European Middle and Upper Palaeolithic radiocarbon dates are often older than they look: problems with previous dates and some remedies.

Higham, Thomas

Introduction

The European Middle to Upper Palaeolithic is widely accepted as being the transitional period over which the final Neanderthals became extinct and were replaced by anatomically modern humans (Mellars 1989, 1999; Zilhao & D'Errico 1999; Zilhao 2006 and references therein). Questions of major importance arise. Did Neanderthals independently develop symbolic and adaptive behaviours before the ~36 500 [sup.14]C BP arrival of modern humans in Western Europe? Did they copy incoming modern human behaviour ('acculturation') whilst the two co-existed in the few millennia prior to Neanderthal extinction? Where, when and how did Neanderthals become extinct and how did their extinction relate to the spatial dispersai of modern human populations? Did Neanderthals and modern humans meet, were they contemporaries in some or all parts of Europe, and did they mate?

The resolution of many of these questions awaits an unambiguous and reliable chronology, bur for 50 years, this has been unattainable. Radiocarbon dates exhibit an asymptotic tendency as they approach the measurement limit, which is itself determined in individual laboratories by careful repeat radiocarbon measurements of material known to be beyond the reach of the technique, i.e. greater than about 60 000 or 70 000 years old (Chappell et al. 1996). Subsequent radiocarbon measurements can never be older than this, since if a radiocarbon date is within two standard deviations of the background value, a 'greater than' age is calculated.

Several scholars have attempted to explore the issues using databanks of radiocarbon dates obtained from archaeological sites (e.g. Bocquet-Appel & Demars 2000, bur see Pettitt & Pike 2001; Joris et al. 2003; van Andel et al. 2003; Dolukhanov & Shukurov 2004; Gamble et al. 2004, 2005; Kuzmin & Keates 2005) and received guidance on how best to view the many available dates (Pettitt et al. 2003). But research undertaken by the Oxford Radiocarbon Accelerator Unit (ORAU) over the last decade throws doubt on the validity of many of these published dates. The principal areas of error lie in the selection of inappropriate samples and the effects of inadequate pre-treatment to remove contamination. This paper addresses some of the problems and illustrates recent progress and a way forward.

Sample selection

It is clear that many samples dated in the past have failed to pass the basic test of bearing witness to the presence of humans. Types of useful samples include cut-marked or humanly-modified bones, human remains, organic artefacts or charcoal from clearly identified features, such as hearths. It is axiomatic that all of these samples must come from secure archaeological contexts, bur even if they do not, they have the distinct advantage over other types of nondescript bone of still being able to provide information regarding the age of human presence. Samples of bone without cut marks ought to be avoided because they may be deposited by animals such as hyena that regularly use the same types of cave and rockshelter sites frequented by humans. Such problems have been particularly acute in the British Paiaeolithic (Jacobi et al. 2006), but are almost certainly more widely applicable. In some contexts it is exceedingly difficult to identify cut marks, since surface etching, degradation, overprinting by animals and deposits of carbonate and sediment can obscure them. Without careful sample selection, no amount of improved pre-treatment chemistry will result in accurate ages.

Contamination

Samples of Middle/Upper Palaeolithic age are particularly susceptible to errors from contamination. Whilst laboratory and chemical background is quantifiable, however, trace contamination derived from archaeological contexts rarely is. It is this material that may be responsible for producing finite dates for samples that are, in reality, older than the [sup.14]C limit. Historically, this has left us a situation in which we now have many radiocarbon ages that are artificially younger than they ought to be.

In some rare instances, it is possible to recognise the scale of error. The Spanish site of El Sidron is important because it contains one of the largest groups of Neanderthal bones ever found: the remains of nine individuals (Fortea et al. 2003). It is interesting for dating because it is likely that the individuals in the site died at the same time. Their bones are covered in innumerable cut marks (Fortea et al. 2003) and some of the associated lithics could be refitted (Santamaria et al. 2010). A series of direct radiocarbon dates from the site are shown in Figure 1 (after De Torres et al. 2010). A wide range of dates is apparent, from 49 000--48 000 BP to 11 000-10 000 BP. OSL dates bracketing the human remains suggest that they were deposited between 47-30 ka cal BP. This suggests, therefore, a significant problem with the radiocarbon ages, particularly the youngest samples. Sensibly, De Torres et al. (2010) consider the youngest ages to be severely underestimating the real age and blame unremoved contamination.

[FIGURE 1 OMITTED]

A second example comes from work on Neanderthal DNA by Krause et al. (2007), in which radiocarbon dates were obtained from a single Neanderthal sub-adult humerus from Okladnikov Cave in the Altai Mountains of Siberia. Despite very well-preserved bone being dated, the results from three laboratories showed great variation, from 29 990 [+ or -] 500 BP (KIA-27011), to 34 860 [+ or -] 360 BP (Beta-186881) and 37 800 [+ or -] 450 BP (OXA-15481). Krause et al. (2007) lend more credence to the Oxford determination than the others, but chose to average the results (to 34 190 [+ or -] 760 BP) to estimate the age of the specimen. It is highly doubtful whether this is a meaningful value because not all of these dates can possibly be accurate.

These results do not inspire confidence, but why is there such variation? Although ancient carbon contaminants are of concern, their presence is virtually insignificant compared with the equivalent proportion of modern contamination. The addition of l0 per cent old carbon to a sample dating to 40 000 BP would yield a date of 40 800 BP, only marginally too old even accounting for the sometimes large error terms on measurements of this age. Even 20 per cent old carbon contamination would only give an age of 41 600 BP. And these are contamination levels that radiocarbon specialists would consider unusually high.

