Radiocarbon dating the appearance of modern humans and timing of cultural innovations in Europe: new results and new challenges
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Journal of Human Evolution 44 (2003) 331–371
Radiocarbon dating the appearance of modern humans and
timing of cultural innovations in Europe: new results and
new challenges
Nicholas J. Conard *, Michael Bolus
Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters, Universität Tübingen, Schloss Hohentübingen, 72070 Tübingen,
Germany
Received 1 February 2002; accepted 18 November 2002
Abstract
New radiocarbon dates from the sites of Bockstein-Törle, Geißenklösterle, Hohle Fels, Hohlenstein-Stadel,
Sirgenstein, and Vogelherd in the Swabian Jura of southwestern Germany indicate that the Aurignacian of the region
spans the period from ca. 40–30 ka BP. If the situation at Vogelherd, in which skeletal remains from modern humans
underlie an entire Aurignacian sequence, is viewed as representative for the region, the dates from the Swabian Jura
support the hypothesis that populations of modern humans entered the region by way of the “Danube Corridor.” The
lithic technology from the lower Aurignacian of Geißenklösterle III is fully developed, and classic Aurignacian forms
are well represented. During the course of the Aurignacian, numerous assemblages rich in art works, jewelry, and
musical instruments are documented. By no later than 29 ka BP the Gravettian was well established in the region. These
dates are consistent with the “Kulturpumpe” hypothesis that important cultural innovations of the Aurignacian and
Gravettian in Swabia predate similar developments in other regions of Europe. The radiocarbon dates from
Geißenklösterle corroborate observations from other non-archaeological data sets indicating large global fluctuations
in the atmospheric concentrations of radiocarbon between 30 and 50 ka calendar years ago. These fluctuations lead to
complications in building reliable chronologies during this period and cause the “Middle Paleolithic Dating Anomaly”
and the “Coexistence Effect,” which tend to exaggerate the temporal overlap between Neanderthals and modern
humans.
2003 Elsevier Science Ltd. All rights reserved.
Keywords: Neanderthals; Homo sapiens sapiens; Chronostratigraphy; Swabian Jura; Middle and Upper Paleolithic; Cultural
innovations
Introduction
Over the last two decades considerable evidence
* Corresponding author
E-mail addresses: nicholas.conard@uni-tuebingen.de has accumulated indicating that modern humans
(N.J. Conard), michael.bolus@uni-tuebingen.de (M. Bolus). evolved in Africa, while fossil hominin remains
0047-2484/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0047-2484(02)00202-6
332 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
from Europe document the in situ evolution of rich research tradition stretching from the middle
Neanderthals out of earlier archaic populations of the nineteenth century until today has produced
(Arsuaga et al., 1997; Bräuer, 1984; Bräuer and the best record of these changes. The Swabian Jura
Smith, 1992; Rightmire, 1989; Smith and Spencer, of southwestern Germany has been a major center
1984). The evidence from ancient DNA studies of Paleolithic research that is closely connected
(Krings et al., 1997; Ovchinnikov, 2000; Scholz with excavations of O. Fraas, R. R. Schmidt, G.
et al., 2000a,b) supports this hypothesis, although Riek, R. Wetzel, J. Hahn and others. Particularly
the interpretation of these data remains contro- renowned are the sites in the Ach and Lone
versial (Templeton, 2002). Finds from scores of Valleys, such as Vogelherd, Hohlenstein-Stadel,
European sites that provide Middle Paleolithic Geißenklösterle, and Hohle Fels (Fig. 1), that
deposits and the chronostratigraphic framework document the oldest universally accepted figurative
for the development and pan-European distribu- art and musical instruments (Conard and Floss,
tion of Neanderthals are also consistent with the 2000; Hahn, 1986; Hahn and Münzel, 1995;
Out of Africa hypothesis. The central question at Müller-Beck et al., 2001). These finds belong to the
present is not if modern humans evolved outside Aurignacian period and are accompanied by the
Europe, but by what route they arrived, when, and earliest stratified remains of modern humans in
by what processes their populations expanded Europe (Churchill and Smith, 2000a,b; Riek, 1932,
across Europe. Closely related to the question of 1934). This paper reports a series of AMS radio-
the advent and spread of anatomically modern carbon dates on bones from Middle Paleolithic,
humans is the source and timing of the spread of Aurignacian and Gravettian deposits of the
fully modern behavior as demonstrated by the Swabian Alb and examines their implications for
complex technology and symbolic communica- the arrival and spread of modern humans and fully
tion recorded in many early Upper Paleolithic modern cultural behavior in Europe. In Swabia
assemblages. These questions form the stage for hominin remains are known from Middle and
intense international debate in contemporary Upper Paleolithic contexts. Although human
paleoanthropology (d’Errico et al., 1998). The fossil material is not abundant in Swabia, so far
recent dating of European Neanderthals and Neanderthal remains have been found only in
Middle Paleolithic artifact assemblages in Iberia association with Middle Paleolithic artifacts and
(d’Errico et al., 1998), Crimea (Chabai, 2000; modern humans exclusively with Upper Paleolithic
Marks and Chabai, 1998; Pettitt, 1998), Croatia artifacts (Table 1). Thus, as a working hypothesis,
(Smith et al., 1999), the northern Caucausus we assume for the purpose of this paper that
(Golovanova et al., 1999) and Georgia (Adler, modern humans rather than Neanderthals made
2002; Adler and Tushabramishvili, in press) to the Aurignacian assemblages of southern
around, and in some cases after, 30 ka BP raises Germany.
