A new manuscript of Pua Ara Hoa 'a Rapu from the archives of William Mulloy, Part 2: The manuscript in the context of known Rapanui manuscripts

Following the publication of the photographs from Manuscript H labeled with Pua Ara Hoa's name in the first part of the paper, we report the results of comparative analysis of Manuscripts H, A, C, and E. All three sections of Manuscript H–the list of rongorongo signs, the list of place names, and the list of the kings from Hiva and Easter Island–show remarkable differences with versions appearing in other manuscripts, suggesting that Manuscript H was possibly written on an earlier date. The Manuscripts A, C, and E could have been partially based on Manuscript H or another similar document.

Siguiendo la publicación de las fotografías del Manuscrito H inscrito con el nombre de Pua Ara Hoa, continuamos con una análisis comparativo de los Manuscritos H, A, C, y E. Las tres secciones del Manuscrito H–la lista de signos rongorongo, la lista de topónimos, y la lista de los reyes de Hiva y Rapa Nui-muestran considerables diferencias con las versiones que aparecen en otros manuscritos, sugiriendo que el Manuscrito H fue posiblemente escrito en una fecha temprana. Los Manuscritos A, C y E podrian estar parcialmente basados en el Manuscrito H o en algún otro documento similar.

Introduction

In the first part of this paper (Horley & López Labbé 2014), we reproduced previously unpublished photographs of Rapanui Manuscript H, inscribed with the name of Pua Ara Hoa 'a Rapu. Since publication of that paper, we received extraordinary feedback from Patrick C. McCoy: "The MS. was photographed [on Rapa Nui] by Herb Pownall during the 1968 [Archaeological] survey [of Easter Island initiated by William Mulloy] and Ahu [Ko te Riku and Ahu Tahai] restoration project" (McCoy, pers. comm. 2015).

In Bulletin No. 1, corresponding to the Easter Island field season of February–July 1968, Mulloy writes the following: "Photographic work for both projects was accomplished by Pownall, Chief of the Photo Service of the University of Wyoming and his son Paul. Since August 1st the photographic work has been under the charge of Rafael Rapu H" (Mulloy 1997:5) and "Pownall was required to divide his time between survey and restoration projects in addition to designing and constructing the photographic laboratory and maintaining film development and printing" (Mulloy 1997:6). Alas, we were unable to find any additional information concerning the possible provenance of the manuscript.

The second part of the paper, presented here, focuses on the analysis of three main sections of Manuscript H: the list of rongorongo signs accompanied with comments / interpretations in the Rapanui language, the list of place names, and the list of the kings.

Rongorongo Signs from Philippi's Tracings of the Santiago Staff

Every known Rapanui manuscript contains fragments of rongorongo script, either a copy of Jaussen's list (Manuscripts A-D, F, G), lists of glyphs with tentative interpretations inspired by it (Manuscripts A, B), or rongorongo passages that may coincide with those of the inscribed tablets (Manuscript E). Manuscript H is not an exception–it has a list of rongorongo characters with their tentative interpretations on pages 7 and 8 (Horley & López Labbé 2014:Figures 1 & 2). Many signs are drawn upside-down on purpose, which is surprising for a "rongorongo dictionary." The very same list of glyphs also appears in Manuscript A:

"The text in Figures 111-121 [in Heyerdahl & Ferdon 1965] of the Atan manuscript [also known as Manuscript A] is associated with copies of glyphs occurring on the Santiago stick [sic, Santiago staff]. This text proves to be more connected and coherent, and has been preserved in a better state that the text associated with the glyphs in Figures 96-107 [that represent Jaussen's list] of the same manuscript. The sentences of this text are arranged to look like interpretations of the associated glyphs, and are arranged in two columns, as in the copy of Jaussen's dictionary"

(Kondratov 1965:405).

The Santiago staff was presented by Jean-Baptiste Dutrou-Bornier to the officers of the corvette O'Higgins in 1870 (Fischer 1997:455) and several years later was deposited in the Museo Nacional de Historia Natural, Santiago de Chile. Rodulfo Philippi, then director of the museum, published the detailed tracings of the Santiago staff in 1875, splitting it into separate lines. However, there was a problem with a line that runs about three quarters of the length of the staff. Philippi correctly guessed that this unusual line might signal the end of the text (Philippi 1875:678), yet his drawings took a safer approach by presenting the unusual shorter line merged together with a neighboring line, resulting in bands of straight and upside-down glyphs (Figure 1a, line XII, segments b-d). The line following the unusual "doubled line" in Philippi's tracing had to be shown upside-down, reflecting the break of the inverse boustrophedon pattern. It is worth noting that two other plates (with Philippi's line numbers I-X) have all signs straight.

Philippi's last plate for the Santiago staff is a good candidate for a source of the upside-down signs appearing in Manuscripts A and H. Kondratov (1965:406) was completely correct in his observation that the order of the signs in the list of Manuscript A did not follow that on the staff. The signs are drawn roughly, as if being based on a quick copy made by a person who had only brief access to Philippi's tracings. Such rapid sketching makes the exact identification of the signs complicated. The most reasonable matches are shown in Figure 1, side-by-side with Philippi's tracings. We tried to improve the initial sign identification made by Kondratov (1965:413-415). To simplify the discussion, each glyph at the bottom of the figure was numbered and oval marks were added to possible similar signs in Philippi's tracings. As one can see from the figure, every sign from the list can be associated with Santiago staff glyphs. The large number of upside-down signs fits the mixture of straight and upside-down lines in Philippi's tracing. It is important to emphasize that the calligraphy of the glyphs from Manuscript H is more similar to that of rongorongo in comparison with sign shapes seen in Manuscript A.