The effect of more recent contamination is dramatically different; add 10 per cent modern carbon to the sample and the age is 18 000 BP. Even 0.5 per cent of modern contamination gives a wholly distorted age (35 600 BP). Therefore, in assessing the reliability of radiocarbon ages from the Palaeolithic, it is usual to consider older ages as being more likely to be closer to the 'true' age than younger ones. The younger dates from El Sidron or Ohkladnikov mentioned above, therefore, are almost certainly erroneous and the reason is probably due to non-autochthonous carbon.

Remedial measures--bone

About half of the available corpus of [sup.14]C dates from the European Palaeolithic are dates of bone, the material targeted for dating being almost always the protein collagen, which has a well-defined structure. Most facilities pre-treat bone in a similar manner using a two-step process, the first removes the mineral fraction (hydroxyapatite, about 80 per cent of the bone by weight), and the second stage gelatinises the bone collagen using a technique first applied by Longin (1971). This process brings the collagen into solution and separates it from insoluble residues.

Since 2000, the ORAU has applied an additional ultrafiltration step to the process, based on Brown et al. (1988). An ultrafilter is a type of molecular sieve, which separates protein chains above and below a certain molecular weight (MW). Since an undegraded polypeptide chain from a collagen molecule weighs 110kD (a Dalton is defined as 1/12th of the mass of one [sup.12]C atom) we use a Vivaspin[TM] 30kD ultrafilter that traps these larger particles above the filter, and separates lower MW components below it using centrifugation (Bronk Ramsey et al. 2004). One would expect this to result in a better quality of collagen, since degraded amino acids, salts, insoluble particles and lower MW contaminants pass through the filter.

Earlier work by our group (Bronk Ramsey et al. 2004; Higham et al. 2006a & b; Jacobi et al. 2006) has shown that the use of ultrafiltration often (but not always) results in older ages for bones than those previously obtained. This suggests that pre-treatment is a very important influence on accuracy and that previous determinations, less rigorously pretreated, can sometimes be affected by residual contamination. Confidence in this diagnosis is increased because the samples we tested were dated in the same laboratory, in which similar protocols operated. The differences can, therefore, be ascribed largely to pre-treatment variation.

By way of example, dates from a bone or antler point excavated at the British site of Hyaena Den (Wookey Hole, Somerset) are given (Table 1). The first date obtained was prepared in 1991 using an ion-exchanged gelatin method to extract collagen (denoted as code AI). The result was 24 600 [+ or -] 300 BP. In 2004, two other dates were obtained, using the gelatinisation (AG), and the ultrafiltration (AF) methods, respectively. The AF method produced the oldest and probably more correct date of the three. Other examples obtained in the last 5-10 years show similar patterns (see Higham et al. 2006a; Jacobi et al. 2006), and these downward revisions appear to make far more archaeological sense. A bone or antler point in the British Isles at 24 600 BP appeared an unlikely proposition owing to the extremely cold climate of the time. The new earlier result fits neatly alongside other evidence for osseous point manufacture during the Evolved Aurignacian of Europe.

Another example is provided by the Palaeolithic site of the Geissenklosterle, in the Swabian Jura of Germany, which is important for the chronology and dispersai of the first anatomically modern humans (represented by the Aurignacian) in Western Europe (Hahn 1988; Conard & Bolus 2003, 2008). Conard and Bolus (2003, 2008) have published radiocarbon results from the site that show a wide variation and are inconsistent with the stratigraphic sequence. Two different explanations have been proposed to account for this. First, that the results are caused by mixing and movement of material between the principal Aurignacian horizons (AHII and AHIII) (Zilhao & D'Errico 1999), which themselves are an amalgamation of what may be a series of discrete, brief occupations (Teyssandier et al. 2006). Second, that the variation in date is due to significant variations in the production of radiocarbon, as attested to in the datasets obtained by Beck et al. (2001) and Voelker et al. (2000) through this period (see Conard & Bolus 2003, 2008 who term this the 'Middle Palaeolithic Dating Anomaly').

It is important to say immediately that the latter explanation can essentially be addressed in the light of more recent work on the calibration of radiocarbon between ~25 000-55 000 cal BP, which shows that the large discrepancies in [sup.14]C production cannot be duplicated (Hoffman et al. 2010), and that the Beck et al. (2001) dataset was affected by a dead carbon influence. Insofar as the first explanation is concerned, we decided to test the reliability of the previous corpus of dates by redating using ultrafiltration. The results are shown in Table 2. For each sample of animal bone, there are two radiocarbon determinations. The first radiocarbon determination in each case was obtained in the 1990s, mostly using gelatinisation. The second is a new ultrafiltered gelatin date from the same bone. With some exceptions, the differences between the two are, once again, often dramatic, with the original dates underestimating the real age. This suggests strongly that the reason for the variation in the original radiocarbon series is not due to changes in radiocarbon production, as previously mooted, but lies instead in a combination of factors amongst which contamination must be considered alongside taphonomic or post-depositional influences.

The next challenge is to demonstrate that these newer results are accurate and not themselves underestimating the true age. In the absence of ancient bone of known age, radiocarbon laboratories attempt to quantify the background limit for the small traces of [sup.14]C that are inevitably taken up during the processing of the material. During the combustion stage, for instance, we estimate that 0.0007 [+ or -] 0.0010mg modern carbon is added. This figure, based on repeat combustions and measurements of radioactively 'dead' material, is subtracted from the final radiocarbon measurement. In addition, small amounts of modern radiocarbon are also taken up during the several steps of the chemical pre-treatment of the bone, which must also be accounted for. To quantify this, the ORAU use a >70 000 BP bison bone, from the Ash Bend site in Alaska (Brock et al. 2010), analysed multiple times to produce a background correction subtracted from all bone dated in the laboratory. These measurements enable us to determine the maximum dateable age for bone at the ORAU of just under 50 000 BP (Wood et al. 2010). It follows that a failure to account for this in the radiocarbon dating of old bone would result in ages that are, once again, too young by varying degrees.

