The Azilian Epipaleolithic Industry of Spain and France

In this post we present information, sourced from official publications, on the Azilian archaeological industry of the Epipaleolithic period of northern Spain and southern France.

Painted pebble – Mas d'Azil

Painted pebble – Mas d’Azil

From the paper titled “Chronostratigraphy of the Pleistocene/Holocene boundary: The Azilian problem in the Franco-Cantabrian region“, by Lawrence Guy Straus, 1985, we read among other things:

Conclusions Assemblages labelled as ‘Azilian’ can be found in deposits dating from as old as 12,000 B.P. to as recent as 9,000 B.P. They can span, as an ensemble, the Allerød, Dryas III and Preboreal pollen zones or periods, and thus straddle the Pleistocene/Holocene boundary as it is traditionally (albeit arbitrarily) defined.

It is clear that there is considerable variability and disaccord among the different methods used to date the Azilian deposits.

Site-by-site and level-by-Ievel there are significant differences among the dates derived from the various methods. Only very rarely is there agreement among the 14C determinations and both the pollen and sediment ‘dates’. In only a few more cases is 14C in agreement with either pollen or sediments. However even in these cases of apparent agreement there are instances of equivocal 14C dates or problematical Azilian attribution. There are many more instances where palynology and sedimentology are in apparent agreement but are either in disaccord with 14C dates or lack radiocarbon controis. Whether in cases of one-way dating, two-way or threeway chronological concordance, as in disaccord, the dates for assemblages assigned to the Azilian culture-stratigraphic unit clearly do span the Allerød, Dryas III and Preboreal, with more problematical tails to the distribution in the Dryas II and Boreal. Frequently radiocarbon dates are younger than palynological ‘dates’, particularly with regard to deposits in northern Spain palynologically attributed to Allerød.

There is considerable evidence of temporal overlap on the one hand with ‘Magdalenian’ assemblages (including ones with cylindrical section harpoons) and, on the other hand, with ‘Mesolithic’ ones. Of course part of this overlap may be due to the many uncertainties and errors which are inherent to the radiocarbon method. However, as the Magdalenian, Azilian and Mesolithic are all part of a cultural continuum, with few clear-cut, abruptly-marked differences in technology, real overlap is to be expected. Indeed cases of physical association of such supposedly temporal diagnostics (‘fossil directors’) as cylindrical- and flat-section harpoons prove this point quite dramatically.

There is no such thing as a ‘typical’ Azilian assemblage. As in all Upper Palaeolithic industries, for example, there is considerable inter-assemblage variability both in terms of ‘fossil director’ types (harpoons, decorated pebbles) and in terms of relative frequencies not only of Azilian points and short endscrapers, but also microliths, endscrapers, burins and other major tool classes in general. There is admittedly a fair degree of circularity involved in the assignment of assemblages to the Azilian (or to any other culture stratigraphic unit arbitrarily concocted by prehistorians).

The Azilian is undeniably in some way related to the major readjustments in human adaptations which came about at the time of (and in part, due to) the glacial/interglacial transition which occurred during the period centered on the date of 10,000 B.P. The nature of this relationship is, however, far from clear.

Azilian assemblages of artifacts and fauna do indeed lie in between the Magdalenian and Mesolithic in composition and in their behavioral implications. The Azilian clearly belonged to two worlds and can genuinely be viewed as ‘transitional’. One can but only speculate that the apparently abrupt change in artistic activity coeval with the Azilian, might be a reflection of new forms of social organization. Indeed, it is likely that the adaptations to post-Pleistocene conditions may have involved considerable readjustments of social organization, including territorialism, at best only inadequately reflected by the standard sorts of data and analyses archaeologists use in the study of the Azilian.

Fragment 317 with a bifacial ornamentation side A) head of aurochs surrounded by radiating lines; side B) head of aurochs

Fragment 317 with a bifacial ornamentation side A) head of aurochs surrounded by radiating lines; side B) head of aurochs

From the paper titled “Epipaleolithic and Mesolithic adaptations in Cantabrian Spain and Pyrenean France“, by Lawrence Guy Straus, 1991, we read:

Abstract The period of deglaciation from ca. 13,000 to ca. 9000 B.P. along the northern edge of the Cantabrian Cordillera and Pyrenees was characterized by marked climatic and environmental oscillations, culminating in the establishment of interglacial conditions. While along the Cantabrian coast, late Upper Paleolithic groups had long been developing diversified systems of adaptation, fully exploiting the wide range of food resources of that narrow but ecologically varied region (notably red deer and marine mollusks), Magdalenian hunters along the southern edge of the Aquitaine basin were becoming increasingly specialized in the hunting of one medium-size game species, reindeer. Thus, while the artifact industries and artistic traditions of the two adjacent regions along the forty-third parallel developed along similar lines in the Magdalenian and Azilian, and despite a common montane specialization in ibex hunting, the changes that came with the end of the Last Glacial affected the human groups of the two regions very differently, as reflected in the early Mesolithic records of Vasco-Cantabria and Gascony, respectively.

