ArticlePDF Available

The Archaeology of Ritual: The Sanctuary of Pan at Banias/Caesarea Philippi


Abstract and Figures

From the third century B. C. through the fifth century A. D. a sanctuary to the Greek god Pan existed at the mouth of the Jordan River. At its founding, the sanctuary served as a rural shrine for the local pagan population; when it was abandoned, it had long been the city shrine of Caesarea Philippi, whose population included Jews and Christians. While the cult's longevity is generally seen as reflecting the stability of local religious life, fundamental historical changes suggest that cult rituals must have changed over time. Abundant ceramic remains provide evidence for reconstructing those changes. In Hellenistic times, worshipers from nearby settlements brought local household pottery in which they made dedicatory meals, suggesting that they spent some time at the site. When the sanctuary became a civic shrine in the first century A. D., simpler dedications-such as lamps-became common. By the second century, impressive buildings and sculpture transformed the sanctuary into a formal site, and private rites seem to have been abandoned. Individual patronage resumed, however, in the third and fourth centuries, as indicated by the presence of several thousand lamp dedications. The cult's popularity at that period is impressive, although the character of the dedications indicates that worship was essentially passive. The sanctuary was abandoned by the mid-fifth century. No evidence exists for its purposeful destruction, although by that time the shrine housed a pagan cult in an increasingly Christian city.
Content may be subject to copyright.
The Archaeology of Ritual: The Sanctuary of Pan at Banias/Caesarea Philippi
Author(s): Andrea M. Berlin
Bulletin of the American Schools of Oriental Research,
No. 315, (Aug., 1999), pp. 27-45
Published by: The American Schools of Oriental Research
Stable URL:
Accessed: 07/04/2008 17:10
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact
Department of Classical and Near Eastern Studies
University of Minnesota
Minneapolis, MN 55455
From the third century B.C. through the fifth century A.D.
a sanctuary to the Greek
god Pan existed at the mouth of the Jordan River.
At its founding, the sanctuary served
as a rural shrine for the local pagan population; when it was abandoned, it had long
been the city shrine of Caesarea Philippi, whose population included Jews and Chris-
tians. While the cult's longevity is generally seen as reflecting the stability of local
religious life, fundamental historical changes suggest that cult rituals must have
changed over time. Abundant
ceramic remains
provide evidence
for reconstructing
changes. In Hellenistic times, worshipers from nearby settlements brought local house-
hold pottery in which they made dedicatory meals, suggesting that they spent some time
at the site. When the sanctuary became a civic shrine in the first century A.D.,
dedications-such as lamps-became common.
By the second century, impressive
ings and sculpture transformed
the sanctuary into a formal site, and private rites seem
to have been abandoned. Individual
patronage resumed, however, in the third
centuries, as indicated by the presence of several thousand lamp dedications. The cult's
popularity at that period is impressive, although the character of the dedications indi-
cates that worship was essentially passive. The sanctuary was abandoned by the mid-
fifth century. No evidence exists
for its purposeful destruction, although by that time the
shrine housed a pagan cult in an increasingly Christian city.
tudents of religion need to abandon the
notion of 'essence,' of a unique differ-
entium... as well as the socially impos-
sible correlative of a community constituted by a
systematic set of beliefs. The cartography
messier. We need to map variety, ... which appears
as a shifting cluster of characteristics that vary
over time" (Smith 1982: 18).
For a period of 700 years-from the early third
century B.C. through the fifth century A.D.-there
was a cult and sanctuary
at the mouth of the Jordan
River dedicated to the Greek god Pan. When the cult
was founded no one lived in the environs; it was a
rural shrine, which the ancient historian Polybius
referred to as "the Panion"
for its religious associa-
tion (Histories 16.18.2, 28.1.3). By the time the cult
was abandoned, its surroundings
had changed con-
siderably: the sanctuary had long been integrated
into the busy urban
landscape of Caesarea Philippi,
the city that Herod Philip established in 2 B.C. in the
plain immediately below the old shrine. The sanctu-
ary's longevity has been an unremarked
non, seen as proof of the popularity of the cult and
the stability of local religious life. However, even a
cursory outline of changes in the region's history
and the sanctuary's
situation suggests that although
the name and nominal deity remained
the same, cult
ritual must have changed greatly over time. Here I
offer a description of those changing rituals, in the
hope that they will provide contours for the sort of
map Smith envisioned (1982: 18).
Literary sources provide little information about
the sanctuary,
and none concerning activities or rit-
uals; aside from the simple topographical reference
Fig. 1. Map
of Palestine.
provided by Polybius, only Josephus (A
JW 1.21.3) and Eusebius (Ecc. Hist. 7
mention the Panion. Six seasons of
(1988-1994) under
the auspices of the ]
uities Authority,
however, have cleared
terrace and produced ample archaeologi
including 6 structures;
100 crates of tel
tiles; about 200 kg of stone inlay fragr
2500 fragments of animal bones; 10 c
with tiny pieces of blown and moldma
of about 10 altar
fragments of marble and limestone sc
2000 kg of pottery.1
Of these finds
remains are the most informative for re
ritual practices, for three reasons. First,
securely dated regardless of their findsp
thus be associated with their respective
use. Second, ceramics exist from every I
sanctuary's life, and so provide a coi
and continuous category of evidence. Tl
remains are especially illustrative of the offerings
of individual worshipers, since they constitute the
most readily available and least expensive type of
Banias- J. Z. Smith has described offerings, the physical
CasareaPhilippi residue of ritual, as "linguistically impoverished"
(1987: 102). Their recovery and explanation
is, how-
ever, a crucial component of the study of religious
life in antiquity. The rituals that they reflect are,
A OF as R. MacMullen has put it, the mechanisms "by
[JLEE which religion was established close to the center
"niat of daily life and therefore forced... on people's
(MacMullen 1981: 36). Studies of offer-
ings at Greek sanctuaries
have traditionally
on their evidence for economic life and political co-
hesiveness, rather than on ritual significance (Snod-
grass 1989-1990; Morel 1989-1990; Burkert
Nevertheless, careful analysis of the sanctuary's
ramic assemblages by period can provide insights
on a host of fundamental
issues, including the char-
acter of individual
the nature of ceremonies
and offerings; the status of the deity (or deities) to
whom those offerings were made; and the relation-
ship between the sanctuary, the city of Caesarea
Philippi, and the larger region, including Mt. Her-
mon, the Golan Heights, and the Hula Valley. More-
over, this large body of offerings from a pagan
that remained operative through the birth
and early empowerment of the Church
vividly doc-
uments MacMullen's
(1981: 134) that "the
enormous thing called paganism ... did not one day
tnt. 15.10.3, just topple over dead."
7.17) briefly
he sanctuary
ical remains, The Sanctuary
of Pan is located beneath the cliffs
roof of Mt. Hermon, at modem Banias in the northwest-
nents; about ern corner of the Golan Heights (fig. 1). The sanc-
artons filled tuary lies on a 250 foot long narrow terrace, at one
de glass; 20 end of which is a huge natural cave, and beneath
about 200 which is a ravine from which one of the sources of
ulpture; and the Jordan
issues forth. The name and location of the
,the ceramic Sanctuary of Pan have been known since antiquity
constructing because of carved niches and inscriptions visible in
they can be the rock face above the springs. Explorers in the
)ots, and can 19th and early 20th centuries identified
the site, cop-
e periods of ied the inscriptions, surveyed the visible remains,
of the and collected what carvings and artifacts
there were.
mprehensive Detailed exploration and excavation were not, how-
ceramic ever, pursued, because the narrow terrace immedi-
Fig. 2. Plan of the Sanctuary
of Pan at Banias/Caesarea
Philippi (after MaCoz
ately beneath the niches was strewn with boulders,
and enormous rock fall from an earthquake
in 1837
had blocked the entrance and floor of the grotto. Ar-
chaeological survey and salvage excavations were
conducted in and around the ancient city center after
1967. The sanctuary terrace itself was the site of
active excavations from 1988 to 1994 (MaCoz
1995; 1996).
