Content uploaded by Peter Fabian
Author content
All content in this area was uploaded by Peter Fabian on May 08, 2016
Content may be subject to copyright.
340
Y,
Goren
and
P. Fabian
Elu:d
Ruheibq o
Nirsqno
Fig.
L Map of the
Roman-Byzantine
cities
of the Negev
desert.
Fig.
2
unpublished photograph
of the
oboda potter's
workshop
during
excavation,
view
to
the southwest
(courtesy
A. Negev).
3,11
The
Oboda
potter's
workshop reconsidered
Yuval
Goren and
Peter Fabian
History of
i€search
Study of the
Roman-Byzantine cities of the Negev desert
has been a focus of archaeological
research
for
nearly a century.
While much of the early
research has focltsed
on
architecture,
agricultural installations,
water-supply,
and lrade
ror.rtes,
the
typology,
lechnology,
and dis-
tribution of the pottery
vessels
found in
these
cities were relatively neglected prior
to
the
sun eys and excavations by
J.
H. Iliffe and N. Glueck.t
The
latter's
work
drew attention to the
elaborated Nabatean
Painted
Fine Ware
(NPFW),
found
also on several sites in
Jordan,
inch.rding Petra. An important
landmark in the study of NPFW occurred nearly
a
half-century
ago with ihe discovery of a potter's workshop at Oboda
(Avdat)
by A. Negev. His narrative is
important for our discussion:
Given
the
enormous
quantities of Nabatean painted
$'are,
I had little doubt that such an installation
existed on the site.
To
this search I devoted much of the
winter
and spring of 1959, but to no avail
...
Then came a stloke of luck. For some reason or other, one day I came to
Oboda
much earlier than
usual, an hour or so before sunrise. It was a
pleasant
early
June
morning and,
with
time on
my
hands,l again
$'ent to
the eastem outskirts of the
town in
search of
the potter's workshop.
Sudden
ly,
when
the first
rays
of the rising sun glinted off the ground, my eye rvas siruck by a segmeni
of a
small circle made of field
stones
cove.ed with
Brey-brown
moss. With my
small
pick
I could easily
make out a
circle
some two meters in
diameter. Cleaning
off the shallow topsoil produced
thick
black
jshe'.
Frrrekal
I
had dscovered
a potter's kiln.2
The
excavation of the
workshop
yielded enormous quantities of pottery. Dating the
workshop
was
based on the assemblage of 49 coins found
in it, which included
two Hellenistic coins,
16
Nabataean
(of which
12
were
firmly attributed
to
Aretas IV),
one
coin
of the
late lst-early 2nd
c. A.D., and two Late
Roman
(3rd-4th
c.).
With no obvious stratification, the
dating
of the
entire installation to the
1st
c. A.D.
was
based on the br.rlk of
the
identifiable coins. Other
evidence
for
the dating came from oil lamps, terra sigillata
(including
significant amounts
of
Easlern
Sigillata
A
[ESA]
vessels),
and other
wares. Due
to
its abundance in
the
workshop,
Negev
suggesteds that the
ESA
too was produced
locally. Recognizing
the
importance
of
this
discovery, Negev
dedicated
his
doctoral dissertation to the chronology
of
NPFW, under the
supervision of M. Avi-Yonah.{
Negev's
idea
that NPFW, as
well
as
the
ESA
pottery
found, were made locally in lhis
workshop,
and his
dating,
were
adopted uncritically at the time.
However,
the discovery
in
1979
of pottery
kilns
at az-Zurrabah near Petra
added another dimension to the understanding
of NPFW. Excavations by the Department
of
Antiquities
of
Jordan
led
to the unearthing
of
5
pottery
kilis datirg to the 1st through 6th
c.
A.D., as well as parts of a large workshop
dated
to
the 4tir, Seven other
kilns
and parts
of several
workshops were discovered dr.lring
later
construction
projects
in the
vicinity,
making the industrial area at az-Zrrtrabah
one of the
larg-
est and the
longest-operating workshop
zones known in the
southern
Levant.5
A research project in the 1980s
using instrumental Neutron Activation Analysis
(NAA)
was
set up to provide additional information
on the provenance of the NPFW and ESA from Israel
and
Jordan.6
The research focused
on these wares as well as
on clays and sr,rpposed "unfired
A. Negev, The Nnbnfaenn pottet's
uarkshop at Oboda
(Bonn
1974) 48 nn.2
3;
J.
H.lliffe, Nabat..nll
pattery
fronr
the Negeo
(QDAP
lll, 1933)
132-35;
N. Glueck, Erploratlon n eastern Palestire I
(AASOR
1934)
1
ff.
A. Negev, T/re fafe Hellenisttc
and Early Roman pottery ofNnbntlean
Oboda:
final
rcport
(Qedem
22,1986)
xiii-xiv.
Negev
(supra
n-l).
A.
Negev, I/le Nab'?fean
painted
pottcry of Oboda and tlre chronology
0f
the
Nabntean
piinted p\tterV
gl\.D.
diss., Hebrew
Univ. of
Jerusalem)
[Hebrew,
English abstract].
