Text and Manuscripts of the Old Testament


The present situation

Fifty years ago research into the history of the OT text seemed to have reached a dead end in two important directions. On the one hand, the collations that Kennicott and De Rossi had published in the latter part of the 18th cent. with their comparison of the readings of hundreds of MSS had shown extremely few variations of any importance in the consonantal text of the OT. It was therefore felt that one could be quite certain as to the precise text of the OT as it existed a thousand years ago. On the other hand, there seemed to be no way of tracing it with much certainty to an earlier period. Nearly all of the extant MSS were written after a.d. 1100. Very few could be safely attributed to an earlier date, and none to a date before c. 900. The history of the Heb. text previous to that time seemed to be largely a matter of conjecture, with little reason to hope that further light would ever be thrown upon it.

This situation has now greatly changed. The discovery of the DSS in 1947 has made available a great corpus of ancient material that can throw light on the history of the OT text. Equally important is the availability of other new material, some of which had been discovered prior to that time, but had as yet been comparatively little studied. A great number of OT MSS had been collected in the 19th cent. through the efforts of Abraham Firkovitch and were available in the Leningrad library, but results from their study were little known to the western world. Nearly 200,000 fragments of Heb. and Aram. MSS of many kinds had been taken from a Cairo Geniza to museums and libraries in the W, but the study of these materials had been little more than begun.

Another source of new information, not previously available, is the OT codex that was long treasured by the Sephardic synagogue in Aleppo. Fifty years ago many scholars were convinced that this MS had actually been written by the famous Masorete, Aaron Ben Asher, and that its text was therefore the most important evidence for the true MT. Yet it was not available for study, since its keepers steadily refused to allow it to be photographed or carefully researched. After the burning of the Aleppo synagogue in the riots of 1948, it was feared that this irreplaceable codex had been destroyed. Later on it was learned that about three-fourths of it had been rescued and taken to Jerusalem, where it is now being intensively studied.

A brief survey of the history of the Hebrew text

The period between the writing of the books and the destruction of Jerusalem in A.D. 70.

Until the discovery of the DSS in 1947 the only direct evidence about this period was the little that might be gleaned from a comparison with the Samaritan Pentateuch or with the LXX. This will be discussed below under “Versions.” At this place it will be sufficient to point out that we are here dealing, not with the history of ordinary books, but with a very special situation. It is the belief of Christians that these books were considered as sacred books from the very time of their writing. God so inspired the writers as to keep them from error in what they wrote. He selected men whose background, experience, and personality would fit them for preparing the sort of book that He desired. Then He guided these men in their selection of material; He revealed to them certain new facts and ideas; and He so directed their activity that in their selection of words from their own vocabularies they should be kept free from saying anything that would involve errors of fact, of doctrine, or of judgment. According to the Christian view these books were given by their authors to the people of God as divine writings which should be carefully guarded and intensively studied.

This is indeed the view that is contained in the Pentateuch, which says that Moses ordered that a copy of its contents should be preserved in the very Holy of Holies (Deut 31:26), and that another copy should be constantly before the king in order that he might study it carefully and direct his activities in accordance with it (17:18, 19).

Since the books were regarded as possessing such a transcendent importance, they would have been preserved with great care. This does not mean that no errors could possibly enter their text. The Bible contains no promise of absolute accuracy in text preservation, as the books would be copied and recopied through the centuries. It is impossible for a human being to copy any book without making at least a few errors. Even the best of proofreaders may overlook some of these mistakes, and a few may have found their way into the official copies preserved by the heads of the nation. Yet there can be no doubt that in these official copies, prepared and checked with extreme diligence, errors would be kept to a minimum.

It would be somewhat different in the case of the many copies made for individuals and distributed widely throughout the land. While there is no direct contemporary evidence about this, the discovery at Khirbet Qumran (excavated 1951-1956) in the neighborhood of the Dead Sea where a room was laid out as a Scriptorium in which copies were constanly being made for the use of the members of this ascetic colony, shows how widely the copies of the sacred Scriptures must have been disseminated during the centuries immediately before the time of Christ. Most likely the Scriptures were similarly distributed during the preceding centuries. It would often be expensive to make copies, and the people in a distant part of the land, instead of going to Jerusalem and paying to have a copy made directly from the official text, might sometimes have to content themselves with a copy made from the one possessed by the local leaders. Thus errors would inevitably creep into these copies in the course of time and eventually it would be natural that schools of MSS would develop, as occurred later in the case of the NT.

Previous to the discoveries at Qumran there was no Heb. evidence that such a development had occurred, but the variations in the text of the LXX and in the Samaritan Pentateuch made it appear probable that such text families did exist in various parts of the country, even though the errors in the official text preserved at Jerusalem would doubtless be kept to an extreme minimum. This question will be further discussed below (see IX. 3).

The period from the destruction of Jerusalem to about A.D. 900.

During this period renewed emphasis was placed upon the Scriptures. With the loss of the Temple and the capital city, the Jews might have entirely lost their identity, had it not been for the great stress laid upon their religious unity, and upon the OT as the source from which it flowed. Groups of rabbis gathered in various sections of Palestine to discuss problems connected with the OT and to arrive at conclusions that they could defend in their relations both with other Jews and with outsiders. One of their primary aims was to preserve the integrity of the sacred books. During the previous centuries and during a substantial part of this period the men who were particularly concerned with this task were generally referred to as סוֹפְרִים, “scribes.” Eventually the specialists in this work came to be designated by another name בַּעֲלֵי הַמַּסּוֹרֶת, “masters of tradition,” or simply Masoretes.

Akiba, one of the leading rabbis at the beginning of this period, had stressed the importance of using tradition as “a fence about the law” to protect its integrity. The scribes undertook to do this by counting the number of letters, the number of words and the number of vv. in each section, marking the middle letter and the middle word of each, and noting peculiar forms or other facts related to this purpose. There is very little contemporary evidence about their precise activities, although some of the discussions of the rabbis have been reported in the Talmud. Many of the marks that the scribes inserted at various places in the Scripture, and some of their marginal notes were included in the later works of the Masoretes, even though the meaning and purpose of some of them had been forgotten by that time. These activities will be discussed more fully below.

It is not known when the term Masorete began to be used, but by a.d. 800, the men who were particularly devoting themselves to the care of the Scriptures were called by this name, instead of scribes. They had many problems with which to deal, one of the principal ones being that of indicating the exact pronunciation of the words, and the way they should be read in the services, particularly since originally only the consonants were written. Especially between 800 and 900 a great deal of effort was expended on this task, and the result attained received such general acceptance that most of the previous MSS were quickly superseded. Since the Masoretes had done their work so well, after c. 900 this title was no longer applied to individuals, and the Heb. text came to be designated as the MT.

