Close examination of the photograph accompanying this report suggests that there are two types of petroglyphs present on the Fariseu panel: impact-produced figures of very slight grey patination, and more recent, superimposed abrasion-produced figures of very light, practically white colour. Both types have been cut into a grey to dark-grey (Munsell 10YR4/1 to 10YR6/1) patina, which thus predates all of the rock art. The very recent appearance of the petroglyphs is acknowledged in the IPA report.

The Fariseu site is close to the inflow of the Côa river into the Pocinho reservoir and has therefore experienced significant sediment deposition for the past 17 years, during which time it was submerged. Even much further downstream, at Canada do Inferno, three metres of sediment has accumulated as a result of the Pocinho dam. Beneath that very recent deposit, dated inscriptions of the 17th century were excavated in 1995, centring around A.D. 1620. There have been several instances in the Côa valley (e.g. at Barca and Penascosa) where petroglyphs were found covered by sediment, but in all cases so far this was either dam sediment of the last 17 years, or recent fluvial sand deposit. It is to be noted that most of the Côa rock art sites are still flooded frequently.

The IPA report does acknowledge, however, that direct dating of the rock art is desirable, which so far had consistently been rejected by the IPA team. It also concedes, indirectly, that the purported stone tools cannot be used for dating the rock art, by reporting that they are of colluvial origin. While this renders their dating potential for the rock art entirely worthless, the Fariseu researchers nevertheless claim to have provided secure ‘archaeologi-cal dating’. This is clearly premature even as a claim of archaeological rather than scientific dating. Until analytical evidence has been provided for the Fariseu sediment deposit, in the form of carbon isotope, TL or OSL analysis, that deposit has not been dated at all. Even if the presumed stone tools were of the Gravettien, which remains to be demonstrated, this would clearly be of no relevance to the age of the colluvium in which they were found. Even after the radiocarbon results from that deposit are available it must be cautioned that they may not refer to the event of sediment formation at all, but may be from material predating that event (e.g. slope wash).

All of this should be self-evident, which raises once again the question of the calibre of the work conducted by the IPA since the Côa valley rock art came under its jurisdiction. This work amounts to a catalogue of unsubstantiated claims, usually trumpeted through the mass media and presented without any adequate scientific data; and to a series of admitted failures in site management. The wholesale scrubbing of engraved panels ‘with wooden tools and river water’ to remove all lichens (thus eliminating the possibility of lichenometry) and accretionary mineral deposits (eliminating the possibility of using them for direct dating) ‘for political expediency’ (Zilhão 1996), the banning from the project of all researchers considered sceptical and the related xenophobia (Swartz 1997a, 1997b), and the surreptitious procedures of research and reporting of results have become the hallmarks of this work, the financial costs of which have been enormous to the Portuguese taxpayer. It needs to be emphasised that the Côa research project has cost millions of dollars already, and has so far not produced much credible scientific evidence. Its initial claims that the Penascosa terrace was Pleistocene and would yield Palaeolithic occupation evidence was abandoned after Zilhão et al. (1997) conceded that Watchman’s preliminary TL dates were correct, and that the terrace was of recent millennia. The IPA team has even discovered a remnant of a genuine Late Pleistocene terrace at the Penascosa site, but it is 40 m above the river, clearly indicating that the river was then 40 m above its present elevation. This only confirms the predictions of earlier researchers that the prospects of discovering any Pleistocene sediments at or near the floor of the lower Côa valley are very discouraging, because this is a very young valley experiencing rapid downcutting. It is perfectly possible that Pleistocene occupation evidence exists on the steep hill slopes high above the present river, perhaps in pockets of surviving terrace deposits. It is equally possible that debris from such deposits have found their way into colluvial (secondary) sediments on the valley floor. But what is becoming increasingly clear from the desperate IPA campaign to locate in situ Palaeolithic occupation evidence near the valley floor is that there is none to be found. A most intensive search conducted over several years has not yielded any such evidence. Nor should we expect to find in situ Palaeolithic material in what are, apparently without exception, late Holocene sediments. The obvious corollary is of course that most of the rock panels on which the Côa petroglyphs occur, particularly those near the present river’s elevation, could therefore not have existed prior to the Holocene, because the river was at a higher level in the Pleistocene.

It follows that, despite huge and very costly research efforts, no archaeological evidence has been recovered in the course of the IPA Côa campaign that can be linked securely to the rock art. Since this was the exclusive purpose of this massive campaign it can only be described as an unqualified failure. We remind the Director of IPA that he has a moral obligation to contribute credible data concerning the age of the Côa petroglyphs, because it was due to his decision that the Côa petroglyph panels were systematically denuded of all their lichens and accretions (Jaffe 1996). We remind him that in so doing he has effectively eliminated most direct dating approaches, particularly lichenometry (which would have almost certainly provided highly accurate and reliable estimates of rock art age) and analyses of accretionary mineral skins. So far he has pursued only unproductive strategies to make good his mistake, and the Fariseu project is merely the latest fiasco in a disastrous campaign.

If we were genuinely interested in the possible age of the Fariseu petroglyphs, several avenues could be explored. For instance, the pictures themselves provide clues. The uppermost ‘horse head’ in the above image includes a line carved across the animal’s muzzle, right at the base of the mouth. This marking seems to indicate the presence of a bridle which connects in a neat curve to where the bit would be located. This detail seems to imply that the animal was domesticated. Most archaeologists reject Bahn’s (1980, 1981) suggestion that horses were domesticated in the Pleistocene, in fact domestic horses are not thought to appear in Portugal before mid-Holocene times. Until we have solid evidence for domesticated horses in the Gravettian, this image is not likely to be of that period.