K.
cAmr,
ADAI
35
(1991)
313-21, id. and A.
Momani,
AD,4/
43
(1999)
175-94.
J.
Gunneweg, I. Perlhan and
J.
Yellin,
T/re pra@nieflce, tltpology €r chronolag!
of Eastern Terftt
Sigilhtn
342
Y.
Goren and
P. Fabian
ESA" vessels
from
lhe potler's
workshop
at Oboda, submitted
for
NAA at both
the
Archaeo-
metry
Unit
of the Hebrew University
of
Jerusalem
and at the Lawrence Berkeley Laboratory.
While
conclusive evidence
for
the marufacture of the
NPFW repertoire
at
Petra was
provided
by chemical analysis, NAA showed that the chemical composition of the NPFW
vessels found
at Oboda natched that of the kiln wasters from
Petra
but not the Iocal clavs
at Oboda.
Fr.rrther,
the majority of the
ESA
vessels found in and around the Oboda
worl-shop
had chemr
cal compositions
in
agreement
with
the reference materials
from E
Cyprus,
while
the
minority
matched
Anatolian
reference groups.
J.
Cunneweg and his colleagues
concluded that none of the
NPFW and ESA
vessels
found at
Oboda
was
made
locally,
and that the vessels produced
by the
kiln were
probably of the
Byzantine
period.z That so
much
other early ESA pottery was found
within
the
kiln was
explahed by the fact
that
the kiln
stood on
a layer
of discarded pottery,
as seen by them
in
the
nearby
baulk. Although an attempt was made
by
K.
"Amr8
to
moderate
these
conclusions by suggesting that Nabataean fine
wares
could have
been
made locally in
Jor-
dan as
well as in Israel, Cunneweg was
able
to show
categorically that,
by
any statistical
approach, all the Nabatean fine ware
that
was analyzed at the two laboratories
was
imported from
somewhere
around Petra
solely,9
Petrographic
studies of the
Nabataean pottery
from Petra
only confirmed this conclusion.r0
In an attempt to rationalize the NAA results, Negevlr attempted to
suggest
an alternative
solution
to
this unexpected conundrurn. In
his
final publication
of the Late
Hellenisfic and
Early Roman poltery
assemblage from Oboda, he
wrote:
I do
not
doubt that
the clay
from
which
all ETS
I and ETS
ll
pottery was
fashioned originated
in
eastern Cyprus
(or
perhaps western Anatolia for ETS Il), but I doubt very much whether these
centers furnished the
finished products
... This
would imply
that ships
went
empty to the Palestinian
coast. As no ancieni
vessel
could have made this voyage safely
without
a proper ballast, I suggesi
that
... ihis ballasi consisted
of sacks or baskets
of eastern
Cypriote
clay,
to be dispersed overland
and there made into the much favored
eating
and ddnking vessels ... it is inconceivable that the
Nabatean painted eggshell ware could have survir.ed intact a
joumey
on camel back to
Oboda
and
other sjtes in the Negev
...
If the qualiiy of the clay
was
one of the secrets of Nabaiean painted and
plain pottery, then bringing
supplies of clay,
even from the reSion of Petra
(about
a three-day trip
from Oboda), could
not have
been a difficult and
costly matter. The fact
that
a
Nabatean
potier's
norkshop did exist at Oboda for about half a century cannot be deniedi
what
else could
it have
produced?
With
that,
the enigma of the
Oboda
portery
workshop
reached a stalemate. Although the
results
of the NAA study
have never
been serior.rsly
challenged again
(nor
could they be by
most
archaeologists),
neither was
Negev's
far-reaching hypothesis questioned
explicitly.
Thrrs, in a
recent
essay
on Nabataean minor
arts,
R. Rosenthal-Heginbottom maintains,
with
no reflection
of the debate, that NPFW
was
produced at both Petra and Oboda.l2
In the 1990s, P. Fabian, together with T.
Gini-Erichson,
conducted several seasons of excava-
tions in
Oboda
at the Roman army camp located only a few hundred
meters north of the potter's
workshop,
as
well
as in the Roman quarter lying
farther west
and
northwest
of
it.r3In the
(Qedem
17, 1983);
J,
Gunneweg,
L Perlman and F. Asaro,
"The provenience,
typology and
chronology
of
Nabataean painted fine ware,"
/,RGZM
35
(1988)
115-45.
Ibid.
K.
cAmr,
"lnstrumental
Neutron Activation
Analysis
of pottery and clay from the Zurrabah kiln com-
plex,"
ADAI 30
(1986)
319-28; id., The
pattery
frant
Petra: o Neutrcn
Actialtion
Anolysis slrrdy
(Oxford
1.987).
J.
Gunneweg,
"Nabataean
painted pottery: an answer to
published
INAA data," Acta RCRF 29
(1997)
341-48.
B.
J.
