Jacob Mikanowski at Cabinet Magazine presents a fascinating look at a mysterious script whose key was lost in frustratingly recent times:

Of all the literatures in the world, the smallest and most enigmatic belongs without question to the people of Easter Island. It is written in a script—rongorongo—that no one can decipher. Experts cannot even agree whether it is an alphabet, a syllabary, a mnemonic, or a rebus. Its entire corpus consists of two dozen texts. The longest, consisting of a few thousand signs, winds its way around a magnificent ceremonial staff. The shortest texts—if they can even be called that—consist of barely more than a single sign. One took the form of a tattoo on a man’s back. Another was carved onto a human skull.

Where did the rongorongo script come from? What do its texts communicate? No one knows for sure. The last Easter Islanders (or Rapanui) familiar with rongorongo died in the nineteenth century. They didn’t live long enough to pass on the secret of their writing system, but they did leave a few tantalizing clues. The island’s spoken language, also called Rapanui, lives on, but today it is written in a Latin script and its relationship to rongorongo is unclear. So far at least, no one has successfully connected one with the other. To this day, rongorongo remains a puzzle, an enigma, and a mirror for the folly of those who try to solve it. […]

The glyphs of rongorongo are unique. So is the manner in which it was written and read. In fact, it was not written, but carved. Its scribes used shark’s teeth to inscribe its symbols on wooden tablets. Wood is scarce on Easter Island, and most of these inscriptions were made on pieces of driftwood. One decorated an oar. A second, a beam. A third, a statue of a bird. However, most of the surviving examples of rongorongo decorate square tablets. These appear to have been written from bottom to top, and were read following a pattern called the reverse boustrophedon. Boustrophedon is a Greek word meaning “in the manner of an ox,” and scripts written in it move like an ox plowing a field, reversing direction with each line.

How did the Rapanui go from signing their names to a bogus document of annexation with simple drawings to creating a complex writing system incorporating hundreds of signs? We will most likely never know. First contact with Europeans was a trauma, one to which the Rapanui responded with incredible creativity and cultural ferment. But peering into the history of how this happened is unfortunately almost impossible, for it means looking through the scrim drawn by a holocaust. […]

The process of fantastical translation began early, with one Dr. Allen Carroll, a Sydney physician who may have been the illegitimate son of the Duke of Norfolk and who believed the tablets to be the work of pre-Inca South Americans. Carroll’s translations take the form of fulsome paeans to unnamed gods, and sound a bit like bowdlerized versions of the rubaiyat of Omar Khayyam: “To those who are our Guardians, oh give ear to us in your temple. You are our protectors. … Ye gods.”

Carroll was the father of the so-called “rongorongo fringe.” Many more followed in his footsteps. Most share a single, driving idea: the conviction that the Rapanui could not have invented their script on their own. The islanders’ writing was necessarily the mark of an older, greater civilization, they believed. All agreed it must have come from somewhere else. Where exactly it came from was the only thing left to debate. […]

The most sustained effort to decipher rongorongo rooted in professional linguistics took place in St. Petersburg. The dogged work of the Russian school helped to sort the script into a clear inventory of signs. Through the use of internal analysis and statistical comparison, its members hoped to place the study of the script on a firm scientific basis. But their careful analysis yields results that nonetheless sound insane—in Irina Fedorova’s translation, one text ends: “yam, yam, taro, taro, he cut a tuber of yam, he took a tuber of taro, a tuber, a tuber, he dug up, he cut, he cut, taro, turi sugar-cane.”

I have elided a bunch of crackpot attempted decipherments (compare the endless Voynich nonsense) and will send you to the link for them, as well as for the remarkable story of Katherine Routledge, who gathered ethnographic data in 1914 from some of the few Rapanui elders who survived the holocaust of the 1860s and eventually went mad, and much else. Thanks, Trevor!


  1. He’s being too kind to Fischer. True, Fischer has done some very thorough historical research on rongorongo, and when published, his book was the most detailed treatment available for things like rongorongo traditions and tablet provenance. However, his purported readings of the tablets are as obviously fanciful as those of his many predecessors quoted in the article.
    A number of people have done valuable work on the script, before and after Fischer: Barthel, Guy, Pozdniakov, Horley, Wieczorek, and others. So far, the only trustworthy results are still at the level of comparing parallel passages, defining the sign inventory, and the like. Even the most cautious claims of identifying the meaning of any individual glyph are too speculative to be convincing.

  2. Thanks, I was hoping somebody who knew more than I would comment!

  3. yam, yam, taro, taro, he cut a tuber of yam, he took a tuber of taro, a tuber, a tuber, he dug up, he cut, he cut, taro, turi sugar-cane

    Let me fix the formatting.

    yam, yam,
    taro, taro,
    he cut a tuber of yam,
    he took a tuber of taro,
    a tuber, a tuber,
    he dug up,
    he cut, he cut,
    taro, turi

    Even in English translation this chant has a rhyme and a rhythm, the meaning and purpose are pretty straightforward too. (the writer wants gods to grant better harvest this year and hopes to achieve this result by reciting some folk song about mythical heroes of the past. Classical Polynesian magical thinking)

  4. Michael Eochaidh says:

    Well, Fischer did also “decipher” the Phaistos Disk.

  5. Yeah, superficially at least, that “translation” evokes old Māori waiata to me too.


    Lots of repetition, allusion, and ellipsis in that style. So while clearly we have no insight into the soundness of methods and whether this is in fact a genuine translation, it’s not actually “insane”.

  6. The “repetitiveness” of the Māori chant is typical for work songs, marine and otherwise (“Yo, ho, up she rises” etc.) The “yam, yam, taro, taro” thing fits no known genre, at least not one that Fedorova demonstrated.

    Most or all of the bad RR interpretations go like this: squint, until you imagine the different glyphs remind you of various concrete objects. As needed, imagine that the glyphs that look like hands are engaged in some actions; those will be your verbs. Combine them all into a text. Handwave away the lacjk of grammatical morphemes of any kind; a good excuse is that wood was precious, so they wrote in telegraphic, grammar-free style to conserve space. String your words along, and fill in the translation with whatever grammar will stick the pieces together.

    In the end you’ll get a chant, a prayer, or even a genealogy. It won’t make any sense, because the Polynesians were mysterious like that.

    Quite a bit of the Rongorongo corpus consists of long strings of short phrases, similar but not identical. They could be lists of names, or genealogies, or something else. Nobody knows. It is those “lists” which provide the most fertile ground to imaginary translations.

  7. The “yam yam” sounds like that thing Google Translate does on one through thirty repeated copies of a single letter.

  8. The “yam yam” sounds pretty much exactly like a viral pop song to me, of the kind intended for dance clubs. I can all but hear the underlying techno beat. “Sugar cane!” is sang in a high-pitched female voice as the chorus.

    @Y and others: I’m most curious about the question of origin—is it credible that it may be truly autochthonous, without stimulus diffusion from abroad?

  9. Trond Engen says:

    The quoted passage seems to say that it was part of a cultural transformation in response to the arrival of Europeans.

  10. What language could it be? It’s either Polynesian or (following Heyerdahl) some South American language. That’s a fairly small set of possibilities.

    The concept of writing most likely came from Europeans because the Polynesians didn’t have writing and the South Americans only had knotted string. So more likely it must be from the early European contact period.

    If Polynesians suddenly invented writing, what would they write? We know they liked chanting, genealogies and mythological and historical stories. If the writing has a lot of repetition, it’s probably one of the first two possibilities. Everyone wants it to be some kind of history where we could find out the real story about the Longears and the Shortears and who cut down the last tree.

    I suspect that either “yam yam” or “son of so-and-so” are the most likely results if anyone succeeds in deciphering it. I kind of favour the genealogy idea because why would anyone write down chants that everybody knows? But genealogy, that’s a family thing that needs to be preserved in these hectic times.

  11. David Marjanović says:

    Handwave away the lacjk of grammatical morphemes of any kind; a good excuse is that wood was precious, so they wrote in telegraphic, grammar-free style to conserve space.

    Grammatical morphemes weren’t written in Old Chinese either. They have to be reconstructed from characters that have two pronunciations and two meanings nowadays.

    Indeed, as long as a script is pictographic enough, grammatical morphemes that don’t have an easily drawable meaning can’t be drawn in the first place… the trick of using pictures for homophones is a later step, and the trick of making signs up probably comes even later.

    why would anyone write down chants that everybody knows?

    Didn’t the Cherokee do that pretty much immediately? Or am I confusing them with the Cree? Also, if writing is for solemn occasions (as the passing Spaniards may have demonstrated), rather write down something really important if wood is scarce.

    who cut down the last tree

    Allegedly the Spaniards. The speed of this whole “end of the world as we know it” seems to have been overinterpreted.

  12. ə de vivre says:
  13. Me, I was thinking of badger badger mushroom.

  14. Yam taro tethera.

  15. Did any of the scholars who sought to decipher Rongorongo have any serious knowledge of comparative Polynesian linguistics? My understanding is that Modern Rapanui is heavily influenced by Tahitian and thus quite unlike what Rapanui must have been like when Rongorongo was actively used. Furthermore, if the script was used to represent a sacred register of some kind, then the gap between this “Rongorongo Rapanui” and the present-day language may be quite pronounced, So much so that it seems to me that any serious attempt at deciphering Rongorongo must take comparative Polynesian linguistics as a starting point, reconstructing early Rapanui in order to focus upon the script itself.

    Also, on the origin of the script: within the article it is taken for granted that this was due to the diffusion of the idea of writing as a result of contact with the first Spanish expedition. The WP article on the script, however, points to the fact that other instances of diffusion involved much closer and longer-lasting contact. I wonder: could the trigger for the birth of Rongorongo have been something other/more than the first Spanish expedition? A single castaway literate European sailor, stuck on the islands for years if not decades, could very well have been the real catalyst.

  16. ə de vivre says:

    If contact with Spanish writing was what prompted Rongorongo’s creation, shouldn’t it also be an alphabetic script (or at least a syllabary)? Are there any examples of a culture being in contact with alphabetic writing and coming up with a logographic system for themselves?

  17. David Eddyshaw says:

    Cherokee has a syllabary, the creation of which was sparked off by Sequoiah’s encounter with the Latin alphabet. Apparently he couldn’t read the alphabet as such, and created his syllabary de novo once he discovered that written symbols could represent language.

  18. Marja Erwin says:

    “Are there any examples of a culture being in contact with alphabetic writing and coming up with a logographic system for themselves?”

    It’s not uncommon for specialist writing in cultures with alphabetic scripts: hobo signs, military symbols, and so on.

  19. maidhc: Pace Heyerdahl (and others), there’s no evidence for any prehistoric South American presence on Rapanui, linguistic, cultural, or genetic. The language of rongorongo would have been an ancestor of modern Rapanui.

    Etienne: Kieviet’s recent comprehensive Rapanui grammar estimates that about 10% of the total vocabulary in his database is of Tahitian origin. However, he says, even though Tahitian influence on the language started in the 1880s, recorded texts don’t show much Tahitian vocabulary until the 1920s. So there’s a fair amount of record of pre-Tahitian Rapanui. In any case, Rapanui grammar and phonology remained distinct from those of Tahitian even post-contact.

    In the early 19th century many European and American boats stopped by the island for very brief trade visits. If there were any castaways, they wouldn’t be there for long. I haven’t read of any in the early period, and I would suppose that if there were, their story would be well-publicized.

    Of course there was an elevated register in Rapanui, as in other languages. A little of it remains in some traditional texts. But even in languages where the elevated register is well documented (e.g. Hawaiian and Tuamotuan), it is not that alien, though it relies a lot on esoteric metaphor.

    David M: Any thoughts as to why early Chinese omitted grammatical morphemes? Did it have anything to do with the logographic nature of the script? Rongorongo, whatever it is, is not logographic. A good guess is that it’s a mixed syllabic script (a few dozen frequent signs, with a scattering of others).

    There’s no decisive answer as to whether Rongorongo existed before European contact. There are reasonable but weak arguments both for and against. Only one tablet was carbon-dated, and it is of 19th century age.

  20. Trond Engen says:

    The Europeans could have shown the trick of reading a sacred or official text from a roll or something without explaining the mechanism. Or maybe the trick of marking property with signs signifying the owner’s name. I’d think a shipwrecked sailor would have meant a thorough explanation of the alphabetic system as well.

  21. why early Chinese omitted grammatical morphemes

    Sumerian, for the matter, works the same way.

  22. ə de vivre: In answer to your question (“Are there any examples of a culture being in contact with alphabetic writing and coming up with a logographic system for themselves?”), one example that comes to mind is the Yupik script: while the final product was a syllabary, it was originally a logographic script, whose creator (Uyaquq), tellingly, was never exposed to any writing system except the alphabet.

    So in principle Rongorongo (whether it was logographic, syllabic or some mix of the two) could likewise be due to cultural diffusion from a European source, possibly (a) castaway(s), as I wrote above. Which, in turn, opens the door to an intriguing possibility: could the “Rongorongo” register of Rapanui contain loanwords or proper nouns drawn from the language of the castaway(s)?

  23. David Marjanović says:

    Sequoyah himself, too, started out by devising parts of a logographic script, then quickly gave that up as too cumbersome and came up with an incomplete syllabary. All his prior exposure to writing was the Latin alphabet, and that’s what most of the glyph shapes are taken from.

    More later.

  24. There is no evidence in early Rapanui of any late loanwords. In contrast, work by Geraghty and Tent has uncovered quite a number of Dutch loanwords going back to the 17th century, which had spread throughout many parts of Polynesia in advance of actual European contacts.

    I can’t argue with certainty that Rongorongo did not have any grammatical markers. I’m just saying that without them it becomes much harder to distinguish a decipherment from noise.

  25. Y: I was referring to the “Rongorongo” register of Rapanui when I speculated that it might contain European loanwords, not to (ordinary spoken) Rapanui. And the reason I brought up the possibility was because (especially if grammatical markers were indeed not written) the presence of known vocabulary items might well ease the task of deciphering Rongorongo.

  26. marie-lucie says:

    Etienne: What type of vocabulary items are the Dutch loanwords? If the purpose of the rongorongo writings was ceremonial or historical, their vocabulary is unlikely to include recent borrowings from a completely different culture.

  27. m.-l., that was me who mentioned the Dutch loanwords… They are mostly trade items. See these three articles by Geraghty (an expert on Fijian and Polynesian) and Tent (an expert on early Dutch):

    Early Dutch loanwords in Polynesia
    More early Dutch loanwords in Polynesia
    Exploding sky or exploded myth? The origin of Papālagi

    This last one is actually a Malay word transported by the Dutch. It came up here at LH before. Astonishingly, Polynesian pāpalangi ‘White person’ has nothing to do with the Frank-derived words of Europe.

  28. Marie-Lucie: it was Y who referred to the early Dutch loanwords in Polynesian, not I, and from what I know of them they refer mostly to introduced objects (boxes, needles…) in the context of immediate trade.

    But any cultural diffusion leading to the birth of a writing system, by contrast, can be expected to be a little more intimate, and it would be unsurprising (although not inevitable, nota bene!) to find European loanwords in Rapanui (possibly in a specialized “Rongorongo” register) to refer to writing, writing instruments, letters and related topics. Indeed, as an example, consider Dutch itself: “schrijven”, its verb for “to write”, was borrowed from Latin “scribere”.

  29. Etienne, it would be nice to have an identifiable word in Rongorongo. Even the final decipherment of Linear B depended, as I recall, on identifying one place name in the written material. However, the statistical study of RR is far behind where that of Linear B was, even in Kober’s time. In particular, there is no clear indication of word division anywhere.

  30. The glyphs in Merahi metua no Tehamana are imaginary.

  31. marie-lucie says:

    Y and Etienne, I apologize, both of you mentioned the Dutch words and I guess I went by the last mention.

    My point was that the borrowings were likely to be for imported objects or acts and therefore unlikely to occur in texts that appeared to have been composed or at least used in some traditional cultural contexts. When a missionary tried to get people to translate some of the inscribed panels, two “readers” gave very different translations, probably because they (or at least one of them) did not now how to actually “read” the writing but (assuming they were not bluffing) could yet recite from memory what may have been actual traditional compositions.

    If the Rapanui people did meet Dutch sailors, those sailors (or at least their officers) must have been the ones seen writing and reading what was written, even if they did not try to write the local language, let alone teach the people to do so. The rongorongo characters look superficially closer to handwriting than to print. It might be interesting to compare them with Dutch handwriting styles of the period of contact.

  32. It was one reader, Metoro, who gave two different versions at dfferent times. Also, the Rapanui of his “reading” was full of recent Tahitian borrowings (that was after most of the Rapanui population had been living in Tahiti for a while). That is another reason to suspect that what he read had nothing to do with the text.

  33. Athel Cornish-Bowden says:

    Once in Valparaíso I came across an item about Easter Island with an explanation in Spanish and English. The English was approximate at best and a place called “Shavesnui” had me stumped until I looked at the Spanish, The translator had obviously looked up “rapa” in a Spanish-English dictionary and found “shaves”.

  34. Whenever Rongorongo is discussed, I wonder two things:

    What is the stylistic relationship between the Rongorongo glyphs and other Rapanui carvings that are known to be representational? For example, there are stone carvings that are known to depict the annual Bird Man ceremony.

    Is a short-lived form of writing like this really that unusual, especially among indigenous groups after their first limited contact with literate explorers? I am curious about this because Rapanui has received a grossly outsized amount of attention because of a feature of the island (the moai) that is utterly unrelated to the Rongorongo writing. Would comparable developments in writing have been found, say, on some other Polynesian islands, if they had been pored over by interested outsiders to the same extent?

  35. There are very clear parallels between Rapanui rock art and Rongorongo iconography, though there are no rongorongo texts in the rock art at all.