Selection of engraved pieces from Le Rocher de l'ImpeÂratrice

Selection of engraved pieces from Le Rocher de l’Impératrice

From the paper titled “The Late Upper Paleolithic-Mesolithic-Neolithic Transitions in Cantabrian Spain“, by Lawrence Guy Straus, 2009, we read:

Abstract Although broad-spectrum subsistence began during the Solutrean (ca. 20,000 ¹⁴C BP) in Cantabrian Spain, and there was much continuity in technology and settlement between the Magdalenian and Azilian, there were dramatic changes in human use of the postglacial landscapes of this Atlantic region after ca. 9,000 ¹⁴C BP. Interrupting a Terminal Paleolithic trend toward increased utilization of the montane interior of the region, the Mesolithic was mainly a coastal phenomenon. Although the Magdalenian-Azilian transition did include disappearance of cave art, the marked adaptive break came after the traditional end of the Pleistocene, with concentration of Mesolithic sites along the Holocene shore and emphasis on marine resource exploitation. An exception was the interior Basque Country, where there was significant human occupation of the upper Ebro Basin. While Mesolithic hunter-gatherers of that Ebro region were quick to incorporate Neolithic cultigens, domesticated animals, and ceramics into their lifeways—by at least 6,500 ¹⁴C BP—their coastal Cantabrian neighbors continued to exist as fisher-gatherer-hunters until as recently as ca. 5,700 ¹⁴C BP, when the interior was again much used by people.

Azilian point - Tourasse Cave

Azilian point – Tourasse Cave

From the paper titled “A comparison of the Lithic industries from two Azilian sites in Aquitaine: how to interpret different degrees of technical simplification?“, by Celia Fat Cheung, 2014,  we read among other things:

This study compares Azilian technologies and their variability in southwestern France through the analysis of two archaeological assemblages. The first site, Abri Pagès (Pages rockshelter), south of the Dordogne River, is located in a context with relatively good access to siliceous raw materials. The second site, the Grotte de Troubat (Troubat Cave), is located on the Pyrenean piedmont near flint sources with nodules that are smaller than in the preceding context.

Conclusion With this comparison of two sites, several elements associated with the technical simplification of these industries were confronted. Broadly, these two assemblages share several features: at both sites, blades and bladelets were manufactured from cores using a stone hammer, and the preparation of these cores was minimal. Most of the short blades, wide bladelets, elongated flakes and regular flakes were retouched to make endscrapers, laterally retouched blades or elongated flakes (sidescrapers) and backed mono-points, sometimes with a retouched base.

Though both assemblages are associated with simple production techniques, this technical simplification is realized in a different manner. In Quercy, blade-bladelet manufacturing remains frequent, while in the Pyrenees, other methods are employed, such as flake manufacturing and anvil flaking. Blade manufacturing is thus not frequent at Troubat, being replaced by smaller products, such as bladelets, elongated flakes and flakes. What do these differences mean? Do they represent circumstantial adaptations by identical groups, or on the contrary, a regionalization perceptible in their cultural practices?

This microlithization (or miniaturization) in the Pyrenean reduction strategies concords with tools that are also smaller than those in the Quercy region. The procurement of raw materials is different as well, concerning both flint, which is smaller, and the use of stones other than flint. In the scenario in which the same groups would have occupied the Pyrenees and Quercy region, this constraint would have had little influence on the tool kit: the tools and reduction strategies could have been similar. However, the transformation of the products is different.

The backed points are indeed smaller, but they also have some speciic features, especially in the fabrication of a second back. The microlithism of endscrapers also provides information on these Pyrenean particularities: in this case, the morpho-dimensional reduction is clearly intentional. It is probable that these endscrapers were hafted and that their size was thus dictated by other modalities, such as those related to their attachment to a handle. The small dimensions of these tools in the Pyrenees thus indicates that there were distinct practices in these two regions.

These differences in the tools are associated with differences in the manufacturing strategies, thus confirming a regionalization of this culture. The Pyrenean strategies are not found in sites north of the Aquitaine, where small tools are also much less abundant. Blade manufacturing is relatively infrequent in the Pyrenees, but remains preponderant in Quercy. Even if the constraints of the manufacturing strategies are low in both assemblages, this difference in the size of the products is important to our understanding of the variability of Pyrenean strategies. Bladelet manufacturing requires little standardization in core preparation, in contrast to blade manufacturing, which, even when it is simple, must follow more precise technical rules. This emancipation from the constraints of blade manufacturing thus distinguished the practices in Pyrenees from those in Quercy, and could also explain the originality of the use of anvil percussion in the Pyrenees.

