Introduction
In recent years, a growing number of accurate dates for Palaeolithic rock art has been obtained by the application of AMS radiocarbon measurements on charcoal used in the production of art (Clottes et al., Reference Clottes, Valladas, Cachier and Arnold1992; Valladas et al., Reference Valladas, Cachier, Maurice, Bernaldo de Quirós and Clottes1992; Clottes, Reference Clottes1994; Ochoa, Reference Ochoa2011), and by uranium-thorium measurements on calcium carbonate flowstones associated with it (Aubert et al., Reference Aubert, Brumm, Ramli, Sutikna, Saptomo and Hakim2014; Pike et al., Reference Pike, Hoffmann, García-Diez, Pettitt, Alcolea and de Balbín2012, Reference Pike, Hoffmann, Pettitt, García-Diez and Zilhão2017). These new techniques provide absolute dates in the case of the former, assuming an association between the production of the charcoal (which the date reflects) and the use of it to create art, and minimum or maximum ages for the art when dated flowstones are stratigraphically related to the art. They complement the existing relative techniques, such as examining the theme and style, which still provide a useful framework (Lorblanchet & Bahn, Reference Lorblanchet and Bahn1993; Alcolea & Balbín-Behrmann, Reference Alcolea and Balbín-Behrmann2007). Together, these constitute significant advances in our understanding of the timing, nature, and early development of art, a phenomenon which emerged during the Palaeolithic. However, all these techniques have their limitations, reflected in the controversies over perceived discrepancies between ‘traditional’ and chronometric dates, and over the reliability, pertinence, or interpretation of chronometric measurements (Fortea, Reference Fortea2000; Pettitt & Pike, Reference Pettitt and Pike2007; Pettitt & Zilhão, Reference Pettitt and Zilhão2015). It goes without saying that holistic approaches should be followed when evaluating any chronological evidence relevant to the subject.
The situation may, at first sight at least, seem more straightforward in the case of portable art, especially when it is recovered from reliable stratigraphic contexts with distinct archaeological associations, and for which accurate chronologies have been established. If one can do this with confidence, chronologically constrained sequences of art objects, such as the long sequence of the mid to late Upper Palaeolithic engraved and painted plaquettes from the cave of El Parpalló in Spain (Villaverde, Reference Villaverde1994), can be used to reconstruct thematic/stylistic frameworks. Here, we refer to previous debates on the efficacy of using such artefacts with reference to the artistic development in the European early to middle Upper Palaeolithic.
The western European Gravettian (c. 34–33 ka cal bp to c. 24–23 ka cal bp) coincided with a markedly cold climate which became increasingly humid towards the end. It is well understood typologically; its lithic assemblages, for example, typically contain abundant points and micro-points named after the eponymous site of La Gravette, complemented to various degrees by other type fossils that for some time have been rightly used to define specific phases, such as an early phase with Noailles burins, and a second phase with Isturitz points and backed elements (Otte, Reference Otte2013). In recent years, however, an even greater diversity in the Gravettian lithic toolkit has been recognized for northern Spain, to the extent that this now hinders a techno-typological classification in which a diachronic development can be unequivocally observed (Rasilla & Straus, Reference Rasilla, Straus and Fano2004; Arrizabalaga & De la Peña, Reference Arrizabalaga, de la Peña, Heras, Lasheras, Arrizabalaga and Rasilla2012; Bradtmöller, Reference Bradtmöller2015). Despite this, shared lithic forms and raw material movement patterns nevertheless reveal contact between northern Spain and south-western France (Tarriño & Elorrieta, Reference Tarriño, Elorrieta, Heras, Lasheras, Arrizabalaga and Rasilla2012), justifying regional research perspectives that typically combine the two.
In Europe, north of the Pyrenees, Gravettian portable art includes the so-called ‘Gravettian Venus’ figurines, examples of this broad cannon being recognized in regional pockets from south-western Siberia to the Pyrenees (Delporte, Reference Delporte1979; Duhard, Reference Duhard1993). So far, however, no exemplars have been found in northern Spain. Aside from these, figurative representations on bones or stone plaquettes are rare in the period. Non-figurative images are, by contrast, relatively common. These were produced on bone tools and less frequently on stone objects, and generally consist of series of lines or simple ‘signs’ that are sometimes repeated (Abaunza, Reference Abaunza2015). Objects of personal adornment made on shells and red deer relict canines are also common (Álvarez-Fernández, Reference Álvarez-Fernández2005, Reference Álvarez-Fernández2007, Reference Álvarez-Fernández2010; Álvarez-Fernández & Jöris, Reference Álvarez-Fernández and Jöris2007).
