Re-dating the oldest known figurative cave art.

The rock-art left behind by prehistoric cultures has the potential to provide us with insights into the lives of these long-vanished peoples. Dating this art is notoriously difficult. In the past few decades the predominant approach has been the analysis of uranium and its decay products within flowstone layers partially covering ancient art. Flowstone is formed by the deposition of calcium carbonate onto surfaces by evaporating water; typically water that has flowed through limestone deposits then run out onto a surface such as a cave wall or cliff face before evaporating. The most obvious examples of this are stalagmites and stalactites, though many caves have an interior surfaces covered by flowstone. Where these flowstone deposits occur in caves with paintings they will often overlay the artwork, which means that if the flowstone can be dated, then a minimum age for the art can be established (as the art cannot be younger than the flowstone that overlays it).

This method has been used to date ancient art in many parts of the world, including Western Europe, Island Southeast Asia, and Russia. Notably, a hand-stencil in Spain has been dated to 64 800 years before the present, which implies that it must have been made by a Neanderthal artist, although the reliability of the data used in this study has been questioned. The oldest date we currently have for figurative cave art comes from Sulawesi, Indonesia, where an image of a Warty Pig, Sus celebensis, at Leang Tedongnge in the Maros-Pangkep karst has been given a minimum age of 45 500 years.

The methods used to date flowstone samples to date have relied upon dissolving a sample to form a solution, something which will tend to homogenise the sample, averaging the age of multiple layers overlaying a piece of art, and therefore producing a younger age estimation for the art than if the oldest of these layers could be isolated. 

In a paper published in the journal Nature on 3 July 2024, a team of scientists led by Adhi Agus Oktaviana of the School of Humanities, Languages and Social Science at Griffith University, the Pusat Riset Arkeometri of the Badan Riset dan Inovasi Nasional, the Griffith Centre for Social and Cultural Research, and the Center for Prehistory and Austronesian Studies, and Renaud Joannes-Boyau of the Geoarchaeology and Archaeometry Research Group at Southern Cross University, present the results of a study which used laser-ablation uranium-series dating to provide more accurate dates for the Leang Tedongnge image, as well as other examples of cave art in the same region. 

The laser-ablation method enables the targeting of an area 44 μm in diameter, within a laboratory environment. This enables far smaller samples to be collected than with previous methods, which typically involved grinding a sample from the rockface with a rotary tool. With the laser ablation method it is possible to take a polished thin section of rock, and target points upon that, which is both cheaper and less destructive than traditional methods, as well as far more accurate, as it enables specifically targeting the layer of rock directly overlaying the pigment.

Map of the study area. (a) The Indonesian island of Sulawesi, showing the location of the southwestern peninsula (area inside rectangle). (b) South Sulawesi, with the limestone karst area of Maros-Pangkep indicated by blue shading. The locations of cave sites with dated Late Pleistocene rock art were as follows: 1, Leang Bulu’ Sipong 4; 2, Leang Karampuang; 3, Leang Tedongnge; 4, Leang Timpuseng. Oktaviana & Joannes-Boyau et al. (2024).

At the Leang Bulu’ Sipong 4 site in the Maros-Pangkep karst, a 4.5 m wide panel on the rear wall of a cave depicts a number of Human, or Therianthrope (Human-Animal hybrid) figures, interacting with Warty Pigs and Anoas (Dwarf Buffalo), Bubalus sp.. The figures are holding some form of objects, possibly spears or ropes. The artwork may depict a hunting scene, or possibly a visual representation of a myth.

Dated rock art panel at Leang Bulu’ Sipong 4. (a) Photostitched panorama of the rock art panel. Ther, Therianthrope. (b) Tracing of the dated rock art panel showing the results of laser-ablation uranium-series dating. (c) Transect view of the rock art sample BSP4.5 after removal from the artwork, highlighting the paint layer and the three integration zones (ROIs) and associated age calculations. (d) Laser-ablation-multicollector inductively coupled plasma mass spectrometry imaging of the BSP4.5 thorium²³²/uranium²³⁸ isotopic activity ratio. Oktaviana & Joannes-Boyau et al. (2024).

The imigary at Leang Bulu’ Sipong 4 is covered by four distinct speleothems (flowstone deposits), which have previously been dated to minimum ages of 35 100 years, 43 900 years, 40 900 years, and 41 000 years. The new data obtained by Oktaviana & Joannes-Boyau et al. revises these dates to minimums of 27 600 years, 39 600 years, 39 500 years, and 48 000 years. Most of these ages are similar or older to the previously obtained ages, though one younger date was obtained, possibly because Oktaviana & Joannes-Boyau et al. were careful to avoid areas showing clear signs of post-depositional alteration. 

Since the speleothems overlie the rock art, the oldest speleothem must still be younger than the art, giving a minimum age. The art was previously dated to a minimum of 43 900 years old, but Oktaviana & Joannes-Boyau et al.'s data raises that minimum age to 48 000 years, an increase of over 4000 years, older than the previous oldest dated art at Leang Tedongnge.

At the Leang Karampuang site, again in the Maros-Pangkep karst, a ceiling panel depicts three Human or Therianthrope figures interacting with an Animal, probably another Warty Pig, although the preservation here is poor due to surface exfoliation (flakes breaking off the surface of the limestone), and extensive overlying coralloid growths (small nodes of calcite, aragonite or gypsum that form on surfaces in caves). The image executed in red, and comprises the large, Pig-like Animal, 92 x 38 cm, in side view, with an infill pattern of stripes or lines, consistent with depictions of Pigs and other Animals elsewhere in South Sulawesi. There are other Pig depictions within the Leang Karampuang cave, although it is uncertain if they are the same age as the dated example. The Pig is surrounded by three Humanoid figures. The largest of these is 42 v 27 cm and lacks legs; it has both arms extended and appears to have a rod-shaped object in its left hand.  The second figure measures 28 x 25 cm and is located directly in front of the Pig, with its head in front of the Pig's snout. The final figure, measuring 35 x 5 cm, is upside-down relative to the other figures, with its legs splayed out away from the Pig and one hand reaching towards the Pig's head. A possible fourth figure may have once been present between the first and third figures. There are also at least three hand stencils on the same panel, two which appear to be contemporary with the Pig, plus one darker one which is partially overlain by the Pig, and presumably, therefore, pre-dates it.