One comes to the conclusion that the passages written with the glyphs were not intended to be exact translations by noting that similar-looking glyphs are accompanied by different text, as in the case of the upside-down bird glyph with a zig-zag object interpreted as kikita kikita ki raro (Manuscript A, sign #5), poki tae maiare (Manuscript H, sign #8), and he ika ku toparia ana te haoa (Manuscript H, sign #18). A separate upside-down bird that is interpreted as poki vahine (sign #12) and poki manu (sign #14) in Manuscript A; the interpretations from Manuscript H are even more divergent. The compound sign #20 consists of two elements, with the first being similar to sign #19, suggesting that the interpretation of sign #20 will start with a part corresponding to sign #19. Yet, both captions have essentially the same meaning:

Sign #19: E rua       kahui              o te korotea

                     Two bundles of banana of korotea variety (that)

                            i kumi

                            grew long

Sign #20: He rua              maika              mo te koro

                     Two banana for the feast/father, (which) grew long,

                            i kumi i kumi te kahui

                            grew long, (in) bundles

The differences between the phrases "interpreting" the signs from Manuscripts A and H are summarized in Table 1, where one can see, for example, word substitution: in #10, repa ("young man") changes to tangata ("man") and in #12, tamavahine ("female") changes to poki vahine ("daughter"). The phrase o roto i te taua "migrates" from glyph #9 in Manuscript H to sign #7 in Manuscript A. The most interesting–in our opinion–are the following changes. The word nee in #6 of Manuscript A is erroneous, while me'e ("thing") in Manuscript H is correct. In #18, Manuscript H gives the phrase he ika ku toparia ana te haoa ("the victim was torn by a wound") with ku-ana forming a proper verbal frame, while kua from Manuscript A has a distinct function (Englert 1948:464). Also, the word kumi ("long") in Manuscript H appears in Rapanui dictionaries while kuni, from Manuscript A, does not. The passages for signs #10 and #11 use the particle te (corresponding to the definite article) in Manuscript H, which is given as e in Manuscript A. These observations make it tempting to speculate that Manuscript H, with its more coherent text and more accurate rendition of rongorongo signs (in comparison with Manuscript A), may possibly indicate that Manuscript H (or another similar document) might have served as the source for the creation of Manuscript A.

The interpretations of the signs do not form contiguous narration, yet some meaningful groups can be established. For example, signs #2-6 are accompanied by the text he tangata e ruku ana, ki roto ki te pu. He tangata here ruku. Kikita kikita ki raro koe, me'e kiki noho koe, which is very reminiscent of the description of a confrontation between Hotu Matu'a and Oroi. Oroi made a loop with a rope (Manuscript E, Tori 3: he hakapu i te taura) on the road on which Hotu Matu'a would be walking, and hid nearby to ambush the

Figure 1.

Tracings of the Santiago staff: top, published by Philippi (1875) with the spacing between the lines increased to accommodate glyph numbers; bottom, drawings of the glyphs from Manuscripts H and A

(Heyerdahl 1965:Figures 119-122, images courtesy of the Kon-Tiki Museum).

king. Hotu Matu'a saw the loop, stepped on the rope, and fell down on purpose to lure Oroi out of his hiding place. Though not following Manuscript E (Barthel 1978:354) word-for-word, the interpretations supplied for signs #2 and #3 in Manuscript H describe a similar situation: he tangata e ruku ana (a man dived) ki roto ki te pu (into the hole/loop). As Oroi rushed from his hiding place to attack Hotu Matu'a with a bone dagger, the king killed his adversary with a spell: kikita kikita taviri tavara ki raro koe ka mate (Manuscript E, Tori 4) clearly identifiable with the interpretation of sign #5. Barthel approximately translates the spell as ". . . turn around as though gripped by a dizzy spell and fall to the ground, die!" (Barthel 1978:198), and also supplies a version thereof recorded by Campbell: ka kikikiki ka harara koe with the translation "remain stiff forever" (Barthel 1978:199). The latter version is very similar to the interpretation of sign #6, me'e kiki nōho koe ("you remain a stiff thing"). The other (though smaller) meaningful groups of sign interpretation phrases includes poki tae maiare / poki ma[i]are (Manuscript A, glyphs #8, 9), poki vahine, poki manu, and poki tane (Manuscript A, glyphs #12, #14, and #15).

Table 1.

Text associated with the signs from the Santiago staff. The differences between two lists are given in boldface.

The List of Place Names

The second part of Manuscript H records a lengthy list of place names spanning from page 8 to page 10 (Horley & López Labbé 2014: Figs. 2-4), and these probably once occupied part of the now-missing pages 11-12. The list is preceded by the text: "He vaenga tini o te henua te kona ne(i) ko Hanga Ra'u 'a te ariki. A Mako'i Ri(ngi)-ringi 'a Hua Tava te kī nei i tataku nei (h)e 'ing(oa) o te kainga" that can be translated as "The very middle of the land is that place, Hanga Rau of the King. To Mako'i Ringi-ringi of Huatava it was told to count the names of the places." In Manuscript E, an essentially similar list of place names is preceded with a lengthier introduction:

"Then Ira continued to speak to Makoi: "Tomorrow, when it grows light, set out and name the places beginning with Apina." Makoi replied, "How shall I give the names?" Again Ira spoke, "In Hiva are the names that are to be taken to name (the places of the new land)." It grew light and Makoi got up. He set out and came to Apina. When he arrived there, he gave the name "This is Apina Iti, this is Rapa Kura." He went on and came to Hanga O Ua. He gave the name "This is Hanga O Ua of the beautiful wave (vave renga)." Makoi went on, giving names, until he made a (complete) circle around both sides (of the island). In Apina Nui a stone (maea) was erected, saying that the naming was done on a (round) trip during a single day"

(Barthel 1978:61-62).

Thus, both Manuscript E and H assign the task of compiling a list of place names to Mako'i (Ringi-ringi), son of Hua Tava. The resulting list of place names contains about sixty entries in Manuscript E; the total number of place names once written in Manuscript H is unknown due to the missing pages 11 and 12. The cross-reference of both lists was of crucial importance, because in some cases the inscription of Manuscript H was too damaged to be transcribed without a priori knowledge of its possible contents.

The very similar list of place names was also recorded by Englert. His unpublished Manuscript DE-145 from the collections of the Biblioteca William Mulloy (Figure 2) has the following introduction:

"We are giving in continuation the list of some place names of the coast, starting with Apina Nui and following to the west, north, east and south of the Island. These are the names that Señora Maria Atan [Mariana "Te Rora" Atamu Niare (1905-1971), C. Moreno Pakarati, pers. comm. 2014], of the Pakarati family, remembers having heard from her mother Hilaria Pakomio, now deceased. We did not try to accompany them [the names] with a translation, because, as happens with proper place names, they contain obscure expressions and moreover seem to refer to personal episodes that are not known today. In some [cases], as for Ahu Tongariki for which it speaks about cooked rats, an ironic note is [definitely] present. It is desirable that during the works of monument restoration a [special] importance will be given to record the largest possible quantity of the ancient place names for the entire island, because the memory of these names will disappear in very few decades"

(Englert n.d.a; translation from Spanish by the authors).

The archives of Englert also contain Manuscript DE-095, which contains two slightly different preliminary lists that were merged together to form DE-145. Importantly, there is a short introduction narrating the assignment of the place-naming task (Englert n.d.b): "He kī a Ira ki a Raparenga: Ka-oho koe ko tou mahingo ka-nape te ingoa o te kainga ka-varivari-ró te aro mai Apina Iti, ana haka-mata i te ingoa o te kainga" that can be translated as "Ira said to Raparenga: go with your people, put the names of the land; go around the island from 'Apina Iti when starting [to assign] the names of the land." This is an interesting turn, because, according to the lore, Ira and Raparenga belonged to the same family among the seven explorers sent to Rapa Nui to verify the dream voyage of Hau Maka (Barthel 1978:54):

Sons of Hau Maka	Sons of Hua Tava
Ira (first born)	Kuukuu (first born) Ringiringi Nonoma Uure
Raparenga (last born?)	Mako'i Ringiringi (last born)

Thus, in the versions of Manuscript E and Manuscript H, the leader Ira, of Hau Maka's part of the team, gave the order to survey the land to the youngest

Figure 2 (part A).

The original document is kept in the Archives of the Capuchin Order, Santiago de Chile). The figure is composed of two manuscript pages. On the left page, the initial letters are missing due to the page having been cut. They have been reinserted as type characters.

List of Rapa Nui place names recorded by Englert (manuscript DE-145, reproduced courtesy of Biblioteca William Mulloy.

Figure 2 (part B).

The original document is kept in the Archives of the Capuchin Order, Santiago de Chile). The figure is composed of four manuscript pages. On the lower right page, the initial letters are missing due to the page having been cut. They have been re-inserted as type characters.

List of Rapa Nui place names recorded by Englert (manuscript DE-145, reproduced courtesy of the Biblioteca William Mulloy.

son of Hua Tava; in Englert's version, Ira assigned this task to Raparenga, the younger brother of his own family. Another mention of family-related interactions in the explorer's party is recorded within the place name list itself. For Hanga te Pau (the bay in which Ahu Vinapu is located), Manuscript H says (Figure 4, lines 4-6):

He óti te ki nei ka tahi painga o te kainga.

The talk about the first part of the land finished.

1 Ko te tomonga o Ira he rua tini o te kainga.

Ko Hanga té Pau o Ira.

1 The landing place of Ira is the second center of the land.

Hanga te Pau of Ira [emphasis added by the authors].

Keeping in mind that the place naming was done by Hua Tava's family member, this comment seemingly establishes a certain kind of boundary. The other copies of the "place name list by Mako'i" do not contain such a bold formulation; for example, Manuscript E says (Barthel 1978:88):

Ko Hanga te Pau a Ira.Hanga te Pau of Ira.

He tini o te kainga a Hanga te Pau.The center of the land at Hanga te Pau.

To simplify further discussion of the place name lists, we present three copies of the lists studied side-by-side (Table 2) with place name orthography according to Alarcón (2008:43-45). Each list entry is marked with a number with prefixes "H", "E", and "T" standing for Manuscript H, Manuscript E, and the Tradition recorded by Englert. The general numbering is added for reference; it was also used for plotting the locations of place names on the map (Figure 3).

Each list entry in Manuscript H is numbered in a cumbersome way. The place names are clustered in groups of ten and a number from 1 to 10 is used to refer to the names within the cluster. However, the scribe apparently had complications with numbers exceeding ten, so that the notion of a large number was expressed by using an "improper" number at the very beginning of each group, so that the third group features the numbering "4, 2, 3, 4, 5 . . . 10" (Horley & López Labbé 2014:Figure 3, lines 7-17), the next group goes on with "5, 2, 3, 4, 5 . . . 10" (Horley & López Labbé 2014:Figure 3, lines 18-28). Yet, within this special system, the numbering was reset to "1" at Hanga te Pau, perhaps to highlight the notion of a boundary. Under these circumstances, it was deemed too complicated to use the numbering proposed by the scribe, so that a new continuous numbering system was introduced starting with H01 for Hanga Ra'u 'a te Ariki. Some list entries were too damaged to be transcribed entirely. In the particular case of H56 (Horley & López Labbé 2014:Figure 4, line 23), we are sure about the presence of the line (as it appears close to the beginning of the group), yet we were unable to distinguish any part of the phrase with certainty. The passage "mu ka tao ka" that seemingly appears on page 8, line 25 (Horley & López Labbé 2014:Figure 2) actually belongs to line 29 of page 10 (Horley & López Labbé 2014:Figure 4), so that the corresponding text appears in entry H61 (Table 3). The reason for this is that a large triangular-shaped piece is missing from the leaf (pages 7 & 8).

In Manuscript E, each place name is listed with a number (Barthel 1978:322-324). The numbering is continuous and starts with E01, Apina Iti. In two cases, the scribe decided to change the order of the sites, possibly to keep the proper arrangement along the route taken by Mako'i:

29 Ko te Rano a Raraku

30 Ko Oparingi a a Uuri

31 Ko Motu Kumu Koka a Kaoa

[starting a new page of Manuscript E]

31 Oparingi a Uuri

32 Motu Humu Koka a Mare a Kaoa

33 Hanga Maihiku a Papa Hakakiva

42 E tai e Puku Hotake e a te tini i uta he Hare Rourou Koveka

43 Vai Ngaere a Puku Hehaheha

[starting a new page of Manuscript E]

44 E tai e Teho e. Ka tao taau ngu. E Po e Kiko e, ka tutu tou oone

45 Vai Ngaere a Puku Hehaheha

The number 55 is missed altogether (Barthel 1978:324). As a consequence, the list contains only 57 entries though its last entry being numbered as "60". Despite this, corrections to the numbering system used in Manuscript E are minor, so this numbering was retained in Table 2.

It is worth mentioning that Manuscript E presents a second copy of the same toponym list in a form of pata'uta'u recited by Mako'i to the string figure made by Ira (Barthel 1978:62). The few spelling differences between the main list and pata'uta'u are presented in curved brackets in Table 2. The final version of the place list recorded by Englert (Figure 2) does not contain any numbers; therefore, it was numbered throughout from 1 to 63. In the cases when an entry for a certain list includes two entries for other lists, they are marked with the letter "a" (see, for example, place numbers 35 and 37 that correspond to the same entry E57, Table 2). We plotted the locations of the place names on the map (Figure 3) using several publications for reference (Barthel 1962, 1978; Charlin Ojeda 1947; Cristino et al. 1981; Englert 1948; Hotus 2007). In total, we were able to locate over 70% of the places. The first impression from the figure is that the density of place names significantly increased in

Figure 3.

Location of the place names mentioned in Manuscripts E, H, and the Tradition recorded by Englert. The numbers on the map correspond to those in the left column of Table 3. The locations of toponyms mentioned in one list only are marked with an open circle. The place names listed in all three sources are marked as a solid circle. Some instances when the second part of the place name represents a nearby location are marked with arrows.

Table 2.

The list of place names complied by Mako'i Ringiringi 'a Hua Tava.

Figure 4.

King lists from Routledge's field notes: a) the list of Ure Vae Iko, recorded from words of Veri Tea 'a Tea (Routledge n.d.:Reel 2(2), p.391); b) the list of "the kings who came with Hotu Matua" provided by Kapiera (Routledge n.d.:Reel 2(2), p.376); c) the list recorded from words of Kapiera and Porotu (Routledge n.d.:Reel 2(2), p.328). Images Copyright the Royal Geographical Society with Institute of British Geographers.

Digital images courtesy of the Pacific Manuscripts Bureau.

the vicinity of Hanga Roa, with approximately 20 toponyms mentioned–about one third of the total number of place names appearing in Manuscript E or those recorded by Englert.

Table 3.

List of kings from Hiva.

Such a marked preference towards the Hanga Roa shore area can be explained by several factors. One can assume that the surroundings of Hanga Roa were historically more densely populated so that people produced more toponyms here. Also, while the list of place names may be of considerable antiquity, it could have undergone modifications upon being passed verbally from one generation to another. During the late 19th to early 20th century (the period when Pua Ara Hoa's manuscripts were probably written), the main population of Rapa Nui lived in Hanga Roa. Therefore, it is natural that the names of places near Hanga Roa were more familiar and were remembered with greater ease.

Barthel made an interesting comment upon studying the list of toponyms from Manuscript E:

"Locating the names on a good map of Easter Island, one quickly recognizes the omission of a number of names. Important cult places, such as Ahu Tepeu, Ahu Hekii, and Ahu Vinapu are missing, as are Anakena, the royal residence, and Orongo, the center of the birdman cult. This might argue against a political or legal function of the list"

(Barthel 1978:91).

The absence of 'Orongo from the list is understandable, because the cult of the birdman was established during the later period of Rapa Nui's history. At the same time, the absence of a royal residence in 'Anakena–which is present in Hau Maka's dream voyage described in the same Manuscript E–indeed seems strange. Both remaining sources analyzed in the paper contain 'Anakena, albeit in different forms. Manuscript H mentions it as Hanga Ra'u (which is another name for 'Anakena Bay; Sutterle 2012:53), while traditions collected by Englert provide the name given by the soul of Hau Maka, Hanga Morī a 'One Tea. It is important to emphasize that in Manuscript H, 'Anakena is first on the list, as one would expect for the royal residence. The further place names are named along the route going around the island in a counter-clockwise direction. In contrast, Manuscript E and the Tradition recorded by Englert trace the coastline in a clockwise direction starting with 'Apina Iti or 'Apina Nui, respectively. This shift is explainable if one considers the narrative of Manuscript E-the seven explorers arrive to the island, follow the path taken by the soul of Hau Maka, return to the coast of Hanga Roa where Ira assigns Mako'i a task to walk around the island giving names to the places. In this setting, it is logical to start naming with nearby 'Apina rather than returning all the way to 'Anakena. The tradition recorded by Englert matches the name system of Manuscript E closely. In contrast, the list of place names in Manuscript H is given independently of any narrative, resulting in a natural shift of the "initial place" to 'Anakena.

The general construction of the place names in the "list of Mako'i" contains several names joined with the particle "a" marking possession: #1 Hanga Ra'u 'a te ariki "Hanga Rau of the King", #26 'Ōkahu 'a Uka Ui Hetu'u "'Ōkahu of the Girl who watches stars." Sometimes there are three names joined in the same fashion: #45 Ko Motu Kaviti 'a Turu Kupenga 'a Haka Viri. The same notion may be seemingly expressed in a different way: #36 'Ata ahi-ahi tō'ou, e Puoko Honu ē "The evening shadow is yours, the Head of the Turtle" (Puoko Hōnu is the name of a rock standing in the sea in front of Ahu Makere, near 'Apina Nui). In general, it is difficult to establish the nature of the second name. In some cases, it looks as a personal name, as in #70 Ko te Rano 'a Raraku "The volcano of Raraku," while in others, it seems to be purely descriptive: #30 Hanga o Uo 'a Vave Renga "Hanga o Uo of a beautiful surf" (Barthel 1978:76). Surprisingly, there are place names for which the second name is also a (generally nearby) place name, such as: #29 Hanga Roa 'a Tuki Tukau (Motu Tuki Tukau is located on the coast north from Hanga Roa o Tai Bay); #63 Koe-koe 'a Nuahine Rima Roa ika ra(hi?) ("The old woman with a long arm" is known from lore, e.g., Métraux 1940:370-371); at the same time, the place Nuahine Rima Roa exists to the northeast of Koe-koe Cape, (Figure 3), and also #82 Ko Ma'unga Tea-tea 'a Pua Katiki (Pua Katiki is the summit of the Poike peninsula on which Ma'unga Tea-tea is located (see Figure 3). Such pairing of toponyms may represent the traces of "place hierarchy", as in the latter example: Ma'unga Tea-tea of Pua Katiki, because it belongs to the locality "governed" by the summit of Poike. On the other hand, mentioning two names together marks the boundaries of a locality rather than a spot. The vestiges of this type are also found in the list, usually giving a reference point at the shore, tai or highland, 'uta: #62 E tai e Puku Hōtake ē, a te tini 'i 'uta te Hare Rou-rou Kōveka (Puku Hōtake is towards the shore, the house Rou-rou Kōveka is in the middle towards the inland, according to Barthel 1978:86).

Several place names are even more complicated, featuring lengthy additions (Barthel 1978:82) that can be possibly related to "chants or fragments of folklore" (Fedorova 1988:81, note 17). In the general numbering used in Table 2, these are the places: #53 Ma'unga Marengo (Mt. Orito), #57 Renga Havini, #58 Teho, #59 Hanga (west of Hanga Te'e, Manuscript H only), #61 Ohunga (east of Hanga Te'e, Manuscript H only), #66 Huareva, #70 'Oroi, #71 Ahu Tuta'e, and #79 Tongariki. When these extended toponyms were known only from Manuscript E–albeit appearing in two copies, the list of place names proper and the pata'uta'u chant–it was complicated to make a judgment about their nature. The discovery of Manuscript H changes the situation. It is written by the same person as Manuscript E, but it presents a distinct version of the list with many additional toponyms; the "survey route" starts from 'Anakena and goes around the shore in an opposite direction to that of Manuscript E. However, despite these differences, only the same few places are associated with extended names in both Manuscript E and Manuscript H, discarding the idea of a possible "contamination" with folklore data as mentioned by Fedorova. It could also be said that Manuscript H clarifies–to a certain degree–the presence of extended names by introducing a kind of boundary at Hanga te Pau as he rua tini o te kainga, "the second center of the land". Almost immediately after that begins the series of the extended toponyms with Ma'unga Marengo (Mt. Orito). The extended names follow the entire south coast up to Ahu Tongariki. Remarkably, Manuscript H features two additional extended toponyms for the vicinity of Hanga Te'e that are absent from both Manuscript E and the Tradition recorded by Englert. It should be stressed that the text added to extended toponyms is definitely related to the surroundings of the place–such as mentioning 'Oroi being cooked in the earth oven for the region of 'Oroi (Table 2, #70) or "call[ing] out to mother (over there), to Motu Roa" (Barthel 1978:87) for Hue Renga Havini (Table 2, #57). The "long islet" Motu Roa indeed exists in front of Ko Renga Havini (Cristino et al. 1981:Quadrant 6, Vaihu). Thus, the phrases added to extended toponyms seem to code the place name information in an indirect way, resorting to the references from the folklore if needed. Such notable differences in the style of toponyms used for the different regions of the island are extremely interesting and require much additional study.

King Lists

The records of the royal genealogies contribute considerably to ethno-historical studies. The common ancestors in genealogies of Polynesian islands may reveal migration patterns and connections between the ruling dynasties. On Rapa Nui, the situation is complicated:

"If Easter Island had kingly genealogies similar to those recorded for other parts of Polynesia, most of her mysteries might be solved. It would be possible to determine the approximate date of the first landing of Polynesian immigrants to the island, and thus to relate its traditional history to that of some other island. The early missionaries and travelers who inquired about the past of Easter Island accepted as genealogies mere lists of kings. Each author who got a list of Easter Island kings was convinced that he had the correct sequence of ariki-mau (supreme chief) who succeeded to power by order of primogeniture. But there are several lists which coincide neither in names nor in number of kings. A comparison of these documents gives one the impression that the informants remembered the beginning of the genealogies and of course the last kings, who lived in recent times, but forgot or confused many of the intervening names."

(Métraux 1940:88-89).

Under these circumstances, each list of kings is important, as it may potentially contain new information or reinforce already known data. In this regard, Manuscript H is interesting by featuring two king lists–one naming seven rulers of Hiva (Horley & López Labbé 2014:Figure 5, upper part) and continuing on to Rapanui kings (Horley & López Labbé 2014:Figure 5, upper & lower parts). We present these king lists in Tables 3 and 4, side-by-side with the lists known from the other manuscripts.

Table 4.

The list of Easter Island kings from different Rapanui manuscripts.

The line before the first list states he ariki o Hiva, "kings of Hiva." The list contains seven numbered entries starting with "Tangaroa, the king of Hiva." The very similar list appears in Manuscript E, starting with 'Oto 'Uta as father of Tangaroa. The two last names in Manuscript H are Ta'ana Harai and Hotu Matu'a. In Manuscript E, both of them are split in two, producing the names of four rulers: Harai, Ta'ana, Matu'a, and Hotu. In this way, the list of names, which records seven generations in Manuscript H is expanded (with the addition of 'Oto 'Uta) to ten generations in Manuscript E, supposedly to obtain the "ideal magic number" of ten (Barthel 1978:11). The reduced list of six kings appears in Manuscript F, written by Gabriel Hereveri (Table 3, right column).

After the list of kings, Manuscript H adds:

ó Marae Tohia. te ariki nei. ehitu.

These are seven kings of Marae Tohia.

In Manuscript E, the description is elaborated further (Alarcón 2008:13):

o Ma'ori te 'ariki nei, e tahi te 'angahuru. Ko Ma'ori te 'ingoa o te kāinga, ko Marae Renga te 'ingoa o te ma'ara noho o te 'ariki nui. Ko Marae Tōhia te rua ma'ara noho o te 'ariki.

These are ten kings of Ma'ori. Ma'ori is the name of the land; Marae Renga is the name of the place selected for residence of the great king. Marae Tohia is the second place selected for residence of the king.

According to Englert (1948:22), Hotu Matu'a lived in Marae Renga and Marae Tōhia was the residence of his sister, Queen Ava Rei Pua.

The text of Manuscript H goes on with "[the ariki] o te Pito o te Kainga" (in line with the previous title "he ariki o Hiva"), which is followed by the list of 30 names starting with Hotu Matu'a's son, Tu'u Maheke. As the bottom part of the page is missing, we actually have only list entries 1, 2 (very fragmented), and also the names from 20-30 appearing on the next page. Comparison with other manuscripts (Table 4) brings forward several thought-provoking details. The lists are coherent, except for three kings #23-25 in Manuscript H–Te Pito o te Kainga, Pure Nahe, and Hau Moana. Manuscripts A and C have only two kings in this part of the list–Honga and Te Kena. It is worth mentioning that the king Hau Moana occurs earlier in Manuscript A (it also appears in Roussel's, Thomson's, and Métraux's lists on positions #16, #44, and #9, respectively; see Métraux 1940:90). Therefore, it is most likely that the king list from Manuscript H omitted Hau Moana in his predicted position #8. The names Te Pito o Te Kainga and Pure Nahe do not occur in other king lists known to the authors. There is an interesting detail concerning the replaced names of Honga and Te Kena mentioned by Englert (1948:78): "Honga and Tekena do not seem to have been Ariki henua, rather than descendants of Hotu Matu'a by lateral line" (translation from Spanish by the authors).

The last numbered entry of the kings list from Manuscript H gives the name of Rokunga, who was actually a birdman (Englert 1948:78). A careful study of the photograph shows that the scribe first wrote "ko Rókuga, a Gaara," erased the patronymic, and wrote the corrected version "a Kai Makoi" outside the erased area. The two additional lines below the list mentions the second Ngaara in relation with Kai Makoi and Rokoroko he Tau; this part is embedded inside the king lists from Manuscripts A and C. The list of Manuscript C continues with the names of important Rapanui persons who appear in genealogies of the islanders published by Englert (1948:54-63; 66-68) and Hotus (2007).

It is important to emphasize that the lists of Rapanui kings from Manuscripts A, C, and H coincide with each other except for a couple of names. The very same list of kings was recorded by Silva Olivares (Butinov & Knorozov 1957:42) and Métraux from:

"the daughter of Ure-Vaeiko, one of Thomson's informants. Besides the native memory, my list is based on a fragmentary record that my friend Tepano had inscribed in a notebook many years ago"

(Métraux 1940:89).

Barthel (1959:79) published the genealogy of the kings recorded by Gabriel Hereveri in Manuscript F, which we present here in Table 5. It features a good correlation with the versions known from Manuscripts A, C, and H for early kings. Curiously, the king Miru is mentioned twice–for the first time as the son of Tu'u Maheke and then as the son of Hetuke (Table 5). Another paper by Barthel (1961) discusses two genealogies reported by Lanyon-Orgill (1960), recovered from papers by Palmer and Routledge. Palmer's genealogy contains only 13 names, the majority of which are well-known from the other documents. The notable additions are the kings Mirumiru and Tuutuu-maha (Barthel 1961:138) that follow Miru, son of Tu'u Maheke.

The most interesting situation, in our opinion, concerns the second king list published by Lanyon-Orgill, which was supposedly:

"derived from the Routledge expedition and the original was among the papers placed on deposit in the British Museum in 1920; it is listed in the catalogue of those papers which Mrs. Routledge herself compiled, but is not otherwise documented. During a search of the Museum's records in 1959 I failed to find the item in question, but fortunately I have a manuscript copy of it, derived from a photostat which was prepared for me by the Museum in 1940.

Table 5.

The lists of Easter Island kings recorded by Gabriel Hereveri and Katherine Routledge.

As the original cannot now be traced, this important genealogy is here printed. It is an important document because it is considerably longer than any of the other known genealogies, even that recorded by Paymaster Thomson"

(Lanyon-Orgill 1960:9).

The "Routledge king list" published by Lanyon-Orgill contains 69 names. Barthel's analysis of this document shows that the name order generally follows Thomson's list of 57 kings, "intercalated" with names on Métraux's list of 30 kings (Barthel 1961:138-139), with only a few names "swapped". Several of the king names listed by Lanyon-Orgill can be traced to the list compiled by Jaussen, Roussel, and Palmer. Few names are totally new: Atua Mata, Ure Mata, Tikitena, Te Teratera, Morumoru, and Punapuna. It goes without saying that the authors became intrigued by the existence of the "Routledge list of 69 kings," as no such document was ever published by Routledge herself. The only information she provides is:

"There are various lists of the succession of chiefs, counted from the first immigrant, Hotu-matua. The oldest lists are those given by Bishop Jaussen and by Admiral Lapelin, which contain some thirty names. Thomson gives one with fifty-seven. In our day there was admittedly much uncertainty about the sequence, but the number was said to be thirty, and two independent lists were obtained. All these categories differ, though they contain ma[i]nly the same names, particularly at the beginning and end" (Routledge 1919:241), adding a footnote on the same page: "Thirty is, however, a very favourite number: cf. the folk-tales."

We performed a search in online databases of the British Museum in hopes of finding the document mentioned by Lanyon-Orgill, but without success. A careful study of Routledge's field notes (Routledge n.d.) revealed two large name lists, one containing 30 and the other containing 22 names. We reproduce these lists here for the first time, thanks to the kind permission of the Royal Geographical Society (with the Institute of British Geographers) as Figure 4 and Table 5. The longer list of 30 names (Figure 4a) is prefaced by a passage attributing it to Daniel Ure Va'e Iko (the Elder interviewed by Thomson about reading of rongorongo tablets). The list is said to be written from the words of "Neri Tea Atan [sic, Veri Tea 'a Tea, Fischer 1997:130], wife of Te Haha." The second shorter list (Figure 4c) is associated with the names Kapieri (Kapiera Reva Hiva, Fischer 1997:128) and Porotu (Hongi 'Atua 'a Ure Au Viri, Fischer 1997:131). It is important to note that Ure Va'e Iko's list has a column of numbers on the right with a title "Kapieri's list," while the Kapiera / Porotu list has a similar column entitled "Ure Vae Iko' list" (Figures 4a and 4c, respectively). The discovered documents fit perfectly into the notion of two lists with approximately 30 names (Routledge 1919:241), strongly suggesting that these were the principal king lists recorded by Routledge. There are a few short fragments, such as the sequence "Hotu Matua–Ngaara–Oroi–Maurata–?–Liloroko (added: Valparaiso)" accompanied with a caption "Kings as per Juan" (Routledge n.d.:Reel 2(2):234) and "Te Kena Orne–Tupa–Ko Te Kena A Ong–Tupa a Ota–Atamu Te Kena Maurata (added: chief)" with a caption "Late R[apanui] Kings" (Routledge n.d.:Reel 2(2):378), neither of which represents the entire reigning sequence of Rapanui 'ariki mau. In view of this evidence, the "Routledge list of 69 kings" announced by Lanyon-Orgill raises several questions. Indeed, if Routledge would have recorded such a long list of the kings, she might have mentioned it elsewhere in her publications. The presence of cross-referenced columns in each list (Figures 4a & 4c) suggests that the two lists considered form the complete data set analyzed by Routledge. It is also suspicious that the "Routledge list of 69 kings" was studied from the manuscript copy of the original that disappeared (Lanyon-Orgill 1960:9). Taking into account these observations, the authors are inclined to think that the "Routledge list of 69 kings" as published by Lanyon-Orgill should be treated with caution until more evidence about its existence and authenticity can be found.

In analyzing the king lists of Ure Va'e Iko and Kapiera / Porotu (Table 5), one can confirm that king names recorded by Routledge are consistent with those appearing in Rapanui Manuscripts (Table 4), neglecting the minor spelling mistakes such as "Nui Tupahotu" in place of "Riu Tupahotu" or "Te Hetuki" in place of "Hetuke". Several king names are provided with previously unknown "secondary kings" (e.g., "Ko Kao Aro Aro" / "Ko Kao Erua" as well as "Ko Toko te Rangi" / "Ko Toko te Rangi Iti iti"). The names "Ko Mahuta Ariki" and "Ko Tupu Tua Mahomea" from the list of Ure Va'e Iko were not previously recorded. The Kapiera/Porotu list features the names "Ko Riri Katea", "Tangaroa", "Ko Rongarongo" (Korou a Rongo?), "Vaka Tangata", "Ko te Kena One-one", "Papue", and "Ngata hora". Three former names were known from the king list recorded by Thomson. The names "Rokunga" (last in the list of Ure Va'e Iko) and "Ngata Hora" (added in pencil after the list of Kapiera / Porotu) were recorded by Routledge as the names of birdmen.

Surprisingly, Routledge's field notes contained yet another list of Rapanui kings (Figure 4b) labeled "The kings who came with Hotu Matua." It is tempting to interpret it as a list of local chiefs who were subordinate to Hotu Matu'a. The list includes the names of Nuku, Hinelilu (Hinariru), Ataata, Te Varu Re Nea (Renga?), Hotu Matu'a himself, Tukoihu (Tu'u ko Ihu), Horka (Horia or Hoka?), Pupu he Tai, and Te Marumaru.

Therefore, a comparative analysis of king lists from Manuscripts A, C, H, and F confirms their general agreement, with only few names not matching between the manuscripts. The same small differences suggest that the lists were not copied verbatim from one manuscript to another. In contrast, the lists recorded from the living memory differ to a larger degree. Previously unpublished records from the field notes of Katherine Routledge provide new material. At the same time, it can be said that the "Routledge list of 69 kings" published by Lanyon-Orgill and Thomson's list of 57 kings should be viewed with caution due to in-mixture of mythical and historical names (Métraux 1940:94).

Discussion about the Dating of the Manuscripts

All known Rapanui manuscripts contain only two written dates. Manuscript B features "1865" penciled on its front page, which:

"is obviously too early to reasonably represent the year of origin of this manuscript. Although Eyraud discovered the [rongorongo] tablets on arrival in 1864, Zumbohm did not reach Easter Island until 1866, and brought the first tablet to Jaussen in 1868. It seems likely, however, that the number 1865 may indicate the hypothetical year of origin suspected by the person who pencil-marked the copy in Spanish"

(Heyerdahl 1965:363).

The second date is 1936, accompanying the Moon calendar in Manuscript A (Heyerdahl 1965:Figure 128). The night names from Manuscript A differ from those in the Polynesian Moon calendar (Kondratov 1965:416, Table 4), yet the total number of nights–30–fits well with the length of the synodic month (Horley 2011:20). Two records connected with Moon observations in Manuscript C (Heyerdahl 1965:Figure 149, pencil notes page bottom; Figure 158, notes at page bottom) were also shown to correspond to the year 1936 (Guy 1992:122), setting a reasonable upper limit for the manuscripts' updates prior to their discovery by the Norwegian Archaeological Expedition.

Van Hoorebeeck (1979:278) made a sound observation that the use of paper with printed marks in Spanish on the pages of Manuscript A (Heyerdahl 1965:Figures 112, 114 & 136, bottom of page) suggests that its creation date was after 1888; this argument holds also for Manuscript H. At the same time, we find it difficult to confirm that the manuscripts (at least, the surviving ones) were "first written down in the 1890s" (Fischer 2005:148). The main problem lies in the fact that the majority of the manuscripts include copies of Jaussen's list (Jaussen 1893:259-270). Van Hoorebeeck (1979:186) reasonably asks how much time it took for Jaussen's posthumous paper (published in Paris) to reach Rapa Nui, suggesting that it may have been Routledge who brought copies of Jaussen's list to the island. Other evidence also seemingly points to Routledge–the rongorongo passages in Manuscript E, according to Barthel, present excerpts from the recto side of the London tablet copied after Routledge (1919:Figure 98), and the recto side of Aruku Kurenga.

One can try to estimate the date of writing of the Pua Ara Hoa manuscripts with two principal approximations: the "strict solution", assuming direct participation of Pua Ara Hoa in manuscript creation, and the "non-strict solution" for which the writing might have taken place after Pua Ara Hoa passed away, using his name to legitimatize and validate the recorded tradition (C. Moreno Pakarati, pers. comm. 2014). In this paper, we tried to find a "strict solution", looking for the earliest plausible date for the writing of the Pua Ara Hoa manuscripts. To do so, we establish a set of constraints connected with Manuscripts B, E, and H that bear Pua Ara Hoa's name:

1. 1. Manuscript B contains a copy of Jaussen's list published in 893;

2. 2. Manuscript E contains rongorongo excerpts from the recto side of the London tablet, the recto side of Aruku Kurenga;

3. 3. Manuscript H contains signs from the Santiago staff copied from Philippi's 1875 paper showing the glyphs with mixed up/down orientation;

4. 4. Routledge met Pua Ara Hoa in 191, yet his name did not appear in Estella's census of 1918.

It is true that Routledge brought several photographs of rongorongo artifacts with her; some of these–the London tablet and the London reimiro–were published in her book (Routledge 1919:Figures 98 & 115, respectively). Routledge's field notes (Routledge n.d.) contain photographs of both sides of the Mamari tablet and the verso side of a plaster cast of the Small Santiago tablet. Routledge abundantly cites Thomson, so it is possible that she might have shown a copy of his report with pictures of the Aruku Kurenga tablet to the islanders (Thomson 1891:Plate 49). She was also perfectly aware of Jaussen's work (Routledge 1919:247). With this information, we can satisfy constraints #1 and #2 listed above. The main problem occurs with Philippi's tracings of the Santiago staff. We were unable to find any mention of them in Routledge's book or in her field notes. The time factor is also quite problematic. Even if Routledge brought all the required materials in 1914, it would allow only a couple of years–at best–for the completion of Manuscripts B, E, and H with Pua Ara Hoa's participation or supervision.

More complications surface in comparative analysis of the manuscripts. For example, the toponym list of Manuscript H contains a larger number of names and starts with 'Anakena–a natural choice of an initial point of all the places of the land. In contrast, in Manuscript E the list is blended into a larger narrative starting with a less prominent site of 'Apina Iti. This seemingly suggests that the list in Manuscript E may be a re-worked version of an earlier list, for which the toponym list of Manuscript H is a possible candidate. Another line of indirect evidence consists of the awkward numbering system used in Manuscript H, when the scribe had difficulties writing numbers above ten. In contrast, in Manuscript E, the two-digit numbering is fluent, but the scribe has problems with three-digit page numbers–the notion of "105" is given as "Tori 5" translated as "above 5" by Barthel (1978:355). The list of kings from Hiva in Manuscript H also underwent certain changes in Manuscript E. These subtle details incline the authors to think that Manuscript E was possibly based–at least, in part–on Manuscript H or another similar document, and that a considerable period of time might have passed between the writing of these documents.

There is a footnote in Kondratov's paper regarding how the tracings of rongorongo objects from Santiago might have reached Rapa Nui:

"In 1911 reproductions of these tablets were brought back from Chile to Easter Island by Knoche (1925, p. 242), who showed them to the elder natives in a futile attempt to get a translation . . . In 1924 drawings of the signs on the Santiago stick were published by Brown (1924, pp. 88, 91, 92, 96) . . . It may be presumed that Brown's reproduction, like those of Knoche, came back into the hands of Easter Islanders"

(Kondratov 1965:405, Editorial footnote by Heyerdahl & Ferdon).

It is worth mentioning that both options were acceptable at the time of writing for the second volume of reports of the Norwegian Archaeological Expedition because the only piece of evidence with the Santiago staff tracings was Manuscript A with the date 1936 on one of its pages. The situation changes with the discovery of Manuscript H, suggesting that the tracing of the Santiago staff could have reached the island at a much earlier date so that Pua Ara Hoa could have accessed them. This constraint essentially discards the line of MacMillan Brown, who arrived to Rapa Nui in 1922 (Fischer 2005:183).

In turn, Walter Knoche stayed on Rapa Nui at a much earlier date, on April 13-25, 1911 (Fischer 1997:122). Knoche was interested in rongorongo and made some brief inquiries:

"The copies of the Easter Island script that I brought from the Museum in Santiago [emphasis ours] were allegedly interpreted by two of the oldest men. Since the interpretations, however, were completely different and the extent of content recited by each individual stood in no relation to that of writing, one can assume that the knowledge of the script is unknown to the current inhabitants, especially as the entire environment [group of people] at the recitation explained that the ancients knew nothing about it, but that each of them told one of the many traditions, which were also known to the others without the aid of writing"

(Knoche 1925:242-3, translation from German by the authors).

The list of references in Knoche's book includes Jaussen's 1893 paper with the sign list (Knoche 1925:7) and also the account of Don Felipe González published by Hakluyt Society under the edition of Bolton Glanvill Corney (Knoche 1925:6), which features photographs of both sides of the London tablet (Corney 1908:Plate facing p. 128). Thomson's Easter Island report does not appear in Knoche's reference list, but is cited several times (Knoche 1925:240, 247, 254, & 306). Thus, in our opinion, Knoche's visit satisfies all four constraints mentioned: Knoche came to the island in 1911 when Pua Ara Hoa was in better health, he presumably brought to the island copies of the rongorongo inscriptions from the Santiago Museum (copied to Manuscript H), Jaussen's 1893 paper with the sign list (copied to Manuscript B), as well as Thomson's 1891 report and Corney's 1908 book with photographs of the Aruku Kurenga and the London tablets, the excerpts of which made their way into Manuscript E.

It is quite possible to envision the following scenario: Knoche's interviews with the elders might have catalyzed the idea of recording traditions, and perhaps it was even Knoche who left a couple of new Chilean notebooks on the island. After several manuscripts were composed by / under the dictation of Pua Ara Hoa, the writing tradition attracted more people, resulting in several (though considerably similar) scribal hands observable on the pages of Manuscripts A and C. Jaussen's list was eventually reproduced several times in Manuscripts A, C, D, and G. The signs from the Santiago staff were re-drawn in Manuscript A following the graphics of Manuscript H, with certain changes to sign interpretations. The vestiges of the manuscript tradition were then noticed by Routledge with the discovery of Tomenika's ta'u (Routledge 1919:Figure 99). The other manuscripts were kept in secret, though their texts were updated on several occasions, offering a plausible explanation to recent dates seen in records associated with the Moon in Manuscript A. Finally, the existence of multi-page Rapanui manuscripts was discovered in the 1950s by the Norwegian Archaeological Expedition.

Conclusions

The text of Manuscript H includes the interpretations of rongorongo signs copied from Philippi's tracings of the Santiago staff. The sign list is followed by a lengthy list of place names, consistent with the list known from Manuscript E, as well as a list recorded by Father Sebastián Englert. The list from Manuscript H contains a set of extended toponyms that completely coincides with other similar lists, including several extended place names that were not reported before. Due to the absence of photographic documentation for manuscript pages 11 and 12, the total number of place names in this list remains unknown. The lists of kings of Hiva and Rapa Nui from Manuscript H positively agrees with their counterparts recorded in Manuscripts E, A, and C. The final two pages of Manuscript H feature large drawings of the sooty tern (manutara). The number of facts connected with Manuscripts B, E, and H (availability of publication with Jaussen's list of rongorongo glyphs, Phillipi's tracings of the Santiago staff, as well as publications with photographs of Aruku Kurenga and the London tablets) suggest with a considerable probability that the writing of the manuscripts might have been started / catalyzed by the time of the visit of Walter Knoche in April of 1911.