The lack of a clear framework for the development of Gravettian figurative art stands in sharp contrast to our understanding of its technology, which is regrettable given its importance for understanding the earliest development of figurative art. Here, we shall try to begin the process of establishing a basic framework by assessing the quality of the stratigraphic, chrono-cultural, and chronometric data currently available for the few decorated Gravettian objects from the north of the Iberian Peninsula.
Materials
In northern Spain, three portable objects with engraved decoration can be securely attributed to the Gravettian: they come from the caves of El Castillo and Cueva Morín in Cantabria, and from Antoliñako koba in the Basque Country (Figures 1 and 2).
Figure 1. Location of sites with Gravettian portable art in northern Spain: El Castillo (1), Cueva Morín (2), and Antoliñako koba (3).
Figure 2. Gravettian portable art: A) Cueva Morín (Barandiarán, Reference Barandiarán1973); B) El Castillo (Barandiarán, Reference Barandiarán1973); C) Antoliñako koba (Aguirre & González-Sainz, Reference Aguirre and González-Sainz2011: 53, Fig. 9). Reproduced with permission of the original authors and publishers.
El Castillo (Puente Viesgo, Cantabria)
The cave of El Castillo was discovered in 1903, and was excavated by Obermaier between 1910 and 1914 (Cabrera, Reference Cabrera1984; Breuil & Obermaier, Reference Breuil and Obermaier1914; Obermaier, Reference Obermaier1916; Cabrera, Reference Cabrera1984), by Cabrera Valdés and Bernaldo de Quirós between 1980 and 2004 (Cabrera et al., Reference Cabrera, Maíllo, Lloret and Bernaldo de Quiros2001, Reference Cabrera, Bernaldo de Quirós, Maíllo, Pike-Tay and Garralda2005), and by Bernaldo de Quirós since then, although these latter have yet to be fully published. Obermaier excavated strata with a total depth of 22 metres, divided into twenty-six stratigraphic levels dating from the upper Acheulian—later reassigned to the end of the Middle Pleistocene (Álvarez-Alonso, Reference Álvarez-Alonso2014)—to the Chalcolithic (Cabrera, Reference Cabrera1984). The decorated object under discussion here (Figure 2B) was discovered during the excavation of the cave entrance (Cabrera, Reference Cabrera1984: 274), in the Aurignacian α or upper Aurignacian level according to Obermaier's phasing (Reference Obermaier1925: 177), or Cabrera's (Reference Cabrera1984) Gravettian Level 12. It is a quartzite retoucher, with traces of use on the right edge of its upper face.
Almost the entirety of this face is covered by an engraving of an elongated zoomorph (Figure 2B). There are two possible readings of this image, depending on how one interprets the lines, and hence size and shape, of the head. In the first reading, a relatively large head with a large and open mouth, a long, receding forehead, and a small protrusion in the proximal area of the frontal line, which might represent an ear, may be inferred. From there, both dorsal and cervical lines are sinuous, conveying the impression of markedly convex withers (shoulder hump), a concave mid-dorsal region and a rounded rump, the vertical portion of which may correspond to a short tail (if so, slightly forked at its end). A sinuous ventral line essentially mirrors that of the dorsal line, forming a continuous line with the front of a single rear limb. The outline of the forelimbs is drawn with parallel lines, incorrectly drawn as connected to the chest. Overall, the perspective in this reading is in absolute profile. In the second reading, the head is considerably smaller, includes a round ear, behind which the sinuous cervico-dorsal line starts. The rear legs are not represented but the belly and the forelegs are; one of the forelegs extends forwards (the jaw in the first reading), while the other, represented by several parallel lines, is oriented vertically.
The specific reading of this image affects its taxonomic identification. Hence it has traditionally been regarded as a taxonomically indeterminate zoomorph, although some have suggested that, at least according to the second reading, it could represent a feline (Barandiarán, Reference Barandiarán1973: 106). This is certainly possible. In terms of the first reading, however, the sinuosity of its cervical-dorsal line with its marked volume at the hump, the overall elongation of the body, and the head with a wide open mouth are suggestive of a bison, similar to the bisons depicted in sienna pigment in the ‘Panel de las Manos’ of El Castillo (García-Diez & Eguizabal, Reference García-Diez and Eguizabal2008: 73–74). The figure discussed here is a simple outline engraved with fine and narrow lines.