Dolinka, Touards a socio-econo ic histor!/
of
Nabataean Aila
(Aqnbo,
Iordan)
frotfi
the
1st cL'nt
r'!/ BC
through the early 2nd century AD:
ceramic eoidence
from
the Roman Aqabn Projrct
(MA
thesis, N. Caro-
lina
State Univ.
1999\
88-90, 941with references.
Supra
n.2, xix.
R. Rosenthal-Heginbottom,
"Artists
and artisans: refle.tions on Nabatean minor art
-
clay, metal, and
stone," in ead.
(ed.lThe
Nabaleons
i,i tire Negea
(Haifa
2003) 23--29..
P. Fabian,
"Evidence
of earthquake destruction in the archaeological
record
-
the case of ancient Av-
l0
tl
12
13
The
Oboda
Dotter's
workshop reconsidered
343
course
of study of the pottery
assemblage from the army
camp,
Y.
Goren made
a
petrographic
study
of 37
pottery
samples
representing
NPFW,
ESA, and plain wares,
as
well
as
some
Byzan-
tine
jars
from the Roman
quarter.
The results
obtained by
petrotraphy
(a
method
unrelated
to
NAA,
being
mineralogical in principal
rather than
geochemical) were
in full agreement
with
Lhe conclusions of Cunneweg ef ol.
(see
Appendix A
below
for
the inventory
of samples and
summary
of
results); moreover,
evcn the plain wares were
found to be made
of
raw
materials
that do not prevail in the
immediate
vicinity
of Oboda.
A situation
in which
a
highly
specialized
potter's workshop is producing
a
wide variety
of
pottery using only exotic clays,
some
imported
by
ship from overseas, others brought by camel
caravans across deserts and mountains, is unparalleled
in
any study of a
pre-industrial
ceramic
workshop known to us.
Thinking
that this anomaly deserved
further
consideration,
we con-
ducted additional
investigations
of the settirg, contents, architecture, and possible role of the
Oboda potter's workshop.
Pottery production
systems
and provenance
studies
When
discussing the
workability
of ceramic
raw materials,
one should distintuish between
the different types.ra Any ceramic material by definition
is
composed of clay that is expected
to sinter upon
firing. The
other component is
a
non-plastic one commonly termed 'temper'.1s
Ethnographic data indicates that, in
acquiring and
using raw materials,
skilled
potiers tend to
be
more
selective
in
choosing lheir clay and temper.r6
These
studies also show that
potters
located
adjacent to outcrops of high-quality
clays
tend to specialize and
will
dominate the
markets more
easily.
Expert potters
tend to
improve the choice and quality
of
their
raw
materials,
adjusting them
to the
desired
function of each type of
vessel,
A large body of ethnographic Iiterature tends to associate the choice of
raw
materials
first
and foremost
with
the availability
of clay and temper
around the production
site.
A
survey of
the ethnographic data regarding present-day
traditional
potters indicates
that the exploit-
able
territory of clay and
temper
is usually limited by the expenditure of time and the distance
to the
sources
of the raw material. Based
on nLrmerous
far-fh.rng reports
on clay exploitation by
potters,
D.
E. Arnold
concluded that in most cases the exploitable territory does not extend
beyond a range of L0 km from the potter's location.17 This model may lead us
to draw a
circle
with
a
radius of 10 km
around any
given
site in order to establish
its probable
clay and temper
resources. Taking
these
models, we
can suggest a
more precise
definition
for
the term'local'
commonly used by archaeologists
and ceramic analysts.
Yet geographical and political
contexts
in
the
ancient
Near
East
could
potentially have
extended the
range
of exploitable
territory
for
clay
beyond the radius
of
10 km.
Modern
indus-
trialized workshops,
of course, constitute a completely different case.
Specialized
workshops
often acqr"rire their clay from
far
distances,
sometimes
even from overseas. Yet in
these cases
we
are referring to fully industrialized manufacturers
spreading their commodities over
vast
areas, not
to
a modest workshop having
a single
room for
throwing
vessels,
one wheel stand,
and one kiln, serving a few thousand
potential consumers at the
most.
Thus,
while
it may
be
jus-
t.l
l5
l6
dat," in Big Cities World
Conference on Nr.httal Disasfer Mitigatiotl in cotij ,lction
@ith
llte Tenth
lnternational
Semunf on Earthq ake Prcgnostics,
Abstmcts
(Cairo
1996) 25; id.,
"Evidence
of earthquake
destruction in
the archaeological record
-
the case of ancient Avdat,"
Geological Society Atr unl Mett-
l'rg
0elusalem
1998)
21E-26E; id.,
"New
excavations in the Roman ar:rny
camp at Avdat," Proceedings af
the
tralnty set)enth Arclneolagic.tlccngress
(Bar
Ilan Univ. 2001) 78; id., Arttlot
tl
ring
the eslnblishment
of pror\ncia Arabia: tlte Roman
military camp
and its
position
t the easter defense syst?tl
of the
entpire
(Ph.D.
diss., Ben
Gurion Univ.
2005)
lHebrew;
English abstract].
D. E. Arnold, Cerarlic
theory awl ctltural p,'ocdss
(Cambridge
1985) 20.
A.
Shepard, Cernarics
for
the archaeologist
(W
ashington,
D.C.
1965)
6-51.
K.
Nicklin,
"The
location of pottery
manufacture," Man 1.4
(1979)
436,58; Arnold
(supra
n.1,4) 32-57;
P. M. Rice,
"Whiteware
pottery
production
in the Valley
oI Guatemala: specialization
and resource
utilization,"
,lFA
4
(1977)
221-33;
ead., Pottery ntnllsis, a sourccDool
(Chicago
1987\
"177-80.
l7 Arnold
(supra
n.14)
32-60.
344
Y. Goren
and
P. Fabian
W
NATURAI
ROCK
@
sroNt
ml
sRtcKs
Fig. 3. PIan
of the
potter's workshoP at Oboda
(after
Negev).
:l--F-l
0
| 2
3 4
5rn
The Oboda
potter's
workshop
reconsidered
.t1i
Fig.
4. View into
the kiln. A: the
stoke-hole;
B: the level of the innel
perforated
support,floor as discovered
during lhe
excavations
(at
bottom
of the
scale).
tified
lo refer to the
production
of NPFW
and
ESA
as a highly
specialized industry,
this
will
still fail
to explain the diversity
and remote
source of the raw materials
of
all
the plain
wares
from
the site, Such
a situation is quite
abnormal,
The
components
of the Oboda pottels
workshop
A generalized
plan
(better,
sketch)
of the
Oboda
potter's
workshop
was given
by Negev in
his various
reports
(fig.
3).18
Only a few photographs
of the workshop
(or
parts
of it) at the
time
of excavation appeared,le
and the
discussions
lack
any detailed
plan
or stratigraphic
section
of the
kiln,
the room for
the
potter'E wheel,
or the room for working
the
clay. Negev,s
rather parenthetical
comment
that "there was no
obvious
stratification in
the
potter,s
work-
shop"20 raises
the possibility
thaf
during
the excavation no
sections were recorded.
The plan
can be complemented
by the actual remains,
which
are still well preserved,
due to
their
inilus-
ion in Avdat
National Park.
The workshop
complex includes
thiee
main features
extending
over
a modest excavated
area
of 140 rn2.
These include
an enclosure room,
drv-built
of
local
fieldsfones, which
contains
a circular
basin with walls made
of hammer-dressed
stones that is
interpreted
as the room for working
the clay. To fhe east is
a
room
rvith
benches
set along the
inner
walls; in
its
center
is
a
truncated
cylindrical
installation
built of fieldstones,
c.1 m i.r
diameter
and
0.8
m
high, which
is
interpreted
as the base for
the
potter,s
wheel.
To
the east
of
the circular installation
stands
the cylindrical kiln,
still in a
very
good
siate of
preservaLion
(figs.
2
and 4) It is
built of locai
limesione
rocks
set
in
some mud
mortar. The
upper
part
was
built
above ground, while
the lower
part was
sunk, the natural
rock forming
irs iloor.
On its E
wall
is
a
large
stoke-hole (or
fire
mouth),
an
opening used for feeding
the fire.
The lo$,er
part
of the
kiln,
or
combustion-chamber,
was
lined with
crude bricks
set in mortar;
it
consisted
of 5
18
20
Negev
(supra
n.1)
12; id.
(supra
n.2)
xii; id.,
,'Oboda,"
in
E.
Stern
(ed.),
Iie
,rc?o crtcyclopeclin
of archaeo
Logical
cxcooations
in the Holy land
3
Oerusalem
1993) 1155-65.
Negev
(supra
n.1)
pl. 1-2.
Negev
(supra
n.1)
24.
.#
346
Y. Goren and P. Fabian
arches
sr.rpporting
a floor
0.20-0.30 m
thick
which was
built
of bricks. Small holes in
the
floor
conyeyed the
heat from
the combustion chamber
to the upper
space or
the oven, where the
pot-
tcrv was
supposedly placed
for firing.21
Negev's
plan
and description indicate
that the
structure was that of an
updraft kiln.
The
supports and
raised
oven
floor
are
hardly
visible today, yet,
with
the aid of the only publish-
ed pi1otograph of this part
of
the kiln
at lhe time
of the excavation,22 we
could irace the
remains
of this section and
reconstruct
its level. Surprisingly, the
elevated
floor
separating the
combusiion
charnber and the oven is
very
low,
while
the stoke-hole is
unusually big. As a
result,
the stoke-hole opens
below and above the elevated floor, which
crosses it at a point
about one-third up its heitht. Thus
the large
stoke-hole
was
opened both to the
combustion-
chamber and to the oven, permitting
the feeding of both
(fig.