Oktaviana & Joannes-Boyau et al. collected samples from four coralloid growths, one overlying each of the figures, plus one from the Pig. The oldest date came from the coralloid overlying the second figure, with a minimum age of 51 200 years. The minimum dates from the first and third figures were 18 700 years and 44 000 years respectively, while the coralloid growth from the Pig yielded a minimum age of 31 900 years. Thus, if the figures and Pig do represent a single piece of artwork, as seems likely, then the whole scene can be assumed to have a minimum age of 51 200 years, making it the oldest known piece of figurative art in the world. 

Dated rock art panel at Leang Karampuang. (a) Photostitched panorama of the rock art panel. (b) Tracing of the rock art panel showing the results of laser-ablation uranium-series dating. (c) Tracing of the painted scene showing the Human-like figures (H1, H2 and H3) interacting with the pig. (d) Transect view of the coralloid speleothem, sample LK1, removed from the rock art panel, showing the paint layer and the three integration zones (ROIs), as well as the associated age calculations. (e) Laser-ablation-multicollector inductively coupled plasma mass spectrometry imaging of the LK1 thorium²³²/uranium²³⁸ isotopic activity ratio. Oktaviana & Joannes-Boyau et al. (2024).

Oktaviana & Joannes-Boyau et al.'s method shows that the Leang Bulu’ Sipong 4 art is over 4000 years older than the previously determined age, with a minimum age of 48 000 years, while the Leang Karampuang art is at least 51 200 years old. These dates significantly increase the maximum known age of figurative art. The oldest known art dates from the Middle Stone Age of southern Africa, between 75 000 and 100 000 years ago, but this comprises geometric marks carved into ochre nodules. Figurative art is presumed to have arisen within Africa, and from there to have been carried around the world by migrating Humans, but there is currently no evidence for this, and it cannot be excluded that this form of expression arose in another region and then spread back to Africa.

The South Sulawesi art also challenges to long-standing preconceptions about cave art, which have come about largely from studies based upon the extensive European rock-art record. These are that Humans and/or Human-like figures did not appear in rock art until the very end of the Pleistocene, and the other is that narrative compositions were absent from early rock art. 

Three of the oldest dated rock art panels in the world come from South Sulawesi, Leang Karampuang, at least 51 200 years old, Leang Bulu’ Sipong, at least 48 000 years old, and Leang Tedongnge, at least 45 500 years old, all include figures, and all appear to involve interactions between the figures and one-another and/or Animals which imply a narrative context. Another piece of cave art, at Leang Timpuseng, dated to at least 35 300 years before the present, depicts a Pig standing in a painted line, presumably representing a ground surface. This depiction of composed scenes presumably had some communicative function, allowing the telling of a story through a narrative interpretation of the art, probably in conjunction with oral storytelling. Thus, this cave-art can also be interpreted as evidence of the emergence of a consistent form of mythology, many thousands of years before any such evidence appears in Europe.

See also...

Understanding the giant handaxes of the British Lower Palaeolithic.

The Lower Palaeolithic is the earliest stage in the archaeological record from which stone tools are known, and while such ancient tools are rare, they are much more common than the remains of the Hominins which produced them, or any other remains of their activity, thus providing a key insight into the activities of our ancient ancestors. The first bifacial handaxes appear in the African archaeological record about 1.76 million years ago, and are considered to be a significant conceptual leap from the Oldowan core-and-flake technology which preceded them, although whether these axes had an entirely novel function or were a new technological approach to tasks which were already undertaken is unclear. One of the notable manifestations of this technology was the manufacture of 'giant' handaxes, notably larger than other examples of these tools, sufficiently large that archaeologists are divided as to whether they had any practical function at all.

It has been suggested that the purpose of these giant axes may have been social rather than practical, with the production of such tools representing a significant time investment, and therefore being an indicator of patience and commitment to a project. Giant, highly symmetrical handaxes from British sites such as Furze Platt and Shrub Hill have been used to support this hypothesis. An alternative suggestion is that the Hominins involved may have found the manufacture of such giant axes an aesthetically pleasing activity in its own righty, an early sign of artistic expression emerging from the purely functional.

In a paper published in the journal Antiquity on 4 March 2024, Luke Dale of the Department of Archaeology at Durham University, Aaron Rawlinson, also of the Department of Archaeology at Durham University, and of the Department of Britain, Europe and Prehistory at the British Museum, Pete Knowles and Frederick Foulds, again of the Department of Archaeology at Durham University, Nick Ashton, also of the Department of Britain, Europe and Prehistory at the British Museum, David Bridgland of the Department of Geography at Durham University, and Mark White, once again of the Department of Archaeology at Durham University, present the results of a study of giant handaxes from across the British Lower Palaeolithic.

Dale et al. note that symmetry is often cited as a key feature of giant handaxes, but that no previous study appears to have actually looked at whether these axes are actually more symmetrical than other axes from the same period, nor whether the style or frequency of these objects varies over time. 

All giant handaxes are associated with the Acheulean technology, which in Britain was present by Marine Isotope Stage 13, about 500 000 years ago, and may have appeared as early as Marine Isotope Stage 15, about 600 000 years ago, if recent dates for the Fordwich and the Bytham River terraces are correct. The global climate underwent a series of oscilations during the Middle Pleistocene (781 000-126 000 years ago), which in Britain manifested as a series of warm interglacials during which Hominins invaded and settled the landscape, interspersed with colder glacial intervals during which they disappeared. This pattern results in a series of distinct cultural intervals within the British Acheulean, providing a potential opportunity to detect discrete phases in the manufacture of giant handaxes in Britain.

Dale et al. obtained data on 4160 handaxes of all sizes, from 47 sited dating to between Marine Isotope Stage 15 and Marine Isotope Stage 9. For convenience, each (odd) Marine Isotope Stages was considered to represent a warm phase, as well as the cool phases bracketing it, thus Marine Isotope Stage 11 in fact refers to the Marine Isotope Stage 10-11-12 cool-warm-cool cycle; each cycle therefore represents about 100 000 years. The warm phases covered by Dale et al.'s study are MIS 15 (approximately 610 000 to 560 000 years ago), MIS 13 (approximately 524 000-474 000 years ago), MIS 11 (approximately 427 000 to 364 000 years ago), MIS 9 (approximately 328 000-301 000 years ago. Many of the sites in the study are well dated, but others represent chronologically mixed assemblages, such as Dunbridge in Hampshire, where the artefacts were recovered from dredged aggregates, have only been tentatively dated, such as Whitlingham in Norfolk, or are known to have been the subject of collecting biases, such as Furze Platt in Berkshire. Dale et al. recognise that these factors will have an impact on the reliability of their study, but reason that the size of their dataset should offset this. They also note that a further two giant handaxes were discovered at Frindsbury in Kent after the initial daft of their study was submitted, and that these are thought to date to MIS 9.

Handaxes were divided into six classes on the basis of their symmetry, with Class 1 representing near perfect symetry, and Class 6 indicating almost no symmetry.

Dale et al. measured the length of all the 4160 handaxes in their study, splitting them into four categories; small (less than 80 mm in length), average (80-150 mm in length), large (150-220 mm in length) and giant (more than 220 mm in length.

Forty two of the handaxes fell into the 'giant' category, of which 33 can be confidently assigned to the MIS 9 cycle, the final stage of the Acheulian in Britain. A single giant axe was dated to MIS 11, and two to MIS 13 - although one of these is- derived from a MIS 6 gravel bed, and is interpreted as having been reworked from a MIS 13 deposit. Since MIS 9 only represents about 1% of the British Pleistocene record, this concentration of giant handaxes in MIS 9 deposits is interpreted as being significant.

In the 1960s archaeologist Derek Roe developed a system of classifying handaxes upon their shape. The majority of giant handaxes in Dale et al.'s study conform to Roe's Group I (pointed, with cleavers) and Group III (plano-convex), both of which are associated with MIS 9. Notably, handaxes in an assemblage from Broom on the Devon/Dorset border, which included several giant handaxes, while still from MIS 9, did not easily fit into Roe's scheme. The stone tools here were made from chert derived from the Upper Greensand, and tended to be ovate or asymmetrical, unlike other MIS 9 assemblages.

Other than the presence of giant handaxes in the MIS 9 material, the size distribution of handaxes between different interglacials varied little, with the average size of the axes varying only by 10-20%, although the MIS 15 handaxes appear on average slightly larger than those from later assemblages, and the MIS 11 assemblage contained a higher proportion of small handaxes than other stages. However, Dale et al. note that their MIS 15 material came from only two sites, Brandon Fields in Suffolk and Fordwich in Kent, with the material used at Fordwich being elongated pipeflint, which lends itself to the making of larger bifaced tools with a minimal amount of effort. Given this, Dale et al. are reluctant to read much into this data without further MIS 15 material becoming available.

Analysis also suggests that axes from MIS 13 and MIS 9 are on average larger than those from MIS 11, and MIS 9 axes are also slightly larger, on average, than those from MIS 13. Dale et al. are uncertain whether this variation relates to cultural differences between the different populations colonising Britain in each interglacial, and make no further assessment of this data at this time.

A comparison of handaxe length between MIS 15, 13, 11 and 9, with handaxes grouped according to length class. MIS 15 displays a greater proportion of handaxes in the large class, while small handaxes are much more limited. MIS 11 displays a greater proportion of small handaxes. Giant handaxes represent a very small percentage of the total assemblage, emphasising their status as extreme outliers beyond the usual variation in handaxe length. Frederick Foulds & Mark White in Dale et al. (2024).

The majority of giant handaxes come from the Thames Valley or its southern tributaries, although examples are also known from Norfolk Yare, the Axe Valley and Solent River deposits. The majority of sites where they are found are close to primary chalk-flint outcrops, where large flint nodules were available. However, some were found in flint-poor areas, such as Wolvercote and Stanton Harcourt in Oxfordshire, whereas older archaeological sites, such as Boxgrove in West Sussex, and Elveden in Suffolk, are located near excellent sources of large flint nodules, but yield no giant handaxes. 

No giant handaxes available from Stoke Newington in London, one of the UK's best MIS 9 sites, although it is known that the site did yield such axes; as the original excavator of the site, Victorian archaeologist Worthington Smith, illustrated several giant handaxes in his book, Man, the primeval savage, but then gave these away to distinguished visitors to the site; these axes are thought to be either in private collections or lost. 

Geological map displaying the locations of sites with giant handaxes in relation to underlying bedrock geology, coloured as per the British Geological Society scheme. Colours for chalk bedrock are highlighted in the legend. Note that large clasts in overlying superficial deposits largely reflect local geology and ‘exotic’ raw materials would generally be smaller in size. Dale et al. (2024).

Of the forty two giant handaxes in the study, thirty five (or 83%) are pointed, typically conforming to the lanceolate (lance-shaped) or ficron (having a pointed planform shape with biconcave edges) type. These shapes are considered to be part of the same typological grouping as chisel-ended cleavers, a grouping which is thought by many archaeologists to be restricted to MIS 9. Although ficrons are associated with giant axes, not all ficron handaxes are giant; known ficrons range in size from 63 mm to 307 mm.

Examples of giant pointed and ficron handaxes: (A) the ‘Beast of Biddenham’ (length 253 mm); (B) a giant ficron from Canterbury West (length 285mm) (photographs courtesy of the British Museum (A) and The Seaside Museum, Herne Bay (B)). Dale et al. (2024).

Since the overwhelming majority of the giant axes date from MIS 9, Dale et al. examined axes from this stage for their symmetry. Handaxes from this interval were found to vary a great deal in their degree of symmetry, although there was a direct corelation between axe length and symetry, with longer axes tending to be more symmetrical than shorter ones. 

Splitting the axes into groups based upon their size emphasised this trend, with axes in larger categories being more symmetrical than those in smaller size categories. Examination of earlier axes suggested that this tendency for larger axes to be more symmetrical was also present in MIS 11, but not MIS 15 or 13, although the number of axes available from the earlier periods makes this result less reliable. 

However, this tendency for larger axes to be more symmetrical in MIS 11 and 9 deposits may be influenced by another trend, which is for axes in general to be much more symmetrical in MIS 13 and 11, compared to other periods, and in particular to MIS 9. Thus, the increased symmetry seen in the large MIS 9 axes is balanced by a decrease in the symmetry seen in smaller axes, with the smallest axes (less thn 100 mm in length) tending to be roughly and quickly made, with very low symmetry. 

Giant handaxes are often cited as a key feature of Hominin cognitive and cultural development. Dale et al.'s study demonstrates that these items became prevalent in Britain during MIS 9, i.e. immediately prior to the Lower-Middle Palaeolithic transition, 300 000 and 250 000 years ago. This interval has already been identified as one of significant technological innovation, beginning with a non-handaxe phase, which was then followed by a period with handaxes, then a final phase in which a hierachical core-working technology, known as the proto-Levallois was introduced. The linking of giant handaxes to this stage adds another element to this increasing cultural complexity.

The presence of giant handaxes is often independent of the availability of suitable manufacturing materials, with giant handaxes known from areas of Oxfordshire where there are no sources of suitable large flints. This implies that either the raw material or the finished axes were transported for at least 25-40 km to the sites where they were found. This phenomenon has also been observed in France, with a giant handaxe found at Caune de l’Arago in Pyrénées-Orientales Department, more than 30 km from any suitable source of material. This transportation of materials is another new innovation from this phase, with most previous Lower Palaeolithic sites using stone sourced less than 5 km from the location where it was discovered. The increased effort being put into moving giant axes to locations away from the materials from which they were made lends further support to the hypothesis that these were valuable items, although another possibility is that these tools had a long working life, and were subject to regular reworking to maintain a sharp edge, which would explain the plano-convex form of many handaxes, including those at Wolvercote in Oxfordshire. Under this hypothesis, many large handaxes would have been discarded prematurely, before being extensively reduced.

Many of the giant handaxes are ficrons, a type of axe otherwise rare in Britain. Large ficrons are particularly striking objects, apparently crafted with care to maintain the length of the axe while reducing the width of the upper and central portions of the axe. This may have been a way to correct for a lack of symmetry in the original platform (the rock from which the ficron was made), or a sign of the tool having an extended lifespan, with regular resharpening. Another hypothesis is that the width reduction may have been a way to reduce the weight of the tool part of a general Acheulian trend for longer tools to be narrower. The alternative is that the ficron-shape was desired from the outset, and that the extra effort put into producing an axe of this shape is an indicator of the significance of the object to the people who made it. Whatever the truth, there does appear to be a link between the high symetry, large size, and unusual shape of these giant ficrons which appears to show a degree of artistry transcending the merely functional.

It has often been speculated that giant handaxes had no practical function, but experimentation has shown that examples far larger than anything found in Britain can be used as tools. Even the largest giant axes can be used two handed, can be used for digging, or can be used as static cutting tools across which materials could be drawn. Despite this possible functionality, all the British giant handaxes are distinctively formed and highly symmetrical, making it hard to believe that aesthetic considerations played no part in their manufacture. 

The oldest known giant axe is 1.6 million years old, and came from the Olduvai FLK West site in Tanzania. This is long before anything else considered to be even possibly an expression of communal artistic standards, and therefore was presumably made by an individual for reasons that were important to them alone. A more communal approach to aesthetics begins to appear about 500 000 years ago, at which point tools also being to appear which have been shaped to make the most of features such as fossils on the rock cortex. Giant handaxes may also represent an aesthetic development, with the size and symmetry of stone tools being increased to create visually pleasing objects. This implies the people who made these tools were capable in taking pleasure in creating such objects, which would imply an increasing level of cognitive and social development.

Giant handaxes may also have had some symbolic purpose. The manufacture and then apparent discarding of objects which take a degree of forward planning and labour investment may have been a way of marking territory, or signposting important locations to other members of the community.

The ability to produce giant handaxes clearly predates MIS 9; they are known from earlier in the Lower Palaeolithic and across the Acheulean world. A 300 mm-long lanceolate giant handaxe from Caune de l’Arago is thought to date from MIS 14, and other examples from across northern France, at sites including St Acheul and Thennes in the Somme Valley, Vaudricourt in Pas-de-Calais, and Montguillain in Oise Department, have variously been dated to MIS 11 and 9, although none of these dates are considered to be particularly secure. Giant handaxes are also known from the Levant and Middle East, though these tend to be less accurately dated. A 220 mm long example from Qesem Cave in Israel is thought to be between 420 000 and 200 000 years old, while a 265 mm-long basalt handaxe was found at Wadi Dabsa in Saudi Arabia. A site at Porto Maior in north-west Spain was occupied for a long interval across MIS 8-7, by a people of apparent African affinities, who regularly produced giant handaxes, picks and cleavers. Giant handaxes have been found at 23 sites across Africa, including Olduvai in Tanzania, where the largest such axe is 330 mm long and 1.4 million years old, Isenya in Kenya, where the largest axe is 253 mm long and 700 000 years old, and Kalambo Falls in Zambia, where the largest axe is 350 mm long and 300 years old. However, the African Acheulian, with which many sites in southern Europe and the Middle East are also associated, has a greater emphasis on large cleavers, picks, and handaxes, together referred to as 'large cutting tools'.

