Masada

as da
The
T

Siege—
Sieg —
Gwyn Davies

T
he dramatic archaeological site of Masada,
perched on an isolated mesa-top in the Judean
desert above the southwest corner of the Dead
Sea, is justifiably one of Israel’s premier visi-
tor attractions. The thousands of tourists who come here
every year to visit the spectacular ruins of the Herodian
fortress-palace exposed by Yigael Yadin’s famous excava-
tion between 1963 and 1965 are treated by their guides
to an equally stirring account of the sustained resistance
mounted here by a band of determined Jewish fighters
against the implacable might of the Roman Empire. Even-
tually, with their defenses breached and defeat inevitable,
BARON WOLMAN

the defenders are celebrated for choosing mass suicide

over the ignominy of surrender.
 From the
Roman
Viewpoınt
ROMAN SIEGE WORKS
PREVIOUS PAGES: Perched atop an isolated mesa in the
Judean desert, Masada was the site of the final confron-
tation between the Jewish rebels and the Roman Empire
at the end of the First Jewish Revolt. The rebels had
seized King Herod’s cliff-top palace-fortress during the
early days of the revolt and were subsequently joined by
additional combatants, as well as fleeing civilians, after
the fall of Jerusalem in 70 C.E. The Roman army spent
approximately three months besieging the fortress in
order to stamp out the last remnants of the Jewish resis-
tance. According to Josephus, when the Romans finally
breached the Jewish defenses, they discovered that the
Jews had chosen to commit mass suicide rather than sur-
render and be enslaved.
This human tragedy derives much of its impact
from the austere landscape in which it was enacted,
as well as the story of death over dishonor and of
resistance to external oppression. This chronicle
was ready-made to be harnessed in the fashioning
of a suitable ideological narrative in the early days
of the new state of Israel.
Although today the tour buses to the Visitors’
Center park under the ramparts of one of the
Roman siege camps, and travelers who arrive along
the western approach road make their way to the
summit along the side of the Roman assault ramp,
little direct attention is paid to the extensive archae-
ological remains of the military campaign that saw
the fall of Masada. This article considers the siege
not from the heroic viewpoint of the resistance on
the mesa-top itself but, instead, from the rather
more prosaic perspective of the Romans below.1
The Roman operation against the desert fortress
of Masada during the late winter/spring of either
72/73 or 73/74 C.E. represented the final act of
the First Jewish Revolt against Rome. This rebel-
lion, which had begun so calamitously for impe-
rial forces with the rout of Syrian governor Ces-
tius Gallus’s army at Beth-Horon in 66 C.E., had
proved difficult to contain. Extensive and expensive
military campaigns had been necessary to bring the
rebels to heel. And the triumph celebrated at Rome
in 71 C.E. to commemorate the sack of Jerusalem
in the preceding year amounted to a premature
declaration of victory. Our main source for the his-
tory of the revolt is by a former rebel-commander-
turned-quisling, Flavius Josephus. He provides
only limited details of the situation in Judea after
Jerusalem’s fall, but it is apparent that the Roman
authorities were still left with a significant counter-
insurgency problem.
The concentration of legions, auxiliaries and
client forces that had been necessary to carry out
the siege of Jerusalem had been subject to a rapid
draw-down after 70 C.E. to the extent that the
30
new governor of the province, Lucilius Bassus, was
left with only one legion (and some auxiliary regi-
ments) to deal with the remaining threat.2 “Mop-
up” operations in a war that, officially, had already
been “won” were scarcely likely to have earned the
commander much glory, and the poverty of our lit-
erary record probably reflects the prevailing view
that Emperor Vespasian’s son, Titus, had already
accomplished the hard task of eliminating the ene-
my’s capital.
Nonetheless Bassus proceeded in a methodical
manner to reduce the main hostile concentrations
holding out at the fortresses of Herodium (south
of Jerusalem) and Machaerus (east of the Dead
Sea). Thereafter, he surrounded and wiped out a
guerrilla force that had sought refuge in the “For-
est of Jardes” (unidentified, but perhaps located in
Moabite territory south of Machaerus).
As a result of these actions (concluded in 72
C.E.), the last site to offer continuing Jewish resis-
tance was the isolated fortress of Masada.
Rather than pressing ahead to end matters
immediately, Bassus turned to the more urgent task
of reorganizing the province in line with the direc-
tives he had received from Vespasian, a policy that
suggests that the “threat” emanating from Masada
was not considered particularly pressing. The
defenders of Masada, whom Josephus describes as
Sicarii,3 often misleadingly referred to as Zealots,
had clearly decided to rely on the strength of their
position as their best guarantee against any Roman
attack. These Sicarii had seized Masada in the first
days of the Jewish Revolt and, although joined by
some refugees who had fled after the fall of Jeru-
salem, their numbers were limited. Josephus claims
that there were 967 individuals (both combatants
THE ROMAN SIEGE SYSTEM at Masada illustrates the
methodical approach with which the Roman army pur-
sued the remaining Jewish rebels at the end of the First
Jewish Revolt against Rome. Led by Flavius Silva, the
Legio X Fretensis, a veteran military unit that had been
active in Judea since 67 C.E., began the siege operation
against the Masada rebels in 72 or 73 C.E. A circumvalla-
tion wall (red) 2.5 miles long ringed the site. Eight camps
(designated A through H in yellow) laid down around the
wall served as bases and garrison points for the troops.
The Roman headquarters lies, as marked, northwest of
Camp F. Fifteen numbered turrets—or towers—lined the
eastern and northern stretches of the circumvallation
wall (red). An assault ramp (orange) constructed by the
Roman troops exploited the natural terrain in order to
raise a siege tower to the height of the rebels’ defen-
sive wall. The Romans waged literal and psychological
warfare against the Jews—demonstrating through their
terrifying siege works the futility of resistance against the
empire.
J U LY / A U G U S T 2 0 1 4
 FROM RICHMOND, “ROMAN SIEGE-WORKS OF MASADA,” P. 145, FIG. 5. WITH PERMISSION FROM THE SOCIETY FOR THE PROMOTION OF ROMAN STUDIES

