Pigments from the Middle Palaeolithic levels of Es-Skhul (Mount Carmel, Israel)
Introduction
Over the past decade, the systematic use of pigment has been repeatedly mentioned as one of the defining elements of ‘modern behaviour’ (Knight et al., 1995, Watts, 1999, Watts, 2002, Watts, 2009, Watts, in press; Barham, 1998, 2002; McBrearty and Brooks, 2000, d’Errico, 2003; d’Errico et al., 2003, d’Errico et al., 2009a, d’Errico et al., 2009b Henshilwood and Marean, 2003, Hovers et al., 2003, Henshilwood and Dubreuil, 2009) and a proxy for the origin of language (Knight, 2008, Watts, 2009, Watts, in press). Considerable debate, however, surrounds the reality and interpretation of the earliest instances of pigment use and their significance in the origin of high cognitive functions. This paper summarizes the evidence, contrasting interpretations, and presents results of the analysis of unpublished lumps of pigmental material from the early modern human site of Es-Skhul, Israel.
Many cases of pigment collection, curation and use are reported in the scientific literature at early and middle Pleistocene sites in Africa and Europe such as Kapthurin Tuff K4, Terra Amata, and Wonderwerk (see de Lumley, 1966, Leakey, 1958, 1971; Marshack, 1981, Wreschner, 1980, McBrearty, 2001, McBrearty and Tryon, 2006, Beaumont and Vogel, 2006, Cruz-Uribe et al., 2003). However, the supporting evidence has rarely been pictured in publications and analysed in detail. Partial exceptions are the potential pigment found at the base of Tuff K4, Kapthurin formation, Kenya (McBrearty, 2001), the Acheulean site of Beçov in the Czech Republic (Šajnerová-Dušková et al., in press), Twin Rivers in Zambia (Clark and Brown, 2001, Barham, 2002), Sai Island in Sudan (van Peer et al., 2003, van Peer et al., 2004), and Pinnacle Point in South Africa (Marean et al., 2007, Watts, in press).
In Africa pieces of red pigmental material, many bearing clear traces of modification by scraping, grinding, and notching, became a common feature at Middle Stone Age (MSA) sites younger than 180 kya and in particular at MSA 2b/Still Bay, Howiesons Poort and MSA 3 sites (Wurz, 2000, Watts, 1999, Watts, 2002; Henshilwood et al., 2009; Rigaud et al., 2006, Jacobs et al., 2006). Microscopic residues of red pigment have been systematically recorded on shell beads recovered from Middle Stone Age sites in North and South Africa (d’Errico et al., 2005, d’Errico et al., 2009a, d’Errico et al., 2009b; Bouzouggar et al., 2007) and the Middle East (Walter, 2003, Bar-Yosef Mayer et al., 2009). They are interpreted to be the result of deliberate colouring, or from having rubbed against ochred hide, skin, or thread. The presence of ochre on unperforated shells from Grotte des Pigeons, Taforalt is attributed to contact with pigment in a pouch or other container (Bouzouggar et al., 2007). Red pigment residues are also found on lithics from Still Bay, Howiesons Poort, and post Howiesons Poort contexts, and are interpreted as having been either components of adhesives for hafting (Gibson et al., 2004, Wadley et al., 2004, Lombard, 2005, Lombard, 2006, Lombard, 2007, Lombard, 2008) or as having resulted from the use of iron oxide nodules as knapping tools for thin Still Bay points (Soriano et al., 2009). Similar residues are also present on Blombos bone tools from layers dated to 75 kya (Henshilwood et al., 2001) and are thought to have been incorporated in the polish during tool use from either the users’ hands or ochre stained hide.
In the Middle East, 71 fragments of red pigment, some bearing traces of use and heating, come from the Mousterian levels of Qafzeh, Israel, and are dated to ca. 100 kya (Hovers et al., 2003: 494; Godfrey-Smith and Ilani, 2004; Bar-Yosef Mayer et al., 2009). This site has yielded burials attributed to Anatomically Modern Humans (AMH), but no pigments have been recorded in clear association with the skeletons.
