When they came to the crowd, a man came up to Jesus and knelt before Him.
Berean Standard Bible · CC0
Greek — tap a word ↓
Kai elthontōn pros ton ochlon anthrōpos prosēlthen autō gonypetōn auton
The Fallible · Synthetic · Study Bible
The Boy with a Demon
Matthew 17:14–18 — The Boy with a Demon. Each verse below carries the full apparatus: the Berean Standard Bible, the vocalized original (tap any word), and a parsed breakdown of every term transcribed from the interlinear. Synthesized commentary, canonical threads, and the reading of Christ gather at the end, over the whole unit.
When they came to the crowd, a man came up to Jesus and knelt before Him.
Berean Standard Bible · CC0
Greek — tap a word ↓
Kai elthontōn pros ton ochlon anthrōpos prosēlthen autō gonypetōn auton
“Lord, have mercy on my son,” he said. “He has seizures and is suffering terribly. He often falls into the fire or into the water.
Berean Standard Bible · CC0
Greek — tap a word ↓
kai Kyrie eleēson mou ton huion legōn hoti selēniazetai kai paschei kakōs gar pollakis piptei eis to pyr kai pollakis eis to hydōr
I brought him to Your disciples, but they could not heal him.”
Berean Standard Bible · CC0
Greek — tap a word ↓
kai prosēnenka auton tois sou mathētais kai ēdynēthēsan ouk therapeusai auton
“O unbelieving and perverse generation!” Jesus replied. “How long must I remain with you? How long must I put up with you? Bring the boy here to Me.”
Berean Standard Bible · CC0
Greek — tap a word ↓
Ō apistos kai diestrammenē genea de ho Iēsous Apokritheis eipen heōs pote esomai meth’ hymōn heōs pote anexomai hymōn pherete auton hōde moi
Then Jesus rebuked the demon, and it came out of the boy, and he was healed from that moment.
Berean Standard Bible · CC0
Greek — tap a word ↓
kai ho Iēsous epetimēsen autō kai to daimonion exēlthen ap’ autou kai ho pais etherapeuthē apo ekeinēs tēs hōras
The verse-by-verse work is done. What follows gathers the whole unit. All three layers below are machine-generated (⚙). Weigh them; they have no authority.
AI synthesis — woven from the public-domain voices above and the original text; generated and fallible.
⚙ The unit opens on a genitive absolute with no named subject, ἐλθόντων, so the reader is set down in the crowd before being told who arrived. Ellicott and Cambridge both reach for Raphael's Transfiguration to name the contrast of glory above and trouble below, and Ellicott then takes the painting apart on a point of fact: "The two scenes did not synchronise." Pulpit agrees and dates the arrival to the morning after a night vision, following Luke 9:37. Pulpit adds the older comparison, that as Moses was absent at Sinai while the people fell into idolatry (Exodus 32), so the nine faltered while their Master and their leaders were on the mountain. The man who meets them is ἄνθρωπος, no name, one of the crowd (Gill, from Mark and Luke), and his posture is the rare verb γονυπετῶν, present tense, still kneeling as he speaks. What the knee means is left unsettled in the sources we hold: Gill reads "homage and worship," Barnes flatly denies religious homage and reads "high respect and earnest entreaty," and Geneva glosses only the custom of supplicants. ⚙ We leave the disagreement standing, because v. 17 will say the faith in this crowd was thin, and an unresolved knee is truer to that than a confident one.
⚙ The father's plea is one aorist imperative, ἐλέησόν, mercy demanded outright, followed by the strangest word in the unit: σεληνιάζεται, "he is moon-struck," a verb that occurs in only two verses of the New Testament, both in Matthew (here and 4:24). Benson relays Dr. Campbell's judgment that the symptoms belong to epilepsy rather than lunacy and that the name reflects "the general sentiments, at that time, concerning the moon's influence." Cambridge concurs and quotes the composite verdict, "the child was a possessed epileptic lunatic." Ellicott catalogues the ancient awe around the disease, "the divine, the sacred disease." Geneva insists that besides the natural sickness "he had a demonic derangement," and Pulpit objects on principle to translations that bury the actual word: "it seems inexpedient to conceal the actual word used." ⚙ We follow Pulpit's instinct in the literal line and keep the moon, because Matthew keeps it: the Evangelist reports the father's vocabulary in v. 15 and supplies his own, δαιμόνιον, only at v. 18. The doubled πολλάκις, once for the fire and once for the water, is the cadence of a parent counting.
