The MT was primarily copied, edited and distributed by a group of Jews known as the Masoretes between the seventh and tenth centuries AD. Though the consonants differ little from the text generally accepted in the early second century (and also differ little from some Qumran texts that are even older), it has numerous differences of both greater and lesser significance when compared to (extant 4th century) manuscripts of the Septuagint, a Greek translation (made in the 3rd to 2nd centuries BC) of the Hebrew Scriptures that was in popular use in Egypt and Palestine and that is often quoted in the Christian New Testament.
The Hebrew word mesorah (מסורה, alt. מסורת) refers to the transmission of a tradition. In a very broad sense it can refer to the entire chain of Jewish tradition (see Oral law), but in reference to the masoretic text the word mesorah has a very specific meaning: the diacritic markings of the text of the Hebrew Bible and concise marginal notes in manuscripts (and later printings) of the Hebrew Bible which note textual details, usually about the precise spelling of words.
The oldest extant fragments of the Masoretic Text date from approximately the ninth century AD, and the Aleppo Codex (the oldest copy of the Masoretic Text, but missing the Torah) dates from the tenth century.
Another Talmudic story, perhaps referring to an earlier time, relates that three Torah scrolls were found in the Temple court but were at variance with each other. The differences were then resolved by majority decision among the three.
The discovery of the Dead Sea Scrolls at Qumran, dating from c.150 BC–AD 75, shows however that in this period there was not always the scrupulous uniformity of text that was so stressed in later centuries. The scrolls show numerous small variations in orthography, both as against the later Masoretic text, and between each other. It is also evident from the notings of corrections and of variant alternatives that scribes felt free to choose according to their personal taste and discretion between different readings. However, despite these variations, most of the Qumran fragments can be classified as being closer to the Masoretic text than to any other text group that has survived. According to Shiffman, 60% can be classed as being of proto-Masoretic type, and a further 20% Qumran style with bases in proto-Masoretic texts, compared to 5% proto-Samaritan type, 5% Septuagintal type, and 10% non-aligned. Furthermore, according to Haas, most of the texts which vary from the Masoretic type, including four of the Septuagint type manuscript fragments, were found in Cave 4. "This is the cave where the texts were not preserved carefully in jars. It is conjectured, that cave 4 was a geniza for the depositing of texts that were damaged or had textual errors."
An emphasis on minute details of words and spellings, already used among the Pharisees as bases for argumentation, reached its height with the example of Rabbi Akiva (d. AD 135). The idea of a perfect text sanctified in its consonantal base quickly spread throughout the Jewish communities via supportive statements in Halakha, Aggada, and Jewish thought; and with it increasingly forceful strictures leading ultimately to the statement in medieval times that a deviation in even a single letter would make a Torah scroll invalid. Very few manuscripts are said to have survived the destruction of Jerusalem in AD 70. This both drastically reduced the number of variants in circulation, and gave a new urgency that the text must be preserved. New Greek translations were also made. Unlike the Septuagint, large-scale deviations in sense between the Greek of Aquila and Theodotion and what we now know as the Masoretic text are minimal. Detailed variations between different Hebrew texts in use still clearly existed though, as witnessed by differences between the present-day Masoretic text and versions mentioned in the Gemara, and often even Halachic midrashim based on spelling versions which do not exist in the current Masoretic text. (Mostly, however, these variations are limited to whether particular words should be written plene or defectively - i.e. whether a mater lectionis consonant to represent a particular vowel sound should or should not be included in a particular word at a particular point).
The current received text finally achieved predominance through the reputation of the Masoretes, schools of scribes and Torah scholars working between the 7th and 11th centuries, based primarily in Palestine in the cities of Tiberias and Jerusalem, and in Babylonia. These schools developed such prestige for the accuracy and error-control of their copying techniques that their texts established an authority beyond all others. Differences remained, sometimes bolstered by systematic local differences in pronunciation and cantillation. Every locality, following the tradition of its school, had a standard codex embodying its readings. In Babylonia the school of Sura differed from that of Nehardea; and similar differences existed in the schools of the Land of Israel as against that at Tiberias, which in later times increasingly became the chief seat of learning. In this period living tradition ceased, and the Masoretes in preparing their codices usually followed the one school or the other, examining, however, standard codices of other schools and noting their differences.
It has been suggested that there never was an actual "Ben Naphtali"; rather, the name was chosen (based on the Bible, where Asher and Naphtali are the younger sons of Zilpah and Bilhah) to designate any tradition different from Ben Asher's. This is unlikely, as there exist lists of places where ben Asher and Ben Naphtali agree against other authorities.