BIBLICAL ARCHAEOLOGY REVIEW

General Headquarters

Masada
ROMAN SIEGE WORKS

31
ROMAN SIEGE WORKS
and civilians) still alive within the fortress on the
eve of its fall.
Bassus fell ill and died while still in harness. This
meant that the reduction of Masada would be left
to his successor, Flavius Silva, who was appointed
to replace him in 72 C.E. Since the exact date of
Silva’s arrival in Judea cannot be determined with
certainty, the date of the Roman siege of Masada
also remains uncertain. Epigraphic evidence sug-
gests that Silva was still in Italy in the spring of
73, so an operation mounted in late winter/spring
73/74 appears likely. After all, considerable time
must have elapsed to allow for the selection of the
new governor, for his travel to Judea, and for prep-
arations to be made before he could march against
the desert redoubt.4
Regardless of the exact date of the operation, it
was clear that Silva would eventually have to tackle
this enemy stronghold. To allow the Masada reb-
els to continue their defiance of Roman imperial
might was unthinkable, and the simple expedient
RAISING THE SIEGE FORT. Silva’s first operational order
was to construct a circumvallation wall around the site
to prevent the rebels escaping. More than 5 feet wide
and 10 feet high, the wall was composed of rough stone
blocks with a rubble core. This section of the wall has
been reconstructed and is located near Wadi Sebbe (see
plan on p. 31), a critical area where the Jews might have
attempted a sortie. The circumvallation wall was built
even over difficult terrain where escape would have been
impossible, further demonstrating to the rebels Silva’s
determination to bring an end to their resistance.
32
of ignoring the defenders until they starved to death
risked the prospect of desperate fighters fleeing
their bolt-hole and rekindling the rebellion in oth-
erwise pacified territories. No matter the difficulties
that a siege might entail, it was essential to bring
matters to a decisive conclusion while the enemy
(the Jews) remained concentrated in one place.
The Romans were well aware of the exten-
sive arsenal holed up within the fortress, lav-
ishly stocked with both provisions and arms. The
capacious cisterns also gave the defenders access
to ample water supplies. In these circumstances,
a passive siege predicated on the deployment of
a relatively small number of troops positioned in
blockade camps around the fortress would not have
been effective. With the defenders capable of sus-
taining themselves for an unknowable duration, the
Roman besiegers would have required a constant
resupply in a hostile environment where foodstuffs
and water would have had to be shipped in at
great expense. Once the decision had been taken
to attack Masada, it was necessary to make a single
concerted effort to bring the siege to a close at the
earliest possible opportunity.
Silva pressed ahead with the conquest of Masada
only after formulating a detailed operational plan.
He was not going to risk a hastily improvised
assault at the end of attenuated supply lines.
Indeed, the Roman governor left little to chance.
First and foremost was his decision to concentrate
the military resources of the entire province on
the task at hand. The main strike force that was
TODD BOLEN/BIBLEPLACES.COM
J U LY / A U G U S T 2 0 1 4

