16 Pages
English

Early Blade Industries in the Levant : The Placement of Douara IV Industry in the context of the Levantine Early Middle Paleolithic - article ; n°1 ; vol.15, pg 215-229

-

Gain access to the library to view online
Learn more

Description

Paléorient - Année 1989 - Volume 15 - Numéro 1 - Pages 215-229
The third excavation at Douara cave in Syria yielded Early Middle Paleolithic assemblages from the geological level IV. The assemblages designated as Douara IV industry are characterized by a large amount of the blade blank production. This paper attempts a technological comparison between the Douara IV and other relevant blade industries such as Pre-Aurignacian, Amudian, Hummalian and Tabun D type of Mousterian. The result has emphasized the technological affinity of Douara IV with Tabun D type of Mousterian, and the author suggests that Douara IV is a regional or local variation of Early Levantine Mousterian blade industry.
Le troisième sondage effectué dans la grotte de Douara en Syrie a livré dans le niveau géologique IV une industrie du début du Paléolithique Moyen. Ces assemblages, appelés industrie Douara IV, se caractérisent par une grande quantité de produits laminaires. On comparera ici l'industrie de Douara IV avec des industries contemporaines : Pré-Aurignacien, Amudien, Hummalien et Moustérien de type Tabun D. Cette confrontation souligne les affinités technologiques de Douara IV avec le Moustérien de type Tabun D et permet de suggérer que Douara IV représente une variation régionale ou localisée de l'industrie laminaire du début du Moustérien Levantin.
15 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 1989
Reads 72
Language English
Document size 3 MB

Yoshihiro Nishiaki
Early Blade Industries in the Levant : The Placement of Douara
IV Industry in the context of the Levantine Early Middle
Paleolithic
In: Paléorient. 1989, Vol. 15 N°1. pp. 215-229.
Abstract
The third excavation at Douara cave in Syria yielded Early Middle Paleolithic assemblages from the geological level IV. The
assemblages designated as Douara IV industry are characterized by a large amount of the blade blank production. This paper
attempts a technological comparison between the Douara IV and other relevant blade industries such as Pre-Aurignacian,
Amudian, Hummalian and Tabun D type of Mousterian. The result has emphasized the technological affinity of Douara IV with
Tabun D type of Mousterian, and the author suggests that Douara IV is a regional or local variation of Early Levantine Mousterian
blade industry.
Résumé
Le troisième sondage effectué dans la grotte de Douara en Syrie a livré dans le niveau géologique IV une industrie du début du
Paléolithique Moyen. Ces assemblages, appelés industrie Douara IV, se caractérisent par une grande quantité de produits
laminaires. On comparera ici l'industrie de Douara IV avec des industries contemporaines : Pré-Aurignacien, Amudien,
Hummalien et Moustérien de type Tabun D. Cette confrontation souligne les affinités technologiques de Douara IV avec le
Moustérien de type Tabun D et permet de suggérer que Douara IV représente une variation régionale ou localisée de l'industrie
laminaire du début du Moustérien Levantin.
Citer ce document / Cite this document :
Nishiaki Yoshihiro. Early Blade Industries in the Levant : The Placement of Douara IV Industry in the context of the Levantine
Early Middle Paleolithic. In: Paléorient. 1989, Vol. 15 N°1. pp. 215-229.
doi : 10.3406/paleo.1989.4497
http://www.persee.fr/web/revues/home/prescript/article/paleo_0153-9345_1989_num_15_1_4497vol. 15/1 - 1989 PALEORIENT,
EARLY BLADE INDUSTRIES IN THE LEVANT :
THE PLACEMENT OF DOUARA IV INDUSTRY IN
THE CONTEXT OF THE LEVANTINE EARLY
MIDDLE PALEOLITHIC
Y. NISHIAKI
ABSTRACT. — The third excavation at Douara cave in Syria yielded Early Middle Paleolithic assemblages from the geological
level IV. The assemblages designated as IV industry are characterized by a large amount of the blade blank production.