⚙ The father brought the boy with προσήνεγκα, the bringing-near verb Matthew uses of the sick carried to Jesus (4:24, 8:16); he did the right thing at the wrong address. The failure is stated in an aorist passive, ἠδυνήθησαν οὐκ, closer to "they were not enabled" than "they could not," and the sources fill that silence in four different ways. Ellicott: they trusted "the wonted formula of their Master's name" and, scribes and disciples together, never thought of gaining the power that might avail. Pulpit, more generously, lists the whole weight on them, the absent Master, the missing three, the scribes' uncompromising disbelief, the memory of the Passion prediction, and concludes that "the conscious weakness in their exorcism nullified its power." Gill splits the blame between the disciples' unbelief and the father's, and answers a Jewish objection by noting that the three chief disciples were not there. Benson says Christ permitted the failure to keep them humble and to glorify His own power. Henry draws the pastoral edge: it is good to distrust ourselves, "but it is displeasing to Christ when we distrust any power derived from him." Poole records the plain narrative fact and no more, that the father, failing with the disciples, "crieth unto Christ for his help."
⚙ The rebuke is built of two vocatives and two questions. ἄπιστος is withheld trust, and διεστραμμένη is not an adjective at all but a perfect passive participle, something twisted and left twisted; Barnes glosses the root as "twisted or turned from the proper direction," as of an eye out of its natural line. Gill says the words "were long ago spoken of their ancestors, Deuteronomy 32:5," and Pulpit independently notes that Moses uses this very word in his great song in the Septuagint. Who is addressed divides our voices sharply, and we do not harmonise them: Gill holds the phrase is never used of the disciples, Benson says the disciples "were not faithless" and the reproof is aimed at the people and the scribes, Cambridge says the disciples "are by no means excluded," Pulpit says it takes in all who were present, "especially the nine." Then twice ἕως πότε, "until when," the old wilderness question; and ἀνέξομαι, to hold oneself up under a weight, which Ellicott reads as the confession that His whole life "was one long tolerance of the waywardness and perversity of men." Pulpit refuses to hear impatience for departure in it and hears grief. ⚙ And then, abruptly, one plural imperative from the same root as the bearing, φέρετέ: bear him here to me. The diagnosis reaches a nation; the remedy reaches one boy.
⚙ Matthew spends nine Greek words on the miracle itself. There is no formula and no invoked name: ἐπετίμησεν, which Barnes rightly calls a word of "reproving and commanding" together, and the δαιμόνιον is out. Pulpit sets the two exorcisms side by side, the faltering one the devil disregarded and this one, where "Jesus orders with the calmness of assured authority, and is at once obeyed." The decisive small fact is a change of voice: what the disciples could not do in the active (θεραπεῦσαι, v. 16) is done to the boy in the passive (ἐθεραπεύθη, v. 18). Ellicott states the outcome without overreach, that "the spiritual power of the Healer had overcome the force, whether morbid or demoniac, which was the cause of his sufferings," and he is candid that those who deny the possession must make assumptions of their own, "and we may well be content to confess our inability to solve them." Pulpit is equally candid in the other direction, granting that for epilepsy no cure is known and resting the miracle on the suddenness and the permanence. Henry turns the whole scene inward: if Satan's bodily possession of one boy astonishes us, what of "his spiritual possession of every son of Adam from the fall."