A reappraisal of the lithic assemblages of El Castillo (Cabrera, Reference Cabrera1984: 396) characterized Level 12 as ‘Upper Perigordian with no further attributions’ (like Bernaldo de Quirós’ (Reference Bernaldo de Quirós1982) description), although two pieces with Solutrean retouch were additionally identified (Cabrera, Reference Cabrera1984: 254, 396). The most recent re-examination (Bernaldo de Quirós et al., Reference Bernaldo de Quirós, Maíllo-Fernández, Castaños and Neira2015) did not conclude that the pieces identified by Cabrera are in fact Solutrean, as one of them is now understood to be the striking platform of a core bearing simple retouching, flattened at its distal end, and the other an endscraper on a retouched blade, neither of which are demonstrably Solutrean. They are, therefore, consistent with the fact that Level 12 is separated from the Solutrean Level 10 by Level 11, an archaeologically sterile yellowish silt at least 50 cm thick (Cabrera, Reference Cabrera1984: 103, 396).
The most recent reappraisal of the archaeological assemblages recovered in Level 12 (Bernaldo de Quirós et al., Reference Bernaldo de Quirós, Maíllo-Fernández, Castaños and Neira2015) does not, therefore, differ from the conclusions of Cabrera (Reference Cabrera1984). Burins are slightly less numerous than endscrapers, and Obermaier noted the presence of a Font Robert point (now missing) (Cabrera, Reference Cabrera1984: 278). As with the other levels in the cave, diagnostic pieces such as this are the most frequent forms (37 per cent). The ‘splintered pieces’ sensu lato are equally of interest because of their phylo-cultural implications: these artefacts appear in Gravettian assemblages of the western Cantabrian region and the Pyrenees and are starting to be considered characteristic of this chronology (Arrizabalaga & de la Peña, Reference Arrizabalaga, de la Peña, Heras, Lasheras, Arrizabalaga and Rasilla2012). Technologically, blades dominate the lithic production, particularly bipolar prismatic methods to obtain blades and bladelets in a continuum between each type of blank.
Recent dates for Level 12, from samples from Obermaier's excavations, curated in New York's Natural History Museum, and others obtained during the cleaning of the section, are internally consistent, placing the assemblage in the middle/second half of the Gravettian (Bernaldo de Quirós et al., Reference Bernaldo de Quirós, Maíllo-Fernández, Castaños and Neira2015): the dates obtained are 24,070 ± 150 bp (Beta-242617: 28,488–27,783 cal bp); 25,520 ± 140 bp (Beta-298431: 30,178–29,240 cal bp); and 25,920 ± 140 bp (Beta-298430: 30,643–29,684 cal bp), calibrated with INTCAL 13 (Reimer et al., Reference Reimer, Bard, Bayliss, Beck, Blackwell and Bronk Ramsey2013) using OxCal 4.2 (Bronk-Ramsey, Reference Bronk-Ramsey2009). We may, therefore, confidently take Level 12, and hence the context of the engraved retoucher, to be an unmixed Gravettian (mid-Upper Perigordian) assemblage.
Cueva Morín (Villaescusa, Cantabria)
Cueva Morín was brought to the attention of the scientific community in letters by Obermaier and Wernert in 1910 (Carballo, Reference Carballo1923). The first excavations in the cave were carried out by Carballo from 1917 to 1919 (Carballo, Reference Carballo1923) and Vega del Sella in 1918 (Vega del Sella, Reference de Vega del Sella1921). Subsequent excavations from 1966 to 1969 were directed by González-Echegaray and Freeman, who identified twenty-two levels in the stratigraphic sequence, containing material from the Mousterian to the Azilian (González-Echegaray & Freeman, Reference González-Echegaray and Freeman1971, Reference González-Echegaray and Freeman1973). Small-scale excavations were later conducted by Maíllo-Fernández and González-Echegaray in 2005 (Maíllo-Fernández et al., Reference Maíllo-Fernández, González Echegaray, Arteaga, Iriarte-Chiapuso, Fernández, Bernaldo de Quirós and Palomera2016), and by González-Urquijo and Weniger in 2008, although this latter is not yet published.