further questions about the nature of the inter- The archaeological record of southern Germany
action between modern and archaic hominins. shows a clear break between the latest Middle
Similarly, the appearance of initial Upper Paleo- Paleolithic including several stratified and
lithic assemblages in the Levant (Azoury, 1986; numerous unstratified Blattspitzen assemblages
Kuhn et al., 1999, 2001; Marks, 1983; Monigal, and the earliest Upper Paleolithic characterized by
2001; Volkman, 1983), central and northeastern Aurignacian assemblages rich in lithic and organic
Asia (Brantingham et al., 2001; Derevianko et al., tools, artworks and ornaments (Bolus, in press;
2001) suggests that new models to explain the Bolus and Conard, 2001; Bolus and Rück, 2000;
population dynamics and the cultural develop- Bosinski, 1967). Thus far no hominin remains have
ments of the early Upper Paleolithic are needed. been found with Blattspitzen assemblages in south-
Much research on the timing of the appearance ern Germany. In the Swabian Jura where the
of modern humans and the development of com- research has been most intense, despite numerous
plex symbolic behavior, a key characteristic of excavations, there are no examples of inter-
cultural modernity, has focused on Europe where a stratification of Middle and Upper Paleolithic
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 333 Fig. 1. Map of Southwestern Germany with the principal sites mentioned in text. Ach Valley: (1) Sirgenstein-(2) Hohle Fels-(3) Geißenklösterle-(4) Brillenhöhle; Lone Valley: (5) Bockstein (Bockstein-Höhle and Bockstein-Törle)-(6) Hohlenstein (Stadel and Bärenhöhle)-(7) Vogelherd. assemblages, and the technology and typology of remains of modern humans in Europe are from the the Aurignacian assemblages show a radical base of the Aurignacian layer V from G. Riek’s departure from all known Middle Paleolithic 1931 excavation at Vogelherd (Churchill and assemblages in the region (Bolus and Conard, Smith 2000a,b; Riek, 1932, 1934). The artworks 2001; Hahn, 1977). and musical instruments from the Aurignacian of Two models developed in Tübingen in connec- the Swabian Jura are among the earliest finds of tion with the sites of the Swabian Jura are the this kind worldwide (Bolus and Conard, 2001; Danube Corridor and Kulturpumpe models Hahn, 1986; Hahn and Münzel, 1995), and current (Conard, 2002a,b; Conard and Floss, 2000; excavations continue to provide new evidence for Conard et al., 1999). The former postulates that cultural innovation during the Upper Paleolithic modern humans rapidly entered the interior of (Conard and Floss, 2000; Conard et al. 2002). New Europe via the Danube Valley. The latter model results from Geißenklösterle and Hohle Fels presents competing working hypotheses to explain have significantly expanded the range of known the early advent of fully modern behavior and the ornament, mobile art, and lithic and organic tools cultural innovations of the Aurignacian and from the Aurignacian and Gravettian of Swabia. Gravettian in the Swabian Jura. The Danube Dates from several sites in the Swabian Jura Corridor model is supported by the independent document Gravettian assemblages including dis- confirmation of the early 14C dates of the tinctive lithic and organic artifacts by 29 ka BP. Aurignacian from Geißenklösterle using TL meas- These rich assemblages predate similar assem- urements on burnt flint (Richter et al., 2000). blages in Europe and correspond to a period when Based on these measurements, some of the earliest the Aurignacian was still widespread in western Aurignacian assemblages in Europe date to Europe (Bosinski, 1989; Delporte, 1998; Djindjian, ca. 40 ka BP and come from the Swabian Jura 1993; Djindjian et al., 1999). Determining the of southwestern Germany. The earliest skeletal route of entry into Europe and the location of
334
Table 1
Human remains from Middle Paleolithic, Aurignacian, and Gravettian deposits of the Lone and Ach Valleys
Site Arch. horizon Fossil Anthropological Archaeological References
determination context
Lone Valley
Hohlenstein-Stadel
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
“Schwarzes diaphysis of a right Neanderthal Mousterian Völzing, 1938;
Moustérien” femur male? adult Kunter and Wahl, 1992
19–20 m premolar modern H. s.? Aurignacian Hahn, 1977
spit 6 young adult
Vogelherd
V (basis) Stetten 1 modern H. s. Aurignacian Riek, 1932;
cranium with mandibula male adult Gieseler, 1937;
2 lumbar vertebrae Czarnetzki, 1983
V (basis) Stetten 3 modern H. s. Aurignacian Gieseler, 1937;
humerus male Churchill and Smith, 2000a
V (basis) Stetten 4 modern H. s. Aurignacian Czarnetzki, 1983
left metacarpal
IV (top) Stetten 2 modern H. s. ? Riek, 1932;
cranium male young adult Gieseler, 1937;
Czarnetzki, 1983
Ach Valley
Sirgenstein
VI left upper canine modern H. s. Aurignacian Schmidt, 1910;
left lower molar adult Schliz, 1912
VI right upper canine modern H. s. Aurignacian Schmidt, 1910;
adult Schliz, 1912
Geißenklösterle
It right upper deciduous modern H. s. Gravettian Hahn et al., 1990
molar child
It deciduous molar modern H. s. Gravettian Haas, 1991
Hohle Fels
II cranial fragment modern H. s. Gravettian
young adult?
II right lower deciduous modern H. s. Gravettian Haas, 1991
molar juvenileN.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 335
centers of cultural innovations during the period and archaeological context are well documented.
when Europe was occupied by both archaic and Samples from earlier excavations bear designations
modern hominids hinges to a significant extent on for archaeological layers and in some cases spits
the reliable dating of late Middle Paleolithic and within archaeological layers, but the specific con-
early Upper Paleolithic deposits. At present, texts of these specimens are less secure than the
despite well documented variations in 14C produc- piece-plotted finds. In practice this means that high
tion in this period (Beck et al., 2001; van der Plicht, resolution three dimensional coordinates are avail-
1999; Voelker et al., 2000), radiocarbon measure- able for all AMS dates from Geißenklösterle and
ments provide the only broadly applicable means Hohle Fels and are lacking for specimens from all
of dating find horizons in the critical period the other sites.
between 30 and 50 ka calendar years BP. The Tables 2 and 3 present the results of these
current study seeks both to test the Danube measurements as well as important previously
Corridor and the Kulturpumpe hypotheses as well published results. While the majority of the dates
as to examine the strengths and limitations of the included here are AMS dates, some conventional
radiocarbon dating in the period before 30 ka BP. radiocarbon dates have also been included. We
have only excluded previous dates from unmodi-
fied cave bear bones, which, in general, appear to
Results have accumulated independent of human activi-
ties, and several dates on mixed bone samples
To address these issues the accelerator radio- submitted by J. Hahn before the advent of AMS
carbon facilities of the University of Kiel, Purdue dating (Housley et al., 1997). Although Richter
and Oxford Universities produced 49 new radio- et al. (2000) suggest that systematic differences
carbon measurements on bones from Bockstein- exist between conventional and AMS dates, in our
Törle, Hohlenstein-Bärenhöhle, Hohlenstein- view when errors in collecting and processing
Stadel, and Vogelherd in the Lone Valley and samples are excluded, conventional and AMS
Geißenklösterle, Sirgenstein, and Hohle Fels in the dates are comparable. Similarly we see no reason
Ach Valley. The majority of the newly dated to assume that dates on carefully prepared samples
specimens showed clear anthropogenic modifica- of bone, antler and charcoal should not be com-
tions including impact fractures and cut marks. parable (Jöris et al., 2001). Thus we include exist-
Several additional dates were obtained directly ing conventional dates along with the previous and
from bone artifacts. With the exception of one date new AMS dates on bone, antler and charcoal in
on reindeer antler, all of the dates were made on Tables 2 and 3. At present there are no data from
well preserved bone, and in every case the yield of the Swabian Jura indicating that large systematic
collagen was significantly high to produce a reli- errors preclude comparisons between these
able date. Ivory artifacts were not dated to maxi- materials. The sample preparation and collagen
mize the comparability between the dates. In extraction using the Kiel, Purdue and Oxford
several cases fresh breaks on the faunal remains protocols successfully removes lipids and carbon-
appear to be the result of anthropogenic bone ates through a combination of the use of organic
cracking for the extraction of marrow, but damage solvents and acid and base washes (Hedges and
by large carnivores cannot completely be excluded van Klinken, 1992; Longin, 1970; Hedges et al.,
as possible agents of modification. Hyenas, the 1989; Grootes, pers. comm. 2000). There is no
main non-human species associated with bone reason to believe that the previous handling of the
cracking (Zapfe, 1939), are absent or extremely specimens or contamination with calcium carbon-
rare in the faunal assemblages under study. The ate led to anomalous ages. This is not a trivial
specimens from fieldwork conducted by Hahn and point since the specimens from sites including
other researchers after 1973 were piece-plotted Sirgenstein and Vogelherd have been handled to
during excavation to the nearest centimeter in varying degrees since their excavation in the first
three dimensions, and thus their exact provenience half of the 20th century.336 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
Table 2
14
C-dates with 1 uncertainties for the Lone Valley sites. The dates from Groningen (GrN) and Heidelberg (H) are conventional
radiocarbon dates, those from Kiel (KIA), Purdue (PL), and Zurich (ETH) are AMS dates
Lab. number Arch. horizon Material Modification Date Cultural group First publication
Lone Valley
Bockstein-Törle
H 4058-3355 VI mixed bone 20 400220 Aurig./Grav. Hahn, 1977
sample
KIA 8956 VI long bone frgt. fresh break 20 990+120/110 Aurig./Grav.
H 4058-3526 VI mixed bone 23 440290 Aurig./Grav. Hahn, 1983
sample
KIA 8953 VI reindeer fresh break 31 530230 Aurig./Grav.
radius-ulna
H 4049-3356 VII mixed bone 26 133376 Aurignacian Hahn, 1977
sample
KIA 8952 VII reindeer fresh break 30 130+260/250 Aurignacian
metatarsal
H 4059-3527 VII mixed bone 31 965790 Aurignacian Hahn, 1983
sample
KIA 8954 VII reindeer femur? fresh break 44 390+990/880 Aurignacian/MP?
KIA 8955 VII horse metapodial fresh break 46 380+1360/1170 Aurignacian/MP?
Hohlenstein-Bärenhöhle
KIA 8967 brown loam rib fresh break 26 080+140/130 Aurignacian?
Hohlenstein-Stadel
KIA 8951 19 m, spit 6 reindeer humerus impact 31 440250 Aurignacian
H 3800-3025 20 m, spit 6 mixed bone 31 750+1150/650 Aurignacian Hahn, 1977
sample
ETH-2877 20 m, spit 6 reind. ulna + wolf 32 000550 Aurignacian Schmid, 1989
astrag.
KIA 13077 20 m, spit 6 reindeer radius fresh break 32 270+270/260 Aurignacian
KIA 8949 19 m, spit 7 reindeer? fresh break 33 920+310/300 Aurignacian
longbone
KIA 8950 19 m, spit 7 elk metatarsal fresh break 36 910+490/460 Aurignacian
KIA 8948 19 m, spit 8 horse? longbone impact 41 710+570/530 Aurignacian?
KIA 8947 19 m, spit 9 horse longbone fresh break 42 410+670/620 Aurignacian?
KIA 8946 19 m, spit 10 reindeer fresh break 39 970+490/460 Aurignacian?
metapodial
KIA 8945 19 m, spit 11 longbone fresh break 40 220+550/510 Aurignacian?
Vogelherd
KIA 8957 IV long bone frgt. cutmarks 26 160150 Aurignacian?
H 4053-3211 IV mixed bone 30 730750 Aurignacian Hahn, 1977
sample
GrN-6662 IV/V charred bone 27 630830 Aurignacian? Hahn, 1977
PL0001339A IV/V horse tibia cutmarks+ 32 180960 Aurignacian
fresh break
PL0001342A IV/V bovid-horse rib cutmarks 34 1001100 Aurignacian
H 8498-8950 V mixed bone 25 900260 Aurignacian? Hahn, 1993b
sample
H 8497-8930 V mixed bone 27 200400 Aurignacian? Hahn, 1993b
sample
H 4054-3210 V mixed bone 30 1621340 Aurignacian Hahn, 1977
sample
H 8500-8992 V mixed bone 30 6001700 Aurignacian Hahn, 1993b
sampleN.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 337
Table 2 (continued)
Lab. number Arch. horizon Material Modification Date Cultural group First publication
GrN-6661 V charred bone 30 650560 Aurignacian Hahn, 1977
H 8499-8991 V mixed bone 31 3501120 Aurignacian Hahn, 1993b
sample
KIA 8968 V tibia impact 31 790240 Aurignacian
H 4056-3208 V mixed bone 31 9001100 Aurignacian Hahn, 1977
sample
PL0001338A V horse tibia cutmarks 32 4001700 Aurignacian
KIA 8969 V reindeer long impact 32 500+260/250 Aurignacian
bone
KIA 8970 V horse long bone impact 33 080+320/310 Aurignacian
PL0001337A V bovid-horse cutmarks 35 810710 Aurignacian
longbone
Although, as discussed below, considerable remains of mixed ages. Hahn at times submitted
variation in atmospheric radiocarbon concen- such mixed samples to avoid destroying finds that
trations have been documented during the Late he considered too valuable to date by conventional
Pleistocene, in our view accurate calibration means. A degree of mixing between strata is not
parameters are not yet available for the critical surprising given that the deposits of the cave were
period between 30,000 and 50,000 calendar years excavated over a period of less then three months
ago (Richards and Beck, 2001). Thus, we present in the summer and fall of 1931 before careful
the radiocarbon ages in years before 1950 AD excavation methods and detailed studies of site
based on the Libby half-life without attempting to formation processes were common. AMS dates of
calibrate them. single bones from layer IV have also yielded dates
of Magdalenian age, indicating that Riek’s excava-
Lone Valley tion techniques did not succeed at rigorously sep-
arating the archaeological units. Several dates,
Dates from the Lone Valley provide new insight including one new AMS date of ca. 26 ka BP,
into the chronostratigraphy of the area. Gustav suggest that a Gravettian component is also
Riek’s excavation from 1931 at Vogelherd (Riek, present at Vogelherd.
1934) documented the rich Aurignacian layers V Robert Wetzel’s excavations at Hohlenstein-
and IV, which contain numerous organic tools and Bärenhöhle running mainly from 1956–1961
a dozen small figurines carved from mammoth yielded a small artifact assemblage from a brown
ivory (Fig. 2). With the exception of one find from loam with isolated Aurignacian elements (Fig. 3)
layer IV, the many split based bone points from including a carinated and a nosed end scraper, six
the site stem from layer V (Hahn, 1977; Riek, carinated burins, and one bone burnisher (Hahn,
1934). The wealth of finds indicates that the cave 1977). Hahn emphasizes the complex taphonomy
was occupied repeatedly during the Aurignacian. and partially reworked stratigraphy of this site.
Although the assemblages from these layers con- The Aurignacian layer overlies a richer Middle
tain some younger materials, all but one of the Paleolithic deposit (Beck, 1999). A single date
eight new AMS dates fall within the expected from the former layer yielded an age of ca. 26 ka
range for the Swabian Aurignacian. The newly BP and represents the first attempt to date the
dated finds from layer V yielded ages between 31 layer. This age lies well outside the expected range
and 36 ka BP. These results tend to predate earlier of the region’s Aurignacian and suggests the pres-
conventional radiocarbon dates from Vogelherd. ence of a Gravettian component at the site. Fur-
The relatively young conventional dates may well ther work is needed to date the small Aurignacian
result from bulk sampling of many small faunal assemblage from this site.338 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
Table 3
14
C-dates with 1 uncertainties for the Ach Valley sites. The dates from Bern (B), Heidelberg (H), and Pretoria (Pta) are
conventional radiocarbon dates, those from Kiel (KIA), Oxford (OxA), and Zurich (ETH) are AMS dates
Lab. number Arch. hor. Material Modification Date Cultural group First publication
Ach Valley
Geißenklösterle
OxA-5157 Ip hare pelvis 24 360380 Gravettian Housley et al.,
1997
OxA-4855 Ir reindeer phalange 27 000550 Gravettian Housley et al.,
1997
OxA-4857 Ir horse rib cutmarks 27 500550 Gravettian Housley et al.,
1997
OxA-4856 Ir horse radius 30 950800 Gravettian Housley et al.,
1997
OxA-5227 Is horse femur 28 050550 Gravettian Housley et al.,
1997
OxA-5226 It reindeer tibia impact 26 540460 Gravettian Housley et al.,
1997
OxA-5229 It mammoth rib cutmarks 27 950550 Gravettian Housley et al.,
1997
OxA-5228 It mammoth rib 28 500550 Gravettian Housley et al.,
1997
OxA-4592 It reindeer phalange 29 200460 Gravettian Hahn, 1995
OxA-4593 It bone 29 200500 Gravettian Hahn, 1995
OxA-5706 Ia red deer antler 29 220500 Gravettian Richter et al., 2000
OxA-5161 Ic reindeer impact 30 300750 Gravettian Housley et al.,
metacarpal 1997
H 4147-3346 IIa mixed bone 30 625796 Upper Hahn, 1983
sample Aurignacian
H 4279-3534 IIa mixed bone 31 525770 Upper Hahn, 1983
sample Aurignacian
OxA-5707 IIa horse scapula impact + 33 200800 Upper Richter et al., 2000
cutmarks Aurignacian
OxA-5160 IIa hare tibia 33 7001100 Upper Hahn, 1988
Aurignacian
OxA-4594 IIa reindeer? humerus 36 8001000 Upper Hahn, 1995
Aurignacian
KIA 8960 IIb mammoth rib impact 29 800240 Upper
Aurignacian
Pta-2361 IIb charred bone 31 070750 Upper Hahn, 1983
Aurignacian
KIA 8958 IIb horse humerus impact 31 870+260/250 Upper
Aurignacian
Pta-2270 IIb charred bone 31 8701000 Upper Hahn, 1983
Aurignacian
OxA-5708 IIb mammoth 32 300700 Upper Richter et al., 2000
cranium Aurignacian
PtA-2116 IIb charred bone 32 680470 Upper Hahn, 1983
Aurignacian
OxA-5162 IIb hare pelvis 33 2001100 Upper Housley et al.,
Aurignacian 1997
H 4751-4404 IIb mixed bone 33 700825 Upper Hahn, 1983
sample Aurignacian
OxA-6256 III reindeer tibia impact 30 100550 Lower
AurignacianN.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 339
Table 3 (continued)
Lab. number Arch. hor. Material Modification Date Cultural group First publication
KIA 8963 III long bone impact 31 180+270/260 Lower
Aurignacian
H 5118-4600 III mixed bone 34 1401000 Lower Hahn, 1983
sample Aurignacian
H 5316-4909 III mixed bone 36 5401570 Lower Hahn, 1983
sample Aurignacian
OxA-5163 III ibex mandible 37 3001800 Lower Housley et al.,
Aurignacian 1997
OxA-4595 III horse femur 40 2001600 Lower Hahn, 1995
Aurignacian
OxA-6629 IIIa reindeer 30 300550 Lower
metatarsal Aurignacian
OxA-6628 IIIa reindeer 30 450550 Lower
metatarsal Aurignacian
ETH-8268 IIIa bone 33 100680 Lower Hahn, 1995
Aurignacian
OxA-5705 IIIa reindeer 33 1501000 Lower
metatarsal Aurignacian
ETH-8269 IIIa bone 33 500640 Lower Hahn, 1995
Aurignacian
OxA-6255 IIIa rhino humerus 32 900850 Lower
Aurignacian
KIA 13075 IIIa reindeer tibia impact 34 330+310/300 Lower
Aurignacian
KIA 13074 IIIa reindeer tibia impact 34 800+290/280 Lower
Aurignacian
ETH-8267 IIIa bone 37 8001050 Lower Hahn, 1995
Aurignacian
KIA 8962 IIIb rib impact 28 640+380/360 Lower
Aurignacian
KIA 8961 IIIb reindeer humerus fresh break 33 210+300/290 Lower
Aurignacian
KIA 13076 IIIb reindeer tibia impact + 34 080+300/290 Lower
cutmarks Aurignacian
KIA 8959 IIIb femur fresh break 34 220+310/300 Lower
Aurignacian
KIA 16032 IIIb roe deer impact 36 560+410/390 Lower
metacarpal Aurignacian
OxA-6077 GH 17 ibex tibia 32 050600 sterile
OxA-6076 IV red deer tibia 33 6001900 Middle
Paleolithic
Hohle Fels
OxA-4599 IIc reindeer antler tool (decor. 28 920400 Gravettian Hahn, 1995
adze)
OxA-5007 IIc reindeer antler tool (decor. 29 550650 Gravettian Housley et al.,
adze) 1997
KIA 8964 IId rib 29 560+240/230 Aurignacian
rhino-mammoth
KIA 8965 IId reindeer antler 30 010220 Aurignacian
KIA 16040 IIe horse pelvis impact + 30 640190 Aurignacian
cutmarks
OxA-4979 III Salix charcoal 27 600800 Aurignacian Housley et al.,
1997340 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
Table 3 (continued)
Lab. number Arch. hor. Material Modification Date Cultural group First publication
OxA-4601 III bone 30 550550 Aurignacian Hahn, 1995
KIA 16038 III reindeer femur impact + 29 840210 Aurignacian
cutmarks
KIA 16039 III reindeer tibia impact 31 140+250/240 Aurignacian
OxA-4980 IV Salix + Betula 28 75050 Aurignacian Housley et al.,
charcoal 1997
OxA-4600 IV reindeer 31 100600 Aurignacian Hahn, 1995
metapodial
KIA 16035 IV horse femur tool 33 090+260/250 Aurignacian
(retoucher)
Sirgenstein
KIA 13079 II bone tool (point) 27 250+180/170 Gravettian
KIA 13080 III bone tool 30 210220 Aurig./Grav.
