" Out of Arabia " and the Middle-Upper Palaeo- lithic transition in the southern Levant

Full text

*corresponding author
49
Quartär 61 (2014) : 49-85doi: 10.7485/QU61_03
“Out of Arabia” and the Middle-Upper Palaeo-
lithic transition in the southern Levant
„Out of Arabia“ und der Übergang vom Mittel- zum Jungpaläolithikum in der
Südlichen Levante
Jeffrey I. R1* & Anthony E. M2
1 Ronin Institute, Montclair, NJ, USA; e-mail: jeffrey.rose@ronininstitute.org
2 Southern Methodist University, Dallas, TX, USA; e-mail: amarks@mail.smu.edu
Abstract - Beginning some 50 thousand years ago, a technological transition spread across the Near East and into Eurasia, in
the most general terms characterized by a shift from preferential, prepared core reduction systems to the serial production of
elongated points via opposed platform cores. The earliest known occurrence of such a technological shift is the Emiran
Industry, whose oldest manifestations are found in the southern Levant. The cultural and demographic source(s) of this
industry, however, remain unresolved.
Looking to archaeogenetic research, the emerging picture indicates a major dispersal of our species out of Africa between
100 and 50 thousand years ago. Ancient DNA evidence points to low levels of admixt ure between Neander thal and pioneering
modern human populations in the Near East. These propositions underscore the significance of the Emiran and beg a
reassessment of its origins. In this paper, we ask whether the Emiran was a local development, a cultural/demographic
replacement, or the fusion of indigenous and exogenous lithic traditions. Our analysis considers the techno-typological
features of the Emiran in relation to late Middle Palaeolithic and contemporaneous assemblages from adjacent territories in
northeast Africa and the Arabian Peninsula, in order to identif y overlapping cultural features and potential antecedents. Parsi-
monious with the archaeogenetic scenario of admixture, the Emiran seems to represent a fusion of local southern Levantine
Mousterian tool types with the Afro-Arabian Nubian Levallois reduction strategy. We propose that Emiran technology is
primarily rooted in the Early Nubian Complex of the Nile Valley, which spread onto the Arabian Peninsula during the Last
Interglacial and developed at the interface of northern Arabia and the southern Levant between 100 and 50 thousand years
ago.
Zusammenfassung - Vor etwa 50.000 Jahren begann ein technologischer Wandel, welcher zuerst im Nahen Osten und folgend
auch in Eurasien greifbar wird. Diese Veränderung wird im Allgemeinen gekennzeichnet durch das Ersetzen von präparierten,
formbestimmenden Abschlagskernen durch bidirektionale Kerne, die der seriellen Herstellung von langgestreckten Spitzen
dienten. Das früheste bekannte Auftreten eines solchen technologischen Wandels ist das Emiran, dessen älteste Erscheinungs-
formen in der südlichen Levante zu finden sind. Der kulturelle und demographische Ursprung dieser Industrie bleibt jedoch
ungeklärt.
Jüngsten archäogenetischen Forschungen zufolge fand die Ausbreitung anatomisch moderner Menschen zwischen
100 und 50 tausend Jahren vor Heute statt. Nachweise durch aDNA deuten auf eine geringfügige Vermischung von Neander-
talern und Pionieren der anatomisch modernen Menschen im Nahen Osten hin. Diese Tatsachen unterstreichen die Bedeutung des
Emiran und verlangen eine Neubewertung dessen Herkunft. In diesem Aufsatz gehen wir den Fragen nach, ob das Emiran eine
lokale Entwicklung, einen kulturell-demografischen Wechsel oder die Verschmelzung von indigen und exogen lithischen Tradi-
tionen darstellt. Die techno- typologischen Merkmale des Emiran werden mit den spätmittelpaläolithischen Industrien der Levante
und den benachbarten Gebieten im Nordosten Afrikas und der Arabischen Halbinsel verglichen um überlappende kulturelle
Merkmale zu erfassen. Des Weiteren sollen, mittels der vorgelegten Analysen, mögliche technologische Vorgänger des Emiran
festgelegt werden. Entsprechend geringer genetischer Vermischung, scheint das Emiran eine Verschmelzung von typologischen
Elementen des südlevantinischen Mousterian und der afro-arabisch, nubischen Levallois Abbautechnik darzustellen. In diesem
Sinne sind die technologischen Wurzeln des Emiran vor allem in dem frühen nubischen Komplexen des Niltals zu suchen, welches
seit dem letzten Interglazial auch in der Arabischen Halbinsel verbreitetet war. Zusammenfassend liegt der Ursprung des Emiran
demnach an der Schnittstelle von Nord-Arabien und der südlichen Levante zwischen 100 und 50 tausend Jahren vor Heute.
Key words - Out of Africa; Out of Arabia; Emiran Industry; Nubian Complex; Middle-Upper Palaeolithic
transition; modern human dispersal
Out of Africa; Out of Arabia; Emiran; Nubischer Komplex; Übergang Mittel- zu Spätpaläolithikum;
Ausbreitung moderner Mensch

Quartär 61 (2014) J. I. Rose & A. E. Marks
50
Introduction: through a prism of paradigms
The following paper builds upon the “Out of Arabia”
modern human expansion scenario proposed by
Marks and Rose (2014). Our initial publication
reviewed archaeological and genetic data to posit an
origin of the Upper Palaeolithic in the southern Levant.
From this synthesis, we concluded that there was some
degree of cultural, hence demographic, input from
populations in the Arabian Peninsula during early
Marine Isotope Stage 3 (MIS 3). The work presented
here expands and refines our hypothesis by providing
a quantitative and qualitative description of the
proposed archaeological scenario using lithic techno-
typological patterning across northeast Africa, Arabia,
and the southern Levant.
As the only extant land bridge out of Africa, the
“Levantine Corridor” is often presumed to have been
a primary conduit of demographic exchange between
Africa and Eurasia throughout the Quaternary. The
transition from the Middle (MP) to Upper Palaeolithic
(UP) in the Levant, which occurred during early MIS 3
around 50 thousand years ago (ka), has been ex plained
by some as an influx of African groups through the
Levantine Corridor bearing early UP cultural features
(e.g., Bar-Yosef 1987; Tostevin 2000; Meignen &
Bar-Yosef 2005; Douka et al. 2013). While it is clear
that both anatomically modern humans (AMHs) and
Neanderthals were present in the Levant prior to the
UP (e.g., Stringer & Andrews 1988; Stringer 1994;
Hublin 2000), the taxonomy of toolmakers and
chronology of occupations (e.g., Shea 2007; Hovers &
Belfer-Cohen 2013) are far from resolved. Due to this
ambiguity, the archaeological record of the Levant
tends to be tied to the prevailing paradigm of modern
human evolution. The region serves as a prism through
which to view these paradigms, guiding and framing
scholars’ views of the biological and behavioral
emergence of our species.
Most of the Initial UP assemblages found in the
vast territory stretching from central Europe to south-
western Asia and to northern Asia are recognized by a
similar stage of technological development involving
the serial production of elongated points via opposed
platform, bidirectional core reduction systems (Kuhn
& Zwyns 2014). The earliest known manifestation of
this technological transition outside of Africa - the
Emiran Industr y - appeared some 50 ka in the southern
Levant and subsequently spread northward (Leder
2013). Given the co-association of anatomically
modern human remains (AMHs) with an Emiran assem-
blage at Ksar Akil in Lebanon, researchers often link
the appearance of this technology during the Initial
UP to the migration of AMHs out of Africa (e.g., Douka
et al. 2013). Yet, despite its significance, the ultimate
origin(s) of the Emiran Industry still remain unknown.
Broadly speaking, there are three possibilities to
explain the appearance of the Emiran: 1) it arose from
an exclusively local technological base, manufactured
by indigenous toolmakers (e.g., Ewing 1947; Garrod
1951; Copeland 1975; Azoury 1986; Bar-Yosef &
Belfer-Cohen 1988; Ohnuma 1988; Demidenko & Usik
1993; Marks 2003); 2) it developed outside of the
Levant from a non-Levantine technological base and
was brought by foreign populations moving into the
region (e.g., Mellars 1996; Tostevin 2000; Meignen &
Bar-Yosef 2005; Shea 2008); or 3) its development was
the fusion of a local tradition influenced by external
stimuli from one or more adjacent regions (e.g.,
Van Peer & Vermeersch 2007; Van Peer et al.
2010; Meignen 2012; Marks & Rose 2014). The
three paradigms of modern human evolution
that accompany these interpretations of the
MP-UP transition in the Levant are, respectively, a
local development from archaic forms to modern, the
total replacement of archaic populations by incoming
modern human groups, and replacement with some
admixture between these two species. Although the
possibility of admixture was considered in the past
(e.g., Ahrensburg & Belfer-Cohen 1998; Hawks &
Wolpoff 2001), only recently has enough empirical
evidence emerged to support genetic exchange
between subspecies. Ancient DNA extracted from
Neanderthal and Denisovan remains in Europe and
Asia indicate that modern humans interbred, to some
extent, with these populations, as well as with archaic
groups within sub-Saharan Africa as recently as ca.
20 ka (Durand et al. 2011; Hammer et al. 2011; Reich et
al. 2011; Skoglund & Jakobsson 2011; Alves et al. 2012;
Meyer et al. 2012; Neves & Serva 2012; Sankararaman
et al. 2012, 2014; Fu et al. 2014; Vernot & Akey 2014).
Some researchers have proposed that the initial locus
of AMH-Neanderthal admixture was in the Near East,
inferred from the distribution of shared Neanderthal
markers among all modern Eurasian populations
(Green et al. 2010; Yotova et al. 2011; Sanchez-Quinto
et al. 2012). This has significant implications for under-
standing the Levantine archaeological record, making
total replacement of local populations by African
emigrants unlikely.
Only recently have we been able to reconsider
early human occupation in the southern Levant in
respect to the adjacent Arabian Peninsula. The
posited significance of the southern dispersal route
out of Africa (e.g., Lahr & Foley 1994; Quintana-Murci
et al. 1999; Kivisild et al. 2004; Metspalu et al. 2004;
Forster & Matsumura 2005; Macaulay et al. 2005;
Ghirotto & Barbujani 2011) served to invigorate
archaeological fieldwork programs throughout Arabia
over the past decade. In stark contrast to the “coasting
out of Africa” model (Stringer 2000; Mellars 2006;
Oppenheimer 2009; Mellars et al. 2013), all of these
new findings in Arabia unanimously suggest that
demographic movements into and out of the Peninsula
were not associated with occupation of the littoral
zone, and were far more complex than previously
considered (e.g., Rose 2006, 2007; Bailey 2009;
Crassard 2009; Rose & Petraglia 2009; Rose & Usik

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
51
2009; Armitage et al. 2011; Rose et al. 2011; Delagnes
et al. 2012; Petraglia et al. 2011, 2012; Usik et al. 2013).
Reviewing these new data, we revisit the origins of
the southern Levantine Emiran Industry, sensu stricto.
As we have previously suggested (Marks & Rose 2014),
the technological trajectories of Arabia and the
southern Levant appear to have been long inter-
twined. The two regions encompass a single but
varied physiographic landmass belonging to the
Saharo-Arabian phytogeographic zone (Al Nafie
2008); logically, discussions of potential population
movements to and from the southern Levant should
include Arabia, as well as Africa.
Background
Recognizing and defining the Emiran
The Emiran Industry was formally defined by Garrod
(1951) and named for the type assemblage excavated
at Mugharet el-Emireh (Garrod 1955). Such assem-
blages bore a mix of both MP and UP diagnostic
features, including MP tools made on classic Levallois
blanks, typical UP tools made on UP blade blanks, and
Emireh points - Levallois points with distinct bifacial
basal thinning (Garrod 1951: 128). Garrod (ibid.: 129)
emphasized that at el-Wad there was “an absence of
forms transitional from one to the other” and
attributed the presence of a prismatic blade
technology to the “invention of a new technique within
an old tradition.” This was the initial definition of the
Emiran. Because of the Emiran’s stratigraphic position
within clearly defined MP-UP sequences, as well as the
presence of both MP and UP tools and blanks, Garrod
(ibid.) interpreted the industry as being both “truly
intermediate” and “truly transitional” between the MP
and U P.
By the 1970s, Garrod’s definition had been
changed: rather than a simple co-association of MP
tools on Levallois blanks and UP tools on UP blade
blanks, the assemblages were described as essentially
consisting of UP tools (e.g., burins, end scrapers and
chamfered pieces) made on elongated Levallois blanks
(Azoury 1986), as well as elongated points from
unipolar cores, which “differ little from those of the
Levalloiso-Mousterian levels below” (Copeland 1975:
337). Aside from the north/south dichotomy in type
fossils, with chamfered pieces in the north (Newcomer
1970) and Emireh points largely in the south (Copeland
2001; Volkman & Kaufman 1983), the Emiran was
conceived as static, with no technological or
typological development through time.
Our understanding of the Emiran developed
considerably in the early 1980s, with the discovery of
the stratified site of Boker Tachtit (Fig. 2: 1). The site’s
state of preservation allowed for large-scale core
reconstructions and detailed descriptions of techno-
logical changes over four consecutive periods
beginning around 50 ka (Marks 1983b). These changes
manifested in a shift from Level 1 at the bottom of the
sequence, which exhibits a standardized, hard hammer
bidirectional Levallois point and blade reduction
strategy utilizing extensive cresting in initial core
shaping (Fig. 1), to Level 2, which shows a co-association
between the bidirectional Levallois point production
system and hard hammer volumetric blade core
reduction, primarily bidirectional with occasional
unidirectional flaking, to Level 3, with a marked shift
away from bidirectional Levallois point cores and to
an increase in unidirectional reduction. Finally, in Level
4, the assemblage is dominated by a hard hammer
volumetric blade strategy, mainly unidirectional, but
with some bidirectional reduction, as well (Volkman
1983). The sequence begins with Levallois point and
blade reduction a nd ends with an entirely no n-Levallois
blade strategy; yet, this change in technology had very
little impact on the morphology of the blanks
produced, and virtually no effect on the tool types,
which throughout were dominated by points and
various types of burins and end scrapers (Marks &
Kaufman 1983).
The technological developments documented at
Boker Tachtit (Volkman 1983, 1989) raised the
question as to when the “transitional” Emiran had
reached a stage it could be perceived as being fully
UP. In this case, the disappearance of Levallois
reduction in Level 4 was the criterion by which the
Initial UP at Boker Tachtit was recognized (Marks &
Ferring 1988). In combining the Üçağizli sequence
with that at Boker Tachtit, Kuhn et al. (2009) lost the
taxonomic distinction between Emiran (Boker Tachtit,
Levels 1 - 3) and Initial UP (Boker Tachtit, Level 4).
Most recently, scholars working on this topic (i.e.,
Leder 2013; Kuhn & Zwyns 2014; Marks & Rose 2014)
have reasserted a strict definition of the industry by
limiting the Emiran to those assemblages technolo-
gically and typologically comparable to Boker Tachtit,
Levels 1 - 3. We consider only those levels from Boker
Tachtit as Emiran.
Leder’s (2013) study of MP-UP transitional assem-
blages in the Levant recognizes two distinct industries:
the Emiran, occurring primarily in Lower Galilee, and a
new industry called “Bokerian,” which Leder (ibid.:
162) describes as “a specific industry within the Levant
that emerged in the Negev desert and later spread
into Lebanon and western Jordan.” The author divides
the Bokerian into seven chronological phases, with
its initial stage linked to Boker Tachtit, Level 1, and
final phase associated with the Üçağizli sequence.
Although his terminology differs from that used in this
paper, we fully agree with Leder’s recognition of a
transitional industry that emerged in the Negev and
subsequently spread northward.
Methods
Evaluating the origins of the Emiran
Part of the problem lies in what actually constitutes an
“origin” or “transition.” Origins are either inherently

Quartär 61 (2014) J. I. Rose & A. E. Marks
52
difficult or extremely easy to recognize and define.
They are easily recognized when they represent an
obvious in situ transformation from one state to
another in a stratified context, (e.g., the passage from
Emiran to Initial UP at Boker Tachtit), or the relatively
sudden appearance of something so different from
what came before that no serious case could be made
for autochthonous developmental change (e.g., the
appearance of the Aurignacian in the Levant). These
situations are extremes, however, and most transitions
tend to be far less obvious and more difficult to define.
Finding the root of a lithic industry first involves
differentiating between continuity and discontinuity.
In this case: does the Emiran industry represent
something new to the region, or is it merely a late
manifestation of the Levantine MP? The detailed
technology of core formation and blank production,
as described in Copeland (1975), is essentially the
same as that in the local MP. Yet, if this same pattern
can be shown to also occur in Africa around that time,
Fig. 1. Early Emiran artifacts from the southern Levant: Levallois points from Boker Tachtit, level 1 (a,d,e-h,l-n) and ‘Ain Difla, levels 1-5
(b,c,o,p); Emireh points from Boker Tachtit, level 1 (i-k); crested blades from Boker Tachtit (s,t) and ‘Ain Difla (q); burins from Boker Tachtit,
level 1 (u,v); endscraper on crested blade from Boker Tachtit, level 1 (r); Levallois point cores from Boker Tachtit, level 1 (w-y,aa-ad) and ‘Ain
Difla (z,ae). Illustrations after Marks and Kaufman (1983: Figs. 5.2, 5.6-5.9, 5.11, 5.18); Demidenko and Usik (1993: Figs. 2-7).
Abb. 1. Steinartefakte des frühen Emiran aus der Südlichen Levante: Levalloisspitzen aus Boker Tachtit, Level 1 (a,d,e-h,l-n) und Ain Difla, Level
1-5 (b,c,o,p); Emireh- Spitzen aus Boker Tachtit, Level 1 (i-k); Kernkantenklingen aus Boker Tachtit (s,t) und Ain Difla (q); Stichel aus Boker Tachtit,
Level 1 (u, v); Kratzer an Kernkantenklinge aus Boker Tachtit, Level 1 (r); Levalloisspitzenkerne aus Boker Tachtit, Level 1 (w-y, aa-ad) und Ain
Difla (z, ae). Zeichnungen nach Marks und Kaufman (1983: Abbildungen 5.2, 5.6-5.9, 5.11, 5.18); Demidenko und Usik (1993: Abbildungen 2-7).