Ben Asher was the last of a distinguished family of Masoretes extending back to the latter half of the eighth century. Despite the rivalry of Ben Naphtali and the opposition of Saadia Gaon, the most eminent representative of the Babylonian school of criticism, Ben Asher's codex became recognized as the standard text of the Bible. See Aleppo Codex, Codex Cairensis.
Considerable influence on the development and spread of Masoretic literature was exercised during the eleventh, twelfth, and thirteenth centuries by the Franco-German school of Tosafists. R. Gershom, his brother Machir, Joseph ben Samuel Bonfils (Tob 'Elem) of Limoges, R. Tam (Jacob ben Meïr), Menahem ben Perez of Joigny, Perez ben Elijah of Corbeil, Judah of Paris, Meïr Spira, and R. Meïr of Rothenburg made Masoretic compilations, or additions to the subject, which are all more or less frequently referred to in the marginal glosses of Biblical codices and in the works of Hebrew grammarians.
By long tradition, a ritual Torah scroll shall contain only the Hebrew consonantal text - nothing may be added, nothing taken away. However, perhaps because they were intended for personal study rather than ritual use, the Masoretic codices provide extensive additional material, called masorah, to show correct pronunciation and cantillation, protect against scribal errors, and annotate possible variants. The manuscripts thus include vowel points, pronunciation marks and stress accents in the text, short annotations in the side margins, and longer more extensive notes in the upper and lower margins and collected at the end of each book.
The Small Masorah consists of brief notes with reference to marginal readings, to statistics showing the number of times a particular form is found in Scripture, to full and defective spelling, and to abnormally written letters. The Large Masorah is more copious in its notes. The Final Masorah comprises all the longer rubrics for which space could not be found in the margin of the text, and is arranged alphabetically in the form of a concordance. The quantity of notes the marginal Masorah contains is conditioned by the amount of vacant space on each page. In the manuscripts it varies also with the rate at which the copyist was paid and the fanciful shape he gave to his gloss.
In most manuscripts, there are some discrepancies between the text and the masorah, suggesting that they were copied from different sources or that one of them has copying errors. The lack of such discrepancies in the Aleppo Codex is one of the reasons for its importance; the scribe who copied the notes, presumably Aaron ben Moses ben Asher, probably wrote them originally.
Hence the Masoretes contributed the Numerical Masorah. These notes are traditionally categorized into two main groups: the marginal Masorah and the final Masorah. The category of marginal Masorah is further divided into the Masorah parva (small Masorah) in the outer side margins and the Masorah magna (large Masorah), traditionally located at the top and bottom margins of the text.
The Masorah parva is a set of statistics in the outer side margins of the text. Beyond simply counting the letters, the Masorah parva consists of word-use statistics, similar documentation for expressions or certain phraseology, observations on full or defective writing, references to the Kethiv-Qere readings and more. These observations are also the result of a passionate zeal to safeguard the accurate transmission of the sacred text.
The Masorah magna, in measure, is an expanded Masorah parva. It is not printed in BHS.
The final Masorah is located at the end of biblical books or after certain sections of the text, such as at the end of the Torah. It contains information and statistics regarding the number of words in a book or section, etc.
Thus (Leviticus 8:23) is the middle verse in the Pentateuch; all the names of Divinity mentioned in connection with Abraham are holy except (Genesis 18:3); ten passages in the Pentateuch are dotted; three times the Pentateuch has the spelling לא where the reading is לו. The collation of manuscripts and the noting of their differences furnished material for the Text-Critical Masorah. The close relation which existed in earlier times (from the Soferim to the Amoraim inclusive) between the teacher of tradition and the Masorete, both frequently being united in one person, accounts for the Exegetical Masorah. Finally, the invention and introduction of a graphic system of vocalization and accentuation gave rise to the Grammatical Masorah.
The most important of the Masoretic notes are those that detail the Kethiv-Qere that are located in the Masorah parva in the outside margins of BHS. Given that the Masoretes would not alter the sacred consonantal text, the Kethiv-Qere notes were a way of "correcting" or commenting on the text for any number of reasons (grammatical, theological, aesthetic, etc.) deemed important by the copyist. [Reference: Pratico and Van Pelt, Basics of Biblical Hebrew, Zondervan. 2001. p406ff]
Since no additions were allowed to be made to the official text of the Bible, the early Masoretes adopted other expedients: e.g., they marked the various divisions by spacing, and gave indications of halakic and haggadic teachings by full or defective spelling, abnormal forms of letters, dots, and other signs. Marginal notes were permitted only in private copies, and the first mention of such notes is found in the case of R. Meïr (c. AD 100-150).
