After the Blast: Recovering Juukan Gorge’s Deep History

A stone cutting tool from Juukan Gorge still carries its handle: a mixture of spinifex resin and kangaroo dung, according to excavation director Michael Slack. Reported as approximately 3,000 years old, it preserves something that stone implements usually lose before archaeologists find them: the material that made a sharp edge usable in the hand. Such survivals bring the practical details of an ancient campsite into focus. They also make the circumstances of this excavation difficult to comprehend.

In its 25 October 2026 report, ABC News describes some 50,000 artefacts recovered during investigations commissioned by the Puutu Kunti Kurrama and Pinikura Aboriginal Corporation. Scarp Archaeology is excavating beneath a protective dome at the damaged shelter. The work belongs to a programme of recovery and rehabilitation controlled by the Traditional Owners, whose relationship with this place survived the destruction of its architecture. [Nicolas Perpitch’s report](https://www.abc.net.au/news/2026-10-25/juukan-gorge-archaeological-dig-uncovering-ancient-secrets/107298466) provides the immediate account of the continuing excavation.

On 24 May 2020, Rio Tinto blasted the Juukan Gorge shelters during expansion of its Brockman 4 iron ore mine in Western Australia’s Pilbara. The parliamentary inquiry subsequently documented the destruction of places of archaeological importance and living cultural significance. Its account makes clear why describing the event simply as the loss of an ancient site is inadequate: the shelters belonged within an existing Aboriginal cultural landscape. [The inquiry’s interim report](https://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Former_Committees/Northern_Australia_46P/CavesatJuukanGorge/Interim_Report/section?id=committees%2Freportjnt%2F024579%2F75133) records both the event and testimony concerning that relationship.

What does 50,000 years mean?

The age claim requires care. Slack and colleagues’ 2024 study combined the 2008 and 2014 excavations at Juukan 2. Radiocarbon measurements, sediment dating and Bayesian modelling placed the beginning of its archaeological sequence between 47,130 and 42,140 calibrated years before present, at 95.4 per cent probability. The study also identified chronological discontinuities. Its results establish great antiquity without resolving an uninterrupted sequence of individual visits. The approximately 50,000-year description in current reporting should therefore be distinguished from this published estimate and from any revised chronology emerging from subsequent work. [The primary study](https://doi.org/10.1016/j.quascirev.2024.108823) remains the essential reference for the earlier excavation.

Archaeological dates answer particular questions. Radiocarbon analysis of charcoal estimates the age of the carbon in burnt wood. It does not automatically date the manufacture of a nearby stone tool. Even charcoal lying within a fireplace may contain wood that was already old when collected. Laboratories remove contaminants through physical cleaning and chemical treatment, but sample selection and archaeological association remain indispensable. The [Waikato Radiocarbon Dating Laboratory’s guidance](https://www.waikato.ac.nz/research/research-services-facilities/radiocarbon-dating/operating-procedures/pretreatment/) explains why identifying short-lived material can help avoid this inherited age.

Optically stimulated luminescence approaches the deposit differently. It measures energy accumulated in mineral grains since their previous exposure to daylight. Under suitable conditions, the result estimates when sediment was buried. Insufficient light exposure before burial or later movement of grains can complicate that relationship. Agreement between independent dating methods is valuable because they measure different processes; neither method removes the need to understand how the deposit formed. [Historic England’s scientific dating guidance](https://historicengland.org.uk/images-books/publications/scientific-dating-pleistocene-sites/scientific-dating-methods/) describes these principles and limitations.

A model can combine dates with the order of archaeological layers to estimate a chronology more effectively than isolated measurements. Its precision nevertheless depends on the evidence and assumptions supplied. A boundary representing the earliest surviving material is not necessarily the date when people first entered the district. Earlier visits may have left no recoverable trace, or their deposits may have disappeared.

Likewise, repeated occupation over many millennia does not require permanent residence. A shelter might have been used seasonally, occasionally or intensively during particular generations. Establishing which pattern prevailed requires the distribution of hearths, tools and food remains, considered against the rate at which sediment accumulated. A long cultural relationship with a place and an archaeologically continuous deposit are different propositions.

