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