Análise Acadêmica do NT

Esta página é o complemento acadêmico do estudo de Redação Teológica do NT. Fornece textos primários gregos, evidência manuscrita representativa por data e tipo textual, dados do aparato UBS/NA28, testemunho patrístico e citações bibliográficas das variantes discutidas no estudo principal. Está organizada por nível de confiança – os casos mais sólidos primeiro – para que o leitor possa avaliar a evidência de forma independente.

Como ler esta página

Nem todas as variantes são iguais. Esta página distingue três níveis:

O leitor deve considerar os casos de Nível 1 com confiança, abordar os casos de Nível 2 com abertura informada e tratar os casos de Nível 3 como argumentos provocadores que requerem mais pesquisa. Apresentamos os três porque a honestidade intelectual exige distinguir o que sabemos do que suspeitamos.

1. Metodologia – Como funciona a crítica textual do NT

The manuscript base

The New Testament is preserved in approximately 5,800 Greek manuscripts, plus roughly 10,000 Latin manuscripts, 9,300 manuscripts in other ancient versions (Syriac, Coptic, Armenian, Georgian, Ethiopic, Gothic, Slavonic), and thousands of patristic quotations – roughly 25,000 handwritten witnesses in total. By raw number of manuscript witnesses, the NT is unusually well attested among ancient literary corpora; Homer's Iliad, the next most-attested, survives in approximately 1,900 manuscripts.

The volume of evidence produces a corresponding volume of variation. The commonly cited range is 300,000–400,000 textual variants across all witnesses. Bart D. Ehrman (Misquoting Jesus, HarperSanFrancisco, 2005, p. 10) observed: "There are more variations among our manuscripts than there are words in the New Testament." The NT contains approximately 138,000 Greek words. However, the vast majority of variants (estimated 70–80%) are trivial: spelling differences, movable nu, itacisms, and word-order variations. Only a small fraction – perhaps 1–2% – affect the meaning of the text in any significant way.

NA28 and UBS5

The two standard critical editions of the Greek NT share the same reconstructed text but differ in their apparatus:

• Novum Testamentum Graece, 28th edition (NA28) – published 2012 by the Institut für Neutestamentliche Textforschung (INTF) in Münster. Contains a comprehensive critical apparatus listing variant readings, manuscript support, and sigla. Designed for scholars and translators who need full access to the textual evidence.
• The Greek New Testament, 5th edition (UBS5) – published 2014 by United Bible Societies. Contains the same Greek text as NA28 but selects only variants significant for translation, providing fuller manuscript evidence and confidence ratings for each decision.

UBS confidence ratings

The CBGM

The Coherence-Based Genealogical Method (CBGM), developed by Gerd Mink at the INTF, was applied for the first time in the Catholic Epistles within NA28. Instead of assigning manuscripts to predetermined text-type families, the CBGM analyzes actual textual agreements and disagreements at every variation unit to construct textual-flow diagrams showing how readings likely developed. About one-third of the 33–34 changes in the Catholic Epistles between NA27 and NA28 moved toward Byzantine readings – a significant shift from earlier Alexandrian preference. The method is being extended to Acts and the rest of the NT. See Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism (SBL Press, 2017).

External evidence criteria

• Age – Earlier manuscripts are generally preferred, all else being equal. A 3rd-century papyrus carries more weight than a 10th-century minuscule. However, a late manuscript can preserve an early reading if its ancestor was early.
• Text-type – Readings attested across multiple independent text-types carry more weight than readings confined to one tradition.
• Geographic distribution – A reading attested in Egypt, Syria, and Italy is more likely original than one found only in Constantinople, because geographic diversity implies independent transmission lines.
• Independence – Manuscripts must be weighed, not counted. Five hundred Byzantine manuscripts copied from one archetype count as one witness, not five hundred. "Manuscripts are to be weighed, not counted" (Bengel, 1734).

Internal evidence criteria

• Lectio difficilior potior ("the harder reading is to be preferred") – Scribes tend to smooth difficulties, not create them. A reading that is theologically awkward, grammatically rough, or potentially embarrassing is more likely original because no scribe would deliberately introduce it. The principle has limits: a reading can be "harder" simply because a scribe made a mistake.
• Lectio brevior potior ("the shorter reading is to be preferred") – Scribes are more likely to add than to omit. Glosses, harmonizations, and liturgical expansions inflate the text over time. Exceptions include homoeoteleuton (eye-skip from similar endings), which produces accidental omissions.
• Transcriptional probability – The reading that best accounts for all other variants as derivative is most likely original.
• Intrinsic probability – The reading most consistent with the author's known vocabulary, syntax, theology, and rhetorical habits is preferred.

Text-types

2. 1 Timóteo 3:16 – ΘΣ vs ΟΣ Nível 1

Texto grego

UBS rating: {C}. NA28 reads: hos ephanerōthē en sarki (ὅς ἐφανερώθη ἐν σαρκί) – "who was manifested in the flesh."

A third reading, ho (ὅ, neuter "which"), occurs in the Western tradition, harmonizing with the nearest antecedent mystērion ("mystery"). It indirectly supports hos as the base reading, since ho likely arose as a grammatical adjustment rather than independently.

