📜 LEGEND · EXP. Nº0270 ACTIVE

The Rainbow Serpent

The Rainbow Serpent is a legend from Ubirr, Kakadu National Park (Northern Territory), AU — a folk story passed down over generations. This file collects its origins, its meaning and how it has been retold.

Where does the legend of The Rainbow Serpent come from?

Location
Ubirr, Kakadu National Park (Northern Territory), AU
Kind of legend
Myth
Coordinates
-12.411, 132.958
Testimonies
0
Last updated
LOCATION

What is the legend of The Rainbow Serpent?

Two words in this file's title are English inventions with a date. "Dream-times" first appears on page 185 of a paper by Francis James Gillen, the Alice Springs telegraph stationmaster, in the 1896 Horn Expedition report; it does not survive into Spencer and Gillen's famous 1899 book, which glosses Alcheringa as "the far past times" and uses the word "dream" five times in 240,000 words, never as a translation. "Rainbow Serpent" was fixed as a pan-Australian category by A. R. Radcliffe-Brown in 1926, and the printed index to volume I of Oceania shows the merger happening: three fieldworkers describing three different regions, filed in 1930 under one heading. The beings underneath are not one. In Bininj Kunwok, Ngalyod is female, the daughter of Yingarna the mother rainbow serpent, and the dictionary speaks of rainbow serpents in the plural, living in deep waterholes, still detectable by marrkidjbu. The oldest evidence is painted: about 6,000 years old, and shaped, one archaeological argument holds, like a small marine pipefish.
The Rainbow Serpent, Ubirr, Kakadu National Park (Northern Territory), Australia
The Commons file does not record which site this is. What it shows is not the coiled snake a reader expects but a thick, arched figure with a cranial protuberance — the horns the Bininj Kunwok dictionary registers as kundjilinj, “the horns of ngalyod”. Image: The original uploader was Digitaltribes at English Wikipedia . · CC BY 2.5

What is the history of The Rainbow Serpent?