In Europe, more than 40 Mousterian sites dated to MIS 6–3 have yielded objects described as blocks of pigment, or stones used to grind or crush pigment (see Bordes, 1952, Bordes, 1972; Demars, 1992; Soressi and d’Errico, 2007). The majority of pigment used by Neanderthals is black. Red pigment is rare but clearly present at some sites such as Cueva de los Aviones and Cueva Anton, in southern Spain, where its use is apparently associated with that of marine shells (Zilhão et al., 2010). Most of these sites date to the end of the Middle Palaeolithic, between 60 and 40 kya, and are attributed to the Mousterian of Acheulean tradition (MTA) or the Charentian Mousterian. At the MTA site of Pech de l’Azé I more than 500 blocks of pigmental material, half of which bear traces of utilisation, and a sandstone slab used to grind pigment, have been recovered (Soressi and d’Errico, 2007; Soressi et al., 2009). The last-known Neanderthals in France and Italy made intensive use of both black and red pigments (Salomon, 2009, Ronchitelli et al., 2009). A case in point is the 18 kg of red and black pigments, often bearing traces of use, found in the Châtelperronian layers of the Grotte du Renne, Arcy-sur-Cure, the largest quantity of pigmental material found so far at a Palaeolithic site. Finally, red pigments appear in Australia during the initial human colonisation of the continent, ca 50 kya (Taçon, 2004; but cf. O’Connell and Allen, 2004). Ochre is also associated with the oldest dated Australian burial, Mungo 3, from the Willandra Lakes site, and dated to ca. 40 kya (Bowler et al., 2003).
The symbolic value attributed by human cultures to pigments, particularly red ochre, is illustrated by many ethnographic examples (Boivin and Owoc, 2004). However, this observation does not demonstrate that red ochre was exclusively used for symbolically mediated actions in the ancient past, nor does it explain how its symbolic use arose. Specific frames of inference have been proposed to define a link between pigment use on the one hand and cultural modernity and language on the other. Berlin and Kay’s (1969) proposition that eleven basic colour words are common to all the languages of the world and that they arose in different languages in a regular sequence (black and white followed by red, green, yellow, blue, and brown) has been used to support the view that colour definitions emerged in the same order during the process that led to modern cognition (Hovers et al., 2003; but see Watts, in press). Henshilwood and Dubreuil (2009) contend that aesthetic and symbolic behaviours are dependent on the highest level of cognitive capacity. If such a capacity is present, as would be indicated by the ancient use of personal ornaments and pigments, they argue that it must be assumed a modern level of language was already in place.
The female cosmetic coalition model (Knight et al., 1995, Knight, 2008, Power and Aiello, 1997) posits that the use of red pigment for body and cloth painting, used to mimic menstrual blood, should be seen as one of the strategies created by females to build a more co-operative society in order to meet higher reproductive costs. Kuhn and Stiner, 2007a, Kuhn and Stiner, 2007b discount an exclusively functional use of pigments by pointing out that the available data supporting this view consist of ethnographic anecdotes, and that the available archaeological data for “practical” uses of mineral oxides in the Upper Palaeolithic indicate that they were employed in hide preparation and as an abrasive in drilling, both of which would also have potential applications to ornamentation. This suggests to them that even if a utilitarian function is attested, concomitant use of pigments for body decoration and as a system of communication becomes inescapable. A similar argument is put forward by Barham (2002) to infer language abilities for the users of pigments found at Twin River, Zambia.
Others (Wadley, 2001, Wadley, 2005a, Wadley, 2005b, Wadley, in press Wadley et al., 2004, Wadley et al., 2009; Klein, 1999) have countered the uniquely symbolic interpretation of early pigment use by testing the utility of ochre in functional tasks, showing that pigments also have utilitarian functions in ethnographically documented societies. They suggest that the presence of pigment residues on MSA artefacts does not necessarily imply symbolic activities.
To test these hypotheses, increased attention has focused on detecting consistencies in early pigment selection, modification, and use. The presence of pigment on MSA shell beads (d’Errico et al., 2005; Bouzouggar et al., 2007; d’Errico et al., 2009a, d’Errico et al., 2009b) and, at Blombos, the choice of ochre as a medium on which abstract designs were engraved (Henshilwood et al., 2009) all seem to contradict an exclusively functional interpretation of early pigment use, as it appears that pigments were directly involved in non-utilitarian behaviours. d’Errico et al. (2003) noticed that a number of MSA and Mousterian pieces of pigment was fully shaped by grinding to produce pointed ends, suggesting that these pieces, similar to objects found in Upper Palaeolithic and more recent contexts, were likely used as crayons or fusains to trace lines and produce designs on a variety of media. Watts, 1999, Watts, 2002, Watts, in press argues against a mere utilitarian function of pigments found at MSA sites by showing that there was a preference for the use of a strong red pigment, even when yellowish or yellowish-brown material was available. A similar argument is made by Hovers et al. (2003) for the pigments found at Qafzeh. Pieces of ochre from this site must have been used in symbolic activities because red ochre is less common than yellow ochre in the area and they appear to have been selected, mined, and used specifically for their hue. This suggested to them that the red ochre was mined far from the site and systematically looked for. Godfrey-Smith and Ilani (2004) reached a similar conclusion, but on different grounds. Their TL analysis of the quartz grains contained in three ochre fragments from Qafzeh revealed that two fragments were heated to a temperature between 300°C and 450°C, which may have reddened their original hue, if the fragments were mainly composed of goethite. This result is interpreted as evidence that Qafzeh AMHs practiced deliberate heat treatment in order to alter the colour of the ochre (Godfrey-Smith and Ilani, 2004: 189). However, recent experiments conducted by Wadley (in press) demonstrate that yellow ochre nodules buried 5 and 10 cm under an open fire in anoxic conditions can become red or shades of red. This indicates that ochre may have been heated accidentally, due to its close proximity to fire, and may explain the predominance of red pigments at sites where fires were repeatedly lit and maintained. Methods to establish whether red ochre pieces found at archaeological sites are composed of natural haematite or haematite produced by heating goethite are now available (Helwig, 1997, Pomiès et al., 1998, Pomiès et al., 1999a, Pomiès et al., 1999b; Chalmin et al., 2003, Chalmin et al., 2004, Chalmin et al., 2006, Chalmin et al., 2008; Salomon et al., 2008), but they have not been applied specifically to the analysis of the earliest pigments. Identification of haematite resulting from heated goethite has been traditionally based on XRD analysis. However, since heated and unheated haematite can produce similar XRD spectra, the identification of haematite produced by heating is now based on the inspection of iron crystals with TEM.
