Archaeological evidence of an ethnographically documented Australian Aboriginal ritual dated to the last ice age

The study addresses how far back in time ethnographically described rituals and intangible heritage can be traced archaeologically, a question central to understanding social transmission and the longevity of oral traditions. A key challenge is that visible material evidence can be copied or reinterpreted over generations, and hermeneutic processes can alter transmitted knowledge. The authors sought evidence of practices that would not have been open to copying because the critical elements were concealed or perishable. At the invitation and under the leadership of GunaiKurnai Elders, the team excavated Cloggs Cave, a secluded site historically used not for habitation but for ritual activities by GunaiKurnai mulla‑mullung. They report two deeply buried, well‑preserved miniature fireplaces each associated with a straight, trimmed Casuarina stick, lightly singed and smeared with animal or human lipids, consistent with nineteenth‑century GunaiKurnai ritual accounts. Ethical protocols were co‑developed with the GunaiKurnai Land and Waters Aboriginal Corporation, complying with AIATSIS and Australian Archaeological Association Codes of Ethics.

Prior work highlights difficulties in tracking intangible heritage archaeologically due to reinterpretation and copying (e.g., Silliman 2009; Sheppard et al. 2004; Aswani 2000). Ethnographies from the nineteenth century (Howitt 1887, 1904; Fison & Howitt 1880; Smyth 1878; Mathew 1925; Elkin 1945) describe GunaiKurnai and neighbouring groups’ ritual practices involving secluded fires, wooden implements (murrawan), Casuarina wood (barn), and the application of human or animal fat, with chanting of victims’ names and specific installations. In Australian archaeology, well‑preserved Pleistocene–Early Holocene wooden artefacts are exceptionally rare; previously, wooden artefacts of this antiquity were mainly known from Wyrie Swamp (Luebbers 1975). Local geomorphological and archaeological studies at Cloggs Cave (Delannoy et al. 2020; David et al. 2021; McDowell et al. 2022) established long‑term ritual use indicators (stone arrangements, broken stalactites, calcite powders, quartz crystals) and the absence of food refuse, aligning with ethnographic accounts of ritual, not residential, cave use.

Fieldwork: In 2020, the team excavated a 50 × 50 cm area (square R31) against the cleaned northeast wall of the 1971–1972 excavation pit at Cloggs Cave. The stratigraphy comprises culturally sterile SU5 (megafauna layer) at depth, overlain by SU4 (22 sublayers, SU4A–SU4V) and SU2 (SU2G–SU2A) ashy anthropogenic layers. Excavation proceeded in arbitrary units (mean 2.3 cm) following stratigraphy, with all sediments bagged by unit without sieving in the field. Sediment DNA, pollen, phytolith and micromorphology samples were collected systematically. In the lab, sediments were wet‑sieved (2‑mm mesh), sorted, and residues curated under GunaiKurnai cultural protocols. Handling used powder‑free nitrile gloves; artefacts were double‑bagged without direct label contact. Chronology: A total of 69 AMS radiocarbon dates from R31 were obtained on charcoal (n=29), herbivore scats (n=33; common brushtail possum), unburnt wood (n=5) and bark (n=2). Ten single‑grain OSL ages were also acquired from nearby stratigraphy. Bayesian modeling of 69 AMS + 2 OSL ages indicates XU43–XU1 span 25,740–1,460 cal BP, with good chronostratigraphic order above the SU4U–SU4V interface. Direct AMS dating of one stick: Wk‑50278, 10,361 ± 30 BP (11,930–12,440 cal BP); the other: Wk‑50968, 9,726 ± 21 BP (10,870–11,210 cal BP). The miniature fireplace contexts were modeled at 11,420–12,950 cal BP (XU11; SU4E) and 10,720–12,420 cal BP (XU8–9; SU4D). Wood identification: Small fragments were examined under reflected light microscopy across transverse, tangential, and radial planes following IAWA criteria (163 features assessed). Anatomical traits (large multi‑seriate rays, diffuse to semi‑diffuse porosity, vessel characteristics, absence of Allocasuarina features) identify both sticks as Casuarina cf. cunninghamiana or C. glauca. Residue analysis: Aseptic low‑power microscopy (Dino‑Lite AM4815ZT, ×25–×85) identified residue locations. Extractions used ultra‑purified water via hydrostatic pressure (Venturi effect) and adapted Picrosirius Red staining, with high‑power microscopy (Leitz Dialux 22, ×200–×400) to characterize residues. Four samples per stick were analyzed, revealing lipid/tissue residues on both sticks, particularly extensive on the XU8–9 stick. Lipid chemistry: Residues were eluted from slides with LC‑MS–grade acetonitrile and analyzed by LC‑MS (Exploris 120 Orbitrap, 12‑min gradient, 60,000 resolution, positive mode) and derivatized for GC‑MS (TSQ 9000; 23.5‑min run). Compound identification used mzCloud, ChemSpider, mzVault, and MassList. Multiple saturated, mono‑ and polyunsaturated fatty acids, glycerides, and dicarboxylic acid degradation products characteristic of animal/human adipose fats were identified. Ethics and governance: Research was initiated by and conducted with the GunaiKurnai Land and Waters Aboriginal Corporation, governed by a memorandum of understanding (23 Oct 2018), Cultural Heritage Permit GKRAP‑19‑0001, and AIATSIS and AAA Codes of Ethics. Indigenous materials remain property of the Traditional Owners.

