INFORMATION TO USERS

INFORMATION TO USERS
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly fi'om the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely afifect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, begiiming at the upper left-hand comer and
continuing from left to right in equal sections with smaU overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zeeb Road, Ann Arbor MI 48106-1346 USA
313/761^700 800/521-0600
A TECHNO-TYPOLOGICAL ANALYSIS OF TOR AL-TAREEQ (WHS 1065)
AN EPIPALEOLITHIC SITE IN WEST-CENTRAL JORDAN
by
Michelle Nanette Stevens
Copyright ® Michelle Nanette Stevens 1996
A Thesis Submitted to the Faculty of the
DEPARTMENT OF ANTHROPOLOGY
In Partial Fulfillment of the Requirements
For the Degree of
MASTER OF ARTS
In the Graduate College
THE UNIVERSITY OF ARIZONA
19 9 5
UMl Number: 1381776
Copyright 1996 byStevens / Mxchelle Nanette
All rights reserved.
UMI Microform 1381776
Copyright 1996, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
2
STATEMENT BY AUTHOR
This thesis has been submitted in partial fulfillment of requirements
for an advanced degree at The University of Arizona and is deposited in
the University Library to be made available to borrowers under the rules
of the Library.
Brief quotations from this thesis are allowable without special
permission, provided that accurate acknowledgment of source is made.
Requests for permission for extended quotation from or reproduction of
this manuscript in whole or in part may be granted by the copyright
holder.
SIGNED:
APPROVAL BY THESIS DIRECTOR
This thesis has been approved on the date shown below:
John W. Olsen
Pjrofessor of Anthropology
m \.
Date
IM
3
ACKNOWLEDGEMENTS
I would like to thank several people who provided assistance and
guidence during my preparation of this thesis. Geoff Clark graciously
allowed me to use the Step C, WHS 1065 lithic collection and to
participate in the 1993 field season of the Wadi Hasa Paleolithic
Project (WHPP). Deborah Olszewski introduced me to the WHPP and was
also very helpful during the early stages of my analyses. I would also
like to thank Mike Neeley for providing me with an advanced copy of the
site report and letting me into the ASU lithics laboratory. Steve Kuhn
was especially helpful providing many insightful and critical comments
on several earlier drafts of this thesis, especially on lithic and
statistical analyses. John Olsen and Carol Kramer were helpful not only
for their careful readings of this thesis but also for academic guidance
during my graduate career. Barbara Mills graciously allowed me to use
space in the archaeology lab and use some of her laboratory equipment,
e.g., calipers and computers. She also provided statistical assistance.
4
TABLE OF CONTENTS
LIST OF FIGURES
5
LIST OF TABLES
6
ABSTRACT
7
CHAPTER 1:
INTRODUCTION
10
History of Research
15
Cultural Sequence and Description
23
Conclusions
39
CHAPTER 2:
TOR AL-TAREEQ AND PLEISTOCENE LAKE HASA
41
Paleoenvironment and Paleolandscape
41
Previous Research
49
Tor al-Tareeq (WHS 1065) - Excavation and Stratigraphy
52
Interpretations
60
CHAPTER 3:
RESEARCH METHODOLOGY
63
Acquistion of Raw Material
64
Core Reduction
65
Manufacture, Use and Discard
67
Sampling Rationale
71
Analysis of Cores and Debitage
74
Analysis of Retouched Tools
77
CHAPTER 4:
LITHIC ANALYSES
79
Debitage
79
Debitage Morphometries
94
Debitage Summary
98
Retouched Tools - Typology and Technology
101
Major Tool Classes - Typology
103
Major Tool Classes - Technology
105
Microliths
128
Microburin Indices
134
5
TABLE OF CONTENTS - ContAnxied
CHAPTER 5:
DISSCUSSION AND CONCLUSIONS
139
Site Formation Processes
13 9
Intra-site Functional Variability
141
Intra-site Variability in Operational Sequences
143
Regional Comparisions of Operational Sequences
148
Conclusions
154
APPENDIX A:
WHS 1065 DEBITAGE ANALYSIS CODING LIST
157
APPENDIX B:
WHS 1065 TOOL AND CORE ANALYSIS CODING LIST
158
REFERENCES
163
6
LIST OF FIGURES
FIGURE 1.1, Major Eipaleolithic sites in the Levant
11-12
FIGURE 1.2, Schematic illustrations of common Epipaleolithic
microliths
27
FIGURE 2.1, Distribution of excavated sites in the Wadi Hasa
drainage basin
42
FIGURE 2.2, Site map Tor al-Tareeq (WHS 1065)
53
FIGURE 2.3, The east profile of Steps B and C
56
FIGURE 4.1, Box plot of core weights in grams by groups
91
FIGURE 4.2, Box plot of the widths of unmodified blade and
bladelet blanks by level
95
FIGURE 4.3, Box plot of the thicknesses of unmodified blade
and bladelet blanks by level
99
FIGURE 4.4, Box plot of the lengths of unmodified blade and
bladelet blanks by level
100
FIGURE 4.5, Box plot of flake tool widths by level
110
FIGURE 4.6, Box plot of flake tool thicknesses by level
Ill
FIGURE 4.7, Box plot of flake tool lengths by level
112
FIGURE 4.8, Box plot of blade and bladelet cool widths by level .... 114
FIGURE 4.9, Box plot of blade and bladelet tool thicknesses
by level
115
FIGURE 4.10, Histograms of blade and bladelet tool widths
for levels C08N-C11N
118
FIGURE 4.11, Histograms of blade and bladelet tool widths
for levels C12N-C15
119
FIGURE 4.12, Histograms of unmodified blade and bladelet blank
widths for levels C08N-C11N
121
FIGURE 4.13, Histograms of unmodified blade and bladelet blank
widths for levels C12N-C15
122
7
LIST OF TABLES
TABLE 1.1, Cultural sequences in the Levant ca. 20,000-10,000 BP .... 24
TABLE 2.1, Correlations of natural and arbitrary levels from Steps B
and C, and Units B and C
55
TABLE 2.2, Radiometric dates from Tor al-Tareeq (WHS 1065)
58
TABLE 4.1, Percentages of completeness categories for flakes,
blades and bladelets by level
80
TABLE 4.2, Percentages of medial and distal fragments classified
as blades and bladelets, and debris by level
80
TABLE 4.3, Debitage and tool percentages by level
81
TABLE 4.4, Ratios and indices of various artifact classes by level .. 81
TABLE 4.5, Percentages of size categories for complete and proximal
flakes, blades, and bladelets by level
84
TABLE 4.S, Percentages of debris size categories by level
84
TABLE 4.7, Percentages of cortex for flakes, blades, bladelets
and debris by level
89
TABLE 4.8, Percentages of cortex for flakes, blades, bladelets
by level
89
TABLE 4.9, Percentages of cortex on cores by level
89
TABLE 4.10, Summary statistics for core weights by groups
89
TABLE 4.11, Percentages of core types by groups
93
TABLE 4.12, Kolmogrorov-Smirnov two-sided probability test results
for blade and bladelet blank widths
97
TABLE 4.13, Percentages of major cool classes by level
104
TABLE 4.14, Row percentages of blank type for major tool classes
by level
107
TABLE 4.15, Pearson chi-square test for independence of blank
type by level and blank type by group for four
major tool classes and all retouched tools
107
TABLE 4.16, Summary statistics for flake tool widths and
thicknesses by level
116
TABLE 4.17, Summary statistics for blade and bladelet tool
widths by level
116
TABLE 4.18, Summary statistics for bladelet tool widths by level ... 116
TABLE 4.19, Kolmogrorov-Smirnov two-sample probability test for
bladelet tool widths by level
116
8
LIST OF TABLES - Continued
TABLE 4.20, Summary statistics for blade and bladelet blank
widths by level
124
TABLE 4.21, Distribution of dorsal flake scars on blade and
bladelet tools by level
124
TABLE 4.22, Distribution of dorsal flake scars on flake tools
by level
124
TABLE 4.23, Percentages of tool platform types by level
127
TABLE 4.24, Percentages of tool retouch types by level
127
TABLE 4.25, Percentages of microliths by level
130
TABLE 4.26, Frequencies of specific microlithic types by level
132
TABLE 4.27, Formulas for microburin indices
136
TABLE 4.28, Percentages of microburins by level
136
TABLE 4.29, Microburin indices by level
136
9
ABSTRACT
A techno-cypological analysis of the chipped stone assemblage from Tor
al-Tareeq (WHS 1065), an Epipaleolithic site in Wadi Hasa, west-central
Jordan, suggests that significant typological and technological changes
occurred during the occupation of this site.
The lowest levels have
reliable radiocarbon dates (ca. 17,000-16,000 BP) and are associated
with very narrow, backed microliths, single platform bladelet and multiplatform flake and blade cores, and use of the microburin technique.
The overlying, undated levels are associated with wide, short, geometric
microliths, bi- and multi-directional flake and blade cores, and absence
of the microburin technique.
These technological and typological
changes, associated with decreased mobility and moister climatic
conditions in the upper levels, were not synchronous.
The trend towards
the manufacture of wide bladelet tools occurred before significantly
wider bladelet blanks were being manufactured.
The techno-typological
characteristics of these assemblages resemble roughly contemporary sites
in the Azraq Basin, northeastern Jordan.
10
CHAPTER 1 :
INTRODUCTION
Current researchers use microlithic techno-typological variability
as the main criterion for distinguishing Epipaleolithic cultures in the
Levant (ca. 20,000/18,000-10,000 BP) (e.g., Bar-Yosef 1991; Byrd 1994;
Goring-Morris 1987, 1995; Henry 1989b).
Although the identification of
prehistoric cultures should rely on multiple artifact classes and lines
of evidence, microlithic tools are abundant and considered to be the
most temporally and spatially sensitive artifact class (Bar-Yosef 1981,
1991; Henry 1982, 1989b).
However, researchers do not necessarily agree
on whether the level of observed variability represents culture groups,
subgroups within a single culture, or environmental adaptations of one
or more culture groups.
In order to help clarify some of these issues, additional technotypological studies and studies of operational sequences used to
manufacture chipped stone assemblages (i.e., raw material acquisition,
core reduction, manufacture, use and discard of lithic artifacts) at
multi-component Epipaleolithic sites in a variety of environmental and
geographic contexts are necessary.
The assumption behind these studies
is that techno-typological and operational sequence approaches to lithic
analysis have the potential to reflect, in conjunction with other data,
different prehistoric culture groups, subsistence strategies, and land
use patterns.
If lithic assemblages from a variety of environmental and
geographic contexts are analyzed, these studies will help fill spatial
and temporal gaps in the archaeological record and allow more plausible
interpretations of culture variability and change to be postulated.
The present study is a techno-typological analysis of a sample of
chipped stone from Tor al-Tareeq (WHS 1065), an early to middle
Epipaleolithic site situated in upper Wadi Hasa, west-central Jordan
(Figure 1.1).
Earlier studies and in-field analyses of the lithic
FIGURE 1.1: Major Epipaleolithic sites in the Levant.
(1) El-Kowm
(2) Yabrud III
(3) Ksar Akil
(4) Azraq Basin: Wadi Uwaynid 18, Uwaynid 14, Wadi Jilat 6
(5) Wadi Hasa: Tor al Tareeq (WHS 1065)
(6) Petra area: Beidha, Wadi Madamagh
(7) Wadi Hisma/Wadi Judayid sites
(8) Ain Mallaha
(9) Hayonim
(10) Kebara Cave, El-Wad, Nahal Oren
(11) Shiokbah
(12) Jericho
(13) Neuville's sites (1934, 1951)
(14) Negev sites
12
l&yiy ^v
0
Palmyra -7
Beirut
Damascus
Mediterranean
«A
Druze
Tel Aviv
El-Jafr
13
assemblage from this site indicate that t^'pclcgical =md technological
changes occurred in the early to middle Epipaleolithic deposits, i.e.,
natural levels 5 and 7, Step C (see Figures 2.2 and 2.3, and Teible 2.1)
(Clark et al. 1987, 1988; Donaldson 1986; Donaldson and Clark 198S).
I
wanted to study the technological and typological variability associated
with this transition (i.e., from a tool assemblage dominated by
nongeometric microliths cuid use of the microburin technique to a tool
assemblage with high proportions of wide geometric microliths and an
almost complete absence of the microburin technique) in order to better
understand why these techno-typological changes occurred.
Particularly,
I was interested in determining to what extent these technological and
typological changes reflect only diachronic change or a combination of
diachronic change, and changes in subsistence and land use strategies in
response to paleoenvironmental change.
Also since techno-typological characteristics and operational
sequences are frequently used to identify cultural variability in the
Epipaleolithic, determining the synchrony of technological and
typological change should enable a better understanding of the nature of
culture change and variability, as reflected in chipped stone
assemblages.
If different cultural groups are present in the northern,
southern, eastern and western portions of the Levant as some suggest.
Tor al-Tareeq is geographically positioned such that the lithic
assemblage from the site may be able inform on the applicability of
these cultural groupings.
Several aspects of this site have been previously reported, e.g.,
preliminary lithic analyses (Clark et al. 1987, 1988; Donaldson 1986;
Donaldson and Clark 1986), surface site structure (Coinman et al.
1989:213-236), paleogeography and paleoenvironment (Schuldenrein and
Clark 1994), and a general site description (Neeley et al. 1995) .
This
14
Study is a reanalysis of the eight lowest excavation levels in
excavation unit "Step C" (see Figures 2.2 and 2.3, and Table 2.1).
Earlier studies (i.e., Donaldson 198S) relied on in-field analyses of
chipped stone debitage and retouched pieces.
In-field analyses were
conducted by several crew members with varying degrees of lithic
expertise.
Even though additional analyses were conducted on the
chipped stone tools after the excavation season had concluded, in-field
analyses did not identify all the tools and special debitage classes
like microburins, within the assemblage.
Therefore, a reanalysis of the
debitage and tool components from this excavation unit is called for.
This study supplements and complements previous research at this
site by (1) identifying debitage and tool categories that may have been
overlooked or misidentified during in-field analyses; and (2) re­
analyzing the techno-typological variability in the debitage and tool
categories in the lowest levels of an excavation unit where previous
analyses indicated typological and technological changes in the chipped
stone assemblage occurred.
These techno-typological changes may be
associated with a transition from an earlier Kebaran component to a
later Geometric Kebaran component (Clark et al. 1987, 1988; Coinman et
al. 1989; Neeley et al. 1995).
As little is currently known about this
transition in west-central Jordan, this analysis will provide
information on lithic techno-typological variability and the operational
sequences used to manufacture chipped stone tools in west-central Jordan
during these periods.
Tor al-Tareeq is particularly important for this
because several reliable chronometric dates have been obtained from the
lowest excavation levels (Clark et al. 1987, 1988).
This opening chapter presents a general overview of Epipaleolithic
research in the Levant.
The general Levantine cultural sequence and the
characteristic attributes used to define major cultural divisions are
15
discussed.
Chapter 2 is a presentation of the paleogeography and
paleoenvironment of the Levant and the upper Wadi Hasa in order to
position Tor al-Tareeq (WHS 1065) and the Wadi Hasa in their regional
and temporal contexts.
Also, previous research at Tor al-Tareeq (WHS
1065) including stratigraphic interpretations and relationships, and
lithic analyses will be discussed.
In chapter 3, a methodological
discussion including the sampling rationale, and typological and
technological approaches to this analysis is presented.
Chapter 4
presents this study's analyses and interpretations of the debitage and
tool components of the chipped stone.
Finally, this paper concludes
with a comparative discussion and interpretation of this site, and the
implications of this research at local and regional scales.
History of Research
The Epipaleolithic period in the Levant dates between ca.
20,000/18,000 and ca. 10,000 BP.
The Levant is defined here as the area
encompassing the modem political states of Israel and Jordan.
This
region is divided into western and eastern areas by the Wadi Araba and
Jordan River, both of which lie in the Rift Valley.
Although some
include the modem political regions of Lebanon and Syria in definitions
of the Levant (e.g., Goring-Morris 1995), there is considerably less
research on the Epipaleolithic in these areas.
Therefore, the
discussion of the Levantine Epipaleolithic presented here will not focus
on these northern regions.
Prior to 1950, research on the Epipaleolithic in the Levant was
dominated by a culture-historical approach in which researchers were
primarily interested in establishing chronological sequences through
artifact seriations.
This early phase of field work was concentrated at
cave sites in the Mediterranean vegetation zone where deep cultural
deposits were thought to be present.
Many of these sites such as Kebara
16
Cave (Garrod 1954; Turville-Petre 1932), al-Watwat, an Nugtah, Wadi
Fallah, and Abu-Usba Caves (Stekelis and Haas 1942, 1952), and Ksar Akil
in Lebanon (Ewing 1949) were located in coastal areas (Figure 1.1).
However, Shukbah Cave in the Carmel Mountains (Garrod 1942) and several
sites in the Judean Hills (Neuville 1934, 1951) were located further
inland, but still far to the east of the present study area.
The transitional nature of the Epipaleolithic was recognized early
on by Dorothy Garrod who first identified the Natufian culture at Wadi
el-Natuf (Garrod 1932).
Influenced by western European concepts and
general cultural sequences, Garrod used the term "Mesolithic" for the
transition between the Upper Palaeolithic and Neolithic periods (1932,
1937; c.f. Lubbock 1865).
She recognized the period by the appearance
of backed and retouched microliths.
Neuville (1951) who was working at
Kebaran sites termed the period Epipaleolithic.
The term Mesolithic,
borrowed from western European archaeologists, was used to denote a
closer association with the Neolithic, while the term Epipaleolithic
reflected a North African bias and implied closer affinities with the
local Levantine Upper Paleolithic sequence (Phase IV) (Neuville 1951).
Since Levantine microlithic assemblages did not have any clear
association either chronologically or culturally with the western
European sequence but did exhibit similarities with microlithic
assemblages identified in North Africa (Perrot 1966; Tixier 1963), the
designation Epipaleolithic eventually replaced the term Mesolithic.
Garrod and Neuville, the most prominent researchers at that time,
were relatively progressive in their methodological and analytical
approaches when compared to their contemporaries.
They employed the
standard archaeological approaches of artifact seriation and
stratigraphy to construct regional culture histories.
However lonlike
other excavators, they supplemented these approaches with stratigraphic.
17
geologic, and faunal studies in order to reconstrijct climatological and
environmental sequences (Henry 1989b:7).
In one study, the relative
proportions of Persian fallow deer {Dawa mesopotamica) and gazelle were
used to reconstruct past climatic conditions (Garrod and Bate 1937).
Although these studies were innovative, they focused on cultural and
environmental sequencing.
Little attempt was made to apply their
climatic and environmental data to test or to explain any cultural or
environmental theory.
Although Childe's (1939) hypothesis for the
origins of agriculture was published shortly after most of these early
reports (Henry 1989b:7), none of the researchers reevaluated their data
to support or refute Childe's hypothesis.
Although data on this early
research had already been published, Childe did not incorporate the
climatological and environmental sequencing data collected by Garrod and
Neuville into his hypothesis (Henry 1989b;7).
Thus, there were two
parallel tracks of Epipaleolithic research, a theoretical approach
focusing on the origins of agriculture and a field oriented culturehistorical approach.
The culture-historical approach emphasizing data collection
continued during the 1950s to early 1960s but new geographical and
environmental situations were explored.
In the western Levant, several
important open air sites dating to the late Epipaleolithic to early
Neolithic were investigated.
They include Ain Mallaha (Perrot 1962,
1966), Jericho (Kenyon 1959) , and Nahal Oren (Stekelis and Yizraeli
1963).
Only Perrot (1962), however, attempted to explain the social and
economic significance of his data (Henry 1989b:7).
In the eastern
Levant, the first field work was conducted with Diana Kirkbride's survey
and test excavations at Wadi Madamagh (Kirkbride 1958) and Beidha
(Kirkbride 1966) in southern Jordan.
is present.
Again a late Epipaleolithic bias
The geographic separation of Kirkbride's sites from those
18
investigated, in the western Levant limited comparisons betv/een these tv/c
geographic areas to general stratigraphic and artifact seriation
correlations (Henry 1989b).
As chronometric dating techniques were not yet well developed,
culture-historical approaches were limited to the relative dating
techniques of artifact seriation and stratigraphy.
Out of necessity,
retouched and backed microliths became the "index fossils" for the
entire Epipaleolithic.
However, this approach can be somewhat circular.
Since microliths were believed to be only associated with the
Epipaleolithic, all sites with microliths were automatically assigned an
Epipaleolithic date, regardless of other artifact classes present.
Furthermore, the microlithic component of some assemblages may not have
been recognized, due to poor recovery techniques.
This would be
especially true if microliths comprised only a small proportion of the
chipped stone assemblage.
After the development of the radiocarbon dating technique, it
became apparent that the mere presence of bladelets and retouched and
backed microliths was no longer adequate for distinguishing the
Epipaleolithic from earlier or later periods (Byrd 1994).
Bladelet
blanks and microliths are found at Upper Paleolithic sites dating as
early as 30,000 BP (Bar-Yosef and Belfer-Cohen 1977; Bar-Yosef and
Phillips 1977; Byrd 1994:206; Gilead 1983, 1988; 1991:121-125; Phillips
1994).
Also, several sites that chronometrically date to the
Epipaleolithic may have predominantly non-microlithic assemblages and
industries and have very few retouched or backed microliths (Garrard et
al. 1994; Garrard and Byrd 1992; Gilead 1991; Goring-Morris 1987) or
contain retouched bladelets that resemble those found in Upper
Paleolithic, i.e., pre-20,000 BP, deposits (Byrd 1994:206).
However,
the presence, absence and typological characteristics of microliths were
19
only cfitsirici 3.vail5t]?ls to 2rsss5i2rch.s2rs st
tinis.
still, even after chronometric dating techniques had been
developed, researchers in the mid 1960s to early 1970s continued to rely
heavily on index fossils.
In fact, the sequence based on microlith
forms was further cemented by systematic lithic analyses which
categorized tools based on a formal type list approach, both in the
Levant (e.g., Bar-Yosef 1970; Hours 1976; Perrot 1966) and North Africa
(e.g., Tixier 1963).
Marked changes in research foci and methodologies
became evident in the late 1970s as systematic regional surveys and test
excavations expanded to geographic areas outside the Mediterranean
vegetation zone, especially into the arid Negev and Sinai (Marks 1976,
1977, 1983 ; Phillips and Mintz 1977).
One goal of these projects was to
study diachronic change in regional land use and settlement patterns
taking into account regional geomorphology and its effect on the
visibility of archaeological sites during a given time period.
Extensive geomorphological and environmental studies enabled the
reconstruction of paleolandscapes from which less biased settlement
patterns could be derived.
One major outcome of this research was Marks
and Friedel's (1977) model of radiating versus circulating settlement
patterns.
This model is based on settlement distribution data.
Marks
and Friedel (1977) suggest that a semi-sedentary radiating settlement
pattern characterized the Negev during moist periods, i.e., the Middle
Paleolithic and Natufian; while a circulating pattern involving a higher
degree of mobility was more typical during drier periods, i.e., the
Upper Paleolithic and early Epipaleolithic.
Multi-disciplinary approaches to Paleolithic research were also
conducted in the eastern Levant, e.g., the Wadi Hisma/Wadi Judayid area
in southern Jordan (Henry 1982, 1983, 1987; Henry and Garrard 1988;
Henry et al. 1983), the Upper Wadi Hasa in west-central Jordan (Clark et
20
al. 1987, 1988; MacDonald et al. 1980, 1982, 1983; MacDonald 19SS), the
Azraq Basin in northeastern Jordan (Garrard et al. 1986, 1987, 1988),
and the Palmyra and El-Kowm Basins (Cauvin 1981).
The impetus of
several of these studies was to test Marks' model of radiating versus
circulating settlement patterns and to determine if such land use
strategies were used in different paleoenvironmental contexts.
If these
strategies were used in different paleoenvironmental situations,
researchers were interested in determining to what degree settlement
patterns were influenced by changing climatic conditions.
Extensive
surveys and excavations were also conducted in the western Levant with
several projects associated either directly or indirectly with the
emergency surveys of the Negev and Sinai (e.g., Goring-Morris 1987).
However as the number of microlithic and non-microlithic
assemblages with chronometric dates increased, it became apparent that
documenting regional and temporal typological variability in lithic
assemblages was no longer sufficient for addressing the new research
questions.
To more adequately study regional land use strategies and
settlement patterns, more technological approaches to lithic analyses
were required (Henry 1989b).
Such technological studies focused on
systematically recording the morphometries and types of blanks (e.g.,
Henry 1973; Marks 1983; Olszewski 1989).
In addition, some analyzed
specific blank attributes such as dorsal flake scar patterns, overall
blank shape, and blank curvatures (e.g., Henry 1973, 1977).
This shift in lithic analysis from a typological to a technotypological approach reflected American trends in lithic research.
These trends were in part responses to the debate between Hordes,
Binford, and others regarding the nature of lithic variability in Middle
Paleolithic assemblages of western Europe (Binford and Binford 1966;
Binford 1973, 1983; Bordes 1961, 1973, 1978; Dibble 1984, 1985, 1987;
21
Mellars 1969; Holland 1977, 1981) .
Due to the reductive nature of
lithic technology, if chipped stone typological variability was
influenced by technological choices made during manufacture and reuse,
any stylistic properties manufactured on the original artifact may have
been altered during use and maintenance (i.e., resharpening) activities
(Neeley and Barton 1994).
In addition, other factors such as the
quality and quantity of raw material, and expediency may have more
influence on artifact foms than stylistic conventions.
Additional
research emphasizing lithic reduction sequences and technology is
necessary if the meaning of formal variability in artifact forms is to
be better understood.
Techno-typological variability in lithic
manufacture can then be related to regional land use and subsistence
strategies, settlement patterns, and economic and social interactions of
Epipaleolithic groups.
These factors provide the building blocks in
understanding changes in social and economic strategies between a mobile
hunter-gatherer lifestyle in the Upper Paleolithic to a sedentary
agricultural lifestyle in the Neolithic.
In an effort to explain both the typological and technological
variability in Epipaleolithic artifact assemblages, numerous taxonomic
classification systems have been developed reflecting a culturehistorical or a time-stratigraphic approach (e.g., Aurenche et al. 1981;
Besancon et al. 1975-7; Byrd 1994:206; Garrard et al. 1994; GoringMorris 1995:141; Moore 1985).
The most widely accepted culture-
historical terms are Kebaran, Geometric Kebaran, Harifian, Mushabian,
and Natufian (Bar-Yosef l991a; Henry 1989b; Valla 1988a) which some
refer to as technocomplexes or cultural complexes (Henry 1989b).
Still,
some researchers are reluctant to apply even these general terms to both
the western and eastern Levant (e.g., Byrd 1994).
However, some sort of
broad Levantine taxonomic scheme is useful in organizing the variability
22
in material culture, especially in the late Bpipaleolithic when there
seems to be increasing regional variability in lithic assemblages.
Much of the variation in nomenclature reflects the interpretive
levels at which these taxonomic schemes are applied.
Classification of
material culture requires selection of arbitrary categories that reflect
varying degrees of technological and typological similarities with other
artifact assemblages (Henry 1989b).
Henry's {1989b) hierarchical
classification scheme for scales of material culture during the
Bpipaleolithic recognizes four levels: complex, industry, phase or
facies, and assemblage.
This hierarchical scheme is based on Clarke
(1968, 1979) but differs from Clarke's in that two independent
classification schemes are used, one for material culture and another
for prehistoric socio-economic data (Henry I989b:80).
This modification
enables material culture to be compared without involving assumptions
regarding the behavior that produced the artifacts (Henry 1989b).
Socio-economic levels (i.e., culture, culture groups and technocomplex)
are based on environmental, economic, demographic, spatial and temporal
evidence, and previously identified material culture units (Henry
1989b:81).
For the Bpipaleolithic, comparisons of material culture are
based on quantitative analyses of chipped stone tools, tool blank
morphology, and microburin indexes.
In the hierarchy of material culture, artifacts are grouped into
an assemblage, the smallest building block, which consists of artifacts
in a cultural deposit believed to be deposited over a single interval,
examples include an occupational horizon at a stratified site or debris
from a limited activity site.
or facies.
Assemblages are combined to form phases
Phases have a very high level of technological and
typological similarity.
Typological similarity includes those
attributes related to form or style and function.
Stylistic attributes
23
should be unrelated to fxinction, and include types of retouch (Henry
1973, 1977, 1989b:83), the position and orientation of retouch on
certain tools (Close 1978), and specific uses of the microburin
technique (Henry 1989b:83; Marks and Simmons 1977).
Industries are
groups of assemblages with very similar "specialized" technological
attributes.
The microburin technique is one example of a "specialized"
technological attribute.
The frequency aind types of major artifact
classes in an assemblage can also be used to reflect core reduction
strategies and other technological choices that can be used to determine
the degree of similarity between assemblages.
An "industry" should
include specific quantitative boundaries, e.g., bladelet width as well
as frequency ranges (Henry 1989b:83).
Industries are grouped together
to form a "complex" based on the rsinge of percentages of blank
production and tool manufacture (Henry 1989b).
A complex, the most
general grouping of material culture, exhibits a "high level of
technological affinity and a comparatively low level of typological
affinity" (Henry 1989b:82).
Cultural Sequence and Description
While there is consideraible continuity in the techno-typologies of
lithic industries throughout the early Epipaleolithic (ca.
20,000/18,000-15,000 BP), some variability in the microlithic
assemblages has been noted.
At the largest scale of complex, four,
early to middle Epipaleolithic complexes are recognized: Kebaran,
Qalkhan, Geometric Kebaran, and Mushabicui.
Several industries have been
suggested within each of these complexes that reflect regional and
temporal variability (Table 1.1).
The Kebaran and Qalkhan Complexes are
roughly contemporaneous and appear to be geographically separate.
The
Kebaran is found predominately in the Mediterranean vegetation zone in
the western Levsint along coastal areas and in upland environments.
This
24
Table 1.1:
Cultural sequences in the Levant ca. 20,000-10,000 BP.
Industry level designations are denoted by complex/industry,
e.g., Kebaran/Early Hanursui. Levcintine regional variants
(industries) are distinguished by vegetation zones and
indicated by (M) - Mediterranecin, (S) - Southern IranoTuranian and Saharo-Arabicin, (E) - Eastern Irano-Turanian
and Saharo-Arabioui, and (EN) - Eastern eUid Northern IranoTuranian and Saharo-Arsibian.
KYA
Climate
Western
Complexes
(Byrd 1994)
Eastern
Complexes
(Byrd
(Henry
1994)
1995)
Levantine
Complexes
(GoringMorris 1995)
10-
Humid
warm
Harifan
11-
Dry
cold
Natufian
Natuf.
eind
related
industries
Final Natuf (M)
Harifian (S)
EciNatuf. (EN)
Late Natufian
Late
Natuf.
Early
Natuf.
1213-
Humid
warm
14-
15-
1
1
1
Dry
cold
1
18-
i
1920Drying
>20
NonMicrolithic
Kebaran
15-
17-
Geom.-jMushabietric 1 an
Kebi
aran {
11
•
Mushabian/
Madam.
Geo.
Keb./
Hamran
Keb­
aran/
E.Hamran
1
Late/Teminal
Upper
Paleolithic
Late Kebaran
Nizzanan
(Parts of M, S,
and E)
• --
1
1
1
i
Early |RaNatuf. 1 monian
1 (S)
1
Geo.
1
metric |MushKebaran| abian
1 (S)
E. Kebaran (M)
Qalkhan (EN)
Nebekian (EN)
>
1
NonNatufian
Microlithic
Qalkhcui
Late/Terminal
Upper
Paleolithic
Masraqan
Late Upper
Paleolithic
25
distribution may in part reflect the relatively dry and cold conditions
at that time.
The Qalkhan Complex (Henry 1995) or the Non-Natufian
Microlithic (Byrd 1994) is foiand in eastern steppic regions, generally
east of the Rift Valley.
Although, it may be present in the northern
Levant as well (Henry 1995).
The Geometric KebarcUi and Mushabian are
roughly contemporaneous complexes.
Both are found in the western and
eastern Levant eind appear to succeed either the Kebarsui or Qalkhan
Complexes depending on the geographic region in which they are foxond.
Brief descriptions of the lithic assemblages, temporal range,
geographical distribution, and some regional industry level variants of
the Late/Terminal Upper Paleolithic Kebaran, Qalkhan, Geometric,
Mushabian Complexes are presented in the following section.
The Late/Terminal Upper Paleolithic (pre-20,000-ca. 16,000 BP)
assemblages encompass considerable lithic variability but are
represented in relatively few (six) archaeological horizons with
reliable chronometric dates (Byrd 1994:208).
Most researchers accept a
terminal date of 20,000 BP for the Late/Terminal Upper Paleolithic
period (Table 1.1).
However in the western Levant, similarities in
lithic tool traditions between Late/Terminal Upper Paleolithic
assemblages and later sites that chronometrically overlap with Kebaran
sites, suggests to some that the Late/Terminal Upper Paleolithic tool
tradition may have continued for another 4,000 years in the western
Levant (Byrd 1994:208).
Generally, these assemblages are characterized
by the presence of some microliths (between 10% and 60% of the total
tool assemblage).
The microliths differ typologically from those found
in contemporaneous Kebaran sites.
Both twisted bladelets typically with
interior lateral retouch, and finely retouched microliths are common,
but generally not in the same assemblage (Byrd 1994:208).
Core reduc­
tion technologies also seem to differ between Kebaran and Late/Terminal
26
Upper Paleolithic assemblages (Byrd 1994; Goring-Morris 1995).
The early Epipaleolithic Kebaran Complex (ca. 19,000-14,500 BP)
was discovered by Turville-Petre in 1931 during his excavations in
Kebara Cave, Mt. Carmel (Turville-Petre 1932), and was later described
by Garrod.
In general, identification of the Kebaran is based on the
high frequency of microliths and the predominaince of certain types of
non-geometric microliths (Bar-Yosef 1975, 1981; see Figure 1.2 for
illustrations of some common Epipaleolithic microliths).
Also, the
microburin technique was infrequently used to segment microliths prior
to backing.
Various subdivisions of the Kebaran have been suggested.
Bar-
Yosef (1970) postulated a scheme that subdivided the Kebaran into four
contemporaneous culture groups.
