The Angel in Gethsemane: Luke 22:43-44

In the garden of Gethsemane, the night before his death, Jesus knelt down and prayed. What happened next is one of the most contested questions in New Testament textual criticism. Did an angel appear and strengthen him? Did his sweat fall like drops of blood? Or did someone add those details to Luke's Gospel centuries later?

The standard critical text of the Greek New Testament (NA28) double-brackets both verses, signaling deep doubt about their authenticity. The committee that made this decision gave it their highest confidence rating. Most modern translations either omit the verses or relegate them to footnotes.

The UPDV restores them. Not because tradition demands it, but because the evidence does. The verses were not added by a later hand. They were written by Luke and then deliberately removed by scribes who found them theologically uncomfortable. The earliest surviving witness to the included text is a Syriac manuscript whose Greek Vorlage — the source text behind the translation — may date to the second or third century, and whose testimony has been catalogued in every critical apparatus for over a century but whose translation technique we found has not been systematically cross-examined.

What the UPDV prints

43 And an angel appeared to him, strengthening him. 44 And he was in great distress, and was praying urgently, and his sweat became like drops of blood falling upon the ground.

This text follows the Curetonianus (syc), one of only two surviving Old Syriac Gospel manuscripts, for both verses. It differs from the standard Greek in two ways: the angel has no qualifier ("from heaven" is absent), and the emotional language throughout verse 44 reflects an older, independent translation tradition. These are not arbitrary departures. They follow the evidence to its source.

The problem with "earliest manuscripts omit"

The case for omission rests on a small group of manuscripts: Papyrus 75, Codex Vaticanus, the original hand of Codex Sinaiticus, and a few others. The standard argument is straightforward: the earliest and best manuscripts omit the verses, therefore the verses were probably not original.

This argument has a problem. P75 is its foundation, and that foundation is less secure than it appears.

When P75 was discovered in the 1950s, its editors dated it to approximately 175-225 CE based on paleographic comparison with other manuscripts. This made it the earliest papyrus witness to Luke's Gospel, and its omission of verses 43-44 became the anchor for the committee's confidence. If the oldest manuscript omits them, they were probably not there.

In 2016, Brent Nongbri published a detailed reassessment in the Journal of Biblical Literature challenging this dating. His argument, in brief: the paleographic parallels used to date P75 were themselves undated, creating a circular chain. The leading papyrologist of the twentieth century, Eric Turner, had quietly moved P75's date to 225-275 CE by 1977, but the field continued citing the earlier date. Nongbri's own analysis of securely dated parallels points to the fourth century. The closest match he found — letters from the 320s CE — provides, in his words, "the closest securely datable paleographic connection known to me for the script of P75."¹

If P75 is a fourth-century manuscript, the chronological argument for omission weakens considerably. Every patristic writer who quotes Luke 22:43-44 — Justin Martyr around 155, Irenaeus around 180, Hippolytus around 220 — predates every manuscript that omits them. The "earliest manuscripts omit" argument becomes "fourth-century Egyptian manuscripts omit" — a very different claim.

Furthermore, P75 and Codex Vaticanus agree ninety to ninety-four percent in Luke and John, far higher than Vaticanus agrees with Sinaiticus. If both are fourth-century products, their shared omission may not represent two independent witnesses but a single editorial decision reflected in two copies of the same localized text type.

Epiphanius and the Orthodox deletion

Epiphanius of Salamis was a fourth-century bishop and tireless cataloguer of heresies. He was no fringe figure — he was one of the most prominent churchmen of his era. In his Ancoratus, he addresses the textual status of Luke 22:43-44 directly:

"It stands in the Gospel according to Luke in the uncorrected copies (and holy Irenaeus bears witness in Against Heresies, against those who say Christ appeared only in seeming); but the Orthodox removed the passage, being afraid, and not understanding its purpose and its strongest argument."²

Epiphanius says five things: the verses were original. They stood in the older, uncorrected manuscripts. Irenaeus had already used them against docetism. Orthodox scribes deliberately deleted them. And the deletion was a mistake — the verses were the strongest proof of Christ's real humanity.

This is not merely a modern inference from scattered data. It is explicit fourth-century testimony from a bishop alleging deliberate scribal removal. The interpolation theory — that someone added the verses to Luke — requires dismissing Epiphanius while accepting the manuscripts he says were tampered with. The deletion theory requires taking his testimony seriously.

