Textual criticism

No original manuscript of any biblical book has survived. What we possess instead are copies of copies, transmitted across centuries by scribes working by hand in conditions that inevitably introduced changes. Textual criticism is the scholarly discipline that examines these surviving manuscripts to determine, as closely as possible, what the original authors wrote.¹ Far from being a hostile enterprise, textual criticism is a foundational tool used by scholars of all religious persuasions, including devout believers, to understand and accurately translate the biblical text.²

The discipline applies rigorous methods developed over centuries of classical scholarship to the unique challenges posed by biblical manuscripts. For the New Testament alone, scholars work with more than 5,800 Greek manuscripts, over 10,000 Latin manuscripts, and thousands more in Syriac, Coptic, Armenian, and other ancient languages.^{1, 3} For the Hebrew Bible, the textual situation is different but equally complex, involving the Masoretic Text, the Septuagint, the Dead Sea Scrolls, and other ancient witnesses.⁴ By comparing these manuscripts systematically, textual critics can identify where scribes made errors, where they made intentional changes, and which readings most likely preserve the original text.

What textual criticism is

Textual criticism, sometimes called "lower criticism" to distinguish it from literary or source criticism, is the science and art of evaluating manuscript evidence to reconstruct an ancient text.¹ The term "criticism" here carries no negative connotation; it derives from the Greek krinein, meaning "to judge" or "to discern." Textual critics are not criticizing the Bible but carefully evaluating the evidence to determine what its authors originally wrote.²

The need for textual criticism arises from a simple historical reality: before the printing press, every copy of a text had to be made by hand. Each time a scribe copied a manuscript, there was opportunity for error. Letters could be misread, words skipped, or lines accidentally repeated. Over centuries of transmission, these small changes accumulated, creating manuscript traditions that diverge from one another in thousands of places.^{1, 5}

Bruce Metzger, widely considered the foremost New Testament textual critic of the twentieth century, explained the goal of the discipline in his classic work The Text of the New Testament: Its Transmission, Corruption, and Restoration: because we no longer possess the autographs—the original documents written by the biblical authors—scholars must evaluate the thousands of existing manuscript copies to reconstruct what the original documents said.¹ This reconstruction is never absolutely certain, but the cumulative weight of evidence allows scholars to determine the original reading with high probability in the vast majority of cases.²

Old Testament manuscript traditions

The textual criticism of the Hebrew Bible involves three major manuscript traditions: the Masoretic Text, the Septuagint, and the Dead Sea Scrolls. Each provides a window into the text at different points in its transmission history, and comparing them reveals both the stability and the variability of the biblical text over time.⁴

The Masoretic Text (MT) is the authoritative Hebrew and Aramaic text of the 24 books of the Hebrew Bible in Rabbinic Judaism. It was primarily copied, edited, and distributed by a group of Jewish scribes known as the Masoretes between the seventh and tenth centuries CE.⁶ The oldest complete manuscript of the Masoretic Text is the Leningrad Codex, dating to 1009 CE, which serves as the base text for modern critical editions including the Biblia Hebraica Stuttgartensia (BHS) and its successor, the Biblia Hebraica Quinta (BHQ).⁷ The Masoretes developed an elaborate system of marginal notes, vowel pointing, and accent marks to preserve the text with extraordinary precision, yet they were working with manuscripts already a millennium removed from the original compositions.⁶

The Septuagint (LXX) is the Greek translation of the Hebrew Scriptures, begun in Alexandria, Egypt, around 280-270 BCE to serve Greek-speaking Jews who had lost proficiency in Hebrew.⁴ For centuries, scholars assumed that where the Septuagint differed from the Masoretic Text, the Greek translators had simply made errors or taken liberties. The Dead Sea Scrolls dramatically changed this picture. Scholars now recognize that many Septuagint differences reflect not translation errors but a different pre-Christian Hebrew text tradition.^{4, 8}

The Dead Sea Scrolls, discovered beginning in 1947 in caves near Qumran, represent the oldest surviving biblical manuscripts, dating from the third century BCE to the first century CE.⁸ These manuscripts predate the Masoretic Text by nearly a millennium and provide direct evidence of textual diversity in the Second Temple period. Approximately 60% of the biblical manuscripts found at Qumran reflect a proto-Masoretic text type, but smaller proportions align with the Septuagint or Samaritan Pentateuch traditions, while others represent independent textual traditions.^{4, 8} This discovery confirmed that by 200 BCE to 70 CE, there was no single uniform Bible text; several textual traditions coexisted.⁴

The Great Isaiah Scroll (1QIsa^a), the most famous of the biblical scrolls, contains all sixty-six chapters of Isaiah and dates to approximately 125 BCE.²⁵ When compared with the Masoretic Text preserved a thousand years later, the scroll is approximately 95% identical, with over 2,600 variants—most being minor spelling differences, word order variations, or scribal corrections.^{25, 26} This demonstrated remarkable transmission fidelity across a millennium while also revealing that some textual diversity existed in the Second Temple period.

The two most important Masoretic manuscripts are the Aleppo Codex and the Leningrad Codex. The Aleppo Codex, produced around 920 CE in Tiberias, was considered the most authoritative text in Judaism and was vocalized by the renowned Masorete Aaron ben Moses ben Asher.²⁷ Tragically, approximately 40% of the manuscript was lost or damaged when the synagogue housing it in Aleppo, Syria was attacked during anti-Jewish riots in 1947.²⁷

Several other textual traditions provide additional witnesses to the Hebrew Bible. The Samaritan Pentateuch, the scripture of the Samaritan community, preserves an independent textual tradition of the first five books and contains approximately 6,000 differences from the Masoretic Text, though most are minor.²⁸ Intriguingly, Dead Sea Scroll fragments of Deuteronomy support the Samaritan reading of "Gerizim" rather than "Ebal" in Deuteronomy 27:4, suggesting this may have been the earlier reading.²⁸ The Targums are Aramaic translations and paraphrases of the Hebrew Bible that originated during the Babylonian Exile, and unlike literal translations, they often include interpretive expansions valuable for understanding ancient Jewish interpretation.²⁹ The Peshitta, the Syriac Bible, was translated largely from Hebrew with some Septuagint influence, probably by the second century CE, and provides an early witness to the biblical text in a language closely related to Aramaic.³⁰

New Testament manuscript evidence

The New Testament is the best-attested work of antiquity in terms of manuscript evidence. Scholars have catalogued more than 5,800 Greek manuscripts, ranging from tiny fragments to complete Bibles, spanning from the second century to the invention of printing.^{1, 3} In addition, there are over 10,000 Latin manuscripts, and thousands more in Syriac, Coptic, Armenian, Georgian, Ethiopic, and other ancient languages.³ This abundance of evidence is both a blessing and a challenge: the more manuscripts available, the more data for reconstruction, but also the more variants to evaluate.

The earliest New Testament manuscripts are written on papyrus, a writing material made from reeds. Among the most significant are P52 (also known as the Rylands Papyrus), a small fragment of John's Gospel typically dated to 125-150 CE, making it the earliest surviving New Testament manuscript.⁹ P66, containing most of John's Gospel, is dated to approximately 125-175 CE, while P75, preserving significant portions of Luke and John, dates to 175-225 CE.¹⁰ P75 is particularly important because its text agrees approximately 83% of the time with Codex Vaticanus, demonstrating a stable transmission line from the late second century into the fourth.¹⁰

Major early New Testament manuscripts^{1, 9, 10}

The great uncial codices of the fourth and fifth centuries are among the most important witnesses to the New Testament text. Codex Vaticanus (designated B), housed in the Vatican Library, dates to approximately 325-350 CE and is generally regarded by textual scholars as the highest-quality Greek New Testament manuscript that survives.¹¹ Codex Sinaiticus (designated by the Hebrew letter aleph), discovered at St. Catherine's Monastery in 1844, dates to approximately 330-360 CE and is the earliest complete copy of the New Testament in Greek.¹² Both manuscripts represent the Alexandrian text type, which scholars generally consider closest to the original.¹³

More substantial than these early fragments are the Chester Beatty Papyri, acquired in the 1930s from an Egyptian dealer. P45 (third century) contains portions of all four Gospels and Acts, P46 (circa 175–225 CE) preserves most of the Pauline epistles, and P47 (late third century) contains chapters 6–17 of Revelation, providing the earliest extensive text of the Apocalypse.³¹ The Bodmer Papyri, discovered in the 1950s, include P66 (containing nearly the complete Gospel of John, initially dated to around 200 CE though recent scholarship has argued for later dates) and P75 (containing substantial portions of Luke and John), whose text closely resembles the fourth-century Codex Vaticanus, suggesting this textual tradition extends back to an early period.^{32, 33}

Beyond the great fourth-century uncials, Codex Alexandrinus (A), dating to the fifth century and now in the British Library, contains nearly the complete Bible, though portions of Matthew, John, and 2 Corinthians are lost.³⁴ Codex Bezae (D), a bilingual Greek-Latin manuscript from the fifth century, is the primary representative of the Western text-type and exhibits a remarkably free text with many unique readings, interpolations, and omissions not found in other manuscripts.³⁵ From the ninth century onward, the minuscule script replaced uncials as the dominant form of Greek writing; over 2,900 minuscule New Testament manuscripts have been catalogued, with Minuscule 33, the ninth-century "Queen of the Cursives," frequently agreeing with Codex Vaticanus.³⁶ Lectionaries—manuscripts containing biblical readings arranged for liturgical use—number approximately 2,400 and predominantly reflect the Byzantine text-type.³⁷ Ancient translations into Latin, Syriac, Coptic, and other languages provide additional witnesses, while over one million quotations in the writings of church fathers offer evidence for the form of the text used in specific times and places.³⁸

Manuscripts are grouped into "textual families" or "text types" based on shared patterns of readings. The four main families are the Alexandrian, Byzantine, Western, and Caesarean.¹³ The Alexandrian text type, represented by Codex Vaticanus and Codex Sinaiticus, is characterized by brevity and is generally considered oldest and most faithful to the original.¹³ The Byzantine text type, which underlies the King James Version, became dominant in the medieval period but is usually regarded as further removed from the original autographs.¹⁴ The Western text type, represented by Codex Bezae, exhibits a tendency toward paraphrase and expansion.¹³

Methods of textual criticism

Textual critics employ two main categories of evidence when evaluating variant readings: external evidence, which concerns the manuscripts themselves, and internal evidence, which concerns the readings within those manuscripts.^{1, 2}

External evidence includes the age of the manuscripts supporting a reading, their geographic distribution, and their textual family. A reading attested by early manuscripts from diverse geographic regions carries more weight than one found only in late manuscripts from a single location.¹ The quality of the manuscripts also matters: some scribes were more careful than others, and some manuscripts show evidence of being copied from earlier, more reliable exemplars.²

Internal evidence involves evaluating the readings themselves to determine which is most likely original. Two famous principles guide this analysis. The first, lectio difficilior potior ("the more difficult reading is stronger"), recognizes that scribes were more likely to replace unusual or difficult words with familiar, easier ones than vice versa.¹⁵ If one manuscript has an obscure word and another has a common synonym, the obscure word is more likely original because a scribe would be unlikely to introduce difficulty. The second principle, lectio brevior potior ("the shorter reading is stronger"), is based on the observation that scribes more often added explanatory material than deleted text.¹⁶ However, this principle is applied with caution in modern criticism because scribes could also accidentally omit text through errors like homoeoteleuton.¹⁶

Perhaps the most important internal criterion is determining which reading best explains the origin of the others. If reading A could easily have given rise to readings B, C, and D through known scribal tendencies, but B, C, and D could not plausibly have produced reading A, then A is likely original.^{1, 2} Modern textual criticism employs what scholars call "reasoned eclecticism," weighing both external and internal evidence without mechanically preferring one over the other.¹⁵

Types of textual variants

Textual variants fall into two broad categories: unintentional errors arising from the physical and mental challenges of copying by hand, and intentional changes made by scribes for various reasons.^{1, 5}

Unintentional errors include several common types. Homoeoteleuton ("similar ending") occurs when two lines end with the same or similar letters, causing the scribe's eye to skip from one to the other and omit the intervening text.¹⁷ Dittography ("double writing") is the accidental repetition of a letter, word, or phrase.¹⁷ Haplography ("single writing") is the opposite: writing once what should have been written twice.¹⁷ Scribes also made errors of hearing when texts were dictated, confusing similar-sounding words, and errors of memory when they held a phrase in mind and wrote it slightly differently.⁵

Intentional changes present a more complex picture. Harmonization occurs when scribes altered one passage to match a parallel passage elsewhere, particularly common in the Gospels where similar stories appear in multiple accounts.⁵ Scribes sometimes added explanatory glosses—brief notes clarifying the meaning of a word or phrase—that later copyists incorporated into the text itself.¹ Theological corrections were made when scribes altered readings they found doctrinally problematic, smoothing out apparent contradictions or strengthening statements about Christ's divinity.^{5, 18} Bart Ehrman's research has documented how scribes altered texts to harmonize conflicting portrayals of Jesus, to de-emphasize the role of women in the early church, and to counter theological positions they considered heretical.¹⁸

Approximate breakdown of New Testament textual variants^{3, 18}

The scale of variation

The New Testament contains approximately 140,000 words in Greek. The number of textual variants among surviving manuscripts is estimated at 400,000 to 500,000—more variants than words in the text itself.^{3, 19} This striking statistic requires careful interpretation. A 2016 study by Peter Gurry in New Testament Studies estimated approximately 500,000 non-spelling variants, while earlier estimates by Bruce Metzger and Bart Ehrman ranged from 200,000 to 400,000.^{19, 18}

The high number is partly a function of the wealth of manuscript evidence: the more manuscripts that survive, the more opportunities to observe variation. A single spelling error copied into a hundred manuscripts counts as a hundred variants.³ Scholars consistently emphasize that the vast majority of these variants—estimates range from 95% to 99%—are insignificant: spelling differences, word order variations, easily recognized scribal blunders, and other changes that do not affect meaning.^{3, 18}

The variants that are both meaningful (affecting interpretation) and viable (having a reasonable claim to being original) constitute less than 1% of the total—fewer than 4,000 of the 400,000 or more variants.³ Even among these, most involve minor points of interpretation rather than major doctrines. No essential Christian doctrine depends on a textually disputed passage, a point acknowledged by scholars across the theological spectrum.^{2, 18}

The Hebrew Bible presents fewer overall variants than the New Testament, partly because fewer manuscripts survive and partly because the Masoretes succeeded in standardizing the text from the medieval period onward.⁴ However, the Dead Sea Scrolls revealed that significant textual diversity existed before that standardization. Comparing Qumran manuscripts with the Masoretic Text shows differences in spelling, wording, and occasionally entire passages or verses.⁸

Notable examples of significant variants

Several passages commonly found in English Bibles are now recognized by textual scholars as later additions to the original text.¹ These examples illustrate both the power of textual criticism to identify interpolations and the complexity of the transmission history.

The Comma Johanneum (1 John 5:7-8) is perhaps the most famous textual interpolation. In the King James Version, the passage reads: "For there are three that bear record in heaven, the Father, the Word, and the Holy Ghost: and these three are one. And there are three that bear witness in earth, the Spirit, and the Water, and the Blood." The words about the Father, Word, and Holy Ghost are found in no Greek manuscript before the fifteenth century and are absent from all early Greek manuscripts, early translations, and early church fathers.²⁰ The passage first appears in the writings of Priscillian, a Spanish sect leader in the late fourth century, and entered the Latin tradition as a marginal gloss that was later incorporated into the text.²⁰ Modern translations, both Catholic and Protestant, either omit the passage or relegate it to a footnote.²⁰

The Pericope Adulterae, the story of the woman caught in adultery (John 7:53-8:11), is absent from the earliest and best Greek manuscripts, including P66, P75, Codex Sinaiticus, and Codex Vaticanus.²¹ The passage contains vocabulary not found elsewhere in John's Gospel and interrupts the narrative flow between chapters 7 and 8.²¹ In manuscripts that do contain it, the passage appears in different locations—some place it after John 7:36, others after John 21:25, and some even place it in the Gospel of Luke.²¹ The first Greek manuscript to contain the passage is Codex Bezae from the fifth or sixth century.²¹ Most modern translations print the story in brackets with a note explaining its textual status.

The longer ending of Mark (16:9-20), which includes resurrection appearances and the famous passage about handling snakes and drinking poison, is absent from Codex Sinaiticus and Codex Vaticanus, both of which end at Mark 16:8 with the women fleeing the empty tomb in fear.²² In the fourth century, Eusebius and Jerome reported that nearly all the Greek manuscripts known to them ended at 16:8.²² The vocabulary and style of verses 9-20 differ markedly from the rest of Mark, and there is an awkward transition between verse 8 and verse 9.²² The vast majority of New Testament scholars conclude that the longer ending is not original to Mark.²²

Luke 22:43–44 describes an angel strengthening Jesus during his agony in the Garden of Gethsemane, followed by the statement that "his sweat was as it were great drops of blood falling down to the ground." While most manuscripts include these verses, they are absent from P75, the original hand of Codex Sinaiticus, Codex Alexandrinus, and Codex Vaticanus—the earliest and generally most reliable witnesses.³⁹ Bart Ehrman argues the verses were inserted to counter Docetism—the belief that Jesus only appeared to suffer—by emphasizing his genuine physical distress, while others note that church fathers including Justin Martyr and Irenaeus quote the passage in the second century, earlier than any manuscript that omits it. Modern critical texts include the verses in double brackets, acknowledging the uncertainty.³⁹

The differences between the Masoretic Text and Septuagint in 1 Samuel 17-18, the story of David and Goliath, illustrate the kind of large-scale variation found in the Hebrew Bible. The Septuagint version is approximately 40-45% shorter than the Masoretic Text, missing substantial portions including David's conversation with his brothers, his second meeting with Saul, and the aftermath of the battle.²³ Scholars debate whether the Masoretic Text contains secondary expansions or the Septuagint reflects a Hebrew text that was later shortened.²³ Additionally, Goliath's height is given as "six cubits and a span" (approximately 9 feet 9 inches) in the Masoretic Text but "four cubits and a span" (approximately 6 feet 9 inches) in the Septuagint, the Dead Sea Scrolls, and Josephus.²³

Modern critical editions

The results of textual criticism are published in critical editions of the biblical text. These editions present a reconstructed text based on scholarly analysis of the manuscript evidence, accompanied by a critical apparatus listing significant variant readings and the manuscripts that support them.¹

For the New Testament, the standard scholarly edition is the Novum Testamentum Graece, commonly known as the Nestle-Aland edition after its most influential editors, Eberhard Nestle and Kurt Aland.²⁴ Currently in its 28th edition (NA28), this text is published by the German Bible Society and edited by the Institute for New Testament Textual Research in Münster.²⁴ The Greek text of NA28 is identical to that of the United Bible Societies' Greek New Testament, Fifth Revised Edition (UBS5), although the two editions differ in their critical apparatus: the UBS edition is designed for translators and focuses on variants important for meaning, while the Nestle-Aland includes a more comprehensive selection of variants for scholarly study.²⁴ Most modern Bible translations are based on the Nestle-Aland/UBS text.²

For the Hebrew Bible, the standard critical edition is the Biblia Hebraica Stuttgartensia (BHS), published in 1977 and based on the Leningrad Codex.⁷ The BHS is being successively replaced by the Biblia Hebraica Quinta (BHQ), which is appearing in individual fascicles beginning in 2004 and offers updated textual notes incorporating Dead Sea Scrolls evidence.⁷ Both editions print the Masoretic Text as preserved in the Leningrad Codex, with critical apparatus noting variants from other manuscripts and versions.⁷

What textual criticism reveals

Textual criticism demonstrates that the transmission of the biblical text was a thoroughly human process, subject to all the limitations and errors that accompany any work done by hand over many centuries.^{1, 18} Scribes made mistakes of eye, ear, and memory. They introduced changes, both accidental and intentional, that accumulated over generations of copying. The text that reached the medieval period and eventually the printing press was not identical to what the original authors wrote.⁵

At the same time, textual criticism reveals the remarkable stability of the text's overall transmission. The vast majority of variants are trivial, and even meaningful variants rarely affect central matters of doctrine or ethics.^{2, 3} The sheer abundance of manuscript evidence allows scholars to identify scribal errors and reconstruct the original text with a high degree of confidence in most passages.¹

The discipline challenges certain claims about biblical preservation while confirming others. It is not accurate to say that the Bible has been "perfectly preserved" in every word, since the manuscripts clearly differ from one another in thousands of places.¹⁸ But neither is it accurate to say that the text is hopelessly corrupt or that we cannot know what the biblical authors wrote. Textual criticism provides a disciplined method for navigating between these extremes, acknowledging uncertainty where it exists while establishing the text on solid scholarly footing.^{1, 2}

Perhaps most significantly, textual criticism invites readers to engage with the Bible as a collection of ancient documents with a real history. The manuscripts were copied by real people in real places over real centuries. Understanding this history enriches rather than diminishes appreciation for the text, revealing the immense labor of preservation while honestly acknowledging its limitations.^{2, 18}