Pedra Furada

[Mikey Brass]

G.M. Santos et al. A revised chronology of the lowest occupation layer of Pedra Furada Rock Shelter, Piau!ı, Brazil: the Pleistocene peopling of the Americas. Quaternary Science Reviews 22 (2003) 2303–2310.

Abstract
The present work revisits the chronology of the archaeologically controversial Pedra Furada Rock Shelter of Southeast Piaui, Brazil, using an improved radiocarbon laboratory pre-treatment and measurements on charcoal samples. The procedure, known as ABOX-SC (acid–base–wet oxidation followed by stepped combustion), has previously been used to secure radiocarbon dates of >40ka for the antiquity of human occupation of Australia and South Africa, and now has been applied to charcoal from the previously dated oldest occupation layer of the Pedra Furada site. Previous radiocarbon dating had obtained only lower limits of 40–45 ka BP for the Pedra Furada basal layer. Nine charcoal samples from well-structured hearths were subjected to the ABOX-SC procedure and their radiocarbon content determined by accelerator mass spectrometry. Measurements on five of the samples returned ages of greater than 56 ka BP, from graphites produced from ABOX pre-treated charcoal combusted at 910 [degrees] C. Two other samples were greater than 50 ka BP. The remaining two samples were essentially completely combusted at 650 [degrees] C, with no material surviving to make a 910 [degrees] C CO2 fraction. Their ages were 41.3 and 47.2 ka BP. Ages obtained from graphites generated from the 650 [degrees] C combusted fraction are considered minimum ages.
2003 Elsevier Ltd. All rights reserved.

[Jacques Cinq-Mars]

G.M. Santos et al. A revised chronology of the lowest occupation layer of Pedra Furada Rock Shelter, Piau!ı, Brazil: the Pleistocene peopling of the Americas. Quaternary Science Reviews 22 (2003) 2303–2310.

Thanks for passing this on, and also for having uploaded the paper "you know where"!

I think it is a good paper -- you know I am biased! The information provided by the authors clearly demonstrates that the Pedra Furada material is, indeed, quite old. They conclude by saying:

Two of the new values have helped to obtain limit ages for the basal Layer in the Pedra Furada Rock Shelter and, so far they are the oldest radiocarbon results obtained for any South America archaeological site.

A rather humble statement in view of the fact that it is actually applicable to both South and North America.

However, with Pedra Furada, the real problem lies elsewhere, i.e., in the inability of too many North American archaeologists to see very far beyond the tip of their fluted points. The best example of this attitude is the "hatchet job" carried a few years ago by Metzer and colleagues and published in Antiquity:

Meltzer, David J., James M. Adovasio and Tom D. Dillehay.1994. On a Pleistocene human occupation at Pedra Furada, Brazil. Antiquity 68:695-714.

Other illuminating titles dealing with the various controversies surrounding Pedra Furada at the time can also be read in:

Bahn, Paul G. 1993. 50,000-year-old Americans of Pedra Furada. Nature 362:114-115.

Bahn, Paul G. and Hansjürgen Müller-Beck. 1991. Concerned Readers. Archæology 44(2):10-11.

Borrero, Luis Alberto. 1995. Human and natural agency: some comments on Pedra Furada. Antiquity 69:602-603.

Delibrias, G. N. Guidon  and F. Parenti. 1988. Stratigraphy and Chronology of the Toca do Boqueirao do Sitio da Pedra Furada. IN: Archaeometry: Australasian Studies1988, ed. by John R. Prescott, p. 153. Adelaide: University of Adelaide.

Dennell, Robin and Linda Hurcombe. 1995. Comment on Pedra Furada. Antiquity 69:604.

Fagan, Brian. 1990. Tracking the First Americans. Archæology 43(6):14-20.

Guidon, Niède, Fabio Parenti, Maria de Fatima da Luz, Claude Guérin et Martine Faure. 1994. Le plus ancien peuplement de l'Amérique: le Paléolithique du Nordeste brésilien. Bull. S.P.F. 91(4-5):246-250.

Guidon, N. and B, Arnaud. 1991. The chronology of the New World: two faces of one reality.

And then, for the more recent stuff, one only has to do a search in the PALANTH Forum!

Jacques

[Mikey Brass]

Thanks for passing this on, and also for having uploaded the paper "you know where"!

I think it is a good paper -- you know I am biased! The information provided by the authors clearly demonstrates that the Pedra Furada material is, indeed, quite old. They conclude by saying: A rather humble statement in view of the fact that it is actually applicable to both South and North America.

As I understand the sites in the area, which isn't much at all, the caves dated by the authors are different to the Pedra Furada cave that has been the center of controversy with its *stone tools*. Instead, I believe this paper is in connection with the one published a few months ago claiming that 14C dates show the occupation was in the Holocene.