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
53
then an African origin for the Emiran is also a
reasonable option. The same may be said for Emiran
typology: are the Emiran type fossils (i.e., the Emireh
point and the chamfered piece) associated with t ypical
UP tools found somewhere in Africa, either with or
without Levallois technology? What of the predomi-
nance of UP tools made on Levallois blanks, does such
a pattern exist in the Levant outside of the Emiran?
Does it exist in Africa, at all?
In this paper, we systematically consider from
whence came the Emiran, in light of techno-ty pological
attributes found in contemporary and preceding
industries from 1) the southern Levant, 2) northeast
Africa, and 3) the Arabian Peninsula. On a very broad
level, Emiran technology (i.e., hard hammer removals,
presence of Levallois per se, etc.) is not very useful for
tracking its origin because these attributes are very
widespread and may represent technological conver-
gence from disparate cultural bases (Kuhn et al. 2009).
Within the Levallois method, however, there are
different reduction strategies that have limited distri-
butions in time and space. Since virtually all MP indus-
tries have some retouched tools, the composition of
tool assemblages and the presence or absence of
specific tool types are valid observations for
comparison when shown to document regional, time-
transgressive patterns.
Since Bordes (1950, 1953) introduced the “total
tool assemblage” as the unit of comparison, it has
become the norm. For this paper, however, we are
concerned with typology mainly at the class level and
across a huge geographic spread. Those factors that
influence total tool assemblages: e.g., site function,
intensity of occupation, availability of raw material,
cave versus open-air sites, etc. simply cannot be
controlled for and add noise to the broader trends of
interest here. In order to remove as much noise as
possible in our study, we exclude most central and
northern Levantine sites from our analysis. This area is
not part of the contact zone between posited popula-
tions expanding from Arabia and/or Africa. Moreover,
most, but not all (Zaidner et al. 2014), of the published
MP sites in the northern Levant are found in karstic
cavities, while in the whole of the southern Levant,
Arabia and the Nile Valley, there is only one thinly
deposited cave site (Sodmein Cave in the Red Sea
Hills) and three shallow rock shelters (Tor Sabiha and
Tor Faraj in Jordan and Jebel Faya in eastern Arabia).
By excluding most Levantine cave sites, we also avoid
the fundamental differences between cave and
open-air occupations. Only some of those central and
northern sites with Tabun-C like assemblages are
included, since they do not occur in the south and
their association with AMH fossils is clearly relevant
here.
Characterizing Emiran technology and typology
In the Emiran, there were a few variations in how
blocks of raw material were shaped into cores,
particularly the initial formation of ridges on both the
major core surfaces and at the core extremities prior
to platform formation (Volkman 1989: Fig. 6-3). The
ridges formed on the major flaking surface and the
two extremities were then struck off, resulting in
classic crested blades, while the ridge formed on the
bottom of the core was left intact throughout subse-
quent core reduction (Volkman 1989: Fig. 6-6). After
the ridge formation on the major flaking surface and
the extremities, the latter were flaked to create
striking platforms, during which a small crested blade
was produced from each end. One platform was used
to strike off the ridge along the major flaking surface,
resulting in a long crested blade (Volkman 1989: Fig.
6-6). The major flaking surface was cleared of what
cortex remained, and then blades were struck from
both platforms to shape the flaking surface, platforms
being rejuvenated as needed and, in the case of
Levallois point production, to establish the appro-
priate Y-arête pattern. In the case of point cores, three
removals were required to set up the point removal.
First, a flake was struck from one end of the core, then
two elongated blanks were struck from the other end,
forming a central arête. Finally, a point was struck from
the first platform.
A large number of reconstructions show variations
on this theme: the number of blanks struck, the
number of platform rejuvenations, other means of
stripping cortex, etc. (ibid.). Yet, the basic two patterns
remain the same for all Levallois cores, whether point,
flake, or blade: bidirectional removals and the initial
use of constructed crests to set up core surfaces, which
comprised the majority of the Level 1 core assemblage
(Fig. 3). Hence, while the predominant use of bidirec-
tional flaking may be a signature characteristic in
tracking the source of the Emiran, the co-association
of these two distinct reduction strategies is also
significant.
The lack of extensive core reconstructions at other
southern Levantine MP sites means that technological
similarities and differences with Boker Tachtit, Level 1,
can only be addressed via the products of its signature
bidirectional reduction strategy. Bidirectional core
reduction, regardless of the specific reduction
strategy, is monitored both through the scar patterns
on the primary working surface of abandoned cores,
and by dorsal scar patterns on byproducts and
endproducts of core reduction. Two kinds of bidirec-
tional scar patterns exist, only one of which indicates
true bidirectional reduction. In the most common
case, a few short scars originate from one end of the
core or blank, while most of the scars were struck from
the other end. In these cases, the platform with a few
scars is considered secondary and the result of core
maintenance when convexities need to be re-estab-
lished (Monigal 2002: 162-165; Mustafa and Clark
2007: 65-68). On the other hand, when dorsal scar
patterns show removals originating from both distal
and proximal platforms that meet close to the

Quartär 61 (2014) J. I. Rose & A. E. Marks
54
mid-point of the flaking surface of the core (or on the
dorsal surface of the blank), then true bidirectional
reduction may be inferred. For our analysis, we
measure the frequency of bidirectional reduction
within each assemblage using an index of bidirection-
ality (IBi), calculated by the number of cores exhibiting
at least one working surface with some form of bidirec-
tional flaking strategy (i.e., preferential or recurrent
opposed platform Levallois, Type 1 Nubian Levallois,
or bidirectional blade cores).
In terms of typology, Emiran assemblages are
characterized by a high percentage of UP tools and
Levallois points (Fig. 3). The predominance of UP
retouched tools alongside Levallois points, however,
does not clearly distinguish the Emiran from Levantine
Mousterian sites. In the south, MP assemblages at Rosh
Ein Mor (Fig. 2: 12), Nahal Aqev (Fig. 2: 7), and ‘Ain
Difla (Fig. 2: 6) all have significant numbers of both
Levallois points and UP type tools. In particular, burins
and end scrapers were found manufactured on
elongated Levallois blanks, in conjunction with MP
tools that were well in the minority (Marks & Crew
1972; Marks & Kaufman 1983; Munday 1976, 1977;
Clark et al. 1997). Even at the late MP sites of Tor Faraj,
top C (Fig. 2: 2), and WHS621 (Fig. 2: 3), UP tools are
more numerous than MP variants (Henry 1995, 1998).
Thus, the dominance of UP tools over MP tools is
consistent across the southern Levant, regardless of
the type of Mousterian industry that is present.
Another important typological aspect of the
Emiran is the percentage of Levallois points, including
Emireh points (Levallois points with bifacial basal
thinning), relative to the number of traditionally
diagnostic MP (Bordes 1961) and UP tools (e.g., de
Sonneville-Bordes & Perrot 1954). In Boker Tachtit
Level 1, the high proportional occurrence is consistent
with Mousterian sites in the southern Levant,
regardless of age (Fig. 3). Thus, Levallois point
production is time-transgressive, typical of both
Emiran and MP assemblages in the region. This trait is
also characteristic of Tabun type D and B assemblages
in the northern Levant (Meignen & Bar-Yosef 2005)
and, as such, ty pifies much of the Levantine Mous terian.
One additional typological characteristic of the
Emiran, particularly in Boker Tachtit Levels 1 and 2, is
right lateral or bilateral nibbling to steep retouch
adjacent to the platforms of Levallois points. On some
it is minor, but the number of points and blade/points
with this modification is striking: 65 % in Level 1 and
59.5 % in Level 2 (Marks & Kaufman 1983). Although
Fig. 2 . Map of Middle P alaeolithic/M iddle Stone Age appr oximate site locat ions included in this s tudy. Sites correspon d
to the reference numbers listed in Figures 3, 8 & 11. DEM courtesy of Yamandú Hilbert.
Abb. 2. Karte zur Lage der in dieser Studie einbezogenen Fundstellen. Die Nummerierung der Fundstellen entsprechen
denen in Abb. 3, 8 & 11. Das digitale Höhenmodell wurde von Y.H.Hilbert erstellt.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
55
this trait is difficult to compare with other assemblages
in the region because it was not consistently recog-
nized, there are data to suggest that it is not time-
transgressive, but only common in Late Mousterian
assemblages, regardless of the Tabun industry type.
Through these signature Emiran characteristics
(i.e., elongated point production via a bidirectional
Levallois reduction strategy, the use of lateral cresting
in core preparation, the significant presence of UP
tool types, right lateral/bilateral nibbling at the base
of Levallois points from hafting, and the production of
Emireh points), we examine lithic techno-typological
patterning in the southern Levant and surrounding
areas from the time period between approximately
130 - 50 ka. In theory, the Emiran should demonstrate
significant overlap, in both core reduction strategies
and tool types, with the industry from which it arose.
Archaeological Evidence
The southern Levant
The Tabun problem
Determining whether or not the Emiran is rooted in
the southern Levantine MP, or comes from an
exogenous source, depends largely upon how one
interprets the geographic and temporal relationship
of Levantine Mousterian industries. Logic suggests
that any argument for or against continuity should
consider the youngest Mousterian and the oldest
Emiran assemblages (Tostevin 2003). This logic,
however, necessitates acceptance of the view that the
MP sequence from Mount Carmel (e.g., et-Tabun and
Kebara) is pan-Levantine, both chronologically and
techno-typologically (e.g., Howell 1959; Bar-Yosef
1980, 1998; Tostevin 2003).
Until the early 1980s, such a view was reasonable,
since little was known of MP assemblage types outside
of Mount Carmel and Lebanon. Taking into account
the sizable body of evidence amassed over the last
three decades, however, numerous scholars have
forcefully argued that the Tabun assemblage type
definitions are not representative of the technological
complexity found across the whole of the Levant (e.g.,
Hovers 1998; Goren-Inbar & Belfer-Cohen 1998;
Hovers 2009; Hauck 2011; Meignen 2011; Leder 2013;
Marks & Rose 2014). As such, our use of the Tabun
terminology should be understood as a necessary evil,
in order to engage with pre-existing literature, and
not a reflection of the actual range of Levantine
Mousterian variability. At this point, there is little
doubt that the MP Tabun industries are neither chrono -
logically linear nor pan-Levantine.
In particular, the southern Levantine sequence
indicates that some form of Tabun D-like Mousterian
lasted longer in the arid margins than it did in the
Mediterranean zone, where there was a major
Assemblage Reference # Pts, MP, & UP
tools (n)
% Levallois
points
% M P
tools
% U P
tools
Blanks
(n) ILam Cores (n) IBi
Emiran
Boker Tachtit, level 1 1a 63 70 233 4274 40 58 91
Boker Tachtit, level 2 1b 155 42 356 9432 29 198 84
Boker Tachtit, level 3 1c 15 87 013 2147 35 34 71
Late Levantine Mousterian
Tor Faraj, top C 2131 78 616 372 26 41 25
WHS 621 3279 36 49 15 4005 19 71 N/A
J444 485 95 4 1 1468 38 30 N/A
Tor Sahbiha 559 90 7 3 1867 37 10 N/A
Ain Dia, all levels 6336 83 516 7228 45 180 30
Nahal Aqev, level 3a-b 7a 55 56 18 26 1938 26 76 11
Nahal Aqev, level 3c-e 7b 200 64 16 20 3531 28 197 6
D40 12 23 70 922 504 29 140 36
Early/Middle Levantine Mousterian
Skhul, B1 81304 787 62564 N/A 1127 N/A
Ksar Akil, level X XVI A 9a 85 082 18 171 24 56 13
Ksar Akil, level X XVI B 9b 73 10 81 10 306 20 90 19
Ksar Akil, level X XVII A 9c 54 069 32 227 25 116 17
Ksar Akil, level X XVII B 9d 125 14 61 26 546 26 110 14
Tabun, Bed 39 10 105 19 71 10 194 29 26 N/A
Qafzeh, levels XXII-XVII 11 107 840 53 1877 967 12
Rosh Ein Mor, all levels 13 2795 64 828 41574 29 1153 11
Fig. 3. Composite data for Levantine sites included in the study. Site reference number corresponds to the map number on Figure 2 as well
as the plot numbers on Figures 13 & 14.
Abb. 3. Au fgeführt sind d ie technischen Daten der Levantin ischen Fundstellen, die in der h iesigen Studie be handelt werden. Refer enznummer der
Fundstellen stimmen mit der Karte in Abbildung 2 und dem Diagramm in Abb. 13 & 14 überein.

Quartär 61 (2014) J. I. Rose & A. E. Marks
56
technological shift seen in the appearance of Tabun
C-type assemblages. The chronology of the southern
Levant raises the possibility that some form of Tabun
D-like industry persisted in the Negev after 80 ka,
thus, may be considered as a potential candidate for
the cultural source of the Emiran.
The Early Levantine Mousterian
In the northern Levant, Tabun D assemblages are
roughly bracketed between 270 and 150 ka. These
ages vary widely, depending upon which dating
technique is used. In the case of TL, the range is
270 - 170 ka, while ESR produces results between 200
and 150 ka (Bar-Yosef 1998). In the same caves, TL and
ESR measurements from Tabun C assemblages range
from 170 ka to 85 ka (ibid.). Given the average TL
(92 ± 5 ka) and ESR (96 ± 13 ka) ages on the Tabun C
materials from Qafzeh (Fig. 2: 11), Tabun C is likely to
be closer to 100 ka than to 170ka.
Dates from the southern Levant are much less
consistent. The Tabun D site of Rosh Ein Mor produced
a Th/U date of 200 ka (Rink et. al. 2003), corre-
sponding with comparable assemblages in the north.
The Tabun D-type assemblage at Nahal Aqev,
however, is younger than ~80 ka, based on two Th/U
dates (85.2 ± 10 ka and 74 ± 5 ka) from a travertine
directly underlying an elongated Levallois point
embedded in the adjacent fossil spring (Schwarcz et
al. 1979). While additional dates are desirable, there is
presently no good reason to reject them (contra Shea
2003).
Within the Early Mousterian assemblage from Rosh
Ein Mor in the central Negev, there are trivial amounts
of bidirectional reduction (Fig. 4). Among the Levallois
cores, 3 % show bidirectional removals on the major
flaking surface (Munday 1976: Table 6-14), while 2 % of
blades have comparable scars (Monigal 2002:
Appendix D). Levallois points exhibit a similar pattern,
with less than 3 % struck from opposed platform cores
(Crew 1976: 83; Henry 1992: Table 11.2). Flat, opposed
platform cores of all types account for 13 %, most of
which were small, heavily reduced, and close to
exhaustion (Monigal 2002: 352). In terms of the
Levallois points with limited retouch adjacent to the
butt - one of the key features of the Emiran - only 8 %
of Levallois points at Rosh Ein Mor had any retouch
and none was similar to those from Boker Tachtit,
(Crew 1976: 100; Monigal 2002).
An undatable surface workshop, the assemblage
from D40 was initially ascribed to the Early Levantine
Mousterian, yet exhibits some degree of bidirectional
reduction (Fig. 3). Although the two sites are in close
proximity, D40 differs from Rosh Ein Mor in that
regard, and might suggest a later date closer to ‘Ain
Difla. Unlike the triangular/sub-triangular point cores
found at ‘Ain Difla, however, the D40 specimens are
mainly squat and ovate (Monigal 2002: 382).
In the Sinai Peninsula, surveys in the north located
a single MP surface scatter with centripetal Levallois
reduction, which may be one of the few Tabun C-type
sites outside the Mediterranean zone (Gilead 1985).
Two assemblages in secondary but stratified position
were excavated at the Split Rock site (Fig. 2: 54) in
central Sinai (Kobusiewicz 1999); OSL ages indicate
the lower horizon falls within MIS 5.3 (~100 ka) and the
upper around MIS 5.1 (~75 ka) (Kobusiewicz et al.
2001). Technologically, the two assemblages are
seemingly distinct; the lower is characterized by a
combination of preferential Levallois a nd non-Levallois
reduction strategies, mainly discoidal, single platform,
and multiplatform. Bidirectional cores occur, but are
rare (9 %). Blade production is minor and comes from
single platform cores. Although no Levallois point
cores were found, four points were present. From the
cores, the upper horizon is dominated by preferential
Levallois flake reduction, accompanied by single and
multiplatform reduction. Again, bidirectional cores
are uncommon (8 %) and only two blades come from
such cores. Typologically, the lower horizon has few
retouched tools aside from notches, denticulates and
simple retouched pieces, insufficient tools to warrant
discussion. It is clear, however, that Levallois point
production was rare (one point and one point core),
while Levallois flake production was considerably
more frequent. MP tools far exceed UP types, with no
fewer than ten sub-types of side scraper (Kobusiewicz
1999: 197).
The Late Levantine Mousterian
In the Negev, the chronologically late Levantine
Mousterian site of Nahal Aqev, dated to ca. 80 ka,
shows a similar technological pattern to the nearby
Tabun D-like assemblage from Rosh Ein Mor (Figs. 3,
5). On the southern Jordanian plateau, over 50
Levantine Mousterian surface scatters and a handful
of rockshelters were found, including the “Tabun B”
assemblages from Tor Sabiha (Fig. 2: 5) and Tor Faraj
(Potter 1995; Henry 1995, 1997, 1998, 2003). Three
TL dates from Tor Faraj, top C range from 43.8 ± 2 ka
to 52.8 ± 3 ka, averaging 48 ± 2.7 ka (Henry 1998),
making it contemporary with the youngest of the
Tabun B assemblages at Kebara Unit VI, dated to
48 ± 3.5 ka (ibid.). AAR measurements on Tor Foraj,
top C and Tor Sabiha, top C provided matching ages
of 69 ± 6 ka (Henry 1995). While correlating TL and
AAR dates is difficult, the AAR dates indicate that the
occupations at Tor Faraj, top C and Tor Sabiha, top C
were essentially contemporary, and based on the TL
dates, were coeval with Boker Tachtit, Level 1.
The ages of the ‘Ain Difla assemblages are open to
interpretation. Three ESR dates from Test A, levels 19
and 20, produced early uptake readings between
88.3 ka and 112.5 ka, averaging 98.8 ± 12.7 k a ,
while late uptake for the same samples gave a range of
142.8 ka to 185.8 ka, with an average date of 161.1 ±
28.2 ka (Clark et al. 1997). One ESR date from Level 12
gave an early uptake date of 114.5 ± 14.2 ka,
while under the assumption of late uptake it was

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
57
165.7 ± 20.5. A TL date from Level 5 gave a reading of
105 ± 15 ka and an ESR date of 102.9 ± 12.9 ka
(Bar-Yosef 1998: Table 1). Although Mustafa and Clark
(2007) bracket the ‘Ain Difla occupations between
90 and 180 ka, the technology and typology of levels
20 to 5 fully support the early uptake readings of ca.
100 ka. The undated strata above level 5, the assem-
blages of which so strongly resemble Boker Tachtit,
Level 1, may be closer to 50 ka.
At the later Tabun B-like sites excavated on the
Jordanian plateau (i.e., Tor Faraj and Tor Sabiha), the
assemblages show a common use of bidirectional
Fig. 4. Early Levantine Mousterian artifacts from Rosh Ein Mor: Levallois points (a-o); sidescraper (p); burins (q-s); endscraper (t);
unidirectional-convergent Levallois cores (u,ab,af,ag); bidirectional Levallois core (ad); centripetal Levallois cores (v,v,ac,ae); single platform
unidirectional cores (x,z,aa). Illustrations after Crew (1976: Figs. 5.3-5.10).
Abb. 4. Steinartefakte des frühlevantinischen Moustérien aus Rosh Ein Mor: Levalloisspitzen (ao); Schaber (p); Stichel (q – s); Kratzer (t);
Unidirektional-konvergierend präparierte Levalloiskerne (u, ab, af, ag); Bidirektional präparierter Levalloiskern (ad); Zentripetal präparierter
Levalloiskern (v, ac, ae); Unidirektionale Platt formkerne (x, z, aa). Zeichnungen nach Crew (1977: Abbildungen 5.3- 5.10).

Quartär 61 (2014) J. I. Rose & A. E. Marks
58
preparation (Henry 1997, 1998), the frequency of
which increases as the point cores are reduced in size
(Henry 2003: 71). At Tor Faraj, 25 % of the Levallois
point cores have bidirectional preparation, while from
53 % to 49 % of the Levallois points have bidirectional
scar patterns, depending upon the publication (Henry
1998 vs. Henry 2003). Groucutt (2014) provides
different calculations for Tor Faraj: Levallois points
with bidirectional scar patterns are said to comprise
only 8 %. The author does not address this disc repancy;
Fig. 5. Late Levantine Mousterian artifacts from the southern Levant: Levallois points from Nahal Aqev (a) and Tor Faraj (b-j); retouched
Levallois points from Nahal Aqev (k-o); burins from Nahal Aqev (p,q); endscraper from Nahal Aqev (r); unidirectional-convergent Levallois
cores from Tor Faraj (s-ac). Illustrations after Munday (1977: Figs. 2.5-2.7); Demidenko and Usik (2003: Figs. 6.1-6.26).
Abb. 5. Steinartefakte des spätlevantinischen Moustérien aus der Südlichen Levante: Levalloisspitzen aus Nahal Aqev (a) und Tor Faraj (b-j);
Retuschierte Levalloisspitzen aus Nahal Aqev (ko); Stichel aus Nahal Aqev (p,q); Kratzer aus Nahal Aqev (r); Unidirektional- konvergierend
präparierte Levalloiskerne aus Tor Faraj (s-ac). Zeichnungen nach Munday (1977: Abbildungen 2.5-2.7); Demidenko und Usik (2003:
Abbildungen 6.1-6.26).

