- The Trou des Blaireaux
- The Grotte de Sy Verlaine
- The Trou des Nutons & the Trou du Frontal
- The Trou de Chaleux
- The Grotte du Coléoptère
- The Grotte du Bois Laiterie
- The Trou da Somme
- The Grotte Walou
INTRODUCTION
During the last decade the chronology of the human settlement of Lateglacial Europe
has been the subject of intensive research (Hedges 1986; Housley 1991).
This paper sets out to discuss one part of this data set -- the dating of the Late
Magdalenian of the north-western Ardennes region of Belgium. The work presented
here forms a part of doctoral research undertaken by the author between 1990 and
1994 (Charles 1994). Further publication relating to the dating of other
archaeological groups (the Creswellian, Tjongerian and Ahrensburgian) present
within this study area during the Lateglacial is planned. Climatic terms such as Bölling will be used throughout the following text following a modified version of Mangerud et al. (1974) which is widely used by Belgian and French archaeologists and palaeontologists. Correlations between this and other schemes for the European subdivisions of the Lateglacial can be found in the preface to Barton et al. (1991) which is presented here as Figure 1.
This paper focuses on the problems of dating the earliest of the Lateglacial archaeological
assemblages found in the Belgian Ardennes -- the Magdalenian. The archaelogical residues
recovered from a series of Late Magdalenian sites in this region (Figure 2)
appear to give us the earliest evidence for the human re-colonisation of this
part of Europe at around the time of the first of a series of rapid climatic
oscilations which mark the end of the last Ice Age in western Europe,
locally known as the Bölling Interstadial phase.
BACKGROUND
The relatively recent development of Accelerator Mass Spectrometry (AMS) has revolutionised our understanding of Upper Palaeolithic archaeology by making it possible to use far more appropriate samples for dating than ever before. By 1990 this technique had only been used to anywhere near to its full potential for the Lateglacial and early Postglacial periods in Britain, via dating humanly modified bone and antler. The available dates for the Belgian Lateglacial were, by contrast, almost all measured from conventional samples (large amounts of material taken from bone, charcoal or wood, often ‘bulking’ a number of specimens to obtain the requisite quantity). In many cases, precise details of samples submitted and their relationship to the archaeology which they purport to date were unclear. This in turn had led to major problem in interpreting the Belgian Late Upper Palaeolithic sequence, which at that time appeared to be at odds with the known archaeological sequences from adjacent areas in Europe.
The most extreme example of this is the radiocarbon dates said to be associated with Tjongerian material. These dates are spread over much of the last 30,000 years. Anomalies were also found within the radiocarbon record for the Magdalenian. If taken at face value this suggests that humans had returned to this region of north-western Europe by roughly 16,000 BP and that Magdalenian settlement might have continued in this region until the 8th millenium BP. The vast majority of these radiocarbon dates come from ‘bulked’ samples of bone fragments or charcoal, where precise contextual relationships between the samples themselves and the archaeological residues recovered from the sites in question was in many cases ambiguous.
The bulking of material from whole layers has major implications for
sample integrity and subsequent interpretation of any dates obtained.
Organic material may accumulate within a geological or archaeological
layer over a considerable period of time due to a range of taphonomic
processes. Consequently, dating samples of ‘bulked’ material can only
give an average of all the individual ages of the elements of which
the sample was actually composed. At best, such dates may give an
indication of the antiquity of at least part of the material within
a layer; at worst, they can be highly misleading, bearing little
relationship to the ‘real’ age of apparently associated archaeological
residues. In contrast to all this, the tiny samples required for AMS
dating can be taken from individual artefacts or modified bones, which
have an unequivocal association with a single episode of human activity
or presence.
A further possibility for dating human presence during the Lateglacial
lies with Thermoluminescence (TL). This is a dating technique which
has only recently been applied to Palaeolithic datasets
(Huxtable & Aitken 1986). At the time of writing only a few dates
associated with traces of Lateglacial human presence are available.
Burnt flint is the raw material usually used from Palaeolithic contexts,
although the technique has also been widely applied to ceramics and
sediments. TL dates are given in calendar years, and as such cannot be
directly compared with uncalibrated radiocarbon dates (which are in
radiocarbon years). TL dates are currently known from four Lateglacial
sites in Belgium: Orp-le-Grand East and West (Magdalenian), Rekem and
Meer IV (Tjongerian).
TL dating of the Belgian Magdalenian is still in its infancy. At only two sites, Orp-le-Grand East and West, (open air lithic scatters lacking associated fauna) has TL dating been successful. The dates (Figure 3) indicate use of these sites during the Lateglacial, with Orp West apparently slightly later than Orp East. The standard deviations attached to the dates are too great to give any precise indication of temporal patterning, although one can be fairly confident that the real ages lie between 14,000 and 10,000 BP.
THE AMS DATING OF THE LATE MAGDALENIAN SETTLEMENT OF BELGIUM
Between 1991 and 1993 two dating projects relating to the human settlement of the north-western Ardennes during the Late Upper Palaeolithic were organised in collaboration with the Accelerator Unit at the Research Laboratory for Archaeology and the History of Art at Oxford University. The destructive element of the AMS technique is so limited as to preclude the usual objections to sampling worked bone or specimens of archaeological or palaeontological interest (cf. Hedges 1986). Given the problems of association between archaeological residues and faunal remains, only specimens of human or humanly modified bones were selected as samples, as the association between these and human activity is unquestionable.Initially the scope of these dating project was limited to those sites whose faunas were already under study as part of a review of the Belgian Lateglacial archaeological record (Charles 1994; Trou de Chaleux, Trou des Nutons at Furfooz, Trou du Frontal at Furfooz, Grotte du Coléoptère and the Grotte de Sy Verlaine) and to bones with a clear archaeological association. As the project developed it became appropriate to expand the scope of the dating project to include sites at which the author was not among the primary investigators (Trou da [sic] Somme, Grotte de la Bois Laiterie and Trou des Blaireaux).
The initial dating project resulted in 6 dates on human or humanly modified bones (Hedges et al. 1993). This was subsequently extended, and a further 11 dates were obtained (Hedges et al. 1994). It is this recent dating work which forms the basis for the following discussion.
The Magdalenian in Belgium
Originally defined on the basis of the stone, bone and ivory formal artefacts found from the site of La Madeleine in south-west France; Magdalenian assemblages have been recognised in Belgium since the pioneering excavations of Édouard Dupont during the 1860s and 1870s (Dupont 1872). The basic typology of the Magdalenian in its ‘core’ area (south-western France) is traditionally divided into six stages, generally accepted as having at least some genuine chronological significance (these are usually noted in Roman numerals), the final stage being subdivided into groups ‘a’ and ‘b’. Recent work by Hemingway (1980) has helped to develop our understanding of the earlier stages of the Magdalenian and their definition, the later stages being fairly well characterised by the presence/absence of certain artefact types and their typology, notably bone and antler ‘harpoons’ (types of barbed point).
| |
|
|
|---|---|---|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
Table 1: Radiocarbon dates available for the Belgian Magdalenian in 1990.
