Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Early Neanderthal constructions deep in Bruniquel Cave in southwestern France


Very little is known about Neanderthal cultures1, particularly early ones. Other than lithic implements and exceptional bone tools2, very few artefacts have been preserved. While those that do remain include red and black pigments3 and burial sites4, these indications of modernity are extremely sparse and few have been precisely dated, thus greatly limiting our knowledge of these predecessors of modern humans5. Here we report the dating of annular constructions made of broken stalagmites found deep in Bruniquel Cave in southwest France. The regular geometry of the stalagmite circles, the arrangement of broken stalagmites and several traces of fire demonstrate the anthropogenic origin of these constructions. Uranium-series dating of stalagmite regrowths on the structures and on burnt bone, combined with the dating of stalagmite tips in the structures, give a reliable and replicated age of 176.5 thousand years (±2.1 thousand years), making these edifices among the oldest known well-dated constructions made by humans. Their presence at 336 metres from the entrance of the cave indicates that humans from this period had already mastered the underground environment, which can be considered a major step in human modernity.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Prices vary by article type



Prices may be subject to local taxes which are calculated during checkout

Figure 1: Ortho-image of the Bruniquel Cave structures.
Figure 2: The calcite cores sampled from the structures.
Figure 3: Uranium-series ages (with 2σ error bars) obtained from the structures.


  1. Mellars, P. The Neanderthal Legacy. An Archaeological Perspective from Western Europe, (Princeton University Press, 1996)

  2. Soressi, M. et al. Neandertals made the first specialized bone tools in Europe. Proc. Natl Acad. Sci. USA 110, 14186–14190 (2013)

    Article  ADS  CAS  Google Scholar 

  3. Soressi, M. & d’Errico, F. Pigments, gravures, parures: les comportements controversés des Néandertaliens. In Les Néandertaliens, Biologie et Cultures (eds Vandermeersch B. et Maureille B. ) Doc. Préhist. 23, Paris, CTHS, 297–309 (2007)

    Google Scholar 

  4. Maureille, B. & Vandermeersch, B. Les sépultures néandertaliennes. In Les Néandertaliens, Biologie et Cultures (eds Vandermeersch, B. & Maureille, B. ) Doc. Préhist. 23, Paris: CTHS, 311–322 (2007)

    Google Scholar 

  5. Villa, P. & Roebroeks, W. Neandertal Demise: An Archaeological Analysis of the Modern Human Superiority Complex. PLoS One 9, e96424 (2014)

    Article  ADS  Google Scholar 

  6. Rouzaud, F., Soulier, M. & Lignereux, Y. La grotte de Bruniquel. Spelunca 60, 27–34 (1996)

    Google Scholar 

  7. Lafon, L. La Grotte de Bruniquel (Tarn-et-Garonne). Inventaire au Sol des Vestiges Fauniques. Thesis, University of Toulouse Paul Sabatier (1996)

  8. Lorblanchet, M. La Naissance de l'Art. Genèse de l'Art Préhistorique (Paris, Errance, 1999)

  9. Hayden, B. Neandertal social structure? Oxf. J. Archaeol. 31, 1–26 (2012)

    Article  Google Scholar 

  10. Cheng, H. et al. Improvements in 230Th dating, 230Th and 234U half-life values, and U-Th isotopic measurements by multi-collector inductively coupled plasma mass spectrometry. Earth Planet. Sci. Lett. 371–372, 82–91 (2013)

    Article  ADS  Google Scholar 

  11. Wainer, K. et al. Millennial climatic instability during penultimate glacial period recorded in a south-western France speleothem. Palaeogeogr., Palaeoclim. Palaeoecology 376, 122–131 (2013)