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
59
until independently confirmed, we adhere to the
primary site reports published by Henry (1995, 1998).
From these reports, it seems that bidirectional point
preparation was a component of the later stages of
the Tor Faraj core reduction strategy.
At Tor Sabiha, too few cores were recovered to be
meaningful, but 51 % of the 52 Levallois points have
bidirectional dorsal scar patterns (Henry 1995, 1998),
as do 26 % of the blades (Monigal 2002: 452), which is
consistent with Tor Faraj. There are the high
percentages of Levallois points with right lateral
retouch (Henry 1995: Figs. 5.13, 5.22), although
retouch in other positions is more common at Tor Faraj
than at Boker Tachtit. Data presented by Groucutt
(2014: Fig. 12) for the same observations are radically
different, perhaps owing to a different use of
nomenclature.
In stark contrast to the other Tabun D-like assem-
blages, and particularly to Tabun itself (Shimelmitz &
Kuhn 2013), the lower and middle levels of ‘Ain Difla
have over 50 % bidirectional cores, while in the upper
five levels they account for 49 %. Blades from the lower
levels (15 - 20) are similar to the cores: 44 % have
bidirectional scar patterns. This falls to 28 % for the
middle levels (6 - 14) but again rises to 43 % in the top
five levels (Mustafa & Clark 2007: Table 7). Thus,
bidirectional preparation of cores is well represented
from the earliest occupations at ‘Ain Difla, making it
distinct from Tabun D-like sites in the Negev, with
which it may temporally overlap. In fact, a study of ‘Ain
Difla’s blade technology led Monigal (2002: 488) to
conclude that, in spite of the temporal discrepancies,
all of the ‘Ain Difla assemblages should be considered
technologically the same as Boker Tachtit, Level 1.
Typologically (Lindly & Clark 1987, 2000), the upper
levels of ‘Ain Difla exhibit the same pattern of
retouched Levallois points as at Boker Tachtit.
While bidirectional core preparation alone might
be found in unrelated contexts, the combination of
bidirectional Levallois point core preparation with the
use of cresting for initial core formation is unusual.
Cresting is common in the Early Levantine UP (Monigal
2003; Davidzon & Goring-Morris 2003); yet, it is rare
in MP assemblages such as Tabun, Unit IX (Shimelmitz
& Kuhn 2013) or Hayonim, Lower E (Meignen 1998:
172), where the production of elongated blanks was
well developed, and even where volumetric cores are
present in some numbers (Marks & Monigal 1995). At
Rosh Ein Mor, there was neither a single core suggest ing
the use of cresting, nor a single crested piece reported
from the debitage (Crew 1976). The same is true for
the Tabun D-like surface sites in the Nahal Zin area
(Munday 1976). At Nahal Aqev (Munday 1977), there
are a small number of core trimming elements, but
these do not indicate cresting, sensu stricto (Monigal
2002: 411). There are a few core trimming elements
reported from Tor Sabiha and Tor Faraj (Henry 1995,
1998), but they do not include crested blades. Rather,
they tend to be “non-diagnostic flake-proportioned
elements that removed a part of a lateral or a distal
end of the core” (Monigal 2002: 449). The only crested
blades known in the southern Levant from a pre-UP
context, aside from Boker Tachtit, are those from
Levels 1 - 5 at ‘Ain Difla (Demidenko & Usik 1993;
Mustafa & Clark 2007). In addition, ‘Ain Difla yielded
six cores with crested backs (Mustafa & Clark 2007:
68-69), a type only known from the early Emiran at
Boker Tachtit (Volkman 1983). Given both the quali-
tative and quantitative similarities between ‘Ain Difla
Levels 1 - 5 and Boker Tachtit, Level 1, it is difficult not
to assign ‘Ain Difla to the Emiran.
Northeast Africa
We define northeast Africa as the territory encom-
passed by the Nile Valley and flanking Eastern and
Western deserts. This region has undergone some of
the most intensive and extensive archeological investi-
gations in all North Africa (for histories of research see
Wendorf & Schild 1976; Van Peer & Vermeersch 1990;
Van Peer 1998; Kleindienst 2006; Vermeersch 2012).
The MP of this area is exceedingly complex, and inves-
tigations spread over almost 100 years have led to
myriad industry names (e.g., Levalloisian, Mousterian,
Denticulate Mousterian, Nubian Mousterian Type A,
Nubian Mousterian Type B, Nubian Middle Palaeo-
lithic, Khormusan, etc.). Van Peer and Vermeersch
(2000, 2007) helped condense this ample array of
taxa by organizing them all into three chronological
stages - Early, Middle, and Late MP - that encompass
two different technocomplexes, based primarily on
variations in the Levallois method (Van Peer 1988,
1992). Although this organization has not been univer-
sally accepted and may require some modification
(Wendorf & Schild 1992; Garcea 2001; Kleindienst
2006), it does provide a working structure based on
specific technological criteria, thus, bringing some
harmony to the cacophony of nomenclatures.
The Nubian Levallois method
In contrast to the Levantine Mousterian, which is
dominated by unidirectional-convergent (Fig. 6: b)
and centripetal Levallois (Fig. 6: a) methods, many
northeast African assemblages exhibit a reduction
strategy based upon a markedly different system of
core preparation, referred to as Nubian Levallois. One
of the key features of this technology is the use of
bidirectional flaking to prepare and to rejuvenate a
prominent median distal ridge (Usik et al. 2013: Fig. 3),
enabling the toolmaker to remove a pointed and
elongated endproduct. The “Type 1” strategy (sensu
Guichard & Guichard 1968) relies on two distal-
divergent, elongated debordant removals to form the
pointed distal end of the core and to establish a
central guiding arête (Fig. 6: c), with further shaping of
the working surface mainly from the proximal end.
Before a point is struck, the flaking surface shows a
clear bidirectional scar pattern and a pronounced
median distal ridge. Nubian “Type 2” cores exhibit

Quartär 61 (2014) J. I. Rose & A. E. Marks
60
lateral-distal preparation to form the pointed tip
(Fig. 6: d), but results in the same prominent distal
ridge. At times, it is clear that both Type 1 and Type 2
preparations were used on the same core, particularly
during stages of rejuvenation, called Nubian Type 1/2
(Chiotti et al. 2007, 2009; Olszewski et al. 2010). In
evaluating the relationship between Type 1, 2, and 1/2
organizational systems and their efficacy as temporal
markers, Usik et al. (2013: 251-252) consider these
organizational systems simply as gradients within the
Nubian Levallois technological spectrum; therefore,
not useful as chronological indicators.
The Early MP
During the Early MP, there are two site groupings. The
first - originally called “Nubian Middle Palaeolithic” by
Guichard and Guichard (1965) - is characterized by
assemblages with classic Levallois and Nubian Levallois
(primarily Type 2) reduction strategies; along with the
façonnage production of bifacial foliates. The second
group is less abundant and lacks both Nubian Levallois
and façonnage technology, referred to by Guichard
and Guichard (ibid.) as “Non-Nubian Middle
Palaeolithic.” Neither group is dated, however, the
weathering and position on the landscape suggest
they are quite old relative to other nearby MP sites.
While Levallois reduction is prominent, Levallois
points are sporadic and there is no tendency toward
consistent blade production; laminar indices often fall
below 10. Tools are largely limited to classic MP side
scrapers, although occasional UP types are found
(Guichard & Guichard 1965: Tables 4 and 5). Their
typological and technological assemblage structures,
limited distribution around the Upper Nile Valley,
and apparent great age makes these industries unlikely
candidates for the immediate progenitor of the
Emiran.
The Middle MP
Two groups exist within the Middle MP: the “Lower
Nile Valley Complex,” characterized exclusively by
centripetal Levallois reduction, and the “Nubian
Complex,” characterized by Nubian Levallois cores
and, to a lesser extent, classic centripetal Levallois
cores (Fig. 7). Both groups occur in the low and high
deserts adjacent to the Nile Valley, but not in any of
the Nilotic silt deposits. Lower Nile Valley Complex
sites are found as far south as the Second Cataract in
northern Sudan, and as far north as the Middle Nile
Valley, as well as in the Western desert. The Nubian
Fig. 6. Schematic of preferential Levallois core preparation strategies mentioned in text: centripetal
(a), unidirectional-convergent (b), Nubian Type 1 (c), and Nubian Type 2 (d). After Rose et al. (2011:
Fig. 2).
Abb. 6. Schematische Darstellung der im Text erwähnt Levalloiskern Präparations-Strategien: Zentripetal
(a), Unidirektional konvergierend (b), Nubischer Type 1 (c), und Nubischer Type 2 (d). Aus Rose et al.
(2011: Fig. 2).

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
61
Complex has a wider distribution; ranging from south
of the Second Cataract to Middle Egypt, as well as the
Western and Eastern high deserts.
The main phase of Nubian Complex occupation in
northeast Africa occurred between ca. 130 ka and 70 ka
(Huxtable 1993; Stokes 1993; Mercier et al. 1999; Van
Peer et al. 2010). In Middle Egypt, the Nubian Complex
site of Taramsa 1, Activity Phase II, has OSL ages as far
back as 117 ka (Van Peer et al. 2010: 228), while on the
eastern edge of Egypt, the Nubian Complex
occupation at Sodmein Cave (Fig. 2: 33) produced two
TL dates averaging 118 ± 8 ka (Mercier at al. 1999).
Thus, it appears that the Nubian Complex was
widespread across Egypt during the Last Interglacial.
In the south, the Nubian Complex includes most of
the Nubian Mousterian Type A and B sites (Marks
1968a), while the Lower Nile Valley Complex include a
few Nubian Mousterian Type A sites, as well as those
called “Denticulate Mousterian” (Marks 1968a).
Nubian Complex workshop sites are abundant in
Middle Egypt on both the low (Vermeersch et al.
2000) and high deserts (Chiotti et al. 2009; Olszewski
et al. 2010), as far north as Nazlet Khater (Van Peer
1988), and as far east as Sodmein Cave in the Red Sea
Hills (Mercier et al. 1999). Actual living sites, however,
with reasonable numbers of retouched tools, are
virtually unknown (Vermeersch et al. 2000; Van Peer
et al. 2010).
Lower Nile Valley Complex assemblages have been
excavated in the Western desert (Fig. 2: 59-62) at Bir
Sahara (Schild and Wendorf 1981) and Bir Tarfawi
(Wendorf & Schild 1980; Close 1993). The Bir Tarfawi
site of BT 14 was classified as Aterian (Wendorf &
Schild 1980) based on the presence of a few bifacial
foliates and pedunculates; however, this attribution
may be questioned, as bifacial foliates are now known
to be a standard element of the Lower Nile Valley
Complex, while pedunculates are known to occur,
albeit infrequently, in a number of non-Aterian
contexts.
Since most MP sites in Upper Egypt tend to be
small workshops with poor samples of artifacts, there
are only a few where the relative frequency of bidirec-
tional cores can be judged beyond just their mere
presence or absence. One such site, El Gawanim I
(Fig. 2: 43), produced 123 identifiable cores, of which
33 % were bidirectionally prepared and, of these, 18 %
were Nubian Levallois (Vermersch et al. 2000: 20-25).
Another nearby site, El Ghineimiya 3 (ibid.: 38-39),
with only 17 cores, had 41 % bidirectional cores, of
which 86 % were Nubian Levallois. Nubian Complex
workshop sites in the high desert of Middle Egypt
show comparably important proportions of bidirec-
tional cores: at ASPS 46a they account for 40 %, at
ASPS 49 they comprise 35 %, and they average 50 %
from a series of random samples taken at various other
locales (Olszewski et al. 2010: 196). Of the 15 cores
recovered at Makhadma 6, 66 % were bidirectional,
primarily Nubian Type 1. (Van Peer 2000: 91-100).
Further south in Sudanese Nubia, the proportional
occurrence of bidirectionally-prepared cores at
Nubian Complex sites varies considerably, having
consistently lower percentages than those from
Middle and Upper Egypt. The highest percentages of
Nubian cores are at 1035 (Fig. 2: 49) with 27 %, at 1038
with 26 %, and at 1010-8 (Fig. 2: 47) with 18 %. Other
Nubian Complex assemblages in northern Sudan
exhibit a mixture of Nubian Levallois cores and
recurrent bidirectional cores ranging between 6 %
and 15 %, with the remaining identifiable types mainly
centripetal Levallois, marginal (platform cores), and
discoidal (Marks 1968a: 287, 291). Lower Nile Valley
Complex sites in Sudanese Nubia lack Nubian
technology, but occasionally exhibit bidirectional
flake cores (Marks 1968a: 209, 215). In the Western
desert, Bir Sahara 13 has neither Nubian Levallois nor
other bidirectional cores (Schild & Wendorf 1981:
102-103), as is the case for site E-72-4 at Dakhla Oasis
(Schild & Wendorf 1977: 111).
Crested blades are not found in the Nubian
Complex assemblages of Sudanese Nubia (Marks
1968a) or in the Lower Nile Valley Complex assem-
blages of the Western desert. They do occur, however,
at two Nubian Complex workshop sites in Upper
Egypt: El Gawanim 1 and Beit Khallaf 3 (Fig. 2: 43, 44).
In both assemblages, there are multiple examples
associated with Nubian cores, although no refittings
link them directly to that particular reduction strategy
(Vermeersch et al. 2000).
The high frequency of Nubian Levallois cores at
Nubian Complex sites might suggest that Levallois
points would be abundant. This, however, is not the
case. Levallois points are rare and usually quite poorly
made (Fig. 7); in Sudanese Nubia they make up a very
small portion of the toolkit, never reaching above 12 %
of the combined points, UP, and MP tool samples
(Fig. 8). The same is true for those Lower Nile Valley
Complex sites with reasonable diagnostic tool counts.
In Middle Egypt, there are so few retouched tools that
this comparison is moot. Among the unretouched
Levallois artifacts from those sites with adequate
samples, the percentage of Levallois points is
extremely low: at ASPS 46a it is 3 % and at ASPS 49 it
is 4 % (Olszewski et al. 2010: 196). At Makhadma 6
(Fig. 2: 42), where only six retouched tools were
recovered from two discrete lithic concentrations, one
was a Nubian point (Van Peer 2000).
The Late MP
The Late MP of northeast Africa is characterized by a
diversification of industries stemming from the
preceding Nubian Complex, including the Khormusan
in Sudanese Nubia (Marks 1968b) and the Taramsan in
Middle Egypt (Van Peer & Vermeersch 2007). In the
Western desert, the Aterian is supposedly rooted in
the Nubian Complex base d on the occasional presence
of Nubian Levallois cores (ibid.). Although Aterian
assemblages in this zone have not been dated directly,

Quartär 61 (2014) J. I. Rose & A. E. Marks
62
Fi g. 7. Nubian C omplex artifac ts from the Nile Valley : Levallois endprodu cts from Abydos (a,i), 1038 (b,e, o), Makhadm a 6 (c,f ), Sodme in Cave (d),
Nazlet Safaha 1 (g,h,j,k,m,n); Nubian Levallois cores from 1038 (p,u), Nazlet Safaha 1 (q), Taramsa Hill, Activity Phase III (r), Makhadma 6 (s),
Abydos (t,v,w), Abu Simbel (x-z). Illustrations after Guichard and Guichard (1965: Fig. 22); Marks (1968a: Figs. 21, 28, 32, 33); Van Peer (2000:
Figs. 4.14-4.16); Van Peer et al. (2002: Figs. 7.20-7.22, 7.46-7.51); Olszewski et al. (2005: Figs. 5, 6); Chiotti et al. (2007: Fig. 11); Van Peer et al.
(2010: Fig. 5.13); Vermeersch (2012: Fig. 2.6).
Abb . 7. Steinartefakte des Nubischen Komplexes aus dem Niltal: Levalloisabschläge aus Abydos (a,i), 1038 (b,e,o), Makhadma 6 (c,f), Sodmein
Cave (d), Nazlet Safaha 1 (g,h,j,k,m,n); Nubische Levalloiskerne aus 1038 (p,u), Nazlet Safaha 1 (q), Taramsa Hill, Activity Phase III (r), Makhadma 6 (s),
Abydos (t,v,w), Abu Simbel (x-z). Zeichnungen nach Guichard und Guichard (1965: Abbildungen 22); Marks (1968a: Abbildungen 21, 28, 32, 33);
Van Peer (2000: Abbildungen 4.14-4.16); Van Peer et al. (2002: Abbildungen 7.20-7.22, 7.46-7.51); Olszewski et al. (2005: Abbildungen 5, 6);
Chiotti et al. (2007: Abbildungen 11); Van Peer et al. (2010: Abbildungen 5.13); Vermeersch (2012: Abbildungen 2.6).

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
63
palaeoenvironmental observations suggest human
occupation in the Western desert was unlikely after
60 ka (Garcea 2001). Khormusan sites were found
stratified in Dibeira-Jer Formation Nile silts or in sands
inter-fingering within them (de Heinzelin 1968). While
34A, which is stratigraphically the earliest Khormusan
site, is undated, 1017 (Fig. 2: 36) has Th/U measure-
ments on associated teeth of ca. 84 ka (McKinney et al.
unpl. data). Site ANW3 (Fig. 2: 35), found in situ in a
downcut of the same formation, has Th/U dates on
wood remains indicating an age between 65 - 62.5 ka
(ibid.), thus, the Khormusan appears to span MIS 5.1
and MIS 4. The Taramsan Industry of the central Nile
is dated to 56.2 ± 5.5 ka (Van Peer et al. 2010), based
on OSL age estimates from Taramsa Hill 1, Activity
Phase IV (Fig. 2: 34).
The Khormusan is distributed around the Second
Cataract (Marks 1968b), and possibly somewhat to the
south. In a break with the earlier Nubian Complex, the
Khormusan utilized a wide range of non-ferrocrete
Assemblage Reference
#
Pts, MP, &
UP tools
% Levallois
points
% M P
tools
% U P
tools
Blanks
(n) ILam Cores
(n) IBi
Late Nubian Complex
Sodmein Cave 33 N/A N/A N/A N/A N/A N/A N/A N/A
Taramsa Hill 1 34 N/A N/A N/A N/A N/A N/A N/A N/A
Nazlet Safaha 1 41a 346 098 2116 3 4 N/A 427 N/A
Nazlet Safaha 2 41b 249 097 38887 N/A 298 N/A
Makhadma 6 42 N/A N/A N/A N/A 284 N/A 16 78
El Gawanim 1 43 8 0 63 38 383 27 123 33
Beit Khallaf 3 44 3 0 0 100 334 28 20 40
1010-8 47 49 245 53 225 417 18
1033, upper 48a 47 11 51 38 1649 10 66 2
1033, lower 48b 29 22 44 33 1569 13 79 6
1035 49 40 545 50 670 8138 34
1036 50 86 061 40 852 345 6
1037 51 79 346 51 299 636 10
1038 52 69 10 44 46 295 11 64 30
121 53 37 067 33 741 776 6
Early Nubian Complex
655 90 753 40 1219 11 100 5
Jebel Brinikol 56 112 273 26 927 13 125 27
Arkin 5 57 33 093 34738 193 100
Sai Island 58 N/A N/A N/A N/A N/A N/A N/A N/A
Khormusan
ANW3 35 173 10 684 2989 30 330 11
1017 36 61 8 3 89 N/A 37 74 9
34D 37 88 4 5 90 4301 23 59 19
Aterian
8708 38 24 442 54 482 912 N/A
8735 39 32 733 60 177 631 N/A
E-78 -11 40 N/A 22 57 22 N/A N/A N/A N/A
Lower Nile Valley Complex
1000 45 20 045 55 749 14 30 0
36B 46 34 064 36 603 12 26 31
Dakhla Oasis E-72- 4 59 25 892 0412 850
Bir Sahara 13 60 29 10 69 21 368 690
Bir Tarfawi 14A,
surface 61 105 584 11 1251 336 3
Bir Tarfawi 14B,
surface 62 58 081 19 2203 255 0
Unknown Middle Palaeolithic
Split Rock, upper 54a 26 577 18 1904 326 9
Split Rock, lower 54b 31 13 26 61 489 234 8
Fig. 8. Composite data for African sites included in the study. Site reference number corresponds to the map number on Figure 2 as well as
the plot numbers on Figures 13 & 14.
Abb. 8. Aufgeführt sind die technischen Daten der Afrikanischen Fundstellen, die in der hiesigen Studie behandelt werden. Referenznummer der
Fundstellen stimmen mit der Karte in Abbildung 2 und den Diagrammen in Abb. 13 & 14 überein.

Quartär 61 (2014) J. I. Rose & A. E. Marks
64
sandstone raw materials that had been washed down
the Nile. Technologically, it initially exhibits a prefe-
rential Levallois reduction strategy that resulted in the
production of classic Levallois flakes from ovate-
shaped cores. These flakes were mainly used as blanks
for UP burins, although a few MP side scrapers were
produced, as were large numbers of well-made
denticulates. Point and blade production, on the other
hand, was quite limited (Fig. 8). A special technique
was used to flake small Nile pebbles (Sellet 1995),
while bidirectional core preparation is quite rare
(6 % - 9 %, including 1.6 % Nubian at site ANW3).
Through the duration of the Khormusan, the Levallois
component remained strong; there was no tendency
toward increasing amounts of blade or bidirectional
core reduction strategies, only the increasing use of
Nile pebbles as a source of raw material.
Compared to the western Egyptian and Libyan
deserts (Ferring 1975; Wendorf & Schild 1992; Garcea
2001), there is a relative paucity of Aterian sites in or
near the Nile Valley. No Aterian has been reported
from the Eastern desert and only a single, poor surface
site, E-78-11 (Fig. 2: 40), was discovered within the Nile
Valley itself (Singleton & Close 1980). Little can be
said of this assemblage since only a Bordian type list of
the tools was published, but, typologically at least, it is
consistent with other Egyptian Aterian assemblages
(Fig. 8). There is a cluster of Aterian surface scatters
and at least one workshop at Kharga Oasis (Caton-
Thompson 1952; Wendorf & Schild 1992). Of eight
surface scatters reported in the Libyan desert near
Dungal Oasis, six had inadequate samples (<60) to
make any observations (Hester & Hobler 1969: 80-81).
Indeed, their attribution to the Aterian was solely
based on the presence of at least one pedunculated
tool (Hester & Hobler 1969: 83). Only two of the eight
sites, 8708 (Fig. 2: 38) and 8735 (Fig. 2: 39), have
reasonable tool assemblages (Ferring 1975: 116) that
permit a view of typological patterning. Unlike the
Khormusan, there is roughly a 40/60 split between MP
and UP tools and Levallois points are extremely rare
(Fig. 8). Neither site exhibits any evidence of bidirec-
tional core reduction; however, this may be due to
small core samples, the generalized typology used, or
the nature of the sites themselves. While no Nubian
Levallois cores were reported from any of these
Aterian assemblages, small numbers of bidirectional
and Nubian cores were found at Aterian sites further
to the west, such as Uan Tabu (Garcea 1999: 173).
During Activity Phase IV at the Taramsa Hill 1
quarry site (Van Peer et al. 2010: 89-117), there is a
shift observed in the core reconstructions from classic
Nubian Levallois production in the preceding Activity
Phase III to a “Taramsa blade production system”
(ibid.: 234). This involved a modification of the Nubian
Levallois reduction strategy to increase the convexity
of the flaking surface, so that a series of elongated
blanks could be bidirectionally struck from the
core without the traditional need to rejuvenate the
convexities of the flaking surface (ibid.: 53). This
resulted in the serial production of large numbers of
elongated blanks (Fig. 9), mainly blades, although a
few typical parallel-sided Nubian points were
produced as well (ibid.: Fig. 6.24, 6.26, 6.28). Since
both Nubian Levallois and Taramsa blade systems are
co-associated, this Activity Phase is considered
technologically “transitional” (ibid.: 241). Among the
tools, fifteen blanks had some retouch, mainly simple,
notched or denticulated, while the only formal tool
was a single end scraper. OSL ages from Activity Phase
IV are between 60 and 50 ka, while ages from the
overlying archaeological stratum, Activity Phase V,
cluster around 40 ka; as such, the Phase V assemblage
cannot have been related to the early development of
the Emiran.
The uppermost Middle Palaeolithic archaeological
level at Sodmein Cave, MP1, produced a small
collection of artifacts that was reported to include two
Emireh points (Fig. 7: d) and two burins (Mercier et al.
1999). The points in question are described as having
“basal thinning on the ventral face” (Mercier et al.
1999: 1340). These specimens, however, do not fit the
definition of Emireh points, which must have bifacial
basal thinning (Volkman & Kaufman 1983; Copeland
2001). Without a strict definition of this type fossil,
there is a risk of expanding the classification to
subsume all marginally retouched Levallois points,
such as those from the “Tabun B” sites in southern
Jordan. Hence, we maintain that the present evidence
precludes classifying level MP1 at Sodmein Cave as
Emiran.
In sum, the full suite of technological characte-
ristics seen in the early Emiran is not found together in
any Nilotic industry. Typologically, the dominance of
UP tools, as well as the emphasis on Levallois points, is
entirely missing from Africa. Yet, the emphasis on
bidirectional preparation that is characteristic of the
Nubian Levallois reduction strategy begs the question:
could this be the source of Emiran opposed platform
Levallois point production? It is certainly plausible
that Emiran core technology may, ultimately, derive
from the Early Nubian Complex in Africa; however,
other Emiran features such as cresting and elongation
are typically not found together in African assem-
blages. Crested blades have only been found at three
Nubian Complex sites, while the lateral retouch (“later-
alization” sensu Van Peer 1991), occurs at two Middle
Nile Valley Complex sites: Nazlet Safaha 1 and 2
(Vermeersch et al. 1990). Crested blades and “later-
alized” Levallois pieces have not been found together
in the same assemblages, but their presence in Egypt
might suggest similar approaches to core formation
and hafting at roughly the same time as those seen at
Boker Tach t it.
The Arabian Peninsula
For the last century, the Arabian Peninsula has been
relegated to terra incognita, and is traditionally not

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
65
Fig. 9. Taramsan artifacts from Taramsa Hill 1, Activity Phase IV: Levallois points (a-h), Levallois blades (i-u), Nubian Levallois cores (v-z).
Illustrations after Van Peer et al. (2010: Figs. 6.21-6.29).
Abb. 9. Steinartefakte des Taramsan aus Taramsa Hill 1, Activity Phase IV: Levalloisspitzen (a-h), Levalloisklingen (i-u), Nubische Levalloiskerne
(v-z). Zeichnungen nach Van Peer et al. (2010: Abbildungen 6.21-6.29).

