Volume 49 1940 > Volume 49, No. 196 > The Easter Island tablet Atua-mata-riri, by Alan S. C. Ross, p 556-563
                                                                                             Previous | Next   

- 556

UP to the present no Easter island tablet has been published in such a way as to render detailed work on the individual signs practicable. In the present article, I gave a ‘diplomatic’ text of the tablet Atua-mata-riri as illustrating a method which might facilitate this type of work.

The Easter island tablet Atua-mata-riri 1 is in the United States National Museum, Washington, where it bears the catalogue number 129773; it was purchased by pay-master W. J. Thomson from a native of the island during his stay there at the end of 1886. Virtually all that is known of it is to be found in Thomson's book, Te pito te henua, or Easter Island (Report of the National Museum 1888-'89, pp. 447-552), to which the reader may be referred for further details (p. 513 ff. and p. 537).

I obtained photographs of both sides of the tablet from the United States National Museum 2 (here reproduced as Plates 1 and 2) and enlargements were then made. From these enlarged photographs I constructed a sign-list which is here presented, again reduced, as Plates 3-5. The sign-list contains a tracing, made from the clearest example available on the tablet, of every sign appearing on the tablet. In the sign-list each sign is accompanied by a number, and a sign may then be referred to by its number, printed in italics. It is thus possible to transcribe the text by means of these italic numbers.

- i
- ii
- 557

All we know of the Easter island script is that, of two adjacent lines, one is inverted with respect to the other (this can easily be seen from the many obvious homo-signs which, in any pair of lines, are arranged either feet to feet or head to head). Since we do not know either the direction of reading or the interrelation of the lines and the faces, it will be necessary, in our transcription, to enumerate the signs in a purely conventional manner. I therefore enumerate the face (arbitrarily) numbered 1 (Plate 1) first and then proceed to the face (arbitrarily) numbered 2 (Plate 2). I distinguish the lines on each face by (arbitrary) letters and the position of each letter on the Plates shows which way up the line is to be read. I (arbitrarily) enumerate the signs from left to right.

A notation of the type 55 (56?) means that it is doubtful which of the two signs is to be read; a notation of the type ?34 means that the sign is defaced but is in all probability 34; ?s means that there are traces which suggest a definite sign but that this sign is either not identical with any other sign on the tablet or else is so defaced that it is impossible to decide which of a number of signs it is (as in the case of a homo-sign of which the right-hand part is missing); a question-mark (?) means that there are traces of signs but that it is quite impossible to resolve these traces; dots (……) indicate a part of the tablet much defaced, but one in which it may nevertheless reasonably be assumed that there once were signs.

The tablet may now be transcribed as follows:

  • A …..?89—17—133—35—96—71—57—61—?s—?
  • B ?129—?—23—17—154—105—37—37—144—49—143—17—141—100—99—25—1—?—?—55 (56?)—?s—?—?s …..
  • C …..?—102—16—128—102—83—19—102—96 (97?)—102—43—102—43—102—30—17—?s
  • D …..?—134—1—1—1—1—?s—? …..?—104—137—158—148—156—85—82—2—159—41—?s …..
  • E …..?34—34—138—84—138—139—70—84—101—85—87—43—147—46—? ….. ?—115—5—69—61—58—17—56—45—58—?
  • F ?—85—59—59—20—47—?s—?—?—?61—75—75—61—98—86—61—119—39—114—114—1—93—77—61—62—60—60—55—57—17—26—?s …..
- 558
  • G ….. 79—?85—?s—?s—55 (56?)—?s—?61—126—65—61—161—31—32—136—61—97—80—149—91—60—91—92—80—130—60—15—60
  • H 44—102—22—102—73—10—74—11—83—131—62—83—131—56—18—94—?s—?s—? …..
  • I ….. ?—?—?—? …..
  • [……………..] 3
  • A …..?s—135—113—94—41—42—?s—162—43—122—55—60——?s—?s—?s—?s—?s—?s—?s—?s
  • B 37—37—37—163—60—61—107—60—152—9—112—151—12—13—43—124—124—72—123—123—90—88—140—85—?s …..
  • C ….. 14—14—76—60—129—24—110—64—33—37—43—89—61—109—109—43—109—150—17—132—60—142—4—60—60—36—111—5—85—?
  • D 127—55—120—27—21—60—50—63—53—89—38—43—108—60—68—66—67—160—60—67—43—?s—?s—?—?s—?s—?s—?78—25—5—?s …..
  • E ….. ?—?s—85—81—? ….. ?—?s—145—1—85—6—24—54—48—157—121—51—40—51—118—28—7—78—?s
  • F ?—?—?—1—83—103—61—17—85—116—17—146—146—153—155—3—117—?s—?s—?s—?s ….. ?s—?s—?s—?s …..
  • G ….. ?—?s—?37—?55—85—41—41—4—60—29—95—17—43—41—?4—60—125—108—106—8—52—83—106—50—41—1—?s
  • H … 43—61—28 ….. ?s—?s—56—?s—56—?s—? …..