In later years the scholars who gave very particular attention to studying the work of the Masoretes, and to endeavoring to continue preservation of the integrity of the text, were called “grammarians” or, more generally, naqdanim, “punctuators.” Many new MSS of the MT were prepared in the following centuries, but in their essential feature, that of the consonantal text, they were remarkably uniform, even though written in widely separated portions of the earth. Students of the OT text and MSS have held various theories as to the relation of the MT to the Scriptures as originally written. Various facets of this problem will be discussed below.

Consonantal writing and indication of vowels

All writing is a rather imperfect means of representing human words and statements. Oral expression contains many features that are not conveyed in writing. Modern languages have introduced punctuation marks, which convey some further idea of the expression that the voice gives. Such marks were unknown in Heb. antiquity. In addition, ancient Sem. writing represented somewhat less of the idea than is done in the writing of modern languages, since in most instances (except in Cuneiform) only consonants were written and there was little or no indication of the vowels. This was not as much of a detriment as it would be in most of the modern languages, since in Sem. languages a root generally consists only of consonants, the function of the vowels being to make the consonants pronounceable and to convey ideas as to the form of speech, the tense, the mood, or some other relationship of the word. However, even Indo-European languages would usually not be difficult to read if only the consonants were written. Thus the sentence, “Have you read this book?” would be understood quite easily if only the consonants were written, as follows: “Hv y rd ths bk?” Similarily it would not be difficult to understand the sentence, “Wl y rd ths bk?” The Eng. reader will inevitably pronounce “rd” in the first of these sentences in one way, and in the second in quite a different way. Context will often show how to pronounce words, or how to interpret them. This is true even when the vowels are written out, as in modern Eng. In both of these sentences, if written fully, the word “read” would be spelled identically, but the reader would pronounce it differently with scarcely a thought of the reason for so doing.

As time goes on, all languages develop and change. Certain consonants may cease to be pronounced at all, though still written as silent letters. In Heb. this occurred in the case of the waw in such words as sws, prob. originally pronounced sawas but eventually contracted to sūs. Similar developments occurred with other letters, and in such cases these letters came to be thought of, not as indicating a consonant, but as representing a long vowel of one type or another. Eventually the practice began of inserting these “vowel letters” to indicate a long vowel wherever it seemed that the word might not be correctly understood without it, even when no contraction was involved. In later Heb. writing considerable flexibility developed as to the insertion of vowel letters, with the result that there are great numbers of variations in this feature in the various MSS. However, most of these variations have no more effect on the meaning than the question whether one writes “honor” as is common in America, or “honour” in accordance with British usage.

After the exile Heb. was gradually displaced by Aram., and eventually came to be used only for religious and literary purposes. It was still used extensively in the synagogue services, and the OT books were also read at home in Heb. Children frequently heard the text read, and there was a tendency to retain orally a tradition as to which vowels would be used at certain places. During the centuries the pronunciation of vowels and consonants naturally changed, often being affected by the particular language used in common speech, whether Aram., Gr., or Arab. Eventually the custodians of the sacred writings began to realize the necessity of finding some better way to indicate precise vowel pronunciation than merely the presence or absence of vowel letters. At the centers of Jewish learning in Babylonia, a system was invented of placing dots or other marks over certain letters to indicate the following vowel. A different, but somewhat similar system was developed in Palestine. Eventually a third system was developed at Tiberias in Pal. which largely substituted marks under the consonants for the previous system of designating vowels by marks above them. This system soon became dominant, and was used exclusively in later MSS and in printed Heb.

Types of MSS

Hebrew MSS are of two types: those for synagogue use, and those for the use of private persons. Synagogue MSS are sometimes so made as to contain simply the portions of the OT selected for public use in the regular worship of the synagogue. The law (the five books of Moses) may be included in one MS, since it is read consecutively Sabbath after Sabbath. In connection with the weekly readings from the law, it became customary to read certain appropriate passages from the second division of the Heb. Bible, called the Haphtaroth, which were selected at a very early time. These selections are sometimes copied together in one scroll. The Book of Esther, which is read at the feast of Purim, and the four other books that are read at special fast or feast days, are placed on separate scrolls and are called Megilloth (scrolls).

The Talmud includes very minute rules for the making of synagogue MSS. They are always in the form of scrolls rather than of codices (like our modern books). Parchment, prepared from the skin of a clean animal, must be used. The text must be written in black erasable ink, without vowels or accents, and must be copied with great care. Even a single letter may not be written without the scribe looking first at the MS being copied. Extraordinary points and any letters of unusual size, position or form must be carefully copied. The MS must be corrected within thirty days after it is written. If four errors are found on any page it must be condemned.

Very few of these synagogue scrolls have become available for scholarly study, since, to avoid the risk of desecration, it was required that they be buried as soon as they had become too worn for regular use.

Private MSS were generally made for study by individuals or for use in the home. Being made by hand, they tended to be rather expensive. Well-to-do people could afford to hire excellent scribes and to have the MS checked with great care. Other MSS may have been copied somewhat more hurriedly. Since according to Jewish law every Jew was expected to have at least one copy of the law in his home, the number of private MSS prepared was far greater than the number of synagogue scrolls.

The private MSS sometimes contained the whole OT but more commonly only a part, and sometimes only a single book. Although occasionally in the form of scrolls, they were usually codices of various sizes. Sometimes they were written on parchment or prepared skins, but more frequently on cotton or linen paper, and generally with black ink. Vowel marks and accents were included. The upper and lower margins and also the side margins often contained Masoretic notes and various readings. Sometimes a commentary by a noted rabbi was placed beside the text. Often a tr. was included, either into Aram. (a Targum), into Arab., or into some other language. Usually the consonantal letters were all written first, the vowels and accents being added subsequently, sometimes by a different person, and generally with a different pen and different ink. Frequently a MS would pass through several hands in the course of preparation. One person might write the consonants, another might append the vowels, a third might go through to correct it, a fourth would add the Masora, a fifth might retouch it after it had become defaced by age or use. Often the initial words or letters were ornamented, and the margin was decorated with pictures of flowers, trees, or animals. The MS generally contained a colophon at the end, telling who performed these functions and giving other information about the MS.

It is often difficult to determine the age of a Heb. MS. Since the MSS were produced in many different areas, and the type of writing often varied from place to place, determination of age by paleography (changes in styles of writing) is generally difficult.

Sometimes the colophon contains a statement of the time when the MS was made, but this is often missing, and when it is included it is not always easy to understand. The years may be reckoned from the creation, from the destruction of the second Temple, from the hegira of Mohammed, or from the era of the Seleucidae (312 b.c.), this last being particularly common. Often the number of thousands of years is omitted, and sometimes even the number of hundreds. In some cases the colophon may have been simply copied from an earlier MS. Fortunately, some MSS give the date according to two or three different eras, and this makes possible a helpful check as to the time that is meant.