A measure of the methodological improvement evident can be seen when we redate bones suspected of being beyond the [sup.14]C limit. In Table 3, the results of dates of bones from the so-called 'Banwell fauna' are shown. This fauna has been suggested by Currant and Jacobi (1997, 2001) as most likely to date to Marine Isotope Stage 5, specifically MIS5b or a (Gilmour et al. 2007), and therefore to be well beyond the radiocarbon limit. The fauna is a cool climate, low species diversity fauna dominated by bison (Bison priscus) and reindeer (Rangifer tarandus). Some of the results obtained in Oxford in the 1990s, however, were younger than ~40-45 ka BE Table 3 also shows results of the same bones redated using ultrafiltration which are almost always 'greater than' ages and/or substantially older than previously. This suggests that the age of the fauna is greater than ~50 ka BP, and it also shows the previous ages to be underestimates again. Once more, an ultrafiltration treatment appears to remove the types of contaminants that appear not to have been removed before and confirm an older, beyond background age, for this fauna.

Redating charcoal

Radiocarbon dates of charcoal from Palaeolithic sites, previously assumed to be largely reliable (Joris et al. 2003), have also seen some important developments in pre-treatment chemistry that hint at potential problems with the previous corpus of dates. The Grotta di Fumane provides an example. This site lies at the southern fringe of the Venetian Pre-Alps within a complex karst system comprising a number of cavities containing a sedimentary succession over 10m thick that spans the Middle to Upper Palaeolithic, from MIS5-2 (Peresani et al. 2008). Previous dates were extremely variable and disclosed little age-depth patterning. Once again, some have suggested that the reason for this may be due to larger than expected variations in the production of [sup.14]C in the atmosphere (Giaccio et al. 2006). To test this, we decided to redate previously analysed samples using a much more rigorous technique.

The routine pre-treatment for charcoal samples is the so-called acid-base-acid protocol (ABA). This method removes carbonates and humic complexes using acidic and basic solutions respectively. Evidence suggests that this is adequate in the majority of cases, particularly for samples less than about 20 ka BE However, when dating older material, several workers have shown that it is not always effective enough and contamination may remain (Chappell et al. 1996; Bird et al. 1999; Turney et al. 2001; Santos et al. 2003; Higham et al. 2008, 2009; Brock & Higham 2009; Douka et al. 2010). In Table 4, samples of charcoal previously dated at the ORAU using the ABA method are compared with those dated using a more rigorous technique called ABOx-SC (acid-base-oxidation: stepped combustion) developed by Bird et al. (1999). The ABOx-SC results are uniformly older, and almost certainly more reliable, than the ABA dates. When the dates are analysed within a Bayesian model calibrated against INTCAL09 (Reimer et al. 2009) they appear to exhibit much greater age-depth consistency (Higham et al. 2009). Samples from the same phase yield closely similar ages. Results from the A2 level of the site, which is the Proto-Aurignacian, all cluster around ~40 000 cal BE Interestingly, in the south of the Italian peninsula, the independently-dated Campanian Ignimbrite (dated at ~39 300 cal BP by Ar-Ar) seals identical cultural horizons beneath it (De Vivo et al. 2001; Giaccio et al. 2006). This suggests that the A2 ages ought to be older than this, and indeed they are (Higham et al. 2009).

[FIGURE 2 OMITTED]

We also obtained new ultrafiltered bone dates from the site and these can be compared with the charcoal ABOx-SC series. The results are shown in a Bayesian model in Figure 2. The overall level of agreement is excellent and there is only one outlier in the assemblage (this result is considered more likely to be caused by sedimentary and cryoturbation processes in the upper parts of the site). Taken together, the level of agreement suggests that the chronometric framework is strong and the results reliable. This good news is tempered by the fact that while we see good agreement now, a comparison with previous published ages (Peresani et al. 2008) shows that around 70 per cent of the published ages appear to be erroneous.

Our experience suggests that one useful indicator of potential problems with bone determinations is collagen yield. When these fall below ~0.5-1 per cent by weight, CN atomic ratios often increase and amino acid profiles indicate greater instability, implying the potential for problematic AMS determinations (e.g. Ambrose 1990; van Klinken 1999). Radiocarbon laboratories ought to fall these types of bone routinely when using gelatinisation pre-treatments. For charcoal, the analytical indicators are much less apparent, but a decline in carbon yield, indicating degradation, is one useful measure and often, when dating old charcoals, low carbon yields are associated with anomalous, younger results.

Conclusions

The tendency of radiocarbon dates to cleave asymptotically to the dating limit means that many European Middle to Upper Palaeolithic results produced over the last 50 years are underestimates of their real age. Sometimes these underestimates are severe. This means that there are significant problems with the current database of results and it is very difficult to be certain that some dates are not prone to undetected error.

Radiocarbon dating has not achieved its potential principally because of limitations in the pre-treatment of samples for dating and an unwise selection of material. At the Fumane site, our analysis suggested that more than 70 per cent of the previous determinations obtained were underestimates of the real age (Higham et al. 2009). What was surprising was that these determinations were mostly obtained in the last 10 years. If this estimate is applied to the entire European database, it suggests that there is a great deal of work to be done to build new chronologies in which we can be confident. The results in this paper show that the problems with decontamination of samples from the Palaeolithic are significant, but can be overcome.

