23 Pages
English

Nahal Ein Gev I : A Late Upper Palaeolithic Site by the Sea of Galilee, Israel - article ; n°1 ; vol.30, pg 25-45

-

Gain access to the library to view online
Learn more

Description

Paléorient - Année 2004 - Volume 30 - Numéro 1 - Pages 25-45
Sont présentés ici les résultats de fouilles qui furent reprises sur le gisement paléolithique supérieur tardif de Nahal Ein Gev l, situé sur la rive orientale du lac de Tibériade. L 'assemblage lithique est caractérisé par un débitage sommaire à éclats grossiers obtenu à l 'aide d'un percuteur dur. Parmi les outils prédominent les burins sur troncatures et encoches clactoniennes. Les autres classes d 'outils n 'apparaissent pas spécialement standardisées mais comportent une petite composante microlithique. Ces résultats recoupent largement ceux obtenus lors des fouilles antérieures et montrent bien que l 'assemblage décrit est représentatif de l 'ensemble du site. Il semble presque certain que les restes humains qui avaient été signalés appartiennent bien au site ; celui-ci devait être un camp de prédateurs comme l 'indique la grande variété des restes d 'animaux recueillis.
Dans le cadre du Paléolithique supérieur tardif l'assemblage appartient à l 'entité Atlitienne comme le montrent aussi bien la technologie que la typologie alors qu 'ils n 'offrent peu, si ce n 'est aucune, ressemblance avec les industries de l'Aurignacien levantin.
This paper presents the results of renewed excavations at the late Upper Palaeolithic site of Nahal Ein Gev 1 on the eastern margins of the Sea of Galilee. The lithic assemblage is characterised by a chunky flake technology produced using a hard hammer. The tool repertoire is heavily dominated by burins on Clactonian truncations and notches. Other tool classes are not especially standardised, and there is a small microlithic component. The results broadly conform to those of the previous excavations and corroborate that the described assemblage is indeed representative of the site as a whole. It seems almost certain that the previously described skeletal remains are indeed integral to the site, which appears to represent a foraging camp as indicated by the broad range offaunal remains. The assemblage is placed within the framework of the later Upper Palaeolithic Atlitian entity. It is concluded that the techno-typolog- ical configuration of this and comparable Atlitian assemblages bear little, if any, resemblance to Levantine Aurignacian industries.
21 pages
Source : Persée ; Ministère de la jeunesse, de l’éducation nationale et de la recherche, Direction de l’enseignement supérieur, Sous-direction des bibliothèques et de la documentation.

Informations

Published by
Published 01 January 2004
Reads 154
Language English
Document size 4 MB

M. Spiers
Daniel E. Lieberman
A. N. Goring-Morris
A. Davidzon
Anna Belfer-Cohen
Nahal Ein Gev I : A Late Upper Palaeolithic Site by the Sea of
Galilee, Israel
In: Paléorient. 2004, Vol. 30 N°1. pp. 25-45.
Citer ce document / Cite this document :
Spiers M., Lieberman Daniel E., Goring-Morris A. N., Davidzon A., Belfer-Cohen Anna. Nahal Ein Gev I : A Late Upper
Palaeolithic Site by the Sea of Galilee, Israel. In: Paléorient. 2004, Vol. 30 N°1. pp. 25-45.
doi : 10.3406/paleo.2004.4771
http://www.persee.fr/web/revues/home/prescript/article/paleo_0153-9345_2004_num_30_1_4771Abstract
This paper presents the results of renewed excavations at the late Upper Palaeolithic site of Nahal Ein
Gev 1 on the eastern margins of the Sea of Galilee. The lithic assemblage is characterised by a chunky
flake technology produced using a hard hammer. The tool repertoire is heavily dominated by burins on
Clactonian truncations and notches. Other tool classes are not especially standardised, and there is a
small microlithic component. The results broadly conform to those of the previous excavations and
corroborate that the described assemblage is indeed representative of the site as a whole. It seems
almost certain that the previously described skeletal remains are indeed integral to the site, which
appears to represent a foraging camp as indicated by the broad range offaunal remains. The
assemblage is placed within the framework of the later Upper Palaeolithic "Atlitian " entity. It is
concluded that the techno-typolog- ical configuration of this and comparable Atlitian assemblages bear
little, if any, resemblance to Levantine Aurignacian industries.