Azilian painted pebbles from the cave of Le Mas d'Azil

Azilian painted pebbles from the cave of Le Mas d’Azil

From the paper titled “Changes in ecosystems, climate and societies in the Jura Mountains between 40 and 8 ka cal BP“, by Christophe Cupillard, 2014, we read among other things:

Early Azilian and Middle Azilian: from 14400 to 13000 cal BP

The first manifestations of Azilian culture, with its backed-curved points, were easily recognised at the open air sites of Champréveyres and Monruz along the Neuchâtel lake where they are dated between 14400 and 14000 cal BP, at the end of the Bölling, GI-1d. Between 14400 and 13800 cal BP, well-dated occupations are scarce and lithic techno-assemblages are not typical and could be attributed to early Azilian. The sites in question are Thaygen (Kesslerloch, layer IIc pro parte), Mont-la-Ville (Mollendruz, layer C5 inf. pro parte), Buttenloch layer B, Meussia layer 3, and Cuiseaux (La Balme, Layer B). During the Alleröd, from the end of GI-1c to GI-1a, corresponding to the Middle Azilian, there is apparently a numerical increase of sites characterized by larger geographical dispersal.

As the well-dated sites are located in the northwestern part of the Jura. These sites are located below 550m a.s.l., except Le Bichon (846m). The 22 available radiocarbon dates from Alleröd belong mainly to the first part of this period, the end of GI-1 and GI-1b. Palynological studies show that all these sites were located below the treeline, in an open pine-birch forest. Rochedane‘s layers C’1 and B show intensive occupations which provided a rich fauna dominated by red deer, an abundant lithic industry, engraved and painted portable art and scattered, well-dated human bones. In contrast, the other sites including Oberlag Layer S, Rigney 2, and Bretonvillers have smaller lithic and bone series and seem to be short-term occupations. The cave of Bichon is a very special case of a bear hunting accident. On the other hand, the study of lithic raw material shows the importance of regional flints, which indicate that the Jura range was no longer a barrier to exchanges.

Late Azilian and Late Epigravettian: from 13000 to 11700 cal BP

For the Younger Dryas, archaeological data are poor and our precise knowledge is based only on four well-dated sites. These occupations in rock shelters are the layer R of Oberlag, the layer A4 of Rochedane, the layer D2 of Gigot at Bretonvillers and the layer III of Longevelle. In addition, apart of layer II of Rislisberghöhle could be included in this small group of sites. The lithic techno-assemblages from sites of Oberlag, Rochedanne, Bretonvillers, and Longevelle have been recently revised. Layers A4 of Rochedanne and R of Oberlag may be attributed to the Late Epigravettian, while layers D2 of Bretonvillers and III of Longevelle may be considered as belonging to the Late Azilian or to the Late Epigravettian.

At Rochedane and Oberlag, the hunted fauna has been well studied: it is temperate and dominated by red deer. Stone industries are made from local resources and the abundance of small geometric points at Rochedane reflects intensive hunting practices with the presence of engraved and painted portable art, where as the three other sites are less documented. These very late Final Palaeolithic occupations are all dated from the first part of GS-1 between 12900 cal BP and 12661 cal BP.

Tablet 741 with bifacial ornamentation side A) complete horse; side B) a special composition of two horses complete horses in axial

Tablet 741 with bifacial ornamentation side A) complete horse; side B) a special composition of two horses complete horses in axial

From the paper titled “Environmental and cultural changes across the Pleistocene-Holocene transition in Cantabrian Spain“, by Lawrence Guy Straus, 2016, we read:

Abstract A review of the cultural evidence from northern coastal Atlantic Spain (a.k.a., Vasco-Cantabria) spanning the late Last Glacial and early Postglacial (from Greenland Interstadial 1 to the mid-Holocene) reveals that some changes may have been related to major climate/environmental changes, while others may be attributed to demographic factors that caused possible resource overexploitation and to historical factors such as the long-term availability of Neolithic domesticates and technology in adjacent regions. The culmination of the warming trend of the Last Glacial Interstadial in the Allerød seems to have been of particular importance in the transition from the classic Upper Magdalenian (with its rupestral and portable art and complex stone and bone technologies) to the Azilian, despite continuity in the main game species and in the process of subsistence intensification. The Younger Dryas, on the other hand, seems to have had little immediate direct repercussion in this region, as the Azilian continued, straddling the Pleistocene-Holocene boundary. On the other hand, the climatically non-dramatic Preboreal-Boreal boundary seems to have seen the abrupt, marked break between the “Epimagdalenian” Azilian and the Asturian coastal shell midden Mesolithic in the western sector of the region. This contrasted with greater technological continuity (albeit with similarities to the Sauveterrian tradition in adjacent SW France) in the Mesolithic of the Basque Country, with no archeological indications that the 8.2 cal kya event had important consequences in this region. Then, some 15 centuries later, came the sudden, but centuries-delayed appearance of Neolithic domesticates and ceramics on the Atlantic side of the Cantabrian Cordillera originating from sources in the Mediterranean environments of the upper Ebro basin and/or southern France. This major lifeway change was possibly finally accepted, within a still mixed economy, in the face of the overexploitation of wild food resources. The “neolithization” of Vasco-Cantabria was finally underway by c. 6.6 cal kya, quickly leading to new human-land relationships characterized by mainly ovicaprine pastoralism, apparently limited cereal agriculture, continued foraging, recolonization of the montane interior and the construction of modest megalithic monuments.