Two of the four words in the title of this file are English inventions with a date, an author and a page number. The beings they were placed on top of are older, more numerous, and mostly female. Start with the second word. In September 1896 the Carlton printers Ford and Son turned out Part IV — Anthropology of the Report on the Work of the Horn Scientific Expedition to Central Australia, edited by Baldwin Spencer. Pages 161 to 196 carry a separate paper by Francis James Gillen, the post and telegraph stationmaster at Alice Springs, who was also Special Magistrate and Sub-Protector of Aborigines. On page 185, explaining how the Arrernte account for fire, Gillen writes that their ancestors lived in the distant past, ulchurringa, "which really means in the dream-times, for this is the manner in which the natives always speak of the long ago". That sentence is the earliest use on record of the English expression. It is plural, it is parenthetical, and it belongs to a stationmaster, not to a professor. The expression then vanishes. Spencer and Gillen's famous monograph of 1899, The Native Tribes of Central Australia, runs to roughly 240,000 words and uses "Alcheringa" 298 times; the word "dream" occurs five times in the whole book and never once as a translation of it. The footnote on page 73 says only "the name applied to the far distant past with which the earliest traditions of the tribe deal", and the glossary repeats it: "Name given to the far past times in which the mythical ancestors of the tribe are supposed to have lived". Far past times, not dream times. When W. E. H. Stanner opened his 1953 essay The Dreaming by crediting "Spencer and Gillen" with having "immortalised" the phrase, he was, as the linguists Cliff Goddard and Anna Wierzbicka put it in 2015, "speaking somewhat loosely": they trace it back past Spencer, past the 1899 book, to Gillen alone, in 1896, on page 185. Behind Gillen stand the Arrernte speakers who gave him the word, and they are named nowhere. Now the label. In 1926 A. R. Radcliffe-Brown published "The Rainbow-Serpent Myth of Australia" in the Journal of the Royal Anthropological Institute (volume 56, pages 19-25), assembling accounts from peoples across the south-east — Kamilaroi, Yuwaalaraay, Wiradjuri, Wongaibon, Weilwan — under a single heading. Catherine H. Berndt, writing the Encyclopedia of Religion entry on the subject in 1987, summarises the position he reached as making the being "perhaps the most important nature-deity", the embodiment of "the creative and destructive power of nature". The moment the category hardened can be read off a printed index. The October-December 1930 issue of Oceania, the journal's first volume, carried three papers on the same theme at once: Radcliffe-Brown on the south-east, Ursula McConnel on North Queensland, A. P. Elkin on the north-west, with Ralph Piddington on the Karadjeri water-serpent alongside. When the volume was indexed, all of them were filed under one entry — "Rainbow-Serpent— in North Queensland .. 347, 348 / in North-West Australia 349-351 / in South-East Australia 342-346" — with a cross-reference reading "Serpent myths. See Rainbow-Serpent". Three regions, three fieldworkers, three sets of beings, one heading. The pan-Australian Rainbow Serpent is visible there being made. The beings themselves are the part the label flattens. In Bininj Kunwok, the language of western Arnhem Land and part of Kakadu, the community language centre's dictionary carries ngalyod as a headword: "Rainbow serpent, daughter of the Mother One Rainbow Yingarna. The rainbow serpent is a central figure in traditional Bininj religion. She is the creator of places and is active during the wet season and is associated with rain and bodies of water. Rainbow serpents live in deep waterholes and rivers. Marrkidjbu are able to discern their presence and control them." Four things in that entry contradict the standard English retelling. She is female — the dictionary's grammar note explains that "the feminine prefix reflects the conception of the rainbow serpent as a generative figure, 'the mother one'". She is a daughter, not a first cause: Yingarna is "Mother rainbow serpent, the mother of ngalyod, really 'biggest one'", a "female creation ancestor, who emerged from the sea to the north and travelled throughout many lands", dropping off particular people and foods in particular places and telling each which language to speak. There is more than one of them: the entry says rainbow serpents, plural, live in deep waterholes. And they are not left in the past — marrkidjbu, men with special power, can still detect and control them. The same dictionary shows how deep the being runs in the vocabulary. Delek, the white ochre used in painting, is defined as kaolin and huntite and then as "the faeces or vomitus of ngalyod", with a named djang or sacred site for it at kudjarnngal near Manmoyi. Larraddi is an intransitive verb that exists for one purpose: "there is a light given off by Ngalyod (the Rainbow Serpent), esp. in the sky in the distance". Djarnkerrk, the Antilopine kangaroo, is "offspring of Yingarna the Rainbow Serpent". The dialect table under ngalmudj lists the regional forms in a row: ngalyod in Kunwinjku and Kuninjku, ngalmudj in Kune, almudj in Kundjeyhmi, the Kakadu language, with bolung given as a synonym. These are not variant spellings of one name. They are the reason the English umbrella had to be invented. The oldest evidence is on rock. In 1996 Paul Taçon, Meredith Wilson and Christopher Chippindale published "Birth of the Rainbow Serpent in Arnhem Land rock art and oral history" in Archaeology in Oceania, analysing 107 painted images. They dated the earliest of them to around 6,000 years ago and argued that the odd anatomy of the early figures — the snout, the projections behind the head, the trailer, the spines on the tail — matches no snake but a small marine fish, the ribboned pipefish Haliichthys taeniophorus of the northern Australian coast, which would have arrived in country as post-glacial seas flooded the plains. At Ubirr in Kakadu, Parks Australia records that the Rainbow Serpent "painted her image on the rock as she passed through Ubirr to remind people of her presence" — the pronoun in the park's own text is her.