The object that concerns us here was recovered from Level 4 in square IXA during the excavations of González-Echegaray & Freeman (Reference González-Echegaray and Freeman1971: 244, fig. 123). It is a shale retoucher (Figure 2A), broken at both ends. Its entire upper face bears an engraving of what is usually interpreted as an anthropomorph. The presumed head of this figure, shown uppermost in Figure 2A, appears to look towards the right, assuming that a small, incomplete circle represents the eye, and a convex protuberance represents the nose; additionally, a single engraved line intersecting the line of the head may represent hair. The figure's trunk is elongated relative to its head, and gradually widens downwards from a relatively thin chest towards the middle (belly) of the figure, from which point at least one lower limb is represented as a prolongation of the trunk. A second limb may be shown by one or two short parallel lines at the extreme right of the image. The upper limbs appear to be represented by highly truncated and sharply flexed appendices. Two converging lines central to the uppermost part of the legs may form an incipient representation of the genitals. In any case, few details have been added. In this reading, the perspective of the drawing is ‘straight biangular’; lines and connections between anatomical regions are rigid, and overall it is a simple outline made of fine and narrow lines.
The lithic assemblage from the Gravettian Level 4 of Cueva Morín has been studied on several occasions (González-Echegaray & Freeman, Reference González-Echegaray and Freeman1971; Arrizabalaga, Reference Arrizabalaga1995; Bradtmöller, Reference Bradtmöller2015). All studies place it, from a techno-typological point of view, in a second phase of the Gravettian, characterized by steep retouching and backed pieces. The assemblage is dominated by burins, projectiles, and backed elements. Endscrapers appear in low percentages, even lower if one excludes their carinated or thick forms (Bradtmöller, Reference Bradtmöller2015). Technologically, the main objective of lithic reduction was the production of blades and bladelets, particularly using uni/bipolar prismatic methods for blades and the pyramidal and carinated endscraper/burin-type production for bladelets.
The dating of the assemblage from the 1966–69 excavations in Cueva Morín was inconclusive (Bernaldo de Quirós, Reference Bernaldo de Quirós1982; Soto, Reference Soto2003: 224–31), although all studies agree that the assemblage can be ascribed to a relatively late stage of the Gravettian from a technological perspective, i.e. to the final or evolved Gravettian ‘in a very late stage’ (González-Echegaray & Freeman, Reference González-Echegaray and Freeman1973: 198, 296), or the ‘upper Perigordian’ (Bernaldo de Quirós, Reference Bernaldo de Quirós1982), and the most recent phase of the Gravettian (Bradtmöller, Reference Bradtmöller2015). Two samples from the 1966–69 excavations have recently been dated: one from the metre square from which the decorated retoucher was recovered, and the other from the adjacent square (Bradtmöller, Reference Bradtmöller2015). The results are statistically of the same age: 23,640 ± 190 bp (Poz-66758: 28,111–27,445 cal bp) and 23,790 ± 190 bp (Poz-66759: 28,307–27,563 cal bp). The internal consistency of these measurements, one of which dates a sample directly associated with the engraved retoucher, together with the absence of any anomalous elements in the lithic assemblage, allows us to assign the object to the middle/second half of the Gravettian with a good degree of confidence.
Antoliñako koba (Gautegiz-Arteaga, Basque Country)
Antoliñako koba was excavated between 1995 and 2008. Its stratigraphic sequence consists of a series of Upper Palaeolithic levels (Aurignacian, evolved Aurignacian, Gravettian, Upper Solutrean, late Lower Magdalenian, Upper Magdalenian, Azilian and ancient Epipaleolithic: Aguirre, Reference Aguirre, Heras, Lasheras, Arrizabalaga and Rasilla2012; Rofes et al., Reference Rofes, García-Ibaibarriaga, Aguirre, Martínez-García, Ortega and Zuluaga2015). The portable art object discussed here was excavated from square 5E of the cave's ‘North Passage’ (Aguirre, Reference Aguirre, Heras, Lasheras, Arrizabalaga and Rasilla2012) at a depth of between 1.65 and 1.70 m (Aguirre & González-Sainz, Reference Aguirre and González-Sainz2011: 46, fig. 3) in Level Sabk, which corresponds to Levels Smbk in square 3H and Lmbk sup. of the cave's entrance hall and ‘West Chamber’. The object is a sandstone hammerstone, bearing traces of damage or removals to one side that have partly damaged its engraved face (Aguirre & González-Sainz, Reference Aguirre and González-Sainz2011). Despite this, the engraved motif seems to have been minimally affected (Figure 2C). The partial figure of a red deer hind has been engraved with a few long, confident lines, consisting of a head drawn with straight frontal and maxillary lines that converge in the area of the mouth, the lower line (the jaw) intersecting with the neck line; a neck and chest that shortens the line of the jaw and consists of a very slight distinction between a convex neck and a concave chest (probably to indicate the throat), and a modestly concave dorsal line. Two straight and more or less parallel ears are represented, each as continuations of the forehead and dorsal lines. Overall, the representational perspective is straight biangular or oblique. The lines and connections between anatomical regions combine a rigid and fluid character, while the engraved outline is simple. Nevertheless, its taxonomic identification is clear.