Methinks this is the start of a new debate on the region :-)

[Jacques Cinq-Mars]

As I understand the sites in the area, which isn't much at all, the caves dated by the authors are different to the Pedra Furada cave that has been the center of controversy with its *stone tools*. Instead, I believe this paper is in connection with the one published a few months ago claiming that 14C dates show the occupation was in the Holocene.

Methinks this is the start of a new debate on the region :-)

Sorry to disagree, but the paper does address some of the Pedra Furada dating issues. As for the early stone tools from this site (and I have seen a few of them), they really don't need to be **.

Jacques Cinq-Mars

[Mikey Brass]

Sorry to disagree, but the paper does address some of the Pedra Furada dating issues. As for the early stone tools from this site (and I have seen a few of them), they really don't need to be **.

Jacques Cinq-Mars

I said I didn't know much about the region ;-))))

With regards to the earlier paper, which dated some of the rock art in the region to the Holocene and which contradicted earlier dating results placing, this correspondence was posted (from someone corresponding with one of the scientists involved in the Holocene-date article) on a yahoogroups list:

"Our radiocarbon dating of the rock paintings does NOT have anything to do with the sediment archaeology and the ground charcoal dates at Pedra Furada. We are not disputing that.  Those ages for Pedra Furada may very well prove to be correct.  I believe that there are similarly old ages (tens of
thousands of years old) for Bastiana for charcoal in the ground, but am not positive.  I hope Dito Morales know about the archaeology of Bastiana.

I believe that there is occupation at both Pedra Furada and Bastiana for long periods of time. (Again, Dito correct me if I am wrong.)

All we are saying by our paper (which I attach here) is that the rock art at Bastiana and the other images we have dated at other nearby shelters (including Pedra Furada) are less than 4,000.  This is in contradiction to Watanabe's result of 30,000 years old for a calcite layer covering some of
the rock art images at Bastiana."

[Daryl Habel]

I said I didn't know much about the region ;-))))

With regards to the earlier paper, which dated some of the rock art in the region to the Holocene and which contradicted earlier dating results.....
[SNIP]

Just to clear away potential confusion, the site mentioned in the quote above is Toca da Bastiana, another rockshelter located within the Serra da Capivara National Park.  A portion of a sample taken from a calcite veneer at Toca da Bastiana was dated using EPR to 27 ka by Baffa Jr (1991).  The remaining portion of the same sample was dated using EPR and TL to 35-43 ka (Watanabe et al. (2003).  However, Rowe & Steelman (2003) followed this with a reported radiocarbon measurement of calcium oxalate extracted from the calcite layer dated by Watanabe et al., which "yielded only 2490 +/- 30 BP for a minimum age" and a "direct" radiocarbon date for "the red paintings associated with the calcite layer...determined to be 3730 +/- 90 BP  using plasma-chemical extraction and AMS radiocarbon dating."  In addition they report four additional dates on paint from Bastiana as well as four more from Toca do Sitio do Meio, Toca do Extreme, and Pedra Furada, all of which fall between about 1200 to 3600 BP.

Rowe, Marvin W. & Karen L. Steelman (2003). Comment on "some evidence of a date of first humans to arrive in Brazil". Journal of Archaeological Science 30: 1349-1351.

Watanabe, Shigueo, Ayta, W.E.F., Hamaguchi, H., Guidon, N.. La Salvia, E.S., Maranca, S. & Baffa Filho, O. (2003). Journal of Archaeological Research 30: 351-354.

Baffa, O. Jr. (1991) Private communication unpublished. (referenced in Watanabe et al. 2003).

The new paper, which is the subject of this thread (Santos et al. 2003) reports dates from  9 charcoal samples taken from "well structured hearths" found in proximity to stone artifacts in the oldest phase (PF 1) of human occupation at Pedra Furada.  These were  subjected to the recently-developed ABOX-SC (pre-treatment) procedure and determined by AMS radiocarbon dating.  Seven yielded determinations of over 50 ka and five yield ages greater than 56 ka.  The remaing two samples were essentially completely combusted at 650 degrees C. and their ages of 41.3 and 47.2 ka are considered minimum ages.

From a previous discussion on this forum there seems reason for optimism about the ability of the ABOX-SC procedure to increase the capability of AMS radiocarbon dating to a limit of about 56 ka.  

Santos et al. (Table 1) shows that two heated pebbles from hearth 59 and one heated pebble from hearth 100 (as well as two heated pebbles from structure 99) were TL tested (reference to Parenti et al. 1990).  The text (p. 2306) says some of  the charcoal samples were taken from closely associated structures which had been previously sampled by TL (presumably those listed in Table 1),  but nowhere that I can find do they give any TL age determinations. I note in the latest news report from ABC:

http://www.abc.net.au/science/news/stories/s990775.htm

a quote from Dr. Michael Bird, from the team that helped develop the ABOX-SC technique, makes a point that TL dating of the hearthstones showed that they "were heated independently from the stones found outside the hearths in the same layer; thus refuting the possibility that the stones were heated by natural fires.  The same point is made in the new Santos et al. paper (p. 2303) with a reference to Bellomo 1993.

While I have no problem accepting the ABOX-SC AMS dating (what do I know about it?), it seems to me they could very easily have made this point about rejecting natural fires much clearer by providing a bit more contextual information on the associations of the new dates on hearths/structures with the ages determined for the previously TL dated hearthstones (even though they do reference this).  

Without access to Parenti et al. 1990 or Bellomo 1993, I'm wondering about the TL dates and hearthstone contexts, and how well  these correlate with the ages and contexts of the ABOX-SC AMS samples taken from the hearths.  If Santos et al. (2003) describe these relationships with anything other than reference to other papers,  I must have missed it (which is possible).

Do  the age determinations for the TL estimates and AMS estimates from associated structures come close to matching?
Any comment would be much appreciated.

Thanks in advance,
Dar

[Jacques Cinq-Mars]

All,

For those of you who have not had the opportunity to read the Santos & al. paper presently under discussion, the following (HERE) is a fairly decent journalistic summary of what the story is all about.

Jacques Cinq-Mars

[AWSX]

Also here is a website interview with some of the authors discussing the results and other discoveries in the area:http://www.athenapub.com/10pfurad.htm

The teeth and skull fragment from Garrincho are particularly interesting for the very robust traits. Finds like this in the Americas have been very rare. This seems to indicate an early immegration of  late Pleistocene humans with very archaic craniofacial traits.

Allan Shumaker

[Jacques Cinq-Mars]

Just a few quick notes, here:

As of the mid-90s, much of the TL work, at Pedra Furada, was essentially carried out in order to determine whether or not "combustion features" were actually hearths. This has been thoroughly discussed by Fabio Parenti in:

Parenti, Fabio, 2001, Le Gisement Quaternaire de Pedra Furada (Piauí, Brésil). Stratigraphie, Chronologie, Evolution Culturelle. Paris: Editions Recherche sur les Civilisations

Also, at about same the time, the chronostratigraphy of the shelter had been worked out by Fabio (again, Parenti 2001), on the basis of an impressive – and coherent – series of 46 C-14 dates obtained from charcoal found in hearths. The dates obtained from the Pleistocene units of the deposit range from 10,000-50,000 kya (uncorrected). The earliest ones (about seven of them) from the two lowest layers fall between 40,000 and 50,000 kya

At present time, to my knowledge, the most pertinent information that would allow one to compare the AMS and the TL (dating) results can be found in:

Valladas, H., N. Merciera, M. Michaba, J. L. Joronb, J. L. Reyssa, and N. Guidon. 2003. TL age-estimates of burnt quartz pebbles from the Toca do BoqueirImageo da Pedra Furada (Piaui, Northeastern Brazil). Quaternary Science Reviews 22(10-13): 1257-1263.

Abstract:

The thermoluminescence technique was used to date 40 burnt quartz pebbles from the lowest layers (PF1 and PF2) of the Toca do BoqueirImageo da Pedra Furada rock shelter (Piaui, Brazil) attributed to Upper Pleistocene period, whose radiocarbon ages exceed 35 ka. Our results suggest that the quartz specimens were burnt between 30 and more than 100 ka ago, but they provide no evidence that the heating was related to human activity.

.

If one takes into consideration Parenti's AMS dates, the Valladas & al. TL ones, and the ones presented by Santos & al., it does appear that what is being dated is certainly as old as 40,000 and perhaps older that 55,000 kya.

The real question with the Pedra Furada material is not so much about the (early) dates; it has to do with whether or not Niède Guidon, Fabio Parenti and colleagues know how to recognize an archaeological site (features, artefacts, etc.) when they step on it. Guess what side I am on!

Jacques Cinq-Mars

[Mikey Brass]

The real question with the Pedra Furada material is not so much about the (early) dates; it has to do with whether or not Niède Guidon, Fabio Parenti and colleagues know how to recognize an archaeological site (features, artefacts, etc.) when they step on it. Guess what side I am on!

Jacques Cinq-Mars

What appears amazing to me, as an outsider to the continent's archaeology, is why would a site whose first occupation layers are dated to 50-40kya have stone tools that resemble Oldowan tools in their simplicity, with other archaeologists being able to dismiss them as geofacts. One would have thought that people at this time period would have been able to make formal stone tools.