The large natural
grotto at one end of the terrace
was, presumably,
the site of the earliest cult activity
(although no remains have been found inside). Con-
struction atop the terrace itself began in the early
Roman period (late first century B.C.), with the first
building placed immediately in front of the cave.2
Successive constructions were added east of that
building, until the entire length of the terrace was
filled (fig. 2). The buildings include (from west to
east, i.e., earliest to latest): (a) an elaborate lime-
stone and marble
propylon, probably
the Augusteion
mentioned by Josephus (Ant. 15.10.3; cf. Roller
1998: 190-92); (b) an open-air court fronting an
artificial cave in the cliff face, called the "Court of
Pan and the Nymphs" (after a later inscription over
a niche above the cave); (c) a tetraprostyle
dedicated in the reign of Trajan or Hadrian, called
the "Temple
of Zeus and Pan";
(d) a second, narrow,
open-air court fronted by a staircase from the road
above the spring, called the "Nemesis Court"
a dedicatory inscription over a niche in the cliff
face at the rear of the court); (e) a building with
three long halls, fronted
by a paved street, called the
Building"; and (f) an apsidal structure
with two tiny back rooms, probably
a shrine or small
temple, called the "Temple of Pan and the Goats"
Almost all of the Hellenistic and early Roman
pottery found at the sanctuary was recovered from
soil deposits around the natural bedrock outcrops
140 -
Fig. 3. Total number of diagnostic ceramic fragments
and vessels found at the Sanctuary
of Pan, by period.
along the edge of the terrace slope. Despite its un-
stratified findspot, most of this material can be
confidently dated on the basis of a ware's known
periods of production
(as in the case of Eastern
illata A, or ESA), or well-dated comparanda.
to late Roman
second to mid-
fifth century A.D.) was found under structures built
on the terrace itself. This material is datable by a
combination of associated numismatic or inscrip-
tional evidence, as well as by comparison with sim-
ilar forms from nearby sites. The amount of pottery
from each period varied greatly; figure 3 indicates
the total number
of identifiable
or vessels
found at the sanctuary.4
Hellenistic Period (Third to Mid-First Century B.C.)
From the topographic reference provided by Po-
lybius, we can infer that the Panion must have been
established by the year 200 B.C. It is thus reason-
able to identify the Hellenistic pottery found on and
below the terrace as the residue of cult activities
there.5 More than 95% of that pottery was found in
soil pockets below the Court of Pan and the Nymphs
and the Temple of Zeus and Pan. These Hellenistic
finds must be taken as representative of the period
as a whole, because the lack of stratified contexts
and the continued production
of many third century
B.C. wares and forms into the second century B.C.
Fig. 4. Total number of diagnostic Hellenistic ceramic frag-
ments found at the Sanctuary
of Pan, by function.
precludes our knowing precisely when vessels were
brought to the sanctuary. The Hellenistic corpus
consists of extremely small, worn fragments that
rarely make a partial profile. In this earliest period
of the sanctuary's
life, before buildings and court-
yards covered the terrace, offerings would have
been periodically cleaned from altars within or im-
mediately in front of the grotto and swept over the
sloping hill, thus accounting for the lack of joining
fragments or restorable vessels. As a consequence,
some of these periods' ceramic remains are probably
missing, and caution must be exercised when com-
paring total quantities.
The Hellenistic ceramic assemblage reflects a
good deal about the character of that period's cult
activities. Over 90% of the offerings consist of cook-
ing vessels, of which most show signs of use, and
table vessels intended for eating or drinking
(figs. 4-
5). While these may have been proffered as dedi-
catory offerings (the cooking vessels presumably
holding some food), a more likely explanation
is that
this assortment
is evidence for dining at the sanctu-
ary. Table and kitchen vessels could be "official"
possessions of a cult. Plates and cups are specifically
cited, for example, on a fourth century B.C. inven-
tory from a sanctuary of Hera in Boiotia, Greece
(Tomlinson 1980: 221-24). Within a cave in the
precinct of Poseidon at Isthmia, fourth century B.C.
cooking pots, casseroles, kraters,
jugs were found
30 BASOR 315
in a pithos in a deep hole dug into the ground
(Broneer 1962: 4-7). At the Panion, however, the
lack of buildings that could have held such goods
makes it more probable that they were brought by
The preparation and consumption of food and
drink at the sanctuary might be characterized as
but it could also be interpreted,
informally, as picnicking. During the Hellenistic pe-
riod, the Sanctuary
of Pan was a rural site, situated
at one of the most abundant
perennial water sources
in the region. There was no city anywhere nearby,
the closest settlements being some small pastoral
encampments and farmsteads in the northern
Heights, on the slopes of Mt. Hermon, and in the
Hula Valley.7 The largest of this area's Hellenistic
period sites was the villa at Tel Anafa, about 6 km
to the southwest, in the northern Hula Valley (Her-
bert 1994). In other words, the sanctuary,
though not
out of the way, was also not especially convenient
to any settled place, and a visit must have required
some time and planning. At least some of the many
vessels found at the site are likely to be the residue
of informal visits to the grotto, the terrace, and the
spring. Ancient authors describe such activities at
elsewhere, for example at the Sanc-
tuary of Aphrodite at Knidos, where "shady trees
[covered] couches for those who wished to feast
themselves there"
Amores 12). The mise-
en-scene of Menander's
Dyskolos comes to mind as
well: At a cave sacred to Pan and the Nymphs in
Phyle (Attica), many people, including slaves and a
cook, arrive laden with food, both to sacrifice to the
god and to laze away the afternoon feasting; as the
hero of the play says, "Look at these thieves at their
sacrifice! They bring in beds, wine by the cask-for
their own pleasure, not the gods"' (lines 447-49;
Vellacott 1967: 78).
The range of wares and forms in the sanctuary's
Hellenistic-period assemblage is parochial; 84% of
the vessels are of "local" manufacture,
with fabrics
that derive either from the northern
Golan Heights,
the slopes of Mt. Hermon,
or the Hula Valley (fig. 6).
Foremost among the local vessels are those made
of spatter painted ware, which were produced in
the Hula Valley (Berlin 1997a: 7-9). Imports con-
sist almost exclusively of Phoenician semifine, BSP
ESA vessels, which
were manufactured
on the Phoenician (and Syrian?)
coast (Berlin 1997a: 9-10; 1997b: 77-78; Slane
1997: 269-71). The only imported
vessels from out-
side the Levant are two Greek island amphoras.
Whether locally manufactured
or imported, exam-
ples of all forms have been found at nearby sites,
an indication that they were readily available from
area markets or peddlers. The preponderance
of lo-
cal, or locally available, vessels is not surprising,
especially at a site so removed from major popula-
tion centers. At other sites in this region, however,
more "exotic" ceramic offerings have been found,
including among other items mold-made bowls, Ital-
ian (Campanian)
table vessels, and painted lagynoi
probably made in Alexandria (Cornell 1997; Slane
1997: 347-48; Berlin 1997a: 42-45). The absence
of such finer
wares at the sanctuary suggests not only
that this period's
dedicants came from the immediate
environs, but also that they were neither well off nor
generous. The Hellenistic sanctuary
may thus be in-
ferred to have been a local and fairly poor cult place.
Early Roman Period (Late First Century B.C.
to Late First Century A.D.)
Towards the end of the first century B.C. two
events occurred that had a significant impact on the
The first was the decision by King Herod
the Great
in 19 B.C. to dedicate a temple to Augustus
at the cult site itself (Josephus Ant. 15.363-64). In
so doing, Herod introduced
not only another
to the sanctuary, but also a certain amount of offi-
cial attention, money, and status. Shortly thereaf-
ter, in 2 B.C.,
Herod Philip, the youngest of Herod's
sons and successors, chose the area below the Pan-
ion springs for the location of his new capital city,
Caesarea Philippi. With this event, the sanctuary
was transformed from a rural to an urban cult, its
prestige and reputation
connected to that of the new
capital city.8
None of the buildings dated to this period (the
Augusteion, the Court of Pan and the Nymphs, and
the Temple of Zeus and Pan) contained pottery that
may be associated with their use. Rather, as with
the Hellenistic ceramic assemblage, that of the early
Roman period is unstratified
and very fragmentary.
Almost all of this pottery was found in soil pockets
in and around the bedrock boulders of the terrace
edge, and it probably
comprises debris cleared from
the cave and/or the terrace.
The quantity and ratio of types within the early
Roman-period assemblage reveals changes in ritual
at the sanctuary.
The most notable difference is the
increase in lamp dedications (fig. 7). Lamps
1999 31
I .1
0 3 6cm
v, /1 9
I, 10
Fig. 5. Hellenistic
table and cooking vessels found at the Sanctuary
of Pan.
1. Bowl, incurved
Single small rim
Spatter painted ware (5YR 7/8), light gray core. Slip fired brownish
orange out
(2.5YR 5/8), dull gray in (5YR 4/1). Berlin
1997a, PW 135, pl. 16.
2. Bowl, incurved
Single small rim
Spatter painted
ware (7.5YR 7/6), fully
fired. Matte brownish
orange slip (2.5YR
5/8) on upper
wall in and out. Berlin
1997a: 72-75, PW 138, pi. 16.
3. Bowl, everted rim
Two joining fragments preserve small section of rim
and upper
wall. Spatter painted
ware (10YR 7/6), fully
fired. Thin, matte gray to brown-gray slip (10YR 4/1-2) in and out. Berlin
1997a: 72-75, PW 141, 142, pls.
16, 76.
4. Bowl base, stamped (501A/5108).
Single small
fragment. Spatter painted
ware (7.5YR 7/4), fully
fired. Polished slip fired
pale pink
out (5YR 7/4),
'^-1- 4
I 11
i . . _ , I _ ,
32 BASOR 315
are, of course, not unusual in sanctuary
(e.g., Broneer 1977; Edlund-Berry
1989-1990: 336).
Nevertheless, during the entire Hellenistic period
few lamps-a total of seven-were ded-
icated at the Panion (fig. 4). During the early Roman
period, on the other hand, lamps were the most pop-
ular category of ceramic offering. Fragments
of 141
lamps have been found, of which over half are im-
ported discus lamps, of Italian, Cypriot, and Syrian
manufacture. Lamps are an easier type of dedica-
tion than a cooked offering. The startling rise in the
number of lamps reflects an increase in brief visits
and more casual gifts. This circumstance may be a
result of the sanctuary's
new status as an urban cult.
In the previous period, when the site was a simple
rural shrine, dedicants traveled some distance to
visit, and would thus have more readily chosen to
linger. Now public buildings replaced the forest just
below the spring,
and a beautiful stone building stood
before the grotto. Some of this period's dedicants
were surely residents of the new city, perhaps stop-
ping at the sanctuary
while in the center of town, and
giving the token offering of a lamp to either Augus-
tus or Pan.
A large number of early Roman-period cooking
vessels were also recovered, indicating that, concur-
rent with the increase in simple lamp dedications,
cooked meals continued to be made as well (figs. 7,
8.4-7). As with the Hellenistic cooking pots and
casseroles, many of these show signs of use, whether
from preparation of a sacrificial offering or of an
The plates, dishes, and bowls that
have been found (fig. 8.1-3) could have served sev-
eral similar functions: for individuals' drinking and
dining while visiting the sanctuary,
for the more el-
egant presentation
of a cooked offering, or as dedi-
cations in and of themselves. The continuation of
dining within the setting of the sanctuary is con-
sistent with Phoenician cult practices elsewhere; for
example, there were small dining rooms within the
second century B.C. complex of the Sanctuary
of the
Syrian Gods on Delos (Starcky 1949: 62-67).
Many more of this period's vessels are imports:
69 percent of the early Roman-period assemblage
was made outside the Golan Heights-Mt. Hermon-
Hula Valley area (fig. 6). Compared
to the late Hel-
lenistic assemblage, of which only 16 percent was
imported, such a figure might support the sugges-
tion that a real shift in the character of the cult and
its dedicants occurred at this time. Some of the im-
ported table wares do come from quite far afield,
such as a single Pergamene
dish, several Italian sig-
illata plates and bowls, and two Cypriot sigillata
vessels; these certainly represent
more generous ben-
efactions than any given previously. The two most
common imports, however, are ESA and Kfar Ha-
nanya ware, which are also common at other sites
in this region (e.g., Tel Anafa, Caesarea Philippi),
Fig. 5. Cont.
orange in (2.5YR 6/8). Palmette
stamp and three parallel
lines of incised dots on inside floor. Berlin
1997a: 75-76,
PW 148, pls. 16, 76.
5. Bowl base, stamped (501/5123/3).
Single small fragment. Spatter
ware, burned
pale grayish
white. Pinkish
orange slip (2.5YR
5/8) in and out.
Single crude palmette stamp on inside floor. Berlin
1997a: 75-76, PW 147, pls. 16, 76.
6. Bowl base, stamped (529/5358).
Single small fragment.
Spatter painted ware (10YR 8/6), fully
fired. Patchy thin orangish brown
slip (5YR 6/8) in
and out. Single palmette
stamp on inside floor. Berlin
1997a: 75-76, PW 146, pls. 16, 76.
7. Necked pointed rim
cooking pot (328/3342/4).
Single small rim
and shoulder fragment
with one handle. Spatter painted ware (5YR 7/6), fully
fired. Thin matte
slip (5 YR 4/2) on rim
and neck. Berlin
1997a: 84-87, 88-89, PW 187-190, pls. 21, 78.
8. Necked triangular
cooking pot (680/7197/45).
25 joining
fragments preserve most of upper half of vessel, both handles. Sandy fabric (2.5YR 6/8), many small
and medium,
white and angular
black inclusions,
fired. Berlin
1997a: 84-87, 90-91, PW
207, pls. 24,
9. Squared lip casserole (328/3388).
Single fragment
and wall fragment
with one handle. Hard,
coarse pinkish-brown
fabric (2.5YR 6/6), wide light
gray core. Burned on exterior rim
and handle. Berlin
1997a: 94-97, 99-100, PW 248, 253, pls. 29, 80.
10. Angled rim
casserole (291/2801/13).
Two joining fragments preserve small piece of rim
and wall. Sandy bright
orange brown
(2.5YR 6/8), many
fine and small rounded
gray and white, medium
angular gray, some medium
and large angular
white inclusions,
many small and medium
voids, narrow dark
gray core. Berlin
1997a: 94-98, PW 240, pls. 28, 80.
11. Angled rim
casserole (328/3337/1).
Single fragment
and wall
Sandy bright
orange brown
5/8), many
fine and small rounded
white and black, some medium
white inclusions, wide light gray core. Berlin
1997a: 94-98, PW 234, pls.
28, 80.
1999 33
Hellenistic Early Roman Middle
Roman Late Roman
Local I Imported
Fig. 6. Ratio of local to imported
diagnostic ceramic frag-
ments and vessels found
at the Sanctuary
of Pan, by period.
and so must have been in regular supply (Slane
1997: 272; Berlin 1997a: 14-15, 30-32; Adan-Baye-
witz 1993: 215-19; S. Israeli, personal communi-
cation). This preponderance
of imported table and
cooking wares probably reflects a diminished local
ceramic industry, and concomitant dependence on
outside manufacturers.
It is this circumstance,
than an upswing in fortune specific to the sanctuary,
that best accounts for the high number
of early Ro-
man-period imports. The Panion remained a rela-
tively simple shrine in this period, similar to the
small sanctuaries found on Mt. Her-
mon (Dar 1993). The few physical accommodations
and the continuation of modest, individual rituals
stand in contrast to contemporary sanctuaries east
and south in the Hauran, such as the theater and
temple complex dedicated to Ba'al-shamin at Seeia
(Si'), the theater and temple complex at Sahr, and
a similar complex at Sur (Butler 1915: 373, 379,
fig. 324 [Si'], 429-30, ill. 371 [Sur],
442-45 [Sahr]).
Middle Roman Period (End of the First to
Second Century A.D.)
By the middle Roman period, the Sanctuary of
Pan had become a religious and cultural fixture for
the population of Caesarea Philippi and its vicin-
ity. The Augusteion and the Court of Pan and the
Nymphs, with its small artificial cave, filled the ter-
race's western side, their architectural
Lamps Table vessels Cooking vessels Storage vessels
Fig. 7. Total number of diagnostic early Roman ceramic
found at the Sanctuary
of Pan, by function.
providing physical evidence of the cult's wealth and
status. At the end of the first century, probably con-
current with the city's centennial, the construction
of the Temple of Zeus and Pan began (MaCoz
1-2). Probably after this temple was completed
during the second century,
several dedicatory
were carved into the western cliff face (behind the
Court of Pan and the Nymphs), and the Nemesis
Court was laid out east of the new temple. These
structures provided both the opportunity and the
incentive for displays of largesse, reflected in the
dedications of imported marble sculptures made
during these years. These include a Nemesis and
a Roma, both colossal; seven life-sized statues of
nymphs and deities, including Asclepios and Apollo;
and smaller statues of Artemis and Hermes (Fried-
land 1997: 272). The buildings, the niches, and the
all provide evidence for steady and gener-
ous patronage
of the sanctuary
during this period. It
is therefore surprising to find remarkably few ce-
ramic remains (fig. 3).
Stratigraphic and historical circumstances pro-
vide a partial
explanation. First, the Temple of Zeus
and Pan and the Court of Nemesis were both con-
structed directly on the surface of the terrace (even
incorporating particularly large bedrock boulders
within their foundations or walls), and situated so
that their front ends ran up to and parallel with the
terrace edge. One small, stratigraphically
intact soil
deposit was found within the pronaos of the Temple
r-v-z^," I 2 /VW it
,, 3
2 J"~~s''
0 3 6 cm
Fig. 8. Early
table and cooking vessels found at the Sanctuary
of Pan.