No doubts exist about the integrity of the Gravettian level at Antoliñako koba, from which samples dated by AMS radiocarbon provide a robust chronology. The proximity to the decorated object of some of the dated samples, obtained from the adjacent square and at a similar depth, arguably makes this object the best-defined chronological marker for Gravettian portable art in the region. The lithic assemblage of Level Lmbk sup./Sabk is dominated by denticulates, followed by burins, backed elements, and sidescrapers. The dates obtained from objects nearest to the engraved hammerstone derive from square 3H in Level Smbk (which corresponds to Sabk and Lmbk sup.) at a depth of approximately 1.70 m and found together with Noailles-type burins: 26,720 ± 180 bp (Beta-251299: 31,156–30,650 cal bp) and 27,100 ± 190 bp (Beta-251300: 31,347–30,853 cal bp). The results are statistically the same and allow us to attribute the object to the early Gravettian with confidence.
Stylistic Characterization of the Objects and its Implications
The three art objects presented here have distinct excavation histories: the example from El Castillo comes from an old excavation of the early twentieth century, the item from Cueva Morín was found in the late 1960s, and the exemplar from Antoliñako koba was excavated in the first decade of the twenty-first century. Nevertheless, these objects can be placed in stratigraphic order in the context of clear techno-typological associations. In the 1960s and 1970s, improvements in excavation procedures were gradually introduced across the Iberian Peninsula (González-Echegaray et al., Reference González-Echegaray, Cabrera and Bernaldo de Quirós1997; Estevez & Vila, Reference Estevez and Vilá1999), involving multidisciplinary methodologies, with particular attention being paid to sedimentary processes and resulting stratigraphies, including the micro-sedimentology of the so-called ‘occupation floors’. By contrast, the excavations of the early twentieth century should, a priori, be treated with a degree of caution. It is necessary to compare the validity and precision of the stratigraphic and cultural sequences that were put forward at the time with those proposed using new analytical procedures; this may reveal possible contradictions and we need to investigate the reasons for these. For our purposes, we can be confident of the chronological and cultural attribution of the three objects under discussion. The lack of any obvious disruptive sedimentary processes in the levels from which they derive, the apparently discrete (uncontaminated) nature of the lithic assemblages, and the consistency of the available AMS radiocarbon dates pertinent to the piece from Antoliñako koba support this and provide a reliable chronological framework for the interpretation of the art featuring on the objects.
With this in mind, we can establish that portable art was present in the Cantabrian region in the early to middle stages of the Gravettian (31,347–30,650 cal bp) with the Antoliñako koba piece; and in the succeeding middle stage of the Gravettian at El Castillo (between 30,643 and 27,783 cal bp) and Cueva Morín (28,307–27,445 cal bp) (Figure 3).
Figure 3. Chrono-cultural attribution of the three pieces of Gravettian portable art discussed. Modified from Barandiarán, Reference Barandiarán1973; Aguirre & González-Sainz, Reference Aguirre and González-Sainz2011: 53, Fig. 9.
We are aware that we are only dealing with three objects and our conclusions are inevitably preliminary. A parsimonious interpretation of the chronology would see these representing just two points in time; an early point at around 31–30 ka cal bp and a later point, when the objects from Cueva Morín and El Castillo overlap at around 30–27 ka cal bp. Alternatively, they may represent a more continuous age range from c. 30–28 ka cal bp. Relatively speaking, the piece from Antoliñako koba can be said to predate the other two examples, and the piece from Cueva Morín may be the most recent.
Several characteristics are shared by all three objects: they are a sketchy expression of the outline anatomy of the depictions concerned, which, except in a few instances (e.g. the eye of the Cueva Morín figure), is limited to the outline of the depictions. They lack any internal details, and focus only on representing the depictions’ basic form. Specifically:
• The outline is mainly represented by a simple line (El Castillo and Cueva Morín), i.e. a figure in profile.
