Matthew Chapter 17 — The Narrow Door, the Jerusalem Lament, and the Great Banquet

Overview

UPDV chapter 17 covers the narrow door (17:1-5), the Lord Lord saying and the house on rock (17:6-10), the many from east and west (17:11-12), Herod's threat (17:13-15), the Jerusalem lament (17:16-18), the Sabbath healing (17:19-24), the parable of seats at the feast (17:25-29), the instruction to invite the poor (17:30-32), and the great banquet parable (17:33-42). This material comes from canonical Matt 7:21-27, Matt 8:11-12, Matt 12:11, Matt 22:1-14, Matt 23:37-39, and Luke 13:23-14:24.

The chapter follows the Q and Lukan sequence from where chapter 16 ended. Chapter 16 concluded with the journey to Jerusalem (Luke 13:22); chapter 17 picks up at Luke 13:23 and follows Luke's order through 14:24. The compiler scattered this material across five different chapters of his Gospel — the Sermon on the Mount (Matt 7), the centurion story (Matt 8), a Sabbath controversy (Matt 12), a parable discourse (Matt 22), and the woes against the Pharisees (Matt 23) — embedding each piece in his own thematic arrangements.

The chapter contains significant L material alongside its Q core. The Herod fox episode (17:13-15), the Sabbath healing narrative (17:19-22, 24), the seats parable (17:25-28), and the instruction to invite the poor (17:30-32) are all L material that Luke placed between Q blocks, providing narrative context for the discourse. The same structural gap principle discussed in chapter 16 applies here.

The Narrow Door (17:1-5)

"And one said to him, Lord, are there few who are saved?" The narrow door passage (17:1-5) comes from Luke 13:23-27, which has a partial parallel in the compiler's Sermon on the Mount at Matt 7:13-14 (the narrow gate) and Matt 7:22-23 (the exclusion dialogue).

Davies and Allison are notably agnostic about whether the narrow gate/door saying itself comes from Q. They open: "Although 7:13-14 has a parallel in Lk 13:23-4, one cannot with great confidence posit a common origin in Q. The differences between the two texts are considerable." They list multiple possibilities — Q with heavy Matthean redaction, M and L independently, or differences between Qmt and Qlk — and explicitly state: "We leave the problem open."

The differences are indeed substantial. Luke has θύρα (thura, "door") while the compiler has πύλη (pulē, "gate"); Luke has ἀγωνίζεσθε (agōnizesthe, "strive") while the compiler has an apodictic imperative; the compiler expands the saying into a full two-ways antithetical structure (broad gate/destruction vs. narrow gate/life) that has no Lukan parallel. Davies and Allison reconstruct the probable original as closer to Luke: "Strive to enter (into life or the kingdom) by the narrow gate (or door)" — a summons to discipleship that the compiler expanded with "conventional accretions associated with the theme of the two ways."

The exclusion dialogue (17:3-5) is clearer. Davies and Allison identify Matt 7:22-23 as Q from Luke 13:26-27, but "a very free redactional construction in the interests of the Matthean context." The compiler changed the subject from unbelieving Jews — who ate and drank in Jesus' presence and heard him teach in their streets — to false prophets who prophesied, cast out demons, and performed mighty works in Jesus' name. He changed Luke's "I do not know where you come from" (οἶδα, oida) to "I never knew you" (ἔγνων, egnōn), "probably so that 7:23a would cover the protracted period of ministry presupposed by 7:22." And he used ὁμολογήσω (homologēsō, "I will confess/declare") instead of Luke's ἐρεῖ (erei, "he will say") "because of its solemnity, public character, and legal sense (which connotes irreversibility)." Davies and Allison are explicit: "Compared with this [Lk 13:26-7], Mt 7:22-3 is secondary." The UPDV restores Luke's Q form — the intimacy of table fellowship replaced by the grandiosity of charismatic performance.

Lord, Lord and the House on Rock (17:6-10)

"Not everyone who says to me, Lord, Lord, will enter into the kingdom of heaven; but he who does the will of my Father who is in heaven." The Lord Lord saying (17:6) comes from canonical Matt 7:21, which is Q material paralleled at Luke 6:46. The house on rock parable (17:7-10) comes from Matt 7:24-27, paralleled at Luke 6:47-49.

