The Bulletin du Cercle d'Études sur l'Île de Pâques et la Polynésie (C.E.I.P.P.) has published in its April-May 1999 issue, No.28, a letter by Jacques Guy arguing that recitation Atua-Matariri is not a procreation chant at all, but memories of "spelling" lessons overheard by Ure Vaeiko in his youth, when he was a servant of king Ngaara (some say his cook), the last king said to have had a knowledge of the writing.
Guy first draws attention to the absurdity of most of the verses of Atua-Matariri as reconstituted by Métraux and to the fact that, apart from Atua-Metua and Tiki-Te-Hatu, none of those gods and goddesses appear elsewhere in Polynesia, or even in other oral traditions of Easter Island. He then recalls the outrage of his Latin teacher at a fellow-student's unfortunate translation (Romulus and Remus, suckled by the Roman she-wolf by means of the brazen awl): "How dare you, how dare you; do you take the Ancients for cretins that you should have them spout such stupidities!" Do we not also take the Ancient Polynesians for cretins, Guy continues, that we should have them spout such stupidities as Moon by copulating with Darkness produced Sun (verse 25), Tail by copulating with Hina-oio produced the crayfish (verse 27), Stinging-fly by copulating with Swarm-of-flies produced the fly (verse 16), etc.? He then takes as example the Chinese character for copper , composed of metal and together , the formation of which can be described, Atua-Matariri style, as Together (), by copulating with Metal (), produced copper (), or, replacing "together", "metal", and "copper" by their Chinese pronunciations, Tóng (), by copulating with Jin (), produced tóng (). Under such a hypothesis the fantastic copulations of Atua-Matariri, and the genitors engendering themselves, become reasonable descriptions of the parts entering into the formation of compound hieroglyphs. At best, Ure Vaeiko would have studied the writing in his youth, but forgotten all, as his "readings" of the tablets demonstrate; as often in such cases, the first lessons learnt are forgotten last, and he would have remembered some of them, the formation rules for some signs. At worst, and that seems to be Guy's opinion, he never knew anything but, being a servant of Ngaara, he would have had the opportunity to overhear "spelling lessons" delivered by him, perhaps without even realizing their true meaning.
Guy's hypothesis is perhaps not so unlikely as it may seem. One detail has been overlooked by all authors, Guy included. Ka is not the mark of the past in the Easter Island language (that is kua), but of the imperative, and the meaning of "ka pû te..." is not "there issued forth the..." nor "produced the..." but "let the... come forth", e.g. By mating Ti with Ta, let the ti-tree come forth (verse 6). There is also the matter of the expression "ki 'ai ki roto..." translated as "by copulating with..." First, ki roto means into; second, the common, vulgar verb for "copulate" is transitive in most, perhaps all, languages, from Latin (futuere) to Chinese (cào ), and 'ai seems to convey precisely the same meaning and have the same function, viz. Englert's dictionary: 'ai coito, hacer coito los animales. (Es expresión grosera.)
Perhaps, then, the fantastic copulations of Atua-Matariri are nothing but mistranslations of formal descriptions of compound signs using a prescribed formula "by X being incorporated (copulating) into Y, let Z come forth". That hypothesis, however, cannot be verified short of deciphering the rongorongo and, as such, will always remain a mere hypothesis (Guy: "Il ne s'agit là que d'une hypothèse qui ne restera sans doute jamais rien qu'une hypothèse"), but it has the merit of crediting the Ancient Polynesians with intelligence and reason ("Je suis toujours persuadé que Atua-Mata-Riri est un abécédaire, parce que je ne puis me résoudre à prendre les anciens polynésiens pour des crétins").