Although in other areas of Europe the Magdalenian has been clearly defined chronologically by the use of radiocarbon dating, at the outset of this review in 1990 there were few radiocarbon dates for the Magdalenian in Belgium (a total of 15 were claimed: see Table 1). Instead, the dating of the Belgian Magdalenian had relied heavily on inter-regional comparisons. These few radiocarbon dates with apparent associations with the Belgian Magdalenian covered a broad temporal span ranging from 7720 ± 110 BP (Lv-1137) to 16270 ± 230 BP (Lv-1385). However, the majority of these fell within the range 12,000 to 14,000 BP.
I shall argue that many of these conventional radiocarbon dates give a ‘rough’ impression of the probable age of the Belgian Magdalenian. However, they cannot pass unquestioned and on closer inspection many, if not all, have a range of problems associated with them.
For example, the youngest of these dates (Lv-1137 7720 ± 110 BP) from the Trou des Nutons at Furfooz, if taken at face value, would appear to be totally inconsistent with current understanding of the temporal occurrence of the Magdalenian across Europe. The date comes from an undocumented sample of bone fragments. It appears to be far too recent (the Magdalenian is usually dated between 18,000 and 11,000 BP), and inspection of the lithics and fauna from the site indicates that this Magdalenian findspot has strong affinities with other Magdalenian locales in the valley (e.g. the Trou du Frontal at Furfooz and the Trou de Chaleux). There are also examples of conventional radiocarbon dates which appear to be too old, most notable amongst these are the dates from The Grotte de Sy Verlaine and the Trou des Blaireaux à Vaucelles. These sites will be discussed in further detail below. What is important here is that one is confronted with a particular problem: if the Belgian dates are accepted at face value, then the Lateglacial archaeology of Belgium differs dramatically from the rest of north-western Europe in its spatio-temporal distribution. This leaves two main options, either to accept that there is something intrinsically ‘different’ about the Belgian Lateglacial archaeological sequence which seperates it from that of the rest western Europe, or to question the assumptions which underly that conclusion. The latter approach is followed here.
If one is justified in rejecting the existing radiocarbon dates for the Magdalenian settlement of Belgium, then one is left in the unenviable situation of being left with no more than the typological structure developed by the Abbé Breuil (1912, 1927 & 1954) and subsequent workers such as Mme de Sonneville-Bordes (1960) for the French Magdalenian from which to deduce chronological patterning. There would also be predictable problems of regional variation to add to the shortcomings of relying on such a strategy.
On the basis of published descriptions of Belgian Magdalenian assemblages, only the later stages of the Magdalenian as recognised in France (V & VI) actually appear to be present in the Belgian Ardennes. Comparisons with other European assemblages ascribed to these stages indicate that the Belgian sites may span a range of perhaps the very late Dryas I/Bölling to Dryas II (roughly 13,000 to 12,000 BP). In fact, this would be consistent with suggestions that the origins and early development of the Magdalenian occurred in southern France and expanded to other parts of Western Europe only in its later stages (cf. Dewez 1992a).
What follows is an overview of particular Belgian Magdalenian sites at which radiocarbon dating has been applied, outlining the perceived problems, detailing more recent AMS work on the chronology and concluding with suggested explanations for some of these problems.
TROU DES BLAIREAUX À VAUCELLES (50° 07' 14" NORTH, 5° 44' 16" EAST)
This cave is situated in the south-western part of Belgium, to the south of Phillippeville. The site itself was first investigated archaeologically in 1904 (de Loë and Rahir 1905). More recent excavations by Claire Bellier and Pierre Cattelain (1983, 1986a, & 1986b) have revealed an important Lateglacial sequence, which is thought to give strong evidence for the earliest human re-colonisation of north-western Europe by “Magdalenian” hunter-gatherers (Straus 1991).| |
|
|
|---|---|---|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
Table 2: Radiocarbon dates from the Trou des Blaireaux
The radiocarbon sequence from this site is one of the most impressive for the Belgian Lateglacial (Table 2). However, a closer inspection of the published details of the site reveals some major problems. The ascription of the archaeological residues from this site to the Magdalenian appears to be based on the dates obtained for the site, rather than the presence of any diagnostic fossiles directeurs. Indeed the lithic industry, as published so far, would appear to consist of undiagnostic and broken backed pieces and debitage found in association with a faunal and floral spectra of Lateglacial age. Almost all the dates from this site come from samples found in three stratigraphically distinct bone accumulations in the lowest excavated level of the site (Couche III). The one exception to this is Lv-1386, which comes from Couche II. Lv-1385 and Lv-1558 come from the lowest concentration in Couche III (stratigraphically the oldest). This concentration appears to be purely faunal and lacks any artefacts (Bellier & Cattelain 1986a, 54). Worked reindeer antlers have been claimed (ibid.). Lv-1309D, Lv-1433 and Lv-1434D come from the middle concentration, approx. 80 cm above the lowest, which is said to have included bones of reindeer, horse, fox, hare, and ibex as well as several flints (ibid.). Lv-1314 comes from the uppermost concentration in Couche III, which had a high abundance of micro-mammal remains; whether any lithics were recovered from this concentration is unreported, though worked antler debitage is again claimed (ibid.). The dates from the Trou des Blaireaux have yet to be formally published in Radiocarbon, and so it has not been possible to determine precise details of the material sampled; where sample details have been reported in publications, the dates have been taken from bulk samples of reindeer antler.
During two research visits to Belgium I was able to visit Bellier and Cattelain in Treignes, where it was possible to examine briefly a substantial part of the faunal assemblage from the Trou des Blaireaux. Reindeer antlers were among the most abundant anatomical elements recovered during the excavation of Couche II and Couche III. The antlers examined from the site did not appear to have been humanly modified in any way (pace Bellier & Cattelain 1986a, 54 & 56) and it was not possible to confirm any of the “traces de coups de silex, cassures par flexion, ...” (ibid., 56) reported by the excavators.
Reindeer antlers which retained their basal portions were all shed, and all specimens were either female or juvenile as previously reported (ibid.; Bellier and Cattelain 1986b, 41). Numerous antler beam and tine fragments were present within the assemblage; however, there seems little to support Bellier & Cattelain’s suggestion (1986a & 1986b) that these specimens are debitage from antler working. It was not possible to locate any specimens which showed clear traces of human modification, such as groove and splinter technique, ‘ringing’, sawing or ‘pecking’. None of the specimens illustrated as humanly worked by Bellier & Cattelain (1983, figure 5) show any unquestionable signs of human modification, and (based on personal observations of the specimens concerned) at least some of the damage illustrated was due to carnivore activity.
In addition to this it seems unlikely that female and/or juvenile reindeer antlers would have been collected as raw material for artefact manufacture -- female/juvenile antler is inappropriate for this function, as it has neither the length nor breadth to provide adequate blanks for effective working. No artefacts were found associated with the lower antler accumulation. Consequently, it seems probable that instead of dating the human re-colonisation of Europe after the Last Glacial Maximum (pace Straus 1991, 271- 272), the two dates from this layer (Lv-1385 & Lv-1558) might instead be thought to give only an indication of the period during which the lowest antler concentration was being formed.
The middle concentration in Couche III is reported to have contained “quelques silex” (Bellier & Cattelain 1986a, 54), however, a direct link between human presence and any of the faunal assemblage remains undemonstrated. The dates associated with this concentration, as with those from the lower concentration, would seem to be solely of faunal relevance. A similar situation appears to be the case with the uppermost concentration in Couche III: it is unclear whether any artefacts were recovered from this context, and again it seems that the dates may only indicate a period of faunal accumulation during Dryas I. Consequently the suggested Magdalenian use of the site during the early Dryas remains unproven. Even the association between the single date from Couche II (Lv-1314) and the reportedly more abundant archaeological remains from this concentration (Bellier & Cattelain 1986b, 39) is dubious. Indeed, it has been demonstrated in a number of cases for the Lateglacial of north-western Europe (for example at the Robin Hood Cave, Creswell Crags, see Charles & Jacobi 1994) that an apparent association between artefacts and faunal remains within an archaeological unit, but which however lack evidence of direct human modification, can not necessarily be taken as proof of contemporaniety. At best, such samples may yield dates indicative of the temporal span over which material might have been incorporated into a sediment; at worst, there may be no real association of any kind between lithics and fauna.
In the case of the Trou des Blaireaux, it appears most likely that, rather than documenting the earliest phases of the human re-colonisation of north-western Europe after the Last Glacial Maximum, the conventional dates instead document the local presence of reindeer during the earliest part of the Lateglacial. At present it is difficult to make any realistic assessment of the extent of human input to the accumulation of the faunal assemblage from this site. A small number of atypical backed blades, most probably Lateglacial in age, were found within both Couche II and Couche III by the present excavators and during previous excavations (Bellier & Cattelain 1986a, 53-55).
An attempt was made to directly date human activity at the site, by
the application of AMS. A single specimen was noted which
showed a possible butchery mark -- a partial ulna of
Equus ferus provenanced to Couche II (VTB D7 5: x =
65.0 cm, y = 25.0 cm, z = 82.1 cm; Figure 4). This was submitted
to the Oxford Accelerator Unit for dating, resulting in OxA-4200
(13330 ± 160 BP).
However, subsequent examination of this specimen led to questions being raised about the validity of the mark initially identified as a butchery mark. It has not proven possible to clarify the situation regarding this mark. The specimen itself was de-mineralised and attempts at examining it with either an optical or scanning electron microscope have not been successful. Consequently, this sample has not helped clarify the position vis-à-vis the Lateglacial human settlement of this region.
If the specimen was modified, then it provides the only clear date for human activity at the site, and suggests that the claims for an exceptionally early human presence in Belgium after the Last Glacial Maximum (circa 18,000 BP), based on the previous radiocarbon dating evidence from the Trou des Blaireaux, might be unfounded. Similarly, the apparent lack of clear traces of human activity within the faunal assemblage would seem to argue for this.
Should the specimen be unmodified, however, the date would only be of interest for its biostratigraphic information, and could not be extrapolated to human activity. Consequently it would shed little light on the question of the human re-colonisation of north-western Europe after the Last Glacial Maximum. As the supposed nature of the sample has been brought into question, it seems wisest for the time being to treat this date solely as one for the local presence of Equus ferus.
It is clear from the preceding discussion that further work on the faunal assemblage is needed to clarify whether or not any of the faunal specimens from Blaireaux can be clearly linked to human activity. An analagous situation can be found in the arguments surrounding Reindeer Cave at Creag nan Uamh (Lawson & Bonsall, 1986a & 1986b). The presence of an accumulation of reindeer antlers, thought to date to the Loch Lomond Stadial, was suggested (ibid.) as evidence for human activity at the site. More recent research on this material (Murray et al. 1993), however, indicates that this is not necessarily the case. As with Blaireaux, the vast majority of the antlers are from female or juvenile reindeer (ibid., 6). The initial dating evidence from the site, based on bulked samples of antler and unmodified individual antlers, has also been brought into question. Further dating work using the AMS technique has suggested a far more complex situation, dates on individual antlers ranging between 22300 ± 240 BP (OxA-3792) and 47900 ± 3600 BP (OxA-3788) give some indication of the temporal spread of material within the accumulation. This has led Murray et al. (1993) to the conclusion that “there is less need to appeal to an agency of collection which behaved purposefully during a short period”(ibid., 7-8). Consequently “the previous inference that the reindeer antlers were introduced to the cave by man during the Loch Lomond Stadial is not supported” (ibid., 9).
It is likely that the situation at Blaireaux could be at least in part resolved by the further application the AMS technique to individual faunal specimens, should appropriate samples present themselves. It remains unclear whether any of the archaeological residues so far published from the site can be definitively ascribed to the Magdalenian, or any other Lateglacial technocomplex. The final report on the recent excavations at the Trou des Blaireaux is currently being prepared by Bellier & Cattelain. It is hoped that the problems raised here will be addressed in the course of this work.
GROTTE DE SY VERLAINE (50° 24' 15" NORTH, 5° 31' 30" EAST)
This cave lies between the two villages of Sy and Verlaine, near the river Ourthe. The cave itself is a small passegeway in the limestone hillside, approx 40 m from the rail line between Liège and Hamoir. There is no chamber to this cave, and the small passageway itself runs out after approx 10 m. The cave is also known as the Grotte de Verlaine and the Trou des Nutons.The single radiocarbon date from the Grotte de Sy Verlaine of 13780 ± 220 BP (Lv-690) is another crucial piece of evidence for the earliest human re-colonisation of north-western Europe after the Last Glacial Maximum. The date was taken from bone splinters excavated by P. Destinez and L. Moreels during May 1888 (Destinez & Moreels 1888), and little documentation, either published or archival, survives in relation to any excavations at the site. The radiocarbon date is usually interpreted as dating human occupation, and consequently the site has been given the cultural label of ‘Magdalenian IV’ (Dewez 1987, 391), presumably because that is the stage usually found at the time of Lv-690. The lithic and worked bone material is certainly typologically Magdalenian and it perhaps corresponds most closely with Magdalenian V within the ‘classic’ sequence (based on the presence of a uniserial harpoon/barbed point made of reindeer antler Nº 1513 Musée Curtis, University of Liège) (Doize 1960). Whilst discussing this problem with Michael Dewez it became apparent that the sole reason for the assignation of this assemblage to Magdalenian IV was indeed the early radiocarbon estimate and we concurred that further radiocarbon dates should be attempted on the material from this site if it were possible. It seemed that an ideal sample would have been part of the harpoon; however after consultation with Professor Otte at the University of Liège it became apparent that the use of this specimen for dating would be highly problematic. Inspection of the harpoon itself revealed that it had been treated at some point in the past with organic preservatives, so that any date obtained from the specimen could be open to doubt.
