You are on page 1of 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274205115

The Architecture of PPNB ʿAin Ghazal, Jordan

Article  in  Bulletin of the American Schools of Oriental Research · July 1984


DOI: 10.2307/1357072

CITATIONS READS
8 227

2 authors:

Edward Bruce Banning Brian F. Byrd


University of Toronto Far Western Anthropological Research Group
215 PUBLICATIONS   2,142 CITATIONS    74 PUBLICATIONS   1,909 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

China Lake Paleoecology View project

Investigations at Ain Ghazal View project

All content following this page was uploaded by Brian F. Byrd on 30 July 2015.

The user has requested enhancement of the downloaded file.


The Architecture of PPNB ʿAin Ghazal, Jordan
Author(s): E. B. Banning and Brian F. Byrd
Source: Bulletin of the American Schools of Oriental Research, No. 255 (Summer, 1984), pp. 15-
20
Published by: The American Schools of Oriental Research
Stable URL: http://www.jstor.org/stable/1357072
Accessed: 30-07-2015 16:21 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

The American Schools of Oriental Research is collaborating with JSTOR to digitize, preserve and extend access to Bulletin of
the American Schools of Oriental Research.

http://www.jstor.org

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
The Architecture of
PPNB 'Ain Ghazal, Jordan
E. B. BANNING BRIAN F. BYRD
Department of Near Eastern Studies Anthropology Department
University of Toronto University of Arizona
Toronto, Canada M5S IAI Tucson, AZ 85721

INTRODUCTION maximum extent-if we make the untested assump-


tion that the density of settlement along the road-
R ecognition of the Pre-Pottery Neolithic cut is representative of the site as a whole-would
site of 'Ain Ghazal might have waited be approximately 350 houses. Further excavation
years more, or might never have occurred, is of course necessary to test this estimate.
were it not for the fact that a road cut for a new
Amman-Zarka highway interchange revealed re- HOUSE PLANS
mains of substantial architecture, including stone
walls and thick plaster floors. Salvage excavations The 1982 excavations were focused primarily
at the site during the early part of 1982 (Rollefson on four excavation units: 3067, 3073, 3079, and
1984) uncovered less than 1% of this large site. 3083. Only in two of the Squares, 3073 and 3083,
However, when the excavation data are used in were large enough sections of houses exposed to
conjunction with information from a photographic show the relationship between the walls and inte-
survey of the architecture exposed in the roadcuts rior areas. In addition, it is only in Square3083
(Rollefson and Suleiman 1983) a number of gener- that we have a large portion of a house plan.
alizations are possible about the architecture of Square 3083, despite being the most complete
'Ain Ghazal. house we have to date, is missing half of one large
'Ain Ghazal was built on the western slope of room, which was removed by the road construction
the Wadi Zarqa, and portions of the settlement (fig. 2). This "east room" is rectilinear, with a
appear to have been terraced into the hillside, door 1 m in width near the center of the western
thereby minimizing the distance to the river and wall (no other doors are preserved), and with a
what were probably abundant resources in the flat small alcove in the northwest corner. A round
alluvium of the wadi bottom. How near to the basin-like hearth is set into the floor directly
bottom of the hill the village extended is now opposite the doorway about 1.15 m from the west
almost impossible to determine, because construc- wall. Its rim is almost flush with the floor surface,
tion of the highway destroyed the eastern edge of although a slight curb appears along the north-
the settlement (fig. 1). The surviving portions of western arc, and its maximum depth is 6 cm. At
the site, however, extend over an area of approxi- the time of its discovery the hearth was filled with
mately 600 m x 200 m. It appears that the density fine ash upon which rested one fist-sized stone.
of settlement varied from one part of the site to The adjoining room directly to the west is only
another. Rollefson and Suleiman (1983) suggest 3.75 m from north to south, in contrast to the
that there were at least three principal clusters of eastern room, which is 4.9 m in length. Although
houses along the north-south axis of the site. poorly preserved and not completely excavated, a
Within these clusters they recorded traces of at number of features are apparent. Based on the
least 54 buildings with 71 associated floors ex- presence of the north and south walls, the room
posed in the roadcuts. A reasonable estimate of appears also to have been rectangular. The low
the total number of houses in the settlement at its (40 cm in height), semicircular wall segment in the

15

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
16 BANNING AND BYRD BASOR 255

to sports
Cor

LEGEND

0 o073 excovation squaress , ,--., embonkmant


-_ ,7 expod 'structures c" a
rock *secrpmnt
- iron fence o o olive grov
wt_ _ r pipe ___ _ orth channel
I o 100
........ field oll -- contour intervl 2m. 4 I t II I I t I
suvtYtrc a* mo *$*.
s .LODHSTOc
- - ite boundary

Fig. 1. Topographicplanof the site.