Quartär 61 (2014) J. I. Rose & A. E. Marks
66
considered part of the Near Eastern MP archaeo-
logical record (e.g., Bar-Yosef 1994, 2000). Within the
last decade, however, new research projects in Yemen
(Crassard 2007, 2009; Crassard and Thiebaut 2011;
Delagnes et al. 2012, 2013), Oman (Rose 2004, 2006,
2007; Rose et al. 2011; Usik et al. 2013; Hilbert et al.
2014; Rose & Hilbert 2014), Saudi Arabia (Petraglia et
al. 2011, 2012; Crassard & Hilbert 2013; Crassard et al.
2013), and the UAE (Marks 2009; Armitage et al. 2011;
Bretzke et al. 2013, 2014) have discovered a wide
variety of MP lithic assemblages, some of which may
be considered as potential demographic and/or
cultural sources of the Emiran.
Early MIS 5
The earliest known post-Acheulean occupation of
Arabia was discovered at the collapsed rockshelter
site of Jebel Faya 1 (Fig. 2: 15) in Sharjah, UAE (Marks
2009; Armitage et al. 2011; Bretzke et al. 2013, 2014).
While its Assemblage C is argued to stem from the
East African MSA (Armitage et al. 2011), neither the
technology nor typology shows any element even
vaguely related to the Emiran. It is characterized,
rather, by the co-association of bifacial, volumetric
blade, and preferential Levallois flake reduction
strategies; a techno-typological package that does not
exist in the Levant. The two stratigraphically younger
assemblages at Jebel Faya, Assemblages A and B
(ibid.), seem to be strictly of local origin, having no
bearing on the Emiran. Other sites around the Gulf
and its hinterlands (Biglari et al. 2009; Scott-Jackson et
al. 2009; Wahida et al. 2009) exhibit similar techno-
logical co-associations as Assemblage C and, as such,
are not relevant to our investigation of the earliest
Emiran.
The most abundant and distinct type of MP on the
Arabian Peninsula is the Nubian Complex, attributed
to a population dispersal from Africa during early MIS
5 (Rose et al. 2011; Beyin 2013; Crassard & Hilbert
2013; Usik et al. 2013). Two matching OSL measure-
ments of ca. 106 ka from the Classic Dhofar Nubian
site of Aybut al-Auwal (Fig. 2: 27) indicate that Nubian
Complex toolmakers were present on the Peninsula by
MIS 5.3 (Rose et al. 2011). Given its affinities to those
assemblages found in Egypt, Usik et al. (2013: 244)
propose the term “Afro-Arabian Nubian
Technocomplex.”
Hints of Nubian Levallois technology were first
found in Arabia in the 1980s, from a handful of cores
recovered at surface sites in Wadi Muqqah (Fig. 2: 25),
western Hadramawt, Yemen (Inizan and Ortlieb 1987).
Some years later, several surface scatters were
reported from Wadi Sana and Wadi Wa’sha in central
Hadramawt (Fig. 2: 26), where Crassard (2007: 7-8)
noted a resemblance between bidirectionally-
prepared Levallois point cores (“B2,” “B3,” and “B4”
types; Fig. 10: s) and Nubian Levallois cores from
Africa. Although over 20 sites were mapped exhibiting
such technology, the assemblages’ small sample sizes
precluded any conclusive determination of industry
type. Between 2010 and 2013, the Dhofar Archaeo-
logical Project mapped over 250 surface sites with
Nubian Levallois technology on the Nejd plateau in
southern Oman, ranging from large-scale workshops
(>2000 artifacts) to isolated points and discarded
Nubian Levallois cores (Rose et al. 2011; Usik et al.
2013). Most recently, additional Nubian/Nubian-
derived assemblages have been reported on
interdunal gravels in the Rub’ al Khali desert (Rose &
Hilbert 2014), near Al Kharj (Fig. 2: 24) in central Saudi
Arabia (Crassard & Hilbert 2013), and in the Al Jawf
(Fig. 2: 18) region of northern Saudi Arabia (Hilbert et
al. unpubl. data). No Nubian Complex assemblages
have yet been found in eastern Arabia. We discount
the single Nubian core reported by Wahida et al.
(2009), which comprises less than 2 % of the total
cores and does not exhibit any of the essential charac-
teristics of Nubian Levallois technology (sensu Usik et
al. 2013).
In Dhofar, Nubian Complex sites are almost exclu-
sively characterized by the standardized production
of large, elongated points (Fig. 10: a-l) via preferential
bidirectional Nubian Levallois core preparation
(Fig. 10: m-o, q-r), typically comprising well over 50 %
of total cores (Rose et al. 2011: Table 3; Usik et al. 2013:
Table 4). In the most extreme case, of the 172 cores
collected from the site of Aybut ath-Thani, 155 (90 %)
were Nubian Levallois. The lowest percentage of
Nubian Levallois cores comes from Jebel Markhashik 1,
where 65 (57 %) of the 115 specimens were classified
as such. Core technology is based on opposed
platform preparation; the only other ty pe of reduction
found in any significant amounts is a simple-unidirec-
tional strategy for the production of elongated blanks,
ranging from 10 % to 20 %. The classic centripetal
Levallois strategy is quite rare, comprising less than
3 % of cores at most sites. Typologically, retouched
tools are uncommon, but when found in reasonable
numbers are weighted toward MP forms, primarily
side scrapers, with a low number of end scrapers on
flakes present. Nubian points are numerous (Fig. 11)
and show some morphological variability: both classic
Nubian points with a pitched shape, as well as true
triangular Levallois points, are common in this industr y.
Late MIS 5 – Early MIS 3
The Nubian Levallois tradition endured in Dhofar for
some time, encompassing at least two separate
technological phases: the Classic Dhofar Nubian and
its derivative, the Mudayyan Industry (Usik et al.
2013). Although the Mudayyan has no absolute dates,
assemblages are consistently found on landscape
surfaces post-dating the Classic Dhofar Nubian, and
consistently exhibit far less patination and chemical
dissolution than artifacts from Classic Dhofar Nubian
assemblages . The industry’s more limited distribution
around fossil springs speckling the Nejd plateau, as
opposed to the Classic Dhofar Nubian, which is

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
67
ubiquitous from the Dhofar escarpment to the Rub’
al-Khali desert (Rose & Hilbert 2014), is suggestive of
diminished mobility in response to environmental
desiccation. Hence, we surmise that the Mudayyan
may coincide with a phase of weakened Indian Ocean
Monsoon activity after 75 ka (e.g., Fleitmann et al.
2003, 2011; Stokes & Bray 2005; Preusser, 2009).
Based on a sample of five seemingly typical
Mudayyan assemblages found on the Nejd plateau,
including Jebel Kochab 1 (Fig. 2: 23), Umm Mudayy 1
(Fig. 2: 21), Umm Mudayy 2 (Fig. 2: 22), Jebel Dahsha
(Fig. 2: 19), and Burj Dakin (Fig. 2: 20), it is clear that the
most prominent reduction strategy is preferential
“Micro-Nubian” Levallois (Fig. 12: a-e), ranging from
Fig. 10. Nubian Complex artifacts from the Arabian Peninsula: Levallois points from TH.236 (a), TH.323 (b), TH.238 (c), Jebel Markhashik 1
(d,e), Aybut ath-Thani (f), Aybut al-Auwal (g,i), TH.173 (h), TH.258 ( j), Mudayy as-Sodh 1 (k), Jebel Sanoora 1 (l); Nubian Levallois cores from
Aybut al-Auwal (m,n), Mudayy as-Sodh 1 (o), Al Kharj 22 (p), Jebel Markhashik 1 (q), TH.323 (r), Wadi Wa’shah (s). Illustrations af ter Crassard
(2009: Fig. 7); Rose et al. (2011: Figs. 9, 10, 14); Crassard and Hilbert (2013: Fig. 7); Usik et al. (2013: Figs. 2, 7).
Abb. 10. Steinartefakte des Nubischen Komplexes aus der Arabischen Halbinsel: Levalloisspitzen aus TH.236 (a), TH.323 (b), TH.238 (c),
Jebel Markhashik 1 (d,e), Aybut ath-Thani (f), Aybut al-Auwal (g,i), TH.173 (h), TH.258 (j), Mudayy as-Sodh 1 (k), Jebel Sanoora 1 (l); Nubische
Levalloiskerne aus Aybut al-Auwal (m,n), Muda yy as-Sodh 1 (o), Al Khar j 22 (p), Jebel Ma rkhashik 1 (q), TH.323 (r), Hadrama wt (s). Zeichnungen
nach Crassard (2009: Abbildung 7); Rose et al. (2011: Abbildungen 9, 10, 14); Crassard und Hilbert (2013: Abbildung 7); Usik et al. (2013: Abb. 2, 7).

Quartär 61 (2014) J. I. Rose & A. E. Marks
68
19 % to 37 % of all cores found in these assemblages
(Rose et al. unpubl. data). At every site, these diminutive
Nubian Levallois cores grade into a bidirectional
Levallois variant (from 4 % to 23 %), in which the medial
distal ridge is flat and the shape of the core ranges
from sub-triangular to ovate, producing ovate
endproducts. These types, in turn, grade into rectan-
gular, opposed platform Levallois cores (from 6 % to
23 %) that enabled the serial production of elongated
pointed endproducts (Fig. 12: f-i). In addition to these
different types of bidirectional core reduction,
Mudayyan assemblages exhibit an entirely separate
reduction strategy (from 13 % to 29 %), in which blanks
were unidirectionally struck from the narrow,
elongated working surface of the core (Fig. 12: k). In
only a few cases (n=8) is there technological overlap, in
which a Nubian Levallois strategy was applied to the
narrow working surface of the core (Fig. 12: j); for the
most part, it seems these reduction strategies were
applied separately (ibid.).
Crested blades, albeit rare (1 % - 7 % of debitage),
were found in four of the Mudayyan assemblages.
While it is possible that these are lateral byproducts
of Nubian Levallois preparation, the presence of six
pre-cores at Jebel Dahsha, exhibiting coarse bifacial
flaking followed by a single blade struck from the
narrow working surface of the core, suggests that
some kind of rudimentary cresting technique was
employed within the Mudayyan blade reduction
strategy (ibid.).
The shift f rom classic Nubian Levallois to Mudayya n
core reduction is characterized, in part, by a change
from preferential Nubian Levallois to recurrent
bidirectional point production systems. In the Classic
Dhofar Nubian (as well as the Late Nubian Complex in
Africa), the Nubian core’s prominent median distal
ridge enabled toolmakers to control distal convexity
and achieve an elongated point. Within some Classic
Dhofar Nubian assemblages, there are occasional
cases in which the distal ridge is flat and generates two
or three points from the opposed platform, rather
than the twisted débordant blades that are typical of
Type 1 Nubian Levallois reduction. In Mudayyan
assemblages, conversely, these aberrant bidirectional
point cores become increasingly common, as Nubian
Levallois toolmakers de-emphasized preferential core
preparation, in favor of a flat flaking surface with
opposed platforms that enabled them to serially
produce points from both ends of the prepared core.
Consequently, rectangular-shaped cores replace the
distinctive triangular/sub-triangular classic Nubian
Levallois morphology. In both industries, the
Assemblage Reference # Pts, MP, &
UP tools (n)
% Levallois
points
% M P
tools
% U P
tools
Blanks
(n) ILam Cores
(n) IBi
Mudayyan Industry
Jebel Dahsha 19 22 18 577 169 16 143 51
Burj Dakin 20 15 47 053 228 953 82
Umm Muday y 1 21 24 58 042 402 25 143 55
Umm Muday y 2 22 39 33 760 128 37 166 73
Jebel Kochab 23 743 057 516 21 157 33
Afro-Arabian Nubian Technocomplex (general)
Al Jawf 18 N/A N/A N/A N/A N/A N/A N/A N/A
Al Kharj 22 24 N/A N/A N/A N/A N/A N/A N/A N/A
Wadi Muqqah 25 N/A N/A N/A N/A N/A N/A N/A N/A
Wadi Sana/ Wadi Wa‘shah 26 N/A N/A N/A N/A N/A N/A N/A N/A
Classic Dhofar Nubian Industry
Aybut al-Auwal 27 102 59 23 19 407 26 297 75
Aybut ath -an i 28 56 38 61 21503 12 157 85
Mudayy as-Sodh 1 29 32 56 22 22 804 16 92 77
Jebel Sanoora 1 30 771 29 0330 38 104 74
Jebel Markhashik 1 31 35 100 0 0 693 22 115 79
TH. 377 32 22 46 27 18 126 34 45 88
Unknown Middle Palaeolithic
Wadi Surdud Complex 14 N/A N/A N/A N/A 4793 22 39 N/A
Jebel Faya, assemblage A 15a 17 029 71 1081 16 72 4
Jebel Faya, assemblage B 15b 37 043 57 1295 19 84 13
Jebel Qaar 1 16 N/A N/A N/A N/A 79 4 9 0
Jebel Katefeh 1 17 N/A N/A N/A N/A 765 398 9
Fi g. 11. Composite data for Arabian sites included in the study. Site reference number corresponds to the map number on Figure 2 as well
as the plot numbers on Figures 13 & 14.
Abb . 11. Aufgeführt sind die technischen Daten der Arabischen Fundstellen, die in der hiesigen Studie behandelt werden. Referenznummer der
Fundstellen stimmen mit der Karte in Abbildung 2 und den Diagrammen in Abb. 13 & 14 überein.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
69
endproducts are Levallois points (from 18 % to 58 % of
tools at Mudayyan sites); however, Mudayyan Levallois
points are categorically smaller (Usik et al. 2013).
In contrast to the Classic Dhofar Nubian, in which
retouched tools are few and poorly made, Mudayyan
assemblages exhibit a higher percentage of tools and
greater range of retouched tool types. Diagnostic MP
types were only found in two assemblages, and
account for less than 7 % of tool types (Figure 11).
UP types comprise the vast majority, accounting for
between 42 % and 77 % of Mudayyan tool
assemblages, including a variety of convex, nosed,
straight, and atypical endscrapers, as well as burins
and perforators (Rose et al. unpubl. data).
There are three published MP sites associated
with ancient lake deposits around the Jubbah palaeo-
lake basin (Fig. 2: 16-17) in northern Saudi Arabia
(Petraglia et al. 2011, 2012). Potentially the oldest,
Jebel Umm Sanman, is roughly bracketed between
100 and 60 ka. Excavations produced only 77 buried
artifacts, mainly non-dia gnostic debitage but including
10 centripetal Levallois cores. At another locality,
Jebel Katefeh 1, the artifact-bearing layer, Unit H,
yielded two clusters of equally probable OSL ages
around 85 - 90 ka or 50 ka. Although 300 artifacts
were excavated from Jebel Katefeh, 270 of these are
chips. An additional 923 artifacts were collected from
the surface and included together with the buried
assemblage counts. In total, there were 39 Levallois
cores, the majority of which have unidirectional-
convergent or centripetal preparation. In addition, a
number of broad-based Levallois points were
recovered (Petraglia et al. 2012: fig. 10). The site of
Jebel Qattar 1 has OSL ages of 75 ± 5 ka. A total of 114
artifacts were recovered, including small discoids and
centripetally prepared Levallois cores, as well as ten
informally retouched tools.
Survey around the Mundafan palaeolake in south-
western Saudi Arabia yielded a surface scatter of MP
artifacts including preferential, centripetal Levallois
cores and their products. In spite of its small assem-
blage size, sufficient cores were recovered for
researchers to observe that the Mundafan assemblage
shows some overlap with the MP material from Jubbah
Fig. 12 . Mudayyan artifacts from Jebel Kochab 1, Jebel Dahsha, and Nubi as-Saghir: Micro-Nubian cores (a-e); bidirectional Levallois cores
(f-i); single and opposed platform blade cores ( j, k); Levallois point (l). Illustrations after Usik et al. (2013: Figs. 15 & 16) and Hilbert et al.
(unpubl. data).
Ab b. 12. Steinartefakte des Mudayyan aus Jebel Kochab 1, Jebel Dahsha, and Nubi as-Saghir: Micro -nubische Levalloiskerne (a-e); Bidirektional
präparierte Levalloiskerne (f-i); Klingenkerne mit Abbaufläche an Schmalseite (j, k); Levalloisspitzen (l). Zeichnungen nach Usik et al. (2013:
Abbildungen 15 & 16) und Hilbert et al. (unpubl. Daten).

Quartär 61 (2014) J. I. Rose & A. E. Marks
70
(Crassard et al. 2013). In all of these cases, not one of
the diagnostic Emiran elements is present, suggesting
the antecedents of the Emiran are unrelated to any of
the Jubbah or Mundafan assemblages.
The only dated evidence for human occupation in
the Arabian Peninsula during MIS 3 comes from the
aforementioned Jebel Faya 1 rockshelter in eastern
Arabia, and from the Wadi Surdud Complex (Fig. 2:
14) in Yemen, where two assemblages dating between
63 and 42 ka were found interstratified within a
six-meter fluvial accretion (Delagnes et al. 2012; Sitzia
et al. 2012). Over 5,000 artifacts were excavated, and
in both archaeological horizons, the most prominent
reduction system was, by far, a simple unidirectional-
convergent strategy producing elongated pointed
flakes and blades (Delagnes et al. 2012: 13). The
excavators assigned both the upper (SD2) and lower
(SD1) assemblages to the Late MP, noting that it is
primarily a non-Levallois strategy, since most striking
platforms (>70 %) are either unfaceted or cortical, and
less than 10 % exhibit any kind of faceting. Elongated
pointed blank production was flexible, grading from
occasional instances of preferential, unidirectional-
convergent Levallois preparation to the more frequent
use of recurrent “frontal” or “semi-tournant” core
exploitation (ibid.: Fig. 12). There are just 25 retouched
pieces (<1 %), none of which is considered a formal
tool. The paucity of retouched tools is not necessarily
a characteristic of the Wadi Surdud industry; rather it
may be due to “the physical properties of rhyolite,
which rendered the transformation of blank edges
difficult and/or unnecessary” (ibid.:14). The Wadi
Surdud assemblages share a single overlapping feature
with the Emiran in the manufacture of elongated
pointed blanks. Their predominantly unidirectional-
parallel/unidirectional- convergent laminar production
system, however, does not suggest any direct
connection to the Emiran.
Discussion
A local origin?
If the Emiran arose from a local, southern Levantine
base, without external influences, we should expect
that its technology and typology could be traced back
into its ancestry. Certainly, the emphasis on the
production of triangular Levallois points and their
associated elongated debitage seems to be the
continuation of a deeply rooted pattern reaching back
to the Early Levantine Mousterian; this is even the case
when the points themselves were not elongated, as at
Tor Faraj and Tor Sabiha (Henry 1997, 1998). The one
significant technological discontinuity between the
Emiran and all other Levantine Mousterian industries,
however, is the use of a bidirectional Levallois strategy
associated with extensive cresting in core preparation
and rejuvenation. This stands out in contrast to the
unidirectional-convergent Levallois strategy and use
of débordant removals to rejuvenate flaking surfaces,
which is ubiquitous throughout the Levantine
Mousterian. I n short, the Emiran assemblages (Bokerian
I, sensu Leder 2013) employed a foreign Levallois
strategy to produce points and some blanks for UP
retouched tools that were fully consistent with
long-term Levantine patterns.
The local, pre-Emiran examples of bidirectional
reduction are mainly found in Middle and Late
Levantine Mousterian contexts across the Arava Valley
in southern Jordan. Although some bidirectional scar
patterns do occur on debitage in northern Levantine
Mousterian assemblages (e.g., Kebara, Units VII, XI,
XII), the cores and Levallois points themselves exhibit
unipolar convergent preparation (Meignen 1995;
Meignen & Bar-Yosef 1988). The “Tabun D-type”
assemblages from D40, ‘Ain Difla, Levels 6 - 20, and
“Tabun B-type” assemblages from Tor Faraj and Tor
Sabiha, all exhibit an unusually high index of bidirec-
tionality among the cores (Fig. 3); in the case of ‘Ain
Difla, at least 50 ka before it is seen at Boker Tachtit.
That this apparently non -Levantine reduction strategy
appears in both Tabun D and B-type assemblages
mainly in southern Jordan suggests it was either intro-
duced from further south or was an autochthonous
development emphasized in this area, perhaps in
attempts to increase core productivity (Henry 1998:
32). On the other hand, the ubiquitous truncated/
faceted cores of the MP virtually disappear in the
Emiran, with only three in Boker Tachtit, Level 1,
comprising less than 1 % of Boker Tachtit cores.
The tool types found in the Emiran match those of
the southern Levantine Mousterian in several aspects.
The high frequency of Levallois points characterizes
most Tabun D and B-type assemblages, which is
visually represented in a ternary plot of the relative
percentages of Levallois points versus MP tools versus
UP tools from each assemblage in this analysis (Fig. 13).
This trend stands out in contrast to some Tabun C
type assemblages, such as those at Ksar Akil (Marks &
Volkman 1986), as well as virtually all the African and
Sinai sites. The significant presence of UP tool types,
as opposed to MP types, is part of a general trend in
the southern Levant, reaching as far back as the Early
Levantine Mousterian. UP tools are also found in
significant proportions in some northern Levantine
sites of very different ages, from the Amudian of
Qesem Cave (Shimelmitz at al. 2011), through Tabun D
at Tabun level 39 (Jelinek 1975), to the Mousterian of
Qafzeh (Hovers 2009). With the exception of some
Tabun C assemblages, this is a widespread Levantine
MP trait. The significant presence of Levallois points
with lateral retouch adjacent to the proximal platform,
on the other hand, is one Emiran characteristic that is
not widespread in the preceding Levantine
Mousterian. In the south, this trait is not present at
Nahal Aqev (<80 ka) and is only found in the uppermost
levels at ‘Ain Difla above Level 5 (<100 ka). Thus, it may
represent either an Emiran innovation or a foreign
trait adopted by the Emiran.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
71
Viewed solely within a Levantine context, the
technological and typological patterning of the early
Emiran might suggest it derived from local innovation
without significant external demographic input.
Meignen (2012), for instance, sees a “stimulus for
new combinations of pre-existing technologies.” If
locally restricted to the southern Levant, one possible
explanation is that this stimulus came from decreased
landscape carrying capacity in the central and
southern Negev during MIS 4 (Vaks et al. 2007;
Frumkin et al. 2011), at which time aridification may
have encouraged higher mobility and a more efficient
system of point and blade production (e.g., Marks &
Freidel 1977). The problem with this scenario,
however, is that bidirectional Levallois point
production was well established at ‘Ain Difla before
the onset of MIS 4, and the extensive use of cresting
that is associated with this reduction strategy is no
more efficient than the removal of débordant blades
to shape and rejuvenate core surfaces.
So, with no clear antecedents in the Levant, do the
three Emiran technological traits that have no deep
Levantine ancestry (i.e., bidirectional core prepa-
ration, the use of cresting, and the presence of lateral
modification on Levallois points), have demonstrable
origins elsewhere? Conversely, do those deeply
rooted Levantine characteristics of the Emiran (i.e.,
elongated Levallois point production and abundant
UP tool manufacture) have comparable analogues in
adjacent areas?
Through the Nile Corridor?
Any model that predicts the dispersal of AMHs out of
northeast Africa directly through Sinai into the
southern Levant, a matter of only a few days walk (in
one case, as much as 40 years), must also accept that
those AMHs would have brought their culture with
them. Therefore, the degree and extent of techno-
logical and typological parallels between Boker
Tachtit, Levels 1 - 3, and Nilotic industries is a crucial
test for the Out of Africa model through the Levantine
Corridor. The African Nubian Complex, sensu latu, is
the most obvious antecedent for Emiran bidirectional
point production. Yet, 100,000 years separate the
initial known manifestation of Nubian Levallois
technology at 150 ka at Sai Island, (Van Peer et al.
2003) and possibly at Arkin 5, near the Second
Cataract (Chmielewski 1968) from the Emiran around
50 ka. To examine their relationship, we must chart the
development of Nubian Levallois technology over this
interval of time.
During MIS 5, the geographic center of the Nubian
Complex in Africa appears to be the Egyptian Middle
Nile Valley and its hinterlands (Van Peer & Vermeersch
2007; Chiotti et al. 2009; Olszewski et al. 2010). The
absence of Nubian Complex sites in Lower Egypt may
be due to the thick mantle of post-MIS 5 sediments
covering more ancient landscapes (Vermeersch 2009:
72). Bidirectional indices are relatively high (Fig. 3),
although, in nearly every case, do not approach the
levels seen at Boker Tachtit, Levels 1 - 3 (Marks 1983a)
Fig. 13. Ternary plot showing relative percentages of Levallois points versus Middle Palaeolithic tools versus Upper Palaeolithic tools.
Abb. 13. Ternärdiagramm zur Gegenüberstellung der relativ prozentualen Anteile von Levalloisspitzen, mittelpaläolithischer sowie jungpaläo-
lithischer Werkzeugformen.

Quartär 61 (2014) J. I. Rose & A. E. Marks
72
or ‘Ain Difla, 1 - 5 (Mustafa & Clark 2007). The
frequency of bidirectional technology in the African
Nubian Complex tends to be the same as that as the
“Tabun D” and “Tabun B” assemblages from southern
Jordan (Fig. 14), which is in contrast to the almost exclu-
sively unidirectional-convergent core preparation
Fig. 14. Scatterplot showing indices of bidirectionality versus elongation.
Abb. 14. Streudiagramm zur Gegenüberstellung der Indices zur Bidirektionalität und Länge Artefakten.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
73
strategies of the classic Tabun D and B assemblages
found in the Mediterranean zone. In this sense, the
late MP material from southern Jordan is techno-
logically closer to the African Nubian Complex than
the Levantine Mousterian of the Carmel.
In both the Nubian Complex and the Lower Nile
Valley Complex, there is a tendency toward the
production of Levallois flakes, rather than Levallois/
Nubian points that are typical of the Emiran (Fig. 7).
Laminar indices are also quite low (Fig. 14) and MP
tools are predominant relative to UP types (Fig. 13).
There is also a strong trend toward the production of
denticulates in Lower Nile Valley Complex assem-
blages, in some cases comprising the majority of the
combined UP/MP tool assemblage (Marks 1992: 242),
a pattern that is absent in the Levantine MP (ibid.).
Given these technological and typological conside-
rations, we find no direct relationship between any
northeast African MIS 5 industry and the early Emiran.
The African data do suggest that the Emiran’s prefer-
ential bidirectional Levallois point product ion strategy
ultimately arose in northeast Africa during late MIS 6.
This reduction strategy became increasingly
widespread during MIS 5, where it spread as far south
as the Ethiopian Rift and east into the Arabian
Peninsula. This wide distribution suggests extensive
cultural contact, either direct or indirect, along the
Nile Valley and across the Red Sea during the Last
Interglacial, when climatic conditions were optimal.
If Nubian Levallois technology reached the K’One
crater in Ethiopia, some 2300 km south of Middle
Egypt, bidirectional reduction at ‘Ain Difla, some
600 km to northeast should not be a surprise. Yet,
while some Nubian-like cores have been sparsely
documented at a few sites (Crew 1975; Vermeersch
2001) the intensive use of Nubian Levallois technology
does not appear to have reached any further into the
Levant. As importantly, the consistent techno-
typological patterns seen among all northeast African
MIS 5 assemblages (i.e., a paucity of Levallois points,
roughly equivalent UP and MP retouched tools, and a
lack of significant laminar production), is not present
at ‘Ain Difla. At most, the high frequency of bidirec-
tional core preparation in Levels 6 - 20, represents the
selective adaptation of an originally African trait.
The three northeast African MP industries
extending into MIS 4 - the Khormusan, Taramsan, and
Aterian - are the most likely African candidates to have
contributed to the development of the Emiran. The
Khormusan’s limited distribution around and south of
the Second Cataract (Marks 1968b), however, makes
any direct connection to the southern Levant highly
unlikely. The only aspect of the Khormusan that
parallels the Emiran is the clear dominance of UP tools,
relative to MP ones (Fig. 13). Yet, when Khormusan
technology is considered, with its emphasis on ovate
Levallois core reduction, paucity of Levallois points,
minor blade component, and rarity of bidirectional
core preparation (Fig. 14), there is no demonstrable
relationship to the Emiran.
Among the Aterian (or Atero-Mousterian sensu
Dibble et al. 2013) of northeast Africa, the ratio of UP
tools to MP ones is only slightly higher than in other
coeval assemblages found in the region. The trivial
production of Levallois points is fully consistent with
other northeast African MP assemblages, and radically
different from the Emiran (Fig. 13). In addition, its
geographic distribution is not documented east of the
Nile Valley and it is barely present in the valley itself. If
those signature aspects of the Atero-Mousterian -
pedunculates and bifacial foliates - had expanded
through the Sinai into the Negev, they would surely
appear in at least one MIS 5 or MIS 4 context
somewhere in the southern Levant, which they do not.
The lithic workshops excavated at Makhadma 6
and Taramsa 1, Activity Phases IV demonstrate that
the Nubian Levallois reduction strategy persisted
through MIS 4 and into MIS 3, while undergoing
technological modifications leading to serial blade
production. Given this technological trajectory, similar
to that of the Emiran, and the age range immediately
preceding it, it is conceivable that this assemblage
could represent the progenitor of the Emiran (Meignen
& Bar-Yosef 2005). Indeed, it does seem to belong to
a technologically transitional stage, where an increase
in flaking surface convexity on Nubian Levallois cores
shifted them to volumetric reduction, permitting
serial production of blades, while maintaining the
necessary flaking surface convexities. In that sense, the
Taramsan blade production system shares two critical
elements with the Emiran: bidirectional reduction and
the recurrent production of elongated blanks.
On the other hand, there are a number of techno-
logical differences between the Taramsan and the
Emiran. The assemblage from Boker Tachtit, Level 1,
includes one example of true prismatic blade
technology (Volkman 1983: 133-136), which never
occurs at Taramsa Hill. At the same time, the bidirec-
tional Levallois point system at Boker Tachtit was
modified so that a greater numbers of blades were
produced both before and after the removal of the
Levallois point. Unlike Taramsan technology, however,
this did not result in increase d flaking surface convexit y
and a shift toward volumetric reduction on the
bidirection al Levallois point cores. Rather, non-Levallos
bidirectional volumetric reduction, first appearing in
Level 1, gradually replaced the bidirectional Levallois
point strategy. Thus, while both the Taramsan and the
Emiran exhibit technological transitions from prefer-
ential, bidirectional Levallois systems to recurrent,
non-Levallois systems, the means by which these
trajectories developed were quite different.
While MIS 4 - MIS 3 data from Egypt show devel-
opmental trajectories heading toward the UP, and
even a few specific traits found in the Emiran, there is
not a single Egyptian assemblage that can be recog-
nized as directly related to the Emiran (Veermersch
2009).

Quartär 61 (2014) J. I. Rose & A. E. Marks
74
Out of Arabia?
From an ecological perspective, hunter-gatherer
populations in Arabia are the most likely candidates to
have contributed to the development of the Emiran,
as there are no natural physiographic borders
separating nomadic groups in the Arabian Peninsula
from those in the southern Levant. Given their location,
should we consider sites such as ‘Ain Difla, Tor Faraj,
and Tor Sabiha as Levantine or Arabian? Certainly,
MP inhabitants of the region were not aware of
crossing from the Arabian Peninsula into the Levant at
the Jordanian border.
By MIS 5.3, perhaps as early as the Last Interglacial,
the African Nubian Complex was widespread in
Arabia, from the Yemeni Hadramawt to the eastern
edge of the Nejd plateau in southern O man. Additional
manifestations of this technocomplex have been found
in the Rub’ al Khali, central Saudi Arabia, and the Al
Jawf basin of northern Saudi Arabia, less than 300 km
southeast of ‘Ain Difla. While most Nilotic Nubian
Complex assemblages tend to have low laminar indices
under 13, the Classic Dhofar Nubian exhibits a wide
range (Fig. 14), overlapping with both low levels of
elongation in Africa (e.g., Aybut ath-Thani and Mudayy
as-Sodh 1), and the more elongated southern
Levantine Mousterian assemblages (e.g., Jebel Sanoora
1, TH.377, Jebel Markhashik 1). Like Africa, the
frequency of MP versus UP tool types is roughly
equivalent, skewed somewhat toward the MP (Fig. 13).
Differing from the African Nubian Complex, however,
Levallois point production is far more common,
although not quite as prevalent as found in the
southern Levant. These Arabian variations on the
Nubian Complex suggest that, even in MIS 5, techno-
logical and typological differentiation was occurring
between Africa and Arabia.
At least five distinct assemblage types have been
documented in Arabia from MIS 5.1 to early MIS 3
(ca. 90 - 50 ka). In the central and southwestern parts
of the Peninsula, both surface and buried sites have
been found around the Jubbah and Mundafan palaeo-
lakes. Although these assemblages have inadequate
tool and core counts to discern detailed technological
or typological patterns, it is apparent that none is
related to the Afro-Arabian Nubian Technocomplex,
and none is a likely precursor of the Emiran. Rather,
the predominantly radial forms of core reduction
resemble those found in the roughly contemporary
Tabun C-type assemblages from the Levant (Crassard
& Hilbert 2013: 1-2), while the short, broad based
Levallois points struck from unidirectional converging
flaking surfaces at Jebel Katefeh 1 are far more
reminiscent of Tabun B-type Levallois points (ibid.)
than anything coeval in northeast Africa.
In eastern Arabia, Assemblages A and B from Jebel
Faya are sufficiently similar to one another to think
they are temporally different manifestations of the
same, as yet, undefined industry. The absence of any
purposeful Levallois or volumetric blade technologies
in either, among other reasons, leads to a reasonable
interpretation that they are strictly local (Marks 2009),
perhaps representing an autochthonous development
associated with the Gulf “Oasis” region of eastern
Arabia (Rose 2010). Consequently, they can be
discounted from this discussion.
The stratified Wadi Surdud assemblages in
western Yemen, dating between 60 and 40 ka, are
radically different from those in central Arabia and
around the Gulf. Delagnes et al. (2012) observe that
the use of an unfaceted, unidirectional-convergent
reduction strategy to produce large numbers of
elongated blanks is superficially reminiscent of those
Tabun D-like southern Levantine Mousterian assem-
blages with high laminar indices, however, they caution
that:
“the Levallois debitage at Wadi Surdud differs signifi-
cantly, both qualitatively and quantitatively, from the
Levantine Levallois debitage. Some affinities between
the SD1 assemblage and the Levantine Mousterian are
still plausible, but would relate more to indirect
temporal and geographical connections between the
two regions, rather than any direct cultural affiliation”
(ibid.: 20).
While SD1 has some Levantine technological
proclivities, the absence of bidirectional reduction,
cresting, and true Levallois points makes it an unlikely
antecedent of the Emiran. Moreover, the SD1 date
ranges overlap with Boker Tachtit, Level 1; it is not
necessarily earlier than the Emiran. That is not to say,
however, that we discount SD1 as representing some
descendant form of the Arabian Nubian Complex.
Virtually nothing is known of the MP in western Arabia,
where any connections b etween Levantine Mousterian
and the Wadi Surdud assemblages might be found.
There are reports of numerous MP and UP sites
mapped by the Comprehensive Archaeological
Survey of Saudi Arabia along the flanks of the Asir
mountain chain, as well as on the Red Sea coastal plain
(Adams et al. 1977; Zarins et al. 1979, 1980, 1981, 1982;
Zarins & Zahrani 1985; Zarins & al-Badr 1986), which
are likely to be fruitful areas of future research.
As discussed in the preceding section, there is
ample evidence indicating that the Mudayyan Industry
is a late aspect of the Classic Dhofar Nubian (Usik et al.
2013; Hilbert et al. 2014). In comparison to the techno-
logical shifts seen in the late Nubian Complex
industries of the Nile Valley (e.g., Taramsan, Safahan,
Khormusan), however, the changes that brought about
the Mudayyan follow a different trajectory. Mudayyan
toolmakers continued to use the Nubian Levallois
strategy, but to produce diminutive points.
Additionally, there was a modification to the prefe-
rential Nubian Levallois system that de-emphasized
the formation of a pronounced median distal ridge, in
favor of a recurrent, bidirectional flaking strategy.
Such cores produced points and blades struck from
opposed, faceted platforms across both a broad
working surface, as well as from the narrow edges

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
75
adjacent to the primary flaking surface. At Jebel
Dahsha and Umm Mudayy 1 and 2, crested blades (2 %
- 7 %) and pre-cores with crest preparation (1 % - 13 %)
were identified among the cores and debitage,
suggesting the presence of a cresting technique in the
Mudayyan (Rose et al unpubl. data). The emergence
of these technologies may well have been an adaptive
response to the more challenging environmental
conditions that beset Arabia after 75 ka that required
smaller armatures and a more efficient flaking strategy.
Given the statistically significant difference between
the large Levallois points that are found in the Classic
Dhofar Nubian, and the diminutive variants found in
Mudayyan assemblages (Usik et al. 2013), it is tempting
to associate this change with the shift in projectile
hunting technology proposed by Sisk and Shea (2011).
Although not an exact match, Mudayyan technological
(Fig. 13) and typological (Fig. 14) patterning is closer to
the early Emiran than to any other assemblage type
considered in this study.
With such parallels between the Mudayyan and
the early Emiran, is it reasonable to claim one led
directly to the other? Not at this time, as the presently
known distribution of the Mudayyan is far from the
Negev. Taking into account the age of ‘Ain Difla, Levels
6 - 20, it is more likely that the Emiran derived, in part,
from a northern Arabian variant of the Nubian
Complex that had expanded into the region by MIS
5.3, if not during MIS 5.5. The recent discovery of
surface scatters with Nubian Levallois technology in
northern Saudi Arabia, just under 300 km southeast of
‘Ain Difla, warrants additional research in this area.
There is also evid ence of non-Nubian demograp hic
interactions between Arabia and the Levant in the
Late MP. At the Jubbah palaeolake, Tabun C-like and
B-like technological features (Crassard & Hilbert 2013)
suggest either cultural diffusion, or southward forays
of Levantine Mousterian groups at times of optimal
climatic conditions. It is likely that there were other
such demographic/cultural overlaps, but still little is
known of the Palaeolithic in northern and western
Arabia.
The Emiran through the admixture prism
The techno-typological patterns we have observed
point to an origin of the Emiran that was neither wholly
rooted in the Levant nor the result of a complete
demographic replacement from groups expanding
out of Africa; rather, the Emiran combines elements of
the Nubian Levallois system with typological elements
from the southern Levantine Mousterian. Our
proposed scenario envisions a zone stretching across
the interface of northwestern Arabia and the southern
Levant, where the territories of Levantine and Arabian
hunter-gatherer populations overlapped during MIS 5.
Bilateral exchange over time resulted in the incorpo-
ration of an Afro-Arabian core reduction strategy
with a Levantine tool-making tradition that extends
back to the Early Mousterian. At the same time, some
late MP Arabian tool-making traditions show a trend
toward Levantine characteristics: the adoption of a
unidirectional-convergent reduction strategy for
blade and point production at the Wadi Surdud
Complex; the production of short, broad-based
Levallois points at Jebel Katefeh 1; and, Tabun C-like
centripetal Levallois reduction strategies around the
Mundafan and Jubbah palaeolakes. Neither the
interior Nubian Complex findspots at Al-Jawf and Al
Kharj, nor the Mudayyan assemblages from the
geographically isolated Dhofar refugium absorbed
any discernable Levantine traits. Rather, the Mudayyan
appears to exclusively be a late expression of the
Nubian Complex - a sympatric development alongside
the Emiran, Safahan, Aterian, and Taramsan Industries.
Palaeoanthropogical evidence suggests African
Nubian Complex toolmakers were modern humans.
AMH specimens have been documented in North
Africa from 150 ka onward (Smith et al. 2007b; Hublin
& McPherron 2012), while no other species has yet
been found there. An AMH child burial was excavated
from an extraction pit associated with Activity Phase
III at Taramsa Hill 1 (Vermeersch et al. 1998; Van Peer
et al. 2010), with a terminus post quem of ca. 70 ka.
Although there is some question as to whether this
specimen was intrusive from a later occupation at the
site, the overlying assemblages from Activity Phases IV
and V are both later stages of the Nubian Complex,
therefore, would suggest demographic continuity at
the site. Given the technological similarity of the
Classic Dhofar Nubian and the Late Nubian Complex
of the Middle Nile Valley in Egypt, as well as the fact
that there is no evidence for prior MP human
occupation in southern Oman, it is reasonable to
associate the distribution of Nubian Complex sites
with a population of AMHs spread across northeast
Africa and the Arabian Peninsula. The presence of
Nubian Levallois technology in southern Arabia and
the Horn of Africa (Clark 1954; Kurashina 1978; Clark
1988; Beyin 2013), as well as northern Arabia and in
the Red Sea hills of Egypt, suggests that early human
groups could have traveled to and from Africa via
both the Arabian and Levantine Corridors.
In the Levant, the taxonomies of Levantine
Mousterian toolmakers are not easy to distinguish. It
was previously thought that Tabun C toolmakers were
AMHs, and that their presence at Skhul and Qafzeh
provided a specific route and window of time for an
African dispersal between 110 - 90 ka (e.g., Bar-Yosef
1987, 1994, 2000; Tostevin 2000). Just as the Tabun
sequence has been found to be non-linear and not
pan-Levantine, the classification of Neanderthal and
modern human specimens as distinct species in the
Levant has been questioned by some researchers
(e.g., Trinkaus 1986; Clark & Lindly 1989). For nearly
thirty years, scholars have urged the decoupling of
Tabun industries to distinguish AMH and Neander-
thals (e.g. Ahrensburg & Belfer-Cohen 1998; Kaufman
2001; Hovers & Belfer-Cohen 2013).

Quartär 61 (2014) J. I. Rose & A. E. Marks
76
Our scenario is in agreement with the proposition
of an intermediate stage of expansion out of Africa,
which led to the partial divergence of African and
non-African populations between 120 - 60 ka,
followed by a later divergence between Europeans
and Asians after 60 ka (Scally & Durbin 2012; Schiffels
& Durbin 2014). The current body of ancient DNA
evidence indicates mult iple admixture events betwe en
AMH and other archaic species (e.g., Fu et al. 2014;
Hellenthal et al. 2014; Sankararam et al. 2014).
Modeling the the per iod of genetic exchange between
AMHs and Neanderthals, Sankarararm et al. (2012)
propose a window between 86 - 37 ka, while Fu et al.
(2014) further refine this timeframe to 60 - 50 ka, using
a fully sequenced AMH femur from Ust’ Ishim in
western Siberia. Stringer (2012: 198-199) also
considers the Near East as the most likely place of
genetic exchange, but questions whether there could
have been an earlier period of mixing:
“…the interbreeding might even have happened when
people like those from Skhul-Qafzeh and Tabun were
in the Middle East 120,000 years ago. If a thousand of
those early moderns mixed with just fifty Neander-
thals and then survived somewhere in Arabia or North
Africa, could they have subsequently interbred with
the Out of Africa emigrants 60,000 years later, and
passed on their hidden component of Neanderthal
genes?”
The initial dispersal of Nubian Complex toolmakers
out of Africa seems to have occurred during early MIS 5,
perhaps in response to the well-documented
“greening” of the Saharo-Arabian phytogeographic
zone (e.g., Anton 1984; McClure 1984; Sanlaville 1992;
Rose 2000, 2006, 2007; Parker & Rose 2008;
Rosenberg et al. 2011; Drake et al. 2013). Although our
understanding of the timing and extent of Late
Pleistocene environmental oscillations remains
somewhat coarse, evidence from speleothems growth
(Vaks et al. 2003, 2007, 2010; Frumkin et al. 2008,
2009, 2011; Fleitmann & Matter 2009; Sorin et al.
2010; Fleitmann et al. 2011; Ayalon et al. 2013), fluvial
deposits (McClaren et al. 2008; Abouelmagd et al.
2014), dune formation (Preusser et al. 2002; Radies et
al. 2004; Stokes & Bray 2005; Preusser 2009), deep sea
cores (Bar-Matthews et al. 2003, 2006; Langgut et al.
2011), and lacustrine and tufa deposits (Smith et al.
2007a; Torfstein et al. 2009; Brookes 2010; Petit-Maire
et al. 2010; Waldmann et al. 2010; Rosenberg et al.
2011, 2012; Parton et al. 2013; Barrows et al. 2014)
indicates that North Africa, Arabia, and the Levant all
experienced greatly increased precipitation during
the Last Interglacial.
After the Last Interglacial, however, the magnitude
of rainfall supplied by Atlantic-Mediterranean
Westerlies versus Indian Ocean monsoon regimes
became asynchronous (Fig. 15), resulting in variable
conditions that may have had both “pushing” and
“pulling” effects on hunter-gatherer mobility patterns
in these regions. Nor th Africa experi enced widespread
drying out after 115 ka (Carto et al. 2009; Blome et al.
2012; Drake et al. 2013); in contrast, speleothems
(Fleitmann et al. 2011) and lacustrine deposits
(Rosenberg et al. 2011, 2012) from southern Arabia
are indicative of a humid episode associated with
sub-stage 5.3 (110 - 100 ka). In the Negev, speleothem
growth did not resume until ca. 90 ka (Vaks et al. 2007,
2010; Frumkin et al. 2009, 2011), while Lake Amora
(Dead Sea) levels show low stands throughout all of
MIS 5 (Torfstein et al. 2009; Waldmann et al. 2010).
By MIS 5.1 (90 - 75 ka), there was, again, greater
activity in the Indian Ocean monsoon, and, to a lesser
extent, Mediterranean storms. Pluvial proxy signals
are found throughout Arabia, northeast Africa, and
the southern Levant. The interior basins of Arabia,
such as Jubbah (Petraglia et al. 2012), Al Jafr (Davies
2005), Khujaymah and Mundafan (Rosenberg et al.
2011), and Mudawwara (Petit-Maire et al. 2010), all
exhibit lacustrine deposition, while cave speleothems
in the Negev indicate a short pulse of humidity. Given
these widespread and generally favorable conditions,
one might expect the expansion of multiple hunter-
gatherer territories and the resulting cultural/genetic
exchange between different groups coming into
contact with one another. The extent to which the
climate deteriorated in the intervening sub-stages 5.4
(120 - 110 ka) and 5.2 (100 - 90 ka) is uncertain, as
sediment accumulation was quite limited. Some
records of dune accumulation, however, point to a
dramatic increase in aridity at these times (e.g., Radies
et al. 2004; Preusser 2009).
During MIS 4 (75 - 60 ka), Frumkin et al. (2008: 365)
observe that “the depositional periods at Oman,
within the South Arabian desert, are almost all chrono-
logically distinct from the wet periods of the North
Arabian desert.” Rainfall regimes across the Arabian
Peninsula and the Levant have a negative correlation
at this time. Arabia was beset by prolonged aridity
caused by the southward displacement of the Inter-
tropical Convergence Zone; records from the Arabian
Sea indicate a period characterized by cooler sea
surface temperatures, low productivity, and increased
aeolian input (e.g., Reichart et al. 1997; Pattan & Pearce
2009; Banakar et al. 2010), corroborated by studies of
dune formation (e.g., Radies et al. 2004; Stokes & Bray
2005; Preusser 2009). Conversley, to the north, Lake
Amora shows a substantial highstand (Waldmann et al.
2009) and speleothem records from the northern
Negev indicate increased humidity (Vaks et al. 2006;
Frumkin et al. 2011). This disparity between rainfall in
the north and south of the Peninsula might have drawn
Arabian Nubian Complex toolmakers northward, into
the region affected by Mediterranean precipitation.
This process might have also been affected by a
subsequent wet pulse documented in eastern and
central Arabia between approximately 60 and 50 ka
(McLaren et al. 2008; Parton et al. 2013), which,
presumably, would have again triggered hunter-
gatherer range expansions.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
77
We suggest that over tens of thousands of years,
genetic and cultural information mingled across
expanding and contracting contextual areas in Arabia
and the southern Levant, driven by the asynchronous
rhythm of Indian Ocean, Atlantic, and Mediterranean
weather patterns. These bidirectional corridors of
exchange could have extended into Africa as well,
hinted at by the seemingly intrusive Hargeisan assem-
blages in the Horn of Africa (Clark 1954; Rose & Usik
2009) and the genetic signature of a back migration
from the Levant into northeast Africa during the early
Upper Palaeolithic (e.g., Cruciani et al. 2002; Olivieri
et al. 2006). We consider the proposed zone of inter-
action in northwestern Arabia as overlapping
“contextual areas,” defined as “an array of adaptive
relationships between natural and socio-cultural
factors within a human habitat” (Weissmüller 1995:
53-57). Such interactions between the contextual
areas presented in this paper might resemble the
meteorological maps of Bantu migration patterns
Fig. 15. Sum probability curves showing the likelihood of wet conditions from MIS 6 to MIS 2 in the northern Negev (Frumkin et al. 2011),
central-southern Negev (Frumkin et al. 2011), the Arabian Peninsula (Drake et al. 2013), and North Africa (Drake et al. 2013).
Abb. 15. Die Kurven der Summenwahrscheinlichkeit stellen die Wahrscheinlichkeit von Feuchtphasen im Zeitraum zwischen MIS 6 und MIS 2 in
der nördlichen Wüste Negev (Frumkin et al. 2011), in der zentral-südlichen Wüste Negev (Frumkin et al. 2011), der Arabischen Halbinsel (Drake
et al. 2013), und Nordafrika (Drake et al. 2013)

Quartär 61 (2014) J. I. Rose & A. E. Marks
78
documented in northern Namibia (Richter et al. 2012:
7), which chart movement not with vector arrows, but
with shifting high and low pressure systems.
For this reason, we will probably never find a
single, direct antecedent for the Emiran. Rather, the
Emiran should be viewed as one manifestation of a
widespread, long-term transformation from Nubian
Levallois technology to recurrent, bidirectional,
elongated point producing technologies that took
place from early MIS 5 to early MIS 3. This is evident in
the Taramsan, Safahan, Mudayyan, and Emiran indus-
tries. The ultimate success of the Emiran (as opposed
to the other three Nubian Complex-derived indus-
tries), which eventually transformed into a true UP and
spread northward into the territories of modern-day
Lebanon and Turkey, seems to rest on relatively
favorable environmental conditions in the southern
Levant throughout MIS 4 and MIS 3 (Fig. 15).
This model of Arabian-Levantine interaction
during the late MP can be verified by archaeological
research in northwestern Arabia and southern Jordan.
Late MP assemblages found within the overlapping
Afro-Arabian Nubian Complex and southern
Levantine contextual areas should exhibit techno-
typological features resembling the early Emiran:
bidirectional core preparation, Levallois point
production, crested blades, a predominance of UP
tools, and, perhaps, even Emireh points. MP surface
sites found around palaeolakes in this zone, such as the
Jafr basin (Davies 2005), Mudawwara depression
(Petit-Maire et al. 2010), and Al Jawf basin (Hilbert et
al. unpubl. data), are all promising starting points for
this endeavor.
Acknowledgements: We are grateful to the Dhofar Archaeo-
logical Project (DAP) team including Dr. Vitaly Usik, Dr. Yamandú
Hilbert, and Mr. Mohammed Jaboob for their integral role in
mapping, studying, and illustrating the Classic Dhofar Nubian
and Mudayyan lithic assemblages. We thank Mrs. Jeanne Marie
Geiling for translating the title, abstract, and captions into
German and Dr. Yamandú Hilbert, Prof. Erella Hovers, and Prof.
Chris Stringer for their valuable feedback while preparing the
manuscript. We are grateful to the Ministry of Heritage and
Culture in Oman for granting permission to carry out this
archaeological fieldwork and providing logistical support for
DAP. From 2009 to 2012, DAP was funded by a UK Early Career
Research Grant (AH/G012733/1) awarded to J. Rose and a 2013
National Geographic Waitt Grant awarded to J. Rose.
Literature cited
Adams, R., Parr, P., Ibrahim, M. & al-Mughannum, A. S. (1977).
Preliminary report on the first phase of the Comprehensive
Survey Program. Atlal 1: 21-40.
Abouelmagd, A., Sultan, M., Sturchio, N. C., Soliman, F.,
Rashed, M., Ahmed, M., Kehew, A. E., Milewski, A. &
Chouinard, K. (2014). Paleoclimate record in the Nubian
Sandstone Aquifer, Sinai Peninsula, Egypt. Quaternary Research
81: 158 -167.
Al Nafie, A. H. (2008). Phytogeography of Saudi Arabia. Saudi
Journal of Biological Sciences 15: 159-176.
Alves, I., Hanulová, S., Foll, M. & Excoffier, L. (2012). Genomic
data reveal a complex making of humans. PLoS Genetics 8 (7):
e10 02 8 37.
Ahrensburg, B. & Belfer-Cohen, A. (1998). Sapiens and
Neanderthals: Rethinking the Levantine Middle Palaeolithic
Hominids. In: T. Akazawa, K. Aoki, O. Bar-Yosef (Eds.)
Neanderthals and modern Humans in Western Asia. New York,
Plenum Press, 311-322.
Armitage, S. J., Jasim, S. A., Marks, A. E., Parker, A. G., Usik, V. I.
& Uerpmann, H.-P. (2011). The southern route “Out of
Africa”: evidence for an early expansion of modern humans into
Arabia. Science 331: 453-456.
Aton, D. (1984). Aspects of geomorphological evolution:
paleosols and dunes in Saudi Arabia. In: A. R. Jado, & J. G. Zötl
(Eds.) Quaternar y period in S audi Arabi a, Vol. II, Sedim entological ,
Hydrogeological, Hydochemical, Geomorphological, and
Climatological Investigations of Western Saudi Arabia. Springer,
Vienna, 275-296.
Ayalon, A., Bar-Matthews, M., Frumkin, A. & Matthews, A.
(2013). Last Glacial warm events on Mount Hermon: the
southern extension of the Alpine karst range of the east
Mediterranean. Quaternary Science Reviews 59: 43-56.
Azoury, I. (1986). Ksar Akil, Lebanon. A Technological and
Typological Analysis of the Transitional and Early Upper
Paleolithic Levels of Ksar Akil and Abu Halka. BAR International
Series 289, Oxford.
Bailey, G. N. (2009). The Red Sea, coastal landscapes and
hominin dispersals. In: M. Petraglia & J. I. Rose (Eds.) The
evolution of human populations in Arabia: Paleoenvironments,
Prehistory, and Genetics. Springer Academic Publishers,
Netherlands, 15-37.
Banakar, V. K., Mah esh, B. S., Burr, G. & C hodankar, A. R. (2 010).
Climatology of the Eastern Arabian Sea during the last glacial
cycle reconstructed from paired measurement of foraminiferal
δ18O and Mg/Ca. Quaternary Research 73: 535-540.
Bar-Matthews, M., Ayalon, A., Gilmour, M., Matthews, A. &
Hawkesworth, C. J. (2003). Sea-land oxygen isotopic
relationships from planktonic foraminifera and speleothems in
the eastern Mediterranean region and their implication for
paleorainfall during interglacial intervals. Geochimica et
Cosmochimica Acta 67: 3181- 3199.
Barrows, T. T., Williams, M. A. J., Mills, S. C., Duller, G. A. T.,
Fifield, L. K., Haberlah, D., Tims, S. G. & Williams, M. F.
(2014). A White Nile megalake during the last interglacial
period. Geology 42: 163 -166 .
Bar-Yosef, O. (1980). Prehistory of the Levant. Annual Review of
Anthropology 9: 101-133.
Bar-Yosef, O. (1987). Pleistocene connections between Africa
and Southwest Asia: an archaeological perspective. Af rican
Archaeological Review 5: 29-38.
Bar-Yosef, O. (1994). Th e Contribution s of Southwest Asia to the
study of the origin of modern humans. In: M. H. Nitecki & D. V.
Nitecki (Eds.) Origins of Anatomically Modern Humans. Plenum
Press, New York, 23-66.
Bar-Yosef, O. (1998). The Chronology of the Middle Palaeolithic
of the Levant. In: T. Akazawa, K. Aoki & O. Bar-Yosef (Eds.)
Neander tals and Modern Humans in Western Asia. Plenum Press,
New York, 39-56.
Bar-Yosef, O. (2000). T he Middle and earl y Upper Palaeolithic i n
southwest Asia and neighboring regions. In: O. Bar-Yosef & D.
Pilbeam (Eds) The Geography of Neanderthals and Modern
Humans in Europe and the greater Mediterranean. Peabody
Museum, Harvard University, Cambridge, 107-156.
Bar-Yosef, O. & Belfer-Cohen, A. (1988). The Early Upper
Paleolithic in Levantine Caves. In: J. F. Hoffecker & C. A Wolf
(Eds.) The Early Upper Paleolithic: evidence from Europe and the
Near East. British Archaeological Reports, International Series
437, Oxford, 23-42.
Beyin, A. (2011). Upper Pleistocene Human Dispersals out of
Africa: A Review of the Current State of Debate. International
Journal of Evolutionary Biology 2011: 1-17.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
79
Beyin, A. (2013). A surface Middle Stone Age assemblage from
the Red Sea coast of Eritrea: Implications for Upper Pleistocene
human dispersals out of Africa. Quaternary International 300:
195 -212 .
Biagi, P. (1994). An Early Palaeolithic site near Saiwan (Sultanate
of Oman). Arabian Archaeology and Epigraphy 5: 81-88.
Biagi, P. & Starnini, E. (2011). Neanderthals at the South-
Easternmost Edge: The Spread of Levalloisian Mousterian in the
Indian Subcontinent. In: K. Biró & M. András (Eds.) Emlékkħönyv
Violának: Papers in Honor of Viola T. Dobosi. Hungarian National
Museum, Budapest, 5-14.
Biglari, F., Javeri, M., Mashkour, M., Yazdi, M., Shidrang, S.,
Tengberg, M. , Taheri, K . & J. Darvi sh (2009). Test excavations
at the Midd le Paleolithic sites o f Qaleh Bozi, southwest of ce ntral
Iran, a preliminar y report. In: M. Ot te, F. Biglari & J. J aubert (Eds.)
Iran Paleolithic. BAR International Series S1968. Archaeopress,
Oxford, 29-38.
Blome, M. W., Cohen, A. S., Tryon, C. A., Brooks, A. S. &
J. Russell (2012). The environmental context for the origins of
modern human diversity: a synthesis of regional variability in
African climate 150,000-30,000 years ago. Journal of Human
Evolution 62 (5): 563-592.
Bordes, F. (1950). Principes d’une méthode d’étude des
techniques de débitage et de typologie du Paléolithique ancien
et moyen. L’Anthropologie 54 (1-2): 19-34.
Bordes, F. (1953). Levalloisien et Moustérien. Bulletin de la
Société Préhistorique Française 50 (4): 226-234.
Bordes, F. (1961). Typologie du paléolithique ancien et moyen.
Delmas, Bordeaux.
Brookes, A. I. (2010). Spatially variable sedimentary responses
to orbitall y driven pluvi al climate during M arine Oxygen I sotope
Stage 5.1, Dakhla Oasis region, Egypt. Quaternary Research 74:
252-264
Bretzke, K., Armitage, S., Parker, A., Walkington, H. &
Uerpmann, H.-P. (2013). The environmental context of
Palaeolithic settlement at Jebel Faya, Emirate Sharjah, UAE.
Quaternary International 300: 83-93.
Bretzke, K ., Conard, N. J . & Uerpmann , H.-P. (2014). E xcavati ons
at Jebel Faya — The FAY-NE1 shelter sequence. Proceedings of
the Seminar for Arabian Studies 44: 69-82.
Carto, S. L ., Weaver, A. J ., Hetheringto n, R., Lam, Y. & Wiebe, E . C.
(200 9). Out of Africa and into an ice age: on the role of global
climate change in the late Pleistocene migration of early modern
humans out of Africa. Journal of Human Evolution 56: 139-151.
Caton-Thompson, G. (1952). Kharga O asis in Prehistor y. Athlone
Press, London.
Chiotti, L., Olszewski, D. I., Dibble, H. L., McPherron, S. P.,
Schurmans, U. A. & Smith, J. R. (2007). Paleolithic Abydos:
reconstructing individual behaviors across the high desert
landscape. In: Z. Hawass & J. Richards (Eds.) The Archaeology
and Art of Ancient Egypt: Essays in Honor of David B. O’Connor.
Supreme Council of Antiquities Press (distributed by American
University in Cairo Press), Cairo, 169-183.
Chiotti, L., Dibble, H. L., Olszewski, D. I., McPherron, S. P. &
Schurmans, U. A. (2009). Lithic technologies of the western
high desert in Egypt (Abydos, Egypt). Journal of Field
Archaeology 34: 307-318.
Chmoel ewski, W. (1968). Early and Middle Palaeolithic Sites near
Arkin, Sudan. In: F. Wendorf (Ed.) The Prehistory of Nubia.
Southern Methodist University Press, Dallas, 110-147.
Clark, G . A. & Lindly, J. M . (1989). Modern Human Origins in the
Levant and Wester n Asia: The Foss il and Archeolog ical Evidence.
American Anthropologist 91: 962-985.
Clark, G. A., Schuldenrein, J., Donaldson, M. L., Schwarcz, P.,
Rink, W. J. & Fish, S. K. (1997). Chronostratigraphic contexts
of Middle Palaeolithic horizons at the ‘Ain Dif la rockshelter
(WHS 634), west-central Jordan. In: H. G. K. Gebel, Z. Kafafi &
G. O. Rollefson (Eds.) The Prehistory of Jordan, II. Perspectives
from 1997. Ex Oriente, Berlin, 77-100.
Clark, J. D. (1954). The Prehistoric Cultures of the Horn of Africa.
Cambridge University Press, Cambridge.
Clark, J. D. (1988). The Middle Stone Age of East Africa and the
beginnings of regional identity. Journal of World Prehistory 2:
235-305.
Close, A . (1993). BT-14, a stratified Middle Palaeolithic site at Bir
Tarfawi, Western Desert of Egypt. In: L. Krzyżaniak, M.
Kobusiewicz & J. Alexander (Eds.) Environmental Change and
Human Culture in the Nile Basin and Northern Africa until the
Second Millennium B.C. Poznań Archeological Museum, Poznań,
115 -12 2 .
Copeland, L. (1975). The Middle and Upper Palaeolithic of
Lebanon and Syria in the light of recent research. In: F. Wendorf
& A. E. Marks (Eds.) Problems in Prehistory: North Africa and the
Levant. Southern Methodist University Press, Dallas, 317-350.
Copeland, L. (2001). Forty-six Emireh points from the Lebanon
in the context of the Middle to Upper transition in the Levant.
Paléorient 26 (1): 73-92.
Crassard, R. (2007). Apport de la technologie lithique à la
définition de la Préhistorie du Hadramawt, dans le contexte du
Yémen et de lArabie du Sud. Ph.D., dissertation, Université Paris,
Paris.
Crassard, R. (2009). The Middle Palaeolithic of Arabia: the view
from the Hadramawt region, Yemen. In: M. Petraglia & J. I. Rose
(Eds.) The Evolution of Human Populations in Arabia:
Paleoenvironments, Prehistory and Genetics. Springer Academic
Publishers, Netherlands, 151-168.
Crassard, R. & Thiébaut, C. (2011). Levallois points production
from eastern Yemen and some comparisons with assemblages
from East-Africa, Europe and the Levant. ERAUL 999, Liège, 1-14.
Crassard, R., Petraglia, M. D., Drake, N. A., Breeze, P., Gratuze,
B., Alsharekh, A., Arbach, M., Groucutt, H., Khalidi, L.,
Michelsen, N., Robin, C. J. & Schiettecatte, J. (2013). Middle
Palaeolithic and Neolithic Occupations around Mundafan
Palaeolake, Saudi Arabia: Implications for Climate Change and
Human Dispersals. PloS One 8: e69665.
Crassard, R. & Hilbert, Y. H. (2013). A Nubian Complex Site
from Central Arabia: Implications for Levallois Taxonomy and
Human Dispersals during the Upper Pleistocene. PloS One. 8:
e69221.
Crew, H. L. (1975). An evaluation of the Relationship between
the Mousterian Complexes of the Eastern Mediterranean: A
Technological Perspective. In: F. Wendorf & A. E. Marks (Eds.)
Problems in Prehistory: North Africa and the Levant. Southern
Methodist University Press, Dallas, 427-438.
Crew, H. L. (1976). The Mousterian site of Rosh Ein Mor. In:
A. E. Marks (Ed.) Prehistor y and Paleoenvironments in the Central
Negev, Israel, vol. I, The Avdat/Aqev Area, Part 1. Southern
Methodist University Press, Dallas, 75-111.
Cruciani, F., Sanolamazza, P., Shen, P., Macaulay, V., Moral, P.,
Olckers, A., Modiano, D., Holmes, S., Destro-Bisol, G., Coia,
V., Wallace, D. C., Oefner, P. J., Torroni, A., Cavalli-Sforza, L.
L., Scozzari, R. & Underhill, P. A. (2002). A Back Migration
from As ia to Sub-Sah aran Afric a Is Support ed by High- Resolution
Analysis of Human Y-Chromosome Haplotypes. Amer ican
Journal of Human Genetics 70: 1197–1214.
Davidzon, A. & Goring-Morris, N. (2003). Sealed in Stone: The
Upper Palaeolithic Early Ahmarian Knapping Method in the
Light of Refitting Studies at Nahal Nizzana XIII, Western Negev,
Israel. Journal of the Israel Prehistoric Society 33: 75-205.
Davies, P. C. (2005). Quaternary Paleoenvironments and
Potential for Human Exploitation of the Jordan Plateau Desert
Interi or. Geoarchaeology: An International Journal 20: 379-400.
Delagnes, A., Tribolo, C., Bertran, P., Brenet, M., & Crassard, R.
(2012). Inland human settlement in southern Arabia 55,000
years ago. New evidence from the Wadi Surdud Middle
Palaeolithic site complex, western Yemen. Journal of Human
Evolution 63 (3): 452- 474.

