• -10 000
  • -9 000
  • -8 000
  • -7 000
  • -6 000
  • -5 000
  • -4 000
  • -3 000
  • -2 000
  • -1 000
  • 0
  • 1 000
  • 2 000

Docteur Jean-Georges Rozoy

Résumé des abréviations utilisées dans les articles : consulter la liste.





By the term Epipalaeolithic the author means all the cultures of hunters-fishermen-collectors previous to the introduction of agriculture and animal breeding and which are characterised by the use, in large numbers (20 to 60% of the tools) of microlithic armatures others than the backed bladelets of Upper Palaeolithic. The term : microlithic armature, which has been precisely defined in previous papers (Rozoy 1967, G.E.E.M. 1969, 1972) is applied with rare exceptions to implements with abrupt retouches, made out of bladelets or thin blades, not exceeding 5 cm in length and above all never more than 4 mm thick.

For 8 years the author has worked principally on problems of chronological and geographical classification of Franco-Belgian Epipalaeolithic industries, classification based principally (owing to the force of circumstances and absence of satisfactory conservation of other evidence) upon the detailed typology of the flints, which are sorted, in every case, by the author himself (Rozoy 1973). The guiding principles of his method is to make use of, throughout modern excavations of sites with the complete conservation of finds, all the conserved elements, in whatever order they may be, no matter whether it is a question of type of station, animals hunted, the style of cutting up the blades or of qualitative or quantitative typological factors. Priority is always given to human data not depending on (or depending very little on) the influence of environment. A great deal of attention is given to elements indicating the internal evolution of each industry.

The ethnological problems : reconstruction of the way of life, nature and motives of its modifications, etc., seemed to the author possible to approach with any hope of results, only with the basis of chronological and geographical classification. In fact it is doubtful, whether the ethnology of an inland human group of the Boreal can be brought to light by established facts about another group living at the coast at Atlantic time.

This classification has imposed a tedious piece of work whose method (Rozoy 1966, 1967, 1968, 1969; G.E.E.M. 1969 and 1972), a partial summary (Rozoy 1971 b) and some detailed elements (Rozoy 1971 c, 1972) have been published and which will be the object of a following monography (Rozoy 1973). The time has now come to consider the problems of all categories posed by these results. Some of them will be found in the following pages.

The Tardenoisian.

This term, previously very general (Clark 1936, 1970; Barrière 1956), must be restricted to the culture which gave birth to it and which is limited to a part of the Parisian Basin : Tardenois and the Paris region essentially. To assimilate Tardenoisian with the trapezia, and vice versa, is a gross methodological error and must be foresaken as well as the assimilation to all industries containing microlithic armatures, but without axes. Instead of dividing the European Epipalaeolithic into two cultures: Maglemosian and Tardenoisian (Clark 1936, 1970) it is necessary to recognise in it multiple human groups which are more or less related, of which Tardenoisian is but one example, and it is not the most complete for we still have no knowledge of his early phases nor of the stages of development since the Upper Paleolithic.

We are aware, in Tardenoisian, of two facies -one without "typical" trapezia (Coincy facies, Daniel 1943, 1948), the other with numerous "typical" trapezia (Montbani facies, Daniel 1933, 1948).

The Coincy facies (fig. 2) is characterised by an approximative balance between 5 classes of armatures : obliquely truncated points, crescents, backed bladelets,(these, sometimes rare sometimes plentiful), scalene or isocele triangles (the latter older) and points with transversal base. The tools of the common stock which are very scarse - the end-scrapers are rare and occasionally absent - are mainly truncated bladelets, retouched bladelets or bladelets with a regular single notch (6th class of the author's list). Whereas the irregularly notched and retouched blades and bladelets (Montbani blades and bladelets) are almost completely missing.

The style of cutting up is that of Coincy (Rozoy 1968 b) -the process of section of the bladelets with microburin is widely used (but not exclusively), producing very regular armatures, which are found in numbers comparable to those of the microburins. The Coincy facies is now well dated from Boreal (Parent 1966, 1969 1971, 1973). The ground of the wrong datings (Hinout 1965) having been established (the late covering over by the dunes at the time of the Neolithic deforestation has selectively conserved the coal and the pollen spectrum contemporaries of this covering over), Coincy facies can be established as being the middle Tardenoisian.

The Montbani facies (Daniel 1948, and fig. 3), which is the recent Tardenoisian (Rozoy 1971c) contains a quarter of tools from the common stock comprising end-scrapers (10% Belgian influence), a quarter of the armatures from the middle period, especially triangles and Tardenois points, a quarter of trapezia, especially asymmetrical, 80% lateralised on the right (95% on the 1eft to the South of the Seine) and a quarter of Montbani blades and bladelets. The cutting up style is that of Montbani (Rozoy 1968 b). The process of section with microburin is always widely used. The most probable dating, by comparison, is at the limit of Boreal and Atlantic (Rozoy 1973).

The final Tardenoisian, characterised at the Allée Tortue (Parent 1967, fig. 4), no longer uses the armatures of the middle period; the Montbani blades and bladelets form half of the tools, and armatures with flat inverse retouch appear, which will prove to be the only element adopted by the later intruders, the "Bandkeramik" culture people.

The final Tardenoisian seems to disappear without sequel with the intrusion of the "Bandkeramik" culture which came (already !) from the East. This regional "Bandkeramik" Neolithic adopts a few arrowheads with flat inverse retouch (fig. 4 n° 6), but all the rest of his stock of tools is imported and there is no "Tardenoisian tradition" in the north of the Parisian Basin (Rozoy 1971 c ; Bailloud 1971).