He has subsequently revised his earlier
groupings, incorporating results from Lebanon (Hours 1976) and more
recent investigations elsewhere (Bar-Yosef and Vogel 1987; Goring-Morris
1995:153).
Group A is represented by narrow micropoints with basal
modifications or truncations smd broad micropoints.
As it is found in
the southern coastal plain, it probably represents a regional variant.
Some suggest that coastal Kebaran sites, many of which are not well
dated, may actually date to the Upper Paleolithic (Byrd 1994:209; Gilead
1991).
The wide distribution of the other three groups B, C, and D
suggests that these groups are not regional variants.
Group B is
represented by curved, backed, and pointed retouched bladelets; some of
which have basal truncations.
Group C contains a combination of narrow
micropoints and obliquely truncated and backed bladelets.
Group D has
obliquely truncated backed bladelets (Kebara points) and narrow, curved
backed bladelets.
Bar-Yosef now recognizes that these groupings
represent both temporal and spatial lithic variability in Kebaran
assemblages.
27
Figure 1.2: Schematic illustrations of common Epipaleolithic microliths
(after Goring-Morris 1987; Henry 1989; Muheisen 1988; Tixier
1963): (a) curved/arched ijacked bladelet, (b) obliquely
truncated and backed, (c) scalene triangle, (d)
trapeze/rectangle, (e) wide trapeze/rectangles, (f)
microgravette, (g)regular microburin, (h) piquant tiedre
microburin, (i) Krukowski microburin, (j) lunate, (k)
Qalkhan point, (1) la Mouillah point.
28
The Kebaran has also been divided into early and late phases based
on the frequency of geometric microliths.
In general, micropoints,
curved or arched backed bladelets, and microgravettes are more common in
the early Kebaran and tend to stratigraphically precede Kebara points
(large, obliquely tnoncated, backed bladelets).
Therefore, in Bar-
Yosef's original scheme (1970), Group C is considered earlier than Group
B.
However, some sites that are dominated by curved micropoints have
relatively late dates which suggests to some that early Kebaran dates
should be extended upwards (Byrd 1994:209; Hovers and Marder 1991).
Although the late Kebaran is dominated by obliquely tr\incated and
backed forms, curved and arched backed microliths still persist in low
numbers in mainy assemblages (Bar-Yosef and Vogel 1987:225) .
A
technological trend is also recognized in a change from fine and inverse
retouch to abrupt and bipolar retouch on the microliths.
It has also
been suggested that the early Kebaran has greater lithic variability
than the late Kebaran {Bar-Yosef 1981:392-393, 1990).
This greater
lithic variability may account for most of the spatial subdivisions of
the Kebaran (Byrd 1994:208).
In general, Kebaran sites are small (usually ca. 25-100 m^ and
rarely larger than 250 m^) and are located in the Mediterranean
vegetation zone (Goring-Morris 1995:153).
During this period, climatic
conditions are considered to be drier and colder than those at present
based on low arboreal pollen frequencies in pollen diagrams throughout
the Levant (Baruch 1994; Bottema 1987) .
In addition, the early
Epipaleolithic and Kebaran Complex are coincident with the last glacial
maximum of the Pleistocene, ca. 22,000-18,000 BP.
As vegetation zones
would have shifted north, southern and eastern Levantine areas would
have been relatively drier and colder, as the Sahara-Arabian vegetation
zone expanded northward.
The steppic, Irano-Turanian zone would have
29
also moved north and may have been somewhat restricted.
The
Mediterranean vegetation zone would also have been restricted, being
found only along coastal areas, the Jordan Valley, and the west-central
and northern Levantine areas.
Essentially no horizontal data exist at Kebaran sites; therefore,
intra-site patterning of features and artifacts is not well studied
(Goring-Morris 1995:153).
Unlike later Epipaleolithic complexes, few
skeletal remains have been recovered from Kebaran sites.
Based on the
presence of bone that appeared to be cremated at Kebara C, it is
suggested that the lack of skeletal remains may be related to cremation
mortuary practices (Goring-Morris 1995:153).
Most sites contain some
marine mollusks, but, generally in low frequencies.
Bone artifacts are
also infrequent, probably due to poor preservation conditions, and
include awls, points, spatulas, and very occasionally, an art object
made on bone (Goring-Morris 1995:153).
Although most researchers use the term Kebaran to refer to early
Epipaleolithic assemblages in the western Levemt, some (e.g., Ferring
1977, 1988; Goring-Morris 1987, 1995; Marks 1976) believe there is
sufficient lithic variability in the earliest Epipaleolithic to warrant
its separation from the Kebaran into the Masraqan industry.
The
Masraqan industry was first described at Masaraq an-Na'aj in the Judean
Desert (Perrot 1955).
It is found throughout the Levant including the
Mediterranean, Irano-Turanian, and Saharo Arabian vegetation zones.
As
indicated in Table 1.1, this industry is coeval with some Late/Terminal
Upper Paleolithic, early Kebaran, Qalkhan or Non-Natufian Microlithic
Complexes.
This industry has been subsumed imder the early Kebaran
Complex in the west, and the Qalkhan or Non-Natufian Microlithic
Complexes in the east.
Generally, Masraqan sites include small sites (ca. 25-250 m^) that
30
have a single hearth and an associated artifact scatter, and larger
sites that have several hearths with each hearth having its own artifact
scatter (Goring-Morris 1995:151).
stones.
Occasionally, hearths are outlined by
Some vegetal remains have been recovered from at least one
Masraqan site, i.e., Ohalo II {Goring-Morris 1995:151).
Gronndstone and
bone points and awls are occasionally present at these sites.
Marine
shells, usually from the Mediterrsinean, are also present, sometimes in
relatively high frec[uencies.
Masraqan lithic assemblages are characterized by elongated, narrow
single platform cores, and evidence of intensive preparatory abrasion,
resulting in carination (Goring-Morris 1995:151).
The core technology
used to manufacture larger blade blanks appears to differ from that used
to manufacture bladelets.
Bladelets are narrow, elongated and thin,
with incurvate, not twisted, profiles (Goring-Morris 1995:151).
Retouch
includes Ouchtata, semi-abrupt, and abrupt types usually along the
length of the blade; the distal and proximal ends of the blade are
commonly not modified (Goring-Morris 1995:151).
The microburin
technique is absent.
In the eastern Levant, more variability in classification schemes
exist (Table l.l).
While some researchers are very clear about the
hierarchial organization of complexes, industries, and phases (e.g.,
Henry 1989b, 1995), others are not and continue to split the early
Epipaleolithic into numerous industries without clearly identifying the
relationship between them (e.g., Goring-Morris 1995).
Given the
plethora of extent classification schemes, some seem reluctant to jump
into the fray opting to use the category Non-Natufian Microlithic until
more studies and radiocarbon dates are generated from the eastern Levant
(e.g., Byrd 1994).
Clearly more research is needed before the sequence
and nature of interaction between groups during the early Epipaleolithic
31
can be hypothesized.
It seems that the identification of numerous
industries has less utility, tending to obscure techno-typological
changes behind nomenclature, than demonstrations of technological
affinities between industries.
This is especially true if these
industries are not particularly well dated.
The Qalkhan Complex (ca. 20,000-15,500 BP) seems to be a very
early Epipaleolithic complex that was first recognized in the Ras enNaqb/Hisma area of southern Jordan (Henry 1982, 1983, 1989b, 1995).
This complex appears to be an arid Icind adaptation parallel to the
Kebaran Complex in the Mediterranean woodlands.
Few chronometric dates
are associated with this complex, especially in southern Jordan.
However, this complex has been identified at Petra in southern Jordan
(Schyle aJid Uerpmann 1988), the Azraq Basin in northeastern Jordan (Byrd
1988; Garrard and Byrd 1992; Garrard et al. 1985, 1986, 1987; Garrard
and Gebel 1988), layers 4-7 at Yabrud III rockshelter in western Syria
(Rust 1950), and perhaps even further north in the El-Kowm Basin of
northeastern Syria (Cauvin 1981; Cauvin and Coqueugniot 1990; Cauvin et
al. 1979) by Henry (1995:38).
Based on stratigraphy and radiocarbon
assays from sites in the Azraq Basin, this complex seems to date between
20,000-15,500 BP.
Although its geographic distribution seems to be
restricted to the eastern steppic regions, it may also be present at
Mahal Lavan lOlOS in the western Negev (Goring-Morris 1995:152).
Qalkhan sites have a bimodal distribution with small sites (ca.
50-200 m^) located in open-air and sheltered situations in southern
piedmont areas, and larger sites (ca. 1,200-1,400 m^) located in more
northern areas (Henry 1995:235).
Although the general climate is
characterized as being dry and cold, climatic conditions in the Levant
were moister than at present.
In southern Jordan, Qalkhan sites contain
only chipped stone artifacts; no features or other artifact classes have
32
been foiind in association with these sites (Henry 1995:38)•
The lithic
assemblages contain relatively long narrow bladelets as well as many
relatively wide bladelet tools which are often wider than the normal
range of microlithic tools.
Therefore, these wide bladelet tools may
also be described as blade tools.
The microlithic assemblage has
Qalkhan points which are shouldered triangles manufactured by the
microburin technique.
The microburin scar on Qalkhan points is
unretouched (Goring-Morris 1995:152).
Large la Mouillah points, double
truncated bladelets, and relatively high proportions of notches axid
denticulates are frequently found in tool assemblages (Goring-Morris
1995:152).
Importantly, this complex represents the first appearance
and consistent use of the microburin technique.
Byrd (1994:210) uses the term Non-Natufian Microlithic, rather
than Qalkhan Complex, to describe lithic variability in the eastern
Levant during the early Epipaleolithic.
All eastern regions exhibit a
development from the production of bladelets manufactured into narrow,
backed and retouched microliths to more geometric, tnoncated and/or
backed microliths.
This period is best represented in the Azraq Basin
where sites have abundant evidence for the use of the microburin
technique (Byrd 1988; Byrd and Garrard 1990; Garrard et al. 1986, 1987,
1988; Garrard et al. 1994; Henry 1988, 1989b; Muheisen 1988).
Earlier
sites in the Azraq area have narrow, finely retouched and backed, and
arched backed bladelets; while younger sites have larger backed tools
with more la Mouillah points and double truncated pieces (Byrd 1988),
and wide geometric, truncated and/or backed microliths (Muheisen 1983,
1988) .
Variability and lack of reliable chronometric dates across the
eastern Levant make Byrd (1994) reluctant to group these assemblages
into a single complex.
For example, in the Jordan Valley, narrow,
33
straight backed and obliquely truncated, non-geometric micrcliths
dominate some early assemblages (Edwards 1987, 1990; Edwards et al.
1988).
However, other assemblages are dominated by unbacked, obliquely
truncated, non-geometric microliths (Bar-Yosef 1970).
In these later
assemblages, straight backed non-geometric microliths occur in very low
frequencies.
Tabaqat el-Bumma in west-central Jordan has four
internally inconsistent dates and an assemblage that contains narrow,
obliquely truncated and backed bladelets, and some micropoints (Banning
et al. 1992) .
These assemblages did not have evidence for the use of
the microburin technique (Bcinning et al. 1992; Edwards 1987, 1990) and
appear to be associated with the earlier Epipaleolithic based on
typological considerations.
Since these assemblages are more similar to
the Kebaran or Late Upper Paleolithic Complexes of the western Levant,
they may be contemporaneous with the western Levantine Kebaran and Late
Upper Paleolithic Complexes (Byrd 1994:211) and some would include them
in Masraqan industry (Goring-Morris 1995).
Ironically, Byrd (1994) also views several southern Jordanian
sites, the same sites used to define the Qalkhan Complex and industry,
as problematic because they have few reliable chronometric dates and
because of typological differences between these southern sites and
northern sites in the eastern Levant.
However, several important
similarities between the lithic assemblages, particularly the early use
of the microburin technique, suggest that these assemblages should be
grouped together into a larger technological complex.
Several industries that appear to have strong affinities and
geographical overlap with the Qalkhan Complex as defined by Henry
(1995), may be included in this complex such as the Nebekian, Qalkhan
and Nizzian.
The Nebekian industry (ca. 20,000-18,000 BP) was
originally described by Rust (1950).
Geographically, it is found east
34
of the Rift Valley.
Despite the fact that asseinblages assigned to this
industry have not been well described, Goring-Morris (1995:152)
distinguishes this industry from others based on its high frequencies of
elongated, narrow and symmetrically curved, pointed arch-backed pieces
which are almost oblique tnancations.
(Goring-Morris 1995:152).
Retouch is abrupt and invasive
The larger tool assemblages are characterized
by truncated pieces and non-standardized retouched pieces (Goring-Morris
1995:152).
Although several assemblages in northeast Jordan have
affinities with the Nebekian (Garrard et al. 1994), Goring-Morris
(1995:152) suggests that several southern Jordanian sites may also be
related, i.e., Wadi Madamagh levels A1-A2, Tor Hamar C, and J431.
However, Henry (1995) assigns these later sites, and many of the
northern Jordan sites that define the Nebekian, into the Qalkhan
industry.
The Qalkhan industry as described by Goring-Morris (1995) includes
most of the characteristics used to define the Qalkhan Complex.
The
criteria for distinguishing the Nebekian from the Qalkhan industry
appear to be that the Qalkhan industry has wide bladelet tools, Qalkhan
points with a retouched microburin scar, large la Mouillah points,
double tr\incated bladelets, and abundant notches and denticulates.
It
is also chronologically later than the Nebekian (Goring-Morris
1995:152).
Henry (1995), however, describes the Qalkhan industry
essentially as he does the Qalkhain Complex, using identical criteria and
an expanded geographical distribution that includes arid regions along
the eastern edge of the Mediterranean woodland zone from southern Jordan
to northeastern Syria.
The Nizzian (ca. 17,000-15,000 BP) includes small scalene and
isosceles triangles made by the microburin technique and microgravettes
('spiky points') (Goring-Morris 1995:152).
This industry is frequently
35
svibsumed under the Geometric Kebaran Complex; yet. radiometric
determinations indicate this industry overlaps with the end of the
Kebaran.
Therefore, it may represent either a regional contemporaneous
variant or a later phase of the Qalkhan (Goring-Morris 1995:155).
The
lithic assemblage contains exhausted cores, usually pyramidal, single
platform cores with some opposed platform cores present.
The tool
assemblage also includes well-made scrapers and burins cind numerous
dihedral burins (Goring-Morris 1995:155).
Scalene bladelets and
intensive use of the microburin technique are common.
It has been
suggested that this assemblage may be a precursor or predecessor to the
Mushabian (Goring-Morris 1995:155).
The middle Epipaleolithic (ca. 14,500-12,500 BP) includes the
Geometric Kebaran and Mushabian Complexes.
Climatic conditions during
this period changed siibstantially and are associated with the retreat of
the glaciers after the last glacial maximum at the end of the
Pleistocene.
In general, climatic conditions throughout the Levant
became relatively warmer and moister than the previous period based on
an increase in the amount of arboreal pollen in pollen diagrams across
the Levant (Baruch 1994; Bottema 1987).
The warmer and moister climatic
conditions during this period favored the southward migration of the
major vegetation zones such that the Mediterranean woodlands, and
steppic environments would have expcind as the Saharo-Arabian zone
contracted or was pushed further south.
The geographical distribution
of middle Epipaleolithic sites reflects these favorable climatic
conditions with an increased number of sites found in desert regions.
The Geometric Kebaran Complex (14,500-13,000 BP) is characterized
by straighter backed, non-geometric microliths than in earlier periods,
and the appearance and widespread use of geometric, trapeze-rectangle
microliths (Bar-Yosef and Belfer-Cohen 1989:462-453; Byrd 1994; Henry
36
1989b:93-94; Valla 1988b).
Bladelet blanks tend to be slightly wider
and longer than blanks fovind in the Mushabian Complex.
In general, the
microburin technique was only occasionally used to truncate microliths
cind there is considerable variability in microburin indices.
When the
technique is used, it is more often associated with the manufacture of
triangles than with trapezes and rectangles (Henry 1989b:93).
The
blanks for rectangular and trapeze microliths were apparently snapped
without the microburin technique.
Well dated Geometric Kebaran sites
are found primarily in the Sinai and Negev (Byrd 1994:209).
As previously mentioned, settlement patterns during the Geometric
Kebaran differ from Kebaran patterns in that there is a noticeable
increase in the number of Geometric Kebaran sites in the desert regions.
Geometric Kebaran sites are generally still found near major water
sources including perennial springs, seasonal playa type settings or
confluences of major drainages (Goring-Morris 1995:161).
This strong
association of Geometric Kebaran sites with major water sources, and
ameliorating climatic conditions at this time suggest to some that the
Geometric Kebaran originally derived from an early Mediterranean
woodland adaptation of the Kebaran (Goring-Morris 1995:161).
Although
one expects both prehistoric and modem groups to inhabit areas close to
reliable water sources, the idea is that Geometric Kebaran sites do not
represent a different cultural adaption to a more arid environment.
Rather, the ameliorated climate during this period would have expanded
the Mediterranean woodland zone and Geometric Kebaran sites would have
been located in environmental situations similar to those occupied
during the Kebaran.
Compared to early Epipaleolithic sites, there is
decreased evidence for the use of marine shell, perhaps due to rising
sea levels at this time (Goring-Morris 1995:161).
The Geometric Kebaran has been divided into early and late phases
37
based on widths of backed microliths.
Narrcv/ r^icrcliths are generally
considered older (Bar-Yosef 1981:397-398); however, recent dates from
Neve David may not support this assumption (Byrd 1994:209).
In southern
Jordan, a regional variant of the Geometric Kebaran Complex, the Hamran
industry, has been recognized (Henry 1989b, 1995).
The Middle Hamran
industry is typical of the Geometric Kebaran Complex with relatively
narrow bladelets from which trapeze/rectangles were manufactured without
the use of the microburin technique.
Sites are typically 320-650
and
contain hearths, basalt pestles, and ornamental shells (Henry 1995:39).
The Late and Final Hamran differ from the Middle Hamran by having
shorter bladelets, the microburin technique, and the appearance of
lunates (Henry 1995:39).
The Late and Final Hamran are distinguished by
the relative frequencies of trapeze/rectangles and lunates.
The Late
Hamran has more trapeze/rectangles than lunates, while the Final Hamran
has more lunates than trapeze/rectangles.
Also, in the Final Hamran,
there is evidence for increased use of the microburin technique, and
some lunates have Helwan retouch.
Based on the appearance of lunates
and Helwan retouch, and more substantial architectural features (i.e.,
stone-lined hearths), these Final Hamran sites show a clear
developmental trend towards the Natufian (Henry 1995:39).
The Mushabian (ca. 14,000-12,800 BP) was defined in the Gebel
Maghara, Northern Sinai and some other sites in the Negev (phillips and
Mintz 1977).
It is characterized by abundant backed bladelets, mostly
arched backed bladelets, scalene bladelets, and la Mouillah points
manufactured with the microburin technique (Phillips and Mintz 1977) .
The prominent use of the microburin technique is seen in the Mushabian's
high microburin indices.
Originally, the high microburin indices led
some to postulate that the origins of this complex were in North Africa
(Bar-Yosef and Vogel 1987; Phillips and Mintz 1977).
Geometric
38
microliths are present but occur in
forms (Henry 1989b:91).
IGV: frequencies
and have variable
In general, the bladelet blanks are relatively-
short and wide.
The settlement patterns and site locations of the Mushabian
Complex resemble those of the Geometric Kebaran in many respects.
However, it differs from Geometric Kebaran settlement patterns in that
many sites do not appear to be as strongly tied to water sources as
Geometric Kebaran sites (Goring-Morris 1995:164).
Mushabian sites are
more frequently located in exposed areas, or areas with good views of
the local landscape.
Such locations may have been advantageous in
monitoring the movement of game in steppic environments.
Many regional variants, i.e., industries, of the Mushabian Complex
have been identified; the most prominent being the Ramonian (GoringMorris 1987, 1995), and the Madamaghan (Henry 1989b, 1995).
The
Ramonian is distinguished from other Mushabian industries based on
differences in raw material usage, single platform cores, pyramidal
cores, and blades and bladelets that are frequently quite narrow and
elongated.
The Ramon point, a concave backed and obliquely truncated
bladelet manufactured using the microburin technique, is present.
The
morphological dimensions of many Ramon points exhibit little variation
which suggests that the morphology of these points may have been
somewhat standardized.
Settlement patterns differ from those found in
preceding periods by an increase in the number of high elevation sites
located outside the Mediterranean woodland zone in more arid
environments.
Most Ramonian sites were located in the Irano-Turanian
vegetation zone and the lowland Saharo-Arabian dune localities in the
Negev.
Since highland sites have generally smaller and more diverse
chipped stone assemblages, it has been suggested that highland and
lowland sites may represent a seasonal pattern of dispersion and
39
aggregation by groups whose Local ranges transected the different
environmental zones.
Ramonian sites frequently have larger tool
assemblages and appear to have separate reduction sequences for larger
tools such as scrapers and smaller tools such as microliths.
Henry (1989b, 1995) has identified the Madamaghan industry within
the Mushabian complex.
Technologically, this industry has relatively
long, wide bladelets, and high microburin frequencies.
Typologically,
there are moderate percentages of points including la Mouillah,
microgravette, and arched backed varieties, and low percentages of
geometric microliths consisting of trapeze/rectangles, and normal and
Helwan lunates.
Although the characteristics of this industry seem to
exhibit good affinities with other Mushsibian sites, it is not clear that
the type site for this industry, Wadi Madamagh, actually dates to the
middle Epipaleolithic.
Although Henry (1989b, 1995) asserts that the
lithic assemblage from the rockshelter in Wadi Madamagh typologically
resembles other assemblages in the Mushabian Complex, the original
interpretation of the site (Kirkbride 1958) and later references to this
site (e.g., Byrd 1994:205) indicate that Henry's assertion is not
universally accepted.
Conclusions
There is considerable typological and technological variability in
early and middle Epipaleolithic lithic assemblages.
As a result,
numerous classification schemes have been postulated that reflect
techno-typological differences in lithic assemblages during this period.
Henry's (1989b) hierarchical scheme of complex, industry, phase,
assemblage is being adopted more frequently.
Still, a number of authors
are reluctant to apply this scheme in both the eastern and western
Levant, or continue to identify numerous groups at only the industry
level.
All researchers agree that additional quantitative technological
40
and typological descriptions of lithic assemblages are necessary from
sites with reliable chronometric dates before this situation can be
satisfactorily resolved.
in this direction.
The analyses presented here should be a step
41
CHAPTER 2:
TOR AL-TAREEO AND PLEISTOCENE LAKE HASA
Tor al-Tareeq (WHS 1065) is located on the north bank of a small
tributary of Wadi Hasa in west-central Jordan (Figure 2.1).
WHS 1065 is
spread over a 812 m^ area and has cultural deposits with an average
depth of about one meter.
Originally, Tor al-Tareeq was interpreted as
a series of superimposed Kebaran (and possibly Natufian) basecamps
(Clark et al. 1987, 1988; Coinman eC al. 1989).
This interpretation
was based on the high density of artifacts in well-defined natural
levels; typological similarities between the WHS 1065 and Kebaran
chipped stone assemblages; and several radiocarbon determinations from
the lower excavated levels at WHS 1065 that cluster between 16,50015,000 BP (Clark et al. 1987, 1988; Coinman et al. 198 9; Neeley et al.
1995).
Several features including hearths, pits, overridden wall
sections, midden deposits under a collapsed rockshelter, and a fossil
spring deposit were used to support this initial interpretation (Clark
et al. 1987, 1988; Coinman et al. 1989).
Recently, a broader
interpretation has been suggested, i.e., that the site may have been
occupied repeatedly for relatively short periods and/or intensively at
several points in time (Neeley et al. 1995).
The temporal and cultural
interpretation of the upper levels has also changed from a possible
Natufian basecamp to several Geometric Kebaran occupations (Neeley et
al. 1995).
Paleoenvironment and paleolandscape
The modem Levantine climate is characterized by long, hot, dry
summers that alternate with shorter, wetter winters.
Annual
precipitation varies between 50-100 mm in the arid regions of the Negev,
Sinai, and southern Jordan to 200 mm in the northern Negev, Jordanian
Plateau, and lower Jordan Valley and to 200-300 mm in the upland and
coastal regions of the northern Levant.
Today, WHS 1065 is located at
42
Figure 2.1: Distribution of excavated sites in the Wadi Hasa drainage
basin (after Neeley et al. 19951
43
815 m asl in the Irano-Turanian steppe environment, a setni-arid zone
that receives about 100-200 mm of precipitation annually.
rainfall usually occurs between December and March.
Modem
However as
precipitation is commonly torrential and localized, the annual amount
and intensity of precipitation varies considerably within and across the
region.
A description of the paleoenvironmental and paleoclimatic
conditions of the Levant during the Late Pleistocene (ca. 25,000-10,000
BP) is presented in more detail here than in the preceding chapter.
An
understanding of paleoenvironmental variability and consistency across
the Levant is helpful in positioning the occupations at WHS 1065 in
their regional, environmental and temporal contexts.
It should become
clear that although the timing may vary between regions, the
paleoenvironmental sequence for the eastern Hasa region is largely
consistent with other Levantine paleoenvironmental reconstructions.
Climatological conditions in the Late Pleistocene produced longterm, multi-annual cycles that brought about fluctuations between
periods of drought and periods of higher humidity (Katsnelson 1964;
Rosenan 1970).
While dry periods probably had somewhat higher humidity
levels than present, humid periods may have had as much as 50-100%
greater humidity (Issar and Bruin 1983:75), perhaps even more.
As a
consequence, precipitation and recharge rates across the Levant were
generally greater and aquifers tended to fill (Issar and Bruins
1983:63), so that, spring activity was probably greater throughout the
Late Pleistocene.
In the Late Pleistocene, WHS 1065 would have been located at the
northwest edge of a large Pleistocene lake (Schuldenrein and Clark
1994).
Several depositional changes at WHS 1065 are present such as the
interfingering of cultural materials with marshy sediments, the
44
deposition of cultural materials over marl deposits, and the movement of
cultural materials by colluviura.
The interfingering of palustrine (pond
deposited) silts and clays, and colluvium indicate that the local
environment changed as lake levels in Pleistocene Lake Hasa fluctuated
between mesic and xeric conditions (Schuldenrein and Clark 1994) .
Sometime between 25,000 and 22,000 BP, climatic conditions became
increasingly arid throughout the Levant.
evidence includes:
Supporting geomorphologic
(1) the initial lowering of Lake Lisan between
21,000 and 18,000 BP (Gat and Magaritz 1980; Kaufman et al. 1992;
Yechieli et al. 1993); (2) the first appearance of eolian sand in
drainages in the Sinai and Negev around 25,000 BP (Goldberg 1986:242);
(3) the absence of dated fluvial sediments between 22,000-17,000 BP
indicating widespread valley erosion in the Negev (Marks 1977); and (4)
an erosional period between 25,000-20,000 BP in the Judayid Basin-Wadi
Hisitia area of southern Jordan (Henry 1986).
Evidence for a drying trend, i.e., lower lake levels and reduced
base levels, appears somewhat later (ca. 20,000 and 17,000 BP) in the
eastern or Upper Hasa (Schuldenrein and Clark 1994:49).
Between 26,000-
20,000 BP, Pleistocene Lake Hasa had high lake levels which are
associated with increased spring activity, and more mesic conditions.
The Late Ahmarian occupation in Wadi Hasa (WHS 618) is associated with a
variety of spring and marsh environments during this period.
The
environmental situation at WHS 1065 is directly associated with
fluctuations in lake levels (Schuldenrein and Clark 1994:40).
Between
ca. 20,000-17,000 BP, the Hasa Marls, which were deposited in subaqueous
conditions, eroded and stream margin environments around Lake Hasa
formed (Copeland and Vita-Finzi 1978; Schuldenrein and Clark 1994:46;
Vita-Finzi 1966).
Evidence from WHS 1065 indicates that this interval
is associated with lower lake levels, drier conditions and erosion of
45
laminar lake beds (Schuldenrein and Clark 1994).
The lack cf riparian
taxa in the pollen spectrum also supports this interpretation of cooler
and drier conditions; this interpretation is consistent with regional
interpretations (Neeley et al. 1995).
Still, conditions at the site
were probably at least seasonally marshy as these deposits are
associated with marl and palustrine silts and clays (i.e., marshylacustrine silts and clays) (Clark et al. 1987, 1988; Schuldenrein and
Clark 1994:47).
Similarly, the onset of drier conditions in
northeastern Jordan occurs sometime after 21,000 BP, 2,000-3,000 years
later than in more southern regions based on evidence from Wadi Jilat 9
in Azrag Basin (Garrard et al. 1987).
Palynological evidence suggests that moist conditions actually
continued across the Levant until ca. 22,000 BP which is closer to the
Azraq situation (Horowitz 1979:341).
However, subsequent palynological
studies from a well dated core in the Hula Basin (Bottema and Van Zeist
1981) suggest that the climatic reconstruction of Marks (1977) and
Goldberg (1981, 1986) in which there is a change to more arid conditions
ca. 25,000 BP, is more appropriate.
Therefore, the "lag" in the onset
of drier conditions in the Azraq Basin and Wadi Rasa is most likely due
to its deeper placement in the northern climatological belt.
As the
climatic belts migrated north during drier periods, the Azraq Basin and
Wadi Hasa would have been affected later than more southern regions.
The rate of environmental change during this period and its
potential affects on human and animal populations throughout the Levant
can be reconstructed to some extent from these data.
As a result of
drier climatic conditions and erosion. Late Upper Paleolithic and early
Epipaleolithic sites (ca. 22,000-18,000 BP) are scarce (Goldberg and
Bar-Yosef 1982; Henry 1986) especially in the southern Levant.
If any
archaeological material was deposited during this period, it would most
46
likely have been eroded.
The few Late Upper Paleolithic sites that do
exist in the Negev and southern Jordan are found associated with eolian
sediments and dunes (Goldberg 1986:242; Henry 1986).
Between ca. 17,000-12,000 BP, a moist period is suggested by (1)
renewed alluviation and aggradation in the northern Sinai aind western
Negev (Goldberg 1986:240); (2) new fan deltas at the junction of major
tributaries; (3) local palustrine basins in the Hasa (Schuldenrein and
Clark 1994) and lower Jordan Valley (Schuldenrein and Goldberg 1981);
(4) high lake level stands at Lake Lisan ca. 14,000 BP (Begin et al.
1985; Druckman et al. 1987); (5) spring heads reaching their highest
level at Wadi al Hammeh ca. 15,000-11,000 BP (Macumber and Head
1991:72); (6) higher water tables and an increase in relative moisture
from lake-like deposits in Wadi Mushabi and extensive gleying in Qadesh
Bamea associated with middle Epipaleolithic sites ca. 14,000 BP
(Goldberg 1981:60-21; 1986:240); (7) well developed paleosols in the
Negev highlands and the Gebel Maghara area of northwestern Sinai by
14,000 BP indicating moist, stable conditions must have prevailed in the
region for at least 3,000-4,000 years (Goldberg 1981); and (8) the
pollen diagram from the Hula pollen core (Baruch and Bottema 1991).
At Wadi Hasa, there is evidence for higher lake levels, an
increase in spring activity, and deposition of tributary alluvium on
eroded surfaces between ca. 17,000-15,000 BP (Schuldenrein and Clark
1994).
This period corresponds to the earliest cultural deposits at WHS
1065 (some of which are analyzed in this study).
In general, high lake
levels are associated with moister climates and transitions to paludal
(pond-like or marshy) environments.
Pollen evidence supports this
interpretation showing a steady increase in the ratio of arboreal to
steppe plants (Neeley et al. 1995).
Wadis Salibiya and Fazael in the
lower Jordan Valley show slightly earlier climatic amelioration (Darmon
47
1987; Leroi-Gourhan 3ind Darrr.cn 19S7) .
However, the northern position of
the lower Jordan Valley would naturally result in this area experiencing
the effects of climatic amelioration earlier than more southern areas as
climatological belts migrated southward.
How much moister was this "moist" period?
Based on the well
developed paleosols in the Negev highlands and the Gebel Maghara area of
northwestern Sinai, climatic conditions may have been considerably
wetter with 500 mm of annual precipitation (2.5 times greater than
present) by 14,000 BP (Goldberg 1981).
In southeastern Jordan, Judayid
504, a rockshelter overlooking a dry lake bed, lies in red sands with
high frequencies of oak, walnut, and conifer pollen (Henry 1982).
This
is indicative of an environment with 200-300 mm of precipitation
annually (again, 2-3 times greater than present) (Henry 1982).
Based on
the Negev data, rainfall gradients must have been depressed about 150 km
to the south and by extension across the entire Levant and Sinai
(Goodfriend and Magaritz 1988) .
Not surprisingly, land use patterns also suggest more mesic
climatic conditions as an increased number of late Kebaran and Geometric
Kebaran sites are found outside the Mediterranean woodland vegetation
zone as reconstructed for the early Kebaran period.
This suggests that
the Mediterranean woodland zone extended further south and east than
previously.
Moister conditions continued in the northeastern Levant
until at least ca. 14,500 BP in the Azraq Basin (Garrard et aJ. 1986).
Unfortunately, the geomorphic stratigraphy of the Judayid Basin-Wadi
Hisma area is not well developed and paleoenvironmental reconstructions
rely heavily on palynological studies at a few locations.
However,
evidence from these locales suggests that moister conditions continued
to approximately 12,000 BP (Henry 1986).
A drying trend with more arid conditions than the previous dry
48
period began about 11,000-12,000 BP and continued to about 10,000 BP.
It corresponds to the Geometric Kebaran in the Negev and Sinai and
Natufian across the Levant.
Several locations provide sedimentological
and palynological evidence for drier conditions at this time including
(1) initiation of the halite deposition in the Rift Valley between
11,000-8,500 BP (Yechieli et al. 1993); (2) a dramatic decrease in
arboreal-nonarboreal ratios from 75% arboreal pollen in 11,500 BP to 30%
by 10,500 BP in the Hula pollen core (Baruch and Bottema 1991); (3)
absence of lake deposition at Wadi al Hammeh ca. 11,000 BP in response
to significant increases in evapo-transpiration rates (Macumber and Head
1991); (4) a highly evaporated sabkha deposit in the northern Negev ca.