This paper attempts a technological comparison between the Douara IV and other relevant blade industries such as Pre-Aurignacian,
Amudian, Hummalian and Tabun D type of Mousterian. The result has emphasized the technological affinity of Douara IV with
Tabun D type of Mousterian, and the author suggests that Douara IV is a regional or local variation of Early Levantine Mousterian
blade industry.
RESUME. - Le troisième sondage effectué dans la grotte de Douara en Syrie a livré dans le niveau géologique IV une industrie
du début du Paléolithique Moyen. Ces assemblages, appelés industrie Douara IV, se caractérisent par une grande quantité de
produits laminaires. On comparera ici l'industrie de Douara IV avec des industries contemporaines : Pré-Aurignacien, Amudien,
Hummalien et Moustérien de type Tabun D. Cette confrontation souligne les affinités technologiques de Douara IV avec le Mous-
térien de type Tabun D et permet de suggérer que Douara IV représente une variation régionale ou localisée de Г industrie laminaire
du début du Levantin.
INTRODUCTION dating techniques such as C-14 and Fission track
ones have been applied to these units. Results of
C-14 technique seem to provide only minimal dates.
The last interglacial and early last glacial period The actual dates for the Middle Paleolithic deposits in the Levant witnessed the appearance of exceptio are probably beyond the limit of the standard C-14 nally varied kinds of stone industries/facies (1). It technique. A fission track date of 75,000 BP, which has been established that the Pre-Aurignacian, Amud is derived from only one sample of burnt barite noian, Hummalian and Tabun D type of Mousterian dule and should be tested by further samples, may were characterized by a strong emphasis on the pro provide some idea about the actual date of unit duction of blade banks. The third excavation season IVB (3). Between unit IIIB and unit IVA is inter-
at the Douara cave in the summer of 1984 produced, bedded an extensive layer consisting of weathered
from its basal Level IV, lithic assemblages with a limestones, which indicates the presence of a potentlarge number of blades, whose stratigraphie position ial time gap. The assemblages from IIIA and IIIB and absolute dates have suggested that the assem are characterized by the dominance of the Levallois blages belong to the period under consideration. This blank production method including multi-directional paper examines the techno-typological relationships core preparation and elaborate platform faceting, of Douara IV industry to other early blade-oriented which are common in the Levantine Mousterian; industries. these units have been assigned to the Levantine
The Douara Cave opens on the southern slope Mousterian (4). On the other hand, Akazawa (5) has
of the Jabal ed-Douara, about 18 km northeast of reported that the IVB assemblage is characterized by
Palmyra, Syria (fig. 1). The cave and its surroun the dominance of prismatic cores and blade blanks
dings have been investigated since 1967 by the To which have uni- and bi-directional dorsal scars and
kyo University Scientific Expedition to Western plain butts. It has been argued that the IVB assem
Asia (2). After the test excavation in 1970, the 1974 blage does not easily match any known Levantine
excavations unearthed a long sequence of stratified Mousterian industry. The original reports (6) have
Middle Paleolithic assemblages from the stratigra stated that the assemblage more closely resembles
phie units IIIA, IIIB, IVA and IVB (fig. 2). Several
(1) COPELAND and HOURS, 1981; JELINEK, 1981, 1982. (3) NISHIMURA, 1979.
(2) SUZUKI and KOBORI, 1970; SUZUKI and TAKAI, 1973, (4) AKAZAWA, 1976, 1979.
1974; HANIHARA and SAKAGUCHI, 1978; HANIHARA and 165. (5) Ibid.; AKAZAWA and HANIHARA, 1983
AKAZAWA, 1979a, 1983; AKAZAWA and SAKAGUCHI, 1987. (6) Ibid.
215 :
:
:
:
:
STRATIGRAPHIC LITHIC DATES EXCAVATIONS INDUSTRIES UNITS (YEARS B.P.)
NEOLITHIC
TO PRESENT
HA
EPI
PALEOLITHIC I IB
1974 IDA SEASON
46. 700 (C-14) 1MB >53.800(C-14)
MIDDLE WEATHERED
LIMESTONES PALEOLITHIC
IVA ■(FEW FLINTS) SYRIA 38.900 C-14) >43.200 C-14 IVB > 52.000 UPPER F.T. 75,000
IVB LOWER >52.OOO(C14)
1984
SEASON IVC
IVD
FIG. 2. - Schematic sequence of stratigraphie units and ar-
cheological levels in the Douara Cave.