    Article  ADS  Google Scholar 

  12. Hublin, J.-J. The origin of Neandertals. Proc. Natl Acad. Sci. USA 106, 16022–16027 (2009)

    Article  ADS  CAS  Google Scholar 

  13. Otte, M. The management of space during the Paleolithic. Quatern. Int. 247, 212–229 (2012)

    Article  Google Scholar 

  14. McBrearty, S. & Brooks, A. S. The revolution that wasn’t: a new interpretation of the origin of modern human behavior. J. Hum. Evol. 39, 453–563 (2000)

    Article  CAS  Google Scholar 

  15. D’Errico, F. The invisible frontier. A multiple species model for the origin of behavioral modernity. Evol. Anthropol. 12–4, 188–202 (2003)

    Article  Google Scholar 

  16. Tchernych, A. P. in Molodova I: A Single Case of Mousterian Settlement in the Middle Dniestr Basin (in Russian) (ed. Goretsky, G.I. & Ivanova, I.K. ) 6–102 (Nauka, 1982)

  17. de Lumley, H. Une Cabane Acheuléenne dans la Grotte du Lazaret à Nice. Vol. 7, Paris, Soc. Préhist. Franç. (1969)

  18. Mania, D. H. et al. Bilzingsleben II. Homo erectus – Seine Kultur und Seine Umwelt. VEB Deutscher Verlag der Wissenschaften, Berlin (1983)

  19. Goren-Inbar, N. et al. Evidence of hominin control of fire at Gesher Benot Ya’aqov, Israel. Science 304, 725–727 (2004)

    Article  ADS  CAS  Google Scholar 

  20. Berna, F. et al. Microstratigraphic evidence of in situ fire in the Acheulean strata of Wonderwerk Cave, Northern Cape Province, South Africa. Proc. Natl Acad. Sci. USA 109, 1215–1220 (2012)

    Article  ADS  Google Scholar 

  21. Roebroeks, W. & Villa, P. On the earliest evidence for habitual use of fire in Europe. Proc. Natl Acad. Sci. USA 108, 5209–5214 (2011)

    Article  ADS  CAS  Google Scholar 

  22. Clottes, J. et al. Les peintures paléolithiques de la grotte Chauvet-Pont d’Arc, à Vallon-Pont-d’Arc (Ardèche, France): datations directes et indirectes par la méthode du radiocarbon. C.R. Acad. Sc. Paris 320, 1133–1140 (1995)

    Google Scholar 

  23. Plagnes, V. et al. Cross dating (Th/U-14C) of calcite covering prehistoric paintings in Borneo. Quat. Res. 60, 172–179 (2003)

    Article  CAS  Google Scholar 

  24. Aubert, M. et al. Pleistocene cave art from Sulawesi, Indonesia. Nature 514, 223–227 (2014)

    Article  ADS  CAS  Google Scholar 

  25. Bednarik, R. The cave petroglyphs of Australia. Aust. Aborig. Stud. 2, 64–68 (1990)

    Google Scholar 

  26. Arsuaga, J. L. Bermudez de Castro, J. M. & Carbonell, E. (Eds) The Sima de los Huesos hominid site. J. Hum. Evol. 33 (special issue)

  27. Onac, B. P. et al. U-Th ages constraining the Neanderthal footprint at Vârtop Cave, Romania. Quat. Sci. 24, 1151–1157 (2005)

    Article  ADS  Google Scholar 

  28. Bégouën, R., Clottes, J., Feruglio, V. & Pastoors, A. La Caverne des Trois-Frères, Paris Somogy- Assoc. L. Bégouën (2013)

  29. Delannoy, J.-J. et al. The social construction of caves and rockshelters: Chauvet Cave (France) and Nawarla Gabarnmang (Australia). Antiquity 87, 12–29 (2013)

    Article  Google Scholar 

  30. Petit, J. R. et al. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399, 429–436 (1999)

    Article  ADS  CAS  Google Scholar 

  31. Rouzaud, F., Soulier, M., Brugal, J.-Ph. & Jaubert, J. L’Igue des Rameaux (Saint-Antonin-Noble- Val, Tarn-et-Garonne). Un nouveau gisement du Pléistocène moyen. Premiers résultats. Paléo 2, 89–106 (1990)