The UPDV places these sayings after the narrow door and exclusion dialogue because they form a thematic unit: the narrow door warns about exclusion; the Lord Lord saying states the criterion for inclusion (doing, not merely professing); and the house on rock illustrates the principle through parable. In Q, the Lord Lord saying and the house parable closed the sermon (both Matt 5-7 and Luke 6:17-49 end with them). The UPDV relocates them from the sermon conclusion to this point because the compiler had already embedded them in his Sermon on the Mount, and here they join the Q material about entrance and exclusion that the compiler scattered.

Davies and Allison identify 7:21 as Q from Luke 6:46: "This last verse is a skandalon for the Griesbach hypothesis. Luke omits precisely what is characteristic of Matthew, as the statistics show. One may infer... that Matthew and Luke independently drew upon the same source, Q." The UPDV retains the compiler's formulation rather than reverting to Luke's simpler "Why do you call me Lord, Lord, and don't do what I tell you?" because the two forms express the same principle — obedience, not verbal profession, determines standing — and the compiler's version, while characteristically Matthean in vocabulary ("kingdom of heaven," "my Father who is in heaven"), does not distort the Q meaning.

For the house parable, Davies and Allison confirm it "terminated Q's sermon on the plain. Both Matthew and Luke have let it retain this function in their gospels." The compiler created perfect antithetical parallelism — "the rain descended, and the floods came, and the winds blew" — repeated identically across both builders, with only the final verb changing (προσέπεσαν, prosepesen, "fell upon" vs. προσέκοψαν, prosekopsan, "struck against"). Davies and Allison identify this heightened parallelism as secondary: "Matthew has revised Q so as to make this verse conform to 7:24. The result is perfect parallelism." But they also note that Matthew preserves the original point better than Luke. Manson's judgment is decisive: "The simpler statement of Mt. is to be preferred... We are to imagine both houses as equally well built. The only difference is that correctly given by Mt.: one was founded on rock and the other on sand." Luke's "because it had been well built" misses the point — the issue is foundation, not construction quality. The UPDV retains the compiler's form because the parallelism enhances without distorting, and Matthew's theological emphasis (foundation choice) is closer to Q than Luke's (building technique).

The storm in the parable is eschatological, not quotidian. Davies and Allison observe that "in the OT the storm often represents God's judgement," and the image recalls Noah's flood, which "is taken up in an attempt to portray the eschatological affliction" (cf. the days of Noah in chapter 16). The house that survives the storm is the obedient life that endures divine judgment.

Many from East and West (17:11-12)

"And I say to you⁺, that many will come from the east and the west, and will sit down with Abraham, and Isaac, and Jacob, in the kingdom of heaven." The east and west saying (17:11-12) comes from canonical Matt 8:11-12, which is Q material that the compiler displaced from its original context at Luke 13:28-29.

Davies and Allison build a strong case for displacement: "In this verse and the next Matthew interpolates a saying which had a different context in Q; see Lk 13:28f." They demonstrate that Luke preserves Q's context: "Lk 13:23-30 consists entirely of Q material, it makes fair sense as it stands, and as Luke is much more apt than Matthew to preserve Q material intact, presumption favours taking Lk 13:23-30 as a block from the sayings source." The compiler relocated it to the centurion story to "highlight the contrast between the faith of a Gentile and the unbelief of 'the sons of the kingdom.'"

The compiler also rearranged the internal order. In Luke, the sequence runs: weeping (28a) → when you see Abraham and yourselves cast out (28b-c) → they will come from east and west (29a) → they will recline (29b). The compiler reversed this to: many will come (11a) → recline with Abraham (11b) → sons of kingdom cast out (12a) → weeping and gnashing (12b). Davies and Allison explain each step of the rearrangement through Matthean redactional logic: the compiler could not open with weeping because "as of yet there is no antecedent for ἐκεῖ," so he started with the arrival, then moved "weeping and gnashing of teeth" to the end — "which accords with his tendency to put the phrase... at the end of sections (13:50; 24:51; 25:30)."

The UPDV retains the compiler's wording here, including "the sons of the kingdom," "the outer darkness," and the "weeping and gnashing of teeth" formula. This creates a tension with the UPDV's practice in chapter 16, where the same "weeping and gnashing of teeth" refrain was stripped from Matt 24:51 as "a concluding redactional refrain." The defense is that in Matt 24:51 the formula is a stereotyped parable ending appended by the compiler to multiple parables (13:42, 50; 22:13; 24:51; 25:30), whereas in Matt 8:12 it describes the specific response of those excluded from the messianic feast — a functional part of the saying rather than a formulaic appendage. But the "outer darkness" (τὸ σκότος τὸ ἐξώτερον, to skotos to exōteron) is a distinctively Matthean phrase appearing only at 8:12, 22:13, and 25:30, and "sons of the kingdom" is Matthean redaction for Luke's direct "you yourselves." A future revision should consider whether the Lukan form of this saying — which Davies and Allison judge closer to Q — should replace the compiler's version.

Davies and Allison's most striking contribution concerns who "the many from east and west" are. Against nearly all commentators, they argue these are not Gentiles but diaspora Jews: the allusion to Psalm 107:3 concerns the return of Jewish exiles; "east and west" in Jewish texts "is frequently associated with the return of diaspora Jews to their land" (Babylon and Egypt); and "the coming of the Gentiles is never conceived of as a judgement upon Israel" in the OT. Their conclusion: "As spoken by Jesus, Mt 8:11f. par. was intended to draw a stark contrast not between unbelieving Jews and believing Gentiles but between privileged and unprivileged Jews." The "sons of the kingdom" are not all Israel but privileged religious leaders — "yet one more example of the conviction that the first will be last, the last first."

Herod's Threat (17:13-15)

"In that very hour there came certain Pharisees, saying to him, Get out, and go from here: for Herod wants to kill you." The Herod fox episode (17:13-15) comes from Luke 13:31-33 and has no parallel in the compiler's Gospel.

This is L material. Bovon identifies it explicitly: "L (rather than Q, since Matthew was unacquainted with this episode) transmitted a brief biographical apophthegm." He concludes that v. 32b is traditional (its vocabulary is "scarcely Lukan"), while v. 33 is a Lukan redactional interpretation, and v. 31, while written by Luke, draws on traditional information. Marshall surveys the full range of scholarly opinion, from those who find the pericope "substantially historical, claiming that its very obscurity spoke against its being a literary invention" to Denaux, who "claims that the vocabulary of the section can all be attributed to Luke."

Bovon's analysis of the fox metaphor is illuminating: "The metaphor of the fox seems to have a double significance: it suggests (a) that Herod resorted to a ruse and (b) that his power was negligible. Often contrasted in antiquity with the lion, the 'fox' (here Herod) did not measure up when he was faced with Pilate or the Roman power, or especially Jesus and the divine power." On the "third day" reference, Bovon argues persuasively that it "designates the end of Jesus' ministry rather than an immediate resurrection. The fact that it did not match the Christian kerygma is a strong indication in favor of the authenticity of the saying." If early Christians had composed this, they would have made the resurrection allusion explicit; its very awkwardness testifies to pre-Easter origin.

The UPDV includes this L passage because it occupies the narrative space between two Q blocks: the narrow door discourse (Luke 13:23-30) and the Jerusalem lament (Luke 13:34-35). The episode also provides the geographical and dramatic link to the lament — Jesus is told Herod wants to kill him, he responds that "a prophet perishes in Jerusalem," and the lament over Jerusalem follows immediately. Without 13:31-33, the connection between the narrow door discourse and the Jerusalem lament would be abrupt.

The Jerusalem Lament (17:16-18)

"O Jerusalem, Jerusalem, that kills the prophets, and stones those who are sent to her!" The Jerusalem lament (17:16-18) comes from canonical Matt 23:37-39, which is Q material paralleled at Luke 13:34-35.

Davies and Allison assign this to Q with high confidence. Their evidence for the two-source theory is twofold: first, the Griesbach hypothesis "requires that Luke changed a saying spoken by Jesus and directly addressed to contemporaries into a reported speech of pre-existent Wisdom" — unlikely given Luke's "lack of interest in the figure of Wisdom." Second, Matthew has secondary details that "call to mind the fate of Jesus and his disciples ('crucify', 'scourge in synagogues')."

On the original Q context, Davies and Allison make a notable argument: they believe Luke moved the lament from its Q position to create a Jerusalem cluster. "Luke is, as a general rule, closer to Q's order. But in this particular case it would seem that Luke moved the unit in order to bring together texts about Jerusalem: 13:22-30 (with its redactional introduction naming Jerusalem) is followed by 13:31-3 ('it is impossible for a prophet to be killed outside of Jerusalem') which is followed by 13:34-5 ('Jerusalem, Jerusalem, the city that kills the prophets')." They suggest that in Q, the lament may have been a continuation of Wisdom's words (Lk 11:49-51), placed in the cluster of polemical sayings — closer to Matthew's placement at the end of the woes.

The UPDV follows Luke's arrangement despite this evidence. This is a case where the UPDV's general principle — follow Luke's Q sequence — conflicts with the specific evidence. If the goal is Q reconstruction, and D&A are right that Luke moved the lament to create a Jerusalem cluster, then following Luke here reconstructs Lukan redaction, not Q. The UPDV's defense is consistency: adopting Luke's order as the default and making case-by-case exceptions risks introducing more editorial judgment than it eliminates. But the reader should recognize that the placement at 17:16-18 (following the Herod fox episode) reflects Luke's narrative design, and that Q may have placed the lament after the woes discourse — closer to the compiler's arrangement at Matt 23:37-39.

The compiler's most significant editorial change is the addition of ἔρημος (erēmos, "desolate") to 23:38. Luke reads: "Look, your house is left to you." The compiler reads: "Look, your house is left to you desolate." Davies and Allison identify the addition as Matthean: the word "probably envisages the destruction of AD 70." The UPDV drops "desolate," following Luke's Q form (17:17). Without "desolate," the saying refers to the departure of the divine presence — "originally a reference to the departure of the Shekinah from the temple," as Davies and Allison put it. They add a remarkable christological observation: "Perhaps the distinction is without a difference, for Matthew identified Jesus with the Shekinah."

Davies and Allison dissent from the scholarly consensus that this passage records God's definitive rejection of Israel. They argue that v. 39 is "best understood not as a pronouncement of condemnation but instead as a promise of redemption: the Messiah will come when his people repent." The text is a conditional sentence: "when his people bless him, the Messiah will come" — supported by Jewish sources on contingent timing of redemption and the standard pattern of sin, punishment, and hope. This reading transforms the lament from a verdict to an appeal: "the prophetic judgements are mingled with affection and Jesus becomes, like Jeremiah, a reluctant prophet."

The Sabbath Healing (17:19-24)

"And it came to pass, when he went into the house of one of the rulers of the Pharisees on a Sabbath to eat bread, that they were watching him." The Sabbath healing of the man with dropsy (17:19-24) comes from Luke 14:1-6, with the sheep-in-the-pit argument (17:23) taken from Matt 12:11.

The narrative framework (17:19-22, 24) is L material — Nolland concludes there is "no clear parallel to this report in the synoptic Gospel materials." But the core halakhic argument has a clear parallel in Matt 12:11, and scholars agree both derive from the same tradition. Bovon calls it "the oldest saying preserved by Luke, one known to Matthew as well, one that must have been pronounced by the historical Jesus." He provides a side-by-side comparison showing the two Greek forms "do not rely on the same Greek source, since the wording is divergent throughout. Instead, they furnish us with two independent translations of the same saying."

The UPDV uses the compiler's form (Matt 12:11) for the argument — "What man will there be of you, that will have one sheep, and if this falls into a pit on the Sabbath day, will he not lay hold on it, and lift it out?" — embedded in Luke's narrative setting. This creates a composite text: L's narrative framework (the Pharisee's dinner, the man with dropsy, the silent opponents) with the compiler's halakhic saying inserted where Luke 14:5 would stand. Bovon confirmed that the two Greek forms are "two independent translations of the same saying," not one copied from the other. The UPDV chose the Matthean form, but the result is a hybrid that never existed in any historical source document — L's scene with the compiler's vocabulary. The reader should recognize this as an editorial construction, not a recovered text.

Davies and Allison note that Luke's version (14:5) probably preserves the more original wording. Luke has υἱὸς ἢ βοῦς (huios ē bous, "son or ox"), which yields Aramaic paronomasia: בְּרָא (bera', son), בְּעִירָא (be'ira', ox), בְּאֵרָא (be'era', well). The compiler changed this to "sheep" for two reasons: "the evangelist wished to make the argument a fortiori" — a son cannot be the lesser in a comparison meant to prove humans outweigh animals — "and in composing 12:9-14 he was probably reminded of the parable of the lost sheep (18:12-14; cf. Lk 15:4-7) and assimilated one passage to the other." The UPDV keeps the compiler's "sheep," but this is a case where the original wording is recoverable with unusual confidence. The Aramaic triple wordplay (bera'/be'ira'/be'era') strongly favors "son or ox" as Q's reading, and the compiler's substitution was motivated by his own editorial program (the a fortiori structure and lost-sheep assimilation), not by a different source tradition. A future revision should consider reverting to Luke's form.

Davies and Allison affirm the saying's authenticity: "The saying presupposes Jewish sentiment (cf. Deut 22:4), contains an expression characteristic of Jesus (τίς ἔσται ἐξ ὑμῶν ἄνθρωπος), is consistent with the several synoptic stories in which Jesus acts provocatively on the sabbath, and harmonizes with his tendency to exalt mercy over his contemporaries' understanding of holiness."

Seats at the Feast (17:25-29)

"And he spoke a parable to those who were invited, when he marked how they chose out the chief seats." The seats parable (17:25-29) comes from Luke 14:7-11, which is L material with no Matthean parallel for vv. 7-10. The concluding saying (17:29 = Luke 14:11) — "everyone who exalts himself will be humbled; and he who humbles himself will be exalted" — is a widely circulating Q logion paralleled at Matt 23:12 and recurring at Luke 18:14.

Marshall notes that a version of vv. 8-11 appears in Codex D after Matt 20:28, "which could be independent of Luke's text" — evidence that this teaching circulated in multiple streams. Bovon treats the passage as authentic teaching that functions at both social and theological levels, arguing "there is no reason to doubt its authenticity."

The exaltation/humbling logion (17:29) is a piece of floating tradition. Marshall identifies it as "a translation-variant of the same Aramaic original," with Luke "giving the more literary version." The UPDV retains Luke's form here. The compiler used the same saying at Matt 23:12 to conclude his woes against the Pharisees; the UPDV relocates that occurrence to this point, following Luke's sequence.

Invite the Poor (17:30-32)

"And he said to him also that had invited him, When you make a dinner or a supper, do not call your friends, nor your brothers, nor your kinsmen, nor rich neighbors." The instruction to invite the poor (17:30-32) comes from Luke 14:12-14 and is L material — Marshall confirms it is "peculiar to Lk." He notes: "while the thoughts would be congenial to Luke, there is no evidence that they are created by him or by the early church."

Nolland observes that the list of beneficiaries — "the poor, the crippled, the lame, and the blind" (17:31) — reappears in the great banquet parable at 17:39 (= Luke 14:21). He argues that Luke drew the list from the parable (14:21) for use here (14:13), not the reverse, creating a deliberate verbal link between the instruction and the parable that follows.

The passage sets up the great banquet by establishing the principle: true hospitality invites those who cannot reciprocate. The guest's beatitude (17:33 — "Blessed is he who will eat bread in the kingdom of God") then triggers the parable that illustrates what happens when the expected guests refuse.

The Great Banquet (17:33-42)

"But he said to him, A certain man made a great supper; and he invited many: and he sent forth his slave at supper time to say to those who were invited, Come; for all things are now ready." The great banquet parable (17:33-42) comes from Luke 14:15-24, which is Q material with a parallel in the compiler's wedding feast at Matt 22:1-14. A third version exists in Gospel of Thomas 64.

Marshall states the scholarly consensus plainly: "A very similar parable appears in Mt. 22:1-14, but with considerable alteration and addition. It is generally accepted that the two parables are variants of one original theme, and considerable progress can be made towards reconstructing a basic form of the parable (which turns out to be very close to the Lucan form)."

The compiler's editorial transformation was massive. He turned a private host into a king (= God), a supper into a wedding feast for his son (= Jesus), one slave into plural servants, and realistic excuses about property and marriage into violent rejection — the invited guests kill the king's servants, prompting the king to send armies and burn their city. This last detail is widely recognized as an allegorical reference to the destruction of Jerusalem in 70 CE. Nolland identifies the changes systematically: the king and son are "a natural allegorical development (the son plays no further role in the parable)"; the violence against servants is "certainly secondary"; and the punitive campaign reflects post-70 theology. The compiler then appended a second parable — the wedding garment (22:11-14) — with its own concluding formula: "many are called, but few chosen."

The UPDV replaces the entire Matthean parable with Luke's form because Luke preserves Q's original shape. Luke's version has a private host, a single servant, three specific excuses (a field, five yoke of oxen, a new wife), and two rounds of gathering replacement guests. Nolland argues that Luke's two sendings — first to "the streets and lanes of the city" for "the poor and maimed and blind and lame" (17:39), then to "the highways and hedges" (17:41) — are more original than Matthew's single sending. Matthew conflated the two rounds, but their trace survives in his own doubled dispatch of servants (22:3-4). Nolland also gives Luke priority for the guest list: "his version fits the priorities of the ministry of Jesus rather better than Matthew's 'as many as you find.'"

The final verse (17:42) — "For I say to you, that none of those men who were invited will taste of my supper" — shifts from narrative to direct address. Nolland identifies it as a Lukan formulation where the host "steps as it were on to the apron of the stage and addresses the audience." Luke identifies the host with κύριος (kurios, "lord") — an allegorical identification of Jesus as the speaker. Bovon sees the two sendings as reflecting the expansion of mission: "the master enlarged the circle of his beneficiaries: he sent his servant out of town, made him leave the alleys for the roads, the town for the countryside, Israel for the Gentiles."

What the UPDV Removes from This Section

Matt 7:14 (how narrow is the gate / few find it): The compiler's two-ways expansion. Davies and Allison reconstruct Q as a single saying (Luke's form), with the broad-way/destruction antithesis as "conventional accretions associated with the theme of the two ways."
Matt 7:22-23 (prophesying, casting out demons, mighty works): Revised to Luke 13:26-27 (eating and drinking in Jesus' presence). D&A: the compiler "changed the subject from unbelieving Jews to false prophets who profess Jesus Christ" — "a very free redactional construction."
Matt 22:1-14 (wedding feast): The entire Matthean parable replaced by Luke 14:15-24 (great supper). The compiler's king, son, murdered servants, destroyed city, and wedding garment are all secondary allegorization. The "many are called, but few chosen" formula (22:14) is removed.
Matt 23:38 — revised: "desolate" (ἔρημος) removed, following Luke's Q form. D&A: "a Matthean addition which probably envisages the destruction of AD 70."

References

Davies, W. D. and Dale C. Allison Jr. A Critical and Exegetical Commentary on the Gospel according to Saint Matthew. 3 vols. International Critical Commentary. Edinburgh: T&T Clark, 1988–1997.
Bovon, François. Luke. 3 vols. Hermeneia. Minneapolis: Fortress, 2002–2012.
Marshall, I. Howard. The Gospel of Luke: A Commentary on the Greek Text. New International Greek Testament Commentary. Grand Rapids: Eerdmans, 1978.
Nolland, John. Luke. 3 vols. Word Biblical Commentary 35. Dallas: Word, 1989–1993.