The MIS 9 giant handaxes from Britain appear to show a different phenomenon, a discrete technological interval in which larger, more symmetrical tools suddenly appeared. In France, the MIS 9 sites in the Somme Valley are the closest available to compare, with the Cagny l’Epinette locality producing 33 handaxes, the largest of which is still less than 170 mm in length, while the Revelles site has produced 52 handaxes, none more that 190 mm long. Ferme de l’Epinette, a MIS 10 site in the Somme Valley has produced 31 handaxes, all less than 180 mm long, while the MIS 9 stage of the Soucy complex in the Paris Basin has produced 178 handaxes, the largest of which are 155 mm long, and no handaxe from the entire complex exceeds 180 mm. The Orgnac Cave site in the Ardèche Department has a sequence of archaeological layers running from MIS 10 to MIS 8, but has produced no handaxes more than 200 mm in length. The MIS 9 layers at Menez-Dregan in Brittany have produced 29 handaxes, all less than 160 mm in length, although this may be because the tools are made from sandstone and microgranite beach pebbles which would not lend themselves to making bigger tools. This lack of large rocks for use as a source of tools may also account for the generally small handaxes found at Atapuerca in Spain.

Thus the production of giant handaxes in Britain appears to be a geographically, as well as chronologically, restricted phenomenon. This fits in with a more modern understanding of the Acheulian Culture, which includes many such small-scale variations, rather than being a single long-lived and widespread monoculture, as was once perceived. 

See also...

A possible rope-making tool from the Late Palaeolithic of Europe.

The lives led by the Palaeolithic peoples are often difficult to understand, due to the limited number of surviving technological items, and the difficulty in understanding how these were used. Rope, twine, and string are known to have been used in the Late Palaeolithic of Europe, by impressions left on clay items, by depictions in art from the period, and by traces of fibres thought likely to have come from these materials. The manufacture of rope would require specific technology to the task, the nature of which is currently not known, either because we have not found the tools required for the task, or because we have not recognised them for what they are. 

In a paper published in the journal Science Advances on 31 January 2024, Nicholas Conard of the Department of Early Prehistory and Quaternary Ecology at the University of Tübingen and the Senckenberg Centre for Human Evolution and Paleoenvironment, and Veerle Rots of the TraceoLab at the University of Liège, propose a model for the production of rope by the early Upper Palaeolithic Aurignacian people of Europe.

Thirteen pieces of worked Mammoth ivory were discovered in a horizon dated to between 40 000 and 35 000 years before the present, during excavations at the Hohle Fels Cave in the Ach Valley of the Swabian Jura of southwestern Germany in 2015, with another two pieces in a section of collapsed material which partially derived from the same horizon. The pieces were fitted together to form a baton 20.4 cm in length, 3.6 cm wide, and 1.5 cm thick, perforated by four holes, each of which was surrounded by a spiral of six precisely carved grooves. The object was obviously carefully made, and the grooves carefully chiselled, although it shows little use-associated wear.

Ivory perforated baton from Hohle Fels Cave, southwestern Germany with four views. Convex surface second from the left, flat surface far right. Distal end up, handle down. Hildegard Jensen in Conard & Rots (2024).

The Hohle Cave was occupied on numerous occasions by both Humans and Cave Bears, both of which disturbed the floor, leading to a degree of time averaging within deposits. The exact age of the baton could not be determined, however, Conard and Rots are confident that it is no more than 40 000 years old and no less than 35 000 years old, based upon dates obtained from Human-modified bones from the same layer. Furthermore, only one of the dates from this layer was older than 36 500 years, making it very likely that the baton comes from the later part of the window. 

Hohle Fels. Perforated baton of Mammoth ivory from Aurignacian layer AH Va at the time of discovery in July 2015. Conard & Rots (2024).

Four other perforated batons have previously been described from Aurignacian deposits within the Swabian Caves. One of these appears to have been almost identical to the Hohle Cave baton, and was found at Geißenklösterle Cave, 2 km downstream of Hohle Cave in the Ash Valley, in 1984. This baton was more weathered and fragmented than the Hohle Cave example, and split into 30 fragments. It was recovered from a layer which also produced a number of Aurignacian tools, and which has been dated to between 42 500 and 37 000 years before the present.

The baton from Geissenklösterle. Conard & Rots (2024).

Another three perforated Mammoth ivory batons were found in layers which produced Aurignacian tools at Vogelherd in the Lone Valley in 1931. One of these has two perforations with spiral grooves, and was interpreted as a gorget. A second baton had a single hole, again with spiral grooves, and was interpreted as a bullroarer. A third baton had at least one hole with spiral grooves, but was broken. Exact dating is not available for this site, although there is no reason to believe that the Aurignacian tools from here came from a notably different time interval from those from other sites in the Swabian Jura.

Location of Aurignacian sites in the Swabian Jura. Conard & Rots (2024).

