This entry consists of two articles, the first on manuscripts of the Hebrew Bible and the second on manuscripts of the Greek New Testament. For discussion of types of manuscripts, see Books and Bookmaking. For discussion of the importance of individual manuscripts in establishing the text of the Bible, see Dead Sea Scrolls and Textual Criticism.

Hebrew Bible


Ancient texts such as the Hebrew Bible were transmitted over the centuries in various forms, at first in scrolls, later in manuscripts of codex format, and in recent centuries in printed editions. Ancient scrolls consist of sheets of leather sewn together, resembling the Torah scrolls still used today in synagogues. A codex consists of any number of sheets (leather, parchment) bound together into a book. Scrolls contained individual biblical books, while codices could contain most or all of the Hebrew Bible. Hebrew biblical texts were first transmitted in scrolls; after the use of codices for the New Testament and other ancient literature, the codex form began to be used as well.

Development of the Biblical Text.

When the composition of a biblical book was completed, in theory there should have been only one master copy of that particular book. In practice, however, even at that stage there may have been several copies which differed from one another, often a great deal; the same is true of relatively modern works as well. These differences increased rather than diminished in the next centuries, as generations of scribes added new details to the manuscripts and scrolls they copied.

At an early stage, biblical Hebrew, like other Semitic languages, was written solely with consonants; in English this would mean that a word like “book” would be written “bk.” The first type of difference between biblical texts was thus in consonants, as is still visible from a comparison of early sources such as the Dead Sea Scrolls. More differences come to light when these texts are compared with the Samaritan Pentateuch and the Septuagint translation in Greek, whose underlying Hebrew text can often be reconstructed.

Many of these different texts fell out of use, especially because the religious groups supporting them ceased to exist or diminished in importance. In the Roman period, however, there were many different “families” of texts in use, as the Qumran texts show; among these was what was to become the official version of the text of the Hebrew Bible (“the received text”) in the centuries to follow, the Masoretic Text.

The Masoretic Text.

After the destruction of the Second Temple in 70 CE, a particular group of texts was emerging, endorsed by and copied by the central stream of Judaism, that of the Pharisees, later to be known as the Masoretic Text (MT). When compared with others, this text‐family had no special characteristics, and the only feature that stands out is its careful transmission, though probably other texts were also transmitted as carefully. The antecedents of the MT, found principally at Qumran, are called proto‐Masoretic, since the later components of that text had not yet been added to it.

A good example of an early proto‐Masoretic source is the shorter Isaiah scroll from Cave 1 at Qumran (1QIsab). Comparison of this scroll with the medieval and modern printed text shows that the consonantal framework of the MT changed little after the first century BCE. This is confirmed by the fragments found from Wadi Murabbaʿât south of Qumran, dating from about 130 CE, all of which are virtually identical with the medieval text. In addition, the biblical text quoted in the Talmud and early midrashim is virtually identical with the later texts, though owing to the imprecision of quotation and the nature of the Talmudic context occasional divergences exist. Thus, around the turn of the era there was a strong tendency to create a single unified text, but this ideal was never fully achieved owing to the vicissitudes of textual transmission. In other words, at no time did there exist one “Masoretic Text”; there always were several different “Masoretic Texts.”

The Masoretic Text known to us today was created in the early Middle Ages, when other components were added to the consonantal framework. Learned scribes involved in the transmission of the biblical text developed and perpetuated in and around the text various kinds of data, including vocalization and cantillation marks as well as a Masoretic apparatus of remarks in the text. These were the Masoretes who derived their name from the Masorah (see below).


The use of consonantal spelling meant that any word in Hebrew could be read in different ways; in English this would mean that the letters “sl” could be read “seal,” “sill,” “sell,” “soul,” “sale,” etc. Such ambiguity was troublesome, especially for an authoritative text; as a result, certain letters (ʾalep, hē, yōd, wāw) were added to the text to indicate which vowels were to be read (“vowel letters”). Later actual vowels were added to the text in the form of newly‐invented graphic symbols, so that from then on there was no confusion regarding the reading of the text. Various vocalization systems were developed, but the system finally accepted by all western Jewish communities was the Tiberian system (from Tiberias) developed by Moshe Ben‐Asher (tenth century CE). A so‐called Babylonian system was adopted by the oriental Jewish communities. In the course of the centuries, however, the Tiberian Ben‐Asher tradition has been more or less accepted by all communities, even though traces of other systems persisted in more remote areas, especially in Yemen.


Other signs were added to the consonants as guidance for the liturgical reading of the text in the synagogue. These signs have both musical value and syntactic importance, since they indicate the relationship between any two words in the text in different gradations of conjunctiveness or disjunctiveness. Various cantillation systems were developed, but finally the Ben‐Asher system was accepted for this purpose as well. The interpretive tradition embodied in the vocalization and cantillation system was established early, but only from the eighth century CE onward were these transmitted in a written form.