The shelter as an ecological record

Juukan also preserves evidence that people were only some of the animals contributing to its deposits. A September 2026 paper by Timothy Churchill and colleagues identifies Tasmanian devil remains from the shelter: a jaw fragment from Holocene deposits and a premolar fragment from much older sediments. Associated charcoal brackets the jaw at approximately 7,300–6,900 years ago; luminescence samples place the older tooth within approximately 47,100–42,500 years ago. These are contextual ages, rather than direct dates on the specimens. [The published analysis](https://doi.org/10.1002/arco.70031) establishes the species’ presence in the Pilbara.

The same study interprets some modified bones as evidence of scavenging, while acknowledging overlapping signatures from different chewing animals. No devil-like classifications occurred in its small Last Glacial Maximum subsample. Human consumption of the devils remains unproven, and the proposal that dingoes disrupted a longstanding relationship between people and scavengers is explicitly hypothetical. The distinction matters: identifying an animal securely does not establish every aspect of its behaviour or disappearance.

For an excavator, this creates a practical problem. A concentration of fragmented bone may include meals discarded by people, material introduced by predators, and older remains subsequently chewed or digested. These processes can operate successively on the same carcass. Taphonomy, the study of what happens to remains between death and recovery, is therefore part of reconstructing human activity, rather than an exercise performed after the human story has been written.

The question is often how several agents interacted. People could bring food into a shelter and leave portions that scavengers later consumed. Animals might redistribute the refuse, altering the apparent boundaries of an activity area. Burning, cut marks, fracture patterns and tooth damage must be examined together. Their relationships can distinguish a human food deposit from material merely found beside human artefacts.

Regional comparisons help prevent a single shelter from becoming a model for an entire landscape. At Boodie Cave on Barrow Island, microscopic sediment analysis demonstrated differences between depositional conditions near the entrance and inside the cave. Investigators identified erosion and reworking while retaining confidence in the broad archaeological sequence. [Ward and colleagues’ 2017 study](https://doi.org/10.1016/j.jasrep.2017.08.012) shows why stratigraphic integrity needs to be assessed spatially, even within one excavation.

Boodie also supplies a geographical caution. Its investigators dated initial occupation to approximately 51,100–46,200 years ago, when Barrow Island remained connected to the mainland. The shoreline then lay farther west. Modern maps can therefore mislead comparisons between coastal and inland sites. Reconstructing movement requires the geography people actually encountered, including land subsequently submerged. [The researchers’ account](https://www.abc.net.au/news/2017-05-19/boodie-cave-reveals-earliest-australians-coastal-lifestyle/8542190) explains that changing setting.

What objects retain, and what numbers conceal

The earlier Juukan study reported starch interpreted as bush potato on grinding implements, and a braided human-hair object approximately 3,500 years old. Its genetic analysis recovered a partial, low-coverage mitochondrial genome, supporting a possible connection with contemporary Indigenous people. That result should not be enlarged into a complete genealogy spanning the shelter’s entire history. Nor should cultural belonging depend on the recovery of genetic evidence. [Slack and colleagues](https://doi.org/10.1016/j.quascirev.2024.108823) provide the scientific basis and its qualifications.

The plant evidence poses its own questions. A residue can identify an activity otherwise poorly represented among durable objects, but its interpretation depends on identification, contamination controls and association with a working surface. Even a secure food residue documents processing on a particular implement; it does not by itself reveal how much of the diet that food supplied. Grinding equipment is evidence of practical knowledge, although the organisation of gathering, preparation and sharing must be investigated separately.

Current reporting also describes marine shell at Juukan. Transport from the coast is a reasonable inference; a particular trading route is harder to establish. An object could pass between several communities or travel with someone moving directly across the landscape. Its date, modification and geological or biological source would help distinguish possibilities. The presence of a distant material establishes a connection more readily than it reconstructs the social arrangements behind it. [The ABC account](https://www.abc.net.au/news/2026-10-25/juukan-gorge-archaeological-dig-uncovering-ancient-secrets/107298466) reports the find.

Large artefact totals require similar restraint. Fifty thousand pieces cannot be translated into a population estimate. One episode of stone working can produce numerous flakes, while an extended visit may leave relatively few durable objects. Excavation area, sieve size, fragmentation and recording conventions all affect the final count. Comparisons become useful when researchers measure material against excavated volume, chronology and the activities that generated it.

This matters especially when comparing a limited test excavation with a larger recovery programme. More finds may represent a larger sample, rather than a previously unsuspected increase in ancient occupation. The stronger questions concern associations: whether tools occur beside fireplaces, whether raw materials change through time, and whether food residues correspond with the equipment interpreted as preparing them.

Recovery has a boundary

Rebuilding the shelter introduces another distinction. CDM Studio describes developing a reconstruction through digital modelling, full-scale prototypes and review by Traditional Owners drawing on memory and lived knowledge. The corporation leads decisions and approvals. Such work can restore a sheltered place and support future visits, but reconstructed architecture cannot recreate ancient deposits or the relationships erased by blasting. [The project description](https://cdmstudio.com.au/project/juukan-2/) sets out the reconstruction process.

Archaeologist Jo McDonald argued after the destruction that heritage management needed stronger Indigenous authority, secure keeping places for recovered objects and conservation planned at landscape scale. Her argument remains relevant to the interpretation of recovery itself. Excavating material successfully is one responsibility; deciding what should survive in place is another. [Her 2020 assessment](https://www.uwa.edu.au/news/article/2020/october/can-a-mining-state-be-pro-heritage-vital-steps-to-avoid-another-juukan-gorge) addresses that institutional problem.

At Juukan, the evidence already establishes ancient Aboriginal use, varied technologies and a shelter shared with other animals across major environmental changes. Continuing excavation can refine the timing and character of those activities. The most demanding task is to connect individual finds to securely dated episodes while distinguishing surviving ancient context from disturbance. Recovery will be judged by how much of that relationship can still be demonstrated, alongside the knowledge and priorities of the people whose place it remains.

Selected Sources and Further Reading

1. Slack, M. J., et al., including the Puutu Kunti Kurrama and Pinikura Aboriginal Corporation. 2024. A 47,000 year archaeological and palaeoenvironmental record from Juukan 2 rockshelter on the western Hamersley Plateau of the Pilbara region, Western Australia. Quaternary Science Reviews 338, 108823. [Primary research](https://doi.org/10.1016/j.quascirev.2024.108823).
2. Churchill, T. J., Slack, M., Matheson, E., Neill, L., Saktura, W. M., and Ashburton, G. 2026. First Record of Tasmanian Devils in the Pilbara: Late Quaternary Subfossils From Juukan Gorge, Puutu Kunti Kurrama Country, and Insights Into Human—Faunal Relationships and Extinction. Archaeology in Oceania, published online 10 September. [Primary research](https://doi.org/10.1002/arco.70031).
3. Ward, I., et al. 2017. 50,000 years of archaeological site stratigraphy and micromorphology in Boodie Cave, Barrow Island, Western Australia. Journal of Archaeological Science: Reports 15: 344–369. [Primary research](https://doi.org/10.1016/j.jasrep.2017.08.012).
4. Joint Standing Committee on Northern Australia. 2020. Never Again, interim report, chapter 1: Juukan Gorge. Parliament of Australia. [Report and testimony](https://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Former_Committees/Northern_Australia_46P/CavesatJuukanGorge/Interim_Report/section?id=committees%2Freportjnt%2F024579%2F75133).
5. Perpitch, N. 2026. How archaeologists are uncovering the ancient secrets of WA’s Juukan Gorge. ABC News, 25 October. [Excavation reporting](https://www.abc.net.au/news/2026-10-25/juukan-gorge-archaeological-dig-uncovering-ancient-secrets/107298466).
6. Ulm, S., Ward, I., Veth, P., and Manne, T. 2017. Remote cave off Western Australia reveals earliest Australians’ coastal lifestyle. ABC News/The Conversation, 19 May. [Researchers’ account](https://www.abc.net.au/news/2017-05-19/boodie-cave-reveals-earliest-australians-coastal-lifestyle/8542190).
7. McDonald, J. 2020. Can a mining state be pro-heritage? Vital steps to avoid another Juukan Gorge. University of Western Australia, republished from The Conversation. [Heritage analysis](https://www.uwa.edu.au/news/article/2020/october/can-a-mining-state-be-pro-heritage-vital-steps-to-avoid-another-juukan-gorge).
8. Historic England. Scientific Dating of Pleistocene Sites: Scientific Dating Methods, including luminescence dating guidance by Geoff Duller. [Methods reference](https://historicengland.org.uk/images-books/publications/scientific-dating-pleistocene-sites/scientific-dating-methods/).
9. University of Waikato Radiocarbon Dating Laboratory. Pretreatment. [Laboratory guidance](https://www.waikato.ac.nz/research/research-services-facilities/radiocarbon-dating/operating-procedures/pretreatment/).
10. CDM Studio. Juukan 2. [Reconstruction project description](https://cdmstudio.com.au/project/juukan-2/).