Mostrando entradas con la etiqueta Flavius Josephus. Mostrar todas las entradas
Mostrando entradas con la etiqueta Flavius Josephus. Mostrar todas las entradas

11/19/2026

Flavius Josephus met a Saul [of Tarsus?]: a descendant of Idumaeans, a plunderer of the weak and an emissary to Nero.

BS"D

I begin with an exercise that seems trivial and is not. Take a sheet of paper and write down everything you know about Paul of Tarsus: that he was born in Tarsus, that he was a Roman citizen, that he was trained in Jerusalem ‘at the feet of Gamliel’, that he was a Pharisee, that he persecuted the church and that, before he was called Paul, he was called Saul. Now strike out every line that you cannot support with something Paul himself wrote.

What survives is little, and it is not what you would expect. Tarsus does not appear in the letters. Gamliel does not appear. Roman citizenship does not appear. The name ‘Saul’ does not appear even once: the one who signs is always Παῦλος (Paûlos, ‘Paul’). Everything else comes from a single witness, the author of Acts, who writes when Paul was already dead and who has a visible interest in his protagonist being, at once, a Jew trained in Jerusalem, a citizen of Tarsus and a Roman by birth (Acts 21:39; 22:3; 22:28). When it is Paul himself who speaks of his origin, he does so in a single line, Philippians 3:5–6:

περιτομῇ ὀκταήμερος, ἐκ γένους Ἰσραήλ, φυλῆς Βενιαμίν, Ἑβραῖος ἐξ Ἑβραίων, κατὰ νόμον Φαρισαῖος, κατὰ ζῆλος διώκων τὴν ἐκκλησίαν (circumcised on the eighth day; of the stock [γένος] of Israel; of the tribe of Benjamin; a Hebrew born of Hebrews; as to the law [νόμος], a Pharisee; as to zeal, a persecutor of the assembly [ἐκκλησία]).

I do not strike out ‘Pharisee’; I underline it in pencil, because later it will have to live alongside what follows. But I pause over the form of the list. A curriculum vitae is written to boast; an enumeration so orderly, so insistent and so defensive is written to answer an accusation. Commentators usually say that the accusation comes from the ‘Judaizers’. I asked myself a different question, and I leave it to you unanswered for now: what kind of man needs to make clear that he was circumcised on the eighth day, that he belongs to a γένος (lineage) beyond doubt and that his tribe is a specific tribe?

And a second question, more uncomfortable: what source earlier than, or independent of, Luke tells you that this man was born in Tarsus, studied with Gamliel and was a Roman citizen? Think about it for a moment before going on.

When you look for it, you discover that the later witnesses do not agree either. Jerome (De viris illustribus 5) knows a tradition according to which Paul’s family was de tribu Beniamin et oppido Iudaeae Giscalis (of the tribe of Benjamin and of the Judaean town of Gischala) and was carried off to Tarsus after a Roman capture. Epiphanius (Panarion 30.16.6–9), gathering Ebionite memory, says that Paul was a Greek, from Tarsus; that he went up to Jerusalem; that he wished to marry the high priest’s daughter; that he became a proselyte in order to achieve it and that, on failing, he wrote against circumcision, the Sabbath and the Torah (for a modern thesis of non-Jewish origin, see Maccoby, 1986). Three biographies incompatible with one another—Luke’s, Jerome’s and the Ebionites’—and none is that of the letters.

And there is a concrete absurdity that Luke did not see coming. Acts 16:37 has Paul claim his Roman citizenship after having been flogged at Philippi without trial. Paul, on the other hand, writes in 2 Corinthians 11:25: τρὶς ἐραβδίσθην (three times I was beaten with rods), the punishment of the lictors, which Cicero regarded as a crime when it fell upon a citizen (scelus verberare, ‘to flog is a crime’, In Verrem 2.5.170). A Roman citizen by birth who three times receives the Roman rods and who never, in any of his letters, mentions the recourse that would have stopped each one of them. Imagine it: the man who does not keep silent about a shipwreck, nor about a basket let down from a wall, nor about his hungers, keeps back his strongest card.

It was then, searching outside the New Testament for some Saul of the first century, that I opened the Greek text of Josephus—Niese’s edition—and came across a sentence that drew from me an uneasy smile. It is November of the year 66; the Roman legate Cestius Gallus has just been defeated and the aristocracy of Jerusalem is calculating. Josephus writes (Jewish War 2.556):

Μετὰ δὲ τὴν Κεστίου συμφορὰν πολλοὶ τῶν ἐπιφανῶν Ἰουδαίων ὥσπερ βαπτιζομένης νηὸς ἀπενήχοντο τῆς πόλεως. Κοστόβαρος γοῦν καὶ Σάουλος ἀδελφοὶ σὺν Φιλίππῳ τῷ Ἰακίμου, στρατοπεδάρχης δ᾽ ἦν οὗτος Ἀγρίππα τοῦ βασιλέως, διαδράντες ἐκ τῆς πόλεως ᾤχοντο πρὸς Κέστιον. (After the disaster of Cestius many of the eminent Jews swam away from the city as from a ship that is being sunk. So Costobar and Saul, brothers, together with Philip son of Jacimus—this man was commander of King Agrippa’s army—slipped out of the city and went off to Cestius.)

Look at the participle: βαπτιζομένης (baptizoménēs), from βαπτίζω (‘to immerse’, ‘to sink’). It is the same root with which the New Testament names the rite of initiation; here it designates a ship going down. A Saul who escapes by swimming from a ship that is being ‘baptised’. And a Paul who, in 2 Corinthians 11:25, boasts:

τρὶς ἐναυάγησα, νυχθήμερον ἐν τῷ βυθῷ πεποίηκα (three times I was shipwrecked; a night and a day I have spent in the deep).

If a missionary presented me with this as proof, I would recommend to him Samuel Sandmel’s essay (1962) on ‘parallelomania’, written precisely to nip this sort of enthusiasm in the bud. A smile is not evidence. But the smile kept me reading, and what came next was no longer amusing.

The first thing I did was to come to terms with the text, because the summaries circulating in Spanish about this Saul repeat, one from another, a datum that turned out to be inaccurate. Josephus names him in three places, and it is worth reading them in Greek (I cite the text of Niese, 1885–1895).

The first is Jewish Antiquities 20.214, under the procurator Albinus (AD 62–64):

Κοστόβαρος δὲ καὶ Σαοῦλος αὐτοὶ καθ᾽ αὑτοὺς μοχθηρὰ πλήθη συνῆγον, γένους μὲν ὄντες βασιλικοῦ καὶ διὰ τὴν πρὸς Ἀγρίππαν συγγενείαν εὐνοίας τυγχάνοντες, βίαιοι δὲ […] καὶ ἁρπάζειν τὰ τῶν ἀσθενεστέρων ἕτοιμοι. ἐξ ἐκείνου μάλιστα τοῦ καιροῦ συνέβη τὴν πόλιν ἡμῶν νοσεῖν, προκοπτόντων πάντων ἐπὶ τὸ χεῖρον. (Costobar and Saul, on their own account, gathered wicked mobs; they were of royal stock and, on account of their kinship with Agrippa, enjoyed favour; but they were violent […] and ready to seize the property of the weaker. It was from that time above all that our city came to be sick, everything advancing from bad to worse.)

The second, Jewish War 2.418–419, at the outbreak of the revolt (summer of 66), when the aristocracy tries to restrain the insurgents:

καὶ πρέσβεις οὓς μὲν πρὸς Φλῶρον ἔπεμπον, ὧν ἦρχεν υἱὸς Ἀνανίου Σίμων, οὓς δὲ πρὸς Ἀγρίππαν, ἐν οἷς ἦσαν ἐπίσημοι Σαῦλός τε καὶ Ἀντίπας καὶ Κοστόβαρος προσήκοντες τῷ βασιλεῖ κατὰ γένος. (and they sent envoys: some to Florus, headed by Simon son of Ananias, others to Agrippa, among whom were the notable Saul and Antipas and Costobar, belonging to the king by birth.)

The third, the one you have already read, Jewish War 2.556, and its continuation in 2.558:

Κέστιος δὲ τοὺς περὶ Σάουλον ἀξιώσαντας ἀνέπεμψεν εἰς Ἀχαΐαν πρὸς Νέρωνα τήν τε αὐτῶν δηλώσοντας ἀνάγκην καὶ τὰς αἰτίας τοῦ πολέμου τρέψοντας […] (Cestius, for his part, sent up to Achaea, to Nero, Saul and his party, who had requested it, to make known their plight and to shift the causes of the war [on to Florus, says the rest of the sentence].)

Observe what Josephus does not say. I took it for granted—as does the table that accompanies Eisenman’s thesis—that Josephus calls Saul and his companions ‘kinsmen of the king’ with the noun συγγενής (kinsman). He does not. Josephus uses three formulae, and none is that title: γένους βασιλικοῦ (of royal stock) and συγγένεια (kinship, an abstract noun) in Ant. 20.214, and προσήκοντες … κατὰ γένος (belonging to the king by birth) in War 2.419. The correction seems minor and I make it in full view because it affects an argument we shall see: that of the ‘kinsmen’ of Romans 16. For συγγενεῖς (kinsmen) belonged, in that period, to the language of the court as much as to that of the family. The First Book of Maccabees speaks of the golden brooch that is customarily given to the kinsmen of kings:

καθὼς ἔθος ἐστὶν δίδοσθαι τοῖς συγγενέσι τῶν βασιλέων (1 Macc 10:89) (as it is the custom to give to the kinsmen of the kings),

and Josephus himself summons to a royal tribunal οἱ φίλοι καὶ οἱ συγγενεῖς βασιλέως (the friends and the kinsmen of the king; Ant. 17.93). The word, therefore, proves blood on neither side: neither the one Josephus does not use for Saul, nor the one Paul uses for six persons in Romans 16. The adjective ἐπίσημοι (notable, ‘marked’), on the other hand, does recur: Josephus applies it to Saul and his companions (2.418), and Paul to Andronicus and Junia, two of those ‘kinsmen’ of his (Rom 16:7: οἵτινές εἰσιν ἐπίσημοι ἐν τοῖς ἀποστόλοις, ‘who are notable among the apostles’). But ἐπίσημος is an ordinary word, and I do not load it with more than it can bear.

And who were Saul’s parents? Josephus does not say. The only explicit thing is that Costobar and Saul were ἀδελφοί (brothers; War 2.556). The rest is reconstruction. The name ‘Costobar’ is an Idumaean theophoric, Qos-gabar (קוס־גבר, ‘Qos is strong’, with the root גבר, g-b-r, ‘to be strong’, as in the name of the Edomite king Qaus-gabri who appears in the Assyrian annals of the seventh century BCE). And of that Costobar I, Josephus says:

Κοστόβαρος ἦν γένει μὲν Ἰδουμαῖος, ἀξιώματος τῶν πρώτων παρ᾽ αὐτοῖς καὶ προγόνων ἱερατευσάντων τῷ Κωζαι· θεὸν δὲ τοῦτον Ἰδουμαῖοι νομίζουσιν. (Ant. 15.253) (Costobar was an Idumaean by birth, of rank among the foremost of them and of ancestors who had served as priests of Kozai; this one the Idumaeans reckon a god.)

He was married to Salome, Herod’s sister:

ἡ δὲ Βερενίκη Κοστοβάρου καὶ Σαλώμης παῖς τῆς Ἡρώδου ἀδελφῆς (Berenice was the child of Costobar and of Salome, Herod’s sister; Ant. 18.133).

Prosopographers infer that the two brothers of 62–66 were grandsons of that marriage, through an Antipater (see Kokkinos, 1998). It is a reasonable hypothesis, supported by the fact that the name reappears a generation later. It is not a datum from Josephus, and it is as well not to forget it when someone presents it to you as such.

Before entering the case, it is worth looking one rung higher up the tree, because there ends what Josephus knows and there begins what was told. Josephus makes of Antipas, who bore the same name as his son Antipater, a leading Idumaean and strategos of Idumaea under Alexander Jannaeus, and does not say who his father was (Ant. 14.8–10). Three traditions fill that gap, and none agrees with the others. Nicolaus of Damascus, historian at Herod’s court, made Antipater descend from Jews who had returned from Babylon; Josephus denies it and explains it as flattery of the king (Ant. 14.9). Julius Africanus, cited by Eusebius (Ecclesiastical History 1.7.11), relates that Antipater’s father was a certain Herod, ἱερόδουλος (temple slave) of Apollo at Ascalon, and that Antipater himself was captured as a boy by Idumaean bandits and raised among them; in his version there is no Antipas as governor, so that it cannot be harmonised with Josephus. Strabo, who has no stake in the dispute, says that the Idumaeans were Nabataeans by origin who, expelled by a revolt, passed over to the Jews and adopted their customs (Geography 16.2.34). Each version has its interest: Nicolaus’s serves the throne; Africanus’s, a Christian polemic that profits from a Herod of low birth (he himself adds that Herod burned the genealogical records to hide his origin, 1.7.13, a report that he links to what the relatives of Yeshu preserved); Josephus’s is the soberest, but it too makes Idumaean Jewishness something acquired by force (Ant. 13.257). Strabo’s is the only one without a stake, and for that reason it weighs more, though it weighs little: he is a geographer repeating what he heard. What survives the comparison is not a lineage but a finding: in no version are Antipas’s ancestors ‘of Israel by lineage’ before Hyrcanus, and those who take an interest in that lineage do so in order to correct it or to deny it. None of this is attested, and that is why the tree draws it in purple. But it explains why, in the first century, the γένος of a Herodian was a public matter, and why anyone wishing to settle that doubt would find it convenient to leave in writing a list like that of Philippians 3:5.