⚙ Offered to be tested. Read the five verses as one sentence about where power is: a father bears his son to the right community and the wrong hands, and the words that fail are not wrong words but borrowed ones. Matthew arranges this around a single verb. In v. 16 the disciples cannot do θεραπεῦσαι; in v. 18 the boy simply is ἐθεραπεύθη. Between them stands a rebuke that is not aimed at nine men who fumbled a technique but at a γενεά described with the Song of Moses' own word for crookedness, and our sources genuinely split over whether the nine are inside that word (Gill and Benson say no; Cambridge and Pulpit say yes), so we will not pretend to settle it. What the Greek does settle is the tone of the two questions. ἀνέξομαι means to hold oneself up under a load, and it is in the future: whatever else the question is, it is not a threat to stop. The One who asks how long He must bear with them is the One whom JFB places here between two announcements of His own death, that is, on His way to bear something heavier. And His answer to His own lament is not a withdrawal but an imperative built from the same root as the bearing: φέρετέ, carry him here to me. ⚙ So the fallible reading we offer is this: the unit is less a lesson in how to obtain power than an exposure of how power is held. It is held personally, by Him, and it is available only at the end of that last dative, μοι. Pulpit's comparison is apt and we borrow it: the prophet's staff in Gehazi's hand raised nothing, and Elisha had to come himself (2 Kings 4:31). We do not press this into a theory about faith as a quantity; v. 20, outside our unit, speaks to that, and Mark's father ("Lord, I believe; help thou mine unbelief," quoted by Poole, Barnes and Ellicott) shows that the faith accepted here was ragged. Mercy came to a kneeling man whose creed was one word short.
Not Scripture, a reading offered for testing: the words that failed were not the wrong words, they were borrowed ones.
AI-generated connections. Each carries a verification badge with a recorded basis; contested links are flagged.
The Verifier's strongest candidate for this unit is Matthew 4:24, and the reason is lexical. σεληνιάζομαι (G4583) stands in only two verses of the New Testament, both of them in Matthew: the summary list of 4:24, where lunatics are named as a class alongside those who suffer badly (G2560 κακῶς, also shared here) and the demon-possessed and the paralysed, and this single case in 17:15. The link is reinforced by προσφέρω (G4374), the bringing-near verb of 4:24 which the father uses in 17:16. Cambridge makes the same connection from the other side: "This is the only special instance of cure in the case of a lunatic. They are mentioned as a class, ch. Matthew 4:24." ⚙ The point is that Matthew's opening summary promised a class of sufferers and his narrative here pays out one member of it.
Matthew 4:24 · Matthew 8:16
basis: Verifier: shared Greek lexeme G4583 σεληνιάζομαι, rare at 2 verses in the New Testament, plus G2560 κακῶς (16 vv) and G4374 προσφέρω (45 vv); Matthew 8:16 shares G2560 κακῶς, G4374 προσφέρω and G2323 θεραπεύω (42 vv). Greek to Greek, same author, no quotation claimed beyond shared vocabulary.
Mark 9:19 and Luke 9:41 report this rebuke with the same rare interjection ὦ (G5599, 17 verses in the New Testament) and the same ἄπιστος (G571), which the Verifier records as the shared basis. JFB, Poole, Barnes, Ellicott, Cambridge and Pulpit all send the reader to Mark for the full account, and Poole prints Mark 9:19 to 27 nearly whole. Luke and Mark omit the second adjective; Matthew alone adds διεστραμμένη. Gill leans on the Lukan form ("bring thy son hither") to argue that the plural imperative of Matthew 17:17 is chiefly addressed to the father.
Mark 9:19 · Luke 9:41 · Luke 9:42
basis: Verifier: Mark 9:19 shares G5599 ὦ (17 vv), G571 ἄπιστος (21 vv), G611 ἀποκρίνομαι (231 vv); Luke 9:41 shares G5599 ὦ, G571 ἄπιστος, G5207 υἱός; Luke 9:42 shares G4334 προσέρχομαι, G1140 δαιμόνιον, G3816 παῖς. Synoptic parallel, Greek to Greek, the rare vocative particle carrying the link.
γονυπετέω (G1120) occurs in four verses only, and the Verifier lists three of them as candidates here: the leper who kneels and says "if you are willing" (Mark 1:40), the man who runs and kneels to ask about eternal life (Mark 10:17), and the soldiers who kneel to mock the King (Matthew 27:29). ⚙ Matthew uses the word twice and only twice: a father kneeling for mercy in 17:14, and a cohort kneeling in derision in 27:29. The vocabulary makes no judgment about sincerity, which is exactly why our voices can disagree about this father's knee: Gill calls it "homage and worship," Barnes says it "did not imply religious homage," Geneva only "as men used to do when making supplication." The word alone never decides a heart.