In this paper we apply these methods to analyze four previously undescribed fragments of pigmental material from the Mousterian layers of Es-Skhul Shelter, Israel, dated to between 100 and 135 kya and discuss the results in the context of the current debate surrounding the symbolic versus utilitarian role played by early pigments.
The shelter of Es-Skhul is located at Mount Carmel, 3 km south of Haifa, in the canyon of Nahal Me’arot (Wadi el-Mughara), some 3.5 km away from the Mediterranean shore (Garrod and Bate, 1937). Excavations by McCown in 1931 and 1932 identified three main layers (McCown and Keith, 1939): Layer A (20–50 cm thick) contained a mixture of Natufian, Aurignacian, and Mousterian stone tools, Layer B (∼200 cm thick) contained all the recovered human remains along with Mousterian stone tools, and Layer C (shallow sandy deposits at the base of the sedimentary sequence) yielded only a sparse lithic industry and no faunal remains. Layer B was divided into two subunits, mainly distinguished by their hardness. The upper hard earth unit B1 resembled “plaster of Paris”, whereas the lower breccia B2 was similar to concrete. The lithics of Skhul Layer B were attributed to the Levantine Mousterian and are comparable to those of Tabun C and Qafzeh (Garrod and Bate, 1937; Bar-Yosef and Meignen, 1992), while the macro-faunal remains in Layer B are similar to those of Tabun C and D (Garrod and Bate, 1937). Nine adult and immature human individuals (Skhul I–IX) were excavated from layer B at Skhul, some very incomplete or poorly preserved. At least three of these (IV, V, IX) were recognised as intentional burials, and Skhul V had a large boar mandible in its arms, perhaps the oldest-known example of grave goods. The material shows morphological variation which has been interpreted as varying along the Neanderthal-modern spectrum (e.g. McCown and Keith, 1939), or as representing an archaic form of AMH (e.g. Stringer, 2002). Dating studies yielded closed system ESR ages on faunal teeth in the range of 55–100 kya (Stringer et al., 1989) and 46–88 kya (McDermott et al., 1993), U-series ages on faunal teeth in the range of 43–80 kya (McDermott et al., 1993), and TL ages on burnt flint in the range of 99–134 kya (Mercier et al., 1993). New ESR and U-series analyses, including direct dates on the human fossils, indicate the best age estimate lies between 100 and 135 kya (Grün et al., 2005, Vita-Finzi and Stringer, 2007). Garrod and Bate reported the presence of four marine shell species (Acanthocardia deshayesii, Laevicardium crassum, Nassarius gibbosulus, Pecten jacobaeus), identified by Connolly and Tomlin, but did not indicate the number of specimens recovered or their stratigraphic provenance (Garrod and Bate, 1937: 224). The marine shells from Skhul were recently located at the Department of Palaeontology, Natural History Museum (NHM), London, and analysed by a multidisciplinary team (Vanhaeren et al., 2006). The Skhul material includes two perforated N. gibbosulus, a valve of A. deshayesii, a fragment of L. crassum, a fragment of a Cyprid, and a fragment of an undetermined shell. The P. jacobaeus specimen mentioned by Garrod and Bate cannot be located. Only the N. gibbosulus shells bear perforations that could have been used for suspension in a beadwork. In order to identify the layer from which these Nassarius originated, sediment matrix adhering to one of them and sediment samples from layers A, B1 and B2 were analysed for mineralogy and chemical composition. Major and trace elements, as well as the hardness of the sediment adhering to the pierced shell, indicate that it comes from layer B1-2.
