Rongorongo (; Rapa Nui: ) is a system of discovered in the 19th century on Easter Island that appears to be writing or proto-writing. Numerous attempts at decipherment have been made, none successfully. Although some calendrical and what might prove to be genealogical information has been identified, not even these glyphs can actually be read. If rongorongo does prove to be writing and proves to be an independent invention, it would be one of very few independent inventions of writing in human history.
Two dozen wooden objects bearing rongorongo inscriptions, some heavily weathered, burned, or otherwise damaged, were collected in the late 19th century and are now scattered in museums and private collections. None remain on Easter Island. The objects are mostly tablets shaped from irregular pieces of wood, sometimes driftwood, but include a chieftain's staff, a bird-man statuette, and two reimiro ornaments. There are also a few which may include short rongorongo inscriptions. Oral history suggests that only a small elite was ever literate and that the tablets were sacred.
Authentic rongorongo texts are written in alternating directions, a system called reverse boustrophedon. In a third of the tablets, the lines of text are inscribed in shallow fluting carved into the wood. The glyphs themselves are outlines of human, animal, plant, artifact and geometric forms. Many of the human and animal figures, such as glyphs and have characteristic protuberances on each side of the head, possibly representing ears or eyes.
Individual texts are conventionally known by a single uppercase letter and a name, such as Tablet C, the Mamari Tablet. The somewhat variable names may be descriptive or indicate where the object is kept, as in the Oar, the Snuffbox, the Small Santiago Tablet, and the Santiago Staff.
Etymology and variant names
is the modern name for the inscriptions. In the Rapanui language
it means "to recite, to declaim, to chant out".
The original name—or perhaps description—of the script is said to have been kohau motu mo rongorongo, "lines incised for chanting out", shortened to kohau rongorongo or "lines for chanting out".
[Englert 1993] There are also said to have been more specific names for the texts based on their topic. For example, the kohau ta‘u ("lines of years") were annals, the kohau îka ("lines of fishes") were lists of persons killed in war (îka "fish" was homophonous with or used figuratively for "war casualty"), and the kohau ranga "lines of fugitives" were lists of war refugees.
Some authors have understood the ta‘u in kohau ta‘u to refer to a separate form of writing distinct from rongorongo. Barthel recorded that, "The Islanders had another writing (the so-called 'ta‘u script') which recorded their annals and other secular matters, but this has disappeared."
[Barthel 1958:66] However, Fischer writes that "the ta‘u was originally a type of rongorongo inscription. In the 1880s, a group of elders invented a derivative 'script' also called ta‘u with which to decorate carvings in order to increase their trading value. It is a primitive imitation of rongorongo." [Fischer 1997:667] An alleged third script, the mama or va‘eva‘e described in some mid-twentieth-century publications, was "an early twentieth-century geometric decorative invention". [Fischer 1997:ix]
Form and construction
The forms of the glyphs are standardized contours of living organisms and geometric designs about one centimeter high. The wooden tablets are irregular in shape and, in many instances, fluted
, and possibly T
), with the glyphs carved in shallow channels running the length of the tablets, as can be seen in the image of tablet G
at right. It is thought that irregular and often blemished pieces of wood were used in their entirety rather than squared off due to the scarcity of wood on the island.
Except for a few possible glyphs cut in stone (see petroglyphs
), all surviving texts are inscribed in wood. According to tradition, the tablets were made of toromiro
wood. However, Orliac (2005) examined seven objects (tablets B
, and reimiro L
) with stereo optical
and determined that all were instead made from Pacific rosewood
; the same identification had been made for tablet M
in 1934. This 15-meter tree, known as "Pacific rosewood" for its color and called mako‘i
in Rapanui, is used for sacred groves and carvings throughout eastern Polynesia and was evidently brought to Easter Island by the first settlers.
[Skjølsvold 1994, as cited in Orliac 2005]
However, not all the wood was native: Orliac (2007) established that tablets N
, and S
were made of South African Yellowwood
and therefore that the wood had arrived with Western contact. Fischer describes P
as "a damaged and reshapen European or American oar", as are A
(which is European ash, Fraxinus excelsior
) and V
; notes that wood from the wreck of a Western boat was said to have been used for many tablets; and that both P
had been recycled as planking for a Rapanui driftwood canoe, suggesting that by that time the tablets had little value to the islanders as texts.
Several texts, including O
, are carved on gnarled driftwood
The fact that the islanders were reduced to inscribing driftwood, and were regardless extremely economical in their use of wood, may have had consequences for the structure of the script, such as the abundance of ligatures
and potentially a telegraphic style
of writing that would complicate textual analysis.
[Fischer 1997:382–383; see also decipherment of rongorongo]
Oral tradition holds that, because of the great value of wood, only expert scribes used it, while pupils wrote on banana leaves. German ethnologist Thomas Barthel believed that carving on wood was a secondary development in the evolution of the script based on an earlier stage of incising banana leaves or the sheaths of the banana trunk with a bone stylus, and that the medium of leaves was retained not only for lessons but to plan and compose the texts of the wooden tablets.
[Barthel 1971:1168] He found experimentally that the glyphs were quite visible on banana leaves due to the sap that emerged from the cuts and dried on the surface. However, when the leaves themselves dried they became brittle and would not have survived for long. [Fischer 1997:386]
Barthel speculated that the banana leaf might even have served as a prototype for the tablets, with the fluted surface of the tablets an emulation of the veined structure of a leaf:
Direction of writing
Rongorongo glyphs were written in reverse boustrophedon
, left to right and bottom to top. That is, the reader begins at the bottom left-hand corner of a tablet, reads a line from left to right, then rotates the tablet 180 degrees to continue on the next line. When reading one line, the lines above and below it would appear upside down, as can be seen in the image at left.
However, the writing continues onto the second side of a tablet at the point where it finishes off the first, so if the first side has an odd number of lines, as is the case with tablets K, N, P, and Q, the second will start at the upper left-hand corner, and the direction of writing shifts to top to bottom.
Larger tablets and staves may have been read without turning, if the reader were able to read upside-down.
According to oral tradition, scribes used obsidian
flakes or small shark teeth
, presumably the hafted
tools still used to carve wood in Polynesia, to flute and polish the tablets and then to incise the glyphs.
(See shark tooth tools
.) The glyphs are most commonly composed of deep smooth cuts, though superficial hair-line cuts are also found. In the closeup image at right, a glyph is composed of two parts connected by a hair-line cut; this is a typical convention for this shape. Several researchers, including Barthel, believe that these superficial cuts were made by obsidian, and that the texts were carved in a two-stage process, first sketched with obsidian and then deepened and finished with a worn shark tooth.
The remaining hair-line cuts were then either errors, design conventions (as at right), or decorative embellishments. Vertical strings of chevrons or lozenges, for example, are typically connected with hair-line cuts, as can be seen repeatedly in the closeup of one end of tablet B
below. However, Barthel was told that the last literate Rapanui king, Nga‘ara
, sketched out the glyphs in soot applied with a fish bone and then engraved them with a shark tooth.
Tablet N, on the other hand, shows no sign of shark teeth. Haberlandt noticed that the glyphs of this text appear to have been incised with a sharpened bone, as evidenced by the shallowness and width of the grooves.
[Haberlandt 1886:102] N also "displays secondary working with obsidian flakes to elaborate details within the finished contour lines. No other rongo-rongo inscription reveals such graphic extravagance". [Fischer 1997:501]
Other tablets appear to have been cut with a steel blade, often rather crudely. Although steel knives were available after the arrival of the Spanish, this does cast suspicion on the authenticity of these tablets.
The glyphs are stylized human, animal, vegetable and geometric shapes, and often form compounds. Nearly all those with heads are oriented head up and are either seen face on or in profile to the right, in the direction of writing. It is not known what significance turning a glyph head-down or to the left may have had. Heads often have characteristic projections on the sides which may be eyes (as on the sea turtle glyph below, and more clearly on sea-turtle petroglyphs) but which often resemble ears (as on the anthropomorphic petroglyph in the next section). Birds are common; many resemble the frigatebird (see image directly below) which was associated with the supreme god Makemake.
[Guy 2006] Other glyphs look like fish or arthropods. A few, but only a few, are similar to petroglyphs found throughout the island.
- Some of the more iconic rongorongo glyphs. The seated man bottom is thought to be a compound.
- (Readings from Barthel (1958). The captions in the right-most column are merely descriptive.)
Oral tradition holds that either Hotu Matu‘a
or Tu‘u ko Iho
, the legendary founder(s) of Rapa Nui, brought 67 tablets from their homeland.
The same founder is also credited with bringing indigenous plants such as the toromiro
. However, there is no homeland likely to have had a tradition of writing in Polynesia or even in South America. Thus rongorongo appears to have been an internal development. Given that few if any of the Rapanui people remaining on the island in the 1870s could read the glyphs, it is likely that only a small minority were ever literate. Indeed, early visitors were told that literacy was a privilege of the ruling families and priests who were all kidnapped in the Peruvian slaving raids or died soon afterwards in the resulting epidemics.
[Cooke 1899:712, Englert 1970:149–153]
Dating the tablets
Little direct dating has been done. The start of forest-clearing for agriculture, and thus presumably colonization, has been dated to circa
[Date ranges are 1200–1250 and 1180–1290. Mann et al. 2008]
implying a date for the invention of rongorongo no earlier than the 13th century. Tablet Q
(Small Saint Petersburg) is the sole item that has been carbon dated
, but the results only constrain the date to sometime after 1680. Glyph 67
() is thought to represent the extinct Easter Island palm
, which disappeared from the island's pollen record circa
1650, suggesting that the script itself is at least that old.
Texts A, P, and V can be dated to the 18th or 19th century by virtue of being inscribed on European oars. Orliac (2005) argued that the wood for tablet C (Mamari) was cut from the trunk of a tree some tall, and Easter Island has long been deforested of trees that size. Analysis of charcoal indicates that the forest disappeared in the first half of the 17th century. Roggeveen, who discovered Easter Island in 1722, described the island as "destitute of large trees" and in 1770 González de Ahedo wrote, "Not a single tree is to be found capable of furnishing a plank so much as six inches 15 cm in width." Forster, with Cook's expedition of 1774, reported that "there was not a tree upon the island which exceeded the height of 10 feet 3 m."
[Flenley & Bahn 1992:172]
All of these methods date the wood, not the inscriptions themselves. Pacific rosewood is not durable, and is unlikely to survive long in Easter Island's climate.
1770 Spanish expedition
In 1770 the Spanish annexed Easter Island under Captain González de Ahedo
. A signing ceremony was held in which a treaty of annexation was signed by an undisclosed number of chiefs "by marking upon it certain characters in their own form of script."
(Reproduction at right)
Several scholars have suggested that rongorongo may have been an invention inspired by this visit and the signing of the treaty of annexation.
[For example, Flenley & Bahn 1992:203–204] As circumstantial evidence, they note that no explorer reported the script prior to Eugène Eyraud in 1864, and are of the opinion that the marks with which the chiefs signed the Spanish treaty do not resemble rongorongo. The hypothesis of these researchers is not that rongorongo was itself a copy of the Latin alphabet, or of any other form of writing, but that the concept of writing had been conveyed in a process anthropologists term trans-cultural diffusion, which then inspired the islanders to invent their own system of writing. If this is the case, then rongorongo emerged, flourished, fell into oblivion, and was all but forgotten within a span of less than a hundred years.
However, known cases of the diffusion of writing, such as Sequoyah's invention of the Cherokee syllabary after seeing the power of English-language newspapers, or Uyaquk's invention of the Yugtun script inspired by readings from Christian scripture, involved greater contact than the signing of a single treaty. The glyphs could be crudely written rongorongo, as might be expected for Rapa Nui representatives writing with the novel instrument of pen on paper. The fact that the script was not otherwise observed by early explorers, who spent little time on the island, may reflect that it was taboo; such taboos may have lost power along with the tangata rongorongo (scribes) by the time Rapanui society collapsed following European slaving raids and the resulting epidemics, so that the tablets had become more widely distributed by Eyraud's day.
[Bahn 1996] Orliac points out that Tablet C would appear to predate the Spanish visit by at least a century.
Easter Island has the richest assortment of
Nearly every suitable surface has been carved, including the stone walls of some houses and a few of the famous mo‘ai
statues and their fallen topknots
. Around one thousand sites with over four thousand glyphs have been catalogued, some in bas-
, and some painted red and white. Designs include a concentration of chimeric
bird-man figures at Orongo
, a ceremonial center of the tangata manu
("bird-man") cult; faces of the creation deity Makemake
; marine animals like turtles, tuna, swordfish, sharks, whales, dolphins, crabs, and octopus (some with human faces); roosters; canoes, and over five hundred komari
(vulvas). Petroglyphs are often accompanied by carved divots ("cupules") in the rock. Changing traditions are preserved in bas-relief birdmen, which were carved over simpler outline forms and in turn carved over with komari.
Although the petroglyphs cannot be directly dated, some are partially obscured by pre-colonial stone buildings, suggesting they are relatively old.
Several of the anthropomorphic and animal-form petroglyphs have parallels in rongorongo, for instance a double-headed frigatebird
) on a fallen mo‘ai
topknot, a figure which also appears on a dozen tablets. McLaughlin (2004) illustrates the most prominent correspondences with the petroglyph corpus of Lee (1992).
[ However, these are mostly isolated glyphs; few text-like sequences or ligatures have been found among the petroglyphs. This has led to the suggestion that rongorongo must be a recent creation, perhaps inspired by petroglyph designs or retaining individual petroglyphs as (Macri 1995), but not old enough to have been incorporated into the petroglyphic tradition. The most complex candidate for petroglyphic rongorongo is what appears to be a short sequence of glyphs, one of which is a ligature, carved on the wall of a cave. However, the sequence does not appear to have been carved in a single hand (see image at right), and the cave is located near the house that produced the Poike tablet, a crude imitation of rongorongo, so the Ana o Keke petroglyphs may not be authentic.
Eugène Eyraud, a lay friar of the Congrégation de Picpus, landed on Easter Island on January 2, 1864, on the 24th day of his departure from Valparaíso. He was to remain on Easter Island for nine months, evangelizing its inhabitants. He wrote an account of his stay in which he reports his discovery of the tablets that year:
There is no other mention of the tablets in his report, and the discovery went unnoticed. Eyraud left Easter Island on October 11, in extremely poor health. Made a fully fledged priest in 1865, he returned to Easter Island in 1866 where he died of tuberculosis in August 1868, aged 48.
In 1868 the Bishop of Tahiti, Florentin-Étienne "Tepano" Jaussen, received a gift from the recent Catholic converts of Easter Island. It was a long cord of human hair, a fishing line perhaps, wound around a small wooden board covered in hieroglyphic writing. Stunned at the discovery, he wrote to Father Hippolyte Roussel on Easter Island to collect all the tablets and to find natives capable of translating them. But Roussel could only recover a few, and the islanders could not agree on how to read them.
Yet Eyraud had seen hundreds of tablets only four years earlier. What happened to the missing tablets is a matter of conjecture. Eyraud had noted how little interest their owners had in them. Stéphen Chauvet reports that,
Orliac has observed that the deep black indention, about 10 cm long, on lines 5 and 6 of the recto of tablet H is a groove made by the rubbing of a fire stick, showing that tablet H had been used for fire-making.
[Orliac 2003/2004:48–53] Tablets S and P had been cut into lashed planking for a canoe, which fits the story of a man named Niari who made a canoe out of abandoned tablets. [Routledge 1919:207]
As European-introduced diseases and raids by Peruvian slavers, including a final devastating raid in 1862 and a subsequent smallpox epidemic, had reduced the Rapa Nui population to under two hundred by the 1870s, it is possible that literacy had been wiped out by the time Eyraud discovered the tablets in 1866.
Thus in 1868 Jaussen could recover only a few tablets, with three more acquired by Captain Gana of the Chilean corvette O'Higgins in 1870. In the 1950s Barthel found the decayed remains of half a dozen tablets in caves, in the context of burials. However, no glyphs could be salvaged.
Of the 26 commonly accepted texts that survive, only half are in good condition and authentic beyond doubt.
British archaeologist and anthropologist Katherine Routledge undertook a 1914–1915 scientific expedition to Rapa Nui with her husband to catalog the art, customs, and writing of the island. She was able to interview two elderly informants, Kapiera and a leper named Tomenika, who allegedly had some knowledge of rongorongo. The sessions were not very fruitful, as the two often contradicted each other. From them Routledge concluded that rongorongo was an idiosyncratic mnemonic device that did not directly represent language, in other words, proto-writing, and that the meanings of the glyphs were reformulated by each scribe, so that the kohau rongorongo could not be read by someone not trained in that specific text. The texts themselves she believed to be litanies for priest-scribes, kept apart in special houses and strictly tapu, that recorded the island's history and mythology.
[Routledge 1919:253–254] By the time of later ethnographic accounts, such as Métraux (1940), much of what Routledge recorded in her notes had been forgotten, and the oral history showed a strong external influence from popular published accounts.
The 26 rongorongo texts with letter codes are inscribed on wooden objects, each with between 2 and 2320 simple glyphs and components of compound glyphs, for over 15,000 in all. The objects are mostly oblong wooden tablets, with the exceptions of I, a possibly sacred chieftain's staff known as the Santiago Staff; J and L, inscribed on reimiro pectoral ornaments worn by the elite; X, inscribed on various parts of a tangata manu ("birdman") statuette; and Y, a European snuff box assembled from sections cut from a rongorongo tablet. The tablets, like the pectorals, statuettes, and staves, were works of art and valued possessions, and were apparently given individual proper names in the same manner as jade ornaments in New Zealand.
[Buck 1938:245] Two of the tablets, C and S, have a documented pre-missionary provenance, though others may be as old or older. There are in addition a few isolated glyphs or short sequences which might prove to be rongorongo. [Fischer 1997]
Barthel referred to each of 24 texts he accepted as genuine with a letter of the alphabet; two texts have been added to the corpus since then. The two faces of the tablets are distinguished by suffixing r (recto) or v (verso) when the reading sequence can be ascertained, to which the line being discussed is appended. Thus Pr2 is item P (the Great Saint Petersburg Tablet), recto, second line. When the reading sequence cannot be ascertained, a and b are used for the faces. Thus Ab1 is item A (Tahua), side b, first line. The six sides of the Snuff Box are lettered as sides a to f. Nearly all publications follow the Barthel convention, though a popular book by Fischer uses an idiosyncratic numbering system.
Crude glyphs have been found on a few stone objects and some additional wooden items, but most of these are thought to be fakes created for the early tourism market. Several of the 26 wooden texts are suspect due to uncertain provenance ( X, Y, and Z), poor quality craftsmanship ( F, K, V, W, Y, and Z), or to having been carved with a steel blade ( K, V, and Y),
and thus, although they may prove to be genuine, should not be trusted in initial attempts at decipherment. Z resembles many early forgeries in not being boustrophedon, but it may be a palimpsest on an authentic but now illegible text. [Fischer 1997:534]
In addition to the petroglyphs mentioned above, there are a few other very short uncatalogued texts that may be rongorongo. Fischer reports that "many statuettes reveal rongorongo or rongorongo-like glyphs on their crown." He gives the example of a compound glyph,
, on the crown of a mo‘ai pakapaka statuette.
[Fischer 1997:543] Many human skulls are inscribed with the single 'fish' glyph 700 , which may stand for îka "war casualty". There are other designs, including some tattoos recorded by early visitors, which are possibly single rongorongo glyphs, but since they are isolated and pictographic, it is difficult to know whether or not they are actually writing.
The only published reference to the glyphs which is even close to comprehensive remains Barthel (1958). Barthel assigned a three-digit numeric code to each glyph or to each group of similar-looking glyphs that he believed to be allographs (variants). In the case of allography, the bare numeric code was assigned to what Barthel believed to be the basic form (Grundtypus), while variants were specified by alphabetic suffixes. Altogether he assigned 600 numeric codes. The hundreds place is a digit from 0 to 7, and categorizes the head, or overall form if there is no head: 0 and 1 for geometric shapes and inanimate objects; 2 for figures with "ears"; 3 and 4 for figures with open mouths (they are differentiated by their legs/tails); 5 for figures with miscellaneous heads; 6 for figures with beaks; and 7 for fish, arthropods, etc. The digits in tens and units places were allocated similarly, so that, for example, glyphs 206, 306, 406, 506, and 606 all have a downward-pointing wing or arm on the left, and a raised four-fingered hand on the right:
- Coding: The first digit distinguishes head and basic body shape, and the six in the units place indicates a specific raised hand.
There is some arbitrariness to which glyphs are grouped together, and there are inconsistencies in the assignments of numerical codes and the use of affixes which make the system rather complex. However, despite its shortcomings, Barthel's is the only effective system ever proposed to categorize rongorongo glyphs.
Barthel (1971) claimed to have parsed the corpus of glyphs to 120, of which the other 480 in his inventory are allographs or ligatures. The evidence was never published, but similar counts have been obtained by other scholars, such as Pozdniakov & Pozdniakov (2007).
For almost a century only a few of the texts were published. In 1875 the director of the Chilean National Museum of Natural History in Santiago, Rudolf Philippi, published the Santiago Staff, and Carroll (1892) published part of the Oar. Most texts remained beyond the reach of would-be decipherers until 1958, when Thomas Barthel published line drawings of almost all the known corpus in his Grundlagen zur Entzifferung der Osterinselschrift ("Bases for the Decipherment of the Easter Island script") which remains the fundamental reference to rongorongo. He transcribed texts A through X, over 99% of the corpus; the CEIPP estimates that it is 97% accurate. Barthel's line drawings were not produced free-hand but copied from , which helped ensure their faithfulness to the originals.
Fischer (1997) published new line drawings. These include lines scored with obsidian but not finished with a shark tooth, which had not been recorded by Barthel because the rubbings he used often did not show them, for example on tablet N. (However, in line Gv4 shown in the section on writing instruments above, the light lines were recorded by both Fischer and Barthel.) There are other omissions in Barthel which Fischer corrects, such as a sequence of glyphs at the transition from line Ca6 to Ca7 which is missing from Barthel, presumably because the carving went over the side of the tablet and was missed by Barthel's rubbing. (This missing sequence is right in the middle of Barthel's calendar.) However, other discrepancies between the two records are straightforward contradictions. For instance, the initial glyph of I12 (line 12 of the Santiago Staff) in Fischer
[Fischer 1997:451] does not correspond with that of Barthel [Barthel 1958: Appendix] or Philippi, [Philippi 1875] which agree with each other, and Barthel's rubbing (below) is incompatible with Fischer's drawing. Barthel's annotation, Original doch 53.76! ("original indeed 53.76!"), suggests that he specifically verified Philippi's reading:
In addition, the next glyph (glyph 20, a "spindle with three knobs") is missing its right-side "sprout" (glyph 10) in Philippi's drawing. This may be the result of an error in the inking, since there is a blank space in its place. The corpus is thus tainted with quite some uncertainty. It has never been properly checked for want of high-quality photographs.
As with most undeciphered scripts, there are many fanciful interpretations and claimed translations of rongorongo. However, apart from a portion of one tablet which has been shown to have to do with a lunar calendar, none of the texts are understood. There are three serious obstacles to decipherment, assuming rongorongo is truly writing: the small number of remaining texts, the lack of context such as illustrations in which to interpret them, and the poor attestation of the Old Rapanui language, since modern Rapanui is heavily mixed with Tahitian and is therefore unlikely to closely reflect the language of the tablets.
The prevailing opinion is that rongorongo is not true writing but proto-writing, or even a more limited mnemonic device for genealogy, choreography, navigation, astronomy, or agriculture. For example, the Atlas of Languages states, "It was probably used as a memory aid or for decorative purposes, not for recording the Rapanui language of the islanders."
[Comrie et al. 1996:100] If this is the case, then there is little hope of ever deciphering it. For those who believe it to be writing, there is debate as to whether rongorongo is essentially logographic or syllabic, though it appears to be compatible with neither a pure logography nor a pure syllabary. [Pozdniakov & Pozdniakov, 2007]
|This basic inventory of rongorongo, proposed by Pozdniakov & Pozdniakov (2007), accounts for 99.7% of the intact texts, except for the idiosyncratic Staff.|
The SMP range 1CA80-1CDBFF is tentatively allocated for encoding the Rongoringo script. However, a detailed proposal has not yet been written.
[ The Unicode Consortium. Roadmap to the SMP]