ChapterPDF Available

The Upper Paleolithic Settlement of the Armenian Highlands

Authors:
  • Heidelberger Akademie der Wissenschaften

Abstract and Figures

Excavations in 2009 and 2010 at Aghitu–3 Cave in the Syunik Province of southern Armenia yield new insights into the Upper Paleolithic settlement of the Armenian Highlands. The site is situated at an elevation of 1601 m in a side valley of the Vorotan River. The river cuts down through Pleistocene basalt flows and provides a corridor for the movement of people and game through the region. Sediments that accumulated in this basaltic cave are composed mainly of silt, clay minerals and volcanic ash. The archaeological layers preserve evidence of periodic human occupations dating to ca. 35–27 000 cal BP. Caves from the Upper Paleolithic were not previously known in Armenia, although contemporaneous sites exist in neighboring Georgia and Iran. The lithic industry at Aghitu–3 is laminar with a strong focus on the production of bladelets made of obsidian and chert. While completely backed pieces are rare, the majority of tools are represented by finely retouched bladelets. The choice of raw material did not affect the desired end products. Our preliminary interpretation is that this distinctly Upper Paleolithic toolkit was oriented towards the production of hunting equipment and was technologically stable over an extended timeframe. The lower assemblage dates to ca. 35–31 000 cal BP and suggests sparse occupation of the cave. Lithic artifacts are few and cluster near small combustion features. The poorly preserved faunal remains of the lower layers do not appear to be associated with the lithic remains. The bones often appear to be gastrically etched, suggesting accumulation by large carnivores such as wolves. On the other hand, the upper assemblage dates to ca. 29–27 000 cal BP and indicates more frequent occupation by humans. In these finely stratified layers, lithic artifacts are numerous, and combustion features are common. The well preserved, but highly fragmented faunal remains from the upper layers exhibit more indications of carcass processing, such as green breaks and impact fractures. Wild sheep and wild goat dominate the faunal assemblage, with horse and hare also present. Combining the faunal identifications with ecological data gained from microfauna, pollen and charcoal, a mosaic landscape comes into focus: grassland on the level basaltic plateau, interrupted by a steep rocky valley sloping down to the Vorotan, where a riparian environment prevails. The data also suggest an environment that was cooler and moister than today, a picture echoed by preliminary micromorphological results showing cycles of freezing and thawing. Thus we interpret these data as evidence for increasing occupation of Aghitu–3 Cave, which served as a temporary hunting camp. While it is clear that the older occupations of the cave were ephemeral, during the time leading up to the last glacial maximum, occupation became more frequent.
No caption available
… 
No caption available
… 
No caption available
… 
Content may be subject to copyright.
5
2012
L
iège
commission
8
Paléolithique supérieur
Actes du Colloque international
de la commission 8 (Paléolithique supérieur) de l'UISPP
Université de Liège, 28–31 mai 2012
Marcel OTTE & Foni LE BRUN-RICALENS
(.)
Modes de contacts et de déplacements
au Paolithique eurasiatiQue
Modes of contact and mobility during the Eurasian Palaeolithic
2014

 '
Marcel Otte
Foni Le Brun-Ricalens
 '

, 
Mary Étienne
Éléonore Simonin

   
François Lacrampe-Cuyaubère (Archéosphère, France)
 Illustrations originales des divers auteurs participant à l'ouvrage
   Benoit Clarys
  Antartis
http://fr.123rf.com/profile_antartis
   Imprimerie Fr. Faber, Mersch, Grand-Duché de Luxembourg
©Éditions
Tous drois réservés
Reproduction interdite sans autorisation
©

 
  
Marcel Otte (ULg, Belgique)
Marcel.Otte@ulg.ac.be
Université de Liège – Place du XX août 7, bâtiment A1
B-4000 Liège, Belgique
©

 
  
Foni Le Brun-Ricalens (MNHA-CNRA, Luxembourg)
Foni.LeBrun@cnra.etat.lu
Centre National de Recherche Archéologique – 241, rue de Luxembourg
L-8077 Bertrange, Luxembourg
 978-2-87985-305-5
&
IF

T U P 
   H
Andrew W. KANDEL Boris GASPARYAN Samvel NAHAPETYAN
Andreas TALLER Lior WEISSBROD
Abstract: Excavations in 2009 and 2010 at Aghitu–3 Cave in the Syunik Province of southern Armenia yield new insights into
the Upper Paleolithic settlement of the Armenian Highlands. e site is situated at an elevation of 1601 m in a side valley of
the Vorotan River. e river cuts down through Pleistocene basalt flows and provides a corridor for the movement of people
and game through the region. Sediments that accumulated in this basaltic cave are composed mainly of silt, clay minerals and
volcanic ash. e archaeological layers preserve evidence of periodic human occupations dating to ca. 35–27 000 cal BP. Caves
from the Upper Paleolithic were not previously known in Armenia, although contemporaneous sites exist in neighboring
Georgia and Iran.
e lithic industry at Aghitu–3 is laminar with a strong focus on the production of bladelets made of obsidian and chert. While
completely backed pieces are rare, the majority of tools are represented by finely retouched bladelets. e choice of raw mate-
rial did not aect the desired end products. Our preliminary interpretation is that this distinctly Upper Paleolithic toolkit was
oriented towards the production of hunting equipment and was technologically stable over an extended timeframe.
e lower assemblage dates to ca. 35–31 000 cal BP and suggests sparse occupation of the cave. Lithic artifacts are few and
cluster near small combustion features. e poorly preserved faunal remains of the lower layers do not appear to be asso-
ciated with the lithic remains. e bones often appear to be gastrically etched, suggesting accumulation by large carnivores
such as wolves. On the other hand, the upper assemblage dates to ca. 29–27 000 cal BP and indicates more frequent occu-
pation by humans. In these finely stratified layers, lithic artifacts are numerous, and combustion features are common. e
well preserved, but highly fragmented faunal remains from the upper layers exhibit more indications of carcass processing,
such as green breaks and impact fractures. Wild sheep and wild goat dominate the faunal assemblage, with horse and hare
also present.
Combining the faunal identifications with ecological data gained from microfauna, pollen and charcoal, a mosaic landscape
comes into focus: grassland on the level basaltic plateau, interrupted by a steep rocky valley sloping down to the Vorotan,
where a riparian environment prevails. e data also suggest an environment that was cooler and moister than today, a picture
echoed by preliminary micromorphological results showing cycles of freezing and thawing. us we interpret these data as
evidence for increasing occupation of Aghitu–3 Cave, which served as a temporary hunting camp. While it is clear that the
older occupations of the cave were ephemeral, during the time leading up to the last glacial maximum, occupation became
more frequent.
Thème I
ASIE
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique

While our knowledge about the Paleolithic settlement of the Caucasus has grown
in the past decades, the archive of excavated and dated sites remains relatively
small. e Georgian Early Pleistocene site of Dmanisi established the region as
a focal point for research into the earliest human migrations out of Africa at
about 1.8 Ma (Gabunia et al. 2000; Lordkipanidze et al. 2005; Ferring et al. 2011).
Continued research programs at several Georgian sites, such as Ortvale Klde and
Dzudzuana Cave, has vastly improved our understanding of the Middle to Upper
Paleolithic transition of this region (e.g. Nioradze & Otte 2000; Bar-Oz & Adler
2005; Bar-Yosef et al. 2006, 2011; Adler et al. 2008; Moncel et al. 2012).
However, compared to Georgia, the picture in Armenia is just emerging. Lower
Paleolithic settlement is documented mainly by open-air sites with Oldowan-
like and Acheulean industries (Klein 1966; Golovanova 2000; Doronichev 2008;
Kolpakov 2009; Gasparian 2010). In northern Armenia, researchers are inves-
tigating the Middle Pleistocene of the Debed river valley (Egeland et al. 2010,
2011), while others are examining the Dzoraget river valley in the region of Lori
(Dolukhanov et al. 2004; Presnyakov et al. 2012). In central Armenia, Middle Pale-
olithic localities with late Mousterian occupations are known through excava-
tions at Yerevan (Yeritsian 1970) and Lusakert (Yeritsian 1975) caves, but the
dating of these sites remains unresolved. Recent work at Angeghakot–1 (Liagre
et al. 2006), Hovk–1 Cave (Pinhasi et al. 2008, 2011; Bar-Oz et al. 2012) and
Kalavan–2 (Ghukasyan et al. 2011) yielded three new Middle Paleolithic locali-
ties. Meanwhile, ongoing excavation projects led by Adler, Yeritsyan & Gasparyan
in the Hrazdan River Gorge, including Nor Geghi and renewed work at Lusakert,
is shedding light on the nature of the Armenian Lower and Middle Paleolithic
(Adler et al. 2009, 2012).
Despite this improved eort to study the Paleolithic, stratified Upper Paleolithic
sequences are extremely rare in Armenia, represented by a single Late Upper Pale-
olithic open-air site, Kalavan–1, dated to 17–16,000 cal BP (Chataigner et al. 2012).
us, the discovery of the Upper Paleolithic site of Aghitu–3 has significance not
only for the settlement of the Armenian Highlands, but also the Caucasus region.
 
In 2009 the field crew began its systemic excavations, orienting the measuring
grid to the 2 by 2 m test pit excavated by a French team in 2003. e team desig-
nated the coordinate system using letters for the x-axis and numbers for the
y-axis (figure 2). A datum point hammered into the rear wall of the cave along
the 9 m north line served as the zero point for measuring depth. In 2009, we
deepened the French team’s test pit to a depth of about 4 m and excavated a 1.5
by 5 m trench directly south of it, resulting in an L-shaped excavation. In 2010, we
excavated the 2 by 4 m area immediately west of the French team’s test pit and
enlarged the L-shaped excavation by incorporating the 2 by 3 m area to the east.
Additionally, we excavated a 2 by 2 m area just outside of the rock wall that runs
more or less parallel to the dripline of the cave. At the end of both field seasons,
the team drew the main profile walls of the excavation (figure 3).
2
3

e Tübingen-Armenian Paleolithic Project (TAPP) began in 2008 as a joint
endeavor of the Institute of Archaeology of the National Academy of Sciences
of the Republic of Armenia and the Heidelberg Academy of Sciences and
Humanities (co-directed by B. Gasparyan and A. Kandel). TAPP’s main focus is
to examine the Paleolithic settlement of the Vorotan River Basin in the Arme-
nian Highlands. e Vorotan flows southeast to join the Arax River along the
border with Iran (figure 1). e valley of the Vorotan represents a significant axis
for movement through the region because high mountain ranges rising above
3800 m surround the basin on both sides, channeling the movement of game
and early humans through the steeply incised Vorotan valley. e Vorotan basin
enjoys a cool temperate climate and contains a mosaic landscape suitable to
early human settlement.
During TAPP’s initial season in June 2008, the team conducted survey around
the town of Sisian in the Syunik region of southern Armenia (figure 1). In Aghitu,
a village about 5 km east of Sisian, the team observed seven caves around the
base of a basalt massif rising 25–30 m above the surrounding landscape. e
flat-topped massif is situated near the terminal end of several basalt flows that
emanated from Mt. Bugdatapa at 126–111 ka (Ollivier et al. 2010). Archaeolog-
ical remains atop the massif date to the Middle Bronze Age, Hellenistic and Medi-
eval periods (Kroll 2006; Cherry et al. 2007). In 2003 a French mission surveyed
and tested the seven caves around the base of the massif, but found no Pale-
olithic finds in a 2 by 2 m test pit excavated to a depth of 1.5 m at Aghitu–3
Cave (N39°30’50.5”, E46°4’54.5”). Named according to local convention, the cave
is 11 m deep, 18 m wide, 6 m high and situated at an elevation of 1601 m above
sea level, about 115 m above the Vorotan River. During our field survey in 2008,
we observed obsidian artifacts and mineralized bones on the surface, suggesting
that this cave had potential for excavation. In June 2009 and again in July 2010,
the TAPP team conducted further excavation at Aghitu–3 Cave.
1
(Aghitu-3 Cave) –
Reference map of the Caucasus region
showing Aghitu in the Vorotan River
valley and other sites mentioned in the
text. Legend: 1. Debed river sites; 2. Hovk;
3. Kalavan; 4. Lusakert; 5. Nor Geghi; 6.
Yerevan; 7. Angeghakot; 8. Ortvale Klde;
9. Dzudzuana; 10. Dmanisi (Map: Geral-
dine Quénéhérve).
 1
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 

While our knowledge about the Paleolithic settlement of the Caucasus has grown
in the past decades, the archive of excavated and dated sites remains relatively
small. e Georgian Early Pleistocene site of Dmanisi established the region as
a focal point for research into the earliest human migrations out of Africa at
about 1.8 Ma (Gabunia et al. 2000; Lordkipanidze et al. 2005; Ferring et al. 2011).
Continued research programs at several Georgian sites, such as Ortvale Klde and
Dzudzuana Cave, has vastly improved our understanding of the Middle to Upper
Paleolithic transition of this region (e.g. Nioradze & Otte 2000; Bar-Oz & Adler
2005; Bar-Yosef et al. 2006, 2011; Adler et al. 2008; Moncel et al. 2012).
However, compared to Georgia, the picture in Armenia is just emerging. Lower
Paleolithic settlement is documented mainly by open-air sites with Oldowan-
like and Acheulean industries (Klein 1966; Golovanova 2000; Doronichev 2008;
Kolpakov 2009; Gasparian 2010). In northern Armenia, researchers are inves-
tigating the Middle Pleistocene of the Debed river valley (Egeland et al. 2010,
2011), while others are examining the Dzoraget river valley in the region of Lori
(Dolukhanov et al. 2004; Presnyakov et al. 2012). In central Armenia, Middle Pale-
olithic localities with late Mousterian occupations are known through excava-
tions at Yerevan (Yeritsian 1970) and Lusakert (Yeritsian 1975) caves, but the
dating of these sites remains unresolved. Recent work at Angeghakot–1 (Liagre
et al. 2006), Hovk–1 Cave (Pinhasi et al. 2008, 2011; Bar-Oz et al. 2012) and
Kalavan–2 (Ghukasyan et al. 2011) yielded three new Middle Paleolithic locali-
ties. Meanwhile, ongoing excavation projects led by Adler, Yeritsyan & Gasparyan
in the Hrazdan River Gorge, including Nor Geghi and renewed work at Lusakert,
is shedding light on the nature of the Armenian Lower and Middle Paleolithic
(Adler et al. 2009, 2012).
Despite this improved eort to study the Paleolithic, stratified Upper Paleolithic
sequences are extremely rare in Armenia, represented by a single Late Upper Pale-
olithic open-air site, Kalavan–1, dated to 17–16,000 cal BP (Chataigner et al. 2012).
us, the discovery of the Upper Paleolithic site of Aghitu–3 has significance not
only for the settlement of the Armenian Highlands, but also the Caucasus region.
 
In 2009 the field crew began its systemic excavations, orienting the measuring
grid to the 2 by 2 m test pit excavated by a French team in 2003. e team desig-
nated the coordinate system using letters for the x-axis and numbers for the
y-axis (figure 2). A datum point hammered into the rear wall of the cave along
the 9 m north line served as the zero point for measuring depth. In 2009, we
deepened the French team’s test pit to a depth of about 4 m and excavated a 1.5
by 5 m trench directly south of it, resulting in an L-shaped excavation. In 2010, we
excavated the 2 by 4 m area immediately west of the French team’s test pit and
enlarged the L-shaped excavation by incorporating the 2 by 3 m area to the east.
Additionally, we excavated a 2 by 2 m area just outside of the rock wall that runs
more or less parallel to the dripline of the cave. At the end of both field seasons,
the team drew the main profile walls of the excavation (figure 3).
2
3
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
(Aghitu-3 Cave) – Site plan showing square designations of area excavated. Rock wall runs more or less along the dripline of the cave
(Plan: Dmitri Arakelyan).
 2
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
As we explored the geological stratigraphy during the first field season, excava-
tion proceeded in spits of 10–15 cm in units of one square meter. All finds were
collected in one bag. e excavated sediment was searched by hand for smaller
finds, with an eective recovery of about 5 mm. During the second season we
modified this strategy because we encountered higher find densities. Starting in
2010 we excavated in spits of 2–3 cm thickness in units of one quarter meter. We
piece-plotted single finds larger than 2 cm using meter sticks within the squares
and a line level to measure depth. e team used screens of 5 and 2 mm to sieve
the sediment and collected finds in a single bag. is increased the recovery of
smaller lithic artifacts, faunal remains, small bird bones, microfauna and charcoal.
During excavation we encountered many large basalt blocks with maximum
dimensions up to 1.5 m. Removal of these large blocks necessitated coarser
methods. First the team broke the blocks using a sledge hammer, breaker bar,
chisel and pick. en the basalt debris was removed, so that controlled excava-
tion could continue.
(Aghitu-3 Cave) – Digitized drawings of the west and north
excavation profiles depicting geological layers (GH) and their corresponding
archaeological horizons (AH).
 3
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
 