Subsequently the faunal assemblage was re-examined for butchery evidence and the verification of species identifications, one primary objective of this being the identification of single cut bones which could serve as samples for accelerator dating. The species present tallied with the description given by Destinez and Moreels (1888, CXLV-CXLVI), and the list, as they commented, contains species not usually found in association with Magdalenian (Lateglacial) faunal assemblages -- “le Rhinoceros tichorhinus, le mammouth, l’ours et l’hyène des cavernes” (ibid., CXLV). Many of these are more characteristic of earlier parts of the Pleistocene. Although bones of mammoths (Lister 1991) and bears have been found in Lateglacial contexts across Europe, Lateglacial occurrences of hyaena (Crocuta crocuta) have yet to be convincingly demonstrated and the extent of the distribution of the woolly rhinoceros during the Lateglacial in north-western Europe has yet to be established. Indeed, recent work by the author and R.M. Jacobi (Charles and Jacobi 1994) at Creswell Crags indicates that the local presence of hyaena and woolly rhinoceros in the British Isles does not extend into the latter part of the Devensian. The biostratigraphic distribution of these species in mainland Europe during the Lateglacial remains predominantly undocumented. Consequently the possibility that the faunal assemblage from this site could have accumulated over a considerable period of time was considered strong, especially as few stratigraphic observations were made by the nineteenth century excavators.
A small number of cut bones were present in the faunal collections from the site and eventually a cut left pisiform of horse (Equus ferus) was selected. As none of the faunal remains bore any provenance data beyond the name of the site, it seemed likely that as this specimen showed clear evidence of human modification it carried the greatest chance of dating human activity during the Lateglacial. The resulting date of 12870 ± 110 BP (OxA-4014) is, as expected, significantly later than Lv-690; given the potential complications of the original integrity of the Louvain sample, although the situation remains far from ideal, it does not seem unreasonable to suggest that the accelerator date should supersede it as a date for human presence. It is worth noting in passing that this date is far more in accord with other accelerator dates from the Belgian Magdalenian which will be discussed below.
LES GROTTES DE FURFOOZ: TROU DU FRONTAL (50° 12' 45" NORTH, 4° 57' 30" EAST) AND TROU DES NUTONS (50° 12' 45" NORTH, 4° 57' 29" EAST)
The caves at Furfooz are located by the banks of the river Lesse, and are today part of a National Park. They were amongst the first caves containing intact Pleistocene sediments to be systematically excavated during the latter part of the 19th century. Both the Trou du Frontal and the Trou des Nutons have clear archaeological evidence for Lateglacial human presence. Initial excavations at these sites by Édouard Dupont during the 1860s (Dupont 1872), revealed a wealth of Magdalenian material from a stratigraphically discrete unit termed the 1er niveau ossifère.In 1990 there were two radiocarbon dates believed to be associated with the Magdalenian use of these sites: Lv-1137 7720 ± 110 BP (Trou des Nutons) and Lv-1135 10720 ± 120 BP (Trou du Frontal). Both were substantially later than the usual expectations for the late Magdalenian for this region, which does not normally extend beyond the Dryas II/Alleröd interface at any other known site; again, the relationship between samples submitted and the archaeology they are claimed to date is questionable. Two further conventional dates have since been reported from the Trou du Frontal (Leotard 1993), Lv-1749 12950 ± 170 BP and Lv-1750 13130 ± 170 BP. The samples for both Lv-1137 and Lv-1135 are reported as being ‘cut bone splinters’ (bulked samples), no photographic records were made of the samples, and without these it is impossible to determine precisely how many, if any, were actually cut; in any case, the resulting age determinations merely record a mean age for the bone splinters, and not a date for the Magdalenian archaeology. Indeed, examination of the faunal assemblages from both sites, recovered by Dupont between 1864 and 1865 (Institut Royale des Sciences Naturelles de Belgique collections) clearly shows a mixing of faunas ranging in age from at least the Lateglacial to the very recent past. Dupont left no guide to stratigraphic differentiation within these assemblages after their curation in the museum stores. Instead they are referred to in his museum labels as originating from the 1er niveau ossifère. However, in published accounts of Dupont’s excavations at these sites it is apparent that there was a far more complex stratigraphic sequence which he originally observed and recorded (Dupont 1865, planche I).
The faunas from these two Furfooz caves are similar in many details (Charles 1994) and for the sake of clarity some details will be given here. Both sites have a very high proportion of domesticates within their respective collections -- sheep/goat (Ovis aries and/or Capra hircus) and pig (Sus scrofa). These domesticates are not usually considered to be components of north-western European Lateglacial faunal assemblages, and instead their presence within Lateglacial assemblages is often used as an indication of a possible contamination of Lateglacial material with later prehistoric/historic material. Unfortunately, no key to a more detailed stratigraphic differentiation now exists for these faunal collections.
Later Prehistoric, Roman/Iron Age, Mediaeval, and post-Mediaeval archaeological residues have been recovered from both the Trou du Frontal and the Trou des Nutons. Where butchery traces occur on the ovi-caprids, they have always been made by metal and not stone tools. The recent origin of many of the ovi-caprids is clearly demonstrated by the fact that some of the bones still retained their grease. The Suidae present a slightly more complex situation, the majority of these again being clearly of recent origin. However, it is possible that at least a few of these specimens are earlier or later prehistoric in age. It is currently unclear whether wild boar (Sus scrofa, which cannot be differentiated from domestic pig on purely morphological grounds alone) ever formed part of the Belgian Lateglacial megafauna (see also the discussion below in connection with OxA-4193). Some of the butchery marks found on this species are ambiguous as to whether they were caused by metal or stone tools, suggesting that at least some specimens may be of greater antiquity than others.
Other species recovered from these sites include reindeer (Rangifer
tarandus), horse (Equus ferus), red deer (Cervus elaphus)
and at least one form of bovid (Bos sp.). In assessing these faunas,
attempts were made to identify the elements of the Lateglacial fauna by a
combination of the following factors: species, preservation condition and
butchery evidence. This has led to a much reduced selection of fauna
attributed to the Lateglacial. This methodology was partially checked
by the AMS dating of samples of cut bone which also dated human use
of the caves. In all cases, specimens which were thought to be intrusive
yielded significantly more recent radiocarbon dates.
The two samples eventually selected were from a species which appears to have been very significant within the Lateglacial economy -- Equus ferus. Two further samples were selected from the faunal assemblages because it was of particular interest to know whether they were Lateglacial or more recent. The four samples were thus a cut 2nd phalanx of horse (Fig 5) and a cut left navicular cuboid of red deer from the Trou des Nutons; a cut left metacarpal of horse (Fig 6) and a cut right human tibia from the Trou du Frontal. The results are presented in Table 3.
Trou des Nutons (Furfooz)
| |
|
|
|---|---|---|
| |
|
|
| |
|
|
| |
|
|
|---|---|---|
| |
|
|
| |
|
|
Table 3: Accelerator dates from the Trou du Frontal and the Trou des Nutons, Furfooz.
Both OxA-4195 and OxA-4197 confirm human activity during the Bölling Interstadial phase of the Lateglacial, whilst OxA-4194 and OxA-4196 confirm my suggestion that faunal specimens of significantly different ages are present with at both sites. OxA-4196 also confirms the use of the cave as an ossuary during the Neolithic (human remains are abundant from this site). Although Dupont (1872) originally suggested that these human remains were of Palaeolithic age, this view has not been upheld by many subsequent researchers. A conventional date of GrN-10179 4430 ± 30 BP taken from human rib fragments almost mirrors the accelerator date. There seems to be no reason to treat the human remains as palaeolithic.
Returning to the broader problems of the Lateglacial chronology of Belgium, having discussed the dates at the extremes of the temporal range, one is left with a series of dates for the Magdalenian as shown in Table 1, ranging between 12150 ± 150 BP (Lv-686) and 12990 ± 150 BP (Lv-1569). Although at face value these would seem to fit well within current expectations for the presence of later Magdalenian groups in north-western Europe during the Bölling mild phase, one cannot operate a series of double standards and either accept or reject the validity of radiocarbon dates because they do, or do not, suit one’s preconceptions of the archaeological data. In fact almost all of the remaining conventional dates come from bulked samples, subject to the same problems of interpretation as the dates already discussed above. As there are so few sites directly dated, the remaining sites will be discussed individually.
TROU DE CHALEUX (50° 13' 18" NORTH, 4° 56' 30" EAST)
Chaleux is a relatively large cave, with an almost circular single chamber. Initially excavated by Dupont during the 1860s (Dupont 1872), more recent excavation were undertaken by the Université de Liège during the mid 1980s (Otte & Teheux 1986; Otte & Cabboi 1988). The final publication of this research excavation has recently become available (Otte et al. 1994).The archaeological assemblage recovered by Dupont includes 4,000 struck flints, a wealth of worked antler and bone, perforated semi- precious stones, as well as a number of fossil sponges, teeth, and shells from the Paris basin. The material is unquestionably Magdalenian, and is probably the most important Upper Palaeolithic collection from Belgium. Unfortunately, as with the Furfooz sites discussed above, the uppermost layers of the site have been grouped under the gross designation 1er niveau ossifère. As with the Trou du Frontal and the Trou des Nutons at Furfooz, there is evidence for a limited amount of mixing of faunas of very different age, probably spanning both the Lateglacial and at least parts of the Postglacial. Another interpretative problem exists in the shape of a range of cave dwelling predators present within the assemblage: it is apparent that at least part of the faunal assemblage was accumulated by non-human agencies (demonstrated by carnivore damage on bones), and that the activities of carnivores at the site is unlikely to be concurrent with human presence.
Three conventional radiocarbon dates are known from the site: Lv-1136 12710 ± 150 BP, Lv-1568 12370 ± 170 BP, Lv-1569 12990 ± 140 BP (Otte & Teheux 1986), which span much of the Bölling Interstadial phase. As all these samples were documented as “bone splinters” it was felt appropriate to make a more precise selection of further specimens for accelerator dating, with the following objectives in mind:
1. To test whether Lateglacial human presence might be more temporally restricted than the conventional radiocarbon dates suggest.
2. To test whether certain elements of the fauna from the 1er niveau ossifère were Lateglacial or more recent in age.
| |
|
|
|---|---|---|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
Table 4: Radiocarbon dates from the Trou de Chaleux.
Overall, 4 new dates were obtained. These, alongside the conventional dates for the Trou de Chaleux are presented in Table 4. OxA-4193, taken from a specimen of Sus scrofa, confirms the suggestion made above that material of substantially different ages is incorporated within the 1er niveau ossifère. There is still no clear indication as to whether wild boar was ever part of the local Lateglacial megafauna of north-western Europe, although this must remain a possibility. The other three accelerator dates fall in the region 12,700 to 13,000 BP. These are of a similar order to Lv-1136 and Lv-1569, although as the accelerator dates come from individual cut specimens these are the only dates which can be directly linked to human use of the site. It is interesting to note that the only date lying apart from the cluster of early Bölling dates, Lv-1568, is from a sample documented as “bone splinters”. It is tempting to suggest that this date may relate to faunal rather than human presence at the site, and that human presence may be restricted to the earlier part of the Bölling. However, current information about the spatio-temporal distribution of the Magdalenian in Belgium indicates that humans may well have been present in Belgium throughout the Bölling phase (see below). In fact, Lv-1568 cannot really be used in discussing that point at all, being, as we have seen, merely the average age of a collection of bone splinters of uncertain origin.
GROTTE DU COLÉOPTÈRE (50° 22' 20" NORTH,
5° 32' 15" EAST)
This cave is located by the banks of a small stream, within a small
limestone outcrop. Initial excavations at the Grotte du
Coléoptère by J. Hamal-Nandrin and J. Servais between
1923 and 1924 investigated the interior of the cave, as well as a
portion of the platform. Both Ahrensburgian and Magdalenian material
was recovered from the site in the course of these excavations, and
was reported to occur in stratigraphically distinct units. Most
striking amongst the Magdalenian material was the recovery of 5 uniserial
harpoons; the collection also includes a typical range of Magdalenian
diagnostic lithics (backed blades and bladelets, truncated blades and becs),
worked bone and antler and manuports. More recent excavations by
Professor Michel Dewez have confirmed the original observations of
Hamal-Nandrin & Servais (1925), as well as adding many further
details (Dewez 1975).
| |
|
|
|---|---|---|
| |
|
|
| |
|
|
| |
|
|
Table 5: Radiocarbon dates for the Magdalenian from the Grotte du Coléoptère.
Two conventional radiocarbon dates (Table 5) were linked to the Magdalenian occupation at the outset of research. In line with the majority of other Lateglacial dates for the Belgian Magdalenian, they fall within the Bölling phase. Only one date has any clear sample details (Lv-717 from a “reindeer bone”): it is not recorded, however, whether the bone was modified, or what its precise relationship with the archaeological assemblage was. After an evaluation of the faunal assemblage from the Dewez excavation, it was felt that further dating might be profitable. Professor Dewez and Dr. J-M. Cordy were kind enough to release a modified 1st phalanx of Equus ferus. The resulting date (OxA-3635) appears to be significantly earlier than the two conventional dates, as Table 5 shows.
The new date is the only one which can be directly linked to human activity at the Grotte du Coléoptère during the Bölling. It falls broadly in line with the accelerator dates from Chaleux, the Furfooz caves, and Sy Verlaine. Indeed, the initial accelerator project which yielded 4 dates associated with the Magdalenian (OxA-3632, OxA-3633, OxA-3635 and OxA-4014) suggested that Magdalenian presence in Belgian might have been restricted to the earlier part of the Bölling Interstadial phase. One reason to extend the original project was to test this hypothesis.
GROTTE DE LA BOIS LAITERIE (50° 21' 45" NORTH, 4° 51' 45" EAST)
The Grotte de la Bois Laiterie (referred to in earlier publications as the Trou Burnot) is currently under excavation by M. Otte, J-M. Leotard and L.G. Straus (Otte et al. 1994; Otte et al. 1995). It is located in a limestone outcrop overlooking the river Burnot, between the towns of Dinant and Namur.J-M. Leotard (pers. comm.) suggested that three Magdalenian sagaie bases recovered during his excavations at this site could be useful dating samples for further mapping of the temporal distribution of the Belgian Magdalenian. He also indicated that the site had a complex stratigraphy, including a Magdalenian layer immediately overlain by Neolithic burials, consequently direct dating of local Magdalenian fossiles directeurs might give the best opportunity for dating Lateglacial human presence at the site. We were both keen, therefore, to attempt to date one of the Magdalenian fossiles directeurs from this site, as there could be no subsequent arguments over association. Eventually one double-bevelled sagaie base was selected as a dating sample, provenanced to excavation unit 3A at the site. The resulting date, OxA-4198 12660 ± 140 BP was very much in line with prior expectations. Further AMS dating work has since taken place at the site, the results of which are statistically indistinguishable from the Oxford date (Table 6).
| |
|
|---|---|
| |
|
| |
|
| |
|
Table 6: AMS dates from the Grotte de la Bois Laiterie.
Back to Quick Menu
Verheyleweghen recovered a small amount of typically Magdalenian
material. The recent excavations at the Trou da Somme (Leotard 1988)
have also recovered a range of Magdalenian artefacts. Two separate
caves have been identified by Leotard, which he has termed TDS I and
TDS II. It is TDS II which Leotard refers to as the “most important”,
and which he subsequently refers to as the "Trou da Somme" (ibid., 17).
A piece of worked antler, circular in cross section, which may
well have been part of a sagaie, was submitted for accelerator
dating. OxA-4199 (12240 ± 130 BP) came out slightly more
recent than had been expected. However, few doubts can be raised
about the validity of the date. Rupert Housley (pers. comm.) has
confirmed that that there are no chemical problems with the date,
and there is no indication of the sample’s having been ‘conserved’
or subjected to any other form of contamination. Instead, the date
suggests that human presence was not confined to the earlier part
of the Bölling Interstadial phase.
However, it should be noted that the worked antler fragment cannot
be conclusively identified as a sagaie, and until the archaeological
material is published in detail from the Trou da Somme, the
identification of this site as a Magdalenian find spot is open to
at least some debate. The antler fragment could equally have come
from other artefact types which are not Magdalenian fossiles directeurs.
One such possibility would be antler artefacts similar to the ‘marrow
scoops’ found at Church Hole Cave, Creswell Crags (one of these is
illustrated in Dawkins 1880, figure 57). The two from Church
Hole have been dated to 12020 ± 100 BP (OxA-3717) and 12250
± 90 BP (OxA-3718). Another such ‘scooped’ antler rod is known
from Fox Hole Cave, Derbyshire which has also been AMS dated (OxA-
1493 11970 ± 120 BP). It is interesting to note
that these three dates fall very much in line with OxA-4199 from
Trou da Somme, however no such ‘scooped’ artefacts from Late Upper
Palaeolithic contexts have yet been noted in Belgium. It is not
currently possible to establish the precise nature of the antler
artefact of which the dated specimen from Trou da Somme originally
formed a part.
The end result of this was a series of 9 AMS dates which could be directly
linked to human presence at the sites concerned (Table 7).The date of 13330
± 160 BP (OxA-4200) from the Trou des Blaireaux has been omitted from
this table on two grounds: firstly, that it remains unclear whether the
lithics from the site can be attributed to the Magdalenian, and secondly,
that there is some degree of question over whether or not the sample was
modified.
TROU DA SOMME (50° 13' 38" NORTH, 4° 52' 45" EAST)
The Trou da [sic] Somme lies in a rocky outcrop known as the
Roche-al-Rue, on the left bank of the river Meuse approx. 4.5 km
west-south-west of the Trou de Chaleux. It was from this locality
that J. Verheyleweghen (cited in Dewez 1987 and Leotard 1988)
recovered Magdalenian material. Dewez (1987, 245) reports
that many of the small limestone outcrops at Roche-al-Rue are
very close to the Dinant-Givet railway, and that the construction of
this line involved the destruction of many of these shelters. It is
currently unclear whether the Trou da Somme is the same site as that
excavated by Verheyleweghen, or whether the latter was one of the small
shelters destroyed during construction of the railway.GROTTE WALOU (50° 35' 15" NORTH, 5° 40' 30" EAST)
This collapsed cave site in the Magne Valley is currently under
excavation by Professor Michel Dewez. Magdalenian archaeological
residues have been claimed from the site (Dewez 1992b), alongside
Creswellian. The associated lithics from the site have not yet been
published in any form, and the precise nature of Lateglacial use at
the site is currently unclear. Two radiocarbon dates have been
published as having Magdalenian associations: Lv-1582 13030
± 140 BP and Lv-1593 13120 ± 190 BP. Both are
from samples of bone fragments, and there are as yet no published
archaeological associations. The lithics from the Lateglacial layers
have yet to be published. Until further details are known about the
archaeological assemblages from the Lateglacial layers at the Grotte
Walou the significance of these two dates remains unclear, but given
the nature of the samples, one cannot have much confidence that they
accurately date a Magdalenian occupation. OVERVIEW OF THE BELGIAN MAGDALENIAN
From the above discussion, it will be clear that many of the fundamental
assumptions surrounding the radiocarbon chronology for the Magdalenian in
Belgium, as known in 1990, are open to grave doubt. As has been described,
there were as many problematic dates as there were ones in line with broad
expectations, and as double standards cannot be used when assessing the
nature of precise contextual links between radiocarbon dates and the
archaeological assemblages they are though to be associated with, it was
decided that further radiocarbon work on the Belgian Magdalenian was a
priority, ideally on samples which could be unequivocally linked with
human presence. The most logical way of doing this, was the application
of AMS dating techniques to bones from stratified contexts which had clear
evidence for human modification.
| |
|
|
|
|---|---|---|---|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
| |
|
|
|
Table 7: AMS dates on humanly modified bones from Magdalenian contexts in the Belgian Ardennes.