deployed was Legio X Fretensis, a veteran forma-
tion that had been active in Judea since 67 C.E.
onward. The Tenth Legion had participated in
repeated siege operations (including Jerusalem and
most recently at Machaerus) giving unrivaled expe-
rience in the engineering challenges of a siege.
Although the paper strength of the legion may
have approximated 5,500 men, it is unlikely that
its full complement was deployed at Masada, as
some troops must have been held back in super-
visory functions elsewhere in the recently pacified
province. This would have been particularly true
if the operation had taken place in 72/73 rather
than 73/74, as there would also have been less
time for casualties incurred in earlier actions to
have been replenished.
Alongside the citizen soldiers of the legion, Silva
would also have made use of auxiliary cohorts
drawn from provincial garrisons and may also
have had contingents furnished by local client rul-
ers. Although it is impossible to provide an accu-
rate calculation, it is likely that the total man-
power deployed in the Masada operation included
between 7,000 and 10,000 men. It is also likely that
Silva drafted Jewish corvée laborers to act as por-
ters manning his supply lines. Even indirect par-
ticipants in the operation such as these, however,
would have increased the logistical demands on the
besieging force, so they would have been kept to a
necessary minimum.
Masada is girded by precipitous cliffs and flanked
by deeply cut ravines. It had been chosen as one of
King Herod’s most spectacular fortress-palaces,5 osten-
sibly designed to protect the realm against Cleopa-
tra’s overweaning ambitions. Although provided with
a simple casemate wall that ringed the summit, bol-
stered by occasional towers, the real problem for any
attacker was overcoming nature itself.
Preparatory to storming the fortress, Silva would
have been required to undertake extensive engi-
neering preliminaries (1) to prevent the escape
of any defenders and (2) to enable his troops to
mount an attack against the enemy wall.
Modern scholars can be thankful that the iso-
lation of the site and the aridity of the environ-
ment allow an unrivaled opportunity to examine
the material remains of Silva’s siege operation and
to correlate the surviving archaeological record
with Josephus’s narrative. Indeed, Masada gives us
our most complete surviving siege system of the
ancient world.
The investigation of the Roman field works at
Masada has been on a much more limited scale
than the extensive excavations conducted on the
mesa-top itself. In the early 1960s, Shmaria Gutman
BIBLICAL ARCHAEOLOGY REVIEW
ROMAN SIEGE WORKS
GWYN DAVIES
The Wood at Masada—
Where Did It Come From?
Where did the Romans get the wood to brace the main body of the stone-and-
earth-filled assault ramp at Masada? A well-known study conducted by research-
ers from the Weizmann Institute of Science began with the assumption that the
Romans gathered the wood in the immediate area. The researchers analyzed the
carbon and oxygen isotopic composition of the ramp’s timber remnants (which
were primarily Tamarix jordanis) as well as the isotopic composition of modern
samples of T. jordanis from the Masada region and from sites in the central
Negev and the Judean foothills.1 Comparing the ancient and modern samples, the
Weizmann researchers concluded that the wood from the first-century ramp came
from an environment that was more humid than what exists in the Judean desert
today and proposed that the region experienced significant climate change over a
period of two millennia.
A recent study conducted by researchers from the University of Haifa,
however, concludes that the Romans must have imported the wood from else-
where.2 Examining the botanical, archaeological and textual evidence, the Haifa
researchers created models of local wood availability, wood consumption during
the occupation of Masada and the timber needs of the Romans to construct the
siege ramp. The study demonstrated that even if the area around Masada had
more than a normal amount of timber available, this would not have met the
needs of the Romans for their large-scale siege. By the time the Romans arrived
at Masada in 72 or 73 C.E., the entire area would have already been denuded
of trees due to the massive exploitation of local timber from the time of the
cliff-top’s occupation in 150 B.C.E. up to the time of the Roman siege. The Haifa
researchers suggest the Romans may have imported the wood from a more
humid and cooler region, such as from the wadis fed by the Moab Mountains to
the east of the Dead Sea.—R.N.
1 Dan Yakir, Arie Issar et al., “13C and 18O of Wood from the Roman Siege Rampart in Masada, Israel (AD
70–73): Evidence for a Less Arid Climate for the Region,” Geochimica et Cosmochimica Acta 58 (1994), pp.
3535–3539. See also Arie S. Issar and Dan Yakir, “Isotopes from Wood Buried in the Roman Siege Ramp of
Masada: The Roman Period’s Colder Climate,” The Biblical Archaeologist 60 (1997), pp. 101–106.
2 Simcha Lev-Yadun, D.S. Lucas and Mina Weinstein-Evron, “Modeling the Demands for Wood by the Inhabit-
ants of Masada and for the Roman Siege,” Journal of Arid Environments 74 (2010), pp. 777–785.
33
ROMAN SIEGE WORKS

investigated (and reconstructed) Camp A in the

Roman siege system while Yadin conducted test-
pitting in Camp F. In 1995 Benny Arubas, Gideon
Foerster, Haim Goldfus and Jodi Magness under-
took the first modern examination of some of the
other Roman structures. Apart from Magness’s
pottery report, however, the results of these exca-
vations have yet to be published.6 Therefore our
understanding of the Roman siege works at Masada
must be derived from combining the account in
Josephus with a field inspection of its surviving
component elements.
Josephus informs us that Silva’s first opera-
tional order after deploying his men in various
garrison posts around Masada was to ensure the
hermetic sealing of the fortress: Silva “threw up
a wall all around the fortress to make it difficult
for any of the besieged to escape, and posted sen-

DANNY HERMAN
tinels to guard it.”7 This circumvallation wall, much
of which still survives, extends for more than 2.5
miles and circles the entire site.
This was a standard operating technique of the
Roman army and had been deployed elsewhere
during the Jewish Revolt, most notably at Jerusa-
lem, Machaerus and Yotapata. At Masada it com-
prises a wall of rough stone blocks that retain a
rubble core. It still stands up to seven courses high
FIRE POWER. Turret 11 (see plan on p. 31) was one of
15 turrets mounted with catapults positioned along the
northern and eastern stretches of the Roman circumvalla-
tion wall. Ten of the turrets were placed along a section
of relatively flat terrain where the Snake Path descends
from the eastern side of the summit, the most likely path
of a rebel sortie or attempt to escape.
in places. It is more than 5 feet wide and prob-
ably had an original height of about 10 feet. Quite
remarkably, Silva extended it across very difficult
terrain, even in sectors where an escape attempt
would have been next to impossible. The only
gap in the wall lies along the plateau to the south
where a sheer cliff plunges down to the wadi mak-
ing any artificial barrier redundant.
The relentless pursuit of a monolithic build-
ing scheme even at the expense of strict military
necessity suggests that the Romans were con-
sciously adopting a form of psychological warfare by

