Land Earth Sun Moon and the Father of Generations: An Historiography of Words for God in Sara Sikka

The common word in Sara Sikka for the Christian god is amapu , a compound noun meaning ‘father of generations’. There is reason to believe that the word came into use after conver-sions to Roman Catholicism began on Flores in the early sixteenth century but the origin of the word is obscure. An examination of writing about Sikka and its language by early Jesuits and later SVD missionaries on Flores reveals, if not the origin of the word amapu , much about its meaning in relation to the values and cosmology of pre-Christian Sikkanese society, culture and thought. The historiographic evidence also points to how elements of classical Sikkanese traditions have survived in the Christian culture of Sikka. . . . regardless of whether the entity named “God” exists outside his nature sheerly as key term in a system of terms, words “about him” must reveal their nature as words.

deity is called amapu in Sara Sikka, the origin of the word is less certain than I made it out to be. The idea that the word was invented by early missionaries may have come to me from a Sikkanese informant or from someone in the Church in Sikka. It may be what passes for common knowledge in Sikka, or I may have simply assumed that it does not predate the advent of Catholicism in Sikka. However that may be, the origin and history of the word, if it can be established, may well illuminate the history of Catholicism in Sikka and the Christianization of Sikkanese culture. One thing is certain: the idea closest to deity in Sikkanese thought is (or, for most Ata Sikka, was in the past) spoken in the words Nian Tana Lero Wulan, Land Earth Sun Moon. These words were not adopted by Sikkanese Christians for the God of their new faith.
The word amapu is a conjunction of two words. Ama is 'father' and, as a category of kin, includes father's brother(s) and father's father. Pu, a polysemous word rich in connotations in Sikkanese thought, are one's sister's children (man speaking) or brother's children (woman speaking). Mé pu (children and sisters' children [man speaking]/brothers' children [woman speaking]) are one's grandchildren and descendants. Pu plus the nominative -ng in the phrase meng pung means 'grandchildren, descendant(s)'. In Sikkanese folk etymology pu'ang, which means 'base, foundation, source', is related to the word pu. Pareira and Lewis defined amapu as 'Tuhan Allah. Terdiri dari dua kata Ama dan Pung: Bapa pencipta segala' [Lord God. Made up of two words, Ama and Pung: The Father, creator of all] (Pareira and Lewis 1998:4) but we said nothing about the origin of the word. I am still inclined to translate the word into English as 'source father' or 'father of generations'.
The word's meaning says nothing about its origin. A number of possibilities suggest themselves: 1. The word amapu was invented by Christians, either Portuguese priests or by Sikkanese Christians, before the Dutch acquired the island of Flores from the Portuguese in 1859; 2. The word was invented by Christians, either Dutch missionaries or by some Ata Sikka in their congregations, after the Dutch acquired the island of Flores from the Portuguese in 1859; 3. Either a Sikkanese Christian or European missionary translated the name of God from the language of another Christianized community in the eastern Lesser Sunda Islands into Sara Sikka (the language of Sikka); 4. Amapu was the direct translation of the phrase 'father of generations', which could have been in use among either Portuguese or Hollanders; 5. The word was used in Sikka in a context other than Christian belief and practice before the arrival of Europeans in the eastern Lesser Sunda Islands and was then adapted to denote the Christian God.
On the evidence before me, I cannot decide which of the numerous possibilities may be the most likely. I suspect that either possibility 1 or 2 is closest to the mark and that it is quite possible that the word amapu was in use very early in the history of Christianity in Sikka. However obscure the origin of the word amapu in Sikkanese speech may be, its advent in literary Sara Sikka can be dated with fair accuracy.

An Historiography of a Word
If my suspicion is correct, then some time between the appearance of the first Europeans in the Moluccas and the eastern Lesser Sunda Islands and 1873, someone combined two words of Sara Sikka, ama and pu, and gave the new word, amapu, a meaning: God. The name of the neologist is recorded neither in the documents left behind by the Europeans nor in the oral histories of the peoples of Sikka with which I am acquainted. Nor is the date of this bit of lexical bricolage recorded. Furthermore, the word may have been in use in some way and with some meaning by Ata Sikka before the arrival of Europeans. But the specifics of the word's origin we may never know.3

Cornelius Jakobus Omtzigt's Norang Me Me Sikka ko Mangerai
What can be known about this case of theological (in the sense of words for and about God) innovation is that it occurred in or some time before 1873.
3 A search of the main works by the early Sikkanese writers on the history and culture of Sikka, D.D.P. Kondi and A. Boer Pareira, reveals that Boer, writing after the 1920s, used the word amapu only once in his main notebook, in a caption accompanying his drawing of a necklace, part of the regalia of the raja of Sikka, in which Boer uses the phrases 'Tuhan Amapu Espiritu Santo' and 'Tuhan Allah Espiritu Santo' (Boer Pareira 2007:20). In these phrases, he implicitly equates Amapu and Allah. In a description of the death of Augustinu da Gama, Kondi, writing in the 1930s and 1940s, refers to 'Tuhan Deos-Amapu', the only use of the word amapu in his longest manuscript (Kondi 2007:54; see Lewis and Mandalangi 2008:116).
That year saw the first appearance of the language of Sikka in print in the form of a small (10.5 × 16 cm), seventy-one page booklet published in Roermond, the Netherlands. The book carried a title in Sara Sikka: NORANG ME ME SIKKA KO MANGERAI.
