The stylistically Paleolithic petroglyphs of the Côa valley (Portugal) are of Paleolithic age

A refutation of their <<direct dating>> to recent times

João Zilhão



8.4. Survival of aurochs, deer and horse in the <<bleak tundra>> of the Côa region

Bednarik (n.d.) argues that, being only 35 km away from the mountain glaciers of Serra da Estrela, <<the [Côa] valley experienced severe periglacial conditions which only very few species could be realistically expected to have survived (...) in this bleak tundra landscape>>; <<aurochs, deer and horse were almost certainly not among them, and ibex only in summer>>. This statement is an excellent illustration of Bednarik's ignorance of the Paleolithic record of Iberia, and of the geographical realities of the present day world. Any postcard from Switzerland suffices to refute the concept that the landscape 35 km away from a mountain glacier corresponds necessarily to a <<bleak tundra>>. A simple look at a map will also show that most Cantabrian Upper Paleolithic cave sites were less than 35 km away from the glaciated mountains of Northern Spain but, notwithstanding, contained faunal assemblages entirely made up of the animals which, according to Bednarik, would not have survived in the Côa region.



8.5. Ibex as a chronological marker

Bednarik (1995a) also states that <<by 11,000 BP, ibex does not seem to be present in low- altitude occupation deposits, but it would continue to occur in the mountainous areas, so it is not a chronological marker>>. This statement needs correction. Magdalenian faunal assemblages are rare in Portugal, and all come from cave sites. In the examples known -- all located in Estremadura -- ibex, chamois and horse are entirely absent and the fauna is dominated by red deer and rabbit with marginal amounts of wild boar and roe deer. This is particularly the case at the cave of Caldeirão (Fig. 5), which was excavated by the present author and is located at a distance of 20 km from the top of Serra d'Aire, the highest elevation of Estremadura (678 m). At this site, the absence of horse and wild caprids contrasts markedly with their abundance in the underlying Solutrean levels (Póvoas et al. 1992; Zilhão 1995a). This, together with other paleoenvironmental indicators, namely the rodent and the land snail faunas, suggests that, after 16,000 BP, the low altitude limestone elevations of Portuguese Estremadura were already covered by forest, and that alpine species and horse were not present in those areas. Given the proximity of the Côa region to the Meseta and its drier climate, it is quite possible, however, that ibex may have survived into the Magdalenian in this more inland part of the country, although such a survival is hard to conceive after 9500 BP. By then, even the Serra da Estrela, which reaches the highest elevation in Portugal (2000 m) and is located some 50 km south of the southernmost occurrences of Paleolithic petroglyphs in the Côa, was already covered by a Quercus pyrenaica forest up to an elevation of 1600 m (Mateus and Queirós 1993). Since ibex is not known to inhabit these forests, it seems very unlikely that it survived in the Côa region after the end of the Upper Paleolithic and, therefore, that the representations of the animal found in the valley art postdate the Tardiglacial/Early Holocene boundary. It should be stressed, in any case, that ibex is not represented in the faunal inventories from the late Holocene archaeological sites (for instance, those from the Neolithic or the Chalcolithic) of the Douro basin (Jorge 1993).



8.6. Use of metal tools

Bednarik (1995b) categorically states that one stylistically Paleolithic anthropomorphic figure engraved in a panel from Ribeira de Piscos (one that was not analyzed in the framework of the <<direct dating project>>) had been made with a metal tool, which obviously excluded a Paleolithic age for this figure. Francesco d'Errico, a world authority on this kind of problem, contributed the following comments on this issue (in litteris, August 23, 1995): <<Experimentation carried out by me and other colleagues shows that stone tools with sharp points can produce engraved lines which display none of the features characteristic of stone tool use and that can be hardly distinguished from those produced by metal points>>. <<My unpublished technological analysis of Fornols-Haut engravings (...) shown that clues demonstrating the use of lithic implements are present on these figures but that they occur rather rarely. The research was carried out in the laboratory using high resolution casts of the engraved panels examined by mean of low-angled light as well as a scanning electron microscope. If such an analysis had to be limited to field observations the majority if not all of these clues would have remained unnoticed>>. These statements seem sufficient to demonstrate that the categorical nature of Bednarik's conclusion as regards the tool used to engrave the human figure from Ribeira de Piscos is totally without basis.



8.7. Absence of archaeological context

Bednarik (1995a) states that the Paleolithic, Epipaleolithic, or even Mesolithic, settlements closest to the Côa art sites are more than 120 km away and that, therefore, there would be no archaeological context for that art, if it were indeed Paleolithic. The fact that he uses this as part of the argument against the Côa art being of Paleolithic age (Bednarik n.d.) is intriguing since, contrary to what he had to say on last glacial faunas, this time he does equal <<absence of evidence>> with <<evidence of absence>>. In this case, however, the equation is not appropriate, in the first place because, although still few and far between (due to the traditional bias towards caves and the fact that there are no limestone outcrops in the area), several Upper Paleolithic sites have been found in the Spanish Meseta since the early 1980's. Bengoechea et al. (1986) and Fabian (1986), for instance, report Solutrean and Magdalenian open air settlement sites in the Valladolid and Salamanca provinces, across the border from Portugal (Fig. 5). So, if the issue of archaeological context is looked at from a regional perspective, it is quite clear that interpretation of absence is not what is at stake here.

