Content uploaded by Dušan Mihailović
Author content
All content in this area was uploaded by Dušan Mihailović on Jan 23, 2025
Content may be subject to copyright.
Middle Paleolithic assemblages
from Crvena Stijena: insights
from the 1954–1963 excavations
Dušan Mihailović
ISBN 978-86-6427-343-5
FACULTY OF PHILOSOPHYFACULTY OF PHILOSOPHY
UNIVERSITY OF BELGRADEUNIVERSITY OF BELGRADE
Faculty of Philosophy, University of Belgrade │ 2024
Middle Paleolithic
assemblages
from Crvena Stijena:
insights from the
1954–1963 excavations
Dušan Mihailović
Middle Paleolithic assemblages from Crvena Stijena:
insights from the 1954–1963 excavations
Prehistoric settlements in caves and rock-shelters
of Serbia and Montenegro
Fascicule III
Dušan Mihailović
Belgrade 2024
Publisher
Faculty of Philosophy, University of Belgrade
Center for Archaeological Research
Čika Ljubina 18–20, Beograd 11000, Srbija
www.f.bg.ac.rs
For the publisher
Danijel Sinani,
Dean of the Faculty of Philosophy
Reviewed by
Robert Whallon
Steven Kuhn
Editor
Marko Janković
Translation by
Predrag Radović
Cover photo
Marianne van Twillert-Wennekes
Cover design by
Ivana Zoranović
Design by:
Dušan Mihailović
Set by
Dosije studio, Belgrade
Printed by
Službeni glasnik, Belgrade
Printed in
150 copies
ISBN 978-86-6427-343-5
The research was conducted as part of the project
“Neanderthal and Early Modern Human Interactions in the Central Balkans” (NEEMO 7746827),
funded by the Science Fund of the Republic of Serbia.
This book is the third volume in the series of monographs
“Prehistoric Settlements in Caves and Rock-Shelters of Serbia
and Montenegro,” published by the Center for Archaeological
Research of the Faculty of Philosophy in Belgrade. The second
volume in this series featured Upper Paleolithic and Mesolithic
artifacts from the Crvena Stijena rockshelter in Montenegro,
while this volume focuses on Middle Paleolithic material. The
analysis of materials from Basler’s excavations (1960–1963) was
carried out in 2007 during more recent investigations of Crvena
Stijena. I am deeply grateful to Robert Whallon, the director
of those excavations, for allowing me to dedicate myself to this
work. The analysis of lithic materials was greatly supported by
Zvezdana Vušović-Lučić, former curator of the County Museum
Nikšić, Bojana Mihailović, curator of the National Museum of
Serbia (artifact photography), and Gordana Grabež (database
processing). The analysis of Brodar’s material (excavated in
1958) was completed in 2012 after the material was returned
from Ljubljana to the County Museum Nikšić. I received valu-
able assistance from Predrag Radović, who translated the man-
uscript, and Sofija Dragosavac, who prepared the illustrations.
I extend my warmest gratitude to everyone who contributed to
the analysis of the materials and the preparation of the manu-
script for printing.
| 7
Contents
9 | Introduction
10 | General Information about the Site and Previous Research
16 | Previous Interpretations and Open Questions
17 | Material and Methods
21 | Basler‘s Assemblages
21 | Raw Materials
22 | Products of Flaking
25 | Tool s
47 | Brodar’s Assemblages
47 | Raw Materials and General Structure of the Main Categories of Artifacts
49 | Products of Flaking
52 | Tool s
59 | Middle Paleolithic sequence of Crvena Stijena
59 | Occupation intensity
65 | Reduction sequence
68 | Variability of tools
71 | The Middle Paleolithic of Crvena Stijena in a regional context
72 | Crvena Stijena and the Beginnings of the Levallois Method in the Paleolithic of
Southeastern Europe
73 | Crvena Stijena and the Question of the Distribution of Bifacial Backed Tools
74 | Crvena Stijena and the issue of Adriatic Charentian
75 | Crvena Stijena and the Micromousterian Question
75 | Crvena Stijena and the Middle to Upper Paleolithic Transition
79 | Climatic and demographic factors of technological variability
79 | Middle Paleolithic
81 | Middle to Upper Paleolithic Transition
83 | Conclusion
85 | Plates
113 | References
| 9
Introduction
The Crvena Stijena rockshelter in Montenegro has attracted scientific interest since the initial
excavations (1954–1959) led by Alojz Benac, Borivoje Čović, and Mitja Brodar (Benac and
Brodar, 1957, 1958), with Đuro Basler’s subsequent work (1960–1964) solidifying the site’s importance
(Basler, 1975). Crvena Stijena is established as a unique Paleolithic site in the region, featuring an
extensive sequence of over 30 Paleolithic layers exceeding 20 meters in thickness. Layers XXXI–
XII produced particularly remarkable materials, including a rich assemblage of Middle Paleolithic
lithic artifacts. Furthermore, the lower layers of the site contain ash deposits that are more than a
meter thick. At the time, no comparable sites had been discovered in the northern Mediterranean,
and similar findings were only known from the Levant (Jelinek et al., 1973; Meignen et al., 2007).
The findings suggest that Neanderthal communities repeatedly inhabited Crvena Stijena over a long
period, likely using it as a base camp during various phases of occupation.
The research methodology employed during the 1950s and 1960s excavations of Crvena Sti-
jena did not meet today’s standards, and the findings from the site have never been fully published.
However, it would be unfair to suggest that the early researchers of Crvena Stijena did not strive to
meet the high standards of their time in both their research methods and the publication of their
findings. Upon reaching the Paleolithic layers, the initial excavation team invited Paleolithic spe-
cialist Mitja Brodar from the Archaeological Institute in Ljubljana to conduct further excavations.
Brodar led meticulous archaeological excavations of the upper layers of the Paleolithic sequence,
uncovering more than 3,000 flint artifacts. This material remains the most significant and repre-
sentative Middle Paleolithic sample from the earlier excavations at Crvena Stijena. Brodar published
his research reports in several papers (Brodar, 1957a, 1957b, 1962). Sadly, subsequent excavations
were less meticulous. In his eagerness to reach the bedrock, Đuro Basler and his team excavated
an area that nearly encompassed the entire floor of the rockshelter. Basler’s team excavated over 10
m of deposits in just three years, removing all layers containing Upper Paleolithic material. Unlike
Brodar, who collected all the excavated artifacts, Basler retained only a small selection of charac-
teristic finds and discarded the remainder. However, it is important to note that Basler consistently
aimed for an interdisciplinary approach in his research of Crvena Stijena. He entrusted the sedi-
ment analysis to K. Brunnacker (Brunnacker, 1975), the faunal analysis to M. Malez (Malez, 1975),
and the analysis of the raw materials used for knapping to J. Pamić (Pamić, 1975). The collection of
papers presenting the new results was published in 1975 in Nikšić. Unfortunately, this monograph
was released in a limited print run, resulting in the inventory of findings from Crvena Stijena being
poorly known to the academic community for a long time. Few authors have attempted to examine
the Crvena Stijena findings within a broader regional context based on the published data (Ivanova,
1979; Kozłowski, 1992, 2002; Whallon, 1999).
Although the original documentation from earlier research has not been preserved, the pub-
lished data and labels on the artifact bags show that both A. Benac and Đ. Basler conducted their
excavations with full stratigraphic control. The labels on the bags and the artifacts provide informa-
tion not only about the layer from which the finds were recovered but also about the depth at which
they were excavated. The stratigraphy of the layers—which can still be observed in the preserved
profile today—is accurately represented in the profile sketch published by Basler. Interestingly, while
10 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Basler did not typically employ Bordes’ method (which was already common at the time), he se-
lected materials for retention based on Bordes’ criteria (Bordes, 1961), preserving all artifacts on the
list, including cores. This created the conditions for a revised analysis of the archaeological material
from earlier excavations.
Thanks to Robert Whallon, the director of recent excavations at Crvena Stijena (2004–2015),
and Zvezdana Vušović-Lučić (formerly a curator at the County Museum Nikšić), we had the op-
portunity to conduct a detailed analysis of the artifact collections from earlier excavations dur-
ing the recent fieldwork. We published preliminary results of our analyses in several papers
(Mihailović and Whallon, 2017; Mihailović et al., 2017a), but we have yet to provide a compre-
hensive presentation. This book will do just that, following a similar approach to how we pre-
viously showcased findings from the Upper Paleolithic and Mesolithic (Mihailović, 2009a). We
believe that the material from earlier excavations should be published in detail, not only to clarify
uncertainties surrounding the artifact assemblages and individual finds but also because a revised
analysis of the data from the earlier excavations at Crvena Stijena can contribute to a broader
understanding of technological changes in the Middle Paleolithic in Southeast Europe. This is es-
pecially important in light of emerging questions regarding the nature of Neanderthal occupation
along the eastern Adriatic coast.
General Information about the Site and Previous Research
Crvena Stijena is located on the left bank of the Trebišnjica River (now Lake Bileća), in the ter-
ritory of the village of Petrovići, in the far west of Montenegro (Fig. 1). The rockshelter, set at an
elevation of approximately 700 m and facing southwest, has an entrance that is 26 m wide and was
about 25 m deep before the excavations began (Fig. 2). As the excavations progressed, the retreat of
the inner wall of the rockshelter resulted in the base expanding to cover an increasingly larger area
(Fig. 3). Basler examined the layers to a 20–30 m depth without reaching the bedrock. Although the
layers in the lowest part of the stratigraphic sequence contained finds, further excavations could not
continue due to the confined working space (Fig. 4). These layers will become accessible for inves-
tigation only once the interior excavation pits are expanded, as they are situated 8–10 m below the
level where the Paleolithic layers first emerge.
Figure 1. Geographical position of the Crvena Stijena rockshelter.
Introduction | 11
In the initial phase of research in 1954 and 1955, Alojz Benac and Borivoje Čović from the Na-
tional Museum of Bosnia and Herzegovina in Sarajevo examined layers containing cultural remains
Figure 2. Entrance to Crvena Stijena. (Photo: Jamie Clark)
Figure 3. Plan of the rock shelter with the areas of Brodar’s and Basler’s excavations and with
the positions of profiles A and B indicated (modified after Basler, 1975).
12 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
from the Neolithic (II and III), Mesolithic (IV–V), and Upper Paleolithic (VI–X) periods (Benac
and Brodar, 1958). By 1956, the excavation had reached the level of volcanic ash (X), and work be-
gan on the Middle Paleolithic layers. In 1958, Brodar opened a 3-square-meter test pit at the bottom
of the rockshelter to facilitate a detailed examination of these layers (Brodar, 1962). He investigated
several Middle Paleolithic layers (XII–XVIII) by removing mechanical excavation levels within the
geological layers. Layers XII–XVI were removed in 2–4 excavation levels, while layer XVIII (around
120 cm thick) was removed via as many as nine excavation levels (Brodar, 1962). Brodar attributed
the findings from these layers to the Micromousterian.
Basler investigated the Middle Paleolithic layers over an area of at least 90 m2. The specific
method he employed during the excavations remains unknown. However, many residents from
Petrovići and nearby villages participated in the excavations, and, similar to Brodar’s work, it is
likely that the sediment was not screened. Due to the expedited nature of the research and selective
collection of finds, Basler gathered relatively few artifacts—only 23 to 61 artifacts per layer.
Basler’s and Brodar’s layer designations largely align, though uncertainties arise regarding layers
XVIII to XX. Brodar mentions that layer XVIII “contained a thicker accumulation in the middle”
(Brodar, 1962: 16), and the depth of the pit he opened (11.7 m) corresponds to the depth at which
layer XXI was reached during Basler’s excavations (Fig. 5). It is possible that the layers excavated by
Basler were situated at a higher level. However, there is no doubt that Basler’s stratigraphic observa-
tions are valid, as confirmed by Brunnacker’s sedimentological studies and subsequent analyses of
the eastern and southern profiles of Basler’s excavation pit.
Brunnacker’s analyses of the sediments were based on granulometric examinations, including
particle size and shape analysis, as well as mineralogical and chemical analyses of carbonates, iron
oxides, and phosphates (Brunnacker, 1975; Morley, 2017). Based on these analyses, Brunnacker at-
tributed layers XXXI–XXV to the Penultimate Glacial Period (PGP), layer XXIV to the Riss-Würm
interglacial, layer XXIII to the period between the Last Glacial Period and the Amersfoort inter-
stadial, and layers XXII–XVII to the Amersfoort and Brörup interstadials (i.e., a pluvial phase with
humid climate). In contrast, the layers above layer XVII were assigned to the ‘high glacial’ stage of
the Würm glaciation, marked by a relatively cold and dry climate.
Basler’s and Malez’s interpretations of the chronology of the Middle Paleolithic layers at Crvena
Stijena (Basler, 1975; Malez, 1975) relied heavily on Brunnacker’s observations, as did all subsequent
analyses. Basler classified the collected artifacts into various Mousterian facies (more on this later),
while Malez conducted a paleontological analysis and compiled a list of identified taxa by layer.
Figure 4. Stratigraphic profile of Crvena Stijena (after Brodar, 2009, Fig. 36).
Introduction | 13
The excavations at Crvena Stijena were renewed in 2004 through a collaboration between the
University of Michigan in Ann Arbor (R. Whallon), the Archaeological Research Center of Montene-
gro in Podgorica (M. Cerović, L. Saveljić-Bulatović, M. Baković), and the Center for Culture in Nikšić
(Z. Vušović Lučić). Experts from the National Museum of Serbia in Belgrade (B. Mihailović) and the
Faculty of Philosophy at the University of Belgrade (D. Mihailović) also participated in the research.
In recent excavations, a significant portion of the scree cone at the front of the rockshelter was
removed, creating space for work in the eastern section. Here, Middle Paleolithic layers were excavated
across approximately 20 m2. Because the excavated area could not be directly connected to the interior
profile of Basler’s excavation pit, the identified layers could not be reliably correlated with those from
previous excavations. This uncertainty applies to layer M1, positioned directly below the tephra, and
layer M5, which corresponds either to layer XV or, more likely, to layer XVII (Mihailović et al., 2017b).
In contrast, layers M2 and M3 were confidently identified as layers XII and XIII, respectively.
Figure 5. Comparative view of the western profile from Brodar’s excavation (A) and the
eastern profile from Basler’s excavation (B) (modified after Brodar, 2009, Fig. 35).
14 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
R. Whallon’s research, from 2004 to 2015, included an extensive dating program for Middle
Paleolithic layers. Six samples were dated using the electron spin resonance (ESR) method—two
from layer XXIV and four from layer XX—while four samples of burnt flint were dated using the
thermoluminescence (TL) method, three from layer XXIV and two from layer XX. Additionally, ra-
diocarbon dating was applied to one sample from layer M3a1 (i.e., layer XIII) and two samples from
layer M2c (i.e., layer XII).
The dating results were inconsistent. ESR dating yielded an average LU age of 78.3 ± 0.3 ka for
layer XXIV and 48.3 ± 2.4 ka for layer XX, while TL dating of burnt flint produced an age of ~70 ka
for one sample from layer XXIV and 52.7 ± 6.6 ka for another. The slightly younger date for the sec-
ond sample is attributed to the presence of limestone blocks in layer XXIV, which may have resulted
in “a gamma dose rate significantly different from the one measured on adjacent soft sediments”
(Mercier et al., 2017: 142). A sample from layer XX was dated using TL to 65.5 ± 14 ka. Radiocarbon
dates obtained for layers M3a1 and M2c covered a broad range from 45 to 49 ka cal BP, while opti-
cally stimulated luminescence (OSL) dating yielded ages of 44.2 ± 3.4 ka for layer XIII, and 43.2 ±
3.2 ka and 37.6 ± 2.9 ka for layer XII.
Several chronological models were proposed based on the obtained absolute dates—along with
sedimentological, paleontological, and archaeological indicators. Through extensive sedimentologi-
cal investigations that included analyses of the fine sediment fraction and the application of sedi-
Figure 6. Eastern and southern profiles of Basler’s sondage. (Photo: Jamie Clark)
Introduction | 15
mentological, geochemical, mineral magnetic, and microscopic techniques, Mike Morley identified
three lithofacies: III, which contained layers of coarse, angular to sub-angular limestone gravels; II,
which encompassed anthropogenic layers of ash and debris (XXIV–XIV—see Fig. 6); and I, charac-
terized by coarse limestone gravel beds associated with layers XIII and XII (Morley, 2007; Baković
et al., 2009). Lithofacies III has been attributed to MIS 6, lithofacies II to MIS 5a/4, and lithofacies I
(XIII and XII) to MIS 3 (Fig. 7). The volcanic ash layer overlying the deposits of lithofacies I (spe-
cifically layer XI) has been reliably identified as Campanian Ignimbrite (CI) tephra (Morley and
Woodward, 2011).
Figure 7. Comparison of alternative interpretations of the chronology of the Crvena Stijena Middle Paleo-
lithic sequence and its possible correlations with Marine Isotope Stages (MIS) (after Whallon and Morin,
2017, Fig. 19.1).
16 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
In contrast to M. Morley, Eugène Morin and Marie-Cécile Soulier’s faunal analysis suggested
that only the lowest layers (i.e., XXV through XXXI) should be attributed to MIS 5 and MIS 4, while
all other horizons should be attributed to MIS 3 (Morin and Soulier, 2017). Their interpretation was
based on several indicators: the stable structure of faunal and floral remains throughout the entire
sequence; the fact that some ESR and OSL dates place layer XXIV in MIS 3; the presence of horses
in layers XXVII–XXV, which may indicate a cooler climate, combined with the presence of ther-
mophilous species (such as tortoises and roe deer in layers XXVII–XXXI) which suggest a warmer
interval (MIS 5).
The radiocarbon dates for the upper two layers of the Middle Paleolithic sequence were simi-
larly inconsistent. Layer M3a1 was dated to more than 49,257 cal BP, while layer M2c yielded two
dates: 45,289 cal BP and 45,535 cal BP (Mercier et al., 2017). A single previous date for layer XII
spans a range from 46,104 to 42,902 cal BP (Vogel and Waterbolk, 1972). It remains to be seen
whether the slightly younger date for this older sample is due to the lack of ultrafiltration during
pretreatment or because the upper part of layer XII was not captured in recent excavations.
The possibility of layer XII being of a more recent age is further supported by OSL dates for
layers XIII and XII, with layer XIII dated at 44.2 ± 3.4 ka BP and two dates for layer XII at 43.2 ± 3.2
ka BP and 37.6 ± 2.9 ka BP. Mercier et al. (2017) concluded that layer XIII likely accumulated during
the cold Heinrich 5 (H5) event and that the M2c level (corresponding to the middle section of layer
XII) can be correlated with the onset of Greenland Interstadial 12 (GI-12). Layer M1, which may
correspond to the upper section of layer XII, was deposited shortly after the Campanian Ignimbrite
(CI) eruption around 40 ka (Giaccio et al., 2017).
Previous Interpretations and Open Questions
The challenges in identifying and interpreting Mousterian assemblages from Crvena Stijena
are significant in their own right but also raise broader questions regarding the variability of
Mousterian culture across Southeast Europe. In contrast to Brodar, who broadly classified the
assemblages from layers XII–XVIII as Micromousterian (Brodar, 1962), Basler distinguished sev-
eral distinct Mousterian facies, assigning specific names to some of them (Basler, 1975). Basler
attributed artifacts from layers XXXI–XXIX to pre-Mousterian, XXV–XVIII to proto-Mousteri-
an, XXIV–XXIII to Mousterian, XXII–XXI to Pontinian, XX–XIX to Mousterian with triangular
points, XVIII to Pontinian, XVII–XIV to Mousterian, XIII to Denticulate Mousterian, and XII to
late Mousterian.
S. Ivanova (1979) classified the finds from Crvena Stijena into the so-called “Pinios–Crvena
Stijena А” facies, J. K. Kozłowski characterized the material from the lowest layers of the site as a
distinct type of Mousterian—Crvena Stijena–Karain E (Kozłowski, 1992, 2002), while Mihailović et
al. (2017a) attributed the Mousterian assemblages from Crvena Stijena to the Typical Mousterian of
Crvena Stijena type (layers XXXI–XXIV), Pontinian/Charentian (layers XXII–XX), and late Mous-
terian/Micromousterian. A considerable number of named stone tool industries (NASTIES) have
accumulated here (Shea, 2014).
In critiquing Bordes’ and the ‘typological’ approach to classifying Mousterian industries in
the Balkans, Tamara Dogandžić (2023) sought to implement a technological framework for their
differentiation. Considering the technological trends of the Middle Paleolithic at Crvena Stijena
and other sites in Southeast Europe, she proposed that after MIS 5, earlier industries character-
ized by thick flake tools and Levallois blanks, which were used for extended periods (often sharp-
ened) were replaced by industries dominated by discoid technology, featuring numerous single-
use tools. However, Dogandžić also did not shy away from referencing ‘Micromousterian’ trends,
emphasizing that changes in flaking methods are at least partially responsible for the microlithic
character of some industries (Dogandžić, 2023). Ruka et al. (2023) also adopted a technological
approach, citing discoid technology as the dominant knapping method in the Late Middle Paleo-
lithic of Southeast Europe.
Introduction | 17
Indeed, defining lithic assemblages presents a significant challenge with no satisfactory solu-
tion currently available—whether in the context of the Middle (Frick, 2020) or Upper Paleolithic
(Pesesse, 2017)—especially considering that many names for Paleolithic techno-complexes were
established long ago based on research in Western Europe (primarily France) and some continue
to be regarded as distinct cultural entities. This situation has led to numerous, largely unpro-
ductive debates regarding the ‘terminological’ problems associated with identifying Mousterian
facies. However, the question arises as to whether the naming of entities and their associated be-
havioral patterns should be entirely abandoned (Shea, 2014), especially if technological attributes
are interpreted in a manner similar to how variations in tool assemblages were once understood—
without considering the circumstances that influenced their emergence and later use. Regardless
of whether we intend to use Bordes’ terminology or local names (or decide not to use either), it
is evident that other factors have also influenced the variability of Mousterian: the mobility of
communities, the duration and nature of the occupation, economic behaviors, and demographic
and climatic conditions. A true understanding of the factors contributing to the variability of
these industries cannot be achieved if technological manifestations are examined within narrow
regional frameworks, without considering long-term trends and the geographic distribution of
technological phenomena.
Crvena Stijena could have been a crucial site for studying the longue durée in the Middle Paleo-
lithic of the Mediterranean—spanning from MIS 6/5 to MIS 3—yet its potential is hindered by
the outdated approach to excavation, incomplete collections, and a lack of sufficient contextual in-
formation regarding the site and its findings. Despite these challenges, studying long-term trends
remains feasible, particularly given recent research at several Middle Paleolithic sites in the Balkans
and the continuation of microarchaeological work at Crvena Stijena in 2016, led by G. Tostevin
and G. Monnier from the University of Minnesota. These studies have already produced significant
results (Jambrina-Enríquez et al., 2019; Monnier et al., 2020; Rodríguez de Vera et al., 2020; Bradák
et al., 2021; Jones et al., 2021; Lambrecht et al., 2021), opening new pathways for understanding Cr-
vena Stijena within a broader regional framework. New discoveries have also advanced our under-
standing of the Middle to Upper Paleolithic transition in the Balkans and northern Mediterranean.
Research shows that Quina and Levallois technologies appeared remarkably early in the Balkans
(Mihailović et al., 2022a, 2024) and that Quina Mousterian was prevalent not only in the subal-
pine zone (Delpiano et al., 2022) but, according to some researchers, also along the eastern Adriatic
coast (Karavanić and Banda, 2023). Furthermore, in the western Mediterranean, sites have emerged
where both potential remains of anatomically modern Homo sapiens and Upper Paleolithic artifacts
connected with the Initial Upper Paleolithic appear remarkably early (Slimak et al., 2022). These
findings offer a transformative perspective on the Middle to Upper Paleolithic transition across the
northern Mediterranean.
Our planned analyses aim to address questions related to different phases of the Middle Paleo-
lithic, aligned with the previously outlined classification (Mihailović et al., 2017a). According to
this classification, three types of Mousterian industries are found at Crvena Stijena: 1) industries
from the lower layers (XXXI–XXIV), where the “Typical Mousterian” is present and documented at
various sites, both along the coast and in the interior of the Balkans; 2) industries from the middle
stratigraphic complex (layers XXII–XVIII), where the Quina component is more prominent; and
3) industries from the upper stratigraphic complex (layers XVIII–XII), which show Micromouste-
rian features. Particular attention will be devoted to the previously documented Uluzzian elements
(Mihailović and Whallon, 2017) in light of recent discoveries concerning Uluzzian practices in the
Balkans (Kaczanowska et al., 2010) and the Italian peninsula (Ronchitelli et al. 2018; Villa et al.,
2018; Peresani et al., 2019; Delpiano et al., 2024; Higham et al., 2024).
Material and Methods
The following chapters present the lithic assemblages collected by Đ. Basler and M. Brodar dur-
ing the early excavations at Crvena Stijena (Figs. 8 and 9). Our analytical approach was tailored to
18 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Figure 9. Crvena Stijena: Excavations of Pleistocene layers. (Photo: JU Muse-
ums and Galleries Nikšić, archive)
Figure 8. Crvena Stijena: Excavations of Holocene layers. (Photo: JU Muse-
ums and Galleries Nikšić, archive)
20 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
the assemblages’ preservation state (Figs. 10 and 11), optimizing their interpretive potential. Brodar’s
assemblages (3,337 artifacts) were analyzed from both technological and typological perspectives,
allowing for a detailed examination of tool structure and characteristics. In contrast, Basler’s assem-
blages (764 artifacts) were analyzed primarily in terms of cores, tools, and selected artifacts classi-
fied according to Bordes’ typological framework. Quantitative data presented in previous publica-
tions (Mihailović and Whallon, 2017; Mihailović et al., 2017a) were reviewed and, where necessary,
revised. Following this reassessment of available materials, some tools were reclassified. All correc-
tions are noted in the footnotes below the text.
Raw materials were identified solely based on general macroscopic characteristics, as detailed
characterization and sourcing still need to be completed (Ćulafić et al., 2017). For the Basler col-
lection, a comprehensive overview is provided, encompassing cores, characteristic flakes, and tool
assemblage structures across all layers. However, quantitative data on raw materials and products of
knapping are offered only for the lowest layers (XXXI–XX). We believe that Brodar’s material from
layers XVIII–XII is much more relevant for reconstructing knapping technology than Basler’s mate-
rial from the same layers (Mihailović and Whallon, 2017).
The analyses were conducted to provide as much insight as possible into the organization of
lithic technology and the operational sequence involved in artifact production. Conclusions regard-
ing mobility and settlement patterns during particular phases were drawn based on the general
structure of raw materials, primary artifact categories, and the presence of cortex on artifacts. Due
to the incomplete sample and the lack of data on the volume of excavated sediment, we were unable
to calculate the so-called whole-assemblage behavioral index (WABI) of Clark and Barton (2017),
which would have provided a more precise understanding of the techno-economic behavior of the
communities.
The chaîne opératoire analysis was conducted using general parameters (cortex, dorsal surface,
and platform attributes) alongside criteria for identifying Levallois (Boëda, 2013), discoid (Terradas,
2003), and Quina (Turq 1989; Bourguignon, 1996, 1997) methods, with additional consideration
given to criteria for other reduction methods—including Kombewa (Tixier and Turq, 1999), and
the “Asprochaliko method” (Papaconstantinou, 1989). The so-called truncated-faceted pieces (TFP)
were identified based on the Dibble and McPherron (2007) criteria, though they were predominant-
ly classified as tools (Shalagina et al., 2015). The typological analysis of tools was carried out based
on standard classification criteria, largely following Bordes’ criteria. Laplace’s criteria (Laplace, 1972)
were used to analyze retouch.
| 21
Basler‘s Assemblages
Raw Materials
The raw materials used for artifact production are classified as follows: a) high-quality raw
materials (chalcedony and opal), b) medium-quality flint (mostly gray and beige in color), and c)
low-quality flint and siliceous rock. In some cases, the raw material could not be identified due to
factors such as burning, patina, or desilicification. Consequently, the analysis was restricted to the
assemblages from layers XXXI–XX.
The analysis revealed that artifacts made of low-quality gray-beige flint dominate in layers
XXXI–XX, with their proportion varying from 60% to 73% in layers XXXI, XXX, XXV, and XXI,
51% to 52% in layers XXIV and XXII, and just 37% in the assemblage from layer XX (Table 1). In
the latter assemblage, the proportion of medium-quality raw materials (mostly gray flint) is 51.6%.
High-quality raw materials appear in proportions of no more than 6% and are found only in layers
XXX, XXIV, and XXII.
Table 1. Raw-material structure: Basler’s assemblages
from the layers XXXI–XX.
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX
Chalcedony/Opal n0
0.0
1
3.6 0 0 0 0 0
0.0
4
5.9 03
2.5 03
4.8
%
Flint and medium-
quality chert
n24
32.0
10
35.7 0 1 1 1 2
18.2
23
33.8 344
37.3
2
20.0
32
51.6
%
Low-quality chert and
silicious rocks
n51
68.0
17
60.7 6 2 3 0 8
72.7
35
51.5 262
52.5
6
60.0
23
37.1
%
Indeterminate n0
0.0
0
0.0 0 0 0 0 1
9.1
6
8.8 09
7.6
2
20.0
4
6.4
%
Total n75 28 6 3 4 1 11 68 5118 10 62
A slightly higher number of finds (more than 50) in layers XXXI, XXX, XXIV, and XXII sug-
gests that the site was more frequently inhabited during these phases, while occupation in the other
phases (up to layer XX) was likely very brief. This is further supported by the higher proportion of
heavily burnt and charred artifacts, which were found in slightly higher proportions (10% or more)
only in layers XXXI, XXX, XXV, XXIV, and XXI (Table 2).
22 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Table 2. Basler’s assemblages: occurrence of charred (1), totally burned (2),
partially patinated (3), totally patinated (4), partially desilified (5)
and totally desilified (6) artifacts.
XXXI
(75)
XXX
(29)
XXIX
(6)
XXVIII
(3)
XXVII
(4)
XXVI
(1)
XXV
(19)
XXIV
(79)
XXIII
(5)
XXII
(133)
XXI
(11)
XX
(78)
1n7
9.3
4
13.8 1 0 0 0 1
5.3
7
8.9 04
3.0 04
5.1
%
2n0
0.0
0
0.0 0 0 0 0 1
5.3
6
7.6 08
6.0
2
18.2
5
6.4
%
3n6
8.0
4
13.8 0 0 0 0 0
0.0
5
6.3 02
1.5
1
9.1
5
6.4
%
4n1
1.3
1
3.4 0 0 0 0 0
0.0
2
2.5 13
2.2
1
9.1
1
1.3
%
5n9
12.0
2
6.9 0 0 0 0 0
0.0
2
2.5 03
2.2
0
0.0
0
0.0
%
6n0
0.0
1
3.4 0 0 1 0 0
0.0
1
1.3 00
0.0
1
9.1
0
0.0
%
In layer XXII, which yielded a slightly higher number of finds, the proportion of low-quality
flint was lower, as was also the case in layers XXIV and XX. If low-quality flint was sourced from
nearby deposits and high-quality flint from more distant ones, it could suggest that Neanderthal
groups were more mobile during these phases of occupation. In any case, there is insufficient evi-
dence to attribute the organization of lithic technology in these phases to a curated model of tech-
no-economic behavior.
Products of Flaking
The presence of cortex on the completely preserved artifacts from layers XXXI, XXII, and XX
suggests that some initial knapping and decortication of cores were carried out on-site. Flakes with
cortex traces are present in layers XXXI and XX at around 20% and in layer XXII at more than 30%
(Tables 3 and 4).
Table 3. Basler’s assemblages: presence of cortex on retouched artifacts (whole specimens).
XXXI
(50)
XXX
(18)
XXIX
(4)
XXVIII
(0)
XXVII
(4)
XXVI
(0)
XXV
(7)
XXIV
(45)
XXIII
(3)
XXII
(90)
XXI
(6)
XX
(52)
Total n12 1 2 0 0 1 4 0 34 110 1
Table 4. Basler’s assemblages: dorsal surface of retouched artifacts (whole specimens):
parallel scars in one direction (1); parallel scars in one direction and cortex (1-c);
multidirectional, diagonal and perpendicular scars (3, 4, 5); multidirectional, diagonal and
perpendicular scars and cortex (3-c, 4-c, 5-c); cortical specimens—with more than
50% of cortex on dorsal surface (6); Kombewa flakes (7), scars removed by retouching (8).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX
1n14
28.0
5
27.8 1 0 0 0 2 14
31.1 016
17.8 312
23.1
%
1-c n5
10.0
0
0.0 1 0 0 0 0 0
0.0 010
11.1 04
7.7
%
3, 4, 5 n23
46.0
11
61.1 1 0 3 0 2 21
46.7 337
41.1 228
53.8
%
3-c, 4-c, 5-c n3
6.0
1
5.5 1 0 0 0 0 2
4.4 012
13.3 14
7.7
%
6n4
8.0 0 0 0 0 0 1 2
4.4 012
13.3 02
3.8
%
7n1
2.0
1
5.5 0 0 1 0 1 1
2.2 02
2.2 0 0
%
8n0
0.0
0
0.0 0 0 0 0 1 5
11.1 01
1.1 02
3.8
%
Total n50 18 4 0 4 0 7 45 3 90 6 52
Basler’s Assemblages | 23
Based on the analysis of the dorsal face attributes (Table 4), it was determined that most of the
cores, from which flakes later used for tool production were removed, were flaked from multiple di-
rections, likely employing the discoid or Levallois centripetal method. However, nearly a third of the
flakes were removed from a single direction—particularly in the lowest layers of Crvena Stijena. The
highest percentage of flakes with parallel negative scars was recorded in layer XXXI (38%), while the
lowest was found in layer XX (around 30%).
Layers XXII and XX differ from the lowest layers (XXXI, XXX, XXIV) even in terms of plat-
form preparation (Table 5). Specifically, prepared (faceted and bifacial) platforms are much more
common in the assemblages from the lowest layers (overall 40–50%) than in those from layers XXII
and XX (overall 15–20%). In contrast, cortical platforms (similar to the dorsal face cortex) are most
prevalent in layer XXII (14.7%), while they are poorly represented in the lowest layers (as well as
in layer XX). Layers XXXI, XXIV, and XXII recorded a slightly higher prevalence of platforms re-
moved via retouch (over 10%). These layers also show evidence of more frequent occupation, as
indicated by the higher percentage of burned artifacts. However, this is not the case for layer XXII
(which also contains many finds), where burned artifacts are present at only 5.1%.
Table 5. Basler’s assemblages: platforms on retouched artifacts (whole specimens and
proximal fragments): plain (1), dihedral (2), faceted (3), edge (4), punctiform (5), c
ortical (6), undeterminable and damaged (7, 8), removed by retouching (9).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX
1n17
32.1
6
31.6 1 0 1 0 2 17
34.0 241
43.2 427
55.1
%
2n10
18.9
2
10.5 1 0 0 0 2 3
6.0 05
5.3 15
10.2
%
3n11
20.7
8
42.1 1 0 0 0 1 14
28.0 012
12.6 15
10.2
%
4n0 0 0 0 0 0 0 0 0 3
3.2 01
2.0
%
5n1
1.9 0 0 1 1 0 0 0 0 3
3.2 11
2.0
%
6n4
7.5
1
5.3 0 0 1 0 1 7
14.0 114
14.7 12
4.1
%
7n01
5.3 0 0 1 0 0 1
2.0 02
2.1 02
4.1
%
8n1
1.9 0 0 0 0 0 0 0 0 5
5.3 03
6.1
%
9n9
17.0
1
5.3 1 1 0 0 1 8
16.0 010
10.5 03
6.1
%
Total n53 19 4 2 4 0 7 50 3 95 8 49
Retouched tools have an average greater length compared to the few unretouched artifacts
(Table 6), which—had the collection not been selectively curated—could suggest that Neanderthal
communities preferred larger flakes for tool production.
Table 6. Average length of unretouched flakes and tools on flakes (Basler’s assemblages).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX
Unretouched
flakes mm 24.0 32.0 33.3 29.0 25.7
Retouched
flakes mm 49.5 43 48.5 51.3 40.3 39.5 35.4
Layer XXXI
Up to 11 cores were found in layer XXXI of Crvena Stijena. Most are made of relatively
poor-quality beige-gray chert, with one core made of high-quality banded chert. There were five
24 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
discoid and five Levallois cores, and one Kombewa-type core on a massive limestone flake. Three
preferential and two centripetal cores were identified among the Levallois cores (Pl. 1: 1–3, 5).
Among the characteristic products of knapping, two preferential Levallois flakes were identified,
while three Levallois flakes (one preferential and two centripetal) were used as blanks for tool
production.
Layer XXX
The material from layer XXX includes one preferential and two centripetal Levallois cores.
Among the characteristic flakes, a single unretouched preferential Levallois flake and a Kombewa
flake (from which a denticulate tool was made) were identified.
Layer XXIX
A centripetal Levallois core and a unipolar burin-like core on a flake fragment were recorded in
layer XXIX. Two retouched flakes of the éclats débordants type were also found.
Layers XXVIII–XXVI
No cores or characteristic flakes were found in layers XXVIII and XXVII. In layer XXVI, one
centripetal Levallois core was recorded.
Layer XXV
One Levallois centripetal core (Pl. I: 4) and one retouched Levallois flake were recorded in
l ay e r X XV.
Layer XXIV
One Levallois centripetal core on the ventral side of a flake was recorded, and one Kombewa
core was removed at a 90° angle relative to the axis of the flake. Among the unretouched flakes, two
preferential Levallois flakes were identified.
Layer XXII
Two of the three cores found are discoid. One Kombewa core with a platform made on flake
truncation was also identified.
Two preferential Levallois flakes, three recurrent Levallois flakes (two of which are retouched),
and one fragmentary Levallois flake (which cannot be classified more precisely) were also identi-
fied. The assemblage also includes two débordant flakes, one Kombewa flake, one naturally backed
knife, and one knife with a non-cortical back and damage on the opposite edge. Among the tools,
two TFPs were identified—one formed on a thick flake fragment, while the other showed a bipolar
orientation of flaking traces.
Layer XX
A preferential Levallois core and a core fragment were found. In addition, two centripetal Lev-
allois flakes, one preferential pointed Levallois flake, and one atypical (Debénath and Dibble, 1994)
Levallois flake with a laterally positioned cortex were recorded.
Layer XVIII
The assemblage includes one unipolar bladelet core for bladelet and laminar flakes, and three
cores on flakes, including two cores on secondarily used tools. Among the tools made from cores
Basler’s Assemblages | 25
are: a sidescraper flaked on both the ventral and dorsal sides from a laterally oriented truncation,
and a sidescraper flaked similarly—from a truncation on the dorsal side and a retouch on the ven-
tral side. An unretouched thick flake contained traces of flaking on its ventral side.
Layer XVII
The recorded cores include a preferential core with traces of diagonally oriented recurrent flak-
ing on its lower side, an exhausted preferential core, an atypical recurrent Levallois core on the ven-
tral side of the flake, and a bipolar core with altered orientation—flaked from one direction on one
side and from the opposite direction on the other side.
Among the unretouched flakes, two recurrent Levallois blades were identified.
Layer XVI
A single Levallois centripetal core was found, with traces of platform preparation along its en-
tire perimeter. In addition, a TFP was also identified (Fig. 20: 3).
Layer XV
The assemblage includes two cores: a recurrent Levallois core flaked from the lower end and
laterally on the upper end, and a centripetal Levallois core on a flake. Three unretouched Levallois
flakes were also found: one flaked from a recurrent unipolar core, two flaked from centripetal cores,
and one retouched preferential Levallois flake.
Layer XIV
Only one core was recorded—a recurrent-type Levallois core with altered orientation featuring
two prepared platforms at a 90° angle. Three unretouched preferential Levallois flakes and three
débordant flakes with partially preserved core edges were identified among the characteristic flakes.
Layer XIII
The assemblage includes only one example of a centripetal Levallois core on a flake and one
irregular globular core.
Layer XII
The material from layer XII includes an atypical centripetal Levallois core and a centripetal core
on a flake.
Too ls
In Basler’s assemblages, sidescrapers overwhelmingly dominate the tool structure in almost all
layers (Tables 7 and 8), reaching 60–70% of the total number of tools in layers XXIV, XXII, and XX.
An exception is layer XVII, where denticulate and notched tools dominate, and layer XV, where
regular retouched flakes are more numerous than sidescrapers. Other tools are neither as prevalent
nor consistently present across all layers.
26 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Table 7. Tools (Basler’s assemblages).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX XVIII XVII XVI XV XIV XIII XII
Retouched
Levallois
flakes
n2
3.3
0
0.0 0 0 0 0 1 0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
2
7.7
0
0.0
1
2.8
0
0.0
0
0.0
%
Mousterian
points
n1
1.6
0
0.0 0 0 0 0 0 2
3.4
0
0.0
3
2.9
0
0.0
2
3.6
0
0.0
0
0.0
0
0.0
0
0.0
1
2.8
1
2.6
1
5.3
%
Sidescrapers n22
36.1
9
39.1 1 0 1 0 6 41
69.5
1
10.0
69
67.0
4
40.0
34
60.7
16
33.3
5
22.7
7
26.9
3
23.1
11
31.4
12
31.6
9
47.4
%
Endscrapers n7
11.5
3
13.0 0 0 0 0 1 2
3.4
0
0.0
5
4.8
1
10.0
1
1.8
3
6.2
3
9.1
1
3.8
2
15.4
2
5.7
3
7.9
1
5.3
%
Retouched
blades
n0
0.0
2
8.7 0 1 0 0 0 1
1.7
0
0.0
0
0.0
0
0.0
2
3.6
2
4.2
2
13.6
0
0.0
0
0.0
1
2.8
2
5.3
2
10.5
%
Backed
knives
n0
0.0
0
0.0 0 0 0 0 0 1
1.7
1
10.0
0
0.0
0
0.0
2
3.6
0
0.0
0
0.0
0
0.0
2
15.4
0
0.0
0
0.0
0
0.0
%
Retouched
flakes
n6
9.8
2
8.7 0 1 0 0 1 8
13.5
1
10.0
9
8.7
1
10.0
3
5.3
8
16.7
2
9.1
6
23.1
4
30.8
7
20.0
6
15.8
1
5.3
%
Notched and
denticulated
tools
n9
14.7
7
30.4 3 1 1 0 0 3
5.1
2
20.0
11
10.7
3
30.0
3
5.3
10
20.8
8
36.4
6
23.1
1
7.7
6
17.1
6
15.8
1
5.3
%
Raclettes n0
0.0
0
0.0 0 0 0 0 0 0
0.0
0
0.0
0
0.0
0
0.0
1
1.8
2
4.2
0
0.0
1
3.8
0
0.0
5
14.3
1
2.6
0
0.0
%
Splintered
pieces
n1
1.6
0
0.0 0 0 0 0 0 0
0.0
0
0.0
2
1.9
0
0.0
0
0.0
0
0.0
0
0.0
1
3.8
0
0.0
0
0.0
0
0.0
0
0.0
%
Trun cat ed
pieces
n4
6.5
0
0.0 0 0 1 0 0 1
1.7
0
0.0
1
1.0
0
0.0
0
0.0
1
2.1
0
0.0
1
3.8
1
7.7
1
2.8
1
2.6
0
0.0
%
Burins n0
0.0
0
0.0 0 0 0 0 0 0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
1
2.6
1
5.3
%
Perforators n8
13.1
0
0.0 0 0 1 0 0 0
0.0
5
50.0
3
2.9
1
10.0
8
14.3
3
6.2
1
9.1
1
3.8
0
0.0
0
0.0
2
5.3
3
15.8
%
Combined
tools
n0
0.0
0
0.0 0 0 0 0 0 0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
1
2.6
0
0.0
%
Tool
fragments
n1
1.6
0
0.0 0 0 0 0 0 0
0.0
0
0.0
0
0.0
0
0.0
0
0.0
3
6.2
0
0.0
0
0.0
0
0.0
0
0.0
2
5.3
0
0.0
%
Tot al n61 23 4 3 4 0 9 59 10 103 10 56 48 21 26 13 35 38 19
Table 8. Sidescraper types (Basler’s assemblages).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX XVIII XVII XVI XV XIV XIII XII
Lateral n14 2 0 0 0 0 3 26 142 122 10 3 6 2 7 6 6
Bilateral n1 2 1 0 1 0 1 3 0 3 0 2 1 2 1 1 1
Transversal n4 3 0 0 0 0 2 7 0 17 2 7 3 1 3 5
Latero-
transversal n4 2 0 0 0 0 0 5 0 5 1 3 2 1 1
Bifacial n0 0 0 0 0 0 0 0 0 1 0 0 1
Tot al n22 9 1 0 1 0 6 41 168 434 16 5 7 3 11 12 9
Quina n0 0 0 0 0 0 1 2 0 14 0 7 4 1 1 2
Layer XXXI
Retouched Levallois flakes
Two retouched Levallois points were found. One is retouched with a semi-abrupt retouch, while
the other bears marginal retouch at the tip.
Mousterian points
A single Mousterian point with an unmodified tip was found (Pl. 2: 1; Fig. 12: 1).
Basler’s Assemblages | 27
Figure 12. Tools from Basler’s collection, layer XXXI. Mousterian point (1), sidescrapers (4–7), perforator (8),
endscrapers (9, 10), truncated tool (12). (Photo: Bojana Mihailović)
28 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Sidescrapers
Among the sidescrapers (22 pieces), lateral pieces dominate, although bilateral, transversal, and
latero-transversal sidescrapers have also been identified (Pl. 2: 2–6; Fig. 12: 2–7, 11). Four pieces
show ventral thinning.
Perforators
The perforators (8 pieces) are mostly retouched with a semi-abrupt marginal retouch, although
some examples have been processed via deep retouch (Pl. 2: 8–10; Fig. 12: 8). Semi-abruptly re-
touched triangular perforators, almost geometric in shape, represent a characteristic type (Pl. 2: 8, 9).
Other tools
Although numerous (9 pieces), denticulate and notched tools are not typologically differenti-
ated. Most are partially retouched with shallow, semi-abrupt retouching. Seven atypical endscrap-
ers were found, featuring either an arched or nosed working edge. Layer XXXI also yielded six
retouched flakes, one splintered piece, four tools with a truncation (Pl. 2: 7; Fig. 12: 12), and one
fragment of a semi-abruptly retouched tool that cannot be more precisely classified.
Layer XXX
Sidescrapers
Among the sidescrapers (9 pieces), there are two lateral, two bilateral, three transversal, two lat-
ero-transversal, and one bifacial specimen (Pl. 3: 1, 2; Fig 13: 1–3).1 Many of these sidescrapers are
convergent, including both pointed bilateral examples and the latero-transversal sidescrapers made
on déjeté flakes. Two sidescrapers were made on débordant flakes, featuring transverse negative scars
on the dorsal side, indicating a shift in the orientation of core knapping. One oval sidescraper was
worked with bifacial retouch and discarded after an accidental overshot blow (Fig. 13. 3).
Endscrapers
Three endscrapers were found. One is formed on a lateral-transverse sidescraper made from a
déjeté flake. The second endscraper, made on a thick flake, shows evidence of thinning on the ven-
tral side, while the third does not show any distinctive features.
Retouched blades
Two retouched blades were found in the layer—a medial fragment with a shallow, semi-abrupt
retouch on its edge (Pl. 3. 3), and a naturally backed knife with a semi-abrupt, deep retouch on the
opposite edge.
Other tools
Other tools, particularly the denticulate and notched tools (7 pieces) and the retouched flakes
(2 pieces), were not heavily modified and clearly served as ad hoc tools.
Layer XXIX
Tools identified in layer XXIX include one sidescraper, a denticulate tool made on an elongated
débordant blade, an elongated denticulate flake inversely retouched with denticulate retouch on one
edge, and a débordant flake with a notch (Pl. 3: 4; Fig. 13: 4).
1 The bifacial sidescraper was not included in the list of tools in Tables 10.3 and 10.4 of Mihailović et al.
(2017a).
Basler’s Assemblages | 29
Layer XXVIII
The knapped tools found in this layer include only one thick blade retouched using inverse-
raised retouch, one denticulate artifact, and one retouched flake.
Layer XXVII
The material from layer XXVII (Pl. 3: 5, 6; Fig. 13: 5) includes a convergent bilaterally re-
touched pointed sidescraper with ventral thinning, tools with an oblique truncation and a partially
retouched back, a perforator with bilateral semi-abrupt retouch, and a denticulate flake.
Layer XXVI
This layer contained no identifiable tools, except for one oval bifacial sidescraper found at the
contact between layers XXVII and XXVI (Pl. 3: 7; Fig. 13: 6).
Figure 13. Tools from Basler’s collection, layers XXX (1–3), XXIX (4), XXVII (5), and XXVII/XXVI (6). Side-
scrapers. (Photo: Bojana Mihailović)
30 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Layer XXV
Among the six sidescrapers found (Pl. 3: 9, 10; Fig. 14: 1–4), three were lateral (one with evi-
dence of ventral thinning), one was an elongated bilateral sidescraper made on a blade, and two were
transversal sidescrapers (one of which exhibited Quina retouch). Other tool types include an end-
scraper made on a thick, ventrally thinned flake, a retouched Levallois flake, and a retouched flake.
Layer XXIV
Mousterian points
Two pieces were identified—one is laterally retouched, while the other represents an asym-
metrical point made on a déjeté flake.
Sidescrapers
– Lateral sidescrapers (26 pieces—Pl. 14: 1–4). The straight lateral sidescrapers were
made on simple flakes, with one sidescraper also made on a Levallois flake. Among
them, five sidescrapers were made on flake fragments. Regarding convex lateral sides-
crapers, one was retouched via Quina retouch, while another was made on a débordant
flake. Three convex lateral sidescrapers (one made on a Levallois centripetal flake) were
partially retouched, and one had a denticulate retouch on the opposite edge. Three
convex lateral sidescrapers were made on flake fragments showing traces of thinning
on one or more edges.
– Bilateral sidescrapers (3 pieces). Three were identified, including one ventrally thinned piece.
– Transversal sidescrapers (7 pieces). They are not standardized. Two are straight, made on
déjeté flakes. One of these features inverse retouch and ventral thinning. Two are convex,
one of which is retouched with Quina retouch and made on a cortical flake. Three sides-
crapers are partially retouched.
– Latero-transversal sidescrapers (5 pieces—Pl. 14: 5). Among these, one is made on a pref-
erential Levallois flake, one is ventrally thinned (from the direction of the retouch), and
one features a Quina retouch. Tree pieces show a slightly convex transverse edge.
Evidence of ventral thinning on some pieces can be characterized as bifacial retouch. A non-
cortical back was observed in several specimens.
Endscrapers
Two atypical, nosed endscrapers were identified—one made on a preferential Levallois flake
and the other on a déjeté flake.
Retouched blades
Only one blade was found, bilaterally retouched with a semi-abrupt retouch.
Retouched backed knife
Only one piece was found, with facial retouch on the ventral side.
Denticulate tools
Three pieces were recorded, one made on a débordant flake.
Perforators
Three perforators were made on thick, asymmetric flake fragments: one shows traces of dor-
sal thinning, while the second is elongated and slightly curved. The second perforator is bilater-
Basler’s Assemblages | 31
ally retouched and nearly qualifies as a Mousterian point, although it displays a back (Pl. 4: 7).
The third represents a double-pointed perforator with a shape exhibiting geometric characteris-
tics (Pl. 4: 8).2
2 Basler’s assemblage from layer XVIII contains two perforators that are not listed in Table 10.3 in
Mihailović et al. (2017a: 158).
Figure 14. Artifacts from Basler’s collection, layers XXV (1–4) and XXIV (5, 6). Sidescrapers (1–4), Levallois
flake (5), pseudo-Levallois flake (6). (Photo: Bojana Mihailović)
32 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Figure 15. Tools from Basler’s collection, layer XXIV. Sidescrapers. (Photo: Bojana Mihailović)
Basler’s Assemblages | 33
Layer XXIII
The collection from layer XXIII contains few tools. The following were identified: a fragment of
a convex lateral sidescraper, a knife on a blade with a semi-abrupt deep retouched back, tools with a
shallow inverse notch, one denticulate tool, and one retouched flake.
Figure 16. Tools from Basler’s collection, layer XXII. Levallois blade (1), Mousterian point (2), sidescrapers
(3–8). (Photo: Bojana Mihailović)
34 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Layer XXII
Mousterian points
Three Mousterian points were found: one typical (Fig. 16: 2), one ventrally thinned, and one
with retouching at the base.
Sidescrapers
– Lateral sidescrapers (42 pieces—Pl. 5: 2–3, 7, 9). Twenty-two straight lateral sidescrap-
ers were found, five of which have Quina retouch. Among them, three sidescrapers were
made on cortically based flakes, two on pseudo-Levallois flakes, and one on a débor-
dant flake. They are not standardized. Cortex is present on as many as nine specimens.
Among the 19 lateral convex sidescrapers, two are classified as Quina type. Three pieces
were made on naturally backed knives, another three on knives with non-cortical backs,
and three on débordant flakes. Four pieces contain cortex on the dorsal side (Fig. 16: 7).
Thinning was observed in one piece. A single concave sidescraper with Quina retouch
and lateral cortex was also found.
– Bilateral sidescrapers (3 pieces). Of the three specimens found, two are of the convergent
type.
– Transversal sidescrapers (17 pieces). Five are of the Quina type, and four are demi-Quina
(Fig. 16: 3–6, 8). One was formed on a Kombewa flake, and two on cortical flakes; one
shows evidence of thinning. In four cases, only part of the working edge is preserved.
– Latero-transversal sidescrapers (5 pieces). Of these, three are of the Quina type, one was
made on a naturally backed knife, while in two pieces, the platform (talon-dos) was used
as the back. Cortex is present on all five specimens.
Endscrapers
Five pieces were found, all atypical. One is classified as a nosed endscraper, and one as a point-
ed endscraper. Two pieces were made on naturally backed flakes—one with a flaked (non-cortical)
back and the other with a cortical back—while one was formed on a ventrally thinned flake with a
cortical platform.
Retouched flakes
Nine flakes with a shallow, semi-abrupt retouch were found. These include a pseudo-Levallois
flake and a knife with a non-cortical back, retouched along the opposite edge.
Denticulate tools
Eight pieces were recorded, two of which were made on recurrent Levallois flakes.
Notched tools
Three flakes with а notch were recorded, one on a centripetal Levallois flake.
Splintered pieces
Two splintered pieces were recorded.
Truncations
A single flake with an oblique truncation on a naturally backed knife was recorded.
Perforators
Three perforators were identified in the assemblage, one of which was formed on a Kombewa flake.
Basler’s Assemblages | 35
Layer XXI
A small number of artifacts were collected from this layer (Pl. 6). Various types of sidescrapers
were identified: a convex lateral, a straight transverse, a convergent transverse, and a latero-trans-
versal sidescraper. Furthermore, an atypical partially retouched endscraper, two denticulate tools,
and one notched tool were found, along with a retouched flake and a borer with a tip formed on the
proximal end of a denticulate, abruptly retouched backed point.
Layer XX
Mousterian points
Two Mousterian points were identified (Pl. 7: 1; Fig. 17: 2).
Sidescrapers
– Lateral sidescrapers (Pl. 7: 3; Fig. 17: 3, 5). Of the 13 straight pieces, two are of the Quina
type, two were made on débordant flakes, and three on pseudo-Levallois flakes. Of the
nine convex pieces, two are of the Quina type. One Quina sidescraper was made on a
knife with a non-cortical back.
– Bilateral sidescrapers (2 pieces). One is pointed, and the other is ventrally thinned.
– Transversal sidescrapers (Pl. 7: 2, 4; Fig. 17: 4). Two of the seven pieces are of the Quina
type. Four pieces are typical, while the others are essentially ad hoc tools.
– Latero-transversal sidescrapers (3 pieces). These include a latero-transversal Quina side-
scraper on a cortical flake, a latero-transversal sidescraper on a knife with a non-cortical
back, and a latero-transversal sidescraper on a débordant flake.
Endscrapers
A single example was found: a latero-transversal sidescraper on a débordant flake featuring an
endscraper working edge—which led to its classification under this tool type.
Retouched blades
Two blades are partially retouched with a semi-abrupt direct retouch, while one is partially re-
touched with a semi-abrupt marginal inverse retouch.
Points with abruptly retouched backs
Two points of this type were found, one of which was made on a partial débordant flake.
Raclettes
Only one raclette was identified—a larger flake bilaterally retouched with semi-abrupt deep
retouch.
Perforators
Eight perforators were found—five on blades and three on flakes (Pl. 7: 5–10).
Other tools
Among the tools that are not typologically differentiated, two denticulate tools were found (one
point and one laminar flake), a notched tool (Fig. 17: 6), and three retouched flakes—one of which
is a pseudo-Levallois point.
36 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Layer XVIII
Sidescrapers (16 pieces)
– Lateral sidescrapers (10 pieces). Most are fragmented. Almost all are convex, with one
concave and one sinusoidal sidescraper found. Two sidescrapers show traces of thinning
on the ventral side (Fig. 18: 2).
– Bilateral sidescraper (1 piece). The sidescraper is made on an elongated, irregular blade
(Fig. 18: 1).
Figure 17. Tools from Basler’s collection, layer XX. Unretouched blade (1), Mousterian point (2), sidescrapers
(3–5), notched tool (6). (Photo: Bojana Mihailović)
Basler’s Assemblages | 37
– Transversal sidescrapers (3 pieces). Two of these three very short transversal sidescrapers
are retouched with Quina retouch, and one with demi-Quina retouch. One of them is
slightly denticulate.
– Latero-transversal sidescrapers (2 pieces). One typical déjeté sidescraper has been identi-
fied, along with one transversely retouched Quina sidescraper with shallow lateral re-
touch; the latter shows evidence of thinning on its back.
Endscrapers
Three atypical endscrapers were found. One is a laterally oriented, carinated endscraper with a
thinned ventral side.
Retouched blades
Two retouched blades were found (Fig. 18: 5), one of which is pointed.
Figure 18. Artifacts from Basler’s collection, layer XVIII. Sidescrapers (1, 2), perforators (3, 4), retouched
blade (5), core on flake (6), retouched flakes (7, 8), truncated blade (9), backed point (10). (Photo: Bojana
Mihailović)
38 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Backed points
The collection also includes one typical abruptly retouched backed point (Pl. 8: 8; Fig. 18: 10).3
Retouched flakes
All eight pieces are retouched with shallow, semi-abrupt retouch (Fig. 18: 7, 8). Among them is
a semi-abruptly retouched pointed débordant flake, clearly produced via the Levallois method.
Denticulate and notched tools
Nine flakes retouched with denticulate retouch were found, along with one flake featuring a
shallow inverse notch.
Raclettes
Two flakes—one very small and one débordant—have been laterally retouched with a deep,
abrupt retouch.
Truncations
This is a partial débordant blade with a slightly convex, abruptly retouched truncation (Fig. 18: 9).
Perforators
These include a point with a bilateral semi-steep retouch (Fig. 18: 3), a borer on a discontinu-
ously denticulate flake, and an asymmetrical borer made on a flake.
Tool fragments
Three fragments of tools with a raised retouch were found.
Layer XVII
Sidescrapers
– Lateral sidescrapers (3 pieces). Two are straight, and one is convex (Fig. 19: 4). One con-
vex lateral sidescraper is made on an atypical Levallois flake (Pl. 8: 4).
– Bilateral sidescrapers (2 pieces). One sidescraper has a convex working edge and a curved
one, while the other is a pointed bilateral sidescraper formed on a débordant flake. Both
edges of the latter are partially retouched—one at the proximal end and the other at the
distal end.
Endscrapers
Three atypical endscrapers were found, one of which is a shouldered endscraper.
Retouched blades
A semi-abruptly retouched blade and a fragment of a convex, abruptly retouched blade were
recorded—both featuring a faceted platform (Pl. 8: 9; Fig. 17: 2, 3).
Perforators
One perforator is laterally oriented, while the other is a sidescraper with a pointed tip posi-
tioned laterally (Fig. 17: 7).
Other tools
Among the typologically undifferentiated tools (Fig. 19: 1, 5, 8), eight denticulate tools and two
retouched flakes were recorded.
3 The point is not listed in Table 10.3 of Mihailović et al. (2017a).
Basler’s Assemblages | 39
Layer XVI
Retouched Levallois flakes
Two retouched Levallois flakes were found.4
4 Of the total number of retouched flakes listed in Table 10.3 of Mihailović et al. (2017a), two have been
reclassified as retouched Levallois flakes.
Figure 19. Artifacts from Basler’s collection, layer XVII. Unretouched flake (1), retouched blades (2, 3), sides-
craper (4), denticulated flake (5), retouched flakes (6, 8), and perforator (7). (Photo: Bojana Mihailović)
40 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Sidescrapers
There are six lateral and one transversal sidescraper (Fig. 20: 1). One of the straight sidescrapers
is made on a regular flake, while the other three are made on débordant flakes (Pl. 8: 5). Two pointed
sidescrapers are made on preferential Levallois flakes.
Endscrapers
An atypical lateral shouldered endscraper, made on an asymmetrical preferential Levallois
flake, was identified.
Retouched flakes
Six flakes showing a shallow retouch were found.
Denticulate tools
One partial, irregular blade with slight denticulation, retouched at the distal end (Fig. 20: 4),
and two denticulate flakes were found.
Notched tools
Three notched tools were identified—one made on an irregular, abruptly retouched flake, an-
other on a Levallois flake, and the third on a Kombewa flake.
Raclettes
One slightly denticulate raclette on a flake was found. The dorsal side of the flake preserves the
platform edge from which the core flaking surface was prepared.
Truncations
One backed truncated tool was identified.
Perforators
A short, pointed Levallois flake with convergent negative scars, retouched at the tip, was classi-
fied as a perforator.
Layer XV
Sidescrapers
Two lateral sidescrapers and one bilateral sidescraper—partially retouched on one edge and
completely retouched on the other—were identified.
Endscrapers
The layer yielded one short endscraper on a flake and one nosed endscraper on a slightly thick-
er rejuvenation flake.
Splintered pieces
One splintered piece was found (Fig. 20: 2).
Abruptly retouched points
Two tools of this type were identified: an arched, abruptly retouched point and a larger segment
(Pl. 8: 10; Fig. 20: 7).5
Denticulate tools (1 piece) and retouched flakes (5 pieces) from layer XV show no distinctive
characteristics.
5 These tools are not presented in Table 10.3 of Mihailović et al. (2017a).
Basler’s Assemblages | 41
Layer XIV
Retouched Levallois flakes
A single retouched, preferential Levallois flake was identified (Pl. 8: 1).
Sidescrapers
– Lateral sidescrapers (7 pieces). Four straight lateral sidescrapers were found. One can be
characterized as marginal with semi-abrupt retouch, another is fragmented, the third is
made on a pseudo-Levallois flake, and the fourth, retouched with demi-Quina retouch,
is made on a thick, ventrally thinned flake (Pl. 8: 6). Also, three relatively small convex
lateral sidescrapers were identified.
– Bilateral sidescrapers (1 piece). The sidescraper is made on a Kombewa flake, showing
ventral thinning on the proximal end.
Figure 20. Artifacts from Basler’s collection, layers XVI (1, 3–5) and XV (2, 6, 7). Sidescraper (1), splintered
piece (2), truncated-faceted piece (3), slightly denticulated tools (4, 5), bladelet (6), abruptly retouched (seg-
ment-like) point (7). (Photo: Bojana Mihailović)
42 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
– Transversal sidescrapers (3 pieces). Two are typical and very short. One is made on a
cortical flake, and the other is retouched with a Quina retouch. In the third case, it is a
convex, inversely retouched sidescraper on the proximal end of a flake.
Endscrapers
Of the two endscrapers in Layer XIV, one is nosed with a laterally oriented working edge, while
the other shows a straight edge. On the latter piece, traces of ventral thinning can be observed, ex-
tending from the retouched edge (part of the retouch has been removed).
Mousterian points
A single example was found (Pl. 8: 2; Fig. 21: 10). The back has been retouched with semi-
abrupt marginal retouch and deep semi-abrupt retouch on one edge, while the other edge displays
a shallow retouch.6
Retouched blades
A single blade with lateral semi-abrupt discontinuous retouch.
Retouched flakes
A total of seven pieces were found.7
Raclettes
Four flakes retouched laterally with a semi-abrupt retouch were found.
Denticulate tools
A total of four pieces were found. They are not particularly distinctive.
Notched tools
These are two blades featuring notches on the edge.
Truncations
One tool of this type is made on laterally retouched flakes.
Perforators
Only one lateral perforator on an asymmetric débitage flake, bilaterally retouched, was found.
Layer XIII
Mousterian points
The collection includes only one Mousterian point (Pl. 8: 3).
Sidescrapers
Aside from the working edge position, the sidescrapers are not highly differentiated. Four
straight lateral, two convex lateral, five transversal, and one latero-transversal sidescrapers were
found (Fig. 22: 1–4, 7–9). Only one sidescraper exhibits Quina retouch, while two show ventral
thinning.
6 The piece is not listed in Table 10.3 of Mihailović et al. (2017a).
7 One retouched Levallois flake, described here, is included among the retouched flakes in Table 10.3 of
Mihailović et al. (2017a).
Basler’s Assemblages | 43
Endscrapers
Three atypical endscrapers were found. Two are nosed, while one displays lateral raised retouch.
Retouched blades
Two irregular bladelets with a semi-abrupt retouch were found (Fig. 22: 10, 11).
Retouched flakes
Of the total six retouched flakes, two are retouched débordant flakes.
Figure 21. Artifacts from Basler’s collection, layer XIV. Blades (1, 2, 4), sidescraper (3), crested blade (6),
notched tool (7), crested bladelet (6), bladelets and blade-like flakes (8, 9, 11, 12), and Mousterian point (10).
(Photo: Bojana Mihailović)
44 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Denticulate tools
These are made on different blanks. Four pieces were found, one of which is on a débordant
flake. Denticulate retouch is generally not well differentiated.
Notched tools
Two notched flakes were identified.
Truncations
One backed truncation was identified. The truncation is oblique, showing a semi-abrupt re-
touch, while one edge is retouched with a semi-abrupt/abrupt retouch.
Burins
One carinated burin was identified, which may have functioned as a core.
Perforators
Two perforators were identified. Both are laterally retouched via deep, abrupt retouch.
Combined tools
One endscraper-burin on a thick flake was found. At one end of the flake, there is a double lat-
eral burin on truncation, while the other end is modified into an endscraper with a straight working
edge.
Tool fragments
Two fragments of tools with a semi-abrupt retouch were found.
Layer XII
Mousterian points
One piece—laterally retouched by semi-steep retouch—was found (Fig. 23: 1).
Sidescrapers
Three straight lateral sidescrapers were found; one, exhibiting Quina retouch, shows traces of
ventral thinning (Fig. 23: 5). One of the two convex lateral sidescrapers is also Quina-type and made
on a cortical flake.
Other types of sidescrapers are represented by a single specimen each, including: a lateral si-
nusoidal, inversely retouched sidescraper thinned from the opposite edge with surface retouch; a
bilateral pointed sidescraper, ventrally thinned; a latero-transversal sidescraper on a thick flake; and
a bifacial sidescraper with additional semi-abrupt retouch on one edge (Pl. 8: 7; Fig. 23: 2).
Endscrapers
One atypical, carinated endscraper made on a flake was found.
Retouched blades
One blade is retouched laterally, and the other bilaterally by semi-abrupt retouch. Both blades
display a faceted platform (Fig. 23: 4, 7).
Retouched flakes
Only one retouched flake was identified (Fig. 23: 10).
Basler’s Assemblages | 45
Burins
A burin was identified on a semi-abruptly retouched truncation of a thick cortical flake, with
one edge further shaped by a semi-abrupt retouch.
Perforators
Three perforators were identified, two of which are formed on thick asymmetrical flakes. Two
are bilaterally retouched, with laterally oriented tips, and one is bilaterally retouched, with its tip
aligned with the flake axis (Fig. 23: 6, 8).
Figure 22. Artifacts from Basler’s collection, layer XIII. Sidescrapers (1–4, 7–9), splintered piece (5), pointed
flake (6), and retouched bladelets (10, 11). (Photo: Bojana Mihailović)
46 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Figure 23. Artifacts from Basler’s collection, layer XII. Mousterian point (1), sidescrapers (2, 5), truncated-fac-
eted piece (3), retouched blades (4, 7), perforators (6, 8), and retouched flake (9). (Photo: Bojana Mihailović)
| 47
Brodar’s Assemblages
Raw Materials and General Structure
of the Main Categories of Artifacts
Like Basler’s assemblages, low-quality raw materials overwhelmingly dominate Brodar’s assem-
blages, with their proportion ranging from 52% to 74%, except in the assemblage from layer XVII,
where they are present at only 21% (Table 9). A slightly lower percentage (around 50%) was re-
corded in layers XVIII and XII, likely due to the higher number of artifacts with undetermined raw
materials, resulting from thermal damage.
Table 9. Raw-material structure: Brodar’s assemblages from the layers XVIII–XII
(all categories, without chips and small fragments).
XVIII XVII XVI XV XIV XIII XII
Chalcedony/Opal n19
3.3
2
14.3
4
3.6
1
2.8
8
2.3
5
2.9
25
2.6
%
Flint and medium quality chert n179
30.1
8
57.1
42
37.5
6
17.1
103
29.3
34
19.8
318
34.1
%
Low-quality chert and silicious
rocks
n336
57.8
3
21.4
58
51.8
26
74.2
225
63.9
127
73.8
481
51.6
%
Indeterminate n47
8.1
1
7.1
8
7.1
2
5.7
10
2.8
6
3.5
109
11.7
%
Total n581 14 112 35 352 172 933
The general structure of artifacts in Brodar’s assemblages from layers XVIII–XII remains re-
markably consistent, regardless of whether the assemblages contain a large or small number of finds
(Table 10). The frequency of cores varies between 3% and 9%, while the presence of tools ranges
from 17% to 34%. A notable deviation is recorded only in the assemblage from layer XII—the larg-
est by far—where cores represent just 3.2% and tools only 16.9%, indicating increased activity in-
tensity (and possibly a longer occupation) at the site, as evidenced by the very high percentage of
burned artifacts in this layer (Table 11).
Table 10. General structure of Brodar’s assemblages
XVIII XVII XVI XV XIV XIII XII
Cores n15
2.6
0
0.0
4
3.6
3
8.6
15
4.3
15
8.7
30
3.2
%
Flakes n351
60.4
9
64.3
72
64.3
25
71.4
228
64.8
92
53.5
685
73.4
%
Tools n180
31.0
5
35.7
35
31.2
7
20.0
101
28.7
58
33.7
158
16.9
%
Chunks n35
6.0
0
0.0
1
0.9
0
0.0
8
2.3
7
4.0
60
6.4
%
Total n581
100.0
14
100.0
112
100.0
35
100.0
352
100
172
100.0
933
99.9
%
+ Chips and small fragments 269 2 42 29 83 29 657
48 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Table 11. Brodar’s assemblages: occurrence of patinated (1), desilified (2), scorched (3)
and totally burned (4) artifacts (without chips and small fragments).
XVIII (581) XVII (14) XVI (112) XV (35) XIV (352) XIII (172) XII (933)
1n4
0.7 00
0.0
0
0.0
6
1.7
7
4.1
15
1.6
%
2n24
4.1 17
6.2
1
2.8
16
4.5
5
2.9
16
1.7
%
3n184
31.7 317
15.2
9
25.7
74
21.0
20
11.6
274
29.4
%
4n62
10.7 04
3.6
2
5.7
15
4.3
3
1.7
127
13.6
%
Products of Flaking
In evaluating the general characteristics of the production process during the deposition period
of layers XVIII to XII, we statistically analyzed most assemblages, except for those from layers XVII
and XV, which contained too few finds.
Based on the presence of cortex, the proportion of artifacts originating from the initial phase of
core reduction was notably higher in the lower layers, showing a continuing decline until it reached
only 8–15% in layers XIV–XII (Table 12). Regarding cortex position (Table 13), flakes with preserved
lateral and laterodistal cortex dominate over those with proximal and lateroproximal (enveloppant)
cortex. Concerning the structure of the dorsal surface, the frequencies for layers XVIII–XII are re-
markably consistent (excluding layer XVI), with 32–35% parallel negative scars, 60–67% multidirec-
tional negative scars, and 3–4% cortical negative scars (Tables 14). Fully preserved Kombewa flakes
are found in very small numbers, with only one or two specimens per layer. This is surprising, given
that cores on flakes prevail among the core types. This scarcity likely reflects that these cores were
exploited to their maximum potential.
Table 12. Brodar’s assemblages: presence of cortex
on retouched and unretouched artifacts (whole specimens).
XVIII (251) XVII (9) XVI (40) XV (15) XIV (201) XIII (84) XII (333)
Total n50
19.9
2 7
17.5 221
10.4
7
8.3
49
14.7
%
Table 13. Brodar’s assemblages: cortex position on retouched
and unretouched artifacts (P = proximal, PL = proximal-lateral, L = lateral,
LD = lateral-distal, and D = distal).
XVIII (251) XVII (9) XVI (40) XV (15) XIV (201) XIII (84) XII (333)
Pn4
1.6 00
0.0 02
1.0
0
0.0
4
1.2
%
PL n2
0.8 01
2.5 1 0 0
0.0
2
0.6
%
Ln27
10.7 22
5.0 0 10 2
2.4
21
6.3
%
LD n0
0.0 02
5.0 02
1.0
2
2.4
5
1.5
%
Dn12
4.8 00
0.0 02
1.0
1
1.2
12
3.6
%
Other patterns n4
1.6 02
5.0 17
3.5
2
2.4
5
1.5
%
Brodar’s Assemblages | 49
Table 14. Brodar’s assemblages: dorsal surface of retouched and unretouched artifacts
(whole specimens): parallel scars (1); multidirectional, diagonal and
perpendicular scars (2); cortical specimens—with more than 50% of cortex
on dorsal surface (3); Kombewa flakes (4).
XVIII XVII XVI XV XIV XIII XII
1n88
35.0
3 24
60
4 67
33.3
27
32.1
115
34.5
%
2n150
59.8
5 13
32.5
8 125
62.2
56
66.7
205
61.2
%
3n11
4.4
0 2
5.0
2 8
4.0
0
0.0
9
2.7
%
4n2
0.8
1 1
2.5
1 1
0.5
1
1.2
4
1.2
%
Total n251 9 40 15 201 84 333
Regarding the striking platform (Tables 15), the structure of the assemblages is also quite
consistent but with greater variability compared to the dorsal side. In most assemblages, plain
striking platforms predominate (37–60%), while dihedral and faceted platforms are represented
by 8–15% each (with the faceted platform making up 30% in layer XVI), and cortical platforms
make up 3–7%. Faceted platforms occur more frequently in retouched artifacts, while cortical
platforms are more common in unretouched ones. The only discernible trend regarding platforms
is the discontinuous decrease in the proportion of flakes with prepared (dihedral and faceted)
platforms, dropping to less than 20% in layer XII, with a nearly identical percentage observed in
layer XVIII.
Table 15. Brodar’s assemblages: platforms on retouched and unretouched artifacts
(whole specimens and proximal fragments): plain (1), dihedral (2), faceted (3),
edge (4), punctiform (5), cortical (6), undeterminable and damaged (7, 8),
removed by retouching (9).
XVIII XVII XVI XV XIV XIII XII
1n150
53.2 620
31.2 3126
55.0
41
42.3
202
46.4
%
2n31
11.0 010
15.6 130
13.1
12
12.4
46
10.6
%
3n24
8.5 219
29.7 627
11.8
14
14.4
38
8.7
%
4n7
2.5 01
1.6 27
3.0
3
3.1
25
5.7
%
5n19
6.7 10
0.0 28
3.5
3
3.1
24
5.5
%
6n32
11.3 15
7.8 112
5.2
14
14.4
37
8.5
%
7-8 n16
5.7 08
12.5 416
7.0
6
6.2
60
13.8
%
9n2
0.7 01
1.6 13
1.3
4
4.1
3
0.7
%
Total n282 10 64 20 229 97 435
A similar trend is observed with heat-treated and burnt artifacts (Table 11), which are most
abundant in layer XII, comprising nearly 45% of the total artifacts, and are found in almost identi-
cal proportions in layer XVIII. A comparable pattern is observed in the dimensions of the arti-
facts, which are smallest in layers XVIII and XII (22 mm on average for unretouched, 24 mm for
retouched), although this is less noticeable because very small artifacts are also dominant in other
assemblages. All these indicators and the overall number of finds suggest an increased occupation
intensity during the deposition of layers XVIII and XII.
50 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Layer XVIII
Levallois cores on flakes
They represent the most numerous category of cores. These are smaller or larger flakes with a
striking surface on the ventral side (Pl. 9: 1–3). Among these cores, only one could be characterized
as preferential, two are unipolar recurrent, and one is centripetal. All are plano-convex in cross-
section and have prepared striking platforms.
Non-Levallois cores on flakes
In two cases, the ventral side is flaked from one direction along the longer axis of the flake (Pl.
9: 4, 5), while one flake is flaked from both ends (Pl. 9: 7). The knapping was performed on a flake
truncation. This category also includes classic Kombewa cores flaked from the side (Pl. 9: 6, 9) and
a core that showed traces of bipolar flaking (Pl. 9: 8).
Discoid cores
Three irregular discoid cores were found (Pl. 10: 1, 2).
Core fragments
A single core fragment was found, and it could not be classified.
Splintered and truncated-faceted pieces—which contained short negative scars and were, there-
fore, mostly classified as tools—could also serve the function of the cores. The flaking of bipolar
pieces was performed directly from the edge of the flakes, while for TFP, flaking was preceded by
the faceting of the platform. There are quite a few of these pieces (12 specimens—Pl. 10: 3–9). One
was formed on a retouched preferential Levallois flake (representing a sidescraper), where the ven-
tral side was flaked at a 90° angle from the retouched edge (Pl. 10: 3).
Among the characteristic flakes, two recurrent Levallois flakes have been identified. Both have
convergent negative scars and a faceted platform (Pl. 11: 1, 2). Additionally, there are two débordant
flakes (Pl. 11: 3), two naturally backed knives, one knife with a non-cortical back, one typical pseu-
do-Levallois flake (Pl. 10: 5), two crested flakes (Pl. 10: 4) and a blade fragment (Pl. 10: 6).
Layer XVII
No cores were identified in this layer. Among the characteristic flakes, two preferential Levallois
flakes were identified (Pl. 11: 7) —one is atypical due to the absence of a faceted platform. An over-
shot flake—more specifically, a fragment of a core—was also found, showing dorsal negative scars
resulting from removing almost regular blades (Pl. 11: 8).
Layer XVI
A preferential Levallois core on a flake fragment was found (Pl. 12: 1), along with a discoid
core featuring faceted platforms on both ends (Pl. 12: 2) and a unipolar core made of tabular chert
displaying two overlapping negative scars on its surface (Pl. 12: 3). A unipolar burin-type core was
also found in layer XVI (Pl. 12: 5), made from a raw material piece with oblique, elongated negative
scars of flake removals.
Three Levallois flakes were found among the characteristic flakes—two with centripetal nega-
tive scars and one with unidirectional convergent negative scars (Pl. 11: 9). One of the flakes, re-
moved via the centripetal method, shows that it was struck from a Levallois Kombewa core (Pl. 11:
10). One débordant flake was also recorded.
Layer XV
A recurrent Levallois core, struck from a flake using a convergent scheme, was also identified
(Pl. 12: 5). The flake shows traces of ventral thinning. Two characteristically flaked bipolar cores on
pieces of raw material were also found (Pl. 12: 6, 7). The platform of these cores forms a sharp angle
relative to the striking surface, while the core’s face contains overlapping negative scars.
Brodar’s Assemblages | 51
Layer XIV
Among the Levallois cores, the following were identified: a preferential Levallois core on a thick
flake (Pl. 13: 1), three centripetal Levallois cores (Pl. 13: 2, 3), and one unipolar recurrent Levallois
core (Pl. 13: 4). Additionally, an angular, double-sided, multi-faceted burin-like core was recorded.
One end of this core shows multiple traces of edge flaking (Pl. 14: 1), while the other end contains
shallow, surface-level traces of flaking. Furthermore, one discoid core (Pl. 14: 3) and six irregular
cores on flakes (Pl. 14: 4, 5) were recorded, one of which has bipolar-oriented scars on both ends of
the flake (Pl. 14: 2).
Truncated-faceted pieces (TFP) are represented by specimens that show traces of faceting on
both the distal end and laterally (Pl. 14: 6–11).
The layer also produced four Levallois flakes (one preferential and three centripetal), one pseudo-
Levallois flake, one Kombewa flake, one débordant flake, and one naturally backed knife (Pl. 18: 1–5).
Layer XIII
Only a single larger centripetal Levallois core was identified (Pl. 15: 1), while all other cores are
small and made on flakes (i.e., Levallois Kombewa cores), most displaying elongated negative scars.
These include: a) one recurrent Levallois Kombewa core with the striking surface on the ventral side
of an elongated flake (Pl. 15: 2) and b) six centripetal Levallois cores on flakes (Pl. 15: 3–6).
Similar to the cores mentioned above are regular Kombewa cores, which (unlike the Levallois
pieces) do not display faceted platforms. Four such pieces were recorded. Most show only a few
negative scars oriented in various directions (Pl. 15: 7), while only a single piece displays unidirec-
tional, transverse negative scars.
One unipolar core on a small cobble and a classic bipolar core on an irregular flake fragment,
with negative scars directed from both ends of the core, were also identified.
Two TFPs were found, with negative scars extending from the retouched edge at the distal end,
covering the ventral side of the flake.
Among the characteristic flakes, a recurrent Levallois flake with convergent negative scars was
found, along with a proximal fragment of a Levallois flake that cannot be more precisely classified,
three pseudo-Levallois flakes, and one Kombewa flake (Pl. 18: 6–9).
Layer XII
The layer XII of Crvena Stijena contains cores of various types:
a) Preferential Levallois cores on thick flakes (6 pieces—Pl. 16: 1–3).
b) Recurrent unipolar Levallois core (1 piece). There are no preparation marks on the strik-
ing surface of this core. At the opposite end, the core features a prepared platform from
which a single flake was struck diagonally (Pl. 16: 4).
c) Centripetal Levallois cores (7 pieces). These are sometimes difficult to distinguish from
bipolar cores, as they often display flaking from both ends.
d) Preferential Levallois Kombewa core (1 piece). It shows evidence of platform preparation
(Pl. 16: 5).
e) Non-Levallois cores on flakes (3 pieces). Negative scars show bipolar orientation (Pl. 16:
6, 7)—in one case, the dorsal side was flaked from one end, and the ventral side from the
other end of the flake. Additionally, a core that could be classified as a TFP was identi-
fied, featuring a single oblique flake removed from a prepared platform. The second TFP
shows negative scars on the distal end of the flake.
f) Unipolar cores for blades and bladelets (Pl. 17: 1–3). These include a core with conver-
gent lamellar negative scars, a small, nearly prismatic core, and a core on a flake that was
struck from a truncation (resembling a shallow burin).
g) Kombewa cores (7 pieces). They exhibit either multiple negative scars oriented in various
directions or one to two negative scars on the ventral side of the flake (Pl. 17: 4–7, 9). The
striking platform is prepared in two pieces.
h) Discoid cores (1 piece). A single, typical example was identified (Pl. 17: 8).
52 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
The following unretouched flake types were identified: two preferential Levallois flakes, four re-
current Levallois flakes, two centripetal Levallois flakes, three pseudo-Levallois flakes, three débor-
dant flakes, and several Kombewa flakes (Pl. 18: 10–15). Among the retouched tools, two preferen-
tial Levallois flakes were recorded, one atypical. A smaller number of blades were also identified,
along with flakes that showed dorsal negative scars from removing blades and bladelets (Pl. 18:
16–18).
Too ls
As in Basler’s assemblages, sidescrapers prevail in most layers of Brodar’s assemblages. Sides-
crapers generally account for 30% to 40% of the assemblages, except in layers XIII and XII, where
undifferentiated tools on flakes (denticulate and notched tools, retouched flakes) increase to 35–
45% (Table 17). Perforators, raclettes, and endscrapers are better represented in the upper layers of
the sequence (10–15%), while other tool types are poorly and unevenly represented.
Table 16. Average length of unretouched flakes
and tools on flakes (Brodar’s assemblages).
XVIII XVII XVI XV XIV XIII XII
Unretouched flakes mm 22.6 33.8 21.3 20.5 23.6 24.4 22.4
Retouched flakes mm 24.5 51.0 27.0 24.7 27.4 28.6 24.0
Table 17. Tools (Brodar’s assemblages).
XVIII XVII XVI XV XIV XIII XII
Retouched Levallois flakes n0
0.0
0 1
2.9 12
2.1
0
0.0
3
0.7
%
Mousterian points n0
0.0
0 0
0.0
0 1
1.0
0
0.0
1
0.7
%
Sidescrapers n66
41.1 111
32.3 133
35.8
10
18.2
46
30.0
%
Endscrapers n11
6.7 01
2.9 08
7.4
3
5.4
5
5.3
%
Retouched blades n7
4.3 01
2.9 01
1.0
2
3.6
3
2.0
%
Retouched flakes n27
16.6 05
17.6 010
10.5
12
21.8
34
22.7
%
Notched and denticulated
tools
n19
11.6 25
11.8 115
15.8
15
25.4
22
14.7
%
Raclettes n4
2.4 01
2.9 26
6.3
6
10.9
3
2.0
%
Splintered pieces n15
9.2 01
2.9 05
5.3
0
0.0
1
0.7
%
Truncated pieces n2
0.0 03
8.8 04
4.2
1
1.8
3
2.0
%
Burins n5
3.1 10
0.0 02
2.1
1
1.8
5
3.3
%
Perforators n5
2.4 02
5.9 15
5.3
2
3.6
13
8.7
%
Combined tools n1
0.6 00
0.0 00
0.0
0
0.0
2
2.0
%
Tool fragments n3
1.8 13
8.8 13
3.1
2
1.8
11
6.0
%
Total n163 5 34 7 95 55 150
Brodar’s Assemblages | 53
Layer XVIII
Sidescrapers
Table 18. Sidescraper types (Brodar’s assemblages).
XVIII XVII XVI XV XIV XIII XII
Lateral n47 1 6 1 21 4 28
Bilateral n5 1 4 1
Transversal n10 3 9 2 14
Latero-transversal n5 1 4 2
Bifacial n
Total n67 1 11 1 34 10 45
Quina n2
The collection from layer XVIII contains numerous and typologically diverse (but often quite
fragmented) sidescrapers (Pl. 19: 1–9; Table 18). They occur on various blanks, including Levallois
(5 pieces), debordant flakes, and knives with both natural and non-cortical backs.
– Straight lateral sidescrapers (31 pieces). Among these, one Quina sidescraper was identi-
fied. They are not standardized; most are formed on short flakes, though some appear
on blades. Some are intensively retouched, while others are retouched marginally and
partially.
– Convex lateral sidescrapers (11 pieces). These are primarily made on very small flakes
and are difficult to distinguish from straight sidescrapers, as the edge is usually slightly
curved.
– Concave lateral sidescrapers (5 pieces). One of the five shows the Quina retouch.
– Bilateral sidescrapers (5 pieces). Of the five, two are made on thick, elongated flakes, while
one pointed piece (retouched only at the upper end) is close to being classified as a point.
– Transversal sidescrapers (10 pieces). These are not standardized, with only a few typical
examples identified, including one demi-Quina sidescraper on a partially cortical flake
and one sidescraper on a preferential Levallois flake.
– Latero-transversal sidescrapers (5 pieces). These are also made on irregular flakes.
Endscrapers
All 11 pieces are atypical and made on irregular flakes and flake fragments (Pl. 19: 10–13).
Three nosed endscrapers are present.
Retouched blades
Most commonly exhibit a partial, shallow, and semi-abrupt retouch and are not standardized.
Seven pieces were found (Pl. 20: 1, 2). One could be classified as a pointed blade (Pl. 20: 3).
Retouched flakes
A significant number of flakes—at least 27 pieces, as it was sometimes difficult to distinguish
marginal retouch from damage—exhibit edges slightly modified by shallow marginal retouch. One
is a retouched pseudo-Levallois flake.
Denticulate tools
Nine pieces were identified, one of which is a slightly denticulated Levallois flake.
Notched tools
Ten pieces were found. Some notches are more pronounced, while others are less defined, ap-
pearing both along the edge and at the distal end of the flakes.
54 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Truncations
Two truncated tools were found in layer XVIII.
Raclettes
One of the four recorded pieces is a typical raclette retouched along the entire edge. Other
raclettes show a lateral, semi-abrupt/abrupt retouch.
Burins
Of the five burins found (Pl. 20: 4–6), two are transversal, two are lateral, and one is angular.
The lateral burins are formed on flake truncations.
Perforators
Four perforators were found. In one case, the point is aligned with the flake axis; in two cases,
it is positioned laterally (Pl. 20: 8); and in one case, it is located at the proximal end of the flake.
In most cases, the tip is retouched along a single edge. However, in two instances, it is bilaterally
worked with a semi-abrupt retouch.
Combined tools
A single piece features an endscraper at one end and a perforator at the other (Pl. 20: 7).
Tool fragments
Three fragments of semi-abruptly retouched tools were found but could not be more precisely
identified.
Layer XVII
Retouched Levallois flakes
A single short Levallois flake of the centripetal type was found, partially retouched with mar-
ginal retouch.
Sidescrapers
One larger, slightly convex sidescraper was identified, with a tip sharp enough to potentially
classify it as a perforator (Pl. 21: 1).
Denticulate tools
Two denticulate tools were found, both formed on blades. One was made on a large irregular
blade, while the other was made on a small blade and laterally retouched with semi-abrupt denticu-
late retouch (Pl. 21: 3), resembling a knife with an abruptly retouched back.
Burins
One burin with an oblique, semi-abruptly retouched truncation (Pl. 21: 2).
Tool fragments
A single fragment of a semi-abruptly retouched tool was found in layer XVII.
Layer XVI
Retouched Levallois flakes
One retouched centripetal Levallois flake was found.
Brodar’s Assemblages | 55
Sidescrapers
– Straight and convex lateral sidescrapers (6 pieces—Pl. 21: 4–6). Three straight lateral si-
descrapers were identified—one made on a blade and two on flakes. One of the three
convex lateral sidescrapers was formed on a cortically-backed flake.
– Bilateral sidescrapers (1 piece). A single bilateral sidescraper shows traces of ventral thinning.
– Transversal and latero-transversal sidescrapers (4 pieces). Three transversal sidescrapers
were identified, one of which is laterally pointed and formed on a débordant flake. Ad-
ditionally, one small latero-transversal sidescraper was identified.
Endscrapers
A single, slightly pointed endscraper on a thick blade was found.
Perforators
One of the two perforators is retouched laterally, while the other is retouched bilaterally (Pl. 21:
8). One is made on a Levallois flake.
Other tools
The remaining tools can be grouped. These include a raclette, two truncations, and one tool
with an abruptly retouched truncation and back (Pl. 21: 7). Two denticulate tools, one notched tool,
and six retouched flakes were also found. Additionally, three tool fragments were identified.
Layer XV
This layer yielded very few tools, including one simple straight lateral sidescraper (Pl. 21: 9),
a bilaterally retouched Levallois blade with raised retouch (Pl. 21: 10), a borer with inverse surface
retouch (Pl. 21/14), two raclettes (Pl. 21: 11, 12), a small denticulate flake (Pl. 21: 13), and a typo-
logically indeterminate tool fragment.
Layer XIV
Retouched Levallois flakes
Two retouched Levallois flakes were found—one recurrent with convergent scars (Pl. 22: 1) and
one centripetal.
Mousterian points
A single smaller Mousterian point was identified (Pl. 22: 10).
Sidescrapers
These are not typologically differentiated (Pl. 22: 4–9). Straight lateral sidescrapers (6 pieces)
are not particularly characteristic, while the convex sidescrapers (14 pieces) are relatively small in
size or are sidescraper fragments. One sinusoidal lateral sidescraper was also identified. Bilateral
sidescrapers (4 pieces) show traces of ventral thinning. Transversal sidescrapers (9 pieces) are simi-
larly not standardized.
Endscrapers
All endscrapers (7 pieces) from layer XIV are atypical (Pl. 22: 11). Some display lateral retouch,
while others are ventrally thinned. One pointed, nosed endscraper is made on a thick blade, thinned
on the backside (Pl. 22: 12).
Burins
One lateral burin on a flake was found, along with a double burin—double-faceted in retouch
on one end of the tool and multiple-faceted on the other (Pl. 23: 1, 2).
56 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Perforators
Five pieces were identified, displaying notable typological diversity (Pl. 23: 3, 4). These include
two lateral perforators, two with tips aligned along the tool’s axis and one with a curved tip. Some
pieces also exhibit additional retouching on one or both edges.
Other tools
There is no need to describe in detail other tool types from layer XIV in detail as they are
largely undifferentiated within their categories. These include raclettes (6 pieces—Pl. 23: 5, 6), den-
ticulate tools on various blanks (11 pieces—Pl. 23: 7–9), which include one made on a Levallois
flake and one on a blade, notched tools (4 pieces), and truncations on irregular flakes (4 pieces—Pl.
23: 10, 11). Additionally, there is a single blade fragment with discontinuous marginal retouch (Pl.
23: 12), one tool fragment, and retouched flakes (10 pieces)—which include one débordant flake,
one naturally backed knife, and three non-cortical backed knives.
Layer XIII
Sidescrapers
Four straight lateral sidescrapers were identified (Pl. 24: 1–4). Additionally, there are four lat-
ero-transversal sidescrapers (Pl. 24: 5, 6)—one of which is made on a pseudo-Levallois flake—and
two fragmented transversal sidescrapers.
Endscrapers
Three endscrapers were identified. One is short and ventrally thinned, another is arched and
slightly denticulate, while the third—which could be classified as nosed—is formed on an elongated
flake (Pl. 24: 7–9).
Truncations
One semi-abruptly retouched truncation was identified. There is a shallow surface retouch on
one of its edges (Pl. 24: 10).
Burins
One angular burin, made on a flake fragment, was found.
Perforators
One of the two perforators is made on a very small flake, oriented laterally, with its tip re-
touched on both edges (Pl. 24: 11). The second perforator, made on a slightly larger flake, is also
bilaterally retouched.
Other tools
Six raclettes (Pl. 24: 12), twelve retouched flakes (including retouched knives with both natural
and non-cortical backs), five notched tools (Pl. 24: 13), and nine denticulate tools (two of which are
made on pseudo-Levallois flakes) were identified. Additionally, a single tool fragment was found.
Layer XII
Retouched Levallois flakes
One retouched preferential Levallois flake was identified.
Mousterian points
One large, typical Mousterian point was found (Pl. 25: 1).
Brodar’s Assemblages | 57
Sidescrapers
These are not standardized and are better characterized as ad hoc tools (Pl. 25: 2–9). Among
them, 16 straight lateral sidescrapers were identified, one of which was made on a pointed Levallois
flake of the centripetal type, one on a pseudo-Levallois flake, and one on a ventrally thinned flake.
Eight convex lateral sidescrapers were also found, one of which is also ventrally thinned. There are
also four concave lateral sidescrapers, one fragmented bilaterally retouched sidescraper (featuring
one convex edge with direct retouch and the other straight edge with inverse retouch), as well as 14
transversal sidescrapers and two latero-transversal sidescrapers (one of which is ventrally thinned).
Endscrapers
Of the eight endscrapers identified, two are atypical, nosed endscrapers on flakes (Pl. 25: 10),
four are short endscrapers, and two are endscrapers on irregular, somewhat larger flakes (Pl. 26:
1–7).
Retouched blades
Three retouched blades were identified: the first is partially retouched by semi-abrupt marginal
retouch (Pl. 26: 8), the second is bilaterally retouched by semi-abrupt retouch (Pl. 26: 9), and the
third is partially retouched by abrupt retouch, resembling a naturally backed knife.
Raclettes
Three raclettes were found, one of which was made on a preferential Levallois flake (Pl. 26: 11, 14).
Truncations
There are three truncations: one is a tool with an oblique convex truncation formed on a Lev-
allois blade (Pl. 26: 13), one is a blade with an oblique truncation (Pl. 26: 12), and another one is a
truncated tool made on an flake, also displaying lateral semi-steep retouch (Pl. 26: 14).
Burins
Of the five burins found, one is multiple lateral (Pl. 27: 2), one is the double (with multiple
blows on one end and a single-blow on the other—Pl. 27: 1), one is a single-faceted burin, one is a
multiple-faceted burin on truncation (Pl. 27: 4), and one is a latero-transversal déjeté burin made on
a débordant flake (Pl. 27: 3).
Perforators
A total of 13 perforators were identified (Pl. 27: 5–7). Their tips are either aligned with the
flake axis or positioned laterally, and they are mostly bilaterally retouched. Most are made on very
small flakes, though some are made on bilaterally retouched elongated flakes. One of the bilaterally
retouched perforators displays ventral thinning.
Combined tools
The following combined tools were recorded: a perforator combined with an endscraper (i.e.,
the endscraper is partially retouched and has a straight edge) and a curved endscraper combined
with a multiple transversal burin—which may represent a secondarily modified core (Pl. 27: 8).
Other tools
In layer XII, 12 typologically undifferentiated notched tools (Pl. 27: 13), ten denticulate tools
(Pl. 27: 9–12), 34 retouched flakes, and nine tool fragments were recorded.
| 59
Middle Paleolithic sequence
of Crvena Stijena
As previously mentioned, there have been several attempts to identify and categorize the Mid-
dle Paleolithic assemblages from Crvena Stijena and to examine their development within a broader
context (Brodar, 1962; Basler, 1975; Ivanova, 1979; Kozłowski, 1992, 2002; Dogandžić and Đuričić,
2017; Mihailović, 2017; Mihailović and Whallon, 2017). However, all these efforts have been hin-
dered by the fact that artifacts from this site have yet to be fully and comprehensively analyzed. Now
that these analyses have been conducted, the initial observations on the dynamics of technological
changes at Crvena Stijena still appear largely unchanged. There still appears to be a strong continui-
ty in flaking technology and techno-economic behavior at Crvena Stijena, reflected in the consistent
use of the same (low-quality and likely local) raw materials, similar flaking methods, and a relatively
uniform set of tools throughout the entire period of occupation of this rockshelter.
The analyses, however, have also uncovered substantial variation beneath the apparent continu-
ity in knapping technology and techno-economic behavior. This variation is likely better explained
by the duration and nature of occupation—shaped indirectly by climatic, ecological, and demo-
graphic factors—rather than by cultural or social influences. To address this question, we will exam-
ine technological indicators that reveal patterns in occupation intensity, variations in the reduction
sequence, and diversity within tool assemblages.
Occupation intensity
As noted in the introductory chapters, most layers of Crvena Stijena currently lack sufficient
data to determine the duration and nature of occupation or the intensity of activities in specific
phases, particularly with regard to models based on calculated dates and sediment volumes of the
examined layers (French and Collins, 2015; Clark, 2015; Clark and Barthon, 2017). Beyond these
indicators, however, various technological parameters may be relevant for evaluating the duration
and nature of occupation at the site, even when examining artifact collections from earlier excava-
tions. Some of these factors were evaluated by Dogandžić and Đuričić (2017) in their analysis of
lithic artifacts from the upper layers of Crvena Stijena and Bioče. Although their research primarily
focused on comparing the two sites, some of their results are also relevant for understanding shifts
in the techno-economic behavior of the Neanderthal groups that inhabited Crvena Stijena.
Quantity of finds. The quantity of finds in Basler’s collection does not serve as a reliable indica-
tor of the duration or nature of the occupation, particularly in the lower layers of Crvena Stijena.
However, the somewhat higher number of finds in layers XXXI, XXX, XXIV, and XXII, compared to
other layers, suggests that the rockshelter was visited more frequently during these phases—although
this is difficult to assess given that the thickness of layers XXXI and XXIV exceeds 2 m. On the other
hand, the exceptionally large number of finds (2,400 artifacts collected) from layers XVIII and XII in
Brodar’s collection undoubtedly demonstrates that the rockshelter was intensively occupied during
these phases. It remains to be determined whether the overall small number of finds in Basler’s collec-
60 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
tion is due to selective collecting or the spatial distribution of the artifacts. Recent research at Crvena
Stijena has shown that the peripheral areas of the occupied space contain significantly fewer finds
compared to the surface near the inner wall of the rockshelter (Mihailović et al., 2017b).
Occurrence of burned artifacts. In Basler’s collection, data on artifact burning are less reliable
due to the selective nature of the assemblages, but they still offer some insights into the ancient
hominin activities. In the lower layers, the highest proportions of charred and burned artifacts (over
10%) were recorded in assemblages from layers XXXI, XXX, XXV, XXIV, and XXI. In the upper
layers, the highest proportions of charred and burned artifacts (over 40%) were recorded in layers
XVIII and XII.
Occurrence of cores and tools. Evidence of in loco artifact production (flake to core ratio) is
considered one of the main indicators of prolonged occupation in habitations (Kuhn, 1991), while,
conversely, a high proportion of tools (flake to tool ratio) indicates short-term settlement (but see
Figure 25. Hearth in layer M2 during the 2012 excavations. (Photo: G. Ćulafić)
Figure 24. Excavation area in the eastern part of the rockshelter (layer M2/
XII) during the 2012 excavations. (Photo: D. Mihailović)
Middle Paleolithic sequence of Crvena Stijena | 61
Figure 26. Burned artifacts from layers XVIII–XII (Photo: D. Pajović)
Dibble and Rolland, 1992). In this regard, it is worth noting that cores are present in nearly all of
Basler’s assemblages, indicating that settlements were not transitory—at least to the extent that they
involved artifact production. Regarding Brodar’s assemblages, cores are absent only in layer XVII,
with the highest percentages recorded in layers XIII and XV. While the proportion of retouched
tools among total finds cannot be determined for the lower layers, they account for 20% to 30% in
the upper layers, decreasing to only 17% in layer XII.
Degree of core and tool reduction. High intensity of core processing—accessible through the pro-
portional representation of cortex and core trimming elements—may reflect the intensity of activity
and/or the duration of occupation. Similar conclusions can be drawn from the degree of reduction
seen on the working edges of tools—although it should be noted that tools could have been pro-
duced at different locations.
Several ways exist to determine the level of core reduction, including calculating the percent-
age of removed or remaining volume relative to the original blank volume and quantifying cortex
proportions to estimate nodule size (Lombao et al., 2020). Cores were not analyzed from this per-
spective in the present study, partly because Dogandžić and Đuričić (2017) have already attempted
this—albeit on a different sample—and partly because we consider the Crvena Stijena sample and
its raw materials to be somewhat unsuitable for this type of analysis. At Crvena Stijena, raw materi-
als from primary deposits—unsuitable for core reduction—were predominantly used, and most of
the cores were entirely exhausted, making it difficult to identify key attributes necessary for applying
many of the methods mentioned. For these reasons, we limited our analysis to basic indicators of
activity intensity. Given the circumstances (i.e., poor-quality raw materials, incomplete collections,
and the limited area from which the assemblages originate), we believe these indicators provide
more reliable insights into settlement intensity.
More generally, the prevalence of flakes with cortex traces and those with cortical platforms
could indicate the degree of core exploitation. Regarding the lower layers of Crvena Stijena (i.e.,
Basler’s assemblages), a slightly higher prevalence of flakes with cortex traces and cortical plat-
forms was recorded only in layer XXII. In the upper layers (i.e., Brodar’s assemblages), very few
flakes with traces of cortex were found. The trend of decreasing prevalence of cortical flakes
peaks in the layer XII assemblage, where only six whole cortical flakes were found among over a
thousand artifacts, and the total proportion of flakes with cortex traces is under 15%. A similar
pattern is observed with striking platforms, as layer XII also shows the lowest proportion of corti-
cal platforms (8.5%).
Regarding tools, Crvena Stijena documents several instances of old tool recycling, but the exact
number of recycled pieces cannot be reliably determined due to the poorly developed patina. Arti-
facts with traces of secondary modification—aimed at extending the tools’ lifespan—are more nu-
merous. Dogandžić and Đuričić (2017) measured the intensity of tool production and determined
the geometric index of reduction for sidescrapers. While their results are valuable for understanding
62 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
the characteristics of individual assemblages in the upper part of the stratigraphic sequence, they are
less relevant for examining long-term technological trends at Crvena Stijena.
Artifact thinning was documented in nearly all layers of Crvena Stijena, with the highest num-
ber of thinned specimens found in layer XXIV, where sidescrapers thinned on various sides make
up over 10% (i.e., 6 out of 41). Thinning was primarily applied to endscrapers and sidescrapers, in-
dicating that this technique had a practical function, most likely related to hafting the tools. In some
cases, thinning could not be easily distinguished from the secondary utilization of tools as cores.
Cores on tools are documented in both Basler’s and Brodar’s collections from layer XVIII.
Dimensions and fragmentation of artifacts. Both parameters may be linked to activity intensity,
but the causes of stone artifact microlithization during the late Middle Paleolithic are not fully un-
derstood. On the other hand, artifact fragmentation may have been influenced by factors other than
activity intensity, such as raw material quality, sediment composition, and post-depositional pro-
cesses. Nevertheless, we believe that the small dimensions of unretouched flakes in the upper layers
of the sequence (only 22.4 mm in layer XII) indicate intense core exploitation and, by extension,
suggest a somewhat longer occupation of Crvena Stijena.
In Basler’s (selectively collected) assemblages, as many as 79% of artifacts are completely pre-
served. In Brodar’s assemblages, however, the representation of complete artifacts is much lower:
in layers XVIII, XVII, XIV, and XIII, it ranges between 19% and 23%; in layers XVI and XV, it
amounts to 11%; and in layer XII, only 8.6%—indicating an increased intensity of activity in the
rockshelter.
Systematization of data on occupation intensity
across settlement phases
In this section, we will aim to systematize diverse data concerning the duration and nature of
Neanderthal occupation at Crvena Stijena (Table 19).
Table 19. Indicators of occupation intensity and core reduction:
black = highly pronounced, gray = moderately pronounced, white = weakly pronounced.
The flake/core ratio includes both retouched and unretouched flakes,
whereas the flake/tool ratio reflects the proportion of unretouched
to retouched artifacts.
Quantity
of finds
Burned and
charred pieces
Flake/core
ratio
Flake/tool
ratio
Fragmented
specimens
Brodar’s assemblages
XII 933 43% 11.6 4.6 91.4%
XIII 172 13% 7.4 1.7 76.8%
XIV 352 25% 15.5 2.4 79.8%
XV 35 31% 5 3.6 88.6%
XVI 112 19% 12.7 2.1 89.3%
XVII 14 0% 0 0.6 78.6%
XVIII 581 42% 17.8 2.1 80.9%
Basler’s assemblages
XX 62 0%
XXI 10 9%
XXII 118 2%
XXIII 5 0%
XXIV 68 4%
XXV 11 0%
XXVI 1 0%
XXVII 4 0%
XXVIII 3 0%
XXIX 6 0%
XXX 28 10%
XXXI 75 12%
Middle Paleolithic sequence of Crvena Stijena | 63
a) The earliest documented phase of occupation, associated with layers XXXI and XXX,
provides clear evidence of settlement, which might be more extensive considering the
limited excavated area. Communities visiting Crvena Stijena used fire, as indicated by a
moderate percentage of burned artifacts, suggesting a temporary rather than transitory
occupation. The relatively diverse assemblage of cores and tools further implies a longer
stay and a wide range of activities conducted at the site. Additionally, the thinning of
sidescrapers—likely intended for hafting—is already documented during this phase, un-
derscoring the intensity and duration of these occupations.
b) In layers XXVII–XXV, the occupation was ephemeral. Only a few cores and tools were
found (in some layers, none at all), and there are almost no burned artifacts.
c) In layer XXIV, a slightly larger number of artifacts were collected and preserved, with a
small increase in the percentage of burned artifacts compared to the other layers in the
lower complex. Various cores and tools are present, with relatively few cortical pieces.
The tool sizes are small, and sidescrapers often display thinning. These indicators suggest
a heightened intensity of activity within the habitation. However, these indicators alone
likely underestimate the intensity of hominin activity during this phase, as the substantial
deposits of ash and charcoal in this layer suggest a significantly higher activity level. As
a result, the nature of occupation during this phase at Crvena Stijena remains an unre-
solved issue.
d) As indicated by the small number of finds and the lack of burned artifacts, the occupa-
tion at layer XXIII was likely very brief, possibly related to a specialized activity involving
the use of perforators.
e) The nature of occupation in layer XXII differs from that of other layers. While the layer
contains a relatively large number of finds, the percentage of burned pieces is low. Many
cortical flakes were recovered, and Quina sidescrapers, associated with increased com-
munity mobility (Hiscock et al., 2009—but see differing opinions in Rolland and Dibble,
1990), were also identified. Furthermore, thinned pieces and tools with retouched plat-
forms were also found. All this evidence suggests frequent, yet short-term, occupations
during this phase.
f) Layer XXI contains no cores and features few tools, similar to the assemblage from layer
XXIII. However, unlike layer XXIII, sidescrapers are the predominant type. Additionally,
few burned artifacts were found. Based on this evidence, the occupation at the Crvena
Stijena rockshelter was short-term and likely transitory.
g) Layer XX contains a relatively large number of artifacts (similar to layer XXII, though
slightly fewer), including a relatively low percentage of burned pieces. However, the rep-
resentation of cortical artifacts (with cortex found on the dorsal side and platform) is no-
tably smaller. On the other hand, the proportion of high-quality raw materials is higher,
indicating increased mobility.
i) Layer XVIII contains many artifacts (581, excluding chips), making it the second richest
layer. It also features a high percentage of burned pieces (over 40%), a wide variety of
cores and tools, the presence of modified tools, and a microlithic character—all of which
suggest a somewhat longer occupation. Additionally, this layer exhibits a slightly higher
percentage of artifacts with traces of cortex.
j) The assemblages from layers XVII and XV contain few finds, although burned artifacts
are present. Due to the small sample size, it is difficult to assess the character of the oc-
cupation in these layers. In contrast, layer XVI contains a slightly larger number of arti-
facts (112, excluding chips), but the occupation evidenced by this layer appears to have
been short-term, as indicated by the lower percentage of burned artifacts (19%) and the
slightly higher proportion of artifacts with traces of cortex. According to Dogandžić and
Đuričić (2017), the Levallois and laminar components are more prominent in the as-
64 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
semblage from layer XVI—although this remains difficult to ascertain due to the small
sample size.
k) The slightly higher percentage of burned artifacts (25%) and the low representation of
artifacts with traces of cortex (10,4%) in layer XIV suggest that the hominin groups may
have stayed relatively longer during this phase.
l) Layer XIII contains half the number of finds compared to layer XIV, with a very low
percentage of burned artifacts. This suggests a shorter stay at the site despite the low
representation of flakes with traces of cortex and the fact that this layer has the highest
percentage of flakes with platforms removed by retouch (4.1%).
m) Unlike layer XIII, the occupation indicators for layer XII are much clearer. The slightly
lower percentage of cores and tools, the highest number of flakes, a very large number of
chips, the low representation of cortical artifacts, and the high degree of microlithization
all point to intensive workshop activity. The occupation intensity is further highlighted
by the fact that layer XII contains the highest percentage of burned artifacts (43%).
General conclusions on the intensity and nature of occupation
While acknowledging that the analyses conducted do not cover the full range of possible ap-
proaches to studying the material from Crvena Stijena, we present preliminary conclusions about
the settlement of the rockshelter during different phases of the Middle Paleolithic. Overall, phases
of long-term or frequent occupation (XXXI–XXX, XXIV, XXII, XVIII, XIV, and XII) alternate with
phases when the rockshelter was rarely visited. Although archaeozoological material is lacking, it
can be assumed that Crvena Stijena functioned as a base camp during some phases (XXIV, XIV,
and XII) and as a temporary hunting camp during others. There is little evidence to suggest that
the settlement of Crvena Stijena was specialized in nature, with the possible exception of layer XX-
III. The exact nature of occupation evidenced in layer XXIV is yet to be determined. The remains
from layer XXII are particularly interesting, as they suggest frequent, albeit brief, occupations of
the rockshelter.
The evidence on occupation intensity broadly aligns with Morley’s anthropogenic indices (Mor-
ley, 2007; 2017), as recorded in layers with higher charcoal content (XXIV, XXII, XX, and XVIII).
The exception is layer XII, where charcoal is less abundant (Morley, 2017). Nevertheless, technologi-
cal indicators from layer XII suggest frequent, and potentially prolonged occupation at the site. This
discrepancy might not be surprising if the eastern profile of Basler’s pit, where Morley collected
samples, was located on the periphery of the occupied area during the deposition of layer XII. Basler
himself concluded that the final phases of Crvena Stijena’s occupation were merely episodic (Basler,
1975). However, his interpretation relied solely on material from his own excavations, disregarding
the volume of finds from assemblages collected by Brodar.
The stratigraphic sequence, characterized by alternating layers with high charcoal content and
breccia, along with the settlement patterns at Crvena Stijena, suggests that climatic factors signifi-
cantly shaped the duration and nature of occupation at the site. Periods of intensive occupation
largely align with MIS 5 and MIS 3, the timing of which has been established through TL, OSL, and
ESR dating (Mercier et al., 2017) and, to some extent, through faunal evidence (Morin and Soulier,
2017). This refers primarily to layer XXIV, dated to MIS 5a, and to layers XIV–XII, attributed to the
early phase of MIS 3. It is important to note that evidence of occupation is also present in layers
deposited during colder intervals, particularly in layer XXII, which—based on dating and geoar-
chaeological analyses—corresponds to MIS 4. Our analyses, however, suggest that Crvena Stijena
was likely used differently during this phase, with frequent, short-term occupations, and held a dis-
tinct role within the regional settlement system compared to warmer periods. A similar conclusion
could be drawn for layers XXXI and XXX, though it remains uncertain whether these layers (at least
partially) correspond to MIS 5 or if warm-adapted fauna persisted in the Crvena Stijena area even
during glacial periods.
Middle Paleolithic sequence of Crvena Stijena | 65
Reduction sequence
From the layer-by-layer review of the lithic material we provided, it is clear that artifacts flaked
using common Middle Paleolithic methods (such as Levallois, discoid, etc.) are present throughout
the sequence at Crvena Stijena (Tables 20–23). However, it was impossible to trace all stages of their
application—preparation, exploitation, and rejuvenation—in detail, largely due to the use of low-
quality raw materials for knapping. This might also explain the relatively small number of typical
débordant flakes recorded in the assemblages. Arguably more significant than the presence of Leval-
lois and discoid technologies, however, is Crvena Stijena’s potential to reveal all phases of the evo-
lution of the core-on-flake method. Recently recognized as one of the key models of technological
behavior in the Middle Paleolithic of the Mediterranean (Bourguignon et al., 2004; Romagnoli et al.,
2022), the origins and variability of the core-on-flake method remain poorly understood.
Table 20. Core types (Basler’s assemblages).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX XVIII XVII XVI XV XIV XIII XII
Preferential
Levallois 1 1 1 2
Unipolar recurrent
Levallois 1 1
Centripetal
recurrent Levallois 4 2 1 1 1 1 1 1
Preferential
Levallois on flakes
Recurrent Levallois
on flakes 1
Centripetal
Levallois on flakes 1 3 1
Non-Levallois
unipolar on flakes
Non-Levallois
bipolar on flakes 1 1
Non-Levallois
centripetal on flakes 1
Kombewa 1 1 1
Discoid 5 2
Unipolar volumetric 1
Burin-like 1
Globular 1
Core fragments 1
Total 11 3 2 0 0 1 1 2 0 3 0 2 4 4 2 2 1 2 2
Table 21. Core types (Brodar’s assemblages).
XVIII XVII XVI XV XIV XIII XII
Preferential Levallois
Unipolar recurrent Levallois 1 1 1
Centripetal recurrent Levallois 3 1
Levallois undeterminable
Preferential Levallois on flakes 1 1 1 7
Recurrent Levallois on flakes 2 1 1
Centripetal Levallois on flakes 1 3 7
Non-Levallois unipolar on flakes 2 1
Non-Levallois bipolar on flakes 1 2 1
Non-Levallois centripetal on flakes 5 5 3
Kombewa 2 7
Discoid 3 1 1 1
Unipolar and bipolar volumetric 1 1
Burin-like 1 1 2
Globular
Core fragments 1
Total 14 0 4 3 13 11 29
66 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Table 22. Blank types (Basler’s assemblages).
XXXI XXX XXIX XXVIII XXVII XXVI XXV XXIV XXIII XXII XXI XX XVIII XVII XVI XV XIV XIII XII
Preferential
Levallois 3 1 2 2 1 1 2
Unipolar recurrent
Levallois 3 1
Centripetal
recurrent Levallois 5 2 2 1
Levallois
undeterminable 1
Pseudo-Levallois 3 4 1
Kombewa 1
Table 23. Blank types (Brodar’s assemblages).
XVIII XVII XVI XV XIV XIII XII
Preferential Levallois 2 1 2
Unipolar recurrent Levallois 2 1 1 4
Centripetal Recurrent Levallois 2 3 2
Levallois undeterminable 1
Pseudo-Levallois 1 1 3 3
Kombewa 1 1 6
Levallois method
The presence of Levallois products has been documented in almost all layers at Crvena Stijena,
starting from layer XXXI. Different types of Levallois products appear in a stable proportion across
most assemblages. In the stratigraphically lowest layers of the site, only cores and flakes associated
with the use of preferential and centripetal (recurrent) Levallois methods are present— methods
that dominate throughout the entire stratigraphic sequence. Unipolar recurrent products appear
only from layer XXII onwards, but if we include Basler’s and Brodar’s assemblages, they are present
in all layers up to layer XII.
Regarding the Levallois cores, those made on flakes—referred to as Levallois Kombewa—stand
out, first appearing in layer XXIV and subsequently in all layers of the upper part of the strati-
graphic column (XVII–XII). As with Levallois cores on raw material pieces, both preferential and
unipolar, as well as centripetal cores, are present among the Levallois Kombewa cores.
Discoid method
Discoid cores are rare, found only in layers XXXI, XXII, XVIII, XVI, XIV, and XII. A similar
pattern is observed with pseudo-Levallois flakes, which could have been knapped using the discoid
method; a few of these flakes have been identified in layers XXII, XX, and XVIII–XII. In their study,
Dogandžić and Đuričić (2017) highlighted the significant presence of “discoidal flakes” in the up-
per layers of Crvena Stijena, but they did not specify which types of flakes were included in this
category or how many they distinguished.
Quina method
No cores directly associated with the Quina method were found, but a significant number
of “Quinoid” flakes with asymmetrical longitudinal and transverse sections—including cortically
backed flakes—were identified, which could potentially result from its application. These products
were particularly abundant in Basler’s assemblage from layer XXII (although it should be noted that
Brodar’s collection was not analyzed from this perspective). In the structure of Brodar’s assemblages,
“Quinoid” flakes are represented by 22.2% in layer XVIII, 21.4% in layer XVII, 7.5% in layer XVI,
9.4% in layer XV, 20.8% in layer XIV, 21.3% in layer XIII, and 11.4% in layer XII. While the high
percentage of these flakes in Brodar’s assemblages—particularly in layer XVIII—could be linked to
the Quina component, the assemblages from the upper layers cannot. The presence of non-cortical
Middle Paleolithic sequence of Crvena Stijena | 67
backs on flakes from these layers is more likely the result of small cores being knapped, where re-
moving the edge during the process was unavoidable. Given the small size of the cores, the number
of these flakes (as well as pseudo-Levallois flakes) in layer XII would likely have been higher if the
morphometric threshold between chips and flakes had not been set at 15 mm, which is already be-
low the usual standard.
Core-on-flake method (COF)
Cores-on-flakes are present in layers XXII, XVIII, and XIV–XII, appearing in two forms: as
ordinary Kombewa cores (knapped without a clear system) and as the non-Levallois (unipolar, bi-
polar, and centripetal) cores without a faceted platform. During the deposition of layers XXII and
XVIII, knapping was performed on truncations, mostly perpendicular to the axis of the flake (two
tools were also knapped), while systematically knapped Levallois and non-Levallois Kombewa cores
dominate the upper layers of the sequence (see material overview by layers).
Laminar/lamellar knapping
Core knapping aimed at producing blades and bladelets has been documented at Crvena Stijena
in layers XXIX and XVII–XII. However, a systematic approach can only be seen in burin-like cores,
which feature retouching along the flake’s edge. Classic unipolar and bipolar cores, lacking evidence
of preparation, have been identified only rarely—specifically in layer XII. Additionally, the blade
sample is too small to confirm systematic production, particularly given the absence of bladelets.
In the upper layers of the sequence—including those excavated recently—blades resembling those
typically knapped from volumetric cores characteristic of the Upper Paleolithic have been identi-
fied, although no such cores have been found. Their faceted platforms suggest they may also have
been produced from Levallois cores.
General conclusions about flaking technology
The Levallois method dominates the Middle Paleolithic layers of Crvena Stijena, appearing in
nearly all assemblages. It is particularly prominent in the lower layers of the stratigraphic sequence,
where Levallois cores are common, and few other core types have been found. Centripetally and
preferentially knapped cores dominate across all layers, while cores produced using the unipolar
recurrent system come from the upper layers of the sequence. The few identified Levallois flakes
generally match the core types, although this is not always true in individual assemblages. The low
number of Levallois products may be due to the high degree of fragmentation, suggesting that more
may have existed. In contrast, the limited number of débordant flakes is likely due to the small size
of the cores, which prevented frequent rejuvenation during exploitation.
Interestingly, discoid technology was not found to be as prevalent as previous analyses had sug-
gested. This implies that the communities inhabiting Crvena Stijena, despite their tendency for ad
hoc solutions, still had a clear understanding of the technological processes applicable in the later
stages of core exploitation. However, it is important to note that the succession of different methods
in the early and late stages of core reduction has not been definitively established (Dogandžić and
Đuričić, 2017). The Quina method cannot be reliably identified, although some typical products—
such as flakes with cortical and non-cortical backs or thick, wide platforms (talon-dos) serving as
backs—were recorded in layers XXII–XX.
The core-on-flake (COF) concept initially appears as a classic Kombewa method but later un-
dergoes diversification. In their analysis, Doganđić and Đuričić (2017) specifically emphasized the
centripetal method, attributing a substantial number of cores and flakes to its use. Simultaneously,
they identified a relatively small number of COFs, suggesting that the centripetally knapped COFs
documented in this study were instead classified as “centripetal” cores. In the later phases of set-
tlement at Crvena Stijena, there is evidence of integration of COF with other knapping methods
(particularly the Levallois method). This is reflected in the fact that one-third to one-half of the
total number of cores were actually formed on flakes. The proportion of flakes becomes even higher
when accounting for the production of bladelet cores on them. Some truncated faceted pieces could
also be categorized as cores on flakes.
68 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Although a discernible pattern exists, the systematic production of blades and bladelets has yet
to be demonstrated for Crvena Stijena. This issue remains unresolved due to the insufficient and
incomplete sample available for analysis.
Variability of tools
The general structure of the tool assemblages and the changes in their composition have al-
ready been addressed in the previous text and other publications (Mihailović and Whallon, 2017;
Mihailović et al., 2017), so we will not revisit them here. Instead, we will primarily focus on the
variations within the main tool categories. These variations are important not only for determin-
ing the position of the Crvena Stijena Mousterian assemblages within the broader regional context
but also for distinguishing standardized tools from ad hoc implements, which were typologically
classified based solely on formal classification criteria. In analyzing specific tool types, we also con-
sidered other factors—such as the type of substrate, morphometric characteristics, and retouch at-
tributes—without delving into the economy of tool production and reuse, as examined by Doganđić
and Đuričić (2017).
Mousterian points
The few Mousterian points from Crvena Stijena are found in Basler’s assemblages from layers
XXXI, XXIV, XXII, XX, and XII and in Brodar’s assemblages from layers XIV and XII. Due to frag-
mentation, this type is mostly represented by fragments with preserved tips. However, Mousterian
points lacked standardization regarding support, dimensions, and the retouch location and depth.
Typical, bilaterally retouched examples were found in Basler’s assemblage from layer XXII (one of
which shows ventral thinning) and in Brodar’s assemblages from layers XIV and XII. In some cases,
it was difficult to distinguish these points from perforators.
Sidescrapers
Lateral sidescrapers dominate over other types of sidescrapers in nearly all assemblages of Cr-
vena Stijena (Tables 8 and 18). Transversal and latero-transversal pieces are about half as common
and, in some cases, even less frequent than lateral sidescrapers. Most Quina and demi-Quina side-
scrapers are found in the lower part of the stratigraphic sequence (up to layer XVIII), particularly
in layers XXII and XX. However, their presence in the upper layers is also confirmed by Basler’s
assemblages from layers XIV–XII. Sidescrapers from the lower layers are more elaborated and fre-
quently made on standardized blanks, including débitage flakes. In contrast, those from the upper
layers—particularly from layers XIV–XII—are better characterized as ad hoc tools, often made on
fragmented flakes. Thinning was consistently performed and is documented throughout the se-
quence, as previously discussed.
Backed knives and retouched blades
Backed knives and retouched blades appear sporadically and in small numbers throughout the
sequence. A few typical examples have been identified among the knives and points with abruptly
retouched backs (in layers XXII, XVIII, and XVII). The group of backed pieces also includes a “lu-
nate” from layer XV, characterized by an edge that is not continuously retouched; in contrast, the
microlithic lunate discovered in layer M2 (equivalent to layer XII) during more recent excavations
displays a continuous retouch. Unlike abruptly retouched backed knives, retouched blades are not
standardized in shape and size and are often only partially retouched.
Truncations
Several variants of truncated tools occur, including forms resembling raclettes, characterized by
retouching on both the truncation and the back. They share certain similarities with truncated-fac-
eted pieces (TFP), but TFPs are distinct in having truncations that function as platforms for flaking.
Middle Paleolithic sequence of Crvena Stijena | 69
Perforators
Crvena Stijena perforators also form a heterogeneous tool category, with some specimens close
to being classified as Mousterian points or pointed sidescrapers. Notable examples include perfora-
tors from layer XXXI and those from layers XXVII, XXIV, XX, and XVIII, which were made on
elongated blanks.
Other tools
Other types of tools are less relevant for assessing typological variability. Burins and endscrap-
ers are generally atypical. Some exhibit carinated or nosed shapes, including those made on the
pointed tips of convergent and déjeté sidescrapers. Typical raclettes—retouched laterally, bilaterally,
or along the entire perimeter with abrupt and semi-abrupt retouch—occur in the upper layers of
the Crvena Stijena sequence. Denticulate and notched artifacts, as well as simple retouched flakes,
clearly represent ad hoc tools.
General conclusions on tool variability
When considering all the data on the variability of assemblages and tool types, it becomes
clear that the differences between them are relatively minor. However, certain assemblages stand
out: those from the lowest layers of the Crvena Stijena sequence (XXXI–XXX), which, alongside
sidescrapers, also include geometrized perforators; finds from layer XXIV, which feature extremely
thinned sidescrapers; and finds from layers XXII and XXI, which contain transversal sidescrapers.
Finds from layers XX–XVIII exhibit a transitional character, featuring elements typical of both ear-
lier phases (transversal and Quina sidescrapers) and later phases (ad hoc tools, microlithic dimen-
sions, and laminar component). The upper layers of Crvena Stijena are dominated by ad hoc tools
and display more Upper Paleolithic elements. Nevertheless, it remains evident that the lithic assem-
blages from these layers exhibit a distinctly Middle Paleolithic character.
| 71
The Middle Paleolithic
of Crvena Stijena in
a regional context
Figure 27. Middle Paleolithic sites in Southeastern Europe: Divje Babe (DB), Erd (ERD), Krapina (KRA), Vet-
ernica (VET), Vindija (VIN), Mujina Cave (MUJ), Rastuša (RAS), Crvena Stijena (CS), Bioče (BIO), Mališina
Stijena (MLS), Petrovaradin Fortress (PF), Risovača (RIS), Šalitrena Pećina (ŠP), Hadži Prodanova Cave
(HPC), Smolućka Cave (SMO), Cioarei-Borosteni Cave (BOR), Kozja and Mala Cave (KC/MC), Velika and
Mala Balanica (VB/MB), Pešturina (PEŠ), Samuilitca II and Temnata Dupka (SAM/TD), Muselievo (MUS),
Bacho Kiro (BK), Golema Pesht (GP), Asprochaliko (ASP), Theopetra (TEO), Kalamakia (KAL), Lakonis (L),
Klissoura (KLI), Stelida (STE). Sites with Quina assemblages are marked in red, Mousterian sites in black,
sites with leaf points in green, and sites attributed to the Micromousterian in yellow.
72 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
The significance of Crvena Stijena extends beyond the local context, playing a crucial role in
addressing broader regional questions. Among these, the spatial and temporal variability of Mid-
dle Paleolithic industries emerges as particularly important. In terms of spatial variability across
Southeastern Europe, previous research has revealed the following: a) industries characterized by
a pronounced Levallois component are concentrated in the western, eastern, and southern Balkans
(Dogandžić, 2023; Karavanić and Banda, 2023; Darlas, 2007); b) industries featuring bifacially re-
touched tools are found not only in the southern Pannonian Basin but also in Greece (Galanidou
et al., 2016); and c) industries with a distinct Quina/quartzite component occur in the southern
Pannonian Basin, the southwestern Carpathians, and the Carpatho-Balkan massif in eastern Serbia
(Cârciumaru et al., 2000; Mihailović, 2009b; Mihailović et al., 2022b; Karavanić and Banda, 2023).
The geographic distribution of the lithic industries largely coincides with the distribution of
various types of mineral raw materials. Sites with Levallois and bifacial components are clustered
in regions with abundant chert deposits in the Dinarides (Anđelković, 1978), the Balkan and Rho-
dope mountains (Gurova et al., 2022), and Pindus massifs (Biagi et al., 2023), while sites with a
pronounced Quina component are found in areas where chert and other high-quality raw materials
seem to have been less accessible, but where quartz/quartzite was abundant. In areas with extensive
chert formations, industries with a Levallois component appear both in the early and late stages of
the Middle Paleolithic (including Crvena Stijena); the same may be true for quartz/quartzite indus-
tries, provided that some of them are attributed to MIS 3 (which has yet to be demonstrated). In
both cases, however, there are indications that technological variability did not solely depend on the
availability of mineral resources but was also influenced by cultural and demographic factors dur-
ing particular climatic phases. In the following chapters, we will explore this issue in greater detail,
along with the topics of artifact microlithization and the emergence of Upper Paleolithic elements in
the final stages of the Middle Paleolithic.
Crvena Stijena and the Beginnings of the Levallois Method in the
Paleolithic of Southeastern Europe
The beginning of the Middle Paleolithic in Europe was long thought to be primarily character-
ized by the emergence of Levallois technology. According to some interpretations (Moncel et al.,
2020), the Levallois method appeared in Western Europe as early as MIS 10/9, becoming dominant
across most sites by MIS 7. In Eastern Europe, the method seems to have been adopted as early as
MIS 7–6, if not earlier, as evidenced by findings from the Korolevo site in Ukraine (Stepanchuk and
Sapozhnikov, 2010), Mamaia Sat, Mitoc Valea Izvorului, Ripiceni-Izvor, and Abri 122/1200 in Ro-
mania (Balescu et al., 2015; Schmidt et al., 2024), Kozarnika Cave in Bulgaria (Heydari et al., 2022),
and Velika Balanica and Selačka Cave 3 in Serbia (Mihailović et al., 2022a, 2024). The emergence of
the Levallois method in Greece occurred somewhat later, around the end of MIS 6 and the begin-
ning of MIS 5 (Tourloukis and Harvati, 2018). A comparable age has been proposed for sites in the
southern Italian Peninsula (e.g., Poggio, Molare, Cala Blanca), whereas sites in the northern Italian
Peninsula (including San Bernardino) have been attributed to MIS 8/7 (Picin et al., 2013; Carmig-
nani et al., 2021).
Research conducted at Velika and Mala Balanica in Serbia indicated that the onset of the Mid-
dle Paleolithic in Southeastern Europe should not be associated with the appearance of the Levallois
method but rather with the Quina method. Early Quina assemblages—lacking Levallois products—
have been identified in Layer 2 of Mala Balanica and Layer 3 of Velika Balanica, dated to around
290–240 ka (i.e., MIS 9–7; Mihailović et al., 2022a). In contrast, sites dated to MIS 6 contain exclu-
sively Levallois artifacts, while sites from MIS 5 predominantly feature assemblages with both Leval-
lois and Quina products (Mihailović, 2014).
The emergence of the Levallois method in the lowest layers of Crvena Stijena (MIS 6/5) appears
to chronologically align with the basal Mousterian of Asprochaliko (Epirus, Greece), as well as the
lowest layers of other Greek sites—Kalamakia, Theopetra, Klissoura Cave, and potentially Lakonis
(Panagopoulou, 1999; Sitlivy et al., 2008; Darlas and Psathi 2016; Tourloukis and Harvati, 2018). As-
semblages from the lowest layers of the Crvena Stijena sequence show the greatest similarity to the
lower layers of Asprochaliko—Layers 18 and 16, dated to 102 ± 14 ka and 96 ± 11 ka, respectively—
e Middle Paleolithic of Crvena Stijena in a regional context | 73
where a comparable faunal assemblage has been identified, including Stephanorhinus kirchbergensis,
Capra ibex, and Testudo sp. (Bailey et al., 1983). Asprochaliko rockshelter also yielded numerous
sidescrapers and Levallois artifacts but with a pronounced laminar component. Although artifacts
from these sites have not yet been thoroughly analyzed and published, most assemblages are charac-
terized by cores and artifacts produced via centripetal and preferential methods, while those made
with the recurrent unipolar method are less common.
There is a notable similarity between the assemblages from the lower layers of Crvena Stije-
na (including Layer XXIV) and those from the Karain E site in Turkey and sites in Central Asia
(Kozłowski, 2002). Similarities are also observed between Crvena Stijena and certain sites in the Le-
vant. For instance, Nesher Ramla in Israel—dated to MIS 5 and renowned for hominin fossils (Her-
shkovitz et al., 2021)—exhibits nearly all the technological features characteristic of Crvena Stijena:
preferential and centripetal Levallois technology, COF (i.e., Kombewa cores and flakes), TFP, déjeté
and convergent sidescrapers, and artifact thinning (Prévost and Zaidner, 2020).
At this stage, it is unclear whether the parallels between Crvena Stijena and sites in the East re-
sult from population movements (Kozłowski, 2002), cultural transmission, or simply represent cul-
tural convergence. However, it is reasonable to assume that cultural and social interactions existed
between the populations inhabiting the eastern Mediterranean at the end of the Middle and, likely,
into the beginning of the Late Pleistocene. These interactions are evident in the spread of Quina
technology and the habitual use of fire in the Chibanian (i.e., Middle Pleistocene). Whether similar
interactions occurred at the onset of the Late Pleistocene remains to be determined.
Crvena Stijena and the Question of the Distribution
of Bifacial Backed Tools
Recent studies have revealed that the southern limit of the distribution of bifacial backed tools,
characteristic of the Micoquian or Keilmessergruppen (KMG) complex—widespread in the north-
ern Pannonian Basin and Eastern Europe during MIS 5–3 (Bosinski, 1967; Richter, 2000; Delpiano
and Uthmeier, 2020)—extended further south than previously believed. In the Pannonian Basin,
the southernmost occurrences of these industries have been documented in Hungary (Mester et al.,
2020). Tools of this type have also been documented at the Kadar site in northern Bosnia (Brodar,
1953), Petrovaradin Fortress in northern Serbia (Mihailović, 2009b), Kozarnika Cave in western
Bulgaria (Sirakova, 2020), and even at sites in Greece, such as Megalo Karvounari (Galanidou et
al., 2016). Additionally, an asymmetrical sidescraper without a back, potentially linked to this com-
plex, has been identified in layer 6d of Šalitrena Pećina in northwestern Serbia, dated to MIS 4
(Mihailović et al., in press-b).
After the preliminary analysis of the Crvena Stijena materials, we initially interpreted the bifa-
cial and thinned sidescrapers from the site as ad hoc tools, where the bifacial retouch was executed
according to immediate needs and the characteristics of the raw material. However, we have revised
our interpretation following a reevaluation of the collected assemblage and a review of the site’s
chronology. In layers XXX, XXVII/XXVI, and XXV, three atypical oval bifacial sidescrapers (Faust-
keilblätter) were identified—one per layer—while layer XXIV yielded an asymmetrical backed side-
scraper (Keilmesser) resembling those from Central European Micoquian sites (Weiss et al., 2018;
Weiss, 2020). This appears to be more than coincidental, as these layers date to the same period
when this type of industry emerged in Central Europe, and bifacial backed tools have already been
confirmed at sites in the northern Balkans. If similar tools are discovered at other sites along the
coastal zone, we could argue that the distribution of bifacial backed tools extended to the eastern
Adriatic coast. This would not entirely come as a surprise, given the broad distribution of Mico-
quian elements (Frick, 2020) and recent findings indicating that Middle Paleolithic industries with
leaf points spread as far west as northern Italy (Peresani et al., 2022), as far east as Bulgaria (Taneva,
2020), and as far south as Greece (Galanidou et al., 2016) during MIS 4 and MIS 3.
At most sites, including Crvena Stijena, bifacial backed tools are found in contexts where the
Quina component is significantly represented. The widespread use of this tool type is likely linked
to the fact that these tools were frequently hafted (Picin, 2021), as indicated by the traces of thinning
found on many tools from layer XXIV of Crvena Stijena. Of course, only future use-wear analyses
will confirm whether these tools were indeed hafted.
74 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Crvena Stijena and the issue of Adriatic Charentian
The “Charentian question” remains one of the most complex and unresolved issues in studying
technological variability in the Middle Paleolithic. Some researchers still interpret Charentian as a
distinct cultural phenomenon, limited to Western Europe and the period spanning MIS 4 and the
beginning of MIS 3 (see ERC project “Quina World” No. 851793). In contrast to earlier typological
approaches, modern studies of the Charentian are grounded in technological criteria (e.g., Turq,
1989; Bourguignon, 1997; Geneste et al.,1997). The Quina method is linked to the high mobility
of Neanderthal groups during MIS 4 and highlights the use of sidescrapers as curated tools—likely
transported and employed across various locations (Hiscock et al., 2009).
Research at Balanica Cave Complex and other Mediterranean sites has revealed that the Quina
method emerged quite early—during MIS 9 or even earlier—in the Near East, within the Acheulo-
Yabrudian complex (Shimelmitz et al., 2014; Zaidner and Weinstein-Evron, 2016), before gradually
spreading to southeastern Europe and the western Mediterranean (Mihailović et al., 2022a). While
the evidence of its presence in Europe during MIS 6 is scarce, the Quina method becomes the domi-
nant knapping technique in MIS 5, particularly within the so-called Charentian of southeastern and
central Europe (Kozłowski, 2016). By MIS 4, the method dominated assemblages in southwestern
France and remained prevalent until the beginning of MIS 3 (Faivre et al., 2014).
Until about a decade ago, evidence of Quina technology in the Adriatic zone was absent. The
Charentoid assemblages from Crvena Stijena were initially attributed to the Pontinian (Basler, 1975),
while the Charentian aspects of the Theopetra industry were called into question (Darlas, 2007).
However, more recent research has revealed the presence of the Quina component in layers XXII–
XX of Crvena Stijena (Mihailović et al., 2017a) and greater prominence of Quina elements at some
MIS 4 and early MIS 3 sites. In the northern Adriatic, within the subalpine zone, Quina technology
has been found at sites such as the De Nadale and Grotta di Fumane caves (Jéquier et al., 2015; Del-
piano et al., 2022; Peresani et al., 2023), while Karavanić and Banda (2023) have recently reassigned
other sites—originally categorized as other types of Mousterian (including Crvena Stijena)—to the
so-called Adriatic Charentian. The Quina component was also identified in the lower part of the
sequence at Klissoura Cave on the Peloponnese (Sitlivy et al., 2008).
In this context, it is important to note that typical Quina cores have not been recorded in any
layer of Crvena Stijena—not even in layer XXII, where Quina elements are most prevalent. How-
ever, it cannot be said that the assemblages from the middle part of the Mousterian sequence at
Crvena Stijena (XXII–XX) lack Quina component, as they contain a notable number of “Quinoid”
flakes with natural or retouched backs, which were either used right away or employed as blanks
for tool production, primarily sidescrapers. Besides flaking technology, the Quina component is
also prominent in the Crvena Stijena tool structure. In layers XXII and XX, transversal and lateral-
transversal sidescrapers are half as common as lateral sidescrapers, while sidescrapers with Quina
retouch account for the highest proportion (20% of all sidescrapers) across all assemblages. The
Essential Charentian index (IC-ess) is 40 for layer XXII, 31 for layer XX, and 16 for layer XVI-
II (Brodar’s collection). From a typological standpoint, the assemblages from layers XXII and XX
clearly belong to the Quina-type industries.
Considering all of this, we can conclude that the Quina Mousterian is indeed present in layers
XXII and XX of Crvena Stijena—as previously noted in our earlier work on the tools (Mihailović
et al., 2017a)—and that Quina elements are also present, to some extent, in layer XVIII. This is
further supported by evidence of short-term occupation (as we have already discussed), which
aligns with hypotheses about the use of the Quina system during MIS 4 (Hiscock et al., 2009).
However, we believe these data are insufficient to define the “Adriatic Charentian” as a distinct
regional facies. Based on the published data, the presence of the Quina component in collections
from other sites remains highly questionable, particularly given the lack of technological analy-
ses to support claims of its presence. Nonetheless, it can be assumed that the spread of Quina
technology (i.e., increased use of the Quina method) during MIS 4 likely extended beyond the
western Mediterranean and sub-Alpine region to include the Adriatic, and probably the Ionian
and Aegean regions as well.
e Middle Paleolithic of Crvena Stijena in a regional context | 75
Crvena Stijena and the Micromousterian Question
The term “Micromousterian” has such a long history that even Brodar (1962) classified artifacts
from the upper layers of the Crvena Stijena sequence under this facies. During the 1970s and 1980s,
the term fell out of use, with many collections instead classified as “Denticulate Mousterian” until E.
Papaconstantinou revived it in the late 1980s while working on his dissertation based on materials
from Asprochaliko in Epirus (Papaconstantinou, 1989). In recent decades, following a critical reas-
sessment of the criteria for defining Denticulate Mousterian (Thiébaut, 2005, 2010), the term “Micr-
omousterian” has gained renewed use. During this period, Late Middle Paleolithic assemblages from
sites such as Mujina Pećina (Croatia), Crvena Stijena and Bioče (Montenegro), Klissoura (Greece),
and others in the Adriatic-Ionian zone have been directly or indirectly attributed to Micromous-
terian (Đuričić, 2006; Karavanić et al., 2008; Sitlivy et al., 2008; Mihailović, 2014; Dogandžić and
Đuričić, 2017; Vujević et al., 2017).
Most published studies on the Micromousterian, including recent ones, have examined the caus-
es behind the emergence of microlithic industries (Mihailović, 2014; Dogandžić, 2023; Karavanić
and Banda, 2023). In the study focused on the analysis of artifacts from Bioče and Crvena Stijena,
Dogandžić and Đuričić (2017) argued that microlithization was more a result of intense reduction
than the quality of the raw materials. Dogandžić (2023) also proposed that the MIS 5 to MIS 3
transition involved a shift in tool production and use—in the later period, tools that relied on the
modification and further use of the working edge were used less frequently, giving way to a larger
number of expedient tools (Nelson, 1991).
Of the proposed hypotheses, only the one regarding raw materials has been tested. Based on
their analysis of materials from Crvena Stijena and Bioče, Dogandžić and Đuričić (2017) concluded
that the microlithic dimensions of the artifacts were not due to the quality of the raw materials but
rather the use of technological solutions that enabled intensive core exploitation. Our analysis of
materials from Crvena Stijena, however, suggests that the intensive exploitation of cores and COF
technology may be linked to the intensity of occupation—particularly in layer XVIII and the upper
layers of the Middle Paleolithic sequence (XIV–XII). Furthermore, morphometric analysis revealed
that artifacts from Crvena Stijena display smaller dimensions than those from other Micromouste-
rian sites in the Adriatic-Ionian region (Dogandžić and Đuričić, 1917; Karavanić and Banda, 2023;
Papaconsantinou, 1989).
The flaking method employed by Neanderthal groups at Crvena Stijena is similar to that used
at Asprochaliko (Papaconstantinou, 1989) but not identical, as lateral preparation of core platforms
was less frequently practiced. Furthermore, unlike Asprochaliko and other sites in the region, Cr-
vena Stijena yielded a significantly smaller number of pseudo-Levallois flakes. All of this suggests
that the main characteristic of the Micromousterian industries in the Adriatic-Ionian region was
not only the use of discoid technology (Dogandžić, 2023; Ruka et al., 2024) but also the knapping
of cores on flakes. At some sites, the COF method focused more on discoid production, while at
Crvena Stijena, it was more adapted to the Levallois method.
Crvena Stijena and the Middle to Upper Paleolithic Transition
The issue of the Middle to Upper Paleolithic transition in the northern Mediterranean has be-
come increasingly complex in recent years, as two sites in the western Mediterranean—Mandrin
Cave in France, dated to 56,800–51,700 cal BP (Slimak et al., 2022), and Cueva Millán in Spain,
with layer I dated to 44,750–42,850 cal BP (Sánchez-Yustos et al., 2024)—have been assigned to
the early Upper Paleolithic, a phase associated with anatomically modern Homo sapiens (AMHS).
The artifact assemblages from Mandrin Cave and Cueva Millán—attributed to the Neronian and
Arlanzian phases, respectively—feature a more or less pronounced Middle Paleolithic component
(with Mandrin showing less and Cueva Millán more), as well as Upper Paleolithic-type cores and
products, such as pointed elongated flakes and irregular blades struck from conical and carinated
cores. Artifacts from Mandrin Cave are almost entirely Upper Paleolithic in character, while Cueva
Millán, among other finds, contains abruptly retouched tools. Before this, fossil remains of AMHS
76 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
were identified at Uluzzian sites in Italy (Grotta del Cavallo), linking this facies to the early dispersal
of Homo sapiens in Europe (Benazzi et al., 2011). In northern Italy, evidence of continuity between
the late Mousterian and Uluzzian has been noted (Peresani et al., 2016), and Uluzzian components
have also been identified at Crvena Stijena (Mihailović and Whallon, 2017).
The Uluzzian component in the upper Middle Paleolithic layers of Crvena Stijena is evident in
the use of diverse techniques for producing flakes and splintered pieces and in the appearance of
backed tools, including arched points and lunates. However, Middle Paleolithic elements (including
Levallois products) dominate these assemblages, which is uncharacteristic of the Uluzzian. Addi-
tionally, products of bipolar knapping—characteristic of this facies (Moroni et al., 2018; Villa et al.,
2018; Peresani et al., 2019; Rossini et al., 2022; Delpiano et al., 2024)—are not particularly abundant.
Fumane Cave in northern Italy is the only site in the wider region where a comparable situa-
tion has been documented, with transitional elements identified in layer A4 and a significantly more
pronounced Uluzzian component in layer A3 (Peresani et al., 2016). However, the nature of the
remains from layer A4 remains a subject of ongoing debate (Peresani et al., 2019). In this context, it
remains an open question whether the findings from layers XII to XIV of Crvena Stijena should be
attributed to an early (transitional) phase between the Mousterian and the Uluzzian or whether they
result from a disturbed context—particularly considering that the excavations were conducted long
ago and that the possibility cannot be ruled out that the Uluzzian horizon was overlooked during
the excavation process.
Regarding the presence of bipolar technology at Crvena Stijena, it is important to note that
relatively few typical bipolar cores and their flaking products (such as blades and microblades) were
found at the site. However, if we consider that some (completely exhausted) splintered pieces could
have been used as cores, the situation changes: in layer XVIII, the number of splintered pieces and
cores is evenly distributed (15 each), while in layer XIV, 15 cores and five splintered pieces were
found. Additionally, many artifacts classified into different core and tool types based on primary
characteristics also show traces of bipolar flaking. For example, in layer XVIII, dated to before 49
ka cal BP (based on the date obtained for layer M3/XIII), four cores classified as COF were found
to exhibit traces of bipolar flaking, while among the artifacts categorized as splintered pieces, five
examples were identified. This suggests that bipolar technology may have played a more significant
role in flake production at Crvena Stijena than the analysis results suggest.
The issue of laminar technology at Crvena Stijena is thoroughly addressed in the dissertation
by S. Dragosavac (2022), which will be published as a separate volume. For now, it is worth noting
that the proportion of blades and bladelets in the blank structure is 11.4% in layer XVIII, 12.5% in
layer XVI, 12% in layer XV, 5.7% in layer XIV, 7.6% in layer XIII, and 5.8% in layer XII. The pro-
portion clearly differs between the upper and lower layers (when referring to Brodar’s assemblages),
with the laminar component most pronounced in layers XVIII and XVI—a pattern consistent with
the data presented by both Dogandžić and Đuričić (2017) and Dragosavac (2022). According to
Dragosavac (2022), the blades were struck using a soft hammer (both inorganic and organic), most
likely through direct percussion, although some show evidence of indirect percussion.
Without delving deeply into a comparative analysis of the artifacts from Crvena Stijena and
those from the Uluzzian sites in Italy and IUP sites in France and Spain, it is worth noting that the
presence of Upper Paleolithic elements in the upper layers of Crvena Stijena—supported by the very
early dates obtained for layer M2 (46–49 ka cal BP) in 2017, and considering the dates obtained for
Mandrin Cave, Cueva Millán, Bacho Kiro in Bulgaria, and Ranis in Germany (Fewlass et al., 2020;
Hublin et al., 2020; Slimak et al., 2022; Sánchez-Yustos et al., 2024; Mylopotamitaki et al., 2024)—
raises the possibility that interactions between Neanderthals and AMH may have occurred along
the eastern Adriatic coast as early as 45 ka cal BP. Research on layer E in Mandrin Cave suggests that
Upper Paleolithic populations entered the region over 50 ka cal BP (Slimak et al., 2022), likely trave-
ling along the Mediterranean corridor rather than coming from Central Europe. The authors of the
Mandrin Cave study proposed a connection with Ksar Akil and Levantine occurrences—a hypoth-
esis that remains challenging to prove, though evidence from Crvena Stijena apparently supports
this interpretation. It is important to emphasize that the emergence of the IUP in the Mediterranean
did not signify the end of the Middle Paleolithic. At Mandrin Cave, Middle Paleolithic layers were
e Middle Paleolithic of Crvena Stijena in a regional context | 77
identified even in the upper, post-Neronian strata, situated above the Proto-Aurignacian horizon
(Slimak et al., 2022).
Despite the constraints of a modest and non-representative sample, no significant differences
are observed between the technological expressions at Crvena Stijena and those documented at sites
in the western Mediterranean. Both exhibit a similar blend of Middle Paleolithic elements—Leval-
lois, discoid, and core-on-flake (COF) technologies—alongside Upper Paleolithic features, evident in
the presence of unipolar and carinated cores, irregular blades and bladelets, and abruptly retouched
tools. The only feature that appears less prominent at Crvena Stijena compared to those sites (aside
from bladelets, which are sparsely represented, possibly due to methodological limitations of earlier
excavations) is the lack of significant evidence for Upper Paleolithic cores flaked using a convergent
scheme characteristic of the IUP, along with the corresponding flaking products. However, given re-
ports that Arlanzian technology is characterized, among other features, by “the very specific ‘schéma
croisé’ flaking mode, where a ventral surface of a core-on-flake serves as a production plane, fol-
lowing a blade-and-point reduction” (Sánchez-Yustos et al., 2024), and that a similar method was
employed at Mandrin Cave, we find it important to emphasize once again the presence of the COF
method at Crvena Stijena. Notably, unipolar convergent core for bladelet production was identified
among the cores of this type, which fits within this framework.
In conclusion, the Middle to Upper Paleolithic transition in the northern Mediterranean was
characterized by similar technological trends, which should not be seen as isolated phenomena. The
findings from Crvena Stijena suggest a high degree of integration of methods typically associated
with different technocomplexes (Mousterian, IUP, Uluzzian), highlighting the complexity of cul-
tural and social interactions between different populations during the Middle to Upper Paleolithic
transition in this region. These trends are more likely linked to the emergence of AMHS than to the
local development of Middle Paleolithic technology.
| 79
Climatic and demographic factors
of technological variability
While it is widely accepted that climatic factors and the population movements they triggered
played a significant role in technological changes during the Pleistocene, there is limited evidence
on how these factors may have influenced technological shifts in the Middle Paleolithic of the Bal-
kans. It is also important to recognize that the concepts of “migration” or “population movement”
should be applied with caution when interpreting patterns in Paleolithic archaeology (Clark, 1994).
Nevertheless, these factors should certainly be considered in the context of the Balkans, especially
given the region’s historical role as a migration corridor between southwestern Asia and central and
western Europe. Moreover, the Balkans is well-established as a Pleistocene glacial refugium for plant
and animal species and likely for hominin populations as well (Griffiths et al., 2004; Hewitt, 2011;
Roksandic et al., 2023). With its distinct climatic zones and well-defined natural corridors shaped by
its mountainous terrain, the Balkan Peninsula provides ideal conditions for studying the impact of
climatic changes on population movements.
Middle Paleolithic
Contrary to expectations and decades of research, only a few sites with substantial remains have
been discovered on the Balkan Peninsula that could be classified as base camps or multi-occupational
sites, suggesting potential continuity of habitation in a given area. Moreover, very few sites from the
glacial periods have been identified in the Balkans. In this context, it is important to note that for MIS
6, reliable dates are only available for the materials from layer 10b of Kozarnika Cave (Heydari et al.,
2022), while remains from other sites believed to potentially date to this period have either not been
dated (e.g., Crvena Stijena) or the obtained dates encompass warmer phases occurring before or after
MIS 6. A similar situation applies to the Balkan Peninsula during MIS 4. In addition to Crvena Stijena,
layer 6 at Temnata Dupka (Bulgaria) has been dated to more than 67 ± 11 ka (GdTL-254; Bluszcz et
al., 1992). Remains of non-Arctic fauna were found in this layer, and similar ages—corresponding to
the end of MIS 5 and MIS 4—and fauna have been identified in the lower layers of Šalitrena Pećina in
Serbia (Marín-Arroyo and Mihailović, 2017; Dakovic et al., 2019). The lower layers of Klissoura Cave
in the Peloponnese may also date to this period (Starkovich, 2014, Table 1).
Sites with remains from warmer intervals are more numerous, with remains from Crvena Stijena,
Pešturina Cave (Serbia), Krapina (Croatia), Theopetra and Klissoura Cave (Greece), Zobište (Bosnia
and Herzegovina), and other sites all attributed to MIS 5 (Montet White et al., 1986; Mihailović et
al., 2022b). However, precise dates are lacking for many of these sites, and their interglacial ages are
primarily inferred from faunal remains (Miracle, 2007; Starkovich, 2014; Milošević, 2020). A similar
situation applies to MIS 3, with more sites dating to this period than to MIS 4, although their ages
are predominantly determined through radiocarbon (14C) dating rather than methods that can date
older layers. These include Velika Pećina in Kličevica (Croatia), Bioče and Mališina Stijena (Mon-
80 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
tenegro), Bacho Kiro (Bulgaria), and several sites in Greece (e.g., Asprochaliko, Klissoura Cave 1),
though the dates for these sites are somewhat uncertain for various reasons (Karavanić et al., 2018;
Vishnevskiy et al., 2019). For some sites, only one or two absolute dates are available (e.g., Asproch-
aliko, Mališina Stijena), while others have only a minimum age established (e.g., Mališina Stijena)
(Hedges et al., 1990). Some sites are dated over a broad time range (e.g., Bioče, Velika Pećina in
Kličevica), while others face issues with the integrity of layers and the context of the finds, such as
the Serbian sites of Pešturina Cave (layer 3), Hadži Prodanova Cave (layer 5a), Petrovaradin Fortress
(layer 2a), and Tabula Traiana Cave (Alex et al., 2019; Marković et al., 2021; Borić et al., 2021). Only
a few sites—including Mujina Pećina and Crvena Stijena—have yielded reliable, multi-method dates
for MIS 3 (Mercier et al., 2017; Boschian and Gerometta, 2022).
Figure 28. Initial Upper Paleolithic (IUP), Uluzzian and Aurignacian sites in Southeastern Europe and
southern Italy: Mokriška Jama (MJ), Potočka Zjalka (PZ), Šandalja II (ŠII), Šalitrena Pećina (ŠP), Romanești
Dumbravița/Cosava (R/C), Tincova (T), Crvenka-At (C-A), Tabula Traiana Cave (TT), Mala Pećina (MP),
Bukovac/Orlovača (B/O), Baranica (BAR), Kozarnika (KOZ), Temnata/Toplitsa (TD/TO), Bacho Kiro (BK),
Kolomnitsa (KO), Kephalari (KEP), Lakonis (L), Klissoura (KLI), Franchthi (FRA), Cavallo (CAV), Oscurus-
ciuto (OSC), Cala (CAL), Castelcivita (CAS). IUP sites are marked in red, with sites ascribed to the “Proto-
Aurignacian” in green, Aurignacian in black, and Uluzzian in yellow. Crvena Stijena is marked with a star.
Arrows indicate potential directions of population movements.
Climatic and demographic factors of technological variability | 81
Given that Crvena Stijena and other sites in the Balkans show traces of intense settlement at the
end of MIS 5 and the beginning of MIS 4, a hypothesis has been proposed suggesting that as cooling
began, Neanderthal groups may have aggregated in coastal areas, causing a demographic depopula-
tion of the interior of the Balkan Peninsula—a pattern also seen during the Last Glacial Maximum
(Mihailović et al., in press-a). While this hypothesis still requires testing, it is already clear that the
coastal zone was significantly more densely populated than the interior, as indicated by a series of
rich, multi-layered sites in the Adriatic-Ionian region. There are also indications—particularly at
Crvena Stijena—that Neanderthal groups may have persisted in this area even at the peak of the
glacial period (Mihailović et al., in press-a).
If population movements from north to south occurred during MIS 4, they are not clearly re-
flected in the material culture, except for the appearance of bifacial tools in the southern Pannonian
Basin, Posavina, and the sub-Alpine regions during this period, and the spread of the Western Eu-
ropean Quina Mousterian, with its characteristic tool technology and behavioral traits, toward the
eastern Adriatic coast (Delpiano et al., 2022; Karavanić and Banda, 2023). The general character of
MIS 3 industries remains largely unchanged, primarily reflecting the distribution of resources and,
to some extent, cultural regionalization, which aligns with the location of major morphostructural
units in the relief of the Balkan Peninsula (Mihailović and Mihailović, forthcoming). South of the
areas where industries with Levallois and bifacial components occur in the southern Pannonian
Basin and peri-Pannonian region, “Typical” Mousterian industries are also found, though without
bifacial tools. In contrast, the coastal area (unlike the Balkan interior) features a much more promi-
nent microlithic component.
Middle to Upper Paleolithic Transition
Based on the absolute dates obtained for Bacho Kiro, Brno-Bohunice (Czech Republic), and
Ranis, the Initial Upper Paleolithic (IUP) and the first groups of anatomically modern Homo sapiens
(AMHS) appeared in Central Europe as early as 47 ka cal BP (Richter et al. 2009; Hublin et al., 2020;
Mylopotamitaki et al., 2024; Sümer et al., 2024), if not earlier. Proto-Aurignacian (and possibly early
Aurignacian) populations spread around 43–40 ka cal BP (Chu, 2018). In both cases, several ques-
tions arise: when and through which corridors did AMHS settle in Europe, in what numbers and at
what pace, and how did they interact with the local Neanderthal populations?
Regarding Central Europe and the northern parts of the Balkans, it is evident that AMHS set-
tled this part of the continent more than 45 ka cal BP, at a time when Neanderthals were still present
in the region. While it could be assumed that this occurred during warmer phases (Müller et al.,
2011), at least two cases suggest that AMHS appeared during cooler climatic periods, which may
indicate they entered Europe from the steppe regions of Eastern Europe (Nigst et al., 2014; Peder-
zani et al., 2021). Nearly all inland sites have been found within the Danube Basin area, supporting
the hypothesis that most AMHS populations entered Europe via the Danube Corridor (Conard and
Bolus, 2003). Aside from biological (genetic) evidence (Fu et al., 2015, 2016), there are no other (ar-
chaeological) indications of interactions between the local Neanderthal populations and incoming
AMHS groups in this region.
In the coastal zone, the IUP appears very early—before 45 to over 50 ka cal BP in Mandrin
Cave and Cueva Millán and before 44–43 ka in the case of the Uluzzian (Higham et al., 2024)—
although significantly fewer fossil remains reliably attributed to AMHS have been found at these
sites (Benazzi et al., 2011; Higham et al., 2011). However, these industries contain far fewer Mid-
dle Paleolithic elements compared to the IUP in Central Europe, where the Levallois component is
much more prominent. The distinct and pronounced Upper Paleolithic character of the IUP and
Uluzzian in the Mediterranean has been explained (in the case of the IUP) by the rapid arrival of
AMHS into territories previously occupied by Neanderthals (Slimak et al., 2022) and in the case of
the Uluzzian, by a cultural tradition linked to the use of bipolar technology (Delpiano et al., 2024).
In attempts to determine the origins of the Upper Paleolithic populations that settled the western
Mediterranean, some authors have drawn parallels with very distant regions—in the case of the IUP,
with the Levant (Slimak et al., 2022), and in the case of the Uluzzian, with the Late MSA–Early LSA
of Africa (Moroni et al., 2013, 2016, 2018; Ronchitelli et al., 2018; Sano et al., 2019; Delpiano et al.,
82 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
2024). However, due to the lack of sites in the regions between these areas, both scenarios can, at
present, only be considered hypothetical.
While caution is generally advised when attributing different industries to specific hominin
taxa, especially given evidence of hybridization (Finlayson et al., 2023), we believe there is sufficient
evidence to support the early presence of AMHS and the Upper Paleolithic in the Mediterranean.
Moreover, questions remain about how many early Upper Paleolithic technocomplexes from this
period should be attributed distinct cultural traits, given that Middle Paleolithic elements are still
present to some extent (e.g., Cueva Millán, Fumane), whether in the form of Levallois, discoid, or
COF technologies. Conversely, Upper Paleolithic elements, such as laminar technology and differ-
entiated abruptly retouched tools, including lunates, are found in the late Middle Paleolithic of the
Mediterranean, 50–44 ka cal BP. From this perspective, and considering the appearance of Upper
Paleolithic elements at Crvena Stijena over 47 ka cal BP, it seems entirely plausible that technological
transfer and contact occurred between Neanderthals and AMHS, as well as among groups associ-
ated with different Upper Paleolithic technocomplexes, continuing until the eventual extinction of
Neanderthals. Recent analyses of material from Riparo l’Oscurusciuto in southern Italy (Carmig-
nani et al., 2024) further support this hypothesis.
Regarding a slightly later period (43–40 ka cal BP), it is important to note that the spread of the
makers of the Proto-Aurignacian technocomplex seems to have bypassed the eastern Adriatic coast,
as no reliable sites attributed to the Proto-Aurignacian or Early Aurignacian have been identified be-
tween the Peloponnese and Istria. While it is possible that such sites existed but are now submerged,
it is also plausible that the spread of Aurignacian groups toward the sub-Alpine belt and further into
Western Europe occurred along the Posavina corridor rather than the Adriatic corridor (Mihailović,
2020; Dragosavac, 2023). At this point, there is no conclusive evidence to support either hypothesis.
This raises the question of how interactions between Neanderthals and AMHS might have un-
folded. There was likely significant ecological and territorial competition between the two hominin
groups, as indicated by the fact that no contemporaneity has been reliably demonstrated between
Middle Paleolithic and Upper Paleolithic sites at a regional level in Europe, only their succession
(with the possible exception of Mandrin Cave; Slimak et al., 2022). Evidence of interactions is also
scarce. Drawing on assumptions from the push-pull model (Anthony, 1990), we propose that under
competitive conditions, the advance of Upper Paleolithic populations may have driven Neanderthal
groups to retreat to the western Balkans and/or more inaccessible areas of the peninsula’s interior
(Mihailović, 2020). Later, contacts may have occurred due to pull factors, eventually leading to the
assimilation of the Neanderthal population.
Recent research has provided further insights into the chronology of Upper Paleolithic popula-
tions moving from east to west across the Balkans. In contrast, definitive evidence for the retreat
of Middle Paleolithic groups into the interior of the Balkan Peninsula still needs to be provided,
although some compelling indications support this. A key new finding is the exceptionally late dates
obtained from several coastal sites, which lend weight to the hypothesis that the eastern Adriatic
coast may have served as a Neanderthal refugium. However, the chronology of these sites remains
imprecise, and the Middle Paleolithic material from Bioče—reported to have been found above the
Campanian Ignimbrite (CI) tephra layer (Vishnevskiy et al., 2019)—has yet to be published. In this
context, it remains uncertain whether Neanderthals persisted longer in the coastal zone than in the
interior of the peninsula, though this possibility cannot be ruled out, especially considering that
recent findings at the Dalani i Vogël coastal site in Albania include a small number of Middle Paleo-
lithic artifacts, which have been dated between 42,900 ± 1500 and 38,700 ± 1400 years BP via OSL
(Badino et al., 2025).
| 83
Conclusion
Crvena Stijena is one of the numerous multi-occupational sites in the coastal zone of the Bal-
kans, in sharp contrast to the limited number of such sites in the interior of the peninsula. These
sites provide evidence of continuous Middle Paleolithic settlement along the eastern Adriatic coast,
where, quite possibly, population aggregation occurred during glacial periods. To what extent this
continuity in settlement might have influenced technological continuity remains to be determined,
especially since similar patterns have been observed in neighboring regions.
The earliest phase of occupation at Crvena Stijena, likely corresponding to MIS 6/5, was char-
acterized by the use of Levallois technology and the frequent utilization of sidescrapers, which were
often ventrally thinned. The Quina method became increasingly prominent in the middle phase,
likely influenced by Western innovations and prevailing climatic and ecological conditions. The end
of the Middle Paleolithic was marked by the emergence of an expedient techno-economic model,
expressed at some sites through increased use of the discoid method, while at Crvena Stijena, it
manifested as core-on-flake technology adapted primarily to the Levallois method. While unsys-
tematic, the late Middle Paleolithic blade production closely resembled the methods observed in
other regions of the northern Mediterranean.
Although the lithic material collected during the 1950s and 1960s excavations at Crvena Sti-
jena is not a representative sample for addressing key questions about Mousterian technology and
Paleolithic studies in this region, we remain confident that ongoing research holds significant prom-
ise. These new studies, based on microarchaeological excavations of Basler’s trench’s eastern and
southern profiles, have already provided valuable insights into the area’s ecology, hominin activities
related to hunting and fire use, and post-depositional processes.
The interpretative scope of the old material from Crvena Stijena is limited, yet the results of
our analysis can be connected to more recent studies of Middle Paleolithic sites in the surrounding
region and placed within the broader context of the Middle to Upper Paleolithic transition in south-
eastern Europe. We hope our work contributes to a deeper understanding of technological changes
in this region, as a comprehensive knowledge of this area is crucial for evaluating the factors that
shaped the spatial and temporal variability of Middle Paleolithic industries throughout the northern
Mediterranean.
84 |
| 85
Plates
Plate 1. Cores from Basler’s collection, layers XXXI (1–4) and XXV (5). Levallois (1–3, 5), discoid (4). (Origi-
nal drawing by D. Mihailović, redrawn by S. Dragosavac)
86 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 2. Tools from Basler’s collection, layer XXXI. Mousterian point (1), sidescrapers (2–6), truncated tool
(7), and perforators (8–10). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
88 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 4. Tools from Basler’s collection, layer XXIV. Sidescrapers (1–6), perforators (7, 8). (Original drawing by
D. Mihailović, redrawn by S. Dragosavac)
90 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 6. Tools from Basler’s collection, layer XXI. Retouched flake (1), sidescrapers (2–4), perforator (5).
(Original drawing by D. Mihailović, redrawn by S. Dragosavac)
92 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 8. Tools from Basler’s collection, layers XVIII (8), XVII (4, 9), XVI (5), XV (10), XIV (1, 2, 6), XIII (3),
and XII (7). Retouched Levallois flake (1), Mousterian points (2, 3), sidescrapers (4–7), and backed pieces
(8–10). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
94 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 10. Cores from Brodar’s collection, layer XVIII. Discoid cores (1, 2), splintered and truncated-faceted
pieces (3–9). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
Plates | 95
Plate 11. Characteristic flakes from Brodar’s collection, layers XVIII (1–6), XVII (7, 8), and XVI (9, 10). Levallois
flakes 1, 2, 7, 9, 10), débordant flake (3), crested blade (4), pseudo-Levallois flake (5), blade fragment (6), overshot
flake (8). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
96 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 12. Cores from Brodar’s collection, layers XVI (1–4) and XV (5–7). Levallois cores on flakes (1, 5),
discoid core (2), unipolar core on flake (3), burin-type core (4), bipolar cores (6, 7). (Original drawing by D.
Mihailović, redrawn by S. Dragosavac)
98 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 14. Cores from Brodar’s collection, layer XIV. Burin-type core (1), discoid cores (3, 4), irregular cores on
flakes (4, 5), and truncated-faceted pieces (6–11). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
100 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 16. Cores from Brodar’s collection, layer XII. Levallois cores (1–5), non-Levallois cores on flakes (6–8).
(Original drawing by D. Mihailović, redrawn by S. Dragosavac)
102 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 18. Characteristic flakes from Brodar’s collection, layers XIV (1–5), XIII (6–9), and XII (10–18). Le-
vallois flakes (1–3, 6, 10–15), Kombewa flakes (4, 9), naturally backed knife (5), crested blade (7), pseudo-
Levallois flake (8), unretouched blades and flakes (16–18). (Original drawing by D. Mihailović, redrawn by S.
Dragosavac)
104 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 20. Tools from Brodar’s collection, layer XVIII. Retouched blades (1, 2), pointed blade (3), burins (4–6),
endscraper-perforator (7), and perforator (8). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
Plates | 105
Plate 21. Tools from Brodar’s collection, layers XVII (1–3), XVI (4–8), and XV (9–14). Sidescrapers (1, 4–6, 9), bu-
rin (2), denticulated backed blade (3), backed truncation (7), perforators (8, 14), retouched blade (10), raclettes (11,
12), denticulated flake (13). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
106 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 22. Tools from Brodar’s collection, layer XIV. Sidescrapers (1–9), Mousterian point (10), endscrapers
(11, 12). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
108 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 24. Tools from Brodar’s collection, layer XIII. Sidescrapers (1–6), endscrapers (7–9), truncated tool
(10), perforator (11), raclette (12), and notched tool (13). (Original drawing by D. Mihailović, redrawn by S.
Dragosavac)
110 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Plate 26. Tools from Brodar’s collection, layer XII. Endscrapers (1–7), retouched blades (8–10), raclettes (11–
14), and truncated blades (12, 13). (Original drawing by D. Mihailović, redrawn by S. Dragosavac)
| 113
References
Alex, B., Mihailović, D., Milošević, S., Boaretto, E., 2019. Radiocarbon chronology of Middle and Upper Pale-
olithic sites in Serbia, Central Balkans. Journal of Archaeological Science: Reports 25, 266–279. https://doi.
org/10.1016/j.jasrep.2019.04.010
Anđelković, M., 1978. Stratigrafija Jugoslavije, paleozoik i mezozoik. Minerva, Subotica – Beograd.
Anthony, D., 1990. Migration in archeology: the baby and the bathwater. American Anthropologist 92(4), 895–
914. https://doi.org/10.1525/aa.1990.92.4.02a00030
Badino, F., Ruka, R., Pini, R., Frechen, M., Argante, V., Susini, D., Abu El Khair, D., Comolli, R., Mazzini, I.,
Delpiano, D., Çipa, K., Margaritora, D., Gjipali, I., Peresani, M., 2025. Palaeoenvironmental, stratigraphic
and geochronological study of the coastal site of Dalani i Vogël (Vlora, Albania): new evidence for late
Neanderthal occupation and prehistoric archaeology. Quaternary Science Reviews 349, 109111. https://doi.
org/10.1016/j.quascirev.2024.109111
Bailey, G., Carter, P.L., Gamble, C., Higgs, H.P., 1983. Asprochaliko and Kastritsa: Further Investigations of
Palaeolithic Settlement and Economy in Epirus (North-West Greece). Proceedings of the Prehistoric Society
49(1), 15–42. https://doi.org/10.1017/S0079497X00007945
Baković, M., Mihailović, B., Mihailović, D., Morley, M., Vušović-Lučić, Z., Whallon, R., Woodward, J., 2009.
Crvena Stijena excavations 2004–2006, preliminary report. Eurasian Prehistory 6/1–2, 3–31.
Balescu, S., Tuffreau, A., Dobrescu, R., 2015. Mamaia Sat (Romania): A Late Middle Pleistocene Palaeolithic
Site? Anthropologie (Brno) 53(1-2), 157–166.
Basler, Ð. (Ed.), 1975. Crvena Stijena – zbornik radova. Zajednica kulturnih ustanova, Nikšić.
Benac, A., Brodar, M., 1957. Crvena stijena 1955. Glasnik Zemaljskog Muzeja Bosne i Hercegovine u Sarajevu
(Ser. A) N.S. 10, 19–50.
Benac, A., Brodar, M., 1958. Crvena Stijena — 1956: Stratum I–IV. Glasnik Zemaljskog Muzeja Bosne i Herce-
govine u Sarajevu (Ser. A) N.S. 13, 21–64.
Benazzi, S., Douka, K., Fornai, C., Bauer, C.C., Kullmer, O., Svoboda, J., Pap, I., Mallegni, F., Bayle, P., Coque-
relle, M., Condemi, S., Ronchitelli, A., Harvati, K., Weber, G.W., 2011. Early dispersal of modern humans in
Europe and implications for Neanderthal behaviour. Nature 479(7374), 525–528. https://doi.org/10.1038/
nature10617
Benazzi, S., Arrighi, S., Badino, F., Bortolini, E., Figus, C., Lugli, F., Marciani, G., Oxilia, G., Romandini, M.,
Silvestrini, S., Boscato, P., Cipriani, A., Moroni, A., Negrino, F., Peresani, M., Pini, R., Ravazzi, C., Ronchi-
telli, A., Spinapolice, E.E., 2020. Peopling dynamics in the Mediterranean area between 45 and 39 ky ago:
State of art and new data. Quaternary International 551, 1–6. https://doi.org/10.1016/j.quaint.2020.06.029
Biagi, P., Starnini, E., Efstratiou, N., Nisbet, R., Hughes, P.D., Woodward, J.C., 2022. Mountain Landscape and
Human Settlement in the Pindus Range: The Samarina Highland Zones of Western Macedonia, Greece.
Land 12(1), 96. https://doi.org/10.3390/land12010096
Bluszcz, A., Hercman, H., Pazdur, A., Pazdur, M., 1992. Radiometric dating. In: K. Kozłowski, H. Laville and
B. Ginter (Eds.), Temnata Cave. Investigations in Karlukovo karst area. Vol. I . Jagellonia Univeristy Press,
pp. 223–240.
Boëda, É., 2013. Techno-logique et Technologie: une paléo-histoire des objets lithiques tranchants. Prigonrieux,
@rchéo-éditions.
Bordes, F., 1961. Typologie du Paléolithique ancien et moyen. Delmas, Bordeaux.
Borić, D., Cristiani, E., Hopkins, R.J., Schwenninger, J.-L., Gerometta, K., French, C.A., Mutri, G., Ćalić, J.,
Dimitrijević, V., Marín-Arroyo, A.B., Jones, J.R., Stevens, R.E., Masciana, A., Uno, K.T., Richter, K.K.,
114 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Antonović, D., Wehr, K., Lane, C.S., White, D., 2021. Neanderthals on the Lower Danube: Middle Palaeo-
lithic evidence in the Danube Gorges of the Balkans. Journal of Quaternary Science 37(2), 142–180. https://
doi.org/10.1002/jqs.3354
Bosinski, G., 1967. Die Mittelpaläolithischen Funde im Westlichen Mitteleuropa. Böhlau Verlag, Köln – Graz.
Bourguignon, L., 1996. La conception de debitage Quina. Quaternaria Nova 6, 149–166.
Bourguignon, L., 1997. Le Moustérien de type Quina: nouvelles définitions d’une entité technique. Ph.D. disser-
tation, Université Paris X-Nanterre.
Bourguignon, L., Faivre, J.-P., Turq, A., 2004. Ramification des chaînes opératoires: une spécificité du Mo-
ustérien? Paléo, Revue d’Archéologie Préhistorique 16, 37–48.
Bradák, B., Carrancho, Á., Lagunilla, Á.H., Villalaín, J.J., Monnier, G., Tostevin, G., Mallol, C., Pajović, G.,
Baković, M., Borovinić, N., 2020. Magnetic fabric and archaeomagnetic analyses of anthropogenic ash ho-
rizons in a cave sediment succession (Crvena Stijena site, Montenegro). Geophysical Journal International
224(2), 795–812. https://doi.org/10.1093/gji/ggaa461
Brodar, S., 1953. K odkritju kamenih industrij ob Usori. Glasnik Zemaljskog Muzeja Bosne i Hercegovine u
Sarajevu N.S. 8, 225–243.
Brodar, M., 1957a, Crvena Stijena 1955 (Stratum V). Glasnik Zemaljskog Muzeja Bosne i Hercegovine u Sara-
jevu (Ser. A) N.S. 12, 51–56.
Brodar, M., 1957b. Površinski nalazi iz okoline sela Petrovići u Crnoj Gori. Glasnik Zemaljskog Muzeja Bosne
i Hercegovine u Sarajevu (Ser. A) N.S. 12, 57–61.
Brodar, M., 1962. Crvena Stijena 1958. i 1959. Glasnik Zemaljskog Muzeja Bosne i Hercegovine u Sarajevu (Ser.
A) N.S. 17, 15–20.
Brodar, M., 2009. Stara kamena doba v Sloveniji / Altsteinzeit in Slowenien. Self-published, Ljubljana.
Brunnacker, K., 1975. Die Sedimente der Crvena Stijena. In: Đ. Basler (Ed.), Crvena Stijena – zbornik radova.
Zajednica kulturnih ustanova, Nikšić, pp. 171–203.
Cârciumaru, M., Moncel, M.H., Cârciumaru, R., 2000. Le Paléolithique moyen de la grotte Cioarei-Borosteni
(commune de Pestisani, département de Gorj, Roumanie): Étude préliminaire de l’industrie lithique la
question des Moustériens sub-carpathiques et de l’occupation des carpathes. L’Anthropologie 104(2), 185–
237. https://doi.org/10.1016/S0003-5521(00)80045-X
Carmignani, L., Martini, I., Spagnolo, V., Dominici, C., Rossini, M., Scaramucci, S., Moroni, A., 2021. Middle
and early upper Pleistocene human occupations in Southern Italy. A reassessment of the assemblages from
Cala d’Arconte, Capo Grosso and Cala Bianca. Journal of Archaeological Science: Reports 40(A), 103256.
http://dx.doi.org/10.1016/j.jasrep.2021.103256
Carmignani, L., Soressi, M., Ronchitelli, A., Boschin, F., 2024. IUP Technological Signatures or Mousterian
Variability? The Case of Riparo l’Oscurusciuto (Southern Italy). Journal of Paleolithic Archaeology 7, 31.
https://doi.org/10.1007/s41982-024-00196-w
Chu, W., 2018. The Danube Corridor Hypothesis and the Carpathian Basin: Geological, environmental and
archaeological approaches to characterizing Aurignacian dynamics. Journal of World Prehistory 31(2),
117–178. https://doi.org/10.1007/s10963-018-9115-1
Clark, G.A., 1994. Migration as an explanatory concept in paleolithic archaeology. Journal of Archaeological
Method and Theory 1, 305–343. https://doi.org/10.1007/BF02242740
Clark, G.A., Barton, C.M., 2017. Lithics, landscapes & la Longue-durée – Curation & expediency as expressions
of forager mobility. Quaternary International 450, 137–149. https://doi.org/10.1016/j.quaint.2016.08.002
Clark, A.E., 2015. Changes in Land Use and Occupation Intensity at the Onset of the Middle Paleolithic? A
View from Tabun Cave, Israel. In: N.J. Conard and A. Delagnes (Eds.), Settlement Dynamics of the Middle
Paleolithic and Middle Stone Age, Vol. IV. Kerns Verlag, Tübingen, pp. 105–126.
Conard, N.J., Bolus, M., 2003. Radiocarbon dating the appearance of modern humans and timing of cultu-
ral innovations in Europe: new results and new challenges. Journal of Human Evolution 44(3), 331–371.
https://doi.org/10.1016/S0047-2484(02)00202-6
Ćulafić, G., Tostevin, G.. Borovinić, N., 2017. Sources of Lithic Raw Materials near Crvena Stijena. In: R.
Whallon (Ed.), Crvena Stijena in Cultural and Ecological Context. Multidisciplinary Archaeological Research
in Montenegro. Montenegrin Academy of Sciences and Arts, National Museum of Montenegro, Podgorica,
pp. 257–265.
Dakovic, G., Blackwell, B.A.B., Mihailovic, B., Plavsic, S., Qi, J.K., 2019. ESR Dating Herbivore Teeth within
the Mousterian Layers at Šalitrena Pećina, Serbia. Presented at the 84th Annual Meeting of the Society for
American Archaeology, Albuquerque, NM. (tDAR id: 450117)
Darlas, A., 2007. Le Moustérien de Grèce à la lumière des récentes recherches. L’Anthropologie 111(3), 346–
366. https://doi.org/10.1016/j.anthro.2007.04.001
References | 115
Darlas, A., Psathi, E., 2016. The Middle and Upper Paleolithic on the Western Coast of the Mani Penin-
sula (Southern Greece). In: K. Harvati and M. Roksandic (Eds.), Paleoanthropology of the Balkans and
Anatolia. Vertebrate Paleobiology and Paleoanthropology. Springer, Dordrecht, pp. 95–117. https://doi.
org/10.1007/978-94-024-0874-4_7
Debénath, A., Dibble, H. 1994. Handbook of Palaeolithic Typology, Volume I. Lower and Middle Palaeolithic of
Europe. University of Pennsylvania Press, Philadelphia.
Delpiano, D., Uthmeier, T., 2020. Techno-functional and 3D shape analysis applied for investigating the varia-
bility of backed tools in the Late Middle Paleolithic of Central Europe. PLoS ONE 15(8), e0236548. https://
doi.org/10.1371/journal.pone.0236548
Delpiano, D., Zupancich, A., Bertola, S., Martellotta, E.F., Livraghi, A., Cristiani, E., Peresani, M., 2022. Flexi-
bility within Quina lithic production systems and tool-use in Northern Italy: implications on Neanderthal
behavior and ecology during early MIS 4. Archaeological and Anthropological Sciences 14(12), 219. https://
doi.org/10.1007/s12520-022-01684-2
Delpiano, D., Marciani, G., Conforti, J., Lombardo, S., Rossini, M., Santos, M.C., Benazzi, S., Peresani, M., Mo-
roni, A., 2024. Not just a technique! An experimental approach to refine the definition of the bipolar anvil
reduction in the Uluzzian. Archaeological and Anthropological Sciences 16, 202. https://doi.org/10.1007/
s12520-024-02097-z
Dibble, H., McPherron, S., 2007. Truncated-faceted pieces: hafting modification, retouch, or cores? In: S.
McPherron (Ed.), Tools Versus Cores. Alternative Approaches to Stone Tool Analysis. Cambridge Scholars
Publishing, pp. 75–90.
Dibble, H. Rolland, N. 1992. On assemblage variability in the Middle Paleolithic of Western Europe: history,
perspectives, and a new synthesis”. In H. Dibble and P. Mellars (Eds.), The Middle Paleolithic: Adaptation,
Behavior, and Variability. The University Museum, University of Pennsylvania, Philadelphia, pp. 1–28.
Dogandžić, T., 2023. The Middle Paleolithic of the Balkans: Industrial Variability, Human Biogeography, and
Neanderthal Demise. Journal of World Prehistory 36, 257–338. https://doi.org/10.1007/s10963-023-09179-1
Dogandžić, T., Đuričić, L., 2017. Lithic production strategies in the Middle Paleolithic of the southern Bal-
kans. Quaternary International 450, 68–102. https://doi.org/10.1016/j.quaint.2017.03.011
Dragosavac, S., 2022. Laminar technology at the Middle to the Upper Paleolithic transition in the Central Bal-
kans. PhD thesis, Faculty of Philosophy, University of Belgrade, Serbia. (in Serbian, English summary)
Dragosavac, S., 2023. The Aurignacian in northern Bosnia revisited. Studia Praehistorica 17, 1–16. https://doi.
org/10.53250/stprae17.1-16
Đuričić, Lj., 2006. A contribution to research on Bioce Mousterian. Journal of the Serbian Archaeological So-
ciety 22, 179–196.
Faivre, J.-P., Discamps, E., Gravina, B., Turq, A., Guadelli, J.-L., Lenoir, M., 2014. The contribution of lit-
hic production systems to the interpretation of Mousterian industrial variability in south-western France:
The example of Combe-Grenal (Dordogne, France). Quaternary International 350, 227–240. https://doi.
org/10.1016/j.quaint.2014.05.048
Fewlass, H., Talamo, S., Wacker, L., Kromer, B., Tuna, T., Fagault, Y., Bard, E., McPherron, S.J., Aldeias, V.,
Maria, R., Martisius, N.L., Paskulin, L., Režek, Ž., Sinet-Mathiot, V., Sirakova, S., Smith, G.M., Spasov, R.,
Welker, F., Sirakov, N.V., Tsanova, T., Hublin, J.-J., 2020. A 14C chronology for the Middle to Upper Pala-
eolithic transition at Bacho Kiro Cave, Bulgaria. Nature Ecology & Evolution 4(6), 794–801. https://doi.
org/10.1038/s41559-020-1136-3
Finlayson, C., Zollikofer, C., Ponce de León, M., Finlayson, G., Carrión, J.S., Finlayson, S., Giles Guzman, F.,
Shea, J., 2023. Close encounters vs. missed connections? A critical review of the evidence for Late Ple-
istocene hominin interactions in western Eurasia. Quaternary Science Reviews 319, 108307. https://doi.
org/10.1016/j.quascirev.2023.108307
French, J.C., Collins, C., 2015. Upper Palaeolithic population histories of Southwestern France: a comparison
of the demographic signatures of 14C date distributions and archaeological site counts. Journal of Archaeo-
logical Science 55, 122–134. https://doi.org/10.1016/J.JAS.2015.01.001
Frick, J.A., 2020. Reflections on the term Micoquian in Western and Central Europe. Change in criteria, chan-
ged deductions, change in meaning, and its significance for current research. Archaeological and Anthropo-
logical Sciences 12, 38. https://doi.org/10.1007/s12520-019-00967-5
Fu, Q., Hajdinjak, M., Moldovan, O.T., Constantin, S., Mallick, S., Skoglund, P., Patterson, N., Rohland, N.,
Lazaridis, I., Nickel, B., Viola, B., Prüfer, K., Meyer, M., Kelso, J., Reich, D., Pääbo, S., 2015. An Early Mo-
dern Human from Romania with a recent Neanderthal ancestor. Nature 524(7564), 216–219. https://doi.
org/10.1038/nature14558
Fu, Q., Posth, C., Hajdinjak, M., Petr, M., Mallick, S., Fernandes, D.M., Furtwängler, A., Haak, W., Meyer, M.,
Mittnik, A., Nickel, B., Peltzer, A., Rohland, N., Slon, V., Talamo, S., Lazaridis, I., Lipson, M., Mathieson,
116 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
I., Schiffels, S., Skoglund, P., Derevianko, A.P., Drozdov, N., Slavinsky, V., Tsybankov, A.A., Cremonesi,
R.G., Mallegni, F., Gély, B., Vacca, E., González Morales, M.R., Straus, L.G., Neugebauer-Maresch, C., Tes-
chler-Nicola, M., Constantin, S., Moldovan, O.T., Benazzi, S., Peresani, M., Coppola, D., Lari, M., Ricci,
S., Ronchitelli, A., Valentin, F., Thevenet, C., Wehrberger, K., Grigorescu, D.A., Rougier, H., Crevecoeur,
I., Flas, D., Semal, P., Mannino, M.A., Cupillard, C., Bocherens, H., Conard, N.J., Harvati, K., Moiseyev,
V., Drucker, D.G., Svoboda, J., Richards, M., Caramelli, D., Pinhasi, R., Kelso, J., Patterson, N.J., Krause, J.,
Pääbo, S., Reich, D., 2016. The genetic history of Ice Age Europe. Nature 534(7606), 200–205. https://doi.
org/10.1038/nature17993
Galanidou, N., Papoulia, C., Ligkovanlis, S., 2016. The middle Palaeolithic bifacial tools from Megalo Karvou-
nari. In: B. Forsén, N. Galanidou and E. Tikkala (Eds.), Thesprotia expedition III. Landscapes of nomadism
and sedentism. Helsinki, pp. 29–58.
Geneste, J.-M., Jaubert, J., Lenoir, M., Meignen, L., Turq, A., 1997. Approche technologique des Moustériens
Charentiens du Sud-Ouest de la France et du Languedoc oriental. Paléo, Revue d’Archéologie Préhistorique
9(1), 101–142. http://dx.doi.org/10.3406/pal.1997.1230
Gerometta, K., Boschian, G., 2022. Stočari i špilje u Hrvatskoj – novi geoarheološki podaci iz špiljskih sedime-
nata. Vjesnik Arheološkog muzeja u Zagrebu 55(1), 9–41. https://doi.org/10.52064/vamz.55.1.1
Giaccio, B., Hajdas, I., Isaia, R., Deino, A.L., Nomade, S., 2017. High-precision 14C and 40Ar/39Ar dating of the
Campanian Ignimbrite (Y-5) reconciles the time-scales of climatic-cultural processes at 40 ka. Scientific
Reports 7, 45940. https://doi.org/10.1038/srep45940
Griffiths, H.I., Krystufek, B., Reed, J.M. (Eds.), 2004. Balkan biodiversity: Pattern and process in the European
hotspot. Kluwer, Dordrecht.
Gurova, M., Stefanova, E., Andreeva, P., Kecheva, N., Bonsall, C., 2022. Prehistoric Flint Raw Materials and
Artefacts from Bulgaria: The Next Step in Provenancing Balkan Flint. Archaeologia Bulgarica ХХVI(3),
1–30.
Hedges, R.E.M., Housley, R.A., Bronk, C.R., van Klinken, G.J., 1990. Radiocarbon Dates from the Oxford AMS
System: Archaeometry Datelist 11. Archaeometry 32(2), 211–237. https://doi.org/10.1111/j.1475-4754.1990.
tb00468.x
Hershkovitz, I., May, H., Sarig, R., Pokhojaev, A., Grimaud-Hervé, D., Bruner, E., Fornai, C., Quam, R.M., Ar-
suaga, J.L., Krenn, V.A., Martinón-Torres, M., de Castro, J.M., Martín-Francés, L., Slon, V., Albessard-Ball,
L., Vialet, A., Schüler, T., Manzi, G., Profico, A., Di Vincenzo, F., Weber, G.W., Zaidner, Y., 2021. A Midd-
le Pleistocene Homo from Nesher Ramla, Israel. Science 372(6549), 1424–1428. https://doi.org/10.1126/
science.abh3169
Heydari, M., Guérin, G., Sirakov, N.V., Fernandez, P., Ferrier, C., Guadelli, A., Leblanc, J.-C., Taneva, S.K.,
Sirakova, S., Guadelli, J.-L., 2022. The last 30,000 to 700,000 years ago: Unravelling the timing of human
settlement for the Palaeolithic site of Kozarnika. Quaternary Science Reviews 291, 107645. https://doi.or-
g/10.1016/j.quascirev.2022.107645
Hewitt, G.M., 2011. Mediterranean peninsulas: The evolution of hotspots. In: F.E. Zachos and J.C. Habel
(Eds.), Biodiversity hotspots distribution and protection of conservation priority areas. Springer, Berlin, pp.
123–147.
Higham, T., Compton, T., Stringer, C.B., Jacobi, R.M., Shapiro, B., Trinkaus, E., Chandler, B., Gröning, F.,
Collins, C., Hillson, S., O’Higgins, P., Fitzgerald, C., Fagan, M.J., 2011. The earliest evidence for anato-
mically modern humans in northwestern Europe. Nature 479(7374), 521–524. http://dx.doi.org/10.1038/
nature10484
Higham, T., Frouin, M., Douka, K., Ronchitelli, A., Boscato, P., Benazzi, S., Crezzini, J., Spagnolo, V., McCarty,
M., Marciani, G., Falcucci, A., Rossini, M., Arrighi, S., Dominici, C., Devièse, T., Schwenninger, J.-L., Mar-
tini, I., Moroni, A., Boschin, F., 2024. Chronometric data and stratigraphic evidence support discontinuity
between Neanderthals and early Homo sapiens in the Italian Peninsula. Nature Communications 15, 8016.
https://doi.org/10.1038/s41467-024-51546-9
Hiscock, P., Turq, A., Faivre, J.-P., Bourguignon, L., 2009. Quina procurement and tools production. In: B.
Adams and B.S. Blades (Eds.), Lithic Materials and Paleolithic Societies. Wiley, Singapore, pp. 232–246.
Hublin, J.-J., Sirakov, N.V., Aldeias, V., Bailey, S.E., Bard, E., Delvigne, V., Endarova, E., Fagault, Y., Fewlass,
H., Hajdinjak, M., Kromer, B., Krumov, I., Marreiros, J., Martisius, N.L., Paskulin, L., Sinet-Mathiot, V.,
Meyer, M., Pääbo, S., Popov, V.V., Režek, Ž., Sirakova, S., Skinner, M.M., Smith, G.M., Spasov, R., Talamo,
S., Tuna, T., Wacker, L., Welker, F., Wilcke, A., Zahariev, N., McPherron, S.J., Tsanova, T., 2020. Initial
Upper Palaeolithic Homo sapiens from Bacho Kiro Cave, Bulgaria. Nature 581(7808), 299–302. https://doi.
org/10.1038/s41586-020-2259-z
Ivanova, S., 1979. Cultural Differentiation in the Middle Palaeolithic of Balkan Peninsula. In: J.K. Kozłowski
(Ed.), Middle and Early Upper Palaeolithic in Balkans. Zeszyty Naukowe Uniwersytetu Jagiellonskiego,
Warszawa–Krakow, pp. 13–33.
References | 117
Jambrina-Enríquez, M., Herrera-Herrera, A.V., Vera, C.R., Leierer, L., Connolly, R., Mallol, C., 2019. n-Alkyl
nitriles and compound-specific carbon isotope analysis of lipid combustion residues from Neanderthal
and experimental hearths: Identifying sources of organic compounds and combustion temperatures. Qua-
ternary Science Reviews 222, 105899. https://doi.org/10.1016/j.quascirev.2019.105899
Jelinek, A.J., Farrand, W.R., Haas, G., Horowitz, A., Goldberg, P., 1973. New excavations at the Tabun Cave,
Mount Carmel, Israel, 1967-1972: a preliminary report. Paléorient Année 1973 1–2, 151–183.
Jéquier, C., Nicosia, C., Peresani, M., Romandini, M., Delpiano, D., Joannes-Boyau, R., Lembo, G., Livrachi,
A., López-García, J.M., Obradović, M., 2015. The De Nadale Cave, a single layered Quina Mousterian site
in the North of Italy. Quartär 62, 7–21. https://doi.org/10.7485/QU62_1
Jones, D.S., Monnier, G., Cooper, A., Baković, M., Pajović, G., Borovinić, N., Tostevin, G., 2020. Applying
high-throughput rRNA gene sequencing to assess microbial contamination of a 40-year old exposed archa-
eological profile. Journal of Archaeological Science 126, 105308. https://doi.org/10.1016/J.JAS.2020.105308
Kaczanowska, M., Kozłowski, J.K., Sobczyk, K., 2010. Upper Palaeolithic human occupations and material
culture at Klissoura Cave 1. Eurasian Prehistory 7(2), 133–285.
Karavanić, I., Golubić, V., Kurtanjek, D., Šošic, R., Zupanic, J., 2008. Liticka analiza materijala iz Mujine peći-
ne. Vjesnik za arheologiju i povijest dalmatinsku 101, 29–58.
Karavanić, I., Vukosavljević, N., Janković, I., Ahern, J.C.M., Smith, F.H., 2018. Paleolithic hominins and settle-
ment in Croatia from MIS 6 to MIS 3: Research history and current interpretations. Quaternary Internati-
onal 494, 152–166. https://doi.org/10.1016/j.quaint.2017.09.034
Karavanić, I., Banda, M., 2023. The Middle Palaeolithic of South-Eastern Europe. In: A. Ruiz-Redondo and
W. Davies (Eds.), The Prehistoric Hunter-Gatherers of South-Eastern Europe. Proceedings of the British
Academy, 258. Oxford University Press, Oxford, pp. 60–106. http://dx.doi.org/10.1093/9780197267509.0
03.0004
Kozłowski, J.K., 1992. The Balkans in the Middle and Upper Palaeolithic: The Gate to Europe or a Cul-de-sac?
Proceedings of the Prehistoric Society 58(1), 1–20. https://doi.org/10.1017/S0079497X00004059
Kozłowski, J.K., 2002. The Middle and the Early Upper Paleolithic around the Black Sea. In: T. Akazawa, K.
Aoki, and O. Bar-Yosef (Eds.), Neandertals and Modern Humans in Western Asia. Kluwer Academic Publi-
shers, New York, pp. 461–482.
Kozłowski, J.K., 2016. Taxonomy of the Early Middle Palaeolithic in Central Europe. Litikum – Journal of the
Lithic Research Roundtable 4, 23–32. https://doi.org/10.23898/litikuma0016
Kuhn, S.L., 1991. “Unpacking” reduction: lithic raw material economy in the Mousterian of West-Central
Italy. Journal of Anthropological Archaeology 10, 76–106.
Lambrecht, G., Rodríguez de Vera, C., Jambrina-Enríquez, M., Crevecoeur, I., González-Urquijo, J., Lazuén,
T., Monnier, G., Pajović, G., Tostevin, G., Mallol, C., 2021. Characterisation of charred organic matter in
micromorphological thin sections by means of Raman spectroscopy. Archaeological and Anthropological
Sciences 13, 13. https://doi.org/10.1007/s12520-020-01263-3
Laplace, G., 1972. La typologie analytique et structurale: base rationnelle d’étude des industries lithiques et
osseuses. In Banques de données archéologiques, Colloques nationaux du CNRS, pp. 99–142.
Lombao, D., Cueva-Temprana, A., Mosquera, M., Morales, J.I., 2020. A new approach to measure reduction
intensity on cores and tools on cobbles: the Volumetric Reconstruction Method. Archaeological and Ant-
hropological Sciences 12, 222. https://doi.org/10.1007/s12520-020-01154-7
Malez, M., 1975. Kvartarna fauna Crvene Stijene. In: Đ. Basler (Ed.), Crvena Stijena – zbornik radova. Zajed-
nica kulturnih ustanova, Nikšić, pp. 147–169.
Marín-Arroyo, A.B., Mihailovic, B., 2017. The Chronometric Dating and Subsistence of Late Neanderthals
and Early Anatomically Modern Humans in the Central Balkans: Insights from Šalitrena Pećina (Mionica,
Serbia). Journal of Anthropological Research 73(3), 413–447. https://doi.org/10.1086/693054
Marković, S.B., Vandenberghe, J., Stevens, T., Mihailović, D., Gavrilov, M.B., Radaković, M.G., Zeeden, C.,
Obreht, I., Perić, Z.M., Nett, J.J., Lehmkuhl, F., 2020. Geomorphological evolution of the Petrovaradin
Fortress Palaeolithic site (Novi Sad, Serbia). Quaternary Research 103, 21–34. https://doi.org/10.1017/
qua.2020.88
Meignen, L., Goldberg, P., Bar-Yosef, O., 2007. The Hearths at Kebara Cave and Their Role in Site Formation
Processes. In: O. Bar-Yosef and L. Meignen (Eds.), Kebara Cave, Mt Carmel, Israel: The Middle and Upper
Paleolithic Archaeology, Part I. American School of Prehistoric Research, Peabody Museum, Harvard Uni-
versity Press, pp. 91–122.
