Volume 104 1995 > Volume 104, No. 3 > Preliminary evidence for cosmogonic texts in Rapanui's Rongorongo inscriptions, by Steven Roger Fischer, p 303-322
PRELIMINARY EVIDENCE FOR COSMOGONIC TEXTS IN RAPANUI'S RONGORONGO INSCRIPTIONS 1
The present study demonstrates how the three Rapanui rongorongo inscriptions of the “Santiago Staff”, the “Small Santiago Tablet”, and “Honolulu Tablet 1” comprise to a large degree repetitive series of glyphic triads whose first constituent is invariably suffixed by a phallus-like glyph. The glyphic statement of these triads can be epitomised by the formula X1YZn. A nearly identical structure is embodied in the text of Ure Va'e Iko's ancient Rapanui creation chant “'Atua Mata Riri” witnessed in 1886. The discovery of these glyphic triads and their structural correspondence to Ure's cosmogony allows the first scientifically verifiable identification of the genre of a rongorongo inscription. It also suggests a partial phonetic statement.
Rapanui's arcane rongorongo comprises one of the world's few remaining undeciphered scripts (Fig. 1). 2 Any chance to glean rongorongo's meaning from Rapanui informants was lost in the second half of the 19th century with the final annihilation of the already impoverished ancient Easter Island society 3 All attempts to read rongorongo with the aid of mooted extrainsular cognate scripts have also failed, since Rapanui's script is evidently an indigenous elaboration from the last quarter of the 18th century. Kenneth Emory (1972) is generally cited as the first proponent of this viewpoint, though it was voiced as early as the 1870s and has since been repeated by many scholars. The author's own investigations have confirmed that the inspiration for Rapanui's rongorongo indeed lay in the witnessing of the Spaniards' document of annexation in 1770, though in its morphology and function rongorongo is wholly a Rapanui elaboration. The closest an internal analysis of the script has come to an identification of a rongorongo text has been Thomas S. Barthel's recognition in 1956 of some sort of a calendrical passage in the inscription incised on the “Mamari” tablet now located in Rome (Barthel 1958:242-7). Barthel conducted his initial rongorongo research between 1953 and 1956, the results of which were published in 1958. There can be little doubt that the passage on “Mamari” (RR 2a6-9) in some way involves a recitation of the names of the nights of - 304 the moon. However, it is unclear whether this is a list, song, calendar, or something else. A phonetic statement of this brief “Mamari” passage has not been forthcoming in the intervening 39 years.
Figure 1. The Rapanui rongorongo inscription “Small Santiago Tablet” (RR 8v). (Reproduced with the kind permission of the Musée de l'Homme, Paris)
Unhappily, rongorongo has been the victim of fringe scholarship almost from the day its existence was made known to non-Rapanui in 1864. The claims of these fringe scholars have included nearly everything from South American to extraterrestrial origins for rongorongo. Professional scholars who have been actively engaged in the decipherment of rongorongo in the last half of the 20th century have included Thomas S. Barthel of Germany; Viktor Krupa of Slovakia; Yuri Knorozov, Nikolai Butinov, Irina Fedorova, Konstantin Pozdnyakov of Russia; and Jacques Guy of Australia. The author has been investigating rongorongo since January 1989.