Mark 1:40 · Mark 10:17 · Matthew 27:29
basis: Verifier: shared Greek lexeme G1120 γονυπετέω, rare at 4 verses in the New Testament (Mark 1:40 also shares G1410 δύναμαι, 201 vv). A shared-vocabulary link, Greek to Greek; no quotation or literary dependence is claimed.
Gill states that the phrase is drawn from Deuteronomy 32:5, and Pulpit notes the same word is used by Moses in his song in the Septuagint of those who dealt corruptly. The same passage carries the second half of the rebuke: Deuteronomy 32:20 has the LORD hiding His face from a generation in which there is no faithfulness. "How long" is the wilderness formula on God's side of the covenant quarrel: Numbers 14:11, "how long will this people treat Me with contempt," and 14:27, "how long must I bear with this wicked congregation"; Psalm 95:10, the forty years of loathing that generation. ⚙ What is striking is the mouth the words are in. In Deuteronomy and Numbers this is the voice of the LORD over Israel; in Matthew 17:17 it is the voice of Jesus over the crowd, and no Evangelist marks the transfer or introduces it with a citation formula.
Deuteronomy 32:5 · Deuteronomy 32:20 · Numbers 14:11 · Psalm 95:10
basis: Cross-Testament, Greek to Hebrew, so no shared Strong's number can be computed and no verbal tier is claimed: the Verifier returned no lexeme overlap for these refs. The recorded basis is instead the named commentators, Gill ("words, which were long ago spoken of their ancestors, Deuteronomy 32:5") and Pulpit (the same verb in the Septuagint of Deuteronomy 32:5), plus the shared wilderness formula "how long." The LXX-level verbal contact is real but lies outside the Strong's data we hold.
διαστρέφω (G1294) is rare, seven verses, and Philippians 2:15 pairs it with γενεά exactly as Matthew 17:17 does: shine as lights in the midst of a crooked and twisted generation. Acts 13:10 sets the same verb beside the same rare vocative particle ὦ when Paul confronts a sorcerer who will not stop twisting the straight ways of the Lord; Acts 20:30 warns of men speaking twisted things; and Luke 23:2 lays the charge on Jesus Himself, that He perverts the nation. ⚙ The word travels in one direction: it names what has been bent away from a straight thing. We do not claim Paul is quoting Matthew. Both are most likely drawing on the Song of Moses, which is why we tier this structural rather than verbal even though the shared lexemes are rare.
Philippians 2:15 · Acts 13:10 · Acts 20:30 · Luke 23:2
basis: Verifier: Philippians 2:15 shares G1074 γενεά (37 vv) and G1294 διαστρέφω (7 vv); Acts 13:10 shares G1294 διαστρέφω, G5599 ὦ (17 vv), G5207 υἱός; Acts 20:30 and Luke 23:2 share G1294 διαστρέφω. Rare shared lexemes, Greek to Greek, but deliberately not tiered verbal: the likeliest explanation is common dependence on Deuteronomy 32:5 rather than one text quoting the other.
Pulpit states it plainly: "As during the absence of Moses at Sinai the people had fallen into idolatry (Exodus 32), so now, when their Master and their leaders were withdrawn, the nine apostles bad faltered in faith." The figure is fed by the words themselves, since the rebuke is worded from Moses' own song. ⚙ The correspondences are real and worth naming: a mountain of glory, a delegated leadership left below, a collapse in the camp during the absence, a descent, and words of indictment against a crooked generation. The differences are just as real and we will not blur them: Moses comes down in anger and breaks the tablets, and three thousand fall; Jesus comes down in grief, asks how long He must bear with them, and heals the boy. Pulpit's own summary, "in the midst of wrath he remembers mercy," marks the turn.
Exodus 32:7-19 · Deuteronomy 9:12-17
basis: Figural reading of the descent pattern, not a verbal link: cross-Testament (Greek to Hebrew), so no shared Strong's lexeme is available and the Verifier offered none. Attested rather than novel: the comparison is stated by Pulpit at Matthew 17:14. Deuteronomy 9:12-17 is Moses' own retelling of the same descent and is included on that basis, not on a computed one.
