Content uploaded by Manuel Will
Author content
All content in this area was uploaded by Manuel Will on Dec 14, 2020
Content may be subject to copyright.
*corresponding author
Quartär 66 (2019) : 51-80
51
doi: 10.7485/QU66_3
Lithic assemblages from the Middle Paleolithic
of Geißenklösterle Cave provide insights on
Neanderthal behavior in the Swabian Jura
Steinartefaktinventare aus dem Mittelpaläolithikum des Geißenklösterle, Deutschland:
Neue Erkenntnisse zum Verhalten der Neandertaler auf der Schwäbischen Alb
Nicholas J. C*1,2, Viola C. S1,3, Michael B1,4 & Manuel W1*
1 Department of Early Prehistory and Quaternary Ecology, University of Tübingen, Schloss Hohentübingen, 72070 Tübingen,
Germany; email: nicholas.conard@uni-tuebingen.de, manuel.will@uni-tuebingen.de
2 Senckenberg Center for Human Evolution and Paleoenvironment, University of Tübingen, Germany
3 UMR7041, Equipe AnTET, Université Paris Ouest Nanterre La Défense, Nanterre Cedex, France
4 Heidelberg Academy of Sciences and Humanities, Research Project ‘The Role of Culture in Early Expansions of Humans’,
University of Tübingen, Germany
A - The Swabian Jura has long played a crucial role in key debates about the European Paleolithic. One of the best-
known sites, Geißenklösterle Cave in the Ach Valley, has yielded a stratigraphic sequence including both Middle and Upper
Paleolithic find horizons separated by a largely geogenic horizon. Here we present combined techno-economic and attribute
analyses of the lithic artifacts from Middle Paleolithic horizons AH IV-VIII dating between ~90-45 ka BP. The lithic analyses
demonstrate that Neanderthals mainly used the Levallois concept to knap locally available Jurassic cherts and produce small
blanks and tools. Other raw materials occur as isolated artifacts. Apart from various modalities of Levallois technology,
knappers employed Kostenki, bipolar and platform methods. Scrapers and splintered pieces are the most frequent tools,
whereas notches, denticulates and bifacial implements including bifacially backed knives (Keilmesser) and leaf points (Blatt-
spitzen) are absent. Low densities of archaeological finds and the export of selected blanks and tools indicate repeated short-
term occupations of the site in a settlement system character ized by high mobility. Although minor diachronic var iation occurs,
assemblages IV-VIII show a distinct signature that can be attributed to the same general technological and techno-economic
system. Regional comparisons suggest that the Middle Paleolithic assemblages from Geißenklösterle correspond to the
Swabian Mousterian, which is defined by the use of local raw materials, frequent Levallois reduction sequences, multiple
scraper forms, and an almost complete absence of bifacial technology including Keilmesser and Blattspitzen. The upper Middle
Paleolithic assemblages dating to ~50-45 ka BP provide new insights into the behavior and demography of late Neanderthals
prior to the arrival of anatomically modern humans. Overall, the archaeology of Geißenklösterle illustrates a sharp break in
lithic technology, organic artifacts, subsistence strategies, site use and population dynamics between Neanderthals and Homo
sapiens in southwestern Germany.
Zusammenfassung - Seit Beginn des 20. Jahrhunderts spielt die Schwäbische Alb eine zentrale Rolle für die Erforschung des
Mittel- und Jungpaläolithikums in Europa. Das Geißenklösterle im Achtal stellt eine der wichtigsten Fundstellen dieser Region dar.
Vor allem durch die Grabungen von J. Hahn von 1974-1991 bekannt, lieferte die Höhle eine lange und bedeutende archäologische
Abfolge des Mittel- und Jungpaläolithikums, welche voneinander durch einen fundarmen Horizont getrennt sind. Der Fokus der
Forschung auf die jungpaläolithischen Funde des Geißenklösterle führte dazu, dass eine detaillierte Beschreibung der Steinarte-
fakte aus dem Mittelpaläolithikum bisher nicht vorgenommen wurde. Hier stellen wir die Auswertung von fünf Steinartefaktinven-
taren aus dem Mittepaläolithikum (AH IV-VIII) vor, welche auf ~90-45 ka BP datiert sind. Diese Inventare stammen hauptsächlich
aus den Neugrabungen am Geißenklösterle durch N. J. Conard in den Jahren 2001 und 2002, die zum ersten Male die mittelpa-
läolithischen Schichten in größerem Umfang erfassten. Die Analyse der Steinartefaktinventare erfolgte anhand eines kombi-
nierten Ansatzes von Attributanalyse sowie der Auswertung von Abbausequenzen und techno-ökonomischen Aspekten. Unsere
Forschungsziele betrafen insbesondere die Charakterisierung der Neandertaler-Technologie auf synchroner und diachroner
Ebene, die Einbettung der Inventare in das Mittelpaläolithikum der Schwäbischen Alb und Zentraleuropas, sowie Aussagen
hinsichtlich Siedlungsmustern und Demographie. Die lithischen Inventare des Geißenklösterle sind relativ klein (n = 200; >20 mm),
allerdings mit einer Vielzahl an Stücken <20 mm (n = 704) und häufigem Auftreten von Frostbeschädigungen. Die Neandertaler
nutzten vorwiegend den lokal verfügbaren Jurahornstein, um kleine Abschläge und Werkzeuge mithilfe unterschiedlicher Abbau-
methoden herzustellen, wobei Levallois den Hauptteil ausmacht. Andere Rohmaterialien wie Bohnerzhornstein, Muschelkalkhorn-
stein, Radiolarit und Quarzit treten nur als isolierte Artefakte auf. Außer unterschiedlichen Varianten des Levallois-Konzepts
konnte der Abbau von Kostenki, bipolaren und Plattform-Kernen nachvollzogen werden. Die Werkzeuginventare sind durch
Quartär 66 (2019) N. J. Conard et al.
52
Introduction
Since the 1860s generations of archaeologists have
conducted Paleolithic research in the Swabian Jura
with the majority of excavations focusing on the caves
of the Ach and Lone valleys (Fraas 1867; Schmidt 1910,
1912; Riek 1934; Müller-Beck 1983; Wagner 1983;
Hahn 1988; Conard et al. 2015) (Fig. 1). These sites
have provided important archaeological sequences
that span the Middle and Upper Paleolithic (MP; UP).
Most recent work has focused on the UP sequences
and particularly the Aurignacian find horizons from
Fig. 1. Map of Geißenklösterle and other MP sites from the Swabian Jura mentioned in the text.
Abb. 1. Geographische Karte zur Lage des Geißenklösterle und weiterer im Text benannter MP-Fundstellen auf der Schwäbischen Alb.
häufige Schaberformen und seltenere ausgesplitterte Stücke gekennzeichnet, wohingegen gekerbte und gezähnte Stücke sowie
bifazielle Werkzeuge (unter anderem Keilmesser oder Blattspitzen) vollständig fehlen. Die durchweg geringe Dichte an archäolo-
gischen Funden innerhalb des Mittelpaläolithikums, das Fehlen an Befunden und klaren Fundhorizonten, sowie der Export von
ausgewählten Grundformen und Werkzeugen zeigt wiederholte kurzzeitige Belegungen der Fundstelle innerhalb eines Siedlungs-
systems mit hoher Mobilität an, vermutlich durch kleine Gruppen. Obwohl das Mittelpaläolithikum des Geißenklösterle gering-
fügige diachrone Veränderungen aufweist, zeigen Inventare IV-VIII jedoch insgesamt konsistente Signaturen, welche einem
gemeinsamen techno-typologischen und techno-ökonomischen System zugeordnet werden können. Regionale Vergleiche
belegen, dass die untersuchten Inventare mit dem Schwäbischen Moustérien übereinstimmen, welches durch die hauptsächliche
Nutzung lokalen Rohmaterials, den häufigen Abbau durch das Levallois-Konzept, vielfältige Schaberformen und eine fast
vollständige Abwesenheit von Keilmessern und Blattspitzen gekennzeichnet ist. Da die obersten Schichten des Mittelpaläolithikums
auf ~50-45 ka datiert sind, geben unsere Befunde neue Einblicke in das Verhalten und die Demographie später Neandertaler, die
in Südwestdeutschland direkt vor der Ankunft anatomisch moderner Menschen lebten. Zusammenfassend zeigt die archäolo-
gische Abfolge des Geißenklösterle einen scharfen Bruch in der lithischen Technologie, organischen Artefakten, Subsistenzstra-
tegien, Fundplatznutzung und vermutlich Bevölkerungsdichte zwischen Neandertalern und Homo sapiens.
K - Lithic Technology, Swabian Mousterian, Neanderthal mobility, Raw material economy, Levallois
method, Settlement dynamics
Steinartefakttechnologie, Schwäbisches Moustérien, Mobilität der Neandertaler, Rohmaterialnutzung,
Levallois-Methode, Siedlungsmuster
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
53
the region. The early Aurignacian of the Swabian Jura
reflects the first wave of expansion by modern humans
into Central Europe (Conard & Bolus 2003, 2006,
2015; Conard et al. 2003a, 2003b; Higham et al. 2012),
and has yielded early examples of figurative art,
musical instruments and mythical imagery (Hahn 1986;
Hahn & Münzel 1995; Conard 2003, 2007, 2009;
Conard et al. 2009; Kind et al. 2014; Conard & Kind
2019) alongside a wide array of personal ornaments
and numerous examples of innovations among the
lithic and organic artifacts (Hahn 1988; Conard &
Bolus 2003, 2006; Wolf 2014).
This paper presents the MP assemblages from
Geißenklösterle Cave (GK) in the Ach Valley of the
Swabian Jura. The artifacts analyzed here originate
from J. Hahn’s excavations in the 1980s and early 1990s
and from N. J. Conard’s excavation of GK in 2001 and
2002. Hahn published his key findings on the MP
artifacts from the site in his important monograph
from 1988 (Hahn 1988), and we have recently
published a second monograph on GK that provides
many new results focusing on subsistence, dating and
environmental studies (Conard et al. 2019). Under-
standably, Hahn dedicated most of his monograph
and publications on GK to the exceptional UP
sequence. Due to his outstanding work and rigorous
publications, GK is internationally known for its key
Aurignacian sequence with two main find complexes
of archaeological horizons (AHs) II and III. Subsequent
work has done much to inform the scient ific community
about the important Gravettian horizons from the site
(Conard & Moreau 2004; Moreau 2009).
In this paper we present a technological,
typological and techno-economic study of the five MP
horizons of the site. Our main goals were to document
the operational sequences, including methods of core
reduction, knapping techniques, and tool production.
These studies aimed at identifying diachronic and
synchronic variability within the GK sequence. We
consider the cultural and chronostratigraphic
sequence and how it fits with what we know about the
MP of the Swabian Jura and Central Europe. Ultimately,
this work helps us to refine our understanding of the
technology, lifeways and demography of late
Nean derthals living in southwestern Germany prior to
the arrival of anatomically modern humans. One
advantage of having waited to publish the details of
the MP lithic assemblages from GK is that M. Richard
has published new ESR dates from the site (Richard
2015; Richard et al. 2019), which together with earlier
radiocarbon, ESR and luminescence dates makes the
MP find horizons from GK the best dated ones from
the Swabian Jura.
Background on the Middle Paleolithic in the
Swabian Jura
In recent decades, studies of the MP in the Swabian
Jura have received less scholarly attention compared
to the well-known UP. This being said, MP occupations
are known from numerous sites in the region (Fig. 1).
They include Bockstein (Bocksteinloch, Bockstein-
schmiede, Bockstein-Törle), Hohlenstein (Bärenhöhle
and Stadel), Vogelherd and Haldenstein in the Lone
Valley, as well as GK, Hohle Fels, Sirgenstein, and
Große Grotte in the Ach Valley, and Kogelstein located
in the adjacent Schmiech Valley. Less well-known sites
include Heidenschmiede, Irpfelhöhle, Göpfelstein-
höhle and Schafstall II, and rare open-air sites such as
Börslingen, Wippingen, Wittlingen and Winterhalde
(e.g. Burkert et al. 1992; Fisher et al. 2008; Conard et
al. 2012; Çep 2013; Floss & Schürch 2015; Floss et al.
2017). Previous work has generally focused on the
question of Neanderthal technology and climatic
adaptations in Central Europe during the Late Pleis-
tocene. There has been an emphasis on examining the
MP to UP transition, connected to questions about the
arrival of the first modern humans to Europe and the
subsequent extinction and replacement of
Neanderthals.
One of the key findings of previous research is the
consistent documentation of an occupational hiatus
between MP and UP find horizons at several sites,
including GK, Hohle Fels, Sirgenstein and Vogelherd
(Schmidt 1912; Riek 1934; Hahn 1988; Conard 2005;
Conard et al. 2006; Bolus 2011). Assuming, as seems
likely, that Neanderthals always made the MP assem-
blages and modern human always produced the
Aurignacian assemblages, the find horizons left by
Neanderthals and modern humans are separated by
largely geogenic horizons that are either sterile or
very poor in archaeological finds. These observations
support the ‘population vacuum’ hypotheses that
modern humans migrated via the Danube Corridor
into the Swabian Jura when few if any Neanderthals
occupied the region. Initially, we assumed that this
depopulation may have been caused by a cold and
arid climatic phase, most likely the terrestrial equiv-
alent of the Heinrich 4 event ca. 40 ka (Conard 2003;
Conard et al. 2003a; Conard & Bolus 2006). Subse-
quent dating and environmental studies, however,
have shown that modern humans arrived in the region
between 43 and 42 ka calBP in a climatic phase that
was neither remarkably cold nor dry (Higham et al.
2012; Miller 2015; Rhodes et al. 2018; Goldberg et al.
2019). These stratigraphic arguments for a sharp
break in the settlement history of the Swabian Jura are
consistent with a major shift in lithic technology,
organic material culture, subsistence behavior and site
use that suggests that little or no direct contact
between Neanderthals and modern humans occurred
in this region (Conard et al. 2012).
Previous archaeological research shows that the
MP of the Swabian Jura is characterized by variable
but often low levels of occupation intensity by
Nean derthals. Most sites have yielded small MP lithic
assemblages, although exceptions such as the rich and
high-density find horizons of Bocksteinschmiede are
also known in the region. In general, in both the Ach
Quartär 66 (2019) N. J. Conard et al.
54
and Lone valleys, find densities are far lower in the MP
than the Aurignacian find horizons at the same sites
(Conard et al. 2006, 2012; Conard 2011; Bolus 2015).
Mammalian faunal assemblages from the MP cave
sites are frequently dominated by cave bear, medium-
sized carnivores and small ruminants, which normally
reflect natural death assemblages. The anthropogenic
faunas are often rich in remains of horse, reindeer and
ibex, depending on the geographic position of the
site (Münzel & Conard 2004; Krönneck et al. 2004;
Niven 2006; Conard et al. 2012). MP organic tools are
only represented by a few potential bone points and
more numerous retouchers from Vogelherd, Große
Grotte, Sirgenstein and Schafstall (Conard et al. 2006;
Bolus 2015; Toniato et al 2018). Unambiguous features
are rare in the MP deposits of the Swabian Jura, but
notable exceptions come in the form of a fireplace at
Sirgenstein VII/VIII (Schmidt 1912), and concentra-
tions of burnt bone at several sites including Bockstein -
schmiede, Große Grotte and perhaps Hohlenstein-
Stadel (e.g. Wetzel & Bosinski 1969; Wagner 1983;
Beck 1999).
Whereas the MP lithic technology and cultural
stratigraphy of the region is generally well-known,
there have been few more synthetic studies and a lack
of assemblages with absolute chronometric dates
deriving from excavations with modern field
standards. Most of the lithic assemblages in the region
are dominated by locally available raw materials (see
below). The assemblages often contain highly reduced
Levallois cores and diverse debitage products. Tools
of these assemblages are characterized by various
scraper forms, whereas bifacial implements are rare or
absent. The tools generally reflect a low degree of
standardization. These assemblages have been
referred to as the Swabian Mousterian (Schmidt 1912;
Riek 1934; Wagner 1983; Beck 1999; Böttcher et al.
2000; Conard et al. 2012; Bolus 2015). In contrast,
assemblages with high proportions of bifacially
backed knives (Keilmesser) and other bifacial forms are
usually classified as belonging to the Keilmessergruppe
(Micoquian or Pradnikian) (Richter 1997, 2016; Conard
& Fischer 2000; Jöris 2003; Bolus 2015), with the
richest assemblages coming from Bockstein and
Heidenschmiede (Peters 1931; Wetzel & Bosinski
1969; Çep 2014; Çep & Krönneck 2015). The latest MP
is characterized by the leaf points of the Blattspitzen-
gruppe, well-known from sites such as Haldenstein
(Riek 1938; Conard & Fischer 2000; Bolus 2004a, 2011,
2015; Richter 2016).
For all of these technocomplexes, absolute dating
and chrono-cultural correlations have proven to be
difficult. Among other problems, this is due to the
limits of radiocarbon dating and the lack of recent
fieldwork that could apply state-of-the art dating
methods such as OSL, TL or ESR (see Conard & Bolus
2003, 2008; Higham et al. 2012). While most MP
horizons in the Swabian Jura remain poorly dated
(Bolus 2011, 2015), the great majority of currently
known assemblages likely date to the Early and Middle
Würmian (MIS 5d-MIS 3; Conard & Bolus 2008). The
preceding Eemian Interglacial (MIS 5e) might be
represented by the small assemblage from layer IX of
Vogelherd (Riek 1934; Niven 2006) and the MP finds
from the lower travertine of Stuttgart-Untertürkheim
(Wenzel 2007).
Most of the well-known MP assemblages from the
Swabian Jura were excavated between the 1930s and
1960s. While the archeologists leading these excava-
tions adhered to high standards for their time, the
fieldwork, of course, lacked important elements such
as 3D piece-plotting, modern geoarchaeological
analyses and the application of radiometric dating.
Moving beyond the well-known excavations in the
Swabian caves, numerous open-air sites have yielded
MP assemblages, but almost all of these assemblages
originate from undated surface collections (e.g.
Burkert et al 1992; Floss and Schürch 2015). Only
rarely have stratified MP finds been recovered, and
only the late Middle Pleistocene site of Bollschweil in
the Black Forest has been radiometrically dated (Rink
et al. 2002). In order to move research on the MP of
the Swabian Jura and Germany forward, it is of utmost
importance to analyze well-dated assemblages from
excavations conducted in accordance with modern
field methods. Absolute dates are also crucial for
defining the key features and temporal variability of
technocomplexes such as the Swabian Mousterian.
The MP assemblages from GK in the Ach Valley
reported upon here, meet these conditions. Today,
GK, Hohle Fels and Kogelstein remain the only three
stratified MP sequences in the Swabian Jura deriving
from modern fieldwork and associated with absolute
dates, although Kogelstein has yielded only infinite 14 C
ages (Böttcher et al. 2000). The excavations at GK and
Hohle Fels with their excellent recovery and abundant
contextual information provide the most detailed
evidence for site-use and Neanderthal technological
adaptations in the Swabian Jura.
The site of Geißenklösterle
The site of GK is situated in the Achtal – a former
valley of the Danube – located 60 m above the valley
floor at ~550 m a.s.l. (Fig. 2). The cave is part of a karst
system within the Upper Jurassic limestone of the
south-eastern region of the Swabian Jura (Baden-
Württemberg). GK is one of the key Paleolithic sites in
Europe, documenting human occupations during the
Mesolithic, UP (Magdalenian, Gravettian, Aurignacian)
and M P.
Discovered as an archaeological site by R. Blumen-
tritt in 1957, excavations at GK started with the
opening of test trenches in 1963 (G. Riek) and 1973 (E.
Wagner) and were continued by J. Hahn (University of
Tübingen) in 14 field seasons from 1974 until 1991
(Hahn 1988). The systematic field work by Hahn
uncovered deposits of over 4 m in thickness that were
excavated in an area of approximately 45 m2 for the UP
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
55
the early Aurignacian (Fig. 4). The stratigraphic discon-
tinuity between the Aurignacian and uppermost MP
layers is further attested by differences in lithic and
faunal assemblages, a drop in find densities for all
classes of finds at the base of the Aurignacian and
3D-plots of finds showing no overlap between the
mostly horizontal archaeological horizons (Fig. 4;
Conard & Malina 2002, 2003; Conard et al. 2006;
Conard et al. 2019).
The MP deposits amount to ca. 1.0 m, which
include AHs IV-VIII that range in thickness from
10-35 cm. These five layers are characterized by
varying quantities of limestone rubble in a silty matrix
which contained variable, but generally low amounts
of lithic artifacts, modified faunal remains and burnt
bone. The archaeological material did not form clearly
defined find horizons and no archaeological features
were discerned (Fig. 4), as was also the case in later
years at the MP deposits of Hohle Fels (Conard &
Malina 2013). Previous observations on sediments,
artifacts and extensive refitting (Hahn 1988) revealed
that the original positions of stone artifacts and bones
have been moved by cryo- and bioturbation after
primary sedimentation, which also caused edge
damage on lithics (Hahn 1988). Micromorphological
studies have, however, found no mixing between the
MP and UP layers (Miller 2015; Goldberg et al. 2019).
The absence of diagnostic Aurignacian artifacts in the
MP layers and the failure to produce any links between
the Aurignacian and MP strata during refitting of
layers, but opened only a test pit of 4 m2 for the two
uppermost MP levels (AHs IV & V; Hahn 1988) without
reaching bedrock. In 2001 and 2002, N. J. Conard
continued the fieldwork at GK using Hahn’s excavation
grid and stratigraphic designations, but added
systematic 3D piece plotting of archaeological material
with a total station assisted by the EDM program
(Dibble & McPherron 1996) to the field methods. The
new fieldwork focused on the deeper parts of the
deposits (lower Aurignacian III-IIIb and MP layers),
with the aim of recovering the entire vertical stratig-
raphy of the site. These renewed excavations
recovered artifacts from all Neanderthal occupations
(AHs IV-VIII) in 7-10 m2 (Fig. 3) and reached bedrock in
6 m2, uncovering a total thickness of roughly 5 m for
the cave deposits (Conard & Malina 2002, 2003; Miller
2015). The majority of the MP assemblages studied
here (99 % of lithic artifacts) were excavated with
modern field methods by the excavations in 2001 and
2002.
The overall stratigraphy of the site encompasses
23 geological horizons (GHs), among which 20 AHs
could be distinguished (Fig. 2; more details in Hahn
1988; Conard & Malina 2003; Miller 2015; Conard et
al. 2019). The Mesolithic and UP occupations span
AHs I-III (Magdalenian, Gravettian, Aurignacian)
whereas the MP settlements encompass AHs IV-VIII
(GHs 18-23). The MP deposits lie below a largely
geogenic horizon of ca. 20 cm thickness (GH 17; AH
IIIc) that separates the Neanderthal occupations from
Fig. 2. Composite picture of the site and stratigraphy of Geißenklösterle. Left: View of the collapsed cave of Geißenklösterle from afar (red
arrow). Middle bottom: Excavations by J. Hahn (center) with A. Scheer (right; source: Archäologie in Deutschland 1984). Middle top: Excava-
tions into the Middle Paleolithic deposits in 2002 with M. Malina ( left) and L. Giemsch (Photo: N. J. Conard). Right: Composite stratigraphy of
the main profile for the entire archaeological sequence. The MP layers (AH IV-VIII) are highlighted in colors.
Abb. 2. Überblick über die Fundstelle und Stratigraphie des Geißenklösterle. Links: Blick auf die eingestürzte Höhle des Geißenklösterle aus der
Ferne (roter Pfeil). Mit te unten: Ausgrabungen durch J. Hahn (Mitte) mit A. Scheer (rechts; Quelle: Archäologie in Deutschland 1984). Mitte oben:
Ausgrabung in den mittelpaläolithischen Schichten 2002 mit M. Malina (links) und L. Giemsch (Photo: N. J. Conard). Rechts: Stratigraphie des
Hauptprofils (idealisiert) für die gesamte archäologische Sequenz. Die MP-Schichten (AH IV-VIII) sind farblich her vorgehoben.
Quartär 66 (2019) N. J. Conard et al.
56
faunal remains and lithic artifacts (Hahn 1988; Conard
& Bolus 2003, 2008; Teyssandier et al. 2006) further
attest to the integrity of the Neanderthal
occupations.
Given the lack of clear anthropogenic features or
stratigraphic markers, the AHs in the MP do not
represent occupation horizons in a strict sense. They
simply refer to the artifacts from a volume of material
from a geological unit of similar composition. These
AHs can be viewed as palimpsest from different
phases of occupation at GK, and it is best to view the
term AH as a technical term. Unlike in the Aurignacian
deposits, where refitting sequences and other
approaches can be used to critically access the validity
of the AHs (Hahn 1988), the lack of long reduction
sequences and the low densities of cultural material in
the MP horizons makes it difficult to make such
assessments. Additionally, stratigraphic ambiguities
when excavating deposits lacking clear stratigraphic
markers, warn against viewing these AHs as more than
a collection of finds of roughly similar age (Hahn 1988;
Conard et al. 2019).
Absolute dating of the MP layers has proven to be
difficult. Initial absolute dating of the MP horizons by
ESR on teeth from Hahn’s excavations produced a
weighted mean age for AH IV of 43.3 ± 4 ka, placing
the final Neanderthal occupation close to beginning
of the Aurignacian at the site as well as the Hengelo
Interstadial (Richter et al. 2000). Initial radiocarbon
dates for the MP horizons IV, VI, VII, and VIII fell
between 42 and 31 ka BP (Conard & Bolus 2008) but
were considered to be younger than their actual age
due to possible taphonomic reworking and potential
problems related to elevated levels of atmospheric
Fig. 3. Overview on the excavation grid of Geißenklösterle with color indication of the different stratigraphic levels reached (left) (after Conard
et al. 2019). Horizontal find dist ribution for faunal remains and stone tools within the five M P find horizons (right) (after Conard and Malina 2003).
Abb. 3. Messnet z des Geißenklösterle mit farbiger Anzeige der unterschiedlichen stratigraphischen Tiefe, die ergraben wurde (links) (nach Conard et
al. 2019). Horizontale Fundverteilung der Faunenreste und Steinartefakte innerhalb der fünf MP-Schichten (rechts) (nach Conard und Malina 2003).
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
57
radiocarbon (“MP Dating Anomaly”, see Conard &
Bolus 2008; Higham et al. 2012). More recently, the
application of improved AMS 14C dating with an ultra-
filtration protocol has provided uncalibrated dates of
35.5 ± 0.65 ka BP for AH IV and 48.6 ± 3.2 ka BP for AH
VII (Higham et al. 2012, 2014). Using IntCal13
calibration curves (Reimer et al. 2013) and Bayesian
modelling for the lower horizons of GK – taking into
account stratigraphic succession and 14C dates from
AHs III-VI – provide a final boundary for the end of
the MP at the site of 43.8-41.6 ka calBP (Higham et al.
2012, 2014), lying close to the ESR readings from
Richter et al. (2000). Based on these dates, the majority
of the MP sequence (IV-VII) can be constrained to
~52-43 ka BP, indicating late Neanderthal occupations
that mostly fall into MIS 3. The most recent ESR dates
by M. Richard provide a somewhat different result,
yielding older ages for AHs IV-VIII that range between
55 ± 6 and 65 ± 8 ka BP (AH IV) until 73 ± 10 and 94 ±
10 ka BP (AH VII), pushing particularly the lower part
of the sequence well into MIS 5 (see Richard 2015;
Richard et al. 2019). The occupation hiatus (GH 17) is
dated to 48-44 ± 6 ka BP, providing a minimum age for
the MP that is similar to the Bayesian 14 C model. AH
VIII remains to be dated. In sum, the MP sequence is
constrained by several dating methods to between
~94-43 ka BP. While the uppermost layers fall into
early and middle MIS 3 relatively close to the arrival of
the earliest modern humans, the deeper layers appear
to provide a much greater time depth for the entire
sequence extending until MIS 5c/d.
Materials and methods
The Middle Paleolithic assemblages of
Geißenklösterle
The MP assemblages from GK derive from a total of
five archaeological horizons (IV-VIII) that were
excavated in 7-14 m2 with a volume of 9.5 m3 (Fig. 3).
The combined material from Hahn’s and Conard’s
excavations encompass stone artifacts (n = 904), faunal
remains (mammalian fauna: NISP = 1 308; g = 6 719.3),
modified bone (n = 33; e.g. butchery marks on Capra
ibex), burnt bone (141 g) as well as charcoal (n = 1; see
also Münzel & Conard 2004; Conard et al. 2012;
Münzel 2019). Assemblages IV-VIII feature numerous
remains of cave bear (n = 586), some hyena (n = 8) and
other medium-sized carnivores, whereas the anthro-
pogenic fauna at GK is dominated by remains of
reindeer (n = 53), horse (n = 21) and ibex (n = 20)
among others (Münzel 2019). For this study, we
analyzed all lithic artifacts excavated by Conard
(n = 892) and the pieces recovered by Hahn (n = 12
from AHs IV and V; see Hahn 1988). We only recorded
artifacts that could be attributed unambiguously to a
single archaeological layer. These assemblages include
904 stone artifacts – n = 200 larger than 20 mm and
n = 704 smaller than 20 mm, the latter including many
microflakes <5 mm (n = 92) – from all archaeological
horizons (AH IV-VIII) which vary strongly in sample
size (Fig. 5). While layers VI and VII provide reliable
sample sizes (n = 238-477), the topmost (IV, V) and
lowermost assemblages (VIII) only exhibit small
samples (n = 40-85) for artifacts >20 mm. Thus, we
consider layers VI and VII to be the most represent-
ative assemblages to characterize the MP technology
at GK. In relation to sediment volumes, lithic densities
are very low (23.7-146.3 n/m3) and contrast markedly
with the situation in the overlying UP (mean = 684 n/m3;
see Conard et al. 2012). Lithic densities in the MP
correlate strongly with total assemblage size, with
comparatively high values for layers VI and VII
(n/m3 = 112-146) and much lower figures for IV, V and
VIII (n/m3 = 24-47). These values suggest that the
average occupation density is low, but with some
variation.
Fig. 4. Vertical distribution of lithic artifacts (piece-plotted number per AH in brackets) from the lower Aurignacian (AH IIIa, b), largely
geogenic deposits (AH IIIc) and the MP (AH IV-VIII) of Geißenklösterle (modified after Conard et al. 2006).
Abb. 4. Vertikale Fundverteilung von Steinartefakten (Anzahl einzeln eingemessener Stücke pro AH in Klammern) aus dem unteren Aurignacien
(AH IIIa, b), der größtenteils geogenen Schicht (AH IIIc) und dem MP (AH IV-VIII) des Geißenklösterle (modifiziert nach Conard et al. 2006).
Quartär 66 (2019) N. J. Conard et al.
58
Raw material availability in the MP of the Swabian
Jura
Much previous Paleolithic research in the Swabian Jura
has focused on the identification of raw materials and
their sources (Beck 1999; Böttcher et al. 2000; Burkert
2001; Çep & Waiblinger 2001; Burkert & Floss 2005;
Fisher et al. 2008; Çep et al. 2011; Floss et al. 2012; Çep
2013; Floss & Schürch 2015; Herkert et al. 2015). The
lithic raw materials from most MP sites in the region
are dominated by local, predominantly grey and
white, variants of Jurassic cherts ( Jurahornstein), with
lesser quantities of cherts with different col oration
(e.g. Jurassic chert with brown coloration from stratifi-
cation with bean ore called Bohnerzhornstein), radio-
larite (green and red variants), black alpine micro-
quartzite (e.g., from the fluvial sediments of the Alpine
foothills), quartz and others.
Concerning GK, several known sources for the
local light-grey and banded Jurassic chert, such as
Borgerhau, are less than 5 km away from the site (Fisher
et al. 2008; Floss & Schürch 2015). The Jurassic cherts
of the region are accessible in large quantities of
nodules – ranging from roughly golf ball to soccer
ball-size – and are generally of good but varying
knapping quality, appearing in different stages of
sili cification. The sources for the less frequently used
rock types such as Bohnerzhornstein, radiolarite and
quartzite can usually be found within ca. 20 km of the
Swabian caves – often as rounded cobbles from
secondary deposits such as river terraces – with no MP
site having more than 5 % of finds that come from over
20 km away
Methods of lithic analysis
The study of the lithic assemblages proceeded by AH
as principle unit of analysis. All lithic finds >20 mm
(n = 200) were examined individually, regardless
whether broken or not, with quantitative analyses
being performed for the remaining artifacts <20 mm
(n = 704). Concerning methods, we combined quanti-
tative attribute analysis with a more qualitative chaîne
opératoire approach for all lithics >20 mm. Attribute
analysis on debitage products (Auffermann et al.
1990; Hahn 1991; Tostevin 2003; Odell 2004) informs
on technological behaviors by providing quantitative
data of the numerous traces on individual artifacts
that result from the knapping process. Our database
consisted of ~50 discrete and metric attributes on all
debitage products (Fig. 6). Individual stone artifacts
and attributes on these pieces constitute the unit of
analysis in this approach. For broken pieces, we
recorded only preserved attributes and linear
measurements (e.g. platform dimensions on proximal
fragments), with absent characteristics and dimensions
being classified as not assessable (na). Summary
statistics were thus performed on all recorded
attributes, with na-coding not featuring in the sample
size of the respective data analyses. While recorded
on all lithics, we use weight (in g) in further analyses
only for complete pieces. All recorded attributes are
entered into an Access Database, allowing for subse-
quent quantitative and statistical analyses on the
assemblage level or selected samples. Analyses of
collected data were conducted in Excel and SPSS to
calculate measures of central tendencies and
dispersion for individual attributes, and to conduct
further analytical tests. We also identified and
quantified lithic products <20 mm by raw material and
retouch debitage, as this size class was the most
frequent at GK. This approach aids in calculating find
densities, evaluating patterns in the raw material
economy and quantifying the level of on-site tool
production and recycling.
The more holistic and interpretative approach of
the chaîne opératoire – or reduction sequence analysis
– was used to identify the main core reduction
methods and the stages of production, use and
discard of stone artifacts performed on-site (Boëda et
al. 1990; Inizan et al. 1995; Conard & Adler 1997;
Soressi & Geneste 2011). These predominantly quali-
tative and hermeneutic analyses operate on the level
of entire assemblages and raw material units. Due to
the previously established dominance of Jurassic chert
in all MP assemblages at GK (e.g. Conard & Malina
2002, 2003), the approach was principally carried out
on the level of individual AHs. We developed diacritic
schemes based on close reading of individual artifacts
in order to recover more detailed qualitative
Fig. 5. Numerical distribution (n) of analytical categories for single finds (>20 mm) in each AH. Percentage values (%) are calculated only for
single finds (>20 mm).
Abb. 5. Nummerische Verteilung (n) der Einzelfunde (>20 mm) nach Fundkategorie je AH. Prozentangaben (%) beziehen sich ausschließlich auf
die Einzelfunde (>20 mm).
Layer Blank To ol Core Angular debris Hammerstone Manuport Small debitage (<20 mm)1Tot a l
AH IV 10 (59 % ) 2 (12 %) 03 (18 %) 1 (6 %) 1 (6 %) 23 (1) 40
AH V 13 (48 %) 3 (11 %) 3 (11 %) 4 (15 %) 3 (11 %) 1 (4 %) 37 (3) 64
AH VI 28 (51 %) 6 (11 %) 9 (16 %) 4 (7 % ) 7 (13 %) 1 (2 %) 183 (5) 238
AH VII 62 (67 %) 6 (7 %) 9 (10 %) 3 (3 %) 8 (9 %) 4 (4 %) 385 (16) 477
AH VIII 4 (44 %) 0 0 0 2 (22 %) 3 (33 %) 76 (1) 85
Tot a l 117 (59 %) 17 (9 %) 21 (11 %) 14 (7 %) 21 (11 %) 10 (5 %) 704 (26) 904
(N) = Number of identied retouch debitage pieces in brackets
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
59
information on the temporal and structural organi-
zation of removals and hence specific reduction
systems (Dauvois 1976; Inizan et al. 1995; Soressi 2002)
which are of particular interest in MP technologies.
Following a holistic approach to lithic analyses in
which multiple independent sources of evidence
converge to produce more inter-subjective and
reliable results, the findings derived from these
methods are combined and subsequently used for
both intra- and inter-assemblage comparisons. The
analyses aim to identify diachronic and synchronic
variability in Neanderthal technology at the site and
characterize mobility and land-use strategies.
Results
Taphonomy
The size distribution of lithic artifacts (Figs. 5 & 7)
indicates a relation of finds <20 mm to those >20 mm
of almost four to one, with an abundance of small
chipped pieces <10 mm (n = 488). Lithics <20 mm
constitute over half of all pieces in all assemblages,
with AHs VI-VII exhibiting >75 %. These observations
suggest on-site stone knapping by Neanderthals at
GK, though in reduced amount compared to experi-
mental Paleolithic assemblages produced by similar
methods (e.g. Bertran et al. 2012). Considering the
continuous presence of microflakes <5 mm (total
n = 92) throughout the MP sequence (Fig. 7), there has
not been a major impact of size-sorting by tapho-
nomic processes such as the removal of the smallest
particle sizes. The representativeness of the lithic
assemblages also results from the application of
modern methods of excavation and sieving of all
sediments. The MP artifacts demonstrate marked and
frequent evidence for frost and sediment damage (i.e.
cryoturbation) on a total of 69 % artifacts >20 mm,
All lithics Blank Too l Core
Raw material category Blank type Too l t ype Clast type
Raw material variety Diagnostic category # Retouched edges Unied core category
Cortex proportion Completeness Type of retouch MP core category
Edge damage Morphology Location of retouch End product
Patina Bulb Delineation of retouch Length of last end product
Frost damage Contact Point Side of retouch Last action
Fire damage / Burning traces Lipping Distribution of retouch # Removal surfaces
Hertzian cone Orientation negatives on removal surface
Shaered bulb # Plaorm surfaces
Plaorm type Orientation plaorm surfaces
Plaorm form Reason for discard
Plaorm completeness Plaorm preparation
# Dorsal negatives Dorsal reduction
Orientation dorsal negatives Core exhausted
Distal end
Maximum dimension
Maximum length
Maximum & 50 % width
Maximum & 50 % thickness
Exterior plaorm angle
Plaorm thickness
Plaorm width
Fig. 6. Discrete and metric attributes recorded for single finds >20 mm at Geißenklösterle by lithic category.
Abb. 6. M erkmalskatalog der aufgenommenen diskreten und metrischen Attribute für alle Steinartefakte >20 mm am Geißenklösterle nach Kategorie.
Fi g. 7. Histogram of lithic size categories (%) in each AH.
Abb . 7. Histogramm der Größenverteilung von lithischen Artefakten
(in %) je AH.
Quartär 66 (2019) N. J. Conard et al.
60
with these traces being similarly abundant throughout
the entire sequence (67-71 %). Sediment modifications
are also present on 56 % of the bone’s number of
individual specimens (Münzel & Conard 2004). Frost
damage on lithic artifacts is particularly frequent in
the uppermost layers IV and V (26-29 %; see also
Hahn 1988: 102-103; Conard & Malina 2002, 2003).
A higher geogenic input for the uppermost layers
is also suggested by the low proportion of debitage
products <20 mm (~58 %) compared to strata VI-VIII
(77-89 %), that also applies when only considering
pieces <10 mm (IV-V: 48-62 %; VI-VIII: 67-80 %). There
is also a diachronic trend of increasing numbers of
angular debris throughout the sequence – with many
frost shatters – without an associated change in raw
material use, suggesting a higher degree of distur-
bances and taphonomic processes damaging pieces at
the top of the depositional sequence. There is,
however, no evidence for edge rounding or of tapho-
nomic processes introducing or removing artifactual
material. While the excavations did not uncover clear
hearth structures, the use of fire is attested by the rare
occurrence of charcoal (0.12 n/m3), and more frequent
burnt bone (14.2 g/m3; Conard et al. 2012), as well as
traces of heating on 6 % of lithics >20 mm and 5 % of
pieces <20 mm. Heated artifacts and densities for
charcoal and burnt bone are most frequent in the
larger lithic assemblages of VI and VII and virtually
absent in the deepest (IV) and uppermost layer (VIII).
Assemblage composition
The assemblages are characterized by a low number
of finds >20 mm, ranging from a minimum of 9 pieces
(AH VIII) to a maximum of 92 lithics (VII). AHs VI
(n = 55) and VII (n = 92) are the largest and most robust
samples for the sequence. The quantitative analysis of
debitage products for assemblages IV-VIII shows a
relatively homogeneous distribution of products with
little diachronic change (Fig. 5), supported by a
non-significant chi-square test in AH IV-VII for the
numerical distribution of blanks and retouched pieces
(χ 2(df = 3) = 2.30, p = 0.51). Much of the variation can
be accounted for by the very low sample size in the
top and particularly lowest layers (AHs IV and VIII).
Unretouched blanks are the most frequent category
in all layers (44-67 %) except for AH VIII. Retouched
pieces (total n = 17) are absent in AH VIII but increase
throughout the sequence (6.5-11.8 %), with an overall
proportion of tools at 8.5 %. Cores are well-repre-
sented for the middle of the sequence (V-VII: 10-16 %;
total n = 21) but are lacking in the upper- and
lowermost stratum. The consistent increase of angular
debris throughout the sequence, from 0 % in VIII to
18 % in IV, is likely due to taphonomic processes (see
above). A remarkable feature of the MP assemblages
of GK is the high number of cobbles and broken
cobbles (total n = 31), most of which show traces of use
as hammerstones (see also Pop et al. 2018 more
generally). They make up ~16 % of all finds >20 mm,
are found equally frequent throughout most of the
sequence (IV-VII: 12-15 %) and are particularly
numerous in the lowermost assemblage VIII (56 %).
Artifacts <20 mm are abundant in each MP assem-
blage, particularly in the lower sequence (AHs VI-VIII:
77-89 % of all lithics; Figs. 5 & 7). For most of these
layers (V-VIII) artifacts in the size range of 5-10 mm are
the most frequent size class and microflakes (<5 mm)
also occur throughout the deposits. Although early
stages of decortification and core preparation are
under-represented and refitting sequences are
lacking, the presence of all technological products –
blanks, tools, cores, angular debris – document partial
reduction sequences with some knapping taking place
on-site. The proportion of small retouch flakes among
artifacts <20 mm ranges between 1-8 %, throughout
the sequence (total: n = 26; 3.7 %) suggesting occas ional
on-site tool production and curation.
Raw material procurement
All layers exhibit similar proportions of raw materials
(Fig. 8), supported by a non-significant chi-square test
of counts for Jurassic chert vs. all other raw materials in
layers AH IV-VII (χ²(df = 3) = 1.61, p = 0.66). Conforming
to expectat ions from previous studies, locall y acquired
rock types dominate, with no chipped lithics coming
from further than 20 km away from the site. Light-grey
and white Jurassic chert dominates overall (92.9 %)
and in each assemblage with >87 %. Assemblages with
sample sizes n>30 (AHs VI & VII) exhibit between
91.5-95.0 % of this rock type. Chipped pieces of
Jurassic chert show a large range of sizes up to 70 mm
but are on average small (~27.4 mm mean length of
blanks). Artifacts of Jurassic chert frequently possess
cortical surfaces (57 %), which stem predominantly
from primary nodules.
Other raw materials occur rarely (Fig. 8). They
include Bohnerzhornstein (3.6 %; n = 6), Muschelkalk-
hornstein (1.2 %; n = 2), quartzite (1.2 %; n = 2), and
green radiolarite (1.2 %; n = 2). None of these raw
materials occur in every layer of the sequence, they
never reach >7 % in any of the MP assemblages and the
lowest layer is devoid of these rock types. Rare cortical
surfaces are predominantly smooth and rounded,
indicating procurement of cobbles from secondary
river channels and terraces. The raw materials of the
hammerstones (n = 21) provide a deviating picture
from the chipped lithics. Here, Neanderthal
toolmakers mostly used Cretaceous quartzite (62 %),
which is found throughout almost all assemblages.
Other rock types employed for the use as hammer-
stones include sandstone and quartz (14 % each) as
well as rare red radiolarite and quartzite (5 % each).
Blanks
Neanderthal knappers almost exclusively manufac-
tured flakes, dominating overall (94 %) and in each AH
(75-96 %), with blades playing a minor role for the
total assemblages (n = 8; 6 %). The proportion of flakes
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
61
to blades conforms to a homogeneous distribution
throughout the entire MP sequence (χ²(df = 4) = 2.95;
p = 0.57). Considering only AHs with a blank sample of
n>10, blades never account for more than 8 %, and
some appear to be by-products of other reduction
strategies that focused on the manufacture of flakes.
No convergent flakes, crested blades or bladelets
occur among the blanks. A similar pattern is observed
when comparing raw materials: Flakes predominate
for each raw material (>80 %) with 94.4 % of blanks on
the most frequently used Jurassic chert. The few
blades were solely manufactured on Jurassic chert
(n = 7; 5.6 %) and Bohnerzhornstein (n = 1; 20 %).
Most blanks do not exhibit any cortex (44 %), with
a further 36 % showing cortical surface values of up to
a third of the piece (Fig. 9). Blanks with over a third of
dorsal cortex amount to ~20 %. There is only a single
fully cortical flake in the total assemblage. Among all
blanks, facetted platforms amount to ~27 %, indicating
frequent core preparation by the inhabitants of the
site. There is an overall chronological trend of
decreasing platform preparation, with 24-38 % in AHs
VI & VII and markedly lower values in AHs IV & V
(11-14 %). The majority of all platforms is plain (~55 %)
and only few cortical platforms occur (~9 %). A
consistent proportion of ~21-40 % (mean = 30 %) of
blanks is complete (n = 52) with the majority being
broken pieces.
An important feature of the blanks from the MP at
GK is their consistently small size. Overall, knappers
manufactured flakes that are on average only ~26 mm
long, never exceeding 60 mm. Only 9 % (n = 5) of all
measurable blanks are longer than 40 mm. Flakes
throughout the sequence V-VIII remain constantly
small, ranging between a mean of 20.0-25.6 mm, with
AH IV being an outlier with a mean of 38.8 mm.
Considering their overall shape and elongation, most
blanks (56 %) are slightly elongated. There are no
unidirectional temporal trends in elongation
throughout the sequence (Kendall rank correlation
coefficient: τb = -0.527, p = 0.207; n = 5). From a
morphological perspective, most flakes are either of
trapezoidal or rectangular shape (85 %), a pattern
consis tently found in all AHs. A large proportion of
flakes possesses backs, formed either by a cortical
edge or steep removal scars.
Most blanks exhibit only 1-3 dorsal scars (71 %)
with an almost complete absence of pieces with >6
negatives (1 %). Regarding the orientation of dorsal
scars, orthogonal (40.8 %) and unidirectional (36.8 %)
AH IV AH V AH VI AH VII AH VIII Tot a l
>20 mm
Jurassic chert 13 (87 %) 21 (91 %) 43 (92 %) 76 (95 %) 4 (100 %) 157 (93 %)
Bohnerzhornstein 1 (6. 5 %) 1 (4.5 %) 04 (5 %) 06 (4 %)
Muschelkalkhornstein 0 0 2 (4 %) 0 0 2 (1 %)
Quartzite 1 (6.5 % ) 01 (2 %) 0 0 2 (1 %)
Radiolarite 01 (4.5 %) 1 (2 %) 0 0 2 (1 %)
<20 mm
Jurassic chert 22 (96 %) 37 (100 %) 172 (94 %) 366 (95 %) 67 (88 %) 664 (94 %)
Bohnerzhornstein 0 0 1 (0.5 %) 9 (2.5 %) 2 (2. 5 %) 12 (2 %)
Muschelkalkhornstein 000000
Quartzite 1 (4 %) 04 (2 %) 4 (1 %) 4 (5 %) 13 (2 %)
Radiolarite 0 0 6 (3.5 %) 6 (1.5 %) 3 (4 %) 15 (2 %)
Fig. 8. Raw material distribution (n) for lithic finds >20 mm and <20 mm in each AH, flaked pieces only.
Abb. 8. Rohmaterialverteilung (n) der geschlagenen Steinartefakte >20 mm und <20 mm je AH.
Cortex AH IV AH V AH VI AH VII AH VIII To t a l
0 % 8 (53 % ) 9 (39 %) 19 (40 %) 35 (44 %) 3 (75 %) 74 (44 %)
1-33 % 4 (27 %) 8 (35 %) 17 (36 %) 30 (38 %) 1 (25 %) 60 (35. 5 %)
34- 66 % 2 (13 %) 4 (17 %) 8 (17 %) 11 (13 %) 025 (15 %)
67-9 9 % 1 (7 %) 2 (6 %) 3 (6 %) 3 (4 %) 09 (5 %)
100 % 0001 (1 %) 01 (0. 5 %)
Tot a l 15 23 47 80 4169
Fig. 9. Numerical distribution (n) and frequency (%) of cortex proportions in each AH.
Abb. 9. Nummerische Verteilung (n) und Häufigkeit (%) des Kortexanteils je AH.
Quartär 66 (2019) N. J. Conard et al.
62
patterns dominate, with only a few pieces possessing
bidirectional (5.6 %), convergent (2.4 %) or centripetal
(1.6 %) negatives. These observations overall suggest
that the inhabitants of GK followed a unifor m approach
to produce predominantly small flakes of rectangular
shapes with frequent backs and thereby resulting in
asymmetric triangular cross-sections (Fig. 10).
Cores
Only assemblages V-VII have yielded cores (total
n = 21). The majority (85.7 %) is made on Jurassic chert,
while Bohnerzhornstein, Muschelkalkhornstein and
green radiolarite comprise each 4.8 % of the total
cores. In this study, we apply both the unified core
taxonomy by Conard et al. (2004) as well as a standard
MP core taxonomy (e.g. Boëda et al. 1990). According
to the unified core taxonomy, parallel cores are the
most frequent (57 %), followed by platform (24 %) and
bipolar (14 %) types without any clear diachronic
trends. Only Jurassic chert features cores for all of
these categories. The 12 parallel cores at GK all
conform to a Levallois reduction system and include
unidirectional (n = 6), preferential (n = 3), centripetal
(n = 2) and bidirectional (n = 1) modalities (Figs. 11 &12).
Among all core types, only Levallois is found on all raw
materials. The platform cores include single- and
multi-platform types, with the single-platform cores
exclusively knapped on flakes and corresponding
broadly to Kostenki reduction sequences (n = 4;
Figs. 11 & 13). We did not observe inclined cores,
including discoid, or platform types aiming at laminar
production.
The cores predominantly exhibit unidirectional
dorsal negatives (47 %) on their main removal surfaces,
Fig. 10. Levallois core edge f lakes (débordants) and pseudo-Levallois points from the MP horizons of Geißenklösterle (Drawings by S. Boos
and H. Würschem).
Abb. 10. Levallois Kernkantenabschläge (débordants) und Pseudo-Levallois-Spitzen aus den MP-Horizonten des Geißenklösterle (Zeichnungen
von S. Boos und H. Würschem).
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
63
but also a variety of other orientations including
preferential (20 %), orthogonal (13 %), bidirectional
(13 %) and centripetal (7 %) appear. Preparation of
core striking surfaces is common (48 %) consistent
with frequent faceting on blank platforms. The cores
feature few platform and removal surfaces. Both
nodules (n = 9) as well as larger flakes (n = 8) were
predominantly used as original core blanks. Whereas
over three fourths of the cores exhibit cortex, their
extent is mostly between only 1-33 %.
All 21 specimens are flake cores. The cores are
small with a median maximum dimension of only
36 mm (mean: 38.1 mm) and a median weight of 13.7 g
(mean: 18.7 g), conforming to the small size of blanks.
The largest core weighs 95.4 g with a maximum
dimension of 70 mm, whereas the smallest core is only
3.0 g and 20 mm. The Jurassic chert cores (n = 18)
encompass the largest range but are also small and
light on average (median: 35 mm; 11.9 g). The length of
the last products on cores averages 19.9 mm (median:
19.5 mm), with a maximum of 37 mm. The small dimen-
sions, high number of removals and the lack of cortex
on the main removal surfaces suggest that most of the
cores can be considered as being discarded in an
exhausted state on-site. A third of the cores were
abandoned due to knapping accidents impeding a
continuation of the reduction.
Retouched elements
The identification of retouched pieces in the MP of
GK is complicated by post-depositional processes
which led to frequent edge damage (“cryo-retouch”)
on 69 % of all pieces. We thus followed a conservative
approach that only counted chipped lithics as tools
when they exhibit systematic anthropogenic modifica-
tions. We define these traces as continuous and
preferably multiple generations of retouch scars
(preparation, finishing, sharpening) with visible
negative bulbs. Our total count of retouched pieces
(n = 17) is lower compared to previous assessments
(e.g. Conard & Malina 2002). Still, frequencies of tools
are relatively high, ranging between 7-12 % per assem-
blage with exception of the lowermost layer VIII which
lacks retouched elements.
From a typological point of view, various types of
scrapers are the most frequent tool types (82 %),
followed by splintered pieces (18 %) (Figs.14 &15).
The scraper types encompass frequent side scrapers
(n = 9) but also transverse (n = 3), convergent (n = 1) and
end scrapers (n = 1). Interestingly, the MP assemblages
of the site do not feature denticulates, notches or
bifacial implements such as Keil messer. There are also
no clear diachronic trends, with scrapers occurring
and dominating throughout the sequence (Fig. 14).
Tool diversity is low for AHs IV-VII, ranging between
2-3 different types. While all tool types are found on
local Jurassic chert, splintered pieces were more
frequently made on imported raw materials such as
Muschelkalkhornstein and radiolarite.
The inhabitants of GK exclusively selected flakes
for tool manufacturing with a preference for rectan-
gular or trapezoidal morphologies. Often, these
pieces show an asymmetric cross-section with a back
opposite to the retouched edge. Many of the
retouched blanks are broad and wider than long.
Importantly, tools are on average markedly and signif-
icantly larger in maximum size compared to
unretouched blanks (Mann-Whitney U test: U = 491.5;
n = 134; p<0.001), particularly for scrapers that reach a
maximum dimension of close to 70 mm. Here, side
(mean: 36.3 mm; median 30 mm) and particularly
transverse (mean 40.3 mm; median 41 mm) scrapers
are on average between 5-10 mm larger than blanks in
maximum dimension. Splintered pieces constitute the
smallest implements with a mean maximum dimension
of only 27.3 mm (median 28 mm) and never exceeding
29 mm.
The knappers applied retouch predominantly to
the dorsal face of the blanks (75 %) but also in an alter-
nating fashion (25 %). This retouch remains distributed
on the edges of the pieces and does not conform to
unifacial or bifacial shaping. Small stepped and scalar
negatives are the most abundant modification type.
The retouch is often marginal in a convex trajectory
along the edge. Secondary modifications rarely
include several layers of small overlapping negatives
that reach further into the piece. Likewise, retouch
usually covers short parts of the artifact edges, but
some tools exhibit more than one retouched edge
(30 %).
Techniques and methods
The technical act of detaching a flake from a core
constitutes a major variable in technological behavior.
A total of 63 blanks preserve original platform dimen-
sions. They demonstrate a mean platform thickness of
~5 mm in each assemblage (modal value = 4 mm) with
few platforms thinner than 2 mm (1.6 %). Exterior
platform angles (EPA) cluster around 83-88°. Based on
an assessment of all complete blanks and proximal
fragments (n = 72), bulbs are very frequent (94.4 %)
and often strongly developed (69 %) with visible
contact points on over half of the pieces (53 %). Typical
diagnostic features for direct hard stone hammer
percussion like Hertzian cones (14 %) and eraillure
scars (12.5 %) are also present in this sample. Lips
occur in very low frequency (1.4 %). The relatively
high frequency of longitudinal breaks on flakes
(~30 %) is consistent with strong forces exerted by
hard stone hammers that had direct contact with the
core. Based on these traits on the assemblage-level,
blanks in all AHs were predominantly knapped using a
hard hammerstone with direct and internal percussion
a couple of millimeters away from the core edge (e.g.
Pelegrin 2000). In addition, the high frequency of
shattered bulbs (25 %) could tentatively suggest some
application of soft stone hammers (pierre tendre; see
Pelegrin 2000; Roussel et al. 2009).
Quartär 66 (2019) N. J. Conard et al.
64
These observations match well with data on the
abundant hammerstones and hammerstone fragments
throughout the sequence (n = 21). Neanderthals used
cobbles from secondary deposits such as river terraces
encompassing a wide range of raw materials with most
specimens that would be classified as hard (Creta-
ceous quartzite; quartz; red radiolarite; quartzite;
n = 18) and some soft stone hammers (sandstone;
n = 3). Fig. 16 depicts the variety of shape and size in
hammerstones: often they are elongated, relatively
flat and oval in shape. The size of complete hammer-
stones is variable, ranging from 30-206.7 g (median:
55.5 g), 39-63 mm in maximum dimension (median: 52
mm), 29-63 mm in width (median: 42 mm) and 13-38
mm in thickness (mean: 22 mm). Quartz hammerstones
(median: 136.8 g; 66 mm length) are larger compared
to the most frequent Kreidequarzit specimens (median:
52.4 g; 51.5 mm length). Most hammerstones exhibit
one or several pitted surfaces on their edges as macro-
scopic use-wear from their function (see e.g. Pop et al.
2018). Frequent breakage that split the cobble is a
further indication of intense use and then discard of
broken stone hammers at the site.
Regarding knapping methods, Neanderthals
applied a variety of reduction strategies including
Levallois, Kostenki, bipolar and multiple-platform
(Fig.11). Here we provide a more detailed qualitative
and holistic assessment of the different reduction
sequences based on readings of both cores and flakes.
Levallois core reduction is by far the most abundant in
both cores and products, followed by some Kostenki
reduction and rare use of bipolar and other platform
methods. These observations apply particularly to
AHs V-VII, with few diagnostic products for the top-
and lowermost layers (Fig. 17).
The majority of cores in the MP assemblages (12 of
21) are characterized by two hierarchical, asymmetric
and non-interchangeable surfaces, sometimes with
intense preparation of the striking platforms (Fig. 12).
Knappers used both nodules (67 %) and flakes (33 %)
as initial form for these cores. The lower face is often
covered with cortex and only sometimes shows prepa-
ration removals. Knappers prepared the lateral and
distal edges of the core with centripetal removals to
create a convex removal surface. The products
deriving from this system include central flakes (n = 3)
with facetted platforms which have a circular shape as
well as orthogonal or centripetal dorsal scar patterns
and exhibit EPAs >80°. Products removed along the
lateral edges of these cores are very frequent (Fig. 17)
and include both core edge flakes (éclats débordants,
n = 30) and pseudo-Levallois points (dos limités, n = 9;
Fig. 10). The described cores and flakes conform to a
classic Levallois system of reduction (sensu stricto;
Boëda et al. 1990; Boëda 1993) and encompass various
modalities of production. The majority of the cores
and respective blanks demonstrates unidirectional
recurrent removals (cores n = 6), followed by less
frequent preferential (cores n = 3), centripetal (cores
n = 2), and bidirectional (cores n = 1) scar patterns (see
Figs. 11 & 12). The various modalities occur on small
cores (median weight = 13.8 g; range = 9-41.7 g; median
MD = 36.5 mm; range = 31-53 mm) that are all in
advanced stages of reduction, suggesting the appli-
cation of multiple re duction strategies is not associ ated
with decreasing core size.
The second most common reduction method
encompasses a sub-category of platform cores on
flake and amounts to 4 specimens. All of these cores
are knapped on flakes of generally small size (weight:
mean = 13.0 g; median = 4.8 g; length: mean/median
length = 32.0 mm). In this strategy, a single platform is
set up on the proximal end of a relatively thick flake
(9-15 mm) with a steep inverse truncation of the
platform serving as preparation for the subsequent
exploitation on the previous dorsal surface of the
flake. The striking platforms of these cores are mostly
plain. A preparation of the lateral and distal edges
occurred on the dorsal surface to create additional
longitudinal and transversal convexities documented
by small preparation flakes exhibiting a short back and
orthogonal removals. The unidirectional reduction of
end products on the single main removal surface
proceeded along main ridges of the dorsal removals.
Knappers removed small and elongated flakes from
these core types. The configuration, geometry and
Type AH IV AH V AH VI AH VII AH VIII Total
Levallois unidirectional 014106
Levallois preferential 002103
Levallois bidirectional 001001
Levallois centripetal 000202
Kostenki 002204
Bipolar 010102
Plaorm 000101
Discoid 000000
Tes ted 010102
Fi g. 11 . Numerical distribution (n) of core types in each AH.
Abb . 11. Nummerische Verteilung (n) der unterschiedlichen Kerntypen je AH.
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
65
temporal succession of the removals on these cores
are depicted in Fig. 13. The described cores fit broadly
within the definition of the Kostenki method for the
production of small blanks from a dorsal flake surface
(e.g. Demars & Laurent 1989; Dibble & McPherron
2007; Frick 2013). The Kostenki reduction sequence at
GK corresponds to the use of flakes that originate
from a primary production, namely Levallois
reduction, as the basis for secondary production
paralleling the primary one. These flakes transformed
into cores provide a new generation of smaller blanks
according to the principle of ramification (see for
definition Geneste 1991; Bourguignon et al. 2004).
The form of production is considered to be associated
with high mobility and a demand for small tools for
specific tasks (Rios-Garaizar et al. 2015).
Other platform approaches constitute a third but
rare reduction system evident on a single core. This
core has two striking platforms adjacent to each other
and two removal surfaces located orthogonal to each
other, reduced by means of unidirectional, recurrent
removals. The negatives of one of the two removal
surfaces form the platform for the second removal
surface, from which the final flake was removed.
Neanderthal knappers set up the core without prepa-
ration and established non-hierarchical surfaces by
exploiting them successively or alternatively. This
double platform type is reminiscent but not identical
to Quina reduction (e.g. Bourguignon 1997; Delagnes
et al. 2007). The small sample of cores and potential
products for this method preclude unequivocal
assessment. Finally, rare bipolar reduction rounds out
the methods used by Neanderthals at GK. The two
cores of this system express impact scars on opposite
ends with edge crushing and splintering, and an
elongated wedge shape. Small flakes with bidirec-
tional scars are the main products of this strategy. This
being said, bipolar products are rare throughout the
GK sequence (4.5 %).
Raw material economy
Regardless of their small size, most MP assemblages
exhibit products from the majority of the reduction
sequences – small debitage, blanks, tools, cores and
angular debris. The small samples from layers IV
(n = 17) and VIII (n = 9) lack cores, and AH VIII also lacks
tools and angular debris with an overrepresentation
of hammerstones. Assemblages with larger sample
sizes are characterized by more complete reduction
sequences, particularly the richer find horizons VI and
VII. Some on-site knapping, as well as occasional
retouching, are documented by small debitage
products and microflakes <5 mm throughout the
entire sequence. The low values for cortex on knapped
lithics – between 0-7 % pieces with cortical surfaces
>70 % and over three fourths with <33 % (Fig. 9) –
suggest that little decortification took place at the site
which matches the near-absence of large cortical
Fig. 12 . Various modalities of small Levallois cores from the MP horizons of Geißenklösterle. Levallois flake removals in blue (Drawings by S.
Boos and H. Würschem; Photographs by V. C. Schmid).
Ab b. 12. Unterschiedliche Varianten von kleinen Levallois-Kernen des MP aus dem Geißenklösterle. Levallois-Abschläge in blau markiert (Zeich-
nungen von S. Boos und H. Würschem; Photographien von V. C. Schmid).
Quartär 66 (2019) N. J. Conard et al.
66
manuports and tested cores. The average size and
weight of cores of Jurassic chert (median: 35 mm;
11.9 g) lie well below the average dimensions of the
primarily available nodules that range from golf ball to
soccer ball size. The assemblages demonstrate a focus
on the introduction and reduction of prepared cores
in an already advanced state, as well as some
production and modification of blanks.
These patterns are associated with export of
finished blanks and tools. We observed a conspicuous
lack of end-products for the various modalities of the
Levallois reduction system: There are 12 Levallois
cores with multiple negatives of flake removals,
providing evidence for the production of ~20-30
Levallois flakes using a conservative estimate of 2-3
flakes per core. Yet, the combined assemblages only
yield three definite central Levallois flakes (Zielab-
schlag) that were left behind in the MP at GK. In
contrast, the assemblages are characterized by an
abundance of core edge flakes and pseudo-Levallois
points (n = 39), which correspond more to preparatory
operations. Yet, several of these products were also
selected for retouch. This suggests first an export of
some finished Levallois flakes from the site, but also
indicates the frequent discard of asymmetrical
core-edge flakes and their occasional transformation
Fig. 13. Kostenki-like cores and reduction concepts from the MP horizons of Geißenklösterle. Dorsal preparation in blue, final removals in
light-red (Drawings by S. Boos and H. Würschem; Photographs by V. C. Schmid).
Ab b. 13. Kostenki-ähnliche Kerne und Abbaukonzepte aus den MP-Schichten des Geißenklösterle. Dorsale Präparationen in blau, Zielabschläge
in hellrot (Zeichnungen von S. Boos und H. Würschem; Photographien von V. C. Schmid).
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
67
into scrapers with an opposed back that were
discarded on-site (Fig. 18).
A clear pattern of raw material economy can be
inferred from the assemblages. Each AH features more
or less complete reduction sequences for the local
Jurassic chert, indicating its on-site production and
re-sharpening. The relatively low number of blanks in
relation to cores for this raw material (6.2 blanks per
core) also suggests some transport of knapping
products to other places in the landscape (see above).
In contrast, other raw materials of low abundance and
from slightly further away (e.g. Bohnerzhornstein,
Muschelkalkhornstein and radiolarite) are predomi-
nantly present as isolated tools and exhausted cores
that were likely imported by Neanderthals as single
pieces with little to no on-site knapping.
Several lines of quantitative evidence support the
qualitative observations of a differential treatment for
the frequently used Jurassic chert (93 %) compared to
the other raw materials. First, the very low absolute
number of Bohnerzhornstein, Muschelkalkhornstein
and radiolarite for each layer (n = 2-6; >20 mm) that
also applies to small debitage (n = 12-15; <20 mm;
Fig.8) stands out. Small debitage for Bohnerzhornstein
(mean: 1.2 %; range: 0-2.6 %) and radiolarite (mean:
1.7 %; range: 0-3.9 %) is proportionally even less
frequent compared to products >20 mm, whereas
Jurassic chert shows a slightly higher overall proportion
of small chipped pieces (94.3 %). There is a high
frequency of tools for Bohnerzhornstein (17 %),
Muschelkalkhornstein (50 %) and radiolarite (50 %)
compared to Jurassic chert (9 %). Overall,
Layer Side scraper Transverse scraper End scraper Convergent scraper Splintered piece Tot a l
AH IV 1 (50 %) 01 (50 %) 0 0 2
AH V 1 (33 %) 1 (33 %) 0 0 1 (33 %) 3
AH VI 3 (50 %) 1 (17 %) 0 0 2 (33 %) 6
AH VII 4 (67 %) 1 (17 %) 01 (17 %) 06
AH VIII 0 0 0 0 0 0
Tot a l 9 (53 %) 3 (18 %) 1 (6 %) 1 (6 %) 3 (18 %) 17
Fig. 14. Numerical distribution (n) and frequency (%) of tool types in each AH.
Abb. 14. Nummerische Verteilung (n) und Häuf igkeit (%) unterschiedlicher Werkzeugtypen je AH.
Fig. 15. Selection of tools from the MP horizons of Geißenklösterle. The depicted tools encompass various scraper forms except for a splin-
tered piece in the bottom right (Drawings by S. Boos and H. Würschem; Photographs by V. C. Schmid).
Ab b. 15 . Auswahl an Werkzeugen der MP-Horizonte des Geißenklösterle. Die dargestellten Werkzeuge sind unterschiedliche Schaberformen, mit
Ausnahme des ausgesplitterten Stückes unten rechts (Zeichnungen von S. Boos und H. Würschem; Photographien von V. C. Schmid).
Quartär 66 (2019) N. J. Conard et al.
68
Fig. 16. Selection of hammerstones from the MP archaeological horizons of Geißenklösterle (Drawings by S. Boos).
Ab b. 16. Auswahl an Schlagsteinen aus den MP-Schichten des Geißenklösterle (Zeichnungen von S. Boos).
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
69
Bohnerz hornstein, and even stronger Muschelkalkhorn-
stein and green radiolarite are overrepresented
among the tools (including splintered pieces) in
relation to their proportion compared to Jurassic
chert. Higher cortex values (>70 %) occur exclusively
on Jurassic chert with the majority of the other raw
materials showing no or only few cortical surfaces
(<33 %). Blanks on Bohnerz hornstein, Muschelkalkhorn-
stein, radiolarite and quartzite are larger and thicker
(mean length = 32-49 mm; mean thickness = 10-15 mm)
compared to the generally smaller products from
Jurassic chert (mean length = 27.4 mm; mean
thickness = 9.9 mm) which were often more strongly
reduced on-site. These observations demonstrate a
predominantly localized provisioning strategy by the
Neanderthal inhabitants focused on the local light-
gray Jurassic chert, with rare use of other rock types.
In sum, the data support a techno-economic
scenario of several short-term episodes of import,
production, retouching as well as resharpening,
discard, and export. Thus knapping as well as
abandonment of exhausted cores and tools took place
at the site, while Neanderthal toolmakers transported
desired end-products off-site and carried them as
mobile toolkit along from place to place (Porraz 2009),
complying with an individual provisioning strategy to
cope with immediate and anticipated technological
needs (Kuhn 1995, 2004).
Discussion
The MP assemblages of Geißenklösterle: Charac-
teristics and diachronic trends
Our results represent the first comprehensive analysis
of the MP artifact assemblages from GK. These numer-
ically small assemblages derive from multiple archaeo-
logical horizons in a secure stratigraphy, possess
detailed contextual information, and are associated
with absolute dates. Based on these features, the MP
assemblages at GK provide new insights into
Neanderthal technology and lifeways during the Late
Pleistocene of the Swabian Jura.
Various Levallois modalities constitute the most
frequent core reduction strategies documented in all
assemblages. The assemblages, however, feature only
few typical Levallois blanks. Neanderthals also used
Kostenki-like reduction, and rarely exploited bipolar
and platform cores at the site. We, however, have not
identified discoid or laminar reduction at the site. The
knappers predominantly produced small, elongated
flakes of trapezoidal or rectangular shape, often with
asymmetrical cross-sections (Figs. 10 & 18). Only a few
Type AH IV AH V AH VI AH VII AH VIII Total
Levallois end product 001203
Levallois core edge (Débordant)33519 030
Pseudo-Levallois point (Dos limité)014409
Kostenki 001001
Bipolar 121206
Quina 001001
Laminar 000000
Bifacial 000000
Undiagnostic 810 21 40 483
Fi g. 17. Numerical distribution of technologically diagnostic pieces for single finds (>20 mm; except cores) in each AH.
Abb . 17. Nummerische Verteilung von technologisch diagnostischen Stücken innerhalb der Einzelfunde (>20 mm; ohne Kerne) je AH.
Fig. 18 . Summary of the operational sequence discerned from the MP lithic assemblages of Geißenklösterle.
Abb. 18 . Zusammenfassung der Operationskette, abgeleitet aus den MP-Steinartefaktinventaren des Geißenklösterle.
Quartär 66 (2019) N. J. Conard et al.
70
flakes exceed 40 mm in maximum dimension. The
knappers mainly used hard hammer percussion as is
attested by a high number of hammerstones from
local river cobbles. Different types of scrapers are the
most frequent tool types, followed by splintered
pieces. Based on what was discarded at GK,
Nean derthals preferentially selected relatively large
blanks with asymmetrical cross-sections for scraper
forms, and applied retouch along the thinner edge of
the flakes as a working edge opposite to the natural
backing presumably representing a passive or
prehensile part. The assemblages completely lack
notches, denticulates or bifacial implements such as
Keilmesser or Blattspitzen.
Throughout all layers, the inhabitants predomi-
nantly knapped local Jurassic cherts (>90 %) with only
small amounts of other raw materials. The assemblages
demonstrate relatively complete reduction sequences
for Jurassic chert – particularly in AH VI and VII – with
a notable underrepresentation of Levallois blanks. A
low frequency (2-10 %) of small retouch flakes, along
with abundant small artifacts between 5 and 10 mm in
size (n = 396) and less frequent microflakes smaller
than 5 mm (n = 92) attest to some on-site blank and
tool production in the MP horizons of GK (see e.g.
Bertran et al. 2012). This pattern stands in contrast to
less common lithic raw materials including radiolarite,
Bohnerzhornstein, Muschelkalkhornstein and black
alpine microquartzite, which are characterized mainly
by isolated flakes and tools from the distal ends of
their biographies. Small (n = 30) and micro-artifacts
(n = 7) are almost completely absent for these rare raw
materials. Thus in contrast to Jurassic chert, artifacts
on the less frequent rock types were mostly brought
to the site in finished form and then discarded at GK.
The MP archaeological sequence at GK consists of
five AHs, allowing diachronic observations. Although
our analyses reveal some inter-assemblage variation,
there are more similarities than differences in all
principle technological domains and raw material use.
The remaining diachronic differences likely stem from
small sample sizes particularly for layers IV and VIII.
The largest assemblages, AH VI & VII, are similar in
their techno-typological characteristics. Assemblages
IV-VIII from GK appear to represent a stable techno-
typological and techno-economic pattern of behavior
within the MP of the Swabian Jura.
Geißenklösterle in the context of the Swabian
Middle Paleolithic
The MP assemblages from GK offer a well-studied
basis for further comparisons on the local and regional
level. We are interested in how GK fits within previous
work on MP lithic technology in the Swabian Jura and
the general cultural stratigraphy of Germany, Central
Europe and beyond (see next section). Numerous sites
and studies of MP assemblages from the Swabian Jura
allow for comparisons (Schmidt 1912; Peters 1931;
Riek 1934; Wetzel & Bosinski 1969; Hahn 1988; Beck
1999; Böttcher et al. 2000; Çep & Waiblinger 2001;
Conard 2005, 2011; Conard et al. 2006, 2012; Bolus
2011; Çep 2013; Bolus 2015; Çep & Krönneck 2015).
Many of these assemblages have not been excavated
by modern standards and often lack contextual infor-
mation, comparable quantitative data and absolute
dates. Thus we compare strategies of raw material
procurement and techno-typological characteristics
of GK to the most important sites of the Swabian Jura
as far as data permit.
The assemblages of GK are comparable to many
other Neanderthal sites in the Swabian Jura in featuring
small assemblages and a low density of lithic artifacts
(GK: 24-146 n/m3), such as Hohle Fels (33-703 n/m3),
Sirgenstein (~20 n/m3), Kogelstein (203-318 n/m3),
Große Grotte (1-14 n/m3), and Vogelherd (30-320 n/m3;
density values from Conard et al. 2012; see also Riek
1934; Wagner 1983; Conard et al. 2006, Conard 2011;
Bolus 2015). This also applies to other classes of finds.
In this regard, GK contrasts sharply with larger lithic
assemblages such as Bockstein and Heidenschmiede
(Peters 1931; Wetzel & Bosinski 1969; Çep 2014; Çep
& Krönneck 2015). Raw material procurement is
predominantly local at all MP sites of the Swabian Jura.
The overall proportions of Jurassic chert at GK (~93 %)
are similar to Hohle Fels (86-96 %; Conard & Malina
2013) but higher compared to other MP sites such as
Kogelstein (45 % local Jurassic chert; 75 % all Jurassic
chert), Sirgenstein (VII/VIII = 76 %), Hohlenstein-
Stadel (62 %), Hohlenstein-Bärenhöhle (85 %) and
Große Grotte (63-78 %; data from Beck 1999; Böttcher
et al. 2000; Çep 2013: Fig. 4). As is the case at many
other sites in the region (Çep & Waiblinger 2001; Çep
& Krönneck 2015), incomplete reduction sequences
with isolated artifacts characterize the use of rock
types other than Jurassic chert at GK.
Regarding general technological aspects, various
modalities of the Levallois concept dominate most MP
sites of the region, including Hohle Fels, Sirgenstein,
Große Grotte, Hohlenstein and Vogelherd. Another
unifying feature between these sites and GK is the
observation of frequent small and intensely exhausted
Levallois cores. Comparable metric data exists for the
assemblages at Hohlenstein which provide an average
size of prepared cores between ~48-52 mm (Beck
1999: Table 19 & Table 47) with respective cores at
GK being even smaller (mean: 37 mm; range:
25-53 mm). The end-products of these reduction
systems are also of small size: Levallois flakes at GK lie
at an average length ~30 mm and comparable data
from Hohlenstein at 36-38 mm (Beck 1999: Table 8 &
Table 40). In combination, these characteristics are
often used to describe the technocomplex of the
“Swabian Mousterian” (Schmidt 1912; Riek 1934;
Wagner 1983; Beck 1999; Böttcher et al. 2000; Conard
et al. 2006, 2012; Conard 2011; Bolus 2015). While we
originally viewed these highly reduced cores and
debitage products as characteristic of find horizons in
caves like GK and Hohle Fels, preliminary results from
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
71
Fröhle’s doctoral research suggest that such highly
reduced cores are also well represented at open-air
sites in Baden-Württemberg (Fröhle et al. 2019).
Although Kostenki cores were also reported for the
Aurignacian of GK and Vogelherd, here associated
with bladelet production (Hahn 1988: 285, Fig. 21.1;
Hahn 1991), comparable pieces have not been
identified at any other MP sites of the region,
documenting an aspect of variability at these sites.
This study illustrates a further degree of diversity in
reduction strategies at GK, with the presence of
bipolar and platform cores.
With a tool proportion of 8.5 % (>20 mm) GK
compares well to the MP at Kogelstein (7.2 %; calcu-
lated from Böttcher et al. 2000: Table 38 excluding
small debitage <10 mm), Sirgenstein (3.8 % in layer VII/
VII; Çep 1996), and Hohle Fels (3-8 %; Conard &
Malina 2013), but differs strongly from Hohlenstein-
Stadel (23.9 %; Beck 1999), Hohlenstein-Bärenhöhle
(33.9 %; Beck 1999), and Bocksteinschmiede (~20 %;
Çep 2014). Some of these older excavations, however,
did not systematically recover small lithics, resulting in
a bias towards higher frequencies for retouched
pieces. Typologically most similar to GK are MP assem-
blages attributed to the Swabian Mousterian, defined
by an array of diverse scrapers and found at sites such
as Hohle Fels, Sirgenstein, Hohlenstein, Große Grotte,
and Vogelherd. The observation of larger sizes for
retouched compared to unretouched pieces, and thus
a selection of larger blanks for further modifications,
unites the assemblages from Hohlenstein (Beck 1999:
155) and GK. Yet, GK differs from these assemblages
with regard to the occurrence of splintered pieces.
Various bifacial pieces – Blattspitzen or bifacially
backed knives (Keilmesser) – characterize the other
main technocomplexes of the region, the Blattspitzen-
and Keilmessergruppen (Richter 1997, 2016; Conard &
Fischer 2000; Jöris 2003; Bolus 2004b, 2011, 2015)
most notably at Bockstein, Heidenschmiede and
Haldenstein (Peters 1931; Riek, 1938; Wetzel &
Bosinski 1969; Bolus & Rück 2000; Çep 2014; Çep &
Krönneck 2015). While isolated finds of these bifacial
artifacts can be found in almost all MP assemblages of
the Swabian Jura, complicating their use as chrono-
cultural markers (Çep & Krönneck 2015; Herkert et al.
2015), none of the archaeological horizons at GK has
yielded bifacially retouched tools or debitage from
bifacial shaping. The absence of these products at GK
appears to be a distinct aspect of the MP
assemblages.
Our local comparisons indicate that the MP assem-
blages from GK correspond closest to the Swabian
Mousterian (Schmidt 1912; Riek 1934; Beck 1999;
Conard 2011; Conard et al. 2012; Bolus 2015) – synony-
mously denoted as “Albhöhlen Moustérien” by some
(Wagner 1983: 56). The absence of bifacial technology
separates GK clearly from the Blattspitzen- and
Keil messergruppen. The findings from GK, however,
provide additional elements of variability to the
definition of the Swabian Mousterian: the exploitation
of cores corresponding broadly to Kostenki reduction
and to a lesser degree bipolar as well as platform
cores and the occurrence of splintered pieces. A
re-evaluation of other assemblages attributed to the
Swabian Mousterian is required to check whether
these aspects are unique to GK – and, if so, why,
considering the comparable use of raw materials and
other unifying characteristics of assemblages from this
technocomplex.
Unlike the other sites in the region, GK provides a
rare glimpse into the relative and absolute timeframe
of the Swabian Mousterian. Absolute dating at the site
indicates that the technocomplex lasted from MIS 5
into MIS 3. The stability and time depth of the Swabian
Mousterian is also demonstrated by the five consec-
utive archaeological layers with similar techno-
typological characteristics. At GK the Swabian
Mousterian is followed by a largely geogenic horizon,
marking the technocomplex as the final Neanderthal
occupation at the site and potentially lasting until
~45-43 ka BP (Richter et al. 2000; Higham et al. 2012,
2014; Goldberg et al. 2019). In this regard, the MP
from GK has the potential to contribute to the chrono-
cultural stratigraphy of Neanderthals in the Swabian
Jura and Central Europe.
The MP of Geißenklösterle in a broader geographic
context
To what extent is the Swabian Mousterian at GK
comparable to the MP in other regions of Central
Europe? How can we explain the variable presence or
absence of bifacial technological elements in the MP
of southwestern Germany and beyond (i.e. cultural,
temporal or functional)? In the following, we will
restrict ourselves to only the most informative compar-
ative sites relevant to our questions from stratified
contexts, with detailed data on lithic assemblages in
the timeframe ~90-40 ka and a focus on Germany.
One of the most informative sites on MP assem-
blages with and without bifacial technology are the
well-studied and find-rich assemblages from
Sesselfelsgrotte in the Altmühl Valley (Bavaria) close to
the Swabian Jura (Weißmüller 1995; Richter 1997,
2016; Freund 1998). The key technological feature of
the long cultural stratigraphy of the site (23 MP
occupations spanning MIS 5c-3) is the alternation
between occupations rich in bifacial technology and
those with only a few Keilmesser or a complete lack of
these pieces in the G-Complex of MIS 3 (Keilmesser-
gruppen or Micoquian in the sense of a “Mousterian
with a Micoquian option”, MMO; Richter 1997, 2016).
The bottom of the MP sequence (Untere Schichten)
dating to the Early Würmian does not feature bifacial
technology. Richter (1997, 2016; see also Uthmeier
2004) interprets the interstratification of assemblages
with and without bifacial technology in the G-Complex
as different functional and seasonal variants within a
single settlement system (land-use cycles).
Quartär 66 (2019) N. J. Conard et al.
72
Assemblages poor in bifacial artifacts represent
so-called “Initial inventare” deriving from initial,
explorative occupations of a given region followed by
longer and more-specialized settlements (“Konseku-
tivinventare”) with bifacial technology, all belonging to
the same MMO cult ural unit. Importantly, this situation
of marked diachronic change in (bifacial) technology
contrasts with the assemblages at GK which are
consistent in their techno-typological and contextual
characteristics throughout and do not feature any
bifacial tools or debitage from shaping. Differences to
the MMO concept also occur concerning a consistent
procurement of raw materials and with regard to
reduction strategies, with all GK layers showing
evidence for multiple Levallois modalities without a
clear presence of Quina methods. High percentages
of denticulate tools in small assemblages – expected
as markers of Initialinventare (Richter 1997, 2016) – are
also lacking at GK. Thus the consistent and distinct
techno-typological markers for the five assemblages
at GK cannot be easily explained as Initialinventare.
The site complex of Buhlen in Hessen provides
another comparative example. The Obere Fundplatz,
or Upper Site, yielded rich, stratified Keilmesser-
gruppen assemblages (Schichtkomplex III; assigned by
Jöris (2001, 2003) to late MIS 5a; but see Richter 2016)
overlain by layers without bifacial technology (Schicht
II attributed to MIS 3), interpreted as a separate
Mousterian technocomplex (Bosinski & Kulick 1973;
Jöris 2001, 2003). The assemblage from Schicht-
komplex II shows similarities with the sequence at GK
with a focus on various modalities of Levallois
reduction and the occurrence of core edge flakes,
pseudo-Levallois points, and the production of
various types of scraper. The assemblages from the
Untere Fundplatz, or Lower Site, dated to the Early or
Middle Würmian, feature both assemblages without
(“Moustérien” similar to Schicht II) and with some
bifacial pieces (Keilmesser in Fundkomplex 4; see
Bosinski & Kulick 1973), the latter associated with
Levallois and discoid components and a diverse tool
kit including backed knives and denticulates
(Fiedler 2009). The assemblages from Buhlen-4, like in
the sequence from GK, contain artifacts exhibiting
Kostenki reduction (Kostenki-Enden; Fiedler 2009:
29-31; Taf. 45- 51, 56-58) and splintered pieces (Fiedler
2009: Taf. 50). This being said, these assemblages
from Buhlen-4 differ from the MP sequence at GK in
their laminar component, high frequency of backed
bifacial knives (Keilmesser) and rare examples of other
bifacial artifacts. On the whole, the variable presence
of bifacial technology at Buhlen stands in clear contrast
to the cultural stratigraphic situation at GK.
The open-air, lakeside site of Königsaue in Saxony-
Anhalt (Mania & Töpfer 1973; Mania 2002) provides
another instructive example from the MP of Central
Europe for interpreting stratigraphic sequences that
yield assemblages with (Königsaue Levels A and C)
and without or with only few (Königsaue Level B)
bifacial elements (Keilmesser Typ Königsaue). Recently
Picin (2017) interpreted the absence and presence of
Keilmesser as being directly dependent on mobility
systems and site use. All levels are associated with
various modalities of Levallois reduction and short-
term occupations. However, in level B without
Keil messer, Levallois flakes were transported off-site
indicating more residential mobility with repeated
short visits to this lakeshore setting for flint knapping
and other activities according to Picin (2017). In
contrast, levels A and C document bifacial tools and
the export of Levallois cores, which he interprets as
indicators of logistical mobility. An independent study
by Weiss et al. (2017) also demonstrated the techno-
logical conformity and similarities in raw material
procurement patterns of the Königsaue assemblages,
which were unrelated to the presence/absence of
bifacial tools. Importantly, the assemblages from GK,
while containing a few cortical pieces reflecting sparse
primary reduction, lack bifacial technology, and
demonstrate the export of Levallois flakes and the
import of already prepared cores. As discussed above,
we associated the assemblages as being the result of
multiple, short-term occupations of the cave. The
assemblages of Königsaue show general differences
with GK in both core reduction (discoid methods) and
tool assemblages (denticulates and “Fäustel”). Unlike
at GK, Königsaue and the other sites mentioned above
document the interstratification of bifacial and
non-bifacial assemblages.
In many ways there are similarities between the
assemblages from GK and Bosinski’s (1967; quotation
marks in original) “Moustérien” Formengruppe lacking
bifacial technology, with Levallois modalities, mainly
unifacial retouch and a predominance of scraper
types. Bosinski included Balve IV (~60 ka), Buhlen II
(see above) and Kartstein III (MIS 4-3) in this Formen-
gruppe, or technocomplex. Thi s category is reminiscent
of Bordes’ “Moustérien typique” (Bordes 1972) and
similar to the Late (post-Eemian) Mousterian as
described by Conard & Fischer (2000) as well as
Richter’s Initialinventare of the MMO in the “late
Middle Paleolithic” (MIS 3 assemblages; Richter 2016),
all of which lack bifacial artifacts. The assemblages at
GK, however, differ from expected Initialinventare
(see above) and are also unlike Balve IV and Buhlen II
in not featuring retouched points and a laminar
component (e.g. Bosinski 1967; Bosinski & Kulick 1973).
Finally, the Swabian Mousterian assemblages from
GK contrast sharply with the somewhat older assem-
blages from MIS 5 in the Rhineland, like the hilltop
volcanic crater of Tönchesberg (Conard 1992) and the
floodplain deposits from Wallertheim (Conard &
Adler 1997; Conard 2001; Adler et al. 2003). These
sites, while also lacking bifacial elements, document
greater technological diversity in terms of raw material
procurement, more diverse pat terns of lithic reduction
and a more diverse tool spectrum than GK. In these
open-air settings, the brief occupations reveal much
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
73
higher technological variability and major short-term
changes in knapping strategies and tool manufacture,
use, recycling and discard compared to the case of GK
and the other Swabian caves. The record from GK
appears to reflect the presence of more stable behav-
ioral adaptations and perhaps more stable popula-
tions than does the sample from the open-air sites in
the Rhineland. Interestingly, the D-layers at
Hummerich, another volcanic crater in the East Eifel
(Bosinski et al. 1983, 1986; Street 2002), which
post-date most of the find horizons at Tönchesberg
except layer 1B and all of the six find horizons at
Wallertheim, are of similar age to the lower deposits at
GK, contain a record of bifacial and non-bifacial finds,
underlining the diversity of these regional records.
Unfortunately, the geological context of the MP from
Hummerich is reworked and lacks the high chrono-
stratigraphic resolution needed for more systematic
comparisons.
In summary, GK differs most strongly from the
comparative sites and other long sequences of the
German MP in its consistent diachronic techno-
typol ogical signal within the t ime range of ~94-43 ka B P.
The sequence is characterized by local raw material
procurement, use of various modalities of Levallois in
addition to Kostenki reduction, as well as bipolar and
platform methods, abundant scraper forms plus splin-
tered pieces and an absence of denticulates and
bifacial implements. Despite the long timespan
covered by the sequence, and in contrast to Sesselfels-
grotte, Buhlen, Königsaue, and other sites, the archae-
ological record at GK does not feature high intra-site
variability with different technocomplexes, such as an
interstratification of assemblages with and without
bifacial pieces (e.g. MMO), and there is overall no
evidence for bifacial flaking taking place at the site.
Post-depositional and taphonomic arguments can be
ruled out to explain the absence of these pieces – or
their by-products – as modern field methods were
used to excavate the site, allowing detailed geoar-
chaeological work and the recovery of artifacts
regardless of their size. Based on these observations
we conclude that the techno-typological signals at GK
are distinctive from those of many other sites of similar
age.
Implications for Neanderthal behavior, mobility
patterns and demography
In combination with contextual information from
geoarchaeology, zooarchaeology, absolute dating and
field observations (Richter et al. 2000; Conard &
Malina 2002, 2003; Münzel & Conard 2004; Conard &
Bolus 2008; Conard et al. 2012, 2019; Higham et al.
2012, 2014; Richard et al. 2019), the MP lithic assem-
blages of GK offer insights on Neanderthal behavior,
mobility patterns and demography in the Late Pleis-
tocene of the Swabian Jura. Consideration of these
findings contributes to our understanding of the
lifeways of late Neanderthals as well as to the
behavioral and demographic context of the late MP
just before the arrival of modern humans in the region,
which is documented at GK around 43 ka calBP
(Higham et al. 2012).
In terms of raw material procureme nt and economy,
the overall proportions of Jurassic chert at GK are
even higher compared to many other MP sites within
the region. This observation suggests a particularly
strong focus on local raw material procurement from
the close primary and secondary sources by
Nean derthals. With regard to the overall structure of
mobility, similar observation from other MP sites of
the region suggest reduced spheres of residential and
logistical mobility. Neanderthals likely spent most of
the year in the region and usually organized their
economic and social lives on a local scale but with high
residential mobility (Conard et al. 2006, 2012).
As far as transport of raw material is concerned,
the initial phases of decortification took place outside
of GK, with cortex values being low on both blanks
and cores. Knappers imported mostly prepared cores
of Jurassic chert as well as finished pieces for other raw
materials. Production and modification of blanks on
Jurassic chert was a focus during the occupation at the
site as is indicated by relatively complete reduction
sequences and some small debitage, including
occasional retouch flakes. The rarity of end products,
particularly from Levallois reduction, suggests the
export of selected blanks and tools from the site and
could be viewed as an indicator of short-term use of
the site. In sum, the low density of lithic artifacts and
other anthropogenic materials in all the MP find
horizons at GK in combination with a somehow
fragmented reduction chain with common import and
export of finds reflect a settlement system with high
mobility and frequent movement of individuals and
groups.
The low density of cultural materials also helps to
shape our interpretations of the settlement dynamics
of the Swabian MP in general and at GK in particular.
At GK, the archaeological material does not originate
from clearly defined find horizons (Fig. 4), and no
coherent archaeological features could be identified
during excavation or during subsequent analyses. This
is also the case during the MP at Hohle Fels (e.g.
Conard & Malina 2013) but contrasts markedly with
the overlying Aurignacian at both these sites (Conard
& Malina 2002, 2003; Conard et al. 2006, 2012). These
stratigraphic observations are likely explained by
taphonomic alteration and mixing of the MP layers by
geogenic, biogenic and cultural processes, although
the ephemeral and infrequent use of the site by
Neanderthals points to a key role by other tapho-
nomic agents (Conard 2011; Conard et al. 2012, 2019;
Miller 2015). There is also the possibility that the
excavations uncovered only the edges of the MP
occupation centers (Münzel 2019). Moreover, frequent
evidence for cryoturbation of the sediments damaging
both bones (56 %; Münzel & Conard 2004) and lithic
Quartär 66 (2019) N. J. Conard et al.
74
artifacts (67-71 %) is present throughout the entire MP
sequence, but particularly in the uppermost layers IV
and V, which also harbor a high amount of limestone
debris reflecting higher geogenic input (see also Hahn
1988: 102-103; Conard & Malina 2002, 2003).
Analyses of faunal material found similar evidence
for more abundant non-anthropogenic input, particu-
larly from cave bears, higher levels of hominin mobility
and lower population density for the MP compared to
the UP at GK, as well as for Hohle Fels, Sirgenstein,
Große Grotte, and Kogelstein (Münzel & Conard
2004; Conard et al. 2012; Kitagawa et al. 2012; Münzel
2019). In sum, several lines of evidence indicate
repeated, brief knapping episodes during
Neanderthal occupations of GK within a relatively
long period (~94-43 ka BP). Prolonged periods of
abandonment and use of the site by cave bears
resulted in the accumulation of small lithic assem-
blages without recognizable archaeological features.
Burnt bone varies between 8-24 g/m3 in the MP layers
at GK (Conard et al. 2012). The material almost
certainly results from the controlled use of fire by
Neanderthals inside the cave, but taphonomic
processes, including most notably actions of cave
bears, have damaged the context of the burnt faunal
remains recovered during the excavation and subse-
quent analyses. Although there are rare cutmarks on
MP faunal remains and rare anatomical refits, there are
no cutmarks on the burnt bones (Münzel 2019). As is
usually the case for fragments of burnt bone, the
bones cannot be identified to species level. Diverse
biological activities including denning and hibernation
by bears can radically effect the preservation of Paleo-
lithic sites (Fosse et al. 2004; Camáros et al. 2016), and
the well documented and intense use of GK and many
other Swabian caves by cave bears underline their
importance in modifying the archaeological record of
the M P.
The consistent signal of low-density archaeo-
logical material at GK and other sites of the Swabian
Jura generally suggests low intensity occupation of
caves and high residential mobility by Neanderthals
during the Middle Würmian. Most known open-air
sites derive from surface collections, such as
Wippingen, Sonderbuch or Asch, and have yielded
comparatively low find densities of unequivocal MP
artifacts such as isolated Keilmesser (Floss & Schürch
2015). Richer sites including Börslingen or Wittlingen
(Burkert et al. 1992; Floss et al. 2012) are likely quarry
sites, hindering direct comparisons. While it is
difficult to relate these observations directly to
measures of demography, low populations combined
with higher residential mobility in the Swabian Jura
seem the most likely explanation. Additional
evidence for such demographic interpretations
comes from recent paleogenomic studies from the
Neanderthal femur of Hohlenstein-Stadel (Posth et
al. 2017), but also from genetic analyses of other
Neanderthal remains in Eurasia (e.g. Castellano et al.
2014; Prüfer et al. 2014), finding evidence for
inbreeding alongside low genetic diversity.
Several studies using archaeological proxies for
estimating demographic parameters have found a
population increase following the shift from the MP to
the UP (Bocquet-Appel & Demars 2000; Conard et al.
2006, 2012; Mellars & French 2011; Bocquet-Appel &
Degioanni 2013), but such studies face several
methodical problems (e.g. Dogandžić & McPherron
2013; more generally Bocquet-Appel 2008; French
2015). It is also relevant to note that not all MP sites from
the Swabian Jura are archaeological deposits with small
assemblages and low densities, such as is the case at
Bockstein and Heidenschmiede (Peters 1931; Wetzel &
Bosinski 1969; Çep 2014; Çep & Krönneck 2015),
indicating either longer occupations, larger groups
using the sites or slower rates of geological deposition
when the MP find horizons formed. Many European
localities from neighboring regions during similar
temporal frames including Sesselfelsgrotte (Weißmüller
1995; Richter 1997) and Buhlen (Bosinski & Kulick 1973;
Jöris 2001; Fiedler 2009) in Germany or further into the
west at the Vanne-Tal in France (Depaepe 2007), to the
east at Kůlna 7a (Valoch 1988; Neruda 2017), or to the
south at Fumane (e.g. Peresani et al. 2011) exhibit higher
artifact densities and reflect higher occupation inten-
sities than do the find horizons at GK.
Radiometric dates for the MP at GK of ca. 94-43 ka
BP (Richter et al. 2000; Higham et al. 2012, 2014;
Richard 2015; Richard et al. 2019) provide an important
chronological anchor for the region, since reliable
dates are lacking for most other sites in the Swabian
Jura. With the upper portion of the MP assemblages of
GK dating to ~50-45 ka BP, our results provide insights
into the behavior and demography of late Neander-
thals living in southwestern Germany prior to the
arrival of anatomically modern humans. Differences in
the use of the site and potentially population density
are most striking when comparing the MP of GK
directly with the overlying Aurignacian. After a short
occupational hiatus, the Aurignacian at the site is
characterized by different faunal assemblages, new
technologies and novel lithic and organic artifacts,
including a great variety of personal ornaments,
figurative art objects and musical instruments, all
unknown in the preceding MP (Hahn 1988; Conard &
Bolus 2003, 2006, 2008; Münzel & Conard 2004;
Conard et al. 2006, 2015). The clear stratigraphic and
behavioral discontinuity between the uppermost MP
layers and the Aurignacian horizons at GK is further
substantiated by a sharp rise in find densities for all
classes of archaeological materials at the base of the
Aurignacian and no mixing between the archaeological
horizons. The end of the MP and the subsequent
occupational hiatus at GK and many other sites in the
Swabian Jura (Conard et al. 2006) points to a decrease
in occupation intensity associated with a dwindling
population of Neanderthals. The gap between
Neanderthal and modern human occupations at GK
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
75
and other sites suggests that the former might have
abandoned the region, with UP groups expanding
into largely depopulated territory as postulated by
the ‘Population Vacuum’ hypothesis (Conard 2003;
Conard et al. 2003b; Conard & Bolus 2006). In addition,
Neanderthal occupations of the cave sites of the
Swabian Jura with small lithic assemblages and relatively
low amounts of other anthro pogenic materials reflect
more short-term stays within a system of higher mobilit y
compared to modern humans in the Aurignacian of the
area.
The break in archaeological and demographic
signatures between Neanderthals and Homo sapiens
found at GK is similar to the record in southern and
west-central German record of the late MP (Richter,
2016; Uthmeier 2004; Böhner 2008). This pattern
stands in marked contrast with other European regions
(Bolus 2004a; Peresani et al. 2008; Soressi et al. 2013;
Higham et al. 2014), as is exemplified by technocom-
plexes such as the Chatelperronian (e.g. Harrold 2000;
Soressi & Roussel 2014), Uluzzian (Palma di Cesnola
1989; Peresani et al. 2008), and Szeletian (Allsworth-
Jones 1986; Skrdla et al. 2014; Hauck et al. 2016).
While there is heated debate on the purported transi-
tional nature, stratigraphic integrity and makers of
these entities (e.g. Higham et al. 2010; Benazzi et al.
2011; Soressi & Roussel 2014; Zilhão et al. 2015;
Gravina et al. 2018), neither the Swabian Jura nor
southern and west-central Germany (Richter, 2016;
Uthmeier 2004; Böhner 2008) feature transitional
cultural stratigraphic entities beyond the Blattspitzen-
gruppe, which is usually viewed as a cultural devel-
opment arising from the late MP independent of
contacts with modern humans.
Instead, southwestern Germany is home to a late
MP with more typical features as reflected in the
Swabian Mousterian. The UP follows without recog-
nizable interaction in the form of a very early but fully
developed Aurignacian characterized by a vast range
of cultural innovations lacking in the region’s MP.
While the basic structure of the Swabian record is
clear, and highlights the varied spatial nature of the
European Paleolithic record, explaining these regional
differences and gaining a better grasp of the chrono-
cultural stratigraphy of the late MP of the Swabian Jura
remain essential lines of future inquiry. The ongoing
excavations into the stratified MP layers of Hohle Fels
(e.g. Conard & Janas 2018) will provide such an oppor-
tunity for testing and refining results from GK
presented here.
A: This article is dedicated to Rainer Blumen-
trit t (1940-2019) who discovered the site of G eißenklösterle an d
who suppo rted our research in the A ch Valley for many dec ades.
The excavations in 2001 and 2002 were f unded by the Deutsche
Forschungsgemeinschaft, the Landesdenkmalamt für Denkmal-
pflege of Baden-Württemberg and the Alb-Donau-Kreis. We
also thank the city of Blaubeuren, the Gesellschaft für
Urgeschichte, the Urgeschichtliches Museum and the Heidelberg
Cement AG for providing accommodation and workspace for
the excavation team. We want to thank the numerous people
involved in the organization of the excavations as well as their
generous efforts in supporting the Geißenklösterle campaigns.
We also thank the many collaborators who have contributed to
work at Geißenklösterle presented here including Paul
Goldberg, Tom Higham, Maria Malina, Chris Miller, Susanne
Münzel, Maïlys Richard, and Daniel Richter. We are indebted to
Sabine Boos and Heike Würschem for the drawings of stone
artifacts. Finally, we thank two anonymous reviewers for helpful
recommendations.
Literature cited
Adler, D. S., Prindiville, T. J. & Conard, N. J. (2003). Patterns of
spatial organization and land use during the Eemian Interglacial in
the Rhineland: New data from Wallertheim, Germany. Eurasian
Prehistory 1(2): 25-78.
Allsworth-Jones, P. (1986). The Szeletian and the transition from
Middle to Upper Palaeolithic in Central Europe. Oxford University
Press, New York.
Auffer mann, B., Burker t, W., Hahn, J ., Pasda, C. & Si mon, U. (1990).
Ein Merkmalsystem zur Auswertung von Steinartefaktinventaren.
Archäologisches Korrespondenzblatt 20: 259-268.
Beck, D. (1999). Das Mittelpaläolithikum des Hohlenstein, Stadel
und Bärenhöhle, im Lonetal. Dr. Rudolf Habelt GmbH, Bonn.
Benazzi, S., Douka, K., Fornai, C., Bauer, C. C., Kullmer, O.,
Svoboda, J., Pap, I., Mallegni, F., Bayle, P., Coquerelle, M.,
Condemi , S., Ronchitell i, A., Harv ati, K. & Weber, G. W. (2011).
Early dispersal of modern humans in Europe and implications for
Neanderthal behaviour. Nature 479: 525 -528.
Bertra n, P., Lenoble, A ., Todisco, D., Desros iers, P. M. & Sørensen ,
M. (2012). Particle size distribution of lithic assemblages and
taphonomy of Palaeolithic sites. Journal of Archaeological Science
39: 314 8-316 6.
Bocquet-Appel, J.-P. & Demars, P.-Y. (2000). Population kinetics
in the Upper Palaeolithic of Western Europe. Journal of
Archaeological Science 27(7): 551-570.
Bocquet- Appel , J.- P. (2008). Recent advances in paleodemography:
data, techniques, patterns. Springer, Dordrecht.
Bocquet-Appel, J.-P. & Degioanni, A. (2013). Neanderthal
demographic estimates. Current Anthropology 54(8): 202-213.
Boëda, E., Geneste, J. M. & Meignen, L. (1990). Identification de
chaînes opératoires lithiques du Paléolithique ancien et moyen.
Paléo 2: 43- 80
Boëda, E. (1993). Le débitage discoïde et le débitage Levallois
récurrent centripède. Bulletin de la Société préhistorique
française 90: 392-404.
Böhner, U. (2008). Sesselfelsgrotte IV. Die Schicht E3 der
Sesselfe lsgrotte und die Fu nde aus dem Abri I am Sc hulerloch. Späte
Micoquien-Inventare und ihr Verhältnis zum Moustérien. Franz
Steiner Verlag, St uttgart.
Bolus, M . & Rück, O. (2000). Eine Blattspitze aus Wittislingen, Lkr.
Dillingen a. d. Donau (Bayern). Zur südwestlichen
Verbreitungsgrenze spätmittelpaläolithischer Blattspitzen-
inventare. Archäologisches Korrespondenzblatt 30 : 165 -172.
Bolus, M. (2004a). Der Übergang vom Mittel- zum
Jungpaläolithikum in Europa. Eine Bestandsaufnahme unter
besonderer Berücksichtigung Mitteleuropas. Germania 82: 1-54.
Bolus, M. (2004b). Settlement analysis of sites of the Blattspitzen
complex in central Europe. In: N. J. Conard (Ed.) Settlement
Dynamics of the Middle Paleolithic and Middle Stone Age II. Kerns
Verlag, Tübingen, 201-226.
Bolus, M . (2011). T he late Middle Paleolithic and t he Aurignacian
of the Swabian Jura, southwestern Germany. In: A. P. Derevianko
& M. V. Shunkov (Eds.) Characteristic features of the Middle to
Upper Paleolithic transition in Eurasia. Publishing Department of
the Institute of Archaeology and Ethnography. SB RAS,
Novosibirsk, 3-10.
Quartär 66 (2019) N. J. Conard et al.
76
Bolus, M. (2015). The Transition from the Middle to the Upper
Paleolithic in the Swabian Jura, Southwestern Germany.
Anthropologie 53(1/2): 167-179.
Bordes, F. (1958). Le passage du Paléolithique moyen au
Paléolithique supérieur. In: G.H. R. von Königswald (Ed.) Hundert
Jahre Neanderthaler: Neanderthal centenary 1856-1956. Böhlau
Verlag, Köln. Graz, 175-181.
Bordes, F. (1972). Du Paléolithique moyen au Paléolithique
supérieur: continuité ou discontinuité. In: F. Bordes (Ed.) The
origin of Homo sapiens. Proceedings of the Paris symposium.
United Nations Educational, Scientific and Cultural Organization,
Par is, 211-218 .
Bosinski, G. (1967). Die mittelpaläolithischen Funde im westlichen
Mitteleuropa. Böhlau Verlag, Köln, Graz.
Bosinski, G. & Kulick, J. (1973). Der mittelpaläolithische
Fundplatz Buhlen, Kr. Waldeck. Vorbericht über die Grabungen
1966 -1969. Germania 51: 1-41.
Bosinski, G., Kulemeyer, J. & Turner, E. (1983). Ein
mittelpaläolithischer Fundplatz auf dem Vulkan Hummerich bei
Plaidt, Kreis Mayen/Koblenz. Archäologisches Korrespondenzblatt
13: 415–428.
Bosinski, G., Kröger, K., Schäfer, J. & Turner, E.
(1986). Altsteinzeitliche Siedlungsplätze auf den Osteifel-
Vulkanen. Jahrbuch des Römisch-Germanischen Zentralmuseums
33: 97-130.
Böttcher, R., Çep, B., Kind, C.-J., Mörike, D., Pawlik, A., Rähle,
W., Steppan, K., Torke, R., Torke, W. & Ziegler, R. (2000).
Kogelstein: Eine mittelpaläolithische Fundstelle bei Schelklingen-
Schmiechen, Fundberichte aus Baden-Württemberg 24: 7-176.
Bourguignon, L . (1997). Le Moustérien de Type Quina. Nouvelle
définition d´une entité technique. Ph.D. dissertation, Université
de Paris X, Paris.
Bourguignon, L ., Faivre, J.-P. & Turq, A. (2004). Ramification
des chaînes opératoires: une spécificité du Moustérien? Paléo
16: 37-48 .
Burkert, W. (2001). Lithische Rohmaterialversorgung im
Jungpaläolithikum des südöstlichen Baden-Württemberg. Ph.D.
dissertation, University of Tübingen, Tübingen.
Burkert, W. & Floss, H. (2005). Lithic exploitation areas in the
Upper Pale olithic of west a nd southwest Ger many – a comparat ive
study. In: G. Körlin & G. Weisgerber (Eds.) Stone Age – Mining
Age. Proceedings of the VIIIth International Flint Symposium
Bochum 1999. Der Anschnitt Beiheft 19, Bochum, 35- 49.
Burkert, W., Çep, B. , Kind, C.- J., Pasda, C ., Schranz, M ., & Simon,
U. (1992). Wittlingen. Eine mittelpaläolithische
Freilandfundstelle bei Bad Urach. Fundberichte aus Baden-
Württemberg 17(1): 1-110.
Camarós, E., Münzel, S. C., Cueto, M., Rivals, F., & Conard, N. J.
(2016 ). The evolution of Paleolithic hominin–carnivore
interaction written in teeth: Stories from the Swabian Jura
(Germany). Journa l of Archaeologic al Science: Repor ts 6: 798-809.
Castellano, S., Parra, G., Sánchez-Quinto, F. A., Racimo, F.,
Kuhlwilm, M., Kircher, M., Sawyer, S., Fu, Q., Heinze, A.,
Nickel, B., Dabney, J., Siebaue r, M., White, L., B urbano, H. A.,
Renaud, G., Stenzel, U., Lalueza-Fox, C., de la Rasilla, M.,
Rosas, A., Rudan, P., Brajković, D., Kucan, Ž., Gušic, I.,
Shunkov, M. V., Derevianko, A. P., Viola, B., Meyer, M., Kelso,
J., Andrés, A. M. & Pääbo, S. (2014). Patterns of coding
variation in the complete exomes of three Neandertals.
Proceedings of the National Academy of Sciences of the United
States of America 111: 6666-6671.
Çep, B. (1996). Die Steinartefakte der Schichten VIII/VII und VI
des Sirgensteins bei Weiler. In: I. Campen, J. Hahn & M.
Uerpmann (Eds.) Spuren der Jagd - Die Jagd nach Spuren.
Festschrift für Hansjürgen Müller-Beck. Mo Vince Verlag,
Tübingen, 207-216.
Çep, B. (2013). Ausgangsbasis oder Versorgungsstandort?
Raumnutzung im Mittel- und Jungpaläolithikum des Ach- und
Blautals bei Blaubeuren. Quartär 60 : 61-93.
Çep, B. (2014). Das mittelpaläolithische Silexinventar des Bocksteins
im Lonetal (Schw äbische Alb). Vielfa lt der Formen ode r Fortbestand
einer technologischen Idee? Varia neolithica VIII: 79-92.
Çep, B. & Waiblinger, J. (2001). The use of cave and open-air
sites in southern Germany. In: N. J. Conard (Ed.) Settlement
Dynamics of the Middle Paleolithic and Middle Stone Age. Kerns
Verlag, Tübingen, 251-263.
Çep, B. & Krönneck, P. (2015). Landscape and cave use in the
Middle Paleolithic of Bockstein: New results from the lithic and
fauna analysis. In: N. J. Conard & A. Delagnes (Eds.) Settlement
Dynamics of the Middle Paleolithic and Middle Stone Age IV.
Kerns Verlag, Tübingen, 227-251.
Çep, B., Bu rkert, W., & Floss, H. (2011). Zur mittelpaläolithischen
Rohmaterialversorgung im Bockstein. Mitteilungen der
Gesellschaft für Ur- und Frühgeschichte 20: 33-52.
Conard, N . J. (1992). Tönchesberg and its position in the Paleolithic
prehistory of northern Europe. Monographien des Römisch-
Germanischen Zentralmuseum 20, Dr. Rudolf Habelt Verlag.
Conard, N. J. (2001). River terraces, volcanic craters and Middle
Paleolithic settlement in the Rhineland. In: N. J. Conard (Ed.),
Settlement dynamics of the Middle Paleolithic and Middle Stone
Age. Tübingen, Kerns Verlag, 221-250.
Conard, N. J. (2003). Palaeolithic ivory sculptures from
southwestern Germany and the origins of figurative art. Nature
426: 830- 832.
Conard, N. J. (2005). Sind sich Neandertaler und moderne
Menschen auf der Schwäbischen Alb begegnet? In: N. J. Conard,
S. Kölbl & W. Schürle (Eds.) Vom Neandertaler zum modernen
Menschen. Jan Thorbecke Verlag, Ostfildern, 131-152.
Conard, N. J. (2007). Cultural evolution in Africa and Eurasia
during the Middle and Late Pleistocene. In: W. Henke & I.
Tattersall (Eds.) Handbook of paleoanthropology. Springer,
Berlin, 2001-2037.
Conard, N . J. (2009). A female figu rine from the b asal Aurigna cian
of Hohle Fels cave in southwestern Germany. Nature 459:
248-252.
Conard, N. J. (2011). The demise of the Neanderthal cultural
niche and the beginning of the Upper Paleolithic in southwestern
Ger many. In: N. J. Conard & J. Richter, (Eds.) Neanderthal
Lifeways , Subsistence and Technolog y. One Hundr ed Fifty Years of
Neanderthal Studies. Springer, Dordrecht, Heidelberg, London,
New York, 223-240.
Conard, N . J. & Adler, D. S. (1997). Lithic reduction and hominid
behaviour in the Middle Paleolithic of the Rhineland. Journal of
Anthropological Research 53: 147-175.
Conard, N. J. & Fischer, B. (2000). Are there recognizable
cultural entities in the German Middle Palaeolithic? In: A. Ronen
& M. Weinstein-Evron (Eds.) Towards Modern Humans:
Yabrudian and Micoquian, 400-50 k years ago. BAR International
Series 850, Oxford, 7-24.
Conard, N. J. & Malina, M. (2002). Neue Ausgrabungen in den
untersten Schichten des Aurignacien und des
Mittelpaläolithikums im Geißenklösterle bei Blaubeuren, Alb-
Donau-Kreis. Archäologische Ausgrabungen in Baden-
Württemberg 2001: 16 -21.
Conard, N. J. & Bolus, M. (2003). Radiocarbon dating the
appearance of modern humans and timing of cultural innovations
in Europe: new results and new challenges. Journal of Human
Evolution 44 : 331-371.
Conard, N . J. & Malin a, M. (2003). Abschließende Ausgrabungen
im Geißenklösterle bei Blaubeuren, Alb-Donau-Kreis.
Archäologische Ausgrabungen in Baden-Württemberg 2002:
17-21.
Conard, N. J. & Moreau, L. (2004). Current research on the
Gravettian of the Swabian Jura. Mitteilungen der Gesellschaft für
Ur- und Frühgeschichte 13: 2 9-59.
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
77
Conard, N. J. & Bolus, M. (2006). The Swabian Aurignacian and
its place in Eu ropean prehistor y. In: O. Bar-Yosef & J. Zilhão (Ed s.)
Towards a Definition of the Aurignacian. Proceedings of the
Symposiu m held in Lisbon, Por tugal, June 25 -30, 2002, Trabalhos
de Arqueologia 45, Lisbon, 209-237.
Conard, N. J. & Bolus, M. (2008). Radiocarbon dating the late
Middle Paleolithic and the Aurignacian of the Swabian Jura.
Journal of Human Evolution 55: 886 -8 97.
Conard, N. J. & Malina, M. (2013). Grabungen in Schichten des
Moustérien und Gravettien im Hohle Fels bei Schelklingen.
Archäologische Ausgrabungen in Baden-Württemberg 2012: 78-83.
Conard, N. J. & Bolus, M. (2015). Chronicling modern human's
arrival in Europe. Science 34 8(6236): 754 -756.
Conard, N. J & Janas, A. (2018). Fortsetzung der Ausgrabungen
am Hohle Fels und die Entdeckung einer markierten
Mammutrippe aus dem Gravettien. Archäologische
Ausgrabungen in Baden-Württemberg 2017: 52-55.
Conard, N.J. & Kind, C.-J. (2019). Als der Mensch die Kunst
erfand: Eiszeithöhlen der Schwäbischen Alb. Theiss, Stuttgart.
Conard, N. J., Dippon, G. & Goldberg, P. (2003a).
Chronostratigraphy and archaeological context of the
Aurignacian deposits at Geißenklösterle. In: J. Zilh ao & F. d’Err ico
(Eds.) The Chronology of the Aurignacian and the Transition
Technocomplexes. Dating, Stratigraphies and Cultural
Implications. Trabalhos de Arqueologia 33, Instituto Português
de Arqueologia, Lisbon, 165-176.
Conard, N. J, Langguth, K. & Uerpmann, H.-P. (2003b).
Einmalige Funde aus dem Aurignacien und erste Belege für ein
Mittelpaläolithikum im Hohle Fels bei Schelklingen, Alb-Donau-
Kreis. Archäologische Ausgrabungen in Baden-Württemberg
2002: 21-27.
Conard, N. J., Soressi, M., Parkington, J. E., Wurz, S. & Yates, R.
(2004). A unified lithic taxonomy based on patterns of core
reduction. South African Archaeological Bulletin 59: 12-16.
Conard, N. J., Bolus, M., Goldberg, P., Münzel, S. C. (2006).
The Last Neanderthals and first modern humans in the Swabian
Jura. In: N. J. Conard (Ed.) When Neanderthals and Modern
Humans Met. Kerns Verlag, Tübingen, 305-341.
Conard, N. J., Malina, M. & Münzel, S. C. (2009). New flutes
document the earliest musical tradition in southwestern
Ger many. Nature 46 0: 737-740.
Conard, N . J., Bolus M . & Münzel S. C. (2012). Middle Paleolithic
land use, spatial organization and settlement intensity in the
Swabian Jura, southwestern Germany. Quaternary International
247: 236 -245.
Conard, N. J., Bolus, M., Dutkiewicz, E., & Wolf, S. (2015).
Eiszeitarchäologie auf der Schwäbischen Alb. Die Fundstellen im
Ach-un d Lonetal und in ihrer U mgebung. Kerns Verlag, Tüb ingen.
Conard, N. J., Bolus M. & Münzel S. C. (eds.) (2019).
Geißenklösterle: Chronostratigraphie, Paläoumwelt und
Subsistenz im Mittel- und Jungpaläolithikum der Schwäbischen
Alb. Kerns Verlag, Tübingen.
Dauvois, M . (1976). Précis de dessin dynamique et structural des
industries lithiques préhistoriques. CNRS, Fanlac.
Delagnes, A., Jaubert J. & Meignen, L. (2007). Les
technocomplexes du Paléolithique moyen en Europe occidentale
dans leur cadre diachronique et géographique. In: B.
Vandermersch & B. Maureille (Eds.) Les Néandertaliens. Biologie
et cultures. Editions du CTHS, Paris, 213-229.
Demars P.-Y. & Laurent, P. (1989). Types d’outils lithiques au
Paléolithique supérieur en Europe. CNRS, Paris.
Dibble, H . L. & McPher ron, S.P. (1996). A multimedia companion
to the Middle Paleolithic site of Combe-Capelle Bas (France).
University of Pennsylvania Museum, Philadelphia.
Dibble, H. L. & McPherron, S. P. (2007). Truncated-faceted
pieces: Hafting modif ication, retouch, or cores? In: S. P. McPherr on
(Ed.). Tools versus cores: alternative approaches to stone tool
analysis. Cambridge Scholars Publishing, Newcastle, 75-90.
Depaepe, P. (2007). Le paléolithique moyen de la vallée de la
Vanne, Yonne, France: matières premières, industries lithiques
et occupations humaines. Société préhistorique française 41.
Dogandžić, T. & McPherron, S. P. (2013). Demography and the
demise of the Neanderthals: a comment on ‘Tenfold Population
Increase at the Neanderthal-to-Modern-Human Transition’.
Journal of Human Evolution 64 (4): 311-313.
Fiedler, L. (2009). Die mittelpaläolithischen Funde und Befunde
des Unteren Besiedlungsplatzes von Buhlen/ Band II – Die
Steinartefakte: Formen, Techniken, Aktivitäten und kulturelle
Zusammenhänge. Fundberichte aus Hessen – Beiheft 5.2,
Selbstverlag des Landesamtes für Denkmalpf lege, Wiesbaden.
Fisher, L., Harris, S., Knipper, C. & Schreg, R. (2008). Neolithic
chert exploitation on the Swabian Alb (Germany): 2007
excavations at Asch-“Borgerhau”. The Quarry. The Newsletter of
the SAA’s Prehistoric Quarries & Early Mines Interest Group #2:
8-19.
Floss, H ., Hoyer, C., Dutkie wicz, E., Fr ick, J. A. & Poen icke, H.-W.
(2012). Eine neu entdeckte paläolithische Freilandfundstelle
auf der Schwäbischen Alb – Sondagegrabungen in Börslingen.
Archäologische Ausgrabungen in Baden-Württemberg 2011:
71-73.
Floss, H. & Schürch, B. (2015). Paläolithische Oberflächenfunde
von der Blaubeurer Alb. Mitteilungen der Gesellschaft für
Urgeschichte 24 : 81-100.
Floss, H ., Fröhle, S., Sc hürch, B. & Wettengl, S. (2 017). Open air
occupation in a cave dominated archaeological landscape – new
perspectives on the Paleolithic of the Swabian Jura (Southern
Germany). Anthropologie 55: 43-73.
Fosse, P., Besson, J.-P., Laborde, H., Thomas-Cantie, F.,
Cazenave, G., Delmasure, M-C., Leveque, T., Laudet, F.,
Quiles, J. (2004). Denning behaviour of “modern” brown bear
(Ursus arctos, L.) in caves: biological and paleontological
considerations from French Pyrenean sites. Publications du
musée des Confluences Année 2004/2, 171-182.
Fraas, O. (1867). Über die neuesten Erfunde an der
Schussenquelle bei Schussenried. Jahreshefte des Vereins für
vaterländische Naturkunde in Württemberg 2 3: 48-74.
French, J. C. (2015). Demography and the Palaeolithic
Archaeological Record. Journal of Archaeological Method and
Theory 23 (1): 150 -199.
Freund, G. (1998). Sesselfelsgrotte I. Grabungsverlauf und
Stratigraphie. Franz Steiner Verlag, Saarbrücken.
Frick, J. A. (2013). Kostenki-Enden (Dorsalabbau an
Grundformen). In: H. Floss (Ed.) Steinartefakte vom
Altpaläolithikum bis in die Neuzeit. Kerns Verlag, Tübingen,
459-466.
Fröhle, S., Wettengl, S., Floss, H. (2019). Ein neues Großprojekt
- Zur Erfassung des Freilandpaläolithikums in Baden-
Württemberg. Vortrag am 14.11.2019, Abteilungskolloquium
Ältere Urgeschichte und Quartärökologie, Universität Tübingen.
Geneste, J-M. (1991). Systèmes techniques de production
lithique : variations techno-économiques dans les processus de
réalisation des outillages paléolithiques. Techniques et culture
17-18: 1 -35.
Goldberg, P., Miller, C., Conard, N. J. (2019). Geißenklösterle
stratigraphy and micromorphology. In: N. J. Conard, M. Bolus &
S. C. Münzel (Eds.) Geißenklösterle: Chronostratigraphie,
Paläoumwelt und Sibsistenz im Mittel- und Jungpaläolithikum der
Schwäbischen Alb. Kerns Verlag, Tübingen, 25-62.
Gravina, B., Bachellerie, F., Caux, S., Discamps, E., Faivre, J. P.,
Galland, A., Michel, A., Teyssandier, N. & Bordes, J.-G.
(2018) . No reliable evidence for a Neanderthal-
Châtelperronian association at La Roche-à-Pierrot, Saint-
Césaire. Scientific Reports 8: 15134.
Hahn, J . (1986). Kraf t und Aggressio n. Die Botscha ft der Eiszeitk unst
im Aurignacien Süddeutschlands? Verlag Archaeologica
Venatoria, Tübingen.
Quartär 66 (2019) N. J. Conard et al.
78
Hahn, J . (1988). Die Geißenklösterle-Höhle im Achtal bei Blaubeuren
I. Fundhorizontbildung und Besiedlung im Mittelpaläolithikum und
Aurignacien. Konrad Theiss Verlag, Stuttgart.
Hahn, J. (1991). Erkennen und Bestimmen von Stein- und
Knochenartefakten. Einführung in die Artefaktmorphologie.
Archaeologica Venatoria 10, Tübingen.
Hahn, J. & Münzel, S. C. (1995). Knochenflöten aus dem
Aurignacien des Geißenklösterle bei Blaubeuren, Alb-Donau-
Kreis. Fundberichte aus Baden-Württemberg 20 : 1-12.
Harrold, F. B. (2000). The Chatelperronian in historical context.
Journal of Anthropological Research 56: 59-75.
Hauck, T., Rethemeyer, J., Rentzel, P., Schulte, P., Heinze, S.,
Ringer, A., Richter, J., Chu, W., Lehmkuhl, F. & Vogels, O.
(2016 ). Neanderthals or early modern humans? A revised 14C
chronology and geoarchaeological study of the Szeletian
sequence in Szeleta Cave (Kom. Borsod-Abaúj-Zemplén) in
Hunga ry. Archäologisches Korrespondezblatt 46: 271-290.
Herkert, K., Siegeris, M., Chang, J. Y., Conard, N. J., & Floss, H.
(2015). Zur Ressourcennutzung später Neandertaler und
früher moderner Menschen. Fallbeispiele aus dem südlichen
Burgund und der Schwäbischen Alb. Mitteilungen der
Gesellschaft für Urgeschichte 24: 141-72 .
Higham , T. F. G., Jacobi, R ., Julien, M. , David, F., Basell, L ., Wood,
R., Davies, W. & Bronk Ramsey, C. (2010). Chronology of the
Grotte du Renne (France) and implications for the context of
ornaments and human remains within the Châtelperronian.
Proceedings of the National Academy of Sciences of the United
States of America 107: 20234-20239.
Higham, T. F. G., Basell, L., Jacobi, R.M, Wood, R., Bronk
Ramsey, C. & Conard, N. J. (2012). Τesting models for the
beginnings of the Aurignacian and the advent of figurative art
and music: the radiocarbon chronology of Geißenklösterle.
Journal of Human Evolution 62: 66 4- 676.
Higham, T. F. G., Douka, K., Wood, R., Bronk Ramsey, C., Brock,
F., Basell, L., Camps, M., Arrizabalaga, A., Baena, J.,
Barroso-Ruíz, C.Bergman, C., Boitard, C., Boscato, P.,
Caparrós, M., Conard, N.J., Draily, C., Froment, A., Galván,
B., Gamb assini, P., Garci a-Moren o, A., Grim aldi, S., H aesaerts,
P., Holt, B., Iriarte-Chiapusso, M.-J., Jelinek, A., Jordá Pardo,
J.F., Maíllo-Fernández, J.-M., Marom, A., Maroto, J.,
Menéndez, M., Metz, L ., Morin, E., Moroni, A., Negrino, F.,
Panagopoulou, E., Peresani, M., Pirson, S., de la Rasilla, M.,
Riel-Salvatore, J., Ronchitelli, A., Santamaria, D., Semal, P.,
Slimak, L., Soler, J., Soler, N., Villaluenga, A., Pinhasi, R.,
Jacobi, R. (2014). The timing and spatiotemporal patterning
of Neanderthal disappearance. Nature 512, 306 .
Inizan, M. L., Reduron, M., Roche, H. & Tixier, J. (1995).
Technologie de la pierre taillée. Cercle de recherches et d'études
préhistoriques. CNRS, Meudon.
Jöris, O. (2001). Der spätmittelpaläolithische Fundplatz Buhlen
(Grabungen 1966–69). Stratigraphie Steinartefakte und Fauna
des Oberen Fundplatzes. Dr. Rudolf Habelt Verlag, Bonn.
Jöris, O. (2003). Zur chronostratigraphischen Stellung der
spätmittelpaläolithischen Keilmessergruppen. Der Versuch einer
kulturgeographischen Abgrenzung einer mittelpaläolithischen
Formengruppe in ihrem europäischen Kontext. Bericht der
Römisch-Germanischen Kommission 84: 49-153.
Kind, C. J., Ebinger-Rist, N., Wolf, S., Beutelspacher, T. &
Wehrberger, K. (2014). The smile of the Lion Man. Recent
excavations in Stadel Cave (Baden-Württemberg, southwestern
Germany) and the restoration of the famous Upper Palaeolithic
figurine. Quartär 61: 12 9-145 .
Kitagawa, K., Krönneck, P., Münzel, S. C . & Conard, N. J. (2012).
Exploring cave use and exploitation among cave bears,
carnivo res and hominins in the Sw abian Jura, Germ any. Journal of
Taphonomy 10: 439 -461.
Krönneck, P., Niven, L. & Uerpmann, H.-P. (2004). Middle
Palaeolithic subsistence in the Lone valley (Swabian Alb, southern
Germany). International Journal of Osteoarchaeology 14: 212-224.
Kuhn, S. L. (1995). Mousterian Lithic Technology: an Ecological
Perspective. University Press, Princeton.
Kuhn, S. L. (2004). Upper Paleolithic raw material economies at
Ücagızl ı cave, Turkey. Journal of Ant hropological Ar chaeology 23:
431- 4 4 8 .
Mania, D. (2002). Der mittelpaläolithische Lagerplatz am
Ascherslebener See bei Königsaue (Nordharzvorland).
Praehistoria Thuringica 8: 16-75.
Mania, D. & Toepfer, V. (1973). Königsaue. Gliederung, Ökologie
und mittelpaläolithische Funde der letzten Eiszeit. Deutscher
Verlag der Wissenschaften, Halle.
Mellars, P. & French, J. C. (2011). Tenfold population increase in
Western Europe at the Neanderthal-to-modern human
transition. Science 333: 623– 627.
Miller, C. E. (2015). A Tale of Two Swabian Caves.
Geoarchaeological Investigations at Hohle Fels and
Geißenklösterle. Kerns Verlag, Tübingen.
Moreau, L . (2009). Geißenklösterle. Das Gravettien der
Schwäbischen Alb im europäischen Kontext. Kerns Verlag,
Tübingen.
Müller-Beck, H.-J. (Ed.) (1983). Urgeschichte in Baden-
Württemberg. Konrad Theiss Verlag, Stuttgart.
Münzel, S. C. & Conard, N. J. (2004). Change and continuity in
subsiste nce during the Middle and Upp er Palaeolithic in th e Ach
Valley of Swabia (South-west Germany). International Journal of
Osteoarchaeology 14: 225-243.
Münzel, S. C. (2019). Die jungpleistozäne Großsäugerfauna aus
dem Geißenklösterle. In: N. J. Conard, M. Bolus & S. C. Münzel
(Eds.) Geißenklösterle: Chronostratigraphie, Paläoumwelt und
Subsistenz im Mittel- und Jungpaläolithikum der Schwäbischen
Alb. Kerns Verlag, Tübingen, 147-328.
Neruda, P. (2017). GIS analysis of the spatial distribution of
Middle Palaeolithic artefacts in Kůlna Cave (Czech Republic).
Quaternary International 435(Part A): 58-76.
Niven, L. (2006). The Palaeolithic occupation of Vogelherd Cave.
Implications for the subsistence behavior of late Neanderthals and
early modern humans. Kerns Verlag, Tübingen.
Odell, G. H. (2004). Lithic Analysis. Manuals in Archaeological
method, theory, and technique. Kluwer Academic, New York.
Palma di Cesnola, A . (1989). L'Uluzzien: Faciès italien du
leptolithique archaïque. l'Anthropologie 93 : 78 3- 811.
Pelegrin, J. (2000). Les techniques de débitage laminaire au
Tardiglaciaire: critères de diagnose et quelques réflexions.
L’Europe Centrale et Septentrionale au Tardiglaciaire.
Confrontation des modèles régionaux 7: 73–86.
Peresani, M., Cremaschi, M., Ferraro, F., Falguères, C., Bahain,
J.-J., Gruppioni, G., Sibilia, E., Quarta, G., Calcagnile, L. &
Dolo, J.-M. (2008). Age of the final Middle Palaeolithic and
Uluzzian levels at Fumane Cave, Northern Italy, using 14C, ESR,
234U/230Th and thermoluminescence methods. Journal of
Archaeological Science 35: 2986–2996.
Peresani, M., Chrzavzez, J., Danti, A., De March, M., Duches, R.,
Gurioli, F., Muratori, S., Romandini, M., Tagliacozzo, A. &
Trombino, L. (2011). Fire-places, frequentations and the
environm ental setting of t he final Mouste rian at Grotta di Fu mane:
a report from the 2006-2008 research. Quartär 58: 131-151.
Peters, E. (1931). Die Heidenschmiede in Heidenheim a. d. Br. Eine
Studie über das Altpaläolithikum in Württemberg.
Schweizerbart'sche Verlagsbuchhandlung, Stuttgart.
Picin, A . (2017). Short-term occupations at the lakeshore: a
technological reassessment of the open–air site Königsaue
(Germany). Quartär 63: 7-32 .
Pop, E., Charalampopoulos, D., Arps, C. S., Verbaas, A.,
Roebroeks, W., Gaudzinski-Windheuser, S. & Langejans, G.
(2018) . Middle Palaeolithic Percussive Tools from the Last
Interglacial Site Neumark-Nord 2/2 (Germany) and the Visibility
of Such Tools in the Archaeological Record. Journal of Paleolithic
Archaeology 1: 81-106.
Quartär 66 (2019)Middle Paleolithic of Geißenklösterle Cave
79
Porraz, G. (2009). Middle Paleolithic mobile toolkits in short
term human o ccupations: t wo case studies. Eu rasian Prehisto ry 6:
33-55.
Posth, C., Wißing, C., Kitagawa, K., Pagani, L., van Holstein, L.,
Racimo, F., Wehrberger, K., Conard, N. J., Kind, C. J.,
Bocherens, H. & Krause, J. (2017). Deeply divergent archaic
mitochond rial genome provi des lower time boun dary for Afri can
gene flow into Neanderthals. Nature communications 8: 160 46.
Prüfer, K., Racimo, F., Patterson, N., Jay, F., Sankararaman, S.,
Sawyer, S., He inze, A., Renaud , G., Sudmant, P. H., de Fil ippo,
C., Li, H., Mallick, S., Dannemann, M., Fu, Q., Kircher, M.,
Kuhlwilm, M., Lachmann, M., Meyer, M., Ongyerth, M.,
Siebauer, M., Theunert, C., Tandon, A., Moorjani, P., Pickrell,
J., Mullikin, J. C., Vohr, S. H., Green, R. E., Hellmann, I.,
Johnson, P. L. F., Blanche, H., Cann, H., Kitzman, J. O.,
Shendure, J., Eichler, E. E., Lein, E. S., Bakken, T. E.,
Golovanova, L. V., Doronichev, V. B., Shunkov, M. V.,
Derevianko, A. P., Viola, B., Slatkin, M., Reich, D., Kelso, J. &
Pääbo, S. (2014). The complete genome sequence of a
Neanderthal from the Altai Mountains. Nature 505: 43-49.
Reimer, P. J. Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G.,
Ramsey, C. B. & Grootes, P. M. (2013). IntCal13 and Marine13
radiocarbon age calibration curves 0 –50,000 years calBP.
Radiocarbon 55: 1869–1887.
Rhodes, S. E., Ziegler, R., Starkovich, B. M. & Conard, N. J.
(2018) . Small mammal taxonomy, taphonomy, and the
paleoenvironmental record during the Middle and Upper
Paleolithic at Geißenklösterle Cave (Ach Valley, southwestern
Germany). Quaternary Science Reviews 185: 199 -221
Richard, M. (2015). Chronology of human occupations during the
Upper Plei stocene in the Swabian J ura, Germany, and in the S aone
and Rhone valleys, France, using electron spin resonance and
uranium series dating method, ESR/U-Th. Ph.D., dissertation,
Muséum National d’Histoire naturelle, Paris.
Richard, M., Falguères, C., Valladas, H., Ghaleb, B.,
Pons-Branchu, E., Mercier, N., Richter, D. & Conard, N. J.
(2019) . New electron spin resonance (ESR) ages from
Geißenklösterle Cave: A chronological study of the Middle and
early Upper Paleolithic layers. Journal of Human Evolution 133:
133-145.
Richter, D., Waiblinger, J., Rink, W. J. & Wagner, G. A. (2000).
Thermoluminescence, electron spin resonance and 14C-dating
of the late Middle and early Upper Palaeolithic site of
Geißenklösterle Cave in southern Germany. Journal of
Archaeological Science 27: 71-89.
Richter, J. (1997). Der G-Schichten-Komplex der Sesselfelsgrotte.
Zum Verständnis des Micoquien. Franz Steiner Verlag,
Saarbrücken.
Richter, J. (2016). Leave at the height of the party: A critical
review of the M iddle Paleolith ic in Western Cent ral Europe from
its beginnings to its rapid decline. Quaternary International 411 :
107-12 8.
Riek, G. (1934). Die Eiszeitjägerstation am Vogelherd im Lonetal I:
Die Kulturen. Akademische Buchhandlung Franz F. Heine,
Tübingen.
Riek, G. (1938). Ein Beitrag zur Kenntnis des süddeutschen
Solutréen. Germania 22: 147-150 .
Rink, W. J., Kandel, A. W. & Conard, N. J. (2002). The ESR
geochronology and Geology of the open-air Palaeolithic
deposits in Bollschweil, Germany. Archaeometry 44: 635-650.
Rios-Garaizar, J., Eixea, A. & Villaverde, V. (2015). Ramification
of lithic production and the search of small tools in Iberian
Peninsula Middle Paleolithic. Quaternary International 361:
188 -199
Roussel, M ., Bourg uignon, L . & Soressi, M . (2009). Identification
par l’expérimentation de la percussion au percuteur de calcaire
au Paléolithique moyen: le cas du façonnage des racloirs
bifaciaux Quina de Chez Pinaud (Jonzac, Charente-Maritime).
Bulletine Societé Préhistorique Francaise 106: 219-238.
Schmidt, R. R. (1910). Der Sirgenstein und die Diluvialen
Kulturstätten Württembergs. With contributions by E. Koken &
A. Schliz. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.
Schmidt, R. R. (1912). Die diluviale Vorzeit Deutschlands.
Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.
Škrdla, P., Nejman, L., Rychtaříková, T., Nikolajev, P. & Lisá, L.
(2014). New observations concerning the Szeletian in Moravia.
Quartär 61: 87-101.
Soressi, M. (2002). Le Moustérien de tradition acheuléenne du
sud-ouest de la France. Discussion sur la signification du faciès à
partir de l' étude comparée de quatre sites: Pech-de-l'Azé I, Le
Moustier, La Rochette et la Grotte XVI. Ph.D., dissertation,
Université Bordeaux I, Bordeaux.
Soressi, M. & Geneste, J.-M. (2011). The history and efficacy of
the chaîne opératoire approach to lithic analysis: Studying
techniques to reveal past societies in an evolutionary
perspective. PaleoAnthropology 2 011; 33 4- 350 .
Soressi, M. & Roussel, M. (2014). European middle to Upper
Paleolithic transitional industries: Châtelperronian. In: C. Smith
(Ed.) Encyclopedia of global archaeology. Springer, New York,
2679-2693.
Soressi, M., McPherron, S. P., Lenoir, M., Dogandžić, T.,
Goldberg, P., Jacobs, Z., Maigrot, Y., Martisius, N. L., Miller,
C. E., Rendu, W., Richards, M., Skinner, M. M., Steele, T. E.,
Talamo, S. & Texier, J.-P. (2013). Neandertals made the first
specialized bone tools in Europe. Proceedings of the National
Academy of Sciences of the United States of America 110 :
14186-14190.
Street, M. (2002). Plaidter Hummerich. An Early Weichselian
Middle Palaeolithic site in the central Rhineland, Germany. Mit
einem Bei trag von T. v. Kolfschoten . Monographien de s Römisch-
Germanischen Zentralmuseums Mainz 45. Mainz und Bonn.
Teyssandier, N., Bolus, M. & Conard, N. J. (2006). The Early
Aurignacian in central Europe and its place in a European
perspective. In: O. Bar-Yosef & J. Zilhão (Eds.) Towards a
definition of the Aurignacian. Proceedings of the Symposium
held in Lisbon, Portugal, June 25-30, 2002, Trabalhos de
Arqueologia 45, Lisbon, 241-256.
Toniato, G., Münzel, S. C., Starkovich, B. M. & Conard, N. J.
(2018) . Middle and Upper Palaeolithic bone retouchers from
the Swabian Jura: raw materials, curation and use. In: J. M.
Hudson et al. (Eds.) The Origins of Bone Tool Technologies.
Verlag des Römisch-Germanischen Zentralmuseums, Mainz,
251-2 67.
Tostevin, G. B. (2 003). Attribute analysis of the lithic technologies
of Stránská skála II–III in their regional and inter-regional
context. In: J. Svoboda & O. Bar-Yosef (Eds.) Stránská skála:
Origins of the Upper Palaeolithic in the Brno Basin. Peabody
Museum Publications, Cambridge (MA), 77-118.
Uthmeier, T. (2004). Micoquien, Aurignacien und Gravettien in
Bayern. Eine regionale Studie zum Übergang vom Mittel-zum
Jungpaläolithikum. Archäologische Berichte 18. Deutsche
Gesellschaft für Ur- und Frühgeschichte, Bonn.
Valoch, K. (1988). Die Erforschung der Kulna-Höhle 1971-1976.
Moravske Muzeum, Brno.
Wagner, E. (1983). Das Mittelpaläolithikum der Großen Grotte
bei Blaubeuren (Alb- Donau Kreis). Konrad Theiss Verlag,
Stuttgart.
Weiss, M., Otcherednoy, A. & Wiśniewski, A. (2017). Using
multivariate techniques to assess the effects of raw material,
flaking behavior and tool manufacture on assemblage variability:
an example from the late Middle Paleolithic of the European
Plain. Journal of Archaeological Science 87: 73-94.
Weißmüller, W. (1995). Sesselfelsgrotte II: die Silexartefakte der
unteren Schichten der Sesselfelsgrotte: ein Beitrag zum Problem
des Moustérien Teil II. Franz Steiner Verlag, Saarbrücken.
Wenzel, S. (2007). Neanderthal presence and behaviour in
central and Northwestern Europe during MIS 5e. Developments
in Quaternary Sciences 7: 173-193.
Quartär 66 (2019) N. J. Conard et al.
80
Wetzel, R. & Bo sinski, G. (1969). Di e Bocksteinschmie de im Lonetal
(Markung Ra mmingen, Krei s Ulm). Verlag Müller & Gräf f, Stuttgart.
Wolf, S. (2014). Schmuckstücke – Die Elfenbeinbearbeitung im
Schwäbischen Aurignacien. Kerns Verlag, Tübingen.
Zilhão, J ., Banks, W. E., d’Er rico, F. & Gioia, P. (2015). Analysis of
site formation and assemblage integrity does not support
attribution of the Uluzzian to modern humans at Grotta del
Cavallo. PloS one 10(7): e0131181.
5
Quartär 66 (2019)
Inhalt -
Contents
Exploitation of the natural environment by Neanderthals from Chagyrskaya Cave (Altai)
Nutzung der natürlichen Umwelt durch Neandertaler der Chagyrskaya-Höhle (Altai)
Ksenia A. K, Victor P. C, Alena V. S, Maciej T. K, Magdalena K,
William R, Sergei V. V, Sergei V. M & Andrei I. K............................................7-31
Petrological characterisation of the “Tertiary quartzites” from the site of Troisdorf-Ravensberg
(North Rhine-Westphalia, Germany): first insights in Middle Palaeolithic outcrop exploitation
Petrologische Charakterisierung der „tertiären Quarzite“ von Troisdorf-Ravensberg (Nordrhein-Westfalen,
Deutschland): Erste Einblicke in die Nutzung des mittelpaläolithischen Aufschlusses
Alejandro P, Iñaki Y, Andreas P & Erich Cß.......................................................33-50
Lithic assemblages from the Middle Paleolithic of Geißenklösterle Cave provide insights on
Neanderthal behavior in the Swabian Jura
Steinartefaktinventare aus dem Mittelpaläolithikum des Geißenklösterle, Deutschland: Neue Erkenntnisse zum
Verhalten der Neandertaler auf der Schwäbischen Alb
Nicholas J. C, Viola C. S, Michael B & Manuel W.....................................................51-80
Everything lost? Reconstruction of Middle- and Upper Paleolithic occupations at the Felsenhäusl-
Kellerhöhle, Lower Altmühl Valley (Franconian Jura, SE Germany)
Alles verloren? Rekonstruktion mittel- und jungpaläolithischer Begehungen der Felsenhäusl-Kellerhöhle, Unteres
Altmühltal (Fränkischer Jura, SO Deutschland)
Merlin H, Alvise B & Thorsten U.................................................................81-110
Geoarchaeology and geochronology of the Upper Palaeolithic site of Temerești Dealu Vinii,
Banat, Romania: Site formation processes and human activity of an open-air locality
Geoarchäologie und Geochronologie der jungpaläolithischen Fundstelle Temerești Dealu Vinii, Banat, Rumänien:
Natürliche und menschliche Einflüsse auf die Genese einer Freilandfundstätte
Wei C, Stephan P, Adrian D, Thomas A, Nicole K, Alexandru C,
Janina J. B & Philipp S........................................................................................................111-134
A pre-Heinrich Event 3 assemblage at Fumane Cave and its contribution for understanding the
beginning of the Gravettian in Italy
Ein vor das Heinrich 3-Ereignis datierendes Inventar aus der Fumane-Höhle und sein Beitrag zum Verständnis des
Beginns des Gravettien in Italien
Armando F & Marco P...................................................................................................135-154
Quartär 66 (2019)
6
The Gravettian site Meča Dupka (Serbia) and its regional context
Die Gravettien-Fundstelle Meča Dupka (Serbien) und ihr regionaler Kontext
Senka P & Predrag P........................................................................................................155-175
Pigments on Upper Palaeolithic mobile art. Spectral analysis of figurines from Mal’ta culture (Siberia)
Pigmente auf jungpaläolithischer mobiler Kunst. Spektralanalyse von Statuetten der Mal’ta Kultur (Sibirien)
Liudmila Valentinovna L................................................................................................................177-185
One ring to interpret. Bone ring-type adornment from the Epigravettian site Bratčice
(Moravia, Czech Republic)
Un anneau à interpréter. Un ornement en os de type anneau du site épigravettien de Bratčice
(Moravie, République Tchèque)
Zdeňka N , Bibiana H, Petr N & František Z....................................187-200
Hassi Berkane and Late Iberomaurusian Land Use in the Eastern Rif, Morocco
Hassi Berkane und die Landnutzung im späten Ibéromaurusien im östlichen Rif, Marokko
Tay lor O, Helmut B & Gerd-Christian W...............................................................201-223
Kostënki 17 (Spitsynskaya) and Kostënki 6 (Streletskaya): recent fieldwork and new 14C dates
Kostënki 17 (Spitsynskaja) und Kostënki 6 (Streletskaja): jüngste Feldforschung und neue 14 C-Daten
Rob D, Alexander B, Alexei A, Anton L, Andrei S &
Thomas H...................................................................................................................................225-230
Book reviews
Buchbesprechungen................................................................................................................................231-237