JORDAN
collection from the 1984 season's Douara IV (IVB
and IVC), and to compare it with other relevant in
dustries. The industries chosen here to be compared
with Douara IV are those which appeared during the
last interglacial and the early last glacial periods and
100 km are characterized by high blade-indices : Pre-Auri
gnacian, Amudian, Hummalian and Tabun D type
Levantine mousterian. This paper first presents the FIG. 1. - The location of Douara Cave and other relevant sites technological data on the 1984 season's collection referred to in the text.
of Douara IV, and reviews recently accumulated data 1 : Douara, 2 : Jerf Ajla, 3 : El-Kowm, 4 Yabrud, 5 : Ksar Akil,
6 : Bezez, 7 : Abri Zumoffen, 8 : Zuttiyeh, 9 Tabun, 10 Abou on the other blade-oriented industries to determine
Sif, 11 : Rosh Ein Мог, 12 Boker Tachtit, 13 Nahal Aqev. their technological relationships to Douara IV.
the Pre-Aurignacian than the proper Levantine Mous-
terian, while Copeland (7) has attributed the IVB a
MATERIAL ssemblage to a particular faciès of the Early Levantine
Mousterian. She considers that the IVB assemblages
represent an independent cultural group distributed
The material analyzed in this paper are the flint in the northern inland Levant.
artifacts from the lower part of unit IVB and unit The excavations of the 1984 season explored the IVC. The underlying unit IVD, which was not fully deposits of the lower part of the unit IVB and the investigated in the 1984 season, contains too few arunderlying IVC and IVD, and unearthed lithic a tifacts to be adequately analyzed. These stratigraphie ssemblages including a large number of blades. The units have been defined in accord with their charactselective samples of the assemblages have been des eristic sedimentological features such as color, hocribed in a previous report of the author (8). The mogeneity and hardness (9). Both deposits of lower objective of this paper is to analyze the complete units IVB and unit IVC are found in 8 squares within
11 squares (1 m x 1 m) excavated in the 1984 sea-
(7) COPELAND, 1981 : 256, 258; She has tentively called it
Douaran. (9) HIRAI, 1987; The division of the upper and the lower of
(8) NISHIAKI, 1987. unit IVB is only based on vertical differences.
216 :
and are about 15-25 cm thick. The total number the group of Pre-Aurignacian/Amudian/Hummalian son,
of flints from these deposits are 614 and 388 res (PAH group), and the other is that of Early Levantine
pectively, most of which are very small debitage ob Mousterian. The PAH group has close stratigraphie
relationships with the Yabrudian and the Late Acheu- tained by sieving (10). The relatively small amount
of flints is partly due to the limited excavation area, lean, and precedes the Levantine Mousterian (12).
and is partly due to low occupational intensity in The PAH group consists of assemblages from Yabrud
the cave. The discussion in this paper primarily 1:15,13 (13), Tabun E (14), Abri Zumoffen (15) and
concentrates on the technological aspects of flakes Hummal la (16), and the Early Levantine Mouster
and blades (over 3 cm), since cores and retouched ian is mainly composed of Tabun D type assem
tools were found in extremely small number. What blages as defined by Copeland (17) : Tabun D (18),
can be said with some certainty about the latter is Bezez В (19), Abou Sif В, С (20), Laribka (21),
that both assemblages of IVB and IVC contain sin Rosh Ein Mor (22), Nahal Aqev 3 (23), Hum
gle- and opposed-platform cores and a small number mal II (24), Yabrud 1:8-10 (25), Zuttiyeh (26), Ksar
of retouched tools (figs. 3,4). The sample size is ve Akil (27) and Jerf Ajla F (28). The two groups share
ry small. But the sample was obtained by well-cont a general trend in blank production technique and
rolled excavations and is the best material now both display a strong tendency toward blade product
available on the Douara IV (11). ion as in Douara IV. The following part of this sec
tion discusses the technological comparisons
between Douara IVB and IVC and the two groups.
The available data is scarce and the are DATA
mostly based on the published quantitative data.