4A-B). This
is a very
unusual
structure for ceramic updraft kilns,
in which the stoke-hole
usually opens only into the
combustion chambet, and the
oven
is filled
from above
with
the unfired
pottery.23
Civen the
relatively high
firing-temperature for NPFW and ESA
(estimated
at c.800-900"
C), sr.rch a
kiln is
expected to present thermodynamic problems
because of the difficulty of
proYiding
an even temperature across
the entire
imer volume. Indeed,
Negev wrote
that
There
\rere
no disfigured vessels
found in the
Oboda
potter's workshop even
after the installation
had been discovered. Probably
the reason is ihat the Nabatean potter \\'as a master
in controlling
ihe temperature of ihe firing, which
never seemed to
have
exceeded 800 to 900 degrees
centigiade.2a
However,
the absence of pottery
or
kiln-wall wasters in
the workshop is an
unlikely pheno-
menon.
Given
the
estimated firing temperat[res and the delicate nature
of the prodrlcts, each
firing
event
must have
produced at least c.10%
of
spoiled potlery,
either as overfired and
partially vitrified wasters
or
as
twisted vessels. Indeed, in
the
workshops
at
az-Zurrabah
an
enormous quantity of
wasters
in the form
of
twisted and partially vitrified
pottery was
found.2s The absence
of
wasters
at Oboda
thus
requires an explanation well
beyond the state-
ment
that the
potter was
able to avoid this
common
by-product
of the
work.
Another problem that
the
kiln presents
is the lack of any refractory lining
on
its inner
surface, needed since the regular firing
temperature of
far
abovc 750"C would have
severely
affected its exposed limestone walls. At
those tenperatures limestone decalcinates
into
lime.
An extensive use of such a kiln at higher
temperatures
for hourc
must have required lining u,ith
refractory
clays to
protecl
the
walls.
These
in turn
produce kiln-wall wasters,
which
are !rsual-
ly
discarded in a
nearby
dump; then
the imer surface of the kiln is restored
afler each firing
event. ln most
cases, updraft kilns are
constructed oI adobe bricks, fired bricks
or sandstone,
rather than limestone.26 lndeed,
the kilns
al az-Zunabah
(which
were
rectangular in
shape,
not rounded as aL
Oboda)
were
built of mud
bricks
wilhout
any stone elements,
and contained
high
amounts of kiln wall
and
potfery
wasters.2T
Such
wasters
also
prevail
in high quantities
in
the Byzantine
workshops
of the
NW
Negev,26 including the one found at Elusa
and examined
by trs.2e Again, this deficiency is anomalous
for a
fully-functional,
specializcd
potter's work-
shop, and it
cannot
be
overlooked.
Negev
(supra
n.1)
11, pl. 1-2.
Negev
(supra
n.1) pl.2 fig.4.
D. Rhodes, Ki/rs: desigtr, consttuction,
{r
oper|ttotl
(Philadelphia
1968)
9-17.
Neger'
(supra
n.1)9.
cAmr
(supra
n.5); F. Zayadine,
"The
pottery kilns of Petra," in D. F. Hom6s-Fredericq
(ed.),
Potterll o d
patfcrs: plst
nnd
pteserf
(Trlbingen
1986)
185
89.
Rhodes
(supra
n.23) 13.
cAmr
(supra
n.25);
Zayadine
(supra
n.25).
L Israel,
"A
survey of ceramic rvorkshops,
Nahal
Lachish-NahaI
Besor," Hndnshot
Archarclogiat 10A
(1993)
91,-93
[Hebrerv];
id.,
"The
economy
of the
Gaza-Ashkelon
region
in the Byzantine
period in the
light
of the archaeological
survey and
excavations of the'3rd Mile
Estate' near Ashkelo
,,,
Michnfitlitl
8
(1995)
119-32
[Hebrerv].
P. Fabian and Y.
Goren,
"A
new type
ofLate Roman storagejar from
the Negev," in
J.
H. Humphrey
(ed.),
Tha
Ranlan n d Bynntine
Na..' East
3
(JRA
Suppl. 49, 2002) 115-53.
21
22
2i
21
25
26
27
28
The
Oboda
potter's workshop
reconsidered
317
ln
terms
of
its internal
construction
too,
the oboda potter's
workshop
is
somewhat
obscure.
Tradilional
pottery
workshops
throtrghout
the world
have
a relativeiy
large
built-up area,
due
to the need
for
separation
between
the
vessel
construction
area, the
space where vesiels
are
set to dry,
the
painting
or
other
decorating area, the kiln, and
a storage facility.3o
In
the
hot
and
dry climal.e
of the Negev highlands,
pottery must
be dried in the
shade of
a closed
room
in
order
to prevent cracking.
This is
especially true for very
fine vessels
having practically
no
tcmper and made
of very
plastic
clays, as are NPFW
and ESA.
Although,
as it
seems, the
,,poi-
ter's
workshop"
was
not completely
uncovered, the presence
together
of a kiln
and the produc-
lion facility
(the
unusual wheel
support) wilhin the
same
enclosule and at
an extremeiy
short
distance from
one another is unparalleled,
to the besr of
our
knowledge,
by any
specialized
workshop.
While this
situation mitht
occur
in modest household
indr.rstries, it
cannot be the
case
for
a
specialized
factory which
was
importing
its
clay by sea and
by camel
caravans and
was
producing
vast
quantities
of vessels
of
extremely high
quality.