And if they are stone tools, what the heck caused the loss of the formal stone tool manufacturing processes from wherever they arrived from.

[AWSX]

Very good question. Should we expect to find flaked stone tool technology associated with every population of 'anatomically modern humans' found anywhere in the world? (BTW, I am beginning to dislike that AMH label because it does not reflect differences between the 'robust' pleistocene and more 'gracile' holocene populations.)

The OoA proponents would assume that all AMH groups started their journey with an advanced stone tool technology. However would MR proponents require an advance in stone tool technology to accompany the evolutionary process?

If we consider how sparsely the Americas were populated at that time it could be possible that an isolated group could lose technology.

Allan Shumaker

[lagarvelho]

Mike:

I think isolated groups *do* lose technology.  The Tasmanian people lacked some artifacts that the native people on the main Australian landmass had(I can't remember what these were).  I'm not really surprised.  Sometimes people "lose" technology if they wander somewhat out of the orbit of the rest of their neighbors. . . .or the place they've wandered into doesn't have the raw materials they need to maintain that technology.
Anne G

[Jacques Cinq-Mars]

All,

All,

Here is a brief blurb on what has already been discussed regarding the ABOX dating of Pedra Furada. Emphasized in "red" ( in the text) is Dr. Bird's very cautious statement!

Jacques Cinq-Mars

Ancient hearth tests carbon dating

Bob Beale

ABC Science Online

Monday, 17 November  2003 -- News in Science


Rock art at Serra da Capivara National Park, home of the Pedra Furada site in Brazil (Embassy of Brazil, London)

People were keeping warm by a fire in a rock shelter at least 56,000 years ago, according to new analysis of what may be the oldest known human record in the Americas.

This is about 40,000 years earlier than generally agreed for when people first arrived in the Americas.

The international team of researchers dated charcoal from a hearth at the controversial Pedra Furada archaeological site in Brazil and reported its findings in the latest issue of the journal Quaternary Science Reviews.

They used a new technique that pushes back the so-called radiocarbon dating barrier, according to Dr Guaciara dos Santos and colleagues who ran tests at the Australian National University.

Scientists have been polarised about the age of the Pedra Furada site because estimates have been in "profound disagreement" with accepted wisdom about who, when, where and how people first arrived in the Americas. These were supposedly the Clovis people who walked from Siberia into North America across an Ice Age land bridge only 12,000 to 14,000 years ago.

"These dates are good and reliable and there's no reason to doubt them," Dr Michael Bird, a member of the team who developed the new dating technique, told ABC Science Online. "The question goes back to the archaeology. If they are hearths, they are very old indeed."

The site at Pedra Furada, in the Serra da Capivara National Park, is a rich archaeological area of sandstone rock shelters. It contains many prehistoric sites, including hundreds of rock artworks, stone tools and human remains.

Earlier tests on charcoal from the deepest layers of the excavations suggested that it was at least 40,000 years old, the traditionally accepted accurate "barrier" limit of radiocarbon dating. But scientists were still puzzled about the authenticity of the hearths as human artefacts and whether younger carbon sources could have contaminated the samples and skewed the results.

The new study says that thermoluminscence testing of the hearthstones showed that they "were heated independently from the stones found outside the hearths in the same layer; thus, refuting the possibility that the stones were heated by natural fires".

It revises the dates on those earlier charcoal tests using Bird's technique to decontaminate it first. The procedure is known as ABOX (acid-base-wet-oxidation) and involves chemically scouring a fine layer off the charcoal surface.

"[This] reliably removes contamination from charcoal and wood enabling credible radiocarbon dating to about 55,000 years before present," the report said.

Bird said the method had been used in the past two years to secure radiocarbon dates older than 40,000 years for archaeological sites in South Africa and Australia, notably the famous Devil's Lair site in Western Australia, which was redated at up to 50,000 years old.

Radiocarbon dates become progressively less reliable on older material and until the ABOX technique was developed, few scientists would accept their accuracy beyond the barrier limit, he said.

"At 50,000 years you have only about 0.1% of the original radiocarbon present, so contamination with younger material is a major issue," Bird said. "This is a much better way of pre-treating the samples to get rid of any contamination. It's becoming the gold standard in archaeology for getting good reliable dates that you can believe, particularly at these old time scales."

Out of seven Pedra Furada charcoal samples scientists took from the hearth structures in the deepest layers, five were beyond the limit of the ABOX technique itself, returning ages greater than 56,000 years, the report said. Analysis of the final two samples gave finite ages of 53,000 and 55,000 years.