Quartär 61 (2014) J. I. Rose & A. E. Marks
80
Delagnes, A., Crassard, R., Bertran, P. & Sitzia, L. (2013).
Cultural and human dynamics in southern Arabia at the end of
the Middle Paleolithic. Quaternary International 300: 234-243.
Demidenko, Y. & Usik, V. (1993). The problem of changes in
Levallois technique during the technological transition from the
Middle to the Upper Palaeolithic. Paléorient 19 (2): 5-15.
Demidenko Y. & Usik, V. (2003). Into the mind of the maker:
Refitting study and technological reconstruction. In: D. Henry
(Ed.) Neanderthals in the Levant. Continuum, London, 107-155.
Dibble, H., Aldeias, V., Jacobs, Z., Olszewski, D., Rezek, Z., Lim,
S., Alverez-Fernández, E., Barshay-Szmidt, C., Hallett-
Desguez, E., Reed, D., Reed, K., Richter, D., Steele, T., Skinner,
A., Blackwell, B., Doronicheva, E. & El-Hajraoui, M. (2013).
On the industrial attributions of the Aterian and Mousterian of
the Maghreb. Journal of Human Evolution 64 : 194-210.
Douka, K., Bergman, C. A., Hedges, R. E. M ., Wesselingh, F. P. &
Higham, T. F. G. (2013). Chronology of Ksar Akil (Lebanon)
and implications for the colonization of Europe by anatomically
modern humans. PLoS ONE 8 (9): e72931.
Drake, N. A ., Breeze, P. & Parker, A. (2013). Palaeoclimate in the
Saharan and A rabian Deserts dur ing the Middle Palaeolithic and
the potential for hominin dispersals. Quaternary International
30 0: 4 8-61.
Durand, E. Y., Patterson, N., Reich, D. & Slatkin, M. (2011).
Testing for ancient admixture between closely related
populations. Molecular Biology and Evolution 28: 2239-2252.
Ewing, J. F. (1947). Preliminary note on the excavations at the
Palaeolithic site of Ksar ‘Akil, Republic of Lebanon. Antiquity 84:
186 -196.
Fernandes, V., Alshamali, F., Alves, M., Costa, M. D., Pereira, J.
B., Silva, N. M., Cherni, L., Harish, N. Černý, V., Soares, P.,
Richards, M. B. & Pereira, L. (2012). The Arabian Cradle:
Mitochondrial relicts of the first steps along the southern route
out of Africa. American Journal of Human Genetics 90 (2):
347–355.
Ferring, R. (1975). The Aterian in North African Prehistory. In:
F. Wendorf & A. E. Marks (Eds.) Problems in Prehistory: North
Africa and the Levant. SMU Press, Dallas, 113-126.
Fleitmann, D. & Matter, A. (2009). The speleothem record of
climate variability in Southern Arabia. Comptes Rendus
Geoscience 341: 633 -642.
Fleitmann, D., Burns, S. J., Pekala, M., Mangini, A., Al-Subbary,
A., Al-Aowah, M., Kramers, J. & Matter, A. (2011). Holocene
and Pleistocene pluvial periods in Yemen, Southern Arabia.
Quaternary Science Reviews 30: 783-787.
Fleitmann, D., Burns, S. J., Neff, U., Mangini, A. & Matter, A.
(20 03). Changing moisture sources over the last 330,0 00 years
in northern Oman from fluid-inclusion evidence in speleothems.
Quaternary Research 60: 223-232.
Forster, P. & Matsumura, S. (2005). Evolution: Did early humans
go north or south? Science 308: 965-966.
Frumkin, A., Bar-Matthews, M. & Vaks, A. (2008).
Paleoenvironment of Jawa basalt plateau, Jordan, inferred from
calcite speleothems from a lava tube. Quaternary Research 70:
358-367.
Frumkin, A., Bar-Yosef, O. & Schwarcz, H. P. (2011). Possible
paleohydrologic and paleoclimatic effects on hominin migration
and occupation of the Levantine Middle Paleolithic. Journal of
Human Evolution 60: 437- 451.
Fu, Q., Li, H., Moorjani, P., Jay, F., Slepchenko, S. M., Bondarev,
A. A., Johnson, P. L. F, Aximu-Petri, A., Prüfer, K., de Filippo,
C., Meye r, M., Zwy ns, N., Salaz ar-García, D. C ., Yaroslav V. K.,
Keates, S. G., Kosintsev, P. A., Razhev, D. I., Richards, M. P.,
Peristove, N. V., Lachmann, M., Douka, K., Higham, T. F. G,
Slatkin, M., Hublin, J. -J., Reich, D., Kelso, J., Viola, T. B. &
Pääbo, S. (2014). Genome sequence of a 45,000-year-old
modern human from western Siberia. Nature 514: 445-450.
Garcea, E. (1999). Aterian and “Ea rly” and “Late Acacus” from the
Uan Tabu rockshelter, Tadrart Acacus (Libyan Sahara). In: M.
Cremaschi & S. Di Lernia (Eds.) Wadi Teshuinat:
Palaeoenvironment and Prehistory in south-western Fezzan
(Libyan Sahara). Edizioni All’Insegna del Gigllio, Florence,
155-182.
Garcea, E. (2001). A reconsideration of the Middle Palaeolithic/
Middle Stone Age in north Africa after the evidence from the
Libyan Sahara. In: E. Garcea (Ed.) Uan Tabu in the Settlement
history of the Libyan Sahara. Edizioni All’insegna del Giglio,
Rome, 69-96.
Garrod, D. A. E. (1951). A transitional industry from the base of
the Upper Palaeolithic in Palest ine and Syria. J ournal of the Royal
Anthropological Institute 81: 121-129.
Garrod , D. A. E. (1955). T he Mugharat el- Emireh in lower Galile e:
type station of the Emiran industry. Journal of the Royal
Anthropological Institute 85: 141-62.
Ghirotto, S., Penso-Dolfin, L. & Barbujani, G. (2011). Genomic
evidence for an African expansion of anatomically modern
humans by a southern route. Human Biology 83: 477- 489.
Gilead, I. (1985). The settlement patterns at Gebel Maghara
(northern Sinai) in the Upper Palaeolithic. Mitekufat Haeven 18:
67- 69.
Gilead, I. & Grigson, C. (1984). Farah II: a Middle Palaeolithic
open-air site in the northern Negev, Israel. Proceedings of the
Prehistoric Society 5 0: 71 -97.
Goren- Inbar, N. & Belfer-Cohen, A . (1998). The Abilities of the
Levantine Mousterians: Culture and Mental Capacities. In:
T. Akazawa, K. Aoki & O. Bar-Yosef (Eds.) Neanderthal and
Modern Human in Western Asia. Plenum Press, New York,
205-222.
Green, R. E., Krause, J., Briggs, A. W., Maricic, T. & Stenzel, U.
(2010). A draft sequence of the Neandertal genome. Science
328: 710 -722.
Groucutt, H. (2014). Middle Paleolithic point technology with a
focus on the site of Tor Faraj (Jordan, MIS 3). Quaternary
International 350 : 205-226.
Guichard, J. & Guichard, G. (1965). The Early and Middle
Palaeolithic of Nubia: Preliminary Results. In: F. Wendorf (Ed.)
Contributions to the Prehistory of Nubia. Fort Burgwin Research
Center and Southern Methodist University Press, Dallas, 57-116.
Guichard, J. & Guichard, G. (1968). Contributions to the Study
of the Early and Middle Palaeolithic of Nubia. In: F. Wendorf
(Ed.) The Prehistory of Nubia, vol. 1. Southern Methodist
University Press, Dallas, 148-193.
Hammer, M. F., Woerner, A. E., Mendez, F. L., Watkins, J. C . &
Wall, J. D. (2011). Genetic evidence for archaic admixture in
Africa. Proceedings of the National Academy of Sciences of the
USA 108: 15123 -15128.
Hauck, T. (2011). Mousterian technology and settlement
dynamics in the site of Hummal (Syria). Journal of Human
Evolution 61: 519- 537.
Hawks, J . & Wolpoff, M. (2001). T he four faces of Eve: hypot hesis
compatibility and human origins. Quaternary International 75:
41-50.
de Heinzelin, J. (1968). Geological History of the Nile Valley in
Nubia. In: F. Wendorf (Ed.) The Prehistory of Nubia, Vol. 1.
Southern Methodist University Press, Dallas, 19-55.
Hellenthal, G., Busby, J. B. J., Band, G., Wilson, J. F., Capelli, C.,
Falush, D. & Myers, S. (2014). A Genetic Atlas of Human
Admixture History. Science 343: 747-751.
Henry, D. (1979). Palaeolithic sites within the Ras en Naqb Basin,
Southern Jordan. Palestine Exploration Quarterly 3: 79-85.
Henry, D. (1982). The Prehistory of Southern Jordan and
relationships with the Levant. Journal of Field Archaeology 9 (4):
417-444.

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
81
Henry, D. (1983). Adaptive Evolution within the Epipaleolithic of
the Near East. In: F. Wendorf & A. Close (Eds.) Advan ces in World
Archaeology, vol. 2. Academic Press, New York, 99-160.
Henry, D. (1985). Late Pleistocene environment and Palaeolithic
adaptions in the Southern Levant. In: A. Hadidi (Ed.) Studies in
the History and Archaeology of Jordan, vol. 2. Routledge and
Kegan Paul, London, 67-78.
Henry, D. (1988). Summary of prehistoric and paleoenvironmental
research in the Northern Hisma. In: A. Garrod & H. Gebel (Eds.)
The Prehistory of Jordan: The State of the Research in 1986. BAR
International Series 396 (1), Oxford, 7-37.
Henry, D. (1992). Transhumance During the Late Levantine
Mousterian. In: H. Dibble & P. Mellars (Eds.) The Middle
Paleolithic: Adaptation, Behavior, and Variability. University
Museum, University of Pennsylvania, Philadelphia, 143-162.
Henry, D. (1994). Prehistoric cult ural ecology in southern Jordan.
Science 265: 336-341.
Henry, D. (1995). Prehistoric Cultural Ecology and Evolution:
insights from southern Jordan. Plenum Press, New York.
Henry, D. (1997). Cultural and Geologic Successions of Middle
and Upper Palaeolithic Deposits in the Jebel Qalkha Area of
southern Jordan. In: H. Gebel, Z. Kafafi & G. O. Rollefson (Eds.)
The Preh istory of Jorda n II: Perspecti ves from 1996. Studies in E arly
Near Eastern Production, Subsistence, and Environment 4.
Ex Oriente, Berlin, 69-76.
Henry, D. (1998). The Middle Palaeolithic of Jordan. In: D. Henry
(Ed.) The Prehistoric Archaeology of Jordan. BAR International
Series 705, Oxford, 23-39.
Henry, D. (2003). Neander tals in the Levant. Continuum, London.
Hester, J. & Hobler, P. (1969). Prehistoric Settlement Patterns in
the Libyan Desert. Anthropological Papers 42 (4). University of
Utah Press. Salt Lake City
Hilber t Y. H., Marks, A . E., Usik, V. I. & Rose J . I. (2014). Grasping
temporal variability and landscape evolution: spatial analysis of
surface s catters from the N ejd Plateau, Oma n. Paper presented at
UISPP Session A21C, Burgos, Spain, September 4, 2014.
Hovers, E. (1998). The lithic assemblages of Amud Cave:
implications for the end of the Mousterian in the Levant. In:
T. Akazawa, K. Aoki & O. Bar-Yosef (Eds.) Neandertals and
Modern Humans in Southwest Asia. Plenum Press, New York,
143-163.
Hovers, E. (2009). Lithic Assemblages of Qafzeh Cave. Oxford
University Press, Oxford.
Hovers, E. & Belfer-Cohen, A. (2013). On Variability and
Complexity: Lessons from the Levantine Middle Paleolithic
Record. Current Anthropology 54: 337-357.
Howell, F. C. (1959). Upper Pleistocene stratigraphy and early
man in the Levant. Proceedings of the American Philosophical
Society 103 (1): 1-65.
Hublin, J.-J. (2000). Modern-nonmodern Hominid interactions:
A Mediterranean perspective. In: O. Bar-Yosef & D. Pilbeam
(Eds.) The Geography of Neandertals and Modern Humans in
Europe and the Greater Mediterranean, Peabody Museum of
Archaeol ogy and Ethnolog y, Bullet in No. 8, Cambrid ge, 157-182.
Hublin, J. - J. & McPherron, S. (2012). Modern Origins: A North
African Perspective. Springer, New York.
Huxtable, J. (1993). Thermoluminescence dates for burnt earth
samples from Bir Sahara East and a burnt core from Bir Tarfawi.
In: F. Wendorf, R. Schild & A. Close (Eds.) Egypt During the Last
Intergla cial: The Middle Pal aeolithic of Bir Tarfaw i and Bir Sahara
East. Plenum Press, New York and London, 227-228.
Inizan, M. L. & Ortlieb, L. (1987). Prehistoire dans la region de
Shabwa au Yemen du sud. Paléorient 13: 5 -22.
Jelinek, A. J. (1975). A Preliminary report on some Lower and
Middle Palaeolithic industries from the Tabun Cave. In:
F. Wendorf & A. E. Marks (Eds.) Problems in Prehistory. Southern
Methodist University Press, Dallas, 297-31.
Jelinek, A. J. (1981). The Middle Palaeolithic in the southern
Levant from the Perspective of the Tabun Cave. In: J. Cauvin &
P. Sanlaville (Eds.) Préhistoire du Levant. C.N.R.S., Paris, 265-285.
Kaufman, D. (2001). Comparisons and the Case for Interaction
Among Neanderthals and Early Modern Humans in the Levant.
Oxford Journal of Archaeology 20: 219-240.
Kivisild, T., Maere, R., Metspalu, E., Rosa, A., Antonio, A.,
Pennarun, E., Parik, J., Geberhiwot, T., Usanga, E. & Villems,
R. (2004). Ethiopian mitochondrial DNA heritage: tracking
gene f low across and around th e Gate of Tears. American Jour nal
of Human Genetics 75: 752- 7 70.
Kleindienst, M. (2006). On Naming Things: Behavioral Changes
in the Later Middle to Earlier Late Pleistocene, Viewed from the
Eastern Sahara. In: E. Hovers & S. Kuhn (Eds.) Transitions Before
th e Transition. Springer, New York, 13-28.
Kobusiewicz, M. (1999). Excavations at Sinai-20, The Split Rock
Site, Zarnoq Locality. An Archaeological Investigation of the
Central Sinai, Egypt. University Press of Colorado, Niwot.
Kobusiewicz, M., Schild, R., Bluszcz, A. & F. Wendorf (2001).
Reassessing chronostratigraphic position of the Split Rock Site,
Sinai. In: B. Gehlen, M. Heinen & A. Tillmann (Eds.) Zeit-Räume,
Gedenkschrift für Wolfgang Taute, Band 1, Bonn, 227-236.
Kuhn, S. L., Stiner, M. C. & Güleç, E. (1999). Initial Upper
Palaeolithic in south-central Turkey and its regional context: a
preliminary report. Antiquity 7 3: 505- 517.
Kuhn, S. & Zwyns, N. (2014). Rethinking the initial Upper
Paleolithic. Quaternary International 347: 29-38.
Kurashina, H. (1987). An Examination of Prehistoric Lithic
Technology in East-Central Etheopia. Ph.D., dissertation,
University of California, Berkeley.
Lahr, M. M. & Foley, R. A. (1994). Multiple dispersals and
modern human origins. Evolutionary Anthropology 3: 48-60
Langgut, D., Almogi-Labin, A., Bar-Matthews, M. & Weinstein-
Evron, M . (2011). Vegetation and climate changes in the South
Eastern Mediterranean during the Last Glacial-Interglacial cycle
(86 ka): new marine pollen record. Quaternary Science Reviews
30: 3960 -3972.
Leder, D. (2013). Technological and typological changes at the
Middle to Upper Palaeolithic boundary in Lebanon. Ph.D.,
dissertation, Universit y of Köln, Köln.
Lindly, J. M . & Clark, G. A. (1987). A Preliminary lithic analysis of
the Mousterian site of ‘Ain Difla (WHS site 634) in the Wadi Ali,
West Central Jordan. Proceedings of the Prehistoric Society 53:
279-292.
Lindly, J. & Clark, G. A. (2000). The ‘Ain Difla Rockshelter and
Middle Palaeolithic systematics in the Levant. In: N. R. Coinman
(Ed.) T he Archaeology of the Wadi Al- Hasa, West-Central J ordan,
Volume 2: Excavations at Middle, Upper, and Epipaleolithic Sites.
Arizona State University Anthropological Research Papers,
Tempe, 111-117.
Macaulay, V., Hill, C., Achilli, A.,Rengo, C., Clarke, D., Meehan,
W., Blackburn, J., Semino, O., Scozzari, R., Cruciani, F., Taha,
A., Sh aari, N. K ., Raja, J. M ., Ismail, P., Zainud din, Z., Go odwin,
W., Bulbeck, D., Bandelt, H. - J., Oppenheimer, S., Torroni, A.
& Richards, M. (2005). Single, rapid coastal settlement of A sia
revealed by analysis of complete mitochondrial genomes.
Science 308: 1034-1036.
Marks, A. E. (1968a). The Mousterian Industries of Nubia. In:
F. Wendorf (Ed.) The Prehistory of Nubia, Vol. 1. Southern
Methodist University Press, Dallas, 194-314.
Marks, A. E. (1968b). The Khormusan: An upper Pleistocene
Industry in Sudanese Nubia, In: F. Wendorf (Ed.) The Prehistory
of Nubia, Vol. 1. Southern Methodist University Press, Dallas,
315-391.
Marks, A. E. (1983a). Prehistory and Paleoenvironments in the
Central Negev, Israel, Vol. 3. Southern Methodist University
Press, Dallas.

Quartär 61 (2014) J. I. Rose & A. E. Marks
82
Marks, A. E (1983b). The Sites of Boker and Boker Tachtit: A
Brief Introduction. In: A. E. Marks (Ed.) Prehistory and
Paleoenvironments in the Central Negev, Israel, Vol. 3, the Avdat/
Aqev Area, Part 3. Department of Anthropology, Southern
Methodist University, Dallas, 15-37.
Marks, A. E. (1990). The Middle and Upper Palaeolithic of the
Near East and Nile Valley: the problems of cultural
transformations. In: P. Mellars (Ed.) The Emergence of Modern
Humans: An Archaeological Perspective. Edinburgh University
Press, Edinburgh, 57-80.
Marks, A. E. (1992). Upper Pleistocene Archaeology and the
Origins of Modern Man: A View from the Levant and Adjacent
Areas. In: T. Akazawa, K. Akoi, & T. Kimura (Eds.) The Evolution
and Dispersal of Modern Humans in Asia. Hokusen-sha, Tokyo,
229-252.
Marks, A. E. (2003). Reflections on Levantine Upper Palaeolithic
Studies: Past and Present. In: A. N. Goring-Morris & A. Belfer-
Cohen (Eds.) More then Meets the Eye. Oxbow Books, Exeter,
249-264.
Marks, A. E. (2009). The Palaeolithic of Arabia in an inter-
regional context. In: M. Petraglia & J. I. Rose (Eds.) Evolution of
Human Populations in Arabia: Paleoenv ironments, Prehi story and
Genetics. Springer, Dordrecht, 293-309.
Marks, A. E. & Crew, H. L. (1972). Rosh Ein Mor, Central Negev,
Israel. Current Anthropology 13: 591- 593.
Marks , A. E. & Ferri ng, C. R. (1988). The early Upper Palaeolithic
of the Levant. In: J . F. Hoffecker (Ed .) The Early Upper Palaeolithic:
evidence f rom Europe and the N ear East. BAR Inter national Series
437, Oxford, 43-72.
Marks, A. E. & Freidel, D. A. (1977). Prehistoric settlement
patterns in the Avdat/Aqev area. In: A. E. Marks (Ed.) Prehistory
and Paleoenvironments of the Central Negev, Israel, Vol. II: The
Avdat/Aqev Area, Part 2 and the Har Harif. Department of
Anthropology, Southern Methodist University, Dallas, 131-158.
Marks, A. E. & Kaufman, D. (1983). Boker Tachtit: The Artifacts.
In: A. E. Marks (Ed.) Prehistory and Paleoenvironments in the
Central Negev, Israel, Vol. III, the Avdat/Aqev Area, Part 3.
Department of Anthropology, Southern Methodist University,
Dallas, 69-126.
Marks, A. E. & Monigal, K. (1995). Modeling the Production of
Elongated Blanks from the Early Levantine Mousterian at Rosh
Ein Mor. In: H. Dibble & O. Bar-Yosef (Eds.) The Definition and
Interpr etation of Levallois Technol ogy. Prehistory Press, Madison,
267-268.
Marks, A. E., & Rose, J. I. (2014). A century of research into the
origins of the Upper Paleolithic in the Levant. In: M. Otte (Ed.)
Néandertal/Cro-Magnon: la Recontre. Editions Errance, Arles,
221-266 .
Marks, A. E. & Volkman, P. (1983). Changing core reduction
strategies: a technological shift from the Middle to the Upper
Palaeolithic. In: E. Trinkaus (Ed.) The Mousterian Legacy: Human
Biocultural Change in the Upper Pleistocene. BAR International
Series S164, Oxford, 35-51.
Marks, A. E. & Volkman, P. (1985). Technological variability and
change seen through core reconstruction. Proceedings of the 4th
International Flint Symposium, Vol. 1. Cambridge University
Press, Cambridge.
Marks, A. E. & Volkman, P. (1986). The Mousterian of Ksar Akil:
Levels XX VIA through XXVIIIB. Paléorient 12 (1): 5-20.
Marks, A. E., Shiner, J., & Hays, T. R. (1968). Survey and
excavations in the Dongol Reach, Sudan. Current Anthropology 9
(4): 319 -323.
McClure, H. A. (1984). Late Quaternary Palaeoenvironments of
the Rub’ al Khali. Ph.D., dissertation, University of London,
London.
McLaren, S. J., Al-Juaidi, F., Bateman, M. D. & Millington, A. C.
(2008). First evidence for episodic flooding events in the arid
interior of central Saudi Arabia over the last 60 ka. Journal of
Quaternary Science 24: 19 8-207.
Meignen, L. (1995). Levallois Lithic Production Systems in the
Middle Palaeolithic of the Near East: The Case of the
Unidirectional Method. In: H. Dibble, & O. Bar-Yosef (Eds.) The
Definition and Interpretation of Levallois Technology. Prehistory
Press, Madison, 361-380.
Meignen, L. (1998). Hayonim Cave Lithic Assemblages in the
Context of the Near Eastern Middle Palaeolithic: A Preliminary
Report. In: T. Akazawa, K. Akoi & O. Bar-Yosef (Eds.) Neandertals
and Modern Humans in Western Asia. Plenum Press, New York,
163-180 .
Meignen, L. (2011). The contribution of Hayonim cave
assemblages to the understanding of the so-called Early
Levantine Mousterian. In: J. M. Le Tensorer, R. Jagher & M. Otte
(Eds.) The Lower and Middle Palaeolithic in the Middle East and
Neighboring Regions. ERAUL 126, Liège, 85-100.
Meignen, L. (2012). Levantine Perspectives on the Middle to
Upper Palaeolithic “Transition.” Archaeology, Ethnology and
Anthropology of Eurasia 40 (3): 12-21.
Meignen, L. & Bar-Yosef, O. (1988). Variabilite technologique
au Proche Orient: L’Exemple de Kabara. In: L. Binford & J. P.
Rigaud (Eds.) L’Homme de Néandertal, Vol. 4 La Technique.
ERAUL 30, Liège, 81-95.
Meignen, L. & Bar-Yosef, O. (2005). The Lithic Industries of the
Middle and Upper Palaeolithic of the Levant: Continuity of
Break? In: A. P. Derevia nko (Ed.) The Middle to Upper Palaeolithic
Transition in Eurasia: Hypotheses and Facts. Institute of
Archaeology and Ethnology Press, Novosibirsk, 166-175.
Mellars, P. (1996). The Neanderthal Legacy. Princeton University
Press, Princeton.
Mellars, P. (2006). Why did modern humans populations
disperse from Africa ca. 60,000 year ago? A new model.
Proceedings of the National Academy of Sciences of the USA 103:
9381-9386.
Mellar s, P., Goric, K . C., Carre, M ., Soaresg, P. A. & Ri chards, M. B.
(2013). Genetic and archaeological perspectives on the initial
modern human colonization of southern A sia. Proceedings of the
National Academy of Sciences of the USA 110 (26):
10699-10704.
Mercier, N., Valladas, H ., Froget, L., Joron, J. L ., Vermeersch, P. M.,
Van Peer, P. & Moeyersons, J. (1999). Thermoluminescence
dating of a Middle Palaeolithic occupation at Sodmein Cave,
Red Sea Mountains (Egypt). Journal of Archaeological Science
26: 1339-13 45.
Metspalu, M. , Kivisild, T., Metspalu, E ., Parik, J. & Hud jashov, G.
(2004). Most of the extant mtDNA boundaries in south and
southwest Asia were likely shaped during the initial settlement
of Eurasia by anatomically modern humans. BMC Genetics 5: 26.
Meyer, M., Kircher, M., Gansauge, M. T. & Racimo, H. L. F.
(2012). A High-Coverage Genome Sequence from an Archaic
Denisovan Individual. Science 338: 222-226.
Monigal, K. (2002). The Levantine Leptolithic: Blade production
from the Lower Paleolithic to the dawn of the Upper Paleolithic. 2
vols. Ph.D, dissertation, Southern Methodist University, Dallas.
Monigal, K. (2003). Technology, Economy and Mobility at the
Beginning of the Levantine Upper Palaeolithic. In: A. N. Goring-
Morris & A. Belfer-Cohen (Ed.) More than Meets the Eye. Oxbow
Book s, Exeter, 118-133.
Munday, F. C. (1976). Intersite variability in the Mousterian
occupation of the Avdat/Aqev area. In: A. E. Marks (Ed.)
Prehistory and Paleoenvironments of the Central Negev, Israel,
Vol. I, The Avdat/Aqev Area, Part 1. Southern Methodist
University Press, Dallas, 57-68.
Munday, F. C. (1977). Nahal Aqev (D35): a stratified, open-air
Mousterian occupation in the Avdat/Aqev area. In: A. E. Marks
(Ed.) Prehistory and Paleoenvironments of the Central Negev,
Israel, Vol. II, The Avdat/Aqev Area, Part 2 and the Har Harif.
Southern Methodist University Press, Dallas, 35-60.
Munday, F. C. (1979). Levantine Mousterian technological
variability: a perspective from the Negev. Paléorient 5: 87-104 .

Quartär 61 (2014)The Middle-Upper Palaeolithic transition in the southern Levant
83
Mustafa, M. & Clark, J. (2007). Quantifying diachronic
variability: the ‘Ain Difla Rockshelter ( Jordan) and the evolution
of Levantine Mousterian technology. Eurasian Prehistory 5 (1):
47-8 4.
Nami, M . & Moser, J. (2010). La Hrotte d Ifri nAmmar; Tome 2: Le
Paléolithique Moyen. Forschugen zur Archäologie
Aussereu ropäischer Kultu ren, Vol. 9. Reichert Verlag, Wiesbaden.
Neves, A. & Serva, M. (2012). Extremely rare interbreeding
events can ex plain Neanderthal DNA in living humans. PLoS O NE
7 (10): e47076.
Newcomer, M. (1970). The Chamfered pieces from Ksar Akil
(Lebanon). Bulletin of the Institute of Archaeology University of
London 8-9: 17 7-191.
Ohnuma, K. (1988). Ksar ‚Akil, Lebanon. A Technological Study of
the Earlier Upper Paleolithic Levels of of Ksar ‘Akil, Vol. III. Levels
XXV-XIV. BAR International Series 426, Oxford.
Olivieri, A., Achilli, A., Pala, M ., Battaglia, V., Fornarino, S.,
Al- Zahery, N. & Scozzari, R. (2006). The mtDNA legacy of
the Levantine early Upper Palaeolithic in Africa. Science 314 :
1767-17 70.
Olszewski, D. I., Dibble, H. L., Schurmans, U. A., McPherron, S. P.
& Smith, J . R. (2005). H igh Desert Paleo lithic Survey at Aby dos,
Egypt. Journal of Field Archaeology 30 (3): 283-303.
Olszewsk i, D. I., Dibble, H. L ., McPherro n, S. P., Schur mans, U. A.,
Chiott i, L. & Smith, J. R . (2010). Nubian Complex strategies in
the Egyptian high desert. Journal of Human Evolution 59:
188-201.
Oppenheimer, S. (2009). The great arc of dispersal of modern
humans: Africa to Australia. Quaternary International 202: 2-13.
Parr, P., Zarins, J. , Ibrahim, M., Waec hter, J., Ga rrard, A., Cl arke, C.,
Bidmead, M. & Al-Badr, H. (1978). Preliminary report on the
second phase of the Northern Province Survey. Atlal 2: 29-50.
Parker, A. G. & Rose, J. I. (2008). Climate change and human
origins in southern Arabia. Proceedings of the Seminar for
Arabian Studies 38: 25-42 .
Parton, A. P., Farrant, A. R., Leng, M. L., Schwenninger, J. L.,
Rose, J. I., Uerpmann, H. P. & Parker, A. G. (2013). An Early
MIS3 Pluvial Phase Within Southeast Arabia: Climatic and
Archaeological Implications. Quaternary International 300:
62-74.
Pattan, J . N. & Pearce, N. J. G . (2009). Bottom water oxygenation
history in southeastern Arabian Sea during the past 140 ka:
Results from redox-sensitive elements. Palaeogeography,
Palaeoclimatology, Palaeoecology 280: 396-405.
Petit-Maire, N., Carbonel, P., Reyss, J. L., Sanlaville, P., Abed, A.,
Bourro uilh, R., Font ugne, M. & Yasin, S . (2010). A va st Eemian
palaeolake in southern Jordan (29 N). Global and Planetary
Change 72: 368-373.
Petraglia, M., Alsharekh, A., Crassard, R., Drake, N., Groucutt,
H., Parker, A. & Roberts, R. (2011). Middle Palaeolithic
occupation on a last interglacial lakeshore in the Nefud Desert,
Saudi Arabia. Quaternary Science Reviews 30: 1555-1559.
Petraglia, M., A lsharekh, A ., Breeze, P., Clarkson, C . & Crassard,
R. (2012). Hominin Dispersal into the Nefud Desert and
Middle Palaeolithic Settlement along the Jubbah Palaeolake,
Northern Arabia. PLoS ONE 7(11) : e 498 40 .
Potter, J. (1995). Lithic Technology and Settlement Pattern
Variability within the Levantine Mousterian: a Comparison of
Sites WHS 621 and WHS 634 from Wadi al-Hasa. Studies in the
History and Archaeology of Jordan V: Art and Technology
Through the Ages. Department of Antiquities, Amman, 497-504.
Preusser, F. (2009). Chronology of the impact of Quaternary
climate change on continental environments in the Arabian
Peninsula. Comptes Rendus Geoscience 34: 621-632.
Quintana-Murci, L., Semino, O., & Bandelt, H. J. (1999).
Genetic evidence of an early exit of Homo sapiens sapiens from
Africa through eastern Africa. Nature Genetics 23: 437-441.
Radies, D. , Preusser, F., Matter, A. & Man ge, M. (2004). Eustatic
and climatic controls on the development of the Wahiba Sand
Sea, Sultanate of Oman. Sedimentology 51: 1359-13 85.
Reich, D., Patterson, N., Kircher, M., Delfin, F. & Nandineni, M.
(2 011) . Denisova admixture and the first modern human
dispersals into southeast Asia and Oceania. American Journal of
Human Genetics 89: 516-528.
Reichar t, G. J., den Dulk, M., V isser, H. J., van der Weijde n, C. H.
& Zachariasse, W. J. (1997). A 225 kyr record of dust supply,
paleoproductivity and the oxygen minimum zone from the
Murray Ridge (northern Arabian Sea). Palaeogeography,
Palaeoclimatology, Palaeoecology 134: 149-169.
Richter, J., Hauck, T., Vogelsang, R., Widlok, T., Le Tensorer,
J.-M. & Schmid, P. (2012). “Contextual areas” of early Homo
sapiens and their significance for human dispersal from Africa
into Eurasi a between 200 ka an d 70 ka. Quaterna ry International
274: 5-24.
Rink, W., Richter, D., Schwarcz, H., Marks, A. E., Monigal, K. &
Kaufman, D. (2003). Age of the Middle Palaeolithic Site of
Rosh Ein Mor, Central Negev, Israel: Implications for the Age
Range of the Early Levantine Mousterian of the Levantine
Corridor. Journal of Archaeological Science 30 (2): 195-204.
Rose, J. I. (2000). A Reconstruction of Pleistocene Ara bia: the state
of research and avenues for further inquiry. M.A., dissertation,
Boston University, Boston.
Rose, J. I. (2004). The Question of Upper Pleistocene
Connections between East Africa and South Arabia. Current
Anthropology 45: 551-555.
Rose, J. I. (2006). Among Arabian Sands. Defining the Paleolithic
of Southern Arabia. Ph.D, dissertation, Southern Methodist
University, Dallas.
Rose, J. I. (2007). The Arabian Corridor Migration Model:
archaeological evidence for hominin dispersals into Oman
during the Middle and Upper Pleistocene. Proceedings of the
Seminar for Arabian Studies 37: 219-237.
Rose, J. I. (2010). New light on human prehistory in the Arabo-
Persian Gulf Oasis. Current Anthropology 51 (6): 849-883.
Rose, J. I. & Petraglia, M. (2009). Tracking the origin and
evolution of human populations in Arabia. In: M. Petraglia &
J. I. Rose (Eds.) The Evolution of Human Populations in Arabia:
Paleoenvironments, Prehistory and Genetics. Springer Academic
Publishers, Netherlands, 1-14.
Rose, J. I. & Usik, V. I. (2009). The “Upper Palaeolithic” of
southern Arabia. In: M. Petraglia & J. I. Rose (Eds.) The Evolution
of Human Populations in Arabia: Palaeoenvironments, Prehistory,
and Genetics. Springer Academic Publishers, Netherlands,
169-185.
Rose, J. I. & Hilbert, Y. H. (2014). New prehistoric sites in the
southern Rub’ al Khali desert, Oman. Antiquity 88 (381): Project
Gal lery.
Rose, J. I., Usik, V., Marks, A., Hilbert, Y. H., Galletti, C., Parton,
A., Geiling, J. M., Cerny, V., Morley, M. & Roberts, R. (2011).
The Nubian Complex of Dhofar, Oman: an African Middle Stone
Age industry in southern Arabia. PLoS ONE 6 (11): e28239.
Rosenberg, T., Preusser, F., Fleitmann, D., Schwalb, A., Penkman,
K., Schmid, T. W., Al-Shanti, M. A., Kadi, K. & Matter, A.
(2 011) . Humid periods in southern Arabia: Windows of
opportunity for modern human dispersal. Geology 39:
1115 -1118 .
Rosenberg, T. M., Preusser, F., Blechschmidt, I., Fleitmann, D.,
Jagher, R. & Matter, A. (2012). Late Pleistocene palaeolake in
the interior of Oman: a potential key area for the dispersal of
anatomically modern humans out-of-Africa? Journal of
Quaternary Science 27: 13-16.
Sánchez-Quinto, F., Botigué, L. R., Civit, S., Arenas, C., Ávila-
Arcos, M., Bustamante, C. D., Comas, D. & Lalueza-Fox, C.
(2012). North African Populations Carry the Signature of
Admixture with Neandertals. PLoS ONE 7: e47765.

Quartär 61 (2014) J. I. Rose & A. E. Marks
84
Sankararaman, S., Patterson, N., Li, H., Pääbo. S. & Reich, D.
(2012). The date of interbreeding between Neandertals and
modern humans. PLoS ONE 8 : e1002947.
Sankararaman, S., Mallick, S., Dannermann, M., Prüfer, K.,
Kelso, J., Pääbo, S., Patterson, N. & Reich, D. (2014). The
genomic landscape of Neaderthal ancestry in present-day
humans. Nature 507: 354-357.
Sanlaville, P. (1992). Changements climatiques dans la péninsule
Arabique durant le pléistocène supérieur et l’holocène.