    Article  Google Scholar 

  32. Faivre, J.-Ph . et al. In Modalités d'Occupations et Exploitation des Milieux au Paléolithique dans le Sud-Ouest de la France: l'Exemple du Quercy (eds Jarry, M., Brugal, J.-Ph. & Ferrier, C. ), Paléo 4 (suppl.), 231–270 (2013)

  33. Clottes, J. et al. In L’Art des Cavernes. Atlas des Grottes Ornées Françaises (Paris, Ministère de la Culture-Imprimerie Nationale), 540–551 (1984)

  34. Cantalejo, P., del Mar Espejo, M., Ramos, J. & Weniger, G. C. Elementos de iluminación in Cueva de Árdales. Intervenciones arqueológicas 2011-2014 (eds Ramos, J., Weniger, G.C., Cantalejo, P. & del Mar Espejo, M. ) Ediciones Pinsapar, 119–146 (2014)

  35. Brodard, A. et al. In Actes du colloque MADAPCA, Micro Analyses et Datations de l’Art Préhistorique dans son Contexte Archéologique, MNHN-C2RMF, 16-18 novembre 2011. Paléo (special issue) 233–235 (2014)

  36. Bertran, P., Hétu, B. & Texier, J.-P. Fabric characteristics of subaerial slope deposits. Sedimentology 44, 1–16 (1997)

    Article  ADS  Google Scholar 

  37. Lenoble, A. & Bertran, P. Fabric of Palaeolithic levels: methods and implications for site formation processes. J. Archaeol. Sci. 31, 457–469 (2004)

    Article  Google Scholar 

  38. Scollar, I., Tabbagh, A., Hesse, A. & Herzog, I. Archaeological Prospecting and Remote Sensing. Cambridge, Cambridge University Press (1990)

  39. Le Borgne, E. Influence du feu sur les propriétés magnétiques du sol et sur celles du schiste et du granite. Ann. Geophys. 16, 159–195 (1960)

    CAS  Google Scholar 

  40. Maki, D., Homburg, J. A. & Brosowske, S. D. Thermally activated mineralogical transformations in archaeological hearths: inversion from maghemite (γFe2O3) phase to hematite (αFe2O3) form. Archaeol. Prospect. 13, 207–227 (2006)

    Article  Google Scholar 

  41. Carrancho, Á. & Villalaín, J. J. Different mechanisms of magnetisation recorded in experimental fires: archaeomagnetic implications. Earth Planet. Sci. Lett. 312, 176–187 (2011)

    Article  ADS  CAS  Google Scholar 

  42. Jrad, A. et al. Magnetic investigations of buried palaeohearths inside a Palaeolithic cave (Lazaret, Nice, France). Archaeol. Prospect. (2013)

  43. Brodard, A. et al. Thermal characterization of ancient hearths from the cave of Les Fraux (Dordogne, France) by thermoluminescence and magnetic susceptibility measurements. Quat. Geochronol. 10, 353–358 (2012)

    Article  Google Scholar 

  44. Burens, A. et al. Benefits of an accurate 3D Documentation in Understanding the Status of the Bronze Age Heritage Cave ‘Les Fraux’ (France). Int. J. of Heritage in the Digital Era 1, 179–195 (2014)

    Article  Google Scholar 

  45. Stiner, M. C. & Kuhn, S. L. Differential burning, recrystallization, and fragmentation of archaeological bone. J. Archaeol. Sci. 22, 223–237 (1995)

    Article  Google Scholar 

  46. Lebon, M. et al. Application des micro-spectrométries infrarouges et Raman à l’étude des processus diagénétiques altérant les ossements paléolithiques. Rev. Archéométrie 35, 179–190 (2011)