Rabbi Simon ben Pazzi (third century) calls these readings "emendations of the Scribes" (tikkune Soferim; Midrash Genesis Rabbah xlix. 7), assuming that the Scribes actually made the changes. This view was adopted by the later Midrash and by the majority of Masoretes. In Masoretic works these changes are ascribed to Ezra; to Ezra and Nehemiah; to Ezra and the Soferim; or to Ezra, Nehemiah, Zechariah, Haggai, and Baruch. All these ascriptions mean one and the same thing: that the changes were assumed to have been made by the Men of the Great Synagogue.
The term tikkun Soferim has been understood by different scholars in various ways. Some regard it as a correction of Biblical language authorized by the Soferim for homiletical purposes. Others take it to mean a mental change made by the original writers or redactors of Scripture; i.e. the latter shrank from putting in writing a thought which some of the readers might expect them to express.
The assumed emendations are of four general types:
In fifteen passages in the Bible, some words are stigmatized; i.e., dots appear above the letters. (Gen 16:5, 18:9, 19:33, 33:4, 37:12, Num 3:39, 9:10, 21:30, 29:15, Deut. 29:28, 2Sam 19:20, Isaiah 44:9, Ez 41:20, 46:22, Ps 27:13) The significance of the dots is disputed. Some hold them to be marks of erasure; others believe them to indicate that in some collated manuscripts the stigmatized words were missing, hence that the reading is doubtful; still others contend that they are merely a mnemonic device to indicate homiletical explanations which the ancients had connected with those words; finally, some maintain that the dots were designed to guard against the omission by copyists of text-elements which, at first glance or after comparison with parallel passages, seemed to be superfluous. Instead of dots some manuscripts exhibit strokes, vertical or else horizontal. The first two explanations are unacceptable for the reason that such faulty readings would belong to Qere and Ketiv, which, in case of doubt, the majority of manuscripts would decide. The last two theories have equal probability.
The primary set of inverted nuns is found surrounding the text of Numbers 10:35-36. The Mishna notes that this text is 85 letters long and dotted. This demarcation of this text leads to the later use of the inverted nun markings. Saul Liberman demonstrated that similar markings can be found in ancient Greek texts where they are also used to denote 'short texts'. During the Medieval period, the inverted nuns were actually inserted into the text of the early Rabbinic Bibles published by Bromberg in the early 16th century. The talmud records that the markings surrounding Numbers 10:35 - 36 were thought to denote that this 85 letter text was not in its proper place. One opinion goes so far as to say that it would appear in another location in a later edition of the Torah!
Bar Kappara is known to have considered our Torah as comprised of 7 volumes in the Gemorah "The seven pillars with which Wisdom built her house (Prov. 9:1) are the seven Books of Moses". Genesis, Exodus and Leviticus and Deuteronomy as we know them but Numbers was really 3 separate volumes Num 1:1 to Num 10:35 followed by Number 10:35-36 and the third text from there to the end of Numbers. The 85 letter text is also said to be denoted because it is the model for the least number of letters which constitute a 'text' which one would be required to save from fire due to its holiness.
The materials for the history of the first period are scattered remarks in Talmudic and Midrashic literature, in the post-Talmudical treatises Masseket Sefer Torah and Masseket Soferim, and in a Masoretic chain of tradition found in Ben Asher's "Diḳduḳe ha-Ṭe'amim," § 69 and elsewhere.
Next to Ibn Adonijah the critical study of the Masorah has been most advanced by Elijah Levita, who published his famous "Massoret ha-Massoret" in 1538. The "Tiberias" of the elder Buxtorf (1620) made Levita's researches more accessible to a Christian audience. The eighth prolegomenon to Walton's Polyglot Bible is largely a réchauffé of the "Tiberias". Levita compiled likewise a vast Masoretic concordance, "Sefer ha-Zikronot," which still lies in the National Library at Paris unpublished. The study is indebted also to R. Meïr b. Todros ha-Levi (RaMaH), who, as early as the thirteenth century, wrote his "Sefer Massoret Seyag la-Torah" (correct ed. Florence, 1750); to Menahem Lonzano, who composed a treatise on the Masorah of the Pentateuch entitled "Or Torah"; and in particular to Jedidiah Norzi, whose "Minḥat Shai" contains valuable Masoretic notes based on a careful study of manuscripts.
The Dead Sea Scrolls have shed new light on the history of the Masoretic Text. Many texts found there, especially those from Masada, are quite similar to the Masoretic Text, suggesting that an ancestor of the Masoretic Text was indeed extant as early as the 2nd century BC. However, other texts, including many of those from Qumran, differ substantially, indicating that the Masoretic Text was but one of a diverse set of Biblical writings (Lane Fox 1991:99-106; Tov 1992:115).