Saturday, March 21, 2026

Raymond E. Brown on the Historicity of the Paschal Release

  

C. Historicity of the Paschal Release

 

There is a further question. At the time when Jesus of Nazareth was executed, was clemency extended to Barabbas because there was a set custom of releasing a prisoner during Passover? Once again there are fanciful theories that go beyond the Gospel evidence. For instance, Bajsić would make Barabbas the center of interest for Pilate. In his theory the governor knew that the Jewish crowd would want this dangerous revolutionary released; and even though he had to observe the custom of the festal pardon, he was trying to avoid releasing Barabbas. Obviously Jesus of Nazareth was politically harmless, so Pilate tried to offer Jesus to the crowd in order that Barabbas would not get off. While well intentioned, this theory is just as contrary to the Gospel evidence as is the totally fictional approach to Barabbas.

 

Studying the Gospel accounts, one must ask what legal procedure of pardon the evangelists had in mind. In any developed legal system there are distinct forms of clemency, e.g., terminating a trial and dismissing a prisoner because of lack of evidence; suspending a sentence and dismissing the prisoner even when there is sufficient evidence of guilt because there are mitigating circumstances; after a sentence has been passed, granting pardon from punishment—an action often performed by a higher authority. The last of these (pardon), we might note, is often the most difficult to obtain; and in later Roman law there was reluctance to have governors of provinces reverse their decisions. Unfortunately, the evangelists are not precise about Barabbas: Is he in prison because he has not yet been tried, or because he has been tried and is awaiting sentence, or because he has been sentenced and is awaiting execution? Normal Roman procedure would not suggest intervals in the trial, sentence, and execution pattern. In Justinian’s Code (9.4.5) a principle that seems to have gone back to early imperial and even republican times is enunciated: “One who is convicted should quickly undergo punishment.” Nevertheless, for various reasons including political expediency, normal Roman procedure would not always have been followed. How careful we need to be is suggested by Josephus (Ant. 20.9.5; #215), who reports that the procurator Albinus had many imprisoned in Judea who deserved death but had not been executed. Since presumably Pilate would have had to handle capital punishment trials in relation to a riot, judgment on Barabbas might have waited upon Pilate’s arrival in Jerusalem from Caesarea at the time of the Passover feast. The fact that several criminals were executed with Jesus suggests that Pilate was sentencing and executing judgment during the feast. (Of uncertain applicability is the directive later attested in Tosepta Sanhedrin 11.7 that certain types of criminals should not be executed immediately but kept in Jerusalem until they could be executed more publicly at the feast.)

 

How would clemency fit into the Roman legal procedure? Remembering that Rome had just shifted from a republic to an empire, we must be aware that there is uncertainty in our knowledge of procedures in Pilate’s time, especially in a province like Judea. Although some commentators speak of amnesty, the Greek legal term amnēstia appears rather seldom in Latin literature and mostly as a loan word. The true Latin equivalent would be abolitio, most often in the sense of a mass pardon, sometimes setting aside a controversial legal procedure (Quintillian). Indulgentia, again often on a mass scale, tended to cover the nonapplication of a punishment because an appeal had been made; in Jesus’ lifetime it would largely have been extended as an act of the emperor. More applicable to individuals would have been venia, sometimes extended because mitigating circumstances took away guilt, but also extended independently of guilt—indeed to one known to be guilty (poenae meritae remissio: Seneca, De clementia 2.7.1). All things considered, what the Gospels describe might best be classified as venia to be extended either to Jesus before sentencing (leaving aside the value of the case against him) or to Barabbas wherever he was in the legal process.

 

The Gospels differ as to the background of the custom of pardoning; it pertained to the Roman governor according to Mark/Matt, and to “the Jews” according to John. (Because of the Passover setting this custom has come to be known as the privilegium paschale.) Scholars have sought parallels to the custom in attested Greco-Roman practice of clemency and in Jewish practices.

 

Greco-Roman Parallels. (1) Festal amnesties. The Gospel atmosphere pertaining to the custom involves a feast (not specified by Mark/Matt but specified as Passover by John). Grotius (+ 1645) suggested a parallel in the Roman Lectisternia. Livy (History 5.13.7–8) reports that on the first historic celebration of this type of eight-day feast in 399 bc prisoners were unbound, and such was the religious awe inspired by the proceedings no one dared afterward to rechain them. Dionysius of Halicarnassus (Roman Ant. 12.9.10) reports that there was a release of slaves who had been placed under arrest by their masters. The Lectisternia were not celebrated annually but only on the occasion of thanksgiving in time of special stress. It is not clear that the unbindings were repeated, or that the concessions went much beyond parole. Other scholars (J. Gothofredus) appeal to clemency extended at the Greek feast of Thesmophoria (a feast associated with Demeter the lawgiver, insuring the fertility of the earth). The Panathenae, involving a procession to the acropolis in honor of Athena, has also been proposed as a parallel. Merritt (“Jesus” 62, 65) mentions a prisoner release during the greater Dionysia celebrated in early April at Athens since the 6th cent. bc, and at the Greek festival of Kronia (= Roman Saturnalia; see p. 876 below) celebrated widely in December. Scholars have studied Papyrus Tebtunis 5.1–5 where, following the Greek model of the gracious king, the Hellenistic Egyptian monarchs Ptolemy and Cleopatra proclaimed “an amnesty to all their subjects for errors, crimes.…” At most, however, these are mass amnesties of varying scope; none of them is truly parallel to the custom of a judicial pardon granted to an individual by a Roman governor ruling over subject people.