The four pieces analysed in this study (Fig. 1) were located, like the shells, in the Department of Palaeontology, Natural History Museum, London.
Their labels unambiguously indicate that they come from Es-Skhul layer B. It is worth noting that they were considered important enough to be stored separately in a small tin, with a label apparently written by McCown. However, we have no contextual information that can be used to attribute them to subunit B1 or B2 since Garrod and Bate do not report the presence of pigments in layer B. Inspection of 30 pieces of breccia from layer B kept in the same institution has led to the identification of three large clasts enclosing small pieces of red and orange pigmental material (Fig. 2). This discovery demonstrates the presence of potential pigments in Skhul layer B and further corroborates the attribution of the bigger pieces analysed here to this layer.
Section snippets
Material and methods
The four pieces, called from now on Skhul 1–4, were examined and photographed under a reflected-light microscope in order to record anthropogenic and natural modifications and to identify areas suitable for sampling. Sampled areas are indicated in Fig. 1. Sampling was conducted under the microscope with a pointed scalpel and focused on an area covering ca. 1 mm2. Photomicrographs of the sampled area were taken before, during, and after the sampling process for conservation reasons and to leave
Description of the specimens
Skhul 1 is a piece of red material with an egg-like morphology, still partially contained in a piece of hard, thin grained breccia typical of the site’s “B” layer (Fig. 1a). The breccia shows a facet corresponding to the surface of a large object to which it adhered originally, possibly a flint artefact considering the flat smooth aspect of the facet. A small area of this lump of red material, preserved between the two residues of breccia, seems to keep the pristine appearance of the outer
Discussion
Rediscovered eighty years after they were unearthed, the Skhul pigments from the NHM collections, contrary to those from Qafzeh, are not associated with contextual information that could be used to identify meaningful spatial associations with other prominent features of this site (burials, hearths, personal ornaments, etc.). To what extent this small sample is representative of the pigments originally abandoned at the site, or present in deposits at the time of excavation, is difficult to
Conclusions
Integrated analysis of previously undescribed pigment fragments from Es-Skhul layer B sheds new light on the use of pigments by Middle Palaeolithic archaic modern humans in the Middle East. The small sample of Skhul pigments comes from a variety of geological sources, which implies systematic prospecting of raw material outcrops and is consistent with the hypothesis that the three originally yellow pieces were intentionally heated to change their colour. Firmly establishing whether the
Acknowledgements
We would like to thank Robert Kruszynski for facilitating access to the material kept at the Natural History Museum and the pigment sampling, and William Banks for his critical reading of the manuscript. Financial support for this project came from the CNRS-OMLL programme of the European Science Foundation, the Programme ORIGINE II of the Aquitaine Region, the Programme PROTEA of the French Ministry of Higher Education and Research and the European Research Council (FP7/2007/2013)/ERC Grant
References (105)
- et al.
Shells and ochre in Middle Paleolithic Qafzeh Cave, Israel: indications for modern behavior
J. Hum. Evol.
(2009) - et al.
Review of accelerator gadgets for art and archaeology
Nucl. Instrum. Methods Phys. Res. B.
(2004) - et al.
Excavation of buried Late Acheulean (Mid-Quaternary) land surfaces at Duinefontein 2, Western Cape Province, South Africa
J. Archaeol. Sci.
(2003) - et al.
Nassarius kraussianus shell beads from Blombos Cave: evidence for symbolic behaviour in the Middle Stone Age
J. Hum. Evol.
(2005) - et al.
U-series and ESR analyses of bones and teeth relating to the human burials from Skhul
J. Hum. Evol.
(2005) - et al.
An early bone tool industry from the Middle Stone Age at Blombos Cave, South Africa: implications for the origins of modern human behaviour, symbolism and language
J. Hum. Evol.
(2001) - et al.
Engraved ochres from the Middle Stone Age levels at Blombos Cave, South Africa
J. Hum. Evol.
(2009) - et al.
Extending the chronology of deposits at Blombos Cave, South Africa, back to 140 ka using optical dating of single and multiple grains of quartz
J. Hum. Evol.
(2006) - et al.
Diagenesis in prehistoic caves: the use of minerals that form in situ to assess the completeness of the archaeological record
J. Archaeolo. Sci.
(2000) Evidence of hunting and hafting during the Middle Stone Age at Sibudu Cave, KwaZulu-Natal: a multianalytical approach
J. Hum. Evol.
(2005)