The colophon of the Leningrad Codex B 19a says that this MS was prepared (a) in the year 4770 from the creation of the world, (b) in the year 1444 from the exile of King Jehoiacin, (c) in the year 319 “of the empire of the Greeks,” (d) in the year 940 from the destruction of the second Temple, and (e) in the year 399 “of the rule of the little horn.”

According to a common Jewish reckoning of the time of creation the first of these dates would indicate a.d. 1010. The third date, if considered as omitting the figure for 1000, would come to a.d. 1008 (1319 years after 312 b.c.). The fourth statement would indicate 1009. The fifth date, considering the rule of the little horn as referring to the beginning of Islam, would be 1008, according to the Muslim system of dating. The reckoning from the exile of King Jehoiachin does not reach a date comparable to the others. This prob. results from the mistaken statements in the Targum that the whole period of Pers. rule (539-331 b.c.) was really only a single generation. The coincidence of four of the statements seems to make the date of a.d. 1008 quite reasonable for this important MS (the one upon which the text of Biblica Hebraica 3 is based), particularly since it was written in a Mohammedan land.

Divisions of the text

The convenient arrangement of modern Bibles, with chs. and vv. numbered consecutively, was not introduced until the 16th cent. Originally the books of the OT were written without subdivisions and usually without titles. In ancient times it was often the custom simply to name a book by its first words.

The first division made in the Heb. text was prob. into unnumbered verses. This may have been done very early. Verse divisions in recently discovered MSS written shortly before the time of Christ, are almost identical with those in our present Heb. Bible. However, no numbering is indicated and there were no ch. divisions. While in general the verse division was well made this was not always the case. Some vv. contain two separate sentences while others include only a small part of one sentence. The division is rarely as bad as in Psalm 19:4, which includes the last few words of one sentence and the first few words of another.

The next step was the division of all the OT books except the Psalms into 452 sections called parashas or sedarim. There were 154 of these in the Pentateuch. It is said that the Jews in Pal. used to read each of these consecutively in the weekly synagogue services, taking about three years to go through the entire Pentateuch, until the coming of the exiles from Spain at the end of the 15th cent. resulted in changing the custom so as to read the Pentateuch through in one year.

Some of these parasha were felt to mark a more important division in the text than others. Therefore in some cases the last part of the line was left blank, while in the others a short space on the line was considered a sufficient indication. Those parasha that were thought to be important enough to leave the rest of the line unused were called open parasha, while those where only a short space was made were called closed parasha. Eventually the parasha came to be indicated in the MSS simply by inserting the letter ף (for pethuchah) for open parasha or the letter ס (for sethumah) for closed parasha. Later the Pentateuch was divided into fifty-four sections, so that it could be read through in the Sabbath services in approx. one year. The close of each of these larger sections coincided with the end of one of the already marked parasha, and was indicated in the MSS by writing the appropriate letter whether s or ף, three times. This indication is still continued in Heb. Bibles.

Division into chs. was not made until the 13th cent. It was prob. worked out by an Eng. archbishop and originally placed in his Lat. Bible. Very soon the Jews recognized the convenience of these divisions and inserted them in their Heb. Bibles, usually following the exact division made by the archbishop, but occasionally changing it.

Originally 1 and 2 Samuel; 1 and 2 Kings; and 1 and 2 Chronicles were each written as one book. It was only when printed Bibles were made in the 16th cent. that they came to be divided into two separate books.

The work of the scribes

The men who busied themselves with the care and transmission of the Scriptures from the time of Ezra to that of the Masoretes are designated in Heb. MSS as the sopherim. This is usually tr. “scribes” which might suggest that they were merely copyists. However, the normal Heb. word for “write” is kathav which occurs over 200 times in the OT. Saphar properly means “to count.” It is stated in the Talmud that the keepers of the Scripture came to be called “sopherim” because they counted the letters and words in each section of the Scripture. Actually a sopher is not simply a man who writes things down, but one whom makes lists, keeps track of details, and oversees various aspects of affairs. Eventually the designation came to be applied to anyone who devoted himself to legal or literary pursuits.

During the long period before the time of the Masoretes, little information regarding the activities of the scribes in relation to the Scripture was preserved, but a great deal of information about their activities can be inferred from the results that were passed on in the form of marks or marginal notes inserted in the Bible. It is evident that in their endeavor to preserve the text from alteration or addition, they counted the number of words in each section, and also the number of vv. and paragraphs. They sometimes placed marginal notes in their MSS, wrote certain letters in unusual ways, or inserted dots or other marks at various places. These peculiarities were copied in later MSS and preserved by the Masoretes, even though in some cases their purpose may have been entirely forgotten. Various aspects of the work of the scribes may be designated under the following heads:

Letters written large or small.

At thirty-seven places in the OT it became customary to write a letter larger than normal, and in twenty-three places to write one small. In many cases the reason for the large letter is quite obvious: in three instances it is the first letter of a book; in one it is the middle letter of the Pentateuch. In the case of the oft-re-peated “shema” (Deut 6:4) the last letter of the first word (ayin) and the last letter of the last word (daleth) are written large. This might seem to be a peculiar selection of letters to emphasize, and it has been conjectured that these particular letters were selected as forming the Heb. word for “witness.”

In the case of the letters written small it is often difficult to make a reasonable conjecture as to why this was done. It would seem that the Masoretes found it to be customary in the MSS that had come down to them and simply continued the practice without change.

Suspended letters.

There are a few cases where letters are written somewhat above the line. One of these is the ayin in miya’ar (Ps 80:14) which prob. indicates the middle of the Psalter. Another is the nun in “Manasseh” in Judges 18:30. An early tradition says that this nun was suspended above the line to indicate that it had been inserted in order to change the word from “Manasseh” to “Moses,” so that people would not think that this apostate priest had been descended from Moses, but that it was kept suspended in order to indicate that it was, after all, a change in the text. One cannot say that this early tradition is false, but there is no proof that it is true. In some cases the reason for the suspended letter is quite unknown.

Paseq, “divider.”

This was a perpendicular stroke placed between two words, at about forty-eight places in the OT. It seems sometimes to have been placed before or after the divine name to avoid the possibility of its being incorrectly united with a word which, in the opinion of the scribes, should not be too closely tied to it, or to separate two words when there seemed to be a danger of their becoming merged because the final consonant of one and the first consonant of the other were identical.

Extraordinary points.

There are fifteen places where dots are placed over single letters or over a whole word. Ten of these are in the Pentateuch, the others elsewhere in the Bible. They prob. represent questions that the scribes raised about these places in the texts. The marks have been preserved, but the reason for them has been completely forgotten.

Inverted nun’s.

Nine times there is a mark in the MSS that looks like an inverted nun. There is evidence that discussion occurred as early as the 2nd cent. a.d. as to what it means. It has been suggested that the scribes thought that a group of words might have been misplaced, but did not feel justified in making a change. Some authorities question whether the mark is really a nun.

Tiqqune Sopherim.

Itture Sopherim.