The on the spot neolithisation seems to have existed in the southern part of the Parisian Basin : surface site of Sébouville 1 (Rozoy 1971 c, 1972, and fig. 5).

The proportions of tools, in spite of some variations of ethnological origin (Rozoy 1973), provide a sufficiently characteristic picture of each phase, with all the elements of transition from one to the other (fig. 6).

Other cultures - Evolution on the spot.

A good number of other human groups have been individualised in the Boreal and the Atlantic times (Rozoy 1971 b). The list is not complete and there remain a good deal of regions to examine and to prospect. In a work in preparation (Rozoy 1973) the author has been able to recognize four complete chronological series from Paleolithic to Neolithic.

At the Belgo-Dutch border the passage to the Epipalaeolithic takes place in the Dryas III and can be well studied in the series of the B.A.I. Groningen (Bohmers excavations). Starting from the Tjongerian well dated during the Alleröd, the local Ahrensburgian expands without perceptible outside influences, comprising more obliquely truncated points than tanged points. The backed bladelets of the Upper Palaeolithic disappear. Then, passing through a transitional phase (Geldrop III-2) we reach the middle stage with Aardhorst dated about 6700 B.C. with truncated backed bladelets, triangles, obliquely truncated points and numerous microburins. Then it proceeds at least two culture groups with points with covering retouch, Coincy cutting up style and without microburins, dated from the Boreal (Oirschot V) which later develop trapezia made out of bladelets in the Montbanl style, but without Montbani blades and bladelets (Maarheeze, Bohmers 1956; Lommel). These latter tools and the microburin method arise very late (fusion with Tardenoisian). Neolithic (late "Bandkeramik") is intrusive.

In the Birse valley, at the Franco-Swiss border (researches carried out by Pr. Bandi, Bern) we pass from the advanced Magdalenian (Brüggli) to an equivalent of Azilian (Neumühle) without backed bladelets, and to the sequences of Mannlefelsen I and Rochedane (Thévenin 1972 a and b) with straight-backed points, sequences perhaps parallel with Neumühle, then to that of Birsmatten (Bandi 1963 ; Rozoy 1972). On this site the probably Pre-boreal horizon 5 does not involve backed bladelets ; the horizons 4 and 3, parallel to the middle Tardenoisian (but with abundant end-scrapers and retouched flakes), are to be dated from the Boreal (Rozoy 1973), whereas the appearance of typical trapezia (horizon 2) can be situated before the turning point Boreal-Atlantic. The Montbani blades and bladelets, already present in the horizon 3, are nowadays greatly developed (up to 40%) but the typical trapezia, very scarse indeed, appear to have been substituted by harpoons made out of bone or antler. The cutting up style of the horizon 2 is intermediate between those of Coincy and Montbani and similar to that of Montclus (Castelnovian). The sequence reach its height at Liesbergmühle VI (researches by Grüter) with a collection very near to that of Allée Tortue and dated at 4400 B.C.

In the surroundings of Marseille (Researches by Escalon de Fonton) the passage to Epipalaeolithic takes place at the episode of Alleröd in the course of Romanellian / Valorguian (Escalon 1966 a, 1966 d, 1968 a) with short end-scrapers and Istres points (Escalon 1972), the latter resulting from the transformation of backed bladelets of the upper Palaeolithic. From it results Montadian without end-scrapers, but with many retouched flakes and a great variability in the proportions of armatures, to the extent that in the eponymous site of La Montade there were none (Escalon 1951, 1956).

The ancient Montadian (with hyperpygmean crescents) dating from Dryas III seems to give birth to two industries at least, middle Montclusian (dated at the Boreal) and the middle Montadian with unpygmean armatures, which then evolves towards 5800 B.C. in the Castelnovian with end-scrapers, typical symmetrical trapezia, then asymmetrical, microburins and Montbani blades and bladelets, whereas the Montclusian for its part passes into a stage with both triangles and trapezia, of which only the beginning is known at present and does not comprise Montbani blades and bladelets nor asymmetrical trapezia.

Then on the spot evolution of the Castelnovian passes into the "cardial" Neolithic by acculturation on a local basis (Escalon 1966 b and c, 1970, 1971); Guilaine 1971) with a difference of 500 years or perhaps more between the coast and the interior, an interval during the course of which the evolution of armatures continues and completely renews their panoply twice; passage to the Châteauneuf triangle, then to the Montclus arrowhead, with a decrease and disappearance of the microburin.

In the South-West, starting from Azilian (Champagne and Espitalié 1970) dated to Alleröd, the change over Epipalaeolithic takes place in Dryas III and Preboreal, passing through industries with backed bladelets which are not still very well known (Coulonges 1963; Orliac 1972) and which end in Sauveterrian,

The latter, created by Coulonges (1935), can be well studied at Rouffignac (Barrière excavations; Barrière 1965 a, b, c, in preparation). It is dated for its ancient stage at Preboreal (Champagne and Espitalié 1972; Barrière 1965 a). The style of cutting up the blades is particularly rough (Rouffignac style, Rozoy 1968 b). There are many retouched flakes and tools on crude blades; the Rouffignac knives could be a local particularity. There are less truncated bladelets than in Tardenoisian and no regular, partially retouched bladelets, the percentages of armatures are much less with predominance of pygmean triangles (isoceles, then scalenes, the method of microburin has been used quite widely in the beginning, but less and less in the course of evolution.

