Research note · Australian archaeology
In a rock shelter in Amurdak Country, in western Arnhem Land, a Rainbow Serpent stretches more than six meters along the ceiling. Paul Taçon, who has worked on Arnhem Land rock art for forty years, told Phys.org that he noticed the teeth and the tongue before anything else.
It is one of twenty-nine Rainbow Serpent paintings recorded across twenty-one sites at Awunbarna, also known as Mount Borradaile. The concentration is unusual. Of the fifty-two Rainbow Serpent paintings classified as recent across the whole of western Arnhem Land, twenty are inside Awunbarna’s twenty-seven square kilometres. At Namunidjbuk, to the northeast, researchers recorded two across a survey of three hundred and fifty-two sites.
The paintings differ in scale as well as in number. The recent examples are generally larger than the older Yam Figure style Rainbow Serpents, which the authors note began to be made more than six thousand years ago.
In a 2026 study published in Australian Archaeology, Paul Taçon, Sally May, Andrea Jalandoni and senior Amurdak Traditional Owner Charlie Mungulda argue that many of the largest paintings were made after European contact. They interpret the increase in scale and concentration as a response to the cultural upheaval that followed.
If that reading is right, it raises a question that has little to do with serpents in particular. A figure that had been painted in this landscape for thousands of years was suddenly being painted bigger, and more often, in one place. What was it responding to?
What the evidence establishes
The survey itself is substantial. Since 2018 the team has recorded 165 of the 167 known rock art sites at Awunbarna. Rainbow Serpents appear at twenty-one of them, twenty-nine paintings in all.
Placed against the wider region, the figures are these. Western Arnhem Land now has 139 known Rainbow Serpent rock paintings: seventy in the Yam Figure style, fifty-two classified as recent, seventeen transitional. Awunbarna holds twenty of the fifty-two recent ones, 38.5 per cent of the regional total, inside an area you could walk across in a morning.
Rainbow Serpent rock paintings in western Arnhem Land, by style
Recent paintings at Awunbarna: 20 of 52.
Of the twenty-nine Rainbow Serpents at Awunbarna, the authors argue that twenty were painted after European contact.
It matters how that conclusion was reached. These are not twenty independently dated paintings. Rock art of this kind rarely submits to direct dating, and the attribution here rests on inference: stylistic comparison with material whose age is better understood, the layering of images painted over one another, and the characteristics of the pigments, including the use of white pipe clay. The authors also note that the color combinations and infill of the Awunbarna Rainbow Serpents are consistent with rock art dated to the past few hundred years elsewhere in the region.
This is ordinary archaeological reasoning, and it is reasonable. It is also an argument rather than a measurement, and it should be held as one.
The second claim is of a different kind again. That the paintings grew larger and more numerous is a description of the record. That they did so because of what colonization brought is an interpretation of the record. Taçon has put the reasoning plainly: the Rainbow Serpent, as the most powerful of Ancestral Beings, could be invoked through depictions, stories and ceremonies to give people some stability and control over their destiny at a time of rapid change.
That is a coherent explanation, offered by the people best placed to offer it. It is not the same kind of statement as a count of twenty-nine paintings at twenty-one sites.
Whose paintings these are
Charlie Mungulda, senior Amurdak Traditional Owner, is a co-author of the study. Not a source consulted, not a guide who showed researchers where to look; an author, named alongside the archaeologists.
That is worth stating because of how easily this kind of material gets handled. Rock art of great age invites a particular register, in which paintings become clues and the people who made them become a problem to be solved. Nothing about Awunbarna supports that register. The Rainbow Serpent is an Ancestral Being within a living culture, and the paintings sit inside a landscape of stories, places and knowledge that does not reduce to what archaeology can measure. The study itself records that stories about the power of Rainbow Serpents were told at sites carrying recent Rainbow Serpent images, as part of how knowledge passed between generations.
So there are two things in view here, and only one of them is mine to examine. There is a published archaeological argument about chronology and scale, which can be read, weighed and questioned the way any such argument can. And there is a cultural tradition that belongs to the Amurdak, which is not an object of inquiry for an outsider and does not require my interpretation to be complete.
What follows concerns the first of those, and takes the second as given.
Continuity and change
Two facts sit beside each other in this study, and they do not sit comfortably.
The first is duration. Rainbow Serpent paintings have been made in western Arnhem Land for upwards of six thousand years. Seventy of the 139 known examples belong to the Yam Figure style, the oldest of them considerably older than most of what we usually call ancient. Whatever else is true, this is a figure with remarkable persistence.
The second is responsiveness. If the authors are right, the largest and most concentrated Rainbow Serpent paintings in the region were made within the last two centuries, in one place, under one specific historical pressure. The figure did not simply continue. At a particular moment it was taken up differently, painted at six meters where it had been painted at two, painted twenty times in an area where neighboring country holds one or two.
These are not contradictory findings. A symbol can be long-lived and still be used variably; most long-lived things are. But holding both at once is harder than it sounds, because each one pulls against a different and tempting conclusion.
Duration invites the thought that the figure means something stable, that what it encodes sits underneath history rather than inside it, and recurs for reasons that have nothing to do with any particular century. Responsiveness invites the opposite thought: that the figure means whatever a community needs it to mean, and that its apparent persistence is an artefact of our distance from it.
I do not think this study settles that, in either direction. It establishes that at one place, in one period, the scale and frequency of a long-standing image changed sharply, and it offers a well-reasoned account of why. What it cannot do, what no single site can do, is tell us whether that change is evidence of a symbol being adapted, or evidence of a symbol doing precisely what it had always done, under conditions that happened to make it visible.
Those are different claims. The evidence here is compatible with both.
What remains open
Three limits are worth naming, and the third is the one that gave me most trouble.
The chronology is inferred. Twenty paintings are attributed to the post-contact period on stylistic, stratigraphic and pigment grounds. That is a sound method and it is not the same as twenty dates. If the attribution is wrong for even a handful of them, the shape of the pattern changes.
It is one region, and an atypical one. Awunbarna is remarkable precisely because it is unlike its neighbors — twenty recent Rainbow Serpents against Namunidjbuk’s one. Whatever produced that concentration may be local: particular artists, particular shelters, particular histories. A finding about Awunbarna is not yet a finding about western Arnhem Land, let alone anywhere further.
Rainbow Serpent paintings recorded in two neighboring survey areas
The serpents were not alone. This is the detail that most complicates a tidy reading, and it is in the study’s own title. The post-contact increase in scale was not confined to Rainbow Serpents. The same period produced other exceptionally large paintings, including a whale and a giant, and the paper situates the serpents among them rather than apart from them. The study describes the shift as affecting exceptionally large paintings of animals, humans and Ancestral Beings alike; a change in scale that ran across subject matter rather than within one figure.
That last point matters more than it first appears. If what changed was the size of significant paintings generally, then the story is about how a community responded to upheaval through monumentality, and the Rainbow Serpent is an instance of that, not the explanation for it. Anyone arriving at this study looking for the serpent specifically, as I did, has to reckon with the whale.
The question carried forward
What this study establishes is narrow and solid: at one place, over a definable period, a long-standing image was made larger and more often, and there is a well-argued case that this had to do with what was happening to the people making it.
What it leaves open is wider. We still do not know, from this evidence, what sustains a figure across six thousand years, or whether the thing sustained is the figure or something the figure is used for.
So the question I take from Awunbarna is not about serpents.
What allows a symbol to endure when the circumstances in which it is used change?
There are at least three answers available, that its meaning is stable, that its meaning is flexible, that meaning is not the thing being carried at all, and nothing here chooses between them. That seems to me the right place to stop, and the right place to start looking.
Sources
- Paul S. C. Taçon, Sally K. May, Andrea Jalandoni and Charlie Mungulda, “Paintings of Rainbow Serpents and other big creatures of Awunbarna, Northern Territory, Australia”, Australian Archaeology (2026).
https://doi.org/10.1080/03122417.2026.2676505 (opens in a new tab) - “Aboriginal artists painted giant ancestral beings as colonization arrived, study finds”, Phys.org — source of Taçon’s quoted recollection.
https://phys.org/news/2026-08-aboriginal-artists-giant-ancestral-colonization.html (opens in a new tab) - “Massive Rock Art Imagery Recorded in Northern Australia”, Archaeology Magazine, 3 September 2026.
https://archaeology.org/news/2026/09/03/massive-rock-art-imagery-recorded-in-northern-australia/ (opens in a new tab)
The rock art photographs from this study are not reproduced here. The paper is published under a CC BY-NC-ND license, and the images carry cultural considerations beyond what a license addresses.
Hi, this is a comment.
To get started with moderating, editing, and deleting comments, please visit the Comments screen in the dashboard.
Commenter avatars come from Gravatar.