“Omissions of the Scribes.” The Talmud contains a list of seven places where it says that a word not written in the text is to be read, and five places where it says that a word written in the text should not be read. It also mentions five places where it asserts that “and” was omitted by the scribes.


There are about 350 places, not usually marked in the MSS, but contained in Masoretic lists, where a Biblical word is quoted and then followed by another term which is introduced by the word sebir (Aram. for “suppose”). This does not indicate that the scribes thought something was wrong, but simply that they desired to explain the meaning of an unexpected form or word.

Work of the Masoretes

While the origin of the word “Masorete” (also spelled “masorite” or “massorete”) is uncertain, it is generally believed that it is derived from masar, a late Heb. word meaning “to hand down” or “to pass on,” and that from this root was derived the use of the noun Masorah to designate the tradition handed down in order to make a fence around the law. No one knows when the men who were guarding this tradition began to be called Masoretes; after c. 920 the Masoretes were considered to have completed their work, and the name was no longer applied.

For a time there were active groups of Masoretes in both Babylonia and Palestine. In the end, the Babylonian Talmud and the Targums produced in Babylonia seem largely to have won precedence throughout Jewry over those produced in Pal. but in the case of the Masoretes it was different. Although much was produced by the Babylonian Masoretes, the work of the group of Masoretes in Tiberias in Pal. came to be accepted as authoritative throughout the Jewish world.

The names of several of the Tiberian Masoretes have been preserved. The most prominent were members of the families of Ben Asher and Ben Naphtali. Five generations of the Ben Asher family were active from c. a.d. 780 to c. 920.

The tasks performed by the Masoretes may be arranged under four heads. First and most important was the continuation of the work to which the scribes had already devoted much attention, of endeavoring to maintain the integrity of the text of the Scripture. For this purpose they counted the number of letters, words, verses, and parashas in each book and indicated its middle word. They noted all peculiar and unusual forms, indicating how frequently each occurred. There is no way of knowing how much of this material they figured out themselves, and how much had already been determined by the previous scribes. Many notations and special marks put in by the scribes were scrupulously copied by the Masoretes even though sometimes they seem not to have known what meaning the scribes had intended to convey.

A great amount of material of this type was assembled by the Masoretes, and came to be known as the Masorah. Notes placed on the side margins of the columns were called the Small Masorah. Notes at the top and bottom of the page were called the Large Masorah, a name sometimes also applied to the notes placed at the end of each book, which are more usually called the Final Masorah.

Much of this material is written in very condensed language, about half of it using Heb. words and forms, and the remainder using Aram. Both of these languages had become dead languages by the 9th cent. a.d., as far as most of the Jews were concerned.

A second part of the task of the Masoretes was the standardization of the pronunciation of the words in the OT. As time went on, there was a tendency to forget what vowels should be pronounced with the written consonants, and the pronunciation of the words began to vary with the speech habits of the various localities. Ideas of Heb. grammar tended to become confused, as habits of pronunciation changed. The Masoretes set themselves industriously at this highly complicated task. Three systems of vocalization were attempted, but eventually the Tiberian system established itself as supreme. Careful study of the grammar and investigation of the traditions as to pronunciation led in some points to a conscious return to forms and practices that had existed centuries earlier, but had then largely disappeared. It is the contention of Paul Kahle (and he presents some strong arguments in its support), that in other features, such as the double pronunciation of the Begadh-kephath letters, the Masoretes introduced into Heb. pronunciation practices known in Syr. but new to Heb. Performing an immense labor of standardizing the grammar, maintaining the ascertained tradition, and working out a method of indicating precisely how they thought each word would rightly be pronounced, they placed indications of vowel pronunciation on every word of the Heb. Bible.

A third part of the work of the Masoretes involved providing an indication to the reader of the cases in which established tradition favored reading a word in a way that did not seem to fit the accepted consonantal text. It would seem that the Masoretes were determined on no account to alter the consonantal text that had been handed down to them. Yet, there were a considerable number of cases where it was customary to read it in a different way. Thus it had been customary for many centuries not to pronounce the divine name indicated by the tetragrammaton JHWH. For a time it was customary to substitute the phrase “the name.” Well before the time of Christ it had become usual to substitute the word ’adhonai “Lord,” unless that word was already used in conjunction with it, in which case they would substitute ’elohim “God.” In these cases the Masoretes simply put the vowels of ’adhonai or ’elohim on the consonants already in the text. This came to be called a “permanent Qere.” Qere is the imperative of an Aram. word meaning “read.”

Another permanent qere, much less frequent than the one just mentioned, resulted from the fact that in the early part of the Bible there are a number of cases where the third feminine sing. pronoun was written with the vowel letter that properly belonged with the third masculine sing. pronoun. The Masoretes had no way of knowing whether at that early time the masculine pronoun was sometimes used in a common sense, to apply to either gender, as was regularly done with the third pl. perfect of the verb. They knew that the uniform practice of the rest of the OT required the feminine pronoun in these places, and so put before the vowel letter waw the vowel sign that would be there if the vowel letter yodh had been written.

There are about 1300 instances in which the Masoretes have placed on a word a group of vowels that do not exactly fit the consonants as written. Usually this was to indicate a change of tense or something similar. Except for the four words that were permanent qeres (JHWH, hw’, Jerusalem and Issachar), the Masoretes put in the margin the word qere (“read”), followed by the consonants that would normally go with the vowels that had been placed in the text. The consonants that remained in the text were called the kethibh (Aram. “that which is written”).

It is possible that in many cases there was MS evidence for the qere; in such a case the Masoretes may have felt that the preponderance of MS evidence favored the kethibh and therefore retained it in the text, even though indicating that the qere was what should be read. In some cases there is reason to feel that the kethibh gives the better reading, but in most cases the qere seems preferable.

The fourth aspect of the work of the Masoretes may have been even more time-consuming than the other three, but is of less interest to most modern students. This was the indication of cantillation.

For many centuries it had been established practice to chant at least a portion of the synagogue reading of the Scripture. To provide a measure of standardization, the Masoretes invented an extremely complicated system of so-called accents. Most common is the methegh, which ordinarily indicates the secondary accent, but may also indicate any one of a score of other features. Other accents indicate the relation to the preceding or following words, and show the various features of the cantillation.

It is easy to see that with the addition of all these varied vowel marks and accents to the consonantal text, the difficulty of maintaining an exact text was greatly increased. Although all the Masoretic MSS agree remarkably closely in their consonantal text, it is only natural that considerable variety in preservation of these other features should develop.

The Large Masora includes references to a number of MSS which the Masoretes considered as having great authority. As far as known all of these famous pre-Masoretic MSS have perished. One of the most famous of these was the Codex Hilleli which was attributed to a famous rabbi named Hillel who lived at about a.d. 600. Other frequently cited MSS were named after their place of origin, such as Jericho, Jerusalem, Sinai, and Babylon.

