¿ìè¶ÌÊÓƵ

‘I climbed down the rope to the bottom of the shaft, 80 feet below,
and then realised that bees were heading my way in a squadron. I hid behind
some rocks, but they eventually found me, and I had to get out fast.’ Niede
Guidon grimaces at the memory. ‘I certainly climbed up faster than I came
down,’ she jokes, describing how she emerged from the chasm of Sansao, her
head and arms enveloped in Brazil’s extraordinarily aggressive ‘killer bees’.
Despite 200 stings on that occasion, and a long drive over dirt roads to
the nearest hospital, she not only survived but continues to work in the
cave, though now in protective clothing and headgear.

Killer bees are not the only hazard faced by this tough 58-year-old
Brazilian archaeologist. Over the past 20 years she has carried out an amazing
programme of work in Piaui, one of Brazil’s poorest and most arid regions.
In the course of this research, she has found herself at the centre of the
fiercest debate in American archaeology today: when did people first enter
the New World?

Guidon started by studying biology and natural sciences at the University
of Sao Paulo; later she studied prehistory in France. In 1963 she was working
in a museum at Sao Paulo when one day the mayor of the remote little town
of Sao Raimundo Nonato came to the museum and showed her photographs of
some rock art in his area of Piaui.

The art depicted hunting scenes and acts of sex and violence. It was
so different in style to the art with which Guidon was familiar, which was
mostly geometric, that she resolved to investigate.

After a period of exile in France to escape the military dictatorship
which took over Brazil in the 1960s, she visited Sao Raimundo Nonato in
the 1970s. Over the next two decades, she continued to explore, survey and
excavate the area with a growing team of international assistants. Since
her first archaeological explorations by mule or donkey, she has recorded
346 sites in the region, 280 of which contain rock art.

Guidon’s sites are in a part of Piaui that is characterised by cactuses
and soaring sandstone cliffs, with huge rock shelters hollowed in to their
bases. One of the largest shelters, known as Pedra Furada (‘perforated rock’,
after a natural arch nearby), is profusely decorated from one end to the
other with red, and sometimes white, paintings representing humans as dynamic
little stick figures, along with large birds, and animals such as deer,
armadillo and capybara, the world’s largest rodent. Here and there one can
see rarer depictions such as a crab, or a tiny hunting scene. At other sites
there are scenes of sex and violence. Studies of style, motif and superimposition
(where one layer of art rests upon another) led Guidon and her French colleague
Annemarie Pessis to divide the region’s art into three phases, but they
had no idea how old the paintings might be.

So they started to excavate at the foot of the decorated panels in several
sites, most notably at Pedra Furada in 1978. They explored successive occupation
layers, rich in stone tools and charcoal from campfires. Unfortunately,
all bones have totally disintegrated in the sandstone shelters. As she dug
down, Guidon sent off charcoal samples from different layers to the Centre
des Faibles Radioactivites in Gif-sur-Yvette, the foremost radiocarbon laboratory
in France. The first results were not unusual: 6000 years old, even 12 000
years old. Soon, however, she became convinced there had been a mistake;
the results were 23 000 and 29 000 years old. Since then, charcoal aged
32 000 years and even 48 500 years has been uncovered – all the dates emerging
in a coherent sequence. The dates have been confirmed by the Beta Analytic
Laboratory in Florida.

To understand Guidon’s shock, bear in mind that for decades a fundamental
dogma of all archaeology in the New World was that people first entered
America no earlier than 12 000 years ago, at a time when the Ice Age had
lowered sea levels sufficiently to expose an 85-kilometre land bridge across
the Bering Strait, linking Siberia and Alaska. This theory goes back to
the 1930s when Edgar Howard and John Cotter of the University Museum in
Philadelphia discovered long stone projectile points near Clovis in New
Mexico. These Clovis points are finely flaked spear points and were used
to kill prehistoric big game such as bison and mammoth. When radiocarbon
dating became established as an archaeological tool, in the 1950s, tests
revealed that the spear points were about 12 000 years old. The Clovis hunters
were accepted by most archaeologists as the first Americans.