The nomen sacrum mechanism

In uncial Greek manuscripts, theos was routinely abbreviated as ΘΣ with a horizontal overline (a nomen sacrum). The relative pronoun hos was written as ΟΣ. The visual difference between ΟΣ and ΘΣ is a single horizontal stroke inside the first letter – the crossbar that turns omicron (Ο) into theta (Θ). If that crossbar is faint, blotted, added by a later hand, or bleeding through from the verso, the two forms become virtually indistinguishable.

Johann Jakob Wettstein (1730) demonstrated by physical examination of Codex Alexandrinus that the horizontal line inside the theta was not original but bleed-through from the other side of the parchment. His published challenge to the reading contributed to charges of Socinianism and consequences for his clerical career.

Manuscript evidence for ΟΣ (hos)

Versional support: all ancient versions presuppose a pronoun reading (ὅς or ὅ). No ancient version supports theos. Syriac Peshitta uses a relative pronoun; Gothic (c. 350–360), Ethiopic, and Coptic (Sahidic, Bohairic) all presuppose hos.

Manuscript evidence for ΘΣ (theos)

No uncial manuscript earlier than the 8th/9th century (Ψ) supports theos in its original hand. The reading dominates the later Byzantine tradition: Ψ (044, 8th/9th c.), K (9th c.), L (9th c.), P (9th c.), 075 (10th c.), 0150 (9th c.), and 21+ minuscules including 6, 81, 104, 181, 326, 330, 436, 451, 614, 629, 630, 1241, 1505, 1739, 1877, 1881.

Evidência patrística

Supporting hos: Origen (c. 185–254, earliest witness), Epiphanius (c. 310–403), Jerome (c. 347–420), Cyril of Alexandria (c. 376–444), Liberatus of Carthage (d. c. 566). Supporting theos: Gregory of Nyssa (c. 335–395, earliest witness to theos), Didymus the Blind (c. 313–398), John Chrysostom (c. 349–407). No patristic writer prior to the last third of the 4th century testifies to the reading theos.

Estudos-chave

"No uncial (in the original hand) earlier than the eighth or ninth century (Ψ) supports θεός; all ancient versions presuppose either ὅς or ὅ; and no patristic writer prior to the last third of the fourth century testifies to the reading θεός."

– Bruce M. Metzger, A Textual Commentary on the Greek New Testament (2nd ed., 1994), pp. 573–574.

Gordon D. Fee (1 and 2 Timothy, Titus, NIGTC, 1988, pp. 90–92) argues decisively for hos, noting the textual evidence is "almost conclusive" against theos. He observes that the hymnic/confessional fragment naturally begins with a relative pronoun, as is characteristic of such fragments elsewhere (Phil 2:6; Col 1:15; Heb 1:3).

Bart D. Ehrman (The Orthodox Corruption of Scripture, OUP, 1993, pp. 77–78) classifies this as an "anti-adoptionistic" corruption: scribes altered the text to oppose those who claimed Christ was a man but not God.

I. Howard Marshall (The Pastoral Epistles, ICC, 1999, pp. 521–525) supports hos and notes the relative pronoun opening is "typical of early Christian hymnic and confessional material" and that theos "cannot be traced further back than the late fourth century in the patristic tradition."

Philip H. Towner (The Letters to Timothy and Titus, NICNT, 2006, pp. 277–280) accepts hos and argues the confessional fragment "was almost certainly circulating as a recognized piece of tradition before its incorporation here."

3. 1 João 5:7–8 – O Comma Joanino Tier 1

Greek text

UBS rating: {A} – certain the Comma is not original.

Greek manuscripts containing the Comma

Only a handful of late Greek witnesses contain the Comma. The standard list (per Metzger and Wallace):

Manuscripts omitting the Comma

Every Greek manuscript before the 14th century omits the Comma entirely. This includes all papyri, all early uncials, and the vast majority of minuscules – over 500 continuous-text Greek manuscripts of 1 John. Key witnesses: P74 (7th c.), ℵ (c. 350), A (c. 400–440), B (c. 325–350), K (9th c.), L (9th c.), Ψ (8th/9th c.), and all early minuscules (33, 81, 104, 614, 630, 1505, 1739, etc.) and all lectionaries.

Versional evidence

The Comma is entirely a Latin phenomenon. It is absent from all Syriac versions (Peshitta, Harklean, Philoxenian), all Coptic versions (Sahidic, Bohairic), Ethiopic, Armenian, Slavonic, and Arabic. It appears in some Old Latin manuscripts from the 5th–6th century onward and progressively entered the Vulgate tradition through editorial insertion, not through Jerome's original translation.

The Erasmus story

Erasmus's first edition of the Greek NT (Novum Instrumentum Omne, 1516) omitted the Comma because none of his Greek manuscripts contained it. His second edition (1519) also omitted it, drawing criticism from Edward Lee and Diego López de Zúñiga. The traditional story that Erasmus "promised" to include the Comma if a single Greek manuscript was produced is likely apocryphal (Henk Jan de Jonge, Ephemerides Theologicae Lovanienses 56.4, 1980, pp. 381–389; Grantley McDonald, Biblical Criticism in Early Modern Europe, Cambridge, 2016, pp. 150–151). Erasmus included the Comma in his third edition (1522) after Codex Montfortianus was brought to his attention.