AII of
the above considerations, let
alone the absence of any truly local wares,
indicates
that the
supposed
potter's workshop
at
Oboda could
not practically
function
as such.
Its possi-
ble function shor.rld therefore
be re-examined.
The
"Nabatean
dump" and
the date of the potter's workshop
In
the course of his excavations,
Negev excavated a
so-called Nabatean dump, situated
northeast
of
the Byzantine
citadel and
about
halfway betrveen
it and the milifaiy
camp. This
dump
yielded
hundreds of thousands
of
sherds
and
other
finds
of a rather mixed
nature.
The
nunismatic
finds from the
dr"rmp
range
between the Hellenistic period
and the
3rd c.
A.D., with
a majority of
Nabatean coins
ranging
between the
reigns
of Aretas IV and Rabel Il.r1 The
pottery
from the dump dates mainly
to
the
Hellenistic and Early Roman
periods up
to the 1st c.
B.C.,
and includes nearly 20,000
sherds
of NPFW.
Later,
a survey, accompanied
by some
soundings
in
the same dump as well as in
the oihcr
dumps covering the entire
area between the Byzantine town, thc potter's workshop
and the
Roman
military
camp,
was
conducted by P. Fabian. These revealed
an accumulaiion
of similar
finds, including
enormous quantities
of NPFW and ESA, as well
as
plain wares
dating mainly to
the lst
c.
B.C.
to
2nd
c. A.D. These
dr.rmps accumulated over the remains
of the later
(2nd-3rd
c.)
Roman
quarter,
which was
destroyed
by the earthqr.rake of 363
and covered by piles of d6bris
during construction
of the nearby Byzantine
town.32
The
coins
found
h the dumps were as fol
lows: in Dump
1 one Nabatean coin was
foundi Dump 2 contained one
coin
dated
to A.D. 43 and
another
dating to
A.D.200-240;
Dump
4 yielded
one
coin of A.D. 43, one
undated Nabatean
coin,
one
dating to A.D.209-212, two
of the
2nd-3rd
c., ard three unidentified. In
Dump
8,
24
coins
were
found, including
two of lhe 4th c. B.C.,
two
dating to 110-718.C.,7
datng to 9
B.C.-A.D. 40,
two
undated Nabataean, two of the 2nd
c. A.D., one of the 2nd-3rd c.,
one
of
383, and
7
uniden-
tified. In Dump
9, two coins were found,
of
which one dates to 9 B.C.-A.D.
,10
while
tl.re other
was
unidentified. Hence
if the logic
of dating the ddbis of the potter's
workshop
is
applied to the
dump area, due to the lack
of
any
apparent stratigraphy, both areas should be
dated to the
same
period
shce the majority of the
coins in both
belong
to the
1st
c.
B.C.-1st
c.
A.D.
An alternative interpretation
What,
then,
was
the function
of the
alleged
potter's
workshop at
Oboda? In our opinion,
when
the archaeological
data at hand is
considered
objectively, the solution
becomes simplc.
We
should begin the discussion with
one unequivocal fact: this facility had
something
to do
with firirg in a kiln. For the reasons
explained above, this kiln was not
designed to produce
ex-
30
ll
8.g., E. Hasaki, "The
ethnoarchaeological
project of the potters' quarter at
Moknine, Tunisia,
seasons
2000,2002," in
N. Kallala
(ed.\,
Africa.
Noruelle slrie des slances scientit'iques
(Tunis
2006) 127-70.
A,
Negev,
"Stories
in the history of
Avdat," in
The
Etlat Book: the cightee
h congress af ttu Israel
Etploraliott
Socieh/
(Jerusalem
1963) 128; id.
(supra
n.2) xxi.
Fabian forthcomins.
32
Y.
Goren
and
P.
Fabian
Fig.
5 A. The Roman bakery at Mampsis.
Beneral
\
rew.
....-.-..1
-'.
ri
-a
!
t-.i.i;1{t#
....,rt4::i$.
.
'''
/.;
.
'
,
-..ii,
).J
The
Oboda potter's
workshop reconsidered
349
tremely hi8h temperatures, yet its
size and careful
construction indicate
that it
was not merely
a Tabut
(a
simple household
oven)
or the like. The
other components
of the workshop
-
namely
the
basin with
walls
of
hammer-dressed
stones and
the truncated
cylindrical inltal-
lation built
of fieldstones
-
indicate
that
some
production
activity
took place
near
the kiln.
secondly, we
must
put this workshop
in
the context
of
adjacent
structures. It
was found
at the
confines
of the main built-up
area
of the Roman
town, not far from
a
large
army
camp. The rela_
tion
with
the
Roman
town and
proximity
to a camp intended
to
host
many
occupants
(though
not families
maintaining
a normal
household
life)
indicales
that the alleged potter's
work-
shop
shor.rld
have
been a Roman-period
facility thal was intended
to
sen e those inhabitants.
Based
on these hints
and available
parallels,
we
suggest that the alleged potter,s work-
shop
at Oboda
was
a
bakery of the type
referred to by
J.
T. Bakker
as
a
"mill-bakery,,.33
Huge
commercial
bakeries
of this kind have
been
found at Pompeii,
Ostia, and elsewhere in
the
Roman world.3lThe
truncated
cylindrical
installation built ol fieldstones in
the room
adjacent
to the
oven, unparalleled in
potte6' workshops,
is well known
from Roman
bakeries. It is not
intended
to hold
a
potter's wheel
but the stone mill, which was
usually operated
by a mule, a
donkey,
or by
human workers.
The
typical millstone consisted
of two parts:
an immobile,
conical
base
(nrcti)
a\d,
on top of
that, a stone shaped like an hour-glass
(cnfillus).
A
wooden
frame
over
the cafilhrs was
used for rotating
it over the sta. The
grinding
took
place
between
the two parts
that
were
at
a
very
small,
fixed
distance. Indeed, fragments
of basalt
mills oI the
same type are found
today
in
the visitor's
center at the foot
of the site, and
they could have
originated in
the bakery
under discussion.
Needless to say, a
bread oven can
easily be built of
limeslone
fieldstones with no
internal linin&
as the temperature inside
it is not
expected to
exceed
300o C. The stonelined
pit nearby might have
been used as a
silo
for
storing the grain,
Roman
bakeries were
usually mill-bakeries.
Only
in later
antiquity
can one observe a
dis
tinction
between milling
and
baking, after the
much later
introduction of water
mills. Ail of
these
bakeries have much in
common
and
the technical achievement
was quite
considerable.
Major public bakeries in
Ostia are
quite large, covering
a
floor area ranging from
640 to 1525 m2
and
containing 9 millstones
on averate. The Pompeian bakeries
are somewhal
smaller. Small
provincial
bakeries
must
have
been
modest
in size, as h the case
of Oboda. One does not need to
go too far in
the search
for
parallels.
At Mampsis
(Mamshit),
uncovered
by the
same
excavator,
a bakery was partly
unearthed, including
a room
with
a truncaled
cylindrical installation
built of fieldstones. Not far from
it a
complete
basalt
millstone was found
on the surface and
during restoration
of the site it was placed
over this cylindrical installation, while
yet other
millstones were found
in this room
(fig.5
A-B).3s
As
the
remaining
rooms
of the conplex were
not excavated
any
oven
that
might have
been erected
nearby awaits
exposure, but if we look for
parallels
of the same time and in
ihe same region
for
the
"potter's
wheel
room"
at Oboda, the
Mampsis
bakery offers the
best
instance.
We assume
then that the
Oboda "potter's
workshop" was a Roman
mill-bakery. In the
absence
of any stratigraphic section or
description of
the deposits found
during
the excavation,
it
is impossible to
judge
whether
the pottery and other finds
unearthed
in
it
comprise
iri sit
remains or belong
to
the
dumps of
refuse
that
were deposited
over
the entire
area east of the
Byzantine
city
wall
after the construction
of the
citadel, walls,
and churches. The unstratified
pottery,
coins, and other finds inside
the bakery
wiil be pari of
the
later
d6bris
found
all
around the Byzantine
city,
which
can
be used
neither for
dating the installation where it was
found
nor for
assessing the chronology
of
the NPFW. The
bakery
was probably
part
of the
Roman
quarter
and dates
roughly
to the 2nd-early 3rd c.
33
J.
T-
Bakker
(ed.).
The ,nills-bakelies
of Ostin,
description
nnd
interpretation
(Amsterdam
1999).
34 M.
Junkelmann,
Panis
militaris. Die Ernihrung des rbmischen
Soldaten odet der
Crundstolf
der Macht
(Mai.^z
!997) 110-27.
35 A. Negev, The nrchitecture
of Mampsis,
finnl
rcport
IL tlrc Lnle Ro tol1
ttnd Byzatltine periods
(Qedem
27,
1988) 80-81, plan 20, loc.
463, 109, fi+
12.226.
350
Y. Goren and P. Fabian
Conclusion
The story of the "potter's
workshop"
at Oboda
is very
telling, and
serves as
a
warning
to
field
archaeologists, pottery
specialists, and histori-
ans alike.
The quick identification
of the com-
plex in this way apparently resulted from
the
enthusiasm, resulting from
some archaeoloBical
agenda,
to discover that kind of
industry
at the
site. Over the years a growing
accumulation of
data negated
this
interpretation,
yet
it was
never
seriously challenged. Most revealing is
the
tendency to
overlook
the
straightforward results
of scientifically-based NAA
study of the
wares
which
this
workshop
supposedly
produced,
whe-
ther by avoiding their mention or by suggesting
speculaiive and implausible
explanations.
This
was
done despite the
fact
that a simple explana-
tion could readily be found, if
only
it were looked
for. The implications
can be significant in terms
of the historical and cultural interpretation
of a
site.
In
the case under discussion, the question
whether
or nol NPFW and ESA were produced
at
Oboda has significant historical and
cultural
implications
for the
nature,
function,
and ethn!
city of the Roman cities
of the
Negev.
