Neanderthals coexisted with modern humans for up to 5,000 years

New radiocarbon dates point to longer coexistence but earlier extinction

A new study published in the journal Nature suggests that Neanderthals persisted alongside modern humans in Europe for as long as 5,000 years after the arrival of the latter. A team lead by Tom Higham at Oxford obtained 196 AMS radiocarbon dates from 40 sites across Europe, relying on improved techniques to remove young carbon contamination. The results also indicate that the Neanderthals were probably extinct no later than 41,000 to 39,000 years ago.

Although it was once believed that Neanderthals and modern humans had coexisted for up to 10,000 years, work in the middle of the last decade suggested that the overlap was very brief.  The new results represent a reversion to the earlier position, albeit pushed further back in time since it is now believed that modern humans first reached Europe about 46,000 years ago (Higham’s team suggest the date was around 45,000 years ago).

The new dates may resolve long-running controversy over the Châtelperronian culture, with is believed to be of Neanderthal origin but incorporates elements associated with modern human behaviour. The dates indicate that the Châtelperronian began around 45,000 years ago, suggesting that it was influenced by interaction with modern humans. The Châtelperronian comes to an end at about the same time as the Mousterian, about 41,000 to 39,000 years ago
The researchers were unfortunately unable to obtain any dates for remains from Gorham’s Cave, Gibraltar, where Neanderthal survival after 30,000 years ago has been claimed.

The lengthy overlap rules out the in any case improbable hypothesis that modern humans hunted down and exterminated the Neanderthals. It is more likely that a combination of increased competition for limited resources and the Heinrich Event 4 climatic downturn 40,000 years ago was responsible.

Higham, T. et al., The timing and spatiotemporal patterning of Neanderthal disappearance. Nature 512, 306-309 (2014).

Chauvet cave paintings may be far more recent than generally believed

French archaeologists claim that prehistoric artwork thought to be 36,000 years old is actually 10,000 years younger.

Chauvet Cave is located near the village of Vallon-Pont-d’Arc, Ardèche in southern France. The cave was discovered in 1994 by a team of cavers led by Jean-Marie Chauvet, for whom the site was named. It was the most important cave painting find since the discovery of Lascaux by a group of teenagers during World War II. Unlike the 18,000-year-old Lascaux cave paintings, which became a major tourist attraction after the war and deteriorated badly as a result, Chauvet was rapidly taken over by the French government and a strict conservation program was put in hand.

The artwork comprises 425 panels, depicting rhinoceroses, lions, bears, mammoths, horses, bison, ibex, reindeer, red deer, aurochs, muskoxen, panthers, and the earliest-known representation of an owl turning its head through 180 degrees. Hand prints, red dots and a partial image of a woman associated with a bison have also been discovered.

Radiocarbon dates indicating that the paintings are around 36,000 years old are widely accepted. This would date them to the late Aurignacian period, and make them twice as old as Lascaux. Put another way, the radiocarbon dates suggest that Lascaux is separated from Chauvet by the same interval of time that separates it from the first landing on the Moon. However, archaeologists Jean Combier and Guy Jouve have cast doubt on the great antiquity of the Chauvet paintings.

They argue that on stylistic grounds, the Chauvet artwork cannot be associated with the Aurignacian period. Instead, they claim, the artwork shows affinities to that of the more recent Gravettian and Solutrean periods. Therefore the oldest paintings at Chauvet cannot be more than 26,000 years old. The later ones might even be contemporary with Lascaux.

That Chauvet dates to the Solutrean period was the initial impression of Jean Clottes, one of France’s most eminent prehistorians. Clottes made his assessment in 1995, before any radiocarbon dates were available. His dating of the artwork on purely stylistic grounds was subsequently dismissed as ‘foolhardy’ – but could it be that relying purely on radiocarbon dates is equally unwise?

When first introduced in the 1950s, radiocarbon dating revolutionised archaeology and Willard Libby, the American chemist who pioneered the technique, was awarded the Nobel Prize in Chemistry in 1960. However, radiocarbon dating is not infallible. For example, it is very easy for a sample to become contaminated with more recent organic material that will slew results.

In the case of Chauvet, radiocarbon dates were obtained from wood charcoal used as black pigment. However, Combier and Jouve suggest fossil carbon was used as well as charcoal. This was available at Vagnas, a village not far from Vallon-Pont-d’Arc, where there was a quarry yielding lignite and bitumen. A pigment comprising a mixture of fresh charcoal and fossil carbon would present as being significantly older than one containing fresh charcoal alone.