Acknowledgements

This paper is dedicated to the late Dr Roger Jacobi (the British Museum and the Natural History Museum, London) for his collaboration, his comments on an early manuscript and for encouraging me to write this in the first place. The Fumane project is undertaken in collaboration with M. Peresani & A. Broglio (University of Ferrara, Italy), E Brock, R. Wood and K. Douka (all ORAU). I am very grateful to Marco Peresani for his collaboration on the site and our dating work. Fumane dates were funded by a grant to E Brock and T. Higham from the NERC-AHRC NRCF programme. The excavation is managed by the Ferrara and Milano I Universities in the framework of a project supported by the Soprintendenza per i Beni Archeologici del Veneto, CA.RI. Verona Foundation, Comunita Montana della Lessinia, Comune di Fumane. I am grateful to Fiona Brock for her invaluable contribution to this dating work. I also thank Marco de la Rasillas (Oviedo) for comments and suggestions regarding the El Sidron sequence of dates. The samples from the Geissenklosterle were sampled and prepared for AMS dating by Rachei Wood and I am grateful for her invaluable input on this and for her comments on the paper. She is also responsible for the work on the background correction for bone at the ORAU. I am also grateful to Katerina Douka for her comments on an earlier draft. I thank the staff of the ORAU, University of Oxford and the team working on the NERC (funded by grant NE/D014077/1) radiocarbon project at the ORAU, including C. Bronk Ramsey, R. Wood, K. Douka, L. Basell, J. Davies, A. Bowles, B. Emery, M. Humm, W. Davies and P. Leach. I am very grateful to P. Pettitt, J. Zilhao and W. Davies for their comments.

Received: 5 May 2010; Accepted: 20 October 2010; Revised 9 November

References

AMBROSE, S.H. 1990. Preparation and characterisation of bone and tooth collagen for isotopic analysis. Journal of Archaeological Science 17: 431-51. ANDERSEN, K.K., A. SVENSSON, S.J. JOHNSEN, S.O. RASMUSSEN, M. BIGLER, R. ROTHLISBERGER, U. RUTH, M.L. SIGGAARD-ANDERSEN, J.-P. STEFFENSEN & D. DAHL-JENSEN. 2006. The Greenland ice core chronology 2005, 15-42 ka, Part 1 : constructing the time scale. Quaternary Science Reviews 25: 3246-57.

BECK, J.W., D.A. RICHARDS, R.L. EDWARDS, B.W. SILVERMAN, EL. SMART, D.J. DONAHUE, S. HERERRA-OSTERHELD, G.S. BURR, L. CALSOYAS, A.J.T. JULL & D. BIDDULPH. 2001. Extremely large variations of atmospheric [sup.14]C concentration during the last glacial period. Science 292: 2453-8. BIRD, M.I., L.K. AYLIFFE, L.K. FIFIELD, C.S.M. TURNEV, R.G. CRESSWELL, T.T. BARROWS & B. DAVID. 1999. Radiocarbon dating of 'old' charcoal using a wet oxidation-stepped combustion procedure. Radiocarbon 41 : 127-40.

BLAAUW, M., B. WOHLFARTH, J. ANDRES CHRISTEN, L. AMPEL, D. VERES, K.A. HUGHEN, F. PREUSSER & A. SVENSSON. 2010. Were last glacial climate events simultaneous between Greenland and France? A quantitative comparison using non-tuned chronologies. Journal of Quaternary Science 25: 387-94.

BOCQUET-APPEL, P. & P.Y. DEMARS. 2000. Neanderthal contraction and modern human colonisation in Europe. Antiquity 74: 544-52.

BROCK, F. & T.F.G. HIGHAM. 2009. AMS radiocarbon dating of Paleolithic-aged charcoal from Europe and the Mediterranean rim using ABOx-SC. Radiocarbon 51: 839-46.

BROCK, F., T.F.G. HIGHAM, P.. DITCHFIELD & C. BRONK RAMSEY. 2010. Current pre-treatment methods for AMS radiocarbon dating at the Oxford Radiocarbon Accelerator Unit (ORAU). Radiocarbon 52: 103-12.

BRONK RAMSEY, C. 2001. Development of the radiocarbon calibration program OxCal. Radiocarbon 43: 355-63.

--2009. Dealing with outliers and offsets in radiocarbon dating. Radiocarbon 51: 1023-45.

BRONK RAMSEY, C., T. HIGHAM, A. BOWLES & R.E.M. HEDGES. 2004. Improvements to the pre-treatment of bone at Oxford. Radiocarbon 46: 155-63.

BROWN, T.A., D.E. NELSON, J.S. VOGEL & J.R. SOUTHON. 1988. Improved collagen extraction by modified Longin method. Radiocarbon 30: 171-7.

CHAPPELL, J., J. HEAD & J. MAGEE. 1996. Beyond the radiocarbon limit in Australian archaeology and Quaternary research. Antiquity 70: 543-52.

CONARD, N.J. & M. BOLUS. 2003. Radiocarbon dating the appearance of modern humans and timing of cultural innovations in Europe: new results and new challenges. Journal of Human Evolution 44: 331-71.

--2008. Radiocarbon dating the late Middle Paleolithic and the Aurignacian of the Swabian Jura. Journal of Human Evolution 55: 886-97.

CURRANT, A. & R. JACOBI. 1997. Vertebrate faunas of the British Late Pleistocene and the chronology of human settlement. Quaternary Newsletter 82: 1-8.

--2001. A formal mammalian biostratigraphy for the Late Pleistocene of Britain. Quaternary Science Reviews 20: 1707-16.