Name

Gen.

Approx. dates

Relation to Saulus/Paul

Office or title

Main sources

Certainty

Origin of the Idumaeans (exiled Nabataeans)

0

n/a

None directly; origin of Antipas’s people

n/a

Strabo (via Timagenes)

🟪 Unattested tradition

Herod, father of Antipas

0

unknown

Great-great-great-great-grandfather (if true)

Temple slave of Apollo at Ascalon

Julius Africanus, in Eusebius, Eccl. Hist. 1.7.11

🟪 Hostile tradition, uncorroborated

Antipas of Idumaea

1

fl. 103–76 BC (estimate from the reign of Jannaeus)

Great-great-grandfather of Saulus (reconstructed)

Governor of Idumaea under A. Jannaeus

Josephus, Ant. 14.8–10 (does not name his father)

🟩 Attested

Antipater the Idumaean

2

c. 114/113–43 BC

Great-grandfather

Procurator of Judaea

Josephus, War 1; Ant. 14

🟩 Attested

Cypros of Nabataea

2

unknown

Great-grandmother

Wife of Antipater

Josephus, War 1.181; Ant. 14.121

🟩 Attested

Herod the Great

3

c. 73/72–4 BC

Great-uncle (brother of Salome I)

King of Judaea

Josephus, War 1; Ant. 14–17

🟩 Attested

Salome I

3

c. 65 BC–c. AD 10

Grandmother (if Antipater is her son)

Sister of Herod

Josephus, Ant. 15.253–266; 18.133

🟩 Attested

Costobar I

3

executed c. 25 BC

Grandfather (same condition)

Idumaean governor of Idumaea and Gaza

Josephus, Ant. 15.253–266

🟩 Attested

Antipater (son)

4

unknown

Hypothetical father of Costobar II and Saulus

n/a

Kokkinos (1998); Josephus does not name him

🟨 Strong inference

Berenice I

4

fl. c. 10 BC

Aunt (sister of the supposed father)

Wife of Aristobulus; mother of Agrippa I

Josephus, Ant. 18.133

🟩 Attested

Aristobulus

4

c. 35–7 BC

Uncle by marriage

Son of Herod the Great

Josephus, Ant. 18.133; War 1

🟩 Attested

Costobar II

5

act. AD 62–66

Brother of Saulus

Noble of royal stock; kinsman of Agrippa II

Josephus, Ant. 20.214; War 2.418–419, 2.556

🟩 Existence and brotherhood; 🟨 filiation

Saulus

5

act. AD 62–66

Central person

Herodian noble; envoy to Agrippa; sent to Nero

Josephus, Ant. 20.214; War 2.418–419, 2.556, 2.558

🟩 Existence; 🟥 identity with Paul

Agrippa I

5

10 BC–AD 44

First cousin (son of Berenice I)

King of Judaea

Josephus, Ant. 18–19

🟩 Attested

Agrippa II

6

c. AD 27–c. 92/100

‘Kinsman’ according to Josephus; in the reconstruction, first cousin once removed (son of Agrippa I)

Client king of Rome

Josephus, War 2–3; Ant. 20

🟩 Attested

Paul of Tarsus

= Saulus (G5)

fl. 1st century AD (death: 64 or 67–68 according to tradition)

Hypothetical identity with Saulus

Apostle; of the tribe of Benjamin (Phil 3:5)

Letters (Gal 1:13–14; Phil 3:5–6; 2 Cor 11:32); Acts

🟩 Historical person; 🟥 Herodian filiation

Here the case begins, and I set it out as what it is: a thesis that I hold and shall defend with the text. I hold that the Saul whom Josephus describes in the Jerusalem of the sixties—a Herodian aristocrat, plunderer of the weak and ambassador to Nero—is the man whom the letters call Paul, and that the ‘Saul of Tarsus, disciple of Gamliel’ known to the Church is a figure that Luke constructed afterwards. The observations that follow do not weigh the same, and I shall not pretend that they do: I shall call an indication a coincidence with textual support that admits an ordinary explanation, and a hypothesis whatever works only if we already accept the identification. None supports the thesis on its own. What supports it is that the eight converge without forcing the text and that each could have failed: it would have sufficed for Josephus to write another office, another verb or another social class, and the thesis would have fallen. And, so far as I can see from the summaries of Eisenman’s thesis (1997, 2006) that I have had before me—the ‘datum from Josephus / datum from Paul’ table and the exposition that accompanies it—none of the following appears developed in them. If any was already in his books, so much the better: it would be an independent convergence.

First: the two-faced profile. Of everything that Josephus relates of Saul, the list of verbs is very short: he gathers mobs, he abuses the weak, he is sent as ambassador to Agrippa, he flees and asks to be remitted to Caesar. Two offices: that of persecutor and that of ambassador. Now look at how Paul defines himself. On the one hand, what we have already seen: διώκων τὴν ἐκκλησίαν (persecutor of the assembly). On the other, a diplomatic language unusual in the New Testament:

Ὑπὲρ Χριστοῦ οὖν πρεσβεύομεν ὡς τοῦ θεοῦ παρακαλοῦντος δι᾽ ἡμῶν (2 Cor 5:20) (On behalf of Christ, then, we act as ambassadors, as though God were appealing through us)

and, in a letter of disputed authorship, ὑπὲρ οὗ πρεσβεύω ἐν ἁλύσει (on behalf of which I am an ambassador in a chain; Eph 6:20). The verb πρεσβεύω (to be an ambassador) is a direct relative of the πρέσβεις (ambassadors) with which Josephus designates Saul’s mission in War 2.418. In Philemon 9 the manuscripts have πρεσβύτης (old man), but an ancient conjecture, attributed to Bentley, proposes to read πρεσβευτής (ambassador): a difference of a single letter. I do not accept it; I note it because it fits the metaphor of 2 Corinthians 5:20.

It is an indication: the metaphor of the ambassador is well known and Hellenistic diplomacy generated the verb. But Paul had other images to hand—herald, servant, steward—and often uses them. That at the most delicate moment of his defence of reconciliation he chooses that of the man who speaks in the name of one power before another power is what one would expect of someone who had been an ambassador. That last is hypothesis, and I mark it as such.

Second: the vocabulary of violence. Galatians 1:13–14 is the most direct confession that Paul makes of his past:

καθ᾽ ὑπερβολὴν ἐδίωκον τὴν ἐκκλησίαν τοῦ θεοῦ καὶ ἐπόρθουν αὐτήν, καὶ προέκοπτον ἐν τῷ Ἰουδαϊσμῷ ὑπὲρ πολλοὺς συνηλικιώτας ἐν τῷ γένει μου (I persecuted the assembly of God beyond measure and ravaged it, and I advanced in Judaism beyond many of my contemporaries in my lineage).

Three words in that verse ask us to set them beside Antiquities 20.214. The first is ἐπόρθουν, from πορθέω, the Greek verb of plunder: to lay waste a city, to ravage a field (so the lexicon of Liddell, Scott and Jones, 1996, records it); it is not the verb for ‘prosecuting’ someone. Acts repeats it in 9:21, ὁ πορθήσας ἐν Ἱερουσαλὴμ τοὺς ἐπικαλουμένους τὸ ὄνομα τοῦτο (the one who ravaged in Jerusalem those who called upon this name), and in 8:3 adds the method, ἐλυμαίνετο τὴν ἐκκλησίαν κατὰ τοὺς οἴκους εἰσπορευόμενος (he was devastating the assembly, entering house after house). House after house, with mobs, against the weakest: compare the ἁρπάζειν τὰ τῶν ἀσθενεστέρων (to seize the goods of the weaker) of Josephus. Both belong to the lexicon of rapine, not of jurisdiction. And there is a hostile witness of the first rank within the New Testament itself: the letter to the Hebrews, which does not name the persecutor but preserves the economic side of that violence with the same root that Josephus uses:

τοῖς δεσμίοις συνεπαθήσατε καὶ τὴν ἁρπαγὴν τῶν ὑπαρχόντων ὑμῶν μετὰ χαρᾶς προσεδέξασθε (Heb 10:34) (you sympathised with the prisoners and accepted with joy the plundering [ἁρπαγή] of your possessions).

Acts speaks of prisons and of approving deaths; Hebrews recalls the despoiling of property, precisely what Josephus attributes to Saul.

The second word is προέκοπτον (I advanced), from προκόπτω (to advance). Josephus uses the same verb with bitterness: everything ‘advanced’ from bad to worse (προκοπτόντων πάντων ἐπὶ τὸ χεῖρον). And the letter of disputed authorship to Timothy preserves the entire phrase: πονηροὶ δὲ ἄνθρωποι καὶ γόητες προκόψουσιν ἐπὶ τὸ χεῖρον (evil men and charlatans will advance from bad to worse; 2 Tim 3:13). Before giving it weight, that collocation would have to be measured in the Thesaurus Linguae Graecae: it is not a datum until someone measures it, and I say so bluntly. The third is γένος: Paul measures his advance against those of ‘my lineage’; Josephus sees in Saul’s lineage the reason for his impunity. Each coincidence, taken alone, is a weak indication. What seems to me noteworthy is that they pile up in one and the same verse of Galatians, written decades before Josephus put pen to that passage, so that neither could have copied the other.

Third: the inversion. Perhaps it is the observation that most disturbed me, because I did not seek it. Josephus condemns Saul for being ready to ἁρπάζειν τὰ τῶν ἀσθενεστέρων (seize the goods of the weaker). Now open Romans 15:1:

Ὀφείλομεν δὲ ἡμεῖς οἱ δυνατοὶ τὰ ἀσθενήματα τῶν ἀδυνάτων βαστάζειν καὶ μὴ ἑαυτοῖς ἀρέσκειν (We who are the powerful ought to bear the weaknesses of the powerless, and not to please ourselves).

Compare the structures: transitive infinitive, neuter plural and genitive of ‘the weak’. Change a single element, the verb: ἁρπάζειν (to seize) becomes βαστάζειν (to bear, to carry upon oneself). And the subject changes: they are no longer ‘some’ but ἡμεῖς οἱ δυνατοί, ‘we, the powerful’. Οἱ δυνατοί is, moreover, the term with which Josephus ordinarily designates the aristocracy of Jerusalem; for example, in War 2.301: προσελθόντες οἵ τε ἀρχιερεῖς καὶ δυνατοί (the chief priests and the powerful approached). Paul includes himself, without anyone asking him, in the estate of the δυνατοί. There is more. In 1 Corinthians 1:26 he enumerates the social categories of his audience:

οὐ πολλοὶ σοφοὶ κατὰ σάρκα, οὐ πολλοὶ δυνατοί, οὐ πολλοὶ εὐγενεῖς (not many wise according to the flesh, not many powerful, not many well-born),

and in 2 Corinthians 12:10 he rounds off: ὅταν γὰρ ἀσθενῶ, τότε δυνατός εἰμι (for when I am weak, then I am powerful). A theology of weakness written in the vocabulary of nobility. And Josephus’s verb reappears, in the passive and with another sign: ἁρπαγέντα τὸν τοιοῦτον ἕως τρίτου οὐρανοῦ (such a man snatched up as far as the third heaven; 2 Cor 12:2) and ὅτι ἡρπάγη εἰς τὸν παράδεισον (that he was snatched up into paradise; 12:4). Paul speaks of ‘a man in Christ’, whom almost everyone understands to be himself. The man who, in Josephus’s version, ‘seizes’ the weak is, in his own, the one ‘snatched up’.

Indication, not proof: ἁρπάζω and ἀσθενής are common words, and the contrast between strong and weak is a moral commonplace of the period. What is noteworthy is not each word but that Pauline theology of weakness should be, lexically, the mirror image of the accusation that Josephus makes against Saul. It is a pattern that the identification predicts and that non-identity leaves unexplained, although chance remains possible.

Fourth: an enigma of Damascus. This observation is of another kind: it does not add coincidences, it resolves a problem. Acts 9:23–25 says that it was ‘the Jews’ who watched the gates of Damascus to kill Paul. Paul tells another story in 2 Corinthians 11:32:

ἐν Δαμασκῷ ὁ ἐθνάρχης Ἀρέτα τοῦ βασιλέως ἐφρούρει τὴν πόλιν Δαμασκηνῶν πιάσαι με (in Damascus the ethnarch of King Aretas was guarding the city of the Damascenes in order to seize me).

His version is political, not religious: the one who wants to arrest him is a Nabataean official. Why would an officer of Aretas take the trouble? Josephus relates that Aretas IV was at open war with Herod Antipas over the repudiation of his daughter (Ant. 18.109–115) and that he defeated him around the year 36. If the man sought in Damascus about 37–40 (the usual chronology) were a young relative of the Herodians, Aretas had a motive that no Jewish preacher would have. There is another link: Galatians 1:17 says ἀπῆλθον εἰς Ἀραβίαν (I went away into Arabia), that is, to the Nabataean kingdom; and Josephus records that the ancestors of Costobar I had been priests of Kozai, an Edomite deity, and that Salome, his widow, had a Nabataean suitor, the regent Syllaeus (Ant. 16.220–225). The Herodian house was entangled with Nabataea in both directions: that of alliance and that of enmity.

It is a hypothesis, and a conditional one: it works only if we already accept that Paul belonged to that house. But an identification that solves a problem that nobody posed to it—why Aretas?—has a merit that the mere list of coincidences lacks.