Pulpit draws the analogy at v. 17: "The prophet's staff in Gehazi's hand could not awake the dead; Elisha himself must undertake the work (2 Kings 4:31); so if the desired miracle had to he performed, Christ himself must do it." ⚙ Ellicott's reading of the disciples' failure fits the figure exactly, that they trusted a delegated formula of the Master's name without the faith it was meant to express. Both are readings offered by commentators, and 2 Kings is Hebrew against Matthew's Greek, so nothing verbal is claimed.
2 Kings 4:29-35
basis: Cross-Testament figural comparison with no shared Strong's basis possible (Hebrew to Greek) and none computed by the Verifier. Recorded basis: Pulpit's own citation of 2 Kings 4:31 in the comment on Matthew 17:17. Attestation rests on that single named source.
Mark 9:25, quoted by Barnes, Ellicott and Poole, has Jesus add "and enter no more into him," and Pulpit reads the closing hour of Matthew 17:18 as meaning the boy was "never more to fall under the devil's influence." ⚙ It is tempting to put that beside Matthew 12:43-45, where an unclean spirit returns to a swept house with seven others, and to read the prohibition as the answer to that danger. We flag rather than assert it: the Verifier returned no lexeme overlap between this unit and Matthew 12:43-45, the decisive clause is not in Matthew's text of this episode at all but in Mark's, and none of the ten commentators in our brief makes the connection. Verify before using it.
Matthew 12:43-45 · Mark 9:25
basis: No Verifier candidate and no shared lexeme recorded between this unit and Matthew 12:43-45; the load-bearing clause ("enter no more into him") belongs to Mark 9:25, not to Matthew 17:14-18, and no commentator in the brief asserts the link. Offered as a question, not a basis.
AI-generated reading; weigh it against the text.
Every attempted remedy in this episode stops one step short of a person. The father brings the boy near (προσήνεγκα, v. 16) and reaches the disciples; the disciples act and are not enabled; the scribes dispute (Mark 9:14, cited by Ellicott, Barnes, Poole and Pulpit) and heal no one. The unit's remedy is a dative pronoun at the end of a sentence: φέρετέ αὐτὸν ὧδέ μοι, bear him here to me. Pulpit underlines that He tells them to bring the boy "to me, not to the disciples," and compares Elisha, who could not send a staff in Gehazi's hand but had to come himself (2 Kings 4:31). Henry reads the whole scene as "an emblem of Christ's undertaking as our Redeemer" and applies it to parents whose children's souls are under Satan's power: "he is able to heal them, and as willing as he is able." Ellicott gathers all such works into Peter's single sentence, that "He went about doing good, and healing all that were oppressed of the devil" (Acts 10:38). ⚙ The christological claim here is modest and firm: the authority is not transferable as technique. It is personal, and it is exercised.
Acts 10:38 · 2 Kings 4:29-35 · Mark 9:23
In Deuteronomy 32 and Numbers 14 the words "crooked generation" and "how long must I bear with you" belong to the LORD. Gill says the rebuke of v. 17 is worded from Deuteronomy 32:5; Pulpit finds Moses' own Septuagint verb in διεστραμμένη. In Matthew they are on the lips of Jesus, without apology, formula, or any attempt to soften the transfer, and the Evangelist reports it as unremarkable. ⚙ Two things follow. First, the speaker stands where the covenant God stood in the wilderness texts. Second, the verb He chooses for His own posture toward them, ἀνέξομαι, means to hold oneself up under a weight, and it is future: the bearing continues. Ellicott hears in the question the shape of the whole life, "one long tolerance of the waywardness and perversity of men," and Pulpit hears sorrow rather than impatience. JFB fixes the episode's position for us, between the first explicit announcement of His death and the second. ⚙ The Judge who names the generation crooked is the one who is on His way to carry it. That is what keeps the rebuke from being a dismissal, and Pulpit's own phrase closes it: "in the midst of wrath he remembers mercy."
