The Report of the Women (28:1-10)

Cf. Mk 16:1-8; Lk 24:1-12; Jn 20:1-18; Mk 16:9-11. The various resurrection narratives vary considerably in length, focus, and detail. If Q included a resurrection narrative, most of the Gospel writers treated it as one among many; given the many witnesses of the risen Christ (1 Cor 15:6), it is hardly surprising that numerous accounts would exist and different Gospel writers would draw on different accounts. The four Gospels differ in detail, but in all four the women become the first witnesses, and Mary Magdalene is explicitly named as one witness among them (also Gos. Pet. 12:50–13:57). Sanders 1993: 280 may be right to argue that “a calculated deception should have produced greater unanimity. Instead, there seem to have been competitors: ‘I saw him first!’ ‘No! I did.’” But the divergent details suggest independent traditions, thereby underlining the likelihood of details the accounts share in common (e.g., Boyd 1995: 277-78). This fits what we should expect of eyewitness traditions. (Thus, for example, though two eyewitnesses who accompanied Alexander agreed that Callisthenes was indicted, publicly scorned, and died, and though their accounts could be called entirely trustworthy [pany pistoi], they differ even on whether he died by sickness or hanging — Arrian Alex. 4.14.3. The variation in the Gospel accounts is far less significant than this.)

The variation in length of the Gospels’ resurrection narratives (Lk 24 is long though recapitulated briefly in Acts 1; Mt 28 is quite brief) may represent the desire to make optimum use of scroll length rather than leaving a blank space at the end (as sometimes happened — Diog. Laert. 6.2.38). Josephus once seems caught unexpectedly by the end of his scroll (Jos. Apion 1.320); Matthew, approaching the length limit of his standardized scroll (see the Introduction), may hasten to his conclusion. He surveys the testimony of the women (28:1-10), the testimony of the guards (28:11-15), and finally the testimony of the church (28:16-20) related to that of the women.

The Gospel narratives agree that the revelation of Christ’s resurrection began on the first day of the week, after the Sabbath (Mt 28:1; Mk 16:1; Lk 24:1; Jn 20:1; cf. 20:19, 26). Especially in Mark and Matthew, this language makes it clear that the earliest Christians regarded Sunday as a special day celebrating the resurrection (cf. Acts 20:7; 1 Cor 16:2; Justin 1 Apol. 67; Iren. frg. 7), perhaps even “the Lord’s day” (cf. Rev 1:10; Did. 14.1; Vanni 1978; but note Strand 1967; Lewis 1968), but not as a new Sabbath (this developed in the second century and later; cf. Ign. Magn. 9.1; Barn. 15.8-9; Chadwick 1967: 128; Bacchiocchi 1977), which among the earliest Jewish Christians remained on the last day of the week. The tradition is too early to be influenced by Mithraism (pace Cary and Haarhoff 1946: 344), which did not spread widely in the Roman world until the next century (Grant 1986: 40-41);

this simply was the day Jesus’ followers found the empty tomb, the day after the Sabbath. Sunday became the “Lord’s Day” because of the discovery of the empty tomb rather than the reverse.

First, the narrative demonstrates that those of whom society thinks the least are often those whom God sends with his message. Not all testimony was regarded as being of equal merit; the trustworthiness of witnesses was considered essential (CD 9.21-22; 10.1). Most of Jesus’ Jewish contemporaries held little esteem for the testimony of women;

this reflects a broader Mediterranean limited trust of women’s speech and testimony also enshrined in Roman law. By contrast, the guards’ report that the disciples had stolen the body (28:11-15) might command much greater respect then, and in an antisupernaturalistic culture like much of modern academia as well. Indeed, even the disciples in the late tradition in Mark 16:11 did not believe the women — a tradition that may reflect historical reality at this point (Vermes 1973: 40-41; cf. Ps-Philo 9:10). For the early Christians, neither the empty tomb nor the testimony of the women was adequate evidence by itself (cf. Lk 24:22-24); they also depended on the testimony of men for the public forum (1 Cor 15:5-8; cf. Stauffer 1960: 151; Dunn 1975: 126). No one had apologetic reason to invent the testimony of these women, but the Gospel writers may have a profound theological purpose in preserving it.

Matthew lays these two reports, the true and the false, side by side, forcing his audience to declare their choice. The testimony of the women thus becomes a model for the disciples who will follow them (28:16-20). Jesus commissions them as his shaliaḥim — agents or apostled ones (see comment on 10:5) — to bring news of his resurrection to his own disciples. Their faithfulness, like Joseph’s (27:55-61), is laid over against the authorities’ deceitful accusation of deceit (27:62-66); Matthew thereby calls his audience to suffer rejection and dishonor at the hands of the hostile authorities of their own day.

In ancient literature, one woman’s wrong behavior could be held against her gender (Hom. Od. 11.432-34, 436-39, even though Clytemnestra also slew Cassandra in 11.422; Eurip. Orestes 1153-54); likewise, a woman’s prowess or courage could shame men into acts of valor (Virg. Aen. 11.734; Ovid Metam. 8.380-89, 392, 401-2), though they might also allow the woman’s honor (Virg. Aen. 11.846-47; Ovid Metam. 8.425-27). Such examples suggest that ancient readers generally noticed the gender of heroes when they were female; it is thus not unreasonable to suppose that Matthew expects the positive performance of these women to reflect favorably on their gender. But more importantly, as female examples could provoke both women and men, the women in Matthew’s narrative function as a model for disciples, male or female.

Although Matthew is less explicit than Mark about unfinished business in the burial (Mk 16:1), the approach of the Sabbath with sundown (Mt 27:57)

and the return of the women after the Sabbath had ended Saturday at sundown and Sunday’s morning light had come suggest that this was the purpose for which the women came in the first Gospel as well. (Gundry 1982: 585-86 thinks that the women in Matthew arrive Saturday evening, but against this interpretation, epiphōskō in Mt 28:1 means “grow toward dawn.”) Jewish custom permitted women to attend to corpses of either gender, although men could not tend women’s corpses, and women were often permitted to engage in such preparatory burial rituals (for comparison with the Greek world, see Abrahamsen 1986: 121-22).

Second, the narrative reveals God’s power (28:2-3). The angelic revelation exhibits points of contact with biblical theophanies. The description of glory recalls Jesus’ own in 17:2 (cf. Dan 7:9; 10:5-6). In 4 Ezra, the heavenly Zion (10:25-27) has a countenance that flashes “like lightning” (10:25) and at its voice the earth quakes (10:26). Elsewhere, lightnings flash from the fiery face of a mighty angel and earthquakes accompany him (3 Enoch 22:9). (On such appearances see comment on 17:6-8.)

The white clothes could allude to a variety of functions. Priests generally wore linen, including Egyptian priests (Appian C.W. 4.6.47; Plut. Isis 3-4, Mor. 352C; Apul. Metam. 11.10, 23; Lewis 1983: 92), those at the temple of Artemis (Acts of Jn 38), and Jewish priests (Jos. War 5.229; Pesiq. R. 33:10; Yadin 1962: 219). Worshipers wore white or linen in other worship settings (Eurip. Bacch. 112; Paus. 2.35.5; 6.20.3; Diog. Laert. 8.1.33; SEG 11.923 in Sherk 1988: 58; Ramsay 1904a: 386; cf. the change of garments in Olmstead 1959: 511), including the Jerusalem temple (Jos. War 2.1; Ant. 11.327; Safrai 1974/76d: 877) and that of the Therapeutae during worship (Philo Cont. 66). A philosopher like Pythagoras might wear white wool (replaced in later times with linen — Diog. Laert. 8.1.19). Perhaps because white could signify good and black evil (Diog. Laert. 8.1.34; Ovid Metam. 2.832),

converts might wear linen (Jos. and Asen. 14:12/13; cf. Rev 3:4-5; 4:4; 19:8, 14). Pagan deities could appear in white garments (e.g., PGM 4.637-38, 698-99; inscr. in Grant 1953: 16); Jewish angels likewise appeared in linen (Ps-Philo 9:10; Rev 15:6) or white garments (1 Enoch 71:1; 87:2; 90:31-33; 2 Macc 3:26; 11:8; Jannes and Jambres fragments in Chester Beatty Papryrus 16; cf. the exception in late Pesiq. R. 20:4), or clothed in glory (3 Macc 6:18; 1 Enoch 71:1; cf. Adam in Gen. Rab. 20:12). In this context, the angelic or theophanic functions are paramount.

That the angel “sits” on the stone (28:2; cf. Mk 16:5, likely seated elsewhere) is itself a dramatic statement of supernatural triumph since the stone, probably disk-shaped, would not naturally accommodate one sitting on it; single-handedly rolling away the stone also indicated great strength (cf. Gen 29:8, 11). Ancients had other traditions of becoming “like the dead” before a deity’s holiness (Diod. Sic. 4.24.5), but 28:4 reflects the biblical tradition (Rev 1:17; see comment on Mt 17:6; for the idiom more generally, cf. Jos. Ant. 6.306). Although the guards feared for their lives, God had no intention of slaying them. Yet despite this manifestation of his power, they failed to believe (28:15).

Matthew abbreviates Mark’s account yet heightens the miraculous;

in Mark, even the angel can be mistaken for a young man, who is less explicit about the resurrection (Montefiore 1968: 1:402 thus surmises that the Markan account may be historically accurate). Some attribute this heightening of the miraculous to haggadic elaboration of Mark (Beare 1981: 542), but Mark may as easily have toned down other features to condense his conclusion (like his introduction in Mk 1:1-13 or 1:1-15) and focus on vindication by the resurrection sign alone.

Various non-Markan material here recurs in John and Luke-Acts (e.g., 28:6; cf. Lk 24:6), suggesting access to non-Markan resurrection traditions or material in a now-lost ending of Mark (Wenham 1973; Gundry 1982: 590-91).

It is, in fact, difficult to doubt that such other traditions would have existed, given the large number of witnesses to the resurrection (cf. 1 Cor 15:5-7). Some scholars are convinced that one can completely harmonize the stories of the women at the tomb if we grant that the Gospel writers reported only data essential to their distinctive accounts (Hodges 1966); others, while acknowledging that the conviction of the resurrection is early, doubt that our current Easter stories belong to the earliest stratum of tradition (Dibelius 1949: 139). While harmonization approaches become strained when they misunderstand the liberties literary historians sometimes took in recounting secondary details (see the Introduction), they do work harder than more skeptical approaches to make the best possible sense of the data we have. On any account, two matters are plain: (1) the differences in accounts demonstrate that the Gospel writers were aware of a variety of independent traditions; (2) these divergent traditions overlap significantly, hence independently corroborate the basic outlines of the story.

Third, the narrative shows that God is selective in his revelation (28:4-9). Although the guards witnessed God’s power, the angel spoke only to the women; often when people fell before a revelation as if they were dead, the revealer declared, “Fear not.”

But here the angel says “Fear not” to the women, not to the guards who had fainted before him (28:4-5; cf. 28:10; 17:7; Mk 16:6; Dan 10:11-12). Jesus appears directly to the women as well, but not to those who did not believe (28:8-10; cf. Acts 10:41). In view of the prejudice against women’s testimony in antiquity (see above), no one would have invented the testimony of the women attested in all four Gospels; indeed, Paul even omits it.

The initial dependence of the men on the testimony of the women reflects the gospel’s power to transcend gender restrictions (Thompson 1985: 233). When the women met Jesus, they worshiped (28:9) — finally responding as the wise Gentiles had (2:2, 11), yet — again with an ironic touch — before the male disciples did (28:17).

Nevertheless, Jesus does not cast off the male disciples here; he identifies the disciples to whom he is sending them as his “brothers” (12:50; 25:40; cf. a similar tradition in Jn 20:17, via Mary Magdalene).

Because Paul explicitly reports only resurrection appearances, some suppose that the empty tomb tradition was a myth;

Mack, for example, suggests that before the Gospels we have only Paul’s account of “visions” (Mack 1988: 308). But while Paul’s language could include visionary experiences, he is reporting earlier Palestinian tradition in 1 Corinthians 15:3-7, and Palestinian Jews did not speak of nonbodily resurrections. (Dibelius 1971: 191, who doubts that Paul had the empty tomb tradition because he does not mention it, nevertheless recognizes that Paul would have to assume it based on Jewish premises about the nature of the resurrection. But if early disciples assumed it they would have checked, and given the availability of traditions concerning the resurrection apearances, Paul had no reason to mention the far more ambiguous evidence of the empty tomb.) Nor would anyone have persecuted them for simply affirming that they had seen someone who had been dead; apart from the bodily character of the resurrection — the sort that would leave an empty tomb — people would merely assume that they claimed to see a ghost, a noncontroversial phenomenon (cf. comment on 14:26; note on 1:20). Ghosts were “phantasms” that appeared especially at night (Plut. Brutus 36), but this is not what the resurrection narratives report (Lk 24:40; Sanders 1993: 278). Further, very little evidence suggests the plausibility of successive and mass, corporate visions (e.g., Schweizer 1971: 48-49; see especially 1 Cor 15:5-7). Those inventing an empty tomb tradition would hardly have included women as the first witnesses (see above), and “Jesus’ resurrection could hardly have been proclaimed in Jerusalem if people knew of a tomb still containing Jesus’ body” (Schweizer 1971: 48).

Those who witnessed Jesus alive from the dead (e.g., 1 Cor 15:1-8; virtually all the narrative accounts also suggest significant conversation with him, rather than fleeting appearances) were so convinced of the veracity of their claims that many devoted their lives to proclaiming what they had seen, and some died for it; clearly their testimony was not fabricated (Sanders 1993: 280).

Supposed pagan parallels to the resurrection stories are weak (see Aune 1981: 48). Most pagans would have preferred to play down a savior’s human death (Boring 1995: 151, comparing Philost. V.A. 7:14).

Excursus on Mysteries, Resurrection, and Salvation

The Mysteries appear to have been less popular than was once thought (Rives 1998). They apparently influenced some Palestinian Jewish thought in late antiquity, though the exact date of the evidence is unclear. Numismatic evidence indicates some presence of the Mysteries in Palestine (Avi-Yonah 1974/76a: 60; Flusser 1974/76: 1,065-1,100); the influence of a third-century-A.D. Mithraeum in Caesarea (on the Mithraeum see Bull 1974; Lease 1975; Flusser 1974/76: 1099) is unclear, since Caesarea was of mixed population and the date is much later than our period (cf. Charlesworth 1985: 82). Mystery language did infiltrate some forms of Judaism (Philonenko 1975; idem 1965),

but pagan accusations that confused Judaism and the Mysteries (Russell 1932: 338; cf. Reitzenstein 1978: 174-84) do not constitute good evidence that Judaism as a whole made that confusion; Reitzenstein’s claim that “Even in Trajan’s time the Roman Jewish community still … either altogether or in large part worshiped the Zeus Hupsistos Ouranios and the Phrygian Attis together with Yahweh” (1978: 125) lacks the support of subsequent research into Roman Judaism.

The language of the Mysteries clearly infiltrated Christian writers of the second century and later. Tertullian claims that Christianity has the true Mysteries, of which others are poorer and later copies (Apol. 47.14). Such language becomes much more prevalent in the third and fourth centuries A.D. (Eliade 1958: 120).

Yet it is possible that some features of the Mysteries derive from Christianity. As they began to lose devotees to Christians in a later period, the Mysteries could have adopted some features of Christianity; many of the “parallels” in the Mysteries occur only in the later period (Metzger 1955: 10-11; Eliade 1958: 115). That the church fathers understood the Mysteries as “imitation demoniaque du Christianisme” (Benoit 1975: 79-81) may suggest that they, like many early modern students of these cults, read them through the grid of their own Christian background, and the ready-to-hand explanation of demonic imitation may have led them to heighten rather than play down the similarities between the two.

Much of the most specifically Mystery vocabulary is lacking in earliest Christianity: Metzger, following Nock, lists such terms as mystēs, mystikos, mystagōgos, katharmos, katharsia, katharsis, teletē, and so on (1955: 11). What is perhaps more significant is the different perspective on the events described by the two kinds of religions. As Metzger points out (1955: 20),

… the Mysteries differ from Christianity’s interpretation of history. The speculative myths of the cults lack entirely that reference to the spiritual and moral meaning of history which is inextricably involved in the experiences and triumph of Jesus Christ.

Regarding the apostolic and subapostolic literature Metzger (1955: 15) asserts,

In all strata of Christian testimony concerning the resurrection of Jesus Christ, “everything is made to turn upon a dated experience with a historical Person,” [citing Nock] whereas nothing in the Mysteries points to any attempt to undergird belief with historical evidence of the god’s resurrection.

To be aware of this is perhaps to notice the different cultural matrixes in which these religions took root; it would be difficult indeed for a cult rooted in Israelite biblical piety to have ignored a heilsgeschichtliche perspective on history. In this perspective, God’s acts might be celebrated annually in cultic ritual, but they were viewed as unique events secured by the testimony of witnesses and grounded in corporate piety.

Nock points out that, while many of Paul’s hearers may have understood him in terms of the Mysteries, most of the early Jewish-Christian missionaries like Paul had probably had little firsthand exposure to the Mysteries and reflected instead a broader milieu of which the Mysteries were only a part (cf. Nock 1964a: 31; Cadbury 1955: 28).

One area of special comparison between the two, especially in early twentieth-century literature, involves the matter of salvation and dying and rising gods. The motif of dying and rising gods certainly predates the time of Jesus. Just as fertility fled the earth during Demeter’s search for Persephone in the Eleusinian myth,

so it flees during the absence of the Hittite deity Telepinus (ANET 126-28), the Canaanite Baal (ANET 129-42; Bright 1981: 118), and perhaps the man Aqhat (ANET 149-55, especially Ginsberg on 155). The same theme appears in the late-second-millennium-B.C. story of Ishtar’s descent to the netherworld (ANET 108 lines 76ff.; cf. reverse 34ff.). It seems likely that a much older story line or lines stand behind all the regional variations.

Descent to the underworld in such texts need not be permanent. In the epic of Gilgamesh (Gilgamesh 6.97-99; ANET 84), Ishtar forces Anu to comply with her demands by threatening to smash the doors of the netherworld and to raise up the dead so that they outnumber the living, and similarly addresses the gatekeeper of that world with the tale of her descent there (ANET 107, Descent lines 12ff.). In a tale perhaps dating to the first half of the second millennium B.C. or earlier, Inanna is put to death (though she is a goddess), but after three days and nights she is restored, as the food and water of life are sprinkled sixty times on her corpse (ANET 52-57, especially 55).

But the parallels remain problematic. While there is widespread pre-Christian evidence for the account of Osiris’s resuscitation (cf. also Plut. Isis 35, Mor. 364F), he is magically revivified, not transformed into an eschatological new creation; his corpse is awakened through the same potencies as exist in procreation, and he remains in the netherworld, still needing protection by vigilant gods and replacement on earth by his heir (Wagner 1967: 119; on Nile water, see 127-35). Adonis’s death was mourned annually (e.g., Plut. Nicias 13.7; Ovid Metam. 10.710-39), but his rising is not documented prior to the middle of the second century A.D. (Wagner 1967: 171-207, especially 195).

Attis, too, was mourned as dead, but there is no possible evidence for his resurrection before the third century A.D., and aside from the testimony of the Christian writer Firmicus Maternus, no clear evidence exists before the sixth century A.D. (Wagner 1967: 219, 229; for the typical story, see Vermaseren 1977: 91).

Dionysus’s return from death (cf. Otto 1965: 79-80, 103-19) is clear enough, but perhaps in the same category as Heracles’ apotheosis or the wounding of Ares in the Iliad; mortals could be deified and deities could suffer harm.

Even Persephone was taken down to the underworld alive, and Orpheus descended alive to rescue his beloved Eurydice. Frazer’s scheme of the “dying and rising god” has thus come under heavy criticism in recent times (see documentation in Gasparro 1985: 30n.16).

Many Christian writers have asserted, again perhaps through the grid of their own religious understanding, that the Mysteries must have provided salvation through union with dying-and-rising gods (e.g., Conzelmann 1969: 11; cf. Case 1975: 111; Bultmann 1956: 158-59; Ridderbos 1975: 22-29). While there may be some truth in the idea that ancients accepted that a god not subject to death could grant immortality, Burkert cautions that “This multiplicity of images can hardly be reduced to a one-dimensional hypothesis, one ritual with one dogmatic meaning: death and rebirth of ‘the’ god and the initiand” (1987: 100). Much of the evidence is late (e.g., Apuleius, whom Dunand 1975: 58 interprets thus) and/or specifically Christian (e.g., Firmicus Maternus De Errore Profanarum Religionum 22 in Grant 1953: 146). More recent writers are therefore generally more cautious about connecting spiritual salvation (when it appears in the Mysteries) with the dying deity motif (e.g., Davies 1980: 91).

In the Eleusinian rites, the mystēs received the promise of a happy afterlife, but by being pledged to the goddess rather than by being reborn or by dying and rising with the deity (Wagner 1967: 87). The cult of Cybele also does not support the common conclusion, as Gasparro notes (Gasparro 1985: 82). The main problem with the view that many members of the old Religionsgeschichte school, eager to produce “parallels” to primitive Christianity, adduced is that most of the people who turned to the Mysteries already believed in some afterlife in the netherworld anyway; it was merely a happier afterlife in that world that the gods could guarantee.

Those like Bousset who drew such connections (1970: 57) did not take adequate account of the vegetative, cyclical, and seasonal nature of most of the resuscitation rituals (for the vegetative association see, e.g., Gasparro 1985: 29, 43-49; Ruck 1978: 44-45; Guthrie 1966: 55-56). This is a far cry from the earliest Christian picture of Christ’s bodily resurrection rooted in explicit Jewish eschatological hopes — a perspective on the resurrection that Paul affirms is guaranteed by hundreds of eyewitnesses, including himself, and argues, despite his hellenistic audience, is a necessary understanding of resurrection for a true follower of Jesus (1 Cor 15). One would not think that earlier Palestinian Christianity held a less rigorously Jewish perspective than Paul did (cf. Metzger 1955: 19-20; Ring 1944: 228).

While the third day is used for resurrection in the later ritual for Attis and perhaps for Adonis, these may be based on Christian precedents. The third day in the cult of Osiris is most significant, but the traditional Jewish view about the corpse, the use of a “third day” for an interval between two events in close succession in the Hebrew Bible, and the inherent likelihood of some coincidence between a brief period in early Christian tradition and one in the Mysteries qualify its significance considerably. The fixing of the third day in the pre-Pauline formula in 1 Corinthians 15:3, however, weights the case in favor of a Palestinian Jewish-Christian tradition for Jesus’ resurrection prior to any exposure to the cult of Osiris in the hellenistic world (Metzger 1955: 18-19). And while gods could often die in the Mysteries, their deaths were not portrayed as triumphant or meaningful as in many strands of early Christian tradition.

Because Greek religion in general, like many religions in the world,

addressed the survival of the soul after death (e.g., Herod. Hist. 2.123; Plato Phaedo 64CD; 80DE), it should not surprise us that the Eleusis cult promised a happy life in the underworld (Burkert 1987: 21; Grant 1962: 11-12; Mylonas 1961: 268-69; Wagner 1967: 87), that Isis promised patronage and protection (Wagner 1967: 112), and that the Dionysiac Mysteries may have indicated a happy afterlife (Burkert 1985: 293-95; idem 1987: 21-22). But there is little evidence for any future hopes in the cult of Cybele, and certainly not linked with Attis (Gasparro 1985: 84-106, 125; Wagner 1967: 255-56). When the early Christian picture of bodily resurrection derives directly from Jewish eschatological teaching, one casts the net rather widely to make all human hopes for afterlife parallel to it.

The arguments for deification in the Mysteries (cf. Angus 1925: 108-9; Reitzenstein 1978: 200; Avi-Yonah 1978: 42; Tarn 1974: 354-55) would not add much to the idea in hellenistic religion in general, given the apotheosis of famous philosophers (e.g., Cic. De Opt. Gen. Orat. 6.17; De Leg. 3.1.1; De Nat. Deor. 2.12.32) and heroes (Cic. Tusc. Disp. 1.12.28; 2.7.17; De Nat. Deor. 2.29.62; 3.15.39; Herod. Hist. 1.65-66) or the divinization of good men in philosophy (e.g., Plato Rep. 10.611DE; Philost. V.A. 3.18, 29; 8.5; Plotinus 1.2.7, On Virtue; Epict. Disc. 1.intro.12; 1.9.6-11, 22; 1.14.6; 1.17.27; 2.8.10-11; 2.19.26-28; Cic. Par. Stoic. 14; De Re Publ. 6.24.26; De Leg. 2.9.22; 2.22.55; Tusc. Disp. 1.24.56–26.65; De Divin. 1.37.80; Sen. Dial. 1.1.5; Ep. Lucil. 32.11; 48.11; Marc. Aur. Med. 2.13, 17; 3.5-6, 12, 16.2; 4.16; 5.10.2; 5.27; 12.26). But while the doctrine comes to figure highly in Gnostic and Hermetic systems (cf., e.g., Ménard 1980: 149; Jonas 1963: 44-45), it is lacking in early Judaism and Christianity (cf. 1 Cor 3:4).

Excursus on Jewish Resurrection Theology

Pagan notions of the afterlife and myths of risen deities did provide Gentiles a handle for apprehending aspects of early Christian teaching about the resurrection (cf. Lewis 1983: 100), but the Christian teaching remains distinctly Jewish in its origin.