(burnisher)
KIA 13081 IV mammoth rib tool 28 400200 Aurignacian
(burnisher)
KIA 13082 V bone tool (point) 26 730+170/160 Aurignacian
KIA 13083 VI bone tool (awl) 30 360+230/220 Aurignacian
Brillenhöhle
B-492 VII charred bone >25 000 Gravettian Riek, 1973
B-491 VIII charred bone >29 000 ? Riek, 1973
Five new AMS dates on material from Wetzel’s Gravettian sequence and to test whether or not
excavations in 1953, 1955, and 1956 at Bockstein- the region was occupied during the Last Glacial
Törle provide radiometric ages from this site Maximum. Borges de Magalhães’s (2000) reanaly-
which yielded small Aurignacian and Gravettian sis of the assemblages from Bockstein-Törle indi-
assemblages (Hahn, 1977; Wetzel, 1954). Table 2 cates that layer VII belongs to the Aurignacian
includes four previous radiocarbon dates from and that, based on the presence of backed blades
mixed bone samples. The majority of these dates and the absence of Aurignacian artifacts, layers
appears to be too young, perhaps as a result of VI-IV should be attributed to the Gravettian. The
errors in sampling. A reindeer metatarsus that was attribution of layer VII to the Aurignacian rests
cracked open in a fresh state and one mixed bone mainly on the presence of multiple bone points and
sample produced radiocarbon ages between 30– carinated and busked burins (Fig. 4). While single
32 ka BP for the Aurignacian layer VII. An AMS carefully made bone points are known from both
date on a reindeer radius and ulna with fresh the Middle Paleolithic of Vogelherd and the Große
breaks from Layer VI, which Hahn described as Grotte, they are rare prior to the Aurignacian.
either Aurignacian or Gravettian, yielded an age of Similarly, the diverse forms of carinated and
ca. 31.5 ka BP. Three other dates from layer VI busked burins from layer VII are unknown in
including one AMS measurement gave ages the Swabian Middle Paleolithic. Two AMS
between 20 and 24 ka BP suggesting the use of the dates from layer VII yielded ages in excess of 44 ka
site in the period shortly preceding and perhaps BP. Rather than advocate an extremely early
during the Last Glacial Maximum. More sys- Aurignacian, we view these dates as evidence
tematic study of the later phases of the for the poor separation of the Aurignacian and
Swabian Gravettian are needed to establish the underlying Middle Paleolithic deposits from layer
chronostratigraphic framework for the region’s VIII. While at most other Swabian sites a sterileN.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 341 Fig. 2. Aurignacian of Vogelherd cave. (5, 7–9, 15–16): Vogelherd IV-(1–4, 6, 10–14, 17–21): Vogelherd V. (1–2) carinated end scrapers-(3, 5, 7) nosed end scrapers-(4) busked burin-(6) pointed blade-(8) splintered piece-(9–10) laterally retouched blades-(11) carinated burin-(12) double burin on truncation-(13–14) ivory figurines-(15) bone awl-(16) bone decorated on both sides-(17–19) bone points with split bases-(20) retoucher made of a cave bear canine-(21) bâton percé of ivory. After Hahn, 1977 (1–12, 15–21); drawing (13–14): A. Frey.
342 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Fig. 3. Aurignacian of Hohlenstein-Bärenhöhle. (1, 6) carinated burins-(2) double burin-(3) double end scraper-(4) side scraper-(5) retouched blade. After Hahn, 1977. horizon separates the latest Middle Paleolithic and spit in which the Löwenmensch lay, fall in the range earliest Aurignacian layers, this is not the case at of ca. 32 ka BP (Schmid, 1989). New dates from Bockstein-Törle (Wetzel, 1954). this spit confirm these dates. Modified bones from The Aurignacian finds from Ludwig Bürger’s the underlying 20 cm spits 7–11 yielded dates excavation at Bocksteinhöhle in 1883–84 (Bürger, from ca. 34–42 ka BP. Taken at face value these 1892) have not yet been dated. Although this dates along with dates from Geißenklösterle, sug- assemblage contains a split-based point, two per- gest a continual human presence in the region in forated bear canines and typical Aurignacian lithic the critical period around 40 ka. These samples artifacts (Fig. 4), due to the early date of the come from the rear of the cave in an area rich in excavation, little information about the stratigra- anthropogenically processed faunal remains but phy of the site is available (Hahn, 1977; Schmidt, poor in diagnostic artifacts. Middle Paleolithic 1912). finds are lacking, but the cracked bones cannot be Considerable emphasis was placed on dating readily attributed to a cultural group. A single core the deposits from Robert Wetzel and Otto from spit 9 is a small blade core (Fig. 5.18) and Völzing’s excavations before and after World War suggests an association with the Upper Paleolithic. II at Hohlenstein-Stadel, best known for its Middle Paleolithic and Aurignacian deposits. Par- Ach Valley ticularly noteworthy is the presence of the anthro- pomorphic Löwenmensch figurine, which was New dates from the early Upper Paleolithic excavated in 1939 from the sixth 20 cm spit from of the Achtal focused on materials from Robert 20 meters deep inside this tunnel-shaped cave. Rudolf Schmidt’s 1906 excavation at the Wetzel’s excavation also yielded an assemblage of Sirgenstein and the ongoing excavations at Hohle Aurignacian lithic and organic artifacts (Fig. 5). Fels and Geißenklösterle. Beginning with Previous conventional radiocarbon dates, from the Schmidt’s work at Sirgenstein, the now traditional
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 343 Fig. 4. Aurignacian of Bockstein-Törle VII (1–15) and Bocksteinhöhle (16–19). (1–2, 16) busked burins-(3, 6) carinated burins-(4, 8) end scrapers-(5) splintered piece-(7, 10) laterally retouched blades-(9) perforated tooth-(11–12) ivory rods-(13–14) bone points with massive bases-(15) distal bone point fragment-(17) end scraper-burin-(18) perforated cave bear canine-(19) bone point with split base. After Hahn, 1977.
344 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Fig. 5. Aurignacian of Hohlenstein-Stadel. (1–17, 19–20): max. depth 1,20 m-(18): depth 1,60 m-1,80 m. (1–5) perforated fox canines-(6, 8, 12) carinated end scrapers-(7) ivory bead-(9–10, 13) burins on truncation-(11, 17) laterally retouched blades-(14) bone retoucher-(15) burnisher-(16, 19) bone points-(18) blade core-(20) ivory figurine. After Hahn, 1977 (1–7, 9, 13–17, 19), Hahn in Schmid, 1989 (8, 10–12), and Schmid, 1989 (20).
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 345 nomenclature based on designating archaeologi- tions between 1958 and 1960. The current phase cal units with Roman numerals in increasing of excavation has run semi-continuously since order from top to bottom of a sequence became 1977 under Hahn’s and later under Conard and established. Uerpmann’s direction. Here rich Gravettian find AMS dates were taken on samples from two horizons in archaeological complex II have been bone points, two bone burnishers, and a bone awl documented in detail (Conard et al., 2001; Schiegl from the Aurignacian and Gravettian layers at et al., 2001). The layer IId directly below the Sirgenstein (Hahn, 1977; Schmidt, 1910, 1912). rich Gravettian layer IIc yielded a mammoth The dates fall in the range of ca. 27–30 ka BP. ivory figurine similar to those known from the They suggest a degree of mixing between the Aurignacian of the Swabian Jura. Both dated archaeological units, but are consistent with the bones from layer IId produced dates of ca. 30 ka cultural attribution of the assemblages from layers BP and suggest a degree of temporal continuity II–VI to the Gravettian and Aurignacian. While between the Aurignacian and Gravettian (Conard split-based bone points are lacking, bone tools and and Floss, 2000). Dates of materials from the carinated and nosed end scrapers are present in underlying layers III and IV lie in the range of layers III–V (Fig. 6). The dates from Sirgenstein ca. 28–33 ka BP and also suggest a continuous suggest a continual occupation of the Swabian occupation between the Aurignacian and Jura between the Aurignacian and Gravettian. Gravettian. Excavations in 2001 and 2002 demon- Thus far no new dates have been obtained from strated the presence of rich Aurignacian layers the small Aurignacian assemblage, which contains in still deeper deposits, but these finds have yet to two bone points (Fig. 7.1–2), or from the much be dated. Aurignacian finds from Hohle Fels richer overlying Gravettian horizons (Fig. 7.3–35) include diverse carinated and busked burins, from Riek’s excavations at Brillenhöhle from nosed and carinated scrapers, laterally retouched 1955–1963 (Riek, 1973). The only two radiocarbon Aurignacian blades, ivory beads and pendants, dates available at present are conventional and diverse bone tools (Fig. 8). Particularly measurements that provide a minimum age of remarkable is the rich assemblage from archaeo- 25 ka for burnt bone from the Gravettian fireplace logical horizon IV, which contains an ivory of layer VII and a minimum age of 29 ka for burnt figurine depicting a bird, numerous ivory orna- bone from the underlying fireplace at the top of ments and much refuse from ivory working layer VIII. Two pairs of lithic refits between layer (Conard et al., 2002). VII at Brillenhöhle with artifacts from layer IIb at The main Gravettian deposits from Hohle Fels Hohle Fels, as well as six lithic refitting complexes date to 29 ka BP, although two conventional dates between layer VII at Brillenhöhle and well dated on mixed samples of bones submitted by Hahn Gravettian deposits at Geißenklösterle, layers (1981, 1983) have produced much younger ages. Ia, Ib and It, demonstrate the existence of a The Gravettian assemblages from Hohle Fels and Gravettian occupation at about 29 ka BP in other Swabian sites are usually easy to distinguish Brillenhöhle (Scheer, 1986, 1993). Although Riek from Aurignacian assemblages (Hahn, 1992; did not publish the exact position of the two bone Scheer, 1985, 2000). They usually lack typical points from layer XIV, they lay at least 70 cm and Aurignacian scrapers, burins and laterally re- perhaps as much as 170 cm below the fireplace of touched pieces. The Gravettian assemblages are layer VIII. Unless one advocates an unusually high characterized by increased bi-directional opposing rate of sedimentation, the two bone points from platform blade production, Gravette and micro- the Aurignacian layer XIV must be considerably Gravette points, backed blades and bladelets, older then 29 ka BP. and Font Robert points. Ivory tear-drop-shaped The excavations at Hohle Fels have a long pendants, perforated teeth, and bone tools are history dating back to Oscar Fraas’s and Theodor common, while figurative art is thus far limited to Hartmann’s work in the 1870 s and include engravings (Fig. 9). Particularly noteworthy are Gertrud Matschak’s and Gustav Riek’s excava- A. Scheer’s (1986, 1993) successful refittings of
346 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Fig. 6. Aurignacian of Sirgenstein cave. (1–2, 5–6, 9, 11–12, 14, 21): Sirgenstein IV-(3–4, 7, 10, 13, 16, 18–20): Sirgenstein V-(8, 15, 17): Sirgenstein VI. (1–2, 7) carinated end scrapers-(3, 6) nosed end scrapers-(4) pointed blade-(5) end scraper-(8, 12) dihedral burins (9, 13) burins on truncation-(10) borer-(11) ivory bead-(14) carinated burin-(15) bone awl-(16) truncated blade-(17) laterally retouched blade-(18) splintered piece-(19–20) bone points-(21) burnisher. After Hahn, 1977.
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 347 Fig. 7. Aurignacian of Brillenhöhle XIV (1–2) and Gravettian of Brillenhöhle VII (3–35). (1) bone point (with split base?)-(2) bone point-(3–5) Gravette points-(6–10) micro-Gravette points-(11–12, 18) end scrapers-(13, 16) burins-(14–15) burins combined with end scrapers-(17) ventrally retouched bladelet-(9) truncated blade-(20) blade retouched on both ends-(21) fragment of an ivory figurine-(22–24, 26–27) teardrop-shaped ivory pendants-(25, 28) perforated canines-(29) bone tube-(30–31, 35) bone projectile points-(32) bone rod (decorated?) with incisions-(33) bone awl-(34) bâton percé of ivory. After Hahn, 1977 (1–2) and Riek, 1973 (3–35).
348 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Fig. 8. Aurignacian of Hohle Fels near Schelklingen. (1, 9–11, 15, 19, 21): AH III-(2–8, 12–14, 16–18, 20): AH IV-(22–23): AH V. (1) perforated bear incisor-(2) perforated upper eyetooth from red deer-(3) roughout for ivory beads-(4) half-finished ivory bead-(5–7) double perforated ivory beads-(8–9) burins-(10) truncated blade-(11) busked burin-(12–13) disc-shaped ivory beads-(14) pointed blade-(15) carinated burin-(16) double nosed end scraper-(17) blade with Aurignacian retouch-(18) blade pointed at one end and truncated at the other-(19) bone awl with intense polishing-(20) fragment of a bone point-(21) worked mammoth rib-(22) nosed end scraper-(23) end scraper combined with a pointed end. After Conard et al., 2002 (1–7, 11–18, 22–23); drawings by D. Punčochář.
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 349 Fig. 9. Gravettian of Hohle Fels near Schelklingen. (1–4) Gravette points-(5) micro-Gravette point-(6–7) backed bladelets-(8) Font Robert point (9, 13) burins-(10) borer-(11) end scraper-(12) bladelet core-(14) decorated bone point-(15) pointed blade-(16) decorated hare humerus-(17–19) ivory pendants-(20) bâton percé of antler-(21) pendant made of a bear canine-(22) decorated antler adze. After Conard et al., 2001 (1–3, 11–12), Conard and Uerpmann, 1999 (4–5, 7–8, 13–14, 16–19, 21), Conard et al., 2000 (6, 9–10, 15), and Scheer, 1994 (22).