Deuteronomy 32:5 · Numbers 14:27 · Matthew 17:22-23
⚙ A reading we mark as our own. The unit's architecture is a descent: glory on the mountain, then a crowd, a convulsing boy, a failed ministry, an argument, and one hour in which it is undone. Pulpit and Cambridge both invoke Raphael's two panels, and Ellicott insists the two scenes were not simultaneous, which is the point structurally as well as historically: the glory is not the setting of the healing, it is what the healer came down from. Paul's hymn moves the same way, and by a coincidence of vocabulary that we do not press, the one other New Testament verse pairing γενεά with διαστρέφω sits a few lines after it (Philippians 2:15). ⚙ We flag this as a pattern we are drawing rather than one our commentators assert: none of the ten voices in our brief connects the descent from the Transfiguration to Philippians 2. It is offered to be tested, and the sober version of it is simply this, that in Matthew the mountain is never a destination.
Philippians 2:5-8 · Philippians 2:15 · Matthew 17:1-9
The biblical text is the Berean Standard Bible (BSB), public domain (CC0). Hebrew/Greek text, transliteration, morphology and Strong’s are transcribed from the Berean interlinear (CC0) + Strong’s lexicons (PD); the literal renderings, divergence notes, word notes and all synthesis are this tool’s own work (⚙) — fallible; verify them.
Named voices, quoted verbatim from public-domain works:
⚙ Honesty notes for Matthew 17:14-18. Text. The BSB text and the Berean interlinear with Strong's numbers are the sourced base; word indices in the notes are that interlinear's. Several English words in this unit are supplied and the interlinear marks them: "to [Jesus]" at 17:14 (the Greek has only the dative αὐτῷ), "[the boy]" at 17:17 and 17:18 (bare αὐτόν, αὐτοῦ), and "[the demon]" at 17:18, where the dative αὐτῷ is grammatically open. Pulpit records that some have taken that pronoun of the boy; Gill argues at length for the demon from Mark and Luke, and the Vulgate and Oriental versions he cites read simply "he rebuked him." Our literal line keeps the pronoun and puts the identification in the note, because the identification is an inference, however well grounded.
Matthew's compression. Matthew's account is the shortest of the three. Almost every vivid detail our commentators use, the dumb spirit, the tearing, the foaming and gnashing, the wasting, "from a child," "if thou canst do anything," "Lord, I believe; help thou mine unbelief," the boy left as one dead and lifted by the hand, comes from Mark 9:14-29 or Luke 9:37-42, not from these five verses. We have marked such material as coming from the parallels wherever we used it. JFB declines to expound the passage here at all and refers the reader to Mark; Poole's entry on 17:14-16 and 17:17-18 is largely a transcription of Mark and Luke.
What is uncertain. (1) The addressee of "O unbelieving and perverse generation" is genuinely disputed among our sources, and we have not harmonised them: Gill says the words are never used of the disciples, Benson says the disciples "were not faithless" and aims the rebuke at the people and especially the scribes, Cambridge says the disciples are by no means excluded, Pulpit includes them all and especially the nine. (2) The nature of the malady: the Greek word is lunar, the reported symptoms are epileptic, and the Evangelists ascribe the case to a demon. Ellicott, Geneva, Cambridge, Benson and Pulpit all hold both halves together, and Ellicott is explicitly candid that those who deny the possession face difficulties of their own and "we may well be content to confess our inability to solve them." We have not resolved this and do not think the text asks us to. (3) The link between Matthew 17:17 and Deuteronomy 32:5 is asserted by Gill and by Pulpit, not by a citation formula in Matthew, and because it crosses from Greek to Hebrew no shared Strong's basis can be computed; every Old Testament contact in this unit is therefore tiered structural or typological, never verbal. (4) Gill's rabbinic material on "Kordicus" (Mishnah Gittin and its commentators) and his citation of the Vetus Nizzachon are his own secondhand references; we report that he makes them and have not verified the sources. (5) The famous saying about prayer and fasting at Matthew 17:21 lies outside this unit and carries its own well-known textual question; we make no use of it here.
✦ = human, public-domain source, quoted and named. ⚙ = machine synthesis, to be verified. Flagged cross-references are left visible on purpose — the verifier working in the open. “Search the Scriptures daily, whether those things were so.” (Acts 17:11)