Fifth: Philippians 3:5 as a point-by-point answer. Return to the list at the beginning and read it now with the eyes of a Herodian. Each item has beside it a documented jibe against the house of Herod. The first is περιτομῇ ὀκταήμερος (circumcised on the eighth day). The Idumaeans entered Judaism by a condition that Josephus writes without adornment:

ἐπέτρεψεν αὐτοῖς μένειν ἐν τῇ χώρᾳ, εἰ περιτέμνοιντο τὰ αἰδοῖα καὶ τοῖς Ἰουδαίων νόμοις χρήσασθαι θέλοιεν (Ant. 13.257) (he permitted them to remain in the country if they would circumcise their genitals and were willing to use the laws of the Jews).

For an Idumaean, adult circumcision was the founding stigma (on the Judaisation of Idumaea, see Kasher, 1988); ‘on the eighth day’ is its exact negation. The second is ἐκ γένους Ἰσραήλ (of the lineage of Israel). Antigonus, before Antony, says of Herod:

ἰδιώτῃ τε ὄντι καὶ Ἰδουμαίῳ, τουτέστιν ἡμιιουδαίῳ, δέον τοῖς ἐκ τοῦ γένους οὖσι παρέχειν ὡς ἔθος ἐστὶν αὐτοῖς (Ant. 14.403) (to a commoner, an Idumaean at that, that is, a half-Jew, when it is proper to give it to those who are of the lineage, as is their custom).

Note that the word of the accusation is γένος, the same as Paul’s. The third, Ἑβραῖος ἐξ Ἑβραίων (a Hebrew born of Hebrews), answers ἡμιιουδαῖος (half-Jew). And the fourth, φυλῆς Βενιαμίν (of the tribe of Benjamin), answers the suspicion that the house belonged to no tribe. The mockery came from several sides at once. The Babylonian Talmud says it without circumlocution: הורדוס עבדא דבית חשמונאי הוה (Herod was a slave of the house of the Hasmoneans; Bava Batra 3b). Julius Africanus, already cited, makes him descend from a ἱερόδουλος (temple slave) of Apollo at Ascalon (Ecclesiastical History 1.7.11). Three sources with no reason to coincide—a Jewish historian in the service of Rome, the Talmud and a Christian chronographer—point to the same stigma: doubts about the legitimacy of the lineage. And Costobar I, precisely, hid for years the sons of Baba, enemies of Herod (Ant. 15.260–266), a figure whom many identify with the Baba ben Buta of the Talmud (identification disputed).

With that key, Galatians 4:30 sounds different:

ἔκβαλε τὴν παιδίσκην καὶ τὸν υἱὸν αὐτῆς (cast out the slave-woman and her son).

The allegory of Hagar and Sarah turns servile ancestry into the stigma of ‘the present Jerusalem’ and displaces legitimacy from blood to promise. For someone who has carried the rumour of descending from slaves, that inversion is an enormous rhetorical relief. And in the verse following the list of Philippians, Paul himself declares that all this was gain and that he now counts it loss:

ἀλλὰ ἅτινα ἦν μοι κέρδη, ταῦτα ἥγημαι διὰ τὸν Χριστὸν ζημίαν ... δι᾽ ὃν τὰ πάντα ἐζημιώθην (Phil 3:7–8) (but whatever things were gain to me, these I have counted loss on account of Christ ... on account of whom I suffered the loss of all things).

‘I lost everything’ would be, in the mouth of a noble, a literal phrase. This is, by a distance, the most speculative of the observations: any diaspora Jew confronted with those who demanded circumcision could draw up the same list, and Josephus himself opens his Life by enumerating his pedigree (Life 1–6). All I maintain is that the list of Philippians can be read, point by point, as a ‘no’ to each of the known jibes against the Herodians. Hypothesis.

Sixth: the appeal to Caesar, in one and the same technolect. Return to War 2.558. The legate Cestius ἀνέπεμψεν (sent up, remitted) Saul and his party to Nero, because they had requested it (ἀξιώσαντας), with the mission of pleading the αἰτίας (causes, charges) of the war. Compare Acts 25:21 and 25:27:

ἐκέλευσα τηρεῖσθαι αὐτὸν ἕως οὗ ἀναπέμψω αὐτὸν πρὸς Καίσαρα (I ordered him to be kept until I should send him up to Caesar),

ἄλογον γάρ μοι δοκεῖ πέμποντα δέσμιον μὴ καὶ τὰς κατ᾽ αὐτοῦ αἰτίας σημᾶναι (for it seems to me absurd to send a prisoner without also signifying the charges [αἰτίαι] against him).

Same verb (ἀναπέμπω), same addressee (the emperor), same word for what is alleged (αἰτίαι) and the same origin of the procedure: a man who requests (ἀξιόω in Josephus; ἐπικαλοῦμαι, ‘I appeal’, in Acts 25:11) to be taken before Caesar. And Josephus’s destination is Achaea, precisely the territory where the core of the Pauline communities lies: Rom 15:26 (Μακεδονία καὶ Ἀχαΐα), 1 Cor 16:15 (ἀπαρχὴ τῆς Ἀχαΐας, firstfruits of Achaea), 2 Cor 1:1 (ἐν ὅλῃ τῇ Ἀχαΐᾳ, in the whole of Achaea) and Phoebe, deacon of Cenchreae (Rom 16:1), the port of Corinth. The boring explanation exists: Nero toured Greece between the end of 66 and 67 (Josephus, War 3.1), and ἀναπέμπω is the administrative verb for remitting a case to the emperor. I record it as a weak indication. But note it: if the identification were true, Acts would have taken an episode—that of a noble who asks to be remitted to Nero—and displaced it backwards, to the fifties, giving it the form of a trial. It is the hypothesis of displacement; it is not for me to defend it here, only to point it out.

Seventh, and the most technical: the spelling of the name. Niese’s text gives, for this man, three forms: Σαοῦλος in Ant. 20.214—the same with which Josephus writes the biblical king in his books on the monarchy—, Σαῦλος in War 2.418 and Σάουλος in 2.556 and 2.558. I do not know how much of that variation is Josephus’s and how much the copyists’ or the editor’s; the critical apparatus would have to be consulted before saying anything. But the variation is interesting because it reproduces the pattern of Acts. Luke uses the Hellenised form Σαῦλος in the narrative and reserves the indeclinable, Semitising form Σαούλ (Saul) for the heavenly voice (9:4: Σαοὺλ Σαούλ; also 22:7 and 26:14), for Ananias (9:17; 22:13) and for the biblical king (13:21). It does not prove identity; it proves that oscillation between Σαῦλος and Σαοῦλος was normal in first-century Jewish Greek, and it rules out a minor objection: that the ‘Saul’ of Luke and the one of Josephus could not be the same name. And Luke marks the passage to Παῦλος with a formula typical of the second name, Σαῦλος, ὁ καὶ Παῦλος (Saul, who is also Paul; Acts 13:9), precisely in the episode of the proconsul Sergius Paulus; the Herodians, for their part, took Roman names from their patrons (Marcus Julius Agrippa).

Eighth: the nephew. Acts 23:16 mentions ὁ υἱὸς τῆς ἀδελφῆς Παύλου (the son of Paul’s sister), and Eisenman’s thesis, according to the summary that I have before me, identifies him with Julius Archelaus. I add a datum that I do not see in that summary: in the embassy of War 2.418 there travels with Saul an Antipas, and in the prosopographical literature it has been proposed that this Antipas was the son of Cypros, sister of Costobar and of Saul (this would have to be checked directly in Kokkinos, 1998). If so, the nephew by way of a sister would not only exist in Saul’s family: he would travel with him in the embassy that Josephus records, and would be in his orbit when, in War 2.557, he is besieged in the royal palace. I mark it as a fragile hypothesis, resting on a reconstruction that I have not been able to verify.

So far we have asked what Paul would be if he were Saulus. The symmetrical question is missing, the one that most disturbs whoever reads Acts with the Greek in front of him: if Paul was not a disciple of Gamliel, with what materials was that figure constructed? The question is philological before it is historical. ‘Disciple of Gamliel’ appears once in the whole New Testament, in a speech that Luke composes and places in Paul’s mouth, and a composed text can be taken apart piece by piece. What follows are six pieces, from the least to the most revealing, and at the end the hypothesis that joins them. A warning: none of them comes from the letters.

First piece: a formula of Luke, not a term of a school. This is the complete sentence:

Ἐγώ εἰμι ἀνὴρ Ἰουδαῖος, γεγεννημένος ἐν Ταρσῷ τῆς Κιλικίας, ἀνατεθραμμένος δὲ ἐν τῇ πόλει ταύτῃ, παρὰ τοὺς πόδας Γαμαλιὴλ πεπαιδευμένος κατὰ ἀκρίβειαν τοῦ πατρῴου νόμου, ζηλωτὴς ὑπάρχων τοῦ θεοῦ καθὼς πάντες ὑμεῖς ἐστε σήμερον (Acts 22:3) (I am a Jewish man, born in Tarsus of Cilicia, but brought up in this city, educated at the feet of Gamliel according to the strictness of the ancestral law, being zealous for G"d as all of you are today).

Three participles in a chain—γεγεννημένος (born), ἀνατεθραμμένος (brought up), πεπαιδευμένος (educated)—reproduce the scheme with which Greek literature presented a person: origin, upbringing, education (van Unnik, 1962, studied it as a Hellenistic biographical formula). Josephus follows it in his Life: lineage in paragraphs 1 to 6 and, in 10 and 12, the choice of school:

περὶ δὲ ἑκκαίδεκα ἔτη γενόμενος ἐβουλήθην τῶν παρ᾽ ἡμῖν αἱρέσεων ἐμπειρίαν λαβεῖν ... ἐννεακαιδέκατον δ᾽ ἔτος ἔχων ἠρξάμην τε πολιτεύεσθαι τῇ Φαρισαίων αἱρέσει κατακολουθῶν (Life 10 and 12) (when I was about sixteen I wished to gain experience of the schools [αἱρέσεις] that exist among us ... and in my nineteenth year I began to conduct my life in accordance with the school of the Pharisees).

I do not say that Luke copied Josephus. I say that Acts 22:3 has the form of a Greek self-presentation, not that of a genealogy of a rabbinic school.

Then there is παρὰ τοὺς πόδας (at the feet of). Luke uses that locution eight times: Luke 7:38; 8:35; 8:41; Acts 4:35, 4:37 and 5:2 (the money deposited before the apostles), and the two that matter here, 7:58 and 22:3. Matthew uses it once (15:30), and I do not find it in Paul’s letters. In Acts 7:58, the witnesses of the stoning ἀπέθεντο τὰ ἱμάτια αὐτῶν παρὰ τοὺς πόδας νεανίου καλουμένου Σαύλου (laid down their cloaks at the feet of a young man called Saul); in 22:3, Saul is the one who sits at another’s feet. In the first case the formula marks the one who receives what others put at his feet; in the second, the one who submits. It is a symmetry of authority built within Luke’s prose that needs no school behind it. The rabbinic image of discipleship is another: Yose ben Yoezer says וֶהֱוֵי מִתְאַבֵּק בַּעֲפַר רַגְלֵיהֶם (and wallow in the dust of their feet; Mishnah, Avot 1:4), with a verb, מִתְאַבֵּק, that joins ‘dust’ and ‘struggle’ in a single root (אבק, a-b-q) and that Luke does not use. A weak indication: the image of the feet may have come from a Semitic model; but the locution, as it stands, is Luke’s.

Second piece: a template from Josephus. Compare the description that Josephus gives of Simon, son of Gamliel, with the one that Luke gives of the Paul who is Gamliel’s disciple. Josephus (Life 191):

ὁ δὲ Σίμων οὗτος ἦν πόλεως μὲν Ἱεροσολύμων, γένους δὲ σφόδρα λαμπροῦ, τῆς δὲ Φαρισαίων αἱρέσεως, οἳ περὶ τὰ πάτρια νόμιμα δοκοῦσιν τῶν ἄλλων ἀκριβείᾳ […] (this Simon was of the city of Jerusalem, of very illustrious lineage and of the school of the Pharisees, who are reputed to surpass the others in exactness concerning the ancestral laws […]).

And Luke, before Agrippa (Acts 26:5):

ὅτι κατὰ τὴν ἀκριβεστάτην αἵρεσιν τῆς ἡμετέρας θρησκείας ἔζησα Φαρισαῖος (that according to the strictest school of our religion I lived as a Pharisee).

Four boxes recur: the city (Jerusalem), the lineage, the αἵρεσις (school) of the Pharisees and the ἀκρίβεια (exactness) in matters ancestral. It is the vocabulary with which Josephus describes the Pharisees in general: οἱ μετὰ ἀκριβείας δοκοῦντες ἐξηγεῖσθαι τὰ νόμιμα (those reputed to interpret the laws with exactness; War 2.162). And it is a vocabulary that Paul does not use. In the undisputed letters ἀκρίβεια does not appear, and αἵρεσις appears only in the sense of ‘factions’, and among the works of the flesh (1 Cor 11:19; Gal 5:20). When Paul says what he was, he says κατὰ νόμον Φαρισαῖος (as to the law, a Pharisee; Phil 3:5) and ζηλωτὴς ὑπάρχων τῶν πατρικῶν μου παραδόσεων (being zealous for the traditions of my fathers; Gal 1:14). I say this with care: this is Josephus’s generic template for elite Pharisees, and Luke could have taken it without thinking of Simon. But Simon is the only individual to whom Josephus applies it with Jerusalem and lineage in the same sentence, and he is the son of Gamliel. My reading is a hypothesis, and a falsifiable one: if someone shows the same template in Greek earlier than Josephus, it falls. It says that Luke transferred the boxes from Gamliel’s family to his character.