With the establishment of their new system of vowels and accents, the Tiberian Masoretes had originated a type of MS which soon became standard throughout the Jewish world. Yet not all the Masoretes of Tiberias were united regarding every detail of the task. At the beginning of the 10th cent. their differences found expression in two main traditions, one of which bore the authority of the Ben Asher family, the other that of Ben Naphtali. In MSS prepared during the next two centuries marginal notes frequently refer to the distinctive readings of each of these authorities. Soon writers began to make lists of these differences. Most such lists have disappeared, but portions of several copies of a book on the subject have recently been found. Piecing these fragments together, it has been possible to reconstruct the entire list. The book was written by Mishael ben Uzziel and was entitled Kitāb al-Khilaf. It lists 875 differences between the two traditions. Practically all of these concern matters of vocalization or accentuation, nine tenths of them in fact being concerned with the use of metheg or ga’ya (a form of metheg). As to the consonantal text, there was no important difference between the two traditions.

For a time the two traditions continued side by side, but gradually the majority of Heb. grammarians and scholars may have tended to give preference to the Ben Asher readings, though accepting some of those of Ben Naphtali. Then the famous Jewish philosopher, Moses Maimonides (1135-1204), writing on a Biblical subject unrelated to the difference between these two traditions, declared that he accepted as authoritative an OT codex in Egypt that had been vocalized, collated and provided with Masora by Ben Asher. So great was the prestige of Maimonides throughout the Jewish world that this statement is generally considered as having been the cause of the sharp decline in the standing of the Ben Naphtali tradition that occurred soon afterward.

The particular codex of which Maimonides spoke so highly seems later to have been taken to Aleppo and there preserved in the Sephardic synagogue.

Important Masoretic MSS and printed editions

During the centuries from the Renaissance to a.d. 1800 a considerable number of Heb. MSS were collected by various universities and libraries. Yet the amount of such material available today is prob. three times as great as in a.d. 1800. This great increase is due partly to the intensive searches of Abraham Firkovitch, who collected more than 2000 MSS of parts of the OT and placed them in the Leningrad library, and partly to the many Biblical MSS recovered from the Cairo Geniza. It is difficult to compare the number of MSS in various museums and libraries, because sometimes a MS may include the whole OT, while at other times a very few pages may be listed as a separate MS. Much work remains to be done in the study of this material, but the intensive researches of Kahle and others during the first half of the present cent. have already produced remarkable results.

The following MSS, arranged in order of probable date of origin, are now generally considered to be the best sources for knowledge of the Ben Asher text.

(1) The Cairo Codex of the Prophets (some times designated as C). This MS, dated in a.d. 895, contains the entire second division of the Heb. Bible. It was written and pointed by Moshe ben Asher, the next to the last of the famous Ben Asher family. The MS was presented to the Qaraite community in Jerusalem, but was seized by the Crusaders in 1099. Later it was returned to the Jews, and came into possession of the Qaraite community in Cairo. It is written in three columns, with Tiberian vowels and accents.

(2) Leningrad MS Heb. B 3 (designated as P for Petrograd, the former name of Leningrad). This MS, which is dated in a.d. 916, contains only the later prophets. For a long time it was considered to be the oldest dated MS. It uses the Babylonian supralinear system of vocalization, but follows the Tiberian tradition in its vowels and Masora. On a few leaves the Tiberian signs have been substituted for the Babylonian ones.

(3) The Aleppo MS (sometimes designated as A). The colophon of this MS stated that Aaron ben Asher (the son of Moshe ben Asher), who died about a.d. 940, added the vowels and Masora to this MS. It is written on parchment in three columns. At first it was in Jerusalem, then for a time in Cairo, and later it was taken to Aleppo. It is generally considered to be the MS designated by Maimonides as the codex that he regarded as most reliable. For its later fate see XI below. Originally it contained the entire Bible, but about one fourth of it has now been destroyed.

(4) British Museum Or. 4445. This MS, prob. written in the middle of the 10th cent., contains only Genesis 39:20-Deuteronomy 1:33. The name of Ben Asher occurs several times in the margin.

(5) Leningrad MS B-19A (designated as L). The dating of this MS of the entire OT, which was brought from the Crimea by Firkovitch, has been discussed above (under IV). It claims to have been carefully copied in a.d. 1008 from a MS prepared by Aaron ben Moshe ben Asher. Written in three columns, with the usual Tiberian vocalization, it is the MS that is reproduced in Biblia Hebraica 3.

In addition Kahle states that in the autumn of 1926 he found in Leningrad, chiefly among MSS of the Second Firkovitch collection, fourteen Heb. Biblical MSS that can be dated between a.d. 929 and 1121, all containing the text of Ben Asher.

Despite the new evidence about details of the Ben Naphtali text there is as yet no agreement as to the discovery of particular MSS that presented it in a pure form. Among those suggested are Codex Reuchlinianus, which is preserved in Karlsruhe, Germany, and three MSS formerly at Erfurt (E1, E2, and E3). Some scholars, however, maintain that the Codex Reuchlinianus represents a cross between the work of the Babylonian Masoretes and those at Tiberias.

After 1100 great numbers of MSS were prepared. The text soon became rather composite, fundamentally based on Ben Asher, but including a considerable number of divergences, many of which were from the Ben Naphtali tradition. In these centuries the importance of the Masora ceased to be fully realized, and the marginal notes were frequently written in such a form as merely to form figures of animals and other decorative elements.

Soon after printing was invented the Jews began to issue books in Heb. A number of parts of the Bible were printed before 1500. Daniel Bomberg of Antwerp emigrated to Venice and there established a printing press from which he issued many important Heb. books between 1516 and 1549. His first ed. of a Rabbinic Bible, issued in 1516-1517, was edited by Felix Pratensis. It contained the Heb. text with targums and important commentaries in parallel columns. In 1525 this ed. was superseded by the second Bomberg Bible, edited by Jacob ben Chayim of Tunis. Ben Chayim included a large selection of Masora. Unfortunately the ed. seems to have had at his disposal only comparatively late MSS and to have been occupied with a number of other tasks during the fifteen months that he devoted to its preparation. Nevertheless the text of this ed. was standard in the western world until 1937.

Later edd. of the Rabbinic Bible, and other early publications of the OT in whole, or in part, were of less importance.

In the 18th cent. attention was directed to the variations in the consonantal text of the available MSS, most of which concerned vowel letters. In 1776-1780 Benjamin Kennicott published an ed. of the Heb. Bible at Oxford, listing variants from more than 600 Heb. MSS. In 1784-1788 J. B. de Rossi issued at Parma, Italy, an enlarged list of variants with a selection of the more important readings from 1417 MSS and edd. Most of the sources used by Kennicott and de Rossi were comparatively late.