Claims for earlier human occupation of the New World have surfaced sporadically
ever since. Even Louis Leakey, famous for his work on early humans in East
Africa, declared that California had evidence for occupation about 200 000
years ago. Most archaeologists, however, believe that the alleged stone
tools at his site in the Calico Hills are ‘geofacts’, that is, stone shaped
by natural processes. Only one of the claims advanced has stood up to intense
scientific scrutiny.

In the 1970s, James Adovasio of the University of Pittsburgh carried
out a series of impeccable excavations in the rock shelter of Meadowcroft
in Pennsylvania. He produced more than 50 coherent radiocarbon dates stretching
back to 14 500 years ago. This is older than the official start of Clovis
– though not enormously so in archaeological terms – making Meadowcroft
the earliest definite site of occupation in North America so far. Since
people would have had to cross North America to reach the south, so the
argument goes, it follows that people could not possibly have settled in
South America before this date. Yet Guidon’s Brazilian evidence seems to
indicate otherwise.

When Guidon, after waiting to amass more dates, finally published her
claims in Nature in 1986 (vol 321, p 769) there were reverberations throughout
the archaeological world. Scholars immediately divided into three camps:
some longtime supporters of pre-Clovis occupation of the New World, such
as Alan Bryan of the University of Alberta and Robson Bonnichsen of the
University of Maine, were perfectly willing to accept her results; most
archaeologists advised caution while awaiting fuller publication; and some
expressed the gravest doubts about her data. It is the last group, particularly
Thomas Lynch of Cornell University, which has tended to make its voice heard
most loudly, although none of Guidon’s most vociferous critics has been
to see her evidence at first hand. Yet she is not possessive about the sites
of her finds, and is eager for others to come and investigate – even to
dig their own sites in the region.

Radiocarbon dates from a different South American site, Monte Verde
in Chile, have added fuel to the debate. Tom Dillehay of the University
of Kentucky started to excavate at Monte Verde in 1977. He found charcoal
and wood that date to about 13 000 years ago. Even more remarkable, he uncovered
charred wood that has been dated to about 33 000 years ago. Dillehay now
says that he has ‘unequivocal’ evidence that a flake and core (a lump of
stone from which flakes have been removed), dating from about 33 000 years
ago, were made by humans (American Antiquity, vol 56, p 333).

Monte Verde is a waterlogged site, which means that it is excellent
for preserving organic materials that normally decompose. As he excavated,
Dillehay also uncovered pieces of worked wood (probably used as digging
sticks), wooden mortars and wooden spear tips. He also found the remains
of wooden hut foundations and evidence of plants that were probably used
for food and medicine. All this suggests that while the Clovis community
specialised in big-game hunting with sophisticated stone points, the people
here lived by foraging, in particular for plant foods. Dillehay also discovered
crude flaked tools, suggesting that the Monte Verde community used stone
tools with little or no deliberate shaping.

Guidon’s claims tend to be taken less seriously than those from the
Chilean site. This is caused, in part, by her lack of publication (Dillehay’s
first monograph, published by the Smithsonian Institution in Washington
DC, has already appeared), but it is hard not to suspect that her gender
and nationality may also be significant factors in this attitude.

Guidon is an archaeologist with a solid training in geology. She is
contemptuous of critics who claim that her site’s layers are disturbed by
water action, and that her charcoal is natural rather than produced by people.
However, her main evidence, the 12 000 stone implements found at Pedra Furada,
is ambiguous. The upper layers of the site, dating to 10 000 years ago or
less, contain clearly worked tools made of a flint brought from 16 kilometres
away. But in the lower levels, the contentious ones, there are only quartz
and quartzite pebbles, which eroded out of the cliff tops above. Why didn’t
the earlier occupants seek out and exploit the lint used by the more recent
people? (In the Old World, people were using flint hundreds of thousands
of years ago.) Most of the early ‘pebble tools’ could have been produced
by natural breakage. Guidon’s critics, therefore, dismiss all her early
material.