The inventory of assemblages of IVB (lower)
and IVC is shown in Table 1. The apparent diffe 1) Cresting technique rences between the two units, i.e., the presence of
more cores and retouched flakes and blades in IVC
In the PAH group of assemblages, Yabrud may reflect some biased distribution of these arti
1:15 (29) and Hummal la (30) have the cresting facts in the limited excavation area. In fact, the technique for the preparation of their blade cores. technological traits of flakes and blades, which are
The crested blades presented in the reports of the summarized in Tables 2 to 8, are strikingly similar
two sites are typical, although rather robust when between the two assemblages in every attribute. Both
compared with those common to the Upper Paleolitassemblages have a strong tendency for blade blank
hic. The Amudian levels of Abri Zumoffen and Taproduction using uni- and bi-directional cores, some
bun have not been reported to contain crested blakes. of which have clearly non-Levallois features
However, one atypical piece has been found in the (fig. 3:1). Typical Levallois flakes and points are
Beach Industry of Abri Zumoffen which is considefew in number (Table 2). The characteristics of the
red to be equal to the Amudian in its chronological blades and flakes detached correspond to those of
position (31). cores; their dorsal scar patterns are strongly uni- and
bi-directional (Table 3). Platform preparation is In contrast to the PAH group, most of the a
usually made with coarse faceting and with very oc ssemblages of Early Levantine Mousterian group have
casional exterior platform trimming (Tables 4 and 5). not been reported to include crested blades except
The percussion points are chosen behind ridges on
the flaking surface of the cores in many cases, but (12) Amudian is interpreted as a facies of the Mugharan traare also often between two ridges (fig. 5; Table 6). dition by JELINEK (1981, 1982).
The forms of blades and flakes are mostly elongated (13) RUST, 1950; BAKDACH, 1982.
and relatively thick (Tables 7 and 8). These data (14) JELINEK, 1975.
(15) COPELAND, 1983a. strongly indicate that the IVB and IVC assemblages
(16) BESANÇON et al., 1981; HOURS, 1982; BERGMAN are derived from the same industry or Douara IV, and OHNUMA, 1983. and provide a sound basis on which we discuss the (17) COPELAND, 1975, 1981. technological aspects of Douara IV. (18) JELINEK, 1982.
(19)1983a.
(20) NEUVILLE, 1951; SKINNER, 1965.
(21) VANDERMEERSCH, 1966.
COMPARISON (22) CREW, 1976.
(23) MUNDAY, 1977.
(24) BESANÇON et al., 1981.
Two groups of blade industries occurred during (25) RUST, 1950; BAKDACH, 1982.
(26) GISIS and BAR-YOSEF, 1974. the early Upper Pleistocene of the Levant. One is
(27) MARKS and VOLGMAN, 1986.
(28) SCHROEDER, 1969; The unit F is equivalent to White
1 of the Coon's original excavation.
(10) NISHIAKI, 1987 : Table 5.3. (29) BAKDACH, 1982 : 24.
(11) The artifacts are mapped in a three dimentional coordinate (30) BERGMAN and OHNUMA, 1983.
system, AKAZAWA et al, 1987. (31) COPELAND, 1983a 255, pi. Z.7.
217 FIG. 3. - Stone artifacts of the Douara IV assemblages of the 1984 season's excavation. 3,6,7 : IVB, 1,2,5 : IVC, 4 : IVD.
218 :
TABLE 1 for Bezez В (32) and Yabrud 1:9 (33). The presence Inventory of stone artifacts from Douara IVB and IVC of crested blades in the Early Levantine Mousterian
group seems not to be common (34). A careful exa
mination of the Ďouara IV collection of 1984 season Cores 1 Core Fragments* 1 did not reveal reliable crested blades (35). Flakes Retouched Flakes and Blades 3 6 and blades retaining their core-edges do occur and Unretouched and Blades (over 3cm) 90 45
7 of these were found from IVB and 2 in IVC. Howe * including core rejuvenation flakes
ver, they are not considered to be produced by the
technique of cresting (36).
TABLE 2
Blank types of flakes and blades from Douara IVB and IVC
2) Blades (Tables 7 and 9)
Levallois Prismatic Non-Levallois Total Flakes Points Blades Flakes Blades (%) The tendency for blade blank production can be IVB (n = 71 ) 4.2 5.6 32.4 33.8 23.9 99.9 IVC (n=42) 9.5 compared by using the Bordes' blade index or ILam 9.5 26.2 26.1 28.6
(Table 9). Both groups of assemblages are charac
terized by moderately high indices. While the figures
of the index for PAH group are concentrated at
TABLE 3 around 50, those for the Early Levantine Mousterian Dorsal scar patterns on flakes and blades from Douara IVB group vary considerably from 19.5 (Rosh Ein Mor) and IVC to 76.2 (Tabun IX); it should be noted that the de
uni-directional bi- multi- others Total finition of a blade at Tabun is different from other parallelconvergentdirectionaldireçtional (%) sites (37). If the index at Tabun is excluded from 23.9 25.4 25.4 5.6 100.0 IVB(n=71) 19.7 this comparison, the maximum is 65.5 for Jerf Ajla IVC(n=37) 27.0 27.0 21.6 21.6 2.7 99.9
F. The ILams for Douara IV are very close to that
for Jerf Ajla F and are placed into the highest range
of the Tabun D group.
TABLE 4
Butt Types of flakes and blades from Douara IVB and IVC
3) Levallois products (Tables 2 and 9)
Faceted Dihedral Plain Cortex Others Total(%)
26.1 2.9 8.7 100.0 IVB (n=69) 40.6 20.7 There are definitely higher percentages of Le 43.2 10.8 40.5 0.0 5.4 IVC (n=37) vallois products (IL) in the Levantine Mousterian a
ssemblages (Table 9). The percentages are very low
in PAH group. The Pre-Aurignacian assemblage of
Yabrud 1:15 was reported to be composed of only TABLE 5
non-Levallois artifacts (38). On the other hand, the Exterior platform trimming for flakes and blades from Douar
a IVB and IVC Levallois indices are high in the Early Levantine
Mousterian assemblages; however, they show a con present absent Total(%) siderable inter-assemblage variability. The figures IVB (n=69) 2.9 97.1 100.0 are between 8.7 (Nahal Aqev) and 85.0 (Jerf Ajla F) IVC (n=37) 2.7 97.3
to 90.0 (? Larikba). The causes of this variability
probably includes several factors. One relates to the
functional differences among sites such as open-air
sites versus cave sites (39). Other factors probably TABLE 6
Types of exterior platform shapes of flakes and blades from include the different methods of sample retentions
Douara IVB and IVC and the different definitions of Levallois among clas
sifiers. The latter is especially the case for the pro concave convex others Total symmetric asymmetric (%) blematic Levallois blades in the Early Levantine
IVB (n»69) 39.1 21.7 26.1 13.0 99.9 Mousterian (40). These factors, consequently, make IVC (n=37) 37.8 21.6 32.4 8.1
(32) ibid. 290.
(33) BAKDACH, 1982 : Tafel 74.
(34) The terminal Levantine Mousterian which is also consi TABLE 7 dered to be Tabun D type at Boker Tachtit I utilizes a typical cres Length/Width ratios of flakes and blades from Douara IVB ting technique, MARKS and VOLGMAN, 1983. and IVC (35) The 1974 season's collection also has not been reported
to contain crested blades, AKAZAWA, 1979. <1.00 <1.50 <2.00 <2.50 <3.00 <3.50 <4.00 Total(%)
(36) See TIXIER et al., 1980 : 82-83. IVB(n-71) 7.0 12.7 18.3 23.9 14.1 11.3 9.9 2.8 100.0 (37) JELINEK, 1982 : 92. IVC (n-37) 10.8 16.2 16.2 27.0 13.5 10.8 2.7 2.7 99.9
(38) BORDES, 1955. median; IVB: 2.19; IVC: 2.18 (39) GILEAD and GRIGSON, 1984 : 74.
(40) COPELAND, 1983b; MARKS and VOLGMAN, 1983.
219 1
:
:
I
'
:
TABLE 8 exceedingly high proportions of blades, which is the Midpoint-Width/Midpoint-Thickness ratios of flakes and same feature as that of Jerf Ajla F. Frequencies of blades from Douara IVB and IVC Levallois points are relatively low in Douara IV and
< 1. 00 < 2 .00 < 3.00 < 4. 00 < 5. 00 < Jerf Ajla F, and Yabrud 1:10. Many of the Levallois Total(%)
points of Douara IV are elongated and appear not IVB (n»71) 4 0, .0 25.4 4 33 .8 14 .1 100.1 1. 25. IVC (n»37) 7 2, .7 21. .6 6 99.9 2. 10.8 40. 5 21. to be produced by the classic Levallois method des
IVB: IVC: 3. median; :3.75; 88 cribed by F. Bordes (41). The flaking order of dorsal
scars of the points indicates that many are actually
pointed «series blades» (42).
TABLE 9
Bordes' technological Indices for Levantine Early Middle Pa
leolithic assemblages characterized by high Blade-Indices 4) Dorsal scar pattern (Table 3)
IL IF IFs ILara References
21 _ Hummal la 6. 9 52. Copeland 1985 37. Both PAH and Early Levantine Mousterian ,6A .6 _ Tabun ХЦВ75-1 ) 25. .2 Jelinek 1975 5. 0 49. - Abri Zumof f en TA-1 9 2 1983 groups are known to show a strong tendency for uni- 6. 0 26. 7 45. - TA-3 ,3 22 .6 48 .4 11. ibid. Bakdach and bi-directional flaking. Centripetal scar-patterns Yabrud 1:15 .9 982 47.
such as those reflecting the classic Levallois type of Douara IVB lower 42. ,3 40. 6 62. .0 62. •2\ ,2J IVC 45 54 1.1 43 .2 61. .9 flaking are not frequently reported. Some quantita76.2' 56. 3J 61.4 Tabun IX 48.4 Jelinek 1982 tive analyses, however, have established the presence _ Jerf Ajla F 85.0 43.8 65.0 Schroeder 1969 - Abou Sif В 44.9 52.1 Skinner 1965 51.3 of inter-assemblage variability within the flaking ten_ С 34.2 38.4 57.1 ibid. dency. For example, on the basis of his examination Larlkba 90.0 ? 55.3 43.9 51.7 Vandermeersch 1966 - - - Hummal II 47.1 Besancon et al.1 981 of Levallois flakes, Crew (43) demonstrated the Yabrud 1:8 47 78.6 68.7 37.1 Bordes 1955 9 32.2 64.4 55.8 49.8 ibid. overwhelming uni-directionalism in the Abou Sif a10 27.5 68.8 58.8 23.1 Bezez B(G44 1 D44) 70.8 60.1 53.5 30.7 Copeland 1983 ssemblages, and the rather frequent occurrence of Ksar Akil XXVIIIA 11 .8 57.2 23.7 Marks and 78.8 XXVIIIB 13.5 44.6 27.6 Volgraan 1986 67.5 multi-directional flaking in Bezez B. Marks and Nahal Aqev 3 43.1 8.7 57.4 25.3 Munday 1977 Rosh Ein Mor Volgman (44) have reported that 84 % of 581 Le14.6 54.6 34.7 19.5 Crew 1976
vallois blades of Rosh Ein Mor have uni-directional 1 )only Levallois flakes and points 2) calculated with width at midpoint of length dorsal scar patterns. More than 80 % of flakes and 3) Levallois flakes, points and prismatic blades
blades in Abou Sif В and С collection (about 120
pieces) I examined at Paris have uni-directional dor
TABLE 10 sal scars. My examination of the Bezez В collection Proportions of the Levallois components in Early Levantine (about 600 pieces) at Cambridge has shown that the Mousterian assemblages unidirectional patterns occurred on about 50 % of
flakes and blades, and some 30 % of them have mulLevallois (%) Total References Flakes Points Blades (#) ti-directional patterns. Douara IV assemblages seem
Abou sif В 38 * 12 50. 0 158 Jelinek 1982 to be distinguished by higher frequency of the bi35* С 2 63. 1 141 ibid. Ksar Akil XXVIIIA 47 .6 43. 8 8 .6 105 Marks and directional pattern. The pattern occupies 26.8 % in XXVIIIB 46 .6 41 . 4 12 .1 116 Volgman 1 986 Nahal Aqev 3 IVB, 21.6 % in IVC (Table 3), and 40.7 % in the 28 .8 42. 9 28 .4 455 Munday 1977 .1* Tabun IX 23 .5 34. 4 42 655 Jelinek 1982 upper part of IVB in the 1974 season's collecRosh Ein Mor 34. 0 5235 Crew 1976 Larikba 27. 3 88 1982 tion (45). These values are higher than ca. 10 % or Bezez В 52 .9 21 . 3 25 .8 240 Copeland 1983 Yabrud 1:10 18. 2 143 Skinner 1965 less in both Abou Sif В and C, ca. 10 % in the Bezez Jerf Ajla F 29 .4 5. 9 64 .7 34 Schroeder 1 969 B, and ca. 8 % in the Nahal Aqev 3 (46). Douara IVB upper 8. .2 110 Akazawa 1979 lower .7* 10 .0 13. 3 76 30 .9* IVC 21 .1 21 . 1 57 19
* prismatic blades (standardized blades) 5) Platform faceting (Tables 4 and 9)
difficult adequate comparisons of the assemblages The Bordes' IF and IFs (faceting and restricted with the Levallois indices. The Douara IV collection faceting indices) are significantly different between certainly includes some artifacts showing the same the two groups (Table 9). All the assemblages of morphological character with those which have been PAH group show a low frequency of faceted butts described as Levallois by other authors (fig. 4). The of flakes, and the highest IF from Hummal la equals percentages of products, including prismat to the lowest one of Early Levantine Mousterian ic blades, are shown in Table 9. group of assemblages. The IF and IFs for Douara,
Table 10 lists proportions of the Levallois
components in the Early Levantine Mousterian a (41) BORDES, 1961.
ssemblages; again, there is considerable variability (42) COPELAND, op. cit., see points of Type D in AZOURY
1986 87. among the assemblages. Levallois points account for
(43) CREW, 1975. more than half of the products in Abou Sif {Щ MARKS and VOLGMAN, 1987. В and C, while Levallois flakes dominate in Bezez (45) AKAZAWA, 1979 16, Table 8 4. B. Both Douara IVB ad IVC are characterized by (46) MUNDAY, 1977.
220 12
FIG. 4. - Stone artifacts of the Douara IV assemblages 1,2,4,5,7,9,11,13 of the : IVC. 1984 season's excavation. 3,6,8,10,12,14,15 : IVB;
221 :
:
:
:
:
:
:
IVB and one blade with a plain butt in IVC show which are moderately high, fall into the range of the
use of the technique. Early Levantine Mousterian group.
6) Exterior platform shape (fig. 5, Table 6) 8) Flaking Mode
Exterior platform shapes describe relationships The Hummalian blades are reported to have been between the percussion points and the upper surfaces mainly detached by hard hammers (50). «Levallois of cores. The analysis by H. Dibble (47) at Tabun like» or semi-Levallois flakes and blades in the Amudhas demonstrated that the exterior platform shapes ian level of Abri Zumoffen are described as being of flakes and blades changed from the Amudian to flaked by the use of stone hammers (51). Jelinek the Early Levantine Mousterian levels at Tabun. He feels that Amudian blades of Tabun XI were produfound 17.1 % of flakes and blades had butts with ced by stone hammer percussion (52). Cope- concave exterior shapes in the Amudian level and land (53) has stated that flakes and blades of Early 35.6 % in the Tabun D level. The available relevant Levantine Mousterian (her Tabun D type) were evidata from other assemblages are sporadic; 15 % in dently struck off by a stone hammer. According to Hummal la (48), ca. 40 % in both Abou Sif В and these authors, both the PAH and Early Levantine С (Paris collection) and ca. 50 % in Bezez В (Camb Mousterian groups utilize the hard hammer mode. ridge collection). These figures must be skewed, be The flaking mode at Douara was investigated by cause they are calculated for variously selected K. Ohnuma (54), using the method of analysis decollections. They are not contrary to the results from veloped by Ohnuma and Bergman (55). On the basis Tabun, however, and there seems to be a morpholog of examination of the ventral surfaces of flakes and ical difference in exterior platform shapes between blades, he noted a dominance of soft hammer flaking the PAH and Early Levantine Mousterian groups. in Douara IVB and IVC. If based on these results The frequencies at Douara, 39.1 % in IVB and alone, the Douara IV does not match with either 37.8 % in IVC, are similar to those of the Early Le group. Even the terminal Levantine Mousterian at vantine Mousterian group (Table 6). Boker Tachtit I is believed to use a hard hammer
mode (56). Unfortunately, however, the flaking
modes in assemblages other than Douara and Humm
al la have not been submitted to a quantitative ana
lysis using the same standard. Recognizing the
difficulty of determination of flaking mode (57) and
the lack of systematic comparative data, final stat
FIG. 5. -Schematic representation of exterior platform ements must await future examinations.