Third piece: an ‘exact’ persecutor whom Josephus does not know. It remains to see who, in Josephus, persecutes and who protests. The ἀκριβεῖς (exact) appear in the account of the death of James, and not as persecutors but as those who are indignant:

ὅσοι δὲ ἐδόκουν ἐπιεικέστατοι τῶν κατὰ τὴν πόλιν εἶναι καὶ περὶ τοὺς νόμους ἀκριβεῖς, βαρέως ἤνεγκαν ἐπὶ τούτῳ καὶ πέμπουσιν πρὸς τὸν βασιλέα κρύφα (Ant. 20.201) (those who were reputed to be the most equitable men in the city and exact concerning the laws took this badly and secretly sent word to the king).

The king is Agrippa II, the same patron of Saulus, and what they protest is the violence of the high priest Ananus. Those who plunder, on the other hand, are the servants of the high priest Ananias and the bands of the king’s kinsmen (Ant. 20.206–207 and 20.214). In Josephus’s map, the ‘exact’ and the persecutor are two different social types, almost opposites. Paul, in his letters, grounds his persecution in another word: ζῆλος (zeal; Gal 1:14; Phil 3:6: κατὰ ζῆλος διώκων), the same one that he lists, in Galatians 5:20, among the works of the flesh (ἔρις, ζῆλος, θυμοί: strife, jealousy, wrath). Luke fuses the two types in a single phrase (22:3: κατὰ ἀκρίβειαν ... ζηλωτὴς ὑπάρχων τοῦ θεοῦ), and the fusion needs narrative glue, because a disciple of the Gamliel whom Luke himself presents as a protector of the apostles should not persecute them.

The glue is in Gamliel’s single speech:

ἀπόστητε ἀπὸ τῶν ἀνθρώπων τούτων καὶ ἄφετε αὐτούς ... εἰ δὲ ἐκ θεοῦ ἐστιν, οὐ δυνήσεσθε καταλῦσαι αὐτούς, μή ποτε καὶ θεομάχοι εὑρεθῆτε (Acts 5:38–39) (keep away from these men and let them alone ... but if it is from G"d, you will not be able to overthrow them; lest you be found even fighters against G"d [θεομάχοι]).

Θεομάχος is a word of tragedy: Euripides applies it to Pentheus, the king who persecutes the followers of a new god (Bacchae), and the Maccabees know the verb (θεομαχεῖν, 2 Macc 7:19). Now leap to Acts 26:14, where the heavenly voice, the Bat Kol as the messianic ‘Jews’ identify it, speaks to Saul and Luke specifies that it does so τῇ Ἑβραΐδι διαλέκτῳ (in the Hebrew dialect):

Σαοὺλ Σαούλ, τί με διώκεις; σκληρόν σοι πρὸς κέντρα λακτίζειν (Saul, Saul, why do you persecute me? It is hard for you to kick against the goads).

It is a Greek proverb. Aeschylus uses it (Agamemnon 1624: πρὸς κέντρα μὴ λάκτιζε), Pindar (Pythian 2.94) and, above all, Euripides, in the mouth of Dionysus to Pentheus:

θύοιμ᾽ ἂν αὐτῷ μᾶλλον ἢ θυμούμενος πρὸς κέντρα λακτίζοιμι θνητὸς ὢν θεῷ (‘I would sooner sacrifice to him than, in anger, kick against the goads, a mortal against a god’; Bacchae 794–795).

A proverb of Greek tragedy in a voice that, according to the same verse, speaks Hebrew. That the proverb is related to the Bacchae is something the commentaries have already noticed; what I propose, so far as I can see, is to read Gamliel as the piece that installs that motif before Paul appears: Gamliel names the offence (θεομαχία), the disciple commits it and the Lord quotes it in tragic version. The only teaching that Luke attributes to Gamliel is the one that his disciple is going to violate, and the fusion of the ‘exact’ with the persecutor is resolved with a literary irony, not with a biographical datum.

Fourth piece: a borrowed chronology. In that same speech, Gamliel enumerates two rebels:

πρὸ γὰρ τούτων τῶν ἡμερῶν ἀνέστη Θευδᾶς, λέγων εἶναί τινα ἑαυτόν ... μετὰ τοῦτον ἀνέστη Ἰούδας ὁ Γαλιλαῖος ἐν ταῖς ἡμέραις τῆς ἀπογραφῆς (Acts 5:36–37) (for before these days Theudas arose, claiming to be somebody ... after him arose Judas the Galilean in the days of the census).

Josephus recounts Theudas thus:

Φάδου δὲ τῆς Ἰουδαίας ἐπιτροπεύοντος γόης τις ἀνὴρ Θευδᾶς ὀνόματι πείθει τὸν πλεῖστον ὄχλον ἀναλαβόντα τὰς κτήσεις ἕπεσθαι πρὸς τὸν Ἰορδάνην ποταμὸν αὐτῷ· προφήτης γὰρ ἔλεγεν εἶναι (Ant. 20.97) (while Fadus was procurator of Judaea, a certain charlatan [γόης] named Theudas persuades the greater part of the crowd to take up their possessions and follow him to the river Jordan; for he said he was a prophet).

Fadus governed about 44–46; Judas the Galilean rose in the year 6, with the census of Quirinius. Gamliel speaks about the year 33 and mentions a Theudas who had not yet arisen, and places him before Judas, which is the reverse of the real order. Now see how Josephus continues: a few paragraphs after Theudas (Ant. 20.102) he mentions the sons of Judas the Galilean and refers back to the census of Quirinius. Whoever read that passage straight through would find Theudas first and Judas afterwards, and that is just what Luke’s Gamliel does; even the turn λέγων εἶναι (claiming to be) of Acts has its echo in the ἔλεγεν εἶναι (he said he was) of Josephus. That Luke used Josephus is the thesis of Pervo (2006); others prefer a common source. For what matters here it makes no difference: in both cases the speech, the only extensive one that the New Testament gives to Gamliel, was composed with a catalogue of rebels that the historical Gamliel could not have had. He speaks from a library later than himself.

Fifth piece: Paul’s Scripture is in Greek. Second proof: the language in which he cites. The rabbinic tradition attributes to the house of Gamliel a teaching divided between Torah and Greek wisdom (we shall see it shortly, with its cautions); the question is what trace is left in the citations of those who passed through such a house. There are cases in which Paul’s argument works only in Greek. The first:

μὴ οὐκ ἤκουσαν; μενοῦνγε· εἰς πᾶσαν τὴν γῆν ἐξῆλθεν ὁ φθόγγος αὐτῶν (Rom 10:18) (‘Did they not hear?’ Indeed they did: into all the earth went out their voice [φθόγγος]).

It cites Psalm 19:5, whose Hebrew text says בְּכָל־הָאָרֶץ יָצָא קַוָּם (into all the earth went out their line; קַו, ‘line, measuring cord’). The Septuagint seems to have read קוֹלָם (their voice) and translated φθόγγος; Paul’s rhetorical question about hearing depends on that reading. The second:

ἔσται ἡ ῥίζα τοῦ Ἰεσσαί, καὶ ὁ ἀνιστάμενος ἄρχειν ἐθνῶν, ἐπ᾽ αὐτῷ ἔθνη ἐλπιοῦσιν (Rom 15:12) (there shall be the root of Jesse, even he who rises to rule the nations; in him shall the nations hope),

where the Hebrew of Isaiah 11:10 says אֵלָיו גּוֹיִם יִדְרֹשׁוּ (to him the nations shall resort, shall seek him; verb דרש, d-r-sh). The Greek verb ἐλπιοῦσιν (they will hope) is not in the Hebrew, and Paul needs it: the next verse opens with ὁ δὲ θεὸς τῆς ἐλπίδος (the God of hope; Rom 15:13). His argument is a play on words that exists only in the translation. A third case goes in the same direction. In Romans 3:4 he cites ἐν τῷ κρίνεσθαί σε (when you are judged), where the Hebrew of Psalm 51:6 says בְּשָׁפְטֶךָ (when you judge); the Greek puts G"d in the dock, and the objections of 3:5–7 are raised upon that dock, which the Hebrew does not have. And when Paul departs from the Septuagint, he departs towards another Greek form, not towards the Hebrew: in Romans 11:26, ἥξει ἐκ Σιὼν ὁ ῥυόμενος (the deliverer will come from Zion), he follows neither the Masoretic text of Isaiah 59:20 (וּבָא לְצִיּוֹן גּוֹאֵל, a redeemer will come to Zion) nor the Septuagint (ἥξει ἕνεκεν Σιών, he will come for the sake of Zion); his ἐκ Σιών seems to come from Psalm 14:7 in Greek. These cases, and many more, were already collected by Koch (1986) and Stanley (1992), who show that Paul’s Scripture is overwhelmingly the Greek one; it is not my discovery. What is mine is a matter of method: to set that dependence against the phrase ‘educated at the feet of Gamliel’ and against the tradition of the two halves. If Paul studied in that house, what his citations show is the Greek half; what that tradition does not explain is that the Hebrew one does not appear. Josephus, who does leave a Hebrew text of his own (the height of Goliath in Ant. 6.171 follows a form akin to that of 4QSam-a), leaves a trace of the language in which he studied; in Paul, where the Hebrew and the Septuagint diverge, the trace is always Greek. Of the manner in which Paul interprets I say nothing here: what is measured is the language in which he cites, and that is Greek.

Sixth piece: the Hebrew and Aramaic of Gamliel’s house. Let us pass to the languages of the house itself. In the Mishnah, Rabban Gamliel HaZaken has a maxim that is an instruction about teachers:

עֲשֵׂה לְךָ רַב, וְהִסְתַּלֵּק מִן הַסָּפֵק (Avot 1:16) (make for yourself a teacher and withdraw from doubt).

Paul, on the other hand, writes:

οὐδὲ γὰρ ἐγὼ παρὰ ἀνθρώπου παρέλαβον αὐτὸ οὔτε ἐδιδάχθην (Gal 1:12) (for I neither received it from a human being nor was I taught it),

and adds that he did not consult σαρκὶ καὶ αἵματι (flesh and blood; 1:16), a Semitic locution (בשר ודם, ‘a human being’) that was already in biblical Greek before Paul (Sir 14:18). Galatians 1:12 speaks of his gospel, not of his earlier formation, and for that reason I do not use it as proof of anything about the latter. Something similar holds for the two verbs of that verse and of 1 Corinthians 15:3 (ὃ καὶ παρέλαβον, ‘what I also received’): Gerhardsson (1961) tied them to קִבֵּל (received) and מָסַר (delivered), those of the chain of transmission of the Mishnah, but that reading is disputed, and I do not need it here. παραλαμβάνω and παραδίδωμι are in Greek ordinary verbs for transmitting traditions of any kind, not terms exclusive to a school, and they do not say where the one who uses them was trained. What does count is what does not appear: a teacher’s name, a chain formula (‘so-and-so said in the name of such-and-such’) or a technical term of the house of study. What Paul boasts of in Philippians 3:5–6 is a tribe and a zeal, not a teacher.

There is, moreover, a sentence in the letters that Luke could not have read without stumbling:

ἤμην δὲ ἀγνοούμενος τῷ προσώπῳ ταῖς ἐκκλησίαις τῆς Ἰουδαίας ταῖς ἐν Χριστῷ (Gal 1:22) (and I was unknown by face to the assemblies of Judaea that are in Christ).

The man ‘brought up in this city’ (ἐν τῇ πόλει ταύτῃ, Acts 22:3), who according to Acts 8:3 was entering κατὰ τοὺς οἴκους (house after house) to drag off men and women, turns out to have had no known face for the assemblies of Judaea. It is not an insurmountable contradiction (it has been said that ‘Judaea’ does not include Jerusalem), but it forces a choice between the persecutor of houses in Acts and the unknown man of Galatians.

And we come to the piece that I found hardest to believe. The rabbinic tradition preserves Gamliel’s prestige, and also his social profile. On the prestige, the Mishnah sums it up thus:

מִשֶּׁמֵּת רַבָּן גַּמְלִיאֵל הַזָּקֵן, בָּטַל כְּבוֹד הַתּוֹרָה, וּמֵתָה טָהֳרָה וּפְרִישׁוּת (Sotah 9:15) (since Rabban Gamliel HaZaken died, the glory of the Torah has ceased, and purity and separation have died).

פְּרִישׁוּת (perishut, ‘separation’) is the root from which the very name of the Pharisees comes (Aramaic פְּרִישַׁיָּא, Greek Φαρισαῖοι): according to the Mishnah, with Gamliel the Pharisaic ideal dies. Luke picks up that prestige, not a doctrine: νομοδιδάσκαλος τίμιος παντὶ τῷ λαῷ (‘teacher of the Torah, honoured by all the people’; Acts 5:34), where τίμιος stands in for כָּבוֹד.

On the social profile, the Tosefta records an exception to a decree. The Mishnah prohibits teaching Greek:

שֶׁלֹּא יְלַמֵּד אָדָם אֶת בְּנוֹ יְוָנִית (Sotah 9:14) (that no one should teach his son Greek).

But:

הִתִּירוּ לְבֵית רַבָּן גַּמְלִיאֵל לְלַמֵּד אֶת בְּנֵיהֶם יְוָנִית מִפְּנֵי שֶׁהֵן קְרוֹבִין לַמַּלְכוּת (t. Sotah 15:8, ed. Lieberman) (they permitted the house of Rabban Gamliel to teach their sons Greek because they are close to the government).