• The form of these outlines varies, however, from fluid and curving (Antoliñako koba) and extremely fluid (El Castillo) to sharp and angular (Cueva Morín), even on the same image (Antoliñako koba).
• The articulation between anatomical regions can also be fluid and curving (Antoliñako koba) or angular (El Castillo and Cueva Morín), again even on the same image (Antoliñako koba).
• A fluid outline is used to distinguish certain anatomical parts, in the area of the head (Antoliñako koba), throat, nose, and hair (Cueva Morín), and possibly ear (El Castillo).
• A single line is used to represent the whole of the upper or lower half of the torso, e.g. the neck-back-rump, and back leg-chest on the El Castillo object.
• Limbs are placed in anatomically incorrect positions, e.g. at El Castillo, where the forelimbs are placed ahead of their natural position.
• A lack of internal coherence in the size of the anatomical parts creates a disproportion between the parts of the body in comparison to natural (real) proportions. Examples include a long and narrow trunk (or part of it) linked to the figure's longitudinal configuration (El Castillo and Cueva Morín; possibly Antoliñako koba); an exaggerated height of the head (El Castillo) and hind quarters (El Castillo), and relatively short limbs (El Castillo and Cueva Morín).
• The perspective includes absolute profile (El Castillo), straight biangular (Cueva Morín), and straight biangular or oblique (Antoliñako koba).
• The subject may be represented complete or incomplete.
• There is no representation of movement; the limbs are depicted vertical and rigid in articulation.
• No dynamic attitude is shown; instead the animals are represented statically. It is however conceivable that a) the position of the forelimbs of the zoomorph from El Castillo indicates movement; and b) the flexed and raised upper limbs of the anthropomorph from Cueva Morín correspond to an otherwise unidentified attitude.
Discussion
Three objects of portable art is a remarkably low number, given the wealth of excavated assemblages of Gravettian attribution in the Cantabrian region, and the abundance of parietal art in the same region. How can we explain this? In the regions adjacent to northern Spain, portable art objects of Gravettian age with figurative decoration have been documented in the Pyrenees, that is to the east of the Cantabrian region. Here, bone objects bearing non-figurative decoration are relatively common (San Juan-Foucher, Reference San Juan-Foucher, Heras, Lasheras, Arrizabalaga and Rasilla2012), although figurative representations are scarce in comparison with later phases of the Upper Palaeolithic (Barandiarán, Reference Barandiarán1973, Reference Barandiarán1994; Corchón, Reference Corchón1986). Some of the Pyrenean sites that have yielded figurative portable art were excavated some time ago, and they therefore require the same caution over the art's chronological and cultural contexts.
In terms of figurative portable art, the Pyrenean region's female figurines are well known, although it is difficult to establish their specific cultural provenance with confidence, given that they were discovered in old excavations or as fortuitous (unexcavated) finds. In the case of the ‘Venus’ of Lespugue, it was noted from the start that the relevant stratigraphic levels had been disturbed by subsequent historical occupations in the cave's entrance (San Juan-Foucher, Reference San Juan-Foucher, Heras, Lasheras, Arrizabalaga and Rasilla2012). Although de Saint-Périer attributed the whole deposit to the Gravettian, new, ongoing studies suggest that the artefacts recovered from the cave's Palaeolithic levels in fact belong to several Upper Palaeolithic cultural groups (San Juan-Foucher, Reference San Juan-Foucher, Heras, Lasheras, Arrizabalaga and Rasilla2012). The same can be observed for the Grotte du Pape at Brassempouy, from which ten fragmentary ‘Venuses’, including the famous hooded Venus, are, according to Schwab, ‘[…] poorly dated; owing to the somewhat expeditious excavation methods of the time, they are attributed to the Gravettian on stylistically analogical grounds ([…] mal datées, en raison des méthodes de fouille quelque peu expéditives de l’époque, elles sont attribuées par analogie stylistique au Gravettien)’ (Schwab, Reference Schwab2008: 94).
Engraved stone plaquettes recovered from Gargas in the early twentieth century have been recently studied afresh (San Juan-Foucher, Reference San Juan-Foucher, Heras, Lasheras, Arrizabalaga and Rasilla2012; Foucher & San Juan-Foucher, Reference Foucher and San Juan-Foucher2016). Two of these, with zoomorphic engravings (horse, bovid, and felid), were excavated from Level 6, in a ‘Perigordian’ context associated with a hearth and a large number of Noailles-type burins (Breuil, Reference Breuil1953: 44). This level has been identified in recent excavations in sector GPA in hall 1 (Salle 1; Foucher & San Juan-Foucher, Reference Foucher and San Juan-Foucher2016), from where numerous Noailles-type burins have also been retrieved. The first plaquette bears an engraved bovid (probably a bison), on top of which an animal (probably a felid) was engraved, and, opposite this, the hind quarters of a third animal (probably a horse). A second plaquette shows the engraved front quarters of a horse. These representations share at least some of the traits recognized in the three images from northern Spain discussed above: sketchy outlines, disproportionate anatomical representation, and incoherence in the position of anatomical parts and perspective.