In 1869 S. Baer undertook a new publication of sections of the Heb. OT, in the hope of providing a more scientific text. This work was never completed, and the methods used have received severe criticism. In 1908-1926 Christian D. Ginsburg produced an ed. of the Heb. Bible with an elaborate presentation of variants, but followed substantially the Textus Receptus as established by Jacob ben Chayim. In 1906 Rudolph Kittel issued a Biblia Hebraica, followed by a second ed. in 1912. The Ben Chayim text was used. Numberous footnotes were inserted, giving many conjectural emendations. In 1928 C. C. Torrey of Yale University declared: “The apparatus of Kittel’s Biblia Hebraica contains very many readings erroneously supposed to be attested by the Greek version, readings gathered blindly from the commentaries.”

When it was planned to issue a third ed. of the Biblia Hebraica Paul Kahle suggested that the Ben Chayim text be replaced by the text of a Ben Asher MS. Attempts were made to get permission to photograph the MS in the Sephardic synagogue of Aleppo, but its keepers refused to allow it to be photographed or even intensively studied. Therefore Kahle suggested printing the text of Leningrad MS B-19A. The authorities at Leningrad lent the MS to the University of Bonn, and Kahle photographed it and prepared its text and Masora for Biblia Hebraica 3. When issued in 1937, this publication rapidly became standard among western scholars. Unfortunately the footnotes of the second ed. were retained, and these instead of the actual text have often been treated as authoritative. A number of recent Eng. translations have slavishly followed the text contained in these footnotes. Outstanding scholars have severely criticized these footnotes as being largely conjectural and containing a very arbitrary selection of readings from ancient VSS without any really scientific examination of the alleged validity of the suggested readings.

In 1958 the British and Foreign Bible Society issued a Heb. Bible prepared by Norman H. Snaith, largely based upon the critical notes made from Spanish MSS in 1626 by Rabbi Solomon Norzi. Snaith has declared that the text he prepared in this way is very similar to that which Kahle found in the Leningrad Codex.

The Dead Sea Scrolls

The Christian world was thrilled in 1948 when it learned that in the previous year a cache of scrolls had been discovered that gave evidence of coming from the time of Christ and before. Many scholars naturally expected that these scrolls would differ radically from the Heb. text as found in MSS written nearly a thousand years later.

To the Biblical student the most important of these scrolls was the one that is now designated as 1QIsa. This was a beautifully written copy of the Book of Isaiah. The scroll had evidently been used a great deal, as could be seen from the evidence of wear and the fact that in many places where the writing had become obscure, it had been inked over. Immediately on checking into it, it was apparent that though it agreed generally with the MT there were many differences. The scroll had been copied very carelessly. At many points words written by mistake had been erased or crossed out and corrections had been inserted. Changes of a single letter or word were written in the same hand as the MS as a whole; longer corrections were added in a different hand. Letters and words omitted by the copyist were frequently inserted above the line. Sometimes the inserted material runs down the left-hand margin. Close examination shows that at many points the copyist left a space for something that must have been missing, or was not clear in the MS he was copying. In most cases such an omitted portion was evidently inserted later from another MS.

Aside from the obvious scribal errors, there were places where the text seemed to go with the LXX rather than with the MT, and some scholars took this as proof that the LXX text gives a more accurate idea of the Heb. Bible as it existed 2000 years ago than the MT. On further careful investigation, however, it became clear that, though there are a few places where the text agrees with the LXX as against the MT, in the overwhelming majority of readings it agrees with the MT as over against the LXX.

An interesting peculiarity of 1QIsa is that it uses vowel letters far more frequently than the MT. It would seem that the scribe himself, or the scribe of a previous MS from which this was copied, inserted these extra vowel letters in order to give interpretative aid to the reader. Sometimes the pronunciation that these suggest is different from that contained in the Masoretic vocalization. The MT contains the word tartān as the title of an Assyrian officer (Isa 20:1). In the scroll waw is inserted in the first syllable of the word, indicating that it should be pronounced tūrtān. Discovery of ancient Assyrian records has shown that the Assyrian form was turtannu. The tradition of the first vowel in this unfamiliar foreign word became altered in the course of the centuries, but its consonants were correctly preserved.

Another Isaiah scroll, now designated as 1QIsb, was found in Cave 1. It was very difficult to unroll. Eventually, however, this was safely done and it was found to be in bad shape, with many sections missing. Scholars immediately decided that its text was very close to the MT, and so devoted most of their attention to 1QIsa.

In 1952 a number of MSS were found in caves in the Wady Murabba’at, about eleven m. S of Qumran. Many of these were letters that could be precisely dated in the 2nd cent. a.d. There were also copies of a number of sections of the OT which agreed closely with the MT.

In the Qumran area about a dozen of the more than 300 caves that have been excavated have been found to contain scrolls or fragments of scrolls. The greatest number were found in Caves 1, 4, and 11, but no scroll comparable to 1QIsa in size or completeness has yet turned up. Some fairly extensive Biblical scrolls came from Cave 11. Cave 4 produced thousands of fragments representing hundreds of MSS. It was difficult to interpret these scraps of material. First, they had to be carefully humidified so as not to disintegrate when touched. Then they had to be flattened out, and the few words on each of them carefully studied to see whether it was a section of the Bible and if so what part. Eventually portions of nearly 100 copies of Bible books were identified, representing every OT book except Esther. A fragment of Samuel is thought to come from the 4th cent. b.c.

Although the great bulk of the material from the Qumran caves fits very closely with the MT, a few sections have been found to go with the LXX, or with the Samaritan Pentateuch rather than with the MT. This is particularly true of the Book of Samuel which is prob. the least well preserved book in the whole MT. One scroll in Cave 4 seems to have contained a text of Samuel very close to the LXX, and another is thought to have readings superior to both the MT and the LXX.

The scrolls from Murabba’at, representing as they do the group of Jews who were active in the Bar Cochba revolt in a.d. 132-135, followed closely the official text from which the MT has descended. At Qumran the great bulk of MSS also follow this text. Some few differ, as is only natural, since the sectarians at Qumran doubtless included people from many parts of the land, some of whom brought with them scrolls made in their local areas, sometimes carelessly copied, in which scribal errors and changes of text would have become perpetuated through a series of copyings.

The Cairo Geniza

In view of the number of extant copies of the NT and of the LXX from comparatively early centuries, it might seem strange that, apart from the DSS, no Heb. text of the OT written before a.d. 895 has been preserved. During the Middle Ages the Jews were often subjected to fierce persecution and obliged to move from place to place, while some Christian monasteries in the E have been practically unmolested for nearly 1500 years. Even so, the lack of MSS before the time of the Masoretes would be hard to explain apart from the long-established Jewish custom that writings containing the name of God must be protected from desecration; if worn out, or found to contain errors, they are to be entirely removed from circulation. Every synagogue had its geniza (from ganaz “to hide”), which was a room in its cellar or attic in which MSS and documents that were no longer of value could be deposited until a convenient time would be found for burying them in consecrated ground. Worn-out scrolls were generally buried with a scholar.