Many puzzles arise from Omtzigt's booklet, beginning with the booklet's title: Norang Me Me Sikka ko Mangerai. In Sara Sikka, norang means 'is, are; to have, to possess'. Norang is a nominalized reflex of the Sikkanese root nora, meaning 'with, to accompany; and; when (of time)'. Ko (Sara Sikka: or) in the title suggests that Sara Sikka was also known as Manggarai and so the title can be translated into English as 'To Accompany the Children of Sikka or Manggarai' or 'With [that is, for, on behalf of ] the Children of Sikka or Manggarai'. The work, however, is in Sara Sikka and Malay and not in the language of Manggarai, which is spoken far to the west on Flores. Indeed, between pages 18 and 19 of the copy in the library of the Koninklijk Instituut voor Taal-, Land-en Volkenkunde (KITLV) in Leiden, a note written in a modern hand has been inserted: Colfs visited Maumere in 1800, seven years after the publication of Omtzigt's booklet.5 While Colf's glossaries contain many errors, it is clear from them that he, as a traveller through the islands, knew that the lexicons of Manggarai and Sikka were quite different. If the catechism was written by someone less knowledgeable about Flores than Omtzigt, then we might account for the 'ko' in the booklet's title. Even so, anyone with sufficient knowledge of Sikka to compose a work in its language would surely have known that the language of Manggarai was not Sara Sikka.
Steenbrink refers to Omtzigt's 'catechism in "Manggarai", as the Sikka language was called in this period' (the 1870s), in which Omtzigt wrote (Steenbrink 2003:134). Of the confusion between the Sikkanese and Manggarai languages, it is possible to read too much into the word 'ko' in the title of this little catechism. In any case, even if Omtzigt did not know that the peoples of Sikka and Manggarai spoke quite different languages,6 itself an extremely unlikely gap in his knowledge of Flores, he would have known that, as differences go on Flores, the Sikkanese and Manggarai peoples were quite different. While it is possible that Omtzigt was observing a mistaken convention of the time, the presence of this little two-letter word in the title of the first publication in the Sikkanese language is one of many of the minor mysteries that emerge from the history of Flores in the first years of Dutch dominion over the island. 5 It is possible that Colfs saw Omtzigt's booklet in Maumere and may have used it in compiling his own wordlist. In his comments on Colfs's wordlists, Vorderman remarks that Colfs provides information on the language of Sikka not found in Omtzigt and that in Omtzigt's booklet, Sikkanese 'grammaticale constructie verwrongen naar het laagmaleische tijpe' [grammatical structure is reduced to that of low Malay] (Vorderman 1888:135).
6 Wurm and Hattori (1981, Map 40) classify Manggarai as a language of the Bima-Sumba Group of Austronesian Languages whereas Sara Sikka is of the Flores-Lembata (Lomblen) Subgroup. Figure 1. Omtzigt 1873, title page of KITLV copy.7 7 My thanks to Professor Karel Steenbrink, who took time from an undoubtedly busy work schedule to make and send me (2 November 2010) digital photographs of the KITLV's copy of Norang Me Me before I acquired a copy of the book and to Rini Hogenwoning and Nico van Rooijen of the KITLV library for this high resolution electronic scan of the title page of the Institute's copy.
In three lines at the bottom of the title page, which is customarily given to a publisher's imprint, we find the following: ROERMONDA  This note identifies Omtzigt as the author of the book and 1874 as the date of his death. Could this note be mistaken? Unlikely, but a possibility nonetheless. 10 Steenbrink gives us two brief sketches of Omtzigt's life.
Cornelius Omtzigt arrived in the Indies on 24 September 1867. After the required period of six months on Java, he came to Larantuka and went on a pastoral tour in 1869. At an age of more than 40 years, he was the first to concentrate on language studies. He made two longer trips in 1870 and 1871. He composed a simple book for the school which he 8 This would almost certainly have been Carl Ernst Arthur Wichmann (1851-1927), a German geologist who was professor of geology at the University of Utrecht . Wichmann was a member of a Dutch scientific expedition to Sulawesi, Flores, Timor and Roti in 1888-1889. 9 Confirmation of Omtzigt as the booklet's author is found in my copy of the work. It does not have the handwritten notes inserted between pages 18 and 19 of the KITLV copy. However, a very faint inked inscription on the cover can be made out in a hand different from that responsible for the note on the cover of the KITLV's copy: "Sc_ _ _ _ dit [?] _ ? Omtzigt S.J. Mi _ J [?]. in Larentoeka", thus also attributing authorship to Omtzigt. 10 On the inside of the front cover facing the title page of the KITLV, references to three works by L.F. Calon have been written in apparently the same hand as the notes on the title page. The three works are Calon 1890Calon , 1892Calon , and 1895 opened in Maumere. His catechism in 'Manggarai', as the Sikka language was called in this period, was published in 1873 . . . on 18 December 1873 Omtzigt was recognized by the colonial government as the first parish priest of Maumere. This capable man would become the first of a long series of people who died because of the unhealthy conditions in Maumere. He had to leave the mission in Central Flores soon after his nomination and went to a doctor in Surabaya, where he died on 4 July 1874 (Steenbrink 2003:134).
Omtzigt SJ, Cornelius J [akobus]. born in Nieuwer-Amstel on 6 January 1827 and arrived in the Indies on 24 September 1867. He stayed on Flores for several years, and was very successful in Maumere (1868-1873), but he fell ill and departed to Java. He died in Surabaya on 23 July 187[4] (Steenbrink 2003:477).
The printer or publisher of the booklet is given as 'J. J. Romen Typographa'. Above this line is the phrase: Roermonda datoe tana Ollanda (see Figure 1).