Combier and Jouve note that such a mixture would also have a different isotopic signature to that of pure wood charcoal, i.e. the proportions of the stable carbon isotopes carbon-12 and carbon-13 would differ between the two. It would thus be possible to show whether or not the Chauvet dates were suspect. Such an anomaly has been detected at another cave site, Candamo Cave in Spain, although in this case the ‘old’ carbon leeched into the pigment from the limestone walls of the cave through the action of bacteria.

Accordingly, Combier and Jouve suggest that fresh radiocarbon dates should be obtained for Chauvet, and they believe that it extremely important that more than one laboratory carries out the work.


1. Clottes, J., Cave Art (Phaidon, New York, 2008).
2. Combier, J. & Jouve, G., Chauvet cave’s art is not Aurignacian: a new examination of the archaeological evidence and dating procedures. Quartär 59, 131-152 (2012).
3. Combier, J. & Jouve, G., Nouvelles recherches sur l’identité culturelle et stylistique de la grotte Chauvet et sur sa datation par la méthode du 14C. L’Anthropologie ( (in press) doi:10.1016/j.anthro.2013.12.001) (2014).
4. Mellars, P., A new radiocarbon revolution and the dispersal of modern humans in Eurasia. Nature 493, 931-935 (2006).

Did Neanderthals die out before modern humans reached southern Europe?

Study casts doubt on late Neanderthal survival in Iberian Peninsula.

Until fairly recently, it was believed that Neanderthals and modern humans coexisted in Europe for up to 10,000 years, but recent improved radiocarbon dates suggest that this period was far shorter – possibly no more than 1,000 or 2,000 years (Mellars, 2006). Many supposedly-late Neanderthals have now been shown to be much older than first believed. For example, two specimens from Vindija Cave in Croatia were originally thought to be from 32,000 to 33,000 years old (28,000 to 29,000 radiocarbon years BP) (Smith, et al., 1999), but these dates are now thought to be nearer 36,000 to 37,000 years old (32,000 to 33,000 radiocarbon years BP) (Higham, et al., 2006). Similarly, an infant from Mezmaiskaya Cave in the northern Caucasus, once believed to be a late survivor from 29,000 years ago, is now believed to be have lived more like 40,000 years ago (Pinhasi, et al., 2011).

Another factor is that calendar dates from this period might have been systematically underestimated. Radiocarbon dates do not coincide exactly with calendar dates, and the latter must be estimated using calibration data. A recent re-evaluation suggests that the estimated calendar dates for this period should be older than was previously believed. The revised dates suggest that overall, the period of coexistence between Neanderthal and modern human populations within the individual regions of Europe such as western France was fairly brief, possibly no more than 1,000 or 2,000 years (Mellars, 2006).

At the peripheries of Europe, Neanderthals might have persisted for rather longer than elsewhere. Possible late survival is documented from two very different settings: Gorham’s Cave, Gibraltar, and Byzovaya, in the western foothills of the northernmost Urals. Gorham’s Cave seems to have been a favoured location that was visited repeatedly over many thousands of years. Natural light penetrates deep into the cave, and a high ceiling permits ventilation of smoke from the hearths that were repeatedly made there. Neanderthal occupation of the cave continued until 33,000 years ago (28,000 radiocarbon years BP), and possibly until as recently as 29,000 years ago (24,000 radiocarbon years BP), and the site was later used by modern humans right up until Phoenician and Carthaginian times. However, there was a 5,000 years hiatus after the last Neanderthal occupation before the first modern humans took up residence (Finlayson, et al., 2006; Finlayson, et al., 2008). At Byzovaya, a total of 313 stone artefacts have been collected over the years, all reflecting typical Middle Palaeolithic tool production techniques characteristic of Neanderthal Mousterian industries, and ranging from 31,000 to 34,000 years old (Slimak, et al., 2011).

However, a newly-published study has cast doubt on the late Neanderthal survival in the Iberian Peninsula. Researchers used a technique known as ultra-filtration to remove traces of modern contaminants (for example preservatives and glues) from fossil bone collagens (proteins making up the bone matrix) prior to radiocarbon dating. Without this process, it is claimed that the contaminants make samples appear younger than they actually are. For example, a carbon contamination of just one percent will make a 50,000-year-old sample appear to be just 37,000 years old. A total of 215 Neanderthal bones from 11 supposedly-late Neanderthal sites were screened for collagen. Unfortunately, only 27 bones were found to contain enough collagen for radiocarbon dating using the ultra-filtration technique. These were recovered from just two sites: Jarama VI and Cueva del Boquet Zafarraya. The results suggested that the Neanderthal remains from the two sites were at least 10,000 years older than previously believed (Wood, et al., 2013). Should other dates for the Iberian Neanderthals turn out to have been similarly understated, then it would suggest that they died out before modern humans arrived. However, it should be noted that the authors of the Gorham’s Cave report had previously considered and ruled out the possibility of contamination affecting their results (Finlayson, et al., 2008).


1. Mellars, P., A new radiocarbon revolution and the dispersal of modern humans in Eurasia. Nature 493, 931-935 (2006).

2. Smith, F., Trinkaus, E., Pettitt, P., Karavanic, I. & Paunovic, M., Direct radiocarbon dates for Vindija G1 and Velika Pecina Late Pleistocene hominid remains. PNAS 96 (22), 12281–12286 (1999).

3. Higham, T., Ramsey, C., Karavanic, I., Smith, F. & Trinkaus, E., Revised direct radiocarbon dating of the Vindija G1 Upper Paleolithic Neandertals. PNAS 103 (3), 553–557 (2006).

4. Pinhasi, R., Higham, T., Golovanova, L. & Doronichev, V., Revised age of late Neanderthal occupation and the end of the Middle Paleolithic in the northern Caucasus. PNAS 108 (21), 8611-8616 (2011).

5. Finlayson, C. et al., Late survival of Neanderthals at the southernmost extreme of Europe. Nature 443, 850-853 (2006).

6. Finlayson, C. et al., Gorham’s Cave, Gibraltar – The persistence of a Neanderthal population. Quaternary International 181, 74-71 (2008).

7. Slimak, L. et al., Late Mousterian Persistence near the Arctic Circle. Science 332, 841-845 (2011).

8. Wood, R. et al., Radiocarbon dating casts doubt on the late chronology of the Middle to Upper Palaeolithic transition in southern Iberia. PNAS 110 (8), 2781-2786 (2013).

Radiometric dating techniques

A major problem for archaeologists and palaeontologists is the reliable determination of the ages of artefacts and fossils.

As far back as the 17th Century the Danish geologist Nicolas Steno proposed the Law of Superimposition for sedimentary rocks, noting that sedimentary layers are deposited in a time sequence, with the oldest at the bottom. Over a hundred years later, the British geologist William Smith noticed that sedimentary rock strata contain fossilised flora and fauna, and that these fossils succeed each other from top to bottom in a consistent order that can be identified over long distances. Thus strata can be identified and dated by their fossil content. This is known as the Principle of Faunal succession. Archaeologists apply a similar principal, artefacts and remains that are buried deeper are usually older.

Such techniques can provide reliably relative dating along the lines of “x is older than y”, but to provide reliable absolute values for the ages of x and y is harder. Before the introduction of radiometric dating in the 1950s dating was a rather haphazard affair involving assumptions about the diffusion of ideas and artefacts from centres of civilization where written records were kept and reasonably accurate dates were known. For example, it was assumed – quite incorrectly as it later turned out – that Stonehenge was more recent than the great civilization of Mycenaean Greece.

The idea behind radiometric dating is fairly straightforward. The atoms of which ordinary matter is composed each comprise a positively charged nucleus surrounded by a cloud of negatively charged electrons. The nucleus itself is made up of a mixture of positively charged protons and neutral neutrons. The atomic weight is total number of protons plus neutrons in the nucleus and the atomic number is the number of protons only. The atom as a whole has the same number of electrons as it does protons, and is thus electrically neutral. It is the number of electrons (and hence the atomic number) that dictate the chemical properties of an atom and all atoms of a particular chemical element have the same atomic number, thus for example all carbon atom have an atomic number of six. However the atomic weight is not fixed for atoms of a particular element, i.e. the number of neutrons they have can vary. For example carbon can have 6, 7 or 8 neutrons and carbon atoms with atomic weights of 12, 13 and 14 can exist. Such “varieties” are known as isotopes.