Timeline: the documented record of The Rainbow Serpent

  1. September 1896 Francis James Gillen, post and telegraph stationmaster at Alice Springs, writes on page 185 of his paper in Part IV of the Horn Expedition report that the Arrernte ancestors lived in the distant past, ulchurringa, "which really means in the dream-times, for this is the manner in which the natives always speak of the long ago". It is the earliest recorded use of the English expression, in the plural, inside a parenthesis, in a paper about how fire is made.
  2. 1899 Spencer and Gillen publish The Native Tribes of Central Australia without the gloss. In roughly 240,000 words the book uses "Alcheringa" 298 times and "dream" five times, none of them as a translation; the footnote and the glossary both render it as the far past. The English phrase that would conquer the world drops out of the book that made its authors famous.
  3. 1926 A. R. Radcliffe-Brown publishes "The Rainbow-Serpent Myth of Australia" in the Journal of the Royal Anthropological Institute, volume 56, pages 19-25, gathering accounts from south-eastern peoples under a single English heading and treating them as versions of one myth.
  4. October-December 1930 Volume 1, issue 3 of Oceania carries three regional papers at once — Radcliffe-Brown on the south-east (342-346), Ursula McConnel on North Queensland (347-348), A. P. Elkin on the north-west (349-351) — plus Ralph Piddington on the Karadjeri water-serpent. The printed index to the volume files all three under one heading, "Rainbow-Serpent", with the cross-reference "Serpent myths. See Rainbow-Serpent". The pan-Australian category can be watched being assembled in a table of contents.
  5. 1953 W. E. H. Stanner opens his essay "The Dreaming" by crediting the phrase to "Spencer and Gillen", who, he writes, "immortalised" it as "the dreamtime" or alcheringa of the Arunta. The attribution stuck for sixty years and is off by one author and three years.
  6. 1987 Catherine H. Berndt writes the Encyclopedia of Religion entry "Rainbow Snake". She reports Radcliffe-Brown's conclusion that the being is "perhaps the most important nature-deity", and adds the observation that undercuts the category: even among Aboriginal people the name can be applied to snakes with no obvious rainbow connection.
  7. 1996 Paul Taçon, Meredith Wilson and Christopher Chippindale analyse 107 painted images in Archaeology in Oceania, date the earliest Rainbow Serpent figures to about 6,000 years ago, and argue that their anatomy matches the ribboned pipefish, Haliichthys taeniophorus, a coastal fish that would have appeared inland as post-glacial seas flooded the plains.
  8. 2014 Christine Nicholls publishes a four-part series arguing that "Dreamtime" and "The Dreaming" are grossly inadequate English translations and that the words to use are the ones the languages already have: Jukurrpa, Altyerr, Ungud, Ngarrankarni, Manguny, Wongar.
  9. 28 October 2021 The historian Shino Konishi publishes an essay describing what settler Australia did with the being as "intellectual colonisation": domesticating it into something comprehensible and palatable to Western ideas, rather than understanding it. She names Yingarna and her child Ngalyod among the Kunwinjku and the Waugal among the Noongar, and notes that before colonisation there were around 250 Aboriginal languages, each with its own Dreaming stories.

Is there a rational explanation for The Rainbow Serpent?