Patristic evidence

No Greek Church Father ever quotes the Comma. This includes writers with intense interest in Trinitarian theology who would have used it had they known it: Irenaeus, Clement of Alexandria, Origen, Athanasius, Basil of Caesarea, Gregory of Nazianzus, Gregory of Nyssa, John Chrysostom, Cyril of Alexandria, and John of Damascus.

Possible early allusions come from the Latin tradition: Tertullian (c. 200) writes "these three are one" but applies it to the earthly witnesses, not quoting a heavenly witnesses clause. Cyprian of Carthage (c. 250–258) is the earliest candidate for awareness of the concept. Priscillian of Avila (c. 380–385) is possibly the first to have clearly used the Comma text (Raymond E. Brown, The Epistles of John, AB, 1982, pp. 775–787).

Key scholarship

"That these words are spurious and have no right to stand in the New Testament is certain in the light of the following considerations."

– Bruce M. Metzger, A Textual Commentary (2nd ed., 1994), pp. 647–649.

Raymond E. Brown (The Epistles of John, AB, 1982, pp. 775–787): "All recent Roman Catholic scholarly discussion has recognized that the Comma is neither genuine nor authentic."

Daniel B. Wallace consistently finds late manuscripts containing the Comma to be back-translations from Latin, affirming that the Comma "has no place in the Greek text of the NT."

Grantley McDonald (Biblical Criticism in Early Modern Europe, Cambridge, 2016) provides the definitive reception history of the Comma from Erasmus through the modern period.

4. Mateus 24:36 – "Nem o Filho" Tier 1

Greek text

UBS rating: {B}. NA28 reads: Peri de tēs hēmeras ekeinēs kai hōras oudeis oiden, oude hoi angeloi tōn ouranōn oude ho huios, ei mē ho patēr monos.

Manuscript evidence: including oude ho huios

Patristic support for inclusion: Irenaeus (c. 130–202), Origen (c. 185–254), Epiphanius (c. 310–403), Didymus the Blind (c. 313–398), Cyril of Alexandria (c. 376–444), Hilary of Poitiers (c. 310–367), Ambrose (c. 339–397), Augustine (354–430). Versional support: Ethiopic, Armenian, Georgian.

Manuscript evidence: omitting oude ho huios

Supported primarily by the Byzantine majority: E, F, G, H, K, M, N, S, U, V, W, Γ, Δ, Π, Σ, plus the majority of minuscules. Some Alexandrian witnesses also omit: L (019, 8th c.), 33 (9th c.). Caesarean witnesses for omission: f1, 565, 579. Versional: Syriac Peshitta and Harklean, some Coptic, some Vulgate.

The Mark 13:32 parallel

In Mark 13:32, "nor the Son" is retained by virtually all manuscripts. Only Codex X (10th c.) and a few minor witnesses omit it. If scribes were motivated to remove the phrase for theological reasons, they conspicuously failed to do so in Mark. Charles Powell ("The Textual Problem of οὐδὲ ὁ υἱός in Matthew 24:36," bible.org, 2004) analyzed 61 probable harmonizations between Matthew and Mark and found scribes harmonize Mark to Matthew ~70% of the time, making deliberate addition to Matthew by assimilation to Mark statistically unlikely.

Key scholarship

"The omission of the words because of the doctrinal difficulty they present is more probable than their addition by assimilation to Mk 13.32."

– Bruce M. Metzger, A Textual Commentary (2nd ed., 1994), p. 51.

R. T. France (The Gospel of Matthew, NICNT, 2007, pp. 930–932): "The textual evidence favours its inclusion, and the theological embarrassment of the phrase provides an obvious motive for its omission from most later MSS."

W. D. Davies and Dale C. Allison (Commentary on Matthew, vol. 3, ICC, 1997, p. 377) include "nor the Son" and note the omission is "best explained as theologically motivated."

Bart D. Ehrman (The Orthodox Corruption of Scripture, 1993, pp. 91–92) treats the omission as a "proto-orthodox corruption" – scribes removed "nor the Son" to avoid the implication that Jesus lacked omniscience.

5. Marcos 16:9–20 – O final longo Tier 1

Greek text

UBS rates the shorter ending (at 16:8) as {A} – virtually certain. NA28 places 16:9–20 in double brackets [[ ]].

Mark 16:8 ends: kai oudeni ouden eipan; ephobounto gar. – "and they said nothing to anyone, for they were afraid." The final word is gar (γάρ, "for/because") – a postpositive conjunction. Ending a work on gar is virtually unprecedented in Greek literature, though P. W. van der Horst ("Can a Book End with a Gar?" JTS 23, 1972) demonstrated from Plotinus that it is linguistically possible.

Manuscript evidence: ending at verse 8

Also: approximately 100 Armenian manuscripts, several Georgian manuscripts (9th–10th c.), and one 5th-century Coptic manuscript.

Manuscript evidence: including 16:9–20

Over 1,600 Greek manuscripts include the longer ending: A (c. 400–440), C (c. 450), D (c. 400), K (9th c.), W (late 4th/early 5th c. – also includes the "Freer Logion" between 16:14 and 16:15), Θ, Ψ, f13, 33, 565, 700, 892, and 1,500+ Byzantine manuscripts. Versions: Vulgate, most Old Latin, Curetonian Syriac, Peshitta, Bohairic, most Sahidic, Gothic (4th c.).