The majority of dates lie in the earlier part of the Bölling. However a few of these dates tend towards the latter part of the Bölling. This would seem to indicate that we are potentially looking at human presence in the Belgian Ardennes throughout this Interstadial phase, although whether such settlement was continuous or highly punctuated is at present unclear. The Late Magdalenian artefacts recovered from the sites discussed here can all be placed within stage V of the classic typo-chronological sequence of the Magdalenian. At the time of writing we have no direct indications as to whether use of individual sites in this region was highly seasonal or year-round. The research presented here is only a beginning, there is clearly a need for much further work on this data set.
CONCLUSIONS AND PROSPECTS FOR FURTHER RESEARCH
Two main problems relating to the chronology of the Belgian Magdalenian have been highlighted in the preceding pages. Firstly the number of well dated sites, and of actual dates from them, is currently inadequate to sustain any detailed analysis of temporal patterning. Secondly, are the poor quality of the majority of dates already available. This problem is not unique to Palaeolithic archaeology; however, it is within this field that such problems usually become the most acute. Sample selection and documentation are two crucial elements in the production of radiocarbon dates that make a date of lasting use. Some of the problems created by poor sample documentation have been addressed by the two AMS dating projects outlined and discussed above. However, these can only be a beginning.There is a need for further dating work on sites which presently lack any absolute dates, and further dating of collections which have been recently re-evaluated and are currently under evaluation. The dating evidence discussed within these pages has gone some way towards resolving the question of the date of the earliest human re-colonisation of this small region after the Last Glacial Maximum. There seems to be no clear evidence for human presence in Belgium at or around the Late Glacial Maximum itself (c. 18,000 BP). Very early conventional radiocarbon dates, such as those from the Trou des Blaireaux or the Grotte de Sy Verlaine, appear to be extremely dubious, probably because of the nature of the samples used. In the case of the Grotte de Sy Verlaine, one can argue convincingly that the conventional date (Lv-690) has been superseded by an accelerator date (OxA-4041) on a directly modified specimen. Sadly, the situation is left ambiguous in the case of the one accelerator date so far available from the Trou des Blaireaux (OxA-4200). Indeed, the earliest dates which can be seen as reliable are the accelerator dates from the Trou de Chaleux, Grottes de Furfooz, Trou da Somme, Grotte de la Bois Laiterie, Grotte du Coléoptère and Grotte de Sy Verlaine. These all fall at the latter part of Dryas I/Bölling interface. A consistent chronology appears to be hinted at within these, suggesting that the Magdalenian may be restricted to a far more limited period of occupation in Belgium than hitherto expected. Perhaps humans did not move this far north until the milder Bölling phase was beginning. The precise nature of human ‘re-colonisation’ and presence in Belgium (and more generally north-western Europe) after the Late Glacial Maximum and during the Lateglacial is still largely unknown. What is clear is that Magdalenian hunter-gatherers appear to have been the pioneers who moved back into this part of Europe. Whether we are detecting a single move back into the north, or many smaller ‘events’ is currently unknown.
Acknowledgements
I would like to thank the staff of the Oxford Accelerator Unit for their help and advice throughout the two dating projects reported here. I am extremely grateful to Professor Lawrence Straus who gave access to his results from the Grotte de la Bois Laiterie and Mietje Germonpré who gave access to the faunal collections at the IRScNB. Jean-Marie Cordy, Michel Dewez, Marcel Otte and Jean-Marc Leotard permitted samples to be dated from their research collections. Derek Roe supervised the D.Phil. dissertation of which this research formed a part, and Roger Jacobi made many insightful comments relating to this issues discussed in this paper. The research reported here was undertaken whilst the author held a major state studentship from the British Academy. Additional support was given by Christ Church, Oxford and the Meyerstein Fund of Oxford University.BIBLIOGRAPHY
Barton, N., Roberts, A.J. and Roe D.A. 1991. The Late Glacial in north-west Europe: human adaptation and environmental change at the end of the Pleistocene. London: CBA Research Report 77.Becker, B and Kromer, B. 1991. Dendrochronology and radiocarbon calibration of the early Holocene. In Barton, N., Roberts, A.J. and Roe, D.A. (Eds.) The Late Glacial in north-west Europe: human adaptation and environmental change at the end of the Pleistocene, 22-24. CBA Research Report 77.
Becker, B. 1993. An 11,000 year German Oak and Pine Dendrochronology for Radiocarbon Calibration. Radiocarbon 35(1), 201-213.
Bellier, C. and Cattelain, P. 1983. Fouilles au “Trou des Blaireaux” à Vaucelles (Doische - Prov. Namur) Campagnes 1981-1982. Notae Praehistoricae 3, 42-49.
Bellier, C. and Cattelain, P. 1986a. Le Trou des Blaireaux à Vaucelles. Helinium 26, 47-57.
Bellier, C. and Cattelain, P. 1986b. Le Trou des Blaireaux à Vaucelles. Aslira 18, 35-49.
Breuil, H. 1912. Les subdivisions du Paléolithique Supérieur et leur signification. Congrès International d’Anthropologie et d’Archéologie Préhistorique.
Breuil, H. 1927. Oeuvres d’art paléolithiques inédites du Périgord et art oriental espagnol. Revue Anthropologique 101-108.
Breuil, H. 1954. Le Magdalénien. Livre Jubilaire de la Société Préhistorique Française 1904-1954. Les Grandes Civilisations Préhistoriques de la France. Bulletin de la Société Préhistorique Française 51(8), 59-64.
Charles, R. 1994. Food for Thought: Late Magdalenian chronology and faunal exploitation in the north-western Ardennes. D.Phil. dissertation. Oxford University.
Charles, R. and Jacobi, R.M. 1994. Lateglacial faunal exploitation at the Robin Hood Cave, Creswell Crags. Oxford Journal of Archaeology 13(1), 1-32.
Dawkins, W.B. 1880. Cave hunting. Macmillan & Co. London.
Destinez, P. and Moreels, L. 1888. Exploration de la caverne de Verlaine (Luxembourg): Notice préliminaire. Annales de la Société Gèologique de Belgique 15, CXLV-CXLVII.
Dewez, M. 1975. Nouvelles Recherches à la Grotte du Coléoptère à Bomal-sur-Ourthe (Province de Luxembourg). Rapport provisionaire de la prèmier campagne de fouilles. Helinium 15, 105-133.
Dewez, M. 1987. Le Paléolithique Supérieur Récent dans les Grottes de Belgique. Louvain-la-Neuve: Publications d’Histoire de l’Art et d’Archéologie de l’Université Catholique de Louvain LVII.
Dewez, M. 1992a. Le Magdalénien en Belgique, origine et filiation. Le Peuplement Magdalénien. Paléogéographie Physique et Humaine. CTHS Colloque de Chancelade.
Dewez, M. 1992b. La Grotte Walou à Trooz (province de Liège, Belgique), presentation du site. In M. Toussaint (Ed.) Cinq Millions d’années, l’Aventure Humaine. Liège: Etudes et Recherches Archéologique de l’Université de Liège. 311-318.
Doize, R-L. 1960. Le grand Harpon de Verlaine (Belgique). Bulletin de la Société Royale Belge d’Anthropologie et Préhistoire 71, 25-31.
Dupont, E. 1865. Rapport Adressé à M. Le Ministre de l’Intérieur sur les fouilles scientifiques exécutées dans la Province de Namur pendant l’Année 1864. Bruxelles: Deltombe.
Dupont, E. 1872. L’homme pendant les ages de la pierre dans les environs de Dinant-Sur-Meuse. 2e Ed. Bruxelles: Deltombe.
Hamal-Nandrin, J. and J. Servais 1925. Contribution a l’étude du Paléolithique Supérieur en Belgique. La Grotte dite “du Coléoptère” rapport sur les fouilles, 1923-24. Revue Anthropologique Nºs 4, 5 &6, 120-144.
Hedges, R.E.M. 1986. The Future Prospects of Accelerator dating. In J.A.J. Gowlett and R.E.M. Hedges Archaeological Results from Accelerator Dating. Oxford: Oxford University Committee for Archaeology Monograph.
Hedges, R.E.M., Housley, R.A., Bronk Ramsey, C. and Van Klinken, G.J. 1993. Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 16. Archaeometry 35(1), 147-167.
Hedges, R.E.M., Housley, R.A., Bronk Ramsey, C. and Van Klinken, G.J. 1994. Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 18. Archaeometry 36(2), 337-371.
Hemingway, M.F. 1980. The initial Magdalenian in France. Oxford: British Archaeological Report S90.
Housley, R. 1991. AMS dates from the Late Glacial and early Postglacial in North-west Europe: A review. In Barton, N., Roberts, A.J. and Roe, D.A. (Eds.) The Late Glacial in north-west Europe: human adaptation and environmental change at the end of the Pleistocene, 25-39. CBA Research Report 77.
Huxtable, J., and Aitken, M.J. 1986. Dating of European Flint by Thermoluminescence. In Olin, J.S. and M.J. Blackman (Eds.) Proceedings of the 24th International Archaeometry Symposium Washington DC: Smithsonian Institution Press.
Lawson, T.J. and Bonsall, C. 1986a. Early settlement in Scotland: the evidence from Reindeer Cave, Assynt. Quaternary Newsletter 49, 1-7.
Lawson, T.J. and Bonsall, C.1986b. The Palaeolithic in Scotland: a reconsideration of evidence from Reindeer Cave, Assynt. In S.N. Collcutt (Ed.) The Palaeolithic of Britain and its nearest neighbours: recent trends. Sheffield: University of Sheffield, Dept. of Archaeology.
Leotard, J.-M. 1988. Occupation Magdalenienne au Trou da Somme massif de Roche-al-Rue (Waulsort). Notae Praehistoricae 8. 17-23.
Leotard, J.-M. 1993. Tourisme Magdalénien dans la région Dinantaise. Notae Praehistoricae 12, 63-64.
Lister, A.M. 1991. Late Glacial mammoths in Britain. In N. Barton A.J. Roberts and D.A. Roe (Eds.) The Late Glacial in north-west Europe: human adaptation and environmental change at the end of the Pleistocene, 51-59. CBA Research Report 77.
Loë, A. de and Rahir, E. 1905. Fouilles au Trou des Blaireaux à Vaucelles. Bulletin de la Société d’Anthropologie de Bruxelles 24(1), 1-17.
Mangerud, J. Anderson, S.T., Bergland, B.E. and J.J. Donner 1974. Quaternary stratigraphy of Norden, a proposal for terminology and classification. Boreas 3, 26-31.
Murray, N.A., Bonsall, C., Sutherland, D.G., Lawson, T.J. and Kitchener, A.C. 1993. Further radiocarbon determinations on reindeer remains of Middle and Late Devensian age from the Creag Nan Uamh caves, Assynt, NW Scotland. Quaternary Newsletter 70, 1-10.
Otte, M., Baden-Powell, D., Cabboi, S., Charles, R., Cordy, J-M., Gilot, E., Lejeune, M., Leotard, J-M., Noirel-Schutz, C., Patou-Mathis, M., and Schoch, W. 1994. Le Magdalénien du Trou de Chaleux (Hulsonniaux) - Belgique. Liège: Etudes et Recherches Archéologique de l’Université de Liège 60.
Otte, M. and Cabboi, S. 1988. La Grotte de Chaleux, resultats preliminaires de la fouille 1988. Notae Praehistoricae 8, 25-37.
Otte, M. and Teheux, E. 1986. Fouilles 1986 à Chaleux. Notae Praehistoricae 6, 63-77.
Otte, M., Straus, L.G., Lacroix, Ph., Martinez, A., Noiret, P., Léotard, J.-M., Ancion, V and Lopez Bayon, I. 1994. Fouilles 1994 à l’Abri du Pape et à la Grotte du Bois Laiterie. Notae Praehistoricae 14, 45-68.
Otte, M., Straus, L.G., Léotard, J.-M., Gautier, A., López Bayon, I., Lacroix, Ph., Teheux, E., Ancion, V. and Orphal, J. 1995. Rapport préliminaire sur le site Magdalenien final du Bois Laiterie (Wallonie, Belgique). Notae Praehistoricae 15, 11-33.
Sonneville-Bordes, D de 1960. Le Paléolithique supérieur en Périgord. Bordeaux.
Straus, L.G. 1991. Human Geography of the Late Upper Paleolithic in Western Europe: Present State of the Question. Journal of Anthropological Research 47(2), 259-278.