Conference PaperPDF Available

1997-2005 Research in the cave of Goyet (Gesves, province of Namur, Belgium)

Conference Paper

1997-2005 Research in the cave of Goyet (Gesves, province of Namur, Belgium)

B. Demarsin & M. Otte (dir.). Neanderthals in Europe.
Proceedings of the International Conference held in the Gallo-Roman Museum in Tongeren (September 17-19th 2004).
Liège, ERAUL 117 - Tongeren, ATVATVCA 2, 2006, p. 115-134.
1997-2005 RESEARCH IN THE CAVES OF GOYET
(GESVES, PROVINCE OF NAMUR, BELGIUM)
Tongeren Neandertal symposium excursion, 19 September 2004
Michel TOUSSAINT
Direction de l’Archéologie, Ministère de la Région wallonne, rue des Brigades d’Irlande 1, B-5100 Namur. m.toussaint@mrw.wallonie.be
Introduction
The caves of Goyet are one of the main prehistoric sites of
Belgium, as much because of their role in the development of
regional prehistoric studies as for the number of Palaeolithic
facies represented and the wealth of material recovered. The
scienti c importance of the caves has been acknowledged
since 1868, when E. Dupont, a geologist from Dinant (1841-
1911; Twiesselmann 1952), renowned for his excavation
campaign successes in the caves of the Lesse valley, initiated
the rst excavation (Toussaint 2001).
The Goyet cave system is located at Gesves, in the province
of Namur (Belgium; g. 1). It comprises three areas of
major prehistoric interest complemented by adjacent areas
that all stretch along the right bank of the Samson river, at
the con uence of the Strud (Strouvia) stream ( g. 2). The
Lambert coordinates of the centre of the main terrace are: x =
195.71 km; y = 126.20 km (I.G.N. map 48/5).
Numerous archaeological, palaeontological and palaeo-
anthropological artifacts were found in the different caves of
the site but, for the most part, with scant regard for stratigraphy
or plan drawing. Like other emblematic karstic caves of the
national prehistory (e.g. Engis, Spy and Fonds de Forêt) the
caves of Goyet were excavated too early, at a time when
researchers were mainly looking for archaeological material,
at best within some semblance of a sequence of deposits, but
with hardly any interest either in stratigraphic subtleties or
sedimentology and palaeoenvironment (Toussaint & Pirson
in press).
Dupont himself, although a better stratigrapher than his
contemporaries, was more than casual when he described
- 115 -
the layers he identi ed at Goyet, be it in the "Third Cave"
or at Trou du Moulin. Since he had previously done some
remarkable studies of cave deposits, notably at La Naulette
(Hulsonniaux-Houyet), one can assume that he was already
more concerned by his new duties as director of the Natural
History Museum, in Brussels, than by furthering his cave
research. It also seems that at that time, the last period of his
eldwork, technicians did most of the work on his behalf.
At the same period, Dupont would nonetheless include the
tools from the three upper ossiferous layers he thought he
had identi ed at Goyet in his attempt at classi cation of
Belgian prehistoric industries. The elements of this broad
framework, in which precise typology is restricted to the
most characteristic tools, are built on the distinction between
six successive phases: layers of Hastière, Montaigle and Trou
Magrite at the bottom (dating from the "Mammoth Age"),
layer of Goyet with intermediate features, layer of Chaleux-
Furfooz ("Reindeer Age") and last, polished stone age, found
in the sediments from the "present age".
The many later excavations both on the terrace immediately
outside the caves and within the different caves, either by
individual collectors or large national institutions, were also
perfunctory in character; and their context is badly known.
No layer-based inventory of the documents collected by
the successive diggers is available, and no distribution
maps of the nds. The material is scattered in several
private collections, museums and institutions - as well as
outside Belgium. Laboratory analyses, conducted more
than a century after the rst excavations, represent the
only signi cant work (essentially M. Ulrix-Closset (1975),
M. Otte (1979) and M. Dewez (1987) for the archaeology
and M. Germonpré (1996, 1997, 2000, 2001, 2004) for the
palaeontology).
Abstract: Excavated since 1868, generally without any method, the caves of Goyet are comprised of three principal areas of archaeological
interest: the terrace and its caves entrances, the Upper Shelter and Trou du Moulin. They yielded rich occupations from the Middle Palaeolithic,
Aurignacian, Gravettian and Magdalenian, as well as traces from later periods: Protohistory, Roman and Middle Ages. Bone and teeth remains
from the rst excavations (1868-70) and recently identi ed in the collections from the former excavations might be attributed to Neandertal
Man.
- 116 -
Michel TOUSSAINT
This sorry state of affairs prompted the resumption of eld
research at the Goyet caves in 1997 by the "Direction de
l’Archéologie du Ministère de la Région wallonne", in
conjunction with different non-pro t organisations, especially
"l’Association wallonne d’Anthropologie préhistorique"
(Toussaint et al. 1998, 1999, 2004). The aim was to assess
if some stratigraphic and palaeoenvironmental data was still
recoverable from possibly undisturbed sediments on the
terrace and in the numerous local caves, and to make new
archaeological and palaeoanthropological discoveries at this
ill-treated site.
The focus of the last eight years of modern eld research was
on the following aspects:
- a series of 11 stratigraphic trenches (1997-1998) down to
bedrock in the main terrace and in Cave no. 3, where Dupont
made most of his discoveries ( g. 3);
- multidisciplinary research (1977-1999) at the Upper
Shelter, which, unfortunately, greedy collectors looted almost
extensively half a century ago;
- from 1997 onwards, exploration of the newly discovered
cave system located between Trou du Moulin and the Upper
Shelter, which yielded a Neolithic burial, excavated and
studied in situ from 1998 to 2004;
- digging and sieving in the main caves, notably in the
"wastebin chamber", mostly from 2003 through 2005;
- in 2004, beginning of stratigraphic trenches in a newly
discovered cave under the Upper Shelter, with undisturbed ll;
- study of some yet unpublished material, most notably a
protohistoric knife carved in a human radius.
In parallel, colleagues from the Royal Institute of Natural
Sciences of Belgium undertook a reassessment of the abundant
palaeontological material from Dupont’s excavations kept at
their institution:
Figure 1. Location of the caves of Goyet in the Belgian Meuse river basin.
Figure 2. Map of the cliff of Goyet, with location of the main
prehistoric sites. 1, Trou du Moulin; 2, Upper Shelter; 3, Main
terrace and its cave entrances.
- 117 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
- M. Germonpré for animal palaeontology;
- P. Semal for human palaeontology.
The classic terrace and its cave entrances
History
Dupont was the rst to visit the site, mainly the third cave,
where he conducted a huge excavation in 1868. In 1891,
Doctor F. Tihon dug a series of trenches through the terrace.
From 1907 through 1909, the "Service des Fouilles des Musées
royaux du Cinquantenaire" explored the site, particularly
the back ll from the previous excavations and portions of
undisturbed layers in the second and third caves.