    Google Scholar 

  47. Ellingham, S. T. D., Thompson, T. J. U., Islam, M. & Taylor, G. Estimating temperature exposure of burnt bone - a methodological review. Sci. Justice 55, 181–188 (2015)

    Article  Google Scholar 

  48. Rouzaud, J.-N., Deldicque, D., Charon, E. & Pageot, J. Carbons in the heart of energy and environment questions: a nanostructural approach. C. R. Geosci. 347, 124–133 (2015)

    Article  Google Scholar 

  49. Deldicque, D., Rouzaud, J.-N. & Velde, B. A Raman-HRTEM study of the carbonization of wood: a new Raman-based paleothermometer dedicated to archaeometry. Carbon (2016)

  50. Wainer, K. et al. Speleothem record of the last 180 ka in Villars cave (SW France): investigation of a large delta (18)O shift between MIS6 and MIS5. Quat. Sci. Rev. 30, 130–146 (2011)

    Article  ADS  Google Scholar 

Download references


We thank the owners of the cave (Nidauzel association), the French Ministry of Culture & Communication, MCC (DRAC-SRA Midi-Pyrénées, Toulouse), M. Vaginay, P. Chalard, É. Mauduit, the Speleological & Archaeological Society of Caussade (SSAC), CNRS (InEE & InSU), the University of Bordeaux-PACEA, LSCE Gif-s/-Yvette, M. O’Farrell and C. Garrec for editing, V. Feruglio for a drawing. We thank F. Dewilde and F. Mansouri (LSCE) for their assistance with the isotopic measurements, Y. Vanbrabant (Belgian Geological Survey) for his assistance with the cave monitoring and B. Martinez for his help with the topography. We thank S. Mariot and R. Weil (LPS, Paris-XI University, Orsay) for their help in the infrared spectrometry measurements. This work is mainly supported by French MCC (DRAC-SRA Midi-Pyrénées, Toulouse) and in part by the Belgian Science Policy Office. The U-Th dating was supported in part by the U.S. NSF.

Author information

Authors and Affiliations



J.J., S.V. and D.G. coordinated this study; they wrote the article and conducted the field sampling. M.S. participated in the cave discovery and is in charge of the logistical support and cave access. H.Ch. made the U-Th measurements and R.L.E. oversaw and helped to interpret the U/Th dates. D.B. conducted the δ18O and δ13C measurements. C.B. is responsible for the temperature monitoring. H.C., S.D. and X.M. realised the geographical and new topography studies of the cave. F.L.-C. realised the drawings. F.L. realised the magnetism measurements and their interpretation, D.D., D.G. and J.-N.R, the SEM-EDS, FTIR measurements and Raman spectrometry. F.M. participated in the field trips and archaeological survey. P.M. realised the photogrammetric work. C.F. realised the study of fireplaces and heated areas. É.R. participated in the field trips and the coring. F.S. is responsible for the statistical studies of the structure elements.

Corresponding authors

Correspondence to Jacques Jaubert, Sophie Verheyden or Dominique Genty.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Location and map of Bruniquel Cave.

