The Working of Pigment during the Aurignacian Period: Evidence from Ucagizli Cave (Turkey)

Article excerpt

New finds from the Upper Palaeolithic of Anatolia, and the mineralogical analysis of their colours, extends evidence of a precocious interest in pigments from the western European heartland of Palaeolithic painting into the Near East.


The discovery of a new Palaeolithic site in Turkey was made possible by a recent research programme carried out in the southeastern part of this country. Ucagizli is a cave site located on the east Mediterranean coast at the foot of the Cassius mountains (Minzoni-Deroche 1992). Finds from it - a coloured pebble, a small bladelet with some traces of red colour on one edge, and two coarse nodules of a hard black and red material - suggest the manufacture of pigment by Palaeolithic people.

In the decorated caves of the Pyrenees (largo sensu) from the Magdalenian, analysis of pigment has been carried out on the cave paintings and on mobiliary objects found in the excavations. These identified the manufacture of a specific matter from selected minerals previously ground and mixed together (Clottes et al. 1990a), sometimes also with an organic bmnder (Pepe et al. 1990) - evidence of prehistoric intention to manufacture a real paint, each component of which brought its own property. At La Vache (Alliat, Ariege) for instance, a cave settlement occupied during the Magdalenian period, a true oil paint was manufactured. It was used for painting in the sanctuary of Niaux opposite and across the Vicdessos river (Clottes et al. 1990b). Crayons from Lascaux (Dordogne) (Couraud & Laming Emperaire 1979) or from Arcy-sur-Cure (Yonne) (Couraud 1991) show streaks which may be considered evidence that coloured powder was collected by scraping.

The colour of iron oxides and hydroxides - 'ochres' - changes with simple roasting in open air at moderate temperatures. Goethite, yellow iron hydroxide, becomes browner around 250 [degrees] C, before reaching the bright red colour of haematite, a red iron oxide at about 500 [degrees] C. At higher temperatures, the colour turns to purple and, with temperatures over 1000 [degrees] C, to black (Bouchonnet 1977). In a prehistoric open fire, temperatures between 500 [degrees] C and 700 [degrees] C may be obtained and maintained. Archaeological clues to such a transformation have been forwarded by A. Leroi-Gourhan for the Chatelperronian levels of the Grotte du Renne at Arcy-sur-Cure where nodules of various colours, from yellow to deep red, were excavated in a 'series of small hearths which had been used for the roasting of ochre' (Leroi-Gourhan 1962). In fact, prehistoric people tried very early to modify the colour of minerals from yellow to red. As early as the Mousterian, at La Ferrassie (Dordogne), a sandstone pebble was found with combustion traces; it was associated with ash remains and burnt bone scraps, revealing human intent to make a deep red pigment. Aurignacian remains are numerous: pigments probably heated with various colours have been found, for instance, at La Quina, Les Gardes (Charente) (Henri-Martin 1965) and in the Hyaena Gallery at Brassempouy (Landes) (H. Delporte, pers. comm.). In Perigordian V levels at La Ferrassie, raw nodules associated with used artefacts have been found connected with fire remains. A flat yellow piece of fine sandstone (10.4x6.8x2.8 cm) had been heated on both sides - a change in colour was observed); and it shows many use-wear traces by way of spots and streaks (San Juan 1990). Inside the Upper Solutrean levels from the Fourneau du Diable (Dordogne) (Peyrony 1932) red haematite nodules were excavated in association with yellow goethite ones and combustion remains. From the Upper Magdalenien at Abri Pasquet (Dordogne), F. Carre studied a brownish clay level including traces of human occupation which cannot be interpreted further because of the lack of tools on the minuscule surface that could have been exploited. But he found a small structure made of pebbles which broke out in the fire and upon which spots of red ochre and manganese oxide were found, interpreted as an 'area of colouring matter preparation through roasting' (Carre 1983). …