In 1937-38, the Royal Institute of Natural Sciences of
Belgium, taking advantage of the construction of tourist
facilities, carried out extensive excavations, particularly in
the "sheep chamber", located behind the deep chamber called
"wastebin" onto which entrances 4a and 4b open, as well as
in two corridors connecting to the left wall of Cave no. 1.
Throughout the rst half of the 20th century amateur
prehistorians followed each other, between and after research
of the two large national scienti c institutions. They moved
back ll from the rst digs, and, in places, ruined areas of
undisturbed layers without any caution for stratigraphy.
Among them: J. Hamal-Nandrin in 1914; J. le Grand-Metz
between 1914 and 1920; J. Colette and M. Beaulieu between
1920 and 1935; H. Angelroth between 1920 and 1944; L.
Eloy, essentially during the 1940s; D. de Burnonville and M.
Drion from 1950 through 1953.
And nally, a programme of modern research conducted
jointly by the "Direction de l’Archéologie" and "l’Association
wallonne d’Anthropologie préhistorique" has been under way
since 1997.
Data from the former excavations
Stratigraphy
During his excavations, E. Dupont (1872: 105-124) identi ed
" ve ossiferous layers" alternating with "six alluvial layers" in
the ll of the cave he described as "3rd cave of Goyet" ( g. 4).
Such an interpretation of the stratigraphy mixes sedimentary
data with archaeological and palaeontological data.
The study of the two lower layers, which have a rich fauna,
especially in the darker areas of the cave, in uenced the
digger into thinking that at the corresponding periods,
the site was the den of, alternatively, lions and cave bears
and sometimes hyenas. The three upper ossiferous layers,
essentially present near the entrance, associated numerous
knapped ints and bone tools with a varied fauna. They
were, partly in error, attributed to the "Mammoth Age".
Flints from the third layer, the deepest of the upper layers,
were related to those of "Montaigle type", and the ints from
the second layer to the Montaigle and "Trou Magrite" type.
The objects from the rst layer comprise narrow and regular
blades, related, though considered better knapped, to those
from Chaleux and Furfooz. Clearly, the series of tools related
to the three ossiferous layers are far from homogeneous and
this testi es to Dupont’s poor stratigraphic observations at
Goyet.
Figure 3. Goyet, the main terrace and cave entrances: numbering of
the caves and location of the trenches dug in 1997-1999.
Figure 4. Stratigraphic section of the third cave (after Dupont 1872,
g. 12 et 13; captions: 12, 1st through 5th ossiferous layers, scale is
2 mm for 1 m; 13: scale is 1 mm for 1 m.
- 118 -
Michel TOUSSAINT
Prehistory
Notwithstanding the shortcomings of the rst interventions,
several prehistorians attempted to describe with some
precision the sequence of Palaeolithic occupations of the
site, which contains one of the most complete succession of
industries of the north-west of Europe (Ulrix-Closset 1975;
Otte 1979; Dewez 1987). But these attempts suffered from
a number of shortcomings. It is quite probable that the main
Palaeolithic periods identi ed on a strictly typological basis
correspond to the arti cial grouping of several occupation
layers. In this respect, recent excavations in regional caves,
notably Walou (Draily 2004), demonstrated the presence
of several layers, each containing rare artefacts whose
culture could not be identi ed on a strictly typological
basis. Furthermore, numerous documents are dif cult to
attribute to a particular culture, e. g. several bone points
or backed bladelets that may be as much Magdalenian as
Perigordian. Finally, the Middle Palaeolithic classi cations
used, derived from F. Bordes’ work in south-west France,
are obsolete.
According to Ulrix-Closset (1975), there was some
Quina Mousterian material at the bottom of the deposits,
characterized by a limited use of Levallois reduction, by a
large proportion of sidescrapers and by the presence of special
artefacts such as bifacial sidescrapers, limaces, thick convex
side-scrapers and thinned back side-scrapers ( g. 5). The
different bifaces and leaf points might also have belonged to
this industry of Quina type, even though it is not impossible
that they belonged to a second Mousterian group. This may
then be indicative of a Mousterian of Acheulean tradition or
an evolved Mousterian (Ulrix-Closset 1975). The leaf points
might also, according to some interpretations (Otte 1984), be
related to industries known as "with leaf points" that might
characterize the Middle to Upper Palaeolithic transition.
The abundant material typologically Aurignacian from the
second and third ossiferous layers identi ed by Dupont might,
according to Otte (1979), be attributed to two industries ( g.
6). Most of the material would in that case be closer to the
Aurignacian of Spy cave while a less important group would
exhibit features closer to the Trou Magrite Aurignacian.
One (supposedly, but more realistically, several) Gravettian
occupation(s) follow(s) the Aurignacian, characterized by
the abundance of backed bladelets, notably long points of
Gravette type and long bitruncated bladelets, and by the
scarcity of Font-Robert points and points with at retouch
( g. 7).
The last Palaeolithic occupations at Goyet, found in the rst
ossiferous layer of Dupont, date from the Magdalenian, with
possibly several phases represented ( g. 8; Dewez 1987). Two
AMS dates were recently obtained from animal bones presenting
cut marks that were gathered by Dupont in the rst ossiferous
layer of the third cave: GrA-3237, 12.770±90 BP;
GrA-3238,
12.620±90 BP (Germonpré 1997). They might add new
evidence to the theory of the "recolonisation" of the karst
valleys of Wallonia by Magdalenian peoples (Charles 1996).
The Magdalenian lithic material of Goyet is comprised,
among others, of piercers, some of "Chaleux" type, and
backed bladelets. Single and double bevelled sagaies, a
superb harpoon with double rows of barbs, eyed needles and a
beautiful reindeer antler "arrow-shaft straightener" (bâton de
commandement) decorated with shlike motifs, among them
a trout (Twiesselmann 1951), constitute the most interesting
bone pieces.
Finally, several more recent artefacts, notably Neolithic,
Roman or Medieval, testify to the sporadic occupation of the
caves of Goyet after the Palaeolithic.
Palaeontology
Fauna from Dupont’s excavations come up regularly in
palaeontological research. Like the other studies of the
material found at Goyet in the 19th century, both in prehistory
and human palaeontology, this research is biased by the lack
of precision of the stratigraphic context and the confusion
between strata from different periods. The most recent work
focus mainly on cave bear, studied from the point of view
of the timing and length of its dormancy period which, at
Goyet as well as in other Belgian caves, was proven to vary
according to the changing climatic conditions of the Ice
Age (Germonpré 2000, 2004; Germonpré & Sablin 2001).
Some cave bear bones bearing cut marks or traces of ochre
were also identi ed; this suggests brief encounters between
cave bears and Palaeolithic Man (Germonpré 2000, 2001).
Fauna from the Magdalenian layer(s) of the third cave was
also re-evaluated from the perspectives of taphonomy and
archaeozoology (Germonpré 1996, 1997).
Palaeoanthropology
Most of Dupont’s anthropological nds come from the
third cave, more precisely from what he called the "second
layer". Some of these nds have previously been studied,
in particular three fragments of differentmandibles (Hamy
1873), of which one (inventory no. I.R.Sc.N.B 2878-09; g.
9:1-2) has sometimes been compared to the mandible from
La Naulette (Hamy 1873; Walkhoff 1903). All three were also
attributed to the Upper Palaeolithic (Twiesselmann 1971). It
was recently suggested that another small, left, mandibular
fragment found by Dupont ( g. 9:3; inventory no. I.R.Sc.
N.B 2878-08), as well as an upper incisor ( g. 9:4), might be
Neandertal (Semal et al. 2005). The mandibular fragment still
has the two premolars and the rst molar; the mental foramen
is under the anterior portion of the M1.
All these fossils deserve a detailed inventory, modern
anthropological studies and radiocarbon dating, in order to
separate the Holocene fossils, for the most part Neolithic,
from some possibly older fossils.
Radiometric dating
Table 1 lists all the radiometric dates obtained from animal
and human bones found at Goyet, with or without relation
with prehistoric occupations.
- 119 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
Figure 5. Goyet, the main terrace and its cave entrances, Middle Palaeolithic artefacts, former excavations. 1, Levallois
point; 2, convex side-scraper, with atypical back; 3, convergent side-scraper; 4, convex side-scraper, with back thinned by
bifacial retouch; 5, elongated Mousterian point; 6, elongated Mousterian point; 7, Mousterian point, on cortical ake; 8,
disc on ake; 9, Limace; 10, leaf point; 11, subtriangular biface (after Ulrix-Closset 1975).
- 120 -
Michel TOUSSAINT
Figure 6. Goyet, main terrace and its cave entrances, Aurignacian artefacts, former excavations. 1, nose scraper; 2, simple keel-
shaped scraper; 3, atypical keel-shaped scraper; 4, nose-scraper; 5, nosed burin; 6, keel-shaped burin; 7, keel-shaped burin; 8-9,
keel-shaped burins, Les Vachons type; 10, nosed burin (after Otte 1979).
- 121 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
Figure 7. Goyet, main terrace and its cave entrances, Gravettian artefacts, former excavations. 1-3, backed points with
troncated base; 4-6, stemmed points, La Font-Robert type; 7-10, bitroncated backed bladelets; 11-12, reinder double
bone points (after Otte 1979).
- 122 -
Michel TOUSSAINT
Figure 8. Goyet, the main terrace and its cave entrances, Magdalenian artefacts, former excavations. 1-2, piercers, Chaleux
type; 3, harpoon with double rows of barbs; 4-5, eyed needles; 6, mobiliary rock art, ibex head; 7, reindeer antler bâton
decorated with shlike motifs (after Dewez 1987).
- 123 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
Figure 9. Goyet, third cave, "second layer". 1-2, mandible found by Dupont (inventory no. I.R.Sc.N.B 2878-09); 3, another fragment of
mandible also found by Dupont (inventory no. I.R.Sc.N.B 2878-08); 4, left lateral upper incisor (1, after Hué 1937; 2, after Hamy 1873; 3,
courtesy of Patrick Semal, I.R.Sc.N.B; 4, after Semal et al. 2005).
Modern excavations
Seven trenches were dug on the terrace ( g. 3:1-6, 11). Four
others in the third cave; namely, trench 7 at the entrance,
trench 8 inside the ssure no. 3b - which is an accessory
passage to Cave 3 - and trenches 9 and 10 in the middle of
the gallery ( g. 3). These eleven trenches were all dug to
bedrock. Finally, trench 12 explored a vertical ssure at the
west of the entrance of Cave no. 1.
Trenches 1-6 showed that the terrace in its current layout,
at and 4 to 10 m in breadth, is mostly arti cial. It had been
essentially modi ed by back ll moved out of the caves by the
former diggers and by the construction of tourism facilities,
which was attested by the discovery, between trenches 1 and 2,
of the foundations of a 20th century drinking establishment.
In fact, the sections of trenches 2 and 3 reveal that the rock
substratum of the "terrace" slopes down steeply. Such a slope
is not well adapted to sophisticated prehistoric occupations.
Furthermore, in situ sediments in contact with the bedrock,
under the back ll, in these two trenches as well as in Trench
6 whose rock substratum slopes less, were extremely poor,
archaeologically and palaeontologically speaking.
Goyet’s so-called "terraces" of the second, third and fourth caves
abundantly cited in the archaeological literature are therefore for
the most part a mythic construction than a real prehistoric site.
- 124 -
Michel TOUSSAINT
Table 1. Radiocarbon dates from the different sites of the Goyet cave system.
These observations suggest that the former excavations conducted
on the terrace essentially consisted in reworking the back ll
from the inner caves, notably that from the initial excavations of
E. Dupont. The fact that the back ll was strati ed added to the
quantity of material it contained, both archaeological but above
all palaeontological, may easily have misled diggers who used to
work in small areas, regardless of stratigraphy, only interested as
they were in laying their hands on nice objects.
On the other hand, a large ssure located in the oor underneath
the arch, just before the entrance no. 1, did not seem particularly
engaging at rst since the bedrock was visible in places; yet it
yielded some promising archaeological data ( g. 3:11).
Trenches 7, 9 and 10 ( g. 3) brought evidence that the
third cave had been almost completely emptied to bedrock
by Dupont and the numerous diggers who succeeded him,
although sediments very low in archaeological content were
found in ssure 3b ( g. 3:8).
Trench 12 explored the rift in the west wall of the rst cave’s
entrance area, just outside the iron grid; it was lled essentially
by sediments reworked by former explorations. Still, it yielded
some int artefacts and morphologically modern human bones
that might come from a disturbed Holocene burial.
The Upper Shelter
History
The shelter is ca. 50 m north-west of the terrace of the
‘classic’ caves, 12 m above them, 25 m above the alluvial
plain ( g. 2:2). This Gravettian site was excavated without
any method around 1952, then recently re-excavated
with a multidisciplinary perspective by the "Direction de
l’Archéologie" (Toussaint et al. 1999).
Results from the 1952 excavations
An abundant lithic material has been gathered over the years,
as much by L. Keyser, at the time manager of the tourist caves,
then by a series of amateur archaeologists to whom Keyser
gave permission to loot the site on a regular basis, among
them L. Eloy, whose collection is the only one published to
date, 4 decades after the fact (Eloy & Otte 1995).
The brief stratigraphy recorded in 1952 comprises only three
units: blocks fallen from the cliff, a humic layer and, at the
bottom, the archaeological layer "made of a light powdery
sediment loessic in aspect", with " ne strata corresponding to
past human occupation".
The archaeological material ( g. 10) was gathered without
precise measurements or sieving. It includes blades and
bitruncated backed bladelets, backed points with truncated bases
as well as bi-points which make up most of the arrow points, the
latter also comprising some backed bladelets and fragments of
Gravette and microgravette points. A stemmed point of Font-
Robert type and a fragment of point with at retouch are also
reminiscent of the famous Maisières site industry. Common
tools include different burins, composite tools and scrapers
of which some, with thick front, are problematic in that they
suggest either the persistence of Aurignacian typologies or the
presence of an Aurignacian layer underneath the Gravettian.
Lab Results Interest Cave Bone bearing Taxon Cut Industry References
number BP layer marks
of Dupont
OxA-4926 24.440 ± 280 Prehistory Upper Shelter Bovid no Gravettian Eloy & Otte 1995
GrA-3237 12.770 ± 90 Prehistory 3 1 Equus, vertebra yes Magdalenian Germonpré 1997
GrA-3238 12.620 ± 90 Prehistory 3 1 Ovibos, phalynx yes Magdalenian Germonpré 1997
UtC 8957 12.560 ± 50 Prehistory 3, chamber A 1 Equus, MC Magdalenian Germonpré 2001
OxA-8875 2420 ± 40 Prehistory classic caves Homo s. sapiens yes Protohistory Toussaint 2005a
Beta-124825 4410 ± 50 Palaeoanthropology Trou du Moulin Homo s. sapiens no Toussaint 2005b
OxA-10534 5345 50 Palaeoanthropology Upper Shelter Homo s. sapiens no Toussaint 2002a
OxA-5678 1985 ± 50 Palaeoanthropology 3, chamber A 3 Homo sapiens Preud'homme 1995-1996
GrA-9606 35.470 + 780-710 Palaeontology 3, chamber B 4 Ursus spelaeus, metacarpal no - Germonpré & Sablin 2001
GrA-9605 38.770 +1180-1030 Palaeontology 3, chamber A 1 Ursus spelaeus, pisiform no - Germonpré & Sablin 2001
KIA-18986 27.440 ± 165 Palaeontology 3, chamber A 3 Ursus spelaeus no - Germonpré 2002
KIA-16289 34.920 +330-320 Palaeontology 3, chamber A 2 Ursus spelaeus no - Germonpré 2002
GrA-2812 27.230 ± 260 Palaeontology 3, chamber A 1 Crocuta crocuta, calcaneum no - Germonpré 1997
UtC 8958 35.000 ± 400 Palaeontology 3, chamber A 1 Crocuta crocuta, P4 no - Germonpré & Sablin 2001
KIA-13550 10.640 ± 50 Palaeontology 3, chamber A 3 Ursus arctos, mandible no - Germonpré 2001
- 125 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
Figure 10. Goyet, Upper shelter, Gravettian artefacts, recent excavation by the “Direction de l’Archéologie” (after Toussaint
et al. 1999). 1-5, bitroncated backed bladelets; 6-11, troncated backed bladelets, broken; 12, pointed backed bladelet, broken;
13-14, scrapers; 15, core; 16-18, burins.
- 126 -
Michel TOUSSAINT
Fauna included woolly rhinoceros, cave bear, aurochs, red
deer and ibex, but mostly horse and reindeer. AMS dating of a
bone fragment from a "large bovine" yielded 24.440±280 BP
(OxA-4926).
Modern excavations
The stratigraphic sequence of the Upper Shelter observed
after the three years of recent excavations appears more
elaborate than that described during the rst explorations.
These deposits, studied in detail by the geologist S. Pirson (in
Toussaint et al. 1999) are divided into two units differing by
their deposition processes. A series of erosion sequences was
also identi ed.
The rst unit (layers I.1 to I.10 and II.12.2; g. 11) represents
deposits lling a decapitated karstic gallery. Layers I.1
to I.8 represent an evolution in karstic context, with high
energy uviatile deposits (layers I.1, I.2 and I.6.1, I.6.2, I.8)
interspersed with calmer phases, maybe also of a uviatile
origin (layers I.3 and I.4). Layer I.5 is different in that its input
contains deposits probably aeolian in origin, interspersed
between two uviatile phases. Layer I.7 corresponds to blocks
collapsed from the roof of the gallery when the streamway
was active. Layers I.9 and I.10 have probably been deposited
in the karst, before the collapse of the gallery roof; however,
no layer can be directly correlated with this collapse, probably
because of the perturbations produced by the former amateur
excavations and because of strong erosion. A signi cant
hiatus certainly separates this collapse (probably posterior to
layer I.10) and the deposit of layer II.12.1. On the other hand,
Figure 11. Goyet, Upper Shelter, section drawn by the geologist S. Pirson during the excavation of the “Direction de l’Archéologie” (after
Toussaint et al. 1999).
the top layer of the rst unit, II.12.2, is clearly related to a
rock shelter context, as is the second unit.
The stratigraphy of the second unit (layers II.11 to II.16) -
whose stratigraphic links with the rst unit were in great part
truncated by the 1950’s work - starts with a loess-like sediment
containing Gravettian archaeological material (layer II.12.1).
Layer II.13 represents a phase of strong cryoclasty. Finally,
layers II.14 and II.15 are affected by Holocene pedogenesis.
The new research yielded a few bones from large mammals, a
marine mollusc shell (Glycimeris sp.) and abundant samples
of microfauna and terrestrial molluscs. The bone remains are
deteriorated and very small. They essentially belong to: Bos
primigenius, Equus sp., Cervus elaphus, Rangifer tarandus,
woolly rhinoceros and Vulpes vulpes. This little series suggests
a predatory way of life in the context of a gallery forest and
thus, climatic conditions slightly milder than that deduced
from the identi cations based on photographs of about 20
bones from the 1952 excavations (Eloy & Otte 1995).
Several hundred lithic artefacts including dozens of tools
were found in the still undisturbed archaeological layer and in
the back ll from the rst excavations. The intwork consists
of cores with blades, numerous blades and bladelets as well
as akes of various shapes. The classic tools are represented
by scrapers ( g. 10) and burins, essentially on truncations and
dihedral There are also backed pieces with two truncations,
backed pieces with a truncation at one end and a fracture at
the other end, that are probably nothing more than broken
bitruncated pieces, as well as bladelets with pointed back
- 127 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
that might be fragments from bi-points. Some simple backed
bladelets were also found. Typologically speaking, this
material belongs to the Gravettian, like the one from the
previous excavations to which it is closely related.
A small set of teeth and human bones was found during
the latest excavations, notably in a horizontal ssure in the
back wall of the Upper Shelter, more or less at the level
of the super cial deposits that must have lled the site
before its exploitation. These remains belong to at least one
morphologically modern child and one adult. This places
them within the Middle Neolithic: 5345±50 (OxA-10534),
4250-4040 BC after calibration at 1σ.
Trou du Moulin and its new cave systems
History
Trou du Moulin (Mill Cave), sometimes referred to as "Cave
no. 1" or "Mathot Cave", is located downstream of the Upper
Shelter. E. Dupont was the rst to explore the site, apparently
after his important excavations at the "classic" caves of
Goyet but before 1872, when the site is mentioned in his
main book (Dupont 1872). The site was later transformed by
the construction of a shelter during the First World War. In
1948, H. Danthine (1952) had large trenches dug through the
terrace and in the cave itself.
Since 1998, our research team has found several new galleries
and chambers by exploring a small cleft in the right wall of
the main cave (Toussaint et al. 1998).
Data from the former excavations
From the limited information available in E. Dupont’s book
(1872, synoptic table) and the accompanying labels of the
material he found (Danthine 1952) that is conserved at the
Royal Institute of Natural Sciences of Belgium, there were
three ossiferous layers at Trou du Moulin, all belonging
to the "Mammoth age". The rst two layers yielded some
archaeological material, among them a small series of int
artefacts. The purported homogeneity of the groups de ned
by Dupont must, however, be treated with some caution, as
the presence of pottery shards and some human bones found
in the second ossiferous layer suggests. While excavating
Trou du Moulin, H. Danthine found only disturbed sediments
containing various faunal remains belonging to the same
species as those found by the rst digger, as well as a fragment
of a retouched int and a chert point. This poor lithic material
might indicate a short occupation by a "Levallois-Mousterian"
Middle Palaeolithic group (Danthine 1952).
Cave bears, hyenas and some rhinoceros were the most
represented species in these three layers. The so-called
"evidence of human occupation" as identi ed by Dupont in his
" rst ossiferous layer" is limited (Danthine 1952) to ten int
artefacts including 3 retouched blades and one core, a chert
blade and two rolled cobbles of which one would have served
as a hammerstone. The "second ossiferous layer" yielded for
its part two rolled cobbles and 13 int artefacts including
3 scrapers and a long point. Some human bones were also
found in the second of these ossiferous layers. Pottery shards
coming from the two layers make the homogeneity of the
groups as advanced by Dupont disputable.
The child from the new cave systems
The newly discovered (1998) cave systems of Trou du
Moulin extend to the Upper Shelter. A series of chambers
and connecting corridors, forming an approximate square,
constitute the "Central System" ( g. 12). Three long sub-
systems with magni cent concretions start from three of
the four corners of the Central System: "Régal des Fées",
"Atlantide" and "Salle de Cristal".
Bones from a child about 12 years old, probably a boy, were
found in two areas of one of the "Central System" chambers
(Toussaint et al. 2004; Toussaint 2005b). Most of the bones
came from a vertical ssure ca. 2 m high and 30-40 cm wide
near the roof of "Salle de l’Enfant" ( g. 12). Others were
found 4 m below, in the chamber proper, either on fragments
of stalactite curtains and stalagmites covering the oor or
underneath these fragments. No archaeological material
was associated with it. AMS dating of a foot bone yielded a
date of 4410±50 BP (Beta-124825), i. e. 3100-2920 BC after
calibration at 1σ. This ts well in the rich corpus of dated
human bones from the Belgian Meuse basin (Toussaint 2002a).
It corresponds to the beginning of the Late Neolithic.
Since the ssure was too tight to allow access to the skeleton, an
ad hoc methodology had to be developed, comprising precise
measurements with laser surveying equipment of small plastic
landmarks scattered amid the bones, photographs taken with
a digital camera tted to a pole, software correction of the
photographs to recreate a faithful map of the bone distribution
based on the surveyed landmarks, and nally dismantling of
the burial with a 70 cm long articulated pair of pliers.
Initially, the scattered bones seemed in utter disorder. But
after having divided the ssure in sections numbered 1 to
6 (7 being the chamber immediately below the ssure),
from back to front, and having studied the distribution of
the different types of bones according to this plan, some
minimal order became apparent. As shown on gures 13
and 14, most of the cranial fragments were at the back, in
section 1. The maxillae and the mandible were closer to
the front, respectively in sections 3 and 4. Isolated teeth
were in the bottom half of the ssure. Most of the pectoral
girdle bones were in section 3, the rest in sections 2 and 4.
Vertebrae were found everywhere, a few cervical vertebrae
at the back, in section 1, and a concentration in section 3.
The ribs were mostly concentrated in sections 3 and 4. The
upper long limb bones were somewhat grouped in section 3
and in the back of 4 ( g. 14). Most of the hand bones were
in 3. The pelvis parts were in majority in 5 and the front
of 4. The femora, broken, were in section 5. The left tibia
was in section 4 and the right much further back, trapped
by stalagmite B, in section 3. The right bula was also in
section 3. Several bone fragments had fallen in the chamber
down below, section 7.
- 128 -
Michel TOUSSAINT
Figure 12. Goyet, Trou du Moulin, Neolithic grave of the new cave systems. 1, map of the “Central System”, a series of
chambers and connecting corridors forming an approximate square, with location of the Neolithic child burial; 2, general view
of the grave; 3-4, detail views of the grave.
- 129 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
Figure 13. Goyet, Trou du Moulin, Neolithic grave of the new cave
systems: distribution of the different types of bones in six sections
de ned along the longitudinal axis of the grave, from back to front,
section 7 being the chamber immediately below the ssure.
This distribution pattern suggest the body had been lain with
its head at the back of the ssure and its feet near the entrance,
which is supported by the fact that most of the skeletal remains
had their proximal epiphysis towards the back of the ssure,
where the skull was located.
However, two interesting irregularities appeared during
our analysis. First, the femora were turned over, their head
towards the entrance, while the orientation of the tibiae was
more in accordance with what was expected from a body
with its head at the back. Second, the ulnae and radii were
further back in the ssure than the left humerus, like the
majority of the small hand bones; furthermore, the proximal
end of the right ulna was oriented towards the entrance. The
comparison between these two sets of observations provides
solid evidence in favour of a folded position of the body, of
foetal type.
The combination of three intriguing characteristics of the
burial, namely: the body was alone, completely sheltered
from daylight and in an all but unreachable ssure, confers
on this burial a quite speci c quality in comparison with the
other burial sites from the Late Neolithic Meuse basin.
It seems also that the body of the child decomposed in the
open air, as evidenced by the lack of sediments over the bones
as well as the numerous bones displaced out of the initial area
of the burial, the displacement further back of the skull parts,
the displacement of the mandible, the attening of the pelvic
area, fragments from the same bone found separated by a
certain distance, etc. Decomposition occurred in a primary
burial, which is con rmed by the persistence of "unstable
connections" and a certain level of spatial organisation in the
distribution of the bones.
As a conclusion, the following sequence of events can be
proposed. Neolithics decided, after the death of a child about
12 years old, to bury him in a ssure situated 4 m above the
oor, inside the Goyet karstic system. To achieve this, they
had to hold and sometimes drag the corpse through tight
corridors, in order to access the chamber where they hauled
the body in the ssure and laid it to rest with its head at
the back. It seems that the legs and forearms were folded.
No sediment was deposited on the corpse and no natural
sedimentation covered it later. Two stalagmites, formed by
water dripping from the roof of the ssure, froze two small
sections of the burial. Later, small animals disturbed the
distribution of the bones, eroded the ends of some of them
and pushed some bones into the chamber below the ssure.
Seismic tremors later shattered stalactite curtains adorning
the walls of the "Salle de l’Enfant" as well as stalagmites
and stalactites and toppled them on the oor, covering in the
process the rst bone fragments fallen from the burial ssure.
Finally, burrowing animals again pushed some bones out of
the ssure, the resulting fragments then falling over the pieces
of stalactite curtains and other concretions covering the oor
of the chamber.
Post-Palaeolithic archaeological evidence
Apart from several series of teeth and human skeletal remains
reminiscent of Neolithic Meuse basin collective burials, some
post-Palaeolithic prehistoric material found at Goyet was
recently studied. The caves have also yielded traces from
later periods: Protohistory, Roman and Middle Ages, whose
study falls outside the scope of this paper.
- 130 -
Michel TOUSSAINT
Figure 14. Goyet, Trou du Moulin, Neolithic grave of the new cave systems. 1, distribution of the skull fragments; 2, distribution of the upper
limb bones.
- 131 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
The bevelled stag antler mattock
A bevelled and double-perforated stag antler mattock was
found at an unknown date ( g. 15). The two perforations
are parallel; the rst, incomplete, in the axis of the central
antler, and the second, complete, about 1 cm from the
rst one, towards the bevel. The bevel was obtained
by obliquely sectioning the beam; it exhibits several
microtraces. It matches type ba3 in Hurt’s classi cation
(1982). Chronologically, bevelled tools of type Ba seem
to make their appearance around 6500-6400 BP (Smith
1989) or 6100 BP (Crombé et al. 1999) until, depending
on the regions, 5400-5300 BP (Smith 1989) or even
4700 BP (Crombé et al. 1999). They date from the end
of the Mesolithic or the Early and Middle Neolithic.
Without radiocarbon dating, the age of this tool remains
undetermined.
The Protohistoric knife ( g. 16)
A tool manufactured on a human radius was discovered between
1935 and 1945 in the classic cave system (Toussaint 2002b,
2005a). A splinter from the artefact yielded an AMS date of
2420±40 BP (OxA-8875), i.e. between 760 and 400 BC
after
calibration at 1σ and between 770 and 390 BC at 2σ. Such a
result dates the object to the Iron Age, although the size of
the standard deviation due to large plateaux on the calibration
curve prevents greater precision.
The tool was shaped on an adult left radius, perhaps from a male.
The distal end of the bone was removed during tool preparation.
On the distal part, there is a partial longitudinal edge, prepared
by scraping of the palmar and dorsal surfaces and practically
aligned with the prolongation of the interosseous crest, from
which, however, it is easily distinguished. The lateral face of
Figure 15. Bevelled antler tool.
- 132 -
Michel TOUSSAINT
Figure 16. Protohistoric knife from the “classic” caves.
- 133 -
1997-2005 Research in the caves of Goyet (Gesves, province of Namur, Belgium)
the bone is a longitudinal back opposite the prepared edge.
Unworked on all of the central part of the object, this back
was, however, worked on the most distal part to contribute to
the formation of the point.
Strictly speaking, the object cannot then be considered a
dagger. In spite of the relative shortness of the worked edge, it
corresponds rather to the de nition of a knife. Typologically,
the tool is exceptional. On the basis of archaeological literature,
it is the only sharp-pointed tool created on a radius known both
for the prehistoric and protohistoric periods in Europe and
North Africa; the few other tools made from human long bones
are typically on bulae and, more rarely, on ulnae and humeri.
In addition, these comparable tools are characterized primarily
by their point, which often quali ed them as daggers, while
the speci city of the Goyet knife is the association of the point
with a worked edge opposed to a natural back.
Conclusion
The archaeological richness of Goyet caves was already
known back in 1870. The intensity of the careless
explorations this site has undergone has profoundly altered
the sedimentary deposits and the rich prehistoric material it
contains. Data essential to the accurate understanding of the
stratigraphy and palaeoenvironment have been lost forever.
Both palaeoenvironmental and archaeological studies are
consequently seriously penalized, inasmuch as they have to
compile documents most probably originating from different
strata. Barring the discovery of an undisturbed sedimentary
ll, we will never know in detail the different Middle
Palaeolithic, Aurignacian, Gravettian and Magdalenian
industries present at Goyet.
However, during the research undertaken since 1997, partial
layers have been found in situ, for example at the Upper
Shelter, and the virgin area, in the new systems of Trou du
Moulin, where a Neolithic child burial, has been excavated.
Much more, though, must still be done if we are to better
understand the prehistory of the caves of Goyet.
In the eld, several untouched areas should be excavated.
Among them, a ssure in the oor of entrance no. 1 of the
"classic" terrace stills contains undisturbed deposits. By far
the most promising perspectives, however, lie in the pristine
gallery directly under the Upper Shelter; lled with sediments
to the last 60 cm under the roof, its entrance yielded, as much
on the surface as in stratigraphy, lithic material from the
Early Upper Palaeolithic and bones from large Quaternary
mammals. Finally, diverse small caves spread out along the
cliffs of Goyet still contain multiple unexplored burials.
As for laboratory work, analyses of the deposits from the Upper
Shelter must be nalized, particularly the sedimentology
and palynology, and a monograph has yet to be written. The
study of both the collections kept in museums and the private
collections might also bring some surprises, as proven by the
recent re-discoveries of a knife on human bone unearthed half
a century ago as well as a fragment of a mandible found by E.
Dupont around 1870.
Acknowledgments
The author would like to express his deepest gratitude to the persons who
helped throughout the recent studies at the caves of Goyet, particularly:
- the Tasiaux family, proprietor of the Goyet estate forest;
- the authorities of the commune of Gesves, proprietor of the caves of
Goyet;
- Philippe Lacroix, Sylviane Lambermont, Jean-François Beaujean, Jean-
François Lemaire and Louis Bruzzese, technicians at the "Association
wallonne d’Études mégalithiques", for their participation in the excavations
and the illustration of this article;
- Stéphane Pirson, geologist;
- Rebecca Miller, Archaeologist at the Department of Archaeology of the
University of Liège;
- Jean-François Lemaire, for the translation of this text from French.
References
Becker A., Lacroix Ph., Léotard J.-M., Lopez-Bayon I., Noiret O., Quinif Y. & Toussaint M. (1999) - Les grottes de Goyet. Namur, Ministère de la Région
wallonne, DGATLP, Carnet du Patrimoine 26, 52 p.
Charles R. (1996) - Back into the North: the Radiocarbon Evidence for the Human Recolonisation of the North-Western Ardennes after the Last Glacial
Maximum. Proceedings of the Prehistoric Society 62:1-17.
Crombé Ph., Van Strydonck M. & Hendrix V. (1999) - AMS-dating of antler mattocks from the Schelde river in northern Belgium. Notae Praehistoricae 19:111-119.
Danthine H. (1952) - A propos de la grotte n°1 de Goyet. In: Études d’Histoire et d’Archéologie Namuroises dédiées à Ferdinand Courtoy. Gembloux, Duculot,
p. 31-42.
Dewez M. (1987) - Le Paléolithique Supérieur Récent dans les Grottes de Belgique. Louvain-la-Neuve, Institut supérieur d’Archéologie et d’Histoire de l’Art,
Université catholique de Louvain, 466 p.
Draily C. (2004) - Bilan des occupations moustériennes de la grotte Walou à Trooz (province de Liège, Belgique) et essai d’interprétation des couches à faible
densité de matériel lithique. Notae Praehistoricae 24:17-29.
Dupont E. (1872) - L’homme pendant les Âges de la Pierre dans les environs de Dinant-sur-Meuse. Bruxelles, Muquardt, 250 p.
Eloy L. & Otte M. (1995) - Le Périgordien de l’abri-sous-roche de Goyet (Namur, Belgique). Bulletin de la Société royale belge d’Études géologiques et
archéologiques des Chercheurs de la Wallonie 35:25-40.
Germonpré M. (1996) - Preliminary results on the mammals of the Magdalenian upper horizon of Goyet (Belgium). Notae Praehistoricae 16:75-85.
Germonpré M. (1997) - The Magdalenian upper horizon of Goyet and the late Upper Palaeolithic recolonisation of the Belgian Ardennes. Bulletin de l’Institut
Royal des Sciences naturelles de Belgique 67:167-182.
Germonpré M. (2000) - The cave bear and prehistoric man: a brief encounter at Goyet, Belgium. Notae Praehistoricae 20:35-40.
- 134 -
Michel TOUSSAINT
Germonpré M. (2001) - A Reconstruction of the Spatial Distribution of the Faunal Remains from Goyet, Belgium. Notae Praehistoricae 21:57-65.
Germonpré M. (2004) - In uence of Climate on Sexual Segregation and Cub Mortality in Pleniglacial Cave Bear. In: C.G.M. Roel & I. Plug (eds.), The Future
from the Past. Proceedings of the 9
th
Conference of the International Council of Archaeozoology, Durham (August 2002). Oxford, Oxbow Books, p. 51-63.
Germonpré M. & Sablin M.V. (2001) - The cave bear (Ursus spelaeus) from Goyet, Belgium. The vear den in Chamber B (bone horizon 4). Bulletin de l’Institut
Royal des Sciences naturelles de Belgique, Sciences de la terre 71:209-233.
Hamy E.T. (1873) - Sur quelques ossements humains découverts dans la troisième caverne de Goyet, près Namèche (Belgique). Bulletins de la Société
d’Anthropologie de Paris 8:425-435.
Hué E. (1937) - Crânes paléolithiques. Congrès préhistorique de France, compte rendu de la douzième session, Toulouse-Foix, 1936. Paris, SPF, p. 113-294.
Hurt V. (1982) - Les haches en bois de cerf en Belgique: essai de classi cation. Amphora 29:14-24.
Otte M. (1979) - Le paléolithique supérieur ancien en Belgique. Bruxelles, Musées Royaux d’Art et d’Histoire, Monographies d’Archéologie Nationale 5, 684 p.
Otte M. (1984) - Le Paléolithique supérieur en Belgique. In: D. Cahen & P. Haesaets (éds.), Peuples chasseurs de la Belgique préhistorique dans leur cadre
naturel. Bruxelles, Patrimoine de l’Institut royal des Sciences naturelles de Belgique, p. 157-179.
Preud’homme D. (1995-1996) - Ossements humains préhistoriques de Reuviau, Chaleux, Magrite, Goyet et de l’abri des Autours, province de Namur, Belgique.
Mémoire de licence en Histoire de l’Art et Archéologie, Université libre de Bruxelles, 156 p.
Semal P., Toussaint M., Maureille B., Rougier H., Crevecoeur I., Balzeau A., Bouchneb L., Louryan S., De Clerck N. & Rausin L. (2005) - Numérisation des
restes humains néandertaliens belges. Préservation patrimoniale et exploitation scienti que. Notae praehistoricae 25:25-38.
Smith C. (1989) - British Antler Mattocks. In: C. Bonsal (éd.), The Mesolithic in Europe. Papers presented at the Third international Symposium, Edimburgh,
John Donald, p. 272-283.
Toussaint M. (2001) - Les hommes fossiles en Wallonie. De Philippe-Charles Schmerling à Julien Fraipont, l’émergence de la paléoanthropologie. Namur,
MRW, Carnet du Patrimoine 33, 60 p.
Toussaint M. (2002a) - Problématique chronologique des sépultures du Mésolithique mosan en milieu karstique. Notae Praehistoricae 22:141-166.
Toussaint M. (2002b) - Gesves/Mozet: un couteau protohistorique réalisé au départ d’un radius humain protohistorique en provenance des grottes de Goyet.
Chronique de l’Archéologie wallonne 10:237-240.
Toussaint M. (2005a) - Un couteau aménagé dans un radius humain protohistorique découvert aux grottes de Goyet (Gesves, province de Namur, Belgique).
Bulletin de la Société préhistorique française 102(3):625-637.
Toussaint M. (2005b) - La sépulture d’enfant néolithique des nouveaux réseaux du Trou du Moulin, à Goyet (Gesves, province de Namur). Anthropologica et
Praehistorica 116:179-210.
Toussaint M., Becker A. & Lacroix Ph. (1998) - Recherches 1997-1998 aux grottes de Goyet, à Gesves, province de Namur. Notae Praehistoricae 18:33-44.
Toussaint M., Lacroix Ph., Lambermont S., Lemaire J.-F. & Beaujean J.-F. (2004) - La sépulture d’enfant néolithique des nouveaux réseaux du Trou du Moulin,
à Goyet (Gesves, prov. de Namur). Rapport préliminaire. Notae Praehistoricae 24:159-166.
Toussaint M. & Pirson S. (2006) - Neandertals Studies in Belgium: 2001-2005. Periodicum Biologorum 108(3):373-387.
Toussaint M. & Pirson S. (sous presse) - Aperçu historique des recherches concernant l’Homme préhistorique dans le karst belge aux XIX
e
et XX
e
siècles:
archéologie, géologie, paléoanthropologie, paléontologie, datations. In: Société préhistorique française, Actes du colloque du centenaire, Avignon (septembre
2004).
Toussaint M., Pirson S. & Bocherens H. (2001) - Neandertals from Belgium. Anthropologica et Praehistorica 112:21-38.
Toussaint M., Pirson S., Lopez Bayon I., Becker A., Lacroix Ph. & Lambermont S. (1999) - Bilan préliminaire de trois années de fouille à l’Abri Supérieur de
Goyet (Gesves, province de Namur). Notae Praehistoricae 19:39-47.
Twiesselmann F. (1951) - Les représentations de l’homme et des animaux quaternaires découvertes en Belgique. Bruxelles, Mémoire de l’Institut royal des
Sciences naturelles de Belgique 113, 28 p.
Twiesselmann F. (1952) - Notice sur l’œuvre archéologique d’Edouard-François Dupont. In : Études d’Histoire et d’Archéologie Namuroises dédiées à Ferdinand
Courtoy. Gembloux, ,Éditions J. Duculot, p. 18-30.
Ulrix-Closset M. (1975) - Le Paléolithique moyen dans le bassin mosan en Belgique. Wetteren, Universa, 221 p.
Walkhoff O. (1903) - Die diluvialen menschlichen Kiefer Belgiens und ihre pithekoiden Eigenschaften. In: E. Selenka (éd.), Menschenaften (Anthropomorphae).
Studien über Entwicklung und Schädelbau 6:373-415.
... in the largest of the caves. Dupont found the studied mandible in the second of five fauna-rich levels (Dupont, 1872;Toussaint, 2006). Originally, the fossil was thought to be modern human due to its stratigraphic proximity to Aurignacian artefacts, but this has been re-evaluated and it now is accepted to be a Neanderthal (Rougier et al., 2012;Wißing et al., 2015). ...
... Magdalenian level dated to 13 ka (Toussaint, 2006). Mixing is present in all levels and its date was long ambiguous but this has recently been re-evaluated as dating to 44-45.5 ka cal BP (Wißing et al., 2015). ...
Article
Full-text available
The ecology of Neanderthals is a pressing question in the study of hominin evolution. Diet appears to have played a prominent role in their adaptation to Eurasia. Based on isotope and zooarchaeological studies, Neanderthal diet has been reconstructed as heavily meat-based and generally similar across different environments. This image persists, despite recent studies suggesting more plant use and more variation. However, we have only a fragmentary picture of their dietary ecology, and how it may have varied among habitats, because we lack broad and environmentally representative information about their use of plants and other foods. To address the problem, we examined the plant microremains in Neanderthal dental calculus from five archaeological sites representing a variety of environments from the northern Balkans, and the western, central and eastern Mediterranean. The recovered microremains revealed the consumption of a variety of non-animal foods, including starchy plants. Using a modelling approach, we explored the relationships among microremains and environment, while controlling for chronology. In the process we compared the effectiveness of various diversity metrics and their shortcomings for studying microbotanical remains, which are often morphologically redundant for identification. We developed Minimum Botanical Units as a new way of estimating of how many plant types or parts are present in a microbotanical sample. In contrast to some previous work, we find no evidence that plant use is confined to the southern-most areas of Neanderthal distribution. Although the interpreting ecogeographic variation is limited by the incomplete preservation of dietary microremains, it is clear that plant exploitation was a widespread and deeply rooted Neanderthal subsistence strategy even if they were predominately game hunters. Given the limited dietary variation across Neanderthal range in time and space in both plant and animal food exploitation, we argue that vegetal consumption was a feature of a generally static dietary niche
Article
Full-text available
La situation belge a beaucoup favorisé l'occupation au Paléolithique supérieur : matériaux, abris, carrefours. L'histoire récente du jeune royaume a sollicité très tôt la recherche sur cette période. Les pointes foliacées sont d'affinité septentrionale. L'Aurignacien y est intrusif, rapide. Le Gravettien y est très ancien et se diffuse ensuite au Sud-Ouest. Une interruption se manifeste au Pléniglaciaire. Le Magdalénien est analogue à celui du Bassin parisien, mais le Creswellien semble originaire d'Angleterre. Le Tardiglaciaire comprend deux traditions distinctes, Tjongérien et Ahrensbourgien, aux origines du Mésolithique local. The Belgian situation favoured the occupation during the Upper Palaeolithic: raw material, shelter and link between other occupied zones. Because the kingdom of Belgium was young at the time that prehistoric science was being developed, much interest was given to the prehistoric roots of the country. Foliate points have a Northern affinity. The Aurignacian is intrusive and rapid. The Gravettian is quite early and was subsequently diffused to South-West Europe. An interruption occurs during the Pleniglacial. The Magdalenian is similar to that found in the Paris Basin, but the Creswellian seems to have originated in England. The Tardiglacial includes two distinct traditions - Tjongerian and Ahrensburgian - as origins for the local Mesolithic.
Article
Full-text available
Three bones of muskox, horse and cave hyaena of the first, Magdalenian horizon from the cave of Goyet (Belgium) were dated by Accelerator Mass Spectrometry (AMS); the ones of the first two species show traces of human manipulation. The bones belong to the Dupont collections, curated at the royal Belgian Institute for Natural Sciences. The AMS results on the bones of horse and muskox suggest that the Magdalenian occupations at Goyet (horizon 1) date from the very end of the Pleniglacial or the beginning of the Late Glacial (Oldest Dryas). The cave hyaena bone is much older. The dates and faunal assemblage are compared with those from other Belgian Magdalenian sites. The main Magdalenian colonisation of the region can be dated between ca 15200 cal yr BP and ca 14800 cal yr BP.
Article
Full-text available
The Dupont collections curated at the Royal Belgian Institute of Natural Sciences contain a large bone assemblage from Chamber B, horizon 4 of the third cave of Goyet (Belgium). This assemblage dates from the Weichselian (AMS date: 35.470 + 780-710 years BP). Ninety eight percent of the remains are from Ursus spelaeus. The presence of several skeletons and articulated body parts, and the low frequency of gnawing traces and charriage-à-sec indicate that the carcasses were only slightly disturbed. Based on the eruption sequence of the teeth, cubs perished at or soon after birth, at an age of two and a half months and at an age of six-seven months old, when they were leaving the cave at the end of the winter rest. In Belgium, the timing and length of the dormancy period and the birth season varied according to the climatic conditions of the ice age. At Goyet, Chamber B, horizon 4, female cave bears probably began their winter rest in October, gave birth in November or December and left the cave in June.
Article
Full-text available
Bilan des occupations moustériennes de la grotte Walou à Trooz (province de Liège, Belgique) et essai d'interprétation des couches à faible densité de matériel lithique Christelle DRAILY Résumé Outre la couche CI-8 riche en vestiges archéologiques, la grotte Walou a livré de nombreux niveaux du Paléolithique moyen dans lesquels les artefacts sont très peu nombreux. L'interprétation du comportement de l'homme préhistorique est indissociable de l'étude taphonomique de chacun de ces niveaux et de l'étude de la mise en place des dépôts. Dans la majorité des cas, les couches pauvres en matériel lithique sont incomplètes car fortement érodées ou remaniées. Malgré tout, la densité des pièces est si faible par rapport au niveau moustérien le plus dense que l'on peut avancer l'hypothèse d'occupations de courte durée n'ayant pas nécessité d'activité de débitage importante sur le site. Mots-clés : grotte Walou, Paléolithique moyen, contexte karstique, niveaux pauvres, taphonomie.
Article
On several occasions since 1830 Southern Belgium Meuse Basin caves yielded Neandertal remains, some of them having a major influence on the development of palaeoanthropology as a distinct scientific discipline; in particular, the discovery of human fossils in 1829-1830 at Engis, of a mandible in 1866 at La Naulette and of skeletons in 1886 at Spy. Yet to this day the context of these old finds is not well known. But new finds, from the last two decades, at Couvin, Scladina and Walou took place within modern multidisciplinary field and laboratory studies, backed up by stratigraphic positioning, different dating methods and palaeoenvironmental recordings. In parallel, most of the old and recent Neandertal fossils were the object of new anthropological laboratory studies using state-of-the-art technologies, notably 3D reconstruction from computer tomography scans. This article overviews all these contributions, focusing primarily on the work of the last five years.