350 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Gravettian lithic artifacts between Hohle Fels, cobbles A9 and A14, that about 7% of the artifacts Geißenklösterle and Brillenhöhle mentioned have moved between archaeological horizons II above, thereby demonstrating that these sites were and III, and that 60% of the pieces have not moved used contemporaneously around 29 ka BP. No from their original subunit. The remaining 33% of clear stratigraphic or chronological break is visible the material have moved between subunits of between earliest Gravettian and latest Aurignacian horizon III. Of these finds 29% have migrated horizons. upward and 4% have migrated downward (Hahn, Geißenklösterle is the site from the Swabian 1988: 74). In Hahn’s view while the mixing is not Jura that has been the focus of the most lively trivial, the Aurignacian deposits have not lost their debate on the dating of the Aurignacian (Bolus stratigraphic integrity. After carefully consulting and Conard, 2001; Hahn, 1988; Richter et al., the documentation from Hahn’s excavation, we 2000; Zilhão and d’Errico, 1999). Fieldwork by conclude, as Hahn (1988: 84) himself alluded to, Hahn and others from 1973 – 1991 and Conard that “excavator error”—that is the inability of ex- and colleagues starting in 2000 have yielded the cavators to clearly separate stratigraphic horizons— best studied sequence in the region. Excluding contributed to apparent mixing between archaeo- obvious outliers, such as isolated Magdalenian logical horizons. Hahn also consistently refrained dates from the uppermost Gravettian, and dates from correcting false stratigraphic assessments on cave bear bones and other non-archaeological made during excavation. Given that the layers materials, 47 radiocarbon dates are available on of the site are characterized by subtly stratified finds from the Middle Paleolithic horizon IV, deposits of limestone rubble in a silty matrix, it the lower Aurignacian complex III, the upper is indeed difficult to achieve a completely clean Aurignacian complex II, and the Gravettian separation between the archaeological horizons. complex I. Both Hahn’s excavations and the current exca- The lower Aurignacian (Fig. 10) is character- vations have documented the presence of well ized by typical Aurignacian unidirectional blade defined horizontal features including lithic scatters, production and over 200 stone tools, including areas rich in worked bone and ivory, and concen- numerous carinated and nosed end scrapers, trations of burnt bone, ash and ochre. Such burins and splintered pieces, as well as worked features are largely intact and provide additional bone, ivory and antler artifacts (Hahn, 1988). evidence for a lack of large scale mixing between Artworks and flutes, which are present in the the main Aurignacian horizons (Conard and upper Aurignacian, are thus far lacking in the Malina, 2002; Hahn, 1988, 1989). excavation of the lower Aurignacian of horizon Fortunately over 30,000 piece-plotted objects III. The lower Aurignacian has yielded eight including hundreds of refitted finds make it poss- perforated ornaments. ible to plot profile projections of refitting artifacts. The upper Aurignacian (Fig. 11) includes a In response to claims by Zilhão and d’Errico diverse lithic assemblage rich in splintered pieces, (Zilhão, 2001; Zilhão and d’Errico, 1999) that end scrapers, diverse burins, and very few Dufour considerable mixing has occurred between hori- bladelets. Many organic tools, numerous per- zons II and III, we have examined over 30 refitting forated ornaments, four mammoth ivory sculp- complexes. The plots (Figs. 12–14) demonstrate tures and two small bone flutes have been the outstanding context of the Aurignacian finds recovered from horizon II (Hahn, 1986, 1988; from Geißenklösterle and show that only a small Hahn and Münzel, 1995). portion of the finds underwent significant vertical Hahn’s careful refitting and taphonomic studies displacement. Based on taphonomic grounds it reveal a degree of mixing between the stratigraphic is inconceivable that numerous Aurignacian complexes II and III. Hahn (1988: 48–84) describes elements from horizon III have been reworked at length the possible sources of mixing, including downward from archaeological horizon II. cryoturbation, bioturbation and excavation error Similarly, there is no sign of significant reworking and estimates, based on refitting sequences from of Middle Paleolithic materials upward across the
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 351 Fig. 10. Aurignacian of Geißenklösterle, archaeological horizon III. (1–3) perforated fox canines-(4–5) ivory pendants-(6) ivory bead-(7) grooved bone-(8) tool resembling a carinated end scraper-(9–10) carinated end scrapers-(11–12) nosed end scrapers-(13, 17) burins-(14, 21) bone points-(15) end scraper-(16) splintered piece-(18) ivory rod (projectile point?)-(19) worked ivory splinter-(20) blade core with refitted blades. After Hahn, 1988 (1–2, 4–5, 7–21) and Hahn, 1989 (3, 6).
352 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 Fig. 11. Aurignacian of Geißenklösterle, archaeological horizon II. (1–3) end scrapers-(4) pointed blade-(5) laterally retouched blade-(6, 10–11) splintered pieces-(7) busked burin-(8) burin on truncation-(9) truncated blade-(12) antler pendant-(13) Dufour bladelet-(14, 20, 22) ivory figurines-(15–19) ivory beads-(21) bone flute-(23) decorated bone-(24) bone point with split base-(25) bâton percé of ivory. After Hahn 1986, (14, 20, 22) and Hahn, 1988 (1–13, 15–19, 23–25).
N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 353
largely sterile geological layer 17 that separates the organic tools from horizon II include split-
the uppermost Middle Paleolithic horizon IV based bone points made from reindeer antlers,
from the lowermost Aurignacian horizon III. awls and burnishers made from bone, and mobile
Additionally, unpublished micromorphological art, ornaments, and points without split bases
studies by Gerlinde Dippon and Paul Goldberg made from mammoth ivory (Liolios, 1999). Each
(pers. comm., 2002) also refute Zilhão and of these raw materials reflects a unique method of
d’Errico’s claims that considerable mixing has production, which leads to specific tools as end
taken place between the archaeological horizons at products. In contrast, horizon III is somewhat
Geißenklösterle. poorer in organic artifacts and lacks split-
Beyond the taphonomic arguments against sig- based points and mobile art. Within the lower
nificant mixing between Aurignacian horizons II Aurignacian deposit, there is no rigorous corre-
and III, several key arguments against mixing spondence between raw materials, methods of tool
result from technological and typological production, or the finished tool types. Ivory is also,
studies of the deposits (Hahn, 1988; Liolios and by far, the dominant raw material in horizon III
Teyssandier, in press). For example, Hahn and (Münzel, 1999). Furthermore, perforated carni-
Teyssandier have demonstrated that horizon III is vore teeth are the most common form of orna-
characterized by a complete and coherent chain of ment in horizon III and are lacking in the upper
lithic reduction beginning with whole cobbles and Aurignacian of horizon II.
extending to the systematic uni-directional produc- Due to the critical debate over the stratigra-
tion of blades, which were then modified into phic integrity of the Aurignacian horizons at
diverse tool types. Horizon II, on the other hand, Geißenklösterle, we undertook a new dating initia-
is characterized by incomplete chains of lithic tive that included 17 new AMS measurements
reduction and less intense production of blades. from Kiel and Oxford. These radiocarbon dates
The upper Aurignacian is also characterized by a span the period from ca. 29–37 ka BP (Fig. 15).