 

(2) Common practices by imperial officials. There is no doubt that these officials granted pardons. The Pilate of John 19:10 states, “I have the authority to release you and the authority to crucify you.” Cynically, Origen (In Matt. 27:15, #120; GCS 38.254) comments, “So they grant some favors to people whom they subject to themselves, until their yoke over them is made firm.” Pliny in correspondence with the Emperor Trajan (Epistles 10.31–32) reports that by order of proconsuls and legates, condemned criminals who had become intermingled with slaves working in the public service had been released; Trajan wanted this abuse corrected by insisting that sentences could be reversed only by proper authority. In particular, Florentine Papyrus 61.59ff. (LFAE 267–69) narrates an incident in Egypt in ad 85 when G. Septimius Vegetus released to the crowds a prisoner who was guilty of having sent to prison an honest family and who accordingly deserved to be scourged (mastigoun). Josephus (Ant. 20.9.5; #215) reports that in ad 64 the procurator of Judea Albinus, when he heard that Florus was coming to succeed him, in a final effort to gain a name for himself cleared the prisons. He had those executed who deserved death, but for a bribe released those convicted of trifling crimes. Yet all these examples tell us nothing about a regular custom on a feast; at most they are isolated instances of humane behavior. Another Josephus story about Albinus (Ant. 20.9.3; #208–9) is set on a feast (probably Passover) when the procurator had to let ten bandits (lēstai) go in order to ransom the secretary of the high priest’s son, who had been kidnapped by terrorists (sicarii). Obviously, despite the festal setting, no benevolent custom was involved. Under this same general heading might be treated the thesis that the release of Barabbas was in the pattern of “acclamation” or anaboēsis (§31, D3c above). Justinian’s Code (9.47.12) insists, “When the populace speaks viva voce, they are not to be heard; nor should one believe their voices when they desire that someone guilty of a crime be dismissed, or that someone innocent be condemned.” Would such a maxim have bound Pilate? Suetonius, Tiberius 47, tells us that the emperor himself was forced to release a slave who had done well in the arena because the crowds cried out for it. Yet this is far from the custom described in the Gospels. Would release by acclamation have been a regular festal custom? Was it extended to a subject populace when a riot had recently occurred?

 

(3) A special Roman concession to Jews as a safety valve. J. Spencer (1727) thought that the Seleucid kings in the 3d or 2d cents. bc might have begun such a custom and that the Romans continued it. True, 1 Macc 9:70–72; 10:23 give us examples of the release of captives taken in war as part of an attempt by Syrian rulers to gain peace with the Maccabean leaders. Once again, however, this is quite different from releasing a single prisoner from jail at Passover. Other scholars have thought that the Roman emperors might have begun such a custom. Josephus (Ant. 14.10; #185–267) gives a long list of imperial and local Roman concessions to the Jews, beginning with those of Julius Caesar; but none of these concessions mentions releasing a prisoner on a feast. Still another suggestion (or guess) is that Pilate might have introduced the custom to make amends for his heavy-handed behavior and mistakes as prefect (pp. 698–705 above). The weaknesses of all these parallels are obvious, and one is left with the enduring doubt that Roman governors could ever have committed themselves to a custom that would require them to release a killer in the midst of a recent riot in a volatile province.