In the present state of our knowledge any comment on the inscription is clearly impossible. There remains, therefore, only one point for discussion. The tablet Atua-mata-riri is of especial interest by reason of the fact that, as he relates in his book, Thomson obtained from an old man, Ure-vaeiko, a Rapanui text purporting to be a translation of the tablet (Thomson, pp. 520-1). 4 In the past much doubt has been thrown on this translation (and the few others of the same kind), but the fact remains that there is no evidence of an a priori character to enable us to decide whether it is less likely that Ure-vaeiko's translation is what - 559 it purports to be than not. Ure-vaeiko's Rapanui text, which is extremely obscure, is of considerable interest, and Dr. J. Frank Stimson (of Papeete) has recently written a monograph on it (which he is, I understand, about to publish); this includes a translation. 5 He has been kind enough to place a typescript of this monograph at my disposal. I have therefore endeavoured to test whether the signs of the tablet can or can not correspond to Ure-vaeiko's Rapanui text.

It is clear that this problem affords the most general case of the hieroglyphic problem; there is here no trace of simplifying factors such as those which made possible the solution of the only four hieroglyphic systems so far deciphered. In the Easter island script—if indeed it is a script and not, as M. Métraux tentatively suggests, a mere mnemonic system 6—there are neither obvious word-dividers (such as facilitated the solution of the cuneiform—Persian type—, Runic Turkish and Ras Shamra scripts) nor any indication that a group of signs corresponds to some special unit in the translation (such as was afforded Young and Champollion by the Egyptian habit of enclosing a royal name in a cartouche, evidenced on the Rosetta Stone).

This general problem—does an array of unknown signs correspond or not correspond to an array of known semantemes 7 and/or known phonemes? 8—is obviously not generally soluble. A solution can however be hoped for if the method of ‘repetitions’ is tried; that is to say, it may be possible to ascertain, by trial and error, that a repetition in the unknown sign-array corresponds to a repetition in the known semanteme-array and/or in the known phoneme-array.

- 560

But this method breaks down at once in the case of the tablet Atua-mata-riri by reason of the incompleteness of the tablet. There are obvious gaps—at the ends and internally in the lines, further incomplete lines at the edges of the tablet. The method of repetitions cannot be used when one of the arrays is incomplete, for a repetition in one array may no longer appear as a repetition by reason of part of it having been lost.

It is therefore impossible ever to decide on these data whether Ure-vaeiko's text can or can not be a translation of the tablet. But, finally, it must be emphasized that this conclusion is absolutely indifferent: because of it we have no more reason to doubt the genuineness of Ure-vaeiko's translation than to believe in it.

- 561
- 562
- 563
1   According to Dr. Stimson (see below) the word is to be divided, atua ‘god,’ mata ‘eye,’ riri ‘wrath’ and translated, approximately, “god called Eyes-of-wrath.”
2   I should like to take this opportunity of expressing my thanks to the officials of the Museum for the photographs and for information on various points.
3   It is impossible to be certain, without examination of the original, whether there is or is not some trace of another line here. There appears to be some space which may have contained traces of signs.
4   Ure-vaeiko's translation is at all events perpetuated in the name of the tablet for Thomson doubtless christened it Atua-mata-riri because this is the first word of the translation.
5   It should be noted that the English translation of Ure-vaeiko's Rapanui text, given by Thomson (pp. 521-2) is not trustworthy.
6   See H. Lavachery, Ile de Pâques pp. 59-60; no doubt this is what M. Métraux has in mind in his recent article “The Proto-Indian Script and the Easter Island Tablets” (Anthropos xxxiii, 218-39) when he always refers to the Easter Islandscript” (with inverted commas) as opposed to the Indus script (without inverted commas).
7   A word indicating a meaning of an invariable and precise kind.
8   A single sound; or the sign for it.