Abraham Firkovitch was an expert at ransacking old synagogues and their genizas. He was extremely secretive about the source of his material, but Paul Kahle is strongly convinced that much of it came from what later became famous as the Cairo Geniza.

This was the geniza of a synagogue that had been established in a.d. 882 in a building that had formerly been a Christian church and that had thereafter been used as a synagogue for over a thousand years. For many centuries discarded documents were deposited in this geniza. Then through some chance its very existence was forgotten, and the room was even walled up for a time. When it was rediscovered in the 19th cent., some of the material was buried, but this was discontinued when it was learned that antiquarians were willing to pay a substantial price for old documents. Many MSS and fragments of documents were taken from it to various museums and libraries in Europe and America. Finally, in 1896 the Cambridge University Library sent Solomon Schechter to Cairo with authorization to obtain as much of the material as possible, and a great many boxes of dirty fragments of MSS were shipped to Cambridge. Altogether more than 200,000 fragments of written material have been taken from this geniza. This includes documents of many sorts, since even ordinary business contracts, when no longer of value, would be placed in the geniza, if they contained a date or a salutation in which the name of God might have been used. A study of this material is already greatly increasing our knowledge of the history of the life and culture of the areas that centered in medieval Cairo.

Hundreds of Biblical MSS from this source have now become available. Paul Kahle has made an extensive study of many of these, and from them has worked out his theories of the two different groups of Masoretes, one in Babylonia and one in Israel, and has originated many thought-provoking suggestions regarding the history of the Heb. text. Much has already been learned from these documents and much is yet to be gained through further study of them.

The Aleppo MS

As mentioned above, it was the feeling of many scholars that the Aleppo MS was the earliest complete MS of the OT extant, and that its vocalization and Masora had been placed on it by Aaron ben Asher himself. It was therefore a great disappointment to Paul Kahle when he was unable to use this MS as the text of Biblia Hebraica 3. In 1948 the Sephardic synagogue in Aleppo was raided by a local mob and burned, and for a number of years it was feared that the MS had also been destroyed. However, Izhak Ben-Zvi, president of the state of Israel, refused to give up hope that it might have been saved. For a long time he kept trying to find the place where it might be hidden, and often discussed with Sephardic leaders ways and means by which the “discovery” of the venerable MS might be brought about, and its transfer to safety in Jerusalem assured. At last his efforts succeeded, and in 1960 he was able to announce to the world that the codex had been found and brought to the Heb. University at Jerusalem. Unfortunately it had suffered much at the hands of the rioters; before 1948 it had been complete; now about a fourth of it was missing, including nine-tenths of the Pentateuch.

Although the authorities at the Sephardic synagogue at Aleppo had never permitted a Jewish scholar to photograph any part of the MS they had once allowed an Eng. scholar, William Wickes, to photograph a page of it (Gen 26:17-27:30), which he had published in 1887 as the frontispiece of his book on Heb. accents. In 1966 M. H. Goshen-Gottstein of the Heb. University announced the discovery that once another Christian had also secured permission to photograph a few pages of the MS. The Rev. J. Segall, a missionary in Damascus for many years, published in 1910 a book entitled Travels through Northern Syria, in which he included a picture of the pages of the Aleppo MS that contained Deuteronomy 4:38-6:3. Since both of these are portions that were destroyed, the existence of the pictures is a cause for rejoicing. Unfortunately, the photograph by Segall is not clear enough to make it possible to decipher all the details of vocalization, accentuation, and Masora, though the consonantal letters are clearly discernible throughout. Wickes explicitly denied the claims about the MS, and there is no indication in the original publication that Segall had any realization of the special importance of this particular codex. However, there can be little doubt of the genuineness of the MS or of its direct relation to Ben Asher.

Further study of the Aleppo MS will doubtless throw much light in the near future on additional details of the Ben Asher tradition.

Types of error

In copying Heb. MSS the same types of mistakes are apt to occur as in copying any types of MSS. These may be classified into errors of sight, errors of hearing, and errors of memory.

It is well known that MSS were often copied through dictation. Sometimes one man would dictate and many scribes would make copies, as was often done in the case of the Gr. and Rom. classics. This may sometimes have been done at Qumran or elsewhere in the preparing of popular copies of Biblical books. However, it was strictly against regulations that official copies of the Scripture should be made in this way. The scribe was required to look repeatedly at the material he was copying. Consequently errors of hearing must have been almost nonexistent in the official text. Errors of memory would be few, and yet could occur, since one who is copying by sight may sometimes be confused by his recollection and mistakenly put down something different from what he sees.

Errors of sight are apt to be related to similarities of letters. A letter that is written somewhat poorly can easily be mistaken for another letter. Such errors occur occasionally in Biblical MSS. By far the most common error of this type is confusion of daleth and resh. In Heb. MSS these two letters are sometimes written so similarly that it is difficult to distinguish them. Clear evidence that this occurred can be found by comparison of proper names in Kings and Chronicles where there are many cases in which the same name is written with a d in one and with an r in the other. There are cases where a LXX word seems to have no relation to the Heb. text as it stands, and yet, on assuming that the MS used by the LXX tr. had a daleth instead of a resh, or a resh instead of a daleth, a word may sometimes be substituted which exactly corresponds to that found in the LXX.

Other common errors of sight result from haplography (writing a letter or a group of letters only once where it occurs twice) dittography (accidentally repeating a letter or a group of letters), and omission due to homoeo-teleuton (similar ending), where the eye skips from one word to another that ends similarly. Any writer knows how frequently this last type of error occurs in the copying of his own MSS.

In The Text of the OT, pp. 71-73, Ernst Wuerthwein gives many examples of errors of these various types that are to be found in the first Isaiah scroll from Qumran.

Evidence of the versions

Since each of the VSS of the OT such as Septuagint, Targums, Syriac Peshitta, and Vulgate are treated in separate articles, the present discussion will be limited mainly to general matters concerning the relation of VSS to the establishment of the OT text.

In determining the value of a VS the first consideration that naturally occurs is that of age. A tr. into an Indian language made by a recent tr. would be of no value in fixing the original text of the OT. It would merely give evidence as to the text used by the tr. For a VS to be of value it should have been made in ancient times.

The second consideration is that of “immediacy.” When something is tr. from one language into another a large amount of accuracy is inevitably lost. Words do not exactly correspond in the two languages. Often there is great overlapping between different words that might be used. Types of expression, verb forms, and principles of syntax differ radically in various languages. Therefore a tr. will at best give only a general picture of the meaning of the original. When a tr. is made from something that is itself a tr. the divergence from the original is necessarily compounded. Consequently an important factor in determining the value of a tr. is the question of its immediacy. The Vul. tr. made by St. Jerome at about a.d. 400 was directly from the Heb. and therefore may be of great value in helping to fix the exact Heb. text at that time. The Old Lat. text, though tr. some centuries earlier, was not made from the Heb. OT but from the LXX. It is of value in fixing the text of the LXX at the time when it was made, but its value for determining the original Heb. text is greatly diluted.