• Two miniature fireplace installations were excavated in square R31: one in XU11 (SU4E; 43.2–47.5 cm depth) dated 11,420–12,950 cal BP (Bayesian modeled) with a direct AMS date on the stick Wk‑50278 10,361 ± 30 BP (11,930–12,440 cal BP); and one in XU8–9 (SU4D; 39.4–42.5 cm depth) dated 10,720–12,420 cal BP with a direct AMS date on the stick Wk‑50968 9,726 ± 21 BP (10,870–11,210 cal BP).
• Each installation consists of a small, low‑temperature fireplace and a single straight, trimmed Casuarina stick protruding from or adjacent to the ashy area. The XU8–9 fireplace was bounded by eight lightly burnt limestone rocks (159.7–2,106.4 g each), with maximum burnt area 15–20 cm diameter and 3.1 cm ash thickness, containing only 22 tiny charcoal pieces (0.29 g total).
• The sticks (lengths: 39.5 cm and 19.7 cm; diameters: ~1.3–2.5 cm and 0.7–1.2 cm) were deliberately prepared: smaller lateral twigs were cut or scraped flush to create smooth shafts; both tips were lightly charred with charring limited to 1–3 mm depth, indicating very brief burning; one stick shows a hooked proximal end broken green.
• Residue microscopy and lipid analyses detected abundant animal/human fatty residues on the sticks, especially on the XU8–9 stick. Identified compounds include palmitic, stearic, oleic, linoleic, linolenic, palmitoleic acids; animal‑specific fatty acids (e.g., pentadecanoic, myristoleic, margaric acids); multiple mono‑/di‑/triglycerides; and dicarboxylic acid degradation products (e.g., sebacic, adipic, pelargonic acids), consistent with animal/human fat application.
• A single intact, unburnt wombat (Vombatus ursinus) scat was found placed on the edge of the XU11 fireplace; its position suggests intentional placement as part of the installation.
• Other associated finds include numerous small unburnt Casuarina wood fragments (XU8–9) and a small charcoal fragment (0.15 g) identified as Hedycarya angustifolia (native mulberry; used ethnographically for fire drills), with other tiny charcoals likely from the same taxon/family (Monimiaceae).
• The installations’ small size, paucity of fuel/charcoal, carefully trimmed sticks, and fatty residues are inconsistent with cooking/heating fires and align closely with nineteenth‑century GunaiKurnai ethnographic descriptions of secluded ritual practices using Casuarina sticks smeared with fat set before a small ritual fire and empowered through song.
• These finds represent Australia’s oldest known wooden artefacts and demonstrate continuity of a specific GunaiKurnai ritual practice over ~12,000 years (~500 generations).

The study set out to identify archaeological signatures of an ethnographically documented, intangible ritual practice that could not have been perpetuated through simple copying of visible features. The two Cloggs Cave installations meet this criterion: they were constructed in a secluded cave context not used for habitation; key elements (Casuarina sticks with applied fat) are perishable and were rapidly buried, limiting visibility to later observers; and the association with animal/human lipids would have been invisible to casual inspection. The close correspondence between the miniature fireplace‑and‑stick configurations, the use of Casuarina, the application of fat, and the ritual context described in nineteenth‑century GunaiKurnai sources strongly indicates continuity of practice. The convergence of independent lines of evidence—secure chronostratigraphy and AMS dates, rare preservation of wood, residue and lipid chemistry, contextual artefacts (crystals, calcite, standing stone) and ethnographic accounts—supports the inference that very specific ritual knowledge was transmitted with minimal change for approximately 12 millennia. This finding addresses the broader question of the longevity of oral traditions by providing a materially anchored example of long‑term cultural transmission.

Archaeological excavations at Cloggs Cave uncovered two securely dated Late Pleistocene–Early Holocene ritual installations comprising miniature fireplaces and carefully trimmed Casuarina sticks bearing animal/human lipids, closely matching nineteenth‑century GunaiKurnai ritual descriptions. These constitute Australia’s oldest known wooden artefacts and provide strong evidence for the transmission of a specific ritual practice over ~12,000 years. The study demonstrates how perishable and concealed ritual components, when preserved in situ and integrated with residue chemistry and robust chronology, can ground assessments of deep‑time continuity in intangible cultural heritage.

• Inference is based on material correlates of behavior; performances themselves cannot be directly observed, leaving room for interpretive uncertainty.
• Some taxonomic identifications are limited to genus level; Casuarina species could not be distinguished beyond cf. cunninghamiana/glauca via wood anatomy.
• Lipid analyses indicate animal or human fats but do not distinguish specific species or definitively separate human from non‑human sources.
• Minor chronostratigraphic reversals occur at deeper levels (SU5 to SU4U–SU4V interface) due to roof fall, though the contexts of interest above this interface are well ordered.