Although it is not the emphasis there (cf. Schuller 1989), the teaching appears in some Old Testament texts (Is 26:19; Dan 12:2; Ferguson 1987: 439) and probably has early antecedents in Israel’s history, though personalized eschatology appears in texts only after the exile (cf. Wifall 1978). Jesus’ Jewish contemporaries regarded that resurrection as collective and future (Collins 1997: 97).

Not all streams of early Judaism clearly articulate a doctrine of bodily resurrection. The Sadducees denied it (Jos. Ant. 18.16-17; War 2.165),

while the evidence we do have from Qumran supports the likelihood that they accepted it, though we lack concrete evidence (see Puech 1993; Sanders 1992: 370; cf. Ulrichsen 1977). Clearly the Pharisees and their probable successors in the rabbinic movement (e.g., Stemberger 1973; in the Targums, see McNamara 1972: 136) affirmed the doctrine of the bodily resurrection, almost equating belief in it with belief in the afterlife. But the doctrine was much more widely held than among the Pharisees, representing common Judaism, as widespread attestation indicates (e.g., Ps. Sol. 3:12; 15:12-13; 1 Enoch 22:13; 2 Bar. 30:1; Ps-Philo 3:10; Test. Abr. 7B; cf. Test. Jud. 25:1-4; Test. Zeb. 10:2; Apoc. Ezek. intro.), and as the widespread use of Daniel (especially in the LXX) would almost require (Dan 12:2). The Second Benediction of the Amida undoubtedly obtained beyond Pharisaic circles. The use of ossuaries for secondary burial in the first century likely also supports the widespread character of belief in the bodily resurrection (Rahmani 1978; cf. Goodenough 1953: 1:164-77). Sanders is probably right that nearly everyone but the Sadducees affirmed the doctrine (1985: 237; cf. Wright 1992a: 320-34).

The belief was probably less widely held initially in the Diaspora, though some evidence exists.

Some hellenistic Jewish writers, while accommodating the idea to hellenistic notions of immortality (e.g., Ps-Phocyl. 105) and the language of deification (104), also allude to the doctrine of bodily resurrection (102-4). Perhaps after rabbinic Judaism consolidated its influence, the doctrine of a literal, bodily resurrection also became standard in much of the Diaspora (see Garte 1973).

All our early Christian sources unanimously affirm the doctrine of the bodily resurrection of Jesus,

although 1 Corinthians 15 attests that Paul had to deal with Gentiles who could assimilate the Palestinian Jewish doctrine only with difficulty and did not wish to accept it beyond the case of Jesus. Within earliest Christianity, however, there remains no debate about the received tradition that Jesus himself rose bodily, unless one is inclined to count inferences by some modern scholars without explicit supporting evidence. By some point in the second century, however, Gnostics and others who found the notion of a bodily resurrection of any sort incompatible with Platonic metaphysics sought to interpret the early Christian tradition differently (cf., e.g., in Iren. Haer. 2.29).

Ancients commonly reported apparitions of deceased persons (e.g., Apul. Metam. 8.8; 9.31; ARN 40A; see comment on Mt 1:20) or deities, and hence occasionally those of persons who had become immortal (e.g., Plutarch’s reports of Romulus more than half a millennium earlier, in Talbert 1977: 41),

but these are not resurrection appearances. Even the appearance of Apollonius of Tyana, which exhibits some parallels with the Gospel accounts (Philost. V.A. 8.31; Sanders 1985: 320), is not an exception. This story appears in a third-century source — after Christian teaching on the resurrection had become widely disseminated; further, Apollonius proves that he has not died, not that he has risen (Blackburn 1986: 193). Most cultures believe in some form of life after death, and such cultures frequently accept some form of contact with the spirits of the dead or some of the dead. Such phenomena may help explain how ancient Mediterranean hearers may have conceived of Jesus’ resurrection appearances; but to cite them as “parallels” to those appearances, as if they define the latter, stretches the category of parallel too far to be useful. If Jesus rose again, how would the disciples know it and proclaim it if he failed to appear to them?

Ancients could depict the soul as rising to heaven (e.g., Test. Abr. 20A; 7, 14B), and told stories of newly divinized immortals ascending to heaven (e.g., Diog. Laert. 8.2.68; cf. Talbert 1975), as well as traditions about Enoch, Elijah, Ezra, and others thought to have escaped death (e.g., 1 Macc 2:58; 1 Enoch 39:3; 2 Enoch 67:1-3; Gk. Ezra 5:7; Palatty 1986), and on a more regular basis angels (e.g., Tob. 12:20-22; Jos. and Asen. 17:8, MSS; Test. Abr. 4, 8, 9-10, 15, 20A; 4, 7-8, 10B; cf. Jub. 32:20-21). Some early Christian traditions (not Matthew) recounted as a climactic resurrection appearance Jesus’ spatial remove from his disciples in these terms (Lk 24:51; Acts 1:9-11; cf. Jn 20:17; Mk 16:19).

But the difference again concerns the resurrection. To most ancient Mediterranean peoples, the concept of corporal resurrection was barely intelligible; to Jewish people, it was strictly eschatological. Yet once one grants the possibility of a bodily resurrection of Jesus within past history, the appearances follow naturally with or without parallels. In a Jewish framework, Jesus’ resurrection within history must also signify the arrival of the eschatological era in some sense (e.g., Acts 1:3-6; “from among the dead ones” — Rom 1:4; 1 Cor 15:20; Heb 6:5).

The Report of the Guards (28:11-15)

See comment on the guards on 27:55-66, especially 27:62-66. Just as Josephus’s response to the anti-Jewish polemic of Apion has inadvertently preserved the basic outline of anti-Jewish polemic in his day, Matthew’s response to arguments against the early Christian claims about the resurrection preserves what must have been the basic charge of his day: the disciples stole the body (cf. later sources in Stauffer 1960: 144-45; Tert. Apol. 21).

Matthew would hardly invent anti-Christian polemic of this nature, but he would respond to what was already in circulation. Scholars therefore often concur that this story was circulating among his contemporaries (e.g., Overman 1996: 401).

Indirectly this charge testifies that the opponents of Christianity conceded that the body of Jesus was missing and that no simpler explanation (such as the body’s getting lost by deposition in the wrong tomb) was available (also Craig 1984; Meier 1980: 356). Although Paul does not appeal to the empty tomb tradition in 1 Corinthians 15, his account necessarily implies it. Many people in antiquity claimed to see “ghosts,” but for Palestinian Jews “resurrection” meant bodily resurrection and nothing else.

Against some commentators, it is quite difficult to imagine that the disciples would have begun proclaiming the resurrection and the authorities begun opposing them without someone checking the tomb (Craig 1995: 151). Yet the church depended on the testimony of witnesses of the risen Christ, not simply on an empty tomb (Ladd 1974b: 325); the empty tomb tells us about the nature of the resurrection (and of the body and history), but the witnesses attest to its facticity.

Yet the report of the guards is not credible. Graves, especially with expensive contents, were often robbed eventually (Arrian Alex. 6.29.4); stones were rolled away so that graves could be robbed (Char. Chaer. 3.3.1), but not with guards posted (at least not without subduing the guards, normally fatally). Further, whereas tomb robbers normally carried off wealth, carrying off the body was so rare that it would shock those who heard of it (Char. Chaer. 3.3). If the disciples did not protect Jesus while he was alive, surely they would not have risked their lives to rob his tomb after his death (grave robbing was not only impious — e.g., Plut. Mor. 173B — but a capital offense — e.g., SEG 8.13 in Sherk 1988: 52, §27); nor could they have rolled away the massive stone without waking the guards.

Penalties for falling asleep on guard duty could be severe, and guards who claimed to have slept through the stealing of the body, yet suffered no harm, would sound very suspicious. (Thus, for example, a soldier assigned to guard corpses hanging on crosses to prevent burial found the body stolen, and preferred suicide to court martial and execution — Petron. Sat. 112.) Under normal circumstances, people might suppose that they and those who failed to punish them had collaborated in the disappearance of the body, but in this situation those who failed to punish them had too much to lose for anyone to suppose that.

One might argue that someone took the body, but that guards were not actually present. But then why would the ruling establishment circulate a rumor (to which Matthew must respond) that guards were present? The testimony of guards who slept through the theft would be no more credible than the guesses of investigators after a theft. The story makes the most sense if guards had been present, somehow failed to protect the body, and the officials had to strike a deal to cover their embarrassment.

The narrative’s irony announces both God’s power and human weakness. Guards who saw an angel (rather obviously not a disciple) were ready, like Judas (26:15), to betray the truth for money (28:12); like Peter (26:69-75), they were ready to deny the unbelievable to protect their lives (28:14).

At the same time, the guards pretend to have slept through the Messiah’s deliverance (28:15), whereas when Jesus needed them the most, his disciples had slept through his time of testing (26:40-45). Disciples and enemies alike proved weak, but Jesus’ resurrection was an act of God’s power.

The Report of the Church (28:16-20)

Cf. Lk 24:47-49, Acts 1:8, Jn 20:21, and Mk 16:14-18 in Jerusalem; Jn 21:1-23 in Galilee.

In this closing pericope, Matthew recapitulates and develops the most important themes of his Gospel (Bauer 1988: 109-28): because Jesus is the rightful king, his followers must make disciples for him among the nations.

The “mountain” (28:16) recalls the other sites of revelation in the Gospel (5:1; 17:1), and various other features indicate that Matthew has heavily reworked the tradition behind the closing scene of his Gospel to conclude with an important theological message. Some doubt that the commission in the Gospels stems from the risen Jesus (cf., e.g., Bultmann 1968: 289; Bornkamm 1969: 15; S. Brown 1980), but others argue for the authenticity of this passage (e.g., White 1960: 338-45; Beasley-Murray 1962: 77-92), although recognizing that Matthew has heavily redacted his prior tradition to emphasize the focal themes of his Gospel (see Meier 1977b; Brooks 1981; cf. Harrison 1973; Hubbard 1974; Davies and Allison 1992). (That Luke himself is free to vary the wording in his two accounts illustrates the degree of freedom permitted on such matters — Lk 24:47-49; Acts 1:8.) Certainly the Lukan, Johannine, and Matthean commissions, as well as the missionary spirit of Pauline and Markan Christianity, leave little evidence of a Christianity without missionary impetus; that Christianity’s unanimous urge for missions derives from its founder seems inherently likely. And the different forms of the commission suggest its essential accuracy by multiple attestation (cf. Guillet 1982, though perhaps overplaying divergences).

The women offered a true report (28:1-10) and the guards a false one (28:11-15); Matthew’s closing paragraph announces that disciples of both genders in his own community, like the first women disciples, must offer a true report and resist temptations like money and protection to which the guards succumbed.

First, the narrative teaches about faith and unbelief (28:16-17). Matthew does not report the resurrection appearances in Jerusalem (Lk 24:36-49; Jn 20:19-29; probably 1 Cor 15:2, 5, 7); within the logic of his narrative, the disciples believe the women enough to proceed to Galilee without them (although his audience is probably so familiar with the resurrection tradition that they know other events intervened). The resurrection experiences came especially to those who had followed Jesus before (Acts 10:41), but also to some who had been skeptics (1 Cor 15:8-9; Gal 1:14-16; cf. also James in 1 Cor 15:7). Perhaps also due to space limitations, Matthew omits the inability of some of Jesus’ followers to recognize him initially (Lk 24:16; Jn 20:14-15; cf. Tob 5:4-6, 12; 12:19; Ps-Philo 12:1; 61:9; 64:4).

Those who see him (cf. 1 Cor 15:5, 7) worship him (cf. 28:9), which suggests that they recognize him for who he is — the “God with them” (1:23; 28:18-20). Some of them, however, despite seeing him, still doubt (cf. Lk 24:40), although in this context the expression may connote troubled indecision — “littleness of faith” (8:26) — rather than opposition and unbelief (Garland 1993: 266).

Although Matthew stresses this theme less than Mark, he agrees with Mark that disciples often are foolishly unbelieving (6:30; 14:31; 17:20), even after the resurrection. If even seeing is not necessarily believing, disciples ought not to wait to see before they will believe, as if God has not provided enough evidence already.

Second, the narrative teaches about Jesus’ identity (28:18-20). As Daniel 7’s Son of Man, Jesus holds “all authority” (28:18; cf. 7:29; 9:6; Jn 17:2; Dan 7:13-14).

One may contrast here Satan’s offer in 4:8-9; by pursuing obedience Jesus received more than Satan offered (France 1985: 413). Jewish teachers felt that confessing the one Lord by means of the Shema expressed submission to God’s royal authority (m. Ber. 2:5); this passage indicates that such submission requires confession of Jesus (cf. 10:32). Making disciples for Jesus alone and not for themselves (23:8-10) also portrays Jesus as more than a normal rabbi (28:19).

Disciples baptize not only in the name of the Father and the Holy Spirit, whom biblical and Jewish tradition regarded as divine,

but also in the name of the Son. Placing Jesus on the same level as the Father and Spirit makes even more explicit what is implicit in Acts’s “baptism in Jesus’ name” (Acts 2:38; 8:16; 10:48; 19:5; cf. 22:16) — that is, that Jesus is divine (28:19). While many doubt the authenticity of this baptismal formula (e.g., Bultmann 1985: 252; its specific form could be a Matthean or traditional construct summarizing what they felt was Jesus’ idea), it certainly climaxes Matthew’s emphasis on Jesus’ deity (1:23) and authority (28:18).

Still, even the formula could go back to prior Christian tradition or to Jesus himself (cf. Did. 7:1-3).

Albright and Mann (1971: 362) point out that the trinitarian formula is established by the period of our first extant Christian documents (1 Cor 12:4-6; 2 Cor 13:14; see especially Fee 1994: 839-42) and widespread in the church (1 Pet 1:2; 1 Jn 3:23-24; Did. 7:1, 3; Odes Sol. 23:22); Mark already pairs “Father” and “Son,” as do Q (Mt 16:27) and Johannine tradition; and “For all we know, such a saying may have stood in the now-lost ending of Mark.” The formula distinguishes Christian from other forms of Jewish proselyte baptism (see comments on 3:5-9).

One other aspect of this pericope emphasizes Matthew’s high Christology. Jesus’ continuing presence with his followers even after his departure (28:20) suggests his omnipresence — an attribute limited to deity alone (see comment on 1:23; 18:20).

Finally, the narrative teaches the Matthean community about its mission (28:18-20).

Because Jesus has all authority, because he is king in the kingdom of God, his disciples must carry on the mission of teaching the kingdom (10:7). Jesus’ instructions include an imperative surrounded by three participial clauses: one should make disciples for Jesus by going, baptizing, and “teaching.” Conjoined with “discipling,” the “teaching” involves instruction no less serious than Jewish sages customarily provided their students. In the full context of Matthew, it includes making the kind of disciples Jesus made, to carry on his mission of proclaiming and demonstrating his kingdom or rulership.

Jesus’ “authority” (28:18) includes authority to tell his subordinates to “go” (cf. 8:9); the probably independent attestation of the phrase in Mark 16:15 suggests that this was part of Jesus’ final commission. Making disciples involves “going” (28:19), as it had before (10:7). Because “going” is a participle, we could read, “as you go,” essentially, “on your way,” implying that one need not cross cultural boundaries to fulfill this commission (Culver 1968).

But this suggestion misses the parallel between the final commission and the model mission in chapter 10: even while remaining within Galilee, the disciples had to proclaim the kingdom to those who had not yet heard the message (10:7; cf. Mk 1:38). Further, the participle is aorist: perhaps “having gone,” the going being presupposed in the commission; the aorist participle “going” may represent part of the command, the aorist imperative “make disciples,” whereas the two present participles explain how to make disciples (Rogers 1973; cf. White 1960: 127n.3). But this does not require us to subordinate “going” excessively since Matthew often uses this participle in a sense coordinate with the main verb (cf. 2:8; 11:4; 17:27; 28:7; Blomberg 1992a: 431). Thus “going” may be one of the three components here of “making disciples” for Jesus, one that helps explain discipling the nations.

Greek tradition could praise those who made many disciples (e.g., Diog. Laert. 8.1.16); Greek philosophers thought in terms of “conversion” to philosophy (see Nock 1933a) and various cults were propagated by travelers in antiquity (Stambaugh and Balch 1986: 42; for Christians, cf. Acts 8:4; Liefeld 1967: 151; Malherbe 1983: 65). Judaism also spoke of sages with disciples (cf. comment on 4:19; 19:21-22), spoke of discipleship to earlier stages in transmission of the Torah like Moses or Ezra (e.g., Philo Spec. Leg. 1.345; 2.88; b. Sanh. 11a, bar.; Soṭa 48b; Song Rab. 8:9, §3), emphasized sages raising up many disciples (m. ʾAbot 1:1), and sometimes even spoke of persuading large numbers of people to become more peripheral students of Torah (e.g., ARN 26, §54; Qoh. Rab. 11:6, §1); they also separately recognized the conversion of Gentiles (cf. on 23:15; see De Ridder 1971a; idem 1971b). But ancient hearers would, and modern hearers should, recognize a drastic innovation in a command to disciple “nations.” To be sure, the discipling of nations is carried out through baptizing and teaching individuals in those nations; although exceptions to grammatical consistency in antiquity abound, it is probably significant that the object “them” attaching to baptizing and teaching (28:19-20) is masculine (autous) rather than neuter (auta), although “nations” (ta ethnē) is neuter. Nevertheless, the stark command to disciple “nations” implies more than producing disciples for any ancient teacher would, and in contrast to other disciplers Jesus’ followers would not disciple others to themselves (23:8).

All “nations” probably signifies “all peoples,” rather than the modern concept of “nation-states” (McGavran and Arn 1977: 38; for people-groups see rightly Piper 1993: 169-81). Also far from abandoning the mission to his own people, Matthew’s commission represents “peoples” and not simply “Gentiles” (De Kruijf 1993; Saldarini 1994: 59-60, 78-81; cf. Meier 1977a). When Matthew adds “all nations” to a Markan logion (Matt 24:9), he clearly includes Jewish participation (10:17-18; Stanton 1993: 158). Those who prefer to translate “Gentiles,” however (e.g., Hre Kio 1990; Luz 1992), are correct that Gentile peoples are where the emphasis lies, because this statement extends the commission in chapter 10, which was explicitly limited to Jews (see Levine 1988: 191).

On either translation, therefore, what is important to remember is that the Gentile mission extends the Jewish mission — not replaces it; Jesus nowhere revokes the mission to Israel (10:6), but merely adds a new mission revoking a previous prohibition (10:5).

Given the context of his whole Gospel it is most reasonable to suppose that Matthew lays the emphasis here on Gentile peoples, whom his predominantly Jewish-Christian community most needs to be encouraged in evangelizing; they were already deeply engaged with the needs of their own ethnic communities. Matthew elsewhere emphasizes the nations (1:3-5; 2:1-2; 3:9; 4:15; 8:5, 11, 28; 11:21-22; 12:41-42; 15:22; 16:13; 27:54), including the importance of their hearing the message of the kingdom before the kingdom and judgment come (24:14; 25:31-32; perhaps Rom 1:5; 16:26 indirectly reflect the tradition Matthew preserves here).

But wherever God leads particular disciples to carry out this commission (“going”), the text is clear on the other ways one makes disciples. First of all, one baptizes them under the rulership of Christ. Baptism was an act of initiation and conversion (see comment on 3:6), and this text suggests that disciples initiate others into the faith. Baptism was part of Christian practice universally and from the first, and it is unlikely that the early Jewish Christians borrowed John’s practice of proselyte-baptizing fellow Jews without some sort of approval from Jesus (cf. Jn 4:1-2); thus a command of Jesus to baptize is historically likely (Taylor 1997: 299).

But once they are initiated, mature disciples must also build the new disciples into stronger discipleship by teaching them Jesus’ message. The summaries of Jesus’ teachings earlier in Matthew’s Gospel (chs. 5–7; 10; 13; 18; 23–25) work well as a discipling manual for young believers. Various passages in fact epitomize Jesus’ commandments (7:12; 12:7; 22:37-40; 23:23), but the community still requires specific articulation of the older commandments in light of the kingdom (5:17-19; 13:52). Few Jewish teachers would admit proselytes without basic instruction in the demands of Judaism; Jesus’ command to teach his message is at least implicit in sending his followers, and was probably explicit as well (see De Ridder 1971a: 108); here, as in Jewish instruction in Judaism, the process of teaching continues subsequent to the initiation.

The Gospel closes with a promise. As his disciples carry out the great commission, Jesus will be with them to the end of the age (28:20).

The text probably specifies the end of the age because at that time the Son of Man would return in his kingdom — after the nations had heard the good news of the kingdom (24:14), and hence been prepared for the judgment (25:32-36). Until his coming, the disciples would thus need to exemplify Jesus’ reign on earth as it is in heaven (cf. 6:10). If many Christians today have lost a sense of Jesus’ presence and purpose among them, it may be because they have lost sight of the mission their Lord has given them. Those who would be his disciples must prepare the way for his future coming and his kingdom, as John the Baptist did for his first (3:1-3).

. Although the Passion Narrative is critical to Matthew, his Gospel is not simply “a Passion Narrative with an extended introduction,” as some characterize Mark; the speech sections are too significant (Dahl 1995: 64-65).

. A stock arsenal of motifs may have been applied when expanding stories for dramatic purposes; one may thus compare the mother in Maccabean accounts with the Spartan mother Argileonis in Plut. Sayings of Spartan Women, Mor. 240C. Cf. Robbins 1992: 185, following Nickelsburg 1980: 156 on the tradition of a righteous sufferer vindicated by God, but note that while similar story lines illustrate the nuances with which an ancient audience would have heard the story, they do not demonstrate dependence or genetic relationship. Those who take moral (or immoral) stands against the status quo often invite repression, and did so even more often in antiquity; and presupposing that no deity would ever vindicate those aligned with that deity’s cause is no more historically “neutral” than presupposing the reverse. The tradition places Jesus especially within the rejected prophet tradition (cf. Robbins 1992: 186).

. Boring 1995: 156 lists contrasts with the Maccabean martyr accounts: the Gospels avoid sensationalistic details, interpretive speeches by Jesus, a Stoic lesson contrasting reason with emotions (Plut. Whether Vice Is Sufficient to Cause Unhappiness 2; 4 Macc 8:15; though this feature says more about the social context of the Maccabean audience than about any larger genre per se), and “vengeful threats.”

. E.g., though Greek thought readily supplied intelligibility to an atoning martyr tradition (cf. Robbins 1992: 187 following Williams 1975: 137-254), early Jewish Christians could have drawn on the Isaian servant songs (see comment on 20:28). The concept of atonement in general appears in the Hebrew Bible and is widespread in apparently unrelated cultures, and the ancient Mediterranean champion tradition appears in the Hebrew Bible (1 Sam 17; 2 Sam 2:14-16; cf. Yamauchi 1972: 64-65; Gordon 1965b: 262) just as e.g., with Aias and Hector in the Iliad bk. 8, or with Menelaos and Alexandros (Paris) in bk. 3. On the diversity of Jewish martyr stories, see van Henten 1985.

. While the rest of the Gospels foreshadow this climax, that is also the case in some contemporary biographies (Burridge 1992: 199).

. Dibelius 1971: 178-217 thinks that “the Passion story is the only piece of Gospel tradition which in early times gave events in their larger connection.”

. Thus Jewish scholars with no faith commitment to the narratives may also suggest that other Gospels draw on pre-Markan Passion material (e.g., Flusser 1988: 575-87, though his particular reconstruction may presuppose Lukan priority).

. Whether in following a source incompletely or for other reasons, writers occasionally cite as familiar a name not previously mentioned (e.g., Livy 40.55.2); but it is not the usual practice.

. Some view the fleeing young man of Mk 14:51-52 only in terms of his symbolic significance in the narrative (Crossan 1976: 147-48; Fleddermann 1979; Kelber 1979: 77), but Theissen is probably right to find genuine tradition from the early Palestinian church here (1991: 186; cf. Dibelius 1971: 182-83; Stauffer 1960: 121).

. Some of Theissen’s other arguments (1991: 189-97) are weaker. Pesch 1991 argues at length for a pre-Markan passion narrative, especially the very early last supper material, but Hengel 1991: 209-10 may be correct that Pesch is overly optimistic about his ability to reconstruct sources.

. Brown 1994: 56 suspects that Mark may have acquired some of his style from frequent recitation of the Passion Narrative, the way some modern evangelists acquire their style from the KJV; further, Mark may have rephrased the narrative in his own words, especially where his sources were oral. Brown 1994: 554 also emphatically challenges some earlier redaction-critical studies on the trial narrative in Mk 14:55-64 (perhaps Donahue 1976), complaining that though “Mark used earlier material … our best methods do not give us the ability to isolate confidently that material in its exact wording, assigning pre-Markan verses and half-verses from the existing, thoroughly Markan account” (emphasis his).

. On the opposition Jeremiah faced for his “unpatriotic” prophecies, cf., e.g., Jer 26:6-24; Jos. Ant. 10.89-90; angry crowds could also vent their rage on anyone they felt brought them misfortune (Jos. Life 149). A man inside Tyre likewise reportedly prophesied its judgment and faced the charge of being a traitor (Diod. Sic. 17.41.7-8, which may be legendary or repeat Alexander’s propaganda). A Cynic might be flogged for speaking out and apparently ignore it (Diog. Laert. 6.5.91).

. Kelber 1976a: 173 rightly notes that women seeking to anoint Jesus frames Mark’s Passion section (Mk 14:3-9; 16:1-8); although Matthew cuts the second anointing, he preserves the frame involving women’s devotion.

. For how positively this courage would sound to first-century hearers (and how this narrative would thus function epideictically), see comment on 16:21.

. The Romans probably used public religious offices for politics even more so; cf. Taylor 1966: 90-95.

. Rome held the high priests responsible for controlling the masses (Jos. War 2.232-44), and the high priests moved against potential threats to public order when they thought it best to do so (Jos. War 6.300-309; Ant. 20.199-203; Sanders 1993: 266-67).

. Perhaps in part because I find myself more skeptical that religion often changes human nature, I am less sympathetic to their piety than is Sanders 1993: 336. They probably acted in their own self-interest, as well as for the peace, in relations with the Romans (Horsley 1986b). The charges may be stylized, sectarian polemic, as Sanders suggests (and against the priesthood in general he may be right — 1992: 182-89), but one should not dismiss too readily the reasons for the polemic, which appear widely (cf. 1QpHab 9.4-5; Test. Levi 14:1; 2 Bar. 10:18; t. Minhot 13.21 in Avigad 1980: 130; idem 1983: 71; Hengel 1974b: 23); corrupt priesthoods were common targets of polemic in the ancient Near East through the first century (Crocker 1990). Cf., e.g., the servants of the later Ananias who beat poorer priests to seize their tithes (Jos. Ant. 20.181, 206).

. Manson 1961: 16 derives “Sadducees” from a cognate of “Sanhedrin,” but the Zadokite explanation seems more probable (with Lohse 1978: 74 and others).

. Medieval tradition sometimes links them with Karaites (Jacobs 1973: 10), perhaps for polemical purposes against the Karaites; others (aside from those scholars who once dated the Scrolls to the medieval period) have found for the Karaites predecessors in the Essenes (early after the discovery of the Scrolls; cf., e.g., Ginsberg 1953: 81; Kahle 1953; Wieder 1956-57; Fritsch 1956: 86-89).

. Pace Eppstein 1966, even had Pharisaism exercised adequate control to effect excommunication, the Sadducees could not have been excommunicated almost a decade before the temple’s destruction. Worse still, Jeremias 1969: 264-65 takes later rabbinic sources at such face value that he supposes that Pharisaic (proto-rabbinic) authority had stripped the Sadducees of power (cf. Finkelstein 1962: 2:659).

. “Sanhedrin” is a broad rather than restrictive term, applicable also in Greek texts to an informal assembly of advisors (Diod. Sic. 13.111.1) or, frequently, to Rome’s “senate” (e.g., Diod. Sic. 40.1.1; Dion. Hal. 5.70.5; 6.30.2; 6.81.1; 6.85.2; 8.69.2; 9.32.5; 10.2.6; 12.1.14; 12.6.2 [4]; in these texts it appears interchangeably with boulē, a more common term, e.g., in Dion. Hal. 5.71.1; 6.1.1; 6.21.1; 6.81.4).

. Officials could also assemble their own administrative “councils” from among their friends (e.g., Jos. Life 368).

. Overman 1996: 372-73, 385 regards the Sanhedrin as a Roman political institution, although conceding that “some of the local Jewish elite may have been involved.” Yet in cities like Jerusalem Rome ruled through municipal aristocracies — here, pro-Roman Jewish aristocrats.

. M. Sanh. 1:6; cf. Josephus’s Galilean council of A.D. 70 in War 2.570; Life 79, and that of the Zealots in War 4.336, both undoubtedly following the standard contemporary model; the models probably derive ultimately from Mosaic tradition (Exod 24:9; Num 11:16, 24; cf. Ezek 8:11). Josephus also assumed a council of seven judges as a lower court in every city (War 2.571; Ant. 4.214). An odd number to break a tie made sense; as in Roman law (Dion. Hal. 7.64.6), a tie vote would yield acquittal.

. Brown 1994: 348-49 doubts that an exact list of 71 members existed in the first century, suggesting that it merely included elders from distinguished families alongside chief priests, representatives of whom were expected to appear.

. For bibliography on the Sanhedrin, see Safrai 1974/76e: 418 (the section on the Sanhedrin is on 379-400). Josephus generally prefers the term “sanhedrin,” “assembly,” in Antiquities, and boulē, “council,” in War. Later rabbis believed that God supported the decrees of the rabbinic beth din hagadol, great assembly (Ex. Rab. 15:20), on which Israel rightly depended (Song Rab. 7:3, §1; Lam. Rab. 2:4, §8).

. Brown 1994: 58 suggests that Matthew as a former scribe (Mt 13:51-52) may play down scribal opposition, hence in contrast to Mark has chief priests 15 times, elders 7 times, but scribes only 2 times here. This is improbable, however, given his excoriation of scribes with Pharisees in 23:13-36; Matthew may instead simply wish to emphasize that the priestly aristocracy largely engineered the execution, a conclusion our other data for the most part support. Some of the “scribes” may have been Pharisees, but the Pharisees, though represented (Acts 23:6-9), probably were not dominant in the Sanhedrin (Brown 1994: 350-52), despite Josephus’s possible favoritism toward them (Jos. Ant. 18.15, 17; cf. Life 1, 12 and Ant. passim; Brown 1994: 353-56).

. Jn 11:48 probably reflects earlier tradition (see Winter 1961: 37; Vermes 1973: 50).

. Cf. Mek. Pisḥa 14 (Laut. 1:115) on Ex 12:42 (one of several views) in Lane 1974a: 500; traditions in the Passover haggadah in Black 1967: 237; interpretation of the Hallel in t. Ber. 1:10-11; Pesiq. Rab Kah. 27:5; Lev. Rab. 30:5; Pesiq. R. 2:1. Bokser 1985 argues that redemption is more central in early (m. Pesaḥ. 10) than in later texts.

. Some, relying more on (though lowering — cf. Jos. War 6.420) Josephus’s estimates, propose a population of over 200,000 for metropolitan Jerusalem (Byatt 1973); but obviously no one, including Josephus, was counting. Based on water supply, estimates settle between 40,000 and 80,000 (Wilkinson 1974a; Broshi 1975; idem 1978:14; Stambaugh and Balch 1986: 97); modern sociological studies may suggest no more than 35,000 (Rohrbaugh 1991: 133; cf. Jeremias 1969: 77-82, estimating 25,000). During Passover perhaps 300,000 to 500,000 people were present (Sanders 1993: 125-27); Josephus’s estimates (Jos. War 6.423-25) are too high and rabbinic estimates (e.g., t. Pisha 4:15; Lam. Rab. 1:1, §2) are patent exaggerations, but pilgrimage cities today demonstrate how many people can be packed into a city during festivals (e.g., Jochim 1986: 152).

. The supposed relevance of “prodigies-at-meals” to the anointing (Boring 1995: 146 cites Suet. Vesp. 4) falters both because prodigies-at-meals did not constitute a recognizable category of prodigies and because the anointing, while unusual, remains within the bounds of the natural order.

. For hints of the high cost sometimes expended even on normal unguents and perfumes, see Lachs 1987: 401; for ointments used in burials, see, e.g., m. Shab. 23:5 and below on Jesus’ burial.

. If the phrase usually translated “in memory of her” represents a subjective genitive, it could suggest “her memorial” for Jesus (Greenlee 1960), although this is not the most obvious way to take the phrase. Closer is the Jewish expression “of blessed memory” or “May their memory be blessed” (Sir 45:1; 46:11; cf. Gaius Inst. 1.53) and the formula echoed in rabbinic texts — “May soand-so be remembered [before God] for good” (Lachs 1987: 401 cites b. B. Bat. 21a; Shab. 13b; cf. m. Yoma 3:9; Tan. Ha’azinu 1) — also attested in late inscriptions (Carmon 1973: §§184, 186, pp. 83, 86, 186-87, 192-93; Levine 1982: 8-9).

. One may compare honorific inscriptions, which could enjoin public mourning in a prominent woman’s honor (Horsley 1987: 10-17).

. Origen harmonized accounts by suggesting three separate anointings (Origen Matthew 77; Wiles 1960: 16); but variations likely arose during oral transmission (Dodd 1965: 172). The two stories I propose would be either divergent traditions stemming from one event (e.g., Mack 1988: 200), or a second event in which a second woman probably followed the example of the first (cf. similarly 9:20; 14:36). In view of the likely pre-Markan divergence (except in his programmatic scene at Nazareth, Luke rarely takes such liberties as to rewrite an entire Markan narrative from scratch, and the Johannine account probably confirms the independent antiquity of some of its details), and in view of what most often seems accurate preservation of tradition in the early period (though this pattern would not preclude exceptions transmitted in different circumstances), two distinct anointings eventually conflated in the tradition seem more likely. Following a custom of hospitality, another woman may have followed the example of an earlier woman’s anointing.

. Sanders 1993: 126-27 thinks that Jn 12 may represent a composite of Lk 7 and the accounts in Mt 26/Mk 14, or the traditions associated with them. In any case, he considers it “evident that Jesus attracted women who were not ‘followers’, but who admired him.”

. At least some later rabbis interpreted the passage so literally that they doubted that the poor would even cease in the days of the Messiah (Lachs 1987: 401 cites b. Shab. 63a). Even if this interpretation obtained in Jesus’ day, however, it appears irrelevant to his use of the allusion.

. Cohn 1977: 83 links Judas’s betrayal with the tradition that the priestly aristocracy engaged in “undercover activity,” a tradition that Cohn insists originated only after the lifetime of Jesus. Against Cohn, the initiative here lies with Judas, and even Romans without interest in pursuing crimes were forced to investigate once a delator (accuser) brought activity to their attention (on Christians, see, e.g., Pliny Ep. 10.97).

. Cullmann 1956b: 39 suggests that Judas had materialistic ideas concerning Messiahship into which Jesus’ teachings (especially those concerning vicarious suffering and the new covenant from Isaiah, presented more directly in the next paragraphs) could not fit. (Brandon suggests that Judas sought to betray Jesus’ revolutionary intentions, but this speculation hardly fits the extant evidence about Jesus; more likely, he betrayed the inner circle’s view that Jesus was a “king” — Sanders 1985: 309.)

. Although Gen 37:28 reports that Joseph’s brothers sold him for twenty shekels (the correct average price of slaves in the era in which Joseph lived), most manuscripts of Test. Gad 2:3 have thirty (OTP 815 n. a suggests that this is either from Ex 21:32 or a Christian interpolation; I suspect the latter). Hagner 1995: 761 is impressed that “thirty shekels” reflects a Sumerian idiom for a very small sum (following Reiner 1968); but it is hardly likely that first-century Jews would have any contact with Sumerian idioms. Thirty silver coins was not a small sum; by 100 B.C. a silver drachma was worth about 500 copper ones (Select Papyri 2:xxxiii).

. Despite disagreement on the relation to the festival, most commentators agree that the crucifixion occurred on a Friday (Brown 1994: 1350-51). Even by the third century rabbis were not unanimous about trying and executing someone on a Sabbath (p. Sanh. 4:6, §2).

. Those favoring the Johannine dating respond that, whereas the Synoptics regard the meal as a Passover meal (this is “challenged by no one”), this does not decide the historical question (Bornkamm 1969: 132); but then how do Mark and Paul, writing for largely Gentile audiences, conform the narrative so closely to Passover traditions? And if the Synoptics report the disciples actually keeping the Passover but on a “sectarian” date, would sectarians have observed as many other paschal customs as the text suggests? Jeremias admittedly depends on later traditions, but what evidence we do have fits the Gospel narratives and can hardly have derived from them.

. The calendar of Jubilees may have had some impact on public policy in the second century B.C. (Wirgin 1965: 12-17, 42-43) and has some parallels with later rabbinic calendrical halakah (Grintz 1972: 325), but in contrast to what became mainstream Judaism it preserves the older solar calendar (Morgenstern 1955; cf. Marcus 1956: 12), which is why some even (probably wrongly) label it pre-Hasmonean (Zeitlin 1957-58: 224; cf. idem 1939: 8-16). Opposition to the lunar calendar is implied even in the creation narrative (Jub. 2:9-10; cf. 6:36). This places Jubilees much closer to Qumran thought than to Pharisaism (e.g., Brownlee 1951: 32; Baumgarten 1958a; Grintz 1972: 324); Rivkin 1982 even thinks that Jubilees writes polemically against the Pharisaic calendar.

. Some Diaspora Jews may have ignored Palestinian rules (CPJ 3:27-29, §467). Carson 1984b: 529 also mentions the proposal that Pharisees and Sadducees followed divergent calendars (Strack-Billerbeck) or that the Galileans followed the Pharisaic (and Synoptic) one and Judeans the Sadducean (and Johannine) one. But I suspect that a major difference in observance in the temple would have left more trace in extant first-century sources concerning feasts (sources like Josephus).

. One cannot argue this, however, from the lack of mention of purification or lamb; these elements of Passover ritual would be taken for granted (everyone in the Roman Empire expected animal sacrifices and purifications for festivals), and it would be their omission that would have required comment (Sanders 1993: 251).

. Perhaps because slaves (Gen. Rab. 53:13; 93:6; in Rome, Carcopino 1940: 38-39), especially female slaves, often carried water (Test. Job 21:2-3; cf. Jn 2:7; Gen. Rab. 53:13; 93:6; cf. Eurip. Electra 55, 140; in Rome, Carcopino 1940: 38-39), the disciples could readily recognize the particular man in Mk 14:13. Matthew’s abbreviating technique here also omits the fact that the disciples met in an “upper room” (Mk 14:15), the sort of place scholars sometimes met (cf. Dalman 1929: 114). For so many disciples to have met there would have required that the house was in the well-to-do Upper City (cf. Avigad 1976). People could eat in walled courtyards (Safrai 1974/76b: 730), and sometimes they may have eaten Pesaḥ on rooftops (Safrai 1974/76b: 732).

. Jesus’ “appointed time” here (26:18) is also the fateful “hour” (26:45; cf. Jn 2:4; 7:30; 8:20; 12:23).

. Stauffer probably reads too much into the lack of mention of lamb at the last supper, suggesting that Jesus did not have one and associating the later prohibition of lamb — but not the other paschal food — to apostates (1960: 113); but he also believes (following the Johannine chronology) that they celebrated the Passover a day early (ibid. 115). He suggests that they ate roast kid instead (116-17, referring to m. Pesaḥ. 4:4, which does not directly support his case).

. Although dinners worked well for a symposium format (e.g., Plato Symp.; Plut. Dinner of Seven Wise Men; Table-Talk; Athenaeus Deipn.; Trimalchio’s dinner in Petron. Sat.; symposium section of Ep. Arist.; Lk 14:1-24), the Synoptics do little with the possibility at the last supper.

. At a few months’ remove from the grape vintage, all wine (the meaning of “fruit of the vine” in Jewish blessings — 26:29; cf. p. Ned. 7:1, §6; Whitacre 1992: 867) had some alcohol content, although ancients could not yet artificially increase it beyond natural limits through distillation; nonalcoholic grape juice for the Lord’s Supper arose from nineteenth-century concern for recovering alcoholics. We should note, however, that at most meals ancients watered down wine at least two parts water to every part wine (Plut. Poetry 1, Mor. 15E; Bride 20, Mor. 140F; Table-Talk 1.4.3, Mor. 621CD; Diog. Laert. 7.7.184; Casson 1974: 213; Ruck 1978: 41; in Jewish texts, see Sifra Shir. par. 1.100.1.3; b. ʿAbod. Zar. 30a; Num. Rab. 10:8; Safrai 1974/76b: 748), unless they specifically wished to get drunk (Diog. Laert. 10.1.15; Apul. Metam. 7.12; Ferguson 1987: 80); in the latter case, Greeks may also have added some other intoxicants (Wasson, Hoffmann and Ruck 1978: 89).

. Although the antiquity of the afikomen may be questioned, one should note Daube’s interpretation, followed by Carmichael 1991, of the afikomen (symbolically reuniting the Messiah, a fragment of Israel, with his people).

. Jeremias 1966a: 49-50; Lane 1974a: 497; pace Meier 1980: 318. Less probable is the suggestion (Flusser 1973) that the bread-then-wine sequence (Mt 26:26-27; Mk 14:22-23; 1 Cor 11:23-25) follows Essene versus standard Jewish practice, whereas Lk 22:17-19 and Did. 9–10 follow the more common Jewish sequence (1QapGen 22.14-15 even omits bread and wine in Gen 14); Crossan 1991a: 365 differently finds in the wine-after-supper practice Greco-Roman custom for a formal meal. Contrast Oesterley 1925: 164, who argues that the Gospel sequence implies that the meal preceded the blessings, which is not true of the Pesaḥ meal. But the Gospels simply do not narrate the meal.

. Because the early Christians would hardly have invented the disciples’ failure and certainly not a betrayer — rabbis were esteemed for the loyalty of their followers, not their apostasy! — these traditions, especially that of the betrayal, are surely authentic (cf., e.g., Robbins 1976: 30). The emphasis on Jesus’ foreknowledge of the betrayer (26:20-25; cf. Jn 6:64; 13:18-30) may be counterpolemic against opponents who employed this aspect of the tradition to question Jesus’ prophetic aptitude — underlining the “problem” the betrayer presented to early Christian tradition and how dangerous it would have been for Christians to have invented it.

. E.g., Hom. Il. 21.76; Od. 4.534-35; 11.414-20; 14.404-5; Hesiod W.D. 327; Eur. Cycl. 126-28; Hec. 25-26, 710-20, 850-56; Ap. Rhod. 3.377-80; Livy 25.16.6; 39.51.12. Some have also compared the Qumran custom in which members partook of a meal in sequence by rank (1QS 6.4-5; cf. Jos. War 2.130-31), suggesting that we read “dip with me” temporally: by stretching out his hand simultaneously with Jesus, Judas deliberately indicated his rebellion (Fensham 1965; Albright and Mann 1971: 321; cf. Gundry 1982: 527). While this view is plausible (cf. how one respects others by allowing them to grab first — Sir 31:14-18; for other gestures, see, e.g., 1QS 2.19-23; 1QSa 2.11-17; Jos. War 2.132; Taʿan. 20b), it does leave one to question why no one seems to have noticed such an act where it bore this significance (26:22).

. More relevant would be comparison with teachers like Epicurus who instituted monthly memorial meals in their honor (Diog. Laert. 10.18 in Boring 1995: 149-50), minus the pagan associations with sacrifices for the dead. More distantly, cf. rabbinic phrases used in naming the deceased: “May so-and-so be remembered [before God] for good” (Lachs 1987: 401 cites b. B. Bat. 21a; Shab. 13b; cf. m. Yoma 3:9), probably relevant to 26:13.

. Against this thesis is the antiquity of the resurrection tradition (i.e., disciples were not simply “remembering” a departed loved one), and the closer parallel in the biblical paschal tradition to “commemorating” rather than simply retaining in one’s memory (e.g., F. F. Bruce 1971: 111-12; Gregg 1979; Davies 1980: 252; Martin 1982: 147). The language of the Passover celebration assumed the participation of current generations in the exodus event (e.g., m. Pesaḥ. 10:4-5; t. Pisha 8:18; b. Ber. 14b).

. Minor assimilation begins as early as the Pauline churches (Guthrie 1966: 268; cf. Caird 1955: 96). Paul’s kyriakos, “of the Lord,” may allude to typical Roman usage for “imperial” (Deissman 1978: 357). But early attestation of parallels often comes from early Christian writers who interpreted the Mysteries through the grid of their own experience (cf. Campbell 1968: 323), as an “imitation demoniaque du Christianisme” (Benoit 1975: 79-81), and the rites remained quite distinct (Metzger 1955: 15).

. A small following of Dionysus (Henrichs 1982: 160) and Mithraism (which gained prominence in the Empire after Paul; Nock 1964a: 133; Metzger 1955: 13-15) may have been the exceptions that ultimately affected other Mysteries and the later Christian sacramental view.

. Jeremias 1966a: 28 doubts this, but one may compare echoes of the Kiddush blessing in Did. 9, which far more likely betray Jewish liturgical influence than influence later rabbinic tradition; cf. Jos. Life 279.

. Jeremias 1966a: 50-51 contends that Jewish people drank wine only on special occasions, but he appears incorrect here; even if one merely factors in the regular Sabbath Kiddush meals (e.g., t. Ber. 3:8; Tannaitic tradition in b. B. Qam. 69b; Pesaḥ. 102a, bar.; later, b. Shab. 23b; Taʿan. 24a; Safrai 1974/76b: 747; cf. Jub. 2:21), which he concedes, they drank it quite regularly (cf. blessings over wine in meals in 1QS 6.4-5; b. Ber. 33a; 51a). He is, however, probably correct to emphasize that of the many kinds of wines available (on which cf., e.g., b. ʿAbod. Zar. 30a; Paul 1975), red wine was used on Passover, augmenting Jesus’ symbolism (1966a: 53, 290; t. Pesaḥ. 10:1).

. The generally stricter Shammaite school (regarding Passover, e.g., b. Shab. 18b, bar.) prevailed in his day, but our records were preserved by more lenient, rather than stricter, practitioners. For some source-critical suggestions on the Pesaḥ Haggadah, see Finkelstein 1972: 13-120.

. E.g., m. ʿEd. 8:7; ARN 25A; b. Qidd. 30a; Meg. 19b; Nazir 56b; Pesaḥ. 110b; Shab. 108a, bar.; Qoh. Rab. 1:10, §1; cf. Hillel as a disciple of Ezra in Sanh. 11a, bar.; Soṭa 48b; Song Rab. 8:9, §3.

. Paul may even draw the Last Supper tradition from a larger Passion story already available to him (see Borgen 1986-87), although scholars must examine the evidence for this possibility more thoroughly. The appearance of agrapha (Justin Dial. 35; Syr. Didaskalia 6.5.2 in Jeremias 1964c: 76) in Paul’s context (1 Cor 11:19) may support this thesis.

. The genre of a teacher’s departing instructions may be relevant here, but it is much more relevant to the preceding eschatological discourse (cf. Socrates in Xen. Mem. 4.7.1-10; Robbins 1992: 172 on Mk 13); cf. comment there.

. Some scholars think that the “cup” rather than the “blood” is the earlier form of wording (e.g., Dibelius 1971: 207), whereas others defend the reverse position (e.g., Davies 1980: 244-50). Crossan 1991a: 366 compares “poured out” with the language of libation; a common Mediterranean practice, it appears in the OT (for a metaphorical usage with blood, cf. 2 Sam 23:16-17). For its sacrificial connotations (e.g., Lev 1–7, 16) relevant to Passover, see, e.g., Gundry 1982: 528; Carson 1984b: 537; Hagner 1995: 773. On “the many,” see comment on 20:28.

. Missing in the parallels (including 1 Cor 11:23-25), however, “for remission of sins” may represent Matthean interpretation (Sanders 1985: 112), probably transferred from John’s baptism (Mk 1:4, where Matthew, unlike Luke, omits the phrase; cf. also Meier 1979: 184). Although the LXX expresses the phrase differently, “poured out” may translate the Hebrew in Is 53:12; the phrase surely represents the event of violent death (Hurtado 1983: 223 cites Gen 4:10-11; 9:6; Deut 19:10; 2 Kings 21:16; Ps 106:38; Jer 7:6; Mt 23:35). “For many,” a Semitism that very likely dates to the first decade of Jesus tradition (see Jeremias 1965a: 45), may represent rabbim in this section of Isaiah (Jeremias 1966a: 227), which Jeremias applies not only to Israel but to the nations (1966a: 230-31; cf. Is 52:14-15). “Blood of the covenant” seems to suit first-century Palestinian Aramaic (Emerton 1964).

. Martin 1982: 153 argues this from the Greek and Aramaic, though McNamara 1972: 127-28 is surely right that the words here derive from Ex 24:8 and are possible in Palestinian Aramaic (cf. Gundry 1982: 528; pace Deissmann 1978: 337, who reads the term for covenant here primarily in light of Greco-Roman usage for wills). M. Pesaḥ. 10:6 uses the Passover wine as a metaphor for the blood of the covenant in Ex 24:8 (e.g., Hill 1972: 339; Carson 1984b: 537). And while allusions to Jer 31:31-34 and Is 53:11-12 are probable in this narrative, the Ex 24:8 allusion is the most explicit (Allison 1993b: 257-58).

. For various conceptions of the memorial value of Jewish rituals among various ancient Jewish groups, see Boccaccini 1991: 231-39 (Josephus on 242-45).

. Also most Greeks and Romans: Herod. Hist. 1.123, 129; Mart. Epig. 10.4.1; 11.31.2; Isis aretalogy in Grant 1953: 132; Guthrie 1966: 40. Like a good Skeptic, Sext. Emp. Pyrrh. 3.207 regards abhorrence of cannibalism as culturally conditioned and gives examples of toleration (cf. also Stoics in Diog. Laert. Lives 7.1.121, 188; reports in Herod. Hist. 1.73, 119; 3.99; Otto 1965: 113).

. The tradition of the four cups (e.g., m. Pesaḥ. 10:1; t. Ber. 4:8; b. Pesaḥ. 108a-9a; Gen. Rab. 88:5; Ex. Rab. 6:4; Safrai 1974/76b: 748) may precede the superstitious practice of avoiding drinking an even number of cups, though later teachers harmonized these practices (b. Pesaḥ. 109b). (One could drink more than the four required cups — b. Pesaḥ. 107b-8a; p. Meg. 3:5, §1.) The custom probably borrows or is related to the Greek use of cups to signal stages in banquets (Boring 1995: 147 cites Diod. Sic. 4.3.4; IG 12 3.330, 670).

. By this period people often promised by God to abstain from something for a specified period (cf. Acts 23:12; m. ʾAbot 3:13; conjoined with mourning in Gen. Rab. 92:5). One had to specify the precise parameters of one’s vow, however: one who vowed not to use light, e.g., was not even allowed to use starlight (b. Pesaḥ. 2a, bar.). On kinds of vows, cf. McNamara 1983: 197.