DE TORRES, T., J.E. ORTIZ, R. GRUN, S. EGGINS, H. VALLADAS, N. MERCIER, N. TISNERAT-LABORDE, R. JULIA, V. SOLER, E. MARTINEZ, S. SANCHEZ-MORAL, J.C. CANAVERAS, J. LARIO, E. BADAL, C. LALUEZA-FOX, A. ROSAS, D. SANTAMARIA, M. DE LA RASILLA & J. FORTEA. 2010. Dating of the hominid (Homo neanderthalensis) remains accumulation from El Sidron Cave (Pilona, Asturias, north Spain): an example of a multi-methodological approach to the dating of Upper Pleistocene sites. Archaeometry 52: 680-705. doi: 10.1111/j.1475-4754.2009.00491.

DE VIVO, B., G. ROLANDI, P.B. GANS, A. CALVERT, W.A. BOHRSON, F.J. SPERA & H.E. BELKIN. 2001. New constraints on the pyroclastic eruptive history of the Campanian volcanic plain (Italy). Mineralogy and Petrology 73: 47-65.

DOLUKHANOV, P. & A. SHUKUROV. 2004. Colonisation of northern Eurasia by early modern humans as viewed through the evidence of radiocarbon dating, in T.F.G. Higham, C. Bronk Ramsey & D.C. Owen (ed.) Radiocarbon and archaeology: fourth international symposium, St. Catherine's College, Oxford, 9-14 April 2002: conference proceedings: 53-61. Oxford: Oxford University School of Archaeology.

DOUKA K., T. HIGHAM & A. SINITSYN. 2010. The influence of pre-treatment chemistry on the radiocarbon dating of Campanian Ignimbrite-aged charcoal from Kostenki 14 (Russia). Quaternary Research: 73: 583-7. doi:10.1016/j.yqres.2010.01.002

FORTEA, J., M. DE LA RASILLA, E. MARTINEZ, S. SANCHEZ-MORAL, J.C. CANAVERAS, S. CUEZVA, A. ROSAS, V. SOLER, R. JULIA, T. DE TORRES, J.E. ORTIZ, J. CASTRO, E. BADAL, J. ALTUNA & J. ALONSO. 2003. Sidron's Cave (Borines, Pilona, Asturias): initial results. Estudios Geologicos 59: 159-79.

GAMBLE, C., W. DAVIES, P. PETTITT & M. RICHARDS. 2004. Climate change and evolving human diversity in Europe during the last glacial. Philosophical Transactions of the Royal Society of London, Series B 359: 243-54.

GAMBLE, C., W. DAVIES, P. PETTITT, L. HAZELWOOD & M. RICHARDS. 2005. The archaeological and genetic foundations of the European population during the late glacial: implications for 'Agricultural Thinking'. Cambridge Archaeological Journal 15: 193-223.

GIACCIO, B., I. HAJDAS, M. PERESANI, F.G. FEDELE & R. ISAIA. 2006. The Campanian Ignimbrite tephra and its relevance for the timing of the Middle to Upper Palaeolithic shift, in N.J. Conard (ed.) When Neanderthals and modern humans met: 343-75. Tubingen: Kerns.

GILMOUR, M., A. CURRANT, R.M. JACOBI & C.B. STRINGER. 2007. Recent TIMS dating results from British Late Pleistocene vertebrate faunal localities: context and interpretation. Journal of Quaternary Science 22: 793-800.

HAHN, J. 1988. Die Geissenkloterle-Hohle im Achtal bei Blaubeuren I. Fundhorizontbildung und Besiedlung im Mittelpaleolithikum und im Aurignacien. Stuttgart: Konrad Theiss.

HIGHAM, T.F.G., R.M. JACOBI & C.B. RAMSEY. 2006a. AMS radiocarbon dating of ancient bone using ultrafiltration. Radiocarbon 48(2): 179-95.

HIGHAM, T., C.B. RAMSEY, I. KARAVANIC, F.H. SMITH & E. TRINKAUS. 2006b. Revised direct radiocarbon dating of the Vindija G1 Upper Paleolithic Neandertals. Proceedings of the National Academy of Sciences of the United States of America 103: 553-7.

HIGHAM, T.F.G., H. BARTON, C.M.T. TURNEY, G. BARKER, C. BRONK RAMSEY & F. BROCK. 2008. Radiocarbon dating of charcoal from tropical sequences: results from the Niah Great Cave, Sarawak and their broader implications. Journal of Quaternary Science 24: 189-97.

HIGHAM, T.F.G., F. BROCK, M. PERESANI, A. BROGLIO, R. WOOD & K. DOUKA. 2009. Problems with radiocarbon dating the Middle to Upper Palaeolithic transition in Italy. Quaternary Science Reviews 28: 1257-67.

HOFFMANN, D.L., J.W. BECK, D.A. RICHARDS, P.L. SMART, J.S. SINGARAYER, T. KETCHMARK & C.J. HAWKESWORTH. 2010. Towards radiocarbon calibration beyond 28 ka using speleothems from the Bahamas. Earth and Planetary Science Letters 289: 1-10.

JACOBI, R.M., T.F.G. HIGHAM & C. BRONK RAMSEY. 2006. AMS radiocarbon dating of Middle and Upper Palaeolithic bone in the British Isles: improved reliability using ultrafiltration. Journal of Quaternary Science 21: 557-73.

JORIS O., E. ALVAREZ FERNANDEZ & B. WENINGER. 2003. Radiocarbon evidence of the Middle to Upper Palaeolithic transition in south-western Europe. Trabajos de Prehistoria 60(2): 15-38.

KRAUSE, J., L. ORLANDO, D. SERRE, B. VIOLA, K. PRUFER, M.P. RICHARDS, J.-J. HUBLIN, C. HANNI, A.P. DEREVIANKO & S. PAABO. 2007. Neanderthals in central Asia and Siberia. Nature 449(7164): 902-904.

KUZMIN, Y.V. & S.G. KEATES. 2005. Dates are not just data: Paleolithic settlement patterns in Siberia derived from radiocarbon records. American Antiquity 70: 773-89.