RONGORONGO'S SYNTACTIC STRUCTURE IDENTIFIED
For the first time in the 130-year history of rongorongo investigation, the syntactic structure of three rongorongo inscriptions appears to be retrievable. As a result, a partial phonetic statement of these three inscriptions can - 305 also be tentatively posited. The rongorongo glyphs on the entirety of the “Santiago Staff”, on one side of the “Small Santiago Tablet”, and on the one legible side of “Honolulu Tablet 1” seem, at the present stage of investigation, to comprise three separate cosmogonic texts—that is, creation chants. The “Staff”, acquired by the Chileans on Rapanui in 1870, is a two-kilogram sceptre measuring 126 cm x 6.5 cm that once belonged to an Easter Island 'ariki, a leader of a descent group. The “Small Santiago Tablet”, discovered on Rapanui in the same year by the same Chilean visitors, measures 32 cm x 12.1 cm x 1.8 cm. Both wooden artefacts have been deposited since the 1870s in Santiago's Museo Nacional de Historia Natural. “Honolulu Tablet 1”, obtained by the Bishop Museum in Honolulu in 1920 from the collection of J. L. Young, is a fire-damaged and weatherworn piece measuring 31 cm x 12.5 cm x 2.5 cm. 4
THE “SANTIAGO STAFF”
This apparent structural breakthrough in the decipherment of rongorongo was achievable only because the famous “Santiago Staff” is the one rongorongo inscription that marks textual divisions: It includes as many as 97 irregularly spaced vertical lines (Fig. 2). Each glyph on the “Staff” that- 306
begins a new division—that is, each glyph that appears immediately to the right of one of these 97 vertical lines—is suffixed with a phallus-like glyph (Fig. 3). 5 The left-to-right reading direction, evidently a legacy of the Rapanui witnessing the Spaniards' document of annexation in 1770, is borne out not only by reliable informant data from the 19th century but also by an internal analysis of the original inscriptions. Such an internal analysis reveals that parallel passages—that is, textual passages that recur in the same inscription or other inscriptions—continue boustrophedonically (“as an ox plows”), but rotating the tablet 180° at the end of each line, onto immediately following lines only if a left-to-right reading direction is heeded. This does not obtain with a right-to-left reading.
In the subsequent series of glyphs within each division, nearly every fourth, seventh, tenth, thirteenth glyph and so forth is also suffixed with this phallus-like glyph (Fig. 4). Yet out of as many as 98 separate divisions, none ends with a normal phallus-suffixed glyph; 6 nor is a division's penult glyph once suffixed with the phallus-like glyph (Fig. 5). 7 In addition, the divisions contain a minimum of three glyphs, never fewer (Fig. 6). Longer divisions - 307 of glyphs comprise nearly always multiples of three (Fig. 7). The first glyph of such a triad, with one exception (see note 5), sports a huge phallus, as does the prepenult of nearly every division (Fig. 8). Furthermore, glyph repetitions in discrete series also isolate into triads that sometimes repeat the first two glyphs of the preceding triad (Fig. 9). From this, one must conclude that on the “Santiago Staff” the triad structure of glyphs is the minimum textual statement.
Figure 5. “Santiago Staff”: absence of phallus-suffixed glyphs in penult and terminal positions of division.
THE “SMALL SANTIAGO TABLET” AND “HONOLULU TABLET 1”
This recognition on the “Santiago Staff” of a minimum textual statement of three rongorongo glyphs, the first glyph invariably bearing a phallus-like suffix, was achievable only because of the unique presence of vertical division markers on the “Staff”. However, the same structure can be analogously isolated in the texts of two further rongorongo artefacts which lack any sort of division marker. On one side of the “Small Santiago Tablet”, the first glyph in the bottom left-hand corner—the traditional, and analytically confirmed, beginning of a rongorongo inscription—is suffixed with an identical phallus-like glyph (see the bottom left-hand corner of Fig. 1). 8 Here the series of glyphs also divide into regular triads, though not so routinely as in the inscription on the “Staff” (Fig. 10). And here, too, the first glyph of such triads, or of slightly more numerous groupings of glyphs, invariably brandishes a phallus-like suffix. The same phenomenon is also revealed on the one legible side of the weathered “Honolulu Tablet 1” (Fig. 11).- 310 - 311
A FORMULA FOR ONE RONGORONGO TEXTUAL STRUCTURE
On the basis of these observations, it is possible to epitomise the textual structure of the three rongorongo inscriptions on the “Santiago Staff”, the “Small Santiago Tablet”, and “Honolulu Tablet 1” with the formula X1YZn (Fig. 12). In this formula, X designates the glyph to which phallus1 has been suffixed. Y represents a statistically more frequent glyph that follows X, Z a statistically less frequent glyph that follows Y. And n is the constant, here denoting unspecified repetition of the triad structure. Even without external corroboration, this epitomisation X1YZn comprises a significant insight into three important rongorongo inscriptions that could lead to a productive, if admittedly tenuous, series of philological speculations. 9
