Biblical Archaeology Review 35:1, January/February 2009

A New Reconstruction of Paul’s Prison

Herod’s Antonia fortress

The Antonia, the palace/fortress lavishly described by the ancient Jewish historian Josephus at the northwest corner of the Herodian Temple Mount, is not mentioned by name in the New Testament. For a long time, however, it was thought to be the “praetorium” where Pilate questioned Jesus and found him innocent.

The praetorium is also mentioned in Mark 15:15–16, where Pilate, to satisfy the crowd, delivers Jesus to be crucified, and the soldiers lead him away, taking him “inside the palace (that is, the praetorium).” And in Matthew 27:27, the soldiers take Jesus into the “praetorium,” where he is mocked and hailed as King of the Jews.

A praetorium was originally the residence of a praetor—a provincial governor. These New Testament references make it clear that the praetorium they are referring to is part of a palace that is a royal residence. Herod’s palace was not near the Temple Mount. Scholars are generally agreed that it lay on the western edge of the city, south of today’s Jaffa Gate. It no doubt stood just as Herod built it until the Roman destruction of Jerusalem in 70 C.E. Josephus describes the layout in his account of the First Jewish Revolt against the Romans. It would be—and was—an ideal and honorable place to house the supreme Roman authority. A similar phenomenon is known to us from Caesarea, where Herod’s Promontory Palace later became a praetorium. In short, the praetorium was most likely located in Herod’s palace—not the Antonia—and there is thus no New Testament reference to Jesus in connection with the Antonia.

With Paul, however, it is different. When Paul was arrested at the Temple, he was bound in chains and taken to the “barracks,” which was entered via “steps” (Acts 21:34–35). He was later incarcerated in the “barracks” (Acts 21:37). A very good case can be made that the “barracks” referred to was part of the Antonia at the northwest corner of the Temple Mount.

Attacks on the Temple Mount (and Jerusalem) have traditionally come from the north. Steep valleys lie to the east and west. To the south, the terrain descends more gradually but just as deeply through the City of David, the earliest settlement of Jerusalem.

Before the Antonia was constructed at the northwest corner of the Temple Mount, earlier Jewish rulers (the Hasmoneans) had also erected a citadel here called the Baris. Herod replaced the Baris with his own palace/fortress, not only to protect against invaders, but to control whatever was happening on the Temple Mount. Herod named the grand palace/fortress after his protector and friend Marc Antony. (Naming it for Marc Antony tells us that the Antonia was built before 31 B.C.E., when Antony was defeated by Octavian at the Battle of Actium.)

Josephus describes the Antonia in some detail: With a tower at each of its four corners, the Antonia was apparently square or nearly so. Three of the towers were 50 cubits high; the fourth, on the southeast corner, was 70 cubits high. (Although the precise length of a cubit is a matter of scholarly debate, it is about 50 cm [18–20 in] ).

Josephus also tells us that “A Roman cohort was permanently quartered there.” This sounds a lot like the “barracks” referred to in Acts. Moreover, the Antonia was connected by stairs to the Temple Mount, and Acts tells us that the “barracks” were entered by “steps.” So the connection of the Antonia with Paul’s incarceration seems quite secure. But what was the Antonia like?

Unfortunately, practically nothing remains of the structure. What we have are mainly cuts in the bedrock at the northwest corner of the Temple Mount. While these have raised many questions, no scholar has ever seriously questioned the existence of this palatial fortress at this site.

Josephus tells us the Antonia was set on a height: It was built “on a rock fifty cubits high, and all sides precipitous ... This rock was covered from its base upwards with smooth flagstones, both for ornament and in order that anyone attempting to ascend or descend it might slip off.”

At the top of the rock was a wall 3 cubits high, behind which was the “majestic” edifice. It “resembled a palace in its spaciousness and appointments, being divided into apartments of every description and for every purpose, including cloisters, baths and broad courtyards for the accommodation of troops; so that from its possession of all conveniences it seemed a town; from its magnificence, a palace.”