LONGIN, R. 1971. New method of collagen extraction for radiocarbon dating. Nature 230(5291): 241-2.

MELLARS, P. 1989. Major issues in the emergence of modern humans. Current Anthropology 30: 349-85.--1999. The Neanderthal problem continued. Current Anthropology 40: 341-64.

PERESANI, M., M. CREMASCHI, F. FERRARO, C. FALGUERES, J.-BAHAIN, G. GRUPPIONI, E. SIBILIA, G. QUARTA, L. CALCAGNILE & J.-M. DOLO. 2008. Age of the final Middle Palaeolithic and Uluzzian levels at Fumane Cave, northern Italy, using [sup.14]C, ESR, [sup.234]U/[sup.230]Th and thermoluminescence methods. Journal of Arehaeological Science 35: 2986-96.

PETTITT, P.B. & A.W.G. PIKE. 2001. Blind in a cloud of data: problems with the chronology of Neanderthal extinction and anatomically modern human expansion. Antiquity 75: 415-20.

PETTITT, P.B., W. DAVIES, C.S. GAMBLE & M.P. RICHARDS. 2003. Palaeolithic radiocarbon chronology: quantifying our confidence beyond two half-lives. Journal of Archaealogical Science 30: 1685-93.

REIMER, P.J., M.G.L. BAILLIE, E. BARD, A. BAYLISS, J.W. BECK, P.G. BLACKWELL, C. BRONK RAMSEY, C.E. BUCK, G. BURR, R.L. EDWARDS, M. FRIEDRICH, P.M. GROOTES, T.P. GUILDERSON, I. HAJDAS, T.J. HEATON, A.G. HOGG, K.A. HUGHEN, K.F. KAISER, B. KROMER, F.G. MCCORMAC, S.W. MANNING, R.W. REIMER, D.A. RICHARDS, J.R. SOUTHON, S. TALAMO, C.S.M. TURNEY, J. VAN DER PLICHT & C.E. WEYHENMEYER. 2009. Intcal09 and Marine09 radiocarbon age calibration curves, 0-50 000 years cal BE Radiocarbon 51: 1111-50.

SANTAMARIA, D., J. FORTEA, M. DE LA RASILLA, L. MARTINEZ, E. MARTINEZ, J.C. CANAVERAS, S. SANCHEZ-MORAL, A. ROSAS, A. ESTALRRICH, A. GARCIA-TABERNERO & C. LALUEZA-Fox. 2010. The technological and typological behaviour of a Neanderthal group from El Sidron Cave (Asturias, Spain). Oxford Journal of Archaeology 29: 119-48.

SANTOS, G.M., M.I. BIRD, L.K. FIFIELD, F. PARENTI, N. GUIDON & P.A. HAUSLADEN. 2003. The controversial antiquity of the peopling of the Americas: a review of the chronology of the lowest occupation layer in the Pedra Furada rockshelter, Piaui, Brazil. Quaternary Science Reviews 22: 2303-2310.

SVENSSON, A., K.K. ANDERSEN, M. BIGLER, H.B. CLAUSEN, D. DAHL-JENSEN, S.M. DaVIES, S.J. JOHNSEN, R. MUSCHELER, S.O. RASMUSSEN & R. ROTHLISBERGER. 2006. The Greenland ice core chronology 2005, 15-42 ka, Part 2: comparison to other records. Quaternary Science Reviews 25: 3258-67.

TEYSSANDIER, N., M. BOLUS & N.J. CONARD. 2006. The Early Aurignacian in central Europe and its place in a European perspective, in O. Bar-Yosef & J. Zilhao (ed.) Towards a definition of the Aurignacian: 241-56. Lisbon: Trabalhos de Arqueologia.

TURNEY, C.S.M., M.I. BIRD, L.K. FIFIELD, R.G. ROBERTS, M.A. SMITH, C.E. DORTCH, R. GRON, E. LAWSON, L.K. AYLIFFE, G.H. MILLER, J. DORTCH & R.G. CRESSWELL. 2001. Early human occupation at Devil's Lair, south-western Australia 50 000 years ago. Quaternary Research 55: 3-13.

VAN ANDEL, T.H., W. DAVIES, B. WENINGER & O. JORIS. 2003. Archaeological dates as proxies for the spatial and temporal human presence in Europe: a discourse on the method, in T. van Andel & W. Davies (ed.) Neanderthals and modern humans in the European landscape during the Last Glaciation: 21-9. Cambridge: McDonald Institute for Archaeological Research.

VAN KLINKEN, G.J. 1999. Bone collagen quality indicators for palaeodietary and radiocarbon measurements. Journal of Arehaeological Science 26: 687-95.

VOELKER, A.H.L., EM. GROOTES, M.-J. NADEAU & M. SARNTHEIN. 2000. Radiocarbon levels in the Iceland Sea from 25-53 kyr and their link to the earth's magnetic field intensity. Radiocarbon 42: 437-52.

WOOD, R.E., C. BRONK RAMSEY & T.F.G. HIGHAM. 2010. Refining the ultrafiltration bone pre-treatment background for radiocarbon dating at ORAU. Radiocarbon 52: 600-611.

ZILHAO, J. 2006. Neandertals and moderns mixed, and it matters. Evolutionary Anthropology 15: 183-95.

ZILHAO, J. & F. D'ERRICO. 1999. The chronology and taphonomy of the earliest Aurignacian and its implications for the understanding of Neandertal extinction. Journal of World Prehistory 13: 1-68.