If looked at from a local perspective the fact that, until recently, no Upper Paleolithic archaeological sites had been found in the Côa valley (or, more generally, in Northern Portugal) was in itself meaningless because no survey for Paleolithic sites had ever been carried out. In other words, since no one had ever looked for the evidence, any assessment regarding the presence or absence of such evidence would in this case be scientifically illegitimate. Recent developments illustrate the point better than any further considerations on logic. On August 14, 1995, at 10 a.m., after only 20 minutes of the first serious archaeological survey of the Côa valley by experienced Paleolithic archaeologists, an open air camp site was found at Cardina, on a Pleistocene platform ca. 3 km upriver from Penascosa (Fig. 1). The lithic assemblage so far recovered in the test excavations already carried out at this site is mostly made up of flint and rock crystal tools and cores whose typological and technological characteristics are consistent with an Upper or Terminal Magdalenian age (10,000-12,000 BP). Testing carried out in September at a second nearby platform revealed an extremely rich archaeological level at a depth of ca. 1 m below the surface. The thousands of lithic artifacts recovered in the 4 m2 test make up an assemblage that is typical of the Late Gravettian: similar assemblages excavated in several cave and open air settlement sites located in Portuguese Estremadura have already been radiocarbon dated to ca. 22,000 BP. It should be quite clear from these examples that there must be many more sites in the region waiting to be found.



9. Bednarik's arguments on style

The most forceful statements regarding the stylistically non-Paleolithic nature of the Côa valley art petroglyphs in Bednarik's dating report (Bednarik 1995b) are the following:

Regarding bovids

<<several bovids have internal markings on muzzles that do not resemble Paleolithic art>>; <<the horns on bovids do not resemble those on most of the supposed aurochs figures of Paleolithic times, particularly those presumed to be of the Solutrean. They do, however, resemble the forward position and twist found in modern Iberian cattle breeds>>; <<These [modern Iberian cattle breeds] also share the slim head shape found in the Côa figures>>. In sum, given their muzzles, their head shape, and their horns, the Côa bovids are domestic oxen, not aurochsen.

Regarding caprids

<<Even more inappropriate is the description of the caprid-like figures as ibexes. The large and distinctly curved horns of the ibex are not present, instead the horns found in the petroglyphs resemble those of certain domestic goats>>

Regarding equids

<<The line indicating the overlap of the haunch in the horse at Ribeira de Piscos is not normally found in Paleolithic animal pictures>>.

Although it is really hard to be wrong on all counts, the fact is that not one of Bednarik's above quoted affirmations has any correspondence in reality. <<The line indicating the overlap of the haunch>> can be found, for instance, in painted horses from Ekain (Altuna and Apellániz 1978; Apellaniz 1987) or Niaux (Clottes 1984), just to mention two cases. As for the bovids, the comparison in Fig. 6 between an aurochs head from Lascaux and one of those Bednarik claims to have analyzed at Penascosa should be enough to solve the issue concerning <<internal marks on muzzles>> and <<forward position and twist>> of the horns. As regards horns, comparison (Fig. 6) with the engraved slabs from Parpalló (Villaverde 1994) actually confirms the attribution of most of the Paleolithic Côa art to pre-Magdalenian times. In Fig. 6, the <<slim head shape>> of some of the Côa bovids is also compared to that of the famous aurochs of the Grotte de Tête du Lion, in Ardèche, radiocarbon dated (by association with a close-by hearth containing fragments of the pigment used in the painting) to ca. 21,500 BP (Combier 1984). This figure, as well as several of the aurochsen from Pech Merle (Lorblanchet 1984), also share with many of Bednarik's <<domestic>> bovids of the Côa the squarish shape of the muzzle. As for the caprids, the four representations in Fig. 5 (one from Penascosa, one from Canada do Inferno, and two from Quinta da Barca) should be enough to enlighten the reader as to Bednarik's affirmation that the <<large and distinctly curved horns of the ibex are not present>>! Even the more schematic of these horn representations show the characteristic twisted shape that allows the unmistakable identification of these caprids as belonging to the iberian variety of ibex, Capra pyrenaica (cf. Altuna and Apellániz 1978).

In short, the specific conventions mentioned by Bednarik (1995b) conform completely with what is known from well dated Paleolithic art and the criteria he uses to describe the Côa animals as domesticates would imply that bovids and caprids had been domesticated in the Franco-Cantabrian region as early as 20,000 years ago! This is obviously absurd, and deserves no further comment except that, as was the case with Watchman, it seems fair to conclude that Bednarik came to the Côa essentially for two reasons: to prove himself right, that is, to find <<evidence>> of the correctness of his previous statements on the post-Paleolithic age of the stylistically Paleolithic petroglyphs; and to make an innovative point regarding preservation policies for rock art sites since, according to them, <<if the art were to be shown to be post-Paleolithic, its importance would diminish dramatically and the controversy concerning its preservation would be largely resolved>> (quoted from a letter by Bednarik and Watchman offering their services to EDP's board of directors, dated March 24, 1995).

Índice do texto/Parte anterior/Parte seguinte/References/Figures

Início da página/Página principal/Bibliografia


Esta página foi realizada para o Nestcape Navigator 2.0
Última actualização no dia 1 de Junho 96
siljan@gemini.ci.uc.pt