Mercier, N., Rink, W. J., Rodrigues, K., Morley, M., Van der Linden, M., Whallon, R., 2017. Radiometric da-
ting of the Crvena stijena sequence. In: R. Whallon (Ed.), Crvena Stijena in Cultural and Ecological Con-
text. Multidisciplinary Archaeological Research in Montenegro. Montenegrin Academy of Sciences and Arts,
National Museum of Montenegro, Podgorica, pp. 140–149.
118 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Mester, Z., Salvador, P.-G., Szolyák, P., Lamotte, A., 2020. The middle Palaeolithic Bábonyan of North-East
Hungary in the light of recent excavation of the eponymous site at Sajóbábony. Late neanderthals of the
middle Danube basin in central european context: cutural variability, interregional contacts, developmental
capacities, EAA european archaeologist association, Aug 2020, Budapest, Hungary. hal-04406728
Mihailović, D., 2009a. Upper Palaeolithic and Mesolithic Chipped Stone Industries from Crvena Stijena. Faculty
of Philosophy, Center for Archaeological Research, Belgrade.
Mihailović, D., 2009b. Middle Paleolithic settlement at Petrovaradin Fortress. Petrovaradin edition, Vol. II. City
Museum of Novi Sad, Novi Sad.
Mihailović, D., 2014. Paleolit na centralnom Balkanu – Kulturne promene i populaciona kretanja. Srpsko arhe-
ološko društvo, Beograd.
Mihailović, D. 2017. Paleolithic-Mesolithic Crvena Stijena in Relation to Other Sites. In: R. Whallon (Ed.),
In: R. Whallon (Ed.), Crvena Stijena in Cultural and Ecological Context. Multidisciplinary Archaeological
Research in Montenegro. Montenegrin Academy of Sciences and Arts, National Museum of Montenegro,
Podgorica, pp. 205–229.
Mihailović, D., 2020. Push-and-pull factors of the Middle to Upper Paleolithic transition in the Balkans. Qua-
ternary International 551, 47–62. https://doi.org/10.1016/j.quaint.2019.10.010
Mihailović, D., Mihailović, B., forthcoming. Territorial differentiation of Paleolithic technocomplexes in the
Balkans. Zbornik Narodnog muzeja – arheologija.
Mihailović, D., Whallon, R., 2017. Crvena Stijena revisited: The Late Mousterian assemblages. Quaternary
International 450, 36–49. https://doi.org/10.1016/j.quaint.2016.12.026
Mihailović, D., Mihailović, B., Whallon, R. (in press-a), The Question of Neanderthal Refuge in the Balkans.
In: M. Galaty et alii (Eds.), Oxford Handbook of Balkan Prehistory.
Mihailović, B., Mihailović, D., Dragosavac, S., Marković, J., Pajović, D., Silva, L.O., Skinner, A., Roksandic, M.
(in press-b). Late Middle Paleolithic and Early Upper Paleolithic of the Western Balkans: lithic assembla-
ges from Šalitrena Pećina (Serbia). Quaternary International.
Mihailović, D., Mihailović, B., Whallon, R., 2017a. Excavations of Middle Paleolithic-Mesolithic layers. In: R.
Whallon (Ed.), In: R. Whallon (Ed.), Crvena Stijena in Cultural and Ecological Context. Multidisciplinary
Archaeological Research in Montenegro. Montenegrin Academy of Sciences and Arts, National Museum of
Montenegro, Podgorica, pp. 150–204.
Mihailović, D., Mihailović, B., Borovinić, N., 2017b. Newer Excavations – Archaeological Stratigraphy. In: R.
Whallon (Ed.), Crvena Stijena in Cultural and Ecological Context. Multidisciplinary Archaeological Research
in Montenegro. Montenegrin Academy of Sciences and Arts, National Museum of Montenegro, Podgorica,
pp. 132–139.
Mihailović, D., Kuhn, S.L., Bogićević, K., Dimitrijević, V., Marín-Arroyo, A.B., Marković, J., Mercier, N., Miha-
ilović, B., Morley, M.W., Radović, P., Rink, W.J., Plavšić, S., Roksandic, M., 2022a. Connections between the
Levant and the Balkans in the late Middle Pleistocene: Archaeological findings from Velika and Mala Bala-
nica Caves (Serbia). Journal of Human Evolution 163, 103138. https://doi.org/10.1016/j.jhevol.2021.103138
Mihailović, D., Milošević, S., Blackwell, B.A.B., Mercier, N., Mentzer, S.M., Miller, C.E., Morley, M.W., Bogi-
ćević, K., Đurić, D., Marković, J., Mihailović, B., Dragosavac, S., Plavšić, S., Skinner, A.R., Chaity, I.I.C.,
Huang, Y.E.W., Chu, S., Nenadić, D., Radović, P., Lindal, J., Roksandic, M., 2022b. Neanderthal settlement
of the Central Balkans during MIS 5: Evidence from Pešturina Cave, Serbia. Quaternary International 610,
1–19. https://doi.org/10.1016/j.quaint.2021.09.003
Mihailović, D., Kuhn, S., Skinner, A., Blackwell, B., Bogićević, K., Dragosavac, S., Pajović, D., Petrović, A.,
Roksandic, M. 2024. The emergence of Levallois technology in the Middle Palaeolithic of the Balkans: lit-
hic assemblages from Velika Balanica and Selačka Cave 3. Abstracts 2024 ESHE Zagreb, European Society
for Human Evolution, Zagreb, p. 116.
Milošević, S., 2020. Competition between Humans and Large Carnivores: Case Studies from the Late Middle and
Upper Palaeolithic of the Central Balkans. BAR International Series 2961, Oxford.
Miracle, T.P., 2007. The Krapina Paleolithic Site: Zooarchaeology, taphonomy and catalog of the faunal remains.
Croatian Natural History Museum, Zagreb.
Moncel, M.-H., Ashton, N.J., Arzarello, M., Fontana, F., Lamotte, A., Scott, B., Muttillo, B., Berruti, G.L., Nen-
zioni, G., Tuffreau, A., Peretto, C., 2020. Early Levallois core technology between Marine Isotope Sta-
ge 12 and 9 in Western Europe. Journal of Human Evolution 139, 102735. https://doi.org/10.1016/j.jhe-
vol.2019.102735
Monnier, G., Tostevin, G., Pajović, G., Borovinić, N., Baković, M., 2020. Nova istraživanja paleolitskog nalazi-
šta Crvena Stijena, istoijski kontekst. Istorijski Zapisi XCIII(1-2), 71–108.
Montet White, A., Laville, H., Lezine A.-M., 1986. Le Paléolithique du Bosnie du Nord. Chronologie, environ-
ment et préhistoire. L’Anthropologie 90(1), 29–88.
References | 119
Morin, E. and Soulier, M.-C. 2017. The Paleolithic faunal remains from Crvena Stijena.In: R. Whallon (Ed.),
Crvena Stijena in Cultural and Ecological Context. Multidisciplinary Archaeological Research in Montenegro.
Montenegrin Academy of Sciences and Arts, National Museum of Montenegro, Podgorica, pp. 266–294.
Morley, M., 2007. Mediterranean Quaternary Rockshelter Sediment Records: a Multi-proxy Approach to Envi-
ronmental Reconstruction. Ph.D. Dissertation. The University of Manchester, Manchester.
Morley, M.W., 2017. The geoarchaeology of Crvena Stijena: site formation processes, palaeoenvironments and
hominin activity. In: R. Whallon (Ed.), Crvena Stijena in Cultural and Ecological Context. Multidisciplinary
Archaeological Research in Montenegro. Montenegrin Academy of Sciences and Arts, National Museum of
Montenegro, Podgorica, pp. 82–132.
Morley, M.W., Woodward, J.C., 2011. The Campanian Ignimbrite (Y5) tephra at Crvena Stijena Rockshelter,
Montenegro. Quaternary Research 75, 683–696. https://doi.org/10.1016/j.yqres.2011.02.005
Moroni, A., Boscato, P., Ronchitelli, A., 2013. What roots for the Uluzzian? Modern behaviour in Central-So-
uthern Italy and hypotheses on AMH dispersal routes. Quaternary International 316, 27–44. https://doi.
org/10.1016/J.QUAINT.2012.10.051
Moroni, A., Boscato, P., Ronchitelli, A., 2016. The origin of the Uluzzian: the upsetting of a paradigm? Bulletin
de Musée d’Anthropologie préhistorique de Monaco 56, 136–139.
Moroni, A., Ronchitelli, A., Arrighi, S., Aureli, D., Bailey, S.E., Boscato, P., Boschin, F., Capecchi, G., Crezzini,
J., Douka, K., Marciani, G., Panetta, D., Ranaldo, F., Ricci, S., Scaramucci, S., Spagnolo, V., Benazzi, S.,
Gambassini, P., 2018. Grotta del Cavallo (Apulia – Southern Italy). The Uluzzian in the mirror. Journal of
Anthropological Sciences 96, 125–160. https://doi.org/10.4436/JASS.96004
Mylopotamitaki, D., Weiss, M., Fewlass, H., Zavala, E.I., Rougier, H., Sümer, A.P., Hajdinjak, M., Smith, G.M.,
Ruebens, K., Sinet-Mathiot, V., Pederzani, S., Essel, E., Harking, F., Xia, H., Hansen, J., Kirchner, A., Lauer,
T., Stahlschmidt, M., Hein, M., Talamo, S., Wacker, L., Meller, H., Dietl, H., Orschiedt, J., Olsen, J.V., Ze-
berg, H., Prüfer, K., Krause, J., Meyer, M., Welker, F., McPherron, S.J., Schüler, T., Hublin, J.-J., 2024. Homo
sapiens reached the higher latitudes of Europe by 45,000 years ago. Nature 626(7998), 341–346. https://doi.
org/10.1038/s41586-023-06923-7
Müller, U.C., Pross, J., Tzedakis, P.C., Gamble, C., Kotthoff, U., Schmiedl, G., Wulf, S., Christianis, K., 2011.
The role of climate in the spread of modern humans into Europe. Quaternary Science Reviews 30(3–4),
273–279. https://doi.org/10.1016/j.quascirev.2010.11.016
Nelson, M.C., 1991. The Study of Technological Organization. Journal of Archaeological Method and Theory
3, 57–100.
Nigst, P.R., Haesaerts, P., Damblon, F., Frank-Fellner, C., Mallol, C., Viola, B., Götzinger, M., Niven, L., Trnka,
G., Hublin, J.-J., 2014. Early modern human settlement of Europe north of the Alps occurred 43,500 years
ago in a cold steppe-type environment. Proceedings of the National Academy of Sciences of the United States
of America 111(40), 14394–14399. https://doi.org/10.1073/pnas.1412201111
Pamić, J., 1975. Mineralni sastav i petrografija artefakata iz Crvene Stijene. In: Đ. Basler (Ed.), Crvena Stijena
– zbornik radova. Zajednica kulturnih ustanova, Nikšić, pp. 205–209.
Panagopoulou, E. 1999. The Theopetra Middle Palaeolithic assemblages: Their relevance to the Middle Pala-
eolithic of Greece and adjacent areas. In: G.N. Bailey, E. Adam, E. Panagopoulou, C. Perlès and K. Zachos
(Eds.), The Palaeolithic archaeology of Greece and adjacent areas. Proceedings of the ICoPAG Conference,
Ioannina, September 1994, Vol. 3. London, British School at Athens Studies, pp. 252–265.
Papaconstantinou, E., 1989. Micromousterien, les idées et les pierres: Asprochaliko (Grèce) et le problème des
industries microlithiques du Moustérien. Université Paris X-Nanterre.
Pederzani, S., Britton, K., Aldeias, V., Bourgon, N., Fewlass, H., Lauer, T., McPherron, S.J., Režek, Ž., Sirakov,
N.V., Smith, G.M., Spasov, R., Tran, N., Tsanova, T., Hublin, J.-J., 2021. Subarctic climate for the earliest
Homo sapiens in Europe. Science Advances 7(39), eabi4642. https://doi.org/10.1126/sciadv.abi4642
Peresani, M., Cristiani, E., Romandini, M., 2016. The Uluzzian technology of Grotta di Fumane and its impli-
cation for reconstructing cultural dynamics in the Middle–Upper Palaeolithic transition of Western Eura-
sia. Journal of Human Evolution 91, 36–56. https://doi.org/10.1016/j.jhevol.2015.10.012
Peresani, M., Bertola, S., Delpiano, D., Benazzi, S., Romandini, M., 2019. The Uluzzian in the north of Italy:
insights around the new evidence at Riparo Broion. Archaeological and Anthropological Sciences 11, 3503–
3536. https://doi.org/10.1007/s12520-018-0770-z
Peresani, M., Delpiano, D., Hohenstein, U.T., Arnetta, G., Blain, H.-A., Espedito, V., López-García, J.M., Poti,
A., Sauro, U., Visentin, D., Chelidonio, G., Margaritora, D., Tafelmaier, Y., Beltrame, L., Failla, C., Pellegri-
ni, F., Pellegrini, P., 2022. The Middle Palaeolithic site of Vajo Salsone in the Monti Lessini, Italian Alps.
First report on the archaeofauna and lithic assemblage with foliate tools. Alpine and Mediterranean Qua-
ternary 35(1), 5–29. http://dx.doi.org/10.26382/AMQ.2022.01
120 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Peresani, M., Bourguignon, L., Delpiano, D., Lemorini, C., 2023. Quina on the edge. Insights from a Midd-
le Palaeolithic lithic assemblage of Grotta di Fumane, Italy. Journal of Archaeological Science: Reports 49,
103998. https://doi.org/10.1016/j.jasrep.2023.103998
Pesesse, D., 2017. Is it still appropriate to talk about the Gravettian? Data from lithic industries in Western
Europe. Quartär 64, 107–128. https://doi.org/10.7485/QU64_5
Picin, A., 2021. At the onset of the Micoquian in Central Europe: raw material constraints and technological
versatility at Neumark-Nord 2/0 (Germany). Materiale şi cercetãri arheologice (Serie nouã) H-S, 71–84.
https://doi.org/10.3406/mcarh.2021.2203
Picin, A., Peresani, M., Falguères, C., Gruppioni, G., Bahain, J., 2013. San Bernardino Cave (Italy) and the
Appearance of Levallois Technology in Europe: Results of a Radiometric and Technological Reassessment.
PLoS ONE 8(10), e76182. http://dx.doi.org/10.1371/journal.pone.0076182
Prévost, M., Zaidner, Y., 2020. New insights into early MIS 5 lithic technological behavior in the Levant: Nes-
her Ramla, Israel as a case study. PLoS ONE 15(4), e0231109. https://doi.org/10.1371/journal.pone.0231109
Richter, D., Tostevin, G., Škrdla, P., Davies, W., 2009. New radiometric ages for the Early Upper Palaeolithic
type locality of Brno-Bohunice (Czech Republic): Comparison of OSL, IRSL, TL and 14C dating results.
Journal of Archaeological Science 36(3), 708–720. https://doi.org/10.1016/J.JAS.2008.10.017
Richter, J., 2000. Social memory among late Neanderthals. In: J. Orschiedt and G.-C. Weniger (Eds.), Nean-
derthals and Modern Humans – Discussing the Transition. Central and Eastern Europe from 50.000 – 30.000
B . P. Universität zu Köln, pp. 123–132.
Rodríguez de Vera, C., Herrera-Herrera, A.V., Jambrina-Enríquez, M., Sossa-Ríos, S., González-Urquijo,
J., Lazuén, T., Vanlandeghem, M., Alix, C., Monnier, G., Pajović, G., Tostevin, G., Mallol, C., 2020. Mi-
cro-contextual identification of archaeological lipid biomarkers using resin-impregnated sediment slabs.
Scientific Reports 10, 20574. https://doi.org/10.1038/s41598-020-77257-x
Roksandic, M., Radović, P., Lindal, J., 2023. The complex picture of the Chibanian hominin record at the cro-
ssroads of Europe and Asia. In: A. Ruiz-Redondo and W. Davies (Eds.), The Prehistoric Hunter-Gatherers
of South-Eastern Europe. Proceedings of the British Academy, 258. Oxford University Press, Oxford, pp.
36–59. http://dx.doi.org/10.1093/9780197267509.003.0003
Rolland, N., Dibble, H.L., 1990. A New Synthesis of Middle Paleolithic Variability. American Antiquity 55(3),
480–499. https://doi.org/10.2307/281279
Romagnoli, F., Chabai, V., Gravina, B., Hérisson, D., Hovers, E., Moncel, M.-H., Peresani, M., Uthmeier, T.,
Bourguignon, L., Chacón, M.G., Di Modica, K., Faivre, J.-P., Kolobova, K., Malinsky-Buller, A., Neruda, P.,
Rios Garaizar, J., Weiss, M., Wiśniewski, A., Wragg Sykes, R., 2022. Chapter 10 – Neanderthal technologi-
cal variability: A wide-ranging geographical perspective on the final Middle Palaeolithic. In: F. Romagnoli,
F. Rivals and S. Benazzi (Eds.), Updating Neanderthals. Understanding Behavioural Complexity in the Late
Middle Palaeolithic. Academic Press, Cambridge, pp. 163–205.
Ronchitelli, A.M., Moroni, A., Boscato, P., Gambassini, P., 2018. The Uluzzian 50 years later. In: K.S.P. Val-
de-Nowak (Ed.), Multa per gentes et multa per saecula. Alter Radosław Palonka, Kraków, pp. 71–76.
Rossini, M., Marciani, G., Arrighi, S., Pereira Santos, M.C., Spagnolo, V., Ronchitelli, A., Benazzi, S., Moroni,
A., 2022. Less is more! Uluzzian technical behaviour at the cave site of Castelcivita (southern Italy). Journal
of Archaeological Science: Reports 44, 103494. https://doi.org/10.1016/j.jasrep.2022.103494
Ruka, R., Bassetti, M., Bertola, S., Delpiano, D., Gjipali, I., Peresani, M., 2023. Discoid technology in between:
The Middle Palaeolithic site of Istraishta in the Quaternary context of the Korça Basin, Albania. 29 Decem-
ber 2023, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-3799762/v1]
Sánchez-Yustos, P., Marín-Arroyo, A.B., Arnold, L.J., Luque, L., Kehl, M., López-Sáez, J.A., Carrancho Alon-
so, Á., Demuro, M., Sanz-Royo, A., Buckley, M., Maíllo-Fernández, J., Cuartero-Monteagudo, F., Llama-
zares-González, J., Ruiz-Alonso, M., Luelmo-Lautenschlaeger, R., García-Soto, E., Alcaraz-Castaño, M.,
2024. Initial Upper Palaeolithic lithic industry at Cueva Millán in the hinterlands of Iberia. Scientific Re-
ports 14, 21705. https://doi.org/10.1038/s41598-024-69913-3
Sano, K., Arrighi, S., Stani, C., Aureli, D., Boschin, F., Fiore, I., Spagnolo, V., Ricci, S., Crezzini, J., Boscato,
P., Gala, M., Tagliacozzo, A., Birarda, G., Vaccari, L., Ronchitelli, A., Moroni, A., Benazzi, S., 2019. The
earliest evidence for mechanically delivered projectile weapons in Europe. Nature ecology & evolution 3,
1409–1414. https://doi.org/10.1038/s41559-019-0990-3
Schmidt, C., Veres, D., Muratoreanu, G., Cosac, M., Niță, L., Vasile, Ş., Şerbănescu, G.S., Bartok, I., 2024.
Evidence for the oldest Middle Palaeolithic cave occupation in the Romanian Carpathians. Journal of Qu-
aternary Science, https://doi.org/10.1002/jqs.3667
Shalagina, A.V., Krivoshapkin, A.I., Kolobova, K.A., 2015. Truncated-faceted pieces in the Paleolit-
hic of northern Asia. Archaeology, Ethnology and Anthropology of Eurasia 43(4), 33–45. http://dx.doi.
org/10.17746/1563-0102.2015.43.4.033-045
References | 121
Shea, J.J., 2014. Sink the Mousterian? Named stone tool industries (NASTIES) as obstacles to investigating
hominin evolutionary relationships in the Later Middle Paleolithic Levant. Quaternary International 350,
169–179. https://doi.org/10.1016/j.quaint.2014.01.024
Shimelmitz, R., Kuhn, S.L., Ronen, A., Weinstein-Evron, M., 2014. Predetermined flake production at the
Lower/Middle Paleolithic boundary: Yabrudian scraperblank technology. PLoS ONE 9(9), e106293. https://
doi.org/10.1371/journal.pone.0106293
Sirakova, S., 2020. Middle Palaeolithic leaf points from Bulgaria and their analogous in Southeast and Central
Europe. Sofia.
Sitlivy, V., Sobczyk, K., Karkanas, P., Koumouzelis, M., 2008. Middle Palaeolithic industries of Klissoura cave,
Greece. In: A. Darlas and D. Mihailović (Eds.), The Palaeolithic of the Balkans. BAR International Series
1891, Oxford, pp. 39–50.
Slimak, L., Zanolli, C., Higham, T., Frouin, M., Schwenninger, J.-L., Arnold, L.J., Demuro, M., Douka, K., Mer-
cier, N., Guérin, G., Valladas, H., Yvorra, P., Giraud, Y., Seguin-Orlando, A., Orlando, L., Lewis, J.E., Muth,
X., Camus, H., Vandevelde, S., Buckley, M., Mallol, C., Stringer, C.B., Metz, L., 2022. Modern human in-
cursion into Neanderthal territories 54,000 years ago at Mandrin, France. Science Advances 8(6), eabj9496.
https://doi.org/10.1126/sciadv.abj9496
Starkovich, B.M., 2014. Optimal foraging, dietary change, and site use during the Paleolithic at Klissoura Cave
1 (southern Greece). Journal of Archaeological Science 52, 39–55. https://doi.org/10.1016/j.jas.2014.08.026
Stepanchuk, V., Sapozhnikov, I., 2010. Middle and Upper Pleistocene of Ukraine: Synopsis of Palaeolithic finds
with special reference to patterns of peopling and cultural development. Arheologia Moldovei 33, 13–24.
Sümer, A.P., Rougier, H., Villalba-Mouco, V., Huang, Y., Iasi, L.N., Essel, E., Mesa, A.B., Furtwaengler, A.,
Peyrégne, S., de Filippo, C., Rohrlach, A.B., Pierini, F., Mafessoni, F., Fewlass, H., Zavala, E.I., Mylopota-
mitaki, D., Bianco, R.A., Schmidt, A., Zorn, J., Nickel, B., Patova, A., Posth, C., Smith, G.M., Ruebens, K.,
Sinet-Mathiot, V., Stoessel, A., Dietl, H., Orschiedt, J., Kelso, J., Zeberg, H., Bos, K.I., Welker, F., Weiss, M.,
McPherron, S.J., Schüler, T., Hublin, J.-J., Velemínský, P., Brůžek, J., Peter, B.M., Meyer, M., Meller, H., Rin-
gbauer, H., Hajdinjak, M., Prüfer, K., Krause, J., 2024. Earliest modern human genomes constrain timing
of Neanderthal admixture. Nature. https://doi.org/10.1038/s41586-024-08420-x
Taneva, S., 2021. On the Middle Palaeolithic leaf points from Bulgaria (South East Europe). Materiale şi cer-
cetări arheologice (Serie nouã) H-S, 125–139. http://dx.doi.org/10.3406/mcarh.2021.2206
Terradas, X., 2003. Discoid flaking method: conception and technological variability. In: M. Peresani (Ed.),
Discoid Lithic Technology: Advances and Implications. BAR International Series 1120, Oxford, pp. 19–31.
Thiébaut, C., 2005. Le Moustérien à denticulés: Variabilité ou diversité techno-économique ? Université
d’Aix-Marseille I – Université de Provence.
Thiébaut, C., 2010. Denticulate Mousterian: Myth or reality? Acta Universitatis Wratislaviensis 3207, 345–386.
Tixier, J., Turq, A., 1999. Kombewa et alii. Paléo, Revue d’Archéologie Préhistorique 11, 135–143. https://doi.
org/10.3406/pal.1999.1174
Tourloukis, V., Harvati, K. 2018. The Palaeolithic record of Greece: a synthesis of the evidence and a research
agenda for the future. Quaternary International 466(A), 48–65. https://doi.org/10.1016/j.quaint.2017.04.020
Turq, A., 1989. Approche technologique et économique du faciès Moustérien de type Quina : Étude prélimi-
naire. Bulletin de la Société préhistorique française 86(8), 244–256. https://doi.org/10.3406/bspf.1989.9390
Villa, P., Pollarolo, L., Conforti, J., Marra, F., Biagioni, C., Degano, I., Lucejko, J.J., Tozzi, C., Pennacchioni, M.,
Zanchetta, G., Nicosia, C., Martini, M., Sibilia, E., Panzeri, L., 2018. From Neandertals to modern humans:
New data on the Uluzzian. PLoS ONE 13(5), e0196786. https://doi.org/10.1371/journal.pone.0196786
Vishnevskiy, A.V., Pavlenok, K.K., Kozlikin, M.B., Ulyanov, V.A., Derevianko, A.P., Shunkov, M.V., 2019. A
Neanderthal Refugium in the Eastern Adriatic. Archaeology, Ethnology and Anthropology of Eurasia 47(4),
3–15. https://doi.org/10.17746/1563-0110.2019.47.4.003-015
Vogel, J.C., Waterbolk, H.T., 1972. Groningen Radiocarbon Dates X. Radiocarbon 14(1), 6–110. https://doi.
org/10.1017/S0033822200001016
Vujević, D., Perhoč, Z., Ivančić, T., 2017. Micro-Mousterian in Northern Dalmatia. Quaternary International
450, 50–67. https://doi.org/10.1016/j.quaint.2016.11.019
Weiss, M., 2020. The Lichtenberg Keilmesser – it’s all about the angle. PLoS ONE 15(10), e0239718. https://
doi.org/10.1371/journal.pone.0239718
Weiss, M., Lauer, T., Wimmer, R., Pop, C.M., 2018. The Variability of the Keilmesser-Concept: a Case Study from
Central Germany. Journal of Paleolithic Archaeology 1, 202–246. https://doi.org/10.1007/s41982-018-0013-y
Whallon, R., 1999. The lithic tool assemblages at Badanj within their regional context. In: G. Bailey, E. Adam,
C. Perlès, E. Panagopoulou and K. Zachos (Eds.), The Palaeolithic archaeology of Greece and adjacent are-
as. Proceedings of the ICoPAG Conference, Ioannina, September 1994, Vol. 3. London, British School at
Athens Studies, pp. 330–342.
122 | Middle Paleolithic assemblages from Crvena Stijena: insights from the 1954–1963 excavations
Whallon, R., Morin, E. 2017. Eleven years of research at Crvena Stijena: Synthesis of the results. In: Crvena
Stijena in Cultural and Ecological Context. In: R. Whallon (Ed.), Crvena Stijena in Cultural and Ecologi-
cal Context. Multidisciplinary Archaeological Research in Montenegro. Montenegrin Academy of Sciences
and Arts, National Museum of Montenegro, Podgorica, pp. 450–455.
Zaidner, Y., Weinstein-Evron, M., 2016. The end of the Lower Paleolithic in the Levant: The Acheulo-Ya-
brudian lithic technology at Misliya cave, Israel. Quaternary International 409(B), 9–22. https://doi.or-
g/10.1016/j.quaint.2015.12.080
CIP – Каталогизација у публикацији
Народна библиотека Србије, Београд
903.01”6325/633”(497.16)
MIHAILOVIĆ, Dušan, 1962–
Middle Paleolithic assemblages from Crvena Stijena :
insights from the 1954–1963 excavations / Dušan Mihailović.
- Belgrade : Faculty of Philosophy, University, Center for
Archaeological Research, 2024 (Belgrade : Dosije studio). –
122 str. : ilustr. ; 30 cm. – (University of Belgrade, Faculty of
Philosophy, Center for Archaeological Research ; vol. 22)
Tiraž 150. – Napomene i bibliografske reference uz tekst. –
Bibliografija: str. 113–122.
ISBN 978-86-6427-343-5
а) Археолошка налазишта -- Црвена стијена
б) Археолошки налази -- Црвена стијена
COBISS.SR-ID 161328649
ISBN 978-86-6427-343-5ISBN 978-86-6427-343-5
9 788664 273435 FACULTY OF PHILOSOPHY
UNIVERSITY OF BELGRADE
FACULTY OF PHILOSOPHY
UNIVERSITY OF BELGRADE