Thomas Higham, Oxford Radiocarbon Accelerator Unit, Research Laboratory for Archaeology and the History of Art, University of Oxford, Oxford OX1 3QY, UK

Table 1. Radiocarbon determinations from a bone or antler point
excavated from the Hyaena Den, Wookey Hole (see Higham et ad 2006a and
Jacobi et al. 2006). Stable isotope ratios are expressed in [per
thousand] relative to vPDB and nitrogen to AIR. Mass spectrometric
precision is [+ or -] 0.2 [per thousand] for carbon. Gelatin yield
represents the weight of gelatin or ultrafiltered gelatin in
milligrams. %Yld is the percent yield of extracted collagen as a
function of the starting weight of the bone analysed. %C is the carbon
present in the combusted gelatin. C:N is the atomic ratio of carbon to
nitrogen. At ORAU this is acceptable if it ranges between 2.9 and 3.5.
* Denotes a solvent extraction prior to collagen preparation.

 Radiocarbon Std Chemical Used Yield
OxA age BP error prep code (mg) (mg) %Yld

13323 30 240 380 AG * 480 26 5.4
13803 31 550 340 AF * 568 11.7 2.1
3451 24 600 300 AI 260 9.4 3.6

 [[delta].sup.13]C
OxA ([per thousand]) C:N

13323 -19.3 3.4
13803 -19.3 3.4
3451 -20.7 nd

Table 2. Radiocarbon determinations from the Geissenklosterle,
Germany. There are two determinations from each bone, an AG
(gelatinisation treatment) and AF (gelatinisation and
ultrafiltration). See caption to Table 1 for details of the analytical
data.

OxA Sample i.d. Species

5707 (AG) AH IIa GK-41 Equus ferus, scapula

21656 (AF)

6076 (AG) AH IIIc GK43, Cervus elaphus, tibia
 657.17, 2430

21657 (AF)

6077 (AG) AH IIIc GK44, alpine Capra ibex, left tibia
 57.17, 2389

21658 (AF)

6256 (AG) AH III GK-48 Rangifer tarandus, tibia

21659 (AF)

5229 (AG) AH It GK-38 Mammuthus Primigenius, rib

21660 (AF)

5161 (AG) AH le GK 32- Rangifer tarandus, metacarpal
 Ica

21661 (AF)

OxA Radiocarbon age BP Used (mg) Yield (mg) %Yld

5707 (AG) 33 200 [+ or -] 800 720 29.7 4.1

21656 (AF) 33 000 [+ or -] 500 540 25.9 4.8

6076 (AG) 33 600 [+ or -] 1900 500 20.3 4.1

21657 (AF) 39 400 [+ or -] 1100 480 20.4 4.2

6077 (AG) 32 050 [+ or -] 600 520 12.6 2.4

21658 (AF) 38 300 [+ or -] 900 420 12.6 3

6256 (AG) 30 100 [+ or -] 550 560 4.9 0.9

21659 (AF) 35 050 [+ or -] 600 480 12.1 2.5

5229 (AG) 27 950 [+ or -] 550 1000 11.3 1.1

21660 (AF) 27 960 [+ or -] 290 1040 22.5 2.2

5161 (AG) 30 300 [+ or -] 750 500 8.7 1.7

21661 (AF) 32 900 [+ or -] 450 300 12.2 4.1

 [delta][sup.13]C
OxA %C ([per thousand]) C:N

5707 (AG) 43.8 -20.5

21656 (AF) 45.8 -20.9 3.2

6076 (AG) 40.8 -22.3

21657 (AF) 43.9 -19.4 3.1

6077 (AG) 39.6 -19.4

21658 (AF) 44.2 -18.3 3.1

6256 (AG) 41.3 -19.1

21659 (AF) 44.1 -18.9 3.2

5229 (AG) 42 -20.4

21660 (AF) 41.5 -20.4 3.2

5161 (AG) 41.7 -19.1

21661 (AF) 44.2 -18.3 3.1

Table 3. AMS radiocarbon determinations of bones from Banwell Bone
Cave mammal assemblage-zone sites previously AMS dated in Oxford
(Higham et al. 2006a). Once again, AF denotes ultrafiltered gelatin
determinations, AG denotes a filtered gelatin determination whilst the
term Al denotes ion exchanged gelatin. See caption for Table 1 for
details of the analytical parameters. ([dagger]) Duplicate
measurements. Ages greater than the current limit of 50 000 BP reflect
the fact that the background in the laboratory at the time these ages
were determined was assessed to be lower than now.

Element/species OxA Radiocarbon age BP

Windy Knoll, Derbyshire

Bison priscus, radius OxA-4579 37 300 [+ or -] 1100
 OxA-15001 > 51 700

Steetley Quarry, Nottinghamshire

Bison priscus, metacarpal OxA-2846 > 44 700
 OxA-15000 > 53 200

Brean Down, north Somerset

Canis lupus, humerus OxA-4582 41 200 [+ or -] 1600
 OxA-15002 > 52 700

Ash Tree Cave, Derbyshire (clay)

Bison priscus, cervical vertebra OxA-7736 > 41 500
 OxA-15003 > 57 700

Ash Tree Cave, Derbyshire (clay)

Bison priscus, metatarsal OxA-13800 > 54 100

Bone fragment OxA-13801 > 56 500

Bone fragment OxA-13802 52 800 [+ or -] 3100

Banwell Bone Cave, north Somerset

Bison priscus, calcaneum OxA-14136 > 59 500

Bison priscus, calcaneum OxA-14137 52 700 [+ or -] 1900
 ([dagger])
 OxA-14138 > 53 900
 ([dagger])

Hunter's Lodge Inn Sink, Somerset

Bison priscus, scapula OsA-13566 >54 800

 [delta][sup.13]C
Element/species Method C:N ([per thousand])