Patristic evidence

No knowledge of the longer ending: Clement of Alexandria (d. c. 215) never quotes it. Origen (d. c. 254) never cites it. Eusebius (d. c. 339) states in Ad Marinum that "accurate copies" end Mark at 16:8 and that 16:9–20 is "not contained in all copies" and is "found in few." Jerome (d. 420): "Almost all the Greek copies do not have this concluding portion."

Knowledge of the longer ending: Irenaeus (d. c. 202) quotes Mark 16:19 in Adversus Haereses 3.10.6 – the earliest explicit quotation, c. 180 AD. Tatian (d. c. 185) apparently included material from the longer ending in his Diatessaron (c. 170).

Internal evidence: non-Markan vocabulary

The longer ending contains numerous words and constructions absent from Mark 1:1–16:8:

• poreuomai – 3x in 16:10, 12, 15; never in Mark 1–16:8 (29x in Matthew, 51x in Luke)
• theaomai – 16:11, 14; never in Mark 1–16:8
• meta tauta – 16:12; frequent in Luke and John, absent from Mark
• synergeo – 16:20; absent from all Gospels; Pauline term
• bebaioō – 16:20; foreign to Synoptic Gospels
• epakoloutheō – 16:20; foreign to Synoptic Gospels
• thanasimōn – 16:18; NT hapax legomenon

Mary Magdalene is reintroduced in 16:9 as though a new character, despite appearing in 15:47 and 16:1. The passage reads as a summary drawn from all four Gospels, not as original Markan narrative. It lacks Mark's characteristic vivid detail, historical present tense, and paratactic kai-chains.

Key scholarship

"The earliest ascertainable form of the Gospel of Mark ended with 16:8. … The vocabulary and style of verses 9–20 are non-Marcan."

– Bruce M. Metzger, A Textual Commentary (2nd ed., 1994).

James A. Kelhoffer (Miracle and Mission, WUNT II/112, Mohr Siebeck, 2000) argues the longer ending is a 2nd-century composition reflecting early Christian missionary theology. R. T. France (The Gospel of Mark, NIGTC, 2002) concludes 16:9–20 "cannot be accepted as part of what Mark himself wrote." N. Clayton Croy (The Mutilation of Mark's Gospel, Abingdon, 2003) argues Mark's original ending extended beyond 16:8 but was accidentally lost.

6. Lucas 3:22 – A voz batismal Nível 2

Greek text

UBS rating: {B} (for the standard reading).

Evidência manuscrita

Standard reading: P4 (3rd c.), ℵ (4th c.), A (5th c.), B (4th c.), K, L (8th c.), W (4th/5th c.), Θ (9th c.), Ψ (8th/9th c.), f1, f13, 33, Majority Text. Virtually all extant Greek manuscripts.

Western reading: D (Codex Bezae, c. 400) – the only Greek manuscript to preserve this reading. Also: Old Latin (a, b, c, d, ff2, l), dating 2nd–4th century.

Patristic evidence for the Western reading

Though attested in only one Greek manuscript, the patristic spread is remarkable and difficult to explain if the reading is purely secondary:

Key scholarship

Bart D. Ehrman (The Orthodox Corruption of Scripture, 1993/2011, pp. 62–67) argues the Psalm 2:7 reading is original to Luke based on: (1) transcriptional probability – scribes would change the adoptionist-sounding reading to match Mark, not vice versa; (2) intrinsic probability – the reading fits Luke's Christological pattern; (3) geographic breadth of patristic attestation across the Mediterranean.

Bruce M. Metzger (A Textual Commentary, 1994): "The Western reading, 'This day I have begotten thee,' which was widely current during the first three centuries, appears to be secondary, derived from Psalms 2:7."

François Bovon (Luke 1, Hermeneia, 2002) acknowledges the Western reading is "very old" and reflects early baptismal theology. Joseph A. Fitzmyer (The Gospel According to Luke I–IX, AB, 1981) judges it secondary but provides extensive discussion. I. Howard Marshall (The Gospel of Luke, NIGTC, 1978) notes the Western reading is "interesting as reflecting a very early interpretation of the significance of the baptism."

Theological note

The Psalm 2:7 quotation is a coronation formula – it declares the king's installation, not his origin. In the ancient Near Eastern context, "today I have begotten you" means "today I publicly declare you as my Son and heir," not "today you came into existence." Jesus was always the Son of Abba. The baptism is the public declaration – the moment the Father announces the relationship to Israel. The variant does not imply adoptionism (that Jesus was merely human before baptism); it preserves the coronation announcement that later scribes replaced with a blander affirmation to avoid theological controversy.

Why this is Tier 2, not Tier 1

Only one Greek manuscript preserves the reading. In standard text-critical method, that is extremely weak external evidence. What makes this Tier 2 rather than Tier 3 is: (1) the patristic attestation from 155 AD across multiple geographic regions; (2) the transcriptional argument – no scribe would create an adoptionist-sounding reading by modifying the standard text; (3) Ehrman's case, while not consensus, is serious scholarship published in peer-reviewed monographs. The variant is best understood as documenting an early Christological diversity that was later standardized.