Perforated batons have also been recorded from a range of other Upper Palaeolithic sites across Europe, although the four-holed batons from Hohle and Geißenklösterle are distinctive, with no examples from elsewhere, and the two known specimens coming from a limited geographical range and chronological interval. Notably, these are also among the oldest known examples of perforated batons. Later examples, from the Aurignacian and Gravettian periods, before the Last Glacial Maximum, and the Magdalenian period, after the Last Glacial Maximum are found across Central Europe, and are generally single-holed and made from Reindeer antler. The majority of these have smooth holes, but some have carved grooves similar to those seen in the Swabian Batons. The nature of these batons has remained obscure, with most interpretations regarding them as ritual objects, and the grooves, where present, as being symbolic in nature. Other interpretations have suggested that they were used for straightening arrow shafts, or working leather.

Macro- and microscopic images of the ivory perforated baton from Hohle Fels and residue evidence. (A) Ivory perforated baton. (B) Plant tissue extracted from Lochstab (transmitted-light microscopy). (C) Possible tracheid extracted from Lochstab (transmitted-light microscopy). (D) to (K) Details of the grooved holes according to their position on the artifact. Images are taken on the main fragmented parts of the Lochstab before refitting. Both faces of each hole are depicted in the order as depicted in (A) (distal hole: (D) and (H), note fracture in groove on (H); left central hole: (E) and (I); right central hole: (F) and (J); proximal hole/near handle: (G) and (K)). Hildegard Jensen, Dries Cnuts, & Veerle Rots in Conard & Rots (2024).

The discovery of the Hohle baton led Conard and Rots to re-examine the Geißenklösterle baton, and the extreme similarity of these objects led then to conclude that these were more likely to be tools, which would require the maker to stick closely to a predetermined pattern, rather than ritual objects, where some artistic license might be expected. The holes appear to be intended to have something threaded through them, which led Conard and Rots to hypothesise that they might have been used to make rope or twine, with the groves being used to align fibres in some way. Twine can be made by hand, simply by twisting fibres together, so Conard and Rots decided to experiment with the making of rope, using a replica of the baton, made from bone since ivory was not readily available (an attempt was first made to replicate the baton in wood, but it was found the grooves broke off too easily). This was then replaced with a baton made from Warthog tooth ivory.

Experimental pieces in bone (top) and Warthog tooth ivory (bottom). Conard & Rots (2024).

Conard and Rots began by threading different materials through the holes of the modern baton, experimenting with Deer sinews, as well as Flax, Linum, Hemp, Cannabis, Cattail, Typha, Linden, Tilia, Willow, Salix, and Nettles, Urtica. The baton proved to be ineffective for the processing of sinew, Flax, Nettles, and Hemp, but more positive results were achieved with Cattail, Linden, and Willow. The batton proved to be particularly effective for processing Cattail, crushing the hard outer surface of the stems, which produced (edible) starch as well as usable fibres. Many modern peoples are known to make rope from Cattail, and ropes made from this material have been found in a variety of archaeological contexts (albeit later than the Aurignacian). Cattail can also be used for food, cordage, and basketry. Pollen has not been studied in any Aurignacian layers within the Swabian Jura, but Cattail pollen is known from Gravettian layers at Hohle Fels, and the climate during thr Aurignacian would have been favourable for this Plant. Willow is known as charcoal from the Aurignacian layers at Hohle Fels.

Use of the ivory artefact for Tilia (left) and for Typha (right). Conard & Rots (2024).

The baton could be used to manufacture thin ropes, but was not essential, as these are not difficult to make by hand. However, experimentation with the use of all four holes on the baton found that it was particularly useful for making thicker, stronger ropes from two to four strands. By maintaining a regular thickness of each strand, it enabled the manufacture of long lengths of rope. The grooves help to break down and orient leaves while maintaining torsion, the tool can then be run over the strands at a regular speed, causing them to combine automatically into a rope as a result of the twisting tension. One person is needed to hold the tool and one to hold each strand, so three to five people are needed to make a rope with two to four strands, with the number of holes used determining the number of strands and therefore the thickness of the rope. The experiments found that it was possible to produce 5 m of strong, supple rope in 10 minutes.

Pulling the Typha through the holes and rotating fibres into strands by hand. Conard & Rots (2024).

Conard and Rots have demonstrated that the four-holed Aurignacian artefacts from Hohle Fels and Geißenklösterle are efficient rope-making tools, providing an answer to the question of how rope was produced in the European Upper Palaeolithic. Rope and twine are used for a wide variety of purposes by modern Humans, and all modern hunter-gatherer populations are known to use rope and twine for a wide variety of purposes, making it unlikely that the Aurignacian culture could have managed without it. The methodology demonstrated by Conard and Rots requires multiple people to co-operate on the manufacture of rope, and is therefore also indicative of a culture in which people had a well-developed sense of community, further underlining how this ability to co-operate would have been advantageous .

The manufacture of a three-stranded rope from Typha. Conard & Rots (2024).

See also...

Nautilus shell beads from the End Pleistocene and Early Holocene of Alor Island, Indonesia.

Body ornamentation appears to have been a key piece of Human behaviour from the dawn of our species, and in Island Southeast Asia traces of such ornamentation, or other forms of artistic expression, is often the first evidence of Homo sapiens moving into an area. For a long time, rock art and shell beads from this region have automatically been assumed to be of Holocene age, but in the past decade it has become clear that many such items are much older, and that Modern Humans may have been active in the area as much as 50 000 years ago.

In a paper published in the journal Antiquity on 15 August 2023, Michelle Langley of the Australian Research Centre for Human Evolution and School of Environment and Science at Griffith University,  Shimona Kealy of Archaeology and Natural History and the ARC Centre of Excellence for Australian Biodiversity and Heritage ar the Australian National University, Mahirta of the Departemen Arkeologi at the Universitas Gadjah Mada, and Sue O’Connor, also of Archaeology and Natural History and the ARC Centre of Excellence for Australian Biodiversity and Heritage ar the Australian National University, report the discovery of crafted beads made from the shells of Nautilus pompilius from Alor Island in the Eastern Lesser Sunda Islands, which can be directly dated to about 12 000 years before the present.

The beads were found in Makpan Cave, a lava tube between the modern villages of Halmin and Ling Al on the southwest coast of the island. The entrance to the cave is currently 386 m from the shoreline, and 37.5 m above sealevel. Since Alor Island is extremely steep sided, the cave would have been within easy walking distance of the sea, even during Pleistocene low sealevel stands.

Location of Makpan on Alor Island, Indonesia (A)–(B) and site context showing main excavation pit outlined with range poles (C)–(D). Shimona Kealy in Langley et al. (2023).

Excavations at the cave were carried out by a joint team from the Australian National University and Universitas Gadjah Mada in June-July 2016. A 2 m x 2 m trench was excavated in four blocks (A, B, C, and D) in a series of 5 mm 'spits', down to a depth of 1.5 m, then a single block (B) was further excavated down to a depth of 3.5 m, where beach sand sterile of artefacts was found. Because only block B was excavated tor the full extent of the occupation, Langley et al. concentrate on this block in their study.

Stratigraphic sections for the upper 2 × 2m excavation at Makpan. Langley et al. (2023).

Dating of the sequence was based upon charcoal where this was found, and marine shells where it was not. Occupation of the site was divided into five phases. Phase 1 began about 43 076 years before the present, and lasted for about 28 000 years, incorporating spits 68-58. Phase 2 covers the Terminal Pleistocene, beginning from about 13 965 years before the present, and incorporates spits 57-37. Phase 3 is a substantial shell midden comprising spits 36-21, believed to have been laid down at the Pleistocene–Holocene transition, starting from about 11 805 years before the present. Phase 4 covers the Early-Middle Holocene, starting from about 10 430 years before the present, and comprising spits 20-9. Phase 5 comprises Neolithic deposits, starting from about 3663 years before the present, and comprises spits 8-1. There are some detectable gaps in this record, most notably one of about 3500 years between the end of Phase 2 and the start of Phase 1, with a shorted gap between Phase 3 and Phase 4. The rate at which sediments were deposited appears to have changed over time, with the highest sedimentation rate during Phase 3 and the lowest during Phase 1.

Stratigraphic sections for the lower 1 × 1m square B excavation at Makpan. Langley et al. (2023).

Langley et al. recovered at total of 577 shell beads and bead fragments, of which 36 were deemed to be complete or near complete (i.e. more than 75% of the original bead was present. Of the 36 complete or near complete beads, 33 are of an ovoid shape, with two central holes, while the remaining three are roughly circular with a single hole at their centre.

‘Intact’ (more than 75% intact) beads recovered from Makpan. Bead provenance (Square and Spit) is listed below each artefact, along with their analysis identifier shown in parentheses. Michelle Langley in Langley et al. (2023).

The majority of the beads were found within the midden layer, and are estimated to be dated to between 11 805 and 11 223 years before the present. However, some beads were found within Phase 4, potentially as late as 7284 years before the present, as well as in phases 2 (11 000-13 500 years before the present) and 1 (21 500-40 500 years before the present). The layers within the deposits at Makpan are generally considered to have good stratigraphic integrity, with little movement between the layers, however, small beads would have a high potential for working themselves downwards in loose deposits, so, since the lowest two beads came from spits 63 and 60 (likely to be more than 20 000 years old, and potentially significantly so), Langley et al. decided that it would be prudent to directly date these before including them in the study. Unfortunately, it was not possible to relocate either of these beads, which had apparently been lost since the excavation. 

A direct radiocarbon date was therefore obtained from the third and fourth lowest beads found, a fragment and whole bead recovered from spit 58, at the end of Phase 1. This produced ages of 12 800-11 400 years for the fragment and 12 440-11 810 for the whole bead, slightly younger than a marine shell from spit 57 (the beginning of Phase 2), which was dated to 13 276-12 930 years before the present. This supports an End Pleistocene age for the beads, and suggests a revision of a dating system where Phase 2 starts with spit 57 may need to be revised, given that direct dates have not previously been obtained from spits 58 or 59, and the similarity between the deposits of phases 1 and 2, which makes differentiating a boundary difficult.

All of the beads appear to have been made from the shells of Nautilus pompilius, a Cephalopod Mollusc with an nacreous aragonitic shell which typically lives 200-300 m beneath the surface, but can survive as deep as 800 m. Nautilus are occasionally eaten by Humans, but are more often collected for their distinctive shells, which have an attractive orange, brown, and white pattern on their outer surface and an  iridescent nacreous interior.  Nautilus are caught by fishermen in the Philippines today, who use Bamboo and Rattan traps lowered to depths of 150-200 m. It is possible that the ancient inhabitants of Alor Island caught Nautilus, either intentionally or as a bycatch from other fishing activities (they are known to have been proficient fishermen), but more likely that they were collected from beaches, where they still wash up today.

Nautilus pompilius shell reaches to around 200mm in length, providing a large quantity of nacreous shell for material culture production. Michelle Langley in Langley et al. (2023).

All the intact and fragmentary beads were examined under a light microscope to look for traces of manufacture and use. Nacre is also known to have been used to make fishhooks at Makpan, so it was particularly important to be able to differentiate between bead fragments and fishhook fragments. Usefully, the examined fishhooks were all found to have been made from the shells of marine Gastropods of the genera Rochia and Turbo. Thus, the beads tend to be smaller and flatter than the hooks, which have a distinctive plano-convex shape. 

Typical fragments of two-holed beads. Scale bar is 1mm. Michelle Langley in Langley et al. (2023).