middle room is almost certainly a later addition. possibly the remains of a mat, lay just north of
The feature was built on top of the plaster floor, the bowl.
and yellow clay patching was applied to create a The excavation of Square3067 and 3079 ex-
smooth curve from the old floor up the feature, posed only small areas of floor space with no
covering the seam. This low, curved wall may associated walls. Despite the limited exposure, the
have functioned as a bench or the foundation for two squares provided evidence of a consistent
a windscreen. A second feature of this partially placement of hearths over time. Area 3067 had
excavated room in Square 3083 is a square stone three superimposed floors separated from top to
and mud bin in the northeast corner of the room. bottom by 61 cm and 13 cm respectively. The
This feature was devoid of artifacts, and its func- latest floor (Phase III) had two replasterings while
tion is unclear. the Phase II and I floors had three and no replaster-
In Square 3073 information on house plans is ings respectively. Hearths were present in the same
more limited, since this was a step-trench intended restricted area in all but the upper two replaster-
to reveal stratigraphic information rather than to ings of the Phase III floor. A similar situation
expose extensive areas of domestic activity. Four occurs in Square 3079 where the four replasterings
distinct phases of domestic occupation were ex- of the floor each have a hearth located one almost
posed, and it is in the second phase that we have on top of the other.
the most information about the plan of a house.
Unfortunately only a narrow strip 75 cm wide and Construction Details
4.4 m long survived road construction damage
(fig. 3). On this small remnant of the floor, how- Construction of some of the houses in the earlier
ever, a large stone bowl rested halfway along the phases of occupation began with the excavation
west wall, while a fragment of charred textile, of a shallow, flat-bottomed pit. In two cases, 3073

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
1984 ARCHITECTURE OF CAINGHAZAL 17

succeeding phase, which reused the same wall,


also has a posthole halfway along the wall in a
similar position. Unfortunately both ends of the
Phase III floor were destroyed during later occupa-
tion at the site. The construction technique seems
to have been similar in both these phases, although
their respective floors are separated by a half-
meter of deposit. In Phase II the floor plaster,
while curving up this wall, also curved up and
around the posts to a height of at least 25 cm.
Completed walls received an interior coating of
mud plaster and were finished by a thinner layer
of white plaster. In many instances this was very
well preserved, to a height of at least 40 cm in
Phases II and III of Square 3073 and about 70 cm
against wall 011 in Square 3083, Phase II. Some
walls, like the floors, show evidence of replastering.
In some cases collapsed sections of the walls
reveal the white plaster still adhering to wall
fragments that fell forward. To date we have no
evidence of painted designs or patterns on the
walls but in some cases, such as 3073 Phase II, the
plaster had been painted entirely red.
The floors of the houses at CAin Ghazal are
quite accomplished examples of construction. The
plaster floors were often very thick, ranging from
3 cm to 15 cm, and were often built in two parts.
A thick, hard matrix base (Munsell 10YR7/2),
Fig. 2. Planof Square3083, March1983. often incorporating fragments of earlier destroyed
floors and small gravel, was laid down and then
Phase II and 3083 Phase II, this leveling cut followed by a thin half-centimeter upper laminum
through an earlier floor, down to the sterile pleisto- or wash of fine plaster. Often the floors were then
cene clay. painted red. Finally the floors were burnished.
The builders of CAin Ghazal houses erected Where a floor met the base of a wall, invariably
walls of natural stones directly on the ground or the floor curved upward to a height of approxi-
on a previously levelled area often without using a mately 10 cm, giving the corners of rooms a
foundation trench. Mud mortar bonded these distinctly bathtub-shaped appearance. Except in
stones together, creating walls up to 60 cm thick. the earliest phases (3067, 3073, 3083), all the
They were constructed primarily of large stones excavated floors showed evidence of at least one
arranged along the exterior lines to present a major replastering. The best example of this, to
regular, nearly vertical face. Smaller stones filled date, is the superimposition of three floors in
in the gaps between the irregular large ones. Many Square 3067. Here distinct layers of red-painted
of the excavated walls, some of which are preserved finish reveal five phases of replastering.
to a height of I m, lasted long enough to be One final feature of the floors at CAinGhazal
reused at least once. Wall 035 in Square 3073, for requires mention. Whenever preservation permit-
instance, was associated with the floors of both ted, excavation revealed sunken plaster hearths
Phases II and III. set into the floors of each house. The hearths
The western wall of Phase II in Square 3073 ranged in size from approximately 70 cm in dia-
was associated with a row of three wooden posts, meter to an oval 80 cm by 90 cm. These were
each about 13-17 cm in diameter, which lined the integral parts of the floor, not later additions. The
wall. The posts presumably supported primary rims were usually flush with the floor surface,
roof beams and reduced stress on the walls. The gently curving downward into bowl-shaped depres-

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
18 BANNING AND BYRD BASOR 255

sions 10-14 cm deep. In some cases there was a


low rim or lip around the hearth. The latest
hearths in 3073 and 3079 have such raised rims,
and in the case of 3079 this rim was built up of
clay and stones and then plastered over. In cases
where the floor had been finished with red paint
the hearth showed similar treatment, even though
it was destined to be filled with ash that would
cover most of the decoration. The majority of the
hearths excavated in 1982 did not contain ash, but
burned areas often occur in the hearths that lack
ash, and one hearth, 020 in Square 3079, contained
carbonized seeds, charcoal, and firecracked rock.
The inhabitants probably cleaned the hearths fre-
quently.