11,000 BP (Magaritz 1986); (5) archaeological sites in eolian deposits
in the Azraq Basin defined at Kharaneh 4, Phase C and D (Garrard et al.
1987) and at Judayid 2 in Wadi Judayid (Henry 1986) ; and (6) massive
cycles of colluviation and slope erosion in the lower Jordan Valley
(Schuldenrein and Goldberg 1981) and Wadi Hasa at WHS 1065 (Schuldenrein
and Clark 1994).
Land use patterns also suggest increased aridity as
Geometric Kebaran and Natufian sites in the southern Levant and Sinai
have limited distributions during this period.
In sum, during the end of the Pleistocene, major fluctuations in
climatic conditions created corresponding fluctuations in depositional
and erosional sequences.
Moist conditions prevailed before 25,000 BP
and are associated with Late Upper Paleolithic occupations in alluvial
sediments.
This was followed by a significant drying trend between
25,000 and 17,000 BP and is associated with a decrease in the number of
Late Upper Paleolithic and Kebaran occupations found in most southern
Levantine areas.
While the dry period affected the entire Levantine
region, its effects were more significant in the south where massive
erosion and deposition of eolian sediments occurred.
A brief return to
49
moister conditions between 17,000 and 12,000 BP marks a period of human
expansion into previously arid regions and the development of paleosols
at a number of locations in the Negev aind Sinai.
A drier period between
12,000 and 10,000 BP is indicated by the lowering of Lake Lisan, the
accumulation of evaporite deposits as shallow lakes dried, the incision
of most wadi systems across the Levant and Sinai, and the deposition of
eolian sands in more localized regions.
The timing of the climatic fluctuations and subsequent changes in
geomorphic stratigraphies differs slightly between regions.
In general,
southern regions were more significantly affected by increasing aridity
and had longer arid periods than northern areas.
During these drier
periods, human occupation of the region was restricted to springs,
marshes and lake margins especially in southern regions.
Previous Research
WHS 1065 was discovered in 1982 by Burton MacDonald during the
Wadi Hasa Survey (WHS) (MacDonald et al. 1980, 1982, 1983; MacDonald
1988).
Between 1979 and 1983, the WHS conducted a fairly systematic
survey on the south bank of Wadi Hasa drainage system (Figure 2.1).
The
Wadi Hasa drainage begins in central Jordan, near Qa el-Jinz and flows
west to its mouth at the Dead Sea near As-Safi.
During the WHS, over
1074 sites were identified; 222 of these were lithic sites from the
Lower, Middle, and Upper Paleolithic, Epipaleolithic, and Prepottery
Neolithic periods (Clark et al. 1987, 1988; MacDonald et al. 1983;
MacDonald 1988).
The number of Paleolithic sites identified during the
WHS suggested that additional work, focusing specifically on the
Paleolithic, had the potential to yield valuable information on land use
and settlement patterns of Paleolithic groups in this region.
Although the WHS recorded prehistoric sites, its research design
emphasized identifying historic period sites.
Therefore, many
50
geomorphological, topographic and environmental situations adjacent to
Wadi Hasa were not necessarily explored (Clark ec al. 1987).
As current
and past geomorpological conditions can significantly affect the
distribution, preservation and visibility of Paleolithic sites, these
conditions need to be considered when interpreting the distribution of
cultural and temporal periods in the WHS data.
Although some
geoarchaeological research relating landforms to archaeological sites in
the Wadi Hasa had been conducted (Copeland and Vita-Finz 1978; VitaFinzi 1964, 1966, 1982), the dating of the landforms relied heavily on
artifacts thought to be temporally diagnostic (Schuldenrein and Clark
1994:33).
Only one radiocarbon determination from the lowest terrace in
the present Wadi Hasa drainage system was obtained.
The aggradation
event associated with this lowest terrace (called the Hasa Formation)
dated to ca. 8,000 BP (Copeland and Vita-Finzi 1978).
Additional research was clearly needed to (1) obtain radiocarbon
dates to secure the cultural and environmental sequence in central
Jordan, (2) reconstruct land use and settlement patterns around Wadi
Hasa, and (3) evaluate the existing settlement pattern and land use
models.
Two lines of research that would supplement the WHS and
previous geoarchaeological work were necessary including:
(l)
systematically surveying a variety of geomorphologic and environmental
contexts around Wadi Hasa that had not yet been surveyed; and (2)
defining and refining the environmental and cultural sequence for the
Paleolithic period in central Jordan.
In 1984, the Wadi Hasa Paleolithic Project (WHPP) was developed in
order to gather data to address the additional research topics mentioned
above.
The goals of the WHPP were to "(1) acquire adequate samples of
lithic assemblages, (2) recover faunal, floral and other kinds of
paleoenviromental information, (3) undertake a geomorphological study of
51
the east Hasa drainage, (4) establish the beginnings of a radiocarbon
chronology for west-central Jordan, and (5) map the extent of
Pleistocene Lake Hasa with which most of the archaeological sites are
associated" (Clark et al. 1987:19).
Once the environmental and cultural
sequences for west-central Jordan were developed and settlement
distribution data collected, these data could be used to evaluate two
prominent settlement system models:
Marks' model for paleoenviromental
change and human adaptation (radiating verses circulating settlement
systems) developed for the central Negev highlands (Marks and Friedel
1977; Mortensen 1972) and Henry's model of transhumance developed for
southern Jordan (Henry 1982, 1983).
Marks' radiating settlement pattern has a sedentary to semisedentary residential camp surrounded by small, limited activity sites
during relatively moist climatic periods (Marks cind Friedel 1977) .
These moist periods would correspond to expansions in the Mediterranean
vegetation zone.
A circulating settlement pattern consists of small,
relatively similar multi-purpose sites.
This pattern is associated with
more xeric conditions and the formation of the Irano-Turanian vegetation
zone or dry siab-desert and steppe environments.
Henry's transhumance
model suggests that Pleistocene hunter-gatherers moved between different
elevational zones on a seasonal basis (Henry 1982, 1983).
This seasonal
movement affected group size and composition and would ultimately be
reflected in the size and character of the site.
Therefore, larger
residential sites are located at lower elevations in the winter while
smaller sites tend to be found at higher elevations during the summer.
Thus, a "radiating" settlement system was employed in the winter, a
"circulating" system characterized the summer settlement system.
This
model differs from Marks and Friedel's (1977) model in that settlement
pattern changes occur on a seasonal basis; one settlement pattern does
52
not persist through thousands of years in any given mesic or xeric
period.
To gather the appropriate environmental and artifactual evidence,
the WHPP conducted a systematic survey along the north bank of Wadi Hasa
in 1984, 1992 and 1993 (Clark 1992; Clark et al. 1987, 1988, 1992;
Coinman et al. 1989).
In addition, six Paleolithic sites identified
during the WHS were tested in 1984.
Five of these sites had buried,
stratified deposits (Clark et al. 1987, 1988; Coinman et al. 1989).
Additional excavations at two Epipaleolithic sites (WHS 1065 and WHS
784) were conducted in 1992 and 1993, respectively (Neeley et al. 1995;
Olszewski et al. 1994).
The Late Upper Paleolithic to late
Epipaleolithic sequence was primarily developed from the excavations at
three Upper or eastern Hasa sites:
WHS 618 (a late Ahmarian site), WHS
1065 (a middle Epipaleolithic site), and WHS 784 (a late Ahmarian and
middle-late Epipaleolithic site) (Figure 2.1).
Both WHS 1065 and WHS
784 are collapsed rockshelters on the southern beink of Wadi Hasa.
Tor al Tareeq (WHS 1065) - Excavation and Stratigraphy
A description of the excavations conducted at WHS 1065 in 1984 and
1992 by the WHPP is useful in understanding how natural levels 5 and 7,
Step C (the chipped stone material from these levels are analyzed in
this study) relate stratigraphically and culturally to the other levels
excavated at the site.
The 1984 research conducted at WHS 1065 included
an intensive surface collection in which 95% of surface artifacts were
collected (Coinman et al. 1989) and the excavation of a 44 meter long,
stepped trench through the middle of the site (Coinman et al. 1989;
Clark et al. 1987, 1988).
The trench, excavated in eight 5x1 meter trenches (Steps A-H)
and one 4x1 meter trench (Step I) (Figure 2.2), was placed
perpendicular to the slope through what appeared to be the heaviest
53
Fiqxire 2.2: Site mao of Tor al-Tareeq (WHS 1065) (after Neeley et a l .
1995).
TVPRNIF RFFFINANL
Rf.DRfH'K MORTARS
m i l
J
CIVRSUM RI! V <SOT rtn i RRRRI^*
N\V OIJAI)
RT)RIINI.R. <NOT CDI I.RRRRNI
(INTL (
EXMSRN BEDROCK
RACKDIRR
50
WHS SITE 1065
rONfOI'R tNTI-RVAI. I M
MIDPt-Nd ATPI
IN M
SCALE
<0 N
MARI.S
54
concentration of surface artifacts (Clark et al. 1987, 1988).
Unless
natural stratigraphic levels were discovered, the steps were excavated
in 10 cm arbitrary levels.
The arbitrary levels were later combined
into natural stratigraphic levels based on sedimentological differences
in the excavation profiles.
As might be expected, not all of the
natural levels were identified in each excavated step.
A correlation of
the natural levels of Step B, Unit B, Unit C and Step C, and the
arbitrary levels from Step C is presented in Table 2.1.
drawings of Steps B and C are presented in Figure 2.3.
Profile
The arbitrary
and natural levels used in this analysis are presented in bold type.
In
this chapter, levels will be used to refer to natural stratigraphic
levels unless otherwise indicated.
Over 160,000 pieces of chipped stone
were collected during the 1984 field season (Neeley ec al. 1995).
The best stratigraphic sequences and all of the radiocarbon dates
on charcoal were obtained from Steps A, B, and C (Clark et al. 1987,
1988, 1992; Neeley et al. 1995:23).
Step A contained a wall fragment
probably associated with levels 1 and/or 2 (Neeley et al. 1995:23).
Natural levels 1 and 2 are dominated by larger tools and contained veryfew microliths (less than 10%) compared to levels 3 and 4 (62-65%).
majority of microliths in levels 3 and 4 were non-geometric; the most
abundant non-geometric microlith were straight backed and curved or
arched backed bladelets which comprised 84% of the microlithic types
from Step A (Neeley et al. 1995:23).
No geometric microliths were
recorded.
In Step B, non-geometric microliths comprise 50-60% of the
retouched tools in all levels (Clark et al. 1987, 1988; Neeley et al.
1995:23).
However, level 5 contained the highest percentages of non-
geometric microliths.
Geometric microliths were foiand in low
frequencies (less than 4%) in all levels except level 1 (11%
The
55
Table 2.1:
Correlations of natural and arbitrary' levels from Steps B
and C (excavated in 1984) and Units B and C (excavated in
1992) (after Clark et al. 1987:59 and Neeley et al.
1995:Table 2). The lithic assemblage from arbitrary levels
8-15 in Step C (in bold type) are the focus of this
analysis. Note that the natural levels do not necessarily
correspond between excavated steps and units.
Natural
Level
Unit B
(1992)
Natural
Level
Step B
(1984)
Sediments
Natural
Level
Step C
(1984)
Arbitrary
Level
Step C
(1984)
Natural
Level
Unit C
(1992)
I
1
mixed
surface
deposits
1
1
I
grey ashy
intrusive
2
2
-
-
grey
intrusive
3
2
-
-
-
-
-
-
-
-
-
-
III
2
grey
midden
III
3
grey
calcreted
midden
IV
4
grey/brwn
to brown
silt/sand
-
-
-
-
-
-
-
-
V
5
-
-
?4
II
-
-
-
-
II
4
3-7
II
grey/brwn
to brown
silt/sand
5
8-10
II
cobble
layer
6
compact
light tan
silt
7
-
11-15
-
III, IV
Figure 2.3:
The east profile of Seeps B and C (after Clark et al.
1987, 1988)
WHS SITE 1065
N60
STEP B EAST PROFILE
P
FS
"si
. BEDROCK
®
UNEXCAVATED
N55
STEP C
POTHOLE
EAST PROFILE
BACKOIRT
N50
UNEXCAVATED
RODENT
1.0 METER
SCALE
57
microliths).
Since the straight and arched backed bladelets are the
most dominant type of microliths throughout the levels and "there is
some consistency and continuity between these levels", they relate
broadly to other industries in the Kebaran Complex (Neeley et al.
1995 :24).
Step C differs from Step B in that it contained relatively higher
frequencies of geometric microliths in the upper levels (levels 1-5).
Level 7 was the only level that contained percentages of geometric (4%)
and non-geometries (54%) microliths similar to those of level 5, Step B
(2%, 58%) (Neeley et al. 1995:24).
Based on the typological
similarities and the greater abundance of geometric microliths within
levels 1-5, two temporal occupations are suggested, a stratigraphically
higher occupation (levels 1-5) related to a Geometric Kebaran component,
with levels 4-5 containing the highest frequencies of the atypical, wide
(Hasa) lunates and bitruncated bladelets, and a stratigraphically lower
occupation (level 7) related to industries in the Kebaran Complex (Clark
et al. 1987, 1988; Neeley et al. 1995:24).
There are several similarities between the basal levels of Steps B
and C.
Both basal levels were comprised of a compact, light tan silt
that appeared to be an in situ deposit (Clark et al. 1987).
step C also
contained two hearths (Features 4 and 5) in the lowest level (level 7).
Level 7 contained the highest relative frequencies of bitruncated
bladelets and has arched backed bladelets that occur in frequencies
similar to those of straight backed bladelets.
noted in Step B (Neeley et al. 1995:25) .
This trend was also
Six charcoal determinations
from the lowest levels in Steps A, B, and C yielded dates that range
from 15,580 to 16,900 BP (Table 2.2) (Clark et al. 1987, 1988;
Schuldenrein and Clark 1994:34; Neeley et al. 1995).
The lower two-thirds of the trench (Steps D-I) had a 20 cm layer
Table 2.2:
Aae (BP)
9,010 + :100
11,280 + 290
15,580 + 250
15,860 + 430
16,570 + 380
16,670 + 270
16,790 ± 370
16,900 + 500
Radiometric dates from Tor al-Tareeq (WHS 1065) (Clark et
al. 1987, 1988; Neeley et al. 1995).
Level
Bill
Bill
B4
B4
C7
BIV
C7
A4
Material
Soil/sediment
Soi1/sediment
Charcoal/hearth
Charcoal/hearth
Charcoal/hearth
Charcoal/hearth
Charcoal
Charcoal
Lab. No.
Beta-57898
Beta-57899
UA-43 92
UA-43 94
UA-4390
Beta-57900
UA-4393
UA-43 91
59
of artifacts in a colluvial deposit that was derived frcm upslcpe (Clark
et al. 1987, 1988; Coinman et al. 1989).
In contrast to Steps A-C, the
upper levels of Steps D, E, and F have denser concentrations of
artifacts than the lower levels.
Since these upper levels include some
geometric microliths, they probably correspond to the Geometric Kebaran
occupation (Neeley et al. 1995:26).
component.
Steps D-I appear to lack a Kebaran
Therefore, the Kebaran component seems to be confined to the
upper slope (Steps A, B, C) (Neeley et al. 1995:26).
Steps D, E, and F
were probably beneath Lake Hasa during the earliest occupation of the
site (Clark et al. 1987, 1988; Neeley et al. 1995:24).
Steps G, H, and I are comprised of marls with small gravel and
sand lenses from fluvial and colluvial depositions (Neeley et al.
1995:10).
These marls and calcareous deposits are overlain by sands and
colluvial debris derived from upslope (Neeley et al. 1995:10).
The lack
of artifacts in a marl and calcareous deposit suggests that this area
was below Lake Hasa during relatively moist periods with high lake
levels.
In 1992, two extension units (each 2 x 2 m) were excavated at WHS
1065 in order to further define the Kebaran and possible Geometric
Kebaran occupations, and isolate/define a possible Natufian component
(Clark et al. 1992; Neeley et al. 1995:4).
One extension. Unit B, was
placed east of Step B and the other, Unit C, west of Step C (Figure
2.2).
Over 40,000 lithic artifacts were collected in addition to fauna,
shell and groundstone (Neeley et al. 1995).
In Unit B, five natural
levels were excavated which seem to show a continuation of the
stratigraphy and Kebaran component noted in Step B in the 1984
excavations (Neeley et al. 1995) (see Table 2.1).
densities are high throughout Unit B.
were excavated (Neeley et al. 1995:9).
In general, artifact
In Unit C, four natural levels
Unlike Step C, the four natural
60
levels excavated in Unit C appear to be colluvial deposits (Meeley st
al. 1995).
Unit C differs from Unit B in that Unit C seems to have less
preservation of distinct natural levels from the Kebaran occupation(s)
along the lake margin.
The stratigraphy in Unit C is probably less
distinctive because of greater colluviation in this portion of the site
(Neeley et al. 1995).
The major differences between Unit C and Step C are that the
features and living surfaces identified in Step C, levels 7 (east
profile) and 4 (west profile), are not present in Unit C, levels II-IV
(Neeley ec al. 1995) .
Despite the lack of features and continuous
living surfaces between Step C and Unit C, the chipped stone assemblage
in both excavated areas exhibit similarities in the frequency and types
of artifacts represented (Neeley et al. 1995).
It is believed that only
these two areas have adequate evidence for differentiating an earlier,
Kebaran occupation (Unit C, levels III and IV, and Step C, level 7) and
a later, Geometric Kebaran occupation (Unit C levels I and II, and Step
C, level 5) (Neeley et al. 1995) .
Interpretations
Both the faunal assemblage and pollen profile indicate the
presence of an open woodland/steppe mosaic environment throughout most
of the occupation of the site (Clark et al. 1987, 1988; Schuldenrein and
Clark 1944:39; Neeley et al. 1995).
The faunal assemblage which
contains tortoise, gazelle, auroch, equids, ovicaprines, and avifauna
indicates diversity in the natural environment (Neeley et al. 1995).
The pollen profile includes cattail iTypha), riparian willow iSalix),
alder {Alnus), oak (Quercus) , Chenopodiaceae (usually greater than 50%
in most samples), and grasses (Clark et al. 1987, 1988; Schuldenrein and
Clark 1944:39; Neeley et al. 1995).
Although there are changes in these pollen percentages, steppic
61
pollen is dominate throughout the entire occupation of the site v.'hich
probably reflects vegetation both at the site and in adjacent areas as
airborne pollen from the surrounding landscape is deposited.
However,
the later occupations at the site, i.e., those occupations associated
with greater frequencies of geometric microliths, have increased
proportions of riparian taxa.
These riparian taxa occur relatively
rapidly suggesting that the later occupations at this site enjoyed
warmer, more mesic climatic conditions (Neeley et al. 1995).
The
absence of riparian taxa in the earlier occupations suggests drier,
cooler climatic conditions.
Based on these data, the margins of
Pleistocene Lake Hasa would have provided a broad range of lake and
marsh resources that differed considerably from the surrounding steppe
environment (Neeley et al. 1995).
This diversity was probably a major
factor for prehistoric groups to return repeatedly to the Hasa basin
during the Epipaleolithic (Neeley et al. 1995).
Radiocarbon determinations clustering between 15,900-15,000 BP
place the occupation(s) of the lowest levels as contemporary with the
Late/Terminal Upper Paleolithic (pre-20,000-ca. 15,000 BP) and the
Kebaran (ca. 19,000-14,500 BP) periods in the Levant (Bar-Yosef and
Vogel 1987; Byrd 1994).
However, the overall character of the lithic
assemblage, specifically the high proportion of narrow backed bladelets,
suggests that most of the cultural deposits are associated with the
Kebaran Complex (Bar-Yosef 1975, 1981; Clark et al. 1987, 1988; Neeley
et al. 1995).
Typological changes in the microlithic component of the
flaked stone assemblage, specifically an increase in the number of
geometric microliths in the upper levels of some of the excavated units,
suggest a later Epipaleolithic occupation possibly associated with the
Geometric Kebaran Complex (Clark et al. 1987, 1988; Neeley et al. 1995).
The occupations at the site probably moved up and down the slope
62
following the rise and fall of the lake levels (Schuldenrein and Clark
1994).
This interpretation is based on the fact that the oldest Kebaran
levels are upslope and younger Geometric Kebaran levels are downslope
(Neeley et al. 1995:12).
Although lake levels clearly fluctuated over
time, the occupational surfaces and in situ deposits may have been
scattered along the upper slope without regard to the lake levels
(Neeley et al. 1995:12).
This is based on the Kebaran levels in Step B
having yoimger radiocarbon determinations than levels in Steps A and C
which have the greatest concentrations of artifacts possibly associated
with the Geometric Kebaran and Kebaran Complexes (Neeley ec al.
1995:12) .
In either case, Tor al-Tareeq was occupied repeatedly because
of its location near the margins of Pleistocene Lake Hasa.
63
CHAPTER 3:
RESEARCH METHODOLOGY
Chipped stone material comprises the largest artifactual component
of Paleolithic sites; therefore, it provides an important line of
evidence for examining variations in activities and functions between
and within prehistoric settlements.
As a result, many inferences and
interpretations regarding human adaptions during the Paleolithic are
based on typological and technological analyses of chipped stone
material.
Several systematic typological and technological approaches
have been employed in the study of Epipaleolithic chipped stone
assemblages in the Levant.
A recently developed approach, the chains
operatoire or operational sequence, is applied in this study.
This
approach, largely adapted from French studies (e.g., Lemonnier 1983,
1990), is based on identifying and tracking an artifact through all
stages of production from the acquisition of raw materials to core
reduction and the manufacture, use, and final discard of the artifact
(Bar-Yosef 1991b).
A similar approach, employed by some American
archaeologists, is termed the life history approach (e.g., Schiffer
1975:46-48; Rathje and Schiffer 1982:84-89).
An important assumption behind this approach is that operational
sequences used to manufacture chipped stone artifacts reflect technical
traditions and were learned behaviors "passed from one generation to the
next" (Bar-Yosef 1991b:320).
If various production strategies and
sequences can be identified, Epipaleolithic groups can be distinguished
more reliably and relationships between contemporaneous groups better
understood (Bar-Yosef 1991b:322).
Many archaeologists working in the
Levant have already observed that there is a relationship between
reduction strategies and subsistence issues, mobility practices and
settlement patterns (e.g., Byrd 1988; Garrard et al. 1987; Henry 1989a,
1989b; Marks 1976, 1977, 1983; Marks and Volkman 1983).
Additional
64
studies will enable more meaningful interpretations cf regional
variability to be generated.
Acquisition of Raw Material
The use life of an artifact begins with raw material procurement
which is influenced by the abundance, availability, quality, and size of
the raw material.
Researchers frequently note the ubiquity of lithic
raw material sources, mainly chert or flint, throughout the Levant
(e.g., Goring-Morris 1995:143).
Therefore, raw material constraints are
generally not considered to be significant in influencing core reduction
strategies.
However, the quality and size of the raw material varies
throughout the Levant and some preferences in lithic material between
geographic regions have been noted (Goring-Morris 1995:143).
During
mesic climatic periods, a denser vegetation cover would have obscured
some of the available raw material.
Although there would be less ground
cover during xeric periods and lithic raw material would have been more
visible, there is little evidence for human occupation in these arid
regions during most of the Late Pleistocene.
Therefore, suitable raw
material may not have been as ubiquitous as some suggest.
There is clearly some relationship between the size of the core
and the size of the flake produced.
Tools manufactured on larger blanks
such as scrapers and burins, may be constrained by the dimensions of the
core.
However, the size and quality of raw material is not always
directly related to the size of the debitage (Bar-Yosef 1991b:322).
Bladelets in many assemblages were produced from varying sizes eind
qualities of raw material.
Also, the reduction strategy for
manufacturing bladelets was not necessarily determined by the size of
the raw material (Bar-Yosef 1991b:322).
In Epipaleolithic assemblages,
bladelet dimensions are somewhat standard and there are only small
variations between contemporaneous assemblages (Bar-Yosef 1970, 1991;
65
Henry 1989b).
By definition, bladelets are less than 1.2 cm wide and
have a length to width ratio of 2:l (Tixier 1963).
Some of the
potential variability in morphological dimensions is limited by this
definition.
Regardless of this definition, the fact that bladelets
exhibit some uniformity despite variations in core quality and
morphology suggests that Epipaleolithic knappers had a preconceived
notion on the appropriate dimensions of bladelet blanks (Bar-Yosef
1991:322).
Core Reduction
Several systematic methods and techniques have been employed in
the study of Epipaleolithic core reduction strategies, and tool
manufacture, use, and discard.
Core reduction strategies represent
primary lithic technologies, while blank modification and tool
manufacture represent secondary lithic technologies.
Initially, many of
these techniques were pursued individually and focused on only one step
in the manufacturing process.
The chaines operatoires approach
incorporates many of these earlier methods and techniques.
Primary lithic technology or core reduction includes initial core
preparation (including the removal of cortical material), preparation of
a striking platform, flake or blank removal, and core rejuvenation.
The
refitting of lithic artifacts to the core from which they were removed
is an extremely informative, albeit extremely time consuming, method
with which to study primary lithic technology and core reduction
strategies.
Refitting studies have been used to document the transition
between the Middle and Upper Paleolithic showing that cores, flakes,
blades, and elongate Mousterian points were produced during the
reduction of single nodules (Marks 1983; Marks and Volkman 1983) and in
interpretations of site formation processes and discard patterns in the
Upper Paleolithic of Southern Sinai (Phillips and Gladfelter 1991;
66
Phillips 1991).
Refitting studies have been employed less frequently cn
Epipaleolithic assemblages; although, there are some notable exceptions
(e.g., Hamifagash IV, western Negev cited in Goring-Morris 1995:156).
Sites with greater variability in lithic raw material and small sites
that appear to represent a single chipping episode will be more easily
refit than large, multi-component sites which have little variability in
lithic raw material.
As Tor al-Tareeq falls into this later category
and much of the lithic assemblage was fragmentary, this type of study
was not considered for this analysis.
Since core reduction or primary lithic technology includes the
initial core preparation, preparation of a striking platform, flake
removal and core rejuvenation, this study records the number, type cind
specific attributes of debitage and cores (see below for definitions of
debitage classes and Appendices A and B for coding sheets used in this
analysis).
study.
There are several assumptions associated with this type of
Generally, assemblages that contain high proportions of cortical
flakes are interpreted as representing initial reduction activities.
The ratio of cores to debitage and the percentage of core rejuvenation
elements in an assemblage are used as proxy measures for the relative
amount of core reduction occurring at a site.
Low core to debitage
ratios in an assemblage suggest either intensive core reduction
activities (more debitage produced from a single core) or the
importation of debitage to the site.
High proportions of core
rejuvenation elements in an assemblage generally indicate more intensive
lithic reduction as the striking platform of the core was rejuvenated so
that additional flakes and blades could be removed.
Assemblages with
more intensive core reduction strategies are frequently interpreted as
being more conservative, i.e., lithic raw materials were used more
conservatively.
Intensive core reduction has also been associated with
67
an increase in mobility (Henry 1989a; Kuhn 1994) .
Secondary lithic technologies include those reduction activities
used to modify a blank and are usually associated with tool manufacture.
The most frequent observations about secondary technologies are the type
of retouch used for tool manufacture.
Although similar edge angles may
be produced with different types of retouch, retouch types are often
used to define specific tool classes.
For example, scrapers are partly
defined by their steep angled retouch while microliths are associated
with abrupt retouch.
Another common secondary technology in some
Epipaleolithic assemblages is the microburin technicjue, a method for
truncating or segmenting a bladelet whereby a bladelet is notched and
snapped.
Microburins are the waste or end product of this technique and
are generally considered to be an intermediate stage in the manufacture
of some microlithic tools (Henry 1974; Tixier 1973).
Manufacture, Use and Discard
Analyses of retouched tools have also been approached from a
variety of avenues relating to the typology (form and function) and
technology (primary and secondary manufacture) of the artifact.
Retouched tools are traditionally associated more with typological,
rather than technological studies.
Early analyses of chipped stone
tools involved categorizing them based on a formal type list (e.g., BarYosef 1970 and Hours 1974 in the Levant).
In such typological studies,
typological variability was identified but interpreting the variability
was more problematic.
Some believed that typological variability within
an artifact class could be used to identify individual culture groups
(e.g., Bar-Yosef 1970; Hours 1974).
As researchers recognized the
importance of raw material constraints, tool maintenance activities,
curation, and discard behavior on tool morphology, the assumptions
behind typological approaches were reevaluated.
Although typologies are
68
useful for providing a standard set of terms that can be used to
facilitate commimication between researchers, interpretations of the
variability represented in tool typologies, especially those relating to
functional and stylistic variations, have been modified.
Functional variability in lithic assemblages may be reflected to
some extent in the relative frequencies of tool classes such as scrapers
and microliths.
Current functional interpretations for selected stone
tool types are based on ethnographic analogy, microwear analysis of
prehistoric tools, experimental research on microwear, and evidence for
prehistoric hafting and use.
Some of these functional interpretations
for the major tool types are presented here so that the interpretive
biases of past and present researchers will be made obvious.
Backed
microliths are generally thought to be associated with either weapons or
cutting plant products like grasses (J. Clark 1954,- Clarke 1976; Curwen
1941).
Geometric microliths were hafted as projectile points on wooden
arrows in Predynastic and Dynastic Egypt (J. Clark et al. 1974; J. Clark
1975-1977).
In addition, use wear studies of chipped stone material
from el Wad, Ain Mallaha, and Abu Hureyra suggest that geometric
microliths were hafted on bone or wood shafts and had polish from
cutting meat (Anderson-Gerfaud 1983:78-85; Buller 1983:110-112).
Smaller geometries from Abu Hureyra were interpreted as having been used
on the tips of shafts and as barbs (Anderson-Gerfaud 1983) .
At el Wad
and Ain Mallaha, nongeometric microliths were used on meat probably as a
composite tool hafted in wood (Buller 1983).
At el Wad (Garrod and Bate
1937) and Wadi Hammeh 27 (Edwards 1990), nongeometric microliths were
recovered from bone sickles and were probably used for plant processing.
Therefore, nongeometric microliths are associated with either hiinting or
plant processing, and may be associated with other activities as well.
Based on ethnographic analogies, scrapers were probably used for
69
hide processing (Hayden 1979; Nissen and Dittemore 1974) or v.'ccdv.'crking
(Gould et al. 1971; Hayden 1977:182).
Burins were probably used as
shavers and engravers of wood, antler and bone (Hayden 1977:185; Keeley
1980; Newcomer 1974; Seminov 1964).
It has also been suggested that the
bit may have been hafted onto wood or bone shafts, and that the burin
blow was one manner of blunting the end of the burin for hafting (Buller
1983:109-110; Mortensen 1970).
Burins at Beidha, are sometimes
retouched on the ends with very small burin blows (Byrd 1987:102).
If
the burin was hafted, a burin blow on the lateral edge of the blank may
also have served to narrow the hafted end of the piece.
Retouched
pieces, notches, and denticulates were probably used for a variety of
tasks.
Denticulates were probably used for cutting sinew and other
tough materials that would cut more easily with a serrated edge.
Unretouched pieces that exhibit evidence of utilization were probably
also used for a variety of tasks as an expedient tool.
Typological variation within individual tool classes such as
scrapers, burins and microliths is attributed to stylistic variability,
if such variability is thought unrelated to the function of the
implement.
Typological variations may be produced by the overall
morphology, extent, invasiveness, and positioning of retouch on the
modified blank.
Some stylistic variability in Epipaleolithic
assemblages can be attributed to different temporal and cultural
periods, e.g., lunates with Helwan retouch in the Natufian (Henry 1977,
1989b).
Some suggest that stylistic variability between contemporaneous
lithic assemblages can be used to identify different social groups
moving about the landscape (e.g., Bar-Yosef et al. 1992; Close 1978,
1989; Goring-Morris 1987; Henry 1977, 1989b, 1995).
Others question the
ability to identify prehistoric ethnicity in chipped stone assemblages
because of the difficvlty isolating lithic attributes encoded with
70
social symbols (Clark 1989, 1991).
While it is relatively easy to identify attributes that might
encode stylistic information, interpreting the meaning of this
variability is less obvious.
Much of the stylistic variability used to
distinguish ethnicity in the Epipaleolithic is associated with the
morphology of microlithic tools.
Most would agree that microliths are
frequently hafted either singly as points or knives, or together as
composite tools.
If the shape and/or retouch of microliths encode
social information, this social information may be intended for
distinguishing group identity (emblemic style) or individual identity
(assertive style) (Wiessner 1983:257-258).
If this were the case, much
of the social information would be "lost" when the microlith is hafted.
In addition, much of the typological variation within the microlithic
tool class would be completely unrecognizable from distances greater
than two or three meters away (c.f. Wobst 1977).
Therefore, it is
unlikely that microlithic variability is associated with group
recognition or identification.
The typological variability may also
represent emic notions of how to "correctly" retouch a microlith.
Although these notions could be established ethnographically, they are
difficult to demonstrate archaeologically.
Therefore in this study,
more emphasis is placed on variations in the proportion of tool classes
than on variation within a single tool class.
The view here is that technological variability in microliths and
other chipped stone tools has potential to inform on social groups if
one accepts that technological approaches to core reduction and tool
manufacture are learned behaviors.
Technology of retouched tools
includes the preferential selection of blanks for tool manufacture.
Several studies have been conducted on the selection of blanks for tool
manufacture, (e.g., Henry 1973; Marks 1983; Olszewski 1989).
Secondary
71
technology includes the microburin techniqu- snd manner of retouch.
Although the type of retouch may not vary significantly within a major
tool class, secondary technologies may indicate technological choices
used in the manufacturing process to compensate for other factors such
as expediency and raw material (specifically quality, size and shape of
the raw material) constraints.
The discard patterns in the assemblages analyzed here will be
identified by core to debitage and debitage to tool ratios, and
percentages of major tool classes.
These patterns will then be compared
to other contemporaneous levels which will help determine intra-site
functional variability.
Contemporaneous levels will be determined
either by reliable radiocarbon determinations and stratigraphic
positioning, or by a combination of stratigraphic positioning and
typological similarities in the microliths {i.e., the relative
proportions of geometric and non-geometric microliths).
Although the
later approach is less reliable, the absence of radiocarbon
determinations in some levels prohibits the use of the first approach.