GWYN DAVIES

34 J U LY / A U G U S T 2 0 1 4

ramming home to the rebels the message that they
were now completely isolated without any hope of
relief or escape. This message might also have served
as a salutary reminder to any client forces present of
the implacable determination of the Roman state to
pursue rebels to the very ends of the world and to
inflict due retribution for their defiance.
Along the eastern and northern stretches of the
circumvallation wall, the besiegers added a series of
15 turrets on which could be mounted light cata-
pults. Ten of these turrets are positioned along the
circumvallation wall where it runs beside the rela-
tively flat plain at the base of Masada’s eastern cliff.
This is the side on which the Snake Path descends
from the summit; these turrets would have pro-
vided fire support in the event of a rebel attempt at
a sortie. The remaining turrets seem to have been
positioned to provide flanking fire in the event of
any attempt to break out along the stream-beds of
either the Wadi Nimre or Wadi Sebbe.
Eight camps (designated A though H) to house
the besieging force are distributed around Masada.
The largest of these, Camps B and F, include nearly
5 acres each. Presumably these two camps served as
the bases for the soldiers of the Tenth Legion, with
each camp exerting command and control functions
over the eastern and western sectors of the cir-
cumvallation wall.8 Camp B would have been best
suited to act as the main depot for the entire sys-
tem (where supplies transported along the Dead Sea
could have been stored), while Camp F would have
acted as Silva’s operational headquarters from which
the assault preparations would have been directed.
BIBLICAL ARCHAEOLOGY REVIEW
ROMAN SIEGE WORKS
MOUNTING THE ASSAULT. Spread over 5 acres, Camp F,
seen in the middle ground, was one of the two largest
Roman camps at Masada. Positioned at the base of the
western side of the cliff, Camp F probably served as a
command center. Camp E, discernible to the left of Camp
F, had two gateways through its front defenses, which
allowed for rapid reinforcement if the rebels attempted
a sortie. The great assault ramp on the western slope of
the cliff (see pp. 28–29) is the most conspicuous surviv-
ing evidence of the Romans’ pursuit of the Jewish rebels.
Building upon a natural spur that abuts the mountain,
the Romans constructed a ramp composed of stone and
earth reinforced with timber bracings. The ramp was
designed to raise the ironclad siege tower into position to
fire against the rebels.
The 1995 excavations at Camp F recovered a few
examples of high-quality pottery and glass largely
concentrated in one area, suggesting that Silva’s
personal quarters were located here.
The discrete patterning of the scatter of stones
within the walls of the camp allows us to roughly
reconstruct the camp’s internal buildings. Given the
temporary nature of the operation and the local
scarcity of timber, these Roman buildings would
have consisted of tent units raised on top of dwarf
stone walls designed to increase internal head-
room and to improve the circulation of air. Most
of these structures represent the tent lines of the
troops themselves. Some other “buildings” have
also been identified, for example, a probable hos-
pital in Camp B.
Apart from the main legionary bases (Camps B
and F), the six other camps in the system (roughly
an acre in size) would likely have served as garrison
35
ROMAN SIEGE WORKS
points for auxiliary units manning the circumvalla-
tion wall. Five of these camps butted up directly
against the circumvallation wall, while Camp C,
likely the base for a cavalry unit tasked with patrol-
ling the most vulnerable sector, was recessed imme-
diately behind the encircling circumvallation wall.
It seems that each of these smaller camps was also
designed to perform various specialist functions.
Camp E, for example, is located behind the Roman
assault ramp; it had two gateways through its front
defenses, presumably to allow for the rapid rein-
forcement of those working on the ramp in the event
of any serious hostile sortie. Camp H, isolated on
the scarp edge above the Wadi Sebbe, was primarily
designed as an observation post enjoying sweeping
oversight across the Masada mesa-top. The irregular
“key-hole” configuration of Camp G suggests that
the Romans may have extended this particular base
while the siege was in progress, perhaps because
its original design did not allow for adequate super-
vision of the streambed of the Nahal Masada.
One other enclosure attached to the back of the
circumvallation wall, located west of the assault
ramp, is usually referred to as “the Engineer-
ing Yard.” It is a leveled platform where traces
of metal working activities have been detected
by Ian Richmond in the early 1960s.9 It is likely
the place where the ironclad Roman siege tower
was put together. Josephus claims this siege tower
was nearly 100 feet high. This engine, presumably
transported to the site in prefabricated segments,
would have required a secure site for its assembly,
and it would have made sense to select a place only
a short distance from the point at which it would
eventually be deployed. Perhaps appropriately (if
archaeologically unfortunate) the National Park
Authority has chosen to locate its modern works
and storage depot at exactly this site, so there is
little chance of recovering any further details of
this Roman establishment.
Various other structures, mostly one-to-three-
roomed tent units, can be found scattered along the
hillside in the vicinity of Camps E and F. These are
probably canabae, civilian settlements supporting
the military. The Masada canabae have been vari-
ously interpreted as the premises of camp followers
engaged in providing services for the Roman siege
force or, perhaps more likely, as the accommoda-
tions for the Jewish corvée. Although most of these
appear to be very simple structures, some units
biblicalarchaeology.org/Masada
Visit us online to download the new free eBook Masada:
The Dead Sea’s Desert Fortress.
36
enjoy a more complex plan, including hallways and
dining areas, suggesting that they may represent the
communal mess facilities for these draftees. The rel-
atively flimsy nature of most of these remains and
their exposure to run-off as well as their proxim-
ity to the modern son-et-lumière complex make it
urgent that some of the more vulnerable examples
of these canabae be investigated archaeologically
before further evidence is lost.
The careful disposition of these camps and ancil-
lary structures evidences the methodical and deliber-
ate nature of Silva’s campaign to suppress the rebels
on Masada. However necessary these measures may
have been for isolating the defenders and for making
the besiegers’ task more feasible, the construction of
the circumvallation and its related works were mere
preliminaries to the direct approaches that would
be necessary if Masada itself was to be secured. For
that a siege ramp was needed.
The most striking surviving image of the Roman
siege works at Masada is the great assault ramp (or
agger) that Silva’s men threw up against its western
slopes. This was the only flank on which it might
have been constructed, not least because it is the
location of a natural spur (called the Leuke, or
“white promontory,” by Josephus10) that abuts the
isolated mountain. The underlying geology certainly
exaggerates the seeming Roman construction of the
ramp; the artificial deposition of material consists
only of about 90 feet above bedrock; in short, much
of the ramp consists of bedrock.* The manipula-
tion of the environment that it entails nevertheless
reminds us of the Roman determination to expunge
the last vestiges of the Jewish rebellion.
The ramp is almost 750 feet long and about
650 feet wide at its base, narrowing as it climbs to
about 150 feet at the top, which is about 240 feet
above its base.11 Oddly enough, this is still 42 feet
short of the Masada summit.
Josephus notes that the ramp was once crowned
by a stone platform. This stone cap or platform
would have served as a reinforced base for the
siege tower that housed the Romans’ battering ram
and would have raised this engine to within strik-
ing range of the defenders’ casemate wall. Although
no trace of this platform remains in situ today, the
extensive talus (the stone field) on both sides of the
ramp may represent the debris that in time top-
pled from this structure. Early visitors to the site
remarked on this feature of the ramp.
The main body of the ramp is composed of
stone and earth reinforced by timber bracing that
C O N T I N U E S O N PA G E 7 0
*See Dan Gill, “It’s a Natural—Masada Ramp Was Not an Engineering
Miracle,” BAR, September/October 2001.
J U LY / A U G U S T 2 0 1 4
Roman Siege Works
continued from page 36
still projects in places above the surface.
These timbers of tamarisk and other arid-
environment tree species seem to have
formed a box-revetment filled with com-
pacted earth and small stones. Presumably
this revetment would have been situated
on a series of ascending parallel steps that
the Roman engineers would have hacked
out of the bedrock as they proceeded
to “bench” each side of the spur, much
as modern engineers sometimes scarp
the sides of highway cuttings today. The
resulting double flight of steps ascending
the slope would then have accommodated
the box-revetment. With the flanks of the
spur now stabilized, the intervening space
between the timber revetments would be
filled with dumped material that was then
consolidated by ramming.
The entire point of this ramp was to
AUTHORS
Gwyn Davies
(“The Masada
Siege,” p. 28) is
associate professor
of history at Flor-
ida International
University and
codirector of the
Davies Yotvata Roman
Fort Project. Spe-
cializing in Roman field works, he has
published numerous articles on the
topic as well as the book Roman Siege
Works (Stroud: Tempus, 2006).
Theodore Feder
(“Parsing ‘The
Parting’ Painting,”
p. 42) is an art
historian and
president and
founder of Art
Resource, the
world’s largest
Feder
photo archive of
fine art, as well as president of the
Artists Rights Society, both in New
York City. He is author of Great
Treasures of Pompeii and Herculaneum
(New York: Abbeville Press, 1978).
70
serve as an approach path for the siege
equipment to be deployed directly against
the defenders’ casemate wall. The angle
of the ramp had to be carefully calculated
to ensure that the siege tower could be
raised into position. The siege tower was
an ironclad structure nearly a hundred
feet tall, according to Josephus. It was
designed to resist incendiary attack by the
rebels. The siege tower contained the ram
housing, artillery for clearing the defend-
ers from their parapets and extensible
gangways to allow the storming party
access to the enemy defenses. Once the
ramp had been completed and a timber
track laid on its surface, the siege tower
would have been elevated with a winch in
short incremental steps. If the defenders
made no serious attempt at a disruptive
sortie (and Josephus does not mention
any), the tower’s elevation from the base
of the ramp to its summit may have taken
only a few hours.
Emanuel Tov
(“Original Bible,”
p. 48) is J.L. Magnes
Professor of Bible
Emeritus at the
Hebrew University
of Jerusalem, Israel.
He served as editor-
Tov in-chief of the offi-
cial International
Dead Sea Scroll Publication for nearly 20
years, during which 32 volumes of Dead
Sea Scrolls were published. Professor Tov
has recently published a revised and
expanded third edition of his Textual Crit-
icism of the Hebrew Bible (Fortress, 2011).
Patricia Smith
(“Infants Sacri-
ficed?” p. 54) is
director at the Lab-
oratory of Bio-
anthropology and
Ancient DNA, Fac-
ulties of Medicine
Smith and Dentistry,
Hadassah-Ein
Karem, and Joel Wilbush Professor
Emeritus of Medical Anthropology at
The Hebrew University in Jerusalem.
Although recent accounts have been
published suggesting that the assault ramp
had not been completed by the time that
Masada fell to Silva,12 this hypothesis
seems very unlikely. To begin with, if
the ramp was incomplete, it is not clear
how the Romans would have succeeded
in breaching the enemy defenses which
they clearly did. Secondly, it is surely not
coincidental that the location of the breach
in the defenders’ casemate wall lies imme-
diately above the modern summit of the
ramp. Thirdly, the distribution of stone
ballista projectiles recovered from an arc of
loci within the fortress suggests the field of
fire that would have come from catapults
mounted on a siege tower; the siege tower
would not have been raised into position
had the ramp not been completed.
Despite the impressive scale of these
Roman works at Masada, there is little
reason to suppose that the siege would
have taken much more than three months
to bring to a conclusion.
Josephus states how Masada’s pri-
mary defenses were first overthrown
by ram-strikes. Then the rebels’ hastily
extemporized secondary defenses were
consumed in an incendiary attack. That
night, with further resistance now impos-
sible, Josephus relates how the defenders,
encouraged by their leader Eleazar Ben-
Yair, entered into a suicide pact rather
than face the inevitability of retribution
when Roman forces entered the for-
tress at dawn. When the Roman troops
entered, they met only silence. a
1  For a more detailed account of the Roman
siege system employed at Masada, see Gwyn
Davies, “Under Siege: The Roman Field Works
at Masada,” Bulletin of the American Schools of
Oriental Research 362 (2011), pp. 65–83.
2 It seems that after the fall of Jerusalem, Judea
was organized in an unusual way so that the
provincial governor was also, simultaneously, the
commander (or “legate”) of its one legion garrison.
3 A violent sect who engaged in the assassination
of their many political enemies in the pursuit of
their messianic goals.
4  The alternate view of 72/73 is largely based
on the discovery of a fragmentary papyrus from
the Masada summit (presumably attributable
to the Roman garrison left in place after the fall
of the fortress) addressed to one Julius Lupus,
possibly the same individual appointed as Pre-
fect of Egypt in February 73, implying that the
fortress must already have fallen to the Romans
by that date. However, it is equally possible that
this was a copy of a letter (or even an original
that was never sent) written by an author who
belonged to the post-conquest Masada garrison.
For more details see Hannah M. Cotton and
J U LY / A U G U S T 2 0 1 4
Joseph Geiger, Masada II: The Yigael Yadin
Excavations 1963–1965, Final Reports: The Latin
and Greek Documents (Jerusalem: Israel Explo-
ration Society, 1989), pp. 21–23.
5 See Amnon Ben-Tor, Back to Masada
(Jerusalem: Israel Exploration Society, 2009)
for a lively summary of the eight volumes of
reports that have so far been published from
Yadin’s excavations.
6 Jodi Magness, “The Pottery from the 1995
Excavations in Camp F at Masada,” Bulletin of
the American Schools of Oriental Research 353
(2009), pp. 75–107.
7 Josephus, The Jewish War, VII.276, H. St. John
Thackeray, ed. and trans., Loeb Classical Library
(Cambridge, MA: Harvard Univ. Press, 1927).
8 The steep escarpment edge of the rift valley
would have provided a natural delimitation
between the two operational sectors, although
a carefully engineered zigzag track exiting from
Camp D served as a link between the two zones.
9 Ian A. Richmond, “The Roman Siege-Works
of Masada, Israel,” Journal of Roman Studies 52
(1962), p. 154.
10 Josephus, Jewish War VII.305.
11 Josephus seriously overstates the dimensions of
Silva’s agger. See Josephus, Jewish War VII.307.
12 For the most accessible version of this argu-
ment, see Benjamin Arubas and Haim Goldfus,
“Masada. The Roman Siege Works,” in Ephraim
Stern, ed., The New Encyclopedia of Archaeo-
logical Excavations in the Holy Land, vol. 5
(Jerusalem: Israel Exploration Society, 2008),
pp. 1937–1940.
Forgotten Kingdom
continued from page 41
“Archaeology, Urbanism and the Rise of the
Israelite State,” in Walter E. Aufrecht, Neil
A. Mirau and Steven W. Gauley, eds., Aspects
of Urbanism in Antiquity, From Mesopotamia
to Crete (Sheffield: Sheffield Academic Press,
1997), pp. 172–193; Lawrence E. Stager, “The
Patrimonial Kingdom of Solomon,” in William
G. Dever and Seymour Gitin, eds., Symbiosis,
Symbolism and the Power of the Past: Canaan,
Ancient Israel and Their Neighbors from the
Late Bronze Age Through Roman Palaestina
(Winona Lake, IN: Eisenbrauns, 2003), pp.
63–74; Raz Kletter, “Chronology and United
Monarchy: A Methodological Review,”
Zeitschrift des Deutschen Palastina-Vereins
120 (2004), pp. 13–54; Amihai Mazar, “The
Spade and the Text: The Interaction Between
Archaeology and Israelite History,” in H.G.M.
Williamson, ed., Understanding the History
of Ancient Israel (Oxford: Oxford Univ. Press,
2007), pp. 143–171.
3 Among the book’s many other distortions, I can
list here only a few: (1) Finkelstein claims car-
bon-14 dates have corrected the dates of Ramses
III (p. 24). Actually they are exactly the same.
(2) Finkelstein claims Shechem was destroyed
at the end of the Late Bronze Age (p. 22). The
excavators have emphasized that it was not. (3)
Finkelstein claims that Tell Keisan, Tel Kinrot,
Tel Re ov, Yokneam and Dor were all “Canaan-
ite city-states” (p. 30). But “city-state” is never
defined, and at least two that are so claimed are
Phoenician, one is probably Aramaic, and none
BIBLICAL ARCHAEOLOGY REVIEW
would actually qualify as a city-state. (4) Fin-
kelstein claims that there are dozens, even hun-
dreds, of carbon-14 dates supporting the “low
chronology” (p. 33); in the latest Megiddo report
(Megiddo IV), there are three published for the
pivotal Stratum VA/IVB, and if anything they
support the conventional chronology. (5) Finkel-
stein claims that Jerusalem in the tenth century
B.C.E. was a poor village with no monumental
architecture (p. 43). Even Finkelstein’s colleague
Nadav Na’aman disagrees with him, as nearly
all archaeologists do. (6) Finkelstein radically
challenges conventional dates by putting the
Iron I/IIA transition in the second half of the
tenth century B.C.E. (p. 64). That’s scarcely later
than most, and even earlier than Amihai Mazar’s
“modified” conventional chronology. Finkelstein
claims that Hazor X was destroyed in the late
ninth century (840–800 B.C.E.), as confirmed by
carbon-14 dates (pp. 75, 122). But no evidence is
cited for this, and excavator Amnon Ben-Tor dis-
agrees. (8) Finkelstein claims that those scholars
who see Jerusalem as an early state capital are
“desperate,” Bible-based people (p. 80). That
tells us who is really desperate. (9) For the
view that the Field III city gate at Gezer dates
to the ninth century B.C.E., Finkelstein cites
me (William G. Dever et al., “Further Excava-
tions at Gezer, 1967–1971,” Biblical Archaeologist
34 [1971], p. 103). I never said anything of the
sort—quite the opposite. (10) Finkelstein says
that Megiddo in the ninth century B.C.E. was
“set aside for breeding and training horses”
for chariotry (pp. 113; 133–135). Some of his
own staff members (and others) dispute the
famous “stables” in Megiddo IV. (11) Finkelstein
claims that Tel Masos near Beersheba was the
center of a far-flung “desert polity” in the tenth
century B.C.E. (p. 126). But the relevant Stratum
II follows a massive destruction of the walled
town, and the scant remains consist of only a
few tattered houses. There hardly seems any
point in continuing. Finkelstein simply does not
care much about facts, as many have long since
concluded.
Burke Endnotes
1  While Finkelstein acknowledges that Hazor,
a major site in his analysis of the Iron Age
northern kingdom, was “probably the most
important city-state in the north” (p. 21),
neither its Late Bronze Age nor Iron I phases
are discussed, presumably because it would
complicate the highland-centered interpretive
framework he offers. The weakness of this
analysis is the mistaken assumption that chap-
ter one establishes a Braudelian longue durée
perspective (as explicitly stated but only in the
concluding chapter), when in fact this analysis
does not meet those criteria.
2  For example, at the start of one particular
paragraph we are told that the transition from
Iron I to Iron IIA “should probably be fixed
... in the beginning of the second half of the
tenth” century (i.e., 950 B.C.E.; p. 63). This is,
however, substantially later than Finkelstein’s
low chronology start date of 920 B.C.E. by 30
years, or it is half the distance between the
start date for Iron IIA in the so-called Low
Chronology date (920 B.C.E.) and that of
the Modified Conventional Chronology (980
B.C.E.). (Keep in mind that such seemingly
small decadal shifts in the chronology is what
we are fundamentally talking about, whether
in connection with the shortening of David
and Solomon’s reigns as raised by the Bibli-
cal tradition—to less than the 40 years each
assigned to them—or in the shifting of the start
dates of Iron IIA later.) However, at the end
of the same paragraph we are asked to accept
that Finkelstein would place the transition
between 940/930 B.C.E. (a figure seemingly
grabbed out of thin air), conceding 10 to 20
years on the 920 date for no explicitly stated
reason (p. 94). Attentive readers will wonder
what they are missing, given that three differ-
ent dates are suggested for the start of the Iron
IIA (i.e., 950, 940/930 and 920). The answer
would be a litany of relevant publications that
are not discussed.
71