3] See the list in C. H. Toy, A Critical and Exegetical Commentary on the Book of Proverbs (Edinburgh: T. & T. Clark, 1899), 476, note. For a more recent assessment of the verse, see D. C. Snell, "The Most Obscure Verse in Proverbs," Vetus Testamentum 41 (1991): 350-56.

[4] See Robert Morgenthaler, Statistik des Neutestamentlichen Wortschatzes (Zürich: Gotthelf-Verlag, 1958), 25.

[5] According to information kindly supplied by Frederick W. Danker (letter, December 10, 1989), only two dozen words (not including proper names) have not been found elsewhere.

[6] For further information see the chapter entitled, "How Many Times Does Occur outside the Lords Prayer?" in Historical and Literary Studies, ed. Bruce M. Metzger (Leiden: Brill, 1968), 64-66. After this chapter was published a search has been made in the Thesaurus Linguae Graecae but without discovering any other occurrence of the word.

[7] See Bruce M. Metzger, Manuscripts of the Greek Bible: An Introduction to Greek Palaeography (New York: Oxford University Press, 1981), 31-32.

[8] This punctuation is given by Westcott and Hort in the margin of their edition (1881).