At the end of the 2010 season we refined our geological understanding of the site.
After studying the profiles, the team expanded the six geological horizons (GH)
identified in 2009. e resulting ten lithostratigraphic units are described below
and correlate with six archeological horizons (AH) (figure 3). Using the Tübingen
tradition, we assign Arabic numbers for GH and Roman numerals for AH.
e surface layer consists of a 10–15 cm of very loose, dry, dusty, light gray, organ-
ic-rich, clayey silt with occasional angular fragments and rounded basalt cobbles
up to 10 cm. GH 1 has a completely anthropogenic character, rich in modern
refuse, animal dung and charcoal. e base of GH 1 consists of very compact
dung layers with white mineral laminations that will be the subject of future sedi-
mentological analyses. e surface layer yielded a few obsidian artifacts, some
mineralized bone fragments and occasional ceramic sherds.
e next layer consists of a 10–20 cm thick, uniform, compact, dry, grayish-brown
silt with frequent angular basalt fragments up to 15 cm. AH II yielded surprisingly
little modern refuse, and is anthropogenic in origin. Below this level, modern
debris and ceramic finds are infrequent. AH II contained a few obsidian artifacts,
rare mineralized bone fragments and some ceramic sherds.
e transition to the underlying 60–80 cm thick layer is very distinct, and the
sediment changes from dry to moist. GH 3 is a geogenic yellowish-brown, finely
laminated, clayey silt containing abundant fragments of weathered platy basalt
up to 30 cm. Large boulders in parts of GH 3 appear to represent a phase that
we refer to as the “upper rockfall”. Finds from AH III included plentiful obsidian
artifacts, ample charcoal and well preserved, dark, mineralized bone fragments
coming from four occupation horizons designated from AH IIIa at the top to AH
IIId at the bottom. Several intact combustion features consisted of a reddish-
brown compact layer, underlying a black layer containing charcoal, and topped
by a white ashy layer. Block samples taken from these features were collected for
laboratory studies of the micromorphology of the sediment.
We subdivided the former GH 4 into four new geological layers which we
renamed GH 4–7. AH IV can be regarded as archaeologically sterile despite the
presence of a few finds. is layer yielded good samples of micromammals.
e transition to this 15–30 cm thick layer was clear and marked by a brown,
finely laminated, clayey silt with some sand. GH 4 contained some weathered
basalt fragments 5–15 cm in size, but fewer large boulders than GH 3. is layer
yielded a few lithics and some well-preserved faunal remains.
e transition to this 10–20 cm thick layer was clear and marked by a light brown,
finely laminated, clayey silt with some sand. GH 5 contained some basalt frag-
ments, but no large boulders. is layer yielded no lithics or faunal remains.
4
Stratigraphy 4.1
GH 1/AH I 4.1.1
GH 2/AH II 4.1.2
GH 3/AH III 4.1.3
GH 4–7/AH IV 4.1.4
GH 4 4.1.5
GH 5 4.1.6
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
e transition to this 5–10 cm thick layer was clear and marked by a brown, finely
laminated, clayey silt with some sand. GH 6 contained some basalt fragments,
but no large boulders. is layer yielded no lithics or faunal remains.
e transition to this 20–25 cm thick layer was clear. Based on changes in the
amount of basalt debris, we subdivided GH 7 into three parts (figure 3). is
layer yielded no lithics or faunal remains. e three finer subdivisions were GH
7a–7c:
GH 7a–light olive brown silt with little angular basalt debris, 10 cm thick
GH 7b–light olive brown silt with frequent angular basalt debris, 5–10 cm
thick
GH 7c–light olive brown silt with little angular basalt debris, 5–10 cm thick.
e transition to this 20–30 cm thick, fine grained deposit is very distinct. Based
on changes in color and grain size, as well as increased moisture content, we
subdivided GH 8 into five easily recognized parts (figure 3). GH 8 yielded a few
obsidian artifacts, yellowish-brown moderately preserved faunal remains, and
abundant charcoal. e five finer subdivisions were named GH 8a–8e:
GH 8a–light gray sand layer, 2–4 cm thick
GH 8b–reddish brown clayey silt with abundant charcoal, 3–5 cm thick
GH 8c–yellowish brown clayey sandy silt, 10 cm thick
GH 8d–reddish brown clayey silt with less charcoal than GH 8b, 3–5 cm thick
GH 8e–variably thick, distinct dark gray sand layer, 3–10 cm thick.
Large, angular basalt boulders up to 1.5 m represent a phase that we refer to as
the “lower rockfall”. e spaces between these boulders are filled with a sandy silt
matrix. is layer appears to thicken towards the dripline of the cave, reaching
a maximum thickness of about one meter. is layer is sterile, yielding no lithics
or faunal remains.
e transition to GH 10 is distinct (figure 3). is 180 cm thick layer was marked
by the predominance of fine sediment, mainly silt, with varying amounts of
clay and sand. e organic content of the sediment appears to increase, as does
moisture. AH VI yielded many obsidian artifacts, much charcoal and numerous
yellowish-brown, moderately preserved faunal remains. GH 10 reached its
maximum depth at 414 cm below datum, or about 350 cm below the ground
surface. A few intact combustion features consisted of a reddish-brown compact
layer, underlying a black layer containing charcoal, and topped by a white ashy
layer. Block samples taken from these features were collected for micromorpho-
logical studies.
GH 6 4.1.7
GH 7 4.1.8
GH 8/AH V 4.1.9
GH 9 4.1.10
GH 10/AH VI 4.1.11
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
From the 2010 season, five samples from AH III were sent to Kiel for radiocarbon
dating using accelerated mass spectrometry. e new results from two bone and
three charcoal specimens show uncalibrated dates ranging from ca. 24–22 000
BP for the upper find horizon AH III (figure 4). Four samples analyzed in 2009
resulted in uncalibrated dates of ca. 30–27 000 BP for the lower find horizons of
AH V and VI (Kandel et al. 2012). Using OxCal version 4.1.7 (Bronk Ramsey 2009),
the radiocarbon dates calibrate between ca. 35–27,000 cal BP (Riemer et al. 2009),
placing the dates firmly in the early part of the Upper Paleolithic.
Ten samples of loose sediment from GH 8 and GH 10 were analyzed for paly-
nological remains (Kandel et al. 2012). Wood charcoal was observed in seven
of the samples. While the four samples from GH 8 contained no pollen, five of
six samples from GH 10 yielded small quantities of pollen from the genera Pinus
(pine), Betula (birch), Quercus (oak), and Centaurea (knapweeds), as well as the
chicory subfamily (Cichorioideae). e presence of Botryococcus, a water dwelling
species of green algae, was confirmed in three samples from GH 10, suggesting
the presence of standing water nearby.
Charcoal remains are abundant in many of the excavated layers and 118 discrete
samples were collected during excavation, with many additional samples
collected during sediment screening. While a detailed study of the charcoal is
under way, identified samples include only those sent for radiocarbon dating.
e identified samples suggest that people brought poplar or willow (Populus/
Salix) to the cave as fuel for burning (figure 5). ese riparian species likely grew
nearby in the Vorotan valley.
 
In the following section we limit our discussion to finds collected in 2009 and
2010 coming from the Paleolithic layers AH III, IV, V and VI. We excluded finds
from AH I and AH II because these layers include modern debris and appear
mixed. We also exclude collected finds resulting from profile cleaning or collapse
if these finds spanned more than a single AH. Here we present the results of 2216
finds, including 1970 chipped stone artifacts and 128 large mammalian remains
(figure 5).
Radiocarbon dating 4.2
Botanical remains 4.3
5
     14   14   (1 σ)14   (2 σ)
IIIA –1,09 CHARCOAL
Populus
or
Salix
KIA-43242 22900 ± 180 28036 – 27060 28163 – 26895
IIIB –1,19 CHARCOAL
Populus
or
Salix
KIA-43241 22630 ± 300 27779 – 26905 28066 – 26300
IIIC –1,31 BONE Sheep/goat metatarsal KIA-43238 23140 ± 130 28174 – 27786 28470 – 27647
IIID –1,37 CHARCOAL Indeterminate KIA-43243 23880 ± 150 28955 – 28421 29280 – 28253
IIID –1,48 BONE Equid radius R KIA-43240 23960 ± 120 29024 – 28530 29265 – 28391
VB –2,38 CHARCOAL
Populus
or
Salix
KIA-39640 27110 + 170/ –160 31235 – 31455 31138 – 31588
VI –2,77 BONE Wolf radius L KIA-39642 27120 ± 170 31240 – 31459 31140 – 31595
VI –3,50 BONE Wild sheep or goat femur KIA-39643 28680 ± 200 32814 – 33442 32249 – 34087
VI –4,04 CHARCOAL Indeterminate KIA-39641 30210 + 180/ –170 34665 – 34934 34570 – 35094
Aghitu-3 Cave. Summary of AMS radiocarbon dating results from the Leibniz Laboratory in Kiel, Germany showing provenience, materials
dated, lab numbers, as well as uncalibrated (uncal) dates and calibrated (cal) age ranges before present (BP).
 4
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
e entire lithic assemblage from AH III-VI consists of 1970 chipped artifacts,
but for the purpose of this analysis, we do not consider the 633 chips, or 32 % of
the lithic artifacts smaller than 10 mm. Furthermore, this analysis highlights and
compares data from the two main find horizons, AH III and VI. For complete-
ness, we present data from AH IV and V, but these assemblages are too small to
provide meaningful interpretations about behavior.
e two main raw materials present in the assemblages are obsidian (85.2 %)
and chert (14.4 %). e remaining raw materials (0.4 %) include dacite and an
unknown stone (figure 6). e sources of these raw materials are presently under
study. It is interesting to note that obsidian is much more common in AH III
(88 %) than in AH VI (57 %). Accordingly, AH VI contains a greater proportion of
chert artifacts.
e obsidian used by the inhabitants of Aghitu–3 is of extremely high quality.
It is variable in color, ranging from glassy translucent to smoky gray, and some
variants are matte gray or opaque red. Banding, streaking and speckling are
common features of all obsidian varieties. e diversity in color and texture
suggests several sources for the obsidian. e closest known primary source of
obsidian is represented by the volcanoes of the Vorotan Group about 30–40 km
northwest of Aghitu (Fouloubey et al. 2003; Liagre et al. 2006; Cherry et al. 2010).
Secondary sources of rounded obsidian pebbles likely stem from deposits in the
Vorotan valley and are documented by the presence of brown cortex on 13 % of
the obsidian finds.
e chert also has excellent knapping characteristics, with a uniform microcrys-
talline structure and glossy texture. e chert exhibits a high variability in color,
ranging from dark brown, red and orange through yellow, beige, gray, white
and green.
Stone artifacts 5.1
Raw material 5.1.1
     %
LITHICS
Blank 891 2 4 95 992 50,4 %
Retouched tool 196 2 4 34 236 12,0 %
Core 18 -- -- 2 20 1,0 %
Angular debris (chunks) 80 -- -- 9 89 4,5 %
Small debitage (chips) 594 -- -- 39 633 32,1 %
LITHIC subtotal 1779 4 8 179 1970 100 %
Tool index (excluding chips) 16,5 % 24,3 %
FAUNA
Small mammal (SC1) 3 5 -- 5 13 10,2 %
Small-medium mammal (SC2) 14 7 2 58 81 63,3 %
Large-medium mammal (SC3) 12 2 4 8 26 20,3 %
Large mammal (SC4) 2 -- 1 2 5 3,9 %
Canis lupus -- -- 1 2 3 2,3 %
FAUNA subtotal 31 14 8 75 128 100 %
CHARCOAL 93 4 12 9 118
TOTAL FINDS 1903 22 28 263 2216
Aghitu-3 Cave. List
of main find categories with break-
down of lithics, fauna and charcoal by
archeological horizon.
 5
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
e sources of the chert are presently unknown, but unpublished geological
studies indicate outcrops 8 km west of Aghitu near Brnakot, and also near Goris,
about 25 km to the east. Secondary sources of chert likely stem from pebbles
deposited in the Vorotan valley and can be identified by the presence of cortex
on 17 % of the chert finds.
e chipped lithic assemblage is strongly oriented towards the production of
laminar products on volumetric cores. Blanks come mainly in the form of blade-
lets with widths less than 10 mm, and to a much lesser degree, blades with widths
greater than 10 mm (figure 7). Of the blanks 62 % in AH III are laminar, while the
tendency in AH VI is even more pronounced with 78 % laminar blanks. An even
higher proportion of formal tools are made on laminar products, with 84 % in
AH III and 97 % in AH VI. ese data clearly document a lithic reduction strategy
focused on making laminar blanks that are overwhelmingly bladelets.
e presence of angular debris comprising about 7 % of the entire assemblage, and
especially the 633 chips smaller than 10 mm indicate that stone knapping took
place at Aghitu–3. e argument for on-site reduction is further strengthened by
the presence of cortical surfaces on 13 % of obsidian and 17 % of chert artifacts.
However, of those artifacts with cortex, the vast majority are covered by less than
50 % of cortex. is illustrates that later stages of decortification, as represented
by the artifacts with less than 50 % cortex coverage, occurred on-site. us, the
early stages of reduction must have occurred elsewhere. A notable dierence in
the presence of cortex on artifacts can be seen when comparing the assemblages
of AH III and AH VI. In AH III cortical pieces comprise 14 % of the assemblage, but
only 3 % in AH VI. is dierence seems to indicate a change in reduction strategy
between these phases of occupation, or possibly diering lengths of occupation.
Blank production and technology 5.1.2
      %
OBSIDIAN
No cortex 907 4 8 78 997 87,5 %
Cortex (< 50 %) 140 -- -- 2 142 12,5 %
Cortex (> 50 %) -- -- -- -- 0 0 %
OBSIDIAN subtotal 1047 4 8 80 1139 100 %
FLINT
No cortex 101 -- -- 58 159 82,8 %
Cortex (< 50 %) 26 -- -- 2 28 14,6 %
Cortex (> 50 %) 5 -- -- -- 5 2,6 %
FLINT subtotal 132 0 0 60 192 100 %
Dacite 2 -- -- -- 2
Unknown 4 -- -- -- 4
TOTAL 1185 4 8 140 1337
Obsidian index 88,4 % 5 7,1 % 85,2 %
Flint index 11,1 % 42,9 % 14,4 %
Cortical index 14,4 % 2,9 % 13,1 %
Aghitu-3 Cave.
Frequency of lithic raw materials
showing distribution of cortical pieces
and cortical index by archeologi-
cal horizon.
 6
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
So far 18 cores from AH III (figure 8) and two cores from AH VI (figure 9) have
been recovered, most of which are small, highly reduced, single platform bladelet
cores (figure 10). While the remaining cores have double or multiple platforms,
they are also aimed at producing bladelets. e predominance of platform
cores mirrors the prevalence of laminar blanks in the assemblage, although the
length of the blanks appears to be much longer than the cores themselves. e
presence of crested blades, core tablets and other preparation debris in both
layers confirms that core preparation and maintenance occurred, although the
smaller assemblage of AH VI contains fewer examples. As excavation continues
we expect that the sample size will increase.
Of the 709 blanks with proximal preservation (complete flakes and proximal
fragments), plain striking platforms are most common, followed in frequency
by indeterminate, shattered, punctiform and faceted butts (figure 11). In AH III
pronounced bulbs of percussion are visible on 7 % of blanks, while the incidence
of shattered bulbs is 7 % and bulbar scars are present on 28 %.
Aghitu-3 Cave.
Overview of lithic technology focus-
ing on blank selection in tool produc-
tion by archeological horizon.
 7
Lithic technology III IV V VI n %
Core 18 -- -- 2 20 1,5 %
Flake 367 -- 2 26 395 29,5 %
Blade 149 1 1 6 157 11,7 %
Bladelet 496 2 3 93 594 44 ,4 %
Core tablet 1 -- -- -- 1 0,1 %
Preparation flake 44 1 -- 3 48 3,6 %
Crested blade primary 7 -- 1 -- 8 0,6 %
Crested blade secondary 3 -- -- -- 3 0,2 %
Burin spall 20 -- 1 1 22 1,6 %
Angular debris 80 -- -- 9 89 6,7 %
TOTAL 1185 4 8 140 1337 100 %
Lithic blanks III IV V VI n %
Laminar blanks 675 3 6 100 784 63,8 %
Non-laminar blanks 412 1 2 29 444 36,2 %
TOTAL 1087 4 8 129 1228 100 %
Laminar blank index 62,1 % 77,5 %
Lithic tool blanks III IV V VI n %
Flake 25 -- -- 1 26 11,0 %
Blade 19 -- -- 1 20 8,5 %
Bladelet 146 2 4 32 184 78,0 %
Angular debris 3 -- -- -- 3 1,3 %
Heat spall 1 -- -- -- 1 0,4 %
Core 2 -- -- -- 2 0,8 %
TOTAL 196 2 4 34 236 100 %
Laminar tool index 84,2 % 97,1 %
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
In AH VI, pronounced bulbs of percussion were observed on 5 % of blanks, while
shattered bulbs were seen on 3 % and bulbar scars on 12 %. Together, these char-
acteristics suggest a lower striking intensity for AH VI compared to AH III. ese
data are consistent with the presence of overhanging lips, which were observed
in 15 % of blanks in AH III, contrasted to 33 % in AH VI. e trend shown by
overhanging lips confirms that AH VI not only shows a lower striking intensity,
but also the possible use of diuse force. Another dierence can be seen in the
degree of dorsal reduction, which was observed in 33 % of artifacts in AH III, but
only 14 % in AH VI. is suggests that knapping in AH VI produced blanks that
required less preparation than in AH III.
(Aghitu–3 Cave, AH
III) – Platform cores (1–6, 8–13, 15–17), core
fragment (7) and scraper on broken core
(14). Raw material: chert (1, 10, 14–15);
obsidian (2–9, 11–13, 16–17) (Illustration:
Elham Ghasidian).
 8
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
Lithic core type III IV V VI n %
Single platform 14 -- -- 1 15 75 %
Double platform 3 -- -- 1 4 20 %
Multiple platform 1 -- -- -- 1 5 %
TOTAL 18 0 0 2 20 100 %
(Aghitu–3 Cave, AH VI) – Polishing stone (1), platform cores (2–3), laterally retouched blade-
lets (4–14, 16–18, 20–26) and unretouched bladelets (15, 19). Raw material: metamorphic (1); chert (2, 4, 6–8, 10–11,
13, 15–16, 18–19, 21–23); obsidian (3, 5, 9, 12, 14, 17, 20, 24–26) (Illustration: Elham Ghasidian).
 9
(Aghitu-3 Cave.
Distribution of core types by archeolog-
ical horizon.
 10
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
In summary, the data suggest that the striking intensity in AH VI was lower than
in AH III, but neither assemblage appears to be produced solely by hard hammer.
Based on the striking attributes, both assemblages appear to result from the
application of diuse force, such as soft hammer or indirect percussion. None-
theless, the knapping characteristics of AH III suggest the use of a more forceful
technique compared to AH VI.
Retouched tools constitute a sizable proportion of the assemblage from AH III
(17 %) and even more so in AH VI (24 %) (figure 5), keeping in mind that the
percentage of tools excludes the 633 chips smaller than 10 mm. As mentioned
before, the vast majority of retouched pieces were made on laminar blanks, 84 %
in AH III and 97 % in AH VI (figure 7). e manufacture of laminar blanks, or
more precisely bladelets, was the single most important aspect of lithic reduction,
and based on their high degree of retouch, these bladelets were clearly geared
towards the production of tools (figure 12).
In both AH III and AH VI the most common tool forms are bladelets that are
finely retouched on one, or sometimes both, lateral edges (figures 9 and 13).
e bladelets are often twisted to the right, but this attribute was not observed
systematically during analysis, so that the nature of twisting must remain a
hypothesis for now. e intensity of retouch is very fine and shows a consistent
pattern. e degree of retouch is high, ranging from 50–100 % of a given lateral
edge and can therefore be described as continuous.
Retouched tool typology 5.1.3
Lithic blank preservation III IV V VI n %
Complete 378 1 1 62 442 36,0 %
Proximal 236 2 2 27 267 21,7 %
Medial 246 -- -- 28 274 22,3 %
Distal 227 1 5 12 245 20,0 %
TOTAL 1087 4 8 129 1228 100 %
Proximal index 56 % 69 %
Lithic butt III IV V VI n %
Plain 255 2 -- 52 309 43 ,6 %
Previous negative 18 -- -- 2 20 2,8 %
Fracture plane 3 -- -- -- 3 0,4 %
Punctiform 46 -- -- 4 50 7,1 %
Faceted 48 -- -- 2 50 7, 1 %
Cortical 7 -- -- 1 8 1,1 %
Shattered 64 1 1 7 73 10,3 %
Indeterminate 173 -- 2 21 196 2 7, 6 %
TOTAL 614 3 3 89 709 100 %
Lithic striking attribute III IV V VI n
Bulb of percussion 7,2 % -- -- 4,5 % 48
Shattered bulb 6,5 % -- -- 3,4 % 43
Bulbur scar 2 8,0 % -- -- 12,4 % 185
Lip 14,8 % -- -- 32,6 % 120
Dorsal reduction 32,9 % -- -- 13,5 % 215
Aghitu-3 Cave.
Review of blank preservation, as well
as butt and striking characteristics by
archeological horizon.
 11
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
is similarity in the character of the retouch suggests that the tools were created
on laminar blanks in a standardized fashion for a similar purpose. However, use
wear studies have not yet been conducted, so that we cannot discern the inten-
tions of their makers.
A very low proportion of the bladelets show more invasive forms of modifica-
tion such as semi-abrupt retouch and rare examples of backed pieces (figure 12).
Nonetheless, these forms are much less common. ese more invasive methods
of retouch may indicate an alternate form of use or hafting, or may simply repre-
sent a part of the full spectrum ranging from fine retouch through semi-abrupt to
fully backed pieces. Finally, other tool forms are rare, but include various scrapers,
notches and burins, among other types (figures 8 and 9).
e finds excavated from Aghitu–3 include 128 large mammalian remains
(figure 5). For now, the specimens have not been identified to genus or species
level. Rather we used a preliminary classification system based on the live weight
of an animal (e.g. Brain 1974; Klein et al. 1991) to establish four animal size classes:
(SC1) 5–20 kg, small fauna, hare to fox size; (SC2) 20–100 kg, small-medium fauna,
wild sheep and wild goat size; (SC3) 100–300 kg, large-medium fauna, equid size;
and (SC4) 300–1000 kg, large fauna, aurochs size. We used these size classes to
establish a picture of the distribution of the assemblage as a whole.
e results show that SC2 predominates with 63 % the assemblage, and most
specimens come from AH VI. e next most frequent size class is SC3 with 20 %,
and most examples are found in AH III. Remains of SC1 (10 %) and SC4 (4 %)
correspond to much smaller proportions of the assemblage. e presence of wolf
(Canis lupus) (2 %) in AH V and VI indicates that carnivores were active at the site.
Most of the fauna are moderately well preserved with good surface preservation.
Preliminary taphonomic observations from AH III show three bones broken in
a fresh state (green break), and two bones appear to be burned. Much of the
fauna from AH III consists of well mineralized and well preserved shaft fragments
of long bones that could not readily be identified. e size classes in AH III are
also more evenly distributed, and together these characteristics suggest a fauna
that was accumulated by humans. In AH VI, on the other hand, four bones show
evidence of biting or chewing by carnivores, and 30 appeared etched, possibly by
the gastric juices of a carnivore.
Large mammalian fauna 5.2
Lithic tool type III IV V VI n %
Laterally retouched tool, fine 119 1 4 27 151 64,0 %
Laterally retouched tool, semi-abrupt 33 1 -- 4 38 16,1 %
Laterally retouched tool, backed 9 -- -- 1 10 4,2 %
End retouch 4 -- -- -- 4 1,7 %
Scraper 15 -- -- -- 15 6,4 %
Tanged point 1 -- -- -- 1 0,4 %
Burin 5 -- -- -- 5 2,1 %
Notch 6 -- -- 1 7 3,0 %
Denticulate 1 -- -- -- 1 0,4 %
Splintered piece 3 -- -- 1 4 1,7 %
TOTAL 196 2 4 34 236 100 %
Fine lateral retouch index 60,7 % 79,4 %
Aghitu-3 Cave.
Frequency of tool types by archeolog-
ical horizon.
 12
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
e fauna from AH VI includes many etched but otherwise complete small bones,
such as phalanges, patellae and vertebrae. Furthermore, the fauna from AH VI
appear skewed towards smaller (SC2) mammals. is preservation combined
with the size class distribution suggests a fauna accumulated by carnivores.
us, our first impression of the fauna from Aghitu–3 is that humans and carni-
vores played dierent roles in accumulating the assemblages. While the fauna
from AH III appears to be accumulated by humans, carnivores were more
involved in the collection of fauna in AH VI. As the excavation continues, we
plan to conduct thorough zooarchaeological analyses, including detailed tapho-
nomic studies, to identify the species present and further assess the degree of
anthropogenic and biogenic modifications.
(Aghitu–3 Cave, AH
III) – Retouched tools including scrapers
(1–2, 4–5, 7–10), splintered piece (6), burins
(12, 13, 17) and laterally retouched blade-
lets (11, 14–15, 18–20), platform core (3) and
core trimming element (16). Raw material:
obsidian (1–4, 6–7, 10–11, 13–16, 18–20);
dacite (5); chert (8–9, 12, 17) (Illustration:
Elham Ghasidian).
 13
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
e finds excavated from Aghitu–3 include a well preserved sample of 441 micro-
faunal specimens belonging to small mammals, birds, fish and amphibians. e
microfaunal samples collected from Aghitu–3 were identified at the Zinman
Institute for Archaeology at the University of Haifa and compared to collections
at the Natural History Museum in Vienna. e seven genera of identified micro-
mammals included pika, voles, hamsters and jerboa (figure 14). To gain a more
detailed chronological view, we separated the assemblage into two groups from
AH I-III and AH IV-VI. e composition of micromammalian species suggests
that the climatic conditions of the Upper Paleolithic were generally cooler than
at present. is is based on the high abundance of voles in the lower part of the
sequence in comparison to their rarity in the upper part (figure 14).
More than 30 specimens of bird are present including bone and eggshell, but
these have not yet been analyzed. On first glance, the avian fauna appear to be
composed of small species that may have nested on the roof of the cave, just as
swallows do today. In addition, one fish mandible (Salmo trutta) and four speci-
mens of amphibians have been identified.
Here we provide a brief overview of the post-Paleolithic layers to complete the
picture (Kandel et al. 2012). Up until now, our discussion of AH I and II has referred
to inside the cave, where these layers average 20–30 cm in thickness. But outside
of the cave, these layers extend to a depth of over 2 m and contain more pottery.
While the pottery assemblage as a whole consists mostly of non-diagnostic body
fragments, some diagnostic pieces are present, including rim, neck-rim, wall-rim,
shoulder, base and body-base fragments. e analysis of these sherds represents
the best means to examine the post-Paleolithic history of the site.
e typological distribution of the vessels includes goblets, jars, bowls and pots,
with one example of household ceramic (oven/tile). e majority of pottery
fragments can be attributed to Medieval times (IV-XIII centuries AD) when,
according to the 13th century Armenian historian Stepanos Orbelyan, Aghitu
was a flourishing town. Two sherds are typical of the Achaemenid to Hellenistic
period (VIII century BC-III century AD), and the oldest sherd is represented by
a single fragment dating to the Early Bronze Age (first half of the III millennium
BC) belonging to the Kura-Araxes culture. We attribute these remains to the
settlements situated on the massif above the cave (Cherry et al. 2007), which
likely made use of the caves underlying the massif.
Micromammals, birds,
fish and amphibians
5.3
Post-Paleolithic 5.4
   ( -)  ( -)
Pika 3 14
Vole 1 18
Migratory hamster 1 1
Water vole -- 5
Golden hamster 4 14
Mole vole 1 5
Jerboa 3 5
TOTAL 13 62
Aghitu-3 Cave.
Summary of identified microfaunal
remains divided into two main strati-
graphic groups based on archeological
horizon (MNI = minimum number
of individuals).
 14
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
e micromammal assemblage can be considered characteristic of the steppic
Armenian Highland region as described by Vereschagin (1959). All of the taxa
present in the assemblage can be considered typical of steppe environments
(Nowak 1999). ey can also be associated with the Southwest Asian mamma-
lian complex that is dominant in the Lesser Caucasus today. Similar assemblages
in terms of the taxonomic composition have also been described recently by
Hashemi et al. (2006) from a number of Upper Paleolithic cave sites in north-
western Iran and support a close environmental and faunal anity of the two
regions during the Upper Paleolithic. Furthermore, the distribution of micro-
mammals indicates that the climate of ca. 35–31,000 cal BP was cooler than today.
Aghitu–3 appears to have parallels with other Caucasian sites, most notably
350 km to the northwest in Georgia. e well-studied sequence of Dzudzuana
Cave is typologically and chronologically the closest companion to Aghitu–3
(Bar-Yosef et al. 2006, 2011), and layers D (ca. 34–32 000 cal BP) and C (ca. 27–24
000 cal BP) are its best analogs. e Upper Paleolithic sequence of Ortvale Klde
provides another favorable comparison with layers 4c (ca. 38–34 000 cal BP), 4b
(ca. 32–28 000 cal BP) and 3 (ca. 26–25 000 cal BP) (Adler et al. 2008). Further
afield in Russia, layer 1a (ca. 34–32,000 cal BP) of Mezmaiskaya shows a diverse
array of bone tools and personal ornaments (Golovanova et al. 1999, 2010). In
Iran the upper sequence (ca. 30 000 cal BP) of Yafteh (Otte et al. 2011) can also be
invoked. ese sites show similar trends in chronology, technology and typology
which will be further examined to test the hypothesis of regional links among
these assemblages.
is glimpse of Aghitu–3 adds to our archaeological knowledge of the Caucasus
region and supports our hypothesis that small and highly mobile groups returned
to the site repeatedly over a period of at least 8 000 years. Such short, low inten-
sity occupations hint that Aghitu–3 served as a temporary seasonal camp used
by hunter-gatherers. e results from the field seasons of 2009 and 2010 give us
reason to believe that continued research at Aghitu–3 will provide answers to
questions about the first modern inhabitants of Armenia and add to the growing
spectrum of knowledge about the origins of the Early Upper Paleolithic.
acknowledgments
We are grateful to the Institute of Archaeology and Ethnography of the National
Academy of Sciences of the Republic of Armenia and the ocials from the Ministry
of Culture of the Republic of Armenia for their assistance with permitting and logis-
tics. We also thank the villagers of Aghitu for their help and patience. Hans-Peter
Uerpmann helped classify the fauna, Diana Zardaryan categorized the pottery,
Chris Miller analyzed the micromorphology, and Katleen Deckers identified the
charcoal samples sent for radiocarbon dating. Smbat Davtyan measured the local
topography, Geraldine Quénéhérve and Dmitri Arakelyan created the maps, and
Elham Ghasidian illustrated the stone artifacts. Project funding came from the
Heidelberg Academy of Sciences and Humanities, the Institute of Archaeology and
Ethnography of the Armenian National Academy of Sciences and the Armenian
Branch of the Gfoeller Foundation of America Corporation.
  
e dating of layers AH VI to ca. 35–31 000 cal BP, AH V to ca. 31 000 cal BP and
AH III to ca. 29–27 000 cal BP places occupation of Aghitu–3 firmly within the
early part of the Upper Paleolithic. Although Upper Paleolithic artifacts have
been documented at other sites in Armenia (e.g. Fourloubey 2003), the age of
these sites remains unknown. erefore, the well stratified and dated assem-
blages of Aghitu–3 are unique in Armenia.
Since raw material resources are not located at the site, we expect a conserv-
ative approach to knapping at Aghitu–3. With primary raw material sources
located 30–40 km away, the presence of cortex on some pieces indicates that
secondary sources such as Vorotan river gravels were also exploited. However,
the low frequency of cortical pieces in the assemblage and the low degree of
cortex covering those pieces indicate that primary reduction began elsewhere,
perhaps at the raw material sources or other occupation sites within the settle-
ment system. Despite this generally conservative approach, diachronic trends
show a flexibility in behavior. For example, the proportion of obsidian and chert
is more evenly distributed in AH VI, but obsidian dominates the younger AH
III. is variability may reflect changes in preference or indicate connections to
dierent parts of the landscape.
e conservative approach to knapping is also reflected in the continuity observed
in the lithic assemblages over time. From the bottom of AH VI to the top of
AH III the lithics appear standardized in terms of their typology and technology.
People consistently manufactured bladelets that were finely retouched on one, or
sometimes both, lateral edges. is production chain appears to be independent
of raw material selection, which is not surprising given the high quality of raw
materials available. Such laminar tools may represent insets that were hafted as
arrowheads, although use wear studies will be necessary to confirm this hypoth-
esis. is straightforward approach enabled the people who lived at Aghitu–3
to produce a standardized toolkit. e paucity of cores further supports the
hypothesis that both obsidian and chert were used judiciously, as does the small
and highly reduced nature of the cores. In fact, many of the laminar blanks are
longer than the cores, underlining the ecient approach to knapping.
A surprising aspect of many Paleolithic sites in Armenia is their high elevation,
and Aghitu–3 (1601 m) is no exception. Many Armenian sites are situated above
1000 m, such as Lusakert–1 and 2 (1417 m), Kalavan–2 (1630 m), Angeghakot–1
(1800 m) and Hovk–1 (2040 m). us high elevation sites do not preclude settle-
ment. Situated around 40°N latitude, southern Armenia has a temperate climate
today. Water resources are plentiful, and the volcanic nature of the soils makes
them productive for plant life. Preliminary data from Aghitu–3 suggest that this
environmental backdrop also extended into the past. Judging from the diversity
of species identified so far, the setting was ideal for humans and fauna alike.
Paleoenvironmental studies of the fauna from Aghitu–3 are still underway, but
preliminary data indicate that a small but varied sample of fauna accumulated
on site. Taphonomic analysis suggests that both humans and carnivores acted as
accumulators. e relatively low quantity of fauna suggests that humans lived at
the site for short periods, perhaps seasonally. e presence of carnivore remains,
small nesting birds and ample microfauna also supports the hypothesis that
human activities in the cave were not too intense and did not last too long.
6
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
e micromammal assemblage can be considered characteristic of the steppic
Armenian Highland region as described by Vereschagin (1959). All of the taxa
present in the assemblage can be considered typical of steppe environments
(Nowak 1999). ey can also be associated with the Southwest Asian mamma-
lian complex that is dominant in the Lesser Caucasus today. Similar assemblages
in terms of the taxonomic composition have also been described recently by
Hashemi et al. (2006) from a number of Upper Paleolithic cave sites in north-
western Iran and support a close environmental and faunal anity of the two
regions during the Upper Paleolithic. Furthermore, the distribution of micro-
mammals indicates that the climate of ca. 35–31,000 cal BP was cooler than today.
Aghitu–3 appears to have parallels with other Caucasian sites, most notably
350 km to the northwest in Georgia. e well-studied sequence of Dzudzuana
Cave is typologically and chronologically the closest companion to Aghitu–3
(Bar-Yosef et al. 2006, 2011), and layers D (ca. 34–32 000 cal BP) and C (ca. 27–24
000 cal BP) are its best analogs. e Upper Paleolithic sequence of Ortvale Klde
provides another favorable comparison with layers 4c (ca. 38–34 000 cal BP), 4b
(ca. 32–28 000 cal BP) and 3 (ca. 26–25 000 cal BP) (Adler et al. 2008). Further
afield in Russia, layer 1a (ca. 34–32,000 cal BP) of Mezmaiskaya shows a diverse
array of bone tools and personal ornaments (Golovanova et al. 1999, 2010). In
Iran the upper sequence (ca. 30 000 cal BP) of Yafteh (Otte et al. 2011) can also be
invoked. ese sites show similar trends in chronology, technology and typology
which will be further examined to test the hypothesis of regional links among
these assemblages.
is glimpse of Aghitu–3 adds to our archaeological knowledge of the Caucasus
region and supports our hypothesis that small and highly mobile groups returned
to the site repeatedly over a period of at least 8 000 years. Such short, low inten-
sity occupations hint that Aghitu–3 served as a temporary seasonal camp used
by hunter-gatherers. e results from the field seasons of 2009 and 2010 give us
reason to believe that continued research at Aghitu–3 will provide answers to
questions about the first modern inhabitants of Armenia and add to the growing
spectrum of knowledge about the origins of the Early Upper Paleolithic.
acknowledgments
We are grateful to the Institute of Archaeology and Ethnography of the National
Academy of Sciences of the Republic of Armenia and the ocials from the Ministry
of Culture of the Republic of Armenia for their assistance with permitting and logis-
tics. We also thank the villagers of Aghitu for their help and patience. Hans-Peter
Uerpmann helped classify the fauna, Diana Zardaryan categorized the pottery,
Chris Miller analyzed the micromorphology, and Katleen Deckers identified the
charcoal samples sent for radiocarbon dating. Smbat Davtyan measured the local
topography, Geraldine Quénéhérve and Dmitri Arakelyan created the maps, and
Elham Ghasidian illustrated the stone artifacts. Project funding came from the
Heidelberg Academy of Sciences and Humanities, the Institute of Archaeology and
Ethnography of the Armenian National Academy of Sciences and the Armenian
Branch of the Gfoeller Foundation of America Corporation.
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique

ADLER D.S., BAR-YOSEF O., BELFER-COHEN A.,
TUSHABRAMISHVILI N., BOARETTO E., MERCIER N.,
VALLADAS H. & RINK W.J. (2008) – Dating the demise:
Neanderthal extinction and the establishment of modern
humans in the southern Caucasus. Journal of Human Evolu-
tion 55:817–833.
ADLER D.S., PINHASI R., GASPARIAN B., TUSHABAR-
AMISHVILI N. & YERITSYAN B. (2009) – Neanderthal
life-ways in Southern Caucasus. In: N. Catto (ed.), e Nean-
derthal Home: Spatial and Social Behaviors, Abric Romani
(1909–2009). International workshop, Tarragona-Capellades
(Spain), 6–9 October 2009, Institut Catala de Paleoecologia
Humana i Evolucio Social, p. 125–126.
ADLER D.S., YERITSYAN B., WILKINSON K., PINHASI R.,
BAR-OZ G., NAHAPETYAN S., BAILEY R., SCHMIDT B.A.,
GLAUBERMAN P., WALES N. & GASPARIAN B. (2012)
e Hrazdan Gorge Palaeolithic Project, 2008–2009. In: P.
Avetisyan & A. Bobokhyan (eds.), Archaeology of Armenia in
Regional Context. Proceedings of the International Conference
dedicated to the 50th Anniversary of the Institute of Archaeology
and Ethnography, September 15–17, 2009, Yerevan, Armenia.
NASRA Gitutyn Publishing House, Yerevan, p. 21–37.
BAR-OZ G. & ADLER D.S. (2005) – Taphonomic history of
the Middle and Upper Paleolithic faunal assemblage from
Ortvale Klde, Georgian Republic. Journal of Taphonomy
3(4):185–212.
BAR-OZ G., WEISSBROD L., GASPARIAN B., NAHA-
PETYAN S., WILKINSON K. & PINHASI R. (2012) Tap ho -
nomy and zooarchaeology of a high-altitude Upper Pleisto-
cene faunal sequence from Hovk–1 Cave, Armenia. Journal of
Archaeological Science 39:2452–2463.
BAR-YOSEF O., BELFER-COHEN A. & ADLER D.S. (2006)
e implications of the Middle-Upper Paleolithic chronolog-
ical boundary in the Caucasus to Eurasian prehistory. Anthro-
pologie 44(1):49–60.
BAR-YOSEF O., BELFER-COHEN A., MESHEVILIANI T.,
JAKELI N., BAR-OZ G., BOARETTO E., GOLDBERG P.,
KVAVADZE E. & MATSKEVICH Z. (2011) – Dzudzuana:
An upper Palaeolithic cave site in the Caucasus foothills
(Georgia). Antiquity 85:331–349.
BINFORD L.R. (1984) Faunal Remains from Klasies River
Mouth. Academic Press, Orlando.
BRAIN C.K. (1974) – Some suggested procedures in the
analysis of bone accumulations from southern African
Quaternary sites. Annals of the Transvaal Museum 29:1–8.
BRONK RAMSEY C. (2009) – Bayesian analysis of radio-
carbon dates. Radiocarbon 51:337–360.
CHATAIGNER C., GASPARYAN B., MONTOYA C.,
ARIMURA M., MELIKYAN V., LIAGRE J., PETROSYAN
A., GHUKASYAN R., COLONGE D., FOURLOUBEY
C., ARAKELYAN D., ASTRUC L., NAHAPETYAN S.,
HOVSEPYAN R., BALASESCU A., TOMÉ C. & RADU V.
(2012) – From the Late Upper Palaeolithic to the Neolithic
in north-western Armenia: Preliminary results. In: P. Avetisyan
& A. Bobokhyan (eds.), Archaeology of Armenia in Regional
Context. Proceedings of the International Conference dedi-
cated to the 50th Anniversary of the Institute of Archaeology
and Ethnography, September 15–17, 2009, Yerevan, Armenia.
NAS RA Gitutyn Publishing House, Yerevan, p. 52–63.
CHERRY J.F., FARO E.Z. & MINC L. (2010) – Field Survey
and Geochemical Characterization of the Southern Armenian
Obsidian Sources. Journal of Field Archaeology 35(2):147–163.
CHERRY J.F., MANNING S.W., ALCOCK S.E., TONIKYAN
A.V. & ZARDARYAN M.H. (2007) – Radiocarbon dates for
the second and first millennia B.C. from southern Armenia:
Preliminary results from the Vorotan Project 2005–2006.
ARAMAZD, Armenian Journal of Near Eastern Studies 2:52–71.
DOLUKHANOV P., ASLANIAN S., KOLPAKOV E. & BELY-
AEVA E. (2004) – Prehistoric sites in Northern Armenia.
Antiquity 78 (www.antiquity.ac.uk/projgall/dolukhanov301).
DORONICHEV V.B. (2008) – e Lower Paleolithic in
Eastern Europe and the Caucasus: A reappraisal of the data
and new approaches. PaleoAnthropology 2008:107–157.
EGELAND C.P., GASPARIAN B., ARAKELYAN D., BYERLY
R.M., NICHOLSON C.M. & ZARDARYAN D. (2011)
Multiperiod archaeological reconnaissance in the Debed river
valley, northeastern Armenia. Antiquity 85 (www.antiquity.
ac.uk/projgall/egeland329).
EGELAND C.P., NICHOLSON C.M. & GASPARIAN B.
(2010) – Using GIS and ecological variables to identify high
potential areas for paleoanthropological survey: an example
from northern Armenia. Journal of Ecological Anthropology
14:89–98.
FERRING R., OMS O., AGUSTÍ J., BERNA F., NIORADZE M.,
SHELIA T., TAPPEN M., VEKUA A., ZHVANIA D. & LORD-
KIPANIDZE D. (2011) – Earliest human occupations at
Dmanisi (Georgian Caucasus) dated to 1.85–1.78 Ma. Proceed-
ings of the National Academy of Sciences 108:10432–10436.
The Upper Paleolithic settlement of the Armenian Highlands — Andrew W. KANDEL
et al
. 
FOURLOUBEY C., BEAUVAL C., COLONGE D., LIAGRE
J., OLLIVIER V. & CHATAIGNER C. (2003) – Le Paléolith-
ique en Arménie: état des connaissances acquises et données
récentes. Paléorient 29(1):5–18.
GABUNIA L., VEKUA A., LORDKIPANIDZE D., SWISHER
C.C., FERRING R., JUSTUS A,. NIORADZE M., TVAL-
CHRELIDZE M., ANTON S.C., BOSINSKI G., JÖRIS O., DE
LUMLEY M.A., MAJSURADZE G. & MOUSKHELISHVILI A.
(2000) – Earliest Pleistocene hominid cranial remains from
Dmanisi, Republic of Georgia: Taxonomy, geological setting,
and age. Science 288:1019–1025.
GASPARIAN B. (2010) – Landscape organization and
resource management in the Lower Palaeolithic of Armenia.
Turkish Academy of Sciences Journal of Archaeology (TU-BAAR)
13:159–183.
GHUKASYAN R., COLONGE D., NAHAPETYAN S., OLLI-
VIER V., GASPARYAN B., MONCHOT H. & CHATAIGNER
C. (2011) – Kalavan–2 (north of Lake Sevan, Armenia): A
new late Middle Paleolithic site in the Lesser Caucasus.
Archaeology, Ethnology & Anthropology of Eurasia 38(4):39–51.
GOLOVANOVA L.V. (2000) – Late Acheulian of the
northern Caucasus and the problem of transition to the
Middle Paleolithic. In: D. Lordkipanidze, O. Bar-Yosef & M.
Otte (eds.), Early humans at the gates of Europe. Proceedings
of the first international symposium, Dmanisi, Tbilisi (Georgia,
September 1998). ERAUL 92:49–65.
GOLOVANOVA L.V., HOFFECKER J.F., KHARITONOV
V.M. & ROMANOVA G.P. (1999) – Mezmaiskaya Cave: A
Neanderthal occupation in the northern Caucasus. Current
Anthropology 40:77–86.
GOLOVANOVA L.V, DORONICHEV V.B. & CLEGHORN
N.E. (2010) – e emergence of bone-working and orna-
mental art in the Caucasian Upper Paleolithic. Antiquity
84:299–320.
HASHEMI N., DARVISH J., MASHKUR M. & BIGLARI F.
(2006) – Rodent and lagomorph remains from late Pleisto-
cene and early Holocene caves and rock shelter sites in the
Zagros region, Iran. Iranian Journal of Animal Biosystematics
2:25–33.
KANDEL A.W., GASPARYAN B., BRUCH A.A., WEISS-
BROD L. & ZARDARYAN D. (2012) – Introducing Aghitu–3,
the First Upper Paleolithic Cave Site in Armenia. ARAMAZD,
Armenian Journal of Near Eastern Studies 6(2):7–23.
KLEIN R.G. (1966) – Chellean and Acheulean on the terri-
tory of the Soviet Union: A critical review of the evidence
as presented in the literature. American Anthropologist
68(2–2):1–45.
KLEIN R.G., CRUZ-URIBE K. & BEAUMONT P. (1991)
Environmental, ecological, and paleoanthropological impli-
cations of the Late Pleistocene mammalian fauna from Equus
Cave, Northern Cape Province, South Africa. Quaternary
Research 36:94–119.
KOLPAKOV E.M. (2009) – e Late Acheulian site of Dash-
tadem–3 in Armenia. PaleoAnthropology 2009:3–31.
KROLL S. (2006) – Southern Armenia Survey (Syunik),
2000–2003 (Achaemenid-Hellenistic Period, Aghitu Fortress,
Sisian, 1640 m, N39 30’49.7” E46 4’51.2”). ARAMAZD, Arme-
nian Journal of Near Eastern Studies 1:19–49.
LIAGRE J., GASPARYAN B., OLLIVIER V. & NAHAPETYAN
S. (2006) – Angeghakot 1 (Armenia) and the identification
of the Mousterian cultural facies of “Yerevan points” type in
the southern Caucasus. Paléorient 32(1):5–18.
LORDKIPANIDZE D., VEKUA A., FERRING R., RIGHT-
MIRE G.P., AGUSTI J., KILADZE G., MOUSKHELISHVILI
A., NIORADZE M., PONCE DE LEÓN M.S., TAPPEN M. &
ZOLLIKOFER C.P.E. (2005) – e earliest toothless hominin
skull. Nature 434:717–718.
MONCEL M.-H., PLEURDEAU D., TUSHUBRAMISHVILI N.,
YESHURUN R., AGAPISHVILI T., PINHASI R. & HIGHAM
T.F.G. (2012) – Preliminary results from the new excavations
of the Middle and Upper Palaeolithic levels at Ortvale Klde-
north chamber (South Caucasus Georgia). Quaternary Inter-
national (doi.org/10.1016/j.quaint.2012.09.012).
NIORADZE M.G. & OTTE M. (2000) – Paléolithique
supérieur de Géorgie. LAnthropologie 104:265–300.
NOWAK R.M. (1999) Walker’s mammals of the world (6th
edition). e Johns Hopkins University Press, Baltimore.
OLLIVIER V., NAHAPETYAN S., ROIRON P., GABRI-
ELYAN I., GASPARYAN B., CHATAIGNER C., JOANNIN
S., CORNÉE J.-J., GUILLOU H., SCAILLET S., MUNCH P. &
KRIJGSMAN W. (2010) – Quaternary volcano-lacustrine
patterns and palaeobotanical data in southern Armenia.
Quaternary Research 223–224:312–326.
OTTE M., SHIDRANG S., ZWYNS N. & FLAS D. (2011)
New radiocarbon dates for the Zagros Aurignacian from
Yafteh cave, Iran. Journal of Human Evolution 61:340–346.
PINHASI R., GASPARIAN B., NAHAPETYAN S., BAR-OZ
G., WEISSBROD L., BRUCH A.A., HOVSEPYAN R. &
WILKINSON K. (2011) – Middle Palaeolithic human
occupation of the high altitude region of Hovk–1, Armenia.
Quaternary Science Reviews 30:3846–3857.
 UISPP — Liège, mai 2012 — Modes de contacts et de déplacements au Paléolithique eurasiatique
PINHASI R., GASPARIAN B., WILKINSON K., BAILEY R.,
BAR-OZ G., BRUCH A., CHATAIGNER C., HOFFMANN
D., HOVSEPYAN R., NAHAPETYAN S., PIKE A.W.G.,
SCHREVE D. & STEPHENS M. (2008) – Hovk 1 and the
middle and Upper Paleolithic of Armenia: A preliminary
framework. Journal of Human Evolution 55:803–816.
PRESNYAKOV S.L., BELYAEVA E.V., LYUBIN V.P.,
RODIONOV N.V., ANTONOV A.V., SALTYKOVA A.K.,
BEREZHNAYA N.G. & SERGEEV S.A. (2012) – Age of the
earliest Paleolithic sites in the northern part of the Armenian
Highland by SHRIMP-II U-Pb geochronology of zircons from
volcanic ashes. Gondwana Research 21:928–938.
REIMER P.J., BAILLIE M.G.L., BARD E., BAYLISS A., BECK
J.W., BLACKWELL P.G., BRONK RAMSEY C., BUCK C.E.,
BURR G.S., EDWARDS R.L., FRIEDRICH M., GROOTES P.M.,
GUILDERSON T.P., HAJDAS I., HEATON T.J., HOGG A.G.,
HUGHEN K.A., KAISER K.F., KROMER B., MCCORMAC
F.G., MANNING S.W., REIMER R.W., RICHARDS D.A.,
SOUTHON J.R., TALAMO S., TURNEY C.S.M., VAN DER
PLICHT J. & WEYHENMEYER C.E. (2009) – IntCal09 and
Marine09 radiocarbon age calibration curves, 0–50,000 years
cal BP. Radiocarbon 51:1111–1150.
VERESCHAGIN N.K. (1959) e Mammals of the Caucasus:
e History of the Fauna. Academy of Science, Moscow.
YERITSIAN B.G. (1970) Paleolithic cave site of Yerevan
and its place among the earliest sites of Caucasus. Ph.D. esis,
University of Moscow (in russian).
YERITSIAN B.G. (1975) – e new Lower Paleolithic
cave site Lusakert 1 (Armenia). Kratkiye Soobshcheniya o
Dokladakh i Polevykh Issledovaniyakh 141:54–67 (in russian).
... Measuring 18 m wide, 11 m deep and 6 m high, this cave seemed to have the potential to yield older sediments (Figures 1 and 2). Therefore, between 2009 and 2013 the TAPP team conducted archaeological excavations at Aghitu-3 ( 2014). The cave is situated at 1601 m above sea level at the base of the basalt formation. ...
... Within this in mind, we present here the fi rst " true " Upper Paleolithic sites of Armenia, which have only been recently uncovered and evaluated. Therefore, between 2009 and 2013 the TAPP team conducted archaeological excavations at Aghitu-3 ( 2014). The cave is situated at 1601 m above sea level at the base of the basalt formation. ...
Chapter
Full-text available
While the Armenian Highlands have benefited from a longer history of research into the Early and Middle Paleolithic occupations of this region, its Upper Paleolithic settlement has only recently begun to come into focus. With this brief contribution we summarize new archaeological data from two high elevation sites that together span the majority of the Upper Paleolithic. These well stratified sites in the Armenian Highlands benefited from the use of modern excavation methods. The first evidence for modern human behavior is seen at Aghitu-3 Cave in Syunik Province of southern Armenia, while cultural remains from the late Upper Paleolithic are documented in the north at the site of Kalavan-1, located in the wooded montane landscape north of Lake Sevan. We hypothesize that any hominin who entered the Armenian Highlands had to solve the problem of how to survive in this high altitude environment. Under modern conditions, which we view as analogous to an interglacial, the climate is continental, exhibiting a large fluctuation between summer and winter temperatures. This high altitude region, much of it above 2000 m, is blanketed by snow during the winter and well into spring. During glacial periods, alpine ice sheets would have covered a considerable portion of the region, providing a significant impediment for human habitation. Such environmental hurdles would surely have imprinted on the early inhabitants of the region and facilitated the solutions that led to their survival at these high altitudes.
... Current research focuses on Sefunim Cave in Israel, with a large assemblage of shell ornaments, on Jebel Faya (UAE), with its well-stratified archaeological sequence and detailed environmental contexts (Bretzke et al. 2013), and Suhailah (UAE), with evidence for Middle Pleistocene occupation. Another focus is on Aghitu-3 Cave in Armenia, a site which yielded evidence for the early manufacture of complex clothing (Fig. 14) and an early Upper Paleolithic relationship with wolves that we tested by means of morphometric, genetic and isotopic studies (Fig. 15) ( Gasparyan et al. 2014; Kandel et al. 2014). ...
Technical Report
Full-text available
The research center “The Role of Culture in Early Expansions of Humans” (ROCEEH) is a long-term project of the Heidelberg Academy of Sciences and Humanities. ROCEEH aims to reconstruct the trajectory of human evolution in Africa and Eurasia between 3 million and 20,000 years before present. The project focuses on three different fields of expansion: the spatio-temporal and taxonomic expansion of the various hominin species, the expansion of the ecological environment and resource spaces, and most central, the expansion of hominin performances and cultural capacities (Fig. 1). An interdisciplinary team of archaeologists, paleoanthropologists, paleobiologists, geographers and database specialists addresses the dynamics and interactions of the different fields of expansions to gain a better understanding of the different roles of developing cultural capacities on the unique pathway towards “becoming human”. With workplaces at the Eberhard Karls University of Tübingen and the Senckenberg Research Institute in Frankfurt/Main, Germany, the research center will operate from 2008 until 2027. The evaluation of ROCEEH in 2016 provides good grounds to take preliminary stock of the current status of the project and its results.
Article
Kites in Armenia were recently discovered, and investigations into their construction, typology and dating are ongoing. With these discoveries, it has become necessary to investigate a series of unsolved questions. In order to test the functions of kites, we conducted a synthesis describing the occurrence and habitat range of Late Pleistocene and Holocene wild ungulates in Armenia. Wildlife is discussed by emphasizing animal behavior and distribution, along with the hunting strategies adopted by the communities that inhabited Armenia. In spite of the fact that wild ungulates did not contribute largely to the daily meat intake or to the major raw products needed by humans since their domestication (around 6000 cal. BC), wild goats, gazelle and red deer were the animals most frequently hunted in Armenia in different time periods and in a variety of landscapes. Hypotheses put forward suggest that these preferences might be linked to using kites as traps for herds at different seasons of the year and on different altitudes, between 3000 and 500 BC.
Article
Full-text available
New work from the Caucasus is revolutionising the timing and character of the shift from Neanderthals to early Modern humans in Eurasia. Here the authors reveal a powerful signal of that change from excavations at Mezmaiskaya: the abrupt appearance of a well-formed bone industry and ornaments.
Article
Full-text available
Late Acheulean sites on the Northern Caucasus are found from the Black Sea coast (Mis Kadosh) in the west to Northern Ossetia (Hutor Popov 1 in the Terek River valley) in the east (fig. 1). A total of nearly 40 locations (Autlev, 1961; 1981; Golovanova, 1986; Formozov, 1965) are known here now. The majority of them are situated in the Belaya River basin. This partly reflects the fact that most field research was conducted in the area, but it may also be due to specific conditions of the Paleolithic remains taphonomy in the region. Three local groups of the Late Acheulean sites are of particular interest now. They are concentrated in a relatively small area of the Northwestern Caucasus, and include the Abadzeh and Khadjoh groups in the middle part of the Belaya River valley, and the Abin group near Krasnodar.
Article
Full-text available
New work from the Caucasus is revolutionising the timing and character of the shift from Neanderthals to early Modern humans in Eurasia. Here the authors reveal a powerful signal of that change from excavations at Mezmaiskaya: the abrupt appearance of a well-formed bone industry and ornaments.
Article
This article is a preliminary attempt to explain the settlement of Armenia in the Lower Palaeolithic period, and to estimate the factors that played a decisive role in the 'exploitation'of specific ecological niches and natural landscapes. In recent years, new and highly interesting finds and sites of the Lower Palaeolithic have reinvigorated study in this field. A number of projects have focused on the study of the environment, the sources of raw material, and reconstructions of the palaeogeographical situation and ancient climate. Although many questions remain-cultural frontiers, seasonality, the character of hunting specializations of specific fauna-we remain hopeful that excavations planned for 2010-2011 at the newly discovered Lower Palaeolithic sites in the gorge of the Debed River and in the Ararat Depression may augment our understanding on these matters.