THE RAPANUI CHANT 'ATUA MATA RIRI
However, there is external corroboration. One stormy night on Rapanui in 1886, the American paymaster William Thomson, with generous assistance from a friendly bottle of spirits, was able to elicit from old Ure Va'e Iko—who was perhaps the most important Rapanui informant in the 19th century—the creation chant 'Atua Mata Riri. As I have been able to confirm in an independent study (Fischer 1991, 1992), Ure's creation chant is premissionary in origin but linguistically contaminated. It is likely that the chant, as Ure Va'e Iko had professed, was an original rongorongo composition that had been sung by the Rapanui tuhunga tā, the erstwhile experts in the script. (The Modern Rapanui phrase for this that is currently in vogue, ma'ori rongorongo ‘script expert’ is evidently a late 19th-century invention, since Modern Rapanui ma'ori was Old Rapanui tuhunga and since Modern Rapanui rongorongo [in its meaning of ‘script, incised artefact’, a demonstrable Mangarevan loan from 1871] was Old Rapanui tā.) 10 The chant is a cosmogony:“'Atua Mata Riri ki 'ai ki roto ki 'a Taporō: Ka pū te poporo”— “God Mata Riri [i.e. Angry Eyes] copulated with Sweet Lime [a Tahitian loan]: There issued forth the poporo plant”; “'Atua Matu'a ki 'ai ki roto ki 'a Pipiri Hatu 'ā: Ka pū te miro”—“God Parent [a missionary loan] - 312 copulated with Compacted Sand: There issued forth the tree” and so forth in a list of 41 fanciful copulations and issues (many of which are evidently Ure's own recent inventions or redactions). Here, one again encounters the structure X1YZ. The copulator is X; the phrase “copulated with” is superlinear1; the copulatee is Y; and the issue of the copulation is Z. It is a common rhetorical structure for ancient Polynesian cosmogonies and genealogies. According to John Chariot (1985:172): “Genealogies can take different forms. The most naturalistic is two-source, male and female: A + B → C…” It exactly reproduces the shared glyphic structure of the “Santiago Staff”, the “Small Santiago Tablet” and “Honolulu Tablet 1”.
In Fig. 13, one will notice that both 'Atua Mata Riri and 'Atua Matu'a are X, the subject of the phrase. The sequence ki 'ai ki roto ki ‘copulated with’ occurs in Ure Va'e Iko's chant 41 times. It is one of two invariable sequences that are repeated throughout the creation chant, and it is reminiscent of the phallus-like suffix on each X glyph that fronts a triad in the three cited rongorongo inscriptions. Here, Y is the two objects Taporō and Pipiri Hatu, those being copulated with. And Z is the issue of the union: the poporo plant and the miro ‘tree’. In this way, the X1YZ structure that had been identified in the three rongorongo inscriptions apparently finds replication in this creation chant performed by Ure Va'e Iko on Rapanui in 1886. Though Ure's chant admittedly is partially contaminated, the authenticity of its ancient rhetorical structure is beyond question. For this reason, and for its structural similarity to the identified rongorongo formula, it is probable that Ure's creation chant shows the three rongorongo inscriptions on the “Santiago Staff”, the “Small Santiago Tablet” and “Honolulu Tablet 1” to be similar premissionary Rapanui cosmogonies.
Why not genealogies? one might ask. First, in a genealogy the issue (Z) would have to be the subject (Z=X2) of each subsequent triad. However, this succession occurs only three times in Rapanui's longest rongorongo inscription, the “Santiago Staff”; 11 and it does not occur at all in the inscriptions on the “Small Santiago Tablet” and “Honolulu Tablet 1”. Second, the genealogi-- 313
cal formula XYZZAB BCD and so forth is unattested in Rapanui oral tradition. The cosmogonic formula X1YZ is attested for Rapanui—by Ure Va'e Iko.
As further evidence for cosmogonic texts in these three rongorongo inscriptions, I should like to point out that Ure Va'e Iko sometimes repeated the subject X in the Z position. This repetition is rare in Polynesian genealogies. 12 However, it is frequent in Polynesian cosmogonies. Here, the repetition yields the alternative formula X1YX (Fig. 14). One will notice in Fig. 15 that the sequential triads on the “Small Santiago Tablet” reproduce the same formula X1YX (this formula is also occasionally repeated on the “Santiago Staff”). If one juxtaposes the two—with Ure's chant on the left-hand side and the rongorongo from the “Small Santiago Tablet” on the right-hand side, as one can see in Fig. 16—very little difference in their structure is evident. Only Ure's second repetitive sequence ka pū te ‘there issued forth the’ is missing in the rongorongo text. Perhaps this is because one could “mark” an understood tautology in the rongorongo by means of a zero glyph; that is to say, the tautology's graphic statement is made redundant by the formulaic succession of remaining glyph slots.