broader spectrum of raw materials and a higher Assuming that there are no significant systematic
proportion of more distant raw materials. In con- errors between radiocarbon labs, the new dates
trast to horizon III where complete reduction document relatively early Aurignacian occupations
chains are well documented, horizon II includes in the region. Six 14C dates from three accelerator
more numerous examples of the presence of blades and one conventional lab fall in the range between
that were produced outside the area of excavation 36–40 ka BP. A conventional radiocarbon date
(Liolios and Teyssandier, in press). Teyssandier from a mixed bone sample of 36,0003560 ka BP
also stresses that the lithic artifacts within the (Hahn, 1988) has been excluded from Table 3 due
many sub-units of horizons II and III are balanced to its large standard deviations These early dates
from a typological point of view. Thus carinated are roughly consistent with the mean age of
and nosed scrapers dominate the tool assem- 40.21.5 ka BP based on Richter et al.’s (2000) six
blages in the lower Aurignacian sub-units, while thermoluminescence dates on burnt flints from
splintered pieces, pointed blades, and laterally horizon III. These TL dates range between 38.3
retouched blades are limited almost exclusively and 44.7 ka and have standard deviations between
to the stratigraphic sub-units of the upper 2.1 and 5.6 ka. Based on the taphonomic and
Aurignacian deposits. These clear technological archaeological arguments mentioned above, we
and typological signatures would not be visible if find no basis, at present, to reject these six radio-
significant mixing had occurred at the site. carbon dates. Additionally, the dates cannot be
An examination of the organic artifacts in the rejected simply on the basis of their ages, because
Aurignacian horizons at Geißenklösterle also most forms of contamination or deviations in
documents coherent patterns in the assem- atmospheric radiocarbon content would tend to
blages that are inconsistent with high levels of yield ages that are too young rather than anoma-
taphonomic mixing (Liolios, 1999; Liolios and lously too old. While Hahn did not systematically
Teyssandier, in press; Münzel, 1999). For example, record data on anthropogenic modification ofFig. 12. Geißenklösterle. Aurignacian refitting group A9. Core of Jurassic chert with refitted blades and flakes.
Fig. 13. Geißenklösterle. Aurignacian refitting group A11. Refitted blades of black alpine quartzite.N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 355
Fig. 14. Geißenklösterle. Aurignacian refitting group A16. Refitted blades and flakes of Jurassic chert, among them two tools.
Fig. 15. Geißenklösterle. The stratigraphic position of the AMS radiocarbon samples.
these samples, he deliberately dated specimens quent work by S. Münzel (1999) demonstrates that
from species including horse, reindeer and ibex these species played an important role in the
that he viewed as important game animals. Subse- hunting economies of the Upper Paleolithic at356 N.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371
Geißenklösterle. Noteworthy is the variation in the sions (Baumgartner et al., 1998). Studies of 14C in
dates from the Aurignacian of Geißenklösterle and North Atlantic planktonic foraminifera (Voelker
the fact that the two available AMS radiocarbon et al., 2000), Japanese varves (Kitagawa and van
dates from the Middle Paleolithic lie between 32 der Plicht, 1998), and stalagmites in the Bahamas
and 34 ka, while four ESR dates on ibex and (Beck et al., 2001) show variations in the abun-
rhinoceros teeth yield a mean age of 43.34.0 ka dance of atmospheric radiocarbon during OIS 3.
(Richter et al., 2000). The individual ESR dates Voelker et al. (2000) and Beck et al. (2001) have
range from 35.7 to 52.7 ka with standard devia- documented changes in 14C concentrations in con-
tions of between 4.8 and 7.7 ka. In the following nection with geomagnetic minima and perhaps in
section we address possible explanations for these association with changing patterns of ocean circu-
observations in more detail. lation (Figs. 17, 18). The changes in North Atlantic
Finally the AMS dates for the rich Gravettian planktonic foraminifera reflect extreme peaks in
14
layers of Geißenklösterle (Scheer, 1989, 1993, C production that correspond to temporal off-
2000) range almost exclusively from ca. 27–30 ka sets of more than 6,000 years and perhaps as much
BP with many dates falling in the vicinity of 29 ka as 10,000 years. Given that the marine signal for
14
BP (Fig. 16). No AMS dates on anthropogenically C variation is attenuated due to reservoir effects,
modified material from Gravettian layers postdate we must expect even greater fluctuations in radio-
26 ka BP: The only two radiocarbon dates for the carbon concentrations in terrestrial archives
Gravettian after 26 ka BP are a single conventional including archaeological sites. Richards and Beck
mixed bone sample submitted by Hahn (1983) (2001) have arrived at similar conclusions based on
prior to the advent of routine AMS dating, and an their analysis of the radiocarbon record from a
AMS date on a hare pelvis. This and other bulk stalagmite in the Bahamas.
bone dates must be viewed with caution, since Unfortunately, these major global fluctuations
Hahn often dated numerous, somewhat scattered, in radioisotope production, transport and deposi-
highly fragmentary bones that he considered tion lie in the period from 30 to 50 k calendar years
expendable. Thus at Geißenklösterle, little if any ago when modern humans arrived in Europe and
secure evidence exists for late Gravettian occupa- numerous important innovations of the Upper
tions. While the Lone Valley has yielded few Paleolithic occurred. This period is also of great
Gravettian assemblages, the Ach Valley has pro- importance for studying the late MSA and early
duced multiple Gravettian assemblages with the LSA in Africa and cultural processes including
most intense period of occupation focusing around those related to the early population dynamics in
29 ka BP. Australia and other parts of the Old World. Thus
researchers must view radiocarbon dates in this
period with great caution. The full extent of the
The “Middle Paleolithic Dating Anomaly” and the chronostratigraphic problems researchers face
“Coexistence Effect” when using 14C must be acknowledged before
better interpretations of the origins of modern
The data summarized above demonstrate that humans in Europe can be developed. This problem
the seemingly mundane aspects of building local is particularly acute in the period around 40 ka
chronostratigraphic sequences must be addressed calendar years ago, near the time of the Laschamp
with rigor and in the knowledge that simple magnetic excursion and the probable arrival of
answers cannot be expected. The available data modern humans in some parts of Europe (Laj
show enormous fluctuations in the production and et al., 2002). Similar, but apparently less extreme
deposition of radioisotopes in various media over variations in radiocarbon production occur several
the period from 30–50 k calendar years ago. thousand years later, perhaps in connection with
Recent studies document major peaks in the pro- the Mono Lake excursion.
duction of 36Cl and 10Be in connection with the For years researchers in Tübingen sought to
Mono Lake and Laschamp geomagnetic excur- explain the odd patterns in the radiocarbonN.J. Conard, M. Bolus / Journal of Human Evolution 44 (2003) 331–371 357 Fig. 16. Gravettian of Geißenklösterle. (1–2) micro-Gravette points-(3–10) backed bladelets-(11) flechette-(12, 14) end scrapers-(13, 15–17) burins-(18) refitted blade sequence-(19–22, 24–26) ivory pendants-(23) perforated fox canine-(27) bone demonstrating the use of groove and splinter technique-(28-29) needle-like tool fragments of bone-(30) burnisher-(31) shaft segment of antler-(32) antler point with incisions. After Hahn et al., 1985 (1–10, 12, 14–15, 17–18, 20, 22, 25–31) and Scheer, 1989 (11, 13, 16, 19, 21, 23–24, 32).
You can also read
