ChapterPDF Available

The 2000 Ma’aloula Paleolithic Survey.

  • Heidelberger Akademie der Wissenschaften
Nicholas J. Conard
Tübingen 2006
Published by the
Institut für Ur- und Frühgeschichte
und Archäologie des Mittelalters
Universität Tübingen, Germany
In cooperation with
Kerns Verlag
Tübingen Publications in Prehistory reflects the work of a cooperative project
between the Department of Early Prehistory and Quaternary Ecology of the
University of Tübingen’s Institute for Pre- and Protohistory and Medieval
Archaeology and Kerns Verlag to provide the results of current research in
prehistoric archaeology and all its allied fields to a broad international audience.
Inquiries about publications or orders can be directed to:
Kerns Verlag
Postfach 210516, 72028 Tübingen, Germany
Fax: 49-7071-367641 Tel: 49-7071-367768
All rights reserved. Institut für Ur- und Frühgeschichte
und Archäologie des Mittelalters, Universität Tübingen, Germany.
Layout & Design: Knut Bretzke & Andrew W. Kandel.
Cover design by Kerns Verlag
© Nicholas J. Conard
ISBN: 3-935751-08-7
Printed in Germany
Chapter 14 The 2000 Ma’aloula
Paleolithic Survey
N. J. Conard, A. W. Kandel, A. E. Dodonov, A. Abdulrachman
The 1999 Ma’aloula Paleolithic Survey, which is part of the Tübinger Damaskus Ausgrabungs- und
Survey Projekt (TDASP), succeeded in locating 201 Stone Age localities and documenting many
occurrences of Lower, Middle, Upper and Epipaleolithic material in a variety of geographic settings
(Chapters 12 & 13, this volume). These results were surprising since Paleolithic sites in the survey area
had not been documented previously. The initial work focused on establishing the rough distribution
and abundance of occurrences from each cultural group. One of the main goals of the 2000 survey was
to improve our understanding of the geological setting and the evolution of the local landscape. With
this goal in mind, emphasis shifted from documenting ever more Paleolithic localities to gaining more
systematic observations about the geological setting of the sites. This report summarizes some of the
central results of the geological research and presents the distribution of localities for the main
archaeological periods.
General Geological Setting
The Ma’aloula region is located on the northwestern flank of the Palmyride (Tadmoride) Mountains
close to the Anti-Lebanon orogenic zone. Tectonically, the region is situated in the southern portion of
the Nabk Anticline along its northwestern edge. Paleogene limestone and marl with thin lenses of flint,
as well as Neogene conglomerate, marl and sandstone are widespread in the survey region. Quaternary
deposits fill wadis and form a relatively thin cover atop the Paleogene and, to a lesser extent, Neogene
bedrock, as depicted in the Geological Map of Syria from 1964. Upper Cretaceous limestone and
sandstone with occasional beds of flint are represented only in the southeastern corner of the research
Survey Methods
During September 2000 the survey team chose field routes along the main drainages that transect the
highlands and lowlands, including Wadi Ma’aloula, Wadi Jaba’deen and Wadi George Jaffa. Other
mapping routes focused on the area formed by Paleogene and Neogene rocks that create the higher
topographic relief. Geological and geomorphological observations were systematically recorded at
over 100 points. At these points stratigraphic, lithological and facial analysis were performed.
Occasional paleopedological observations were documented, and at three exposures, palynological
samples were collected.
N. J. Conard, A. W. Kandel, A. E. Dodonov, A. Abdulrachman
Despite the many exposures of Pleistocene deposits, the very coarse nature of the deposits makes
environmental interpretations based on sedimentary and pedological studies difficult. Similarly these
coarse sediments were inappropriate for paleomagnetic studies. In general no fossilized fauna or flora
could be recovered. Only within a lacustrine deposit in the upper portion of Wadi Ma’aloula could a
useful molluscan sample containing at least five fresh water species be recovered from the silts of a
small lake that followed the contours of the wadi in the lower area above the Oligocene cliffline.