The Montbani blades and bladelets appear here before the trapezia, on supports of the Rouffignac style, and assure the continuity of the sequence with the recent Sauveterrian with trapezia; these are firstly symmetrical (and coexisting with the armatures of the middle stage), then asymmetrical (Martinet trapezium, G.E.E.M. 1969) and with the Montclus style. This industry may no longer be called Tardenoisian as its tools are different both qualitatively and quantitatively, and also from the point of view of the style, this recent Sauveterrian is dated for the appearance of the trapezia at 5850 +/- 50 B.C. and its evolution seems to go on a long time before the neolithisation, whose modalities are still badly known (Lacam 1944) but contain Montclus arrowheads.

It is still worth noting, without exhausting the list of cultures, that the Breton Téviécien (Péquart 1937 and 1954 with numerous trapezia and cutting up in the Montclus style, cut without microburins, cannot be assimilated with Tardenoisian because its composition, both qualitative and quantitative, is different. Its derivation from a regional middle stage has been established, with bipartition, for the evolution seems very different in the North and in the South of the mouth of the Loire (Rozoy 1971 b). The Téviécien is dated to 4625 +/- 350 (Rollando 1965).

At last Beaugencien (Rozoy 1972) is another nowadays still badly known culture of the middle Loire region, where a middle stage is known, with pygmean armatures, and a late stage passing, on the spot, into the Neolithic. In this region neither typical trapezia nor Montbani blades and bladelets are known. In place of trapezia we have armatures derived from the Tardenois point. The microburin method is used much more intensively than elsewhere, the number of microburins reaches 5 or 6 times that of the founded armatures, and this (contrary to other regions) continues in the old and perhaps the middle Neolithic.

The "Law of the sands".

It has long been claimed that "Tardenoisian" people - in broad sense of the word - lived only on the sands (Octobon 1923). Some saw in this the proof of their African origin (Vignard 1954 and personal ccmmunications in the same way as multiples interventions at the S.P.F.), other saw there the "ill-adaptation to coping with forests" (Clark 1936, 1970). In fact Tardenoisian people and their contemporaries lived and hunted on every sort of ground : clayish-calcareous (Les Richoux, Rousseau 1967 ; Canneville, Debruge 1910), limy (Vieilles, Cahen 1913 ; Belloy-sur-Somme, Rozoy 1973): schistous (Roc La Tour, Rozoy 1973); besides, most of the islets of sand, particularly those of Tardenois (Parent 1971, p. 137 and 180; Parent 1962) are not widespread enough to permit the survival of a group, even small, living on hunting and fishing; the rapid exhaustion of game would quickly lead (in a few weeks) to an exodus from the sandy area. And, of course, the hunters have no need to "coping with forests", for these are their natural habitat.

The "law of the sands" (Octobon 1923; Clark 1936, 1970) is in fact that of the commodity of research and has only an objective basis of the most restricted kind, if indeed it has any at all. On non-sandy ground it is difficult to sieve, the flints are stuck in the earth, and it would be necessary to wash ; besides, the station is only discovered in general during a Neolithic prospection, and this have given rise to imaginary facies originating from mixtures.

We have seen that the Neolithic with Tardenoisian tradition, limited to the south of the Parisian Basin, comprise only very evolved armatures with flat inverse retouch ; further South it is a matter of other industries, Beaugencien (Rozoy 1973) and others, but always with evolved armatures (Rozoy 1971 a), those of the middle stage (crescents) not being present except in the case of a mixture, which is frequent (Cordier 1958, 1964). Thus, realising the number of the mixed stations, and bearing in mind the greater difficulty for prospection and chiefly for sieving outside the sands, with stronger evidence again for proper excavation, it appears that the occupation of the soil is nearly equal whatever may be the nature of the ground. In any case the main interest in the sands is to constitute glades, and this was a restricted advantage at Boreal, for the pine and birch forest is much more open and lighted and contain more ground vegetation (therefore more game) than that of Atlantic. It was the same, indeed, in Preboreal, in Dryas III and Alleröd.

The qualitative and quantitative composition of the industries from the caves and shelter sites are the same as those from open-air sites, excavated in the same region. It is the case,for example, for Remouchamps (cave) and Geldrop III-l (open-air) or in Tardenois for the caves of Le Lendemain and Le Troglodyte, compared to St.Pierre-les-Nemours or Larchant (Rozoy 1973). There are quantitative variations of ethnological origin at the heart of the same industry, denoting some specialisation of work; they have a bearing especially, on the entire relationship between armatures and tools of the common stock, that is to say probably between hunting and domestic activity, and are evident at the heart of the same type of station : Les Fieux and Rouffignac, for Sauveterrian, for example, indeed even inside a site: Chaintréauviile (Rozoy 1973).


Where it is known (Birsmatten, Rouffignac, Montclus, Téviec), food is based upon wild boar and red deer (and roe deer), with a more appreciable proportion of small animals and birds to Atlantic (Birsmatten). The rabbit is particularly plentiful in Romanellian /Valorguian (Escalon 1968). Fish is very strongly represented at Montclus which is on the bank of the Cèze river (however it is possible that in other sites the conditions of preservation were not propitious), and the shellfood at Téviec, but in both cases we find also the red deer and the wild boar, and it is worth noting that there is more to eat on one of these mammals than on a great number of fishes or shellfish.