. Lane 1974a: 508 suggests that Jesus abstained from the fourth cup, having participated in and interpreted the third cup that signified redemption in the Passover liturgy (cf. Lachs 1987: 408; also Stauffer 1960: 117, though he places the vow after the first cup — 115). Fredriksen 1988: 115 thinks Jesus’ meal with his disciples possible because gospel traditions sometimes depict the kingdom as a messianic banquet and because Essenes and others reportedly celebrated meals of eschatological import. But that Jesus would keep the Passover fits what we may surmise from his loyalty to Torah in the traditions, and his special interpretation of it is consistent with what we may surmise about his deliberate provocation of his martyrdom (see comment on 16:21).

. With Daube 1963: 45; Ellington 1979; Meier 1980: 321; Carson 1984b: 538; Hagner 1995: 774; cf. Philo Spec. Leg. 2.148. Probably pilgrims also sang en route to festivals (Sanders 1992: 128-29, especially citing Herodotus).

. Brown 1994: 123 thinks Mark’s and Matthew’s audiences would think instead of Christian hymns (Eph 5:19; Col 3:16); but it seems likely that at least Matthew’s community would recognize the use of hymns in the paschal context, especially if the community was, as I think, located near Palestine. Even if most of the later paschal liturgy was not in use or common use in this period, praise was standard in eating the Passover (Jub. 49:6; cf. Philo Spec. Leg. 2.148). The exact sequence of the Hallel in the liturgy was probably not standardized by Jesus’ day (cf. the differences among rabbis in t. Pisha 10:9; b. Pesaḥ. 118a).

. See Brown 1994: 217 especially on the scandal the account would represent to “educated Greco-Roman pagans,” whose models of martyrdom included Socrates and others facing death calmly; but also among Jews (the Maccabean martyrs; also Jos. War 1.653; 2.153; 7.417-18). That Mark elsewhere mentions prayer only at 1:35 and 6:46 suggests that this prayer is not a Markan creation (Brown 1994: 150). Sanders 1993: 264 also regards the prayer the tradition attributes to Jesus as “perfectly reasonable”; hoping to escape death, he nevertheless was ready to face whatever God wished. For further discussion of the historicity of the Gethsemane tradition, see Green 1992b.

. The threefold repetition of the prayer (26:44) may be stylistic or derived by inference from Mk 14:41; an echo of the threefold test in 4:1-11 (cf. Thiemann 1989: 336n.13) is less likely.

. Brown 1994: 125 believes that Zech 14:4 is the direct background for Acts 1:9, 12; Mk 13:3 but that the Passion Narrative echoes David’s ascent in 2 Sam 15:16, 23, 30; his connection between David’s ascent and the “house of David” in Zech 13 (128) may be more strained. If the location on the Mount of Olives in our passage (26:30) recalls for the reader the location of Jesus’ end-time discourse, it appropriately sets the scene for his impending suffering in the context of the world’s powerlessness and his ultimate triumph (cf. Zech 14:4).

. The Johannine parallel provides multiple attestation, probably independent of the Markan account (Dewey 1976: 106; Brown 1994: 611-13); for a full defense of the authenticity of the denials, see Brown 1994: 614-21, noting especially that the earliest Christians would not have invented a negative story about Peter (615). Brown suspects that the basic historical fact is retold in an imaginative way (620-21). Given other sets of three in the Passion Narrative (e.g., he finds them sleeping three times), the three denials could fit an early storytelling pattern (Brown 1994: 11-12) but are probably authentic (Brown 1994: 613-14).

. Many follow here Kosmala 1963. Later rabbis’ claim that no fowl existed in Jerusalem (m. B. Qam. 7:7; Lachs 1987: 412) is fanciful nostalgia. Most references to cockcrowing in antiquity, including the context in the Gospels, omit connection with night spirits (although for superstition and cockcrowing, cf. Petron. Sat. 74); hence we may reject the suggestion of Derrett 1983b. And though “cockcrow” was the name of the third Roman night watch (customary signals marked night watches — Sall. Jug. 99.1), it is unlikely that Jesus would have referred to the more obscure trumpet blast at the 3 a.m. changing of the Fortress Antonia guard than to a rooster (pace Hunter 1965: 169; cf. Mounce 1985: 259); in any case, one could not expect Matthew’s audience to catch this allusion.

. Cf. Brown 1994: 137, noting that Mark’s “second” cockcrow would be associated with the dawn, as the second cockcrow normally was (citing Aristoph. Ecclesiazusae 30-31, 390-91; Juv. Sat. 9.107-8; Ammianus Marcellinus Res Gestae 22.14.4).

. Matthew may conform the reading to his reading of the LXX (cf. Gundry 1982: 530) or scribes may have conformed that version of the LXX to Matthew.

. Marxsen views this promise in Mark as the parousia, related to but not identical with the resurrection (in Mk 14:28; 16:7; cf. 1969: 75, 90, 111-16); objections to his position are summarized in VanderBroek 1983: 55-65.

. As such, it does not reflect much “Persian or Hellenistic influence” (pace Meier 1980: 325), although hellenistic influence elsewhere affected early Jewish and Christian language (“flesh” indicated weakness and consequent susceptibility, e.g., in Plut. Pleasant Life Impossible, 6, Mor. 1090EF; 1QS 11.9, 12; see Best 1965: 52). This is no radical contrast between body and soul, as in Plut. Isis 78, Mor. 382F; cf. Epict. Disc. 2.23.20; 3.7.2-3; Marc. Aur. Med. 2.2; the components in Sifre Deut. 306.28.2. Even Paul’s usual usage reflects biblical and Jewish tradition in which humanity is flesh (Ps 78:39; Jub. 5:2; Sir 28:5; Test. Job 27:2/3; Test. Jud. 19:4; Zeb. 9:7) and wholly distinct from God’s Spirit (Gen 6:3; Is 31:3; Jub. 5:8; see Flusser 1988: 64-65).

. Cf. 1 Pet 4:7; 5:8; F. F. Bruce 1972a: 71n.14; Jeremias 1972: 44, 55; Garrett 1998: 91-94; pace Barrett 1967: 47. But Brown 1994: 196, following Lövestam, is right to find “the combination of watching and praying in Psalms” (e.g., Ps 63:6; 77:6; 119:148; also 1QS 6.7-8; Lk 2:37; Acts 16:25).

. Brown 1994: 154 relates Mk 14:34 (Mt 26:38) to Ps 42:6-7 (LXX Ps 41), which 1QH 8.32 also echoes; Davies and Allison 1997: 565 point out that Jewish tradition assigned this psalm to David in the context of Absalom’s betrayal. Dodd 1965: 71 suggests multiple attestation for Ps 41 in Mk 14:36; Jn 12:27; and Heb 5:7. The “troubling” of the soul, however, is idiomatic; if Ps 41:11; 42:5 LXX (Ps 42:11; 43:5 MT) reflect the exact wording of the text, analogous wording appears in Ps 6:3; Gen 41:8 (cf. Jn 12:27).

. Kiley 1986 also finds an allusion in the Markan form of the prayer to Ps 116:4’s entreaty, “Save my life,” but the verbal allusions by themselves, at least, are inadequate.

. Ps 11:6; 60:3; 75:8; Is 29:9-10; 51:17, 21-23; 63:6; Jer 25:15-29; Lam 4:21; Zech 12:2; cf. Ps. Sol. 8:14-15; in the ancient Near East, Albright and Mann 1971: 327. The image could signify death in general as in texts betraying more hellenistic influence, perhaps alluding to Socrates’ execution: Test. Abr. 1:10; 16A; Cic. Tusc. Disp. 1.40.96; Greek Anth. 7.96. But this source was less available to the early Christians than the LXX and far less available to Jesus (though Brown 1994: 169 also finds the cup of suffering in Targ. Neof. Deut 32:1; Neof. Yer. II, and Fr. Targ. Gen 40:23; Mart. Is. 5:13; he also notes “taste death” in Heb 2:9; Jn 8:52; 4 Ezra 6:26, but this was idiomatic — e.g., Sib. Or. 1.82; cf. Marc. Aur. Med. 9.2; Gen. Rab. 21:5; Lev. Rab. 18:1; Pesiq. R. 48:2; Lane 1991: 49). On the nature of literal cups at banquets, see Varro 5.26.121; Athenaeus Deipn. 11.781c-503f; Safrai 1974/76b: 742-43.

. That the Gentile churches may have treated it as an ecstatic palindrome (Aune 1980: 1550) is rendered less probable by the translation accompanying it (Rom 8:15; Gal 4:6); for early Christian liturgical usage, see Jeremias 1964b: 64-65. The supposition that it goes back to the Lord’s Prayer (e.g., Hunter 1961: 50; Ridderbos 1953: 158), however, remains undemonstrated (see, e.g., Lull 1980: 67; comment on 6:9).

. “Abba” is not exclusively children’s language (Vermes 1993: 182), although the frequency of its use in that domain produces nuances of affectionate intimacy (Witherington 1990: 218). It provides some insight into Jesus’ relationship with his Father (see Dunn 1975: 21-26).

. In the terms of traditional Christian theology, by becoming fully human, Jesus experienced the full human dread of death; because the Son is distinct from the Father, his own desire might differ from the Father’s, though he was ready to submit to the Father. The text probably tends, however, more toward a so-called “monothelite” interpretation than much of later Christendom adopted. Some philosophers emphasized leaving decisions to deities rather than praying as one wanted, as in popular piety (Burkert 1985: 75; van der Horst 1995), but resignation to God’s will was a Jewish conception no less (1 Sam 3:18; 2 Sam 10:12; 16:10-12). Even some nonphilosophers regarded praying for one’s life as less than heroic (Longinus Sublime 9.10). Meanwhile, the element of the Son’s choice conflicts with the frequent Greek tradition that stressed the futility of trying to escape one’s prophesied appointed fate (Diod. Sic. 15.74.3-4; cf. Sen. Ep. Lucil. 27.3); on Fate, see comment on 10:29-31.

. Moule 1965: 118 suggests that this is military language summoning soldiers to advance (also Gundry 1982: 536); while this suggestion is too specific, it does reflect the strength of the summons. As in Jn 14:31, the summons is to join Jesus in his mission of suffering according to the Father’s will.

. Stoics could also suggest resigning oneself to God’s will, though this was intended in the sense of fate (e.g., Epict. Disc. 2.16.42; 2.17.22). One should compare Jesus’ prayer with the ancient praise of courage in the face of death (see comment on 16:21).

. On the temple watch, see Safrai 1974/76d: 872-73; Cadbury 1979: 300; on their captain, the segan, and his authority and high-priestly connections, see, e.g., b. Pesaḥ. 14b; 47a; Tamid 27b-28a; Acts 4:1; possibly “the king’s captain” in Jos. War 1.652.

. Klausner 1979b: 337; Lane 1974a: 524, citing b. Pesaḥ. 57a, bar.; t. Menaḥ. 13:21; for use by Pilate’s soldiers, cf. Jos. War 2.176; Ant. 18.61; other Romans, Jos. War 2.326. Stauffer 1960: 120 thinks that carrying nonadorning weapons violated Sabbath law (m. Shab. 16:4), hence that the expedition does not occur after the Passover meal. Even assuming that the Levite police would normally have observed this rule (which is unlikely), however, military emergencies suspended normal Sabbath regulations (1 Macc 2:40-41), and the chief priests may have been willing to construe the potential revolutionary Jesus in these terms.

. Cf., e.g., Stauffer 1960: 119; Winter 1961: 30, 44; Hunter 1965: 166; Anderson 1976: 327; Cohn 1977: 78; F. F. Bruce 1980b: 9; Blomberg 1992a: 397; Brown 1994: 248-51; cf. Jn 18:3, 10, 31. Blinzler 1959: 9 complains about critics who suppress (as inauthentic) evidence “inconvenient” for the thesis.

. See further Catchpole 1971: 149; Blinzler 1959: 64-65; Bammel 1984c: 439-40; cf. Wilkinson 1978: 126-27. Some have argued that the Synoptic trial narratives increase Jewish guilt as against Roman (cf. Townsend 1979: 77, on John; Zeitlin 1962; Winter 1961), others that some historical data point to the Jewish leaders’ responsibility (cf. Bammel 1984c: 445; Betz 1988); for a full summary of views, see Blinzler 1959: 3-21; and especially Catchpole 1971. The narratives make the most sense if both Jewish leaders and Romans were involved (so also Brown 1994: 250, though basing this view partly on the involvement of Roman troops in the arrest; cf. 1 Thess 2:15). Cohn 1977: 75 doubts that any but people of the “lower strata” would have joined this mission; but I suspect that the temple police would have followed the orders of the sagan, a member of the priestly aristocracy.

. In cases of lesser danger, Romans often settled for executing ringleaders of a revolt rather than a whole populace (e.g., Dion. Hal. 3.40.3; 5.43.2); the Jewish officials would probably be still more inclined toward mercy to fellow Jews.

. The early Christians probably extended to one another (Rom 16:16; 1 Cor 16:20; 2 Cor 13:12; 1 Thess 5:26; 1 Pet 5:14; Justin 1 Apol. 65; cf. Acts of Jn 78) the practice of (generally) lip-kissing (light, not passionate; e.g., Eurip. Phoenician Maidens 1671; Virg. Aen. 12.434; Ovid Metam. 2.356-57, 430-31; 10.362) as a matter of familial intimacy (Gen 33:4; Song 8:1; Tob 7:6; 10:13; Jub. 22:10-11; 26:21; 31:21; Lk 15:20; Test. Sim. 1:2; Jos. and Asen. 8:4/3 in view of 7:8/11; Num. Rab. 9:9; Song Rab. 8:1, §1; Eurip. Androm. 416; Virg. Georg. 2.523; Ovid Metam. 4.222, 334; 10.525; Longus 4.22-23; Dion. Hal. 7.7.1; Plut. R.Q. 6, Mor. 265B; Ach. Tat. 1.7.3; cf. Dupont 1992: 285). It was also extended to close friends (Hom. Od. 21.224-27; Ovid Metam. 2.430-31; Artem. Oneir. 2.2) and respected teachers or others (even pupils) who brought one great joy (1 Esdr 4:47; t. Ḥag. 2:1; ARN 6A; 13, §32B; b. Ḥag. 14b; p. Ḥag. 2:1, §4; Hor. 3:5, §3; Rosh Hash. 2:9, §2; Soṭa 13a; Pesiq. Rab Kah. 1:3; Qoh. Rab. 6:2, §1; 9:5, §1; Pesiq. R. 23/24:2; Ap. Rhod. 1.313; Arrian Alex. 4.11.3; Epict. Disc. 1.19.24; Char. Chaer. 2.7.7). Serious kissing could indicate more, however (Jos. and Asen. 8:5-7/4-7), and Roman custom did limit even marital kissing in public (Plut. Bride 13, Mor. 139E; cf. Dupont 1992: 112-13; Brown 1994: 255; Median custom in b. Ber. 8b). Because phileō could mean “love” as well as “kiss,” one could play on both nuances (Diod. Sic. 9.37.1).

. Robbins 1992: 189 cites Dio Chrys. Or. 3.86, 114: the king depends on his friends’ loyalty, and discovering that they are his enemies is his greatest pain. Greek and Roman texts often enough regard the treachery of deceitful betrayal as deplorable (e.g., Polyb. R.R.E. 1.7; 3.52; Jos. War 2.450-54).

. Cf. Cullmann 1956b: 40-41: At last the disciples realized that Jesus was not a revolutionary Messiah and would not fight. Flight is not always wrong (10:23), but abandonment of Jesus was apostasy (10:28, 32-33). Davies and Allison 1997: 566 helpfully compare Ahithophel’s plan to make David’s troops flee (2 Sam 17:2), but the Gospel tradition probably depends on a broader principle here (see Matt 26:31).

. Josephus praises himself by emphasizing the loyalty of the Galileans to him (Jos. Life 84), though as an aristocrat he would have disdained “populist” support had he lacked it.

. Angels can fight for saints (Davies and Allison 1997: 513 cite 2 Kings 6:15-17; 1QM 7.6; 13.10; Ps-Philo 27:10; 61:8; 2 Bar. 63:5-11; 4 Macc 4:10) and sometimes appear with swords (ibid. cite Num 22:23; Josh 5:13; 1 Chron 21:16, 30; 3 Enoch 22:6). Some Jewish texts portray angelic armies arranged in ranks (see, e.g., 1 Enoch 69:3; 75:1; 3 Enoch 5:2; Gk. Ezra 1:7; cf. Jos. and Asen. 14:7; Test. Abr. 1.13; 2.1; 14A; 14B; Test. Mos. 10:2; 3 Bar. 11:4, 6, 8; 2 Enoch 22:6 J; 33:10 A, J; Gk. Ezra 1:4; 4:24; Apost. Const. 8.12.27; Delcor 1955: 374; cf. the armies of heaven in 2 Kings 6:16-17; Josh 5:13-14; 1QM 7.6; 12:8; the alleged apparitions in 2 Macc 5:2-4; Jos. War 6.298-99; Tac. Hist. 5.13). Cf. the common Jewish notion of guardian angels in, e.g., Ps-Philo 11:12 (custodes); 59:4; Sifre Num. 40.1.5; Gen. Rab. 44:3; Philo Gig. 9; cf. Sent. Sext. 32. Matthew lays special stress on angels (20 times in his Gospel; Brown 1994: 276).

. Brown 1994: 268-29 finds little evidence in the text to explain why a disciple would have a sword, except Lk 22:35-38, which may derive from a source; he notes that Essenes carried weapons during travel because of bandits (Jos. War 2.125). After the temple cleansing, it is logical to think that some disciples expected trouble (Lk 22:27-28). Given the Lex Iulia de maiestate (Justinian Dig. 48.4.1, cited in Brown 1994: 274), Jesus may have allowed the presence of swords to facilitate his rapid conviction as one dangerous to the public order (Lk 22:37). That Matthew and John independently agree (without Mark) on the command to sheath the sword suggests that this command derives from pre-Gospel tradition (Brown 1994: 86; unless John depends on very recent oral tradition derived solely from Matthew).

. On like-for-like repayment of misdeeds, cf., e.g., Demosth. Against Zenothemis 6; Diod. Sic. 20.65.2; 20.101.3; Aul. Gel. 7.4.4; Sir 27:25-27; Jub. 4:32; 35:10-11; 37:5, 11; 1QpHab 11.5, 7, 15; 12.5-6; Ps-Philo 44.9-10; m. ʾAbot 2:7; Mek. Nez. 18.55-58; Sifre Deut. 238.3.1; ARN 27, §56; b. ʿAbod. Zar. 17b; Sanh. 108b; Bonsirven 1964: 110; Finkelstein 1972: 219; for another, unrelated culture, see, e.g., Whisson 1964: 289.

. See Cullmann 1956b: 42, 46; Argyle 1963: 206; Reicke 1974: 146; France 1985: 377; Hagner 1995: 797; cf. Harvey 1982: 31; Meier 1980: 330; Sanders 1993: 67.

. Brown 1994: 433 observes that “ancient literary accounts of famous trials” usually include polemic or bias (he would not, however, call the narrative as a whole a “fiction,” as in Mack 1988: 295). Trial scenes can also provide suspense in a plot, e.g., in Chariton Chaereas, toward the beginning, and later before the king of Persia.

. Rabbinic attestation of a religious trial of Jesus (Stauffer 1960: 225; cf. Herford 1966: 78-83) is late and probably derivative, hence I do not admit it as independent evidence; but a Jewish hearing is essential. Even Winter, who emphasizes an arrest and agenda set by the Romans (1961: 30, 147), recognizes that Pilate would have expected the high priest’s aides to prepare the case (1961: 29).

. Weeden’s thesis that the pre-Markan trial tradition belonged to Mark’s “divine man” Christology opponents (Weeden 1976: 129) is fanciful; no trial narrative leading to execution (contrast Apollonius of Tyana’s disappearance) would easily fit such a Christology.

. According to the Mishnah, only a portion of the Sanhedrin need have been present; cf. Mk 15:43; but contrast Mt 27:57, underlining the extent of the opposition perhaps in a manner more relevant to Matthew’s audience. The absence of a defending attorney for Jesus despite rabbinic practice (Blinzler 1959: 135) could represent violation of the rules, but it probably reflects the lateness of the rules (cf. Blinzler 1959: 142-43).

. So also reportedly Egyptian custom (Diod. Sic. 1.77.2). Diod. Sic. 12.12.2 thus considers particularly merciful a law that merely shames false witnesses so much that they flee a city. Cf. also the distaste for perjury in Sall. Catil. 16.2; Speech of Philippus 15; Invective against M. Tullius 2; Cic. Invective against Sallust 5.14; Propertius Eleg. 3.6.20.

. Gentile sages also admonished checking out all claims (e.g., Phaedrus 3.10.5-6). Stauffer 1960: 206 contends that Jewish law permitted the use of “secret witnesses” only in the case of a “misleader”; but it remains doubtful whether the Sanhedrin’s leaders would have observed this rule.

. Greek rhetoric often preferred arguments from probability and internal consistency (which were frequent; see, e.g., Demosth. On the Embassy 120; Against Pantaenetus 23; Arist. Rhet. 1.15.17, 1376a; Dion. Hal. 3.35.5-6; 11.34.1-6; Jos. Apion 1.219-20, 267, 286; 2.8-27, 82, 148; Life 342, 350; Acts 26:8) to witnesses (see Kennedy 1980: 20-21), but the effective testimony of witnesses was nevertheless adequate to convict (Dion. Hal. 8.78.3). Any proofs were, however, better than mere assertions (Jos. Ant. 17.131).

. Jn 2:19 could represent a more original form of the saying; “raise up” (egeirein) fits the image of building a temple (Sib. Or. 3.290).

. Pace Weeden 1976: 121-29, who links it to purveyors of a proposed divine man Christology. What Jesus meant by the new temple in the logion is more problematic; Brown 1994: 438-43 classifies the main positions as follows: (1) the Christian community (Cullmann, Donahue, Gaston, Juel, Sweet, Vielhauer; 440-41); (2) an eschatological sanctuary (Jeremias, Pesch, Sanders; 441-43); (3) the body of Christ raised on the third day (Cole, Gärtner, Gaston again, Lamarche, Nineham, Simon; 443).

. It is doubtful that first-century Jewish custom prescribed that “the examining judge,” here the president of the Sanhedrin, “had the right and even the duty to intimidate the defendant” to make him lose composure (so Stauffer 1960: 123-24, citing p. Soṭa 1:16d; b. Soṭa 1:4). If the high priest’s motives are less than noble, it is in violation rather than fulfillment of Jewish custom. Bowker 1973: 49-50 thinks that Jesus’ refusal to cooperate corroborates their opinion that he is a “rebellious elder.”

. Cf. also Trites 1977: 187. This is a “testimonial oath”; Judaism later borrowed the custom of swearing on the Bible from the Roman-Christian practice (see Cohen 1966: 710-33). The supposed exorcistic link here (e.g., van Tilborg 1972b: 79) represents too narrow an application of the term and is without merit (Gundry 1982: 544; even in Mk 5:7 it is the demon rather than the exorcist who employs it).

. Witherington 1990: 258-59 also defends the saying by the criterion of dissimilarity (later Christians did not combine Ps 110:1 with Dan 7:13-14), by the criterion of embarrassment (the prophecy was unfulfilled in that they did not see him exalted in their lifetimes), and, in my opinion a much stronger argument here, the criterion of consistency: Jesus elsewhere uses both texts cited here (Mk 12:36; 13:26).

. Cf. further here Kingsbury 1983: 122; Perrin 1976a: 95; Hooker 1983: 58-59. Stauffer 1960: 125 and Schoeps 1961b: 161 suggest from the Markan version that Jesus may have originally uttered the sacred name “I am” (ani [we] hu), leading to the condemnation of blasphemy and echoing (Schoeps) “the assertion of His divine sonship … behind the ‘I am’ sayings.” While explaining the charge of blasphemy (a mere claim to Messiahship cannot — Ladd 1974b: 168), this proposal does require an interpretive leap on the part of the high priest, especially given the absence of this formula at most points (Sanders 1990: 65). Nor does a variant reading of the LXX explain the blasphemy charge here (see Pace 1984, against F. F. Bruce).

. That Dan 7 and Ps 110 are here in view, see Dodd 1961: 91; Ellis 1977: 203.

. The phrase could refer to the future (cf. Betz 1968: 90) but probably, as in Luke, begins in the present in some sense (cf. Tasker 1961: 256; F. F. Bruce 1972b: 176; Glasson 1962: 54-59; Brown 1994: 500-504). Thus (pace Meier 1979: 192, who thinks the phrase addresses the Matthean community in particular) Matthew and Luke probably share a common oral (hence the smooth paraphrase) tradition here apparently unknown to Mark (given his theology of the cross, Mark probably would have retained it had he known it). The language ultimately demands the parousia, however, and not merely a heavenly exaltation (Cullmann 1950: 149; McArthur 1958: 158; Argyle 1963: 207; Gundry 1982: 546).

. Although some read Jesus’ “from this time forth” statement and other evidence as a denial that Jesus foresaw an interim between his comings (cf., e.g., Barrett 1967: 74, against Kümmel), Jesus’ presence-of-the-kingdom parables (ch. 13), mission for his disciples, and other factors militate against this suggestion (cf., e.g., Goppelt 1964: 92). Jeremias notes that Jesus rejected forecasts of dates and allowed that God could shorten or lengthen times of distress (Lk 13:6-9; 18:7-8; Mk 13:20; Jer 18:7-10; Jeremias 1971: 139); against “consistent eschatology,” Cullmann notes that the completed victory of the past produces rather than replaces futurist anticipation (1950: 87).

. This title is widely recognized, e.g., by Moore 2:355; Klausner 1979b: 197; Marmorstein 1968b: 161; Smith 1951: 35n.2; Barrett 1966: 72; Jeremias 1971: 10; more references are found in Davies and Allison 1997: 529. The NIV’s “Mighty One” captures the sense; Lk 22:69 adds “of God” to “power,” retaining but clarifying the circumlocution (cf. Test. Abr. 4A).

. Perhaps guarding against convictions based merely on dominant legal interpretations, later rabbis distinguished between violations of Torah, which could be capital offenses, and those of scribal teachings, which could not be (Sifre Deut. 154.2.1). But his hearers would surely regard Jesus’ words as at least presumptuous (Davies and Allison 1997: 534 parallel Test. Job 33:3-9).

. Overman 1996: 388 suggests that “son of God” had political connotations, as a title of the emperor. Matthew’s audience could thus have associated this claim with the impending charge of treason.

. Greek culture also understood that human courts had to prosecute those who insulted the gods or their laws (e.g., Ps-Demosth. Or. 59, Against Neaera 126).

. Brown 1994: 517 shows that non-Jews would also understand, citing the examples of Licinius Regulus and Augustus, who tore clothes in mourning (Dio Cassius R.H. 54.14.1-2; 56.23.1).

. 27:51 may perhaps represent God’s rending for true blasphemy as opposed to human rending for false blasphemy in 26:65, but had Matthew intended the connection to be obvious, he may have more likely used the same word.

. The rabbinic practice may respond to the earlier Roman custom by which the elders voiced their views first (Dion. Hal. 7.47.1; though this changed — Aul. Gel. 4.10; 14.7.9); but the Sanhedrin, not constrained by later Mishnaic rules, may well have followed general Mediterranean practice (cf. 1QS 6.10).

. Although the Mishnah itself is later (Blinzler 1959: 142-43), this rule does suit the spirit of Jewish law in general. Rabbinic rules prohibited convicting one of blasphemy until he had first been warned, but teachers of the law were exempt from requiring such warning (Blinzler 1959: 136, following Strack-Billerbeck); this particular rule, however, sounds much like later rabbinic leniency toward the accused rather than a common Jewish custom of the first century.

. Jewish men could mock their enemies no less thoroughly than Romans; cf. Jos. Life 323.

. Matthew follows Mark on the beating by the Sanhedrin (who are ultimately responsible), but despite other reports of abuse by high priests, Winter 1961: 21 believes that Luke’s account of beating by jailers may more accurately reflect a pre-Markan Passion Narrative (Lk 22:63-65), although Luke probably also takes over material from Mark here, which he then rearranges (Lk 22:67-71).

. Brown 1994: 641 believes that the contrast between Peter and Judas emphasizes that Judas was guilty of Jesus’ blood but Peter was not guilty. But why would Matthew need to stress this?

. Theissen’s parallel with Acts 12, however, is fatally flawed; as Theissen admits, Peter did not deny his faith under Agrippa I. Indeed, there remains no evidence that anyone supposed that he had; rather, that account fits Acts’s theme of miraculous escapes (Acts 5, 16).

. Given the other eschatological connotations in the language of the disciples’ testing (e.g., 26:41), the “outcome” here may point beyond Jesus’ suffering to the birth pangs of the Messiah to be suffered by Jesus’ followers preceding his return (24:6). This is, however, somewhat speculative; the expression is perfectly good Greek.

. Naturally she could accuse him to others, and Jewish tradition seems to have suspected women in particular of gossip (b. Qidd. 49b; Gen. Rab. 70:11). But the witness of women was also more suspect (e.g., Jos. Ant. 4.219; see further below), and she had less direct power to hurt him than did other actors in the narrative.

. The supposed Aramaic mistranslation here is untenable (see Boyd 1956).

. Malina and Rohrbaugh 1992: 160-61 contend that Peter’s deception to maintain his honor comported with Mediterranean values, and the difficulty is that Peter gave his word to Jesus that he would not deny him (26:35). But another problem is that Jesus demands that his followers honor him publicly, so gaining honor before his Father’s tribunal, and that this must take absolute precedence over honor in this world (cf. 10:32-33). Cf. even ARN 27A: The honor of one’s master (and one’s disciple) should be as dear to one as one’s own.

. Analogously, lack of fluency in a second language could give someone away (see Horsley 1989: 31).

. For Upper Galilee as a hotbed of revolutionary fervor, cf. Vermes 1973: 46-48; Vermes 1984: 4-5. Jesus ministered especially in Lower Galilee.

. Cf. Lindars 1961: 120; Gundry 1982: 556-57; Meier 1980: 339; Brown 1994: 651. Jerome cites an apocryphal copy of Jeremiah that included these words (cf. Lachs 1987: 423), but some Jewish Christians had undoubtedly created that document to address the “problem” in Mt 27:9. Luther felt that Matthew’s “error” here was inconsequential (Stanton 1995a: 8), but while Matthew may have been capable of such an error, it is unlikely that one occurred in this case. Given his ability to retranslate the entire Hebrew text based on revocalization (I have examined only the point most germane to the sense of our passage), it is unlikely that Matthew simply got his attribution wrong.

. But cf. Gundry 1975: 202: “The connection with and the typological interpretation of Jer 19 would never have suggested itself apart from a prior tradition of the potter’s field.”

. Overman 1996: 382 finds a specific allusion to 1 Sam 26:11; 2 Sam 1:16 (opposing killing “the Lord’s ‘christ’”), but if Matthew intended so narrow an allusion, one would expect clearer echoes of that phrase here.

. E.g., Gen 4:10; 9:5-6; Num 35:33; 2 Sam 3:28-29; 4:11; 21:1; 1 Kings 2:5-6; Middle Assyrian Laws Tablet A 10 in ANET 181; see on 23:35-36; for the guilt of locales, cf. Deut 21:1-9; Paus. 10.5.3; Aqhat in ANET p. 154; Hammurabi 24; Moyer 1969: 67-69, 120. See other references in Brown 1994: 641, including Philo Spec. Leg. 1.204. Van Unnik 1974: 44-57 rightly notes the betrayal of innocent blood here, but less probably he argues that Judas’s hanging himself resolved the curse because the priests would not let him atone for his sin by sacrifice. Philonenko 1993 finds in righteous blood an allusion to 1 Enoch 47:1, 4.

. Van Tilborg 1972b: 88 regards 27:4 as Matthew’s most significant contribution to the tradition here, painting the authorities as more guilty than Judas; Gundry 1982: 558 concurs with the increase of their guilt but denies the reduction of Judas’s. Stauffer’s conclusion that the recantation of a witness could not alter an established verdict (1960: 128, citing m. Sanh. 6:23 [perhaps he means m. Sanh. 5:5, though its point is somewhat different]; b. Sanh. 43a) is undoubtedly irrelevant to the thought of first-century Sadducean officials.

. This may be implied in their “see to it”; cf. Marc. Aur. Med. 5.25. In at least some Jewish traditions, a suicide could turn away God’s wrath from the person on whom God had pronounced judgment (Qoh. Rab. 7:12, §1; 9:10, §3).

. Regularly in the texts, e.g., Acts 16:27; Soph. Women Tr. 721-22; Demosth. 3d Philippic 62; Diod. Sic. 12.19.2; 16.45.4-5; 25.17.1; Tac. Ann. 1.61; 3.42; 4.25; 6.23-26, 38-40; 11.37-38; 12.8, 22; 13.1, 25, 30; 15.57, 63-64, 69; 16.11, 14-15, 17; Suet. Aug. 27, 53, 67; Tib. 45, 61; Nero 49; Otho 9, 11; Dio Cassius R.H. 17.15.4; 18.4.6; 19, fr. in Zonaras 9.21; 48.44.1; 51.15.3; 57.18.10; Appian C.W. 1.8.74; 1.10.94; 2.14.98-99; Livy 26.15.13-15; 41.11.4-6; Corn. Nep. 20 (Timoleon), 1.6; 23 (Hannibal), 12.5; Epict. Disc. 2.1.19; 3.8.6; Paus. 9.17.1-2; 9.25.1; Jos. Apion 1.236. Taylor 1966: 167 also cites Plut. Cato Min. 66-72. Though (at least for Romans) falling on a sword was often preferred, hanging oneself appears in various localities (e.g., Diod. Sic. 2.6.10; 20.71.4; Philo Mut. 62; Apul. Metam. 1.16).

. 4 Macc 17:1; cf. the analogous b. Giṭ. 57b (Montefiore and Loewe 1974: 260); Jos. Life 137; the Sicarii at Masada (Jos. War 7.320-406); cf. Goodblatt 1995. Although the narratives portray Judg 9:54; 1 Sam 31:4; 2 Sam 17:23 as shameful deaths of wicked men, the doers probably perceived their suicides as tragic acts of valor in keeping with the heroic horror at dishonor of the epic age (Soph. Ajax 864-65, 898-99, 907-8; Oed. Rex 1237-66; Eurip. Androm. 811-13; Helen 136, 200-202; Hippol. 776-81; Virg. Aen. 10.681-82; Ovid Metam. 13.388-92) or its romantic pathos (Soph. Antig. 1175-77, 1234-36, 1282-83, 1315-16; Women of Trachis 891, 923-35; Ap. Rhod. 1.1063-65; Virg. Aen. 4.663-65; Ovid Metam. 4.119-20, 147-66; 6.271-72; 10.378-81; 11.783-84; 14.80-81; Parthenius L.R. 10.4; 11.3; 31.1-2; cf. Hom. Il. 18.33-34; Eurip. Phoenician Maidens 1455-57; Appian R.H. 6.2.12).

. Cf. the prayer of the condemned, “May my death atone for all my sins” (m. Sanh. 6:3; cf. Ps-Philo 25:7; 27:15; t. Ber. 6:17; Kip. 4:8-9; 6:17; Sifra A.M. par. 2.176.1.5-6; ARN 29; 39A; b. Ber. 60a; Shebu. 13a; p. Sheb. 1:6, §5; Pesiq. Rab Kah. 11:23; 25:3).

. E.g., Hesiod W.D. 221; Livy 40.19.11; Longinus Sublime 44.9; Appian C.W. 2.3.23; 2.4.24; Sall. Speech of Philippus 6; Speech of Macer 5; Corn. Nep. 6 (Lysander), 3.2; 15 (Epaminondas), 4.1-3; Dio Cassius R.H. 54.16.1; Quint. 4.1.21; Ex 23:8; Deut 16:19; 27:25; 11QTemple 51.16-18; Jos. Apion 2.207; Life 79, 196; Pesiq. Rab Kah. 15:9. People in authority, especially judges, were expected to be impartial (Demosth. On the Crown 1-2; Isoc. To Nicocles 18, Or. 2; Cato the Elder 13 in Plut. Sayings of Romans, Mor. 198F; Sir 42:2; 1 Esdr 4:39; Ps-Phocyl. 9-12, 137; ARN 36A; on deities, Plato Alcib. 2.150A), and God’s impartiality was often extolled (Sir 35:12-13; Wis 6:7; 2 Bar. 13:8; 44:4; Ps-Philo 20:4; Test. Job 4:7-8; 43:13/10; Sifre Num. 42.1.2; Sifre Deut. 304.1.1; Rom 2:11; see most fully Bassler 1982: 7-119).

. Cf. similarly the hangings in Philo Mut. 61-62 and Ovid Metam. 10.378-81; Parthenius L.R. 11.3; 31.1 (though this was a common method — Soph. Antig. 1221-22; Oed. Rex 1237-66; Ap. Rhod. 1.1063-65; Virg. Aen. 12.600-603; Aul. Gel. 15.10.1-2; Herodian 7.9.4); Brown 1994: 644 cites Jos. War 3.369; m. Sanh. 10:2; Semahot ‘Ebel Rabbati 2.

. The field in the tradition may suggest Matthew’s use of the “potter’s field” from Zech 11:13 and possibly Jer 19:6 (Gundry 1982: 556). Surprisingly, Benoit 1973: 189-207 concludes that where Matthew’s and Luke’s accounts diverge, Matthew is more apt to be correct. Cf. Papias Fg. 3 (from Oecumenius), possibly dependent on the Acts account but perhaps attesting a fuller tradition. Those who wish to harmonize the accounts can point to ropes breaking in unsuccessful suicide attempts (e.g., the spoof in Apul. Metam. 1.16; cf. Petron. Sat. 94; hopefully no one will appeal to the possibility of one’s surviving his innards being spattered — b. Ḥul. 56b-57a), though hanging and falling are normally distinguished (Lucan C.W. 2.154-58). Bowels coming out represents a horrid fate (2 Macc 9:5-9; Jos. Ant. 9.103; cf. Acts 12:23; Herod. Hist. 4.205; Jos. War 1.656; Ant. 17.167; 19.346, 350; Test. Job 20:8-9; b. ʿErub. 41b). That the wicked bought a field with Jesus’ bloodprice (27:7) might suggest an ironic contrast with 13:44, where one gives up all one has to acquire the kingdom (though buying fields was common and the language could be coincidental).

. Ancients would regard offering money as a bribe to falsify another’s witness as despicable (Demosth. Against Meidias 107); for this offense, Athenian law prescribed being stripped of one’s citizenship and property (Demosth. Against Meidias 113), and current Roman law removed officeholders who had achieved office through bribery (Dio Cassius R.H. 54.16.1). The high priests repeat the offense in 28:12-15.

. Ancients were sensitive to such portrayals of hypocrisy (e.g., posthumous concern for another one injured while alive — Babrius 14.5) and would negatively regard the hypocrisy of religious pretense coupled with irreligious motives (e.g., Jos. Life 75; Juv. Sat. 2, 4 in Stewart 1994).

. Roman governors were noted for gaining wealth, sometimes at their clients’ expense (Stambaugh and Balch 1986: 65). The older Roman tradition of avoiding taxes to pay for public works (Badian 1972: 30) or displeasure at Julius Caesar’s spending public monies on a purpose different from that for which they were allotted (Appian C.W. 2.6.41) might have given him pause to consider his decision, but he probably assumed that he followed safe Roman precedent: Augustus and others paid for workers on aqueducts from public and imperial treasuries (Frontinus De aquis 2.89-101, 116-18, in Jones 1970: 207), and the use of public money would have been expected (Jos. Life 199) had it not been from the temple treasury.

. Plato Euthyphro 5; Diog. Laert. Lives 6.2.72-73; Paus. 6.11.2-3; Artem. Oneir. 3.3; Lucian Phalaris 1; Zeus Catechized 18; Zeus Rants 25; Apul. Metam. 11.10; Char. Chaer. 1.5.25; Athenaeus Deipn. 12.523; Tert. Apol. 15.7; cf. Polybius 9.28-39; Strabo Geog. 4.1.13; Tac. Ann. 15.45; Suet. Nero 32; Dio Chrys. 31st Disc. §148; for Jewish attitudes against desecrating pagan shrines, see Ex 22:27 LXX; Sib. Or. 2.14; Jos. Ant. 4.207; Apion 2.237; Philo Mos. 2.205; Spec. Leg. 1.53; ARN 31, §66B; b. Sanh. 64a, bar.; cf., somewhat less tolerantly, Deut 7:25-26; p. Giṭ. 5:9, §8. On temple robbery, see also comment above on 25:14-30.

. Thus the repetition of the death sentence in 27:1 is hardly merely a haggadic echo of the repetition in Jer 26:8, 11 (pace Doeve 1954: 187-88).

. Cf. Hor. Sat. 1.1.9-10 — at cockcrow; Epodes 2.1.103-5; Mart. Epig. 3.26.1-3; Plut. R.Q. 84, Mor. 284D; 3 Macc 5:26; Carcopino 1940: 151; Friedländer 1965: 1:86-93, 207; Cary and Haarhoff 1946: 149; Clarke 1994: 475.

. Prefects in Roman Egypt visited local municipalities only a few days a year, and addressed 700 to 750 petitions in a ten-hour day, better than one a minute, though clerks processed many of these (Lewis 1983: 190). Since he was present primarily to bolster security during the festival, it is doubtful whether Pilate would make himself that available. I omit here discussion of attempts to harmonize the Markan and Johannine chronologies, the latter of which probably fits into John’s theological Passover motif; for discussion, see Brown 1994: 958-59.

. Although Pilate may well have taken counsel with his accessores and comites, as in Acts 25:12 (Brown 1994: 716; cf. O’Rourke 1971: 174-75; Ferguson 1987: 51), the Gospels omit them as unworthy of mention; the governor issued the final decision and stood responsible for it (his responsibility appears also in Tac. Ann. 15.44.3). Yet despite designations of authority consistently applied to Pilate in this Gospel, his submission to the crowds ironically underlines his powerlessness (Weaver 1996: 191-95).

. Attributing the motive of envy to opponents (in stories, e.g., 3 Macc 6:7; Test. Dan 1:6; Acts of Paul 3:15; b. Yoma 71b) does double duty rhetorically: it praises or exonerates the one they accuse while diminishing the opponents’ status. Josephus charges that some Gentile writers ignored Israel due to envy (Apion 1.213, 222, 225), that his fellow aristocrats opposed him due to envy (Life 204), and that his own good fortune produced jealousy (Life 423), though God protected him (Life 425). Political biography regularly lists envy as a motive for hostility (e.g., Herodian 3.2.3; Corn. Nep. 5 [Cimon], 3.1; 8 [Thrasybulus], 4.1-2; 12 [Chabrias], 3.3; 14 [Datames], 5.2; 15 [Epaminondas], 7.1; 18 [Eumenes], 7.2; 10.2; 23 [Hannibal], 1.2). Given Jesus’ populist support, however, the alleged motive is not implausible.

. Although some (e.g., Meier 1980: 342) therefore assume that the gesture is not historical — Pilate was probably not so culturally sensitive as to adopt a Jewish gesture — it remains quite possible, because pagans also used it. (Blinzler 1959: 217-18 cites Virg. Aen. 2.719; Soph. Ajax 654; Herod. Hist. 1.35. Brown 1994: 834 adds Hom. Il. 6.266-68.) Pilate’s protestation of his innocence echoes the wording of the righteous Daniel in Sus 46 (Gundry 1982: 565).

. Matthew probably wrote after A.D. 70, when many Jewish pietists struggled with the question, “Why has God so punished us?”; Matthew intends to supply a response (Brown 1994: 29). Later anti-Semites abused a text whose proclamation of judgment focused on the generation of the temple’s destruction, i.e., Matthew’s own (Hare 1979: 38).

. See Sherk 1988: 40, §39A; Smallwood 1976: 145; Brown 1994: 336-37. Although scholars once thought that the officials brought Jesus to Pilate in the Fortress Antonia, most now concur that Pilate was residing in Herod’s palace (Schürer 1961: 181; Lane 1974a: 548; Wilkinson 1978: 140; Brown 1994: 705-10), as ancient sources suggest (Pilate in Philo Leg. Gai. 299; Florus in Jos. War 2.301, 328). Herod’s palace is farther from the temple but not a difficult trek.

. For the responsibilities of a governor, see, e.g., Justin. Dig. 1.16.4-13 in Jones 1970: 180-83.

. The Docetic idea of a wraith as substituted for Jesus on the cross, followed in a popular interpretation of the Qur’an, derives from hellenistic mythology, e.g., Eurip. Helen (following the Recantation of Stesichorus); Hom. Il. 5.449-53.

. Although the severest form of execution Pharisaic law acknowledged on the basis of the Hebrew Bible was stoning (b. Sanh. 49b-50a), Jewish rulers had used crucifixion before the Roman period. Under Roman rule, however, all official, public executions belonged to the Romans. Even the Essenes toned down capital sentences from Moses’ law (CD 12.2-5), while also destesting Gentile executions in the holy land (CD 9.1). The apparently Jewish execution in b. Sanh. 43a depends on Christian tradition, though preserving the crucifixion’s association with the Passover season.

. Harvey 1982: 12; Overman 1996: 380-81, 387; e.g., Jos. War 2.75, 241, 253, 306; 3.321; 5.449; Ant. 20.102. It also appears as a fitting end for other military enemies (e.g., Diod. Sic. 2.1.10; 25.5.2; Jos. Ant. 12.256; 13.380) and for the most horrid crimes (Apul. Metam. 3.9).

. An intermediate position is that Romans rarely delegated capital authority, but that the Roman governors were authorized to do so (O’Rourke 1971: 174-75); but whatever governors of some provinces may have wished to do, it is inconceivable that Pilate would have shared this authority with the Sanhedrin. Brown 1994: 339 correctly observes that executions required ratification by the Sanhedrin in Jos. Ant. 14.167; while this datum is undoubtedly relevant, we should note that it describes the time of Herod the Great, not direct Roman rule.

. Theissen 1991: 189-93 thinks that the Passion Narrative presupposes that the Sanhedrin had capital authority, and thus reads its own milieu’s circumstances of A.D. 41-44, under Agrippa I, into the narrative. This proposal falters on two counts: First, the logic of the Passion Narrative actually presupposes that the Sanhedrin lacks capital authority; why else would they hand Jesus over to Pilate? Second, Agrippa I was a client king who would hold capital authority — but he was not the Sanhedrin. Besides, Johannine tradition preserves the Sanhedrin’s lack of capital authority in the last decade of the first century (Jn 18:31-32).

. Later rabbinic rules allowed the defendant to speak first in a capital case (t. Sanh. 7:2), but even if some Jewish teachers held this view in Jesus’ day, Pilate would have operated under Roman procedure.

. On governors being tried for abusing power, especially executing innocent people (particularly Roman citizens), see Pliny Ep. 2.11, in Jones 1970: 192-95.

. Brown 1994: 695-705 ultimately concludes, as I do, that most of the Gospel portrait fits what we know of Pilate from the other sources once all has been taken into account.

. Tiberius reportedly viewed even negative remarks as maiestas (e.g., Dio Cassius R.H. 57.9.2; 57.19.1; 57.23.1-2; cf. Caligula — 59.11.6), leading to many false accusations (57.4.5-6); some later emperors also suffered paranoia (Herodian 1.13.7). Among Romans treason was the greatest crime (Dion. Hal. 8.80.1; see comment on 26:15 regarding traitors).

. Reasons for his dismissal in A.D. 36 are debated; Krieger 1992b rejects Josephus’s explanation.

. In Roman tradition ideal matrons were unobtrusive, but many privately influenced their husbands to appropriate public action, often more effectively than Pilate’s wife here (Stambaugh and Balch 1986: 111; cf. negatively p. Sukk. 5:1, §7). After Augustus (cf. Suet. Aug. 24) Roman governors could take wives with them (cf. Tac. Ann. 1.40; 2.54-55; 3.33-34; contrast Severus in Herodian 3.2.5), and many prominent women were interested in Judaism (Jos. War 2.560; Ant. 20.195; cf. Acts of Pilate 2.1; CIJ 1:384, §523; Blinzler 1959: 217; Brown 1994: 804; Leon 1960: 256). Jewish people also told stories of women who influenced powerful relatives not to punish God’s people (Lachs 1987: 428 cites b. Taʿan. 24b). On dreams, see excursus on Mt 1:20.

. Elsewhere in early Christianity, cf. Acts 2:17, but Luke’s narrative emphasizes the more dramatic “visions” (e.g., Acts 7:55; 9:3-4; 10:3, 10; 18:9; 22:6-7, 17-18; 26:19; cf. Acts 27:23; 2 Cor 12:1).

. For ancient literary portrayals of heroism, see comment on 16:21. Cohn 1977: 328-29 thinks that Jesus was innocent but pleaded guilty, perhaps to earn martyrdom and fulfill his prophecies. Some church fathers compared Jesus’ silence to that of an innocent lamb (Gregory Nazianzen Oration 29, On the Son 20; Augustine Tractates on John 116.4; ACCS, 222-23; cf. Is 53:7).

. Theissen 1991: 196 links this story with the Caligula crisis, when he thinks more of the populace would have sided with the “bandits” than with Christians (citing Tac. Ann. 12.54.1; Jos. Ant. 20.5, 97, 102); but this is hardly the only period in the first century in which that would be the case, and Jesus would be less popular than Barabbas to those prone to revolution (22:21) and, probably more to the point here, less popular with most of the Jerusalem masses than the priestly authorities were.

. This freedom may call into question the supposed official report cited by Eusebius (H.E. 2.2) and Tertullian (Apol. 21:24; cf. 5:2), which may depend on an earlier Christian forgery that the Christians assumed to be accurate (probably as in Justin 1 Apol. 35, 48; pace Stauffer 1960: 145-46).

. Pilate could have abolished a preexisting custom, but given previous conflicts with the people (e.g., Jos. War 2.174, 177) and the dangers of popular unrest at festivals (e.g., Jos. War 2.224) he probably would not have done so (though its lack of attestation in Josephus may suggest that someone eventually abolished the custom). Politically prudent rulers in the east presumably often continued festival traditions begun by their predecessors (e.g., Diod. Sic. 17.16.3).

. P. Florentinus 61.59ff., in Deissmann 1978: 269; Blinzler 1959: 207; Lane 1974a: 553; also Livy 8.35.1-9; the later practice of pardoning criminals at Easter (Cod. Theod. 9.38.3-4, 8), however, is probably dependent on the Gospels (Cohn 1977: 167). Blinzler 1959: 207, 218-21 argues for the custom of a paschal release of a prisoner in m. Pesaḥ. 8:6 (following Chavel’s 1941 article; cf. also Schnackenburg 1982: 3:252); but see Bammel 1984c: 427, who argues more persuasively that the text merely indicates the special Jewish desire to free prisoners at this time. Because we have little evidence for the customs of the period and because Roman governors were not required to continue previous customs (Ferguson 1987: 50; hence probably abolishing this one in time), it is not surprising that we have little direct evidence for the practice. Governors might also release prisoners in acceding to terrorist demands (Jos. Ant. 20.208-10). Ancient Mediterranean officials could even proclaim selective mass amnesties if they wished (P. Tebt. 5.1-13, 118 B.C.).

. Matthew or a later scribal tradition may heighten the contrast by including Barabbas’s real name (“Barabbas” is Aramaic for “son of the father”), “Jesus” (a common Jewish name in the period — Acts 13:6; Col 4:11; Jos. Ant. 6.129; 12.239; 20.234; Life 66-67, 134; Ep. Arist. 49; R. Joshua in Tannaitic texts; etc.); so Meier 1980: 341; Gundry 1982: 561. The reading’s difficulty may favor its authenticity (see Metzger 1971: 67-68), though the best texts omit it and geographical distribution seems slightly against it; the matter is not easily decided.

. Philosophers considered themselves most fit to rule (e.g., Quint. 2.17.28; Diog. Laert. 7.1.122; cf. CD 6.6; ARN 44, §124B; Gen. Rab. 93:2; Deut. Rab. 2:33), and many philosophers and moralists writing on the topos of kingship emphasized the ruler’s need for wisdom (e.g., Plato Rep. 5.472; Isoc. To Nicocles 10-11, 29, Or. 2; Plut. Sayings of Romans, Cato the Elder 8, Mor. 198F; cf. Prov 8:15; the Epistle of Aristeas’s symposium section; Sifre Deut. 161.2.1).

. Cf. also references in Meeks 1986: 57: Epicurus Fr. 43; Themistius Oration 26; Vatican Sayings 58; Ex 4:30-31; 5:21-23; fickleness was negative (Sall. Jug. 56.5). The ruling class usually could sway the masses (Saldarini 1994: 38), although deep-rooted popular convictions were no more easily removed then than today (Paus. 2.23.6; Jos. Ant. 13.298; 18.17). Contrast the motif of being well-liked by the people, who advocate one’s case, as a sign of one’s great nobility (e.g., Char. Chaer. 1.1.10; 1 Sam 14:45; Jos. Ant. 6.128; Life 303). Switching sides was sometimes a matter of political discretion, however (e.g., Jos. Ant. 15.193; Bowersock 1965: 85).

. Jn 19:13 necessarily occurs at the bema; capital sentences, unlike others, had to be pronounced from there (Blinzler 1959: 240, following Mommsen). It was natural to go outside if one were to speak with the multitude (Jos. War 2.172).

. Some also venture a chronological explanation for the transformation of the crowds, namely, the passage of six months between Jesus’ act in the temple (at Sukkoth) and his arrest (at Pesach); so Manson 1961: 79. Yet it seems inconceivable that the authorities would have allowed Jesus to survive in public six months after the act in the temple.

. Contrast the Roman people’s swearing (hence involving curses against themselves) specifically for an institution to be ordained forever (Dion. Hal. 6.89.4).

. See, e.g., the laments of Schnackenburg 1996: 263; Garland 1993: 258. Although guilt could be communicated transgenerationally (Ex 20:5; 34:7; Deut 5:9; 23:3-4; 1 Sam 15:2-3; 2 Sam 21:1-9), people were not to effect transgenerational punishments except under direct divine orders (Deut 24:16; 2 Kings 14:6; 2 Chron 25:4; Dion. Hal. 8.80.1).

. Although many contemporary scholars overplay the Gospels’ emphasis on “Jewish” guilt (e.g., Brandon 1967: 282), all four Gospels do emphasize the Jerusalem leadership’s responsibility more than that of Pilate. This apologetic works, however, because they present the leaders’ hostility as an internal Jewish dispute — which requires that the texts be read as the Gospel writers assumed they would be, Jesus and the Christians comprising a movement within Judaism. Cf. Reicke 1974: 184n.44 on the anachronism of the modern question: “it was Jews [Judeans] and Romans who killed him, not Israelis and Italians.”

. So, e.g., Albright and Mann 1971: 345; Matera 1989; Saldarini 1994: 33; Overman 1996: 383-84. Even the second century letter of Mara bar Serapion associates the exile of Israel with the condemnation of Jesus. Although not from a Christian (Boring 1995: 124-25), it undoubtedly testifies to a view circulated by Christians. Rabbinic texts often connect the suffering of A.D. 70 with judgment for sins, but never with Jesus’ execution (Cohn 1977: 219-20).

. Romans did not limit the number of lashes, hence sometimes victims not even sentenced to death died or were disabled under cruel supervisors (Cic. Verr. 2.4.39.85; Philo Flacc. 75); Josephus had opponents scourged “until their entrails were visible” (War 2.612) and reports a procurator laying bare a man’s bones, though the man survived (War 6.304; Blinzler 1959: 223).

. For coercitio as part of preliminary examinations, cf. Lake and Cadbury 1979: 282-83. Josephus adapts such discipline (Life 335).

. Independently attested by Jn 19:1, although the sequence there differs (Brown 1994: 852). Some argue that Jn 19:1 represents a lesser form of scourging (cf. F. F. Bruce 1980b: 15; Blinzler 1959: 224; Bammel 1984c: 440-41, though Blinzler and Bammel go too far in separating this from the crucifixion historically). Blinzler distinguishes forms of scourging thus: “an inquisitional torture [Acts 22:24; probably Jos. War 4.304], as a death sentence (fustuarium, primarily a military punishment [Hor. Sat. 1.2.41-42], as an independent police chastisement [P. Flor. 61; Jos. War 2.269; cf. Dig. 48.2.6; Philo Flacc. 75], and as the introductory stage to execution after the sentence of death [War 2.306, 308; 5.449; 7.200, 202; Livy 33.36]” (Blinzler 1959: 223). But these distinctions may have eluded Matthew’s audience anyway (Brown 1994: 851).

. E.g., Jos. War 2.306-8; 5.449; Livy 2.5.8; 9.24.15; 10.1.3; 26.40.13; 33.36; 41.11.8; Appian R.H. 3.9.3; Cic. Verr. 5.62.162; Dion. Hal. 3.40.3; 5.43.2; 7.69.1; 9.40.3-4; 12.6.7; 20.16.2; 20.17.2; Arrian Alex. 3.30.5; Klausner 1979b: 350; cf. Lucian Dead to Life/Fishermen 2.

. One could also be scourged while bound to the cross itself (Dio Cassius R.H. 49.22.6), though this may have worked best with the upright stake lacking a crossbeam. Victims were normally stripped naked before being beaten (Longus 2.14).

. On the shame of crucifixion, see further Brown 1994: 946-47, and especially Hengel 1977. The “equally negative portrayal” of Roman mockers here with Jewish mockers earlier invalidates the modern “claim that the Gospels excuse the Gentiles but condemn the Jews” (Brown 1994: 877).

. Matthew’s “whole” and “around” might also intentionally echo Ps 22:16 (Cope 1976: 103), though the correspondences are too few to be sure.

. Commentators cite Philo Flacc. 36-39; CPJ 154, 158; Plut. Pomp. 24; Dio Cassius R.H. 15.20-21; cf. also Winter 1961: 102-3. Robbins 1992: xxvi, 189 helpfully supplies another parallel from Persian behavior at the Sacian festival (Dio Chrys. Or. 4.67-70), but lays too much emphasis here to the exclusion of other parallels.

. Cary and Haarhoff 1946: 149; cf. in general Mart. Epig. 4.14; 11.6.2; Carcopino 1940: 250-53; Grant 1977: 82-83.

. Anti-Judaism was common in parts of the Greek East, especially Greek-speaking Egypt (e.g., CPJ 1:24-25; 2:36-55, §153; 3:119-21, §520; Philo Flacc. 1, 47, 85; Jos. Apion bk. 2; Sib. Or. 3.271-72), and also at times in Rome, especially in response to Jewish successes in attracting Roman converts (Hor. Sat. 1.5.100-101; Juv. Sat. 14.96-106; Quint. 3.7.21; Tac. Hist. 5.1-5; Persius Sat. 5.179-84; for more general Roman xenophobia, cf., e.g., Rhet. ad Herenn. 3.3.4; Cic. De Leg. 2.10.25). For more detail, see Whittaker 1984: 85-91; Sevenster 1975; Daniel 1979; Meagher 1979; Gager 1983.

. “Purple” could mean scarlet (e.g., Rev 17:4; 18:16; Appian C.W. 2.21, §150; Brown 1994: 866; cf. Dupont 1992: 260), though the Gospel tradition probably preserves it for its symbolic value, both to the soldiers and Jesus’ later followers. Egyptian gentry in nome capitals purchased green, red, and especially blue apparel (Lewis 1983: 52-53). Derrett 1990 improbably suggests an allusion to Is 1:18, but somewhat more likely (although the LXX employs a different term) also to Is 63:2, which depicts God coming in red.

. See, e.g., Blinzler 1959: 226-27; Jeremias 1971: 78; Lane 1974a: 559; Anderson 1976: 339; Hill 1979: 52; Carson 1984b: 573; Brown 1994: 866; citing 1 Macc 10:20, 62, 69; 11:58; 14:43-44. A person of status who loses prestige may then be insulted by cowards (Phaedrus 1.21.1-2).

. With Matthew’s “led away” here (27:31), cf. also Dion. Hal. 12.6.7, where one is apēchthēsan for crucifixion.

. Artem. Oneir. 2.56; Plut. De Sera Numinis Vindicta 9, Mor. 554AB; Char. Chaer. 4.2.7; 4.3.10; Stauffer 1960: 135; Reicke 1974: 185; Brown 1994: 913.

. That Simon was coming in from the “field” (Mk 15:21) cannot mean that he was coming from work — first, work was prohibited on the feast day (Lev 23:7-8) and, second, one would hardly finish working in the fields at 9 a.m. The expression may thus mean that Simon had miscalculated his transportation and arrived from Cyrene late for the feast (cf., e.g., Acts 20:16; Sen. Ep. Lucil. 53.1-3; 57.1; though the most difficult season — cf., e.g., Acts 27:9; Jos. War 1.279-80; 2.203; 4.499; Ach. Tat. 8.19.3 — had probably ended a month before); more likely, it simply means “from outside the city” (Dalman 1929: 100-101), where he had spent the night, perhaps in a nearby town.

. Brown 1994: 914 questions whether the Romans would force someone when Josephus says they did not force subjects to break their own laws (Apion 2.73), but this objection lays too much weight on Josephus’s propaganda; Josephus employs legal precedents apologetically (cf. Rajak 1984).

. Including a Jewish name in Acts 4:6; CIJ 1:lxvii; 1:59, §85; 1:149, §210; 1:157, §219; 1:288, §370; 2:27, §764; 2:249, §1217; 2:274, §1284; CPJ 1:xix; 3:168-69; in the feminine, CIJ 1:13, §8; 1:102, §144; 1:436, §606.

. E.g., 10:2, 4; 13:55. Fitzmyer (1974: 105-12) regards it as the most popular name of the period. The Greek “Simon” readily matches the biblical and later Jewish “Simeon” (cf. Cullmann 1968: 100; Dalman 1929: 5).

. Matthew changes the myrrh to “gall” to connote the taste depicted in Ps 69:21 (Blinzler 1959: 252-53n.27), possibly under the influence of an Aramaic wordplay (Gundry 1982: 569). The generic Hebrew and Greek sense of “gall” as covering whatever is poisonous or (in this case) bitter (e.g., Hom. Il. 16.203) may allow Matthew to make this change (Gundry 1975: 202). The righteous sufferer of Psalm 69 apparently represents Israel in exile (Ps 69:33-36), which is appropriately applicable to Jesus in Matthew (cf. 2:15-18).

. But pace Bammel 1984c: 442, it need not follow that “This detail… points to a Jewish execution”; actual Jewish executions were undoubtedly rare, and if some showed special mercy during a Jewish execution, how much more during a Roman one? Cf. the Christian allusion in Gk. Ezra 2:25.

. Because myrrh had multiple uses, others doubt that the tradition’s audiences would take the soporiferous effect for granted (so Brown 1994: 941) and interpret the action as part of the mockery, based on its bad taste (e.g., Carson 1984b: 575; Brown 1994: 943). Matthew’s “gall” would probably be less clear whether or not his audience would recall Markan tradition. In any case, noting that ancients sometimes mixed bitter substances with their wine (Pliny N.H. 14.19, §109), Brown 1994: 943 observes that “the Matthean picture is not absurd, even if its readers would recognize the action in the present context as hostile.” Myrrhed wine may also imply expensive scented wine (Pliny N.H. 14.15.92), perhaps seized from a well-to-do paschal pilgrim. That ancient Greeks were aware of narcotics is clear, e.g., in Hom. Od. 4.219-26; cf. Ruck 1978: 42.

. Thus, e.g., one man is bound to a fig tree and anointed with honey so that the ants devour him, but this, too, is called a cross (cruciatum); Apul. Metam. 8.22; cf. Prometheus’s fetters (Mart. Epig. 7; Lucian Prometheus 2). Thus positions varied, but for evidence for one probably common position, see Tzaferis 1985: 52-53. Before the Roman conquest, following hellenistic (e.g., Jos. Ant. 12.256) and Persian (Esth 9:25; De Vaux 1961: 159) practice, Jewish executions had also adopted hanging by crucifixion (e.g., Jos. War 1.97; Ant. 13.380; 4QpNah 1.7-8; Sifre Deut. 221.1.1; p. Sanh. 6:6, §2; cf. 11QTemple 64); though some read the later practice back into earlier times (Ps-Philo 55:3), however, the Israelites originally hanged corpses posthumously (cf. Gen 40:19) and only till nightfall, limiting the shame (Deut 21:23; m. Sanh. 6:4).

. See Artem. Oneir. 2.56; Plaut. Mostell. 2.1.12-13; Lucan C.W. 6.545, 547; m. Shab. 6.10; Lane 1974a: 564; Jn 20:25; cf. Lk 24:39. Cf. Diod. Sic. 25.5.2 (if prosēloō here means “nailed,” as it often does); also the skeleton recovered at Givat ha-Mivtar — F. F. Bruce 1980b: 18, though original reports about the ankle nail(s) have been revised (Stanton 1989a: 148; Kuhn 1978); on the wrists, see Yamauchi 1982: 2; Tzaferis 1985: 52.

. Artem. Oneir. 2.61; Brown 1994: 870 adds Dion. Hal. Rom. Ant. 7.69.2; Val. Max. Fact. 1.7.4; Jos. Ant. 19.270.

. Brown 1994: 870 thinks that the Gospels may “reflect a local concession,” noting that Jos. War 2.246 and Ant. 20.136 do not mention Celer’s disrobing; but this would be an argument from silence. (Brown, citing Melito of Sardis On the Pasch 97 in favor of nakedness and Acts of Pilate 10.1 in favor of a loincloth, ultimately doubts that we can know either way — 953.) Nakedness was probably the rule of thumb (in public Roman punishments, e.g., in Dion. Hal. 7.69.2; in non-Roman executions, e.g., in Jos. Apion 1.191; 2.53).

. Edersheim: 608 insists that ancient Judaism did interpret Ps 22 messianically (citing Yalq. on Is 60); cf. also Pesiq. R. 36:2; 37:1. Whether this interpretation existed before the time of Jesus is unclear (Longenecker 1975: 156 notes its use five times in 1QH to suggest that it may be messianic, but this is not absolutely clear), and certainly other interpretations existed (e.g., Midr. Ps. 22:6 applies to Esther — Bowman 1975: 136); in any case, though many parallels with Ps 22 in the Passion Narrative are noteworthy, they also correspond with what we genuinely know of crucifixion.

. Brown 1994: 955 notes that the law itself exempts the clothing the condemned is wearing, but acknowledges that such rules may not have been followed in the first century. I would add doubts that anyone would have restrained provincial soldiers from such seizure (especially given the abuses of requisitioning from persons not condemned).

. This practice stemmed from the custom of plundering the slain on the battlefield (cf., e.g., 1 Sam 31:8; Joel 3:2-3; 2 Macc 8:27-28; Virg. Aen. 11.193-94; Polyb. R.R.E. 9.26; Dion. Hal. 3.40.3; 3.56.4; 6.29.4-5; and throughout ancient literature).

. The location of the charge identifies the shape of the cross as in Christian tradition, rather than the T or X-shaped crosses also used; mass executions sometimes simply employed scaffolds (on various forms of crucifixion, Brown 1994: 948 cites, e.g., Sen. Consol. 20.3; Jos. War 5.451; on the four-armed cross, the crux immissa, see Iren. Haer. 2.24.4; Tert. Ad Nationes 1.12.7). The upright stakes were ten feet at the highest, more often closer to six or seven feet so that the man hung barely above the ground, with a seat (sedile) in the middle (Blinzler 1959: 249; Reicke 1974: 186); animals sometimes assaulted the feet of the crucified. The Romans could employ high crosses to increase visibility for significant public executions (Suet. Galba 9.1), and given the reed here (27:48; Mk 15:36), Jesus may have been slightly higher than usual (Blinzler 1959: 249); Brown 1994: 948-49 guesses seven feet.

. Boring 1995: 151-52 compares Acta Appiani 33 (second or third century A.D.), in which the martyr receives the mark of distinction “he claims, even if only mockingly.” Whereas many details of martyr stories may be relevant, however, this one is not; in view of the many acts of martyrs with which Jesus’ passion could be compared, a minor parallel involving such ridicule is easily enough coincidental.

. Cullmann 1956b: 42-43; Reicke 1974: 186; Brown 1994: 963 cite Suet. Calig. 32.2; Dom. 10.1; Dio Cassius R.H. 54.3.7; 54.8; Tert. Apol. 2.20; Euseb. H.E. 5.1.44; cf. b. Sanh. 43a. The posting of the accusation on the cross is not well-attested, either because those describing crucifixion had already mentioned its being carried (Bammel 1984b: 353) or because the practice was not in fact standard, although, given the variations among executions, in no way improbable (Harvey 1982: 13). Blinzler 1959: 254 thinks the tablet was “in black or red letters on a white ground.”

. Josephus usually uses “robbers” of revolutionaries (e.g., Jos. War 4.138); Jn 18:40 probably also uses it thus. The term can elsewhere retain its usual sense, of course (e.g., Jn 10:1; CPJ 1:157-58, §21; Ach. Tat. 2.16.2; 2.18.5; 3.9.3; 3.10.1; cf. Epict. Disc. 1.18.5; MacMullen 1966: 255). The conjunction of these terms via Josephus’s aristocratic language suggests support for the thesis of “social bandits”; see under Revolutionaries in the Introduction.

. Because crucifixions were intentionally public events, passersby were historically certain to be there. The Romans crucified near highways for optimum warning value (Dig. 48.19.28.15; Harvey 1982: 12; Brown 1994: 1026-27); those passing by would often learn of the grounds for execution from the other spectators.

. “Pass by” may here recall such sorrowful predicaments as in LXX Lam 1:12; 2:15 (Wilkinson 1978: 154). “Wagging heads” reflects Ps 22:7 and Lam 2:15 (so also, e.g., France 1985: 396).

. That members of the Sanhedrin would want to see that what they had set in motion was plausible, and Joseph of Arimathea’s testimony could strengthen the claim for a reliable source (Brown 1994: 1027-28). Brown doubts, however, that they would show up on Passover; certainly on the eve of Passover even the aristocratic priests would be occupied in the temple while lower priests slaughtered the lambs. But I have argued that this is the day after the slaughter of lambs, and would the priestly aristocracy have considered viewing an execution (which they had devised earlier the same day) to be “work” improper on a feast day? (Even by the third century rabbis were not unanimous about trying and executing someone on a Sabbath — p. Sanh. 4:6, §2.) In any case, however, we should not think that they were present long after being up most of the night.

. Gundry 1975: 203 cites Edersheim’s view that Jewish teachers cited Scripture loosely enough that a genuine citation of Ps 22:8 is possible; but Gundry’s other solution, adapting a different text to recount others’ perspective (ibid., following R. Gordis in HUCA 22), is more likely.

. Boring 1995: 157-60 rightly contrasts Jesus’ pain here with the standard traditions of glad martyrs (2 Macc 6:18–7:42; Jos. War 2.119-20; b. Ber. 61b), adding that Matthew’s signs, Luke’s psalm of trust (Lk 23:46 citing Ps 31:5), and certainly John alleviate some of the starkness of the Markan portrait.

. Gundry 1975: 210; Hurtado 1983: 264, and others are probably right that the Ps 22 and 69 allusions reflect a typological application to Jesus as the righteous sufferer. For a fuller discussion of the use of Ps 22 in the Passion Narratives, see Brown 1994: 1455-64. Augustine insightfully suggests that Jesus adopts the psalmist’s voice, identifying with humanity fully here (Letters, 140 to Honoratus 5-6; ACCS, 233-34).

. Cf., similarly, e.g., Hagner 1995: 844. But contrast the explicit Johannine emphasis that Jesus’ Father remains with him in his passion (Jn 16:32).

. The psalm portrays the righteous sufferer, not specifically identified as the Messiah (though cf. Soggin 1965 on corporate suffering of a king for his people). God could turn away from viewing the defiled (Philo Spec. Leg. 1.167). Translating the Aramaic “praised” instead of “forsaken” (cf. Cohn-Sherbok 1982) misses the context as well as demanding too much knowledge of early audiences concerning the Aramaic rendering of Ps 22:1.

. Given the practice of Jewish Scripture citations, Jesus may actually have quoted more than the sample preserved (cf. Stauffer 1960: 140).

. Rabbinic tradition, which could here reflect ideas already current in the first century, emphasizes that Elijah (who had never died in the OT — 2 Kings 2:11) would periodically rescue sages in distress or aid them in their needs (cf. b. ʿAbod. Zar. 17b; Taʿan. 21a; p. Ketub. 12:3, §6; Pesiq. Rab Kah. 18:5; Gen. Rab. 33:3; Dalman 1929: 205-6; cf. Bamberger 1969: 308).

. Laborers and soldiers in the ancient Mediterranean found wine vinegar better for quenching thirst than water (Ruth 2:14; Plut. Cato Major 1.13; P. Lond. 1245.9; Lane 1974a: 573; Blinzler 1959: 255). Reed mace (Typha) was probably not accessible or strong enough, but most true reeds (Arundo and Phragmites) would have served the purpose of the “reed” here (Hepper 1992: 70).

. Some used pennyroyal or mint stored in vinegar to revive those who had fainted (Pliny N.H. 20.54.152); but these were probably not available.

. E.g., m. Soṭa 9:15; p. ʿAbod. Zar. 3:1, §2; Pesiq. Rab Kah. 27:1; Gen. Rab. 59:4; 62:4; Dalman 1929: 220; with less spectacular signs for “ordinary” persons, see ARN 25A.

. Brown 1994: 1113-14 adds 4 Macc 9:20 (cf. the analogous Mart. Pol. 16.1); Jos. Ant. 17.167 (eclipse of the moon); b. Ber. 61b; ʿAbod. Zar. 18a; and on 1043, 1114 cites considerable Greco-Roman evidence as well (Virgil Georg. 1.472-90 — including earthquakes; Plut. Romulus 27.6; Caesar 69.4; Ovid Fasti 2.493; Cic. De Re Publ. 6.22; Pliny N.H. 2.30, §97; Dio Cassius R.H. 51.17.4-5; 60.35.1; Jos. Ant. 14.309; one may add Hom. Il. 16.459; Ovid Metam. 11.44-47). The Pythagoreans attributed earthquakes to “a mass meeting of the dead” (Thom 1994: 104-5), but the context cited above is closer. Consistent with his antisupernaturalism, Bultmann 1968: 282 regards all these as “pure novelistic motifs.”

. Also recalled in later texts: Pesiq. Rab Kah. 5:9; Ex. Rab. 14:3; Lev. Rab. 6:6; Num. Rab. 15:12; Song Rab. 2:13, §1; Pesiq. R. 17:7; cf. Rev 8:12; 16:10.

. Already objecting to the Gospel tradition some two decades after the event, one Thallus reportedly attributed the darkness to a coincidental eclipse, which Julius Africanus (ca. 221) disputes (Goguel 1948: 91-92; F. F. Bruce 1980a: 113). The darkness could be local (“the whole land”; Hill 1972: 354 prefers the “whole earth,” but all Mediterranean antiquity would have recalled a three-hour worldwide solar eclipse, and no early Christians could have gotten away with fabricating one). An unanticipated solar eclipse for three hours (most last a few minutes at most, like one eclipse for under two minutes in A.D. 29 — Brown 1994: 1041-42) is improbable, but the report suggests that Diaspora Christians were circulating this tradition of the Passion Narrative even before the early 50s.

. For this view, cf. Ps-Philo 44:6-8; m. ʾAbot 1:2, 18; b. Ned. 32a, bar.; b. Sanh. 113b, bar.; p. ʿAbod. Zar. 2:1, §1; Taʿan. 4:2, §13; Pesiq. Rab Kah. 19:6; Pes. Rab Kah. Sup. 1:11.

. While affirming the historical core of the signs tradition (1995: 750), Hagner 1995: 851 thinks that the resucitation of the saints makes better theological than historical sense. He suggests that the tradition may have stemmed from some tombs in the area opened by the earthquake.

. The Messiah apparently leads white-robed, resurrected saints into Jerusalem in the Dura Europos synagogue panel on Ezekiel (Grassi 1965a: 163; cf. also the earthquake in that scene; cf. Ezek 37).

. Some patristic and later interpreters reported these two earthquakes as if they were one (see Swart 1993). For adding earthquakes, cf. Jos. Ant. 6.27 (though cf. the list of Jerusalem tremors in Amiran, Arieh, and Turcotte 1994).

. Though the outer veil appears in postbiblical sources (see the evidence in Faber van der Meulen 1985), it surely existed in Herod’s temple; but would sources later than the earliest Jerusalem Passion Narrative know this? In my opinion, the specification of “the” curtain (27:51) strongly favors the inner curtain, the OT one; the outer curtain was less prominent than the barricade.

. Mark’s interpretation may represent a theophany, the unveiling of the hidden God near the cross; cf. Chronis 1982.

. As in Ezek 9:3; 10:4-18; cf. 2 Bar. 8:1-2; 64:6; 3 Enoch 5:14; Sifra Qed. pq. 8.205.2.1; par. 4.206.2.6; Sifra Behuq. pq. 6.267.2.6; Sifre Num. 1.10.3; Sifre Deut. 258.2.3; 320.2.1; ARN 34, 38A. For judgment on the temple here in Mark (cf. 13:2; 14:58; 15:29), cf. Hurtado 1983: 257; Kingsbury 1983: 132; Matera 1982: 139; undoubtedly Matthew continues this motif. Daube 1973: 25 seeks a parallel with Elijah, supposing that the rending of the veil responds to the crowd’s mockery; but the parallel is not obvious enough to carry much weight.

. Dibelius 1971: 195 cites the tradition that the temple doors miraculously opened forty years before the temple’s destruction (Jos. War 6.293-96; b. Yoma 39b; p. Yoma 43c; cf. Tac. Hist. 5.13), although he thinks the tradition circulated only after 70 (ibid. 195n.3; cf. Carson 1984b: 580; France 1985: 400n.2). Rabbinic legend also laments blood found on the veil after Jerusalem’s fall, though these may simply be stains from the Yom Kippur sprinkling (t. Kip. 2:16; cf. m. Yoma 5:4). Perhaps more relevant theologically, as one tore one’s clothes on hearing blasphemy or mourning the temple’s defilement (m. Sanh. 7:5; cf. b. Moʿed Qaṭ. 25b-26a; p. Moʿed Qaṭ. 3:7, §§7-8; Sanh. 7:6, §7), God rent the veil in response to the blasphemy involved in crucifying his Son (cf. 26:65). Some have also proposed a connection with the rending of heavens in 3:16 (on Mk 1:9-11; 15:38, cf. Motyer 1987).

. For a centurion supervising even an execution by a swordsman, see, e.g., Sen. Dial. 3.18.4.

. The emphasis on recognition in the cross remains clearer in Mark; Matthew’s signs (omitted in Mark) might actually lead a Gentile to recognize Jesus as a son of God (Boring 1995: 160 cites Plut. Agis and Cleomenes 39; Suet. Julius 88-89).

. “Truly” echoes 14:33 (which Matthew may have derived from the parallel to this text in Mk 15:39); “this was” derives from Mark (“this man was”) but on the Matthean level echoes Matthean redaction in 3:17; 27:37. Sim 1993 ingeniously suggests that the soldiers’ “confession” is actually only a recognition of their own sin and defeat, anticipating sorrow at the day of judgment; but this proposal misses the greater emphasis on the Gentile mission in Matthew.

. In Mark, it sends the same message as the rent veil, ending the messianic secret (F. F. Bruce 1972a: 21) and climaxing Mark’s probable christological introduction (1:1; F. F. Bruce 1980a: 45), associating Sonship with the cross. “Son” is not articular but (whatever the centurion’s intention) the evangelists hardly mean it in a watered-down pagan sense (Martin 1972: 131; Bratcher 1956); predicate nominatives often lacked definite articles (Metzger 1952: 125; in Diognetus, cf. Meecham 1952), though one wishing to emphasize the article would include it (Harner 1973: 87).

. The centurion in Mk 15:39 points to the conversion of Gentile Christians (Stock 1978).

. The centrality of the women and the guards closely connects this passage with the resurrection narrative in 28:1-15 (see Reeves 1993: 11).

. Cf. Liefeld 1967: 240: “A preacher who enjoyed the association of women was thus open to suspicion and criticism” (cf. Iren. Haer. 1.13.1, 3; 1.23.2.4; Lucian Runaways 18, on pp. 239-41). Perhaps because of their weaker ties to the established social order, women were often thought more susceptible to various religious movements (e.g., Dion. Hal. 8.39.1; Juv. Sat. 6.540-50); Greek culture also provided them more participation in the religious sphere (Gould 1980: 50). Women could support religious movements as patrons (cf., e.g., Pomeroy 1975: 200-201; Gardner 1986: 239-40; perhaps CIJ 1:384, §523; 2:10, §741; 2:20, §756); but their support tended to reflect negatively on those movements among their critics, including with early Pharisaism (Sanders 1993: 109; Ilan 1995).

. Strength and courage were typically masculine traits (Hom. Il. 20.251-551; Phaedrus 4.17.6). Ancient texts regularly praise women’s courage when it appears (e.g., Appian R.H. 2.5.3), but usually remark on how unusual it is (4 Macc 15:30; Arist. Pol. 3.2.10, 1277b; Dion. Hal. 4.82.3; 6.92.6; Diod. Sic. 5.32.2; 10.24.2; Livy 2.13.6; 28.19.13; Appian R.H. 7.5.29), or depict women’s courage as “manliness” (2 Macc 7:21; 4 Macc 15:23; 16:14; Diod. Sic. 17.77.1; 32.10.9; Apul. Metam. 5.22); conversely, cowardly men were taunted as “women” (Hom. Il. 7.96; 8.163; 11.389; 16.7-8; Virg. Aen. 9.617; 12.52-53; Dion. Hal. 9.7.2; 10.28.3; Diod. Sic. 12.16.1; 34/35.2.22; Aul. Gel. 17.21.33; cf. Hom. Il. 22.124-25; Aristoph. Lysistr. 98). “Courage” is literally “manliness” (e.g., 1 Macc 2:64; Arist. E.E. 3.1.2-4, 1228ab; Dio Cassius R.H. 58.4.6; Diod. Sic. 17.45.6; 40.3.6; Theon Progymn. 9.22; Crates Ep. 19; Char. Chaer. 7.1.8). On women as economic patrons (as here), see note above.

. Wire 1991 suggests that women were marginalized even in Matthew’s “scribal” community, but Perkins 1991 responds that Matthew’s community may not be wholly scribal, and objects to Wire’s use of Chinese examples to interpret Pharisaic/Matthean scribalism.

. The rabbis allow it in theory (e.g., m. Sanh. 6:3), but one incident proved so noteworthy that it recurs, in legendary form, in many later texts (e.g., Sifre Deut. 221.1.1; p. Sanh. 6:6, §2). Reported capital charges for women most often involved adultery (Ilan 1996: 159-62). Men were crucified far more frequently (e.g., Diod. Sic. 17.46.4; Jos. Apion 2.267).

. See, e.g., Witherington 1984; cf. Stanton 1974: 147; Harrington 1985; Borsch 1990; Kopas 1990; Schineller 1992; Knoch 1994.

. Matthew may connect a righteous Joseph protecting Jesus at the beginning and end of his Gospel (Davies and Allison 1997: 656). But even if Matthew had written his Gospel first and started the tradition about Joseph reported elsewhere in Gospel tradition, one should not think that he derived the name from the Joseph of 1:19–2:23 or that the tradition was borrowed by Joseph from Genesis (50:7-8). “Joseph” was a relatively common Jewish name in the Roman period (e.g., CPJ 3.182-83; in a Jerusalem ossuary, CIJ 2:281, §1291; see more fully Davies and Allison 1988: 182).

. Arimathea, from Ramathaim-zophim (1 Sam 1:1), was about 20 miles northwest of Jerusalem (Euseb. Onomasticon 32; Lane 1974a: 579); given the proximity, Joseph would certainly have been present at Jerusalem for the whole feast even if still residing in Arimathea; but the expression probably denotes merely his place of origin, allowing him to function in Jerusalem’s council. Insufficient information exists to demonstrate whether the high priests would have welcomed him to the trial (Anderson 1976: 350 thinks he belonged only to a local “council,” though this was certainly not the impression Mark wished to offer in 15:43), or why he would have preferred this tomb nearer to Jerusalem than to the residence of his ancestors.

. Washing the corpse was standard preburial practice in Mediterranean antiquity (e.g., Hom. Il. 18.345; 24.582; Virg. Aen. 6.219; 9.487; Eurip. Phoenician Maidens 1667; Ovid Metam. 13.531-32; Apul. Metam. 9.30; Acts 9:37), and anointing appears to be frequent as well (e.g., Hom. Il. 18.350-51; 24.582; Virg. Aen. 6.219; Mart. Epig. 3.12; Test. Abr. 20A; for ointments in embalming, Hagner 1995: 758 cites P. Oxy. 736.13; Artemidorus 1.5; Gen 50:2 LXX); for incense, e.g., Virg. Aen. 6.224-25; Ovid Metam. 2.626.

. Although the radiocarbon dating seems against it (Stanton 1995a: 119-20, noting the three independent carbon 14 tests, each claiming 95% certainty) and the colors are known from medieval artists’ pigments (cf. Thompson 1985: 238-43, who surveys both sides), traces of Palestinian plant fibers and early first-century Judean burial customs suggest elements of accurate portrayal in the Shroud of Turin. For a thorough and well-documented survey of scientific data for the latter, as well as scientific evaluations on the contamination of the radiocarbon sample, see Borkan 1995. If the Shroud dates from 1260 to 1390 as the radiocarbon tests suggest, it points to remarkable technology.

. In 1994: 1264-65, Brown argues that the Synoptics probably think of a single cloth whereas John has multiple wrappings.

. Taylor believes that the Markan chronology confirms the Johannine tradition here (1952: 601); Jeremias observes that one could buy necessary food for Passover even on the Sabbath, but pay later (Jeremias 1966a: 77; m. Shab. 23:1).

. Weeden doubts that women would seek to anoint a corpse decomposing this long (1971: 104), but see Craig 1981: 184: “In point of fact, Jerusalem, being 700 meters above sea level, can be quite cool in April” (cf. also Jn 18:18); the body remained in the tomb only a day and two nights, and “a rock-hewn tomb in a cliff side would stay naturally cool”; perhaps some aromatic spices had already been left with the body as well.

. Spices would diminish the stench and could be sprinkled on the bier or burned during the funeral procession (Meyers and Strange 1981: 97-98), but they were not used as preservatives. Jewish burials in this period did not seek to preserve the corpse; rather, they expected the flesh to rot off the bones for one year, after which the person responsible would inter the corpse in an ossuary (cf. m. Sanh. 6:6; m. Pesaḥ. 8:8; Moʿed Qat. 1:5; Meyers and Strange 1981: 28). For the use of spices at funerals, see Jos. Ant. 17.199; War 1.673; m. Ber. 8:6; Safrai 1974/76b: 776; among Gentiles, Herodian 4.2.8.

. Many Greek philosophers constituted notable exceptions (Epict. Disc. 4.7.31; Diog. Laert. 6.2.79; Stowers 1986: 142-43), though even their own disciples often disobeyed their instructions (Socratics Ep. 14; Diog. Laert. 6.2.78).

. In pagan thought, the dead could not enter Hades without burial (Hom. Il. 23.71; Virg. Aen. 6.365-66; cf. Lucan C.W. 1.11), and some thought unburied corpses more susceptible to abuse by witches (Lucan C.W. 6.626).

. If the Mishnah reflects general first-century Jewish practice here, Jewish courts granted criminals obscure burials in a common place, but then expected the gathering of the bones to the place of one’s ancestors a year later (m. Sanh. 6:6), meaning that the bones were kept track of even in the common place, not scattered (Brown 1994: 1209-11). Dishonorable burial was burial nonetheless (Jos. Ant. 4.202, 265).

. Brown 1994: 1207-8 shows that Justin. Dig. 48.24 reports Roman law as early as Augustus allowing relatives to bury the corpse, but refusing it for maiestas (treason); but he rightly observes that magistrates made these decisions themselves in the provinces (cf. Cic. Verr. 2.5.45, §119; Philo Flacc. 83-84). Brown doubts that Pilate would hand over the corpse if he admits the charge of maiestas, which he may not personally do (1208-9). The Jewish officials would probably not object to the burial, however (unlike later in Mart. Pol. 17.2; but even then bones might be handed over — Mart. Pol. 18.1-2), and without opposition Pilate was free to act as he pleased. He had settled matters adequately for the chief priests.

. Burying the dead despite prohibitions against this practice (Eurip. Suppliants; Demosth. Or. 60, Funeral Speech 8; Tob 1:17-20; 2:8) or in the face of danger (Dio Cassius R.H. 57.18.1) functions as a model of courage in ancient texts.

. A rock-hewn tomb might be particularly expensive (Hagner 1995: 858 cites Is 22:16 LXX). As noted above (Excursus on the burial), extravagant burials were reserved for the most noble and wealthy. For historical data on Jesus’ burial, see in greater detail Green 1992a.

. Hachlili 1980: 239; Hachlili and Killebrew 1983a suggest a window perhaps as narrow as A.D. 10-70 (cf. p. Moʿed Qaṭ. 1:5, §§4-5). But some evidence suggests a less significant use for more than a century later (Goodenough 1953: 1:114; cf. Rahmani 1982; idem 1986). For Jewish loculi in Rome, cf. Leon 1960: 59.

. That he was buried fits the culture as well as pre-Pauline tradition (1 Cor 15:4). It is quite difficult to account for the view attributed to Crossan (respected cochairman of the Jesus Seminar) by Ostling 1994 as cited in Craig 1995: 142, namely, that Jesus’ corpse was merely covered with a little dirt and probably eaten by wild dogs, a view that not only runs counter to the consensus of scholarship (no crime in itself) but appears to reflect inadequate familiarity with Jewish burial customs of the period. Even among Greeks, it could seem unthinkable that one would not at least have provided mass graves to enemies slain in battle (Paus. 1.32.5); dogs and birds normally ate the unburied (Hom. Il. 22.42-43, 71-76, 335-39, 353; 23:21; 24:211, 411; Od. 21.363-64; 22.476; Soph. Antig. 697; Eurip. Phoenician Maidens 1650; Virg. Aen. 9.485; Lucan C.W. 7.825-35; 2 Sam 21:10; in a crucifixion, see Llewelyn 1998: 1-3, §1), but as I have noted, even among pagans public sympathy often favored burial and Jews would require it.

. Tomb architecture changed radically after Jerusalem’s fall (Goodenough 1953: 1:84-89), and the Protestant site’s skull shape postdates the first century (Brown 1994: 938-39).

. On the latter, see Brown 1970: 899; idem 1994: 1279-83; cf. Blinzler 1959: 251-52; Smith 1967; Ross 1976; Riesner 1985.

. See, e.g., Brown 1994: 1282; cf. Blinzler 1959: 251-52; for archaeological data, see the notes in Josephus 1982: 338-40, on Jos. War 5.148-55.

. That they call Pilate “Lord” or “Sir” (26:63) is to be expected, but the contrast between this title that suppliants apply to Jesus (e.g., 8:2, 6, 8, 21) and their portrait of Jesus as a deceiver (27:63-64) may also sustain the conviction that many Judeans after 70 might share, that the priestly aristocracy were too beholden to the Romans.

. The opposite suggestion of France 1985: 409, that a fiction created too long after the event for anyone to refute it would have little apologetic value, is equally weak.

. The chief priests would not need Pilate’s approval merely to station their own guards (Gundry 1982: 584; France 1985: 405; cf. Brown 1994: 1295). Using Roman guards rather than their own would leave the Jewish authorities less open to the charge of tampering with evidence; for Matthew, this conveniently provides another example of collaboration between the Jerusalem authorities who opposed Jesus and the Romans.

. Brown 1994: 1301-10 contends for a pre-Matthean popular story about the guards that Matthew has interwoven with the story of the women at the tomb; he believes that the Gospel of Peter, which uses the same story, includes more oral development. Meier 1991a: 117 finds in Gos. Peter 30 the same phrase as in Mt 27:64b and notes that “disciple” is particularly Matthean (73 times in Matthew; 2 times in the Gospel of Peter; not at all in Crossan’s supposed Cross Gospel) — strongly suggesting the Gospel of Peter’s direct dependence on Matthew here.

. It already existed in other areas, such as the northern coast of the Black Sea (Blawatsky and Kochelenko 1966) or farther to the east (cf. Cumont 1975; Francis 1975); it later spread widely in the Roman army (Daniels 1975; Gager 1975: 134; Serban and Baluta 1979; Koester 1982: 1:372-74; Burkert 1987: 7, 42), but even then it remained limited to particular parts of the Empire (Frank 1932: 49-50; Nock 1937: 113; Daniels 1975: 273; Bianchi 1979: 879).

. See, e.g., Jos. Ant. 4.219; m. Yebam. 15:1, 8-10; 16:7; Ketub. 1:6-9; t. Yebam. 14:10; Sifra V. D. Deho. pq. 7.45.1.1; cf. Lk 24:11; Keener 1992: 162-63; Baumgarten 1957; Hooker 1983: 119. Some, though not all, condemned listening to women more generally (e.g., Jos. Ant. 18.255; Syr. Men. Sent. 118-21, 336-39; Maccini 1996: 63-97 rightly notes that the responses are more diverse than in legal contexts). Ilan 1996: 227 thinks that in practice the non-Pharisaic legal system “often” required women’s witness; even if this is overstated, women could testify concerning various matters, and some views of 1QSa 1.10-11 suggest that Qumran was more open to the practice than the Phari-sees were (Ilan 1996: 163-66).

. Hesiod W.D. 375; Livy 6.34.6-7; Babrius 16.10; Phaedrus 4.15; Avianus Fables 15-16; Justin. Inst. 2.10.6 (though contrast the earlier Gaius Inst. 2.105); Plut. Publicola 8.4; Gardner 1986: 165; Kee 1980: 89. Many men regarded women as gullible (cf. Philo Prob. 117; Juv. Sat. 1.38-39).

. Corpses were used for magic (Lewis 1983: 96), and people suspected that witches sometimes stole bodies for magic (see especially the tale of Telephron in Apul. Metam. bk. 2; in other cultures, e.g., Mbiti 1970: 261); corpses that died violent deaths were considered particularly potent for magic (PGM 1.248-49; 2.49-50; 4.342-43, 1390-95, 1402-3, 2211-17; 57.5-6; 58.5-9; 67.21; 101.1-3; Ap. Rhod. 4.51-53; Lucan C.W. 6.538-68). If Jesus’ enemies considered him a magician (12:24), some Jewish leaders may even have anticipated the theft of the body as in 27:64. In less severe cases, tombs generally settled for divine threats against robbers (e.g., IG 3.1417 in Grant 1953: 9). Both tying rope from a cross (Pliny N.H. 28.11.46) and iron that had been pounded through the hands (Lucan C.W. 6.545-47) were used in witchcraft (as a superstitious cure in m. Shab. 6:10).

. Dunn also notes the tradition that stresses Peter first (1 Cor 15:5; cf. Lk 24:34), which Farmer and Kereszty 46 regard as a pro-Petrine tradition.

. The sudden reintroduction of Pharisees in Matthean material charging Jesus as a “deceiver” (27:62-64) undoubtedly fits a charge of Matthew’s time (Stanton 1993: 178-79). Their knowledge of his resurrection claim presumably alludes to 12:40 (likewise involving Matthean redaction; see Stanton 1993: 82).

. The subtext of the Iliad, probably the most widely circulated literary work in the ancient Mediterranean, was that male warriors were fighting because of women, such as Helen and Briseis; cf. especially Il. 9.339-42.

. Lachs 1987: 436 notes that “evening” can mean “about evening” and that burial on the day of death was essential (b. Moʿed Qat. 28a; Sifre Deut. 221).

. Noting the tradition in Semahot 8:1, Longstaff 1981 suggests that relatives often observed the tomb for three days to be sure the person was genuinely dead (cf. Safrai 1974/76b: 784-85; but could the practice have arisen from anti-Christian polemic?); given the brutality of crucifixion, however, this should not have been necessary. Final acts of devotion the next morning would not be uncommon (Char. Chaer. 3.3.1; cf. t. Shab. 17:19 in Safrai 1974/76b: 774; Pesiq. Rab Kah. 11:23).

. Some features of Jesus’ appearances may also be theophanic (Smyth 1975, although some of his material, like the Moses parallel, may be overstated). Some ancients also had nonsupernaturalistic explanations for lightning (Stoics in Diog. Laert. 7.1.153). But all the canonical accounts are restrained by comparison with second-century apocryphal accounts (e.g., Gos. Peter 9:35–11:44; Carson 1984b: 588).

. White also signified joy (p. Rosh Hash. 1:3, §27; cf. Gregory the Great Homilies 21; ACCS, 243) and black mourning (Jos. and Asen. 10.8-9/10; 14:12; Aristoph. Frogs 1337; Dupont 1992: 260; Ovid Metam. 8.777-78; Herodian 4.2.3; but cf. Plut. R.Q. 26, Mor. 270DE) or death (Hom. Od. 11.32-33), hence the burial clothes of the righteous (Ps-Philo 64:6; cf. Test. Abr. 20A; Life of Adam 48.1; Apoc. Mos. 40.1-3; b. Ber. 18b); black is negative in Hom. Il. 1.103; Marc. Aur. Med. 4.28.

. Ephraim Isaac, an Ethiopian translator of 1 Enoch, points out that in 1 Enoch 87:2 white suggests the image of purity in Ethiopic (OTP 1:63n.) Against some modern assumptions, these associations with color derive from day/night divisions (e.g., night magic in Lucan C.W. 6.624; the netherworld in Hesiod W.D. 154-55), not human pigment (e.g., Philo Creation 29 views air as black without light). White is associated positively with the spirit world in various traditional African societies (Mbiti 1970: 73, 277; Isichei 1995: 64).

. Some details, like the invitation to “see the place where he lay” or “they laid him” (functioning to bolster the eyewitness claim), are essentially the same in both (28:6; Mk 16:6); Greek regularly used keitai for lying in a tomb, often in Diaspora Jewish tomb inscriptions (e.g., CIJ 1:8, §4; 1:12, §§6-7; 1:14, §§10-11; 1:15, §§12-13; 1:16, §§14-15; 1:17, §17; 1:19, §20; 1:21, §23; 1:23, §28; 1:24, §30; 1:26, §35; 1:30, §42; 1:31, §45; 1:32, §§46-47; 1:35, §§51-52; 1:36, §53; 1:37, §§55-56; 1:38, §58; 1:39, §§62-63; 1:49, §78; 1:52, §79; 1:56, §81; 1:60, §86; 1:62, §88; 1:66, §93; 1:69, §97; 1:70, §§99, 100; 1:74, §105; etc.).

. The literary connection some draw between Mk 14:50-52 and 16:5 (e.g., Jenkins 1983) lacks adequate foundation (Brown 1994: 299-300).

. The sudden ending of Mk 16 with 16:8 may be original, however, fitting some ancient narration patterns; though in some cases, e.g., Ps-Philo, the ending may be lost, one may also compare abrupt original endings, e.g., in some of Plutarch’s speeches (Fame of Athenians 8, Mor. 351B; On the Fortune of Alex. 2.13, Mor. 345B; Fortune of Romans 13, Mor. 326C; Uneduc. Ruler 7, Mor. 782F); Isoc. Demon. 52, Or. 1; Demetrius On Style 5.304; Lucan C.W. 10.542-46; Herodian 8.8.8. See especially Magness 1986 for more ancient literary parallels; for consistency with Markan style, especially a final gar, cf. Boomershine and Bartholomew 1981; Mus. Ruf. 12 (van der Horst 1974: 308). An abbreviated conclusion allows Mark to retain the centrality of the cross without actually playing down the resurrection (cf. also Thompson 1985: 225). (Mark nevertheless points to resurrection appearances beyond his narrative; e.g., Anderson 1976: 353; Rhoads and Michie 1982: 42; Hooker 1983: 120; pace Weeden 1971: 44-45; Kelber 1979: 87.) Farmer 1974 even makes a noteworthy case on external (3-75) and internal (79-103) grounds that Mk 16:9-20 has more support for being the original ending than often accepted; though it runs counter to scholarly consensus, his thesis is well-supported and worthy of more attention than it has received.

. Raphael reveals himself and says “Fear not” in Tob 12:17; for other parallels, see comment on 17:7. In this passage, the women not only should not fear but should receive news of great joy (28:8), which fits discovery of or entrance into the kingdom (13:44; 25:21, 23; cf. 2:10).

. “Worship” or “prostration” could be offered not only to deities or spirits (e.g., PGM 13.704-5; to God in 3 Macc 5:50) but to humans in this period, not only by Gentiles (e.g., Polyb. R.R.E. 15.1; Char. Chaer. 5.2.2) but by Jews as well (Ruth 2:10; Jos. and Asen. 5:7; Test. Jos. 13:5); but at this point in Matthew, the term clearly connotes worship of Jesus as divine (cf. 1:23; Jn 9:38).

. Eastern Christian tradition regarded Mary Magdalene as isapostolos, “equal to the apostles” (Davies and Allison 1997: 637). That both the angel (28:7) and Jesus (28:9) offer the same commission to Galilee probably represents the weaving together of two traditions (not necessarily challenging the reliability of either), though the second could represent pro-Galilean redaction of tradition.

. Weeden 1971: 102 is among those who doubt that the empty tomb tradition precedes Mark; his claim that there is no “hard evidence that the early church ever knew of Jesus’ grave’s being empty” suggests that it did not occur to him that anyone would have checked — an idea as unlikely in Roman antiquity as today. Boyd 1995: 275 rightly questions whether Mark was inventing 16:1-8 as apologetic — aside from pre-Markan Semitic expressions in the passage, its conclusion with the women’s fear and silence is hardly apologetic, and it lacks mention of corroborating attestation from Joseph of Arimathea or others.

. Likewise, against the unanimous witness of extant evidence, from earliest to latest, he supposes that the resurrection was a late myth originated by Christians not in Jewish Palestine but in northern Syria and Asia (1993: 2). Evidence for early tradition for the site of the tomb (see comment on 27:57), the largely Palestinian evidence for Jewish belief in the resurrection, the extreme unlikeli-hood of a Diaspora movement becoming more Palestinian or Judaized in the anti-Judaism of parts of the Greek East, and so forth render his suggestion incredible.

. Deities periodically “manifested” themselves to mortals in Greek tradition, sometimes in sleep and sometimes as apparitions (e.g., Eurip. Bacch. 42, 53-54; Ael. Arist. Or. 48.41; Apul. Metam. 11.3; Ach. Tat. Clit. 7.12.4; Char. Chaer. 2.2.5; 2.3.5; reports in Grant 1953: 9-13, 123); in unrelated cultures, see Wolf 1958; Mbiti 1970: 105-12 passim; for more concrete effects of angelic manifestations in hellenistic Jewish tradition see Tob 12:19, 22; 2 Macc 3:24-26 (cf. God in 2 Macc 3:30). Paul’s language in 1 Cor 15 applied in the LXX especially to revelations of God or angels (cf. Bar. 3:37; Sib. Or. 1.200; Bartsch 1980). But from the late hellenistic age “epiphanies” of Greek gods usually meant the activity of a deity rather than its appearance (Nilsson 1948: 106; Diod. Sic. 5.62.4; 11.14.3-4; Dion. Hal. 8.56.1-3; it is primarily these that witnesses attest — cf. Grant 1986: 66, 54-55, 64-65), though appearances in personal dreams and visions occur (e.g., PDM 14.74-9195, 98-102, 169); further, Jesus “appeared” to his followers in Acts 1:3 but there provided concrete proofs of his physicality (cf. Lk 24:39-40). Finally, Paul himself distinguishes between the Easter appearances and mere visions (cf. 1 Cor 9:1; 15:8; 2 Cor 12:1-4; Goppelt 1980: 18-19).

. See, e.g., Dibelius 1971: 18-20; Gerhardsson 1961: 299-300; Barrett 1967: 1-2; Conzelmann 1975: 251; Hunter 1961: 15-17; Fuller 1971: 10-11; Webber 1983; Fee 1987: 722.

. Nor is there historical merit to the old “swoon” theory (that Jesus was not yet dead, hence revived enough to act “resurrected” but then died somewhere unknown); crucified persons did not simply revive: Josephus had three friends taken from crosses, and despite medical attention, two died (Jos. Life 420-21). Apul. Metam. 10.11 cites a drug to simulate death (cf. also Diog. Laert. Lives 8.2.61), but his book is full of magic herbs that can do almost anything, here accommodating the story line (cf. the similar plot device in Ach. Tat. 3.15-21; 5.18.2; 7.6.2). Further, if one could revive, one would still be trapped within the tomb, which would lead to death (Char. Chaer. 1.4.11-12; 1.8).

. Conditions in first-century Judea and Galilee were not those that produced the seventeenth-century Messiah Sabbetai Zevi, many of whose followers failed to be deterred by his apostasy (Grayzel 1961: 516; Bamberger 1962: 240), and some even by his death (Scholem 1973: 920; Greenstone 1906: 225-30). Aside from different social conditions, knowledge of the Christian belief in Jesus’ resurrection and redefinition of Messiahship could provide later messianic movements a model for redefining the messianic mission in a manner that did not exist prior to Jesus.

. Some less than persuasive parallels could be adduced. Jos. War 6.297-99 reports that people saw heavenly chariots moving through the clouds and surrounding cities (cf. 2 Kings 6:17; 2 Macc 3:24-26; 4 Macc 4:10-11; Sib. Or. 3.805-8), and priests heard voices in the temple; Horsley and Hanson 1985: 182-84 regard these as collective fantasies, but they could also be (1) true (which we regard as unlikely but which a Post-Enlightenment perspective cannot simply dismiss) or (2) the sun playing tricks on eyes at dusk or (3) borrowed from Flavian propaganda (Saulnier 1989) or (4) Josephus’ own propaganda (he is the only extant witness concerning witnesses apart from sources dependent on him). Other reports of such prodigies are almost standard and stereotypical (e.g., Appian C.W. 4.1.4; Livy 24.10.10; 24.44.8), and often include the warning that imagination may be involved (e.g., Herodian 8.3.8-9) or that the apparitions remained only in the distance (Livy 21.62.5); some opined that such reports tended to circulate and multiply among the superstitious (Livy 21.62.1; 24.10.6; 29.14.2). Some reports imply celestial phenomena we would consider natural (Livy 25.7.8; 29.14.3; 32.8.2; 41.21.12-13; 43.13.3), and the most striking accounts of such apparitions are poetry recycling literary motifs and citing no witnesses (Lucan C.W. 1.572-73). In any case, this phenomenon is quite different from meeting again a person one had personally known, which the Gospel accounts stress (probably authentically, though developed further in later tradition; see, e.g., Epistula Apostolorum passim).

. For a fuller defense of the empty tomb traditions, see Craig 1981; idem 1985; idem 1995: 146-52; Ladd 1963; on the bodily character of the resurrection, see Craig 1980: 47-74.

. Ancients also recognized that willingness to die proved one’s commitment to what one testified and that it did not spring from ulterior motives (e.g., Dio Cassius R.H. 63.11.2–12.1; 58.4.5-6). Josephus cites Jews’ willingness to die for the law (Apion 1.42-43).

. Admitting historical evidence favoring Jesus’ resurrection is not purely the domain of Christian apologetic; e.g., orthodox Jewish scholar Pinchas Lapide, while doubting that the resurrection proves Jesus’ messianic or divine identity (connected though this has traditionally been to the resurrection; see Rivkin 1971: 398), nevertheless finds the evidence for his resurrection compelling (see the review in Kennedy 1985). Without addressing Jesus’ resurrection appearances, Vermes 1973: 41, another Jewish scholar closely acquainted with the primary evidence, opines that “the only conclusion acceptable to the historian” must be that the women actually found the tomb empty. “Those scholars who are unable to believe in an actual resurrection of Jesus admit that the disciples believed it” (Ladd 1974: 320).

. Willoughby 1929: 225-62 tries to compare Philonic language with the conversion language of the Mysteries, but like Godwin 1981: 78-83, he tends to generalize too much. More nuanced is the approach of Wolfson 1968: 1:27-36, cf. 1:101 (Philo adapts their language but denounces them as religious alternatives).

. Others have pointed to similarities between Mithraism and Christianity (cf. Gervers 1979, though qualifying on 598; cf. Gager 1975: 132-34; note the contrast stressed by Mattingly 1967: 5). For example, Deman 1975 notes the later link between the twelve apostles and the twelve signs of the zodiac; but the twelve apostles in earliest Christian tradition stem from the twelve tribes (though Judaism had already linked the tribes with the zodiac in that period). The closest true parallels address only later Gentile Christianity, as it assimilated into the guise of a broader Roman cultural context.

. Manson 1963: 64-65 stresses the moral contrast between the Mysteries (where moral ideals were irrelevant) and Christianity (cf. Carcopino 1940: 138-39).

. Cf. Nock 1933b: 136 for Christianity’s “Oriental” nature but lack of “Oriental” trappings. This is not to suggest that many other Greco-Roman cults could not be distinguished from one another, but rather to point out that the originating cultural matrix of Christianity was different enough, and earliest Christianity’s monotheism rigorous enough, to disallow the degree of assimilation that could characterize most of the cults.

. E.g., Ovid Metam. 5.564-71. Burkert argues that Persephone’s connection with the nature cycle must go back to pre-Greek, perhaps Neolithic times, because the real facts of Mediterranean vegetation suggest an interpretation earlier than the one the Greeks themselves held (Burkert 1985: 160). Whether or not his argument is accepted as persuasive, it is clear that Persephone’s return from the underworld precedes the apostolic proclamation of Christ’s resurrection by many centuries.

. Some have likewise claimed that Marduk died and rose again in some sense (Klausner 1979a: 103, though I have not noticed this in Enuma Elish in Heidel 1951: 18-60).

. The Adonis tradition itself was Semitic and imported into Greek religion from an early period (Burkert 1985: 176-77 thinks perhaps as early as the sixth century B.C.). Apart from the resuscitation it is clearly early; in the Greek bucolic poets, cf., e.g., Women at the Adonis Festival (third century B.C.), a lament for Adonis perhaps by Bion, and The Dead Adonis (Greek Bucolic Poets, LCL 176-95, 386-95, 480-83).

. E.g., Hom. Il. 5.382-404; 5.388; on the death of Pan in Plut. Mor. 419.17, see Borgeaud 1983. Dionysus returns in the spring in poetic fragments (ca. 1 B.C.) in Select Papyri 3:390-93.

. Some utterly unrelated cultures also supply examples of resurrection legends (e.g., the Sonjo myth in Mbiti 1970: 251), although without the sort of historical attestation surrounding the case of Jesus. But given the transcultural interest in life after death, one need not suppose an organic connection among all such accounts except when they are geographically close and the story line is substantially similar.

. Cumont’s view of astral immortality (Cumont 1922: 91-109; cf. Reitzenstein 1978: 64-65; Dahl 1977: 17; Avi-Yonah 1978: 40-41) is much broader than the Mysteries and thus should not be directly linked to them (Gasparro 1985: 98). On regeneration in the Mysteries and Christianity, see Keener 1997b: 143-46. The doctrine of bodily resurrection also appears in the Hebrew Bible earlier than it is attested in Persian texts (Yamauchi 1990: 456-57, 461; cf. 409; for immortality, however, cf. Olmstead 1959: 40, 100-101).

. Though the language appears in some Jewish sources (e.g., Philo Mos. 1.279), the same authors advocate monotheism, and other Jewish texts may polemicize against divinization (Jub. 3:19; 10:20, following Genesis; Ep. Arist. 211, 263; Sib. Or. frg. 1.1-2; especially Philo Virt. 172; Ex. Rab. 8:2; cf. Urbach 1979: 1:252). The second-century church, like Gnosticism, came much closer to accommodating such language than first-century Christians could (cf. Tatian 7; Taylor 1945: 206, citing Irenaeus and, more to the point though later, Athanasius).

. Mack makes Jesus’ resurrection purely mythical (1988: 112-13) by wrongly equating immortality in the Wisdom of Solomon with “resurrection” in 2 Maccabees, wrongly interpreting eschatological narratives about Christ’s resurrection as if they were eschatological allegory, and wrongly taking the Spirit in a purely hellenistic sense instead of its Jewish usage in the NT (on this, see Keener 1997b: 7-8 with ibid. 8-13).

. Rabbinic texts often emphasize that the Sadducees, unlike the Pharisees, denied the teaching and hence held no place in the coming world (e.g., m. Sanh. 10:1; ARN 5A; 10, §26B; cf. b. Sanh. 90b). The doctrine of the resurrection was particularly relevant in the context of martyrdom (2 Macc 7:9, 11; 14:46); those inclined to defend the honor of martyrs hence would take serious offense at the denial (rabbinic texts also suggest moral consequences for denying resurrection and judgment, which they viewed together).

. The supposed resurrection of the Teacher of Righteousness is based on inference from a reconstructed text (cf. 4QpPs 37 fr. 2.2-4 in Dupont-Sommer 1973: 272), which other scholars have reconstructed quite differently.

. This is true though Josephus, adapting his depiction of Jewish “sects” to Greek schools like Pythagoreans and Middle Platonists, portrays Pharisaic confidence in more acceptable hellenistic terms suggesting reincarnation (Jos. Ant. 18.14; War 2.163; 3.374; Apion 2.218).

. They condemned a few others for its denial besides explicit Sadducees (e.g., p. Sanh. 10:2, §11). Other texts regularly defend the resurrection long after the Sadducees themselves had ceased to be an issue (e.g., Lev. Rab. 27:4; Lam. Rab. 3:23, §8), but that the rabbis would engage in “textbook apologetics” (not uncommon in some more traditional religious circles today) would not be surprising, given the variety of hypothetical legal situations they also surveyed.

. Cf. also the graffito in Greek at Beth She’arim: “Good fortune in your resurrection” (Finegan 1969: 208). Ossuaries belong especially to the Roman Imperial period and the pre-Israelite Chalcolithic period (see Silberman 1991).

. Some evidence exists in contemporary Egyptian Judaism, but Philo himself never mentions the doctrine (Wolfson 1968: 1:404). The Samaritans may well have accepted it, though our evidence here is late (see McDonald 1964: 376).

. One cannot cite the widespread use of crosses on early ossuaries, which probably are simply markings for the placement of the lids (R. H. Smith 1974). Is Gustafsson 1956 more helpful?

. Boring 1995: 163-64 cites Romulus’s apotheosis appearance to Proculus Julius in Livy 1.16.2-8; Plut. Romulus 28; Numa 11.3; Ovid Fasti 2.500-9, and notes that Justin 1 Apol. 21 made an apologetic comparison between Jesus’ resurrection appearances and pagan understanding of imperial apotheosis.

. Because of Heracles’ apotheosis people searched only vainly for his corpse (Diod. Sic. 4.38.3-5); other deified persons like Aeneas also “disappeared” (ēphanisthē, Diod. Sic. 7.5.2), as did Moses in Jos. Ant. 4.326. Boring 1995: 162-63 also compares the first-century-B.C. traditions of Romulus’s ascension (Ovid Metam. 14.805-51; De Viris Illustr. 2.13; Plut. Numa 11.2), even by horses and carriage (Ovid Fasti 2.475-510; cf. 2 Kings 2:11-18), and Job’s children in Test. Job 39:8– 40:4.

. Departing from Mark, Matthew exercises more literary freedom in the narrative about the guards, which he alone includes; the language is particularly Matthean (Fenton 1977: 451). Four times in the Gospel Gentiles use the term “Jews” here used by Matthew (28:15; Saldarini 1994: 34-35); but Matthew, like Josephus, applies it to part of his people, often applying “the gentile designation ‘Jews’ for that part of the people he opposes” (Saldarini 1994: 36); Jews regularly call other Jews by this title in Diaspora inscriptions (Overman 1996: 401).

. I do not deny that some may have conceived of various forms of afterlife in different terms; but the term “resurrection” indicated a bodily phenomenon.

. Other factors also militate against supposing that the disciples stole the body. Jewish scholar Geza Vermes notes (1973: 40) that “From the psychological point of view, they would have been too depressed and shaken to be capable of such a dangerous undertaking. But above all, since neither they nor anyone else expected a resurrection, there would have been no purpose in faking one.”

. For keeping watch over corpses, often to prevent burial, see also Soph. Antig. 217; Plut. Agis and Cleomenes 39; cf. 2 Sam 21:10. Guards who fell asleep could be scourged or beheaded (Eurip. Rhesus 812-19, 825-27). Roman soldiers who fell asleep during night watch were normally beaten; if a soldier survived, he was banished from family and country and deeply shamed (Dupont 1992: 126). One could sometimes strike deals, however, and some careless sentries escaped execution (Tac. Hist. 5.22; Brown 1994: 1311, though he does not argue for historicity). The punishment for falling asleep at night watch was so dreadful that some might desert to escape it (Assyrians in Diod. Sic. 2.18.8), though Roman custom also made desertion of one’s post a capital offense (Polyb. R.R.E. 1.17; Livy 24.37.9; similarly, for desertion in battle, cf. Dio Cassius R.H. 48.42.1-2; Dion. Hal. 3.30.7; 6.9.4; Livy 30.43.13; though less severely, Appian R.H. 7.7.43; peoples who showed mercy to deserters were apparently exceptional — Diod. Sic. 1.78.1; 12.16.1).

. Roman guards from whom a prisoner escaped could be executed (Justinian Cod. 9.4.4; Lake and Cadbury 1979: 139; cf. Acts 16:27-28; 27:42), and while the priests would not execute guards (Acts 5:22-24), some Jewish authorities followed that stricter practice (Acts 12:19). Later tradition reports that the chief of the temple police was authorized to beat or even set aflame any watchmen found sleeping (m. Mid. 1:2, where one rabbi testifies that the latter actually happened to his uncle; b. Tamid 27b-28a). Beare 1981: 543 is correct that Roman guards would not normally first report to Jewish officials, but Gundry rightly responds that they would do so in this extraordinary circumstance if Pilate had placed them under the priests’ supervision on this assignment (Gundry 1982: 592).

. Because sleeping on duty was a punishable offense, one would doubt that the guards normally would publicly admit it (Beare 1981: 543); but because the missing body could spell their execution either way, they have reason to please the authorities, who promise conditional amnesty (Gundry 1982: 585). Further, Matthew emphasizes that the authorities paid a substantial sum of money; the narrative probably means a “large” sum, but whether it refers to large (heavy, hence expensive) coins, cf. Thompson 1962: 310. “Persuading” Pilate could easily mean bribing him (Embassy to Gaius 38; Gundry 1982: 593), though if high priests used temple funds rather than their own it would represent serious hypocrisy (Jos. Ant. 18.60-62; War 2.175-77).

. Sanders 1993: 278 suggests that the disciples “fled to Galilee and then returned to Jerusalem”; Luke-Acts is Jerusalem-centered, so Luke had reason to omit the Galilean pilgrimage. Marxsen 1969: 111-16 thinks that Mk 16:7-8 (the rendezvous in Galilee — 28:10) refers to the parousia rather than resurrection appearances in Mark, but this is hardly likely (cf. Lüdemann 1980: 166-67).

. Michel 1995: 49 emphasizes totality in the closing lines: “all” authority (28:18), nations (28:19), Jesus’ precepts (28:20), always (28:20).

. Parkhurst 1979 suggests that the object of doubt is the high Christology implied in worshiping him (for another nontraditional reading cf. Grayston 1984); while this makes sense of the following context, it lacks sufficient clarity in its immediate context. In any case, ancients viewed skepticism with regard to deities or holy men as potentially dangerous (b. B. Bat. 75a; Pesiq. Rab Kah. 18:5) or at least humiliating (Epidauros inscr. 3-4 in Grant 1953: 56-57); doubtlessness was virtuous (Sifra Shemini Mekhilta deMiluim 99.5.13).

. Most scholars recognize the influence of Dan 7 here (e.g., Fuller 1971, also comparing Mt 11:27; Jn 3:35; Michel 1995: 45-46; Ellis 1974: 22; Meier 1980: 369; Garland 1993: 267); despite notable differences between the two texts, the allusion seems likely (cf. 9:6; 24:30; 26:64; pace Vögtle 1964; Gundry 1982: 595). If edothē is an ingressive aorist, it suggests that Jesus’ reign began at the resurrection (Jeremias 1971: 310). Malina and Rohrbaugh 1992: 24 find an echo of 2 Chron 36:23, which closes the Hebrew Bible, just as Mt 1:1 echoes Genesis; the allusion seems strongest in v. 18, but closer verbal parallels elsewhere make the connection at best hypothetical.

. Jewish tradition did not articulate a doctrine of the Spirit’s personality or distinction from the Father, as Christian theology came to do (e.g., Jn 14:16-17, 26; 15:26; 16:7-15; see Keener 1991b: 233-65). Baptism “in the name of the Spirit” may remind believers that this baptism alluded to an experience greater than any that accompanied John’s (cf. 3:11).

. Jesus’ divinity is explicit in Luke’s theology of baptism in Jesus’ name, Acts 2:38 fulfilling Joel’s prophecy recorded in 2:21. Lachs 1987: 442 compares “into the name” with the Heb. leshem, “for the purpose of,” but rabbinic evidence for “baptism in the name of heaven [i.e., God]” may be more fruitful (Abrahams 1917: 45). Cf. repentance “in the name of the Most High God” (Jos. and Asen. 15:7); salvation “in the name of the Lord of Spirits” (1 Enoch 48:7); for a summary of various uses of “name” in the Tenach, see, e.g., Keener 1993: 299-300. Our earliest evidence for Christian communities shows that initiatory baptism was already in place (see Meeks 1986: 99).

. Davies 1993: 207 misses this motif, regarding Jesus’ inclusion here as merely that of a human being whose Sonship was “exemplary for other human beings.”

. Schaberg 1982: 335-36 finds it in a traditional midrash behind Matthew’s redaction, based on the triad of Ancient of Days, one like a son of man, and angels in Dan 7. She doubts, however, that it is fully trinitarian, even in Matthew, thinking the Spirit is assimilated to angels and a human but heavenly figure accepted into the divine council (336-37); Dan 7 does, after all, appear in v. 18 (Allison 1993b: 265-66). Meier 1980: 371 more plausibly accepts the trinitarian sense in Matthew, though also viewing the formula as pre-Matthean (probably from his community’s liturgy); trinitarian language certainly predates this Gospel (e.g., in Paul). France 1985: 415 argues from the baptism in Jesus’ name in Acts that the explicit trinitarian formula is the church’s language (cf. Carson 1984b: 598, who also doubts that this represents the ipsissima verba); but if trinitarian language is as early as Paul (see Fee 1994: 839-42) there is no reason why it cannot have originated with the risen Christ, so the question of the language’s origin should remain open.

. Cf. Athenag. Plea 24; Acts of Jn 94.1-2; in Justin, Kaye 1912: 54; Chadwick 1964: 287; much later, Sib. Or. prol. 15-16. Though regarding both as deity, the early Christians did not confuse Jesus with the Father the way that many forms of ancient mysticism could coalesce persons (e.g., PDM 14.349). Proposed earlier pagan trinities (e.g., Diop 1974: 194) are just subgroups within polytheistic pantheons (e.g., Book of Dead Spells 1; 142), which could as easily come to more or less than three in other cases. The Christian doctrine can hardly derive from such sources; those who formulated the established Christian doctrine did so inductively from NT texts that in many cases had not themselves emphasized the triadic nature so clearly. Later Judaism accommodated the Christian trinity for Gentile Christians as shittuf, partnership, while denying it to Jews for whom they claimed God held a higher standard (Schoeps 1963: 16-17; Berger and Wyschogrod 1978: 33; Jocz 1979: 13; Borowitz 1980: 32; Falk 1985: 33-35).

. This assumes, of course, that Mk 16 did not originally end with 16:8; it may have done so, however (for abrupt endings, see, e.g., Ps-Philo; Isoc. Demon. 52, Or. 1; Plut. Fame of Athenians 8, Mor. 351B; and other references above).

. Jewish teachers, who heavily emphasized traditioning through disciples (e.g., ARN 1A; 1, §2B) and disciples replacing them (Pesiq. R. 51:2, bar.), recognized that prophets, like teachers, trained disciples to succeed them (Sir 46:1; 48:12; Test. Mos. 1:7; 10:15; Mek. Pisha 1.150-53; even patriarchs, in Jub. 19:17), though they did not always succeed in maintaining the same level of power (Pesiq. Rab Kah. 24:18). Smith 1978: 113 parallels Jesus’ sending his disciples (e.g., 10:1) with magic being taught and its powers imparted to others, but this ignores the abundance of nearer parallels. Drawing parallels between characters was standard practice (e.g., Polyb. R.R.E. 10.2; Plut. Theseus 1.2), and those who did so normally located parallels in their abundant data or traditions rather than inventing them (Plut. Sertorius 1.1).

. They continue Jesus’ message earlier in the Gospel (4:17; Scaer 1991). Some scholars find in the sequence (preamble, demand, promise) a covenant format, or echoes of biblical commissioning language (De Ridder 1971a: 178; Hubbard 1974; Smyth 1975; Carson 1984b: 591-92). Additional precedents for authorization-commission are also possible, though probably more distant; e.g., in the final form of the Pentateuch God commissions Adam and Noah to “subdue” the earth (Gen 1:28) because all was beyedkem nitanu, “given into your hands” (Gen 9:2), which prefigures Israel’s commission for conquering the promised land (cf. perhaps Jub. 6:5). For comments on farewell discourses (directed toward Mk 13), see Robbins 1992: 174-75.

. Cf., e.g., “Going, learn what …” Hos 6:6 means (Mt 9:13), but contrast, “I coming will heal him” (8:7), where the participle is significant whether one reads the sentence as a statement or a question.

. Simply because Jewish-Christians perhaps could not witness in some synagogues after 85 would not end their witness among their acquaintances (see Hagner 1996: 43). It is doubtful that Matthew would think that the church proclaims the kingdom as the “new Israel” (as in Roux 1956: 294).

. Cf. likewise Argyle 1963: 222, arguing from the universality of the early Christian practice; pace Bultmann 1958: 153, who improbably attributes the practice to the hellenistic church (ignoring the common function between Christian and proselyte baptism shared via the Baptist; see comment on 3:5-6).

. Unlike Luke (Lk 24:51; Acts 1:9-11), Matthew does not narrate the ascension (though implied in Mt 24:30; 26:64; Mk 13:26; 14:62; Jn 20:17) because he stresses Jesus’ continuing presence and reign; cf. Kingsbury 1975: 32. The final pericope of his Gospel proleptically anticipates the parousia (Meier 1979: 37-38); “until the termination of the age” could imply imminence (cf. Sib. Or. 3.756-57) as easily as delay.

↑ Back to top