Josephus goes on:

At the point where it impinged upon the porticoes of the Temple, there were stairs leading down to both of them, by which the guards descended; for a Roman cohort was permanently quartered there, and at the festivals took up positions in arms around the porticoes to watch the people and repress any insurrectionary movement. For if the Temple lay as a fortress over the city, the Antonia dominated the Temple, and the occupants of that post were the guards of all three; the upper town had its own fortress—Herod’s palace.

All in all, according to Josephus the Antonia was “a crowning exhibition of the innate grandeur of [Herod’s] genius.” The Antonia was apparently Herod’s first building project. He did not begin the renovation and expansion of the Temple and its enclosure until around 20 B.C.E.

With no archaeological remains, however, we are left with only some rock cuttings to learn more. The most important is the remains of a moat that separated the Antonia from the hill to the northeast (today’s Muslim Quarter of the Old City). This moat marks the northern boundary of the Antonia.

About 20 meters (65 ft) south of the moat that marks the Antonia’s northern boundary, is a cut 12 meters (40 ft) deep into the bedrock. The cut is 39 meters (128 ft) long and marks the northern boundary of the Temple Mount as it exists today. One might suppose that this cut marks the southern boundary of the Antonia, but to my mind this cannot be. At only about 30 meters (100 ft) south of the moat marking the northern boundary of the Antonia, this leaves too narrow a strip for the Antonia, especially in light of Josephus’s glowing description.

The Antonia must have extended farther south. I believe it extended onto the area of the present Temple Mount. I am not the first to make this suggestion. In the late 1970s it was proposed by the Israeli scholar E.W. Cohn. His suggestion was not very favorably received at the time, however; perhaps my demonstration here will give it wider acceptance.

An important clue lies in a bulge in the height of the bedrock at the northwest corner of the Temple Mount. Herod leveled the Temple Mount to a height of 738 meters (about 2,420 ft) above sea level (except for the platform in the center for the Temple). But at the northwestern corner of the Temple Mount, it is 740 meters above sea level, an increase of more than 6 feet. Why was this area not leveled to 738 meters? Perhaps because there was already a building there: the Antonia.

When Herod began his rule, the Temple Mount was a much smaller platform. As is well known, he extended it on the north, south and west. (He could not extend it on the east because there the terrain drops steeply into the Kidron Valley.) And we have clear evidence of the extensions.

As I said, prior to the construction of the Antonia, the Hasmonean rulers of Judea had built a fortress known as the Baris (Bira in Hebrew, “fortifications”) on the same site that was to be later occupied by the Antonia. It was the logical place for a fortress to protect the city from an enemy force approaching from the north (the traditional route) and to control the crowds on the Temple Mount to the south. (Indeed, I believe there were even earlier fortifications here built by the Seleucid and Ptolemaic authorities, but that is another subject for another time.)

I don’t believe Herod would have completely dismantled the Baris before constructing the Antonia. I believe he incorporated elements of the Baris into his construction, especially its water cisterns and cellars.

Another important clue to the puzzle of the location of the Antonia consists of the remains of an ancient north-south water channel, cut deeply into the bedrock west of the Temple Mount. About 60 meters (195 ft) south of the northwest corner of the Temple Mount, the channel turns sharply to the east. But it is blocked by the western wall of the Temple Mount. The channel must have therefore been dug before Herod built his extended western wall of the Temple Mount. This water channel apparently served the Baris of the Hasmonean era when the Temple Mount was smaller and the Baris was north of the Temple Mount. With Herod’s enlargement of the Temple Mount, the site of the (former) Baris (and now the Antonia) was partially on the expanded Temple Mount.

The final clue is the western wall of the Temple Mount itself. About 65 meters (213 ft) south of its northwest corner, the western wall takes a 3-meter (10-foot) jog to the east; that is, it goes east by a 90-degree turn and then proceeds south again. The only explanation for why the western wall “sticks out” for the first 65 meters is because the Antonia was there and extended out that far. Only where it began south of the Antonia could Herod’s extended western wall come in to where he wanted it.