7. Elevações de títulos cristológicos Tier 2

Across the NT manuscript tradition, scribes repeatedly replaced earlier titles for Jesus with later formulations that moved toward identifying him with the one God of monotheism. The issue is not whether Jesus is divine – he is the Son of the Most High – but whether scribes collapsed the distinction between Jesus and his Father by upgrading titles toward co-identity rather than sonship. In the selected cases below, the direction is consistently toward conflation. This does not mean every NT variant is a theological upgrade – NT transmission also includes harmonization, homoeoteleuton, itacism, and other non-theological changes. But in passages where Christological titles are at stake, the pattern is well documented.

Bart D. Ehrman (The Orthodox Corruption of Scripture, ch. 3) catalogs these systematically. James R. Royse (Scribal Habits in Early Greek New Testament Papyri, Brill, 2008) demonstrated that early scribes tended toward omission rather than addition – meaning when a longer, more Christologically explicit reading appears in later manuscripts, the shorter reading is more likely original. Gordon D. Fee ("The Use of Greek Patristic Citations in NT Textual Criticism," Studies in the Theory and Method of NT Textual Criticism, Eerdmans, 1993) provides methodological grounding for evaluating such patterns.

A note on what counts as "elevation"

Not every change in this table is theologically false. "Son of God" is an accurate title for Jesus – he is the Son of the Most High. The problem is not the title itself but what it replaced and what it moved toward. When "the Chosen One of God" becomes "the Son of God," the change erases a title that preserves the relationship (chosen by someone higher) in favor of one that later theology read as co-identity with the Father. When "the Holy One of God" becomes "the Christ, the Son of the living God," the change imports the Matthean confession wholesale – harmonizing the text toward a single, uniform Christological formula. The scribal instinct was not to invent lies but to resolve ambiguity in the direction that post-Nicene theology demanded.

Why this is Tier 2

Each individual variant above has strong textual support. The pattern is real. However, some critics argue individual cases (particularly Mark 1:1 and John 1:34) may involve harmonization or scribal error rather than deliberate theological upgrading. The cumulative weight is compelling, but the monocausal explanation ("scribes upgraded titles for theological reasons") is stronger for some examples (John 6:69, Eph 3:9) than others (Mark 1:1, where accidental omission is also plausible).

8. Hebreus 2:9 – chariti theou vs chōris theou Nível 3

Greek text

The two readings differ by a single word: chariti ("by grace") vs chōris ("apart from"). The visual similarity in uncial script (ΧΑΡΙΤΙ vs ΧΩΡΙΣ) makes accidental interchange plausible.

Manuscript evidence for chariti theou

Supported by virtually all Greek manuscripts across all text-types: P46 (c. 200, earliest witness to Hebrews), ℵ (c. 330–360), A (c. 400–440), B (c. 325–350), C (c. 450), D (Claromontanus, c. 500), K, L, 33, 81, Majority Text. Versions: Peshitta, Sahidic, Bohairic, Vulgate (most), Old Latin (most), Ethiopic.

Manuscript evidence for chōris theou

Direct Greek manuscript support is very weak: 424^c (margin, 11th c.), 1739 (margin, 10th c. – marginal note reads: "chōris theou – so Origen and other ancient writers"), 0121b (Codex Ruber, 10th c., body text), and at least one Vulgate manuscript (absque Deo).

Manuscript 1739 is significant because it is a 10th-century copy of a 4th-century exemplar that itself derived from a probable 2nd-century archetype. Günther Zuntz (The Text of the Epistles, British Academy/OUP, 1953) demonstrated its close textual kinship with P46 and B, giving its marginal attribution to "Origen and other ancient writers" real weight.

Patristic evidence for chōris theou

Origen (d. c. 254) knew and accepted chōris theou, interpreting it as universal scope – Jesus tasted death for everything "apart from God," meaning God alone was excluded from the experience of death. Theodore of Mopsuestia (d. 428), Ambrose (d. 397), and Fulgentius of Ruspe (d. 533) also knew the reading. The chōris reading circulated widely in patristic commentary from the 3rd century – far earlier than the 5th-century Nestorian controversy.

The scholarly case for chōris

Bart D. Ehrman (The Orthodox Corruption of Scripture, 1993/2011) argues chōris is original and chariti is an "orthodox corruption" – scribes changed it to prevent separationist Christians from using the text to claim the divine Christ departed before the crucifixion. Zuntz (The Text of the Epistles, 1953) also considered chōris possibly original on internal grounds. Paul A. Hartog ("The Text of Hebrews 2:9 in Its Patristic Reception," Bibliotheca Sacra 171:681, 2014) traces the patristic reception and demonstrates the chōris reading predates the Nestorian controversy by centuries.

Why this is Tier 3

The direct Greek manuscript evidence for chōris is extremely weak – marginal notes and one minor codex, against the entire manuscript tradition including P46, our earliest witness. Lectio difficilior is a strong internal principle, but it cannot override overwhelming external evidence without caution. Harold W. Attridge (Hebrews, Hermeneia, 1989) retains chariti. William L. Lane (Hebrews 1–8, WBC, 1991) discusses the variant extensively but adopts chariti. Most critical commentaries follow the manuscript evidence. Ehrman's case is serious and the patristic evidence is genuinely early, but this remains a minority position. The variant is important as evidence of early theological pressure around Jesus' death – but it should not be cited as a proven scribal change without heavy qualification.