Résumé
Sont présentés ici les résultats de fouilles qui furent reprises sur le gisement paléolithique supérieur
tardif de Nahal Ein Gev l, situé sur la rive orientale du lac de Tibériade. L 'assemblage lithique est
caractérisé par un débitage sommaire à éclats grossiers obtenu à l 'aide d'un percuteur dur. Parmi les
outils prédominent les burins sur troncatures et encoches clactoniennes. Les autres classes d 'outils n
'apparaissent pas spécialement standardisées mais comportent une petite composante microlithique.
Ces résultats recoupent largement ceux obtenus lors des fouilles antérieures et montrent bien que l
'assemblage décrit est représentatif de l 'ensemble du site. Il semble presque certain que les restes
humains qui avaient été signalés appartiennent bien au site ; celui-ci devait être un camp de prédateurs
comme l 'indique la grande variété des restes d 'animaux recueillis.
Dans le cadre du Paléolithique supérieur tardif l'assemblage appartient à l 'entité Atlitienne comme le
montrent aussi bien la technologie que la typologie alors qu 'ils n 'offrent peu, si ce n 'est aucune,
ressemblance avec les industries de l'Aurignacien levantin.,
Nahal Ein Gev I :
a late upper palaeolithic site by
the Sea of Galilee, Israel
A. Belfer-Cohen, A. Davidzon, A.N. Goring-Morris, d. Lieberman and M. Spiers
Abstract : This paper presents the results of renewed excavations at the late Upper Palaeolithic site of Nahal Ein Gev 1 on the eastern
margins of the Sea of Galilee. The lithic assemblage is characterised by a chunky flake technology produced using a hard hammer. The
tool repertoire is heavily dominated by burins on Clactonian truncations and notches. Other tool classes are not especially standardised,
and there is a small microlithic component. The results broadly conform to those of the previous excavations1 and corroborate that the
described assemblage is indeed representative of the site as a whole. It seems almost certain that the previously described skeletal
remains2 are indeed integral to the site, which appears to represent a foraging camp as indicated by the broad range offaunal remains.
The assemblage is placed within the framework of the later Upper Palaeolithic "Atlitian " entity. It is concluded that the techno-typolog-
ical configuration of this and comparable Atlitian assemblages bear little, if any, resemblance to Levantine Aurignacian industries.
Résumé : Sont présentés ici les résultats de fouilles qui furent reprises sur le gisement paléolithique supérieur tardif de Nahal Ein Gev
l, situé sur la rive orientale du lac de Tibériade. L 'assemblage lithique est caractérisé par un débitage sommaire à éclats grossiers obtenu
à l 'aide d 'un percuteur dur. Parmi les outils prédominent les burins sur troncatures et encoches clactoniennes. Les autres classes d 'outils
n 'apparaissent pas spécialement standardisées mais comportent une petite composante microlithique. Ces résultats recoupent largement
ceux obtenus lors des fouilles antérieures et montrent bien que l 'assemblage décrit est représentatif de l 'ensemble du site. Il semble pres
que certain que les restes humains qui avaient été signalés appartiennent bien au site ; celui-ci devait être un camp de prédateurs comme
l 'indique la grande variété des restes d 'animaux recueillis.
Dans le cadre du Paléolithique supérieur tardif l 'assemblage appartient à l 'entité Atlitienne comme le montrent aussi bien la technolo
gie que la typologie alors qu 'ils n 'offrent peu, si ce n 'est aucune, ressemblance avec les industries de l'Aurignacien levantin.
Key-Words : Southern Levant, Late Upper Palaeolithic, Atlitian, Lithic techno-typology.
Mots Clefs : Levant Sud, Paléolithique supérieur tardif, Atlitien, Techno-typologie lithique.
More than half a century after the broad outlines of the lithic Stage V of Neuville3 or within the "Atlitian" entity of
Levantine Upper Palaeolithic sequence were defined, the later Garrod4 (fig. 1). Accordingly the assemblage obtained from
part of the Upper in the region remains poorly the initial excavations at Nahal Ein Gev I was referred to as
documented and understood. Originally the cultural entities "Levantine Aurignacian C" or Neuville's Stage V5, although
pertaining to this period were incorporated in Upper Palaeo- it was clear to all (including the original excavator) that the
3. NFUVILLE, 1951.
1 BAR-YOSEF 1973 4- Garrod and Bate, 1937.
2. ARENSBURG, 1977. 5- BAR-YOSEF, 1973.
Palčorient, vol. 30/1 p. 25-46 С CNRS ÉDITIONS 2004 Manuscrit reçu le 14 janvier 2004, accepté le 10 juin 2004 26 A. belfhr-Cohen, a. davidzon, a.n. goring-Morris, d. lieberman and M. Spiers
assemblage in question differs significantly from that of layer 1993 FIELD SEASON
С at el-Wad, i.e. the type-assemblage of Garrod's "Atlitian"
Since the vicinity of the site had been substantially altered entity. The Atlitian was considered to represent the culminat
in the intervening years due to the construction of a cattle ion or latest phase of the Upper Palaeolithic "Aurignacian"
enclosure and various fences and gates, it was difficult to pretradition, as summarized by Copeland6. Still, as has recently
cisely locate the previous excavation areas. Accordingly, a been discussed, an increasing awareness has developed con
series of 19 1 x 1 m test pits was excavated upslope from cerning the complexity of Upper Palaeolithic developments
where the previous excavations were thought to have centred, within the Levant7. Details of the lithic assemblage deriving
adjacent to the bulldozer trench, mostly at intervals of 5 m from renewed investigations at Nahal Ein Gev I below are fo
northwards towards the base of the nearby Tel Nuqeib llowed by a discussion of the site within the broader context of
(figs 2-3). The 1 x 1 m test pits were excavated in 10 cm spits later Upper Palaeolithic developments in the region.
to depths of 50-100 cm with the aim of elucidating the spread
of the site and the location of potentially in situ deposits. In
addition, a short geological test pit measuring 3.0 x 0.5 m was
LOCATION AND HISTORY OF RESEARCH excavated in the vicinity of the original bulldozer trench
downslope to the southwest. Subsequent examination of air
photos and visits by participants of the first project during the The site of Nahal Ein Gev I is located on the northern bank
of Nahal Ein Gev, some 160 m below sea level, ca 3 km east
of the shore of Lake Kinneret - the Sea of Galilee (fig. 1 )8. The
site was originally discovered in 1971 eroding from a bul
ldozer trench immediately to the south of the biblical Tell
Nuqeib during a survey in the area9. A small test excavation of
5 x 5 m2 was initiated at that time to recover the remains of an
articulated human skeleton, which had been exposed within
the trench section. The associated finds included Upper Palae
olithic lithic material, faunal remains, as well as later proto-
historic pottery and lithics. The lithic assemblage was assigned
to the late "Levantine Aurignacian" on techno-typological
grounds. Comparative fluoride tests at the time indicated that
the skeleton (of a 30-35 years old female) most likely predated
the Epipalaeolithic period10. A further test pit of 3 m2 was
later placed some 30 m away to the southwest, which also
yielded similar materials. These initial investigations thus indi
cated the presence of a substantial open-air late Upper Palaeol
ithic occupation. However, it remained unclear as to the
extent of the original occupation, as well as the degree of
admixture deriving from the biblical Tel Nuqeib located
immediately to the north-northwest of Nahal Ein Gev I.
6. COPKLAND, 1975.
7. Belfer-Cohen and Goring-Morris, 2003.
8. Nahal Ein Gev 1 is located between Israel Grid coordinates 21165-
24335 and 2 1 1 7-24345. The site is situated in a cow-pasture adjacent to an avo
Fig. 1 : Map of the Levant showing the location of Atlitian and cado orchard of Kibbutz Ein Gev.
9. BAR-YOSEF, 1973. related late Upper Palaeolithic sites. Insert : Map of the Galilee and
10. ARENSBURG. 1977. Carmel area showing the location of Nahal Ein Gev I.
Paléoricnt. vol. 30/1. p. 25-46 с CNRS EDITIONS 2004 :
:
Ein Gev I A Late Upper Palaeolithic Site by the Sea of Galilee, Israel 27 Nahal
Fig. 2 : Fifew o/V/ze excavation of Nahal Ein Gev I from the southeast.
Note the location of the various test pits in relation to the animal
corral. The lower slopes of Tel Nuqeib are visible in the background.
course of the season confirmed the location of the previous
excavation area with the human remains, immediately SSE of
the geological test trench.
In light of these tests systematic excavations then focused
on the upslope area, where the Palaeolithic layers were most
likely to be preserved, so that a contiguous area of 9 m2 was
partially excavated (see fig. 3). The excavation was con
ducted in 10 cm spits since no vertical differences were
observed within the occupation horizon (see below). Almost
all sediments were dry sieved through 2 mm mesh screens and
selected units were wet sieved. Although flotation was
attempted, no botanical remains were obtained because of the
nature of the sediments. Geological samples were taken for
micro-morphological analyses.
Fig. 3 : Plan of the test pits and main excavation in relation to the
area of previous investigations at Nahal Ein Gev I.
STRATIGRAPHY
ties of large pebbles, cobbles and boulders. Upper Palaeolithic The profiles provided by the test pits present a generally
lithics were present in some quantity, together with many potuniform picture of the stratigraphy at the site1 1 . In areas where
sherds and occasional proto-historic and Epipalaeolithic lithrecent bulldozer activity has not stripped it away, an upper
ics, but few bones. This material appears to primarily represent horizon (Level I) was noted, varying in depth from 20-100 cm,
slopewash from the adjacent Tel Nuqeib into which eroded and of generally loose grey, sandy matrix, sometimes with quanti-
deflated Upper Palaeolithic and occasional Epipalaeolithic
material has been incorporated. The base of this level is more
11. BAR-YOSEF, 1973, reported two archaeological layers a more dis compact and bedded, with increasing quantities of calcium carturbed upper layer with 14.5 tools/m3 and numerous ceramic fragments, and a
bonate nodules. Sometimes it incorporates fine gravel lenses. lower layer with 24.8 tools/m3 and much fewer ceramic fragments. The results
of the 1 993 season suggest that the mixtures reported by Bar-Yosef were The underlying Level II is represented by compacted and
caused by contamination of level II in the lower part of the site from re-surfac cemented sandy colluvium, with some clay and numerous caing and bulldozing. We found little evidence for ceramics or recent material in
lcium carbonate nodules, and is up to 50-60 cm thick. This rests level II in the northern part of the site where level I (the sandy fill) was undis
turbed. unconformably upon a heavily cemented calcium carbonate rich
Paléorient, vol. 30/1, p. 25-46 @ CNRS EDITIONS 2004 :
;
'
28 A. belfer-Cohen, a. davidzon, a.n. Goring-Morris, d. lieberman and M. Spiers
conglomerate/calici layer (Level III) of unknown thickness. occupation, and that it may well have been intentionally buried to
one side of the site rather than within the site itself. Indeed this Bones, thickly encrusted with concretions, and accompanied by
quantities of Upper Palaeolithic lithic material were recovered may actually reflect Levantine Upper (and even early Epi-) Pal
from Level II in some of the northernmost (i.e. upslope) squares. aeolithic burial tradition norms, which were basically off-site, or
at least off to one side of the site, e.g. Ohalo II 3. This may explain Nevertheless, the contact between Levels 1 and II was
sometimes fuzzy, and mixing of both levels was noted in the (relative) paucity of human remains from both periods.
some areas of the site. This probably reflects biogenic proc
esses, since numerous animal burrows were discerned FAUNAL REMAINS
throughout Levels I and II, as well as recent agricultural and
other activities. Considerable size sorting of both potsherds A total of 502 bone fragments were recovered in the present
and lithics were noted down the length of the slope, with excavations, of which 72 have been identified to species and
larger items being concentrated in the upper area. element (table 1). With the exception of a few bones from
While the contiguous excavation area in grid squares recent fill layers, almost all the bones are covered with a thick
T102/103,S/U104, S/T/U 105, S/U 106, and test pit W 106 pro crust of CaCO3 that will have to be removed to allow further
vided numbers of faunal remains and abundant, large lithic analysis of the material. Preliminary analysis of the faunal
items, the material does not appear to be in primary context, material indicates the presence of the following mammalian
although it has not been washed to any considerable distance. remains (in decreasing order) Gazella gazella, Cervus elap-
Indeed, only in the uppermost, partially excavated test pit in hus, Bos primigenius, Dama mesopotamica, Capra ibex, and
grid square W97 does it seem likely that material is archaeol Vulpes vLilpes. Also present were small quantities of avifauna
ogical ly in situ in primary context. and microfauna. The lack of pig (Sus) remains is surprising, but
Accordingly, it appears that the extensive spread of arte it is likely that some of the unidentified fragments are of pigs.
facts over the site, encompassing an area of at least 2 000 m2, Most of the skeletal elements are appendicular, but there
is largely the result of post-Upper Palaeolithic erosion, slope- are a few axial and cranial fragments, including 10 mandibu-
wash and dispersion of part of the site down the slope to the lar fragments with teeth.
south, in the direction of the wadi (gradient ca l:10).12 This
appears to have both preceded and been pene-contemporane-
LITHIC INDUSTRY ous with the occupation of Tel Nuqeib, from at least the Early
Bronze Age through to sub-recent Arab settlement. Such a
The following detailed description is based on the sample process would also account for the presence of small numbers
deriving from the main excavation area (9 m2) only. However, of clearly intrusive Epipalaeolithic microliths.
the lithics from nine of the 1 9 surrounding test pits (grid squares In summary it appears that the centre of the original occu
pation was located in the general northern area of the excavat Table 1 : NISP of species represented in the present excavation.
ions and that most of the archaeological materials recovered
Species n % from the test pits excavated by us (and also previously) mostly
Gazella gazella 21 29 derive from erosion and slopewash, even though they largely
conform to the rest of the Upper Palaeolithic finds plus pottery. Cervus elaphus 17 24
Unfortunately, no charcoal was recovered for radiometric Bos 14 19
dating. Nevertheless fluoride testing indicated that there is a Dama mesopotamica 7 10
significant difference in fluoride content between the fauna Capra ibex 2 3
recovered from the excavation and recent bones collected for Vulpes vulpes 1 1
comparative analysis. Small mammal 6 8
However, it should be noted that the recovered skeleton Medium mammal 2 3
(which was articulated) is likely to date to the Upper Palaeolithic Large mammal 1 1
Bird 1 1
12. Few objects were found in horizontal position rather, artefacts and Total 72 99 bones appear to be clumped in random orientations in dense concentrations
that represent small hollows or fills in the slope, with larger artefacts and bones
found towards the top of the slope and smaller ones towards the bottom. 13. NADEL and HERSHKOVITZ, 1991 ; NADEL, 1995.
Paléorient, vol. 30/1, p. 25-46 С CNRS EDITIONS 2004 :
:
:
:
Nahal Ein Gev I A Late Upper Palaeolithic Site by the Sea of Galilee, Israel 29
G120, L140, L130, LI 10, L100, T110, T123, T137, and W97 assemblage with a low level of finish. Both flakes and blades
- see fig. 4) were also studied for comparative purposes. The feature thick striking platforms with prominent bulbs of per
study revealed much lower artefact densities, accompanied by cussion. There is a virtual absence of abrasion from the strik
larger quantities of intrusive potsherds, as compared to the ing platforms down the removal surface. Hinging is quite
main excavation area. However, the overall composition of the common. It seems that a hard, direct hammerstone was
lithic samples was very similar to the main sample. This demo applied, usually without any preparation. It seems unlikely
nstrates that the latter is indeed representative of the assem that these characteristics simply reflect the constraints of the
blage as a whole, rather than reflecting intra-site variability. locally available raw materials.
The lithic assemblage from the main excavation area comp Typologically, the assemblage is quite homogeneous with
rises 3 045 artefacts, and includes 115 cores, 389 tools and very few obviously intrusive elements. Among the latter we
2 541 debitage items, excluding chips, which were weighed can perhaps include two sickle blade fragments, one bearing
but not counted (table 2). Most of the raw material is flint cob hafting signs, and some of the non-geometric microliths : two
bles and some pebbles of mediocre quality (cherty, non-homog truncated bladelets, two backed and truncated bladelets, 10
eneous, and often with numerous limestone veins), which backed broken bladelets and six micro-awls (these are
appear to be of local origin, frequently exhibiting a light grey included within the type list described below)14.
patina. Many artefacts are covered by carbonate encrustations
(probably deriving from the natural sandy sediment).
The overall impression provided by both the debitage and
CORES (figs 4-7) tools, technologically speaking, is of a coarsely knapped
The 115 cores include 14 broken items. Details are pro
vided in tables 3-6. Half of the cores have a single striking
platform, while those with two striking platforms (opposed or
at 90°) comprise 24 % of the total. 70 % of the cores are made
on chunks and 75 % were used to produce flakes, or both
flakes and blades. Over 80 % of the cores retained cortex to
varying degrees, with half retaining 25-75 % (table 5).
Striking platform preparation tends to be minimal and it
seems that the removal of a single primary flake was generally
sufficient (e.g. fig. 4 : 1-2). Indeed, sometimes, when the cor-
Table 2 : Tool and debitage frequencies (not including 3 intrusive
items - see appendix).
n % Category
Tools 386 12.7
Cores 115 3.8
Debitage 2,106 69.2
435 14.3 Debris (chunks)
Total 3,042 99.9
Tool Core ratio 43
Debitage Core ratio 18.3 Tool ratio 5.4
14. All the micro-awls appear on a different and rolled raw material and
they are concentrated around grid square T5, which supports their interpreta
tion as intrusive. Fig. 4 : Nahal Ein Gev I. Single platform cores.
Paléorient. vol. 30/1. p. 25-46 © CNRS EDITIONS 2004 ;
,
A. BELFER-COHEN, A. DAVIDZON, A.N. GOR1NG-MORR1S, D. LlF.BF.RMAN and M. SPIERS 30
tex around the location of the future striking platform was being exhausted, while others were cast away after the
smooth, there was no preparation at all. When cores were appearance of hinges. Some cores bear evidence of multiple
modified on flakes, the ventral face of the flake was used as a blows, usually on the cortex but there are also cores that sub
striking platform (fig. 6:2). Some cores were discarded after sequently were used as hammerstones. On the more standard
ized cores the striking platforms were as wide as possible, Table 3 : Core typology.
even though they were never actually rounded into pyramidal
On On types, i.e. "W-type" as opposed to "N-type" cores15. Thus the Type n % Indet. Chunk Flake backs of the cores always retained cortical cover to facilitate
- Single striking platform 58 50.4 49 9 gripping while knapping. Most of the core types are quite
- 12 Two opposed striking 15 13.0 3 homogeneous and the presence or absence of the characterist
platforms ics detailed above is quite consistent. That the cores are quite 90° - 14 12.1 10 4 striking platforms standardised can be seen also from the metric values and the
- > Two striking platforms j 2.6 2 1 average dimensions of all the cores together which are very
- - 3.4 4 Amorphous 4 similar to those of each particular core type (table 6).
Varia 7 6.0 3 3 1 The Tool/Core ratio (1/3.4) is very similar to the one
- - Broken 14 12.1 14 observed for the first excavation project (1/3.9). It is interest
Total 115 99.6 80 20 15 ing to observe that the average sizes of the cores, as well as
those of the debitage blanks, are too small to produce most of
Fig. 6 : Nahul Ein Gev I. Single platform cores. Note that 2 is on a flake.
Fig. 5 : Nahal Ein Gev I. 1 3, single platform cores ; 2, opposed
platform core. 15. See MARDER. 2002 DAVIDZON and GORING-MORRIS. 2003.
Palcoricnt, vol. 30/1. p. 25-46 r CNRS EDITIONS 2004 :
Nahal Ein Gev I A Late Upper Palaeolithic Site by the Sea of Galilee, Israel 31
Table 4 : Core typology according to blanks produced.
For For For Type n % Mixed Indet Flakes Blades Bladelets
- Single striking platform 58 50.4 25 7 1 25
- - - Two opposed striking platforms 15 13.0 7 8
90° - - - 14 12.1 striking platforms 8 6
- - - - > Two striking platforms 3 2.6 3
- - - Amorphous 4 3.4 3 1
- - - - Varia 6.0 7 7
- - - - 12.1 Broken 14 14
Total 115 99.6 46 7 1 40 21
Table 6 : Core dimensions.
Type n Length Width Thickness
5.1 5.1 Single striking platform 58 3.8
Opposed striking platforms 15 5.8 4.8 3.1
90° 14 5.4 5.0 3.7 striking platforms
> Two striking platforms 3 5.9 5.3 4.0
Amorphous 4 6.0 5.6 4.0
Varia 7 4.8 4.5 2.4
Cores on flakes 17 5.0 5.1 3.3
Cores on chunks 80 5.3 5.0 3.7
Cores for flakes 46 5.3 5.2 3.6
Cores for blades 8 6.3 5.0 4.1
Mixed cores 40 4.8 3.7 5.3
Average Total Cores 101 5.3 5.0 3.6
SD 1.6 1.2 1.1
Range 2.4-13.8 2.4-8.7 1.4-6.9
the tool categories ; this indicates that most cores were thorFig. 7 : Nahal Ein Gev I. Massive opposed platform core (note scale).
oughly exploited before discard. The high cortex coverage
Table 5 : Cortex on cores. retained on most of the cores enables the reconstruction of
their original sizes and, indeed, it seems that most cores could 25- Cortical < 25 % Type 0% 50-75 % furnish at most three to four tools, as indicated by the Tool/ 50% Flake
Core ratio (and see tables 6 and 10). - Single striking platform 19 17 14 8
4 1 Two opposed striking 3 6 1
platforms
90° DEBITAGE 2 4 3 3 2 striking platforms
- - > Two striking 1 1 1
- - - Amorphous 1 3 The debitage comprises 2 106 items, with the flakes being
- Varia 1 1 3 2 by far the predominant category (table 7). The standard devia
Total 7 30 33 18 13 tions for most of the debitage metric attributes tend to be rela-
Paléorient. vol. 30/1. p. 25-46 C' CNRS ÉDITIONS 2004 :
:
:
32 A. BELFER-COHEN, A. DAVIDZON, A.N. GORING-MORRIS, D. LlEBERMAN and M. SPIERS
Table 7 : Detailed frequencies of debitage categories Table 8 : Summary of the metric attributes of the complete debitage items.
Debitage n Type n Length Width Thickness %
291 13.8 no 5.8 4.4 2.0 Primary Elements Primary Elements
Flakes 1.043 49.5 SD 1.5 1.3 0.7
142 3-10 1.5-8.2 0.7-4.1 Blades 6.7 Range
Bladelets 212 10.1 Flakes 308 2.8 2.6 0.7
Core Trimming Elements 204 9.7 SD 1.6 1.0 0.4
Burin Spalls 213 10.1 Range 0.8-8.8 1.0-6.8 0.2-2.8
5.6 Mbt 1 0.0 Blades 35 2.3 1.0
Total 2,106 99.9 SD 1.9 0.8 0.5
Range 2.8-11.0 1.3-4.2 0.4-2.7
2.6 CTE Core 1.8 Bladelets 25 0.8 0.3
Flake Blade 7.3 SD 0.7 0.2 0.2
Flake Blade/let 2.9 Range 1.6-3.8 0.4-1.2 0.1-0.7
4.3 Blades & Bladelets 60 1.7 0.7
2.1 SD 0.9 0.5
Range 1.6-11.0 0.4-4.2 0.1-2.7 tively high (table 8). As with the dimensions of the complete
tools, most blades tend to be rather short and squat, while the
bladelets are short and very thin. Although there is a wide
range of dimensions for the flakes, most of the complete items is a clear preference for flake blanks for tools. Apart from the
are actually quite small. Specific flake debitage types associ non-geometric microlith class, the only other exception is the
ated with burins are described below, together with that tool retouched items category, where the tools made on blades
class. equal those made on flakes (table 9). Of the 372 tools for
Blank production was carried out using a hard hammer- which blanks could be identified, 71 % are made on flakes,
stone. Preparation of the striking platform is nominal, with 18.5 % on blades and 9.6 % on bladelets. 25 % of the tools are
crude blows creating the appropriate angle of removal. In patinated, half are broken, and only a tiny fraction is burnt.
most cases decortication was minimal, and cortex often comp Summaries of the tool measurements by blank type are
letely covers the proximal end16. The bulb of percussion is presented in table 10. The standard deviations are relatively
very prominent, and in some cases, there are two bulbs of per large, with significant differences between the average sizes
cussion, a split bulb of percussion, and the striking platform is of the largest and the smallest items in the assemblage. Nonet
wide and oblique. While the bulb of percussion is not really heless, it seems that the averages do indeed reflect the chunky
prominent, it tends to be diffused over the ventral surface, cre nature of most of the tools, with most blade blanks approachi
ating a sort of a "belly". ng flake proportions. Each of the main tool categories reflects
a quite even distribution of sizes, a matter than can be
explained in several ways. This includes the notion that there TOOLS
is but one single reduction sequence for almost all the tool
classes in the assemblage, with each tool type being modified The 1969 London symposium Upper Palaeolithic type-list
from virtually every kind of blank. Or, alternatively, there was was used to analyse the tool assemblage17. The general tool
specialization focusing on the production of one particular categories are presented in table 9, while the individual tool
tool class, namely the burins, with all the other tool categories, types are detailed in the Appendix. of lesser importance, being modified ad hoc on secondary By far the dominant tool category is that of the burins products deriving from the main reduction sequence. The pro(37 %), followed by retouched items (16 %). In general there nounced morphological variance observed within most tool
categories, contrasting the degree of standardization di1 6. While the "cortex" is sometimes limestone, it is more often the natural
splayed within the burin blanks category, strengthens the latter rolled surface of the original wadi nodule.
17. HOURS, 1974. hypothesis.
Paléoricnt, vol. 30/1, p. 25-46 С CNRS EDITIONS 2004