But Guidon has uncovered other evidence. There are a few pieces from
her most ancient layers, 5 metres down, which were almost certainly produced
by people more than 30 000 years ago, according to Hansjurgen Muller-Beck
of Tubingen University, a leading expert in Ice Age studies, who visited
the site last year. Pedra Furada has also yielded 110 ‘campfires’, concentrations
of charcoal within crude arrangements of stones, which are difficult to
explain away. Guidon has made similar finds of charcoal and tools, with
similar dates, at other shelters in the region showing that Pedra Furada
is not a freak site. At Sitio do Meio, for example, charcoal from a hearth
was dated to about 14 000 years ago. And charcoal taken from Caldeirao dos
Rodrigues was found to date from about 18 000 years ago.

There is also the matter of the rock paintings. Pieces of the sandstone
walls constantly flake off and fall, even today. At Pedra Furada, some of
the pieces which fell in ancient times still have painted figures on them.
So dating the layer onto which they fell provides a minimum age for the
art. Pedra Furada has one slab with a little red human figure from a layer
which dates from at least 10 000 or 12 000 years ago. A slab from a lower
level, dating from 17 000 years ago, bears two red lines which are very
probably from a painted figure.

Most amazing of all is an excavation at a rock shelter called Perna
where, as the occupation layers were stripped away, red figures were exposed
on the walls of the shelter; they had survived burial for thousands of years.
Dates from the layers that masked them indicate that they too must have
been made at least 12 000 years ago. They match those of Pedra Furada in
style, and constitute the oldest known paintings in America.

It now seems likely that America’s first settlers came down the coast
from Siberia in boats, rather than across the land bridge and through an
ice-free corridor in Canada. Recent work in northern Australia indicates
that this voyage would have posed few problems even as far back as 61 000
years ago.

Last year, Richard Roberts of the University of Wollongong, together
with Rhys Jones and M. Smith of the Australian National University, working
at a site called Malakunanja II in the Northern Territory, dated some sandy
deposits associated with stone tools. They used thermoluminescence dating
to find the age of the deposits, a technique which gauges the age of sediments
by measuring the quantity of light released when minerals are heated. The
dates obtained suggested that people first entered Australia 45 000 or even
61 000 years ago. Since the continent has not been joined to Southeast Asia
by a land bridge at any time during the past 3 million years, the shortest
journey would have entailed eight sea voyages, seven of them about 32 kilometres
long, and the final lap to Australia at least 88 kilometres. Clearly, people
at that time were already competent seafarers and possessed sturdy craft.
Their traces in North America were probably limited to the immediate coastal
area, and would have been drowned by the subsequent rise in sea level at
the end of the Ice Age, about 10 000 years ago.

There is a slight chance that evidence may emerge in the north to support
Guidon’s claims. In a small cave on a military reservation in New Mexico’s
Chihuahua desert, Richard ‘Scotty’ MacNeish, an American archaeologist,
claims to be uncovering remains which might rival Pedra Furada in antiquity.
MacNeish is a former bush pilot who studied the origins of agriculture and
maize cultivation in the Tehuacan Valley in Mexico in the 1960s. He has
been excavating at Pendejo cave, whose unfortunate name is a Spanish word
for the male member, and has come across layers containing ash, bones of
extinct animals and stone tools. He also claims to have found a clay-lined
hearth.

MacNeish sent samples of charcoal to the University of California at
Riverside and bone to the Lawrence Livermore Laboratory to be dated. The
lowest levels, well over 24 000 years old according to the first radiocarbon
dates, include many pieces of stone from the cave roof. While most, if not
all, of these pieces are geofacts, MacNeish is convinced that some were
brought in from outside or shaped by human hand. Robson Bonnichsen, who
has visited the site, has expressed cautious interest in one or two pieces.
MacNeish himself believes that his new site counters all previous objections
of critics, but the initial reaction among more conservative archaeologists
is that they have heard this many times before, and sometimes from this
same excavator.

MacNeish continues with his excavation, and his site – being far more
accessible than Guidon’s – will be visited and assessed by the most vehement
critics. It remains to be seen whether Pendejo will prove valid, and become
the site that finally breaks the ‘Pre-Clovis barrier’ in North America.
In the meantime, Brazil is where the dates come from.

Paul Bahn is a freelance writer on archaeology who visited the Piaui
region with the BBC in 1990