All the beads from Makpan are small, with intact specimens having a maximum width of between 4.4 mm and 11.6 mm. The overall shape of the bead, oval or circular, appears to have been achieved by controlled flaking, followed by grinding of the outer surface, though a significant proportion were never ground and retain a scalloped outer surface. The perforations were made with a handheld unifacial drill, which left striations on the shell surface, which do not form a full circle and have a wavey trajectory. This drilling was done from the inner, nacreous side of the shell, resulting in chips breaking away from the outer surface as the drill broke through. The striations left by the drilling suggest a stone tool was used.

Features of the Makpan Nautilus two-holed beads. (A) Near-complete example for Square B, Spit 18 with red ochrous residue between and in perforations; (B) & (C) Chipping on the outer shell surface created during unifacial drilling. From Square B, Spits 36 and 23, respectively; (D) & (E) Striations and red ochrous residue on back of bead from Square B, Spit 23; (E) Hand-held drilling of perforations on bead from Square B, Spit 23; Restriction of red ochrous residue to between the two perforations on bead from Square B, Spit 36; (G) polish on edge of bead from Square B, Spit 32. Scale bars are 1mm. Michelle Langley in Langley et al. (2023).

Surprisingly for small objects recovered from a site in a tropical environment, use wear and evidence of polishing can be observed on the beads. The high points of the outer (non-naceous) side of the beads are well polished, and the outer edges of the beads appear well rounded. The perforations are also well rounded, showing notching associated with a thread or string, with a distinct wear-notch between the two holes on the double-holed beads, as well as often having traces of a red, haematite-based pigment. This suggests that the beads were sewn, with their shiny nacreous side out, onto some form of soft fabric as a form of appliqué.

Indications of the mode of attachment on the two-holed beads. (Above) Examples of larger fragments from two-holed beads; (Below) mid-region of two-holed beads showing concentrations of residues and notches (indicated by red arrows). (A) Artefact 131; (B) Artefact 117; (C) Artefact 155. White scale bars are 1mm. Michelle Langley in Langley et al. (2023).

Many of the fragmentary beads appear to have broken across the centre part of the bead, and from the perforations to the outer edge by the shortest route. This supports the idea that they were sewn tightly to a cloth surface, creating a pressure point in the centre. Several beads also show signs of being delibarely ground or scratched with an rock on their exposed surface, possibly to increase their reflectivity. 

Langley et al. were able to split the double holed oval beads into four distinct groups, which they term A, B, C, and D. Type A have a distinctive elongated ellipse shape, while types B, C, and D are closer to round in shape, but fall into distinct size categories, Type B having a mean width of 10.5 mm, Type C having a mean width of 7 mm, and Type D having a mean width of 5.5 mm. Type A, B, and D beads were found throughout the sequence, but type C beads were only found in spits 30-34 (within the End Pleistocene shell midden).

Nautilus shell beads first appear at Makpan between 12 800 and 11 400 years ago and persist for about 5000 years. These beads present the oldest know evidence for appliqué decoration in Island Southeast Asia, although pigment-stained fragments of Nautilus shell have been found at Asitau Kuru on Timor-Leste which date back to at least 42 000 years ago. Further afield, the use of red pigments for both body decoration and for other forms of art is documented in Southeast Asia from a range of Pleistocene and Holocene sites, and such pigments appear to have been adopted by Homo sapiens in Africa and the Levant early in the history of our species. The Makpan beads appear to represent a new example of a persistent habit of combining red pigments with light or bright coloured objects.

Appliqué beads are widely documented in the Eurasian Palaeolithic archaeological record, using both shell and ivory as a material. Examples of this include the Arene Candide 1 burial from Italy, dated to between 28 000 and 27 000 years before the present, which had a cap with appliqué shell beads, and the Sunghir individuals from Vladimir Oblast in southern Russia, dating from between 33 000 and 30 000 years before the present, which had garments covered with ivory beads, both of which are attributed to the Gravettian Culture. The Gravettian is the oldest known culture to which the use of appliqué can confidently be attributed, although some older Aurignacian sites, such as Hohle Fels in Germany and Solutré in France have yielded beads which could have been used in appliqué decoration.

Although the Makpan beads are much later than the Eurasian examples, they form part of a shared cultural tradition found across several islands in southern Wallacea. Similar double-holed beads have been found on Kisar Island to the east of Alor Island, and Timor-Leste to the south, with one bead from Matja Kuru 2 on Timor-Leste directly dated to between 10 110 and 9629 years before the present, and another from Here Sorot Entapa on Kisar Island dated to 12 019–12 400 years before the present. This shows that the decorative tradition was shared between at least three islands in the region during the End Pleistocene and Early Holocene. Making garments covered with appliqué beads in this way would have required a considerable investment in time and labour, and such effort across multiple sites implies a shared worldview and culture across these sites, and the transmission of social information between these sites. 

The date at which these beads first appear correlates with the first appearance of obsidian tools on these islands; made with a material not found on any of the islands, and comes only very slightly after the appearance of fishhooks in the area, itself a manifestation of an increased investment in gaining resources from the sea. Polished shell adzes, considered another part of this shift to utilizing maritime resources, first appear on Obi Island in the Maluku islands in north-eastern Wallacea, and reach Timor by the early Holocene. These adzes are a rare part of these assemblages, with the oldest examples identified by pieces of shell which flaked from them during drilling, making it likely that the technology was in use quite a while before the oldest known examples which have been found.

It is likely that a trade  and social network which passed along technologies such as appliqué beads and materials such as obsidian would also have involved some genetic exchange, important for isolated communities on small islands, and such interchange of marriage partners across islands would have further deepened cultural ties and furthered the exchange of ideas, something potentially more important than the exchange of physical goods between islands, and which has been supported by recent genetic studies of End Pleistocene populations in Wallacea.

See also...

Follow Sciency Thoughts on Facebook.

Follow Sciency Thoughts on Twitter.