Additionally, the excavations at Baaz Rockshelter (Chapters 4-6, this volume) promise to provide
much environmental data for the Epipaleolithic and early Neolithic of the region.
Geological and geomorphological studies were combined with archaeological observations. In
this way the archaeological finds were used as rough chronostratigraphic markers for many open-air
deposits and surfaces that were otherwise not datable.
Quaternary Stratigraphy
The Quaternary stratigraphic subdivisions used here are those of the international stratigraphic scale.
Accordingly, the Neogene-Quaternary boundary is drawn at 1.8 Ma based on the Vrica section in
southern Italy. The Lower Pleistocene covers the range from 1.8–0.8 Ma, while the Middle Pleistocene
covers the range 0.8–0.13 Ma and the Upper Pleistocene from 0.13–0.01 Ma. The Holocene includes
the last 0.01 Ma. This is not a trivial remark since the geological mapping conducted by Soviet
scholars in the 1960s used the Russian chronostratigraphic scheme in which the Lower Pleistocene
corresponded to 0.8–0.4 Ma and the Middle Pleistocene to 0.4–0.13 Ma.
In the survey region, Middle Pleistocene, Upper Pleistocene and Holocene deposits can be rec-
ognized on the basis of geomorphological and archaeological data (Fig. 1).
The Middle Pleistocene is represented by conglomerates that are widespread to the southeast of
the Ma’aloula-Jaba’deen cliffline. These conglomerates are comprised of boulders, pebbles and gravel
with a carbonaceous, coarse sand matrix. They are poorly bedded and unsorted, and contain angular
and rounded pieces. The conglomerates occasionally contain thin lenses of sandy silt with gravel and
pebbles. The conglomerates are of proluvial origin and formed during phases of highly seasonal pre-
cipitation and surface erosion. This poorly sorted, aggrading debris, or proluvium in the Russian termi-
nology, formed over large surfaces during the Middle and Late Quaternary. Proluvium can be equated
most closely with the term fanglomerate used by western geologists, but the authors favor the Russian
term because it is genetically more precisely defined. Upper Pleistocene deposits are characterized by
wadi fill, proluvium and slope colluvium. Fine stratigraphic markers are rare, and only rough strati-
graphic divisions can be achieved in the open-air settings within the region.
Geomorphological Units
The geomorphological units that form the basis for the analysis of prehistoric settlement in the
Ma’aloula region are summarized in Fig. 2. These units correspond closely to the topographic zones
defined in Chapter 13 of this volume and include: 1) the highland plateau formed on Pliocene
conglomerates with thick calcrete at the top tilting slightly downward toward the northwest; 2) the
highland hills formed on Pliocene conglomerates; 3) the resistant Oligocene limestone that forms the
Ma’aloula-Jaba’deen cliffline; 4) the slope at the base of the cliff covered by Late Pleistocene and
Holocene colluvium; 5) the lowland hills formed on Eocene marl and limestone and Middle
Quaternary conglomerates; and, 6) the lowland plain formed on the Middle Quaternary proluvial
conglomerates tilted slightly downward toward the southeast
Collection Procedure
Survey routes covered all of the geomorphological units. The routes usually represent a transect that
was walked in a work day running from 6 am to 2 pm. On several occasions shorter routes and
scattered points were surveyed in the afternoons or as circumstances permitted.
The 2000 Ma’aloula Paleolithic Survey
Figure 1. Ma’aloula Paleolithic Survey. Geological cross-section A-B through the Ma’aloula
hills depicting the location of the six geomorphological units (Scale 5:1, vertical to
When the survey team, which usually consisted of two people, but at times included either three or one,
spotted lithic artifacts, a collection ensued. This collection usually lasted ca. 20 minutes, but depending
on the setting could last anywhere from five minutes to an hour. During collection an attempt was
made to determine the spatial distribution of the artifacts and to define the geographic setting of the
locality. The crew plotted the position of the locality using GPS technology with the uncertainty of the
measurement usually falling between 5-15 m. The elevation of each locality was measured using an
altimeter which was calibrated daily to 1400 m on the lower terrace of the Convent of St. Takla. The
abundance of flint and non-flint raw material was assessed as being absent, low, medium or high.
Similarly, the abundance of artifacts was categorized. When possible the density of artifacts was
assessed for each cultural period identified. These and other parameters were recorded on pre-printed
TDASP survey forms to facilitate the systematic collection of data. Sketch maps and photographs were
made of many, but not all, localities. The spatial distribution of finds was assessed on a scale from
discrete to continuous. While some lithic scatters were relatively small and covered only a few hundred
square meters, other localities covered well over a hectare and showed no obvious boundaries. Under
the latter conditions, which most often prevailed in the lowland areas, at least 200 m were left between
collection points. While the 219 localities reported here all reflect areas where lithic artifacts could be
spotted on the surface, they do not always represent areas with well-defined spatial borders. Many of
the localities produced material from multiple cultural periods. In the present report, areas from which
collections were made are referred to as localities, and collections from a single, cultural group within
a locality are referred to as sites.
Since the volume of the available storage facilities and our processing capacities were limited, a
selection strategy was needed to reduce the number of artifacts removed from the localities. Ceramic
and post-Epipaleolithic artifacts were rare in the survey region, but were collected when encountered.
After a period of collection either Conard or Kandel described the artifacts present using the
checklist on the survey form and recorded additional comments. Usually one or two dozen artifacts
were taken to the field laboratory for closer study. At the most interesting localities, as many as 50
finds were collected; at very low density localities, often only a few artifacts were collected. Even this
limited collection procedure led to the collection of hundreds of kilograms of specimens. All the finds
from the Ma’aloula Survey, the Mediterranean and Damascus surveys (Chapter 12, this volume), and
Baaz Rockshelter are housed or displayed in the Museum of Deir Atieh.
Late Pleistocene and Holocene proluvium (rock debris) (Q3-4)
Late Pleistocene - Holocene slope colluvium (Q3-4)
Middle Pleistocene conglomerates, proluvium (Q2)
Alternating Pliocene conglomerate and marl (N2)
Calcrete at the top of Pliocene deposits
Oligocene limestone forming cliff line (P3)
Eocene marl with thin flint layers (P2)
Upper Cretaceous limestone with flint beds (K2)
Baaz Rockshelter
Highland plateau Highland Cli Cli Lowland hills Lowlands
hills line slope
0 1 2 3km
N. J. Conard, A. W. Kandel, A. E. Dodonov, A. Abdulrachman
Figure 2. Ma’aloula Paleolithic Survey. Areal extent of the six geomorphological units.
While an attempt was always made to collect representative examples of each cultural group, non-
diagnostic finds were also saved. Cores are strongly represented in the collections, since they could
often be placed within an archaeological period. Debitage and tools are also well represented in the
collections. As is often the case with surface assemblages, the distinction between natural edge damage
and anthropogenic retouch was not always clear. At localities with moderate and high artifact densities,
countless diagnostic finds were left in the field. This was particularly the case at Levallois localities
where diagnostic artifacts were often present in great numbers. Survey crews recovered handaxes from
eleven localities in the Ma’aloula region, but never more than one per locality.
Criteria for Defining Archaeological Periods
Given that GPS technology has largely eliminated the problems traditionally associated with plotting
sites identified on surveys, the biggest problem the TDASP team faced was assessing the artifacts
collected from each locality and defining temporal and cultural units of analysis. In many cases
individual finds could be placed in archaeological units without difficulty. This was the case, for
example, with Lower Paleolithic handaxes, Levalloisian cores and flakes, classic Upper Paleolithic
cores, Epipaleolithic lunates or Neolithic projectile points. Nonetheless, many flint artifacts are non-
diagnostic, and at some localities diagnostic forms were rare.
The 2000 Ma’aloula Paleolithic Survey
Generally, the survey sought only to distinguish between the major phases of cultural development in
the Ma’aloula area. At a simplistic level one could consider the most fundamental divisions to be the
Lower, Middle, Upper and Epipaleolithic. While using the results from Yabroud (Rust 1950) or El
Kowm (Le Tensorer & Muhesen 1997) as guides, a finer classification would be possible. However,
we decided not to subdivide the Lower Paleolithic, the Levalloisian Middle Paleolithic, the Upper or
Epipaleolithic at the current phase of analysis.
The biggest problem at this stage of the study is to decide how to classify finds that do not fall
neatly into one of the main categories. In many regions one could use the degree of patination, rolling
or other secondary characteristic to help establish a relative chronology. In the TDASP survey, these
techniques could be used at some localities, but were often of little help since many Lower and Middle
Paleolithic artifacts are in fresh, unpatinated condition. The problem of classification is particularly
acute for the Yabroudian, “pre-Aurignacian” as defined by Rust, and the Hummalian. These assem-
blages are often seen as being transitional between the Lower and Middle Paleolithic and are often dif-
ficult to identify in mixed surface assemblages, that are the norm in the Ma’aloula region. Thus far,
these assemblages have been classified as non-Levalloisian Middle Paleolithic, but more work is
needed to distinguish between the documented variation in these assemblages and to better distinguish
between Lower, Middle and Upper Paleolithic reduction strategies. Another problem centers on the
Upper Paleolithic that is well represented in the surface collections. Here one runs into difficulty
clearly distinguishing the finely made pre-Upper Paleolithic single and multiple platform blade cores
from the Upper Paleolithic cores. Similarly, the larger Epipaleolithic blade cores are not always easy to
distinguish from Upper Paleolithic cores. Here the rich assemblages of cores, debitage and tools from
Baaz Rockshelter serve as a guideline for defining the Epipaleolithic. A good Upper Paleolithic
sequence from a well-excavated context in the region would help to alleviate this problem.
The Distribution of the Paleolithic Open-air Sites
At this stage of the analysis, several conclusions can be made about the distribution of sites in the
Ma’aloula region. First it is surprising that so many Paleolithic localities exist at all. Based on many
casual observations and the results of the systematic survey, it is certain that many sites within the
survey area remain undocumented, and that the routes walked during the survey covered only a small
fraction of the total area. Nonetheless the multiple transects through all of the geomorphological zones
should provide a reliable sample of the population of localities in the region.
The 57 Lower Paleolithic occurrences are present in all of the geomorphological zones except
along the cliffline (Fig. 3). They are most common in the highland hills and the lowland hills, with
fewer sites in the lowlands, and still fewer on the slope below the cliffs. This situation is in part
explained by the heavy erosion and redeposition along the cliffline and slope. Here, any finds dating to
the Middle Pleistocene would have been buried in the thick proluvial and slope colluvial deposits. The
highland plateau has not yielded Lower Paleolithic sites. Of the 11 sites that produced handaxes, seven
are from the lowland hills, 3 are from the highland hills, and one is from the lowlands. These land sur-
faces must be highly stable to preserve the handaxe and Lower Paleolithic core and flake assemblages.
Occasionally, the survey team recovered Lower Paleolithic finds from near wadi bottoms. These and
younger Middle Paleolithic finds were usually surprisingly fresh considering their presence in redepos-
ited gravel.
Levalloisian sites are by far the most abundant in the Ma’aloula region, and these sites are often
the richest (Fig. 4). While the survey team often was unable to securely categorize non-Levallois finds
into cultural groups, Levallois cores and flakes are easy to recognize. The 159 Levallois sites include
examples from all geomorphic regions, with the lowland hills, highland hills and highland plateau pro-
viding the vast majority of the Levallois occurrences. The lowlands yielded sites, but in a lower density
than that documented in the zones mentioned above. Sites are rare on the slope and absent on the clif-
fline. Again geological and anthropogenic processes have reworked and buried nearly all the sediments
in these areas, thereby destroying the archaeological sites. Many Levallois sites produced a high or
medium density of artifacts. A wide variety of Levallois cores and flakes are present in the survey
N. J. Conard, A. W. Kandel, A. E. Dodonov, A. Abdulrachman
region. While many sites yielded large numbers of cores, few sites produced numerous Levallois flakes
in the absence of cores. Thus, more often than not, cores seem to be somewhat over represented. As
one would expect there is a clear positive correlation between artifact and raw material abundance.
More systematic analysis of large collections from the sites is needed to better address questions
related to the use of different spots in the landscape and to reconstruct Middle Paleolithic settlement
dynamics in the region.
Figure 3. Ma’aloula Paleolithic Survey. Distribution of Lower Paleolithic
sites and sites with handaxes.
The survey team identified 132 distinct Upper and Epipaleolithic sites. These sites are spread across
the landscape and covered all geomorphological regions (Fig. 5). The presence of Upper and
Epipaleolithic flint artifacts along the cliffline and the upper reaches of the slopes below the cliffline
documents an intense use of the caves and rockshelters of the region during the Upper and
Epipaleolithic. Unfortunately, nearly all of these caves have been emptied by natural and
anthropogenic processes. Baaz Rockshelter is the best preserved site in the region and promises to
yield a good sequence for the Epipaleolithic. The as of yet unexcavated Louis and Kaus Kozah (see
chapters 7 & 8 this volumen) caves may provide cultural stratigraphic sequences for the Upper
Closing Remarks
In the coming years the TDASP team will work to refine the results reported here and will continue
excavations to gain additional samples of stratigraphy for controlling the typological and technological
assessments of the artifacts collected during the survey. Since many of the results reported here are
based on the analysis of artifacts while in the field, the results must be viewed as preliminary. In the
near future a study season should be planned to analyze the survey finds in more detail. Additionally
we hope to gain further chronostratigraphic and environmental data in the coming years to help
complete the picture of the changing patterns of settlement and landuse in the Ma’aloula survey region.
The 2000 Ma’aloula Paleolithic Survey
Figure 4. Ma’aloula Paleolithic Survey. Distribution of Levalloisian
Middle Paleolithic sites.
Figure 5. Ma’aloula Paleolithic Survey. Distribution of Upper Paleolithic
and Epipaleolithic sites.
Finally, surveys of this kind are of particular importance in the more populated regions of Syria. With
each passing year more of the landscape is being altered by agriculture, building, and industrialization.
Over the two years of the Ma’aloula Survey, the team witnessed the terracing and plowing of many
hectares of new agricultural land for fields, orchards and vineyards. All of the villages in the survey
N. J. Conard, A. W. Kandel, A. E. Dodonov, A. Abdulrachman
area are expanding rapidly, thereby destroying many of the sites that would be promising for future
study. While most of the caves and rockshelters in the region have long been emptied, the few sites
with sediments are being damaged at a fast rate. If one can extrapolate from our experience in the
Ma’aloula area, thousands of Stone Age sites are being damaged or destroyed in Syria every year, the
vast majority of them remaining unnoticed and undocumented.
We wish to thank the General Director of Antiquities for supporting this research, Prof. J.-M. Le
Tensorer for scientific advice, and the nuns of St. Takla Convent for providing the team with room and
board during the 2000 field season. This research was funded by the University of Tübingen.
Le Tensorer, J.-M. & Muhesen, S. 1997. Les premiers hommes du désert syrien. Paris: Muséum
National d’Histoire Naturelle.
Rust, A. 1950. Die Höhlenfunde von Jabrud (Syrien). Neumünster: Karl Wachholz Verlag.
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.