Windy Knoll, Derbyshire

Bison priscus, radius AI -20.1
 AF 3.2 -20.8

Steetley Quarry, Nottinghamshire

Bison priscus, metacarpal AI -22.2
 AF 3.2 -20.6

Brean Down, north Somerset

Canis lupus, humerus AI -19.8
 AF 3.2 -19.5

Ash Tree Cave, Derbyshire (clay)

Bison priscus, cervical vertebra AG 3.3 -20.9
 AF 3.2 -20.6

Ash Tree Cave, Derbyshire (clay)

Bison priscus, metatarsal AF 3.3 -20.4

Bone fragment AF 3.3 -20.4

Bone fragment AF 3.3 -20.2

Banwell Bone Cave, north Somerset

Bison priscus, calcaneum AF 3.2 -20.3

Bison priscus, calcaneum AF 3.2 -20.6
 AF 3.1 -20.7
Hunter's Lodge Inn Sink, Somerset

Bison priscus, scapula AF 3.2 -20.6

 [delta][sup.15]N
Element/species ([per thousand]) Wt.% coll.

Windy Knoll, Derbyshire

Bison priscus, radius 1.4
 4.6 10.1

Steetley Quarry, Nottinghamshire

Bison priscus, metacarpal 8.3
 9.4 2.7

Brean Down, north Somerset

Canis lupus, humerus 1.5
 10.5 1.7

Ash Tree Cave, Derbyshire (clay)

Bison priscus, cervical vertebra 5.6 6.2
 6.6 2.6

Ash Tree Cave, Derbyshire (clay)

Bison priscus, metatarsal 8.8 3.7

Bone fragment 9.9 6.4

Bone fragment 10.0 2.4

Banwell Bone Cave, north Somerset

Bison priscus, calcaneum 10.8 14.8

Bison priscus, calcaneum 11.1 6.0
 10.6 3.5
Hunter's Lodge Inn Sink, Somerset

Bison priscus, scapula 8.8 3.6

 Pretreat.
Element/species yield (mg) %C

Windy Knoll, Derbyshire

Bison priscus, radius 6.8 61.8
 94.5 42.6

Steetley Quarry, Nottinghamshire

Bison priscus, metacarpal 25 44.1
 14 43.3

Brean Down, north Somerset

Canis lupus, humerus 5.0 48.0
 12.1 41.8

Ash Tree Cave, Derbyshire (clay)

Bison priscus, cervical vertebra 75.6 31.2
 25.6 42.1

Ash Tree Cave, Derbyshire (clay)

Bison priscus, metatarsal 30.0 46.8

Bone fragment 47.7 47.1

Bone fragment 17.0 43.3

Banwell Bone Cave, north Somerset

Bison priscus, calcaneum 59.0 41.2

Bison priscus, calcaneum 35.0 41.7
 14.6 41.1
Hunter's Lodge Inn Sink, Somerset

Bison priscus, scapula 17.9 43.2

Table 4. Radiocarbon ages of charcoal from the site of Grotta di
Fumane, Italy. The dates are divided into two columns, one ABA ages
treated with the routine acid-base-acid preparation, the other ABOx-
SC ages, treated using that method. Determinations in each level come
from the same homogeneous sample of charcoal. The differences are
attributed to the different pre-treatment chemistry applied. The OxA-
X prefix is given for the sample from Layer A5 that produced a very
low %C value which is not usually expected for this type of material.

 [delta][sup.13]C
 ABA ages %C ([per thousand])

Proto Aurignacian

Lyr A2, sq. 97d 30 650 [+ or -] 260 58.5 -25.2
 (OxA-11347)

Lyr A2/struc. 18 33 380 [+ or -] 210 64.4 -24.7
 (OxA-19525)

Lyr A2/struc. 16/ 32 120 [+ or -] 240 44.2 -24.5
lev. B (OxA-19413)

Lyr A2/struc. 17 32 530 [+ or -] 240 60.7 -25.6
 (OxA-19411)

Lyr A2, sq. 107i 31830 [+ or -] 260 44.3 -23.3
 (OxA-11360)

Mousterian

Lyr A5, sqs. 85, 86, 33 700 [+ or -] 600 62.3 -22.1
95, 96 (OxA-6463)

 36 860 [+ or -] 700 60.7 -21.2
 (OxA-18199)

Lyr A5 sq. 88i, 34 500 [+ or -] 270 58.8 -24.6
3789/struc. III (OxA-19410)

Lyr A5 + A6, sq. 90 38 800 [+ or -] 750 42.1 -23.8
 (OxA-8022)

 38 250 [+ or -] 700 66.3 -24.2
 (Ox-A-8023)

 39 500 [+ or -] 330 60.4 -24.2
 (OxA-17567)

 39 490 [+ or -] 350 57.8 -24.5
 (OxA-17568)

 [delta][sup.13]C
 ABOx-SC ages %C ([per thousand])

Proto Aurignacian

Lyr A2, sq. 97d 35640 [+ or -] 220 80.6 -22.5
 (OxA-17569)

Lyr A2/struc. 18 35850 [+ or -] 310 40.5 -23.8
 (OxA-19584)

Lyr A2/struc. 16/ 34180 [+ or -] 270 48.6 -24.7
lev. B (OxA-19414)

Lyr A2/struc. 17 34940 [+ or -] 280 61.8 -24.2
 (OxA-19412)

Lyr A2, sq. 107i 35 180 [+ or -] 220 74.5 -21.7
 (OxA-17570)

Mousterian

Lyr A5, sqs. 85, 86, 40150 [+ or -] 350 74.4 -21.1
95, 96 (OxA-17980)

Lyr A5 sq. 88i, 41650 [+ or -] 650 24.4 -23.0
3789/struc. III (OxA-X-2275-45)

Lyr A5 + A6, sq. 90 40460 [+ or -] 360 62.1 -24.4
 (OxA-17566)