Roermond is a town and a diocese in the southeast of the Netherlands. Roermonda and Ollanda, or Olanda, are the names of the town and nation rendered in Sara Sikka. The phrase datoe tana, which could be mistaken for Malay, is a rendering of the Sikkanese detu tana. This is confirmed by Omtzigt's entry 'datoeng | rata' [datoeng | level, flat] (Omtzigt 1873:30), that is, the Sara Sikka word detung meaning 'level (rata)' or 'low, below [as a valley]); valley', which in turn suggests that Omtzigt's /a/ in datoeng is the sound /ə/ or /ʌ/ represented orthographically by e or ĕ in common modern spelling. Thus, Omtzigt's datoe is detu. The representation of detu as datoe occurs throughout Omtzigt's booklet, but not in the work of Calon, a later student of Sara Sikka who represents the word's vowel as e. While datoe does not occur in Calon's later glossaries of Sara Sikka, the word dettoe does, in various forms: Dettoe, ? Lerro dettoe, 12 uur's middags. Dettoen, effen, gelijk; vallei (Calon 1891:296 Calon, who wrote about Sara Sika 20 years after Omtzigt, must have been familiar with Omtzigt's booklet and was clearly somewhat puzzled by Omtzigt's rendering of the word; hence, the question marks. The phrase datoe (or detu) tana is grammatically a solecism,11 and would properly be tana detung (valley, level, or flat land), hence: 'Roermond of the flat (low) land of Holland'.
The point to be taken from this comparison of texts is the clear advance in the missionaries' understanding of the Sikkanese language from Omtzigt's to Calon's treatments of it.
Omtzigt's booklet is divided into two parts. The first part (pp. 1-21), carries the title (p. The book closes with a chart of Arabic numerals for arithmetic and a table of multiplications (pp. 70-1).
Throughout Norang Me Me, Omtzigt identifies the Christian God as 'Amapu (Deos)' and, whenever he uses the word Amapu, he follows it by Deos in parentheses. I believe the parenthetical mention of Deos, a word likely in widespread use in the nineteenth century among Christians in the eastern Lesser Sunda Islands, is significant.12 Omtzigt organized the glossary in his 1873 booklet in two columns, one of Sara Sikka words headed 'Mang.' (Manggarai, that is, Sara Sikka) and the second 'Mal.' [Malay] consisting of translations of Sara Sikka words. There are several Sikkanese phrases in which the word Amapoe [Amapu] appears. Among them are the following. Mang[garai]. Atabiang appang depo prenta Amapoe (deos).
As I have noted, each occurrence of amapu in Omtzigt's booklet is followed by the Portuguese Deos in parentheses. It appears that Omtzigt was not translating amapu as deos in his Norang Me Me for Sikkanese readers of his book. The first school was established in Maumere in the rajadom 12 Deos was the spelling of the word in Portuguese until the mid-to late nineteenth century. The [o] in Deos would be pronounced /w/. Modern spelling (in Portuguese, Deus) takes us closer to this pronunciation by using [u]. My thanks to Professor Alan N. Baxter, an authority on language contact involving Portuguese in Southeast Asia, for these notes on deos/deus in a personal communication via email, 2 November 2010.
13 I have here translated the Malay, which does not translate Omtzigt's Sara Sika well in all respects.
of Sikka in 1874, the year of Omtzigt's death and the book was published in 1873, when there would have been very few, if any, literate Ata Sikka. Omtzigt thus must have bracketed Deos after Amapu with the intention of translating amapu as deos for other priests and missionaries (all of whom were Europeans in Omtzigt's time). The implication is that Deos, the Portuguese word for God, was in common use among Europeans, including priests, at least in eastern Flores and likely throughout those areas of the eastern Lesser Sunda Islands influenced by the Portuguese. Because the book was written in Sikkanese, it would have been intended for priests and for teachers, local people who would staff the schools which would soon be opened in Maumere and elsewhere in the district and who would be trained as teachers of religion. And, perhaps, for Ata Sikka once they became literate.
It is worth recalling that Omtzigt arrived in Maumere in 1868 or 1869 (in Steenbrink's account, 2003:134 and in Wichmann's note), only ten years after the Treaty of Lisbon, the accord between Portugal and the Netherlands that transferred the eastern Lesser Sunda Islands-except for the eastern half of Timor-to the Netherlands. The Sikkanese would have known Deos as the Christian word for God from Portuguese times. The parenthetical translation of Amapu and Deos in Omtzigt's writing hints, if not implies, that he wished to establish a different and, possibly, a new word for God in the Church in the rajadom of Sikka and, further, that he wrote for a readership already acquainted with the word Deos. But whether Omtzigt himself invented the word or he adopted one already in use in Sikka cannot be determined, even if this speculation about his intention in writing his booklet is correct.
Although it is brief and rudimentary, Omtzigt's book is important and of extraordinary interest for a number of reasons. Firstly, Norang Me Me Sikka ko Mangerai is historically and linguistically noteworthy because it is, to my knowledge, the first appearance of the Sikkanese language in print. It also provides an insight into the Malay in use by Catholic missionaries on Flores in the 1870s, at least, when speaking with local people and children. Secondly, its errors help us to gauge the knowledge of the missionaries of the time of the languages and cultures of the peoples of Flores among whom they worked. Thirdly, close study and analysis of the subject, contents, structure, and linguistic form of the book would tell us something about what missionaries of Omtzigt's time were teaching the people of Sikka.

Lambertus Franciscus Calon's Works on the Language of Sikka
L.F. Calon SJ was born on 16 April 1852 in the small town of Biervliet in the Dutch province of Zeeland. Schooled at the gymnasium in Katwijk, he attended the University of Leiden, followed by theological studies at Maastricht. He spent a year in England and in 1887 was posted as a Jesuit missionary in the Netherlands Indies, where he was pastor of Maumere. Steenbrink (2003:145) notes that: In 1892 the mission of Central Flores was established in three permanent stations. There were three priests in Maumere, the gifted language researcher Calon as head of the parish, together with Luypen and Schweitz.14 Calon died on 14 September 1893. The editor of his last published work on Sikka (Calon 1895) noted the following: Almost simultaneously with the receipt of this fourth contribution to the study of Sikkanese the sad news came to us that on the 14th September 1893 in Makassar death relieved the indefatigable and skillful writer from his consuming job. 15 Calon published four works on the Sikkanese language: a 30-page glossary of Dutch words translated into Sara Sikka (Calon 1890); an 80-page glossary of Sara Sikka words translated into Dutch (Calon 1891); a sketch grammar of Sara Sikka with Sikkanese texts (Calon 1892); and his longest work on the Sikkanese language, Bijdrage tot de kennis van het dialekt van Sikka (A contribution to knowledge of the dialect of Sikka) (Calon 1895). The last includes a Sara Sikka to Dutch glossary of 1,404 entries and a number of texts in the Sikkanese language. In addition to the four works on Sara Sikka published under Calon's name, Steenbrink also notes a hymn-book published in 1893 'with texts in Sikka (by Calon), Latin, and also Portuguese, for the mission in Central Flores' (Steenbrink 2003:138).
14 The other two stations were at Koting in the central saddle of Sikka and in the village of Sikka on the south coast.
Calon's first glossary of Sara Sikka includes a clue, the merest of hints, that the Sikkanese word amapu may have had a pre-Christian meaning: he defines the Dutch beeld [English: image, statue, figure of speech] as amapoe (Calon 1890:503).16 This definition is interesting because the peoples of Sikka seldom, if ever, represented iconically the metaphysical entities of their early religion, neither graphically nor in sculpture. In his 1891 Sikkanese to Dutch wordlist, Calon translates 'Amapoe' as 'God' and adds a meaning for another word: 'ámapoe, alle soort van beeld of prentje. ima ámapoe, s[oort] schelp (conus)', [ámapu, any kind of image or picture, engraving. ima ámapoe, a kind of sea shell (conical)]. To my knowledge, there are not two forms of the word in contemporary Sara Sikka.17 Calon may well have been correct to distinguish the two words. That said, in contemporary Sikkanese speech ama (father), an element in amapu, is a word different from 'ama [ʔama], which means skin, husk, or shell. In his first wordlist (1890), ima18 is 'schelp' [sea-shell] and in 1895 he gives ima ámapoe the same meaning. It is significant that ámapoe as shell drops out of his 1895 glossary. I have never heard skin or shells referred to as 'amapu. Amapu (God) begins with a soft a, whereas 'amapu (shell) begins with a prevocalic glottal stop.
I am inclined to think that the form of amapoe as shell was an error, but that amapoe as image, that is, as a figure of speech, was quite possibly correct. In neither wordlist does Calon indicate how the word amapu was used, either to mean image or as a figure of speech. If it was used as a figure of speech, the word may have been used in pre-Christian times, but not likely as a trope for the old Sikkanese deity, as will become clear, because the Sikkanese deity was represented as a complementary combination of feminine and masculine elements whereas Amapu, 'Father of generations', is distinctly and exclusively masculine. 16 The 1890 glossary was Dutch to Sara Sikka. Later in the same work Calon translates the Dutch word God as Amapóe, with an upper case initial letter and an acute over the o [ó].
17 My knowledge of Sara Sikka is not entirely comprehensive. The varieties I know best are those of Sikka Natar, on the south coast of Kabupaten Sikka, and the dialect of the language spoken in Tana Wai Brama in the Tana 'Ai region of the Kabupaten. Language is as variable from one domain of Sikka to another as are culture and social organization.
A further comment on Calon's definition of amapu as image: Father Frans Meyer is well known among Churchmen in Sikka as the author of a massive, but unpublished, Sikkanese-Dutch dictionary. The work was typed at least twice. One typescript is dated 'Ruteng [Manggarai, West Flores], 23 Djanuari 1964' and is identified as 'Tjetakan kedua' (second printing, typing or edition) by the unidentified copyist of its 'Voorwoord', which otherwise reads the same as an earlier version of the work. In my copy of the earlier typescript, the 'Voorwoord' is signed only as 'De Uitgever', but someone has pencilled a note: '(P. Geldens)? Welk jaar? 1937-'38'. Against the entry for 'Amapoe in my copy, someone has handwritten the following note: ' 'Amapoe hala: afgod. {eigenlijk bestaat er geen woord voor afgod of afgodsbeeld, omdat de Sikaneze geen beelds vereerden, geen afgodsbeelden hadden}', ['Amapu hala: idol. {Amapoe is an error; actually there is no word for idol or idol's image, because the Sikkanese worshipped no idols and had no idol images]. The observation of this unknown writer agrees completely with what I know of the people of Sikka.
A final word about Calon's work is merited. While amapu appears as a keyword in Calon's glossaries of Sara Sikka, it is otherwise scarce in his published work. References to a few Christian themes appear among the 302 sora (Sara Sikka: pantun, verses) and some work chants in Calon's Bijdrage tot de kennis van het dialekt van Sikka (Calon 1895:69, no. 176 As all different seed wherever they may be.19 Among these 1895 texts, the word amapu appears not once.

Paul Arndt's Works on Sikka
Fifty-eight years after Omtzigt's Norang Me Me and 36 years after the last of Calon's publications on Sara Sikka, the priest and ethnologist, Paul Arndt SVD,20 included the word amapu in his grammar of Sara Sikka (Arndt 1931) and in his larger work, Mythologie, Religion und Magie im Sikagebiet (Arndt 1932). Arndt begins his Grammatik der Sikasprache with an acknowledgment of Calon's work in Sikka: 'Hier hatte bereits Calon, S.J. in den achtziger und neunziger Jahren mit der Erforschung der Sprache begonnen und einige Wörterverzeichnisse und eine kleine Grammatik geschreiben, die in der vorliegenden Arbeit benutzt wurden' (Arndt 1931:2)  Regarding the c of Camapu, Arndt explains that the c is a 'stimmhaftes h' (Arndt 1931:4), a 'voiced h'. What Arndt here describes are words that begin without the glottal stop with which many Sara Sikka words begin. This reports a mishearing of Sikkanese speech: Sikkanese words beginning with a vowel are either pre-vocalically glottalized, as are all words of Bahasa Indonesia beginning with vowels, or are 'soft'. Because it is phonemic, the presence of a glottal stop determines the meaning of a word as different from the word without the initial glottal. Thus, in the orthography Mandalangi and I used in our Sara Sikka dictionary, ama (father) has no initial stop, whereas 'ama (skin) begins with an initial glottal stop, /ʔ/ (Pareira and Lewis 1998:4). Arndt marks the lack of an initial prevocalic glottal, as in ama (father), with the c. I have elsewhere commented on marking the lack of a phonemic sound orthographically (Lewis 2010:41-2). Hence, in Arndt's transcription, Cama is ama, ama is 'ama, and Camapu is Amapu (without a prevocalic glottal stop).
Numerous references to Amapu and Déot occur in Arndt's Mythologie, Religion und Magie im Sikagebiet, but not as equivalents, as in Omtzigt's Norang Me Me Sikka. Rather, Arndt provides texts in Sara Sikka in which the two words appear to represent distinct entities. The first is an origin myth, which begins on the first page of the book. He provides interlineal translations into German, as in the following excerpts from his book in which Amapu and Déot are mentioned (Arndt 1932:5 An immediate question that arises about this and other mythic texts in Arndt's book concerns the origin of various words in this particular text and elsewhere. Nipa and ri'a, for example, are not Sikkanese words, but are from the language of Lio, the people of the region to the west of Sikka, nipa meaning snake and ri'a21 meaning strong, mighty, large, majestic, exalted. Déot (Dé'ot, in Arndt's orthography) is Portuguese, and 'Amapoe, while Sikkanese, is the term here under consideration.
Second, the myth is, as are others in Arndt's book, in the form of a prose narrative. While it exhibits semantic concordances, Arndt did not set the texts as parallel lines in couplet form, and his transliterations thereby do not reflect the poetic form of the ritual language in which the Ata Sikka narrate almost all Sikkanese myths. Ritual language is a poetic genre of Sara Sikka in which the lines that make up couplets and quatrains are structured with a rigorous parallelism between word meanings. Transcribing ritual language texts in parallel form was already the practice of Jonker in his writing on Rote22 and the early Sikkanese writers, D.D.P. Kondi and A. Boer Pareira (Lewis and Mandalangi 2008). Even Arndt himself in other work on Flores sets out texts of ritual language in parallel couplet form.23 This is a puzzle. It is possible that Arndt's informant told him the myth as 21 Arndt uses the [ ' ] to indicate that two vowels are not a diphthong but that the second is sounded after a stop (Arndt 1933:6 Arndt (1954) and his texts from the Solorese archipelago (Arndt 1951). a prose narrative rather than as a recitation of ritual speech. If so, this and references to foreign ideas such as references to Adang and Ewang (Adam and Eve) in Arndt's texts may mean that he was dealing with already acculturated informants and syncretized narratives.
The difficulty is one that arises in all of Arndt's work, as valuable as it is. Throughout his work on Sikka, Arndt fails to provide us with the provenance of the material he presents. Specifically, he seldom tells us where, when, under what circumstances, and from whom he acquired the information he records. As modern ethnographers of Sikka and Flores generally have learned, there is considerable variation in the details of social organization, ritual practice, speech, and myth within socio-cultural groups that can be mistaken as sharing in all respects a single culture and language. Such variation is found among the regions and dialects of Sikka, among the numerous old Sikkanese tana (ceremonial domains), village clusters, among individual villages, and even among the people of different clans and descent groups that comprise villages. Although it may be difficult to determine which is more authentic, the texts of myths derived from interviews with ritual specialists can differ substantially from myths as they are spoken in ritual performances. Without knowledge of the specific provenance of reported customs and texts, close analysis of ethnological material can be difficult. Thus, for example, the occurrence of non-Sikkanese words such as nipa and ri'a in texts which are represented as Sikkanese would be less of a puzzle if we knew where and from whom the texts originated.
The specifics of provenance become acutely important with the passage of decades. A great deal of Arndt's material differs significantly from that I have recorded in Sikka and Tana Wai Brama since 1977 and the results of David J. Butterworth's research in Krowé since 2005 (see Butterworth 2008). It is difficult to know whether the differences of Arndt's material and ours is the result of culture change (including the loss of myths and certain ritual practices in the past 80 years) or because he acquired his information in villages not covered by more recent ethnographic research. However that may be, Arndt's material must be cited with circumspection.
With regard to the first mythic narrative in Mythologie, Religion, und Magie im Sikagebiet, the suspicion immediately arises that Arndt has given us a text that synthesizes three traditions, the Sikkanese, Lionese, and Christian. Since Arndt tells us nothing of the origin of the text, it is difficult to know.
25 Orinbao (Father Piet Petu SVD) wrote a whole book about the concept of the island of Flores as 'Nusa Nipa', the Isle of Snakes or Snake Island, which he identified as 'Nama pribumi nusa Flores' [the indigenous name of the island of Flores] (Orinbao 1969). 5. Déot Amapu: God (as feminine) and God (as masculine), a phrase of paired terms such as those of Sikkanese ritual language 6. Dé'ot Nipa Ri'a Tana nora 'Amapu Maro Wulang: Déot as the great snake of the earth, the feminine realm, joined with the complementary Father of Generations of the firmament and moon.
Much can be made of these excerpts from Arndt's Sikkanese texts, but I will identify one point in particular, which will be taken up later. The phrase Déot Amapu represents God as a classificatory duality of concepts, one masculine and one feminine. This dualism can be seen in all of Arndt's citations of Sikkanese myth. The fourth is interesting because it reverses the analogy: Déot : feminine :: Amapu : masculine I shall return to this theme. Arndt himself was interested in the semantic associations of the two terms for God, Déot and Amapu. In the early part of his book, he makes the following remark: 'Amapu und Dé'ot sind Gottesnamen, . . . die Du'a Dé'ot scheint mir eine Personifikation des Mondes zu sein; denn nur den Mond habe ich als duä, Frau, bezeichnet gefunden; . . . [Amapu and De'ot are God's names, . . . Du'a De'ot seems to me to be a personification of the moon; because I have only ever found the moon referred to as du'a, woman; . . . . ] (Arndt 1932:26). Here, Arndt equates Déot and Amapu as genderized synonyms, two words, one whose referent is masculine and the other feminine, for a single entity, the moon. I believe that Arndt here made an error. I would argue that the two words are semantic complements and that, when spoken in parallel, they denote God as a dualistic entity whereby one element of His duality is masculine and one is feminine and that, together, they signify the whole of the universe.26 This dually gendered God would have been fully in keeping with the cosmology of the old religion of Sikka.
The first man arose upward from the navel (the center) of the earth. The people who came out first and preceded the others, were Déot (God) Nipa Ria Tana and Amapu (God) Maro Wulang. All things have come from the navel of the earth. After De'ot Nipa Ria Tana and Amapu Lero Wulang emerged, the whites (Europeans) also came forth; they were white as the moon.
A confounding of Sikkanese myth with Christian myth rendered in the Sikkanese language and in a Sikkanese cultural frame occurs a number of times in Mythologie, Religion und Magie im Sikagebiet. One example is in the section titled 'II. Humankind A. The first people 1. Creation myths'. Here we find the following curious narrative, not as a translation of a Sikkanese text into German, but as Arndt's own account, as if he were paraphrasing something he had heard from Ata Sikka (Arndt 1932(Arndt :1970 wished it so. The first humans came out from the earth and were stronger than today's people. Thus Lero Wulang thought, there should be people, and they should have power over everything in the world. After Lero Wulang had thought about it, he said: People should come forth! The first human was a man. Then Lero Wulang thought: If there is no woman, the people cannot increase in number. Therefore let a woman also emerge from the earth. The first pair of humans was Adam and Eve. This is clearly the Biblical story of the creation of Adam and Eve, with two creators identified as Niang Tana and Lero Wulang rather than as God. As the story unfolds, it is, however, the masculine of the dual deity that creates the first people.27 It cannot possibly be an indigenous Sikkanese myth, but can only be a recasting of the origin story in the book of Genesis in the categories of Sikkanese thought. Once again, the missing provenance is crucial for understanding the tale. It is unlikely that Arndt himself would have composed a rendering of the Genesis story in Sara Sikka and then inserted it into a longer text that he represents as Sikkanese. More likely, it was taken down, perhaps verbatim, as a transcript of a tale he heard from an Ata Sikka. Without notes and commentary on the story's origin, Arndt's text, we may regret with frustration, is more perplexing than illuminating. Arndt himself was puzzled by the terms amapu and déos (déot) in his texts. He devotes two full pages to the question, but draws no firm conclusions about their origin and meaning. On the terms amapu and déos, Arndt (1932:187) writes: Noch dunkler wird die Sache mit Lero Wulang durch die beiden namen 'Amapu und Dé'ot (Dé'os); sie kommen oft als Namen für das (oder die) höchsten Mesen28 vor; sie werden bald für sich allein, bald verbunden (Dé'ot-'Amapu), bald wieder in Verbindung mit anderen Namen für gottliche Wesen gebraucht. So spricht Subu aus Nita in seiner Schöpfungsmythe und auch sonst von 'Amapu Maro Wulang, 'Amapu Duni'a, Dé'ot Nipa Ri'a Tana; sonst ist bereits vorgekommen 'ai Dé'os Lero Wulang, 'Amapu Ni'ang Tana, auch Dé'ot 'Amapu Lero Wulang; réta Wulang Dé'ot deri, oben am Himmel wohnt Dé'ot; 'Amapu deri réta wulang; 'Amapu wohnt oben im Himmel.
Etymologically, amapu is composed of 'ama, father, and pu, descended [to descend, be descended from], origin, ancestor. Thus pu occurs in the phrase du'a pu Mo'a; du'a = woman, lady; mo'a = man, old man; pu = grandfather; pu mo'a also = grandfather; du'a pu mo'a = male and female ancestors; mé pung, offspring; . . . Arndt then cites a prayer of offering from Ojang, near the border between Kabupaten Sikka and Kabupaten Flores Timur, which mentions pu rua (Sara Sikka, two or double origins (Arndt 1932:187): Ama a'un pu a'un, ama a'un pu rua, my father is my ancestor (or source), my father of two origins29 . . . He makes the following remark about the reference to two sources, that is, double ancestry (Arndt 1932:187- Pu rua, double origins, probably refers to the two principles, Lero Wulang as father and Niang Tana as mother. The real meaning of amapu seems not to be much appreciated anymore, since one uses the expression, dunia, the earth, which is still perceived as a female; Amapu Dunia. If Amapu Wulang is also Amapu Du'a, then Amapu Wulang (the moon) is the female (feminine) Amapu. . . .
In the region of Sika the Christians [have called] sacred figures and images amapu and the pagans have followed their example. As God's name they used Déos.. . .
There is little doubt that the word Déos comes from the Portuguese. As God's name it is now known throughout the region of Sikka. But it is strange that instead of just Déos (Déot) the native word Amapu passed into the language of the Church to denote God, the Supreme Being, while the Church's language is still rich in other Portuguese expressions. Old people in Sikka still use Déos.
From this passage in Arndt (1932), three further clues about the connotations of amapu, if not about the word's origin, can be extracted. First, at the time he wrote, it appears that the Sikkanese conceived of God as possessed of a dualistic nature as father and mother. Arndt clearly found some difficulty with the logic of the dual attributes of God. Second, he confirms Calon's earlier report that amapu referred to sacred figures and images, but that the word Déos was reserved for the Christian God. Third, he notes that, while old people in Sikka still used the word Déos for God, the word amapu had passed into the language of the Church in Sikka, which he found puzzling. An implication of the third clue is that Arndt thought of amapu as a 'native word [that] passed into the language of the Church', but he does not say clearly that he thought of the word as predating the arrival of Christianity in Sikka.
Some of this confusion can be clarified by considering the nature of the dualistic representation Nian Tana Lero Wulan in the old religion, ceremonial system and ritual language of Sikka.

Nian Tana Lero Wulan, Land Earth Sun Moon
A myth of origin from the domain of Wai Brama tells how land emerged from the world's waters through an act of creation, that is, how an undifferentiated world became differentiated into the sea and the land and how, subsequently, the heavens were separated from the face of the earth (see Lewis 1982 and1988:45-50).
Each of these separations created dual but complementary divisions in the world. By complementary I mean that, rather than standing in opposition to the other, each division completes or makes perfect the other and that the two divisions comprise a whole, in this case, the whole that is the universe. The most important of the primordial divisions was that between the firmament, the realm of the sun and moon, and the land and earth, the realm of human life. The land and earth, perhaps because of their fecundity and human beings' dependence upon them, were further classified as feminine, while the sun and moon in the heavens were classified as masculine.
The ruling principle of the universe came to be that things come in pairs that represent classificatory but complementary differences. Thus human beings are male and female, fathers and mothers; brought together, they produce children who are sons and daughters, sisters and brothers. The plane of human existence is Land and Earth, a duality which itself encodes gender. The Earth is feminine but incorporates Land, a masculine element, just as Sun and Moon are masculine, but Moon incorporates a feminine element in an otherwise masculine domain and category. Thus is the whole of the universe, and thought about it, structured as complementary pairs. A significant feature of the thought and cosmology of the people of Wai Brama is that the separations that brought about the divisions of the world were the work of an entity whose name is ineffable, and who is thus unknowable to humanity. An agent was at work (the ritualists of Wai Brama would agree), but who or what it was is not, and cannot be, known. The Tana 'Ai myth of the creation is spoken in the largely passive voice of ritual language and as narratives rich in events and things done but poor in causes and agents.
A fundamental and striking implication lurks in the cosmology of the people of Wai Brama. If the creator (think of him or her or it as the Great Separator) is without a name and can only be known through its attributes, then it cannot be addressed by name and thus cannot be known, at least not directly. If it cannot be known, then it can neither be a deified person nor personified as a deity. Nian Tana Lero Wulan, Land Earth Sun Moon, were creations; they cannot together be a deity, for if that were so, then the deity created itself-perhaps not impossible for a god, but a logical scandal nonetheless. This idea makes as little sense to Wai Brama ritualists as the idea that God created himself would have made to the early Catholic missionaries in Sikka. But for the people of Wai Brama, it follows that no one can have a personal relationship with or to a deity, that is, to God. The religion of the peoples of Tana 'Ai, is-and that of pre-Christian Sikka, wasrich in myth and ritual, but modest in theology and profoundly agnostic.
Nian tana lero wulan (to be clear) are not spirits, but manifestations, not of a deity (as I represented them in my early writing), which cannot be known, but of creation itself, the categories that subsume everything in the world and which are, thus, the world itself. Each of the named elements is a metaphor, a synecdoche, to be precise, of the whole of the universe. Synecdoche is a figure by which a more comprehensive term is used for one less comprehensive or a less comprehensive term used for one more comprehensive term, such as whole for part or part for whole, genus for species or species for genus. Thus, in Wai Brama, to say 'nian tana' in ritual language is immediately to evoke 'lero wulan', which must be spoken for a couplet of ritual language to be complete. In Tana 'Ai and what I think of as classical, that is, pre-Christian, Sikkanese thought, no more, no agent outside of the world, no super-worldly agent beyond the elements of creation is required.

The Gender of God in Sikka
If in the classical Sikkanese language and world view all things come in pairs, including the manifestations of an unspeakable deity, the God of the new religion was radically different, for He, as a father, was male and was represented metaphorically as masculine. His name did not immediately evoke a feminine counterpart. In terms of the cosmology and ways of thought of the people of Sikka, a God who was Father required a feminine element to be complete. Whatever the origin of the word Amapu, the Father of Generations, once it came into use, a second word, Déos, a gift of the Portuguese, was already at hand for the missing feminine element of the deity. It is perhaps for this reason that a feature of talk about God that Arndt noted was current as Sikka became Christian and before it became fully Catholic. That was the pairing of Amapu and Déos as Déos Amapu, a usage that today has all but disappeared in Sikka and which I have never heard in Tana 'Ai. As Arndt understood, Déos represented a feminine God or the feminine half of God whereas amapu was God's masculine side. The two terms combined fulfilled the requirement of traditional cosmology that all things, including God Himself (and Herself) should consist of paired elements. Moreover, the words might be spoken with a syntagmatic order, Déos Amapu, that preserved the feminine before masculine order of the feminine ina nian tana before the masculine ama lero wulan in the older cosmology. At the same time, more acculturated Sikkanese could speak of Amapu Déos, the word order thus reflecting the primacy of the masculine in Christian thought. It is thus quite possible that in the decades during which Sikka became fully Catholic and thought and society altered to reflect the conversion, this echo of the classifications of the classical Sikkanese world view lingered.
However that may be, the institution, theology, and the liturgy of the Church provided a place and a means of expression of the feminine, as it did for otherwise widely disparate Catholic peoples of the world. The veneration of the Virgin Mary as the mother of Jesus Christ is almost universal in Sikka, where gabungan kontas, neighborhood rosary groups, are a feature of social and religious life in villages throughout the Kabupaten. I cannot here undertake a detailed survey and analysis of the Marian cult in Sikka, but I can note that participants in the gabungan kontas30 associations in Sikka are almost exclusively women and children of the villages.
The anthropologist Edmund Leach once made a distinction that Sikka's gabungan kontas exemplify well. Leach (1968:1) distinguished philosophical or theological religion from practical religion, noting that: the gap between the theology of the higher philosophers . . . and the religious principles which guide the behavior of the ordinary church-goer may be very wide indeed.
Roman Catholic theology originates in the upper levels of the hierarchical Church and in the writings of theologians and scholars. In contrast, ritual devotion to the Blessed Virgin, in Sikka and elsewhere, while sanctioned and often encouraged by the Church, is religious practice on the ground, where Catholic villagers developed and carry on its traditions. If the Christian trinity-the Father, the Son, and the Holy Ghost-is represented in language as masculine, the elements of the classical Sikkanese cosmos were represented by gendered, complementary, and mutually evocative elements whereby the cosmos was represented in language and conceived as divisible into two great gendered, classificatory realms. One is then tempted to explain the prominence of devotion to the Virgin Mary in Sikkanese Catholic practice as a preservation of the complementarity of the feminine and masculine in the new religion. One might note also the pairing of Adang and Ewang (Adam and Eve), a masculine and feminine pair, in Sikkanese ritual language whereas the Christian deity-Tuhang Amapu or Tuhan Amapu Déos-is, in modern times, conspicuous by His absence from the form of representational expression that most reveals Sikkanese classification, cosmology and culture.
Thus, I reach the somewhat disappointing conclusion that the origin of the word amapu in Sara Sikka cannot be determined with the evidence at hand. It may have been, and likely was, an invention, but when and by whom remain unknown. Whether the innovation occurred in Portuguese times, which would mean the term could date to as early as the sixteenth century, or the innovator was one of the later missionaries under the Dutch regime after 1859-perhaps Father Omtzigt himself-cannot be determined. Nor can the possibility that the word arose among the people of Sikka themselves be discounted. Some speculations can be made, however.
First, despite their lack of provenance, Arndt's 1932 texts of Sikkanese origin myths can be taken as suggesting that the term was in use long enough to have been incorporated into the mythology of at least one region of Sikka. The occurrence of the paired terms Amapu // Du'a Déot in the Arndt texts from 1932 suggests that, in incorporating the foreign deity into Sikkanese thought, at least one Sikkanese ritual poet paired two terms for a single deity and, at the same time, by nominating Déot as Du'a (the honorific for a woman in Sara Sikka) and amapu as 'father', preserved the gender duality of the indigenous cosmology and ritual system. This usage was not heard in the 1970s and thereafter in my own research in two regions of Sikka. The gendered deity of Arndt's texts parallels that of the earliest literary records of the Sikkanese language of Omtzigt's Norang Me Me Sikka ko Mangerai. That said, only once do we find the phrase 'Amapu-Deos' in the voluminous works of the Sikkanese writers Boer and Kondi, who were contemporaries of Father Arndt. Here we must tread with care because a lack of evidence cannot be taken as evidence of a lack.
With the adoption of Bahasa Indonesia as the more common language of the mass and the Church, God has become Tuhan Allah. The word amapu is less often heard in contemporary Sikka, and déos more seldom still.31 Even so, in 1993 a ceremony was held in Maumere to reconcile two factions of the Da Silva family, the family of the old rajas of Sikka. The presiding officiant, a retired school teacher famous for his knowledge of Sikkanese culture, intoned an invitation to the gathered families and participants: 'Mari kita bersama serukan Lovado Bénsang de Deos' (Let us together recite the Praise for the Blessing of God) (Lewis 2006:324).
Second, the prominence of the Marian cult throughout the Catholic communities of the Regency of Sikka adds weight to the inference that the peoples of Sikka were inclined toward viewing the Catholic God as a juxtaposition of the masculine and feminine elements of the godhead as a single deity, as their manifestations were represented in the indigenous 31 It is, however, worth noting that the fourth printing of Lalang seu: Surat ngadji Sara Sikka (1964) used Amapu as the word for God in 1964, as did the presumably somewhat more recent Cantate (Lalang seu) (n.d.). religion, and despite the Catholic representation of God in the three persons of the Trinity.
Whether or not the word amapu predates its first appearance in writing is undecidable and the word's provenance is unclear. Thus, while we do not have the word's history, we do have its historiography. The question then becomes, was amapu in use before 1873 and, if so, how long before? Before the arrival of the Portuguese, their priests, and their God? In other words, was amapu the deity of the Ata Sikka before they became Christians?
Consideration of these old names for the manifestations of the deity in Sikka, like the consideration of God Himself, leads to mysteries. They are, however, mysteries that lead to history-of a sort: not the history of the word amapu (much less the history of the culture that produced it), but its publication in writing is historically factual. We are left with something akin to Burke's remark: whether or not amapu existed before the word was written down, the facts of its having been written and what early writers wrote about it reveal their nature as historiography.
What I suggest here is not surprising: changes in religious practice (that is, in the liturgy of religious ritual), belief, and spiritual commitment are one thing, but changes in a people's fundamental conceptions of the cosmos and humankind's place in it, are something more profound. The old cosmos of the distant ancestors of the Ata Sikka lingered in the new religion and abides among their descendants today.