Sampling Rationale
Previous analyses of the Tor al-Tareeq lithic assemblage indicated
the presence of the microburin technique in the site's earliest levels,
(Step B level 5, Unit B level V, Step C level 7, and Unit C levels III
and IV) (Clark et al. 1987, 1988; Donaldson 1986; Neeley et ai. 1995).
In stratigraphically younger levels (Step C level 5 and Unit C level
II), there was an increase in the proportion of geometric microliths
accompanied by a significant decrease in the frequency of the microburin
technique (Clark et al. 1987, 1988; Donaldson 1986; Neeley et al. 1995).
I wanted to study the technological and typological variability
associated with this transition (i.e., from a tool assemblage dominated
by nongeometric microliths and use of the microburin technique to a tool
72
assemblage with, high proportions of wide geometric microliths and an
almost complete absence of the microburin technicjue) in order to better
understand why these techno-typological changes occurred.
Particularly,
I was interested in determining to what extent these technological and
typological changes reflect only diachronic change or a combination of
diachronic change, and changes in subsistence and land use strategies in
response to paleoenvironmental change.
This transition was best represented in natural levels 5 and 7 of
Step C (Neeley et al. 1995).
I reanalyzed all of the lithic material
(debitage, cores and retouched tools) from the northern half of natural
levels 5 and 7 in Step C.
Natural level 5, which is characterized by
grayish brown to brown silt/sand sediments, corresponds to arbitrary
levels COS-CIO (Table 2.1).
Natural level 6, an intervening cobble
layer, probably indicates an erosional period of unknown duration and
lies between natural levels 5 and 7 but has no associated artifacts
(Clark et al. 1987, 1988).
Natural level 7, a compact light brown
sediment, corresponds to arbitrary levels C11-C15.
As the references to arbitrary and natural levels, and units and
steps can be confusing, I want to clarify that most of the following
discussion refers to Step C arbitrary levels 8-15 (i.e., C08N-C15, see
Table 2.1).
Arbitrary levels will always be referred to by the step,
level, and provenience (north, south, or undifferentiated).
Therefore,
level C08N refers to Step C, arbitrary level 8, northern half of the
excavated level; level 014 refers to Step C, arbitrary level 14, and the
combined northern and southern halves of the excavated level.
The
corresponding natural levels in Step C will be referred to as natural
levels 5 and 7.
This step and these specific levels were chosen for various
reasons.
First, the technological and typological transition was best
73
represented in Step C natural levels 5 and 7.
Althaugh the general
character of the assemblage from Step C is known from in-field analyses
and a sample of the microliths has been previously analyzed (Clark et
al. 1987:52-67; Donaldson 1986), a thorough description of this
assemblage has not yet been conducted.
Also, since the microburin
technique is not usually associated with Kebaran lithic industries
(e.g., Bar-Yosef 1970) a strict late Kebaran terminology (Clark et al.
1987, 1988; Neeley et al. 1995) does not seem appropriate for the
assemblage from Step C, natural level 7.
Even using the general term
Kebaran Complex, though certainly chronologically acceptable, does not
significantly aid in sorting out the relationship of this site to others
in the eastern Levant.
Levels COBN-CIO exhibit an increase in the
proportion of geometric microliths which suggests an association with a
different, later microlithic complex, perhaps the Geometric Kebaran
(14,500-13,000 BP) or Mushabian (14,000-12,5000 BP) as defined by
Phillips and Mintz (1977) .
Based on the wide, atypical lunates in the
upper levels of Step C (specifically levels C08-C10) , it has been
suggested by some that the upper levels may be associated with a
Natufian component at the site (12,800/12,500-10,500 BP)(Clark et al.
1987, 1988; Coinman et al. 1989).
Epipaleolithic cultures are largely
defined (both temporally and spatially) on the basis of this technotypological variability in the flaked stone assemblage.
Therefore,
identifying and interpreting this variability at the site level is
necessary before occupations of the site can be placed in a regional
context.
Furthermore, the in-field analyses were conducted by people with
different levels of expertise.
As a result, many retouched pieces and
microburins were not consistently identified.
The lowest natural levels
in Step C (levels 5 and 7) had high artifact densities and are
74
considered to be in situ.
Finally, level C13 has two reliable
radiocarbon determinations on charcoal, one from an associated hearth
(feature 5) dated to 16,570 + 380 BP {UA-4390) and another from charcoal
found in level C13 dated to IS,790 + 340 BP (UA-4393) (Clark et ai.
1987, 1988) .
This situation facilitates comparisons between this study
and studies on contemporaneous assemblages from other Levantine areas.
During the 1984 excavation, the northern and southern areas of
these levels were excavated and collected separately.
Since the north
end produced more artifacts, I analyzed only the lithics recovered from
the north end of Step C (levels C08N, C09N, C11N-C13N).
however, three exceptions.
There were,
Although the north and south ends of level
CIO were excavated and collected separately, this provenience
information was not recorded on the three bags of lithics recovered from
that level.
analysis.
I selected the largest bag (by volume) from level CIO for
Also as the artifact density decreased significantly in
levels C14 and CIS, the north and south ends were not collected
separately.
Therefore, I analyzed all of the lithics collected from
these levels.
While the uncertainties with sampling prohibit direct
comparisons of artifact frequencies (counts) between levels, the
relative frequencies can still be compared.
A total of 7952 pieces
(7093 pieces of debitage and 859 retouched pieces) were analyzed for
this study.
Analysis of Cores and Debitage
The technological analysis of the debitage is concerned with
primary (core reduction) and secondary (retouch and microburin
technique) lithic technologies.
First, the chipped stone was sorted
into major debitage, core, and tool categories.
potential blank for tool manufacture.
Debitage represents any
Debitage was sorted by major
blank categories into flakes, blades, bladelets, microburins, burin
75
spalls, core rejuvenation elements, debris and shatter (for debitage
coding list see Appendix A).
Flakes represent any blank with a single interior surface
(Sullivan and Rozen 1985:759) and does not contain any of the
characteristics used to define the other debitage classes.
In this
analysis, all complete flakes and proximal flake fragments are counted
as flakes.
are wide.
Blades are flakes that are at least twice as long as they
Although blades frequently have parallel sides, they do not
necessarily have to.
Bladelets are blades with widths less than 1.2 cm
and lengths generally less than 3 cm (Tixier 1963) .
The division
between blades and bladelets is arbitrary; therefore, "blades" will be
used to refer to both blades and bladelets unless otherwise specified.
For both blades and bladelets, the number of dorsal flake scars was
recorded.
A first order blade is a blade with one ridge (a high point
between two flake scars) on its dorsal surface.
A second order blade
has two or more parallel ridges on its dorsal surface.
Burin spalls are
generally small flakes with triangular or rectangular cross-sections and
are removed during burin manufacture.
Core rejuvenation elements are
any flake or blade that was removed to prepare the core for additional
flake removals such as crested blades.
distal flake fragments.
Debris consists of medial or
This category is also referred to simply as
"flake fragments" in other studies (e.g., Sullivan and Rozen 1985).
The
shatter category, as used in this analysis, consists of any angular
fragment produced during lithic manufacture that does not have a single,
discemable, interior surface.
This category has also been referred to
as "debris" by some researchers (e.g., Sullivan and Rozen 1985:759).
other analyses, the shatter category sometimes combines shatter and
debris categories as they are defined here (e.g., Clark et al. 1987,
1988).
In
76
Microburins, as previously mentioned, are formed through the
intentional notching and snapping of bladelets.
Based on the appeareince
of the microburin scar, three types of microburins are recognized,
regular, piquant tiedre, and Krukowski (Tixier 1963).
Although part of
the debitage, analyses of microburins are frequently presented in
discussions of retouched tools.
of the debitage.
Here, microburins are counted as part
However, specific technological and typological
attributes of microburins such as type of striking platform, length,
width, and thickness were recorded during the attribute analysis of
retouched tools (see Appendix B for a list of coded attributes).
For
each debitage class, the number, presence of cortex, and completeness
category (complete or proximal, medial, and distal fragment) were
recorded.
Additional analyses, mainly recording length, width, and
thickness of blade and bladelets, were also conducted.
Cores include all pieces from which three or more blanks have been
removed.
Core fragments are cores that appear fractured.
During the
analysis of the debitage, the frequency of cores and core fragments was
recorded.
Additional information (core type, amount of cortex, lithic
raw material, and maximum length, width, thickness and weight) was also
recorded during the attribute analyses (see Appendix B for a list of
coded attributes).
Much of the chipped stone material is incomplete or fragmentary.
Therefore in an effort to learn more about the geologic integrity of
these deposits, some of these debitage categories (flakes and debris)
were further subdivided into four length categories (<1, 1-2, 2-3, >3
cm).
The assumption is that breakage rates and size categories may be
used as proxy measures for the amount of post-depositional disturbance
of these assemblages.
Archaeological deposits sxibjected to greater
post-depositional disturbance may exhibit an increase in the proportion
11
of fragmentary debitage.
Also, these data may be useful in interpreting
patterns in the debitage cuid tool components of the assemblage.
Analysis of Retouched Tools
In the analysis of retouched tools, I recorded technological
(primary and secondary lithic technologies) as well as typological
(formal) attributes of retouched and utilized tools in each assemblage.
Therefore, an attribute analysis was conducted on all of the retouched
and utilized pieces from levels C08N-C15.
The coding sheets used in
this analysis are a combination of those employed by several
researchers.
The tool typology with some modifications is based largely
on Tixier (1963) and Goring-Morris (1987)• while the technological
attributes are based on Phillips and Yerkes (1979).
In the technological analysis of retouched and utilized pieces,
the primary technological attributes recorded for each retouched tool
and utilized piece included original blank type, amoimt of cortex,
distribution of dorsal flake scars, platform type, lithic raw material,
length, width, thickness and weight.
Several secondary technological
attributes (placement, orientation and type of retouch) were also
recorded.
The typological analysis included assigning each artifact to a
major tool class.
The classes used were endscrapers, burins, multi­
purpose tools, retouched blades, tnancated flakes and blades,
nongeometric raicroliths, geometric microliths, retouched flakes,
retouched pieces, notches and denticulates.
Bach of these major tool
classes was further divided into subtypes based on the type of blank,
location and type retouch, or the overall morphology of the tool.
In
addition, the shape of the distal end and lateral edges were recorded
for each tool.
Although some of this information is encoded in the
subdivisions of specific tool types, this information was recorded
78
separately so that variations between artifact classes could be mere
easily accessed and quantified.
Not all of the attributes recorded are analyzed in this study.
Many of the attributes recorded relate to secondary lithic technologies
and/or subtle typological variations.
Although this analysis focuses on
primary lithic technologies associated with core reduction activities
and tool typology, some of the secondary lithic technologies and
typological variation exhibited considerable consistency between levels
and did not show significant variability within the small sample size.
All of the data were entered into dBase IV version 4.0 during analysis.
Later, these files were imported into SYSTAT version 5.04 for data
manipulation and statistical analyses.
79
CHAPTER 4:
LITHIC ANALYSES
Technological and typological attributes of the debitage and tool
components of the chipped stone assemblage from stratigraphic levels
COSN-CIS were analyzed according to the method presented in the previous
chapter.
Chipped stone debitage represents both primary and secondary
lithic technologies {Goring-Morris 1987:48).
Primary technology
includes core preparation and blank removal, while secondary technology
includes s\absequent blank modification such as the microburin technique,
i.e., purposefully sectioning blanks for later use.
Chipped stone tools
also represent technological (blank selection) and typological
(stylistic) choices.
This chapter presents a technological description
and analysis of the chipped stone debitage followed by technological and
typological analyses of the chipped stone tools for levels C08N-C15.
Debitage
The classification of debitage was hampered by the fragmentary
nature of the chipped stone assemblage.
Only 8.55-16.34% of the chipped
stone flakes, blades, and bladelets in each stratigraphic level are
complete (Table 4.1).
of each assemblage.
Proximal debitage fragments make up 25.42-34.41%
The remaining debitage (53.38-66.10% of each
assemblage) is comprised of medial or distal fragments.
The
completeness of cores, core fragments, core rejuvenation elements,
microburins, and burin spalls was not recorded.
These categories occur
only in low frequencies, 4.28% of the entire chipped stone assemblage
(Table 4.3).
Fragmentation may have affected the recognition of some
categories such as burin spalls since identification of burin spalls
requires a triangular cross section and a hinged flake scar termination
which may be absent in fragments.
Other categories can be recognized
despite their degree of completeness.
be small (1-2 cm) and complete.
Therefore, burin spalls tended to
Since there is little evidence for
80
Teible 4.1:
Percentages of completeness categories for *flakes, blades
and bladelets by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
Complete 1 8.55
16.34
10.91
11.43
12.25
9.29
12.41
8.47 | 10.90
27.45
33.45
33.58
34.38
34.41
30.08
8.82
8.63
7.08
14.00
7.60
10.65
47.39
47.00
47.91
39.38
48.70
46.87
25.42 | 32.57
|
3.39 | 10.26
|
62.71 | 46.28
100.0
306
100.0
834
100.0
551
100.0
800
100.0
1119
100.0
798
100.0
177
I
Proximal |32.60
fragment|
Distal
|15.38
fragment|
Medial
|43.47
fragment
TOTAL
100.0
NUMBER
865
C15
TOTAL
I
100.0
5450
* Completeness categories for cores, core fragments, core rejuvenation
elements, microburins and burin spalls were not recorded.
Table 4.2:
Percentages of medial and distal fragments classified as
blades and bladelets, and debris by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
Blades andl20.04
bladelets|
Debris
|79.96
23.26
27.59
25.74
20.61
18.57
20.70
76.74
72.41
74.26
79.39
81.43
79.30
20.5l| 21.81
{
79.49| 78.19
100.0
172
100.0
464
100.0
303
100.0
427
100.0
630
100.0
459
100.0
117
TOTAL
NUMBER
100.0
509
C15
TOTAL
100.Q
3081
81
Debitage and tool percentages by level.
C08N
C09N
24.70
19.55
Blades/
15.13
bladelets
Core rejuv. 1.57
Flakes
Microburins
•
Burin spall
CllN
C12N
C13N
24.09
23.93
to
Table 4.3:
CIO
92
22.18
20.00 19.43
22.95
15.89
20.02
16.07
13.65
13.57
14.44 14.57
15.29
1.83
1.77
1.84
1.80
1.12
1.75
.00
1.56
1.35
1.03
2 .01
1.51
.40
1.07
00
00
16
.00
.13
09
•
61
•
35
17
•
41
•
27
-
12
•
C14
•
CIS
TOTAL
Debris
35.39
26.88
29.76
27.61
29.10
3 0 . 27
28.89 37.65
30.29
Shatter
12.35
14.66
12.05
12. 76
16.65
19.29
21.35 23 .89
16.39
1.22
1.15
86
1.03
65
•
16
.00
.73
80
1.60
1.29
71
•
56
.00
.79
Cores
•
51
Core frags.
•
26
•
81
•
•
•
Tools
7.13
12.22
7.35
9.69
5.67
6.43
7.46
3.24
7.31
Tool frags.
2.61
5.91
2.39
4.17
3.86
3.78
3 .73
.81
3 .50
100.0
247
100.0
7952
TOTAL
NUMBER
Table 4.4:
100.0
1150
100.0
491
100.0
1129
100.0
815
100.0
1165
100.0
1695
100.0
1260
Ratios and indices of various artifact classes by level.
Flake:Blade
Debitage:Core
Shatter:Debitage
Fragments:Debitage
C08N
1.63
47.9
0.30
0.85
C09N
1.23
18.8
0.38
0.70
CIO
1.20
23.86
0.26
0.64
CllN
1.49
17.65
0.30
0.64
C12N
1.90
18.30
0.39
0.67
C13N
1.64
28.65
0.50
0.78
C14
1.38
53.0
0.56
0.76
CIS
1.33
0.00
0.69
1.09
82
burins in the chipped stone tools (Table 4.13). it is doubtful that the
fragmentary nature of the assemblage would have significantly affected
the recognition of burin spalls.
Since the preponderance of flakes, blades, and bladelets in each
level are fragmentary, statements about the frequency of blade and flake
blank manufacture and the selection of blanks for tool manufacture could
be affected by the treatment of this fragmentary material.
The
classification of the broken pieces used in this analysis follows Byrd
(1987:91).
If a broken piece still had a length to width ratio that was
2:1 along its striking axis, it was classified as a blade or bladelet.
If the length to width ratio was less than 2:1, it was placed either
into proximal or medial and distal fragment categories.
Complete blcinks
and proximal blank fragments were used to determine the total number of
flake and blade blanks in each assemblages (Table 4.3).
The medial and
distal fragments with a length to width ratio less than 2:l were not
counted among flake and blade blanks since multiple medial fragments may
be produced from a single blank.
The inclusion of medial and distal
fragments in the overall flake and blade counts might artificially
inflate this category.
The number of flakes recorded in this analysis may be more
"upwardly biased" than other classification schemes.
Medial
blade/bladelet fragments with 2:1 length to width ratios were counted as
blades/bladelets.
However, if the length to width ratio was less than
2:1, the proximal end could be considered a flake.
Therefore, some of
the proximal blank fragments may actually have derived from blades and
bladelets, not flakes.
However, this situation would be equally
problematic if proximal blank fragments with length to width ratios of
less than 2:1 were assigned to the blade/bladelet category.
Some may
assign any parallel-sided, proximal blank fragment to a blade and
83
bladelet category.
This too can be problematic as the proximal ends of
some flakes may have parallel sides at the proximal end.
In addition,
the length of proximal fragments influences the appearance of the sides
of the flake.
Although 72.41-81.43% of the medial and distal fragments
were debris, i.e., not classifiable to a specific blank category, 18.5727.59% of the medial and distal fragments had length to width ratios of
at least 2:1 (Table 4.2) and were classified as blades/bladelets.
Some variability in the frequencies of debitage classes exists
between levels, but there is considerable consistency overall (Table
4.3).
The proportions of flakes (19.43-25.92%) and blades and bladelets
(13.57-16.07%) vary only slightly between levels.
Blades and bladelets
comprise slightly greater percentages of the total assemblage in levels
C08N-C11N than in levels C12N-C15.
Bladelets comprise approximately
one-third (30.1-33.3%) of the total debitage blanks, i.e., flakes,
blades, and bladelets, from levels C10-C13N (Table 4.5).
In the
remaining levels, the frequency of bladelets is slightly higher in
levels C09N, C14 cind C15 than in level C08N (25.5%) .
Blades comprise
almost twice the proportion of the total assemblage in the upper levels
(C08N, ClO-CllN) as in the lower levels (C12N-C15)(Table 4.5).
Again,
C09N has an anomalously low percentage of blades and a high percentage
of bladelets.
The variability in levels C09N and CIS may be
attributable to the relatively small samples from these levels.
Debris comprises approximately one-third (27.61-37.65%) of each
assemblage (Table 4.3).
Microburins are rare in the upper levels {C08N-
ClO) but markedly more abundant in levels C11N-C14.
Microburin
frequencies and their relationship to geometric and non-geometric
microlith frequencies will be discussed in greater detail below.
Burin
spalls are rare in all levels, but slightly more abundant in levels
C08N-C11N.
Cores, core fragments and core trimming elements are
84
Table 4.5:
Percentages of size categories for complete and proximal
flakes, blades, and bladelets by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
C15
Total
Flakes
<1 cm
13 54
17.24
16.67
19.94
25.60
28.05
24.88
25.00
21.60
1-2 cm
31 44
25.86
24.90
29.45
27.98
26.57
23.96
20.24
26.96
2-3 cm
12 01
8.05
7.63
5.83
7.59
4.95
7.14
5 02
4 .02
5.42
4.60
4.34
2.48
2.07
4.76
3.95
Blades
12 45
5 .17
12.45
10.74
3 .47
6.94
4.38
5.95
8.55
Bladelets
25 55
39 .66
32.93
29.45
31.02
31.02
37.56
36.90
31.93
100.0
174
100.0
498
100.0
326
100.0
461
100.0
606
100.0
434
100.0
84
100.0
3041
>3 cm
TOTAL
NUMBER
100.0
458
Table 4.6:
<1 cm
7.14
7.50
Percentages of debris size categories by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
CIS
|43.49
40.91
39.88
40.44
50.44
61.60
56.32
49.46
49.56
43 .18
37.50
44.00
37.76
30.02
37.91
46.24
37.73
12.88
16.37
12.00
9.73
6.63
5.49
4.30
9.96
1
1-2 cm i40.29
Total
j
2-3 cm [12.29
j
>3 cm
1 3.93
3.03
6.25
3.56
2.06
1.75
.27
.00
2.74
TOTAL
100.0
407
100.0
132
100.0
336
100.0
225
100.0
339
100.0
513
100 .0
364
100.0
93
100.0
2409
NtMBER
85
relatively rare in all levels.
This result is consistent with in-field
analyses of the debitage (Clark at al. 1987:59).
Levels C09N-C12N have
slightly higher percentages of cores, core fragments, and core trimming
elements than other levels.
assemblage (Table 4.3).
Shatter varies between 12.05-23.85% of each
The lower levels (C10-C15) have higher
percentages of shatter than the upper levels (C08N-C09N).
Differences in the frequencies of blade and bladelet blanks
between levels are affected by the relative frequency of other debitage
and tool categories in that level (Table 4.3).
Flake to blade ratios
may be a better comparative measure because variation in other debitage
classes will not affect these ratio values (Table 4.4).
However, if
bladelet blanks are preferentially selected for tool manufacture, their
numbers would be reduced in the debitage.
Tool blanks were not included
in comparisons of blank frequencies (Table 4.4).
In-field debitage
analyses of natural levels 5 and 7 from Step C indicated that nearly
equal numbers (0.8 and 0.7 flake to blade ratios, respectively) of
flakes and blades were manufactured (Clark et al. 1987; Donaldson 1986).
Earlier studies suggested that blade manufacture was more heavily
emphasized in the earlier occupations of the site, specifically, natural
level 2 of Step C and natural levels 3-4 of Step B (Clark et al. 1987;
Donaldson 1986).
Similar results were obtained in this analysis for
levels C09N and CIO.
However, levels C08N and C12N produced almost
twice (1.77-1.85) as many flakes as blades.
Levels CllN and C13N-C15
produced 1.25-1.48 times as many flakes as blades (Table 4.4).
Since the assemblage is so fragmentary, flakes, flake fragments,
and debris were further classified into 1 cm size categories (Tables 4.5
and 4.6).
It was hoped that this size classification scheme might be
able to inform on (1) the degree to which post-depositional disturbance
and breakage rates differed between levels, (2) the relationship between
86
the number and size of the flakes, flake fragments, and debris, and the
number ajid size of whole blade blanks, and (3) the potential effect of
these values on flake to blade ratios.
First, the frequencies of
blades, bladelets, and flakes were determined (Table 4.3).
Then,
complete and proximal flake categories were further subdivided into
four, 1 cm size classes (Table 4.5).
The unclassifiable debris was also
divided into four, 1 cm size classes (Table 4.6).
The size distribution pattern of complete and proximal flake
fragments differs slightly from the pattern found in debris.
Complete
and proximal flakes in the 1-2 cm length category comprised the majority
of the flake blanks (24.9-31.4%) in levels C08N-C11N (Table 4.5).
Levels C13N-C15 have only slightly more <1 cm size flakes (by 0.9-4.8%)
than 1-2 cm flakes.
In contrast, the <1 cm debris is consistently the
most abimdant size category (50.4-61.2%) with the 1-2 cm debris (29.146.2%) being the second most abimdant size category.
However, the
relative cibiandance of these categories varied between levels.
In levels
C12N-C14, the <1 cm size debris was 12.68-31.58% more aibundant than in
other levels (Table 4.6).
In levels C08N-C11N and C15, the difference
between the <1 cm and 1-2 cm size debris was less pronounced varying
only 2.7-3.7% between levels.
The 2-3 cm and >3 cm size debris are also
more abundant in levels C08N-C11N thsin levels C12N-C15.
Although the
patterning of 2-3 cm size debris is not present in complete and proximal
flakes, there is a trend for slightly more >3 cm size flakes in all
levels than >3 cm size debris.
There is an inverse relationship between the length of debris and
the relative abundance of debris in all levels.
Interestingly, the
lowest levels (C12N-C15) have at least 10-20% more <1 cm size debris
than the upper levels (CGSN-CllN).
This break between levels CllN and
C12N in the proportion of small debris is also present between these
87
levels in the proportion of small sized flakes.
Based on this
information, it appears that the lowest levels have more small debris
either as a result of debris being eroded down slope or being broken
into smaller fragments in these levels.
The original blank morphology may influence its susceptibility to
post-depositional breakage.
In general, blade and bladelet blanks in
levels C12N-C14 are narrower and very slightly thinner than in levels
C08N-C11N (see discussion on debitage morphometries and Figures 4.2-4.6
below).
These data cannot be used to support a causal relationship
between blank morphology and general size of debris.
However, they do
suggest that assemblages with narrow and thin blanks are more
susceptible to post-depositional stress than wide and thick blanks.
Additional comparisons of flake widths and thicknesses from each level
would provide additional support for this relationship (see section on
the morphometries of flake tools below).
High frequencies of shatter usually indicate more primary core
reduction (Clark et al. 1987, 1988; Sullivan and Rozen 1985); however,
extreme weather conditions in desert environments may inflate the
frequency of shatter as surface scatters have "baked and cooled for
millennia in the desert environments of the Near East" (Clark et al.
1987:54).
Shatter percentages reported in previous analyses (Clark et
al. 1987, 1988) are considerably higher than those reported here.
From
in-field analyses, shatter was "more common in the excavated sample than
on the site surface (42.9% verses 34.2%)" (Clark et al. 1987:59).
Shatter to debitage ratios from in-field analyses produced values
greater than one for all natural subsurface levels except Step C, level
4 (0,76) which may have been exposed to surface weathering processes for
a longer period of time than the other natural levels (Clark et al.
1987, 1988) .
88
In this analysis, the amoiant of shatter varied between 12.0523.89% of each assemblage (Table 4.3). The proportion of shatter is 2-4%
lower in levels C08N-C11N than levels C12N-C15.
Also, shatter
percentages increase progressively between levels C12N and C15.
Shatter
to debitage ratios for levels C08N-C15 indicate lower ratios for levels
C08N-C12N and higher ratios for levels C13N-C15 (Table 4.4).
Differences between in-field amalyses and this cinalysis are probably the
result of differences in the definition of shatter between the two
studies.
Clark's (ec al. 1987) shatter category seems to include both
shatter and debris.
If the debris and shatter categories presented in
this study are combined, values more similar to those presented by Clark
(et al. 1987) are obtained.
If increased amounts of shatter represent increased frequencies of
primary core reduction as have been suggested by some (Sullivan and
Rozen 1985; Clark et al. 1987), then more primary core reduction may
have occurred in the lower than in the upper levels.
If this is true,
other indicators of primary core reduction, such as the frequency of
debitage with cortex, should be higher.
However, a comparison of cortex
frequency in the debitage and debris by level indicates that the
frequency of cortex varies little between levels (Tables 4.7 and 4.8).
Only level C13N has a higher percentage of cortex (by 8.76-16.03%) than
other levels, suggesting that more initial core reduction occurred in
this level than in other levels.
This conclusion is supported somewhat
by an analysis of the amount of cortex remaining on cores.
The majority of cores (50-100%) in all levels had some cortex.
However, there is some variability in the amount of cortex (Table 4.9).
Approximately 50% of the cores in levels C08N and C09N were noncortical; no cores had more than 50% cortex.
In contrast, levels ClO-
C13N had low frequencies (1-2) of cores with 51-99% cortex.
Levels
89
Table 4.7:
Percentages of cortex for flakes, blades, bladeletS; and
debris by level.
CIS
C09N
CIO
CllN
C12N
C13N
No cortex 79.88
78.76
76.86
78.40
79.50
71.58
76.57
77.40
76.88
Cortex
20.12
21.24
23.14
21.60
20.50
28.42
23.43
22.60
23 .12
TOTAL
NUMBER
100.0
865
100.0
306
100.0
834
100.0
551
100.0
800
100.0
1119
100.0
798
100.0
177
100.0
5450
Table 4.8:
C14
TOTAL
C08N
Percentages of cortex for flakes, blades, eind bladelets
by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
C15
TOTAL
No cortex 78.60
79.89
78.11
79.45
75.27
63.86
73.27
72.62j 74.32
Cortex
21.40
20.11
21.89
20.55
24.73
36.14
26.73
27.38| 25.68
TOTAL
NUMBER
100.0
458
100.0
174
100.0
498
100.0
326
100.0
461
100.0
606
100.0
434
100.0
84
Table 4.9:
C09N
No cortex
42.86
50.00
38.46
1-50% cortex
57.14
50.00
.00
.00
51-99% cortex
TOTAL
Table 4.10:
NUMBER
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
100.0
3041
Percentages of cortex on cores by level.
C08N
N
I
100.0
7
100.0
6
CIO
C12N
C13N
.00
25.00
27.27
.00
29.31
46.15
85.71
66.67
63.64 100.00
62.07
15.38
14.29
8.33
100.0
13
CllN
100.0
7
100.0
12
9.09
100.0
11
C14
.00
100.0
2
TOTAL
8 .62
100.0
58
Sununary statistics for core weights by groups: (1) levels
C08N-C10, (2) levels C11N-C13N, and (3) level C14.
GROUP 1
26
10.1
157.7
147.6
46.2
36.1
GROUP 2
30
12.1
154.7
142.6
41.2
28.3
GROUP 3
2
23.6
45.7
22.1
34.7
34.7
90
C11N-C13N had more cores with some cortex (1-50% cortex) and fev/er ncncortical cores than levels C08N-C09N.
Level C14 has such a small
sample that the significance of the frequency of cortical cores in this
level cannot be evaluated.
These data suggest that more initial core
reduction occurred in levels C10-C13N than in levels C08N and C09N.
Core weights were used to access the relative size of cores.
The
assumption is that more intensive core reduction should result in
smaller cores.
Since over 90% of the entire chipped stone assemblage is
comprised of grayish brown chert, core weights can be used as proxy data
for relative core size.
If different lithic materials were being used,
significcuit variability in core weights might reflect differences in raw
material densities, not relative core sizes.
Since the frequency of
cores in all levels was low, levels were combined into three groups
which reflect natural levels and changes in artifact densities.
Group i
corresponds to levels C08N-C10, group 2 to levels C11N-C13N, and group 3
to level C14.
A box plot of core weights by group was produced in order
to determine if core sizes varied significantly between groups (Figure
4.1).
Based on the overlapping notches in each group, these cores
cannot be separated into two statistically different populations.
The
median values of core weights, while very similar, are slightly greater
in group l (levels C08N-C10) than in group 2 (levels C11N-C13N) (Table
4.10).
Based on core weights alone, the relative intensity of core
reduction between levels cannot be determined.
The relative intensity of core reduction in each level can be
further evaluated by comparing debitage to core ratios between levels
(Table 4.4).
Levels C08N and C14 have the highest debitage to core
ratios, 2-3 times higher than other levels.
Levels CIO and C13N have
only slightly greater values than the other levels.
Still, these ratios
suggest that more core reduction occurred in levels C08N and C14 than in
91
Fiaure 4.1:
Box plot of core weights in grams by groups: (1) levels
C08N-C10, (2) levels"cilN-C13N, and (3) level C14.
200
iOU
=
100
d)
'I
S'rangraDhic group
92
levels CIO and C13N.
Levels CIO and C13N had only slightly more
secondary core reduction than other levels.
Alternatively, core
reduction activities may have occurred at another portion of the site.
Therefore, the debitage to core ratios presented here may represent only
the relative proportion of secondary core reduction activities at this
location.
Although cores occurred in relatively low frequencies in all
levels, some general trends by groups (levels C08N-C10 and levels CllNC14) are present.
These trends reflect differences in lithic reduction
techniques between stratigraphic groupings.
The most ubiquitous core
types include single platform blade, bi-directional blade (a combination
of opposed, bipolar, and double platform blade cores), multiple platform
flake, and multiple platform flake and blade cores (Table 4.11).
Interestingly, group 2 (levels C11N-C14) has about 11% more single
platform blade cores, and about 10% more multi-platform flake and blade
cores than group l (levels C08N-C10).
In contrast, bi-directional cores
comprise roughly 21% more of the core assemblage in group 1 than in
group 2.
Single and multi-platform flake cores occur in slightly higher
percentages in group l than in group 2, while multi-platform amorphous
cores occur only in group 2.
In the debitage, flake to blade ratios suggested that more flakes
than blades were being manufactured in all levels.
However, blade cores
are the most abundant core type in both stratigraphic groups.
Although
flake cores occur in all levels, they occur in slightly higher
percentages (by about 8%) in group 1 than in group 2.
This study
suggests that the lithic assemblage in group 2 was more likely to be
manufactured from single platform blade and multi-platform flake and
blade cores than group 1.
Group 1 was more likely to have to have been
manufactured from either flake or blade cores.
Combination cores (i.e..
Table 4.11:
Percentages of core types by groups: (i) levels
CO8N-CI0T (2) levels cilN-ci4T
Group 1
Group 2
Platform type
Single platform blade
23 08
34.38
29.31
Opposed platform blade
11 54
6.25
8.62
Bipolar platform blade
7 69
.00
3 .45
Double platform blade
11 54
3 .12
6.90
Single platform flake
7 69
3 .12
5.17
Double platform flake
3 .85
9.38
6.90
15 38
6 .25
10.34
.00
3 .12
1.72
15 .38
25.00
20.69
.00
9 .38
5.17
3 85
.00
1.72
100 00
26
100.00
32
100.00
58
Multiple platform flake
Opposed platform
flake and blade
Multiple platform
flake and blade
Multiple platform
amorphous
Other
TOTAL
NUMBER
TOTAL
94
flake and blade cores) in group l occur in low percentages.
Debitage Morphometries
During the Epipaleolithic, there is a trend towards the
manufacture of increasingly wide and short bladelet blanks (Henry
1989b) .
Since chronometric dates have only been obtained from level
C13N and bladelet blank morphology is somewhat temporally sensitive, the
morphometries of blade and bladelet blanks were recorded.
In addition,
morphometric data can inform on the size criteria of blanks selected for
tool manufacture and potentially inform on the issue of postdepositional breakage.