The figurative portable art from Isturitz Cave, comprising twenty-three plaquettes, was recovered during the excavations of Passemard and de Saint-Périer (Passemard, Reference Passemard1944; de Saint-Périer & de Saint-Périer, Reference de Saint-Périer and de Saint-Périer1952) and these have also been recently re-evaluated (Rivero & Garate, Reference Rivero and Garate2014). Nine plaquettes come from Level III/C; eight and a half from Level IV (the other half of a broken example was retrieved from Level IIIa); one and a half from Level IIIa; and one from Levels II and F2inf (probably catalogued incorrectly). These levels have been attributed to the Aurignacian (Level SIII), Gravettian (Levels III/C, IV), Solutrean (Levels IIIa, F2), and middle Magdalenian (Level II) (Passemard, Reference Passemard1944; de Saint-Périer & de Saint-Périer, Reference de Saint-Périer and de Saint-Périer1952). Due to the relatively poor quality of the excavation methods employed by unskilled labourers, without paying careful attention to the stratigraphy and the sedimentary processes, researchers who have studied the lithic artefacts have subsequently reported several problematic areas (Esparza, Reference Esparza1995). Recent studies (Simonet, Reference Simonet2009; Goutas & Lacarrière, Reference Goutas, Lacarrière, Heras, Lasheras, Arrizabalaga and Rasilla2013) have revealed the mechanical disturbance of at least one of the cave's two Gravettian levels, i.e. III/C, which ‘attests to important contamination from higher (Solutrean and Magdalenian) layers (témoigne de contaminations importantes en provenance des couches sus-jacentes (solutréennes et magdaléniennes)’ (Goutas & Lacarrière, Reference Goutas, Lacarrière, Heras, Lasheras, Arrizabalaga and Rasilla2013: 568). Additionally, human remains from Level III/C have been directly dated by AMS radiocarbon to 14,640 ± 50 bp (17,995–17,648 cal bp), indicating a degree of intrusion of materials from the overlying late Upper Palaeolithic Level II (Henry-Gambier et al., Reference Henry-Gambier, Normand and Petillon2013). Because of these issues, objects attributed to Levels III/C, II, IIIa, SIII, and F2inf should not be included in any assessment of specifically Gravettian graphic activity as they are not stratigraphically reliable.
By contrast, fewer doubts exist over the objects assigned to Level IV at Isturitz, although some authors do not rule out the possibility that the Gravettian contamination in other levels derives from this Level IV (Simonet, Reference Simonet2009: 22): ‘[…] Although Gravettian elements belonging to Level IV may be found in other layers, the opposite is rarer. A quick examination of the lithic assemblage confirms that the contamination of Level IV is in keeping with a loss of part of the assemblage […] This would indicate that the lower Gravettian level is apparently very homogeneous ([…] Mais si des éléments gravettiens du niveau IV peuvent se retrouver dans d'autres couches, l'inverse est plus rare. Un rapide examen de l'assemblage lithique confirme que la contamination du niveau IV va dans le sens d'une perte d'une partie de la collection […]. Il en ressort que le niveau inférieur de Gravettien paraît très homogène).’ The formal and stylistic characteristics of the representations from this level, particularly the incompleteness of the figures (bovids, cervids, and indeterminate), share a similarity with the portable art of northern Spain and with Gargas (France). However, other elements (such as anatomical details and the fluid nature of the outline used to create anatomical parts) are not found in those places. A rib with a motif representing the front half of a horse and the hindquarters of a possible bovid was also found in Level IV at Isturitz (de Saint-Périer & de Saint-Périer, Reference de Saint-Périer and de Saint-Périer1952; San Juan-Foucher, Reference San Juan-Foucher, Heras, Lasheras, Arrizabalaga and Rasilla2012).