The hidden gradient

The critical apparatus of the NA28 lists Syriac witnesses among those that include the verses: "syc.p.h**" — the Curetonianus, the Peshitta, and the Harklean. What the apparatus does not capture is that these witnesses do not all say the same thing.

Four distinct forms of verse 43 survive:

Form	Text	Witness
Absent	—	S, P75, Vaticanus, Sinaiticus*
Bare angel	"an angel appeared to him"	Curetonianus, Epiphanius (Ancoratus 31)
Angel of the Lord	"an angel of the Lord appeared to him"	Epiphanius (Ancoratus 37, Adv. Haer. 59)
Angel from heaven	"an angel from heaven appeared to him"	Peshitta, standard Greek, majority

The standard apparatus collapses forms 2, 3, and 4 into a single "include" category. This obscures the most important evidence: the earliest surviving included text does not match the standard Greek. It is shorter. Simpler. Less theologically developed. And it comes from an independent translation tradition that we can verify.

The Curetonianus and translation technique

The Curetonianus (designated syc or C) is one of two surviving Old Syriac Gospel manuscripts, generally dated to the fifth century but reflecting a Greek Vorlage — the source text from which it was translated — considerably older, possibly second or third century. It was first published by William Cureton in 1858.

At Luke 22:43, the Curetonianus reads simply: "And an angel appeared to him, strengthening him." No "from heaven." No "of the Lord." Just an angel.

The natural objection is that a Syriac translator might have shortened the text — dropped "from heaven" for stylistic reasons or through carelessness. But the Curetonianus shows no such tendency in the extant Gospel evidence. Across twenty-six surviving passages where its Greek Vorlage contains a comparable qualifier — including twenty instances of "from heaven" — the manuscript preserves the modifier every time. At Matthew 28:2, where the Greek has both "angel of the Lord" and "from heaven," the Curetonianus renders both. The four angel passages are not an isolated sample. They are part of a broader pattern spanning three qualifier types — "from heaven," "of God," "of the Lord" — none of which the Curetonianus ever drops.³

The bare "angel" at Luke 22:43 is therefore best explained as a feature of its Vorlage, not a Syriac simplification. The most natural explanation is that the Curetonianus translator was looking at a Greek text that had no modifier. Not "angel of the Lord." Not "angel from heaven." Just "angel." In the extant comparable cases, the translator renders such qualifiers faithfully.

Verse 44: an independent translation

The Curetonianus does not merely differ from the standard text at verse 43. Its rendering of verse 44 is independently translated throughout, using different vocabulary from the Peshitta at every point of comparison:

Element	Curetonianus	Peshitta	Greek
Emotional state	"in great distress"	"in fear"	"in agony"
Prayer adverb	"urgently"	"earnestly"	"more earnestly"
Verb form	"was falling" (continuous)	"fell" (simple)	present participle (ongoing)

The Curetonianus renders the Greek ἀγωνία (agonia, "agony") as "great distress" — arguably a better rendering than the Peshitta's "fear," since the Greek word denotes intense struggle, not dread. The continuous "was falling" matches the Greek participle more accurately than the Peshitta's simple past tense. At every turn, the Curetonianus is independently translating from its own Greek source, not copying or abbreviating the Peshitta.

This is not one manuscript disagreeing with another over a single word. This is an entire verse rendered in different vocabulary, different syntax, and different verbal aspect — showing independent translation from a distinct Greek Vorlage behind the Curetonianus. That Vorlage included the verses but in a shorter, less developed form than the standard Greek tradition.

The vocabulary fingerprint

The strongest argument against authenticity has been vocabulary. Bart Ehrman, in his influential 1983 article in the Catholic Biblical Quarterly, counted three words in these two verses that appear nowhere else in the New Testament: ἀγωνία (agonia, "agony/struggle"), ἱδρώς (hidros, "sweat"), and θρόμβοι (thromboi, "drops/clots"). An 11.5% hapax legomenon rate in twenty-six words, against Luke's average of 1.1%. Surely a different author.⁴

But Ehrman also documented something he did not follow up on. In a footnote, he observed that Luke's habitual phrase for angels is "angel of the Lord" — seven times in Luke-Acts — while "angel from heaven" appears zero times. Luke never writes "angel from heaven."⁵ This means the standard Greek text of verse 43 contains a phrase its supposed author never uses. The reading Epiphanius quotes — "angel of the Lord" — matches Luke's vocabulary perfectly. The Curetonianus preserves the simplest form: a bare angel with no qualifier at all. The standard text's "angel from heaven" is a phrase its supposed author never uses.

The three hapax legomena, meanwhile, are not as foreign to Luke as they first appear. All three have pedigree in the Greek Old Testament that Luke demonstrably drew from:

Agonia appears at 2 Maccabees 3:14 and 15:19. Hidros (sweat) appears at Genesis 3:19 in the Septuagint — "by the sweat of your face ... to the ground" — the same two elements (sweat + upon the ground) that appear in Luke 22:44. Thromboi (drops/clots) is educated literary Greek found in Aeschylus and Plato, as well as medical writers. Luke, traditionally identified as a physician, would have known the term.⁶

More importantly, two rare vocabulary items in these verses appear only in Luke-Acts in the entire New Testament:

The verb ἐνισχύω (enischyo, "to strengthen") appears eighty-six times in the Septuagint but only twice in the New Testament — both in Luke-Acts (22:43 and Acts 9:19). It is the technical term for angelic strengthening in Daniel 10:18-19, where an angel strengthens Daniel during prayer — the precise Old Testament parallel to the Gethsemane scene.

The ἐκτενής (ektenes) word group applied to prayer appears only in Luke-Acts in the entire New Testament. Luke 22:44 uses the comparative form ἐκτενέστερον (ektenesteron, "more earnestly/urgently"); Acts 12:5 uses the adverb ἐκτενῶς (ektenos) in an identical narrative pattern: crisis, earnest prayer, angel appears, divine intervention. An interpolator would have to independently deploy two rare words that only Luke uses anywhere in the New Testament, in the same narrative template Luke uses them in. These are the marks of Luke's hand, not the usual profile of interpolation.⁷

The intensity shift: from grief to combat

Why does Luke use unusual vocabulary here? Because he is rewriting a tradition that used different vocabulary, and the rewriting is visible.

Matthew and Mark both describe Jesus in raw emotional vulnerability at Gethsemane. Matthew 26:37 says he "began to be sorrowful and sore troubled." Mark 14:33 says he "began to be greatly amazed and sore troubled." Both report Jesus saying, "My soul is exceeding sorrowful, even unto death." Luke omits all of this — every word of grief, every expression of emotional vulnerability.

In its place, Luke puts combat language. Not sorrow but struggle (agonia). Not comfort but military strengthening (enischyo, the Daniel 10 verb). Not desperate crying but disciplined intensification (ektenesteron). And the grief that Matthew and Mark place on Jesus, Luke relocates to the sleeping disciples: "he found them sleeping for sorrow" (22:45). Jesus is not breaking down. He is fighting harder under maximum load.⁸

This is not a random substitution. It is a systematic editorial pattern that Jerome Neyrey identified in 1980: Luke consistently removes the language of emotional vulnerability from Jesus throughout the passion narrative. Grief — λύπη (lype) — in Stoic philosophy and the Greek Old Testament implies defeat, irrationality, and sin — a cardinal passion unworthy of the righteous sufferer. Luke strips it from Jesus and gives it to the disciples. The angel does not comfort a griever; it strengthens a combatant.⁹

An interpolator borrowing from the Synoptic tradition would naturally carry over the grief vocabulary — "sorrowful," "sore troubled," "exceeding sorrowful unto death." Instead, these verses use a completely different register that independently matches Luke's known editorial tendency. From the patterns we have seen in textual interpolation, that kind of systematic register replacement — vulnerability to combat, with sorrow relocated to the disciples — points to an author making consistent editorial choices across an entire scene, not a later hand inserting material.

Hebrews 5:7 and the underlying tradition

There is an independent witness to this tradition that predates any surviving manuscript. The Epistle to the Hebrews, generally dated before 70 CE, describes the Gethsemane prayer:

"Who in the days of his flesh, having offered up prayers and supplications with strong crying and tears unto him that was able to save him from death, and having been heard for his godly fear."¹⁰

"Having been heard" presupposes a divine response to the prayer. Without verses 43-44, no gospel records any evidence that the Gethsemane prayer was answered. With the angel, the match is striking: intense physical prayer, and then a divine response — proof of being heard.

But notice the vocabulary. Hebrews says "tears." Luke says "sweat like drops of blood." Hebrews says "strong crying." Luke says "prayed more urgently." Hebrews describes grief and supplication. Luke describes combat and struggle. The tradition is the same; the vocabulary is completely different.

This is the fingerprint of authorial redaction. Luke took a known tradition — intense prayer, physical distress, divine response — and rewrote it through his combat filter. Tears became sweat. Crying became disciplined intensity. Being heard became a visible angel strengthening for battle. The "unusual" vocabulary that the interpolation theory treats as evidence of foreign authorship is evidence of exactly the opposite: Luke's characteristic editorial hand reshaping a tradition independently attested in Hebrews decades before any manuscript of Luke existed.¹¹

The deletion model

The textual history of Luke 22:43-44 most plausibly reconstructs as follows:

Luke wrote the verses. The angel was unqualified — a bare "angel," as the Curetonianus preserves. The emotional register was combat, not grief — consistent with Luke's treatment of the entire Gethsemane scene. The tradition behind the verses was already known when Hebrews was written, probably before 70 CE.

Within the first two centuries, the verses circulated widely. Justin Martyr quoted verse 44 around 155 CE, and his wording is not that of the standard text: he uses κατεχεῖτο (katecheito, "was poured down") rather than Luke's καταβαίνοντες (katabainontes, "coming down"). That difference makes secondary harmonization to the later standard form unlikely. Irenaeus cited them around 180 CE. Hippolytus discussed them at length around 220 CE. A Greek papyrus fragment designated 0171, from Hermopolis Magna in Egypt, contains the end of verse 44 and dates to approximately 300 CE — proving the verses were present in Egyptian manuscripts as well.¹²

At some point in the third or fourth century, scribes in the Alexandrian tradition removed the verses. Epiphanius tells us why: they feared the passage implied Jesus was weak, playing into the hands of docetists who denied Christ's real humanity. The irony, as Epiphanius himself noted, was that the verses were the strongest argument against docetism. But the deletion stuck in that textual stream and was inherited by P75 and Codex Vaticanus. Even without Epiphanius's testimony, the case for originality would stand on the vocabulary fingerprint, the Curetonianus translation evidence, and the redactional coherence with Luke's treatment of the entire Gethsemane scene. Epiphanius tells us why the deletion happened; the other evidence tells us that it happened.

Remarkably, when Epiphanius quotes the verses in this very passage — the passage where he discusses the deletion — he writes simply ὤφθη ἄγγελος ἐνισχύων αὐτόν: "an angel appeared, strengthening him." A bare angel, no qualifier. The same reading the Curetonianus preserves. Two independent witnesses in different languages and different centuries point to the same unqualified text.

Whether by restoration or by parallel transmission in multiple streams, the broader included tradition shows the angelic reference in expanded forms. Epiphanius himself uses "angel of the Lord" in his other quotations (Ancoratus 37, Adversus Haereses 59). "Angel from heaven" appears in the Peshitta and the majority Greek tradition. The direction of change is the same: qualifiers accumulate. In the traditions we can examine, angelic qualifiers are consistently preserved rather than reduced — but readily expanded, making the heavenly origin explicit.

The standard critical text likely preserves the last and most expanded form of a verse whose earliest recoverable text was simpler, shorter, and more characteristically Lucan than what any Greek manuscript now contains.

The evidence that waited

The textual-critical framework governing Luke 22:43-44 was established by Westcott and Hort in 1881 — before the Old Syriac evidence was available. When F.C. Burkitt published both Old Syriac witnesses in 1904, he noted that the Curetonianus had a bare angel and that Ephrem's commentary matched the same reading. He put it in a footnote. It appears not to have received sustained follow-up.¹³

For over a century, the framework has remained almost exclusively Greek-centric, allowing the Old Syriac evidence to be catalogued in every critical apparatus but never deeply cross-examined. The bare angel has been visible in the data since 1904. What has been missing is not data but attention — a systematic comparison of translation technique across the Curetonianus that reveals what the bare angel at Luke 22:43 actually means. That comparison is what this article provides, and it is what changes the textual question from "should these verses be included?" to "which form of these verses did Luke write?"

---