Width and thickness measurements were recorded
for all blade and bladelet blanks in the debitage.
Length measurements
were recorded only for complete blade and bladelet blanks.
A notched box plot of the widths of unmodified blade and bladelet
blanks (Figure 4.2) shows that widths in levels C12N-C14 are
statistically different with a 0.5% confidence interval from widths in
levels C08N-C10, CllN and CIS.
In notched box plots, notches represent
95% confidence intervals for the sample median.
If box plot notches
overlap, then the two samples cannot be separated into two statistically
different populations.
Surprisingly, statistical differences between
bladelet widths did not occur only between the natural levels identified
during excavation (i.e., levels CIO and CllN).
The unmodified blade and
bladelets blanks in levels C08N-C10 and C15 are wider than levels C12NC14.
Level CllN has unmodified blade and bladelet blanks with widths
between those foimd in levels CIO and C12N.
Since I expected less statistically variability in the box plots,
I wanted to determine if the shape of the distribution of bladelet
widths between adjacent levels also had statistically significant
variability.
Therefore, a Kolmogorov-Smimov two sample test was
conducted to more accurately determine the statistical difference
95
Fioure 4.2:
Box plot of the widths of untnodified blsde and blsdslet
blanks by level.
40
30
E
E
- 20
n
10
0
j_
Level
96
between the shape of the distribution of bladelet widths in each pair of
adjacent levels (Table 4.12).
This test is appropriate for these data
because blade and bladelet widths are not normally distributed, metric,
and derived from two independent samples.
The null hypothesis that the
two samples are the same Ccin be rejected for all of the tests except
those between levels CIO and CllN.
Since the probability between CIO
and CllN (O.OOl) lies below 0.01, the null hypothesis can be rejected.
Therefore, levels CIO and CllN probably do not come from the same
population.
Again, differences in blade and bladelet blank widths can be seen
in levels C08N-C10 and C11N-C13N.
This division coincides with the
natural levels identified in this excavation unit.
Although the median
width value was greater in level CllN than in levels C12N-C14, the shape
of the width distribution for level CllN does not differ statistically
from levels C12N-C14.
The median width value and shape of distribution
of unmodified blade and bladelet widths in level CllN suggests that
lithic manufacture in level CllN may be transitional reflecting a change
in lithic manufacture from the manufacture of narrow bladelets in levels
C12N-C15 towards the manufacture of wide bladelets in levels C08N-C10.
No significant differences in the thickness of blade and bladelet
blanks (Figure 4.3) between levels are discemable.
Initially, it was
thought that if a relationship between width and thickness could be
established, i.e., wide bladelet blanks are also thick, then thickness
measurements on tools could be used as proxy data for determining the
width of the original blank.
However, there is little variation in the
thickness of blade and bladelet blanks.
The variability that does exist
does not have a consistent relationship with the width of blade auid
bladelet blanks.
Although blade and bladelet blank lengths vary, no clear trend is
97
Table 4.12:
C09N
CIO
CllN
C12N
C13N
C14
C15
Kolmogorov-Smimov two-sided probability test results for
blade and bladelet blank widths.
C08N
C09N
0.026
. . . 0.038
CIO
CllN
C12N
C13N
C14
0.001
0.093
0.558
0.454
0.065
98
apparent (Figure 4.4).
However, this result is certainly influenced by
the very small sample size in each level.
The median values for length
of blade and bladelet blanks are somewhat shorter in the upper levels
(C08N-C11N), than in the lower levels (C12N-C13N).
However all of the
notches of the box plot overlap indicating that the different levels
cannot be statistically separated into different groups.
between levels C13N and C14, is present.
narrower range than other levels.
One exception,
The lengths in C13N have a
The relative amount of post-
depositional disturbance and fracturing may be an influencing factor as
the above discussion on flake, flake fragment, and debris size classes
has suggested.
Debitage Summary
These analyses indicate that technological differences exist
between levels C08N-C10 and levels C11N-C13N.
The sample from levels
C14 and C15 are small and may produce questionable, unrepresentative
results.
Based on the data presented above, the lower levels (CllN-
C13N) have higher flake to blade ratios than those presented in earlier
studies, suggesting differences in the treatment of fragmentary debitage
can significantly affect flake to blade ratios.
High flake to blade
ratios in the lower levels may be influenced by greater postdepositional stresses and breakage rates in the lower levels.
Lower
levels (C11N-C13N) have higher percentages of short flakes, flake
fragments, and debris than upper levels (C08N-C10).
This result was
partially influenced by the narrow and thin blade and bladelet blanks in
the lower levels.
It was suggested that similar morphometric trends
(i.e., narrow and thin dimensions) would also be present in flakes and
flake fragments (see the discussion on the morphometries of flake tools
below).
However, debris to debitage ratios are lower in levels C09N-
C12N than in levels C08N and C13N-C15 indicating that relatively few
Picture 4.3:
Box olot of the thicknesses of unmodified blade and
bladelet blanks by level.
20
15
E
E
c
S 10
0
c
X
o
r.
1-
5
0
Level
100
Fiaure 4.4:
O
Box plot of the lengths of unmodified blade and bladelet
blanks by level.
40
-
m
Level
101
blanks were broken in the middle levels.
Debitage to core ratios are higher in levels C08N, C14 and CIS
than in levels C09N-C13N which suggests that these levels may have more
intensive core reduction than middle levels.
However based on the
frequency of cortex and shatter to debitage ratios, levels C13N-C15 have
more primary core reduction.
It was suggested that variability in these
ratios may reflect relative amounts of surface exposure since extreme
temperature fluctuations in desert environments may increase the
frequency of shatter in an assemblage.
Core reduction strategies and the morphometries of blade and
bladelet blanks varied between levels.
In general, the lower levels
(C11N-C15) had more single platform bladelet cores and combination
(flake and blade) cores.
Blade and bladelet widths were shown to be
statistically different between the upper (C08N-C10) and lower (CllNC15) levels.
However, thicknesses and lengths of blade and bladelet
blanks were not statistical different between levels.
Retouched Tools - Typology and Technology
All tools from levels C08N-C15 were placed into general
typological categories based on Tixier's (1963, 1974) Epipaleolithic
typology.
Some modifications, mostly compressing some of the tool
types, were made following Goring-Morris (1987).
At the most general
level, typologies may reflect fxmctional categories between major tool
classes including scrapers, burins, notches and denticulates, and
microliths.
Variations within categories are most likely to exhibit
stylistic influences as seen in the shape of microliths.
Although raw
material constraints may influence variation within a category (Close
1978:3), many stylistic attributes of Epipaleolithic tool assemblages
have been used to infer social groups within a chronological period
(e.g., Henry 1989b).
Throughout the Epipaleolithic, simple end
102
scrapers, angle burins, and notches and denticulates on blades and
bladelets are the most common forms in the major tool classes (Henry
1989b).
In this analysis, more emphasis is placed on general tool
categories than on stylistic differences within a single category.
Although variation within general tool classes is likely to exhibit
stylistic influences and responses to raw material constraints, the
sample of non-microlithic tools from levels C08N-C15 was relatively
small, i.e., less than 30 non-microlithic tools per level.
Generally,
this sample size is insufficient for statistically significcint
comparisons.
In previous analyses, excavation levels were typologically
analyzed then grouped by natural levels when preliminary statistical
analyses were conducted (Clark et al. 1987, 1988).
Some discemable
changes between natural levels 5 and 7 existed in the microlithic
component of the assemblage.
The most notable differences included the
proportions of geometric and non-geometric microliths, and differences
in the microburin index (Clark et al. 1987, 1988; Donaldson 1986).
However before lumping excavation levels into natural levels for my
analyses, I wanted to first determine if the sedimentological chainge
used to distinguish natural levels occurs at the same stratigraphic
position as typological and technological changes in the tool
assemblage, i.e., between levels CIO and CllN.
This is important since
combining levels might mask some typological and technological
variability in the assemblage.
As techno-typological changes are also
used to explain cultural variability, determining the synchrony of
techno-typological and paleoenvironmental changes should aid in the
interpretation of culture change during this period.
To facilitate
analysis and discussion, the typological and technological attributes of
major tool classes will be presented.
This will be followed by a
103
discussion of the typological and secondary' technological attributes of
microlithic tools.
Major Tool Classes - Typology
All levels have relatively low percentages of retouched tools
(3.24-12.22%) and tool fragments (0.81-5.91%) (Table 4.3).
No clear
trends are apparent in the proportion of tools between levels, although
the tool percentages are slightly higher, by about 1-4.5%, in levels
C08N-C11N and C14, than in levels C12N-C13N.
Similarly, no patterned
variability in the proportion of major tool classes by level is present
(Table 4.13).
Scrapers, carinated tools, burins, multi-purpose tools,
retouched and backed blades, truncated pieces, utilized blanks, and
"other" types, each comprise 4% or less of the entire tool assemblage.
Notches and denticulates, and retouched flakes comprise slightly higher
percentages of the total tool assemblage (10.71-12.46%) and are slightly
more abiondant in levels C08N and C09N than in lower levels.
Retouched
flakes comprise 5-10.5% more of the assemblage in levels C08N-C11N, C14
and C15, than levels C12N-C13N.
In general, notches and denticulates,
and retouched flakes comprise greater proportions of tools in upper
levels (C08N-C11N) than in lower levels (C12N-C14).
Tool fragments form the second largest tool category (Table 4.13) .
In Table 4.3, the number of tool fragments varied only 1-2% between
levels.
However, Table 4.3 presents the percentages of both the
debitage and tool components of the assemblage.
Since tools and tool
fragments comprise only a small proportion of the entire assemblage,
some tool variability was masked as the result of the larger sample used
in Table 4.3.
Except for level C09N, levels C11N-C14 had 4-16% more
tool fragments than levels C08N and CIO.
The relatively high
percentages of tool fragments in the lower levels resemble the trend
found in the debitage; however with a smaller sample size, this trend
104
Table 4.13:
Percentages of major tool classes by level.
C08N
Scrapers
C09N
CIO
CllN
C12N
C13N
C14
CIS
TOTAL
3 .57
3.37
1.82
6.19
2.70
1.16
2.84
10.00
3 03
Carinated
.00
1.12
.00
1.77
.00
.58
.71
.00
58
Burins
.89
.00
.91
.00
2.70
.58
.00
.00
70
Multiple
.89
tools
Ret./bckd.
.89
blades
Trunc­
.89
ations
Microliths 33 .93
.00
2.73
1.77
.00
.58
.71
.00
93
1.12
3 .64
2.65
.90
1.16
.71
10.00
1 63
3 .37
.91
4.42
3 .60
5.78
2.13
10.00
3 26
22.47
38.18
22.12
36.04
26.75
24.82
40.00
29 34
18 .75
23 .60
2.73
11.50
9.01
14.45
9.93
.00
12 46
9.82
12.36
13 .64
14.16
11.24
8.18
11.50
6.31
2.89
5.67
3 .57
.00
10.00
4.42
.00
4.05
5.67
.00
4 07
.00
.00
.91
.88
.00
3 .47
.00
.00
93
Ret./bckd. 23 .21
fragment
NUMBER
112
TOTAL %
100
21.35
16 .36
18.58
34.23
34.10
27.66
20.00
25 84
89
100
110
100
113
100
111
100
173
100
141
100
10
100
859
100.0
Notches/
dentic.
Retouched
flake
Retouched
piece
Utilized
blank
Other
3.57
4.50
3 .47
19.15
10.00
10 71
, 00
6 52
105
appears more pronounced in the tool component.
Microliths comprise the most abundant tool category in all levels
(ca. 22.12-38.18%) (Table 4.13).
For now, microlithic tools will be
treated as one of the major Cool classes.
Stylistic variation within
this tool class and its implications will be discussed in greater detail
below.
Major Tool Classes - Technology
Some technological variability exists in the selection of blanks
for some tools.
Blank selection criteria for major tool classes (i.e.,
scrapers, truncations, notches and denticulates, and utilized pieces)
were cinalyzed by determining variations in the proportion of blanks in
each tool class by level (Table 4.14).
Several of the major tool
classes were not analyzed because they either occur in low frequencies
(e.g., carinated tools, burins, and multiple tools) or their blank type
is explicit in the tool type (e.g., retouched and backed blades,
microliths, and retouched flakes).
In order to determine if specific
blanks were being preferentially selected for tool manufacture, a
Pearson chi-square test was applied to four tool categories (i.e.,
scrapers, trvincations, notches and denticulates, and utilized pieces).
Pearson chi-square is an appropriate test for these data because the
data are nominal and in a symmetrical, polytomy table.
However, one
major problem with this independence test is that it is strongly
affected by small sample sizes.
In order to decrease the number of
sparse cells in this analysis, level C15 which has a very small sample
was excluded.
In addition, the Pearson chi-square test was conducted on
the frequency of blank type by level (i.e., C08N-C14) and by group.
Levels were grouped into group 1 (natural level 5, levels C08N-C10) and
group 2 (upper portion natural level 7, levels C11N-C13N) in order to
decrease the number of sparse cells and to compare differences between
lOS
natural levels.
For truncations, and notches and denticulates, comparisons of
blank type by level and by group produced probability values greater
than 0.05 (Table 5.15).
If the stcuidard confidence level of 0.05 is
used, the null hypothesis that the rows and column totals in the table
are independent caimot be rejected.
Therefore, there is no statistical
difference in bleink selection either by level or by group for
truncations, and notches and denticulates.
For scrapers and utilized pieces, comparisons of blank type by
level and by group produced probability values less than 0.05.
Therefore, the null hypothesis can be rejected and blank selection for
scrapers and utilized pieces by level and by group is statistically
dependent.
Scrapers in levels C08N-C10 are more often manufactured on
blades, while scrapers in levels C11N-C15 are more likely to be
manufactured on bladelets and flakes (Tables 4.14 and 4.15).
The
scrapers on bladelets were small thumbnail scrapers and foiand only in
levels CllN and C12N.
This is interesting because simple end scrapers
on blades and bladelets are usually the most common scraper type in
Epipaleolithic assemblages (Henry I989b:85).
Utilized pieces are
manufactured more frequently on flakes in levels C08N-C11N than in
levels C12N-C13N.
Despite the grouping of levels, these significemce
tests may still be suspect as some sample sizes remained low.
In order
to determine if these statistical differences reflect a general trend in
retouched tools, a Pearson chi-square test was conducted on the blank
types of all retouched tools by level and by group (Tables 4.14 and
4.15) .
Blank selection for retouched tools by level is statistically
significant.
In contrast, blank selection for retouched tools by groups
is not significant indicating that blank selection does not vary between
upper (C08N-C10) and lower levels (C11N-C13N).
107
Table 4.14:
Row percentages of blank t^'pe fcr major tccl classes by
level. (Level CIS is excluded in order to decrease the
number of sparse cells.)
C08N
C09N
CIO
100.0
0
0
66.7
0
33 3
100.0
0
0
14.3
71.4
14.3
33 .3
33 .3
33 .4
0
0
0
0
100 .0 100.0
(Number = 22)
36.4
27.3
36.3
100.0
Truncations
0 33.3
Blade
Bladelet
0
0
Flake
! 100.0 66.7
100.0
0
0
60.0
0
40.0
90.0
66,7
25 .0
25 .0
0
0
50.0
10.0
33 .3
(Number = 28)
64.3
3.6
32.1
100.0
Notches/dents.
35.0
Blade
25.0
Bladelet
Flake
J 40.0
50.0
15.0
35.0
100.0
0
0
46.2
15.4
38.5
50.0
10.0
40.0
40.8
20.4
38.8
100.0
Utilized oiece
Blade
0
Bladelet
25.0
75.0
Flake
0
0
0
36 .4
27.3
36.4
20.0
40.0
40.0
0
0
0
All retouched tools
Blade
15 1 27 8
Bladelet
57 5 40 7
Flake
J 27 4 31 5
18 9
57 0
24 1
17 2
46 7
36 0
12.9
6 6 .1
21.0
Scraoers
Blade
Bladelet
Flake
TABLE 4.15:
C12N
C13N
39.1
26.1
34 .8
(Number =
C14
14.3
28 .6
57.1
103)
33 .3
50.0
66.7
0
50.0
0
(Number = 34)
TOTAL %
1 32.4
1 29.4
! 38 .2
100.0
25 5
9 .3
118.26
59 2
44.2 |53 .02
15 3
46.5
1 28.72
(Number = 527) 100.0
Pearson chi-square test for independence of blank type by
level and blank type by group* for four major tool classes
and all retouched tools.
Scrapers
blank type by level
blank type by group
Truncations
blank type by level
blank type by group
Notches and denticulates
blank type by level
blank type by group
Utilized
blank type by level
blank type by group
All retouched tools
blank type by level
blank type by group
*
CllN
VALUE
24.270
9.090
DF
12
2
PROB
0,019
0.028
13.806
2.122
12
2
0.313
0.346
11.064
0.087
12
2
0.523
0.993
11.787
27.559
8
2
0.000
0.000
36 . 76
0.965
14
2
0.001
0.617
Groups are group 1 (levels C08N-C10) and group 2 (levels C11N-C13N).
** More than one-fifth of fitted cells were sparse (frequency < 5).
Therefore, significance tests are suspect. Boldface type indicates
statistically significant values.
108
Since significance tests for the preferentially selection of
blade, bladelet, and flake blanks for tool manufacture by levels and by
groups are frequently suspect, perhaps the size and shape of blanks
selected for tool manufacture would exhibit significant variability.
Therefore, morphometric data were recorded on all tools in order to
determine if blanks with a certain range of measurements were being
preferentially selected for tool manufacture.
Nomtially, researchers
emphasize microlithic tools and bladelet blanks in Epipaleolithic
assemblages.
As previously mentioned, the definition of bladelets
places an arbitrary division between blades and bladelets such that
bladelets have lengths less than 4 cm and widths less than 1.2 cm
(Tixier 1963).
While this bladelet definition is now conventional, I
wanted to analyze the widths of both blades and bladelets together so
that changes in width distributions would more accurately reflect
sample-wide width distribution data.
Therefore, tools are grouped by
blank types into blade/bladelet and flake tool categories.
Furthermore,
since many major tool categories occur in low frequencies, analyses of
morphometric variations within tool classes would likely be heavily
influenced by small sample sizes.
As there is a trend throughout the Epipaleolithic towards
increasingly wide and short bladelet blanks and the resulting
microlithic tools (Henry 1989b:93; Neeley and Barton 1994:282),
variability in blank and microlithic tool morphology might reflect
regional trends and help to identify temporal variations within the
assemblage.
I was interested in determining whether this trend towards
wide and short blanks is found in both the blade/bladelet and flake
components of this assemblage.
However as the lithic assemblage at WHS
1065 is 98% incomplete or fragmentary, the utility of blank morphology
data, at least for length measurements and width to length ratios, is
109
limited.
Blade/bladelet blanks and tools are more likely to break
perpendicular to their length than parallel to it when siibjected to
compaction stresses and colluviation.
Unmodified blanks and retouched
tools having a "rounder" morphology would be more likely to break around
the edges rather than perpendicular to the length of the blank.
Since
tool widths and thicknesses are more likely to be preserved than tool
lengths, analyz .ng the widths and thicknesses of tools in these levels
enables a small aspect of overall blank morphology to be investigated.
Length measurements are analyzed only for flake tools because blade and
bladelet tools are not present in sufficient quantities to enable
statistically meaningful results.
Flake tool widths, thicknesses, ajad lengths were plotted in
separate notched box plots (Figures 4.5-4.7),
In the notched box plot
of flake tool widths (Figure 4.5), all levels have overlapping notches.
Therefore, they cannot be separated into to statistically different
populations.
However, flake tools in levels C08N-C11N have wider median
values than levels C12N-C14 and range between 19.8-25.9 mm.
The mean
for flake tool widths in levels C08N-C11N ranges between 20.46-24.70 mm.
The mean and median widths in levels C12N-C14 ranges between 17.4S-X9.19
mm and 14-16.5 mm, respectively.
Level CIS has a very small sample size
(two pieces) and will not be discussed further.
The plots of flake tool thicknesses and lengths exhibit similar
trends to those found in flake tool widths; although, specific values
vary.
Flake tool thicknesses have fairly confined ranges (Figure 4.6).
Median thickness values are slightly higher (by about 1 mm) in the upper
levels (C08N-C11N) than in the lower levels (C12N-C14).
Flake tool
lengths again seem to show a break between upper levels (C08N-C11N) and
lower levels (C12N-C14) (Figure 4.7).
Figure 4.5:
Box plot of flake tool widths by level.
DO
40 r
E 30
C
on
W
/ 1
y
\ I
U/'
I
'
1
'J
I \
/
i_l
ri
^ A
I \
1C r
r,\0,
Q\6
Levei
Ill
Fiqure 4.6:
Box plot of flake tool thicknesses by level.
Level
Figure 4.7:
Box plot of flake cool lengths by level.
J
i
1
1
!
:1
i
I
r
1
i
)
:
i
i
i
:
i
i ' 1
«
1
-'inti
Hi ; \1 i !H\ n r-! .' 'i un :1
U n M H ' : n ^ n \
-
•.J^
i
i
m
:
i
J
1
*
1
;
:
--
1^,
'j •
Level
O
'V./
113
Figures 4.8 and 4.9 are notched box plots of blade and bladelet
widths and thicknesses by levels.
In Figure 4.8, the confidence
intervals (notches) in levels C08N, C09N, CIO, and CllN overlap with
each other but do not overlap with levels C12N, C13N and C14.
Levels
C12N, C13N and C14 also have notches that overlap with each other.
This
suggests that tool blank selection (at least for widths) in levels C08NCllN is statistically different with a 95% confidence interval from
lower levels {C12N-C14).
quartile.
Level CIS has a notch which is below the lower
Although its notches overlap with all of the other levels,
this is probably due to the small, variable sample (8 pieces) from this
level,
Summary statistics for the blade and bladelet tool widths reflect
the same trend, showing the most pronounced break in median values
between levels CllN and C12N (Tcible 4.17) .
However, mean values exhibit
a more gradual change with the greatest difference between levels CIO
and CllN.
In order to determine the extent to which outliers and
stragglers in level CIO influenced these results, another set of summary
statistics was generated for bladelet tool widths (Table 4.18); blades
were excluded.
Again, a clear break can be seen between levels CllN and
C12N.
A series of Kolmogorov-Smimov two sample probability tests were
conducted on bladelet tool widths to determine if the overall shape of
the distribution is similar between adjacent levels (Table 4.19).
This
non-parametric test is especially appropriate here because the data are
metric, not normally distributed, and derived from two independent
samples.
The null hypothesis that the two samples are the same, can be
rejected for all of the tests except that between levels CllN and C12N.
Since the probability level between levels CllN and C12N (0.007) lies
below 0.01 the null hypothesis can be rejected.
Levels CllN and C12N
114
Figure 4.8;
Box plot of blade and bladelet tool widths by level.
50
40
I 30
S
I 20
n
10
A
^ R
u^n
V
0
Level
115
Ficmre 4.9:
Box plot of blade and bladelet tool thicknesses
by iWel. (Outliers greater than 25 mm have been removed.)
20 r
-
i;
0?
0 r
i
^
f
J
'-J
u
Level
W
n
116
Tcible 4.16;
Summary statistics for flake tool widths and thichnesses
by level.
Width:
N of Cases
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
C08N
24
6.2
42.9
36.7
24.7
26.0
C09N
25
9.3
42.1
32.8
21.2
20.8
CIO
23
5.5
46.0
40.5
20.9
21.2
CllN
38
4.8
36.9
32.1
20.5
19 .8
C12N
17
6.1
35.6
29.5
17.6
14.0
C13N
17
9.9
33 .5
23 .6
17.5
15 .2
Thickness:
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
C08N
2.3
16.7
14.4
6.9
6.6
C09N
2.3
14.4
12.1
5.9
5.0
CIO
1.6
20.7
19.1
5.7
5.6
CllN
1.5
49.0
47.5
7.4
6.1
C12N
2.2
14.0
11.8
5.7
4.5
C13N
2.1
12.2
10.1
5.4
4.3
Table 4.17:
Summary statistics for blade and bladelet tool widths
by level.
C14
44
9.0
39.1
30.1
19.2
16.5
C14
1.9
15.3
13 .4
5.3
4.8
CIS
2
19.7
24.2
4.5
22.0
22.0
C15
5.7
6.6
0.9
6.2
6.2
N OF CASES
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
C08N
83
4.1
45.4
41.3
12 .6
11.1
Table 4.18:
Summary statistics for bladelet tool widths by level.
N OF CASES
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
C08N
56
4.1
15.0
10.9
9.2
9.4
Table 4.19:
LEVEL
C09N
CIO
CllN
C12N
C13N
C14
C15
C09N
57
3.1
32.3
29.2
12.7
11.4
C09N
37
3.1
14 .6
11.5
8.8
9.1
CIO
78
3.9
42.5
38.6
12.9
11.4
CIO
51
3.9
13 .1
9.2
9.4
10.4
CllN
69
3.2
24.3
21.1
10.9
10.6
CllN
51
3.2
16.3
13.1
8.9
9.2
C12N
87
3.0
23.9
20.9
9.0
7.6
C12N
69
3.0
23 .5
20.5
7.2
6.7
C13N
145
2.3
24.5
22.2
8.5
6.6
C13N
114
2.3
12.6
10.3
6.2
5.4
C14
86
2.5
21.9
19.4
8.0
6.0
C14
68
2.5
13 .9
11.4
6.1
5.2
C15
8
4.4
17.9
13 . 5
9.4
7.9
C15
6
4.4
9.9
5.5
6.8
5.9
Kolmogorov-Smimov two sample probability test for bladelet
tool widths by level.
C08N
0.411
....
C09N
CIO
CllN
C12N
C13N
C14
0.273
0.373
0.007
0.073
0.626
0.459
117
probably do not come from the same population.
Blade and bladelet tool thicknesses exhibit a similar trend in
that upper levels {C08N-C11N) have thicker blade and bladelet tools than
lower levels (C12N-C14) (Figure 4.9).
But, the confidence intervals for
blade and bladelet tool thicknesses suggest that the null hypothesis,
i.e., that the samples are from the same population, cannot be rejected.
Complete blade and bladelet tools occur in such low frec[uencies that a
statistically significant sample was not availcible for determining
variations between width and length.
In sum, blade/bladelet and flake tools all exhibit the same
general trends in dimensions.
Levels C08N-C11N have wider and thicker
tools than levels C12N-C14 for both blade/bladelet smd flake tools.
This change is interesting because the interval of change does not
correspond with the stratigraphic break between natural levels 5 and 7.
This suggests that lithic tool manufacture changed before the erosional
event, as marked by natural level 6, an erosional cobble layer.
This
result differs from the result for unmodified blade/bladelet widths.
Unmodified blade/bladelet widths change significantly between levels CIO
and CllN suggesting that wide blade/bladelet tools were manufactured
before primary lithic manufacture produced significantly wide
blade/bladelet blanks.
Histograms of blade and bladelet tool widths were generated in
order to determine the distribution and possible modalities in the data
(Figures 4.10 and 4.11).
It was already known that wider tools were
manufactured in levels C08N-C11N than in levels C12N-C14 from the
notched box plots.
histograms?
So what new information is gleaned from the
Levels C08N-C11N have similar multi-modal distributions
with modes occurring at roughly 5-6 mm and 11-12 mm.
height of each mode varies between levels.
However, the
Generally, a higher
118
Figure 4.10:
Histograms of blade and bladelet tool widths for
levels C08N-C11N.
020
0.20
5 015
CD
•X
LU
Q.
5
f(T
o
Q.
§ Q05
LL
JE
M
20
2G
ClO
C08N
015 -I
tr
<
03
CC 0.10 -I
111
005
u
20
C09M
26
0
0
11
119
Figure 4.11:
Histograms of blade and bladelet tool widths for
levels C12N-C15.
015
020
0 ;c ^
X
!0
0 !o ^
in
x 0.10
0
i 005
Cl
n
iX
iX
a.
'1
- 005 ^
ir
I
10
U
20
U
20
CU
C12N
n
! I
; I
t~n
c:
iij
0 :0 •
02-J
y
I—
cr
u
n
!i li n
M i
j05 - i i
:r
i i
T
!i
f n:
r!!
nii
! i
LL
n j ^11
Ml
20
C13f-i
26
20
14
C15
26
120
proportion of 11-12 mm wide bladelet tools than 5-6 mm wide bladelet
tools are present in the upper levels (C08N, ClO and CllN).
Levels
C12N-C14 all have right skewed distributions with the greatest mode at
about 3 mm.
Although Kolmogorov-Smimov two sample probability tests
identified significant variation between the distribution of
blade/bladelet widths in levels C08N-C11N and C12N-C13N, the histograms
suggest an even more pronounced difference between central tendencies of
blade/bladelet widths than is indicated by summary statistics and
notched box plots.
Bladelets from levels C08N-C11N are commonly 11-12
mm wide while bladelets from levels C12N-C14 are commonly 3 mm wide.
The lower mean aind median width values for levels C12N-C14 reflect the
increased proportion of very narrow microliths in these assemblages,
rather than a decreased range in microlith widths.
The bimodal distribution of blade/bladelet tool widths is not due
to the original blank morphology.
Unlike blade and bladelet tools,
unmodified blade/bladelet widths are not bimodally distributed but have
slightly right skewed distributions in all levels, except level C15
which is discounted in this analysis because of its small sample size
(Figures 4.12 and 4.13).
This suggests that two major width classes of
blade and bladelet tools were preferentially manufactured in the lower
levels.
The very narrow bladelet tools are either backed on one edge,
backed on two edges, or backed and retouched.
To manufacture these very
narrow bladelet tools, multiple retouch episodes, i.e., heavy backing,
may have been necessary in order to narrow the width of the tool.
Since secondary lithic reduction (i.e., retouch) for retooling or
resharpening purposes may significantly alter the original blank
morphology, the number, distribution and orientation of dorsal flakes
scars on blade/bladelet and flake tools should help determine potential
differences in core reduction strategies and blank selection.
The
121
Fiaure 4.12:
Histograms of unmodified blade and bladelet blank widths
for levels C08N-C11N.
026
020
tc
lU 015 0.
z
0
I-
er 0.10a.
0
0
J 005
U
20
26
-f=^
~r~
32
U
C08N
20
26
G2
CIO
03
tr
<
CD
tr
lU
Q.
02 -
2
Q
§
cc
a.
U
20
C09N
28
—T"
32
a
U
20
C11N
i~i
26
JTL
32
122
Pi
4.13:
Hiscograms of unmodified blade and bladelet blank widths
for levels C12N-C15.
03 n
It
<
m
tr 02
z
0
(-
(t
ai
£
M=qU
20
28
14
32
C12N
20
26
32
20
26
32
CU
r
03
a.
<
iD
(T
UJ
(L
02 -
CE
2 0-1
0
rr
G.
14
20
C13N
26
32
S
14
C15
123
assumption is that blade and bladelet tools would have one cr r^.cre
dorsal flake scars parallel to the striking cixis of the original blank.
Flake tools would be more likely to have one or more dorsal flake scars
that were not parallel or a mix of parallel and non-parallel dorsal
flake scars.
If wide tools in levels C08N-C11N reflect differences in
blank selection, this would be an important technological difference
between the assemblages.
As expected, the majority of blade and bladelet tools had dorsal
flake scars parallel to the striking axis of the flake (Table 4.21).
Interestingly, the number of dorsal scars parallel to the striking axis
varied between levels.
Levels C08N-C11N had about twice as many blades
and bladelets with one dorsal flake scar parallel to the striking axis
than levels C12N-C15.
Although variable, tools with more than three
dorsal flake scars parallel to the striking axis are the most common
flake scar pattern in all levels.
The second most common category is
two flake scars parallel to the striking axis.
Except for level C13N,
which has many tools with more than three, parallel dorsal flake scars,
little variability exists between levels.
Not only did level C13N
produce higher proportions of single platform bladelet cores than other
levels (Table 4.11), it also has high frequencies of very narrow
microlithic tools.
These narrow microliths are too narrow, in most
instances, to have three or more dorsal flake scars.
The proportion of blade and bladelet tools with three or more
parallel, dorsal flake scars, and mixed parallel and non-parallel dorsal
flake scars is substantially higher in levels C08N and C09N than in
lower levels.
These values may be related to the relatively wide blanks
that are manufactured in the upper levels.
Although parallel flake
scars on dorsal surfaces are frequently present, their presence or
absence is not the determining criterion.
As sample sizes are
124
Table 4.20:
Summary statistics for blade and bladelet blank v/idths
by level.
C08N
174
3.6
36.0
32.4
10.9
10.0
N OF CASES
MINIMUM
MAXIMUM
RANGE
MEAN
MEDIAN
Table 4.21:
C09N
81
3.6
22.0
18.4
9.3
8.7
CIO
227
3.7
28 .6
24.9
10 .5
9.8
CllN
129
3.2
33 .1
29.9
9.5
8.3
C12N
149
2.8
29.0
26.2
8.1
7.3
C13N
279
2.7
25.2
22.5
8.0
7.4
C14
182
2.6
20.6
18.0
8.0
7.1
Distribution of dorsal flake scars on blade and bladelet
tools by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
CIS
N/A
•
00
.00
.00
.00
1.18
.00
.00
.00
Cortex
-
00
.00
.00
.00
1.18
.71
1.20
.00
7.27
5.13
1 parallel 6 .67
scar
2 parallel 26 .67
scars
3+parallel 52. 00
scars
3+ not
5 . 33
parallel
Multiple
9 .33
mixed
TOTAL %
100.0
NUMBER
75
Table 4.22:
C15
37
3.4
15.2
11.0
8.3
8.9
9.22 20.48
TOTAL N/%
.17
.51
.00
9.78
29.09 29.49
26.64
45.45 56.69
54.97
5.45
6.41
12.73
1.28
100.0 100.0
55
78
. 00
68
85
4 .26
.00
1.42
3 .61
.00
LOO.O
141
.00.0
.00.0
83
. 00
3.71
4.22
100.0
593
8
Distribution of dorsal flake scars on flake tools
by level.
C08N
C09N
CIO
CllN
N/A
.00
4.00
.00
.00
Cortex
. 00
4.00
.00
2.63
1 parallel 20.83
.00
8.00
scars
2 parallel 16.67 32.00 30.43
scars
3+parallel 12 . 50 20.00 34 .78
scars
. 00
1 not
.00
.00
parallel
Multiple
37.50 24.00 17.39
not paral.