Nomadism is certain (because of the exhaustion of the game reserves), but it is not less certain that it finds place in a restricted area, sometimes recognizable, for example by the distribution of the Wommersom sandstone-quarzite, whose radius is in the order of 60 to 80 km (with points at 120 km, but without heavy concentrations (Hamal-Nandrin 1913 ; Ophoven 1948). It is worth noting that the Wommersom site is not central for this distribution, which underlines the human factors in question : the people who knew Wommersom went willingly to Oirschot and the surroundings (80 km), but hardly crossed the Meuse (35 km) and thus, never beyond Liège (40 km) ; to the West they hardly went past Braine-le-Comte (54 km).

Numerous are the examples of these boundaries, in most cases purely human and without natural obstacles. Thus we have the very long duration of cultures without end-scrapers in the whole of Paris Basin, with a late diffusion into this sphere of "mistletoe leaves" and end-scrapers coming from Belgium ; the non-penetration of the Montbani blades and bladelets into Belgium, except very late ; the absence of trapezia at Beaugency (nor of Montbani blades and bladelets). The replacement of the late Montclusian by Castelnovian at Montclus, cannot be explained otherwise than by the existence of stable human groups on their native soil, and holding to their own traditions, which does not hinder recognising eventually contacts or influences. In the latter case (Montclus) moreover, a shifting of the border must have taken place, which does not mean a migration, far from it.

More often, nevertheless, we can see a continuous variation in case of a shifting of ground in one period. Thus in the Boreal the proportion microburins/armatures passes from 5 to 6 at Beaugency to 4 in Seine-et-Marne, 2 around Paris, 1 in the Tardenois, falling to nearly 0 in North-Brabant. The boundaries were then very penetrable.

Who began?

Similar phenomena, though not identical, happen independently at points far apart. Such is the case for the widespread recourse to microlithic armatures which start in North-Brabant in the course of Tjongerian dated to Alleröd and enlarges in Ahrensburgian at the turn of Alleröd and Dryas III (Geldrop I : 9010 +/- 95 B.C.). Near Marseille, it is during the course of Romanellian /Valorguian dated to Alleröd and following a mechanism similar to that of Tjongerian (transformation of backed bladelets into points). Azilian of the South-West, likewise in Alleröd, sees, like TJongerian and Romanellian / Valorguian, with short end-scrapers, the diminution of burins and a tendency toward microlithisation and the multiplication of points, which will enlarge in the course of Dryas III but with points with straight backs, whereas those of Romanellian / Valorguian are spindle-shaped and those of Tjongerian with a curved back. In Dryas III the first will carry on with the straight backs, the second will move to the obliquely truncated points and the tanged points, the third to the hyperpygmean crescents. Other places - nowadays less well dated - are known, each one with its own style and types and its evolutive sequence.

In the same way, in Boreal appeared everywhere (but not in Montadian) backed bladelets, mostly truncated, which do not derive from those of the Upper Palaeolithic and are a reinvention, with a particular style.

Likewise, the appearance of the many and typical trapezia: it is dated to 5720 +/- 110 B.C. in Hatert (North Brabant) (confirmed by Nijnsel II and Kuykgestel to 5870 and 5550 B.C.), to 5850 +/- 50 in Rouffignac, before 5570 +/- 240 in Châteauneuf (Escalon 1967). But these were not the very first, which seems to be dated at Montclus circa 5800 - 5900 B.C. The dating at Birsmatten (H 2) to 5250 +/- 600 is also compatible with these dates, and 5700 is more likely if we take the animal environment and the sediments into account.

Many other examples could be quoted.

Each one of these novelties, of course, is but the reflection of an action made by each one with his own style. It were not the armatures which spread, but the tool (made of wood ?) of which they were the complement: probably the bow and the arrows at Alleröd-Dryas III (the very last Magdalenian people are not longer using propellers), but what else is there in the Boreal for the backed bladelets ? And what else again before the beginning of the Atlantic age, to explain the trapezia ?

It is very difficult under these conditions to say who began even if it is a matter of diffusion or simultaneous invention. We can only assure that these people who at the same time adopted (or invented) the same novelty had reached very comparable technical and economical levels; otherwise they could not have taken advantages of that new fact.

Regional disparities.

The general scheme is very univocal: strong regression of the burins, shortening of the scrapers, systematic change for microlithic armatures, then to trapezia and notched blades. But there are neither two areas nor two cultures where that scheme is followed in the same way. In the whole of France and Belgium the very strong regression of the burins is the rule (fig. 8), but we know that in Maglemosian (probably for making bone points with or without barbs) a fairly big amount of them will be kept in use for a long time.

Likewise the shortening of the scrapers: in Romanellian / Valorguian they are made out of blades, are very small and flat: those of Geldrop or Aardhorst are much bigger, in the Birstal (Birsmatten H 5) they are made out of flakes. The Montadian people will give them up, the people from Brabant will stick to them. The trapezia do not occur in Beaugency nor in the Retz area, nor the Montbani blades and bladelets, which Brittany, too, will never know. In Birstal, they are many but there are the trapezia which are very little developed. No harpoons made of antler, in Castelnovian neither at Rouffignac nor Téviec. At Monclus the trapezia change to the transverse arrowhead with covering retouches, whereas Tardenoisian discovers the Sonchamp point in the South of the Seine, but its north part at that time amalgamates itself with the Belgian group which has only just given up its "mistletoe leaves". These are only a few examples.

It is clear that there was at every time several equivalent possible solutions and that the survival did not require the trapezia (for the Beaugency people did without them).

These regional disparities should make us use very cautiously the "intercultural elements" (Kozlowski 1969) for the datings.