a, Bruniquel Cave (marked with a star) is located in the southwest of France, south of the calcareous plateaus of Quercy, east of the Aquitaine Basin. Its entrance (165 m above sea level) overlooks the Aveyron valley, a tributary of the Tarn on the right bank of the Garonne and down from the Massif Central (base map courtesy of M. Jarry). b, Bruniquel Cave in the Aveyron valley. Orange: Lower Palaeolithic site; red: Middle Palaeolithic sites; green: early Upper Palaeolithic; blue: late Upper Palaeolithic (Magdalenian). Circles indicate caves, vertical lines indicate rock shelters and squares mark open-air sites. *Decorated caves. In this area within a 30 km zone around Bruniquel Cave, fifteen major Palaeolithic sites are known. The oldest known human occupations in this region are those of the Igue des Rameaux (Tarn-et-Garonne), a karstic sinkhole where lithic material was associated with a recent mid-Pleistocene fauna, dated from marine isotope stages 9 to 5 (ref. 31). A Middle Palaeolithic, stratified open-air site is also present at La Rouquette-Puycelsi (Tarn) upstream on the nearby Vère River32. The other sites are all attributable to the Upper Palaeolithic, representing the Aurignacian, Gravettian and Solutrean periods, but mainly the Magdalenian period with three decorated caves: Travers de Jannoye, La Magdeleine-des-Albis (Penne, Tarn) and Mayrière (Bruniquel, Tarn-et-Garonne)33 (base map, courtesy of StepMap GmbH, modified by J.J.). c, Topography of Bruniquel Cave. The cave consists of a 10–15 m wide and 4–7 m high corridor, currently known to be 482 m long. Beyond the narrow entrance passage (filled porch), there are no major topographic difficulties until the chamber containing the structure at 336 m from the unobstructed entrance. Currently, no other access has been identified, laterally or at the other end. In this latter case, a second obstructed entrance would be at least 295 m from another slope. Sources: Structure drawn by M. Soulier and F. Rouzaud, 1992; topography realized by Protée-Expert & Get in Situ, 2015; Digital Elevation Model generated with 1957 aerial photography IGN, public domain).

Extended Data Figure 2 Bruniquel Cave structures.

a, General view of the main structure (structure A) with superposed layers of aligned stalagmites (speleofacts) Photo courtesy of É. Fabre, SSAC. b, Example of speleofacts accumulated over three or even four horizontal levels. c, Stalagmites (speleofacts) placed vertically against the main structure (structure A) in the manner of stays. d, e, Two examples of short back stalagmites serving as sustaining pieces. f, Summary of the metric data of the structures.

Extended Data Figure 3 Fireplaces and heated areas.

a, Examples of a fireplace on the main structure. Note the reddened, blackened and fissured stalagmites34. The structure in this location (top) is covered by white, more recent and still active stalagmites. The heated areas on the speleofacts correspond to the red and grey colours, as well as fissuring and superficial spalling. These scars are similar to thermal alterations studied in the cave of Chauvet-Pont d’Arc (Ardèche)35. In our current stage of observation, the study of their distribution enabled us to identify a well-preserved fireplace in structure A, as well as structures that have been disturbed by processes that remain to be determined (structures D and E, for example). b, Numbers per structure of heated areas, thermic spalling, fissured spots and blackened elements (that is, speleofacts) and soot.

Extended Data Figure 4 Statistics of the speleofacts.

a, b, Kernel density estimates for the dimensions (a, length and b, diameter) of speleofacts across the different structures. Structure A can be distinguished from the others by the presence of very large speleofacts. Such speleofacts are not present in structure D, and only rarely in structure B. Structure C, despite its very small size, is worth considering due to the large dimensions of its speleofacts. Structures E and F, with only a few speleofacts and no specific features, are not represented here. A Kruskal–Wallis test conducted on the structures represented here shows a significant difference between the median length and median diameter across structures (P < 0.05). A post hoc analysis of the diameter with Hochberg’s adjustment method, distinguishes structure C from the three others. c, The weight of the speleofacts is estimated by the following formula: πD2/12 × (1 + d/D + d2/D2) where D is the maximum diameter, d the minimum diameter, L the maximal length, and ρ the calcite density. These weights can be roughly estimated by considering them as truncated cones. As their maximum length L, maximum diameter D, and minimum diameter d are known, their volume can be easily estimated (Extended Data Table 1). Their weight is then obtained by multiplying the previous quantity by the calcite density ρ, which is comprised between 2.5 and 2.8 g cm−3 depending on its porosity and detrital contamination. Minimal weights are obtained using a density of 2.5 g cm−3. d, The figure shows the mean weights and their 95% confidence interval in each structure. eg, The orientation data (Schmidt diagram36) of the speleofacts in the three main structures (A, B, D) are very similar (e, structure A; f, structure B; g, structure D) and do not show any preferential direction. The distance to the centre of the circle represents the slope; the distribution of the speleofacts is isotropic and mostly planar. This confirms in all cases that such orientation and slope patterns cannot be due to natural processes related to water flow, mass flows or other gravitational processes37, which in any case would not have resulted in the current geomorphology of the cave in this sector.

Extended Data Figure 5 Magnetic survey above the structures.

Red circles: main recognized hearths. The magnetic survey aims to reveal the locations that were heated, including hearths or smaller fireplaces through the detection of magnetic anomalies. The first archaeological applications of this prospection method are for the location of heated archaeological structures (see pages 422–519 of ref. 38). The magnetic properties enhancement by heating was first demonstrated for soils39,40,41, and then on substrate of caves42,43,44. In this type of hydromorphic environment, iron is present as nonmagnetic or weak magnetic FeOOH minerals, such as goethite (see pages 375–421 of ref. 38). In these conditions, temperature elevation above 200–250 °C induces dehydration of the FeOOH, present in clay material, to Fe3O4 (magnetite) which is a strong magnetic mineral43. The increase of magnetic susceptibility induced by heating offers similar information than thermoluminescence methods43. In the present case, a magnetic susceptibility increase beyond a factor of two was observed after heating a clay sample of the cave. Therefore, the heated clay-like material, even if present only in small amounts in speleothems, acquired a sufficiently high magnetization to generate a local earth magnetic deformation, also called an anomaly. As this deformation decreases when the source distance increases (see pages 422–519 of ref. 38), a larger anomaly with a medium intensity might reveal a hearth under the stalagmitic floor (between structures B and C), calcite being magnetically nearly neutral (diamagnetic). The realization of magnetic survey at high spatial resolution for detection of paleohearths in prehistoric cave is a recent innovation44. The magnetic field explored above the structures was over one metre thick, with a dual sensor G858 Geometrics magnetometer with an extended cable. A 360° prism was inserted between both sensors, which were superposed at a distance of 0.22 m. These elements were hung at the end of a telescopic boom pole and fixed on a tripod. 3D geolocation measurements were ensured by tracking with a Trimble S8 total station following the 360° prism. This apparatus allows coverage of a volumetric space up to 5 m from the operator with ten measurements per second while controlling the space covered44. Extended Data Fig. 5 presents the results of the magnetic measurements. Altitude contour lines (8.5 cm distance interval) are extracted from photogrammetric data. The magnetic intensity point cloud is a bottom view of the magnetic field intensity gradient, that is, the difference in magnetic field intensity as measured between the bottom and top sensors. As the local past and present magnetic field have an inclination of ~63° down, a magnetic source generates a dipolar local deformation of the magnetic field with a negative anomaly to the north and positive to the south38. In Extended Data Fig. 5, a dipole corresponds to a blue and red spot aligned approximately north–south. The majority of the main dipoles of metric dimension observable are mostly associated to fire traces (reddened, blackened calcite) observed on the horizontally positioned stalagmites, for example, the heated zone of the structures D and E. Increases of magnetic viscosity, known as a fire marker42, are observed in such zones. Some places present split positive anomalies, for example, places located on structure D, indicating twin core fires or non-contemporaneous fires. The main measured dipole is located to the west of structure B at the border of a zone covered by a calcite layer and near a char concentration zone, which suggests the occurrence of a hearth underneath the flowstone. Some visible heated zones did not reveal any magnetic anomaly, indicating that the substratum at these places was heated below 200–250 °C. The most tenuous dipoles located on the flat ground surface may reflect the changing nature of the substratum, rather than any heating. Indeed, the weak magnetic contrast between clay material and calcite material can be the source of a weak anomaly. An alternative explanation is the presence of a heated zone underneath a thick stalagmitic floor, the distance between source and measurement mitigating the anomaly38. For example, an anomaly located at midway between structures B and C. Complementary analysis of the spatial distribution of the clay material must be realized to determine which hypothesis is correct.

Extended Data Figure 6 Burnt bone fragments.

Three black fragments (a, b, c) were analysed with a scanning electron microscope energy dispersive spectrometry probe (SEM–EDS) (e, f), fast Fourier infra-red, FTIR (d) and Raman spectrometry (g, h, i). FTIR analyses were made at the Laboratoire de Physique des Solides (LPS), Paris-XI University, Orsay by S. Mariot on a Nicolet iS50 ABX spectrometer. Raman spectroscopy was performed with an Invia spectrometer from Renishaw and the atomic spectrometry was performed with a FE-SEM Zeiss Sigma equipped with an EDS probe at the École Normale Supérieure, Paris, France. a, A 6.7-cm-long piece of burnt bone (Br-SE-Os) trapped between stalagmite elements in structure E (Extended Data Fig. 5) was almost completely covered by calcite except on its medullar side. Three layers were sampled for uranium-series dating (green, red and blue marks) (Extended Data Table 2). The bone with the 5-mm-thick calcite crust was cut longitudinally and the calcite was sampled along deposition layers, starting at the internal surface after removing the bone material. Three thin discontinuities marked by thin brownish layers separate the deposits into three calcite layers from which three 230Th samples were taken (Extended Data Table 2). Except the middle sub-sample, which was contaminated by detrital elements (high 232Th concentration), 230Th ages given by the other two sub-samples are in stratigraphic order and in agreement with the age of the structures. This demonstrates that humans introduced this bone before 180.9 ± 20.3 ka. Note the elongated medulla cells of the bone and their deep black colour, suggesting that the collagen was carbonized at a temperature between 300 and 400 °C45,46. Note that the burnt bone was covered by a reddish and blackened speleofact (Extended Data Fig. 5), due to the heat. d, FTIR spectroscopy (blue spectrum on the black part of the bone, green spectrum on the grey part of the bone, red spectrum on the overlying calcite crust and grey spectrum on a modern char) show well-characterized PO43− absorbance peaks, suggesting that the bone was burnt; such as the slightly more individualized peak at ~618 cm−1; and the splitting factor (SF) calculated with the heights of the 603 and 565 cm−1 peaks, which are here relatively high (4.6 to 4.8) and typical of burnt bones47. g, Raman spectrometry displays two well-defined peaks at 1,580 cm−1 and at 1,350 cm−1, characteristic of char, demonstrating that it was burnt48,49. b, Sample Br-SB7 is a 3 mm large black fragment found trapped in the core of Br-stm-SB7 (Fig. 2). This fragment is situated just below the base of the regrowth dated to 175.2 ± 0.8 ka, and just above the ancient surface of the ‘old’ stalagmite (whose layers have been dated to 222.4 ± 5.8 ka). h, Raman spectra of this black fragment display two well-defined peaks at 1,580 cm−1 and at 1,350 cm−1 characteristic of char carbon49,50. e, SEM–EDS shows the presence of phosphorous, in addition to carbon, suggesting that it is a burnt bone fragment, similar to the larger bone piece (a). Because it is trapped in the dated calcite core, it also demonstrates that the fire occurred before 175.2 ± 0.8 ka. c, A black aggregate of millimetre-sized fragments (Br-PS92), mainly burnt bones of 1–3 cm was collected in 1992 by F. Rouzaud in the char concentration zone near structure B (Extended Data Fig. 5), and analysed recently. i, As with the previous samples, the Raman spectrum is typical of char carbon with vibrational bands at 1,580 cm−1 and 1,350 cm−1. f, The SEM images (back scattered mode) show a blend of at least three phases at the micrometre scale. The elemental analyses performed by EDS on each of these phases allow their attribution to a carbonaceous component (the EDS spectrum shows a major peak of carbon), a phosphorous component (the three major peaks (Ca, P and O) strongly suggest a phase belonging to the apatite family), and a clay component (attested by the coexistence of the three major peaks Si, Al, O), respectively. The Raman spectra demonstrate that the carbonaceous component is a char48,49, that is, a carbonaceous solid resulting from the heat treatment of an organic precursor. These results confirm that the char concentration zone near structure B was most probably a hearth, and that humans burned bones on the clay-like soil of the cave.

Extended Data Figure 7 Calcite core stable isotope results.

a, b, Stable isotope measurements (calcite δ18O and δ13C) were made on parts of cores extracted from the structure to check the coherency of the isotope signal with an already published time series from speleothems from the Villars Cave (Dordogne)50, located 100 km to the northwest of Bruniquel Cave. The results reveal a good match between the average δ18O of regrowths after 176 ka and the Vil-car1 flowstone stable isotopes. This is also true for the sample that covers marine isotope stage 5e, with a much lower amplitude change, however. The Bruniquel core δ13C signal appears more variable, possibly due to a greater sensitivity of the vegetation density to climatic changes or to detrital contaminations, which are probably close to the discontinuity at the base of the regrowths (b). Higher resolution measurements combined with more uranium-series dating will allow the construction of short palaeoclimatic time series and more detailed observations of climatic variations. Today, the Structure Chamber has an extremely stable temperature of 12.68 ± 0.02 °C (two times the standard deviation of the temperature values measured during one year with a time step of 1.5 h) compared to the outside temperature over the same period (13.2 ± 8.8 °C). These results indicate the current confinement of the cave environment, important for isotopic studies.

Extended Data Figure 8 Human appropriation of the underground environment:

above, the specific task sequence in Bruniquel Cave (a). Below, replacement within the general context of various indicators of modern behaviour (b). a, Chaîne opératoire (task sequence) of the construction of the structures in Bruniquel Cave. This type of construction implies the beginnings of a social organization: this organization could consist of a project that was designed and discussed by one or several individuals, a distribution of the tasks of choosing, collecting and calibrating the speleofacts, followed by their transport (or vice versa) and placement according to a predetermined plan. This work would also require adequate lighting. The construction of such a structure, involving the placement and arrangement of speleofacts, supposes a minimum degree of skill, since architectural techniques such as inserting wedging elements between two rows of speleofacts (Extended Data Fig. 2d, e), or placing stays to act as a buttresses (Extended Data Fig. 2c), appear to have been used. We evaluated the number of speleofacts used (approximately 400), as well as their combined weight (between 2.1 and 2.4 tons), but not yet the number of hours necessary to realize the structures. This will require long and complex experimental procedures that will be undertaken in future research. The complexity of the structure, combined with its difficult access (335 m from the cave entrance), are signs of a collective project and therefore suggest the existence of an organized society that was already on the path to ‘modernity’. Until now, no site of this age, attributed to Neanderthals—even late ones—or early modern humans has been associated with such activities in an underground space. b, A multiple species model for the origin of behavioural modernity in Europe. Modified from ref. 15, to which was added the ‘Deep Cave Occupation’ and ‘Bruniquel Cave’.

Extended Data Table 1 Speleofacts: definition and archaeometry
Extended Data Table 2 Speleothem 230Th dating results

Related audio

Supplementary information

Supplementary Information

This file contains Supplementary Methods, Supplementary Tables 1–2 and Supplementary Data. (PDF 322 kb)

3D-model of the structures in Bruniquel Cave

The 3D-model clearly showing the different types of structures: two annular ones (with superposed layers of stalagmites), which are the most impressive constructions, and four smaller stalagmite accumulation structures (especially two in the centre of the main structure A). (MOV 25689 kb)

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Jaubert, J., Verheyden, S., Genty, D. et al. Early Neanderthal constructions deep in Bruniquel Cave in southwestern France. Nature 534, 111–114 (2016).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing