English Français

This paper aims to improve documentation of the rongorongo tablet VI 4878 from the Berlin Ethnological Museum. For the first time (to the best of our knowledge), we present the detailed digital photographs of the entire artifact almost at the original scale. The computer enhancement of these images helped to produce improved tracings that surpass the tracings published by Adolf Bastian (1883), Thomas Barthel (1958) and Steven Fischer (1997) by the number of documented glyphs. Analysis of the parallel passages shows that the Berlin tablet shares glyph sequences with so-called “Great Tradition” (Large Santiago/Large and Small St. Petersburg tablets), as well as with Aruku Kurenga tablet. Some particular sign groups from the Berlin tablet can be seen on the tablets Tahua, Echancrée, the Small and the Large Washington tablets.

Cet article vise à améliorer la documentation de la tablette rongorongo VI 4878 du Musée ethnologique de Berlin. Pour la première fois (à notre connaissance), nous présentons des photographies numériques détaillées de l’artefact complet à l'échelle quasi-originale. L’amélioration de la qualité de ces images sur ordinateur a contribué à produire des tracés d’un plus grand nombre de glyphes que ceux publiés par Adolf Bastian (1883), Thomas Barthel (1958) et Steven Fischer (1997). L'analyse des passages parallèles montre que la tablette de Berlin partage les séquences glyphiques de ce que l’on appelle la "Grande Tradition" (tablettes Grand Santiago / tablettes du Grand et Petit Saint-Pétersbourg), ainsi qu'avec celle d’Aruku Kurenga. Certains groupes particuliers de signes de la tablette de Berlin se trouvent aussi sur les tablettes Tahua, Échancrée, et sur la Petite et la Grande tablettes de Washington.

Mots-clés :

île de Pâques, rongorongo, tablette de Berlin

Keywords :

Easter Island, rongorongo, Berlin tablet

Documentation of the tablet

Analysis of the parallel passages

Conclusions

PDF 5,7M Signaler ce document

1The rongorongo script of Easter Island (Rapa Nui) is a unique writing system developed in Polynesia. The oral traditions of the islanders « assert that Hotu-Matua, the first king, possessed the knowledge of this written language, and brought with him to the island sixty-seven tablets containing allegories, traditions, genealogical tables and proverbs relating to the land from which he had migrated » (Thomson, 1891: 514). However, the search of possible trans-insular origins of the script did not reveal any results supporting the external origin of the script. At the same time, the iconography of rongorongo signs pronouncedly matches the Rapa Nui petroglyphs (Fedorova, 1983: 45, fig. 2; Lee, 1992: 126-128; Macri, 1996: 184, fig. 19; Horley and Lee, 2008: 114, fig. 6), providing the evidence to Easter Island development of the script.

2The existence of inscribed objects was first witnessed in 1864 by the first Rapa Nui missionary Eugène Eyraud (Fischer 1997: 12), but it was Tepano Jaussen, the Bishop of Tahiti, who realized the scientific value of the discovery and started the search for the survived tablets in 1869 (ibid., 22). All known best-preserved rongorongo artifacts were collected in 1870; the tablets discovered afterwards were damaged to a considerable degree.

3The tablet in Collections of the Berlin Ethnological Museum (Museum für Völkerkunde) has a special position in rongorongo corpus. It was collected as a consequence of Geiseler’s Expedition that arrived to Easter Island in 1882 onboard the ship Hyäne. One of the foremost priorities of the Expedition, set by then-director of the Museum für Völkerkunde Adolf Bastian, was to procure more information about the inscribed tablets and, if possible, to acquire some of them (Fischer, 1997: 78-79, also 581: note 18). The principal Geiseler’s informant on Rapa Nui, Alexander Paea Salmon, negotiated with the islanders about rongorongo artifacts and was able to locate several tablets. As the owners refused to sell the inscribed artifacts to the Expedition members, Salmon promised to buy the tablets after departure of Hyäne and to send them to Gustav Godeffroy, the German Imperial Consul at Tahiti, for further shipment to Berlin. Salmon managed to acquire two inscribed tablets but sent them to Valparaiso to Heinrich August Schlubach, the German Consul, who was the husband of Salmon’s niece.

4One of these tablets was sent to Bastian and arrived to Berlin on April 27, 1883 (Fischer 1997: 80). It was deposited to the Ethnological Museum under the catalog number VI 4878 (ibid., 496). This artifact is usually referred in the literature as “Tablet O” / “the Berlin tablet” (Barthel, 1958: 27), “RR22” (Fischer, 1997: 494). It was also nicknamed “the boomerang” due to its particular shape (Imbelloni, 1951: 101).