The claim worth examining is not whether a serpent made the rivers. It is the sentence our own file used to carry: that this is "one of the longest continuously told stories on earth". That sentence fuses two different things. One is a sequence of images on rock, which archaeologists can order and roughly date. The other is an oral narrative, which nobody has recorded from before the nineteenth century. Taçon, Wilson and Chippindale's estimate of about 6,000 years is a claim about when a recognisable kind of picture first appears in the western Arnhem Land sequence — not a laboratory date on the pigment of a named painting, and not evidence that any particular story told today is the story that image illustrated. Shino Konishi, writing in 2021, states the strong version carefully, as an argument that Rainbow Serpent stories are "among the world's oldest continuous religious traditions"; that is an inference from the art, and it is worth saying so. The second thing to doubt is the singular. Catherine Berndt, in 1987, noted that the name is applied to snakes with no obvious rainbow connection at all, and the beings gathered under it do not all have serpent bodies or rainbow associations. The Bininj Kunwok dictionary uses the plural — rainbow serpents, in the waterholes — and distinguishes Yingarna from her daughter Ngalyod. The 1930 volume of Oceania shows the merger being performed editorially: three fieldworkers describing three different regions, filed under one index heading. Whatever unity the category has was in part supplied by anthropologists writing in English between 1926 and 1930, and it is fair to ask how much of the "same being everywhere" is a finding and how much is a filing decision. The pipefish argument deserves the same treatment. It is a resemblance argument — the anatomy of the early painted figures matches a living fish better than it matches a snake — and it is proposed as a hypothesis about where an image came from, not as a claim that anybody ever mistook a fish for an ancestor. This entry has not read the 1996 paper itself; it is behind a paywall, and the figures given here come from summaries of it. Finally, a caution that matters more than the others. Aboriginal knowledge is not uniformly public. Bininj Kunwok has words for the distinction: kubolkdjamun is a "sacred place, restricted place", and darrinj means "in the public domain, non-secret knowledge or a place which is open to general public access". Everything in this entry is drawn from published dictionary headwords, peer-reviewed archaeology and the national park's own signage and website — the darrinj side of that line. Where a fuller account exists, it belongs to the people who hold it.

Can you visit The Rainbow Serpent?

There is no single place to visit, and that is part of the point: the beings gathered under this English label belong to many different countries, from the Kimberley to Cape York to the Swan River. The one publicly signposted site where Australia's park service names a painted Rainbow Serpent on the rock is Ubirr, in the East Alligator (Erre) region of Kakadu National Park, Northern Territory. Parks Australia's own map point for it is -12.411359, 132.957727; that is about 32 km from Jabiru, the park township, and about 230 km from Darwin, both measured in a straight line, so allow more by road. Ubirr is reached on a sealed 2WD road and the walk through the galleries is short, with a steeper climb to the Nadab lookout at the end. The main gallery also holds X-ray paintings of fish, waterfowl, mussels, wallabies, goanna and yams, most of them from the last 1,500 years, a thylacine, and a contact-period figure of a European in shirt and boots with his hands in his pockets, painted around the 1880s. Burrungkuy (Nourlangie), further south in the park, is open year-round and is the other major art site. A park pass is required and is easiest bought online in advance. As advertised by Parks Australia in August 2026, it is valid seven days and priced by season: in the tropical summer (1 November to 14 May) $25 adult, $12.50 child aged 5-15, $19 concession, $65 family; in the dry (15 May to 31 October) $40, $20, $30 and $100. Northern Territory residents are exempt. Prices and dates change — check before you travel. Seasons matter more than hours here. In the tropical summer the water rises, opening times shorten, and the road to Ubirr can flood; when Magela Creek is up, the Guluyambi cruise is the way across. Parks Australia publishes a live Kakadu Access Report, and it is the thing to read the morning you set out rather than the week before. Arnhem Land proper, east of the East Alligator River, is Aboriginal freehold land and is not open to casual visitors: entry requires a permit from the Northern Land Council, applied for in advance. Within Kakadu, some places are closed to visitors altogether and are marked as such; the sensible rule is that the signage at a site is the traditional owners speaking, and it is the authority, not a suggestion.

