Research in Anakalang, West Sumba (1986–87) and in The Netherlands (1988) was funded by the U.S. Department of Education Fulbright-Hays Doctoral Dissertation Research Program, the Joint Committee on Southeast Asia of the American Council of Learned Societies, the Social Science Research Council, and the Wenner-Gren Foundation for Anthropological Research. Additional fieldwork in Anakalang (1993) was supported by the Southeast Asia Council of the Association for Asian Studies. I am grateful to the Lembaga Ilmu Pengetahuan Indonesia and Universitas Nusa Cendana for their sponsorship. Earlier versions of this article were presented at Northern Illinois University (1989), the University of Wisconsin (1989), and at the meetings of the Association for Asian Studies (1990). I have benefited from the comments of Nina Kammerer, Marshall Sahlins, Rafael Sanchez and Greg Schrempp. This research depended on the help of many Anakalangese. The contributions of Umbu Padda Buli Yora and Umbu Siwa Djurumana have been of particular importance for this article.

¹ For some other dimensions of religious change in Anakalang, see Keane, W., “The Spoken House: Text, Act, and Object in Eastern Indonesia”, American Ethnologist 22,1 (1995): 102–124CrossRefGoogle Scholar, and “Materialism, Missionaries, and Modern Subjects in Colonial Indonesia”, Conversion to Modernities: The Globalization of Christianity, ed. van der Veer, P. (New York: Routledge, forthcoming in 1995).Google Scholar

² See Tapp, N., “The Impact of Christianity upon Marginalized Ethnic Minorities: The Case of the Hmong”, Journal of Southeast Asian Studies 20,2 (1989): 70–95CrossRefGoogle Scholar; Kipp, R.S., Dissociated Identities: Ethnicity, Religion, and Class in an Indonesian Society (Ann Arbor: University of Michigan, 1993)CrossRefGoogle Scholar; the essays in Indonesian Religions in Transition, ed. Kipp, R.S. and Rodgers, S. (Tucson: University of Arizona Press, 1987)Google Scholar; Conversion to Christianity: Historical and Anthropological Perspectives on a Great Transformation, ed. Hefner, R.W. (Berkeley: University of California Press, 1993)CrossRefGoogle Scholar; and Asian Visions of Authority: Religion and the Modern States of East and Southeast Asia, ed. Keyes, C.F., Kendall, L. and Hardacre, H. (Honolulu: University of Hawaii Press, 1994).Google Scholar For a discussion of boundaries and identity involving Hinduism and Islam in Java, see Hefner, R.W., Hindu Javanese: Tengger Tradition and Islam (Princeton: Princeton University Press, 1985)Google Scholar. I address one aspect of Sumbanese “local identity” in an unpublished paper entitled “Knowing One's Place: National Language and the Idea of the Local in Eastern Indonesia”.

³ Thus, while traditions may be reconceptualized with varying degrees of intention and at various levels of society, as argued in Hobsbawm, E., “Introduction: Inventing Traditions”, The Invention of Tradition, ed. Hobsbawm, E. and Ranger, T. (Cambridge: Cambridge University Press, 1983)Google Scholar, it would be a mistake to assume in advance that such reconceptualizations only begin with colonial or other encounters with the West.

⁴ All quotations are in the local language, Anakalangese, unless otherwise indicated as Indonesian or Dutch. For purposes of simplicity, I follow the orthographic conventions used for the Kambera dialect of east Sumba; see Onvlee, L., Kamberaas (Oos-Soembaas)-Nederlands Woordenboek (Dordrecht: Foris, 1984).Google Scholar In Anakalangese pronunciation, the consonants not preceded by m or n are implosives.

⁵ For expressions similar to these, see Onvlee, L.'s early discussion of changes in Sumba, “Tana Mema — Tana Dawa” in Cultuur als Antwoord ('s-Gravenhage: Martinus Nijhoff, 1973), pp. 114–32.CrossRefGoogle Scholar

⁶ Thet the concept of “tradition” perforce involves such juxtapositions is nicely suggested by the fact that Anakalangese use the Indonesian term tradisi — a term obviously of western origins — to name the concept. The indigenous term most frequently translated as tradisi — or as adat (Indo. “customary law”) — is pata, a term whose most important connotations involve formality of action. Religious change is of course only one of the contexts in which epochs can be juxtaposed, and tradition conceptualized: see Anderson, B. O'G. A., “Time of Darkness and a Time of Light: Transposition in Early Indonesian Nationalist Thought”, Perceptions of the Past in Southeast Asia, ed. Reid, A. and Marr, D. (Singapore: Heinemann, 1972)Google Scholar; Volkman, T.A., “Visions and Revisions: Toraja Culture and the Tourist Gaze”, American Ethnologist 17,1 (1990): 91–110CrossRefGoogle Scholar; Steedly, M.M., Hanging Without a Rope: Narrative Experience in Colonial and Postcolonial Karoland (Princeton: Princeton University Press, 1993)Google Scholar; and Pemberton, J., On the Subject of “Java” (Ithaca: Cornell University Press, 1994).Google Scholar For striking parallels to the situation I describe here, see White, G.M., Identity Through History: Living Stories in a Solomon Islands Society (Cambridge: Cambridge University Press, 1991).CrossRefGoogle Scholar

⁷ For further development of this argument in relation to uses of and assumptions about language, see Keane, , “The Spoken House”Google Scholar.

⁸ Hoskins, J.A., “Entering the Bitter House: Spirit Worship and Conversion in West Sumba”, Indonesian Religions in TransitionGoogle Scholar. The colloquial term kafir is commonly used by both Christians and non-Christians, often without, it appears, the usual pejorative connotations.

⁹ Sumba Barat dalam Angka 1986 (Waikabubak: Kantor Statistik, Kabupaten Sumba Barat, 1987).Google Scholar Only five religions are legally recognized in Indonesia — on the census tables, these five are followed by a column for “other” (Indo. lain). Official statistics should, of course, be used with some caution. In addition to the general difficulties of census-taking in Sumba, one cannot draw inferences directly from the category in which people are listed. Christian sympathizers who have not been baptized may be listed as “other”, and there are active ritualists who are officially Christian.

¹⁰ Marapu refers to the spirits of the earliest ancestors. The term has come into general use in Sumba to refer to the entire ritual system and its practitioners. I use the term “follower” with some hesitation — ancestor worshippers do not follow spirits as one follows a prophet, but they do speak of the performing rites as “following” (keri, padutungu) the “tracks” (wewi) left by the ancestors. These figures are from Kantor Statistik, Sumba Barat.

¹¹ Compare for example Abdullah, Taufik, “Modernization in the Minangkabau World: West Sumatra in the Early Decades of the Twentieth Century”, in Culture and Politics in Indonesia, ed. Holt, C. (Ithaca: Cornell University Press, 1972)Google Scholar, and Siregar, S. Rodgers, Adat, Islam, and Christianity in a Batak Homeland (Athens: Ohio University Center for International Studies, 1981).Google Scholar

¹² In fact people do occasionally return to the marapu, especially in response to serious illness or other crises. A person who changes too often, however, especially among different Christian churches, is scorned as feckless and is likely to be excluded from positions of responsibility.

¹³ I know of marapu priests who have let some of their children go to Catholic schools, which probably, though not necessarily, will lead to baptism, while retaining a few sons to carry on the ritual obligations.

¹⁴ Westerners are termed “white foreigners” (jawa kaka). Long pants (in opposition to Sumbanese cloth) are “foreign clothing” (kalabi jawa), the road is “foreign rice-field dyke” (kamotu jawa), government office is distinguished from traditional duties as “foreign duty” (dihi jawa), recently introduced ceremonies, such as the holding of wedding parties, “foreign custom” (na pata jawa). It appears that the actual etymology of the term jawa is from a proto-Austronesian form meaning “far (away)” (see Dahl, O.C., Proto-Austronesian, Scandinavian Institute of Asian Studies Monograph Series No. 15, Lund: Student-litterature, 1973, p. 47).Google Scholar

¹⁵ Keane, W., “Delegated Voice: Ritual Speech, Risk, and the Making of Marriage Alliances in Anakalang”, American Ethnologist 18,3 (1991): 311–30CrossRefGoogle Scholar, and “The Value of Words and the Meaning of Things in Eastern Indonesian Exchange”, Man (ns) 29,3 (1994): 605–629.CrossRefGoogle Scholar

¹⁶ The association of externally-derived money and anti-socialness or danger is a common theme throughout the margins of the cash economy; see Taussig, M.T., The Devil and Commodity Fetishism in South America (Chapel Hill: University of North Carolina Press, 1980)Google Scholar. The association of construction with sacrifice is quite widespread in Southeast Asia, e.g. as reported in Drake, R.A., “Construction Sacrifice and Kidnapping Rumor Panics in Borneo”, Oceania 59,4 (1989): 269–79.CrossRefGoogle Scholar

¹⁷ Witkamp, H. “Enn Verkenningstocht over het Eiland Soemba”, Tijdschrift van het Koninklijk Nederlandsch Aardrijkskundig Genootschap 29 (1912): 744–75.Google Scholar

¹⁸ Wielenga, D.K., “Reizen op Soemba”, De Macedonier 16 (1912): 150Google Scholar; cf. C.E. Lausz, “Memorie van Overgave van de Afdeling Soemba”, Typescript, Algemeen Rijksarchief, The Hague (1919).

¹⁹ G.H.M. Riekerk, “Dagboek”, Typescript, Archives of the KITLV (H1975 [2]), Leiden (1941).

²⁰ The Sumbanese missions produced a voluminous literature. For basic historical materials, see especially (Anon.), Tot Dankbaarheid Genoopt: Gedenkboek ter Gelegenheid van den 25-jarigen Zendingsarbeid op Soemba (Kampen: J.H. Kok, 1927)Google Scholar; Luijendijk, P.J., Zeven Jaar Zendingswerk op Soemba (1939–1946) (Groningen: J. Niemeijer, 1946)Google Scholar; (Anon.) Rapport inzake de Zending op Soemba (Groningen: J. Niemeijer, 1965)Google Scholar; Kapita, Oe. H., Sedjarah Pergumulan Indjil di Sumba (Waingapu: Lembaga Penerbitan Kristen, 1965)Google Scholar; and Sumba di dalam Jangkauan Jaman (Waingapu: Panitia Penerbit Gereja Kristen Sumba, 1976)Google Scholar; E.L., and May, F. (ed.), Die Insel Sumba: Macht und Mythen der Steinzeit im 20 Jahrhundert (Bonn: Hofbauer, 1980)Google Scholar; Haripranata, H., Ceritera Sejarah Gereja Katolik Sumba dan Sumbawa (Ende: Arnoldus, 1984)Google Scholar; Webb, R.A.F.P., Palms and the Cross (Townsville: James Cook University of North Queensland, 1986)Google Scholar; van den End, T. (ed.), Gereformeerde Zending op Sumba: Een Bronnenpublicatie (Alphen aan den Rijn: Raad voor de Zending, 1987).Google Scholar

²¹ Luijendijk, , Zeven Jaar ZendingswerkGoogle Scholar.

²² A brief sketch of the situation in Sumba is given in Webb, R.A.F.P., “The Sickle and The Cross: Christians and Communists in Bali, Flores, Sumba and Timor, 1965–67”, Journal of Southeast Asian Studies 17,1 (03 1986): 94–112.CrossRefGoogle Scholar

²³ Haripranata, , Ceritera Sejarah, p. 376.Google Scholar

²⁴ Locol Indonesian cultures may respond to national and supra-national pressures by focusing on practice at the expense of doctrine or vice versa. Geertz, following Weber, provided an influential illustration in “‘Internal Conversion’ in Contemporary Bali” [see his The Interpretation of Cultures (New York: Basic Books, 1973)]Google ScholarPubMed. Compare the illuminating discussions in J.M. Atkinson, “Religions in Dialogue: The Construction of an Indonesian Minority Religion”, J. Bowen, “Islamic Transformations: From Sufi Doctrine to Ritual Practice in Gayo Culture”, and Hoskins, , “Entering the Bitter House”, all in Indonesian Religions in TransitionGoogle Scholar.

²⁵ I discuss the socio-cultural background of this sense of incompleteness in Signs of Recognition: The Powers and Hazards of Representations in Eastern Indonesia (forthcoming). For the colonial and postcolonial politics of cultural order, see Pemberton, , On the Subject of “Java”Google Scholar.

²⁶ Onvlee, , “Tana Mema — Tana Dawa”, p. 115Google Scholar.

²⁷ E.W.F.J. Waitz, “Bestuurs-memorie van den Gezaghebber van West-Soemba”, Typescript, Algemeen Rijksarchief, the Hague (1933), p. 9.

²⁸ The Akha, in contrast to the potential for millenarianism found in some other highland groups, are content to let their lost book remain in the past; see Kammerer, C.A., “Customs and Christian Conversion among Akha Highlanders of Burma and Thailand”, American Ethnologist 17,2 (1990): 282CrossRefGoogle Scholar. Similarly, the Anakalangese are notably lacking in millenarial impulses and have little interest in eschatology: the pervasive cultural theme is one of movement away from a plentitude that always remains in the past. Compare the Mambai of Timor, who contrast the present and manifest as merely the incomplete “tip” of a prior and hidden “trunk” of complete ritual knowledge. See Traube, E.G., Cosmology and Social Life: Ritual Exchange among the Mambai of East Timor (Chicago: University of Chicago Press, 1986).Google Scholar

²⁹ Compare reformist Minangkabau attempts to derive traditional law (adat) from the Koran (Abdullah, Taufik, “Modernization”, p. 200Google Scholar). It was, of course, a common strategy on the part of missionaries throughout the world to find analogies to paganism that would legitimate conversion to Christianity. In the early colonial Philippines, for instance, non-Christian customs “could then be regarded as part of a familiar series of ‘fallen’ customs that existed before the advent of Christianity” [Rafael, V., Contracting Colonialism: Translation and Christian Conversion in Tagalog Society under Early Spanish Rule (Ithaca: Cornell University Press, 1988), p. 111Google Scholar]. Acts of apparent recognition and claims of repetition pervade both sides of the encounter (see Keane, , “Materialism, Missionaries, and Modern Subjects”, 1995).Google Scholar

³⁰ Verheijen, J.A.J., Het Hoogste Wezen bij de Manggariers (Wien-Mödling: St. Gabriel, 1951).Google Scholar

³¹ This priest's characterization of Luther seems to me to be rhetorically motivated in another way as well — many conservative Anakalangese, with strong support from the bearers of the state's New Order ideology, tend to see “protest” (the Indonesian term protes serving equally to refer to the actions of Luther as to current politics) in purely negative terms.

³² Although I have permission to use the names of several of the people I mention here, in the interest of general discretion, I have altered the names of all living persons.

³³ For the semiotic characteristics of the imagery of paths, widespread in the Austronesian world, see Parmentier, R. J., The Sacred Remains: Myth, History, and Polity in Belau (Chicago: University of Chicago Press, 1987), pp. 109ff.Google Scholar