1. ESA plate (529/5358).
Single rim and wall fragment.
Slane 1997: 307-8, TA Type 23, FW 175, pi. 16; Hayes 1985, form 34 var, pl. 5.7.
2. Italian
plate (289/2779/10).
Single small rim
and wall fragment.
et al. 1990, form
20.4.4, pl. 18.
3. Cypriot sigillata
joining fragments
preserve small section of rim
and wall. Hayes 1985, form P37B, pi. 20.14.
4. Cooking pot (291/2821, 2829/3).
Two joining
fragments preserve small section of rim
and neck. Kfar
ware. Thin brownish
gray slip (5YR
4/2) on exterior.
1993, form
4A, pi. 4A.
5. Cooking pot (291/2811/4, 290/2887).
Three fragments,
two joining,
preserve small section from rim
to shoulder. Sandy orange brown fabric
(5YR 6/8),
fine black and white, small angular
some small and medium rounded white inclusions,
small rounded voids. Adan-Bayewitz
1993, form
4A, pl. 4A.
6. Pan/bowl
Single fragment
from rim to beginning
of floor. Kfar
ware, fully
fired. Bottom
sooty. Adan-Bayewitz
form 1A, pi. 1A.
7. Pan/bowl
Single fragment preserving
rim to beginning
of floor. Kfar
ware, wide gray core. Dark
gray slip (5YR 3/1)
on exterior.
1993, form 1A, pl. 1A.
of Zeus and Pan; but other than this, no soil fills
lay beneath any walls or floors. Further,
period occupants of the Temple of Zeus and Pan
completely dismantled
and rearranged
the cella, thus
destroying that building's interior fills. Thus, with
the exception of the Temple pronaos fill, no pottery
can be directly associated with the sanctuary's
ond century A.D. buildings.
circumstances cannot, however, ac-
count for the overall drop in middle Roman ceram-
ics at the site. After all, with the exception of one
lamp embedded within the Augusteion side wall, no
pottery was found in or directly associated with the
Augusteion or the Court of Pan and the Nymphs
either. Yet 457 ceramic fragments
those structures were found on and below the ter-
race, but only 52 fragments datable to the middle
Roman period (fig. 3). Such a significant decline in-
dicates that small-scale worshipers
essentially aban-
doned the cult.
A breakdown of this period's ceramic finds by
function highlights a further shift in the rituals
practiced by those few worshipers. The assemblage
consists essentially of two categories: lamps and
0 .
.. 5 =J
35 -
30 -
25 -
20 -
10 -
5 -
Lamps Table Cooking Storage
vessels vessels vessels
Fig. 9. Total number of diagnostic middle Roman ceramic
found at the Sanctuary
of Pan, by function.
table vessels (fig. 9). The presence of only one cook-
ing vessel (the handle of an imported pan) reveals
that people no longer dined on site or presented
cooked offerings. The 18 dishes and bowls found
should therefore
be understood
as gifts given in and
of themselves, as were the lamps. Individual
to the sanctuary
thus made only the simplest of ded-
ications, and probably did not linger long. By the
end of the second century A.D., the sanctuary was
at its most visually impressive, the previously empty
terrace filled with buildings. Yet the number and
type of ceramic finds show that the level of individ-
ual involvement in the cult was at its lowest point.
The character of the Panion during this period
seems similar to that of the contemporary
of the Dea Syria of Hieropolis as described by Lu-
cian. He notes that "in the temple are many expen-
sive artifacts, . . . many marvelous things and statues
befitting the gods" (?10), and further that "outside
stand ... myriads of other bronze statues of kings
and priests" (?39; Attridge
and Oden 1976). In addi-
tion to the wealth of objects displayed, his descrip-
tion of the cult and its rituals
highlights the animated
of the statuary
(?10), the wild yet gentle ani-
mals within the temenos (?41), and the sacrifices of
the priests (??42-44). These last have the aspect of
high drama,
with the priest,
in purple,
ing before a very large but silent audience. What is
easily overlooked in his remarkably detailed trea-
tise, however, is that Lucian never mentions a single
rite of or dedication by an individual worshiper.
cult at Hieropolis had the character
of a show rather
than a sanctuary.
The written evidence provided by
Lucian evokes much the same picture as does the
material evidence from the Panion.
Late Roman Period (Third to Fifth Century A.D.)
By the beginning of the third century A.D., the
only remaining open space on the sanctuary
lay at the far eastern edge. The Tripartite
and the Temple of Pan and the Goats were installed
here, connected by a paved path. Both of these con-
tained an identical, and distinctive, corpus of pottery
consisting of a series of small saucer
lamps (fig. 10),
table vessels (fig. 11), serving bowls (fig. 12.4-5),
and cooking vessels (fig. 12.1-3, 6-8). All of these
vessels were made in one or both of two new wares,
one made in the area around
Philippi (Ban-
ias ware), and one made at the site of Khirbet el-
in the northern
Golan Heights (Havarit
1989: 11 [English],
29-30 [site 1.24, Hebrew],
pl. 13 [titles are on p. 244]).9 A large deposit of this
pottery was found below the floor of the Tripartite
Building along with a coin of Julia Maesa (220 A.D.),
indicating that its manufacture had begun by the
early third century A.D.
The late Roman assemblage is astonishingly dif-
ferent from preceding ones. Foremost
among the dif-
ferences is the enormous number of lamps: 2930
fragments or complete vessels (figs. 10, 13). Lamp
dedications comprise 75% of this period's
as compared to 33% of the early Roman period's
(and 62% of the middle Roman;
fig. 14). An increase
of this magnitude must reflect the introduction of
new rites requiring lamps or of a different style of
worship; the available evidence makes the latter
explanation more probable. Rites that would require
lamps, especially in such quantity,
ought to relate to
a mystery cult, or perhaps an oracular shrine. An
example where a sharp rise in the number of lamps
just that comes from the Sanctuary
of Posei-
don at Isthmia. There the exceptional abundance
first century A.D. lamps (as compared
to every other
the popularity
of the nightly myster-
ies necessary for the cult of Palaimon
(Broneer 1977:
2-3, 92). In the region around the Panion, a con-
shrine probably did exist, at the
36 BASOR 315
0 3 6 cm
Fig. 10. Late Roman saucer lamps found at the Sanctuary
of Pan.
1. Saucer lamp, inturned
Single fragment preserving complete profile, about one-third circumference. Banias ware
(5YR 7/6).
2. Saucer lamp, inturned
Banias ware? Overfired
and encrusted.
3. Saucer lamp, ledge rim
Single fragment
complete profile,
Banias ware (5YR
7/6). Burned
on rim.
4. Saucer lamp, delta rim
Two joining fragments preserve about one-quarter
rim and wall. Havarit
(5YR 6/6).
5. Saucer lamp, flanged lip (157/1369/2).
Single fragment preserving entire profile, about one-third
circumference. Banias ware (5YR
7/6). Interior
6. Saucer lamp, thickened rim
Single rim and wall fragment.
Banias ware (5YR 6/6).
second-third century
A.D. temple at Kedesh (Fischer,
Ovadiah, and Roll 1984; Ovadiah, Roll, and Fischer
1993; Magness 1990). The facade of this building
is similar in design to that of the famous shrine of
Apollo at Didyma in that the central portal is raised
and closed off by a grate or screen, making it both
impressive and inaccessible (Aviam 1985). In ad-
dition, there are two side doors flanked by niches
through which libations might be poured; and there
is a spacious forecourt
in which visitors might gather.
At the Panion, however, neither of this period's new
buildings had sufficient audience capacity or provi-
sion for oracular
dissemination. Further,
most of the
almost 3000 lamps from this period's assemblage do
not appear
to have been used at all, which would not
be the case if they had been an integral part
of some
new rite. Their condition, in fact, indicates that they
were ceremonial
rather than functional. These lamps,
like those from previous periods, therefore, most
likely represent simple individual offerings. Their
radically increased quantity thus reveals a sanctuary
far more heavily visited at this time than at any other
in its history. Such enthusiastic small-scale patron-
age is in vivid contrast to the situation in the previ-
ous century, when private dedicants had largely
the cult.
The quantity
of the late Roman-period lamp ded-
ications is not matched by their quality: most (87%)
are miniature saucer lamps (fig. 10), regular-sized
versions of which have been found in the Caesarea
Philippi excavations (S. Israeli, personal communi-
cation), at Tel Dan (Biran 1994: fig. 192.1-2, 4), on
the northern plateau of the Golan Heights at Da-
bura and Dardara
(Hartal 1989: pl. 14.2, 6, 8), on the
slopes of Mt. Hermon at Jebel Somak and Haruya
(Dar 1978: pl. 3.2-4), and at Har Senaim (Dar 1993:
pl. 1.1-9; these are misdated to the Hellenistic pe-
riod). No examples of regular size have been found
at the sanctuary and, conversely, few of the minia-
tures have been found elsewhere. Vessels of both
sizes were made of the local Banias ware, and so
were readily available; the miniatures may well
have been manufactured
specifically for the Panion.
Though finished off with a variety of rim types, these
% 0
J-- 6 9
0 3 6 cm
Fig. 11. Late Roman table vessels found at the Sanctuary
of Pan.
1. Large
carinated bowl (712/7979, 7986, 8002).
Three joining
preserve one-quarter
and wall. Havarit
ware, fully
fired. Upper
wall fired dark brown
stacking in kiln.
2. Small bowl, cut rim
preserve complete vessel. Havarit ware. Exterior wet-smoothed. Firing
crack across bottom.
3. Goblet (729/8166/3).
Six joining fragments preserve complete profile, entire bottom, most of rim and wall. Banias ware. Exterior wet-
4. Small bowl, carinated
(716/8100/2, 726/8238).
preserve complete profile,
entire bottom,
most of rim
and wall. Banias ware. Exterior wet-
5. Small bowl, carinated
(712/7979, 7986).
preserve complete profile,
half circumference
and body, entire base. Banias ware, fully
fired. Exterior
6. Small bowl, carinated
preserve complete profile,
about one-third
rim and wall, entire base. Banias
ware, fully
fired. Neat string-cut
7. Shallow phiale (726/8247).
Single fragment preserves about one-sixth circumference rim to bottom. Banias ware, fully fired. Exterior
smoothed. Finger
impressions around
8. Deep phiale (723/8176/6).
Five fragments,
preserve complete profile, one-quarter
rim to bottom.
Banias ware, fully
fired. Exterior wet-smoothed. Finger
impressions around
9. Deep phiale (699/7571/3, 7584).
fragments preserve complete vessel save for two small holes, partially
restored in plaster. Banias
ware, fully
fired. Finger
impressions around
38 BASOR 315
3 cm 0 3 6cm
Fig. 12. Late Roman
table and cooking vessels found at the Sanctuary
of Pan.
1. Cooking
pot, plain rim
(699/7612/5, 7365).
Three joining fragments
preserve about half circumference
to shoulder. Banias ware, fully
fired. Unused.
2. Cooking pot, plain rim
(712/7974/2, 7944).
joining fragments preserve complete profile,
about half circumference,
one handle, partially
restored in
plaster. Havarit
ware, light
orange slip. Two small burned areas on exterior.
3. Cooking
pot, grooved rim
Nine fragments:
six join to preserve one-quarter
rim to lower wall, one handle; one nonjoining
piece; two joining
wall pieces. Havarit
ware, fully
fired. Kiln blush over one side of rim
and wall.
4. Basin, thickened ledge rim
Single rim and wall fragment.
Hard, dense, light
(2.5YR 6/8).
5. Krater,
ledge rim
join to preserve two large nonjoining pieces: a) one-third circumference
and upper wall, one
b) one-third circumference
foot and lower wall. Banias ware, fully
fired. Exterior wet-smoothed.
6. Knobbed
lid, cut rim
Two joining
fragments preserve complete profile,
circumference. Havarit
ware, fully
fired. Burned on
and lower
outer wall.
7. Cooking
bowl, cut rim
Two joining fragments preserve complete profile,
about one-third
one handle. Havarit
ware, fully
fired. Kiln
blush on upper
8. Cooking
bowl, cut rim
(705/7820, 7822, 7833, 7919).
30 joining fragments preserve three-quarters
to beginning
of bottom. Havarit ware. Smoked on
1999 39
3000 -
2500 -
2000 -
1500 -
1000 -
500 -
0O Lampe
Table Cookng Storage
vses ve s ssels
Fig. 13. Total number of diagnostic late Roman ceramic
fragments and vessels found at the Sanctuary of Pan, by
saucer lamps were otherwise simple, open forms,
reminiscent of pre-Classical pinched lamps. The re-
maining late Roman-period lamps are locally pro-
duced derivatives of imported moldmade discus
lamps, which are marginally more decorative than
the saucer lamps; the types are known as "bilan-
ceolate" and "northern stamped" (Rosenthal and
Sivan 1978: 111-12; Sussman 1989). These "fan-
cier" varieties constitute a mere 13% of a lamp cor-
pus that, in its entirety,
is remarkably
dull. Dedicants
may have visited the sanctuary
in droves during this
period, but they largely presented only the plainest
and least expensive of offerings.
Lamps were probably
not the only vessels to have
been given as offerings. The quantities and condi-
tion of the dining, serving, and cooking vessels sug-
gest that they too were primarily dedications, as
opposed to functional objects. Practically all of the
ceramics, including the cooking vessels, appear to
be new, as if they had
just been purchased
and taken
immediately to the sanctuary.
the ratios be-
tween vessel types is peculiar: there are twice as
many cooking as serving and dining vessels in the
late Roman assemblage, compared to a one-to-one
in both the Hellenistic and the early
Roman assemblages (fig. 14). In those earlier peri-
ods, such vessels represented
the detritus of on-site
dining. The skewed proportions in this assemblage,
Fig. 14. Ratio of functions of diagnostic ceramic fragments
and vessels found at the Sanctuary
of Pan, by period.
coupled with the vessels' unused condition, indicate
that in late Roman times ritual activity was restricted
to dedications alone, as opposed to more communal
The identification of this mass of vessels as ded-
icatory rather than utilitarian sheds light on the func-
tion of the Tripartite Building. Several points are
relevant. First is the position of this building, tucked
behind but accessible from the Temple of Pan and
the Goats. Second, about half of the vessels recov-
ered inside were at least partially restorable. Third,
the vessels appeared
fresh and new, as if they were
placed here soon after they were given. The build-
ing's location, along with the quantity and condition
of its finds, combine to suggest than the Tripartite
Building was used as a repository for dedications,
in effect, a treasury (contra MaCoz 1996: 3). Offer-
ings made in the Temple could have been carted off
for storage here, since, once given over to the deity,
the vessels became sacred and as such had to be
someplace within the temenos. A similar
example of "curating" apparently quotidian objects
within a cult precinct, in this case for up to 400 years,
is attested from the hero shrine of Glaukos at Knos-
sos, Crete (Callaghan 1978).
With the exception of four jar fragments im-
ported from the Aegean, all of this period's 3919
ceramic fragments are purely local productions, ei-
ther from Banias or Khirbet el-Havarit (fig. 6). Such
complete insularity is surprising, for by the later
third and early fourth
centuries A.D.
the pottery trade
went throughout the Mediterranean area (Reynolds
1995; Peacock 1977). During the third through the
40 BASOR 315
fifth century A.D., the busiest producers and ex-
porters of ceramics were the Tunisian
which concentrated on table and cooking wares. The
best known of these is African Red Slip, which is,
in fact, a group name representative
of many sepa-
rate, but visually indistinguishable, production cen-
ters (Kenrick 1985: 341-78). Many Palestinian sites
began to receive African Red Slip wares by the
middle to late fourth century A.D., including Jerusa-
lem (Magness 1993: 181), Caesarea Maritima
1975: 39-40), Tell Keisan (Landgraf 1980), Jalame
(Johnson 1988: 145-68), and Gush Halav (Groh
1990). However, Caesarea
Philippi, the Hula Valley,
and the Golan Heights as far south as Hammat
seem to have been completely isolated. Virtually no
late Roman imported ceramics have been found in
the excavations of Caesarea
Philippi (S. Israeli, per-
sonal communication),
Tell el-Wawiyat
(A. Onn, per-
sonal communication),
or Tel Dan (Biran 1994: 233,
fig. 192; M. Hershkovitz, personal communication),
the northern
and central Golan Heights (M. Hartal,
personal communication), or Hammat Gader (Ben-
Arieh 1997: 356-59). In their dependence on local
wares, then, the sanctuary's
visitors were no different
than the rest of the immediate population. Since the
city and region were so secluded economically (and
socially?), it seems reasonable to conclude that the
Panion received few worshipers from outside this
area. The pottery
indicates that, despite its longevity,
its compelling natural surroundings,
and its aesthe-
tic improvements, in late Roman times the cult was
relevant only to local residents.
The archaeological evidence provides clear testi-
mony that the Sanctuary of Pan was abandoned.
There was no trace of destruction, even though by
late Roman times the Panion was a pagan cult site in
an increasingly Christian city. Rather, the site ap-
pears to have lain, deserted and uncared for, for a
significant period. Eventually people began to use
the terrace again; four discrete architectural
ceramic phases were excavated above the sanctu-
ruins. These later occupations within and above
buildings mean that most of the late Roman deposits
were contaminated by later material. Despite this
situation, however, ceramic evidence from both the
and parts
of the city suggests that the Pan-
ion was probably
by about the middle of the
fifth century A.D.
The most compelling piece of evidence is that the
late Roman assemblage is almost wholly comprised
of vessels made in either Banias or Havarit ware,
production of which ceased sometime in the fifth
century A.D. This alone would not indicate abandon-
ment of the sanctuary,
since nothing precluded the
acquisition and dedication of other wares. In fact,
the assemblage from the next, post-sanctuary, phase
of the terrace contains several other wares, includ-
ing later imported wares such as Cypriot Red Slip
(J. Magness, personal communication). However,
except for the occasional stray fragment, no exam-
ples of any of these later wares were found within
late Roman-period structures.
In other words, none
of these late sixth-early seventh century wares seem
to have been used during the life of the sanctuary
itself. Apparently, after the manufacture of pottery
stopped at Banias and Khirbet el-Havarit, the cult
received no other ceramic dedications. Other kinds
of dedications also ceased. The latest sculptural
ications are two miniature, free-standing fragments
(an Eros torso and the forequarters
of a bovine) that
are dated on stylistic grounds to the fourth century
A.D. (Friedland 1997: 70). An almost complete gap
in the site's numismatic record between the early
fifth and the later sixth century A.D. may be inter-
preted as indicating that no activity occurred on the
site during this period (D. Ariel, personal communi-
cation). The latest inscription from the site dates to
222 A.D. (B. Isaac, personal communication).
Architectural activity and/or maintenance is at-
tested elsewhere in and around the city of Caesarea
Philippi in the fourth century A.D., but by the mid-
dle of the fifth century large parts of the city were
deserted. This includes several villas located just
outside the city proper
(M. Hartal,
personal commu-
nication), as well as some public buildings in the
center of town (S. Israeli and V. Tzaferis, personal
communication). The most significant structure to
have been abandoned was the high-level aqueduct,
on which maintenance
clearly ceased. Excavation of
the channels found them filled with silt and pottery
(Hartal 1995; M. Hartal, personal communication).
All pottery fragments
were of vessels made either in
Banias or Havarit ware, which could only have ac-
before production
of these wares stopped.
If the aqueduct
was abandoned
by the middle of the
fifth century, the city had probably been gradually
losing population over the preceding generation.
The abandonment of the sanctuary and the de-
population of Caesarea
Philippi is of a piece with a
1999 41
general decline throughout
Upper and Lower
Galilee and the Hula Valley from the mid-fourth
the early fifth century A.D. (Adan-Bayewitz
1993: 240-43). Settlements
just outside this region,
however, flourished. At Beth Shearim, in the west-
ern Galilee, evidence exists for a wealthy domestic
and industrial settlement from the mid-fifth to the
A.D. (Vitto 1996: 137-41). At Beth
Shean, just south of the Sea of Galilee, construction
activities in the city center continued with fervor
the fifth and the sixth centuries A.D. (Foer-
ster and Tsafrir
1993: 6-7 [basilica], 22 [shops along
Valley Street],
26 [Roman
colonnade and southwest],
28 [Byzantine "commercial"
Bar-Nathan and
Mazor 1993: 38-41 [the bathhouse],
42-44 [Palladius
Street exedra and shops], 45 [the odeon]). At Hammat
just southeast of the Sea of Galilee, the baths
underwent a large reconstruction
around 450 A.D.,
after which they were active until the middle of the
eighth century (Hirschfeld 1997: 123-43, 478-79).
Sanctuaries received the offerings of various peo-
ples, from private individuals to civic officials to
rulers. Ceramics, being both easily available and in-
expensive, are the best evidence for visitors of the
first category (although others could and did give
them as well); and the ceramics found at the Sanc-
tuary of Pan reveal much about both the cult and
its worshipers.
In its earliest
phases, as a rural
visitors came from nearby
household pottery in which both meals and dedi-
cations were made. Though offerings were poor, and
the ritual casual, the visits were probably of some
duration. The quantity of Hellenistic finds from the
Panion demonstrates
that in those years the cult was
a focus of continuous individual worship. By the
early first century A.D., the sanctuary had become
both a royal and a civic cult spot. Deities and visi-
tors continued to enjoy cooked meals, however,
which is evidence that the terrace retained some of
1Z. MaCoz
directed the excavation
and is responsible
of the architectural
I was respon-
sible for field analysis
and subsequent
cation of the ceramics
the Persian
the late
remains were
in the
or after excavation:
the glass
by Y. Gorin-Rosen,
its rural aspect, and that worshipers must have felt
welcome to linger. The large number
of early Roman
ceramics shows that individuals continued to sup-
port the cult. At the same time its new status as both
a royal and a civic cult spot encouraged a few more
generous offerings, as well as faster, less onerous
By the late second century
the sanctuary
become a showcase. The terrace was adorned with
buildings, open-air courts, and carved niches, all
providing opportunity
and incentive for displays of
wealthy largesse. An abrupt
in the number
middle Roman ceramic offerings, however, reflects
the abandonment of the cult by small-scale worship-
ers. The site was by no means deserted. Rather, it
became formal and fashionable, which apparently
discouraged visits by the common people who had
long sustained it.
In the early third
the cult's
to the decline of individual ritual activity
by commissioning new buildings designed to draw
crowds. As the quantity of ceramic remains attests,
they were successful (fig. 3). This renewed vigor is
of a piece with evidence of vitality from other
shrines in Palestine at the same time (MacMullen
1981: 134). Throughout
the fourth century A.D. the
Panion was a focus of local patronage, and more
people visited than ever before. Their dedications
were, however, off-hand; worshipers
no longer took
the time to prepare
and share a meal. Active involve-
ment gave way to passing, and passive, attendance.
In the final phases of its existence, the Panion must
have had the character more of a spectacle than of
a cult, with its visitors more an audience than dedi-
cants. In the end, despite its spectacular
setting and
elaborate structures, the sanctuary was abandoned.
Some former worshipers surely converted to Chris-
tianity. Others may have turned to one of the new
"ascetic stars"
who rose to religious prominence
ing the late fourth and fifth centuries in Syria (Brown
1971a; 1971b: 80). Ironically,
the period of the sanc-
most abundant remains proved to be its last.
by G. Carver,
the sculpture
by E. Friedland
by B. Isaac,
and the
by D. Ariel.
I thank
the buildings,
B. Isaac
for sharing
his dating
tions of the
D. Ariel
for information
the coins.
42 BASOR 315
2There is some fragmentary
evidence that in the Helle-
nistic period access to the cave was eased by a path and an
improved road bed; a small structure
may have been built
at the foot; see MaCoz
1996: 4-5.
On Pan and his cult in Greece from the sixth through
the fourth century B.C. see Borgeaud 1988. The most ob-
vious similarity between Pan cults in Greece and those in
the sanctuary
discussed here is a predilection for locating
them in or near caves and scenic rural sites (Borgeaud
1988: 49-51).
4The totals indicate the number of individual, nonjoin-
ing, positively identifiable,
diagnostic fragments. Diagnos-
tic fragments (such as rim pieces) that joined to make a
single vessel were counted as one. Individual body sherds
were not counted.
5Persian period vessels have been found, but no posi-
tive evidence exists for the identification of a cult spot
during that period, assertions concerning the antiquity of
the cult or the shrine
(Wilson and Tzaferis
1998: 56, but contra Tzaferis
1992: 132*-33*). The Persian-
period vessels-cooking pots, jars, a perfume
all been found in domestic contexts elsewhere in this re-
gion, and may well represent
a temporary
6On dining in sanctuaries in general, see MacMullen
1981: 36-38.
7For Hellenistic settlement in the Golan Heights, see
Hartal 1989: 7-8 (English), 126-27 (Hebrew); for Mt.
Hermon, see Dar 1993: 17-22; for information about
settlement in the Hula Valley I am indebted to Idan
Shaked, who is conducting an archaeological survey of
this area.
8On the results of excavations in Caesarea Philippi it-
self, see Tzaferis and Israeli 1995; 1996.
9A complete description of both of these fabrics, their
chronology, and the various shapes produced
in them, will
in a chapter
I wrote for Panion I: Excavations at the
Sanctuary of Pan at Caesarea Philippi/Banias 1988-1993
in press).
Eusebius, Ecclesiastical Histories.
Josephus, Antiquities of the Jews.
Lucian, de Dea Syria.
Polybius, Histories.
Ps.-Lucian, Amores.
Adan-Bayewitz, D.
1993 Common
Pottery in Roman Galilee: A Study
Local Trade. Ramat Gan: Bar-Ilan University.
Attridge, H. W., and Oden, R. A., trans.
1976 The Syrian Goddess (De Dea Syria), At-
tributed to Lucian. Texts and Translations,
Graeco-Roman Religion Series 1. Missoula,
MT: Scholars.
Aviam, M.
1985 The Roman Temple at Kedesh in the Light of
Certain Northern Syrian City Coins. Tel Aviv
12: 212-14.
Bar-Nathan, R., and Mazor, G.
1993 The Bet She'an Excavation Project (1989-
1991): City Center (South) and Tel Iztabba
Area; Excavations of the Antiquities Authority
Expedition. Excavations and Surveys in Israel
11: 33-51.
Ben-Arieh, R.
1997 The Roman, Byzantine and Umayyad Pottery.
Pp. 347-81 in The Roman Baths of Hammat
Gader: Final Report, by Y. Hirschfeld. Jerusa-
lem: Israel Exploration Society.
Berlin, A.
1997a Tel Anafa II, i: The Hellenistic and Roman
Pottery: The Plain Wares. Journal of Roman
Archaeology Supplementary
Series 10.2. Ann
Arbor: Kelsey Museum of Archaeology.
1997b From Monarchy to Markets: The Phoenicians
in Hellenistic Palestine. Bulletin of the Ameri-
can Schools of Oriental Research 306: 75-88.
Biran, A.
1994 Biblical Dan. Jerusalem: Israel Exploration
Borgeaud, P.
1988 The Cult of Pan in Ancient Greece. Trans.
K. Atlass and J. Redfield, from French. Chi-
cago: University of Chicago.
Broneer, O.
1962 Excavations at Isthmia, 1959-1961. Hesperia
31: 1-25.
1977 Terracotta
Lamps. Isthmia 3. Princeton:
ican School of Classical Studies.
Brown, P.
1971a The World
of Late Antiquity.
London: Thames
and Hudson.
1971b The Rise and Function of the Holy Man in
Late Antiquity. Journal of Roman Studies 61:
Burkert, W.
1987 Offerings in Perspective: Surrender,
tion, Exchange. Pp. 43-50 in Gifts to the Gods:
Proceedings of the Uppsala Symposium
eds. T Linders and G. Nordquist. Boreas 15.
Stockholm: Almqvist and Wiksell.
1999 43
Butler, H. C.
1915 Ancient Architecture in Syria: Haurdn Plain
and Djebel Haurdn. Publications
of the Prince-
ton University Archaeological Expeditions to
Syria in 1904-05 and 1909, Vol. 2 A, Part 5.
Leiden: Brill.
Callaghan, P.
1978 KRS 1976: Excavations at a Shrine of Glaukos,
Knossos. Annual of the British School at Ath-
ens 73: 1-30.
Cornell, L.
1997 A Note on the Molded Bowls. Pp. 407-16 in
Anafa II, i: The Hellenistic and Roman Pot-
tery, ed. S. Herbert. Journal of Roman Archae-
ology Supplementary
Series 10.2. Ann Arbor:
Kelsey Museum of Archaeology.
Dar, S.
1978 A Survey of the Ancient Settlements of the
South-Western Flank of Mt. Hermon. Israel
Kibbutz Movement. Tel Aviv: Department of
History and Antiquities.
1993 Settlements
and Cults Sites on Mount Hermon,
Israel: Ituraean Culture in the Hellenistic and
Roman Period. BAR International Series 589.
Oxford: Tempus Reparatum.
1989- The Central Sanctuary
at Morgantine (Sicily).
1990 Scienze dell'Antichita 3-4: 327-38.
Ettlinger, E.; Hedinger, B.; Hoffmann, B.; Kenrick, P.;
Pucci, G.; Roth-Rubi, K.; Schneider, G.; von
Schnurbein, S.; Wells, C.; and Zabehlicky-
Scheffenegger, S.
1990 Conspectus Formarum Terrae Sigillatae Ital-
ico Modo Confectae. Materialien zur Romisch-
Keramik 10. Bonn: Habelt.
Fischer, M.; Ovadiah, A.; and Roll, I.
1984 The Roman Temple at Kedesh, Upper Galilee:
A Preliminary
Study. Tel Aviv 11: 146-72.
Foerster, G., and Tsafrir,
1993 The Bet She'an Excavation Project (1989-
1991): City Center (North);
Excavations of the
Excavations and
Surveys in Israel 11: 3-32.
Friedland, E.
1997 Roman
Marble Sculpture
from the Levant: The
Group from the Sanctuary of Pan at Caesarea
Philippi (Panias). Ann Arbor: UMI.
Groh, D.
Hartal, M.
Judaism in Upper Galilee at the End of An-
tiquity: Excavations at Gush Halav and En-
Studia patristica 19: 62-71.
Northern Golan Heights. The Archaeological
as a Source of Regional History. Qazrin:
Israel Department
of Antiquities and Museums
1995 Banias, Aqueduct. Excavations and Surveys in
Israel 13: 1-2.
Hayes, J.
1985 Sigillate Orientali. Pp. 1-96 in Enciclopedia
dell'arte antica. Classica e Orientale. Atlante
delle Forme Ceramiche. Vol. 2, ceramica fine
romana nel bacino mediterranean
(tardo elle-
nismo e primo impero), ed. G. P. Carratelli.
Rome: Enciclopedia Italiana.
Herbert, S.
1994 Tel
Anafa I: Final Report on Ten Years
of Ex-
cavation at a Hellenistic and Roman Settlement
in Northern Israel. Journal of Roman Archae-
ology Supplementary
Series 10.1. Ann Arbor:
Kelsey Museum of Archaeology.
Hirschfeld, Y.
1997 The Roman Baths of Hammat Gader: Final
Report. Jerusalem: Israel Exploration Society.
Johnson, B.
1988 The Pottery. Pp. 137-226 in Excavations at
Jalame: Site of a Glass Factory in Late Roman
Palestine, ed. G. Weinberg. Columbia, MO:
University of Missouri.
Kenrick, P. M.
1985 Excavations at Sidi Khrebish Benghazi (Bere-
nice). Vol. 3, Part 1: The Fine Pottery. Supple-
ments to Libya Antiqua 5. Tripoli: Department
of Antiquities, Secretariat
of Education.
Landgraf, J.
1980 La ceramique Byzantine. Pp. 51-99 in Tell
san (1971-1976), une cite phenicienne en Ga-
lilee, eds. J. Briend and J.-B. Humbert. Orbis
Biblicus et Orientalis, Series Archaeologica 1.
Editions universitaires.
MacMullen, R.
1981 Paganism in the Roman Empire. New Haven:
Yale University.
Magness, J.
1990 Some Observations on the Roman Temple at
Kedesh. Israel Exploration
40: 173-81.
1993 Jerusalem Ceramic Chronology: Circa 200-
800 C.E. JSOT/ASOR Monograph Series 9.
Sheffield: JSOT
1993 Banias. Pp. 136-43 in The New Encyclopedia
of Archaeological Excavations in the Holy
Land, Vol. 1, ed. E. Stem. New York: Simon
and Schuster.
1995 Banias, Temple of Pan-1991/1992. Excava-
tions and Surveys in Israel 13: 2-7.
1996 Banias, Temple of Pan-1993. Excavations
and Surveys in Israel 15: 1-5.
In press Panion I: Excavations at the Sanctuary of Pan
at Caesarea Philippi/Banias 1988-1993. Israel
Antiquities Reports.
44 BASOR 315
Morel, J.-P.
1989- Aspects ltconomiques d'un sanctuaire (Fondo
1990 Ruozzo 'a
Teano, Campanie). Scienze dellAn-
tichita 3-4: 507-17.
Ovadiah, A.; Roll, I.; and Fischer, M.
1993 The Roman Temple at Kedesh in Upper Gali-
lee: A Response. Israel Exploration Journal
43: 60-63.
Peacock, D. P. S., ed.
1977 Pottery and Early Commerce: Characteriza-
tion and Trade
in Roman and Later Ceramics.
London: Academic.
Reynolds, P.
1995 Trade
in the Western
A.D. 400-
700: The Ceramic Evidence. BAR Interna-
tional Series 604. Oxford: Tempus Reparatum.
Riley, J. A.
1975 The Pottery from the First Session of Excava-
tion in the Caesarea Hippodrome. Bulletin of
the American Schools of Oriental Research
218: 25-63.
Roller, D. W.
1998 The Building Program of Herod the Great.
Berkeley: University of California.
Rosenthal, R., and Sivan, R.
1978 Ancient Lamps in the Schloessinger Collection.
Qedem 8. Jerusalem:
of Archaeology,
Hebrew University of Jerusalem.
Slane, K. W.
1997 Tel Anafa II, i: The Hellenistic and Roman
Pottery: The Fine Wares. Journal of Roman
Archaeology Supplementary
Series 10.2. Ann
Arbor: Kelsey Museum of Archaeology.
Smith, J. Z.
1982 Imagining Religion: From Babylon to Jones-
town. Chicago: University of Chicago.
1987 To Take Place: Toward
Theory in Ritual. Chi-
cago: University of Chicago.
Snodgrass, A.
1989- The Economics of Dedication at Greek Sanc-
1990 tuaries. Scienze dellAntichita 3-4: 287-94.
Starcky, J.
1949 Autour d'une dedicace palmyr6nienne
a Sadrafa
et a Ducanat. Syria 26: 43-85.
Sussman, V.
1989 Northern
Oil Lamps and Their Typol-
ogy. Michmanim 4: 23-58.
Tomlinson, R. A.
1980 Two Notes on possible Hestiatoria. Annual of
the British School at Athens 75: 221-28.
Tzaferis, V.
1992 The "God Who is in Dan" and the Cult of Pan
at Banias in the Hellenistic and Roman Peri-
ods. Eretz-lsrael 23 (Avraham
Biran volume):
Tzaferis, V., and Israeli, S.
1995 Banias-1992. Excavations and Surveys in
Israel 14: 1-3.
1996 Banias-1993. Excavations and Surveys in
Israel 15: 5-7.
Vellacott, P., trans.
1967 Menander: Plays and Fragments. Theophras-
tus: The Characters. Baltimore: Penguin.
Vitto, F
1996 Byzantine Mosaics at Bet Shecarim.
New Evi-
dence for the History of the Site. CAtiqot
Wilson, J., and Tzaferis, V.
1998 Banias Dig Reveals King's Palace. Biblical
Archaeology Review 24, no. 1: 54-61, 85.
1999 45
... 211-45). A good example of reconstructing cults from pottery found in a cultic compound in the vicinity of Omrit is in the seminal paper by Berlin (1999) on the archaeology of ritual in the Pan sanctuary at Banias. ...
Full-text available
Excavations of the Roman temple at Horvat Omrit, situated in the foothills of Mount Hermon and the Golan, yielded terracotta figurines dated from the first century BC—first century AD. Some 100 fragments of figurines portray young children standing with arms lifted up from the sides and bent at the elbow, palm turned outward. Although this group is unique in its iconography, it fits in with nearby temples in Phoenicia, where numerous figurines and statues of children were consecrated. Images of children from temples around the Mediterranean are often associated with healing cults and rites of passage. The child figurines from Omrit are examined with regard to their gesture, age, and gender, in order to reconstruct the likely cult that took place in the temple. The picture emerging from the terracottas is of family rites celebrating a crucial threshold in life, when passing from infancy to childhood at around the age of three. This is a vulnerable stage in childhood, since mortality rate among young children was very high in ancient societies, and rites were performed to protect them. These rites have further significance in terms of socialization, in introducing the infant to the family, to the cult, and to society in general.
... The main methodology for ceramic analysis at the site relies on functional division of the different vessel types. Three main functional groups were defined: storage vessels, cooking vessels and tableware (for a similar method, see Berlin, 1999). Cooking vessels include open and closed cooking pots, casseroles and cooking jugs. ...
Full-text available
The Te’omim cave is situated on the western edge of the Jerusalem Hills. The cave consists of the main hall (approximately 50×70 meters with a maximum height of about 10 meters). Several passages lead from the main hall to other small chambers and systems of crevices. The cave detailed description is presented. The Te’omim cave has been researched by archaeologists since the late 19th century till now. Modern excavations discovered multiple artifacts of various periods of time, including Neolith, Bronze and Iron Ages, Roman and Byzantine periods and etc. Recent finds are described in the paper. In the antique times calcite deposits in the cave main hall were quarried to produce “calcite alabaster”. This is the first “alabaster” quarry known in the southern Levant. Quarry faces, toolmarks and unfinished blocks remained, and they are described in the paper.
Intentionally broken “picture” lamps, or Bildlampen , are relatively common at archaeological sites throughout the Roman world. Such lamps typically exhibit a missing central discus. The discus itself – called a lamp “medallion” – often survives, too, and represents further evidence for deliberate lamp breakage. This article explores picture lamps with missing discuses and lamp medallions as a distinct and identifiable artifact group. It also surveys the possible reasons behind their intentional breaking. The work additionally identifies selected findspots where the lighting vessels were broken in rituals, with a special focus on the Shrine of Apollo at Tyre, and examines whether lamp breakage reflects individual choice or collective behavior. In an effort to understand how Roman picture lamps were deliberately broken and the lamp medallions generated for rituals, breakage experiments – drop, impact, puncture, and hammerstone – were conducted on accurate museum-made replicas of Roman picture lamps.
Full-text available
During recent excavations along Antiochia Hippos saddle, a monumental gate dating to the Roman period was unearthed. The monumental gate (propylaeum) served as an entrance to a large extra muros compound built along the western slopes of the saddle. So far, besides the gate area itself, a public bathhouse and a theater were partially exposed in the compound. Since the perfectly preserved bronze mask depicting Pan was unearthed in the propylaeum, and other features were also located within the saddle compound, it is very likely that it served as a sanctuary. The propylaeum, which was the main area of excavations within the compound, is the focus of the present paper.
Full-text available
The Social Archaeology of the Levant - edited by Assaf Yasur-Landau December 2018
Full-text available
The aim of the paper is to explore the way the divine agents were made materially manifest in objects, structures and practices, and what people perceived as a divinity in the rock sanctuary at Caesarea Philippi. How could people interact with the gods to match their demands of divine assistance and protection? How did they imagine the divine at the rock-face? Who came as visitors, and what, then, did the various appearances at the rock and spring sanctuary mean to a citizen of Caesarea or a traveller from a distant place? The agenda is to overcome the prevailing interpretations of features and findings from Caesarea Philippi that proceed through hierarchically ordered material categories. I intend to grasp the significance of the deities’ different appearances, by referring to locations, addressees and recipients. The appearances represent the perceivable net of relations in which the (image-)object takes part as materialisation of the divine, thereby sacralising the place. This perspective is informed by the approach of “Lived Ancient Religion”.
Whether at Panhellenic sanctuaries — Olympia, Delphi, Delos — or at city and ethnos sanctuaries — Lindos, Perachora, Isthmia, the Argive and Samian Heraea, the Athenian Acropolis, Pherae — there is a very marked preponderance in the number of small dedications of geometric and archaic date, by comparison with those of later times. This profligacy of dedication was not a practice of time-honoured antiquity: in most cases, it had itself only come into being in the course of the eighth century BC. We are thus dealing initially with an episode that is roughly co-terminous with the archaic period of Greece. We must immediately concede that it is a phenomenon which relates entirely to preserved dedications. Parallel patterns of dedication, with a climax in the seventh and sixth centuries BC and a decline thereafter, can be observed in other categories of dedication, such as metal armour and weapons. At Olympia, for example, we have also a rich series of helmets which have been well studied.
The site lies on the sloping ground east of the Acropolis, about 300 m. west of the central court of the Minoan palace and 200 m. south of the Unexplored Mansion. In Classical antiquity it must have been on the southern edge of the built-up city; more recently a small olive grove stood there, but this has been increasingly encroached upon by the southward extension of Knossos village. In 1974 house-building operations on the plot of Mr. C. Vlachakis revealed traces of ancient habitation and a cache of Roman sculpture. In the summer of 1976 the sinking of a bothros behind the new Vlachakis house brought to light a stretch of ancient walling and Hellenistic sherds in some quantity. A new excavation was undertaken by the B.S.A. at the request of Dr. St. Alexiou, Ephor of Antiquities in Iraklion and work continued for approximately six weeks. The trenches were then filled in and, later, covered over by a concrete terrace. The present paper contains a detailed study of all the stratigraphy and architectural phases together with a description and interpretation of the Classical and Hellenistic assemblage and occupation. Future papers will deal with the terracottas of this phase and with the earlier and later material.
Tel Anafa II, i: The Hellenistic and Roman Pottery: The Plain Wares
  • A Berlin
Berlin, A. 1997a Tel Anafa II, i: The Hellenistic and Roman Pottery: The Plain Wares. Journal of Roman Archaeology Supplementary Series 10.2. Ann Arbor: Kelsey Museum of Archaeology.
The World of Late Antiquity
  • P Brown
Brown, P. 1971a The World of Late Antiquity. London: Thames and Hudson.