Jewish Parallels. Some have thought that the Romans or Pilate might have had to accept a Jewish custom already in place. The Passover theme of release from Egypt might feasibly have led to a Jewish custom of releasing a prisoner at that feast. Merritt (“Jesus” 59, 61) mentions occasional releases of prisoners as part of the šigū rite of pardon, and points to the influence of the Babylonian calendar and customs on the Jews. Evidence that the Babylonian king released a prisoner on the 6th, the 16th, and 26th days of the 8th month of the year has been adduced to explain why Jews would have moved toward such a practice (Langdon, “Release”). Some of this theorizing, however, neglects the mentality of biblical law which presents its punishments as set by God and therefore not to be dispensed from. Num 35:31 affirms: “You shall accept no ransom for the life of a murderer who is guilty of death, but he shall be put to death.” Heb 10:28 is no less firm: “One who has violated the Law of Moses dies without mercy.…” Thus it is not surprising that the Bible offers no evidence of a Passover-release custom, not even extended by a Jewish king. Was there a change in attitude when serious cases were decided not according to the biblical law but according to the law of the Hellenistic or Roman conquerors? Chavel (“Releasing” 277–78) thinks that in the 1st cent. bc, when there were many political prisoners because of inner Jewish struggles, the Hasmonean priest-kings of Jerusalem introduced this custom of pardoning to please the excitable pilgrim throngs. Josephus (War 2.2.5; #28) tells us that Archelaus liberated those whom his father had imprisoned for the gravest crimes. Such evidence establishes only a possibility; the talmudic literature gives almost an hour-by-hour description of Passover and never mentions the custom; nor do Josephus and Philo mention it. Indirect evidence has been sought in Mishna Pesaḥim 8:6 governing the slaughtering of the Passover lamb (see Blinzler and Chavel). The legal issue concerns those who may and may not be counted as members of a Passover company (ḥaburah), for whom a lamb can be prepared by sacrifice. The discussion includes “someone who has received a promise to be released from prison.” The accountability for this person is a problem, for the promise might not be fulfilled. (TalBab Pesaḥim 91a notes that someone in an Israelite prison is a special case since a temporary release for Passover would be given by Jews.) The citation deals with a series of possible cases, including the sick and the aged (see also Tosepta Pesaḥim 7.11); it concerns someone who might be eligible for permanent or temporary release; it has nothing to do with the right of the populace to have one prisoner released.

 

The conclusion from this discussion of Roman and Jewish amnesty/pardon parallels is that there is no good analogy supporting the historical likelihood of the custom in Judea of regularly releasing a prisoner at a/the feast (of Passover) as described in three Gospels. Already in the early 3d cent. Origen (In Matt. 27:15, #120; GCS 38.253–54) betrayed surprise at such a custom. Luke’s omission of the custom, even though he knew Mark, has been thought to represent an earlier skepticism. Can one reconcile the possible nonhistoricity of the Passover privilege with the existence of a historical Barabbas who was released from prison by Pilate (all four Gospels)? One might suspect that the evangelists (or their predecessors in the tradition) assumed that the release was a reflection of a regular custom when in fact it was an isolated incident. What might have led the evangelists to assume that there was a custom? Wratislaw (“Scapegoat”) suggests that the description might have been influenced by the Israelite liturgical custom of the two goats described in Lev 16:7–22: One goat was let go, the other killed for a sin offering. Other scholars have theorized that already there may have been a Christian custom of pardoning at Easter time. Merritt (“Jesus”) points to general similitude: The existence of various amnesties and pardons in the diverse cultures described above would have made the idea of a regular custom of release at a feast seem plausible to narrators and hearers who had no exact knowledge of Judea ca. 30. Bauer (“Literarische”) gives examples where something that happened was freely explained as coming from a custom.

 

If we lay aside the custom as a secondary development, the historical substratum of the Barabbas incident may have been relatively simple. The following outline could be reconstructed on the basis of the Gospel reports: A man with the name Barabbas was arrested in a roundup after a riot that had caused some deaths in Jerusalem. Eventually he was released by Pilate when a feast brought the governor to Jerusalem to supervise public order. Presumably this took place at the same time that Jesus was crucified, or not far from it, or at another Passover. In any case, this release struck Christians as ironic: The same legal issue was involved, sedition against the authority of the emperor. Although they knew Jesus was innocent, he was found guilty by Pilate, while Barabbas was let go. (As seen in the comment on Mark 15:7, that verse never states that Barabbas rioted or killed. Even if the evangelist judged Barabbas guilty, in a preMarcan stage, closer to the original story, Barabbas’ guilt may not have been established—a fact that would have allowed Pilate to release him.) The storytelling tendency to contrast the released Barabbas and the crucified Jesus by bringing them together at the same moment before Pilate’s “justice” would have been enhanced if both had the same personal name, Jesus.

 

Inevitably, one would have wished a historical investigation to produce more certain results than the above-described likelihoods. (Of course, there is the beneficial effect of showing that some of the plot approaches to Barabbas, often involving antiJewish creation by the evangelists, are fiction with little likelihood.) Yet historical criticism cannot overcome the lack of comparative material, and the Barabbas case is particularly frustrating on that score. Nevertheless, the real import of the Barabbas motif is on another level, namely, the truth that the evangelists wished to convey about the death of Jesus. For them conviction of the innocent Jesus had a negative side, the choice of evil. The story of Barabbas with a basis in fact was dramatized to convey that truth. (Raymond E. Brown, The Death of the Messiah–From Gethsemane to the Grave: A Commentary on the Passion Narratives in the Four Gospels, 2 vols. [The Anchor Yale Bible Reference Library; New Haven: Yale University Press, 1994], 1:814-20)

 

Blog Archive