Decoration

The only type of architectural decoration that


appeared during the 1982 excavations was the
application of red paint to some of the house
floors and walls. The presence of several rubbing
stones and one mortar fragment, all with red
stains on their surfaces, suggests that the hematite
was ground on the site. It was probably suspended
in a liquid for application. Often floors were
completely covered with red paint, although there
are examples of floors with only small areas with
splashed and dripped paint. In one case, 3073
Phase III, the splashes appear as arcs and short
linear alignments of dots. Unfortunately the pre-
served portion of the floor is too small to reveal
any clear patterns. In another case, 3083 Phase II,
the splashed or dripped red paint was restricted to
the area above the burial. At this point in the
research the samples are too limited for us to state
o whether the splotches are intentional or the result
of sloppy work while painting walls, particularly
* in the case of Square 3073 Phase III.

a) Spatial Patterns within Houses


a- To date the excavated sample is still too small
_. for us to say very much about spatial patterning
0
co within the houses at 'Ain Ghazal, but a few
C tentative patterns are emerging. Again Square
cT 3083 offers the most information due to its larger
o exposure of a house plan. One activity area appears
to be present within the doorway and in the
. adjacent part of the west room, where 11 grinding
stones, one of which was stained with red ochre,
i? lay directly on the floor. A sickle blade lay on the

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
1984 ARCHITECTURE OF CAINGHAZAL 19

Area 3079
replastered.The floor in 3079 revealed four hearths
Top Plan
with four sequential burials just to the south of
the hearths (fig. 4). At present, the explanation
for this pattern is unclear, but it appears to relate
to a cultural ritual.

ARCHITECTURALVARIABILITY
IN THE LEVANT

Over the last 30 years, excavations at Jericho,


Beidha, Munhatta, Tell Ramad, Abu Hureyra,
Mureybet, Bouqras, and, to a lesser extent, a
number of other sites have revealed a great deal of
similarity in the late Pre-Pottery Neolithic (Moore
1978). The architecture was quite substantial for
such an early period, at least in the kinds of
settlements that have been excavated to date.
There has been less investigation of more transitory
camps or special-purpose sites in the PPNB (but
0 50cm see Waechter et al 1938; Rollefson and Frolich
I I I I north
1982).
The architectural features that these main sites
Fig. 4. Superimposed burialsand hearths in Square have in common include rectangular buildings of
3079, March1983. stone or mudbrick construction, plastering and
polishing of floors, the use of red or black paint
floor in the same general area. Furthermore, a on floors and walls, frequent replastering and
fist-sized piece of hematite was found slightly rebuilding, the burial of dead beneath floors or in
above the floor next to the center of a low curving open spaces on the site, and dense concentrations
wall. If this feature was a bench, the hematite may of houses with courtyards and lanes separating
have fallen off of it onto accumulated debris them. In these general respects, the 'Ain Ghazal
following abandonment of the house. The floor of architecture is similar to that of other large PPNB
the east room in Square 3083 was devoid of sites.
artifacts, although a thin black layer that covered There are, however, elements of the architecture
much of the floor contained abundant botanical at 'Ain Ghazal that are different from many of the
material. The small alcove in the northwest corner other sites. It appears, at this early stage in the
of House 3083 is also intriguing as a possible research, that the site has less uniformity in house
source of evidence for activity areas. Since there plan than sites such as Jericho (Kenyon 1959,
were no artifacts on the floor in or near this 1960), Abu Hureyra (Moore 1975), Bouqras (Ak-
alcove, however, it is not certain that it functioned kermans et al. 1981), and Tell Ramad (de Conten-
as a storage area. It is possible, however, that the son 1970). This observation is based on the initial
house's last occupants cleaned it out upon aban- excavations and road cut exposures, which show
donment or used it only to store perishables. considerable variation in house and room size,
Finally, there has emerged a pattern of burial and in structural features within the houses. For
placement within the houses at 'Ain Ghazal. Six example, the presence of a curved bench in one
burials were located under the house floors in the domestic room, and the use of posts along the
primary excavation trenches at the site last season, edges of the walls in another building (similar to
and all were placed less than 1.5 m to the south of Hacilar in southwest Anatolia; Mellaart 1970) are
the room's hearth. Although the sample size is atypical features in the Levant. If these observa-
small, the consistency is striking. In addition, the tions are supported by further excavation, there
burials were often placed so close to the hearths will undoubtedly be implications concerning the
that a portion of the hearth was damaged. The growth of the settlement and social organization
entire floor, including the hearth, was then usually within the site (Flannery 1972), one of the largest