shapes. A Concave, В Convex, symmetric, С Convex,
asymmetric, D others.
9) Width/Thickness ratio (Table 8)
Jelinek (58) has revealed a chronological change
of relative thickness of flakes and blades reflected 7) Exterior platform trimming (Table 5)
by their width/thickness ration variances during the
Lower to the Middle Paleolithic in the Levant. He The presence of exterior platform trimming
considered that relatively thinner flakes were deta(overhang removal) technique has been reported at
ched in later periods. His scheme seems quite useful Hummal la (49). The proportion of the occurrences
for Tabun and other sites in the Southern Leis low but the technique was undoubtedly used at
Hummal la. The other available descriptions so far vant (59), while it still remains open to question
published do not discuss this technique in other a whether his ratio works well for sites such as Douara
in the Northern and Central Levant. In fact the ressemblages. The Bezez В collection at Cambridge was
sults of the similar analysis at Jerf Ajla, which is a almost without this technique, while at Tabun D it
does occur (fig. 6:3,6). The Paris collection of Abou
Sif В and С has some evidence for this technique (50) Ibid. (fig. 7:7,9,10). The technique seems, to vary in its (51) COPELAND, 1983a 218. use from assemblage to assemblage in the Early Le (52) JELINEK, 1987.
vantine Mousterian. As for the Douara IV, exterior (53)1975 329.
(54) AKAZAWA, 1987 156. platform trimming is not generally used (Table 5).
(55) OHNUMA and BERGMAN, 1982. Only two flakes, with plain and dihedral butts, in (56) MARKS, 1988.
(57)and BERGMAN (1981 : 169) have recognized
the difficulty to distinguish a mode of "soft hammer" stone from
(47) DIBBLE, 1981. that of soft hammer such as a deer antler.
(48) BERGMAN and OHNUMA, 1983. (58) JELINEK, 1981, 1982.
(49) Ibid. (59)1982 : 94, fig. 10.
222 6 5
FIG. 6. - Stone artifacts of the Tabun D assemblage (Cambridge collection).
cave site near Douara, seem to contradict his scheme. DISCUSSIONS
Schroeder (60) has pointed out that there was a trend
toward the production of relatively thicker flakes in
1) Pre-Aurignacian/Amudian/Hummalian and the upper levels of the Middle paleolithic sequence Early Levantine Mousterian at Jerf Ajla. The width/thickness ratio variances for
Douara IVB and IVC are 2.027 and 4.769 respectiand the vely. Unfortunately, these values are probably ske
following Early Levantine Mousterian of Tabun D wed by the small sample size, while the medians for
type are the early blade industries that appeared in IVB and IVC (Table 8), which may be more reliable
the last interglacial and last glacial periods respectstatistics for such a small sample size, fall into the
ively. Both of them are assigned to the Early Middle lowest range of ratio medians for the Early Levantine
Paleolithic in a recently established framework for Mousterian group defined in the Southern Le
Levantine prehistory (62). Several authors have arvant (61). gued about the relationships between them and seem
to have agreed that the Early levantine Mousterian
of Tabun D type is a derivation of the PAH group
(60) SCHROEDER, 1969 : 385.
(61) JELINEK, op. cit. : 81, Table IX. (62) JELINEK, 1981, 1982.
223