5Once the Berlin tablet « had to be one of the most marvelous rongorongo inscriptions ever produced […] like the “Santiago staff” […] its entire text probably ran contiguously without blank edges in some 26 lines » (Fischer, 1997: 497). However, it underwent significant erosion due to the action of the elements, which erased the majority of its glyphs. This poor preservation state became a considerable obstacle for proper documentation of the Berlin tablet, making it one of the “marginal artifacts” in rongorongo studies. This paper is aimed to improve the situation, providing the high-quality digital photographs of the entire artifact with new tracings based on computer-enhanced images of the tablet.

6For the sake of uniformity, the paper uses Barthel’s nomenclature to address the inscribed artifacts, lines and individual rongorongo signs. Taking into account that Fischer recorded sign traces for two lines before the first line documented by Barthel, it was decided to use Fischer’s line numbering for this artifact (Fischer, 1997: 495-496). All tracings of rongorongo texts shown in the figures were made by the author unless otherwise noted.

Documentation of the tablet

7The first sketch of the Berlin tablet was published almost upon its arrival to the Museum (Bastian 1883: Plate 1.4); the drawing was accompanied with the following caption:

« […] by a kind donation of Mr. Schlubach (earlier [living] in Valparaiso), the Ethnological division of the Royal Museum (by negotiation of the Anthropol[ogical] Soc[iety] in March of [18]83 ) deposited a collection piece of very weathered wood, of which some of already visible signs are copied in their places, while the rest [of inscription] may become clearer only after prior preparation. »

8The adapted version of this drawing is reproduced in fig. 1. The inscription survived in short segments clustered around a large knothole and four fragments located almost at the opposite end of the tablet. The drawing published by Bastian is quite difficult to interpret as it shows many glyphic forms unusual to rongorongo, which has two possible causes: erosion of tablet surface and unfamiliarity of the draftsman with the signs of rongorongo script. The general estimation of glyph number recorded by Bastian overcomes 180, surpassing 90 elements documented by Barthel (1958: 28) and approaching 187 glyphs documented by Fischer (1997: 496).

Figure 1. – First published drawing of the Berlin tablet:

top – general view of the tablet; bottom – line-by-line transcription accompanied with line numbers according to Fischer (1997: 495-496).

9The photographs of the Berlin tablet were first published by José Imbelloni (1951: Plates 8a, general view and 8b, close-up) with the description:

« The tablet from Museum für Völkerkunde of Berlin […] is a large and relatively narrow piece of wood, intensively curved simulating the shape of a boomerang, with the dimensions of: 1.03m long and more or less 0.13[m] wide. Almost all the surface is rot and weathered, in particular the face that bears inscription; [once it] was, as suggested by its aspect, a plank of a boat. Its worst conservation state erased the major part of the signs … but it is still possible to distinguish quite a large number of them, in segments; they have covered the whole length of the tablet arranged in 7 superimposed lines […] A simple calculation deduced from the capacity of a single section allows to affirm that [the tablet] should have contained a total of 1,260 glyphs, and it is really a pity that such an important text founds itself useless for the science […] This artifact was never published. Prof. W. Krickeberg, director of the Berlin Museum, has kindly supplied me with 4 photographs (1 of the general view and the rest corresponding to 3 sections of its length) and I have thought to reproduce them completely, but the difficulty to discern the signs caused me to give up. » (Imbelloni, 1951: 101)

10Thomas Barthel (1958: 27) worked on illustrations from Bastian, Imbelloni, photographs by W. Lehmann and slides from the Berlin Ethnological Museum. He describes the tablet as follows:

« The backside [of the artifact] is completely destroyed, and the front side contains the remains of at least seven lines of characters. Presumably, there were 10 lines with a text of 1200 to 1300 elements – today only about 7% of them can be identified with certainty (at most 90 elements). Even these remains still prove that it is a) a separate text, b) it is a classical style text. » (Barthel, 1958: 28)

11Steven Fischer studied the Berlin tablet in 1991, making many important observations. He proved that the tablet was never used in marine carpentry:

« Imbelloni, Barthel and Helfrich believe the “Berlin Tablet” was originally part of a canoe; however, for this it is missing the necessary lashing holes as one finds on RR16 [tablet S, Large Washington tablet] and 18 [tablet P, Large St. Petersburg tablet]. » (Fischer, 1997: 497-498)

12This point is illustrated with fig. 2 that schematically depicts the aforementioned artifacts. As one can see, both Large St. Petersburg (P) and Large Washington (S) tablets features numerous boreholes along their long edge; additional holes appear at the sides. The photographs of these tablets show that the holes postdate the inscription as they cut through the glyphs. The diameter of the holes is sufficient for passing the lashing cords. There even exist the photographs of Large Washington tablet (Heyerdahl, 1975: Plates 68c and 59c) with a cord passing through one of the holes. Additionally, the geometry of Large St. Petersburg and Large Washington tablets favors their use as canoe planking – the both artifacts are flat, straight and relatively thin: 2cm for tablet P (Fischer, 1997: 483) and 1.6cm for tablet S (ibid.: 472). The Berlin tablet is different – it is considerably thicker (5.2cm, Fischer, 1997: 496) and curved, which makes it unsuitable for canoe planking. Moreover, as pointed out by Fischer, Berlin tablet does not have even a single perforation (fig. 2).

13Fischer’s monograph (1997: 494-495, figs. 67-69) presents close-up photographs of the Berlin tablet. These pictures convey a good impression about the state of the artifact. However, as a meter-long tablet is shown in three 11cm-wide images, the resulting reproduction scale of about 30% is not sufficient to discern individual signs clearly.

Figure 3. – Berlin tablet with rongorongo inscription. Wood, 6 × 103 × 12.5cm. Inv. VI4878

a) Inscribed side; b) weathered side with faint remains of fluting. Asterisks mark the burnt areas.

14A particular difficulty in photographic documentation of the Berlin tablet comes from the combination of the factors – the tablet is large but its glyphs are small and eroded, requiring reproduction of the photographs at actual scale to achieve the sufficient quality to distinguish the signs. Thanks to the exceptional collaboration of the Art Resource (New York) and the Ethnological Museum (Berlin), Claudia Obrocki made a set of 35 Megapixel digital images of the Berlin tablet meeting all the aforementioned requirements. These photographs (figs. 3-7) allow for the first time (to the best of the author’s knowledge) to publish the entire artifact in high image quality. Figure 3 presents the general view of the artifact from the both sides; detailed overlapping close-ups of inscribed side are given in figs. 4-7.

15As one can see, the tablet is much eroded, coinciding with Fischer’s description (1997: 497):

« Bearing some 25 knotholes, it was already greatly weathered and cracked before being incised with glyphs, suggesting that it had originally been a piece of driftwood. The wood is yellowish in colour where exposed; otherwise it is dark brown. Many glyphs are visible inside burnt edges, proving burning occurred (leaving at least five burnt marks) after the glyphs had been incised. »

16The number of cracks is indeed impressive; however, not all of them pre-date the inscription. For example, a long crack passing by the large knothole following the line Oa6 (fig. 5) cuts through the signs, suggesting that it may postdate the inscription. The burnt marks reported by Fischer are comparably small and scattered around the edges of the artifact. To facilitate location, they are marked with asterisks in fig. 3. Two of burnt marks appears on the inscribed side at the beginning of line Oa10 (fig. 3a), the others are discernible on the back side of the artifact (fig. 3b): a burnt knothole at the middle of concave part, a dark spot located about a third of the convex part and yet another spot at the edge, extending to both sides of the tablet at the beginning of line Oa3. In addition to knotholes, cracks and burnt marks, the tablet also features shallow depressions, seven of which follows the line Oa10 (fig. 3a, also fig. 7).

17The back side of the artifact is seriously damaged (fig. 3b). Fischer (1997: 497) suggests that:

« this is the side that must have lain face-down in the soil […] The artifact was once finely fluted, though entire sections of the inscription are now so damaged that even the fluting is often no longer distinguishable. »

Indeed, even back side of the tablet shows some faint traces of fluting, proving that the tablet was most probably completely covered with inscriptions.

18The scholars disagree upon the number of lines composing the Berlin inscription. Fischer (1997: 496) reports that:

« [inscribed] side a [has] 7 visible out of at least 11 original lines, perhaps 12 if the edge have been used; [eroded] side b [once had] 13, perhaps 14 if the edge has been used. »

19Barthel (1958: 27) states that:

« presumably, there were 10 lines [on the inscribed side]. »

Figure 3 shows clear traces of ten lines. However, it should be noted that line Oa1 (as identified by Fischer) starts at the top left part of the artifact (on the side opposite to that with a largest knothole) and goes to the right, which contradicts the common writing order:

« the reading should commence at the lower left-hand corner, on the particular side that will bring the figures erect, and followed as the characters face in the procession, turning the tablet at the end of each line, as indicated. » (Thomson, 1891: 516)

20To fulfill these requirements one has to assume the existence of line Oa0 starting at the side with the largest knothole (i.e., upper right part of the tablet, above the beginning of line Oa2) and going to the left (so that the picture of the tablet in fig. 3a will show the artifact upside-down regarding the proper beginning of the text). In this case, the sequence of lines will follow that proposed by Barthel and Fischer. At the same time, one may notice that the line Oa10 does not reach the edge of the artifact as well. Therefore, most probably there was line Oa11 that commenced below the row of hollow depressions at bottom left side of the tablet as shown in Fig. 3a and following to the right in complete fulfillment of first line identification criterion as cited from Thomson. If this was the case, the existing line order should be reversed.

21However, there is not enough data to offer a solid proof for the correct line sequence of the Berlin tablet: there are neither parallel passages spanning over the neighboring lines nor traces of vertical glyph compression indicating that the scribe was limited from below / above with previously written line. Under these circumstances, one may only hope that proper line sequence could be revealed when rongorongo will be eventually deciphered. Meanwhile, it seems most useful to adhere to the line sequence proposed by Fischer, which accounts for the glyph traces in two lines before the first line documented by Barthel. We also expand the inscription with the line Oa10 that shows the traces of three signs that were not previously documented in the literature.

22Addressing the bad condition of the tablet and considerable complications arising on producing the tracing of its text, Fischer writes (1997: 497):

« It is difficult to draw the glyphs of this piece with reliability, since they are so greatly deteriorated. There are many glyphs that at this juncture simply cannot be adequately transcribed; however, computer enhancement will soon be able to achieve a new transcription, surely providing many more glyphs. »

23Following Fischer’s advice, the close-up images of the tablet (figs. 4-7) were subjected to multi-stage computer image enhancement including local histogram equalization with different processing window size, high-pass filtering and gamma adjustments performed for individual color channels. The resulting images were used to produce new tracings presented in fig. 8. To facilitate the comparison of tracings with the photographs, the beginning of each continuous inscription segment is labeled with number in a circle (with numbers ranging from 1 to 19); the corresponding numbers are also shown in figs. 4-7 with arrows denoting the direction in which the segment goes. This visualization technique appears particularly useful for location of short groups composed of incomplete signs, such as segment 1 (fig. 4) surrounded by completely eroded wood, showing the outline that can correspond to a star glyph 8. Fischer documents here the complete star glyph, also drawing a bird glyph in the same line (Fischer, 1997: 495), the traces of which were impossible to locate in the present study. In the same way, marking the segments in the photographs allows easy location of the glyphic group showing the tail and the wing tips of the bird belonging to line Oa10 (fig. 5, segment 19), surviving on non-eroded portion of wood close to the large knothole.

24The present study documented 230 identifiable glyphs as well as 112 glyph elements that do not allow unambiguous transcription (e.g., legs of anthropomorphic signs without traces of head / hands). On the positive side, the aforementioned glyph elements clearly define the space allocated for particular sign, considerably narrowing the spectrum of possibilities for identification of the glyphs. Upon the eventual decipherment of rongorongo this information will allow to reconstruct the missing parts of the inscription to a considerable degree. Therefore, we can confirm the presence of approximately 340 glyphs and elements thereof, which introduces an improvement upon the previous highest glyph number of 187 recoded for the Berlin tablet by Fischer (1997: 496). The further increase of glyph count can be achieved using the special imaging techniques such as polynomial texture mapping (see Graeme, Martinez and Malzbender, 2010) capable of extracting more information from the eroded sections of the artifact.

25To address the possible changes of the preservation state of the Berlin tablet with time, we have underlined the glyphic passages documented by different authors (fig. 8). If the tablet was in considerably better preservation state in 19^th century, Bastian should have documented more glyphs (or more sections of the inscription) that is currently visible. As one can see, this is not the case. All the passages traced in 1883 (marked with dotted underline) are clearly seen from the modern images, confirming that the tablet did not undergo any deterioration during the past century. Barthel’s documentation recorded only the signs with most prominent outlines, resulting in comparably low glyph number. This fact can be tentatively explain