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
All use subject to JSTOR Terms and Conditions
20 BANNING AND BYRD BASOR 255

in the Levant, and three times the size of the next Huyuk, however, Mellaart (1967) sees a correlation
largest contemporary site known in the southern between male and female burials and their loca-
Levant, Jericho (Rollefson 1984). tions beneath platforms in the houses. At present,
A major focus of future research at 'Ain Ghazal the 'Ain Ghazal sample is too small to study age-
will be to investigate on synchronic and diachronic sex variation in burial patterns, but in the future
levels the following questions pertinent to cultural that should be possible.
stability and innovation expressed in the archi- The site also offers potential for the study of
tectural sphere. (a) To what degree is architectural activity areas. Sites of this time period generally
variability found within single aggregations of have a paucity of floor assemblages (the most
domestic and possibly public structures? (b) Are notable exception is Beidha; Kirkbride 1966, 1968).
there significant differences in house plans and At 'Ain Ghazal two of the four floor areas exca-
building arrangements among the different clus- vated have in situ floor assemblages. If this trend
ters of buildings noted during the initial survey? continues, the site will offer valuable information
(c) Does the possible variability noted in (a) or (b) on subsistence and social patterning during the
reflect sociocultural, economic, or possible func- PPNB.
tional variation among the resident population(s)?
The distribution of burials south of the hearths
in the houses is a unique feature at 'Ain Ghazal. *The authors are indebted to G. O. Rollefson, of the
Although burials under the floors are common in American Center of Oriental Research and Yarmouk
the PPNB, other sites in the Levant have not University, Irbid, Jordan, for his many useful sugges-
reported a similar distribution pattern. At Catal tions and criticisms of earlier drafts of the manuscript.

BIBLIOGRAPHY

Akkermans, P. A., et al. 1967 Catal Huyuk. A Neolithic Town in Anatolia.


1981 Stratigraphy, Architecture and Layout of London: Thames and Hudson.
Bouqras. Pp. 485-501 in Colloques Inter- 1970 Excavations at Hacilar. Edinburgh: University
nationaux du CNRS. Prehistoire du Levant of Edinburgh Press.
598. Paris: Centre National de la Recherche Moore, A. M. T.
Scientifique. 1975 The Excavation of Tell Abu Hureyra in Syria:
de Contenson, H. a Preliminary Report. Proceedings of the Pre-
1970 Tell Ramad, village Syrien des VIIe et VIe historic Society 41: 50-77.
millenaires. Archeologia 33: 69-73. 1978 The Neolithic of the Levant, I and II. Unpub-
Flannery, K. V. lished doctoral dissertation, Oxford University.
1972 The Origins of the Village as a Settlement Rollefson, G. O.
Type in Mesoamerica and the Near East: a 1984 CAinGhazal: An Early Neolithic Community
Comparative Study. Pp. 23-53 in P. Ucko, in Highland Jordan, Near Amman. Bulletin of
R. Tringham, and G. Dimbleby, eds., Man, the American Schools of Oriental Research
Settlement and Urbanism. London: Duck- 254: (forthcoming).
worth. Rollefson, G. O., and Frolich, B.
Kenyon, K. M. 1982 A PPNB Burin Site in Jebel Uweinid, Eastern
1959 Earliest Jericho. Antiquity 33: 5-9. Jordan. Annual of the Department of Anti-
1960 Excavations at Jericho, 1957-58. Palestine quities of Jordan 26: 189-98.
Exploration Quarterly:88-108. Rollefson, G. O., and Suleiman, E.
Kirkbride, D. In press Survey of PPNB Structures at 'Ain Ghazal.
1966 Five Seasons at the Pre-Pottery Neolithic Vil- Annual of the Department of Antiquities of
lage of Beidha in Jordan. Palestine Exploration Jordan 27.
Quarterly:8-72. Waechter, J. d'A., et al.
1968 Beidha: Early Neolithic Village Life South of 1938 The Excavation at Wadi Dhobai 1937-1938,
the Dead Sea. Antiquity 42: 263-74. and the Dhobaian Industry. Journal of the
Mellaart, J. Palestine Oriental Society 18: 172-86.

This content downloaded from 169.237.215.179 on Thu, 30 Jul 2015 16:21:06 UTC
View publication stats
All use subject to JSTOR Terms and Conditions

You might also like