There are four ancient and immediate VSS. These are the LXX, the making of which began at about 230 b.c., the Syriac Peshitta, prob. made in the 5th cent. a.d., though portions may have been made earlier, the Vulgate made about a.d. 400, and the Aram. Targums produced at various times.

Another consideration that should not be overlooked is the difference in the amount of effort devoted to maintaining purity of text. In this regard the care devoted to preservation of the MT is unparallelled in relation to any other book, including the Gr. NT and the various VSS of the OT. The various LXX MSS have diverged in many different directions. A few scholars have devoted innumerable hours to study and comparison of them, but no thoroughgoing system of dividing them into schools or making genealogical charts, similar to what has been done for the Gr. NT, has as yet been devised. It is unlikely that such a scheme will ever be successfully performed, because of the great amount of material involved, and also because of the fact that various books were tr. or copied at different times.

Sometimes it is found that a MS of the LXX includes sections that follow radically different text traditions. If the places at which the text follows one tradition are indicated, and those in which it follows another are indicated, it will sometimes be noted that they occur in groups, and that a division can be made into sections that are of equal length, but quite different from the page divisions of the present MS. It can be proved that the original was copied from a MS from which a considerable number of pages were missing, and that these particular pages were filled in from a different and perhaps inferior MS. In this way, different text traditions become combined in the same MS, and the problem of separating them and defining them is extremely difficult.

Some of the best MSS of the LXX contain great numbers of rather ridiculous scribal errors. Proper names have often been strangely mangled in the course of transmission. The Vaticanus MS, for instance, while generally considered as preserving our best type of LXX text, is itself a very poorly copied MS, and involved a great number of unique textual errors, esp. in proper names.

The LXX text is of greatest help in a section such as Samuel, where the Heb. text has been corrupted far more than in other sections. It is of special significance whenever its reading can be explained by considering the Heb. to have been vocalized differently, or by the assumption that its Heb. prototype had a resh rather than a daleth or vice versa. An interesting instance is Amos 9:12 where the MT reads: “They shall possess the remnant of Edom.” The LXX says: “The remnant of mankind shall seek (the Lord.)” The difference may mainly be explained by assuming a different vocalization in one word, and a difference between resh and daleth in another. In Acts 15:17 James quoted this passage as the clinching argument in a decision by the Jerusalem council. There were present many learned men who would gladly have refuted him if he had not been true to the text. Therefore we can feel quite positive that in his day the original Heb. at this point was like the LXX rather than like the MT.

It should be remembered that the other ancient and immediate VSS were greatly influenced by the LXX, and therefore are often not as strong an argument for the original Heb. as might otherwise be the case. Study of VSS is of great value in learning what interpretation of a passage was common at the time when the VS was made, and sometimes in determining a possible alternative text, which in certain cases may be the true original text. In most instances, however, the MT is far more dependable than the text of the VSS.


It should be noted that the material for the establishment of the OT text is many times as great as that for establishing the text for any other ancient document, except the NT. The agreement in the consonantal text of the various MSS is most remarkable, and the great bulk of the very extensive material that has now been found from before the time of Christ agrees very closely with the consonantal text of the MT. A few fragments give evidence of varying documents in some parts of Israel at that early period, and some of these may represent the text upon which the Samaritan Pentateuch was based or from which the LXX was tr. The feat of copying and recopying the text from the time of Qumran to that of Ben Asher with so little variation is unique in history. The Masoretes did a valuable task in writing down the tradition about the vowels and standardizing it. The fragments from the Cairo Geniza will enable us to reconstruct to a far greater degree than has yet been done the progress of the development of this system, and to see at what places arrangements of vowels or of accents represent the maintenance of an old tradition and at what points they represent the conclusions of the Masoretes. The MS from Aleppo will enable us to know the results of the Masoretic activities more accurately than ever before.

The Qumran MSS and LXX comparisons suggest that in the books of Samuel and Kings a greater amount of distortion may have occurred than in any other books, and here the LXX may make possible a more correct text at certain points. Where the LXX variation can be easily explained on the basis of a natural confusion between Heb. letters it may give a hint of a corruption that has come into the MT, but these cases are comparatively few.

The text has been preserved with a remarkable accuracy but not absolutely so. This represents the intention of the divine Author that we should have such certainty of the text as is possible for no book outside the Bible, and yet that there should always be the possibility in any one verse, taken by itself, that an error has come in. This fact is a useful guard against building important conclusions upon one v. alone. Any one v. may involve an error. Where there is no textual evidence of a variation in two different vv., the probability that each of them conceals an unknown error is so small as to be negligible. It is the divine warning to compare Scripture with Scripture and to study its teachings as a whole, rather than to endeavor to squeeze meanings out of individual words and phrases beyond what is possible in view of the nature of human words.


B. Kennicott, Vetus Testamentum Hebraicum cum variis lectionibus, 2 vols. (1776-1780); G. B. de Rossi, Variae lectiones Veteris Testamenti (1784-1788); C. D. Ginsburg, The Massorah compiled from Manuscripts (1880-1905); Introduction to the Masoretico-Critical Edition of the Hebrew Bible (1897); P. Kahle, Masoreten des Ostens (1913); H. B. Swete, Introduction to the OT Greek (rev. R. R. Ottley) (1914); P. Kahle, Masoreten des Westens, I (1927); II (1930); B. J. Roberts, The OT Text and Versions (1951); P. Skehan, “Exodus in the Samaritan Recension from Qumran,” JBL, LXXIV (1955), 182-187; O. Eissfeldt, Introduction to OT (translated from the 3rd German ed.) (1955); E. Wuerthwein, The Text of the OT (1957); F. M. Cross, The Ancient Library of Qumran (1958); P. Kahle, The Cairo Geniza (2nd edition) (1959); R. D. Wilson, A Scientific Investigation of the OT (reprint with revisions by E. J. Young) (1959); M. H. Goshen-Gottstein, “The Authenticity of the Aleppo Codex,” Textus I (1960), 17-58; C. Rabin, ed., Textus, vols. 1-3, (1960-1963); H. M. Orlinsky, “The Textual Criticism of the OT” in The Bible and the Ancient Near East, G. E. Wright, ed. (1961); M. H. Goshen-Gottstein, “The Rise of the Tiberian Bible Text,” in Biblical and Other Studies, Alexander Altmann, ed. (1963); S. Talmon, Textus, vols. 4-6 (1964-1968); M. H. Goshen-Gottstein, The Book of Isaiah: Sample Edition with Introduction (1965); S. Jellicoe, The Septuagint and Modern Study (1968); B. J. Roberts, “The Old Testament: Manuscripts, Text and Versions,” in The Cambridge History of the Bible, vol. 2, G. W. H. Lampe, ed. (1969).