The physical and chemical properties of various isotopes of a given element vary only very slightly but the nuclear properties can vary dramatically. For example naturally-occurring uranium is comprised largely of U-238 with only a very small proportion of U-235. It is only the latter type that can be used as a nuclear fuel – or to make bombs. Many elements have some unstable or radioactive isotopes. Atoms of an unstable isotope will over time decay into “daughter products” by internal nuclear change, usually involving the emission of charged particles. For a given radioisotope, this decay takes place at a consistent rate which means that the time taken for half the atoms in a sample to decay – the so called half-life – is fixed for that radioisotope. If an initial sample is 100 grams, then after one half-life there will only be 50 grams left, after two half-lives have elapsed only 25 grams will remain, and so on.

It is upon this principle that radiometric dating is based. Suppose a particular mineral contains an element x which has a number of isotopes, one of which is radioactive and decays to element y with a half-life of t. The mineral when formed does not contain any element y, but as time goes by more and more y will be formed by decay of the radioisotope of x. Analysis of a sample of the mineral for the amount of y contained will enable its age to be determined provided the half-life t and isotopic abundance of the radioisotope is known.

The best-known form of radiometric dating is that involving radiocarbon, or C-14. Carbon – as noted above – has three isotopes. C-12 (the most common form) and C-13 are stable, but C-14 is radioactive, with a half-life of 5730 years, decaying to N-14 (an isotope of nitrogen) and releasing an electron in the process (a process known as beta decay). This is an infinitesimal length of time in comparison to the age of the Earth and one might have expected all the C-14 to have long since decayed. In fact the terrestrial supply is constantly being replenished from the action of interstellar cosmic rays upon the upper atmosphere where moderately energetic neutrons interact with atmospheric nitrogen to produce C-14 and hydrogen. Consequently all atmospheric carbon dioxide (CO2) contains a very small but measurable percentage of C-14 atoms.

The significance of this is that all living organisms absorb this carbon either directly (as plants photosynthesising) or indirectly (as animals feeding on the plants). The percentage of C-14 out of all the carbon atoms in a living organism will be the same as that in the Earth’s atmosphere. The C-14 atoms it contains are decaying all the time, but these are replenished for as long as the organism lives and continues to absorb carbon. But when it dies it stops absorbing carbon, the replenishment ceases and the percentage of C-14 it contains begins to fall. By determining the percentage of C-14 in human or animal remains or indeed anything containing once-living material, such as wood, and comparing this to the atmospheric percentage, the time since death occurred can be established.

This technique was developed by Willard Libby in 1949 and revolutionised archaeology, earning Libby the Nobel Prize for Chemistry in 1960. The technique does however have its limitations. Firstly it can only be used for human, animal or plant remains – the ages of tools and other artefacts can only be inferred from datable remains, if any, in the same context. The second is that it only has a limited “range”. Beyond 60,000 years (10 half-lives) the percentage of C-14 remaining is too small to be measured, so the technique cannot be used much further back than the late Middle Palaeolithic. Another problem is the cosmic ray flux that produces C-14 in the upper atmosphere is not constant as was once believed. Variations have to be compensated for by calibration curves, based on samples that have an age that can be attested by independent means such as dendochronology (counting tree-rings). Finally great care must be taken to avoid any contamination of the sample in question with later material as this will introduce errors.

The conventions for quoting dates obtained by radiocarbon dating are a source of considerable confusion. They are generally quoted as Before Present (BP) but “present” in this case is taken to be 1950. Calibrated dates can be quoted, but quite often a quoted date will be left uncalibrated. Uncalibrated dates are given in “radiocarbon years” BP. Calibrated dates are usually suffixed (cal), but “present” is still taken to be 1950. To add to the confusion, Libby’s original value for the half-life of C-14 was later found to be out by 162 years. Libby’s value of 5568 years, now known as the “Libby half-life”, is rather lower than the currently-accepted value of 5730 years, which is known as the Cambridge half-life. Laboratories, however, continue to use the Libby half-life! In fact this does make sense because by quoting all raw uncalibrated data to a consistent standard means any uncalibrated radiocarbon date in the literature can be converted to a calibrated date by applying the same set of calculations. Furthermore the quoted dates are “futureproofed” against any further revision of the C-14 half-life or refinement of the calibration curves.

If one needs to go back further than 60,000 years other techniques must be used. One is Potassium-Argon dating, which relies on the decay of radioactive potassium (K-40) to Ar-40. Due to the long half-life of K-40, the technique is only useful for dating minerals and rocks that are over than 100,000 years old. It has been used to bracket the age of archaeological deposits at Olduvai Gorge and other east African sites with a history of volcanic activity by dating lava flows above and below the deposits.

© Christopher Seddon 2008