Grounds of the transformations.

It is often claimed that the changes in "Postglacial industries" have been the result of the warmer weather, the growth of the forest and the arrival of the red deer in the place of the reindeer (and soon). This forgets the fact that the changes in the industries begin sometimes in the end of Dryas II (La Valduc, Escalon 1966 d), before the end of the very cold period, sometimes in the Dryas III when it occurs again: the Ahrensburgian people were hunting the reindeer with microlithic armatures whereas their ancestors, the Tjongerians, were hunting the red deer with armatures of a more Palaeolithical kind (and less numerous). More than a matter of a sudden change it is one of a tendency to the "microlithisation", augured by isolated trials as early as in Magdalenian (Gare de Couze, Bordes 1964) and achieved over a long period of time which runs from Dryas II to Preboreal.

As for the introducing of the trapezia, which appears everywhere to be earlier than the beginning of Atlantic; the first occupation of Châteauneuf takes place in such a dry weather that Castelnovian people settled in the bed of the old brook and gradually filled it in, without any hint of gullying.

Therefore it is very doubtful that the reason of the transformations is climatic, and in any case this theory needs to be argued more strictly than it has been presented up till now.

The reason of the transformations seems to be more a cultural one and be connected with the discovery of new technics, irrespective of the environment.

The question of the axes.

It is usual to contrast the cultures without axes : Tardenoisien, Sauveterrien, Téviécien, etc. with the northern cultures with flint axes : Maglemosien, Erteböllien, and to infer rather contemptuous ideas about the "miserable" life of the former group. Apart the fact this is a cultivator's way of arguing, who is afraid in the woods and only thinks of cutting down the forest, it is not certain that the facts have been accurately established. A lack of flint axes does not mean necessarily a total lack of tools adapted to the cutting of wood.

We know that in northern cultures, before the neolithisation the amount of axes is always small, and very much less than armatures' one; a few pieces are made of bone (Star Carr, Clark 1954) or of antler (Schuldt 1961; Schwabedissen 1964). Bone or antler axes, if they existed in the cultures "without-axes", will be found in a very small amount, considering the conditions of preservation even in the propitious sites.

And as a matter of fact, that is what happened with the pieces of Birsmatten (Bandi 1963, p. 27 n° I, 2, 3, p. 299 n° 6), of Montclus, of Rouffignac, of Bellefonds (Patte 1971), of Muge (Roche 1960, p. 82 n°1, and 1951, pl. II n°2). The last one is very similar to northern objects (Schwabedissen 1964, p. 392, fig. 7 c) described as "chisels" whereas the former objects could be compared with those from Star Carr and Hohen Viecheln, quoted above (fig. 8, n° 1-8).

Almost all of them come from sites which yielded trapezia, that is to say of the Atlantic period when the forest was thicker. The fragment from Rouffignac is the only one to be dated in the Preboreal (layer 4 c). Birsmatten yielded a good half dozen of them (nine if we include the more doubtful ones). None has been found helved but the helves were almost certainly made of wood, like the hafts of the armatures.

These objects which are all broken (at the same place as the polished axes) are usually described as "skinning tools" (Fellablöser) or "tools for barking" (which implies the likely tanning of skins, but for which we have no positive or negative evidence). In fact the use of tools is always a delicate problem that the discovery of a haft could bring forward. It seems that the medium piece made of antler was only necessary to hold the stone axe, besides, the part made of antler was itself held in (or perforated by) a wooden helve. The direct fitting of the bone axe or antler axe in wood should therefore simplify the process, if not a superiority of the "miserable" men of the forest over the men of the marshy fens, whose "wealth" is only that of a better preservation of their remains. A wooden handle can be preserved only in the peat; therefore it is towards the open-air dwellings at the side of the lakes (especially those in South Germany) that we should turn our attention.

If this interpretation is confirmed, it would diminish greatly the contrast - very artificial - between the northern Epipalaeolithic cultures and the others.


The author expresses his gratitude to the persons and Institutes quoted in this article who readily welcomed him, allowed him to study their material (supplying most of the time excellent working and even lodging conditions), who advised and guided him.

He is also indebted to Mrs Flouest and Mrs Holder for the translation and to Cl. Marolle, who kindly drew all the illustrations, and he would now like to take the opportunity to thank all for their participation.


G. BAILLOUD 1971 - Le Néolithique danubien et le Chasséen dans le Nord et le Centre de la France, dans : Fundamenta, Reihe A, 3, Köln

H.-G. BANDI 1963 - Birsmatten, Basisgrotte. Bern.

C. BARRIERE 1956 - Les Civilisations tardenoisiennes en Europe Occidentale. Paris.

C. BARRIERE 1965a - Le gisement de la grotte de Rouffignac (Dordogne), dans : Actes du Congrès International des Sciences Préhistoriques et Protohistoriques. Rome.

C. BARRIERE 1965b - Les couteaux-faucilles de Rouffignac, "Revista da Facultade de Lettras de Lisboa", 9.

C. BARRIERE 1965c - Flore et faune de Rouffignac (Dordogne), "Annales de la Faculté des Lettres et Sciences Humaines de Toulouse", Nouvelle série, 1.

C. BARRIERE en préparation - La grotte de Rouffignac (fouilles).

A. BOHMERS, A. WOUTERS 1956 - Statistics and graphs in the study of flint assemblages. "Palaeohistoria", 5.

F. BORDES, F. FITTE 1964 - Microlithes du Magdalénien supérieur de la Gare de Couze (Dordogne), dans : Miscellanea en hommage à l’abbé Kenri Breuil. Barcelona.

A. CAHEN 1913 - Contribution à l'étude des "Tout petits si1ex" tardenoisiens de Vieilles. "Bulletin de la Société normande d’Etudes préhistoriques", 20.

F. CHAMPAGNE, E. ESPITALIE 1970 - L'abri du Roc d'Abeilles à Calviac (Dordogne). "Gallia-Préhistoire”.

F. CHAMPAGNE, E. ESPITALIE 1972 - Note sur une datation du Sauveterrien de la grotte des Fieux à Miers (Lot), "Bulletin de la Sooiété Préhistorique Française»

J.G. D. CLARK 1936 - The Mesolithic Settlement of Northern Europe. Cambridge.

J.G.D. CLARK 1954 - Excavations at Star Carr. Cambridge.

G. CORDIER 1958 - Contribution à l'étude préhistorique de la vallée de l'Indre, La station tardenoisienne des Chaumeries, oommune de Murs (Indre). "Bulletin de la Société Préhistorique Française".

G. CORDIER 1964 - Contribution à l'étude préhistorique de la vallée de 1’Indre. La station tardenoisienne de la Blancharderie, commune de Perruson (I.-et-L.). "Bulletin de la Société Préhistorique Française".

L. COULONGES 1935 - Les gisements préhistoriques de Sauveterre-la-Lémance. Dans : Archives I.P.H. 14, Paris.

L. COULONGES 1963 - Magdalénien et Périgordien post-glaciaires. La grotte de la Borie del Rey (Lot-et-Garonne). "Gallia-Préhistoire".

R. DANIEL 1932 - Nouvelles études sur le Tardenoisien français. "Bulletin de la Société Préhistorique Française", 29.

R. DANIEL 1933 - Nouvelles études sur le Tardenoisien français. Stations tardenoisiennes pures. "Bulletin de la Société Préhistorique Française", 30.

R., M. DANIEL 1948 - Le Tardenoisien classique du Tardenois. "L’ Anthropologie", 52.

R. DANIEL, J.G. ROZOY 1966 - Divers types d'armatures tardenoisiennes à base retouchée. "Bulletin de la Société Préhistorique Française".

R. DANIEL, E. VIGNARD 1953 - Tableaux synoptiques des principaux microlithes géométriques du Tardenois français. "Bulletin de la Société Préhistorique Française", 50.

R. DANIEL, E. VIGNARD 1954 - Le Tardenoisien français. Dans : Livre Jubilaire de la S.P.F.

A. DEBRUGE 1910 - La station préhistorique de Canneville près Creil (Oise). Dans : Congrès préhistorique de Franoe, 1909.

M. ESCALON DE FONTON 1956 - Préhistoire de la Basse-Provence. "Préhistoire", 12.

M. ESCALON DE FONTON 1966a - A propos de quelques datations C-14, de la France et d’Italie. "Bulletin de la Société Préhistorique Francaise", 50-51.

M. ESCALON DE FONTON 1966b - Du Paléolithique supérieur au Mésolithique dans 1e Midi méditerranéen. "Bulletin de la Société Préhistorique Française".

M. ESCALON DE FONTON 1966c - Informations arohéologiques. Circonscription de Languedoc-Roussillon. "Gallia-Préhistoire".

M. ESCALON DE FONTON 1966d - Le oampement romanellien de La Valduc a Istres (Bouches-du-Rhône). "L’Anthropologie", 70.

M. ESCALON DE FONTON 1967 - Datation au C 14 du Cardial ancien de Châteauneuf. "Bulletin de la Société Préhistorique Française", 64.

M. ESCALON DE FONTONn 1968a - Le Romanellien de la Baume de Valorgues à St Quentin-la-Poterie (Gard). Dans : La Préhistoire - Problèmes et tendances.

M. ESCALON DE FONTON 1970 - Deux nouvelles datations C-14 pour la fin du Néolithique ancien (épi-Cardial). "Bulletin de la Société Préhistorique Française, 67.

M. ESCALON DE FONTON 1971 - Les phénomènes de néolithisation dans le Midi de la France. Dans : Fundamenta, Reihe A, 3. Köln.

M. ESCALON DE FONTON 1972 - La pointe d'Istres. Note typologique. "Bulletin de la Société Préhistorique Française".

M. ESCALON DE FONTON, G. Dumas 1951 - La grotte de la Montade n° 3. "Revue d’études ligures", 1.

G.E.E.M. 1969 - Epipaléolithique-Mésolithique : les microlithes géométriques. "Bulletin de la Société Préhistorique Française".

G.E.E.M - 1972 – Epipaléolithique-Mésolithique : les armatures non géométriques, 1. "Bulletin de la Société Préhistorique Française".

E. GIRAUD, C. VAOHE, E. VIGNARD 1938 - Le gisement mésolithique de Piscop. "L’Anthropologie", 48.

J. GUILAINE 1971 - La néolithisation du bassin de l'Aude et des Pyrénées méditerranéennes françaises. Dans : Fundamenta, Reihe A, 3, Köln.

J. HAMAL-NANDRIN, J. SERVAIS 1913 - Etudes sur le Limbourg belge préhistorique. Dans : Annales du Congrès de la Fédération archéologique. Gand.

J. HAMAL-NANDRIN 1948 - voir Ophoven.

J. HINOUT 1964 - Gisements tardenoisiens de I'Aisne. "Gallia--Préhistoire", 7.

S.K. KOZLOWSKI 1969 - Le Mésolithique de la Pologne. "Archaeologia Polona", 11.

R. LACAM, A. NIEDERLANDER, H.V. VALOIS 1944 - Le gisement mésolithique du Cuzoul de Gramat. Dans : Archives I.P.H. 21, Paris.

OCTOBON 1923 – La question tardenoisienne. "Bulletin de la Société Préhistorique Française".

M. OPHOVEN, E. SACCASYN DELLA SANTA, J. HAMAL-NANDRIN 1948 - Utilisation à l’Age de la Pierre (Mésolithique) du Grès-quartzite de Wommersom.

E., M. ORLIAC 1972 - Fouilles à la grotte de La Tourasse (St Martory, Haute-Garonne). "Revue de Comminges", 85.

R. PARENT 1962 - Répartition des gisements néolithiques et tardenoisiens dans la région de Fère-en-Tardenois (Aisne). "Annales de la Faculté des Lettres de Toulouse", 5.

R. PARENT 1966 - Nouvelle datation du Tartenoisien du Tardenois. "Bulletin de la Société Préhistorique Française".

R. PARENT 1967 - Le gisement tardenoisien de l'Allée Tortue à Fère-en-Tardenois (Aisne). "Bulletin de la Société Préhistorique Française".

R. PARENT 1969 - Nouvelle datation du Tardenoisien du Tardenois par le C.14 (Sablonnière de Coincy 1’Abbaye). "Bulletin de la Société Préhistorique Française».

R. PARENT 1971 - 72 -Le peuplement préhistorique entre la Marne et l’Aisne. "Travaux de l’Institut d'Art Préhistorique de l’université de Toulouse", 13, 14.

R. PARENT 1973 - Publication en préparation sur La Sablonnière II, Montbani II. "Bulletin de la Société Préhistorique Française".

E. PATTE 1971 - Quelques sépultures du Poitou du Mésolithique au Bronze moyen. "Gallia-Préhistoire", 14.

M., S.-J. PEQUART, M. BOULE, H.V. VALOIS 1937 - Téviec, station -nécropole mésolithique du Morbihan. Dans : Archives I.P.H., 18, Paris.

M., S.-J. PEQUART 1954 - Hoédic. Antwerpen.

R. ROBERT, E. VIGNARD 1945 - Les campements mésolithiques du "Désert d’Auffargis" (Seine-et-Oise). "Bulletin de la Sooiété Préhistorique Française", 42.

J. ROCHE 1971 - Les amas coquilliers (concheiros) mésolithiques de Muge (Portugal. Dans : Fundamenta, Reihe A, 3, Köln.

J, ROCHE 1951 - L'industrie préhistorique du Cabeco d’Amoreira (Muge). Portugal.

J. ROCHE 1960 - Le gisement mésolithique de Moita do Sebastiao (Muge). Portugal. "Archeologie".

Y. ROLLANDO 1965 - Datation du gisement d'Hoédic, "Bull. Soc. Polym. Morbihan", 1285ème séanoe.

G. ROUSSEAU 1967 - La station de surface à l'industrie tardenoisienne des Richoux, commune de Vaux-sur-Lunain (S.et M.)."Bulletin de la Société Préhistorique Française".

J.-G. ROZOY 1966 - voir Daniel et Rozoy.

J.-G. ROZOY 1967a - Essai d'adaptation des méthodes statistiques à 1’ Epipaleolithique (" Mésolithique"). "Bulletin de la Société Préhistorique Française”.

J.-G. ROZOY 1967b - Typologie de l'Epipaléolithique franco-belge. "Bulletin de la Société Préhistorique Française".

J.-G. ROZOY 1968a - Typologie de l'Epipaléolithique ("Mésolithique”) franco-belge. Introduction, outils communs, lamelles à bord abattu. "Bulletin de la Société Préhistorique Française".

J.-G. ROZOY 1968b – L’étude du matériel brut et des microburins dans 1’ Epipaleolithique ("Mésolithique”) franco-belge. "Bulletin da la Société Préhistorique Française".

J.-G. ROZOY 1969 - Typologie de l'Epipaléolithique ("Mésolithique") franco-belge. Reunion of 5 Separates of "Bulletin de la Société Préhistorique Française" 1966, 1967, 1968.

J.-G. ROZOY 1971 a - Microburins et armatures microlithiques dans le Néolithique, "Bulletin de la Société Préhistorique Française".

J.-G. ROZOY 1971b - Tardenoisien et Sauveterrien. "Bulletin de la Société Préhistorique Française".

J.-G. ROZOY 1971o - La fin de l'Epipaléolithique ("Mésolithique") dans le Nord de la France et la Belgique. Dans : Fundamenta, Reihe A, 3, Köln.

J.-G. ROZOY 1972 - L'évolution du Tardenoisien dans le Bassin Parisien. "L’Anthropologie”, 76.

J.-G. ROZOY - 1973 - Les derniers chasseurs. Essai de synthèse sur l’ Epipaléolithique en Franoe et en Belgique. En préparation E. Saccasyn della Santa 1948 - voir Ophoven.

E. SCHULDT 1961 - Hohen Viecheln, ein mittelsteinzeitliohen Wohnplatz in Mecklemburg. Berlin.

H. SCHWABEDISSEN 1964 - Sinngehalt und Abgrenzung des Mesolithikums nach den Forschungsergebnissen im nordlichen Teil des europaischen Kontinents. Dans : 6e Congrès International du Quaternaire, Warszawa 1961. Lodz.

A. THEVENIN, J. SAINTY 1972 a - Une nouvelle stratigraphie du post-glaciaire : l’abri du Mannlefelsen à Oberlarg (Haut-Rhin). "Bulletin de la Société Préhistorique Française".

A. THEVENIN, J. SAINTY 1972b - L'abri de Rochedane, à Villars sous Dampjoux (Doubs). "Revue arohéologique de l’Est et du Centre-Est", 23.

C. VACHE - voir Giraud, Vache et Vignard.

E. VIGNARD 1954 - Sur les civilisations tardenoisiennes en Europe occidentale. "Bulletin de la Société Préhistorique Française".


Fig. 1. - Quoted sites.

Aar - Aardhorst, Abri Cornille - see Istres, Allée Tortue – see Fère~en~Tardenoies, Ba - Baumles, Bg - Beaugency, Be - Belloy-sur-Somme, Birs – Blrsmatten, Bla-La Blancharderie, Bor - La Borie del Rey, Br. C. - Braine-le-Comte, Br. – Brüggli Höhle, Can – Canneville, (Abri) Capeau - see Istres, Chai - Chaintréauville, Ch. M. – Châteauneuf-les-Martigues, Chau – Les Chaumeries, C - Coincy (La Sablonnière), Cu - Cuzoul de Gramat, F - Fère-en-Tardenois (Allée Tortue), Fieux - Les Fieux, Gd - Gand (Port Arthur), Gel – Geldrop, Hat - Hatert, H – Hoédic, I – Istres (Abri Cornille, Abri Capeau, Le Valduc), J – Jägerhaus, Ker - Kerhillio, Lar - Larchant, Le Valduc – see Istres, L - Lautereck, Le – Lendemain, Liè - Liège, Lies – Liesbergmühle, Luy - Luykgestel, Maa - Maarheeze, Mar - Marseille (Abri de St Marcel), Man - Mannlefelsen, Mil - Milheeze, Mont - La Montade, Mb - Montbani, M – Montclus, Moos – Moosbühl, Ne – Neumühle, Oir - Oirschot, Rem - Remouchamps, Ric - Les Richoux, Roc –Roc d'Abeilles, RLT - Roc La Tour, R - Rouffignac, Sa - Sauveterre-la-Lémance, Sé- Sébouville,St P – Saint Pierre les Nemours, Tour - La Tourasse, TV - Téviec, T - Le Troglodyte, Val - Valorgues, V – Vieilles, W - Wommersom.

Fig. 2. - The Sablonnière of Coincy : balanced picture.

Coll. Daniel. One tool out of four (the only end-scraper is included). Perceptible balance between the five classes of armatures. No real trapezia.

Fig. 3. - Montbani 13 : balanced picture.

Coll. Daniel, One tool out of fifty five. Notice the asymmetrical trapezia (15 – 21), the notched blades (22).

Fig. 4. - Allée Tortue : balanced picture.

Coll. Parent. One tool out of twenty four. There are no longer armatures of the middle stage.

Fig. 5. - Sébouville I : Montbani blades and bladelets and armatures.

Coll. Peron. It should be remembered that this is a selection, the armatures only represent 8 % of the material and the Montbani blades and bladelets 33 % ; there are also "tranchets", polished flint axes. etc. The site (if homogeneous) in fact seems basically neolithic but its interest is to show the last phase of evolution of the trapezia and the notched blades.

Fig. 6. - Proportions of tools.

In four specific sites of the Tardenoisien : Coincy , the Sablonnière, 206 tools; Montbani l3, 1500 tools; Park of the ancient castle of Fère-en-Tardenois, 129 tools; Allée Tortue, 631 tools. Coincy is the middle stage : importance of the 7th to the 12th classes of armatures and of the 6th (tools out of bladelets). The "Parc" and Montbani 13 are very similar, the 7th and 8th classes are already fairly limited, the trapezia and the Montbani blades and bladelets well developed, give to the diagram that concavity looking up in an opposite manner to the previous. In the Allée Tortue : the armatures of the medium stage vanish (horizontal line) and the Montbani blades and bladelets progress. Note the appearance of the trapezia with the flat inverse retouches (n° 105).

Fig. 7. - Proportions of tools.

At Kerhillio (sélection Quatrehomme), Hoédic (sélection Péquart), Montbani 13 (sélection Daniel and Rozoy), Kerhillio and Hoédic are two sites near one to another, but Hoédic has more triangles and trapezia, the trapezia of Téviec (n° 103) and other asymmetrical pieces (n° 96 and 97) become more frequent than the symmetrical trapezia (n° 98 to 101). Marked differences with Montbani 13 : especially in the end-scrapers, the worked flakes (classes 1 and 2), the crescents and backed bladelets (classes 8 and 9). The types of trapezia are not the same and, most of all, the Montbani blades and bladelets (classe 14) are almost totally missing in Brittany, and this is becoming more and more the case. Besides that, the microburins, present in the medium stage, have vanished, whereas they survive in the Tardenois area.

Fig. 8. - Axes or not axes ?

That is the question, 1-3 - Birsmatten H 1 (Bandi 1963), 4 - Birsmatten H 2 after checking the objects themselves, 5 - Montclus, layer 12 B (Castelnovien), 6 - Muge (Roche 1960), 7 - Rouffignac 4 c, 8 - Bellefonds (Patte 1971).

Résumé des abréviations utilisées dans les articles : consulter la liste.

© Jean-Georges Rozoy - Tous droits réservés 2016