9. Mateus 28:19 – A fórmula trinitária Tier 3

Greek text

The critical distinction

This is not a textual variant in the normal sense. All surviving Greek manuscripts – every papyrus, every uncial, every minuscule, every lectionary – contain the Trinitarian formula. All extant versional witnesses that preserve this passage contain the formula. By standard text-critical method, the case is closed: the Trinitarian reading is the text.

What follows is a conjectural argument – based on patristic citation practice, internal consistency, and historical probability – that the original text may have been different from what all surviving manuscripts preserve. This is a legitimate category in textual criticism (cf. Kilpatrick, Elliott), but it operates on different evidential ground than standard variant analysis. The reader should weigh it accordingly.

The Eusebius evidence

F. C. Conybeare (Zeitschrift für die neutestamentliche Wissenschaft 2, 1901, pp. 275–288; Hibbert Journal 1.1, 1902) cataloged Eusebius' citations of Matthew 28:19 and identified a pattern:

Pre-Nicene citations (18–21 instances): consistently "Go and make disciples of all nations in my name" – no mention of baptism, Father, Son, or Holy Spirit. These appear in the Demonstratio Evangelica (c. 314–318), Theophania (c. 333), Historia Ecclesiastica (c. 324), commentaries on Psalms and Isaiah, and De Laudibus Constantini (c. 335). In several instances Eusebius comments on "in my name" in ways that suggest it was his text, not merely an abbreviation.

Post-Nicene citations: Eusebius begins quoting the Trinitarian formula only in two works written in his extreme old age – Contra Marcellum and De Ecclesiastica Theologia (both post-335).

The Acts evidence

Every baptism recorded in Acts uses "in the name of Jesus" (or "the Lord Jesus") – never the Trinitarian formula:

• Acts 2:38 – "in the name of Jesus Christ"
• Acts 8:16 – "in the name of the Lord Jesus"
• Acts 10:48 – "in the name of Jesus Christ"
• Acts 19:5 – "in the name of the Lord Jesus"

The Didache counterevidence

The Didache (7:1–3, late 1st or early 2nd century) instructs: "baptize in the name of the Father and of the Son and of the Holy Spirit." This is the strongest counterevidence to Conybeare's thesis. If the Didache is early (which many scholars maintain), then the Trinitarian baptismal formula was in use very early, regardless of whether it was in Matthew's original text.

Key scholarship

Bruce M. Metzger (A Textual Commentary, 1994) retains the Trinitarian formula as original on the basis of unanimous manuscript evidence. He considers Eusebius' citations possible abbreviations or paraphrases.

Daniel B. Wallace defends the Trinitarian formula as original, arguing Conybeare's reading of Eusebius is "faulty" and that Eusebius was abbreviating for rhetorical purposes – a practice documented elsewhere in his works.

Hans Kosmala ("The Conclusion of Matthew," Annual of the Swedish Theological Institute 4, 1965, pp. 132–147) argues the formula is a liturgical addition. George Howard ("A Note on the Short Ending of Matthew," JSNT, 1981) examines the Shem Tob Hebrew Matthew, which has no baptismal formula at all.

Why this is Tier 3

Zero direct manuscript evidence. Zero versional evidence. The argument rests entirely on patristic citation practice (which can reflect abbreviation) and internal coherence (which can reflect developing liturgical practice). Conybeare's thesis is provocative and the Eusebius pattern is genuinely unusual, but by the standard tools of textual criticism, this is a conjecture. It should be presented as such – not as a recovered variant.

10. Evidência de apoio: padrões escribais secundários

Beyond Christological elevation, scribal activity reshaped the NT text along at least four additional ideological axes. These patterns strengthen the overall case for theologically motivated transmission but should be evaluated individually – they include standard textual variants, text-type tendencies, and non-manuscript interpolation hypotheses.

A. Silencing women (textual variants + editorial history)

Romans 16:7 – Junia. The accusative form Iounian is identical for both feminine Iounia ("Junia") and a hypothetical masculine Iounias ("Junias"). Peter Lampe identified approximately 250 examples of the female Latin name Junia in Roman inscriptions. The hypothetical masculine "Junias" has zero attestation in any source. For the first thousand years, every commentator identified the name as feminine – Origen, Chrysostom ("How great the wisdom of this woman must have been that she was even deemed worthy of the title of apostle"), Jerome, Theodoret, John of Damascus. The first masculine reading came from Aegidius of Rome (1245–1316). Erwin Nestle's 1927 edition introduced the masculine accentuation into the critical text. The feminine was restored in NA27 (1993). See Bernadette Brooten (in Women Priests, 1977) and Eldon Jay Epp (Junia: The First Woman Apostle, Fortress, 2005).

1 Corinthians 14:34–35. These two verses ("the women should be silent in the churches") appear after v. 33 in most manuscripts (P46, ℵ, A, B) and after v. 40 in the entire Western tradition (D, F, G, Old Latin, Ambrosiaster). No Pauline manuscript transposes a block this large without structural reason. The verses directly contradict 1 Cor 11:5, which explicitly permits women to pray and prophesy. Codex Vaticanus preserves a distigme and obelos at the end of 14:33, marking the text as suspect. Philip B. Payne (NTS 41, 1995; NTS 44, 1998; NTS 63, 2017) and Gordon D. Fee (The First Epistle to the Corinthians, NICNT, 1987, pp. 697–710) argue for interpolation. Note: many scholars retain these verses as Pauline and explain the transposition differently. The interpolation case is serious but not consensus.

Priscilla name order (Acts 18:26). Priscilla is listed before her husband Aquila in 4 of 6 NT mentions. Older manuscripts place Priscilla first. Codex Bezae and the Byzantine tradition reverse the order.

B. Adding fasting (textual variants)

Later manuscripts systematically add references to fasting alongside prayer:

• Mark 9:29 – "except by prayer" becomes "except by prayer and fasting." Shorter: ℵ, B, 0274, k. Longer: P45^vid, ℵ², A, C, D, K, L, N, W, Θ, Ψ, Majority Text. NA28 reads the shorter text. UBS: {A}. Metzger: "In light of the increasing emphasis in the early church on the necessity of fasting, it is understandable that kai nēsteia is a gloss."
• Matthew 17:21 – The entire verse is absent from ℵ, B, C*, 33, 579, 892*. Likely imported from Mark 9:29 (the longer reading). NA28 omits.
• Acts 10:30 – "I was fasting" added by D and the Byzantine tradition. Absent from P74, ℵ, A, B, C, E. NA28 reads the shorter text.
• 1 Corinthians 7:5 – "to fasting and prayer" in later manuscripts. "To prayer" alone in P46 (c. 200), P11, ℵ*, A, B, C, D, F, G. NA28 reads "prayer" only.

Four passages, four books, one pattern: later manuscripts add fasting to prayer. The direction in these specific cases is always addition, never subtraction.

C. Softening radical grace (textual variants)

• Romans 8:1 – The shortest reading, "There is therefore now no condemnation for those in Christ Jesus" (P46^vid, ℵ*, A, B, C, D*, G, 1739), is unconditional. Later manuscripts add "who do not walk according to the flesh" or the full "who do not walk according to the flesh but according to the Spirit" (ℵ², K, P, Byzantine, Textus Receptus) – transforming grace into a conditional statement by importing language from v. 4.
• Matthew 5:22 – "everyone who is angry with his brother" (P64/67, ℵ*, B, Origen, Jerome, Augustine, Tertullian) becomes "angry with his brother without cause" (eikē) in ℵ², D, K, L, W, Byzantine, Vulgate, Syriac. NA28 reads the shorter text. UBS: {B}. Metzger: eikē was "much more likely added by copyists in order to soften the rigor of the precept."
• Mark 10:24 – "Children, how hard it is to enter the kingdom of God" (ℵ, B, Δ, Ψ, Coptic) becomes "how hard it is for those who trust in riches to enter…" (A, C, D, K, W, Θ, Majority Text). NA28 reads the shorter text. UBS: {B}.
• Revelation 22:14 – "Blessed are those who wash their robes" (ℵ, A, 1006, 1611, Vulgate) becomes "Blessed are those who do his commandments" (046, most minuscules, Syriac, Coptic). The theological direction: from grace (Christ's redemptive work) to works (commandment-keeping). NA28 reads "wash their robes."

D. Anti-Judean intensification (mixed evidence types)

Luke 23:34a (textual variant) – Jesus' prayer "Father, forgive them, for they do not know what they are doing" is omitted by P75 (early 3rd c.), ℵ*, B, D*, W, Θ, 0124, 579, some Old Latin, some Syriac, some Coptic. Included by ℵ², A, C, D², K, L, N, Byzantine Majority. NA28 places it in double brackets. Ehrman (The Orthodox Corruption of Scripture, pp. 107–112) argues that after 70 CE, many Christians interpreted Jerusalem's destruction as divine punishment; a prayer asking forgiveness for the executioners became theologically unacceptable.

The Western text of Acts (text-type tendency) – Eldon Jay Epp (The Theological Tendency of Codex Bezae Cantabrigiensis in Acts, Cambridge UP, 1966) demonstrated that the Western text exhibits "a strongly heightened anti-Judaic tendency" through calculated textual variants: Acts 3:17 (adding ponēron to describe Judean actions), Acts 13:27 (expanding the trial account), Acts 14:2 (intensifying opposition). This is a documented editorial tendency within one text-type, not individual scribal corruption.

1 Thessalonians 2:14–16 (interpolation hypothesis – no manuscript evidence) – The argument rests not on manuscripts (the passage appears in all manuscripts) but on internal, linguistic, and historical grounds. "Wrath has come upon them" uses the aorist, most naturally read as referring to the fall of Jerusalem (70 CE) – but 1 Thessalonians is dated c. 50–51 CE. Paul's position in Romans 9–11 explicitly rejects permanent rejection of Israel. Birger A. Pearson (HTR 64.1, 1971) and Daryl Schmidt (JBL 102.2, 1983) argued for interpolation; the debate remains unresolved and many scholars retain the passage as authentic.

11. Bibliografia

Edições críticas e aparato

• Aland, Barbara et al., eds. Novum Testamentum Graece. 28th ed. (NA28). Deutsche Bibelgesellschaft, 2012.
• Aland, Barbara et al., eds. The Greek New Testament. 5th ed. (UBS5). Deutsche Bibelgesellschaft / United Bible Societies, 2014.
• Metzger, Bruce M. A Textual Commentary on the Greek New Testament. 2nd ed. Deutsche Bibelgesellschaft / United Bible Societies, 1994.
• Omanson, Roger L. A Textual Guide to the Greek New Testament. Deutsche Bibelgesellschaft, 2006.

Corrupção ortodoxa / variantes cristológicas

• Ehrman, Bart D. The Orthodox Corruption of Scripture. OUP, 1993; 2nd ed. 2011.
• Ehrman, Bart D. Misquoting Jesus. HarperSanFrancisco, 2005.
• Royse, James R. Scribal Habits in Early Greek New Testament Papyri. Brill, 2008.
• Wettstein, Johann Jakob. Prolegomena ad Novi Testamenti Graeci. 1730.

Metodologia e método de crítica textual

• Aland, Kurt and Barbara Aland. The Text of the New Testament. 2nd ed. Trans. Erroll F. Rhodes. Eerdmans, 1989.
• Parker, David C. An Introduction to the New Testament Manuscripts and Their Texts. Cambridge UP, 2008.
• Ehrman, Bart D. and Michael W. Holmes, eds. The Text of the New Testament in Contemporary Research. 2nd ed. NTTSD 42. Brill, 2013.
• Wasserman, Tommy and Peter J. Gurry. A New Approach to Textual Criticism. SBL Press, 2017.
• Zuntz, Günther. The Text of the Epistles. British Academy / OUP, 1953.
• Fee, Gordon D. "The Use of Greek Patristic Citations in NT Textual Criticism." In Studies in the Theory and Method of NT Textual Criticism. Eerdmans, 1993.
• Epp, Eldon Jay. The Theological Tendency of Codex Bezae Cantabrigiensis in Acts. SNTS Monograph 3. Cambridge UP, 1966.

Séries de comentários citados

• Attridge, Harold W. Hebrews. Hermeneia. Fortress Press, 1989.
• Bovon, François. Luke 1. Hermeneia. Fortress Press, 2002.
• Brown, Raymond E. The Epistles of John. Anchor Bible. Doubleday, 1982.
• Davies, W. D. and Dale C. Allison. Commentary on Matthew. Vol. 3. ICC. T&T Clark, 1997.
• Fee, Gordon D. 1 and 2 Timothy, Titus. NIGTC. Hendrickson, 1988.
• Fee, Gordon D. The First Epistle to the Corinthians. NICNT. Eerdmans, 1987.
• Fitzmyer, Joseph A. The Gospel According to Luke (I–IX). AB 28. Doubleday, 1981.
• France, R. T. The Gospel of Mark. NIGTC. Eerdmans, 2002.
• France, R. T. The Gospel of Matthew. NICNT. Eerdmans, 2007.
• Lane, William L. Hebrews 1–8. WBC 47A. Thomas Nelson, 1991.
• Marshall, I. Howard. The Pastoral Epistles. ICC. T&T Clark, 1999.
• Marshall, I. Howard. The Gospel of Luke. NIGTC. Eerdmans, 1978.
• Towner, Philip H. The Letters to Timothy and Titus. NICNT. Eerdmans, 2006.

Mulheres e gênero

• Brooten, Bernadette. "Junia…Outstanding among the Apostles." In Women Priests, 1977.
• Epp, Eldon Jay. Junia: The First Woman Apostle. Fortress Press, 2005.
• Payne, Philip B. Man and Woman, One in Christ. Zondervan, 2009.
• Payne, Philip B. "Fuldensis, Sigla for Variants in Vaticanus, and 1 Cor 14.34–5." NTS 41 (1995).
• Payne, Philip B. "Vaticanus Distigme-obelos Symbols Marking Added Text." NTS 63 (2017).

Modificações antijudaicas

• Epp, Eldon Jay. The Theological Tendency of Codex Bezae Cantabrigiensis in Acts. Cambridge UP, 1966.
• Pearson, Birger A. "1 Thessalonians 2:13–16: A Deutero-Pauline Interpolation." HTR 64.1 (1971).
• Schmidt, Daryl. "1 Thess 2:13–16: Linguistic Evidence for an Interpolation." JBL 102.2 (1983).

Cânon e interpolação

• Bartholomä, Philipp F. "Did Jesus Save the People out of Egypt?" Novum Testamentum 50/2 (2008).
• Conybeare, F. C. "The Eusebian Form of the Text Matth. 28,19." ZNW 2 (1901).
• Conybeare, F. C. "The Eusebian Form of the Text." Hibbert Journal 1.1 (1902).
• Croy, N. Clayton. The Mutilation of Mark's Gospel. Abingdon, 2003.
• Hartog, Paul A. "The Text of Hebrews 2:9 in Its Patristic Reception." Bibliotheca Sacra 171:681 (2014).
• Kelhoffer, James A. Miracle and Mission. WUNT II/112. Mohr Siebeck, 2000.
• Kosmala, Hans. "The Conclusion of Matthew." ASTI 4 (1965).
• McDonald, Grantley. Biblical Criticism in Early Modern Europe. Cambridge UP, 2016.
• van der Horst, P. W. "Can a Book End with a Gar?" JTS 23 (1972).