We can find similarities even further afield than northern Spain and the French Pyrenees, notably on a small number of objects from Mediterranean Spain, from the sites of Mallaetes and El Parpalló (Villaverde, Reference Villaverde1994; García-Diez & Ochoa, Reference García-Diez and Ochoa2012a, Reference García-Diez, Ochoa, Heras, Lasheras, Arrizabalaga and Rasilla2012b). Taken together, these confirm the existence of a graphic tradition spread over a wide area during the Gravettian. The absence of figurative portable art in northern Spain in the preceding Aurignacian is also significant (García-Diez & Arrizabalaga, Reference García-Diez and Arrizabalaga2000) and there is a general scarcity or absence of such representations until a late stage of the pre-Magdalenian (García-Diez & Ochoa, Reference García-Diez, Ochoa, Sázelová, Novák and Mizerová2015). New U-Th series minimum ages for Iberian cave art (Pike et al., Reference Pike, Hoffmann, García-Diez, Pettitt, Alcolea and de Balbín2012) have, however, opened a new debate over the dating of the emergence of figurative art in the region. These dates suggest that this may have taken place during the early Aurignacian or even earlier, based on ante quem and post quem dates for the anthropomorph of Tito Bustillo in Asturias. In northern Spain, many examples of parietal art have been traditionally associated with pre-Magdalenian periods, most frequently with the Gravettian. The lack of reliably associated chronometric dates at most of these sites requires us to be cautious, particularly since portable art objects are so scarce, not only in the Aurignacian and Gravettian, but also in the Solutrean. This cannot be ascribed to sampling bias. Numerous excavations of assemblages with good organic preservation belonging to these cultural phases have been conducted, but portable figurative art remains highly elusive.
As it stands, the scarcity of Gravettian figurative portable art in the region is an objective fact. Given this, we propose two hypotheses that will need to be tested by future chronometric dating of Palaeolithic art: a) figurative art, notably on portable objects, was genuinely extremely scarce in northern Spain during the Gravettian; and b) the greater or lesser presence of graphic activity corresponds to a duality in modes of expression in the region, that is portable art (which is very scarce) and parietal art (which is relatively more common, at least according to stylistic arguments that require verification through absolute dating). Perhaps the two may have been functionally complementary, meaning they possibly had different purposes during this period.
Conclusion
We propose here that, irrespective of their specific taxonomic identification, the study of the traits of three objects of portable art from the north of the Iberian Peninsula can be used to establish the first statement on the character of Gravettian art in the region. Our discussion, based on artefacts found in levels that have been reliably dated and characterized chrono-culturally according to their lithic assemblages, demonstrates the existence of figurative graphic activity in northern Spain at Antoliñako koba in the early stage of the Gravettian (31,347–30,650 cal bp). This activity continued in the immediately following stage at El Castillo (30,643–27,783 cal bp) and Cueva Morín (28,307–27,445 cal bp) (Figure 3). It took place either in two distinct cultural phases, or more continuously in a more chronologically constrained range.
The morpho-stylistic characterization of these examples suggests that the graphic conception of this time was based on the depiction of a simplified form that prioritized the animal's outline. This characteristic can be found more widely in the Pyrenees and in Mediterranean Spain; although significant differences can also be observed between the regions, notably the presence or absence of the so-called ‘Venuses’ (absent from northern Spain but present elsewhere, though rare). We suggest, therefore, that figurative portable art was very rare in Gravettian northern Spain. Perhaps its presence or importance was conditioned by the mode (parietal or portable) in which the message or information was preferentially transmitted at any one time. Hence we should not think simply in terms of the Gravettian being rich in art wherever it occurs, and that such art was both parietal and portable. We need to nuance our arguments about its dynamics. Our proposal can be tested with further discoveries in clear contexts and by dating these chronometrically. Advancing the study of both the Gravettian period and Upper Palaeolithic art—its development, dispersion, and the dynamics of the ideas contained within it—requires us to obtain (and publish) new data, especially about cave art.
Acknowledgements
This work was supported by the Prehistory Research Group of the University of the Basque Country (UPV/EHU) and funded by the Basque Government under grant IT-622-13, and by the MINECO project ‘La ruta occidental del poblamiento peninsular durante el Paleolítico medio y superior’ under grant HAR2014-53536-P. Blanca Ochoa's research was made possible thanks to a postdoctoral research grant from the Basque Government (Programa Posdoctoral de Perfeccionamiento de Personal Investigador Doctor). The authors would like to thank the three anonymous reviewers for their comments and the editorial team of European Journal of Archaeology, they helped to improve the original manuscript significantly.