    Early European explorers were very keen on finding “lost civilizations” and early scripts. The decipherment of Egyptian writing and the discovery of cuneiform were very fresh in people’s minds. Missionaries, in particular, were quite aware of the power of writing. Early reports of rongorongo on Easter Island speak of a great deal of tablets in all the houses. No special attention was needed to notice them.

  36. Marie-Lucie: In answer to your observation (“The rongorongo characters look superficially closer to handwriting than to print”), there existed a semi-alphabetical script used in indigenous North America in the nineteenth century, Great Lakes syllabic, which grew out of Roman script handwriting: perhaps at least some Rongorongo symbols grew out of some handwritten Roman letter or letter combination…

    Which, indeed, come to think of it, needn’t even derive from Dutch, Spanish, English or any of the official languages of the European powers which were sending ships into the South Pacific. One thing which Geraghty and Tent’s work reminded me of was that Dutch ship crews at the time were quite multinational, with a large number of speakers of Dutch dialects, Frisian and Low German aboard: similar such linguistic diversity certainly existed aboard other (non-Dutch) European ships, and in some cases literate speakers of non-official languages were crew members: it is not widely known, for instance, that one of the earliest descriptions of the Beothuk in Newfoundland was written in…Breton (Really. And yes, I will give the reference should anyone request so downthread).

    And in fact in some instances Frisian speakers aboard Dutch ships seem to have had some linguistic influence: see for example-

    -which seems to indicate that the linguistic composition of ships’ crews needs to be paid attention to, and indeed, in addition to the Dutch loanwords in Polynesia (linguistic globalization avant la lettre…), perhaps a Frisianism or two likewise made its way to the South Pacific.

    In the context of the birth/creation of Rongorongo, however, here is a question: while there were many languages spoken by sailors in the South Pacific during the relevant historical period, which ones were substantial numbers of speakers (among sailors, nota bene!) literate in? I know that West Frisian was already marginalized if not extinct as a written language at the time, but among Low German speakers would literacy have been in Low or in High German? Or would the ones serving aboard Dutch ships have been literate in Dutch rather than in any non-Dutch West Germanic variety?

  37. David Marjanović says:

    There are very clear parallels between Rapanui rock art and Rongorongo iconography, though there are no rongorongo texts in the rock art at all.

    Some of the common rock glyphs show up as the “signatures” on the Spanish annexation document. I think that document gave people the idea that their glyphs could be used to represent more than individual objects or individual actions during a ceremony.

    among Low German speakers would literacy have been in Low or in High German?

    In the 17th century? Possibly Low, but most likely sailors would’ve been altogether illiterate.

  38. Or would the ones serving aboard Dutch ships have been literate in Dutch rather than in any non-Dutch West Germanic variety?
    Hard to say. In the 19th century schooling in the Lower German speaking areas was already in High German, although most people would probably still have spoken Plattdeutsch in daily life. In the 18th century some Dutch literacy existed in at least the Westernmost parts of Northern Germany; I’ve seen Dutch (not Plattdeutsch) inscriptions on 18th century buildings in Ostfriesland.

  39. Etienne, I don’t need that reference, but I would like the reference you offered in 2014 to Old Gascon as a separate Romance language that was later occitanized.

  40. John Cowan:

    Chambon, Jean Pierre & Yan Greub. 2002. “Note sur l’âge du (proto)gascon”. REVUE DE LINGUISTIQUE ROMANE. 263-264, pages 473-495.

    Hope you find it useful.

  41. I would like the Breton Beothuk reference. Is it well-known? Is it in the HNAI?

  42. If you google “Note sur l’âge du (proto)gascon” the first hit is a link to a pdf of the article.

  43. Y: No, it isn’t in the HNAI, and I doubt it could be called well-known:

    Bakker, Peter & Lynn Drapeau. 1995. “Adventures with the Beothuk in 1787: a testimony from Jean Conan’s autobiography”. In: Cowan, William (Ed.) PAPERS OF THE 25TH ALGONQUIAN CONFERENCE, pages 32-45. Ottawa, Carleton University.

  44. Etienne, thanks! It’s even online.

  45. It’s actually a fascinating read. It’s a diary in verse; unfortunately only the English translation is provided in this article. There’s quite a bit about Jean Conan working very hard at refusing a girl who takes a fancy to him.

  46. marie-lucie says:

    Fascinating indeed! An English translation of a Breton text.

  47. Thanks, Etienne, Hat. Not useful to me, but hopefully useful to my correspondent on Quora.


  1. […] Hat links to a new article on rongorongo, the mysterious and undeciphered script of the Rapa Nui of […]

Speak Your Mind