In and around the text the Masoretes added an elaborate body of notes and signs for the preservation and correct spelling of the text. In the text itself the Masoretes perpetuated all the special elements that were present in the text when the learned scribes decided to allow no more changes and which therefore became part and parcel of it, such as large and small letters, dots under or above the letters, and suspended letters. In addition, in the margin of the text the Masoretes added an elaborate apparatus of notes (Masorah, “transmission”) on individual words in the text, especially on the number of occurrences of certain forms and spellings, meant as guidance for the scribes.

There are many differences among the various Masoretic manuscripts in all these elements. Scholars consider two manuscripts as the most authentic representatives of the Ben‐Asher tradition, since one was apparently vocalized by Moshe Ben‐Asher himself (the Aleppo Codex, partially preserved) and another one (the Leningrad Codex) was copied from it. The text of the Leningrad Codex has been reproduced in the modern edition of the Biblia Hebraica Stuttgartensia.

Emanuel Tov

The Greek New Testament

By 1989 the Münster Institute for New Testament Textual Research had catalogued the number of manuscripts of the Greek New Testament at a total of 5488, in the customary categories: 96 papyri, 299 uncials, 2812 minuscules, and 2281 lectionaries (containing selected passages arranged according to the liturgical year, for use in church services). Most lectionaries are in minuscule script, but the oldest are in uncial. When compared to the numbers of existing manuscripts of ancient classical writers, these numbers are extraordinarily large. However, most of the papyri are fragmentary; some consist of only one or two leaves. Moreover, only 59 manuscripts contain the entire New Testament, and of these only one (Codex Sinaiticus) is uncial; roughly 1500 contain only the Gospels, and the book of Revelation survives in only 287 copies (many of which alternate between sections of text and patristic commentary).


Very few manuscripts were dated by their scribes, and the exceptions tend to be late. Fortunately, secular documents of various sorts carrying dates have survived, enabling paleographers to compare handwriting and ascertain within broad limits the date of a biblical manuscript.

The oldest known New Testament manuscript is a papyrus fragment, ca. 9 by 6.35 cm (3 1/2 by 2 1/2 in), dated to 100–150 CE, which preserves five verses from John 18. P52, as it is called, is now kept at the John Rylands Library in Manchester, England. The oldest substantial portions of the New Testament are the Bodmer papyrus of John (P66), now in Geneva, and the Chester Beatty papyrus (P46) in Dublin and Ann Arbor, which contains ten Pauline letters; both have been dated to ca. 200 CE. The oldest parchment New Testament copies are Codex Vaticanus and Codex Sinaiticus, both from the fourth century. From roughly 300–1000 CE, about three hundred manuscripts remain; and from about 1000–1500 (Gutenberg invented movable‐type printing around 1450), about two thousand copies have survived. Since of all these known copies only fifty‐nine contain the entire New Testament, it is clear that prior to the invention of printing relatively few individuals or even congregations possessed a complete New Testament.

Identification of Manuscripts.

Previously, a manuscript was identified by the name of the owner (e.g., Codex Bezae), the place where it is now preserved (Codex Vaticanus, Codex Washingtoniensis), or the place of its purported origin (Codex Alexandrinus). A simpler, more systematic nomenclature was invented by a Swiss scholar, Johann Jakob Wettstein (1693–1754), who assigned a capital roman letter to each uncial and an arabic numeral to each minuscule. Over time, however, the letters of the alphabet proved too few for the growing number of known uncial manuscripts, so Hebrew and Greek letters came into use; eventually, however, these too proved insufficient, and Caspar René Gregory (1846–1917) proposed that each uncial manuscript be assigned a numeral prefixed by 0. Only for the chief uncials (such as Vaticanus) is the earlier letter system retained.


Early on, Christian communities abandoned use of the scroll in favor of the leaf‐book (codex), which was less expensive and more convenient (see Books and Bookmaking in Antiquity). Over time, parchment slowly supplanted papyrus as the material of choice for books. Several deluxe copies of the scriptures survive from the sixth century; these were prepared for nobles and high ecclesiastics on purple‐dyed parchment, in script written with silver ink (and in some cases gold for initial letters). Once scribes had finished copying the text, artists added illumination and illustrations (miniatures). The text of the Gospels—and eventually that of other New Testament books—was divided into numbered sections; Eusebius (d. ca. 342) collected these into tables that showed parallel passages in four Gospels, in three Gospels, in two, and finally in those sections unique to each gospel. In some cases, scribes provided information as to the total number of lines in a gospel or letter.

In time individual books of the New Testament were provided with a variety of prologues and other prefatory material, supplying certain information about the author, contents, and character of the work. Some prologues are anonymous, while others are attributed to patristic authors (such as Chrysostom, Theodoret, Euthalius). Occasionally, lectionary notations were entered into straight‐text gospel manuscripts. Likewise, musical notations to assist readers in chanting the lessons were placed above the line of text in green or red to contrast with the text's brown or black ink.

Important Manuscripts.

The following lists specify the siglum (identification), date, contents, and place of preservation:


P38: ca. 300; portions of Acts 18.27–19.16; Ann Arbor, University of Michigan Library.

P45: third century; portions of the four Gospels and Acts; part in Dublin, Chester Beatty Library, and part in Vienna, Austrian National Library.

P46: ca. 200; ten letters of Paul; part in Dublin, Chester Beatty Library, and part in Ann Arbor, University of Michigan Library.

P47: third century; Revelation 9.10–17.2; Dublin, Chester Beatty Library.

P52: first half of second century; John 18.31–33, 37–38; Manchester, John Rylands Library.

P66: ca. 200; portions of the gospel of John, part in Cologny‐Geneva, Bodmer Library, and part in Dublin, Chester Beatty Library.

P75: third century; most of the gospel of Luke and two‐thirds of the gospel of John; Cologny‐Geneva, Bodmer Library.

Uncial manuscripts.

χ: fourth century; entire New Testament; London, British Library (Codex Sinaiticus).

A: fifth century; New Testament, with a few lacunae; London, British Library (Codex Alexandrinus).

B: fourth century; Gospels, Acts, letters (lacking 1 Timothy through Philemon, and Hebrews 9.14 through the end of Revelation); Rome, Vatican Library (Codex Vaticanus).

C: fifth century; New Testament (a palimpsest with many lacunae); Paris, Bibliothèque Nationale (Codex Ephraemi).

D: fifth century; Gospels and Acts in Greek and Latin (with lacunae); Cambridge University Library (Codex Bezae).

W: fifth century; four Gospels (with lacunae) in the order Matthew, John, Luke, Mark; Washington, Smithsonian Institution, Freer Gallery of Art (Codex Washingtoniensis).

θ: ninth century; four Gospels (with a few lacunae); Tiflis, Manuscript Institute (Codex Koridethi).

0169: fourth century; Rev. 3.19–4.3; Princeton Theological Seminary Library.

0212: third century; Diatessaron, brief portions of Matthew, Mark, Luke, and John; New Haven, Yale University, P. Dura 10.

Minuscule manuscripts.

1: twelfth century; New Testament, lacking the book of Revelation; Basel, University Library. Related to manuscripts 118, 131, 209, and others as family 1, reflecting the kind of text used at Caesarea in the fourth century.

13: thirteenth century; four Gospels; Paris; Bibliothèque Nationale. Related to 69 and ten other minuscule manuscripts as family 13 (they have John 7.53–8.12 after Luke 21.38), which goes back to an archetype in Calabria.

33: ninth century; New Testament, lacking the book of Revelation; Paris, Bibliothèque Nationale. Sometimes called “the queen of the minuscules,” its text is similar to that in B and χ.

461: 835 CE (oldest dated New Testament Greek manuscript); four Gospels; St. Petersburg Public Library.

565: ninth century; four Gospels; St. Petersburg Public Library; a magnificent purple parchment manuscript written in gold letters.

614: thirteenth century; Acts and letters (lacuna in Jude 3–25); Milan, Biblioteca Ambrosiana. Related to the text of Acts in Codex Bezae.

700: eleventh century; four Gospels; London, British Library. In Luke 11.2 it replaces the petition of the Lord's Prayer, “Thy kingdom come,” with “Thy holy Spirit come upon us and cleanse us.”

1739: tenth century; Acts, general letters, Pauline letters; Mount Athos, Lavra. Agrees frequently with the text used by Origen (d. ca. 254).

2400: thirteenth century; New Testament but without the book of Revelation; Chicago, University of Chicago Library. A splendid manuscript with ninety‐eight miniatures, in a silver case.


The earliest printed Greek New Testament was volume 5 (1514) of the Complutensian Polyglot Bible, which was sponsored by Cardinal Ximénes of Alcalá, Spain; however, the pope did not grant permission to publish it until 1520. The first published Greek New Testament, edited by Erasmus of Rotterdam and issued by Johannes Froben of Basel in 1516, relied on only a few very late manuscripts. For the book of Revelation there was only one Greek manuscript available, and it lacked the last six verses; for them, Erasmus translated the Latin Vulgate into what he supposed the Greek text should read. This type of New Testament text, based on late and sometimes imperfect manuscripts, became the so‐called textus receptus (“received text”). Only toward the end of the nineteenth century, once much earlier and superior New Testament manuscripts became available, could scholars produce more accurate editions.

See Textual Criticism


Bruce M. Metzger