Azilian engraving

Azilian engraving

From the paper titled “Divergence in the evolution of Paleolithic symbolic and technological systems: The shining bull and engraved tablets of Rocher de l’Impératrice“, by Nicolas Naudinot et al., 2017, we read among other things:

The Azilian is a culture of the European Upper Paleolithic. It appears at the end of the Magdalenian around 14,000 years ago and precedes the first Mesolithic hunter-gatherers in the early Holocene. This period is critical to the study of cultural evolution as it is characterized by a major restructuring of hunter-gatherer societies in terms of technology (a decrease in stone and bone tool standardization and a simplification of manufacturing processes), settlement (with changes in mobility patterns), and art, with the development of a unique abstract graphic production that contrasts with the previous Paleolithic but also the final Paleolithic “Laborian” figurative iconography. The role that rapid warming and climatic instability of the Azilian period played in shaping techno-economic and symbolic spheres is debated among scholars.

The Azilian has long been considered an abrupt event, a “revolution” in the Upper Paleolithic. Recent work shows that the development of this culture was not sudden but progressive. This global process of cultural change began during the GI-1e/d (end of the Bølling) with the Early Azilian (EA). However, the Azilian probably finds its roots at the end of the Magdalenian during the GS-2b-a (Oldest Dryas). The paroxysm of this process is the so-called Late Azilian (LA), probably lasting to the first half of GS-1 (Younger Dryas).

A new EA rock shelter in the western extremity of Brittany provides critical data to investigate the tempo of technological and symbolic change during the Azilian. The association of a lithic industry with a rich artistic assemblage of 45 engraved (and sometimes charcoaled) schist stones suggests a clear arrhythmia between symbolic production and technological adaptations. Here the possible techno-economic adaptations to climatic changes appear to have had no direct influence on the symbolic and perhaps spiritual universe of the first “Azilian” people who perpetuated an age-old tradition.



a: Flat-section Azilian harpoon; b: perforated harpoon base; c-d: decorated harpoons

From the paper titled “The earliest double dog deposit in the Palaeolithic record: The case of the Azilian level of Grotte‐abri du Moulin (Troubat, France)“, by Myriam Boudadi‐Maligne et al., 2020, we read:

Abstract It is now largely accepted that the wolf was the first animal to be domesticated during Pleistocene time. Although the exact timing of this event is still the subject of considerable debate, it is generally agreed upon that dogs lived side‐by‐side with humans for at least the last 15,000 years. Recent discoveries from the “Grotte‐abri du Moulin” (Troubat, France) provide crucial new information as to our understanding of the relationships between humans and dogs during prehistory. The site produced a deep stratigraphic sequence demonstrating that the cave was occupied from the Middle Magdalenian (ca. 17.5 ky cal BP) to the Sauveterrian period (ca. 9.5 ky cal BP). Faunal remains from layer 6, attributed to the Azilian culture, included a small Canis. The biometric study of these remains is fully consistent with their attribution to the dog (Canis familiaris). A total of 147 dog remains were isolated and represent a minimum of two individuals. These remains were recovered from a restricted space, and several skeletal elements remained in anatomical connection or in loose connection. The most labile elements were only slightly dispersed, and there is no evidence of human or carnivore modification on these bones. Direct radiocarbon dates obtained from a complete tibia demonstrate the dog remains to be contemporaneous with the Azilian occupation of the cave. The Grotte‐abri du Moulin evidence represents, therefore, the earliest known occurrence of an intentional double dog deposit and evidence for a particularly close relationship between Azilian groups and their dogs.


Harpoon – Mas d’Azil

Research-Selection for NovoScriptorium: Philaretus Homerides

One thought on “The Azilian Epipaleolithic Industry of Spain and France

Add yours

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

Up ↑

%d bloggers like this: