(1) Du Feu 1996:186.

(2) This velar nasal sound, /ŋ/ in the International Phonetic Alphabet, is also written as 'g' or 'ng': I have followed the latter, since there is no uniformity (Roussel/Churchill use 'g', Métraux and Barthel use 'ng', Fuentes, Englert and Du Feu use the IPA-symbol).

(3) This sound is alternatively written as /b/ and /v/. Since Churchill, Englert and Du Feu use /v/, /b/ will only occur in quotations from Fuentes.

(4) The glottal stop is a marked pause between two vowels. It has been and still is the cause of considerable confusion. It was not recorded by the early researchers such as Roussel, Geiseler, Thomson and Routledge. Métraux (1940:32) even went so far as to deny its existence while Englert and Fuentes do not register its presence in word-initial position. In addition, the consonant is written as / ’ / by Englert, as /ʔ/ by Fuentes and as / ' / by Du Feu. To avoid this confusion and because the rongorongo script itself gives no clues for the correct historical phonetics, the glottal stop will be largely ignored in the transliterations presented here – at least for the time being.

(5) Cf. Pozdniakov 1996:297; Macri,1996:185.

(6) Guy (1990:3) has proposed the sign's phonetic use as ta for a defective writing of ta[ne] in the Mamari Calendar.

(7) E.g., Guy 1982:445; Pozdniakov 1996:297.

(8) E.g., Horley 2005:110.

(9) Pozdniakov 2007:13.

(10) Figs. 43-44 were taken from parallel passages in lines Cb1 and Sa1.

(11) This phenomenon will be discussed in detail on a separate page.

(12) Given the fact that the London tablet shows an inexperienced hand at work, it is possible that this mixed method of reduplication was not considered correct. Another complication may have been that the words belong to different sentences and have different meanings.

Signs and syllables


The syllables of Rapanui, the East Polynesian language of Easter Island, are exclusively of the (C)V-type (1). The language has five vowels, /a/, /e/, /i/, /o/, /u/, which can be either long or short, and ten consonants, /ng/ (2), /h/, /k/, /m/, /n/, /p/, /r/, /t/, /v/ (3) and /'/ (the glottal stop) (4). The number of possible (C)V-combinations is therefore 55, i.e., 5 vowels and 50 consonant-vowel-combinations. The fact that more than 99% of the surviving rongorongo texts have been written with a set of some 50 to 60 basic graphemes, strongly suggests that the script is predominantly syllabic, i.e., that most of these signs represent the phonetic values of these (C)V-combinations (figs. 1-3) (5).

As in every writing system, not all of the phonetic characteristics of the language are represented by the script. For Rapanui, the most important of these are stress, vowel length and the consonants /h/ and glottal stop. This means, for example, that the same glyph is used to depict values o, ō, ho and 'o, and, consequently, that there is no distinction between the possessive particle (inalienable) o in the pronoun otaatou, "our" (fig. 4), the first syllable of the adverb hoki, "back" (fig. 5) and the glottalized o in 'oka, "to pierce", "to plant" (fig. 6).

The general CV-pattern is not followed by a small group of signs, which are mostly of the CVV-type with different vowels, such as mai (fig. 7), nei (fig. 8), nui, tou and vae. Special cases are the sign that is used for both na and ana/'ana (fig. 9) and the (C)VCV-structure of the sign that represents the component ina (fig. 10), which is also used as the negative particle 'ina and as hina, "grey-haired".

A number of phonetic values can be written with more than one sign. The most important of these are (h)a, which appears both as "stick" and as "head (with ear(s))" (fig. 11), (h)e, which is either crescent- or lens-shaped (fig. 12), nga which has a "head"- and a "body"-variant (fig. 13) and the "bird"-sign ta (6) which has a long and a short-beaked variant (figs. 1;14), a feature it has in common with the disyllabic "bird-head"-sign tou (fig. 15). Of the short-beaked allograph exists a long- and short-necked version.

An important feature of "shark"-sign ma and "bird"-sign ta is that their "heads" and "bodies" can serve as pars pro toto (e.g., fig. 4). Possibly, the ma-"body" is actually the opened shark mouth turned sideways (fig. 16). In addition to this, certain signs appear diminished in size, reversed, turned upside down and (cross)hatched.

The majority of the basic (C)V-glyphs and all of the disyllabic glyphs, also function as morphemes. The "frigate bird"-glyph ta (fig. 1), for example, is used to write "color", "to color", and the "fish"-sign (fig. 3) represents both the momentary particle ka and the verb ka, meaning "to become very angry" or "to go mad". The few exceptions to this, to the extent that examples of such a usage have not been found, are the syllabic signs for /ku/, /ri/, /ti/, /ve/ and /vi/.

Added: 2010-11-10  Modified: 2017-04-06


Singular and fused signs

Singular glyphs account for roughly 30 percent of the signs in the inscriptions of the corpus as collected by Barthel (1958). The rest of the signs is "fused", i.e., composed of two or more basic glyphs. It is, however, not possible to make a one-on-one comparison between these independent units in the writing and the words in the language. The most important reason for this is that fused signs not only represent words but also word groups consisting of a nominal or verbal lexeme, preceded and/or followed by one or more grammatical markers, occasionally also including adverbs and adjectives. Applying phonetic values of syllabic and disyllabic signs to composite glyphs may therefore not only produce single words such as tama, "child" (fig. 17), or tau, "beautiful" (fig. 18), but also two or more words that are in some way connected together such as tama tau, "beautiful children" (fig. 19).

A number of signs and composites such as the ta-"bird" and the (h)anga-sign, are very suited to serve as "glyph hangers" in fused constructions. A special case is the haka-glyph consisting of a ha-"head" and a body with outstretched arms and legs (fig. 20). It functions exclusively as causative prefix to verbs, e.g., haka-teki-teki, "to be lame(d)" (fig. 21), except for one noun, hakari, "body", e.g., hakari nako, "fat bodies" (fig. 22).

Other glyphs appear to have been designed specifically to "hang on" to other signs. It can hardly be a coincidence that among these are the signs that also represent some important markers of the nominal and verbal frame such as (a)na, no, nei, ra, ro and nei, e.g., angahe ana huru, "when were (they) secluded?" (fig. 23), tau nei, "this beauty" (fig. 24), tehe ra, "will (it) flow?" (fig. 25).  

Although the rules for the fusion process appear to be set by word formation and grammatical structure, this does not mean that all multi-syllabic words appear as composite signs or that markers are always fused with the lexeme. For example, the word toa, "sugarcane", is always written with two separate signs (fig. 26) and the postnominal/verbal particle era (fig. 27) is an independent sign in most instances. Some examples, such as the parallel versions of *hehe, "to dazzle" (figs. 28-29), and rotu, in huru ana rotu tama (na), "Should the children be secluded if (they) protest?" (figs. 30-31), indicate that there was some latitude – at least in certain situations.

These examples show that there are two methods of fusing elementary glyphs which are the most common. The first is connecting two adjacent signs through the use of an "arm", "leg", "wing" or other part, the other is by "stacking" two components on top of each other. It has been suggested that the reading order of stacked glyphs is always from bottom to top as this is suggested by some parallel text fragments (7), However, as there are also examples that indicate the contrary (8), nothing has been proven for the inscriptions in general. If there had been a fixed reading order, it would have severely limited the scribe's combinatory options. Instead, his focus seems to have been on the composition of easily recognizable fused signs and sign groups around a basic set of anthropo- and zoomorphic glyphs (9). To this end, maximum use was made of of the pictorial properties of the syllabary as can be seen in words such as tama, "child" (figs. 17;30-31), which was shaped as a young bird and (h)uru, "to enter", "to seclude" (figs. 30-31;35), which may have used the image of a palm tree to refer to *uru, "grove", a word that occurs in many Polynesian languages,  

As was already mentioned in regard to the "frigate bird"- and "shark"-signs, in the fusing process signs may give up part of their identity, The ra-sign, for example, often loses its "stalk" when it is attached to certain other glyphs (fig. 27). A similar thing happens to the "arm"-sign (h)i (fig. 32) which is sometimes abbreviated to a hand or three "fingers", as in the composite iti, "little" (figs. 33-34). Other glyphs may appear reduced in size, as can be seen in the (h)e-sign in fig. 23 and the ana-sign in fig. 31. Occasionally, components are incorporated in other signs, such as the ura-glyph inside the po-sign (fig. 35), perhaps playfully illustrating the meaning as the phrase reads huru mo po ura, "(They) are secluded to obscure the flames (of the sun)".

Another determining factor in the fusion of signs – and for the choice of allographs in general – appears to be the need to distinguish between near homonyms. For this reason, to'a, "rival", "enemy", "hostile" (fig. 36), is written with a variant of the sign used in toa, "sugarcane" (fig. 26), while ui, "to ask", and ui, "to look", are written with the "hand" of the (h)i-glyph turned to the left and to the right, respectively (figs. 38-39). Even more marked is the difference between tau, "period", "time" (fig. 39), and ta'u, "beautiful" (fig. 18), and between 'ata, "shadow" (fig. 40), and 'ata, "more" (fig. 41).  

In addition to these semantically motivated distinctions, signs may be minimized, tilted, turned upside down or mirrored for aesthetic or practical reasons. Paired anthropo- or zoomorphic signs may for example turn their heads toward each other, as do the birds in tata, "agony" (fig. 42). The two ha-signs reading haha, "to feel", "to walk blindly" (fig. 43), on the other hand, were likely fused into one glyph (fig. 44) to save space (10). The main purpose for writing signs in reverse or upside down, however, appears to be negation of the segment or phrase to which that particular component belongs (11).

The many reduplicated words in the Rapanui language were written in full as shown by the examples of eo-eo, "to smell" (fig. 45); poto-poto, "short" (fig. 46), but they could also be abbreviated. There appear to have been two strategies at the scribe's disposal to save precious writing space.    

In the first method one of the syllabic components can be left unwritten, which means that the phonetic value of the remaining component has to be reduplicated by the reader. For instance, tari , "to carry" (fig. 47), is reduplicated to tari-ta(ri) by writing only the ri-syllable twice (fig. 48). A prominent example is the very frequent occurring glyph for "man", ta-nga(-ta) (fig. 49). This type of reduplication can be most clearly observed in two parallel phrases on tablets Keiti and Small Vienna. The first text has the singular ahi, "fiery" (fig. 50), whereas the second has the reduplicated form, ahi-(a)-hi (fig. 51). The fragment reads "The fiery flames (of the sun) will gaze down (on the neru)!"

Occasionally, the phonetic reduplication of a syllable extends to an adjacent word. This can be observed in the two parallel fragments of figs. 52-53 – both reading pakia ahu, "fattened seals", which appears to be a not very flattering description of the neru. In the compound glyph of the second fragment one of the two (h)a-components of the first text has been dropped. This would only be possible if its phonetic value is supplied by the remaining sign.

In the second method of reduplication – which is used less frequently and only for a small group of glyphs – the double phonetic use of a sign is indicated by horizontal or diagonal hatching. A sign that is very frequently reduplicated in this way, is the "stick"-variant of the (h)a-glyph, representing the interrogative pronoun aha, "which", "what". On the Santiago Staff it appears frequently in combination with benefactive mo, forming mo aha, "for what", "why" (fig. 54). An interesting example of its usage in combination with the first method of reduplication can be observed in the parallel texts of the Small Santiago and London tablets (figs. 55-56).







  o - taa - tou  































 ta-ma  ta-u







haka-ri na-ko




ta-u nei

anga-he ana hu-ru

tehe ra







 to - a


hu-ru ana  ro  -  tu   ta-ma na


hu-ru ana    ro  -  tu       ta-ma







hu-ru mo  po















     ta - u

   a - ta





     ta - ta




    po - to  -  po - to

  e - o  -  e - o





     ta - ri




  ti-ro   a  u-ra a-hi-


ti-ro  a  u-ra  a-hi



 pa-ki-a (a)-hu

         pa-ki-a   a-hu



           a-ra    ro-a      a-ra


       a-ra     a-ra      ha-ro     a-ra   

 mo aha


All the above illustrations are from Barthel (1958), except for figs. 22; 43; 50; 51 (Horley 2010:53); 52; 53 (Horley 2007:28); 55; 56 (Fischer 1997: 440;486).

The scribe of the latter has used cross hatches to reduplicate the a-glyph thereby also implicating the reduplication of the ra-part. Actually, the first ara, "overthere", is the ending of the first sentence. In fig. 55, the next sentence reads ara haro ara, "(their) road is outstretched", and in fig. 56, ara roa ara, "(their) road is long" (12).