The
finds
from this
"workshop"
were
used in al effort to es-
tablish the
chronology of NPFW and the r6le of
Oboda in
the
framework
of Nabatean trade-
routes. These matters should
be
re-evaluated.
Laboratory for Comparative Microarchaeology,
Tel Aviv
University
(Y.C.)
lsrael
Antiquities Authority
and
Ben
Gurion
University
of the
Negev
(P.F.)
z
til
3
F
U
<z
9?
az
E>
9
F
gl
F
F
.l
p
ql
5D
.!
c
E
E
o
c
i
.z
a I
I
U
.1,
t
z
3
2
il
z
i
2
2
g
z
7
g
z
3
z
z
3
2
&
I
-o
'-
!
'6
F
3
z
g
z
3
z
*'
E
in
E
I
U
z
U
J
c
f;
o
T
.i.i i
E
E
J
o
i
o
o
!
a
o
I J
J
3
J
6
E
I
€
E
U
3
cr
o
I
(,
O
t!
LU
?
2
F
.9
.t
U
;
I
a
J
F
e,
N
d E
F
d
\o
\o
Santple
T!
pe
C/nss Mntl ix
T?nt9el 1
T
nDrr
2 Tttltuta 3
Cotnrrcnls
1506
/
7A Rorr Rn1-
1'l llowl
Carinated,
red sliDDed
Chyey micaceous
Quariz
CypftLs ur Aegein
t647 /73 Rtm
Jus
Loess soil
Quartz
NW Neeev
IsotroDic
matdx
1631
/
64 Rm[
Cooking
iar
Lower Cretaceous clav
Ouartz
Feldspal
S Transiordan
L631
/52
Rom.
Bowl
ESA
CIayey micaceous
Quartz
Chert l;eldsDar
Cl.pnrs or Acsean
1157
/31
Rcmr
Rm.5
t*P
Herodian Moza fine
dolomiti(
marl
Limestone
Judean
Mts.
2560
/63
Rom. Rm.
20
Jar
Marl
(undif
ferentiated Limestone
Quartz
Undetermined
Isotropic
matrix.
De<omposed
carbonates
2560
/
63
/
1
Rorn
Rm. 20
Bowl
Loess soil
Quartz
Vegetal
malcnal
NW Negev
Isotropic
matrix
1631/25 Rorn
Cookins
irr
Lowel
Cretaceous clay
Quartz
ARF
S
Transiord,rn
1631/24 Rour.
Iux
Loess
Quartz
NW Negev
1.465
/12
Rorn Rm.
11
Cookingjar Lowcr
Cretaceous clay
Quartz
S Transjoldan
Negev
1986
D.55:399
1647 /22 Rom.
Iar
"Ashkelon'
Loess soil
Ouartz
NW Neeev
v
isotroDic
mat x
1631/20
Rorrl
Bowi
NPFW Lower
Cretaceous clav AIiI:
Quartz
S Transjordan
2560
/26
R(nn,
Rm. 20
Itrs
Essshell Lowe'r
Cretaceous clay ARF
S Transrordan
Nr
)
teml.er
added
1646/16 Rom.
Cooking
iar
Lowcr
Cretaceous clav ARF
Qu.rrtz
S
Transiordan
1523
/71
Rtxn
Juglet
Eesshell Lowcr
Cretaceous clav ARF
Siltstone
5
Transiordan
1630
/13 Rorn.
Ilowl
NPFW Lower
Cretaceuus clav ARI
Quartz
S Transiordan
Nrr temperadded
2560
/
20 Rc'm I{m.20
Bowl
NPFW Lower
Cretaceous
clav ARF
Quartz
S Transiordan
1182
/9
Rom. Rm.
16 Bowl
NPFW Lower
Cretaceous clav ARF
Quartz
S Transiordan
1.432
/ 7 Ror)1 Rm.
1l
lamp
Nabatean Lower
Cretaceous
clay
Quartz
Limeskrle
Chert S Transioldan
3007
/1A
Rorn
Cooking
iar
Lowcr
Cretaceous clay
Quartz
ARF
S
Transiordan
3236/7 Rorn
llm.49
Jug
Negev 1986
o. 102:89
Lower
Cletaceous ciay
Quartz
AI{1.'
S
Transjordan
Isotropic
matrix
1615/3
Rorn
Rnr.16 Bowl
NPFW Lower
Cretaceous clav ARF
Ouartz
S Transjordan
16t9
/
3 Rom
Cookine
iar Lower Cretaceous clav
Ouartz
ARF
S
Transjordan
188/2 Rorn Rnr.16
Lrmp
Nabatean Lowcr
Creticcous cla! ARF
OLl.r
rtz Feldspar
S Transjordarr
I
Molluscs
646/2
|
Rom.
Jug
Loc'ss or Hamra
soil
I Quartz
NW Neqev
1103/1
|
&ry1
Rm.16
|
Bowl
I
NPFW L,)wer
Cretnceous cLiyl ARF
> I fansloroin
I