The Bavli repeats the exception (שֶׁקְּרוֹבִין לַמַּלְכוּת; Sotah 49b; Bava Kamma 83a) and attributes to Rabban Shimon ben Gamliel the recollection that in his father’s house there were a thousand children: חֲמֵשׁ מֵאוֹת לָמְדוּ תוֹרָה וַחֲמֵשׁ מֵאוֹת לָמְדוּ חָכְמַת יְוָנִית (‘five hundred studied Torah and five hundred studied Greek wisdom’). Two philological cautions. First: the Erfurt manuscript of the Tosefta does not say קְרוֹבִין but זְקוּקִין (in need of, dependent upon), as does the Yerushalmi for the house of Rabbi (Shabbat 6:1, 7d: שֶׁהָיוּ זְקוּקִין לַמַּלְכוּת); the reason for the permission oscillates between ‘closeness’ and ‘dependence’ on power. Second, and more important: it is not clear of which Gamliel that tradition speaks, and a frequent reading refers it to Gamliel of Yavne, a generation after Paul; it cannot date anything in his time. What it does do is confirm a profile—an elite Pharisaic house, Greek-speaking and attached to power—that Josephus attests before the year 70 with Simon (γένους σφόδρα λαμπροῦ, of very illustrious lineage).

Observe the word. קָרוֹב means ‘near’ and, said of persons, ‘relative’; it is the Hebrew for what the Greek of Josephus says with προσήκοντες τῷ βασιλεῖ κατὰ γένος (‘belonging to the king by lineage’; War 2.419) and Hellenistic Greek with συγγενεῖς τῶν βασιλέων (1 Macc 10:89). Three languages, a single social category: that of those who are near the throne. Rashi, glossing the ‘Greek wisdom’ of that passage, uses another loanword:

לשון חכמה שמדברים בו בני פלטין (‘language of wisdom that is spoken by the “sons of the palace”’; Rashi on Sotah 49b),

where פלטין is the Greek παλάτιον and the Latin palatium. A caveat: in Tannaitic Hebrew מלכות usually designates Rome, although in the time of Agrippa II it could also designate the Herodian house; the text does not choose.

And now the Aramaic. The Tosefta and the two Talmuds preserve three circular letters of Gamliel HaZaken, dictated to the scribe Yohanan from the steps of the Temple Mount, which follow a fixed mould:

לאחנא בני גלילאה עילאה ולאחנא בני גלילאה תתאה שלומכון יסגא מהודעין אנחנא לכון דזמן ביעורא מטא (t. Sanhedrin 2:6; b. Sanhedrin 11b)

(‘to our brothers of Upper Galilee and to our brothers of Lower Galilee: may your peace increase; we make known to you that the time of the removal of the tithes has come’ [ביעורא]).

In a version of the Yerushalmi, Sanhedrin 1:2, the third letter is addressed לאחנא בני גלותא דבבל ובני גלותא דמדי ובני גלותא דיון (‘to our brothers of the exile of Babylon, and of the exile of Media, and of the exile of Yavan, that is, Greece’). A Pharisaic circular of the Temple, then, written in Aramaic and destined also for the Greek diaspora:

תַּנֵּי. אָמַר רִבִּי יוּדָן. מַעֲשֶׂה בְרַבָּן גַּמְלִיאֵל וּזְקֵנִים שֶׁהָיוּ יוֹשְׁבִין עַל גַּב מַעֲלָה בְהַר הַבַּיִת וְיוֹחָנָן סוֹפֵר הַלָּז יוֹשֵׁב לִפְנֵיהֶן. אָמַר לוֹ רַבָּן גַּמְלִיאֵל. כְּתוֹב. לַאֲחָנָא בְנֵי דְרוֹמָא עילָּאָה לַאֲחָנָא בְנֵי דְרוֹמָא אַרְעַיָיא שְׁלַמְכוֹן יִסְגֶּא. מוֹדַעְנָא לְכוֹן דְּמָטָא זְמַן בִּיעוּרָא. לַאֲפוּקֵי מַעְשְׂרַיָיא מֵעוּמְרֵי שׁוֹבְלַיָיא. וְלַאֲחָנָא בְנֵי גָלִילָא עִילָּאָה וּבְנֵי גָלִילָא אַרְעַיָיה שְׁלַמְכוֹן יִסְגֶּא. מוֹדַעְנָא לְכוֹן דְּמָטָא זְמַן בִּיעוּרָא. תַּפְקוֹן מַעְשְׂרַיָא מִמַּעְטָנֵי זֵיתַיָּא. לַאֲחָנָא בְנֵי גָּלוּתָא דְבָבֶל בְּנֵי גָּלוּתָא דְמָדַי בְּנֵי גָּלוּתָא דְיָוָון וּשְׁאָר כָּל גַּלְוָותָא דְיִשְׂרָאֵל שְׁלַמְכוֹן יִסְגֶּא. מוֹדַעְנָא לְכוֹן דְּאִימְּרַיָּא רַכִּיכִין וְגוֹזָלַיָא דַקִּיקִין וְזִימְנָא דַאֲבִיבָא לָא מְטָא וּשְׁפַר מִילְּתָא בַּאַפַּיי וּבְאַנְפֵּי חֲבֵרַיי מוֹסְפָא עַל שַׁתָּא דָא תַלְתִּין יוֹמִין.

It is taught in a baraita. Rabbi Yudan said: It happened that Rabban Gamliel and the elders were sitting on the steps of the Temple Mount, and Yohanan the scribe was sitting before them. Rabban Gamliel said to him: Write.

‘To our brothers, the people of the Upper South, and to our brothers, the people of the Lower South: may your peace increase (shelamkhon yisge). We make known to you that the time of removal (bi’ur) has arrived, to bring out the tithes from the sheaves of ears.’

‘And to our brothers, the people of Upper Galilee, and the people of Lower Galilee: may your peace increase. We make known to you that the time of removal has arrived. Bring out the tithes from the olive presses.’

‘To our brothers, the people of the exile of Babylon, the people of the exile of Media, the people of the exile of Greece, and all the other communities of the exile of Israel: may your peace increase. We make known to you that the lambs are still tender, and the pigeons still small, and the season of spring (aviv) has not arrived. And the matter seemed good to me and to my colleagues, and we have agreed to add thirty days to this year.’

The greeting שלמכון יסגא (shelamkhon yisge, ‘may your peace increase’) is the formula of the Aramaic chancellery; it is already in Daniel 3:31 (4:1 in other numberings), and the Greek rendered it with the optative of πληθύνω (to multiply): εἰρήνη ὑμῖν πληθυνθείη (‘may peace be multiplied to you’; Theodotion). That formula passes almost intact to three letters of disputed authorship—χάρις ὑμῖν καὶ εἰρήνη πληθυνθείη (1 Pet 1:2; 2 Pet 1:2; Jude 2, where ἔλεος takes the place of χάρις)—and to First Clement. In no letter of Paul. Paul’s greeting is χάρις ὑμῖν καὶ εἰρήνη (‘grace to you and peace’), a hybrid of the Greek χαίρειν and the Hebrew שלום that does not carry the verb of the Aramaic formula. His notice formula, γνωρίζω ὑμῖν, ἀδελφοί (‘I make known to you, brothers’; 1 Cor 15:1; Gal 1:11), has its twin in מהודעין אנחנא לכון, but γνωρίζω is also a formula of Greek letters on papyrus and does not discriminate. The criterion I propose is therefore the greeting, which does discriminate: the chancellery of Rabban Gamliel HaZaken and that of Peter and Jude share an Aramaic imprint that Paul’s does not have. It is an indication a contrario—what does not appear—, it weighs little, and I note it because it can be measured.

Paul, for his part, transcribes three Aramaic words: Ἀββά (abba, ‘father’; Rom 8:15; Gal 4:6), μαρὰν ἀθά (maran atha, ‘our Lord, come’, or ‘has come’; 1 Cor 16:22) and Κηφᾶς (kepha, ‘rock’; Gal 1:18 and other places). All three are vocabulary of the movement: an invocation, a liturgical acclamation and a nickname. None is a school term. None of the names with which the house of study designates its craft passes into his Greek: neither פשר (pesher, ‘interpretation’), nor מדרש (midrash, ‘exegesis’), nor הלכה (halakhah), nor גזרה שוה (gezerah shavah, verbal analogy), nor קל וחומר (qal va-khomer, from the light to the grave). And the steps of reasoning that have been wished to be seen as rabbinic, such as that from the lesser to the greater of Romans 5:15–17 (πολλῷ μᾶλλον, ‘much more’), also do not require a school: their Hellenistic rhetorical parallels were shown by Daube (1949) and Lieberman (1950), so that they are, at most, a hermeneutical influence and not a trace of language. A pupil of Rabban Gamliel HaZaken with the training that Luke says would have a school Aramaic; Paul’s is that of the community which, by his own account, he went to get to know: ἱστορῆσαι Κηφᾶν (‘to get to know Cephas’; Gal 1:18).

Seventh piece, which joins them: from nearness to magisterium. I now join the pieces. To build his scholastic Saul, Luke has at his disposal a Greek formula of biography; a template from Josephus for describing the Pharisaic elite; a Gamliel whose prestige is recalled by rabbinic tradition and whose house, according to that same tradition, enjoyed a singular licence for Greek; a problem to solve—how a Pharisee of Jerusalem writes in the Greek of the diatribe and cites Scripture in its Greek version, and how an ‘exact’ man can be a persecutor—; and a single firm datum from the letters: that he was a Pharisee and zealous. From that material comes ‘Saul, disciple of Gamliel’ without anyone having studied with anyone.

And here is the conclusion that I hold: ‘Gamliel’ is in Luke a social label, not a pedagogical relationship. The rabbinic licence gives the key. In that house one entered Greek on a condition, and the condition, in the Bavli’s formulation, was קרובין למלכות. If a reader who knows that the Pharisees of Jerusalem did not teach Greek to their sons has to be given an explanation of how Paul writes as he does, the only Pharisaic house that Luke can name without sounding absurd is the one that the tradition exempts from that rule; and that house admits to its Greek wing only those ‘near the throne’. Either Luke fabricated a pedagogical provenance for a social profile that the letters let one glimpse, or he preserved, translated into the language of the school, the trace of a kinship that the Saulus of Josephus displays without concealment. In both cases the movement is the same: from nearness to magisterium. Luke converts a social class into a teacher.

These are the three points from which this reading would be attacked: the rabbinic tradition is late and perhaps refers to Yavne; the variant זקוקין is still alive; and this piece, by itself, does not prove Paul’s Herodian kinship: it proves that the label ‘Gamliel’ does not exclude it, and the thesis of kinship explains why Luke chose precisely that one.

What I can say without reservation is what Luke does not say. Look in Acts for a single line of doctrine that Gamliel addresses to Saul. There is none. The only teaching that Luke preserves from Gamliel is that of Acts 5:38–39, addressed to the Sanhedrin, and it says: let them alone, lest they fight against God. A disciple who does the opposite of the only teaching preserved from his teacher, and whose firm datum about his training, in the letters, is κατὰ νόμον Φαρισαῖος and κατὰ ζῆλος διώκων, is a disciple whose teacher is better explained as literary prologue than as biography.

And now the objections that there would be. I do not evade them: each costs something, and I want to say how much.

Chronology. It is the strongest objection, and it has three parts. First: the Delphi inscription fixes Gallio’s proconsulship about 51–52, but it fixes Gallio, not Paul; that Paul appeared before him is said by only one text, Acts 18:12, that is, Luke, the witness whose biography of Paul we are dismantling. The letters are anchored at only two points: Aretas, before the year 40 (2 Cor 11:32), and the collection for Jerusalem, about the middle of the fifties (Rom 15:25–28). After that they say nothing. Second: the ancient chronology of the Church itself leaves the window open. Jerome (De viris illustribus 5) writes:

hic ergo quarto decimo Neronis anno, eodem die quo Petrus Romae, pro Christo capite truncatur (this man, then, in the fourteenth year of Nero, on the same day as Peter, is beheaded at Rome for Christ),

that is, in 67–68, ‘thirty-seven years after the passion’ according to his reckoning. The date of 64, defended by those who link it to the fire of Rome, is a modern reconstruction. With the ancient date, Saulus’s last appearance, sent to Nero in Achaea towards the end of 66 (War 2.558), falls shortly before Paul’s death, not after it. Acts ends about 62 with Paul at Rome (28:30: ἐνέμεινεν δὲ διετίαν ὅλην ἐν ἰδίῳ μισθώματι, he remained two whole years in his own rented dwelling), and later tradition (1 Clement 5:7: ἐπὶ τὸ τέρμα τῆς δύσεως ἐλθών, ‘having come to the limit of the west’) knows a Paul who is free and on the move after that. The window from 62 to 67 is empty of Pauline testimony; it is not contradicted by it. Third, and this is the real cost: Josephus places Saulus at the head of mobs under Albinus (62–64), and Paul, by his own letters, was an apostle before the year 40. My answer is that Ant. 20.214 does not narrate a dated episode but characterises a conduct: Josephus closes it with ἐξ ἐκείνου μάλιστα τοῦ καιροῦ (‘above all from that time’), and the only action of Saulus that he dates with precision is the embassy of 66. It is a possible reading of the text, not the only one. Whoever rejects it will have to explain why Josephus would date that conduct so vaguely; whoever accepts it, that Josephus compresses decades of behaviour into one sentence.

Status. Paul works with his hands: 1 Thess 2:9 (νυκτὸς καὶ ἡμέρας ἐργαζόμενοι, ‘working night and day’), 1 Cor 4:12 (κοπιῶμεν ἐργαζόμενοι ταῖς ἰδίαις χερσίν, ‘we toil, working with our own hands’) and 2 Cor 11:27 (ἐν κόπῳ καὶ μόχθῳ ... ἐν λιμῷ καὶ δίψει, ‘in toil and hardship ... in hunger and thirst’). The objection says that this is not an aristocrat. It has two answers. The first is cultural: working with the hands was a norm of the Pharisaic sages, not a mark of the lower class (יָפֶה תַלְמוּד תּוֹרָה עִם דֶּרֶךְ אֶרֶץ, ‘fine is the study of Torah together with an occupation’; Mishnah, Avot 2:2, attributed to a later Gamliel), so that an elite Pharisee with a trade is normal. The second is the one you have already seen: Philippians 3:7–8, τὰ πάντα ἐζημιώθην (‘I lost everything’), which for a noble would be literal. Neither proves the identification; but with them the objection ceases to count against it.

The name. The link Saul = Paul is supplied by Luke (Acts 13:9: Σαῦλος, ὁ καὶ Παῦλος). In the letters ‘Saul’ does not appear; in Josephus ‘Paul’ does not appear. It is the most serious methodological objection: the thesis needs Luke for the name and at the same time declares him unreliable in all else. The answer is that it does not need him for the substance. What identifies the man is the convergence of vocabulary, of offices and of class between Josephus and the letters themselves (observations first, second and third), and that convergence does not pass through Luke. Luke contributes only the second name. A double name, Semitic and Roman, is normal in the Herodian class; that this name should also be that of the persecutor is what the thesis predicts. The name confirms; it does not support.

Chance. Saul was a common Jewish name (Josephus himself records another, the father of Simon of Scythopolis: Σαούλου τινὸς τῶν οὐκ ἀσήμων, ‘a certain Saul, of the not obscure’; War 2.469), and the Herodians formed a class in which kinships with Agrippa were numerous. For that reason the name alone proves nothing; what counts is what accompanies it: the lineage, the two offices, the vocabulary of Galatians. Sandmel (1962) warned that whoever seeks parallels finds them, and I sought. The defence against that is not to deny it but to expose oneself to failure: that is why I formulated each observation so that it could fall.

The disputed letters. I used Ephesians, 2 Timothy and Hebrews, whose Pauline authorship (or, in the case of Hebrews, whose very authorship) is disputed. If they are discarded, the first and second observations lose part of their support; Galatians, Romans and 2 Corinthians remain, which suffice to sustain the others, but the dependence must not be concealed.

The silence of Josephus. Josephus recounts the death of James / Ya’acov bar Pandira in Antiquities 20.200 (cf. Adversus haereses 78.7.5), and does not link it to Saul: the initiative is that of the high priest Ananus (Janan), who, believing he had a favourable occasion because Festus had died and Albinus was still on the way, καθίζει συνέδριον κριτῶν (convenes a tribunal of judges), brings before it Ya’acov bar Pandira, accused of transgressing the Torah, ‘and some others’, and hands them over λευσθησομένους (to be stoned):


… ἅτε δὴ οὖν

τοιοῦτος ὢν ὁ Ἄνανος· νομίσας ἔχειν καιρὸν ἐπιτήδειο(ν)

διὰ τὸ τεθνάναι μὲν Φῆστον, Ἀλβῖνον δ᾽ ἔτι κατὰ τὴν ὁδό(ν)

ὑπάρχειν· καθίζει συνέδριον κριτῶν· καὶ παραγαγὼν

εἰς αὐτὸ τὸν ἀδελφὸν Ἰησοῦ τοῦ λεγομένου Χριστοῦ Ἰά-

κωβος ὄνομα αὐτῷ· καί τινας ἑτέρους ὡς παρανομη-

σάντων κατηγορίαν ποιησάμενος, παρέδωκε λευ-

σθησομένους·

(»וְחָנָן, בִּהְיוֹתוֹ אִישׁ כָּזֶה, אָמַר בְּלִבּוֹ: עֵת רָצוֹן לִי, יַעַן כִּי מֵת פֶּסְטוֹס וְאַלְבִּינוֹס עוֹדֶנּוּ בַּדֶּרֶךְ. וַיּוֹשֶׁב סַנְהֶדְרִין (עֲדַת שֹׁפְטִים), וַיָּבֵא לְפָנֶיהָ אֶת יַעֲקֹבוֹס (יַעֲקֹב) אֲחִי יֵשׁוּ [»הָאוֹמְרִים עָלָיו מָשִׁיחַ»] וְאֶת אֲנָשִׁים אֲחֵרִים, וַיָּעַד בָּהֶם כִּי עָבְרוּ אֶת הַתּוֹרָה, וַיִּתְּנֵם לְהִסָּקֵל.«)

‘So then, Ananus (Hanan), being such a man and thinking that he had a convenient occasion because Festus was dead and Albinus was still on the road, convenes a tribunal of judges and, bringing before it the brother of Ieshu [who is called Christ], James by name, and certain others, having brought an accusation against them as transgressors of the Torah, delivered them up to be stoned.’

—Flavius Josephus, Jewish Antiquities 20.200, Milan, Veneranda Biblioteca Ambrosiana, Ms. F 128 sup. (gr. 370), 11th century, fol. 349v, digital copy, consulted 20 November 2026, https://digitallibrary.unicatt.it/veneranda/0b02da82800acbda.

Two features of the language weigh more than the absence of a name. The first is the verb: καθίζω is the one for constituting a tribunal, and Josephus repeats it in 20.202, where those who protest to Albinus allege that it was not lawful for Ananus καθίσαι συνέδριον without his consent; the reproach concerns not the substance of the accusation but who had competence to seat the tribunal. The second is the apportioning of schools: Ananus is of the αἵρεσις of the Sadducees, the harshest in judgements (20.199), and those who are indignant are those who passed for ἀκριβεῖς περὶ τοὺς νόμους (20.201), the ‘exact’ of the template we saw in Simon and in Acts 26:5. The persecutor of Ya’acov bar Pandira is, in Josephus, a Sadducee who proceeds by tribunal, and the exact are on the other side: the same separation between the ‘exact’ and the persecutor that we already saw in taking Acts 22:3 apart. Saul is something else. Fourteen paragraphs later, in 20.214, he convenes nothing: συνῆγον μοχθηρὰ πλήθη, he gathers mobs, αὐτοὶ καθ᾽ αὑτούς, on their own. And he does so immediately after Agrippa gave the high priesthood to a Jesus son of Gamliel—Ἰησοῦς ὁ τοῦ Γαμαλιήλου—(20.213; the identification with Yehoshua ben Gamla (יְהוֹשֻׁעַ בֶּן גַּמְלָא) is disputed, but the name, as Josephus writes it, is the same that Luke puts in Acts 5:34). That Josephus does not count Saul among the executioners of Ya’acov bar Pandira is not, then, silence about a fact, but a division of violences: the tribunal for Ananus, the mob for Saul, and a Gamliel in the same passage. I do not say that Luke read Josephus thus; I say that, if he read him, he had a teacher and a persecutor a few lines apart.

Different is the account of the Pseudo-Clementines, and it is where philology yields most, because there there is a nameless persecutor. A prior warning: Frankenberg’s Greek is, as I understand it, a reconstruction made upon the Syriac and Rufinus’s Latin, not a Greek manuscript, so that each word must be checked against the Syriac. The latter calls the man baʿldbābā (ܒܥܠܕܒܒܐ), literally ‘lord of hostility’, the noun with which the Peshitta usually renders ἐχθρός in the parable of the tares; the ‘man’ is already inside baʿl, and the Greek writes it apart, ἄνθρωπος ἐχθρός, the exact turn of Matt 13:28, which the Letter of Peter to Ya’acov bar Pandira also uses for the adversary of his preaching (2.3: ἀνθρώπου ἐχθροῦ). Paul knew that word applied to him: in Galatians 4:16 he picks up the accusation, ὥστε ἐχθρὸς ὑμῶν γέγονα ἀληθεύων ὑμῖν; (have I become your enemy by telling you the truth?), five verses before the allegory of Hagar. The epithet does not seem an invention of his adversaries: they took it from a reproach that he himself preserves. The scene, moreover, does not have the form of a trial. The enemy goes up to the temple μετ᾽ ὀλίγων (Syriac ʿam qalilin, עם קלילין ‘with a few’), ἤρξατο θορυβεῖν τὰ πάντα (Syriac lmazyaʿ kul medem, למזיע כל מדם ‘to shake everything’), snatches a brand from the altar and strikes: mob and tumult, Saul’s manner and not Ananus’s. There is even a thread of vocabulary, a weak indication because ἁρπάζω is a common word: the enemy reproaches the priests for letting themselves be ἁρπάζεσθαι (metḥaṭṭefin) by men led astray after a γόης, and shortly afterwards it is he who ἁρπάσας (ḥṭap, the same Syriac root) the brand. It is the alternation between the one who seizes and the one who is seized that the article has already pointed out between Josephus and 2 Corinthians 12. And the γόης is the same word with which Josephus dismisses Theudas (Ant. 20.97), so that the source speaks the language with which popular prophets of the first century were qualified (the Syriac ḥarrāšā, ܚܪܫܐ, ‘sorcerer’, is a conjectural reading). The text preserved in Die syrischen Clementinen mit griechischem Paralleltext, pp. 74–75, says:

70,1 καὶ τότε ἄνθρωπός τις ἐχθρὸς μετ᾽ ὀλίγων εἰς τὸν ναὸν ἀναβὰς πρὸς τὸν βωμὸν ἐβόα· τί ποιεῖτε, ἄνδρες Ἰσραηλῖται, πῶς προπετῶς 2 ἁρπάζεσθε ὑπ᾽ ἀνθρώπων ταλάνων μετὰ γόητα πλανωμένων; 3 πλὴν τοιαῦτα καταλαλῶν καὶ ἐναντία ἀκούων ὑπὸ Ἰακώβου τοῦ ἐπισκόπου ἐλεγχθεὶς ἤρξατο θορυβεῖν τὰ πάντα, ἵνα μὴ τὰ καλῶς λεγόμενα ἡσύχως δοκιμασθέντα πιστευθείη. 4 ὅθεν μεγάλῃ φωνῇ κατεβόα τῆς τε μωρίας καὶ κατηφείας τῶν ἱερέων καὶ ὀνειδίζων, τί, ἔφη, ὑποστέλλεσθε; 5 τί οὐ τοὺς μετ᾽ αὐτοῦ πάντας εὐθέως ἐπικρατεῖτε; 6 τοῦτο λέγων αὐτὸς πρῶτος προπηδῶν δαλὸν ἀπὸ τοῦ βωμοῦ ἁρπάσας ἤρξατο κόπτειν· 7 οἱ δὲ λοιποὶ τῶν ἱερέων ὁρῶντες ὡσαύτως ἔπραξαν. 8 πολλῆς δὲ φυγῆς γενομένης οἱ μὲν ἐπὶ τοῖς ἑταίροις σφαλλόμενοι ἔπεσον, οἱ δὲ πληγὰς ἔλαβον· καὶ οὐκ ὀλίγοι ἀπέθανον ὥστε πολὺ αἷμα τῶν φονευθέντων ἐχύθη. ἐκεῖνος δὲ ὁ ἐχθρὸς τὸν Ἰάκωβον ἄνωθεν τῶν βαθμῶν ἀπέωσε, πεσόντα δὲ ὡς νεκρὸν κείμενον δεύτερον οὐκ ἔτυψεν. οἱ δὲ

Translation of the Greek text

‘70.1 And then a certain hostile man², with a few others, went up into the temple and, standing at the altar, cried out: “What are you doing, Israelite men? Why are you so rashly 2 being carried off³ by wretched men who have gone astray after a sorcerer⁴?” 3 But, as he slandered⁵ in this way and heard the contrary from James the bishop⁶ and was refuted, he began to throw everything into confusion, so that what was well said, if calmly examined, should not come to be believed. 4 Therefore with a loud voice he cried out against the folly and dejection⁷ of the priests and, reproaching them, said: “Why do you hold back⁸? 5 Why do you not at once overpower all who are with him?” 6 Saying this, he himself, leaping forward first, snatched a brand⁹ from the altar and began to strike; 7 and the rest of the priests, seeing it, did likewise. 8 A great flight having occurred, some fell, stumbling over their companions¹⁰, others received blows; and not a few died, so that much blood of the slain was shed. And that enemy cast James down from the top of the steps¹¹ and, when he had fallen and lay as if dead, did not strike him a second time.’

Notes

¹ ἐξ αὐτῆς: literally ‘from it’. It is probably the adverbial phrase ἐξαυτῆς (‘at once’, with ὥρας understood). The Syriac men hāydēn (‘from then’) confirms the temporal sense.

² ἐχθρὸς ἄνθρωπος: it is the turn of Matt 13:28 (the enemy who sows the tares) and corresponds to Rufinus’s inimicus homo. The text does not name him, but chapter 71 sends him to Damascus, and for that reason most interpreters identify him with Paul. The source would be the hypothetical Anabathmoi Iakōbou (‘Ascents of James’), of Jewish-Christian origin.

³ ἁρπάζεσθε: passive, ‘to be seized, dragged off’. There is a nuance of violence or irresistible seduction.

⁴ γόης: ‘deceiver, sorcerer’. It is the term of ancient polemic against Yeshu as seducer of the people.

⁵ καταλαλῶν: ‘to speak against, to slander’ (cf. Jas 4:11).

⁶ ἐπίσκοπος: the title that the Pseudo-Clementines give to Ya’acov bar Pandira, brother of Yeshu, as head of the church of Jerusalem. It is probably anachronistic with respect to the first century.

⁷ κατήφεια: ‘downcast countenance, sadness, shame’. The Syriac kəryutā has a very close sense.

⁸ ὑποστέλλεσθε: ‘to draw back, to shrink back through cowardice’ (cf. Gal 2:12, where Peter ‘withdrew’).

⁹ δαλός: ‘firebrand, burning log’. Here it is a remnant of the sacrificial fire of the altar used as a weapon.

¹⁰ ἐπὶ τοῖς ἑταίροις σφαλλόμενοι: literally ‘stumbling upon their companions’. It describes the panic of the crowd.

¹¹ βαθμοί: ‘steps, stairs’. It refers back to the title Anabathmoi (‘ascents’) of the source, since the scene takes place on the temple staircase. In 70.8 ἀπέωσε is the aorist of ἀπωθέω (‘to push away, to cast down’).

 


70,1 ܘܗܝܕܝܢ ܒܥܠܕܒܒܐ ܚܕ ܥܡ ܩܠܝܠܝܢ ܣܠܩ ܠܗܝܟܠܐ ܘܩܡ ܥܠ ܡܕܒܚܐ ܘܩܥܐ ܘܐܡܪ܂ ܡܢܐ ܥܒܕܝܢ ܐܢܬܘܢ ܓܒܪܐ ܝܣܪܐܝܠܝܐ܂ 2 ܐܝܟܢܐ ܡܬܚܛܦܝܢ ܐܢܬܘܢ ܡܢ ܒܢܝ ܐܢܫܐ ܡܣܟܢܐ ܕܒܬܪ ܚܪܫܐ ܛܥܝܢ܂ 3 ܐܠܐ ܟܕ ܗܟܢܐ ܡܠܥܙ ܗܘܐ ܘܫܡܥ ܡܢ ܝܥܩܘܒ ܐܦܝܣܩܘܦܐ ܡܠܐ ܕܠܘܩܒܠܗ ܘܐܬܟܣܦ ܫܪܝ ܠܡܙܝܥ ܟܠ ܡܕܡ ܕܠܐ ܢܗܝܡܢܘܢ ܒܡܠܐ ܕܫܦܝܪ ܐܡܝܪ ܟܕ ܒܫܠܝܐ ܐܬܒܚܪ܂ 4 ܡܛܠ ܗܕܐ ܩܥܐ ܒܩܠܐ ܪܒܐ ܥܠ ܣܟܠܘܬܐ ܘܥܠ ܟܪܝܘܬܐ ܕܟܗܢܐ ܘܡܚܪܦ ܗܘܐ ܠܗܘܢ ܘܐܡܪ ܡܢܐ ܡܬܟܠܝܢ ܐܢܬܘܢ܂ 5 ܘܠܡܢܐ ܠܐ ܐܚܕܝܢ ܐܢܬܘܢ ܠܟܠܗܘܢ ܕܥܡܗ ܒܥܓܠ܂ 6 ܘܟܕ ܐܡܪ ܗܠܝܢ ܩܦܙ ܗܘ ܩܕܡܝܐ ܘܚܛܦ ܡܢ ܡܕܒܚܐ ܠܦܝܕܐ ܘܫܪܝ ܡܚܐ܂ 7 ܘܐܚܪܢܐ ܕܟܗܢܐ ܚܙܘ ܘܥܒܕܘ ܗܟܢܐ܂ 8 ܘܡܢ ܒܬܪ ܕܗܘܬ ܥܪܩܬܐ ܣܓܝܐܬܐ ܐܢܫܝܢ ܐܢܫܝܢ ܢܦܠܘ ܥܠ ܚܒܪܝܗܘܢ ܘܐܬܟܫܠܘ ܘܐܚܪܢܐ ܡܚܝܘ ܘܠܐ ܩܠܝܠܝܢ ܡܝܬܘ ܥܕܡܐ ܕܕܡܐ ܣܓܝܐܐ ܕܩܛܝܠܐ ܐܬܐܫܕ܂ ܗܘ ܕܝܢ ܒܥܠܕܒܒܐ ܠܝܥܩܘܒ ܡܢ ܠܥܠ ܡܢ ܕܪܓܐ ܕܚܝܗ ܘܟܕ ܢܦܠ ܣܒܪ ܕܡܝܬ ܗܘ ܘܠܐ ܬܢܝܢܐܝܬ ܡܚܝܗ܂ ܘܗܢܘܢ ܕܝܢ

70,1 והידין בעלדבבא חד עם קלילין סלק להיכלא וקם על מדבחא וקעא ואמר׃ מנא עבדין אנתון גברא יסראיליא׃ 2 איכנא מתחטפין אנתון מן בני אנשא מסכנא דבתר חרשא טעין׃ 3 אלא כד הכנא מלעז הוא ושמע מן יעקוב אפיסקופא מלא דלוקבלה ואתכסף שרי למזיע כל מדם דלא נהימנון במלא דשפיר אמיר כד בשליא אתבחר׃ 4 מטל הדא קעא בקלא רבא על סכלותא ועל כריותא דכהנא ומחרף הוא להון ואמר מנא מתכלין אנתון׃ 5 ולמנא לא אחדין אנתון לכלהון דעמה בעגל׃ 6 וכד אמר הלין קפז הו קדמיא וחטף מן מדבחא לפידא ושרי מחא׃ 7 ואחרנא דכהנא חזו ועבדו הכנא׃ 8 ומן בתר דהות ערקתא סגיאתא אנשין אנשין נפלו על חברהון ואתכשלו ואחרנא מחיו ולא קלילין מיתו עדמא דדמא סגיאא דקטילא אתאשד׃ הו דין בעלדבבא ליעקוב מן לעל מן דרגא דחיה וכד נפל סבר דמית הו ולא תנינאית מחיה׃ והנון דין

Translation of the Syriac text

‘70.1 And then an adversary¹, with a few others, went up to the temple (heikla), stood at the altar and cried out, saying: “What are you doing, Israelite men? 2 How are you being carried off by sons of men, wretched ones, who go astray after a sorcerer²?” 3 But as he spoke thus with contempt and heard from Ya’acov, the epískopo’³, words contrary to his own, and was confounded, he began to stir up everything, so that the word well spoken, which had been examined in tranquillity, should not be believed. 4 Therefore he cried out in a great voice against the folly and the dejection⁴ of the priests and was reviling them and said: “Why do you hold yourselves back⁵? 5 And why do you not seize at once all who are with him?” 6 And when he had said these things, he leapt first, and snatched a torch⁶ from the altar and began to strike. 7 And the other priests saw and did likewise. 8 And after a great flight⁷ had occurred, some fell upon their companions and stumbled, and others were struck, and not a few died, until much blood of the slain was shed. And that adversary cast Ya’acov down from the top of the steps⁸ and, when he fell, thought that he was dead and did not strike him a second time.’

Notes

¹ ܒܥܠܕܒܒܐ (baʿl debābā): literally ‘lord of enmity’, that is, ‘adversary, enemy’. It is the term with which the Peshitta usually renders ἐχθρός (for example in the parable of the tares) and it is also an appellation of the devil. For that reason the Syriac text has a more demonising tone than the Greek.

² ܚܪܫܐ (ḥarrāšā): ‘sorcerer, magician’. It is more strictly ‘wizard’ than γόης, which belongs to the field of the ‘deceiver’. The reading is conjectural.

³ ܐܦܝܣܩܘܦܐ: a direct loan from the Greek ἐπίσκοπος. ܝܥܩܘܒ (Yaʿqōḇ) is the same name as Jacob. The Spanish ‘Santiago’ derives from Sanctus Iacobus.

⁴ ܣܟܠܘܬܐ (sāklūṯā, ‘folly’) and ܟܪܝܘܬܐ (karyūṯā, ‘dejection, heaviness of heart’) correspond to μωρία and κατήφεια. The reading of karyūṯā is conjectural.

⁵ ܡܬܟܠܝܢ (metkalyin): ‘you restrain yourselves, you hold yourselves back’. It corresponds to ὑποστέλλεσθε, but the exact form of the reading is uncertain.

⁶ ܠܦܝܕܐ (lappīdā): ‘torch, firebrand’. It corresponds to δαλός. The reading is conjectural, and it could also be gemrā (‘ember’).

⁷ ܥܪܩܬܐ (ʿarqtā): ‘flight’, with the same sense of panic as φυγή.

⁸ ܕܪܓܐ (darghē): ‘steps, stairs’. In the context it refers back to the source Anabathmoi (‘ascents’). The reading is conjectural.

The aggressor does not receive there the name of Saul. However, in I, 71 it is related that he obtains from the high priest Caiaphas (Qayafa) a commission to persecute the believers in Yeshu and sets out for Damascus provided with letters (ʾegartā):

διηγήσατο οὖν ἡμῖν ὡς ὁ ἐχθρὸς ἐκεῖνος ἐνώπιον τῶν ἱερέων Καιάφᾳ τῷ ἀρχιερεῖ ὑπέσχετο πάντας τοὺς εἰς Ἰησοῦν πιστεύοντας ἀνελεῖν· 4 καὶ εἰς Δαμασκὸν ἐξῆλθεν ἐπιστολὰς αὐτῶν λαβών, ὅπως ἐκεῖ αὐτῷ ἐλθόντι οἱ ἄπιστοι συνεργῶσι τοὺς πιστεύοντας διαφθεῖραι.

(‘He narrated to us, then, how that enemy, before the priests, had promised Caiaphas the high priest to destroy all who believe in Jesus; 4 and he went out to Damascus, having taken letters from them, so that when he came there the unbelievers might help him to destroy the believers.’)

ܘܐܣܦܪ ܠܢ ܕܒܥܠܕܒܒܐ ܗܘ ܩܕܡ ܟܗܢܐ ܐܫܬܘܕܝ ܠܩܝܦܐ ܪܒ ܟܗܢܐ ܕܢܩܛܘܠ ܠܟܠܗܘܢ ܕܡܗܝܡܢܝܢ ܒܝܫܘܥ܂ 4 ܘܢܦܩ ܠܕܪܡܣܘܩ ܘܢܣܒ ܐܓܪܬܐ ܡܢܗܘܢ ܕܟܕ ܢܐܬܐ ܠܬܡܢ ܠܐ ܡܗܝܡܢܐ ܢܥܕܪܘܢܗ ܠܡܘܒܕܘ ܠܡܗܝܡܢܐ܂

ואספר לן דבעלדבבא הו קדם כהנא אשתודי לקיפא רב כהנא דנקטול לכלהון דמהימנין בישו׃ 4 ונפק לדרמסוק ונסב אגרתא מנהון דכד נאתא לתמן לא מהימנא נעדרונה למובדו למהימנא׃

(‘And he told us that that adversary (baʿl debābā) had declared before the priests, to Qayafa the high priest (rab kahana), that he would kill all who believe in Yeshu; 4 and that he went out to Damascus and took letters from them, so that when he arrived there the unbelievers might help him to destroy the believers [in Yeshu]’; cf. Acts 9:1–2.)

There remains the setting. The enemy hurls Ya’acov bar Pandira ἄνωθεν τῶν βαθμῶν (Syriac men darghā, מן דרגא ‘from the top of the steps’), and one usually sees in this the title of the underlying source, Ἀναβαθμοὶ Ἰακώβου. Acts knows those steps with the prefix that the reconstructed Greek loses: Paul arrives ἐπὶ τοὺς ἀναβαθμούς (21:35), the soldiers carry him bodily (βαστάζεσθαι, the verb of Romans 15:1), he speaks from them τῇ Ἑβραΐδι διαλέκτῳ (21:40) and the first thing he says of himself is ‘at the feet of Gamliel’ (22:3). In one source the enemy throws Ya’acov bar Pandira down from the steps; in the other, it is the man rescued on those steps who boasts of his teacher. If the source is earlier than Acts, Luke inverted the scene; if it is later, it is the source that answers Luke. I cannot decide it, and I leave it as hypothesis. Nor does the framework help in the chronology: it places the enemy before Caiaphas (I, 71), in the thirties, and the journey to Damascus with letters seems taken from Acts 9:1–2, so that it is not an independent witness to the Saul of 62–66.

The silence of Josephus weighs, then, less than it seems, and in another direction. Josephus is not silent on who killed Ya’acov bar Pandira / James: he names Ananus, a Sadducee with a tribunal. What he does not say is that Saul was a Christian. That does not prove that they are two men, but neither does it support their identification; to use it as decisive proof in either sense would be to turn an argument from silence into a conclusion that the text does not permit. All I maintain is more modest and can be measured: the tradition hostile to Paul, written in another language and with other interests, did not remember him as a school pupil nor as a judge or member of a tribunal, but as the enemy who stirs up tumult, seizes and strikes, and that profile resembles the Saul of Josephus more than the disciple of Gamliel drawn by Luke. It is an indication, not a proof, and chance remains possible.

An alternative explanation. Pervo (2006) has argued that Acts knows the work of Josephus. If Luke had read Josephus (the Antiquities date from about 93–94), the Saulus of Ant. 20.214 might have given him the name and the profile of his persecutor without its being the same man. That hypothesis explains the ‘Saul’ of Acts, but it does not explain Galatians 1:13–14, written some four decades before Josephus. Either the coincidences of Galatians are chance, or they point to a shared reality. There is no comfortable third option; I opt for the second, because they are too many to dispose of one by one.