Sources consulted

  1. Francis James Gillen, «Notes on some Manners and Customs of the Aborigines of the McDonnell Ranges belonging to the Arunta Tribe, in Report on the Work of the Horn Scientific Expedition to Central Australia, Part IV — Anthropology, ed. Baldwin Spencer, pp. 161-196. The "dream-times" passage is on p. 185. Read in the University of Illinois scan (80,048 words); "dream" occurs twice in the whole volume, of which this is one.» — Dulau and Co., London / Melville, Mullen and Slade, Melbourne, 1896 archive
  2. Baldwin Spencer and Francis James Gillen, «The Native Tribes of Central Australia. Read in the University of Toronto scan (239,931 words): "Alcheringa" 298 occurrences, "dream" 5 and none of them a gloss, "rainbow" 0. Footnote to p. 73 and the glossary entry for Alcheringa are the source of "the far distant past" and "the far past times".» — Macmillan, London, 1899 book
  3. Oceania (Australian National Research Council), «Index to Volume I, Oceania 1930-1931. Source of the single heading "Rainbow-Serpent — in North Queensland 347, 348; in North-West Australia 349-351; in South-East Australia 342-346" and of the cross-reference "Serpent myths. See Rainbow-Serpent".» — Oceania, volume I index, 1930 archive
  4. A. R. Radcliffe-Brown, «The Rainbow-Serpent Myth of Australia, Journal of the Royal Anthropological Institute of Great Britain and Ireland 56: 19-25. Cited for the existence, date and scope of the article; the full text was not read.» — Journal of the Royal Anthropological Institute, 1926 academic
  5. Bininj Kunwok Regional Language Centre, «Bininj Kunwok Dictionary, 2,614 entries. Headwords consulted: ngalyod, Yingarna, ngalmudj, delek, larraddi, djarnkerrk, kundjilinj, marrkidjbu, djang, kubolkdjamun, darrinj. Thirty entries carry the semantic domain "mythology".» — Bininj Kunwok Regional Language Centre, Jabiru, 2026 archive
  6. Paul S. C. Taçon, Meredith Wilson and Christopher Chippindale, «Birth of the Rainbow Serpent in Arnhem Land rock art and oral history, Archaeology in Oceania 31(3): 103-124. Paywalled; the figures used here (107 images analysed, earliest images c. 6,000 years, ribboned pipefish Haliichthys taeniophorus) come from published summaries, not from the article itself.» — Archaeology in Oceania, 1996 academic
  7. Catherine H. Berndt, «Rainbow Snake, entry in the Encyclopedia of Religion (revised bibliography). Source of the summary of Radcliffe-Brown's "perhaps the most important nature-deity" and of the note that the name is applied to snakes with no obvious rainbow connection.» — Encyclopedia of Religion, Macmillan, 1987 book
  8. Cliff Goddard and Anna Wierzbicka, «What does Jukurrpa ('Dreamtime', 'the Dreaming') mean? A semantic and conceptual journey of discovery, Australian Aboriginal Studies 2015/1. Note 2 traces the phrase to "Gillen 1896: 185" and calls Stanner's attribution to Spencer and Gillen "speaking somewhat loosely"; §1 quotes Stanner's 1953 opening.» — Australian Aboriginal Studies (AIATSIS), open access via the ANU Open Research Repository, 2015 academic
  9. Christine Judith Nicholls, «'Dreamtime' and 'The Dreaming': who dreamed up these terms?» — The Conversation, 2014 web
  10. Shino Konishi, «Friday essay: creation, destruction and appropriation — the powerful symbolism of the Rainbow Serpent» — The Conversation, 2021 web
  11. Parks Australia, «Ubirr — Kakadu National Park. Source of the park's own map point (-12.411359, 132.957727), the 2WD access, the Guluyambi crossing of Magela Creek, the X-ray and contact-period paintings, and the sentence that the Rainbow Serpent "painted her image on the rock as she passed through Ubirr".» — Kakadu National Park (kakadu.gov.au), 2026 report
  12. Parks Australia, «Park passes — Kakadu National Park. Prices and season dates as published in August 2026; passes valid seven days, Northern Territory residents exempt.» — Kakadu National Park (kakadu.gov.au), 2026 report
  13. Northern Land Council, «Permits: access to Aboriginal land in the Top End» — Northern Land Council (official), 2026 report

Reviewed by V. Serrano on 2026-08-17 · How we verify our content

📎 Source: Enigma Atlas

Your call, agent: I saw this too 0 Credible 0 Doubtful 0

💬 Field testimonies (0)

No testimonies yet. Add the first.

Add your testimony

🔒 Log in to add a testimony