Multiple
12 . 50
8.00 17.39
mixed
LOO . 0 100. 0 100.0
TOTAL %
NUMBER
24
25
23
7.89
C12N
C13N
C14
00
5.88
.00
.00
1.05
00
6 .82
.00
3.16
5.88 41.18 15 .91
.00
13 .16
5.88 15.90 50.00
20.00
•
5.88
•
21.05
11.76
26.32
17.65 23.53 27.27
.00
42.11
.00
100.0
38
C15
TOTAL
.00
23.68
4.55
.00
1.05
47.06 23.53 18.18
.00
28.95
00 11.36 50.00
8.95
00
11.76
00
•
100.0 100.0
17
17
100.0 100.0
44
2
100.
190
125
relatively small, these results should not be considered conclusive.
Flake tools comprise a greater proportion of the tool assemblage
in the upper levels (C08N-C11N) than in the lower levels {C12N and C13N)
(Table 4.14). As expected, the majority of flake tools had multiple
dorsal flake scars not parallel to the direction of the striking axis.
Many flake tools in level C13N had one dorsal flake scar parallel to the
striking axis of the blank.
Flake tools with two parallel, dorsal flake
scars comprised a lower proportion of the flake tool assemblage in lower
levels (C12N-C14) than in upper levels (C08N-C11N).
Multiple (more than
three) parallel dorsal flake scars are variable in the upper and lower
levels exhibiting no clear trends between levels.
Level C14 had the
only occurrence of a single dorsal flake scar not parallel to the
striking axis of the flake.
This type of dorsal surface is most common
in levels C11N-C12N and C08N.
The occurrence of mixed dorsal flake scar
patterns, i.e., parallel and non-parallel flake scars orientated with
respect to the direction of the blow, is variable.
When mixed dorsal
flake scar patterns occur, they comprise similar proportions of the
dorsal flake scar pattern on flake tools.
However, mixed flake scar
patterns are absent in levels CllN and C13N, perhaps reflecting
differences in the relative amoiant of core reduction in these levels.
In sum, the upper levels have higher percentages of blades and
bladelets with non-parallel or mixed (parallel and non-parallel) dorsal
flake scars than lower levels.
This suggests that core reduction
strategies varied somewhat between the upper and lower levels.
In the
upper levels, blade, bladelet, and flake tools are more likely to have
multiple non-parallel or mixed dorsal flake scars which may reflect the
increased emphasis on manufacturing wide blanks.
Blade or bladelet
technology in the upper levels was still important but perhaps more
variability existed in lithic reduction strategies.
Cores were more
126
likely to be used for both flake and blade mamufacture in the upper
levels.
The lower levels were more likely to have prepared cores that
enabled multiple blades and bladelets to be removed from a single
platform, as seen in the high frequencies of single platform cores and
tools with multiple parallel, dorsal flake scars.
Next, I wanted to determine if this variability in blade/bladelet
and flake tool widths, thicknesses and lengths was reflected in other
lithic manufacturing attributes such as striking platforms and retouch
types.
The results were disappointing.
Both the type of striking
platform and type of retouch do not seem to vary significantly between
levels (Tables 4.23 and 4.24).
The striking platform on tools was
recorded in order to identify potential differences in blank
manufacture.
As might be expected, the majority of tools did not have
the striking platforms preserved as the proximal end of the blank was
often removed by retouch or broken.
Large broken tools were not
selected out with the tool fragments because they were sufficiently
large and complete to be categorized.
Some variability between levels is apparent (Table 4.23).
There
seems to be a very slight increase in the proportion of plain platforms
in levels C08N-C11N.
In contrast, punctiform platforms are more common
in lower levels, especially level C13N, but infrequent in levels CIO,
CllN, and C14.
This reflects an increase in the number of bladelets
removed with the punch technique and an emphasis on narrow blank
manufacture mentioned earlier.
The manufacture of narrow bladelets may
require more controlled blank removal than the removal of wide blanks.
Crushed platforms were also somewhat more abxandant in levels C09N-C10
than in other levels.
assemblages
size.
There were few dihedral platforms in these
Again, level CIS is discounted because of its small sample
127
Table 4.23:
Percentages of tool platform types by level.
C08N
C09N
CIO
CllN
C12N
C13N
Absent
66 .28
67 .14
73 .91
67 39
73 97
75 44
Cortex
1.16
.00
2.17
1 09
4 11
88
12 .79
10.00
15 .22
17 39
6 85
7 02
Faceted
1.16
.00
1.09
1 09
1 37
88
98
.00
.94
Pxinctiform
8.14
10.00
1.09
5 43
8 22
11.40
3 92
.00
6 .75
Crushed
2.33
10 .00
6.52
3 26
2 74
2 63
2 94
.00
4.08
Broken
2 .33
2 .86
.00
3 25
1 37
1 75
5 88
.00
2 .51
Dihedral
5.81
.00
.00
1 09
00
00
00
.00
94
00
1 37
00
00
00
100.0
92
100.0
73
100.0
114
Plain
Other
•
TOTAL %
NUMBER
100.0
86
Table 4.24:
N/A
Very fine
or fine
Semiabrupt
Abrupt or
backing
Mixed ret.
types
Side
scraper
Alternate
Marginal
or util.
Burin blow
TOTAL %
NUMBER
00
•
00
100.0
70
-
00
100.0
92
C14
C15
74 51 75 00
98
TOTAL
71.59
00
1.41
10 78 25 .00
11.62
100.0 100.0
102
8
•
16
100.0
537
Percentages of tool retouch types by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
C15
TOTAL
00
00
1.09
2 17
00
88
.98
.00
.78
1 16
1 43
1.09
2 17
1 37
00
1.96
.00
1.25
17 44
22 86
29.35
25 00
10 96
28 95
30.39 25.00
24.33
43 02
31 43
38.04
29 35
50 68
37 72
29.41
.00
36.26
18 60
31 43
14.13
23 91
21 92
18 42
18 .63 62.50
21.04
12 79
4 29
10.87
7 61
9 59
7 02
7.84 12.50
8.63
00
00
.00
00
00
00
1.96
.00
.31
6 98
8 57
4.35
9 78
1 37
7 02
8 .82
.00
6.75
00
00
1. 09
00
4 11
00
.00
.00
.63
100.0
86
100.0
70
100.0
92
100.0
92
100.0
73
100.0
114
100.0 100.0
102
8
100.0
537
128
There was little variability in retouch types between levels
(Table 4.24).
In general, abrupt retouch occurs frequently (35-55%).
Semi-abrupt retouch (12-25%), and mixed retouch, a combination of semiabrupt, abrupt, and fine retouch (13-35%), occur in similar proportions.
Here, retouch type is more likely to vary between major tool classes
than within them.
For example, microliths frequently have short, semi-
abrupt or abrupt retouch along a lateral edge or tnmcation whereas side
scrapers have long, steep retouch.
Except for microliths, major tool
classes did not occur in sufficient quantities to enable comparisons of
retouch type within one tool class.
Therefore, any variability in
frequencies and percents of retouch type relate more to the frequency of
specific tool categories than technological variability.
Microliths
Microliths comprise the highest proportion of tools (20.22-35.14%)
in all levels.
In previous analyses, some discemable changes between
natural levels 5 and 7 were noted in the proportions of geometric and
non-geometric microliths, and differences in the microburin index.
To
assess the frequency of geometric microliths, two categories of
geometric microliths were identified (Clark et al. 1987, 1988; Donaldson
1986; Neeley ec al. 1995).
Geometric A microliths represent geometric
forms typically associated with Geometric Kebaran assemblages (Henry
1989b).
These include "true" geometric forms such as lunates, trapezes,
rectangles, and triangles.
Geometric B microliths include any microlith
that is truncated (oblique or straight) and backed, and are commonly
associated with non-geometric Kebaran assemblages (Table 4.25).
For comparative purposes, microliths are grouped into general
categories, similar to those used in earlier studies, based on the
presence or absence of backing and the angle of truncation (i.e.,
straight or oblique).
These categories are used because they (1)
129
facilitate comparisons with in-field analyses and between levels by
increasing the number within each category; and (2) reflect general
typological and secondary technological attributes.
These techno-
typological characteristics are better indicators of variation between
levels, as fine typological divisions are more apt to be based on subtle
variations in tool morphology, even a specialist may choose different
typological categories if asked to reanalyze a microlithic assemblage.
In addition, some morphological differences seem to reflect varying
degrees of utilization, a point which will be discussed in greater
detail below.
Backed microliths are the most abundant microliths in all levels
(Table 4.25).
But, they comprise roughly 5-16% more of the microlith
assemblages in lower levels {C12N-C14) than in upper levels (C08N-C11N).
Some variability within specific types of backed microliths exists
between levels (Table 4.26).
Backed, backed and retouched, backed and
notched, and backed and curved microliths exhibit no consistent trends
between levels (Table 4.26).
However, pointed varieties are
consistently more abundant in levels C12N-C14 than in upper levels.
Level CIO has several pointed microliths with backing and/or semi-abrupt
retouch along both lateral edges, and straight, backed pointed
microliths.
Morphological variations of backed and retouched
microliths, that do not clearly "fit" into any category, comprise 2-12%
of the microlith assemblage.
They are most abundeuit in levels CllN,
C12N and C14.
Few non-backed trimcated microliths are present; they comprise
less than 6% of the microliths in each level.
Backed and truncated
microliths comprise slightly higher proportions of the assemblage (210%) than unbacked truncated microliths.
In general, backed and
truncated microliths are more abxindant in upper levels (C08N, C09N,
130
Table 4.25:
Percentages of microliths by level.
C08N
C09N
CIO
CllN
C12N
C13N
C14
C15
TOTAL
Backed/Curved 39 .5
Bladelets
Truncated
2 .6
Bladelets
Oblq. Trune. 13 .1
Bladelets
Backed Trune. 7 .9
Bladelets
Backed Oblq. 28.9
Trune. Bldt
Bi-trunc.
0 .0
Backed Bldt
Atypical
0.0
Trap/Reet.
Atypical
7 .9
Triangle
Atypical
0.0
Lunate
Other
0.0
50.0
38.1
48.0
55.0
66.7
65.7
50.0
52.38
5.0
2.4
16.0
5.0
6.2
2.9
0.0
5.16
5.0
26.2
8.0
7.5
4.1
8.6
25.0
11.11
10.0
0.0
8.0
0.0
2.1
8.6
0.0
4.37
15.0
28.6
16.0
30.0
16.7
11.4
0.0
21.43
5.0
0.0
4.0
0.0
0.00
2.8
0.0
1.19
0.0
0.0
0.0
0.0
2.1
0.0
0.0
0.40
5.0
0.0
0.0
0.0
0.0
0.0
0.0
1.19
5.0
0.0
0.0
2.5
2.1
0.0
25.0
1.59
0.0
4.7
0.0
0.0
0.0
0.0
0.0
0.79
100
23
100
36
100
25
100
40
100
48
100
31
100
4
100
242
TOTAL
N
% Geometries A
% Geometries B
100
35
C08N
7.9
44.7
C09N
15.0
40.0
CIO
4.7
33.3
CllN
4.0
28.0
C12N
2.5
32.5
C13N
4.2
23.0
C14
2.8
22.8
CIS
25.0
25.0
Geometries A = Backed and bi-truneated bladelets, trapeze/rectangles,
triangles and lunates.
Geometries B = Geometries A plus backed and truncated bladelets, and
backed and obliquely truncated bladelets.
131
CllN) than in lower levels.
Unbacked, obliquely truncated microliths
are most abundant in levels C08N and CIO.
Backed, obliquely truncated
microliths are the second most abundcint microlith category in most
levels and have high percentages in levels C08N, CIO, and C12N.
Although wide microliths were already known to be associated with
increased proportions of geometric microliths from this study, both
obliquely truncated, backed microliths and atypical lunates tend to be
manufactured on wide blanks and comprise a greater proportion of tools
in upper levels (C08N-C11N) than in lower levels (C12N-C14).
"True" geometric forms occur in very low frequencies (less than
8%).
As previously indicated, many blade and bladelet tools are
fragments.
While small microlithic fragments have been removed from the
sample, larger incomplete tools were included in this analysis.
Since
recognition of geometric forms almost always recjuires a complete
specimen, backed and truncated fragments may represent one end of a
geometric microlith.
Therefore, the proportion of truncated microliths
may be a useful proxy measure for increased proportions of geometries in
very fragmented assemblages.
Analyses of blade/bladelet tool widths indicated a bimodal
distribution of widths in both the upper (C08N-C11N) and lower (C12NC14) levels.
The separation between modes occurs at about 8 mm (Figures
4.10 and 4.11).
The majority of backed microliths (84-100%) in all
levels are less than 8 tran wide.
Only two types of backed microliths
(curved and backed, and backed and retouched varieties) have equal
proportions of microlith widths greater than and less than 8 mm.
Interestingly, 75% or more of curved and backed, and backed and
retouched varieties of microliths in levels C08N-C11N are wider than 8
mm.
Although some truncated, obliquely truncated, backed and truncated,
and backed and obliquely truncated microliths have widths less than 8
132
Table 4.26:
Frequencies of specific microlithic t^'pes by level.
C08N C09N CIO
Backed
Backed
|4
0
7
Backed & ret. [2
5
l
Backed & notch.\
2
0
0
Curved & backed{
4
3
1
Bck.&ret. variaj
2
1
1
Pointed:
2 edges backed [0
0
3
Straight backed',
0
1
3
Microperforator I
1
0
0
Trxincated (no backing)
Truncated
|l
l
i
Oblique truncation (no backing)
Oblq. trunc.
l4
1
5
Bitriinc .oblq. |l
0
3
Oblq.tr\mc.&ret|
0
0
3
Backed & truncated
Bck.& trunc. |3
2
0
Backed & obliquely truncated
Bck.&oblq. tr\in|
6
2
5
w/ notch
]5
1
7
Bi-truncated & backed
Bi-trunc,& bckj 0
1
0
Geometries
Proto-trapeze |0
0
0
A t y p . l u n a t e j
O
1
0
Atyp. trianglej
3
1
0
Other
|0
0
2
TOTAL NUMBER
38
20
42
CllN
C12N
C13N
C14
0
4
0
3
3
1
7
0
1
4
9
6
2
1
2
l
6
0
5
4
0l22
l|32
0|4
OjlS
0|l7
0
1
1
5
4
0
3
9
0
5
2
0
l|l7
0|20
0|2
4
2
3
1
0|l3
2
0
0
3
0
0
2
0
0
1
0
2
0jl8
1|6
2
0
1
3
Ojll
4
0
12
0
8
0
3
1
0[40
0|l4
l
0
0
1
0
0
0
0
0
1
0
0
1
1
0
0
0
0
0
0
25
40
48
35
C15
0
TOTAL
|
4
0l3
0|l
1 | 4
0|4
0j2
4
252
133
mm, 50% or more of these microliths in all levels are wider than 8 mm.
Therefore, trxincated and obliquely truncated microliths, including both
backed and unbacked forms, comprise the majority of the wide (11-12 mm)
microliths in these assemblages.
Usually "true" geometries are
variations of backed and truncated, and backed and obliquely truncated
microliths; so, the predominance of unbacked truncated and unbacked
obliquely truncated microliths with wide forms may also indicate an
intermediary form between what is considered a non-geometric and a
geometric microlith.
In general, microliths are considered interchangeable tool
components (e.g., Henry 1989b; Henry and Garrard 1988).
An important
question is whether variability in microlithic widths between the
"spiky", very narrow microliths (narrower than 8 mm and commonly 3 mm
wide) and wide microliths (generally wider than 8 mm) in the Step C
lithic assemblages indicate fionctional differences.
Such functional
differentiation would be useful in determining the type of activities
conducted at this site.
In addition, it would help explain the modality
in the distribution of microlithic tool widths.
Some general
differences in utilization and/or retouch on the unbacked edge exist
between wide and narrow forms in levels C08N-C15.
Many wide forms
(i.e., backed and obliquely truncated, and obliquely truncated
microliths) are either heavily utilized and/or notched on the unbacked
edge.
Although many narrow forms (e.g., backed microliths) have retouch
on the unbacked edge, very few narrow forms are notched.
In addition,
the notch on narrow forms is not as deep as the notch on wide forms.
Furthermore, many backed microliths (approximately 30%) are
pointed varieties.
Narrow pointed varieties, especially backed and
retouched microliths, and microliths backed on two edges, may have been
more easily hafted in a manner similar to that suggested by Henry and
134
Garrard (1988:12) for mounting points at Tor Hatnar.
The notch
associated with wide truncated and obliquely trimcated microliths may
indicate that these microliths were not hafted at the tip of the haft
nor aligned longitudinally along the haft's length.
Rather, they may
have been hand-held or hafted singly, along the length of a haft.
The
wide tools with backing on one lateral edge would enable these
microliths to be more easily held.
It is unlikely that these microliths
were hafted in an alignment because the placement of the notch in the
center of the unbacked edge suggests somewhat more controlled use of the
microlith.
I would expect that microliths hafted longitudinally might
produce irregular evidence of utilization on the unbacked edge.
Microburin Indices
The microburin technique is a method of trvincating or segmenting a
bladelet whereby a bladelet is notched and snapped.
to control the position of the snap break.
The notching serves
Together, the notch and snap
produce a characteristic scar on the microburin.
Bladelets segmented
with this technique are usually an intermediate stage in microlith
production.
Later, the scar on the segmented bladelet may be removed as
further modification (usually trioncation) of the bladelet segment
occurs.
Microburins, segmented bladelets that retain the characteristic
snap scar, are the waste or final end products of this technique (Henry
1974; Tixier 1963, 1974).
Various methods have been proposed to calculate the frequency of
the microburin technique (mbt) (Table 4.27).
One index (Imbt) reflects
the use of the technique in the total tool assemblage (Bar-Yosef 1970) .
The reduced Imbt or rimbt (Henry 1974) and the adjusted Imbt or adjimbt
(Marks and Larson 1977) reflect the use of the technique among all
microliths and truncations (pieces where the technique was most likely
to have been used).
The adjImbt reflects the use of the technique only
135
on those items that probably were produced by the technique.
The number of microburins and microburin indices in each level
show a steady increase between levels C08N and C13N (Table 4.28 and
4.29).
Clear "jumps" occur in all microburin indices between levels
C08N and C09N, CIO and CllN, C12N and C13N, and C14 and CIS.
The rimbt
and adj Imbt indicate that the microburin technique was used more in
levels C11N-C14 than in levels C08N-C10.
natural levels 5 and 7.
This division corresponds with
These microburin indices are higher than those
presented in earlier studies (Clark et al. 1987, 1988; Donaldson 1986).
This difference most likely reflects the greater number of microburins
identified in this analysis than in in-field analyses.
Only 60
microburins were recorded during in-field analyses of the lithic
assemblage from all of natural level 7 (Clark et al. 1987, 1988;
Donaldson 1986).
In this study, 8 5 microburins were recorded for only a
portion of natural level 7, i.e., the northern half of levels C11-C13,
and all of levels C14 and C15.
The use of these indices "as cultural markers assumes that
manufacturing debris (microburins) and tools (microliths) were regularly
discarded together in consistent frequencies" (Neeley and Barton
1994:278).
The high microburin indices in the lower levels are
interesting because this technique is not commonly associated with nongeometric assemblages in the early Epipaleolithic.
This suggests that
this technique is more common in the early Epipaleolithic than
previously indicated.
The association of low microburin indices and
geometric microliths in the upper levels is also found in other
Epipaleolithic assemblages (Henry 1989b).
One plausible explanation is
that microburin scars were removed when segmented bladelets were
truncated during the manufacture of geometric microliths.
In this
situation, microburin scars would no longer be identifiable.
136
Table 4.27:
Formulas for microburin indices.
Imbt =
# of microburins
# of microburins + tools
rimbt =
# of microburins
# of microburins + microliths + truncations
adjImbt =
Table 4.28:
Percentages of microburins.
C08N
True
iCnikowski
CIO
CllN
C12N
C13N
C14
C15
Total
66.7
75.0
90.9
83 .3
94.2
90.0
100.0
89.4
0.0
0.0
9.1
8.3
2.9
10.0
0.0
5.9
0.0
33.3
25.0
0.0
8.3
2.9
0.0
0.0
4.7
100
1
100
3
100
4
100.0
Table 4.29:
Imbt
rImbt
adjImbt
C09N
0.0
Piquant
tiedre
Total %
Number
# of microburins
# of microburins + truncations
100
11
100
12
100
34
100
19
100
1
Microburin indices by level.
C08N
1.2
1.6
4.1
C09N
5.0
9.1
23.1
CIO
4.8
5.5
13.3
CllN
13.9
22.4
45.8
C12N
18.2
17.1
40.0
C13N
31.2
34.7
68.0
C14
20.2
28.8
61.3
C15
12.5
14.3
33.3
100.0
85
137
Some have suggested that low microburin indices in Geometric
Kebaran assemblages are one response to a need to use lithic material
more conservatively (Neeley and Barton 1994:280).
This argument is
based on the difference between the mecin microlith length and the mean
blank length in the Geometric Kebaran, Mushabian, and Natufian
industries.
It is suggested that the difference is more pronounced in
the Geometric Kebaran because two microliths were being generated from
each bladelet while in the Mushabian one microlith was manufactured from
each bladelet (Neeley and Barton 1994).
In the Mushabian, the unused
portion of the bladelet was discarded with the characteristic microburin
scar.
In the Geometric Kebaran, both bladelet segments were
manufactured into microliths (Neeley and Barton 1994) .
If this argument is true, I would expect the mean microlith length
in the Geometric Kebaran to be 50% or less of the mean bladelet length.
However, Geometric Kebaran microliths account for 61% of the average
length of unretouched bladelets (Neeley and Barton 1994:280).
In the
Mushabian, backed bladelets account for 80% of the unmodified blank
length (Neeley and Barton 1994:280).
Neeley and Barton's argument seems
to hold for the Mushabian which has about a 1:1 ratio of microburins to
microliths (this would give an rlmbt of about 50).
However, they seem
to argue from negative evidence for the Geometric Kebaran, i.e., fewer
microburins in those assemblages.
They do not offer an explanation why
lithic sources may have been used more conservatively.
I think the
argument could be turned around to suggest just the opposite conclusion,
i.e., lithic material was used more wastefully.
The microburin
technique may not have been used because it was not as effective for
segmenting the wide bladelet blanks used in the manufacture of geometric
microliths.
Before either interpretation is accepted, additional
studies linking changes in land use patterns, subsistence strategies.
138
and conservative or wasteful use of lithic rav.' ^.atsrial CIATQ nscssssiTy'.
139
CHAPTER 5:
DISCUSSION AND CONCLUSIONS
In chis study, techno-typological variability in the debitage and
tool components of the lithic assemblage from Tor al-Tareeq (WHS 1065),
Step C, arbitrary levels C08N-C15 was analyzed.
This variability can be
used to determine differences in the chaine operatoire, or operational
sequence of lithic manufacture in these levels.
Some of the variability
identified in these levels is cyclic and appears to represent functional
changes in the activities conducted, or at least deposited, in these
levels.
Most of the cyclic changes relate in some way to location of
core reduction activities and the intensiveness of core reduction in
each level.
Since much of the assemblage is fragmentary, additional
analyses were conducted in order to identify relative amounts of
depositional and post-depositional breakage and to determine how these
breakage patterns might influence the completeness of the assemblage.
Variability in the operational sequence, specifically core reduction,
and tool manufacture, use and discard, relates to temporal changes in
the lithic assemblage.
Three areas of variability in the lithic
assemblage (site formation processes, site function, and operational
sequence) will be discussed below.
Then, comparisons of the operational
sequences identified in this study will be compared to others in the
region.
Site formation processes
In this study, the frequency of completeness categories, and
debitage and debris size categories, and fragment to debitage ratios
suggest the amount and size of fragmentary debitage and debris in this
assemblage is influenced by the artifact's original size and morphology.
The lowest levels (C12N-C15) have at least 10-20% more very small debris
(lengths less than l cm) and higher percentages of small flakes than the
upper levels (CG8N-C11N).
Therefore, the narrow debitage and retouched
140
cools (particularly microliths) in the lower levels (C12w-C15) may have
been more susceptible to post-depositional stress than wide artifacts in
upper levels (C08N-C11N).
This may account for the relatively high
breakage rates (fragment to debitage ratios) and abundant small debris
in the lowest levels (C13N-C15) of this site.
The abundance of small debitage in the lower levels may also be a
result of downslope movement from upper areas.
Previous analyses of
surface and subsurface deposits at WHS 1065 were conducted in order to
study Che site boundaries, artifact class frequencies, site function,
surface-subsurface congruity, and site formation and disturbance
processes (Coinman et al. 1989).
Structural and compositional
characteristics of a 95% sample of Che site's surface assemblage
(collected in I x 1 m units) were compared to excavated, subsurface
deposics from Steps A-I.
Results from this study suggest chat much of
the spatial patterning at the site can be attributed to downslope
movement of materials (Coinman et al. 1989).
Dense surface deposits are
generally located downslope from dense subsurface deposics.
However, it
is unlikely that this would completely account for the relatively high
percentages of very small debris and flakes in levels C11N-C15.
Levels
C11N-C15 are considered to be in situ deposics, an interpretation that
is strongly supported by the presence of two hearth features in level
C13.
The abundance of small artifacts recovered from lower levels
(C13N-C15) might also be related to the size effect which tends to
transport large artifacts up in a deposit (Baker and Schiffer 1975;
Baker 1978).
Trampling, in combination with the permeability and
texture of the sediments, can also sort surface artifacts into size
classes (Coinman ec al. 1989; Gifford 1978:81-83).
Clark (et al. 1988:261) suggests that lower frequencies of shatter
relative to debitage on the surface and in Step C, natural level 4
141
indicates that these surfaces may hgve been exposed fcr Icngsr periods
of time.
The underlying deposits may have been exposed to chemical and
physical weathering, and other forms of erosion and disturbance.
Since
the shatter to debitage ratio was greater than one for all subsurface
deposits except Step C, natural level 4, Clark (et al. 1988:261)
suggests that this level may have been exposed to the surface for longer
periods of time.
However, Clark (et al. 1988:261) includes debris
(medial and distal flake fragments) in his shatter category.
With a
separation of these categories, it appears that different processes may
be patterning the proportion of debris and shatter in these assemblages.
Levels C08N, C13N, C14, and C15 have higher fragment to debitage ratios
suggesting more breakage in these levels than in other levels.
The
higher percentages of shatter in the lower levels than in the upper
levels may be related either to chemical and physical weathering (Clark
et al. 1987, 1988) or to primary core reduction activities (Sullivan and
Rozen 1985) .
Therefore, although the size effect, collection
strategies, and weathering may influence the size patterning of
artifacts in a deposit (Clark et al. 1987, 1988; Coinman et al. 1989)
artifact size also affects the susceptibility of artifacts to breakage
causing narrow blanks to break into smaller fragments than wide blanks.
Intra-site Functional Variability
Some of the variability identified in levels C08N-C15 represents
functional changes in the activities conducted, or at least deposited,
at the site.
Most of these changes are cyclic, cross-cutting temporal
techno-typological variability at the site, and relate in some way to
the staging and intensiveness of core reduction.
Functional
interpretations of artifact assemblages are based on evidence of primarycore reduction and tool manufacture, and general tool type frequencies
such as the frequency of large, non-microlithic tools and microlithic
142
tools.
Evidence for primary core reduction activities include: core to
debitage, tool to debitage, and shatter to debitage ratios, cortex
frequencies, and to a lesser extent core weights.
Many initial core reduction activities probably occurred off-site,
as evidenced by the relatively low percentages (usually less than 25%)
of debitage and cores with some cortex.
Level C13N differs from other
levels in that approximately 10-15% more of its debitage has some amount
of cortex.
Debitage with completely cortical dorsal surfaces was very-
rare, comprising less than two percent of the entire chipped stone
assemblage from this site (Clark et al. 1987, 1988).
The frequency of
shatter is higher in lower levels (C12N-C15) than in upper levels (C08NCllN) suggesting either more primary core reduction (Sullivan and Rozen
1985) or more intensive weathering of the lower than of the upper
deposits.
However, an increase in the amount of primary core reduction
is not supported by the frequency of cortex on debitage and cores in
these levels.
In this regard, this study and previous studies of this
assemblage (i.e., Clark et al. 1987, 1988; Donaldson 1986; Neeley ec al.
1995) produced similar results.
Interpretations of the intensiveness of core reduction are based
on debitage to core ratios, the relative frequencies of cores and core
trimming elements, and to a lesser extent core weights.
Earlier
analyses already remarked that the frequency of cores and core trimming
elements is very low throughout all excavated portions of WHS 1065
(Clark et al. 1987, 1988).
In levels C08N-C15, core and core trimming
elements each comprise less than two percent of the assemblage.
In
previous analyses, debitage to core ratios indicate slightly more
emphasis on knapping in Step C, natural levels 5 and 5a than in natural
level 7 (Clark et al. 1987, 1988).
However, this analysis indicates
higher debitage to core ratios in levels C08N and C14 suggesting that
143
Che intensity of core reduction activities v.'as probably not constant at
this portion of the site throughout the deposition of natural levels 5
and 7.
More intensive core reduction may have occurred in level C08N
than in levels C09N-C10, and in level C14 than in levels C11N-C13N.
Although the median core weights are higher in the upper levels (C08NClO) than in the lower levels (C11N-C13N), a notched box plot of core
weights indicates that this difference is not statistically significant.
There is little evidence for change in the frequency of tool
manufacture and the frequency of major tool types in this study.
Generally, retouched tools comprise less than 10% of each assemblage.
The proportion of non-microlithic and microlithic tools varies little
between levels.
In the non-microlithic tool category, retouched flakes
and pieces are slightly more abundant in levels C09N-C11N and C14 than
in other levels.
While, notches and denticulates are more abundant in
levels C08N-C09N than in lower levels.
Still, the proportions of these
tool classes are not markedly different between levels.
activity differences cannot be inferred from these data.
Therefore,
Although the
combined results from these analyses do not differ significantly from
earlier studies (i.e., Clark et al. 1987, 1988; Donaldson 1986), this
study does show more variability in staging and intensity of core
reduction in natural levels 5 and 7 than reported in earlier studies
(i.e., Donaldson 1986; Clark et al. 1987, 1988).
Intra-site Variability in Operational Sequences
The operational sequence for the manufacture of chipped stone
begins with the acquisition of raw material.
The majority of artifacts
deposited at Tor al-Tareeq are manufactured from fine-grained grayish
brown chert.
A very coarse grained, fossiliferous limestone is also
present, but is rarely used for tool manufacture.
Although
compositional analyses have not been conducted on the lithic material.
144
it seems that similar lithic sources were used during all occupations of
the site.
Fossiliferous limestone outcrops in several locations in the
immediate vicinity of Tor al-Tareeq and is almost certainly the source
area for this material.
Although the ubiquity of chert throughout the
Levant has been questioned, fine-grained chert is found within a
kilometer radius of the site.
material was scarce.
Therefore, it is unlikely that lithic raw
Thus, from the lithic assemblages analyzed here,
similar lithic procurement strategies were employed throughout the
occupation of the site.
Core and platform types, distribution of dorsal flake scars,
debitage morphometric data, and flake to blade ratios are used to
determine variability in core reduction strategies.
Some technological
changes in core reduction occur between levels CIO and CllN, following
the natural levels; others occur between levels CllN and C12N.
The
upper levels (C08N-C10) have higher portions of bi-directional blade,
bi-directional flake and multiple platform flake cores than lower levels
(C12N-C15).
Dorsal flake scar patterns on debitage from upper levels
(C08N-C11N) show higher frequencies of non-parallel and mixed (parallel
and non-parallel) dorsal flake scars on blades and bladelets, and higher
proportions of flakes with two or more parallel flake scars than lower
levels.
This may reflect greater emphasis on manufacturing wide blanks
and higher percentages of bi- or multi-directional cores in these
assemblages.
In contrast, cores from the lower levels (C11N-C14) are
commonly single platform blade and multiple platform flake and blade
cores.
Some bi-directional flake cores are also present.
The lower
levels have higher frequencies of single platform cores and bladelet
tools with multiple (more than three) parallel flake scars than the
upper levels suggesting a somewhat different manner of blank detachment
may have been employed in the lower levels that enabled multiple blades
145
and bladelets to be removed from a single platform.
However, an
analysis of platform types showed little variability between levels.
The flake to blade ratios did not change significantly between
levels.
The slightly higher ratios for the lower levels (C12N-C13N) may
be influenced by the classification of debitage and the susceptibility
of the narrow bladelet blanks in the lower levels to breakage.
Previous
analyses suggested a one to one relationship for the manufacture of
flake and blade blanks (Clark et al. 1987, 1988).
Certainly, the high
proportion of flake, and flake and blade cores in this assemblage
supports this.
Median widths of unmodified blade and bladelet blanks from upper
levels (C08N-C10) were variable and could not be statistically separated
into different populations.
Median widths from the upper levels were
statistically different, however, from the lower levels (C12N-C14). The
median widths from level CllN are intermediate between the upper (C08NClO) and lower (C12N-C14) levels and could not be statistically
separated from either population, suggesting a change in lithic
manufacture towards wide bladelets occurred at this time.
Despite the
transitional nature of level CllN, the shape of the distribution of
unmodified blade and bladelet widths is statistically different from the
upper (CIO), but not the lower level (C12N).
The next stage in the operational sequence approach is to identify
the primary and secondary technologies used in the manufacture of
retouched tools.
Evidence for these technologies include blank
selection criterion, the frequencies of retouch type, and use of the
microburin technique.
No significant differences in blank selection
criterion for major tool types were noted between levels.
In a combined
analysis of all levels, approximately one-half of all non-microlithic
tools are manufactured on blades suggesting blades and flakes were
146
selected in equal proportions for non-mlcrolichic tool manufacture.
The
median widths of flake, blade and bladelet tools show similar trends to
those already observed in the debitage.
Flake tools in the upper levels
{C08N-C11N) have wider, thicker, and longer median values than those in
lower levels (C12N-C15) but upper and lower levels cannot be separated
into statistically different populations.
Blade and bladelet median
tool widths, however, are significantly wider (ca. 3-5 mm) in upper
(C08N-C11N) than in lower levels (CI2N-C15).