This is the thesis, then, without euphemisms. The Saulus of Josephus and the Paul of the letters share social class, a double office (persecutor and ambassador), the vocabulary with which each names violence, a theology that inverts that violence, an enigma of Damascus that is solved only if the man is a relative of the Herodians, and a destiny, Nero in Achaea, that the ancient chronology of the Church places at a short distance from Paul’s death. Against that stand a chronological objection whose cost I have stated aloud, a chain of names that passes through Luke and a chance that I cannot discount a priori. The explanation that best integrates everything is that it is the same man; the one that does not integrate it has to treat as accidents eight coincidences that support one another.

One hypothesis remains that I do not count among the eight, because it depends on all of them and because it is the one that costs most. It follows the thread of Strabo, for whom the Idumaeans were Nabataeans by origin who, expelled by a revolt, passed over to the Jews and adopted their customs (Geography 16.2.34). Moderns qualify this: they see in the Hellenistic Idumaeans displaced Edomites mixed with Arabs (see Kasher, 1988), but the mixture suffices for what follows. There is, besides, a datum from Josephus that does not depend on Strabo: Cypros, the wife of Antipater and mother of Herod and of Salome, was an Arab of distinguished family (Ant. 14.121; War 1.181), and in his vocabulary ‘Arab’ frequently designates subjects of the Nabataean kingdom. With that, the Saulus of the reconstruction would have Nabataean roots by two routes: through Cypros, his great-grandmother, and through Costobar I, an Idumaean priest of the god Kozai, whose name, Qos, appears also in the onomastics and in the sanctuaries of the Nabataeans. And the ties of the house with Nabataea are no detail: Herod, a fugitive in 40 BC, sought refuge with the Nabataean king Malchus (War 1.274–276); Salome had a Nabataean suitor, Syllaeus, minister of King Obodas (Ant. 16.220–225); and Herod Antipas was married to a daughter of Aretas IV, whose repudiation unleashed the war that Aretas won (Ant. 18.109–115). Aretas thus ceases to be merely an enemy of the Herodians and becomes an aggrieved party within the political kindred. I do not exaggerate: between Cypros and Aretas IV there is more than a century, and I do not claim a close blood relationship; I maintain that a Herodian of that generation had Nabataea in his family history, in his politics and in his enmities.

With that key, Galatians 1:17 ceases to be a gap. Paul does what nobody would expect of a newly called Pharisee: he does not go up to Jerusalem, but ἀπῆλθον εἰς Ἀραβίαν, without saying what he went to do or why, and returns to Damascus, which in those years was under the power of Aretas (2 Cor 11:32). There is an uncomfortable reading: he would not have gone to a foreign country but to the kingdom where his house had relatives, debts and enemies. But the text that most arrests me is Galatians 4:21–31. Paul places Hagar in Arabia (4:25, with textual variants), and Josephus teaches his Greek readers that the Nabataeans descend from Nebaioth, the firstborn of Ishmael (Ant. 1.220–221; Gen 25:13). For a reader of the first century, the children of the slave woman were, first of all, the Arabs. And there is a datum that made me return to the list at the beginning: Josephus says that the Arabs circumcise their sons after the age of thirteen, because Ishmael was circumcised at that age (Ant. 1.214), whereas Isaac was circumcised on the eighth day (Gen 21:4). Look now at Philippians 3:5: περιτομῇ ὀκταήμερος, ‘circumcised on the eighth day’. The first box of the list not only denies the adult circumcision of the Idumaeans (Ant. 13.257); it denies also that of Ishmael, and is, literally, that of Isaac. Then comes the allegory that sets Isaac against Ishmael, and its key verb: ὁ κατὰ σάρκα γεννηθεὶς ἐδίωκεν τὸν κατὰ πνεῦμα, ‘the one born according to the flesh persecuted the one born according to the Spirit’ (4:29). Ἐδίωκεν is the root with which Paul defines himself in 1:13 and 1:23. He applies the allegory to the ‘present’ Jerusalem and to those who harass him, not to himself; but the persecutor of the allegory is the son of the slave woman, and the persecutor of the letters is he. There is one more crossing. The two jibes that, as we saw, fell upon the house of Herod, half-Jew (ἡμιιουδαῖος, Ant. 14.403) and descendant of slaves (Africanus; Bava Batra 3b), are exactly the two components of ‘the son of the slave woman’, who was moreover an Arab. And Ἀραβία appears only twice in the whole New Testament, both times in Galatians and both in passages about provenance: that of his gospel (1:17) and that of the heirs (4:25).

One last link remains, and it is the one that most made me doubt myself, because it comes neither from Paul nor from Josephus but from their enemies. For a Jew of the first century, ‘Idumaean’ and ‘Roman’ were not two ethnicities but two overlapping positions: that of the half-Jew (ἡμιιουδαῖος, Ant. 14.403) whose house owed the throne to Rome. The Herodians bore names from their patrons (Marcus Julius Agrippa), and Herod gave the name of his protector Antony to the Temple fortress (Ant. 15.409). Rabbinic language later sealed the equation: ‘Edom’, the Hebrew name of Idumaea, came to designate Rome (Genesis Rabbah 65:1, of later date), and it has been proposed that it did so in part because Rome governed Judaea by means of an Edomite (Cohen, 1967). I do not say that an Idumaean was a Roman. I say that the insult ‘Edomite’ and the insult ‘Roman’ overlapped, and that whoever wished to deny that a Herodian was of Israel ‘by lineage’ had both at his disposal. Return now to the hostile witnesses. The Ebionites of Epiphanius say that Paul was Ἕλλην, the son of Greek parents (Panarion 30.16.6–9). But in the Greek of the New Testament Ἕλλην is, before it is a nationality, one who is not a Jew (Rom 1:16: Ἰουδαίῳ τε πρῶτον καὶ Ἕλληνι), so that ‘Greek’ is the word with which an Ebionite would write ‘gentile’, and in the Roman East that gentile is not far from the Roman. Their account, moreover, follows the route of the Idumaean: a gentile who circumcises himself as an adult and becomes a proselyte (περιτέμνεται) in order to marry the high priest’s daughter. It is exactly the route that Josephus documents around the Herodian house: Herod married the daughter of Simon son of Boethus and made him high priest (Ant. 15.320–322), and Azizus of Emesa and Polemon of Cilicia were circumcised in order to marry Drusilla and Berenice, sisters of Agrippa II (Ant. 20.139, 145–146). The Ebionite account reads like a court chronicle with the names erased. The Pseudo-Clementines give a second testimony, more oblique, because they attack Paul without naming him: ‘the enemy man’ of Peter’s letter to James (2.3), the one who relies on a vision (Hom. 17.19), the enemy who departs for Damascus with letters from the high priest (Rec. 1.70–71). In his place they put ‘Simon the Magician’, a Samaritan of Gitta, son of Antonius and Rachel, trained in Alexandria in Greek wisdom (Hom. 2.22; Rec. 2.7). The reading Simon = Paul comes from Baur and is defended by Lüdemann (1989); it is disputed, and I use it as what it is: a code of Paul’s adversaries. Look at what that code contains. A father with a Roman name, the same type of patron-name that the Herodians bore. A Samaritan provenance, which occupies in the north the place that Idumaea occupies in the south: another frontier ‘half-Jew’, of disputed Jewish identity. A Greek culture learned far from Jerusalem. And a mother called Rachel, who is the mother of Benjamin: a weak indication, because Rachel is a common name, but precisely the tribe that Paul claims in Philippians 3:5. Greek, Samaritan, son of Antonius: three different labels for a single operation, to deny that he was of Israel by lineage and to assign him by origin to a neighbour, not a brother. If Saulus was a Herodian, those three labels cease to look like scattered inventions and read as three translations of the word that his own house carried stuck to it: Idumaean, that is, Edomite, that is, Roman. The Ebionites and the authors of the Pseudo-Clementines no longer knew who Paul’s family was, but they remembered the stigma. That is hypothesis, and the most fragile of those I have set out.

I harbour no illusions about the cost. Paul says ‘tribe of Benjamin’ and ‘Hebrew born of Hebrews’; the Herodians counted themselves among the Jews; ‘Arab’ in Josephus does not always mean Nabataean; Strabo is a geographer who repeats what he heard; identifying the Arabs with the sons of Ishmael was commonplace among Jews who never set foot in Nabataea, so that any Jewish author could have written Galatians 4; the circumcision at thirteen describes an Arab custom, not an individual; ‘Arabia’ in 1:17 could have been the environs of Damascus; and Paul turns the allegory against his adversaries, not against himself. Upon the last link weigh costs of its own. Epiphanius writes in the fourth century and gathers a polemic; the equation Edom = Rome is later than Paul, and its Herodian origin is a conjecture; Ἕλλην means ‘gentile’ and proves nothing about concrete origin; the identification of Simon with Paul is disputed, and ‘Antonius’ and ‘Rachel’ are common names; and the Ebionites could have fabricated their account from Galatians and Acts without knowing anything about Herod. Nor does it resolve the problem of the rods: if Paul’s house was ‘Roman’ by clientage, the silence of 2 Corinthians 11:25 remains just as strange. None of this proves a Nabataean or Idumaean origin. What it shows is that, if Paul was a Herodian, the origin that the thesis gives him is not that of a Pharisee of Tarsus but that of a man with reasons to want blood to decide nothing. A man whose blood was disputed, half-Jew for some and son of slaves for others, could not win that suit by blood, and he transfers it to another terrain: εἰ δὲ ὑμεῖς Χριστοῦ, ἄρα τοῦ Ἀβραὰμ σπέρμα ἐστέ, κατ᾽ ἐπαγγελίαν κληρονόμοι (‘and if you are Christ’s, then you are Abraham’s offspring, heirs according to the promise’; Gal 3:29). That is why the phrase of Galatians 3:28 sounds different, οὐκ ἔνι Ἰουδαῖος οὐδὲ Ἕλλην, οὐκ ἔνι δοῦλος οὐδὲ ἐλεύθερος (‘there is no longer Jew nor Greek; there is neither slave nor free’): the two pairs are, one by one, the two accusations that weighed upon his house. In that case, the apostle of the gentiles would not have been a Jew who opened the door to the pagans, but a man to whom the door was closed, whom his adversaries, seeking a name for what they could not name, called Greek, Samaritan and enemy, and who decided that no one should be left outside.

And there is a second half of the thesis that does not depend on the first. Even if the reader rejects the identification, what Acts affirms—that Paul was from Tarsus, a disciple of Gamliel and a Roman citizen—remains without support in any letter, and the pieces with which that figure was assembled remain those we saw. Return to the sheet with which we began. If of Paul you know from Luke alone that he was from Tarsus, a disciple of Gamliel and a Roman citizen, and Luke is, according to some specialists, an author who knew Josephus, to what witness do you intend to turn in order to learn who he was, the man who, in the Jerusalem of the first century, called himself Saul?

References

Aland, B., Aland, K., Karavidopoulos, J., Martini, C. M., & Metzger, B. M. (Eds.). (2012). Novum Testamentum Graece (28th rev. ed.). Deutsche Bibelgesellschaft.

Eisenman, R. (1997). James the brother of Jesus: The key to unlocking the secrets of early Christianity and the Dead Sea Scrolls. Viking.

Eisenman, R. (2006). The New Testament code: The Cup of the Lord, the Damascus Document and the Shrouds of Jesus. Watkins.

Gerhardsson, B. (1961). Memory and manuscript: Oral tradition and written transmission in rabbinic Judaism and early Christianity. C. W. K. Gleerup.

Kasher, A. (1988). Jews, Idumaeans, and ancient Arabs: Relations of the Jews in Eretz-Israel with the nations of the frontier and the desert during the Hellenistic and Roman era (332 BCE–70 CE). J. C. B. Mohr (Paul Siebeck).

Koch, D.-A. (1986). Die Schrift als Zeuge des Evangeliums: Untersuchungen zur Verwendung und zum Verständnis der Schrift bei Paulus. J. C. B. Mohr (Paul Siebeck).

Kokkinos, N. (1998). The Herodian dynasty: Origins, role in society and eclipse. Sheffield Academic Press.

Liddell, H. G., Scott, R., & Jones, H. S. (1996). A Greek–English lexicon (9th ed., with revised supplement). Clarendon Press.

Lieberman, S. (Ed.). (1955–1988). The Tosefta according to Codex Vienna, with variants from Codex Erfurt, Genizah mss. and editio princeps (Venice 1521) (Vols. 1–5). Jewish Theological Seminary of America.

Maccoby, H. (1986). The mythmaker: Paul and the invention of Christianity. Harper & Row.

Niese, B. (Ed.). (1885–1895). Flavii Iosephi opera (Vols. 1–7). Weidmann.

Pervo, R. I. (2006). Dating Acts: Between the evidence and the theory. Polebridge Press.

Rahlfs, A. (Ed.). (1935). Septuaginta. Württembergische Bibelanstalt.

Sandmel, S. (1962). Parallelomania. Journal of Biblical Literature, 81(1), 1–13.

Stanley, C. D. (1992). Paul and the language of Scripture: Citation technique in the Pauline Epistles and contemporary literature. Cambridge University Press.

van Unnik, W. C. (1962). Tarsus or Jerusalem: The city of Paul’s youth (G. Ogg, Trans.). Epworth Press.

Cohen, G. D. (1967). Esau as symbol in early medieval thought. In A. Altmann (Ed.), Jewish medieval and Renaissance studies (pp. 19–48). Harvard University Press.

Lüdemann, G. (1989). Opposition to Paul in Jewish Christianity. Fortress Press.