In short, in the early years of his rule, Herod reconstructed the Baris, turning it into the Antonia that Josephus describes. Later, Herod reconstructed the modest Second Temple, which the Jews returning from the Babylonian exile had built in the sixth century B.C.E., and enlarged the Temple Mount. However, on the northern side, he was limited by the fact that the Antonia extended into the new portion of the Temple Mount.

This solution to the location of the Antonia is confirmed by Josephus’s description. If the Antonia had been completely outside the Temple Mount, he would simply have written that it lay north (or northwest) of the Temple Mount. Instead, he wrote: “The tower of Antonia lay at the angle where two porticoes, the western and the northern, of the first court of the Temple met.”

That the Antonia was situated at the “angle of the two porticoes” rather than north or northwest of the Temple Mount is a powerful indication that the Antonia actually extended onto the Temple Mount. This observation is reinforced by Josephus’s statement that the Antonia “impinged upon the porticoes of the Temple.”

He then goes on to say that “at the point where it [the Antonia] impinged on the porticoes of the Temple, there were stairs leading down to both of them.” That there were stairs leading down from the Antonia to the porticoes (or colonnades) of the Temple Mount certainly suggests that the Antonia encroached on the Temple Mount. Otherwise, why would there be separate stairways to each of the porticoes (on the north and west) that abutted the Antonia?

As already mentioned above, Josephus tells us that “the general appearance of the whole [of the Antonia] was that of a tower with other towers at each of the four corners; three of these turrets were fifty cubits high, while that at the southeast angle rose to seventy cubits, and so commanded a view of the whole area of the Temple.” We can get some idea of the visual effect of the Antonia from another Herodian structure that, unlike the Antonia, in large part survives, namely Herodium near Bethlehem. The chief difference is that the Antonia was square (or a rectangle very close to square) and Herodium is round. Herodium also served a somewhat different purpose from the Antonia; it was one of Herod’s summer palaces and a fortress, and Herod was also buried there. We have recently discovered Herod’s burial site at Herodium, but that is another story which I will write about in a future issue of BAR. Both Herodium and the Antonia enclosed a central structure with four towers, one of which was higher than the others. As Josephus says of the Antonia, so it may be said of Herodium: “The general appearance of the whole was that of a tower with other towers at each of the four corners.”

Having steeped myself in the details of Herodian architecture for almost half a century, I sometimes think I am inside Herod’s mind or head, at least architecturally. On this basis, I think I can reconstruct the sequence of events prior to the construction of the Antonia. The first question Herod faced was whether to destroy the older Hasmonean Baris or to integrate it within a new fortress/palace.

Having made the latter decision, he then had to decide on the exact location of the new building. At this stage of his life, he probably already envisioned the future enlargement of the Temple Mount in general terms. The new Antonia was built on the same grid as the adjacent Temple Mount. When Herod later undertook to enlarge the Temple Mount on three sides (all but the east), some adjustments no doubt had to be made to the Antonia as a result of Herod’s northern extension of the Temple Mount. It was probably at this point that the two stairways were constructed from the Antonia to the new porticoes of the Herodian Temple Mount.

There remains one loose end. As I noted earlier, south of the northern boundary of the Antonia (marked by the moat) is a vertical cut in the bedrock 12 meters deep and 39 meters long. This is too far north to be the southern boundary of the Antonia. But who made this cut, and when? It is an elegant and precise cut into the bedrock that runs perfectly along the northern (Herodian) end of the Temple Mount. Of all the various candidates for this cut, the only one that seems plausible is that it was cut by the Umayyad caliph Mu’awiyah, who ruled between 661 and 680 C.E. Mu’awiyah was the first caliph after the Arab conquest to initiate and implement reconstruction projects on the Temple Mount. It is generally recognized that the caliph rebuilt the existing east, west and south walls of the Temple Mount. He most probably also cut 12 meters deep and 39 meters long into the rock on the north side of this huge compound. This cut would not only provide for the reconstruction of the walls but also create a more nearly rectangular shape to the Temple Mount, which I don’t believe existed in Herod’s day.