The fact that superlinear1, the phallus-like glyph, is not logographic like X, Y and Z but semasiographic is probably attributable to the fact that it, too, was understood by the Rapanui scribes of all three inscriptions to be a tautology. However, it was a tautology that was not supposed to be marked by a zero glyph but instead displayed by a suffix glyph, perhaps in order to mark the role of the copulator and/or to emphasise the act of copulation. Here, the suffix glyph would be functioning as a verbal semasioglyph, in contrast to the nominal logoglyphs X, Y and Z. If this observation is correct, then it would evidence an astounding grammatical perception by the premissionary Rapanui people and would characterise their rongorongo script as being functionally far more sophisticated than one hitherto has generally assumed.- 314
Figure 16. A comparison of the X1YX alternative structure of Ure Va'e Iko's creation chant with that of the “Small Santiago Tablet”.
A PARTIAL PHONETIC STATEMENT
For the abovementioned reasons, one should be able to appreciate that, as demonstrated in Fig. 17, the three rongorongo inscriptions on the “Santiago Staff”, the “Small Santiago Tablet” and “Honolulu Tablet 1” now can be reduced to the partial phonetic statement: X ki 'ai ki roto ki 'a Y: Ka pū te Z, or ‘X copulated with Y: There issued forth Z.’ Of course, one cannot expect Ure Va ‘e Iko’ s own chant of 1886, even in part, to be replicated in any of these three rongorongo inscriptions. Only the genre of a cosmogonic text—not the same text—appears to be repeated in all four examples. 13
Figure 17. The partially deciphered phonetic statement of the textual formulae of three rongorongo inscriptions.
What does this tentative discovery signify for rongorongo research, indeed for the larger history of the decipherment of writing systems in general? For the first time since rongorongo's existence was acknowledged by non-Rapanui in 1864, it appears that the genre of no fewer than three of the extremely rare rongorongo inscriptions is distinguishable. This is a most fortuitous breach in rongorongo's erstwhile impenetrability. In the history of decipherment, this apparent advance parallels Alice Kober's identifica- - 316 tion in 1943-50 of inflection in Linear B 14 or Yuri Knorozov's discovery, in the early 1950s, that the Mayan script is logographic. 15 In addition, because these three inscriptions provide a fairly reliable semantic context for an extremely simple and repetitive rongorongo formula, the epigrapher, or specialist in deciphering ancient inscriptions, now has recourse to several decipherment methods which hitherto have been inadmissible for rongorongo for want of a reliable premise. Such methods would be particularly cogent with the “Santiago Staff”, since it comprises no fewer than approximately 2,300 elements and constitutes the longest inscription in Easter Island's limited rongorongo corpus. Also, the epigrapher can now accept with reasonable certainty that rongorongo is a mixed script, one that is both logographic and semasiographic—logographic in the sense that the X in the aforementioned formula is either a morpheme or a whole word or words (such as 'Atua Mata Riri ‘God Angry Eyes’), and semasiographic in the sense that superlinear1—the phallus-like suffix—provides visual communication directly, in this case ki 'ai ki roto ki 'a ‘copulated with’.
Since these inscriptions comprise creation chants, the epigrapher can further begin to isolate categorical groups in order to distil lexical identifications, particularly with the “Staff” inscription. These retrieved lexical values must then be tested against identical glyphs in other rongorongo inscriptions. One should commence this testing process with main signs whose values can already be tentatively posited, pending later adjustments or rejections. In this way, main sign series, such as that in Fig. 18 from the “Santiago Staff”, might finally begin to speak to us again in Old Rapanui. At this juncture I should like tentatively to posit the provisionally deciphered rongorongo sentence based on the above scientific retrieval method: “Te manu mau ki 'ai ki roto ki te ika: Ka pū te ra 'amacr”—“All the birds copulated with fish: There issued forth the sun.”- 317
However, the phonetic values of this reading are, in their entirety, only provisional. Furthermore, no Rapanui chant that reproduces this exact wording is known. So long as no external verification is forthcoming, any such “reading” of rongorongo, even one based on a solid scientific premise as here, will remain tentative, subject to later emendation or rejection. The value mau for ‘all’ was retrieved through the identification of the grasping hand (Old Rapanui ma'u ‘to grasp, seize’) in near-homophonous use and through the system-internal recognition of its function in the “Staff” inscription as a repetitive suffix (i.e. some sort of glyphic qualifier) that is shared with a small number of similar suffixal glyphic qualifiers. There is no evidence that the Modern Rapanui plural marker mau, which is shared with Tahitian and Marquesan, derives from an older *ma'u; that is, that the Old Rapanui plural marker at the end of the 18th century would have been homophonous with ma'u (‘to grasp, seize’). This being the case, one might assume, as Barthel held back in the 1950s with what he called “the principle of partial phonetic suggestion”, that rongorongo signs could also convey meaning in near-homophony and not exclusively in full homophony, while using, as here, the rebus principle. However, this assumption has yet to be scientifically verified.
If this identification of three cosmogonic texts in the rongorongo inscriptions is correct, then Easter Island's rongorongo script would have been much more than a mere mnemonic device that was used between ca. 1780 and 1865 by four generations of premissionary Rapanui singers/priests in order to aid them in recalling only previously memorised texts. The implications resulting from the above preliminary structural retrieval argue that a tuhunga tā, an ancient rongorongo expert, had to physically read the “Santiago Staff”, for example; there can be little question of prior rote memorisation with this particularly voluminous artefact that comprises hundreds of short triads of text. In other words, the Rapanui singers/priests were indeed using a form of writing and, what is more, they apparently were creatively composing in it. Easter Island's rongorongo is the only known indigenous script in Oceania before the 20th century.
This apparent breakthrough is not the decipherment of Easter Island's rongorongo. Notone can yet read the Rapanui inscriptions in their entirety. Nevertheless, the phallus-like suffix glyph that evidently renders the repetitive phrase ki 'ai ki roto ki 'a ‘copulated with’ may represent the first rongorongo glyph to be scientifically retrieved, in that its repetitive function was initially identified through internal analysis and that its phonetic value as then independently supplied by a structurally near-identical, documented rongorongo text from the 19th century. This is at best only preliminary evidence for a partial decipherment of Rapanui's rongorongo - 318 script. There are still many years of hard work ahead before one will finally be able to read the few surviving, yet ever fascinating inscriptions of Easter Island.
I am indebted to the anonymous reviewer, whose excellent suggestions have been incorporated into the final version of this paper. In particular, I should like to express my profound gratitude to Professor Emeritus Thomas S. Barthel of Tübingen, Germany, for his professional inspiration and personal support during the past six years that I have been researching Rapanui's rongorongo script. One generation ago, Barthel's monograph Grundlagen (Barthel 1958) turned rongorongo studies into a science, and the investigation of Rapanui's arcane script has been for Barthel the challenge of a lifetime. Though I may not always agree with his rongorongo interpretations, I will always remain in Thomas Barthel's debt, as both a colleague and a friend.
- 319 Page of endnotes- 320
- 322 Page is blank
1 An abridged version of this paper was presented at the 7th International Conference on Austronesian Linguistics, Noordwijkerhout, Netherlands, 22nd-27th August, 1994. It is based on the author's monograph-in-progress Rongorongo: The Easter Island Script.
2 Adequate to excellent historical overviews about rongorongo can be read in Routledge (1919), von Heine-Geldern (1938), Métraux (1940), Imbelloni (1951), Barthel (1958, 1963, 1971), Heyerdahl (1965), Fedorova (1975), van Hoorebeeck (1979) and Davis-Drake (1988-90). Rongorongo inscriptions are hereinafter catalogued as RR 1, 2, 3 and so forth according to the notation set forth by Fischer (1993). A small a after the numeral indicates the side of an inscription arbitrarily selected by Barthel (1958) as the recto (b for the verso) whenever it is unclear which side commences the inscription; otherwise one has r for recto and v for verso. A numeral after any of these letters indicates the line. Hence, “RR 2v1” means artefact number two, the verso or reverse, line 1.
3 See Métraux (1940), McCall (1976) and Maude (1981).
4 All measurements of rongorongo artefacts cited in this paper were effected in situ by the author.
5 The only exception to this is the first glyph of the twelfth division of RR 10-10 (Barthel: I 7), which apparently is a scribal oversight; this triad comprises its own division. Barthel (1958:280-2) was the first scholar to acknowledge that the suffix glyph could be a phallus; his interpretation derived from the “reading” by Bishop Tepano Jaussen's Rapanui informant Metoro Tau'a Ure on Tahiti in 1873, who regarded a similar glyph combination (i.e. anthropomorph with phallus-like suffix) on Jaussen's rongorongo tablet “Mamari” (RR 2r10) as te tangata ure huki ‘the man with erect penis’. Barthel later interpreted this sign, which he read phonetically as ure ‘penis’, to be ‘descent, patrilineal descent, clan’ (Barthel 1963:400, 432). However, Barthel's interpretation was based on the individual sign and not on its dependent function within the glyphic syntax.
6 The second division in RR 10-2 (Barthel: I 13) ends with a glyph bearing two suffixes, the last of which is a phallus-like glyph; here, it is possible that the phallus is either incorrectly suffixed or fulfilling a different function from that of the majority of such suffixes. The first division in RR 10-10 (Barthel: I 7) ends with a suprafused double phallus, not a suffix. The division after the aforementioned terminates with an independent phallus that is also not a suffix. A strange “piggy-back” arrangement of two glyphs with two phallus suffixes, most unusual for rongorongo, terminates the first long division in RR 10-13 (Barthel: I 10); yet the division that immediately follows comprises only three glyphs, the first of which is suffixed with the phallus-like glyph in normal fashion.
7 In RR 10-8 (Barthel: I 5), the penult of the tenth division comprises an autonomous phallus glyph, not a suffixed glyph.
8 I currently regard this to be the beginning of the “Small Santiago Tablet” inscription. However, Barthel (1958:end plates) designates this as the first line of the reverse. To avoid confusion, I have retained Barthel's traditional numeration with this tablet, designating this beginning line as RR 8v1 (Barthel: Gv1) with the understanding that it might comprise the obverse.
9 Though they fail to recognise this identified X1YZn structure, the two studies of the verso (i.e. recto) of the “Small Santiago Tablet” by Butinov (1960) and Barthel (1963, a critique of Butinov), and the one study of the “Santiago Staff” by Barthel (1990), merit comparative consideration.
10 This point is fully documented in Fischer (n.d.).
11 Searching for X1YZZ1, one finds RR 10-1 (Barthel: I 12), near the end of the first division; RR 10-3 (I 14), the third division and the beginning of the fourth; and RR 10-8 (I 5), in the 14th division. However, each of these repeats the same glyphic sequence. Granted, the apparently abbreviated version of the same sequence—X1YZ1 such as in RR 10-8 (I 5)—is slightly more frequent on the “Staff”; but it is not so frequent as to justify here the identification of a genealogy.
12 It is common in Polynesian genealogies that the son who has been named after the father be given a surname in order to distinguish him from his sire.
13 Commander Ignacio L. Gana who, in 1870, led the Chilean expedition to Rapanui, where he acquired the “Santiago Staff” from the resident Frenchman Dutrou Bornier, later reported that he had, at the time, enquired of the Rapanui people the significance of the “Staff” and that “… they pointed to the sky and the hieroglyphics that [the “Staff”] contained with such respect that I was more inclined to believe that these hieroglyphics recorded something sacred”(Philippi 1875:676). It would not be untoward to regard this performance on Rapanui in 1870 as possible informant evidence for a cosmogonic text on the “Santiago Staff”. However, one should be aware of the general unreliability of all informant material from 19th-century Rapanui; after the labour raids of 1862-3 and the subsequent pandemics, little was left of autochthonous oral tradition and much was borrowed, principally from Mangareva and Tahiti. See Fischer (1994).
14 See Chadwick 1967:35-6.
15 See Coe 1992:145-66.