In addition, the shape of
the width distribution in upper levels statistically differs from that
in lower levels.
These data suggest that the trend towards the
manufacture of wide tools occurred before the trend towards the
manufacture of wide blanks.
Typological differences in the proportion of geometric to nongeometric microliths between natural levels 5 and 7 were already known
to exist from in-field analyses (Clark et al. 1987, 1988,- Donaldson
1986).
In this study, more geometric microliths are present in levels
C08N-C09N than in all of the lower levels suggesting the trend towards
manufacturing geometric microliths occurred slightly later than previous
analyses indicated.
Therefore, relatively wide blanks and tools were
being manufactured before an increase in the percentages of geometric
microlith forms occurred.
Secondary technological attributes of tool manufacture are
represented in retouch type and use of the microburin technique.
In
this study, retouch type did not vary significantly within major tool
types.
Differences in retouch type between levels seems to be
influenced by the proportion of major tool classes suggesting that
variability in retouch type within a major tool class requires a larger
sample than the one used in this analysis.
The frequency of the microburin technique changes markedly between
147
lower levels (C11N-C14) and upper levels (C08N-C10).
The microburin
indices reported here are higher than those presented in previous
studies, reflecting the higher number of microburins identified in this
analysis.
During in-field analyses, only 60 microburins were recorded
in the north and south portions of levels C11-C15.
In this study, 85
microburins were recorded in a sample approximately one-half the size of
the sample used in earlier studies (calculated by the total number of
artifacts recorded during in-field analyses).
As a result, microburin
indices in this study indicate that this technique is much more common
than earlier studies suggested.
In sum, the important features of the operational sequence used to
manufacture chipped stone debitage and tools in the lower levels (C12NC15) includes: (1) frequent use of single platform blade cores,- (2) the
manufacture of flake and blade blanks in similar proportions; (3) the
manufacture of narrow (ca. 7.1-7.4 mm) bladelet blanks; (4) the
manufacture of narrow, retouched microliths which are usually backed or
doubly backed and pointed; (5) the manufacture of slightly wide
microliths with more variable widths; and (6) the frequent use of the
microburin technique.
In contrast, the operational sequence in the
upper levels {C08N-C10) is characterized by (1) higher frequencies of
bi- and multi-directional flake and blade cores; (2) the manufacture of
equal proportions of flakes and blades; (3) the manufacture of wide,
blade and bladelet blanks; (4) the manufacture of wide, obliquely bitruncated and obliquely bi-truncated and backed microliths (ca. 12 mm);
and (5) the infrequent use of the microburin technique.
The lack of synchrony in the technological trend towards the
manufacture of wide tools and wide blanks suggests that primary lithic
technology which is associated with blank production is more resistant
to change or conservative than secondary lithic technologies which is
140
associated with tool manufacture.
Therefore, prirr.ar'/ lithic technology
should be a more reliable indicator of prehistoric culture groups than
secondary lithic technology.
Since flake to blade ratios are similar in
all levels, the techno-typological differences analyzed in this study
probably do not reflect different culture groups.
Therefore, regional
cultural continuity probably exists in this area.
The techno-typological differences between upper and lower levels
probably reflect both diachronic and adaptive changes.
The wider blanks
and tools, and higher frequencies of bi- and multi-direction flake and
blade cores in the upper levels may reflect a more variable lithic
technology and indicate a decrease in mobility in the upper levels.
The
removal of wide blanks would exhaust the core more readily than the
removal of narrow blanks.
Although this trend is not reflected in the
debitage to core ratios, the size effect may influence the frequency of
cores in these deposits.
Also, original core sizes may have varied
between levels which would have enabled more blanks to be manufactured
in some levels.
The bi- and multi-directional flake and blade cores and
the distribution of dorsal flake scars also suggest that more
variability in blank removal exists in the upper levels.
Since the
upper levels are associated with more mesic climatic conditions than the
lower levels, subsistence resources may have been more abundant and/or
diverse enabling a reduction in mobility during the occupation of the
upper levels.
The wide blanks and tools in the upper levels may also
reflect the need for a slightly more robust, durable blank in mesic
environments.
Regional Comparisons of Operational Sequences
Variability in the proportion of non-microlithic and microlithic
tools may relate more to differences in site activities, than to
technological differences in lithic manufacturing activities.
149
Technological differences, specifically core tiype^ flake to blade
ratios, blade and bladelet dimensions, proportion of geometric and nongeometric microliths, and the use of the microburin technique, are
important for identifying similarities in operational sequences between
sites.
Proportional differences in retouched bladelet types and
dimensions are considered diagnostic of the Kebaran, Geometric Kebaran,
early Hamran, Qalkhan and Natufian Epipaleolithic Complexes (Bar-Yosef
1984, 1987; Henry 1983, 1986).
The operational sequence identified in Step C, levels C11N-C15 of
WHS 1065 has some technological attributes that resemble those used in
other early Epipaleolithic assemblages.
Several stratified sites in the
eastern Levant have evidence for the use of the microburin technique and
are found in the Azraq Basin at Uwaynid 18 (trench 1, upper phase),
Uwaynid 14 (late and middle phases), and Jilat 6 (lower, middle, and
upper phases) (Byrd 1980; Garrard et al. 1985, 1986, 1987; Garrard and
Byrd 1992) and in southern Jordan at J405, Wadi Humeima (J406b), J407,
and Tor Hamar (J431) (Henry 1989b, 1995) (see Figure 1.1).
The adjusted microburin index for the Azraq Basin sites ranges
between 20.8 and 54.9 with a mean of 34.8.
Geometric tools comprise a
very small percentage (mean 0.5%) of retouched tool assemblages while
non-geometric microliths dominate them (mean 84%).
Single platform
bladelet cores consistently comprise 60% or more of the core sample
(Byrd 1988:259).
Non-geometric microliths forms include arched backed,
curved pointed pieces, la Mouillah points, and double truncated pieces.
Temporally, there is a typological change in these sites from small,
narrow microliths with arched and backed, curved pointed bladelets at
Uwaynid 14 (Middle Phase) and Jilat 6 (Lower Phase) which are dated to
ca. 19,800 to 18,400 + 350/250 BP) to stratigraphically later
assemblages with longer, thicker, backed bladelets such as robust la
150
Mouillah points or double truncated, backed bladelets (e.g., Uv/a^'nid 14
Upper Phase and Jilat 6 Middle Phase dated to ca. 18,900 to 18,4 00 + 250
BP)(Byrd 1988:260).
Metric data on these assemblages have not yet been
published.
At WHS 1065, the lithic assemblages from levels C11N-C15 differ
from early Azraq assemblages in that they have higher flake to blade
ratios, perhaps early evidence for the trend towards high flake to blade
ratios and wide microliths, and very few triangular microlithic forms.
The use of the microburin technique is often associated with the
manufacture of triangle or pointed microlithic forms (Henry 1995).
Chronometrically, the upper phase of Wadi Jilat 6, dating to ca. 16,700
to 15,470 + 14 0 BP, is contemporaneous with Step C, level C13 at WHS
1065 (Garrard et al. 1994).
Although the assemblage from Jilat 6 has a
high adjusted microburin index, it differs typologically, especially in
the frequency of small asymmetric triangles and microgravette points in
the assemblage (Byrd 1988 ; Garrard et al. 1994).
However, the higher
flake to blade ratios at Jilat 6 are more similar to those found in the
basal layers of Step C at WHS 1065.
Therefore, based on chronomecric
and techno-typological criteria, levels C11N-C15 at WHS 1065 resemble
early Epipaleolithic assemblages in the Azraq Basin.
The restricted microburin index for the southern Jordan sites
ranges between 20.3 and 50.0 with a mean of 33.3 (Henry 1995) .
Cores in
these assemblages also have single, unfaceted platforms on wedge shaped
cores.
Microburins are relatively large compared to those found in
other Epipaleolithic industries (Henry 1995:229).
Typologically,
Qalkhan points, manufactured by the microburin technique, are unique to
Qalkhan industry sites.
Other microliths include narrow (ca. 4.1-8.1
mm) forms with arched backed bladelets, narrow arched backed and pointed
bladelets, and la Mouillah points (Henry 1995).
While there are many
151
techno-typological similarities between the southern Jordan sites and
levels C11N-C15 at WHS 1065, Qalkhan points are not found in the WHS
1065 assemblage.
Henry (1995) considers the Azraq Basin sites and Sabra 3 and Adh
Daman (Schyle and LJerpmann 1988) to be Qalkhan Complex sites based on
the present of Qalkhan points and use of the microburin technique.
Sabra 3 and Adh Daman are surface sites and no chronometric dates have
been obtained from them (Schyle and Uerpmann 1988).
The Azraq Basin
sites have many similarities with southern Jordan sites. However,
Qalkhan style points were only identified by Henry (1995) in the Middle
Phases of Jilat 5.
In Syria, layers 4-7 at Yabrud (Rust 1950) and three
sites in the El-Kowm Oasis (Cauvin ec al. 1979; Cauvin 1981, Cauvin
Coqueugniot 1990) have many techno-typological similarities to Azraq
Basin and southern Jordan sites including narrow, arched backed, pointed
bladelets, scalene triangles, la Mouillah points, possible Qalkhan
points, and use of the microburin technique (Henry 1995:234).
However,
the chronological and typological relationships between all of these
sites are not clear.
Another southern Jordan site, Wadi Madamagh, also has narrow
arched backed and truncated microliths and microburins (Byrd 1994;
Kirkbride 1958) .
Although some radiocarbon determinations have been
obtained from this site (Schyle and Uerpmann 1988:47-52), it is not
clear how these dated deposits relate to Kirkbride's excavation.
Henry
(1989b, 1995) suggests that Wadi Madamagh represents a regional variant
of the Mushabian.
However, this classification does not appear to be
completely accepted (e.g., Byrd 1994)
The assemblages from levels C11N-C15 somewhat resemble "classic"
Kebaran assemblages which are characterized by the presence of abundant
backed bladelets and very narrow curved microliths (Bar-Yosef 1970) .
152
However, the microburin technique which is present in levels C11N-C14 is
not present at Kebaran sites in the western Levant.
Some suggest that
levels C11N-C15 resemble Wadi Hammeh 26 in the northern part of the Rift
Valley based on the high proportions of non-geometric microliths (Neeley
ec a.1. 1995:48).
However, the assemblages from Wadi Hammeh 26 do not
have evidence for use of the microburin technique at this time (Edwards
1987, 1990).
The microburin technique is not consistently used in the
western Levant until the Mushabian (ca. 14,000-13,000 BP).
The
Mushabian, defined in the Gebel Maghara, Northern Sinai and some other
sites in the Negev, is characterized by arched backed bladelets and la
Mouillah points manufactured with the microburin technique (Phillips and
Mintz 1977).
Chronometrically, this industry post-dates level C13 at
WHS 1065 by at least 2,500 years.
The microburin technique, therefore,
appears to be associated with arid and semi-arid environmental
adaptations.
The use of the microburin technique in the early Epipaleolithic
(ca. 19,000-15,000 BP) is unique to the eastern Levant.
Unfortunately,
the only reliable radiocarbon determinations come from the Azraq Basin
and Wadi Hasa.
Although many techno-typological similarities are found
with other sites in southern Jordan, the lack of reliable radiocarbon
dates and deeply stratified deposits at many southern sites make it
difficult to strongly tie these sites chronologically to others in the
region.
Levels C08N-C10 differ significantly from lower levels (C11N-C15)
The appearance of wide obliquely, bi-truncated and backed geometric
microliths in C08N-C10 (termed the broad "Hasa" lunate by Neeley et al.
1995) accompanied by the virtual absence of narrow arched/curved backed,
backed and pointed narrow microliths, higher percentages of geometric
microliths, and little evidence for the use of the microburin technique.
153
indicate different operational sequences for lithic manufacture were
used in the upper levels.
Unfortunately, these levels are undated but
obviously are later stratigraphically than lower levels.
Levels C08N-C10 have more geometric microliths than earlier
deposits.
In the western Levant, Geometric Kebaran microlithic
assemblages are dominated by triangles, trapezes, rectangularly shaped
pieces, and low microburin indices.
Many regional variants of the
Geometric Kebaran have been identified, e.g., the Falitan (Besancon et
al. 1977) and the Hamran (Henry 1983).
Bar-Yosef (n.d. cited in Goring-
Morris 1987:18) has subdivided this period into Al, a group represented
by trapeze-rectangles (most sites), and A2, those sites represented by
triangles (only three sites).
The Al grouping has been further
subdivided into an early phase with narrow bladelets and a late phase
with wide bladelets.
There is a tendency for wide bladelets and
trapeze-rectangles to dominate in southern sites (Bar-Yosef and Phillips
1977).
Although there are some general typological similarities, these
Geometric Kebaran assemblages especially in the Negev and Sinai have
much higher percentages of geometric microliths than WHS 1065, levels
C08N-C10 (Marks et al. 1976, 1977, 1978; Goring-Morris 1987; Neeley et
al. 1995).
Based solely on the relative frequency of geometric microliths, it
has been suggested that levels (C08N-C10) most closely resemble Middle
and Late Hamran assemblages, regional variants of the Geometric Kebaran,
from southern Jordan (Henry 1995; Neeley et al. 1995:47).
Middle Hamran
assemblages are characterized by relatively narrow bladelets, relatively
narrow trapeze/rectangles which were manufactured without the use of the
microburin technique while Late Hamran assemblages have comparatively
short bladelets, intentional use of the microburin technique, and the
appearance of lunates (Henry 1989b, 1995).
In the Final Hamran,
154
trapeze/rectangles are replaced by lunates ar.d there is an increase in
the use of the microburin technique (Henry 1989b, 1995).
Although there
are similarities in the percentages of geometric microliths between
Hamran assemblages and levels C08N-C10 at WHS 1065, the Middle-Final
Hamran has consistent use of the microburin technique and comparatively
narrow microlith forms.
These techno-typological differences suggest
that levels C08N-C10 are not Hamran industry sites.
Unfortunately, none
of these Hamran assemblages have reliable radiocarbon dates.
Typologically, the wide, obliquely, bi-truncated and obliquely bitruncated and backed microliths resembles some of the wide, geometric,
truncated and/or backed microliths and atypically wide trapezes
recovered from phase D at Kharaneh IV in eastern Jordan (Byrd 1994;
Muheisen 1985, 1988; Neeley ec al. 1995).
There is also little use of
the microburin technique in Phase D at Kharaneh IV.
Two apparently
reliable radiocarbon determinations of 15,200 + 450 BP and 15,700 + 160
BP have been obtained in Phase D (Byrd 1994:219).
Based on techno-
typological considerations, the upper levels (C08N-C10) most closely
resemble Kharaneh IV in northeastern Levant, not other variants of the
Geometric Kebaran which are found in southwestern and southeastern
Levantine areas.
Conclusions
Some technological attributes (e.g., flake to blade ratios)
between levels C08N-C15 at WHS 1065 indicate regional cultural
continuity in the west-central Levant.
However, substantial variability
in the operational sequences used to manufacture chipped stone exists
between the upper (C08N-C10) and the lower levels (C11N-C15).
variability reflects both diachronic and adaptive changes.
This
Not all
technological and typological changes between the upper and lower levels
are synchronous.
This study suggests that some primary technological
155
change (specifically the manufacture of wide tools) occurred before sctr.e
secondary technological change (specifically the manufacture of wide
unmodified blanks).
Therefore, primary lithic technology should be a
more reliable indicator of prehistoric culture groups than secondary
lithic technology.
The manufacture of wide blanks and tools, and bi-
and multi-directional flake and blade cores in the upper levels may
indicate a more variable, lithic technology and a decrease in mobility
in the upper levels.
These technological changes appear to be
associated with more mesic climatic conditions at that time.
The
manufacture of wide blanks and tools in the upper levels may also
reflect the need for more robust, durable blanks in mesic environments.
Changes in the frequency of the microburin technique between the upper
and the lower levels may also be associated with changing adaptations to
different environmental conditions, since use of the microburin
technique is frequently associated with arid and semi-arid environmental
adaptations.
The assemblages from both the upper (C08N-C10) and the lower
levels (C11N-C15) at WHS 1055 seem to most closely resemble contemporary
sites in the Azraq Basin.
The lower levels (C11N-C15) resemble Uwaynid
18, Uwaynid 14 and Jilat 6 while the upper levels (C08N-C10) most
closely resemble Kharaneh IV in northeastern Levant.
It would be
interesting to determine if the timing of the techno-typological trends
identified in this study are also found at other contemporary sites in
the eastern Levant, specifically the early Epipaleolithic sites in the
Azraq Basin.
Although others suggest similarities to southern and
western Levantine sites, these techno-typological and temporal
associations are problematic.
Local and regional environmental conditions around Pleistocene
Lake Hasa and the Levant changed from xeric to relatively mesic
156
conditions at about this time.
Some of the techno-t^'pclcgical changes
and changes in the operational sequence of lithic manufacture at Tor alTareeq reflect general regional trends in lithic technology and appear
to be associated with more mesic climatic conditions during the middle
Epipaleolithic.
It has not yet been determined to what extent
environmental change is related to techno-typological changes in lithic
manufacture in other assemblages.
Additional studies that more closely
link lithic, environmental, and subsistence data are necessary in order
to determine if these changes occur in similar paleoenvironmental and
temporal contexts.
APPENDIX A:
WHS 1065 DEBITAGE ANALYSTS CODING LIST
VARIABLE NAME (VARIABLE)
UNIT/LEVEL/DIVISION (UNITLEVEL$)
N - North
S - South
B - Not divided
# - Feature
BLANK TYPE (BLANK)
Flake (Blank Group [BG]) =1)
1 - Flake <1 cm
2 - Flake 1-2 cm
3 - Flake 2-3 cm
4 - Flake >3 cm
Blade (BG = 2)
5 - Blade >3cm - IDC
6 - Blade >3cm - 2DC
7 - Bladelet L<3 cm, W < 1.2 cm
Core rejuventation (BG = 3)
8 - Core rejuvenation blade
9 - Core rejuventation bladelet
10 - Core rejuvenation flake
Microburin (BG = 4)
11 - Regular microburin
12 - Piquant tiedre
13 - Krukowski microburin
Burin spall (BG = 5)
14 - Burin spall
Shatter (BG = 6)
15 - Shatter
Cores (BG = 7 and 8)
16 - Cores ((BG = 7)
17 - Core fragments (BG = 8)
•Debris (BG = 9) (medial and distal flake fragments)
Tools (BG = 10 and 11)
18 - Tool (BG = 10)
19 - Tool fragment (BG = 11)
CORTEX (CORTEX)
0 - Not applicable
1 - Absent
2 - Present
COMPLETENESS (COMPLETE)
0 - Not Applicable
1 - Complete
2 - Bulb present, distal end absent
3 - Bulb absent, distal end pre^rent
4 - Medial segment
COUNT (NUMBER)
FOR BLADES
LENGTH (L) - Measured to the nearest 0.1 millimeter.
WIDTH (W) - Measured to the nearest 0.1 millimeter.
THICKNESS (TH) - Measured to the nearest 0.1 millimeter.
WEIGHT (WT) - Measured to the nearest 0.1 gram.
158
APPENDIX B:
WHS 10<55 TOOL AND CORE ANALYSIS CODING LIST
VARIABLE NAME (VARIABLE)
UNIT/LEVEL/DIVISION (UNITLEVEL$)
N - North
S - South
B - Not divided
# - Feature
BLANK TYPE (BLANK)
Blades (L > 3 cm, W > 1.2 cm)
1 - 1st order (Blank group [BG] = 1)
2 - 2nd order (BG = 1)
3 - Core rejuvenation
4 - Crested
5 - Laminar tablet
Bladelets (L < 3 cm, W < 1.2 cm)
6 - 1st order (BG = 2)
7 - 2nd order (BG = 2)
8 - Core rejuvenation
9 - Laminar tablet
Flajces
10 - Flake
11 - Core rejuvenation
12 - Core tablets
Shatter
13 - Shatter
Microburins
14 - True microburin
15 - Krukowski microburin
16 - Piquant triedre
Burin spalls
17 - Burin spalls
Cores
18 - Single platform, blade
19 - Opposed (perpendicular) platform, blade
20 - Single platform, flake
21 - Opposed platform, flake or flake and blade
22 - Multiple platform, flake and blade
23 - Multiple platform, flake
24 - Bipolar blade
25 - Amorphous, exhausted core
26 - Bipolar, flake
27 - Two platforms, not opposed, flake
28 - Two platforms, not opposed, blade
29 - Single platform on large flake, flake
30 - Other
CORTEX
(CORTEX)
0 - 0 % No cortex
1 - 100 % Cortex present
2 - 51-99 % Cortex present
3 - 1-50 % Cortex present
4 - Cortex present on platform only
TOOL TYPES (TYPE)
0 - Not appliciable
Scrapers (Standard Tool Type [ST] = 1
1 - On a flake
2 - On a retouched flake
3 - Rounded or circular
4 - Thumbnail
5 - Transversal
6 - Sidescraper
7 - On a blade or bladelet
8 - On a retouched blade or bladelet
9 - Ogival
10 - Denticulate
11 - Double
Carinated (ST = 2)
12 - Shouldered or nosed
13 - Broad carinated
14 - Narrow carinated
15 - Lateral carinated
15 - Core scraper
17 - Double carinated
Burins (ST = 3)
18 - Dihedral
19 - Dihedral angle
20 - Angle burin on a break or natural surface
21 - Multiple dihedral burin
22 - On straight truncation
23 - On oblique truncation
24 - On concave truncation
25 - On convex truncation
26 - Multiple on truncation
27 - Multiple on mixed
28 - Beaked
29 - Carinated
30 - Flat faced
31 - Transverse on lateral notch
Multiple tools (ST = 4)
32 - Burin/scraper
33 - Other
Retouched/backed blades (ST = 5)
34 - Partially retouched
35 - Completely retouched
36 - Helwan blade
37 - Backed knife
38 - Curved backed knife
39 - Retouched/back varia
40 - Backed/retouched fragments (ST = 18)
Truncations (ST = 6)
41 - Straight
42 - Concave
43 - Oblique
44 - Backed and truncated or retouched and truncated
Microliths (ST = 7)
45 - Backed
46 - Retouched and backed
47 - Backed and notched
48 - Truncated
49 - Oblique truncation
160
50
51
52
53
54
55
56
57
58
59
60
61
-
Backeci and truncated
Backed and oblique truncation
Bi-truncated and backed
Curved/arched and backed
Curved/arched and backed with basal modification
Pointed - semi-abrupt Sc/or backed retouch on 2 edges
Pointed straight backed microlith (micropoint)
Microperforator
Retouched/backed bladelet varia
Retouched/backed fragment
Humped backed bladelet
Backed obliquely truncated and retouched or notched
on oppposite edge
62 - Bitruncated (obliquely) and not backed
63 - Obliquely truncated & retouched
Geometries (ST = 8)
64 - Trapeze/rectangle
65 - Proto-trapeze
66 - Trapeze
67 - Assymmetrical trapeze
68 - Trapeze with one convex end
69 - Lunate
70 - Atypical lunate
71 - Isoceles triangle
72 - Atypical triangle
73 - Other
Notches and Denticulates (ST = 9)
74 - Retouched notch
75 - Two or more notches
76 - Denticulate
Various (ST = 10)
77 - Retouched flake or piece (ST = 11 for complete flake,
ST = 16 for flake fragments)
78 - Utilized blank (ST = 10)
79 - Other (ST = 12)
8 0 - Microburin (ST = 15)
81 - Core (ST = 14)
PLACEMENT OF RETOUCH (RET_PLACE)
0 - Not applicable
1 - Lateral obverse
2 - Lateral inverse
3 - Distal obverse (or microburin scar)
4 - Distal inverse (or microburin scar)
5 - Proximal obverse (or microburin scar)
6 - Proximal inverse (or microburin scar)
7 - Obverse/inverse
8 - Distal and lateral obverse
9 - Distal and lateral inverse
10 - Distal/proximal
11 - Lateral inverse and distal obverse
12 - Proximal and lateral inverse
13 - Proximal and lateral obverse
14 - Top ridge
15 - Inverse
161
RETOUCH ORIENTATION (RET_ORIENT)
0 - Noc applicable
1 - Left lateral edge
2 - Right lateral edge
3 - Both lateral edges
4 - Left distal (or microburin scar)
5 - Right distal (or microburin scar)
6 - Left proximal
7 - Right proximal
8 - Complete distal
9 - Lateral and distal
10 - Proximal and distal
11 - Complete proximal
12 - Proximal and lateral
13 - Lateral, proximal, and distal
14 - Top ridge
RETOUCH TYPE (RET_TYPE)
0 - Not applicable
1 - Very fine and fine (ouchata)
2 - Semiabrupt
3 - Abrupt (backing)
4 - Mixed
5 - Side scraper retouch
6 - Bipolar
7 - Inverse
8 - Alternate
9 - Helwan
10 - Marginal (utilized, crushed or battered)
11 - Burin blow
DISTRIBUTION OF
0 1 2 3
4
5
S
7
PLATFORM TYPE
0
1
2
3
4
5
6
7
-
FLAKE SCARS(D_FLK_SCAR)
Not applicable/no platform
Cortex/no flake scars
No cortex/flake scars parallel to direction of the blow
(I I to dir. of blow)
Two flake scars (11 to dir. of blow)
Multiple flake scars (|| to dir. of blow)
One flake scar (not || to dir. of blow)
Multiple flake scars (not |1 to dir. of blow)
Mixed flake scars (|j and not jj to dir. of blow)
(PLATFORM)
- Not applicable or missing
- Cortex, plain, not prepared
- Non-cortical - plain, prepared by single blow
- Faceted, several steep splits or facets determines the
angle, not the point of percussion
- Punctiform (pointed)
- Crushed
- Broken
- Dihedral
SHAPE OF DISTAL
0 1 2 3 -
END (DISTAL_END)
Not applicable or missing
Pointed
Blunt and thick
Convex/arched
162
4
5
6
7
8
-
Straight
Oblique
Irregular/wavy
Concave
Microburin scar
SHAPE OF LATERAL EDGES (LATERAL_ED)
0 - Not applicable
1 - Straight
2 - Convex
3 - Concave
4 - Irregular
5 - Straight/convex
6 - Straight/concave
7 - Straight/irregular
8 - Convex/concave
9 - Convex/irregular
MATERIAL TYPE
1
2
3
4
5
(MATERIAL)
- Fine grained chert
- Coarse grained chert
- Mixed - fine and coarse grained chert or with inclusions
- Fossiliferous limestone
- Other
LENGTH (LENGTH) - Measured to the nearest
WIDTH (WIDTH) - Measured to the nearest
O.l millimeter.
0.1 millimeter.
THICKNESS (THICKNESS) - Measured to the nearest
0.1 millimeter.
WEIGHT (WEIGHT) - Measured to the nearest O.l gram.
COMMENTS (COMMENTS $)
163
REFERENCES
Anderson-Geraud, P.C.
1983 A consideration of the uses of certain backed and "lustered"
stone tools from late Mesolithic and Natufian levels of Abu
Hureyra and Mureybet (Syria). In Traces d'Utilization sur
les Outlis Neolithiques du Proche Orient, edited by M.-C.
Cauvin, pp. 77-106. Travaux de la Maison de 1'Orient, no.
5. Maison de I'Orient Mediterranean, Lyon.
Aurenche, 0., J. Cauvin, M.-C. Cauvin, L. Copeland, F. Hours, and P.
Sanlaville
1981 Chronologie et organization de I'espace dans le Proche
Orient de 12000 a 5600 avant J.-C. In Prehistoire de
Levant, edited by J. Cauvin and P. Sanlaville, pp. 571-602.
Colloques Internationaux de CNRS 598, Paris.
Banning, E.B., R.R. Dods, J. Field, I. Kuiut, J. McCorriston, J.
Siggers, H. Taani and J, Triggs
1992 Tabaqat al-Bumma: 1990 excavations at a Kebaran and Late
Neolithic site in Wadi Ziqlab. Annual of the Department of
Antiquities 36:43-62.
Baker, C.
1978
The size effect: an explanation of variability in surface
artifact assemblage content. American Antiquity 43:288-293.
Baker, C., and M.B. Schiffer
1975 Archaeological evidence for the size effect. In Arkansas
Eastman Archaeological Project, edited by C. Baker, pp. 117122. Arkansas Archaeological Survey, Research Report No. 6.
Arkansas Archaeological Society, Fayetteville, Arkansas.
Baruch, U.
1994
The Late Quaternary pollen record of the Near East. In Late
Quaternary Chronology and Palaeoclimates of the Eastern
Mediterranean, edited by 0. Bar-Yosef and R.S. Kra, pp. 103120. Supplement to Radiocarbon, Department of Geosciences,
University of Arizona, Tucson, Arizona.
Baruch, U., and S. Bottema
1991 Palynological evidence for climatic changes in the Levant
ca. 17,000-9,000 B.P. In The Natufian Culture in the
Levant, edited by 0. Bar-Yosef and F. Valla, pp. 11-20.
International Monographs in Prehistory, Ann Arbor, Michigan.
Bar-Yosef, 0.
1970 The Epipaleolithic Cultures of Palestine. Unpublished Ph.D.
dissertation. Department of Archaeology, Hebrew University,
Jerusalem.
1975
The Epipaleolithic in Palestine and Sinai. In Problems in
Prehistory: North Africa and the Levant, edited by F.
Wendorf and A.E. Marks, pp. 313-371. SMU Press, Dallas.
1977
Section A, Mushabi IV. In Prehistoric Investigations in the
Gebel Maghara, Northern Sinai, edited by 0. Bar-Yosef and
J.L. Phillips, pp. 184-193. Qedem 7. Monographs of the
164
Institute of Archaeology, Hebrew University, Jerusalem.
1981
The Epipaleolithic complexes in the southern Levant. In
Prehistoire de Levant, edited by J. Cauvin and P.
Sanlaville, pp. 389-408. Colloques Internationaux de CNRS
598, Paris.
1987
Late Pleistocene adaptations in the Levant. In The
Pleistocene Old World: Regional Perspectives, edited by 0.
Soffer, pp. 219-236. Plenum Press, New York.
1990
The last glacial maximum in the Mediterranean Levant. In
The World at 18,000 BP, edited by C. Gamble and 0. Soffer,
pp. 58-77. Unwin Hyman Ltd., London.
1991a Stone tools and social context in Levantine prehistory.
Perspectives on the Past: Theoretical Biases in
Mediterranean Hunter-Gatherer Research, edited by G.A.
Clark., pp. 371-395. University of Pennsylvania Press,
Philadelphia.
In
1991b The search for lithic variability among Levantine EpiPalaeolithic industries. In 25 Ans D'Etudes Technologiques
en Prehistoire, pp. 319-355. XI Rencontres Internationales
d'Archeologie et d'Histoire d'Antibes. Editions APDC, JuanLes-Pins.
Bar-Yosef, O., and A. Belfer-Cohen
1977 The Lagaman industry. In Prehistoric Investigations in
Gebel Maghara, Northern Sinai, edited by 0. Bar-Yosef and
J.L. Phillips, pp. 42-88. Qedem 7. Monographs of the
Institute of Archaeology, Hebrew University, Jerusalem.
1989
The origins of sedentism and farming communities.
of World Prehistory 3(4):447-498.
1991
From sedentary hunter-gatherers to territorial farmers in
the Levant. In Between Bands and States, edited by S.
Gregg, pp. 181-202. Center for Archaeological
Investigations Occasional Paper No. 9. Southern Illinois
University at Carbondale, Illinois.
Journal
Bar-Yosef, 0., and J.L. Phillips (editors)
1977 Prehistoric Investigations in Gebel Maghara, Northern Sinai,
edited by 0. Bar-Yosef and J.L. Phillips. Qedem 7.
Monographs of the Institute of Archaeology, Hebrew
University, Jerusalem.
Bar-Yosef, O., and A. Killebrew
1985 Wadi Sayakh - a Geometric Kebaran site in Southern Sinai.
Paleorient 10:95-102.
Bar-Yosef, 0., and A.N. Goring-Morris
1977 Geometric Kebaran A occurrences. In Prehistoric
Investigations in Gebel Maghara, Northern Sinai, edited by
0. Bar-Yosef and J.L. Phillips, pp. 115-148. Qedem 7.
Monographs of the Institute of Archaeology, Hebrew
University, Jerusalem.
165
Bar-Yosef, 0., and J.C. Vogel
1987 Relative and absolute chronology of the Epi-Paleolithic in
the Southern Levant. In Chronologies in the Near Ease,
edited by 0. Aurenche, J. Evin and F. Hours, pp. 219-245.
BAR International Series 379, Oxford.
Begin, Z.B., A. Ehrlich, and Y. Nathan
1974 Lake Lisan, the Pleistocene precursor of the Dead Sea.
Bulletin of Geological Survey, Min. Comm. Ind., State of
Israel, No. 63:30.
Begin, Z.B., W. Broeker, B. Buchbinder, Y. Druckman, A. Kaufman, M.
Magaritz, and D. Neev
1985 Dead Sea and Lake Lisan levels in the last 30,000 years:
preliminary report. Geological Survey of Israel Report
29:1-30.
Belfer-Cohen, A., and P. Goldberg
1982 An Upper Palaeolithic site in South Central Sinai.
Exploration Journal 32:185-189.
Besancon J., L. Copeland, and F. Hours
1975-7 Tableaux de prehistoire Libanaise.
a
Israel
Paleorient 3:5-45.
Binford, L.R.
1973 Interassemblage variability - the Mousterian and the
"functional" argument. In The Explanation of Culture
Change, edited by C. Renfrew, pp. 227-254. Duckworth,
London.
1983
Working at Archaeology.
Academic Press, New York.
Binford, L.R., and S.R. Binford
1966 A preliminary analysis of functional variability in the
Mousterian of Levallois facies. American Anthropologist
68 ( :2) :238-295.
Bordes, F.
1961
Mousterian cultures in France.
Science 134 (3482):803-810.
1973
On the chronology and contemporaneity of different
Paleolithic cultures in France. In The Explanation of
Culture Change, edited by C. Renfrew, pp. 217-226.
Duckworth, London.
1978
Typological variability in the Mousterian layers at Peche de
I'Aze I, II, and IV. Journal of Anthropological Research
34 (2) : 181-193.
Bottema, S.
1987 Chronology and climatic phases in the Near East from 16,00010,000 B.P. In Chronologies in the Near East. Relative
Chronologies and Absolute Chronology 16,000-4,000 B.P.,
edited by 0. Aurenche, J. Evin, and F. Hours, pp. 295-310.
BAR International Series 379, Oxford.
Bottema, S., and W. Van Zeist
1981 Palynological evidence for the climatic history of the Near
166
Ease 50.000-6,000 B.P. In Prshlstclrs ds Levant, edited by
J. Cauvin and P. Sanlaville, pp. 111-132. Colloques
Internationaux de CNRS 598, Paris.
Buchbinder B., Z.B. Begin and G.M. Friedmand
1974 Pleistocene algal tufa of Lake Lisan, Dead Sea area, Israel.
Israel Journal Earth Science 23:131-138.
Buller, H.
1983
Byrd, B.F.
1987
Methodological problems in the microwear analysis of tools
selected from the Natufian sites of el Wad and Ain Mallaha.
In Trace d'Utilization sur les Outlis Neolithiques due
Proche Orient, edited by M. Cauvin, pp. 107-126. Travaux de
la Maison de 1'Orient, Maison de 1' Orient, Lyon.
Beidha and the Natufian: Variability in Levantine
Settlement and Subsistence. Unpublished Ph.D. dissertation,
Department of Anthropology, University of Arizona, Tucson,
Arizona.
1988
Late Pleistocene settlement diversity in the Azraq Basin.
Paleorient 14:257-264.
1990
Late Pleistocene assemblage diversity in the Azraq Basin.
Paleorient 14 (2) :257-265 .
1994
Late Quaternary hunter-gatherer complexes in the Levant
between 20,000 and 10,000 BP. In Late Quaternary Chronology
and Paleocliwates of the Eastern Mediterranean, edited by 0.
Bar-Yosef and R.S. Kra, pp. 205-226. Supplement to
Radiocarbon, Department of Geosciences, University of
Arizona, Tucson, Arizona.
Byrd, B.F., and A.N. Garrard
1990 The last glacial maximum in the Jordanian desert. In The
World at 18,000 BP, Vol. 1, edited by C. Gamble and 0.
Soffer, pp. 78-96. Unwin Hyman Ltd., London.
Cauvin, J., M.-C. Cauvin, and 0. Stordeur
1979 Recherches prehistorique a El Kowm (Syrie), premiere
campagne 1978. Chiers de I'Euphrate 2:80-117. CNRS, Paris.
Cauvin, M.-C.
1981 L'Epipaleolithique du Levant. In Prehistoire du Levant,
edited by J. Cauvin and P. Sanlaville, pp. 439-444.
Editions du CNRS, Paris.
Cauvin, M.-C., and E. Coqueugniot
1990 L'oasis d'El Kowm et le Kebarxen Geometrique.
P a l e o r i e n t 1 4( 2 ) : 2 7 0 - 2 a i .
Childe, V. Gordon
193 9 Man Makes Himself.
In
Oxford University Press, New York.
Clark, G.A.
1989 Romancing the stones: biases, style and lithics at La
Riera. In Alternative Approaches to Lithic Analysis, edited
167
by D.O. Henry and G.K. Cdell, pp. 27-50. Archeological
Papers of the American Anthropological Association Number 1.
1991
Paradigms, realism, adaptation, and evolution. In
Perspectives on the Past: Theoretical Biases in
Mediterranean Hunter-Gatherer Research, edited by G.A.
Clark, pp. 441-439. University of Pennsylvania Press,
Philadelphia.
1992
Wadi al'Hasa Paleolithic settlement patterns: Negev and
South Jordan models compared. In Studies in the History and
Archaeology of Jordan, Vol. IV, edited by M. Zaghloul et
al., pp. 89-96. Department of Antiquities, Amman, Jordan.
Clark, G.A., N.R. Coinman, J. Lindly, and M. Donaldson
1987 A diachronic study of Paleolithic and aceramic Neolithic
settlement patterns in the Wadi Hasa, West-central Jordan.
In Studies in the History and Archaeology of Jordan III,
edited by A. Hadidi, pp. 215-223. Department of
Antiquities, Amman, Jordan and Routledge & Kegan Paul,
London.
Clark, G.A., J. Lindly, M. Donaldson, A. Garrard, N. Coinman, J.
Schuldenrein, S. Fish, and D. Olszewski
1987 Paleolithic archaeology in the southern Levant: a
preliminary report of excavations at Middle, Upper, and
Epipaleolithic sites in Wadi El-Hasa, West-central Jordan.
Annual of the Department of Antiquities of Jordan 31:19-78.
1988
Excavations at Middle, Upper and Epipaleolithic sites in the
Wadi al-Hasa, West-central Jordan. In The Prehistory of
Jordan, edited by A.N. Garrard and H.G. Gebel, pp. 209-285.
BAR International Series 396, Oxford.
Clark, G.A., M. Neeley, B. MacDonald, J. Schuldenrein, and K. 'Amr
1992 Wadi al-Hasa Paleolithic project 1992: preliminary report.
Annual of the Department of Antiquities of Jordan 36:13-21.
Clark, J.G.D.
1954 Excavations at Star Carr.- An Early Mesolithic Site at
Seasmer near Scarborough, Yorkshire. Cambridge University
Press, Cambridge, England.
1975-77 Interpretation of prehistoric technology from ancient
Egyptian and other sources, part 2. Paleorient 3:127-150.
Clark, J.G.D., J.L. Phillips, and P.S. Stanley
1974 Interpretation of prehistoric technology from ancient
Egyptian and other sources, part l. Paleorient 2:323-388.
Clarke, D.L.
1968 Analytical Archaeology: Collected Papers of David L.
Clarke, edited by his colleagues. Academic Press, London.
1976
Mesolithic Europe: the economic basis. In Problems in
Economic and Social Archaeology, edited by G. Sleveking,
I.H. Longworth and K.E. Wilson, pp. 449-481. Duckworth,
163
London.
1979
Analytical Archaeology.
Mechuen & Co. Ltd., London.
Close, A.E.
1978 The identification of style in lithic artifacts.
Archaeology 10:223-237.
1989
World
Identifying style in stone artifacts: a case study from the
Nile Valley. In Alternative Approaches to Lithic Analysis,
edited by D.O. Henry and G.H. Odell, pp. 3-26.
Archeological Papers of the American Anthropological
Association Number 1.
Coinman, N.R., G.A. Clark, and M.L. Donaldson
1989 Aspects of structure in an Epipaleolithic occupation site in
West-central Jordan. In Alternative Approaches to Lithic
Analysis, edited by D.O. Henry and G.H. Odell, pp. 213-235.
Archeological Papers of the American Anthropological
Association Number 1.
Copeland, L., and C. Vita-Finzi
1978 Archaeological dating of geological deposits in Jordan.
Levant 10:10-25.
Curwen, W.
1941
Darmon, F.
1987
Some food gathering implements: a study of Mesolithic
tradition. Antiquity 15:320-336.
Analyses polliniques de trois sites Natoufiens (Ancien,
Recent, Final) dans la Region de Salibiya-Fazael.
Paleorient 13:121-129.
Dibble, H.L.
1984 Interpreting typological variation of Middle Paleolithic
scrapers: function, style, or sequence of reduction?
Journal of Field Archaeology 11(4):431-436.
1985
Raw material variability in Levallois flake manufacture.
Current Anthropology 26(3):391-393.
1987
The interpretation of Middle Paleolithic scrapers
morphology. American Antiquity 52:109-117.
Donaldson, M.L.
1986 Preliminary analysis of lithics from WHS site 1065: an
Epipaleolithic site in the Wadi Hasa, Jordan. Unpublished
seminar paper. University of Arizona, Tucson, Arizona.
1991
Historic biases in modern perceptions of the Levantine
Epipaleolithic. In Perspectives on the Past: Theoretical
Biases in Mediterranean Hunter-Gatherer Research, edited by
G.A. Clark, pp. 341-352. University of Pennsylvania Press,
Philadelphia.
169
Donaldson, M.L., and G.A. Clark
1906 WHS site 1065: a Geometric Kebaran site in the Wadi Hasa
(West-central Jordan). Unpublished manuscript, Arizona
State University, Tempe.
Druckman, Y., M. Magaritz, and A. Sneh
1987 The shrinking of Lake Lisan, as reflected by the diagenesis
of its marginal oolithic deposits. Israel Journal of Earth
Sciences 36:101-106.
Edwards, P.C.
1987 Late Pleistocene occupation in the Wadi al-Hammeh, Jordan
Valley. Unpublished Ph.D. dissertation, University of
Sydney.
1990
Kebaran occupation at the last glacial maximum in Wadi alHammeh, Jordan Valley. In The World at 18,000 BP, Vol. 2,
edited by C. Gamble and 0. Soffer, pp. 97-118. Unwin Hyman
Ltd., London.
Edwards, P., S. Bourke, S. Colledge, J. Head, and P. Macumber
1988 Late Pleistocene prehistory in the Wadi el-Hammeh, Jordan
Valley. In The Prehistory of Jordan. The State of Research
in 1986, edited by A.N. Garrard and H.G. Gebel, pp. 525-565.
BAR International Series 396, Oxford.
Ewing, J.F.
1949 The treasures of Ksar Akil.
Thought 24:255-288.
Ferring, C.R.
1977 The late Upper Paleolithic site of Ein Aqev East. In
Prehistory and Paleoenvironments in the Central Negev,
Isreal. Volume II: The Advat Aqev Area, Part 2 and the
Harif, edited by A.E. Marks, pp. 81-118. SMU Press, Dallas.
1988
Technological change in the Upper Paleolithic of the Negev.
In Upper Pleistocene Prehistory of Western Eurasia, edited
by H. Dibble and A. Montet-White, pp. 333-348. University
Museum, Philadelphia.
Garrard, A.N., D. Baird, and B.F. Byrd
1994 The chronological basis and significance of the Late
Paleolithic and Neolithic sequence in the Azraq Basin,
Jordan. In Late Quaternary Chronology and Paleoclimates of
the Eastern Mediterranean, edited by 0. Bar-Yosef and R.S.
Kra, pp. 177-200. Supplement to Radiocarbon, Department of
Geosciences, University of Arizona, Tucson, Arizona.
Garrard, A.N., A. Betts, B.F. Byrd, and C. Hunt
1987 Prehistoric environment and settlement in the Azraq Basin:
an interim report on the 1985 excavation season. Levant
19:5-25.
Garrard, A.N., A. Betts, B.F. Byrd, S. Colledge, and C. Hunt
1988a Summary of paleoenvironmental and prehistoric investigations
in the Azraq Basin. In The Prehistory of Jordan; The State
of Research in 1986, edited by A.N. Garrard and H.G. Gebel,
pp. 311-337. BAR International Series 396, Oxford.
170
Garrard, A.N.. and B.F. Byrd
1992 New dimensions to the Epipaleolithic of the Wadi el-Jilat in
Central Jordan. Paleorient 18(l):47-62.
Garrard, A.N., B.F. Byrd, and A. Betts
1986 Prehistoric environment and settlement in the Azraq Basin,
an interim report on the 1984 excavation season. Levant
18 :5-24 .
Garrard, A.N., B.F. Byrd, P. Harvey, and F. Hivernel
1985 Prehistoric environment and settlement in the Azraq Basin, a
report on the 1982 survey season. Levant 17:1-28.
Garrard, A.N., and Hans Georg Gebel (editors)
1988 The Prehistory of Jordan: The State of Research in 1986.
BAR International Series 396, Oxford.
Garrard, A.M., S. Colledge, C. Hunt and R. Montague
1988b Late Quaternary environmental history during the late
Pleistocene and early Holocene in the Azraq Basin.
Paleorient 14(2):40-49.
Garrod, D.A.E.
1932 A new mesolithic industry: the Natufian of Palestine.
Journal of the Royal Anthropological Society 62:257-269.
1942
Excavations at the cave of Shukbah, Palestine, 1928.
Proceedings of the Prehistoric Society 8:1-20.
1954
Excavations at the Mugharet El-Kebara, Mount Carmel, 1931:
The Aurignacian industries. Proceedings of the Prehistoric
Society 43:155-192.
1958
The Natufian culture: the life and economy of a Mesolithic
people in the Near East. Proceedings of the British Academy
43:155-192.
Garrod, D.A.E., and D.M.A. Bate
1937 The Stone Age of Mount Carmel, Vol. 1.
Oxford.
Claredon Press,
Gat, J., and M. Magaritz
1980 Climatic variations in the eastern Mediterranean area.
Naturwissenschaften 67:80-87.
Gifford, D.
1978 Ethnoarchaeological observations of natural processes
affecting cultural materials. In Explorations in
Ethnoarchaeology, edited by R. Gould, pp. 51-62. Academic
Press, New York.
Gilead, I., and 0. Marder
1989 Geometric Kebaran sites in the Nahal Rut area, western
Negev, Israel. Paleorient 15:123-137.
Gilead, I.
1901a The Upper Paleolithic tool assemblages from the Negev and
Sinai. In Prehistorie du Levant: Chronologie et
171
1'organization de I'espace depuis Iss crlginss jiisqu'3.iL.
No. 598, pp. 331-342. Maison de L'Orient, Lyon.
1981b The Upper Paleolithic in Sinai and the Negev. Unpublished
Ph.D. dissertation, Hebrew University, Jerusalem.
1983
Upper Paleolithic occurrences in Sinai and the transition to
the Epi-Paleolithic. Paleorienc 9:30-53.
1984
Is the term "Epipaleolithic" relevant to Levantine
prehistory? Current Anthropology 25)(2):227-228.
1988
The Upper Paleolithic to Epi-PaleoliChic transition in the
Levant. Paleorient 14(2):177-182.
1991
The Upper Paleolithic period in the Levant.
World Prehistory 5(2):105-154.
Journal of
Goldberg, P.
1976 Upper Pleistocene geology of the Advat/Agev area. In
Prehistory and Paleoenviroaments in the Central Negev,
Israel, Vol. I, edited by A.E. Marks, pp. 25-56. SMU Press,
Dallas.
1981
Late Quaternary stratigraphy of Israel: an eclectic view.
In Prehistoire du Levant, edited by J. Cauvin and P.
Sanlaville, pp. 55-66. CNRS, Paris.
1986
Late Quaternary environmental history of the southern
Levant. Geoarchaeology 1(3):225-244.
Goldberg, P., and 0. Bar-Yosef
1982 Environmental and archaeological evidence for climatic
change in the southern Levant and adjacent areas. British
Archaeological Reports 3-133:399-414.
Goldberg, P., and B. Brimer
1983 Late Pleistocene geomorphic surfaces and environmental
history of Avdat/Havarim area, Mahal Zin. In Prehistory and
Paleoenvironments in the Central Negev, Israel, Vol. Ill,
edited by A.E. Marks, pp. 1-13. SMU Press, Dallas.
Goodfriend, G., and M. Magaritz
1988 Paleosols and Late Pleistocene rainfall fluctuations in the
Negev Desert. Nature 322:144-146.
Goring-Morris, A.N.
1987 At the Edge: Terminal Pleistocene Hunter-Gatherers in the
Negev and Sinai. BAR International Series 361, Oxford.
1995
Complex hunter/gatherers at the end of the Paleolithic
(20,000-10,000 BP). In The Archaeology of Society in the
Holy Land, edited by Thomas E. Levy, pp. 141-168. Leicester
University Press, London.
Gould, R.A., D.A. Koster, and A.H.L. Sontz
1971 The lithic assemblage of the Western Desert Aborigines of
Australia. American Antiquity 36:41-66.
172
Hayden. B.
1977
1979
Stone tool functions in the Western Desert. In Scone Tools
as Cultural Markers: Change, Evolution, Complexity, edited
by R.V.S. Wright, pp. 178-188. Australian Institute of
Aborigine Study, Canberra, Australia.
Snap, shatter and super-fractures: use wear of stone skin
scrapers. In Lithic Use-VIear Analysis, edited by B. Hayden,
pp. 207-230. Academic Press, New York.
Henry, D.O.
1973 The Natufian of Palestine: its material culture and
ecology. Ph.D. dissertation. Southern Methodist University,
University Microfilms, Ann Arbor.
1974
The utilization of the microburin technique in the Levant.
Paleorient 2(2) :389-395.
1977
An examination of the artifact variability in the Natufian
of Palestine. In Eretz-Israel (Vol. 13), edited by B.
Arensburg and 0. Bar-Yosef, pp. 229-239. Hebrew University,
Jerusalem.
1982
The prehistory of Southern Jordan and relationships with the
Levant. Journal of Field Archaeology 9(4):417-444.
1983
Adaptive evolution within the Epipaleolithic of the Near
East. In Advances in World Archaeology Volume 2, edited by
F. Wendorf and A. Close, pp. 99-160. Academic Press, New
York.
1986
The prehistory and paleoenvironments of Jordan: an overview.
Paleorient 12(2):5-26.
1987
Topographic influences on Epipaleolithic land use patterns
in Southern Jordan. In Studies in the History and
Archaeology of Jordan III, edited by A. Hadidi, pp. 21-27.
Department of Antiquities, Amman, Jordan and Routledge &
Kegan Paul, London.
1988
The Epipaleolithic sequence within the Ras en Naqb - el
Quweira area. Southern Jordan. Paleorient 14:245-256.
1989a Correlations between reduction strategies and settlement
patterns. In Alternative Approaches to Lithic Analysis,
edited by D.O. Henry and G.H. Odell, pp. 13 9-155.
Archeological Papers of the American Anthropological
Association Number 1.
1989b From Foraging to Agriculture: the Levant at the End of the
Ice Age. University of Pennsylvania Press, Philadelphia.
1994
Prehistoric cultural ecology in Southern Jordan.
265:336-341.
1995
Prehistoric Cultural Ecology and Evolution:
Southern Jordan. Plenum Press, New York.
Science
Insights from
173
HenrVj D.O., and A.N. Garrard
1988 Tor Hamar: an Epipaleolithic rockshelter in Southern
Jordan. Palestine Exploration Quarterly 120:1-25.
Horowitz, A.
1976 Late Quaternary paleoenvironments of prehistoric settlements
in the Avdat/Aqev area. In Prehistory and Paleoenvironments
in the Central Negev, Israel I The Avdat/Aqev Area, edited
by A.E. Marks, pp. 57-68. SMU Press, Dallas.
1979
The Quaternary of Israel.
1989
Prehistoric cultures of Israel: correlation with the oxygen
isotope scale. In Investigations in the Prehistory in South
Levantine Prehistory, edited by 0. Bar-Yosef and B.
Vandermeersch, pp. 5-18. BAR International Series 497,
Oxford.
Hours, F.
1974
1976
Academic Press, New York.
Remarques sur I'utilisation de listes types pour 1'etude du
Paleolithique superieur et de 1'Epipaleolithique du Levant.
Paleorient 2:3-18.
L'Epipaleolithiquie au Liban. Resultats acquis en 1975.
Preprint from Xleme Congres de I'Union Internationale des
Sciences Prehistoriques et Prothistoriques, pp. 106-131.
Nice.
Hovers, E., and 0. Marder
1991 Typo-chronology and absolute dating of the Kebaran Complex:
implications from the second season of excavations at Urkan
E-Rubb Ila. Journal of the Israel Prehistoric Society Mitekufat Haeven. 24:34-58.
Issar, A., and H. Bruins
1983 Special climatological conditions in the deserts of Sinai
and the Negev during the latest Pleistocene.
Palaeogeography, Paleoclimatology, Palaeecology 43:63-72.
Issar, A., et al.
1980 The character of the climate in the Negev and Sinai between
40,000 and 20,000 Years B.P. In Bat Sheva Seminar on
Approaches and Methods in Paleoclimatic Research, pp. 48-49
(abstract). Weizmann Institute, Rehovot.
Katsnelson, J.
1964 The variability of annual precipitation in Palestine.
Meteorological Bioki B13:163-172.
Arch.
Kaufman, A., Y. Yechieli, and M. Gardosh
1992 Reevaluation of the lake-sediment chronology in the Dead Sea
Basin, Israel, based on new ^^"Th/U dates. QuaternaryResearch 38:292-304.
Kelley, L.H.
1980 Experimental Determination of Stone Tool Uses, A Microwear
Analysis. University of Chicago Press, Chicago.
174
Kenyon. K.
1959
Earliest Jericho.
Antiquity 33:5-9.
Kirkbride, D.
1958 A Kebaran rockshelter in Wadi Madamagh, near Petra, Jordan.
Man 63:55-58.
1966
Kuhn, S.L.
1994
Five seasons at the Pre-Pottery Neolithic village of Beidha
in Jordan. Palestine Exploration Quarterly 98:8-72.
A formal approach to the design and assembly of mobile
toolkits. American Antiquity 59(3):426-442.
Lemmonnier, P.
1983 L'etude des systemes techniques, une urgence de technologie
culturelle. Techniques et Culture, nouvelle serie 1:11-34.
1990
Topsy turvy techniques: remarks on the social
representation of techniques. Archaeological Review from
Cambridge 9:27-37.
Leroi-Gourhan, A., and F. Darmon
1987 Analyses palynoloqiques de sites archeologiques de
Pleistocene Final dans la Vallee du Jourdain. Israel
Journal of Earth Sciences 36:65-72.
Lewarch, D., and M. O'Brien
1981 The expanding role of the surface assemblages in
archaeological research. In Advances in Archaeological
Method and Theory, Vol. 4, edited by M.3. Schiffer, pp. 297342. Academic Press, New York.
Lieberman, D.E.
1993 The rise and fall of seasonal mobility among huntergatherers: the case of the southern Levant. Current
Anthropology 34:599-631.
Lubbock, J.
1865
Pre-historic Times, as Illustrated by Ancient Remains, and
the Manners and Customs of Modem Savages. Williams and
Norgate, London.
MacDonald, B. (editor)
1988
The Wadi el-Hasa Survey, 1979-1983, West-central Jordan.
Wilfrid Laurier University Press, Waterloo, Ontario.
MacDonald, B., G. Rollefson, and D. Roller
1982 The Wadi el-Hasa Survey 1981: a preliminary report. Annual
of the Department of Antiquities of Jordan 26:117-132.
MacDonald, B., E. Banning, and L. Pavlish
1980 The Wadi el-Hasa Survey 1979: a preliminary report. Annual
of the Department of Antiquities of Jordan 24:169-184.
MacDonald, B., G. Rollefson, E. Banning, B. Byrd, and C. d'Annibale
1983 The Wadi el-Hasa Survey 1982: a preliminary report. Annual
of the Department of Antiquities of Jordan 27:311-323.
175
Macumber.. P.G.; and M.J. Head
1991 Implications of the Wadi al-Hammeh sectuences for the
terminal drying of Lake Lisan, Jordan. Palaeogeograpby,
Palaeoclimatology, Palaeoecology 84:163-173.
Magaritz, M.
1986 Environmental changes recorded in the Upper Pleistocene
along the desert boundary, southern Israel.
Palaeogeograpby, Palaeoclimatology, Palaeoecology 53:213219.
Marks, A.E. (editor)
1976 Prehistory and Paleoenvironments in the Central Negev,
Isreal. Volume I: The Advat Aqev Area, Part 1. SMU Press,
Dallas.
1977
Prehistory and Paleoenvironments in the Central Negev,
Isreal. Volume II: The Advat Aqev Area, Part 2 and the
Harif. SMU Press, Dallas.
198 3
Prehistory and Paleoenvironments in the Central Negev,
Isreal. Volume III: The Advat Aqev Area, Part 3. SMU
Press, Dallas.
Marks, A.E., and D.A. Friedel
1977 Prehistoric settlement patterns in the Avdat/Aqev area. In
Prehistory and Paleoenvironment in the Central Negev,
Israel, Vol. II, edited by A.E. Marks, pp. 191-232. Sr^U
Press, Dallas.
Marks, A.E., and P.A. Larson Jr.
1977 Test excavations at the Natufian site of Rosh Horesha. In
Prehistory and Paleoenvironment in the Central Negev,
Israel, Vol. Ill, edited by A.E. Marks, pp. 293-316. SMU
Press, Dallas.
Marks, A.E., and A.H. Simmons
1977 The Negev Kebaran of the Har Harif. In Prehistory and
Paleoenvironments in the Central Negev, Israel, Volume II:
The Advat Aqev Area, Part 2 and the Harif, pp. 233-270. SMU
Press, Dallas.
Marks, A.E., H. Crew, R. Ferring, and J. Phillips
1972 Prehistoric sites near Har Harif. Israel Exploration
Journal 22(2-3) :73-85.
Marks, A.E., and P. Volkman
1983 Changing core reduction strategies: a technological shift
from Middle to the Upper Paleolithic in the Southern Levant.
In The Mousterian Legacy. Human Biocultural Change in the
Upper Pleistocene, edited by E. Trinkaus. British
Archaeological Reports, Oxford.
Mellars, P.A.
1969 The chronology of Mousterian industries in the Perigord
region. Proceedings of the Prehistoric Society 35:134-171.
176
Moore.. A.M . T.
1985 The development of Neolithic societies in the Near East. In
Advances in World Archaeology, Vol. 4, edited by F. Wendorf
and A.E. Close, pp. 1-69. Academic Press, New York.
Mortensen, P.
1970 A preliminary study of the chipped stone industry from
Beidha. Acta Archaeologica 41:1-54.
1972
Seasonal camps and early villages in the Zagros. In Man,
Settlement and Urbanism, edited by P. Ucko, R. Tringham, and
G.W. Dimbleby, pp. 293-298. Duckworth and Co., London.
Muheisan, M.
1983 La Prehistoire en Jourdanie, Recherches sur
1'Epipaleolithique, L'example de gisement Kharaneh IV.
Unpublished Ph.D. dissertation, Universite de Bordeaux,
Bordeaux.
1985
L'Epipaleolithique dans le gisement de Kharaneh IV.
Paleorient 11(2):149-160.
1988
The Epipaleolithic phases of Kharaneh IV. In The Prehistory
of Jordan, edited by A.N. Garrard and H.G. Gebel, pp. 353367. BAR International Series 396, Oxford.
Neeley, M.P., and C.M. Barton
1994 A new approach to interpreting Late Pleistocene microlith
industries in Southwest Asia. Antiquity 68:275-288.
Neeley, M.P., J.D. Peterson, G.A. Clark, and S.K. Fish
1995 WHS 1065 (Tor al-Tareeq), an Epipaleolithic site in the Wadi
el-Hasa, West-central Jordan. Unpublished manuscript. On
file in the Department of Anthropology, Arizona State
University, Tempe.
Neev, 0., and Emery K.O.
1967 The Dead Sea: depositional processes and environments of
evaporites. Geologic Survey of Israel Bulletin 41:1-147.
Neuville, R.
1934 Le prehistorique de Palestine.
1951
Revue Bihlique 43:237.
Le Paleolitbique et le Mesolithique du desert de Judee.
Archives de 1'Institut de Paleontologie Humaine 24. Masson,
Paris.
Newcomer, M.H.
1974 Study and replication of bone points from Ksar Akil
(Lebanon). World Archaeology 6:138-153.
Nissen, K., and M. Dittemore.
1974 Ethnographic data and wear pattern analysis:
socketed Eskimo scrapers. Tebiwa 17:67-88.
a study of
Olszewski, D.I.
1989 Tool blank selection, debitage and cores from Abu Hureyra 1,
Northern Syria. Paleorient 15(2):29-37.
177
Olszewski. D.T., M. Stevens, M. Glass, P..F. Heck, J. Cccper and G.A.
Clark
1994 The 1993 excavations at Yutil al-Hasa (WHS 784), an
Upper/Epipaleolithic site in West-central Jordan.
Paleorieat 20:129-141.
Perrot, J.
1955
Le Paleolithique Superieur d'El-Quseir et de Masaraq an
Na'aj. Bulletin de la Societe Prehistorique Fraxicaise.
52(8):493-506.
1962
Palestine-Syria Celicia. In Courses Toward Urban Life,
edited by R.J. Braidwood and G.R. Willey, pp. 147-164.
Aldine, Chicago.
1966
Le gisement Natufian de Mallaha CEyan), Israel.
L'Anthropologie 70:437-484.
Phillips, J.L.
1991 Refitting, edge-wear and chaines operatoires: a case study
from Sinai. In 25 Ans D'Etudes Technologiques en
Prehistoirs. XI Rencontres Internationales d'Archeologie et
d'Histoire d'Antibes, pp. 303-317. Editions APDC, Juan-LesPins.
1994
The Upper Paleolithic chronology of the Levant and the Nile
Valley. In Late Quaternary Chronology and Paleoclimates of
the Eastern Mediterranean, edited by 0. Bar-Yosef and R.S.
Kra, pp. 169-176. Supplement to Radiocarbon, Department of
Geosciences, University of Arizona, Tucson, Arizona.
Phillips, J.L., and B.G. Gladfelter
1991 The refitting of lithics and site formation processes in the
Upper Paleolithic of Southern Sinai. Paper presented at the
annual meeting of the Society for American Archaeology, New
Orleans.
Phillips, J.L., and Mintz, E
1977 The Mushabian, In Investigations in Gebel Maghara, Northern
Sinai, pp. 149-183. Qedem 7. Monographs of the Institute
of Archaeology, Hebrew University, Jerusalem.
Phillips, J.L., and Yerkes
1979 Unpublished artifact master coding sheet (revised).
Rathje, W., and M.B. Schiffer
1982 Archaeology. Harcourt Brace Jovanovich, New York.
Rolland, N.
1977 New aspects of Middle Palaeolithic variability in western
Europe. Wature 266(5599):251-252.
1981
The interpretation of Middle Paleolithic variability.
16(1):15-42.
Rosenan, N.
1970 Climatic regions.
Atlas of Israel.
Sheet IV/3.
Man
178
Rossingol, M.
1969 CJne sequence climacique du Pleistocene dans la region de la
Mer Morte (Israel). Pollen and Spores XI(3):604-614.
Rust, A.
1950
Die hohlenfunde von Jabrud (Syrian).
Neumunster.
Schiffer, M.B.
1976 Behavioral Archaeology.
Karl Wachholtz Verlag,
Academic Press, New York.
Schuldenrein, J.
1987 Paleolithic archaeology in the Southern Levant: a
preliminary report of excavations at Middle, Upper, and
Epipaleolithic sites in Wadi El-Hasa, West-central Jordan.
Annual of the Department of Antiquities of Jrdan 31:68-70.
Schuldenrein, J., and G.A. Clark
1994 Landscape and prehistoric chronology of West-central Jordan.
Geoarchaeology 9(1):31-55.
Schuldenrein, J., and P. Goldberg
1981 Late quaternary paleoenvironments and prehistoric site
distributions in the lower Jordan Valley: a preliminary
report. Paleorienc 7(l):57-7l.
Schyle, D., and H.P. Uerpmann
1988 Palaeolithic sites in the Petra-area. In The Prehistory of
Jordan: The State of Research in 1986, edited by A.N.
Garrard and H.G. Gebel, pp. 39-65. BAR International Series
396, Oxford.
Seminov, S.A.
1964 Prehistoric Technology.
Adams and Mackay, Cory, England.
Shehadeh, N.
1985 The climate of Jordan in the past and present. In Studies
in the History and Archaeology of Jordan II, edited by A.
Hadidi, pp. 25-38. Department of Antiquities, Amman, Jordan
and Routledge & Kegan Paul, London.
Shott, M.
1986
Technological organization and settlement mobility: an
ethnographic examination. Journal of Anthropological
Research 42:15-51.
Stekelis, M., and G. Haas
1942 Preliminary report on soundings in prehistoric caves in
Palestine (Mugharet al-Watwat; Mugharet an-Nugtah; Mugharet
Wadi-Fallah; Mugharet Abu-Usba; Iraq el-Barud). Bulletin of
the Awercian Schools of Oriental Research 86:2-10.
1952
The Abu-Usba Cave. Israel Exploration Journal 13:1-12.
Stekelis, M., and T. Yizraeli
1963 Excavations at Nahal Oren: a preliminary report.
Exploration Journal 13:1-2.
Israel
179
Sullivan. A.P., and K.C. Rozen
1985 Debitage analysis and archaeological interpretation.
Awericem Antiquity 50(4):755-779
Tchernov, E.
1976 Some Late Pleistocene faunal remains from the Avdat/Aqev
area. In Prehistory and Paleoenvironments in the Central
Negev, Israel, Volume I: The Avdat/Aqev Area, edited by A.E.
Marks, pp. 69-74. SMU Press, Dallas.
Tixier, J.
1963
1974
Typologie de I'Epipaleolithic du Maghreb. Memoires du
Centre de Recherches Anthropologiques. Prehistoriques et
Ethnologiques, Alger, Paris.
Glossary for the description of stone tools with special
reference to the Epipaleolithic of the Maghreb. Newsletter
of Lithic Technology. Special Publication 1.
Turville-Petre, F.
1932 Excavations in the Mugharet el Kebarah. Journal of the
Royal Anthropological Institute 62:270-276.
Valla, F.R.
1988a Epipaleolithic syntheses.
Paleorient 14 (2) :316-320.
1988b La fin de 1'Epipaleolithique au Levant: Les industries a
microlithes geometriques. L'Anthropologie 92(3):901-925.
Vita-Finzi, C.
1964 Observations on the Late Quaternary of Jordan.
Exploration Quarterly 96:19-31.
Palestine
1966
The Hasa Formation:
1:387-390.
an alluvial deposition in Jordan.
1982
The prehistory and history of the Jordanian landscape. In
Studies in the History and Archaeology of Jordan, Vol. I,
edited by A. Hadidi, pp. 23-27.. Department of Antiquities,
Amman, Jordan.
Wendorf, F.
1968 The Prehistory of Nubia, 2 Volumes.
Man
SMU Press, Dallas.
Wiessner, P.
1983 Style and social information in Kalahari San projectile
points. American Antiquity 48:253-276.
Wobst, M.
1977
Stylistic behaviour and information exchange. In Papers for
the Director, edited by C. Cleland, pp. 317-342.
Anthropological Papers No. 61. University of Michigan
Museum of Anthropology, Ann Arbor.
Yechieli, Y., M. Magaritz, Y. Levy, U. Weber, U. Kafri, W. Woelfli, and
G. Bonani
1993 Late Quaternary geological history of the Dead Sea area,
Israel. Quaternary Research 39:59-74.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement