Article

The Indonesian numerals

From Munshipedia, the MBRAS digital historical encyclopedia

The Indonesian numerals

R.J. Wilkinson’s 1895 article in the Journal of the Straits Branch of the Royal Asiatic Society surveys the comparative etymology of Indonesian and Oceanic numerals, drawing on the work of Dr. T.H. Pardo de Tavera and the Rev. D. MacDonald to argue that these numeral systems share a common origin rooted in quinary (base-five) counting and that their wide distribution across the Pacific provides strong evidence for Polynesian migration from the Malay Archipelago.

Summary

Wilkinson’s article is a compact comparative-linguistic review that engages with two recent publications: a 1894 pamphlet by Dr. Pardo de Tavera on the origin of Tagalog numerals and a 1893 paper by the Rev. D. MacDonald in the Journal of the Polynesian Society. The central question is the origin and distribution of the numeral systems shared across Malay, Polynesian, Melanesian, and Malagasy languages. Wilkinson presents a comparative table spanning six languages (Malay, Maori, Tagalog, Malagasy, Formosan, Fijian) to demonstrate the remarkable uniformity of the basic numerals 1–10 across a vast geographic area.

The article then evaluates two competing etymological theories. Pardo de Tavera’s account, which Wilkinson largely endorses, holds that the oldest numerals (2, 3, 4) were originally collective substantives (“a couple,” “a trio,” “a company”) rather than abstract numerals, and that the higher numbers (5–10) were built on a quinary system based on counting on the fingers. The word lima/rima (5) still means “hand” in many Formosan and Polynesian dialects, and the Malay delapan (8) is derived from dua-lepan (“two turned down”), preserving the finger-counting logic. Wilkinson then turns to MacDonald’s more radical proposal that the Oceanic numerals are of Semitic origin, tracing supposed cognates between Arabic, Syriac, and Amharic forms and their Oceanic counterparts. Wilkinson rejects this theory on multiple grounds: the phonological character of Oceanic languages is wholly unlike Semitic; the Arabic numerals are not collective nouns and do not follow a quinary system; and MacDonald’s specific etymological pairings are philologically unsound (e.g., the connection between anam and sitt via butanga is arbitrary, and the selection of delapan as a cognate of t’man ignores the well-established internal derivation from dua-lepan).

The article closes by situating the numeral evidence within the broader debate on Polynesian origins. Wilkinson sides with de Quatrefages’ migration theory, noting that the numerals are used by Melanesian tribes along the routes believed to have been followed by migrating Polynesians from Ceram and Bourou. He acknowledges Wallace’s alternative explanation—that Malay traders carried the vocabulary to the South Seas—but finds it implausible given the primitive condition of Polynesian societies at first European contact.

Key Findings

  • A comparative table of numerals 1–10 is presented across six languages (Malay, Maori, Tagalog, Malagasy, Formosan, Fijian), demonstrating near-identical forms across the entire Indonesian–Polynesian–Malagasy area (p. 99).
  • The word lima/rima (5) still means “hand” in many Formosan and Polynesian dialects; quinary counting systems persist among the inhabitants of Triton Bay (New Guinea), Santo, Efate, Ambrym (New Hebrides), Engano (near Sumatra), and some Formosan tribes (p. 100).
  • The Malay delapan (8) is derived from dua-lepan (“two turned down”), and sembilan (9) from sa-ambilan, both preserving the finger-counting logic of the original quinary system (p. 100).
  • Pardo de Tavera’s etymologies identify the oldest numerals as collective substantives: lua/dua (2) from a root meaning “double or copy”; telo/tolu (3) from a root meaning “triple” connected with tali (rope, triple strand); pat/fa (4) from a root meaning “a complete set” or “company” (p. 100).
  • MacDonald’s Semitic-origin theory is refuted on the grounds that the Arabic numerals are not collective nouns, do not follow a quinary system, and that specific pairings (e.g., anam–sitt via butanga; delapan–t’man) are philologically untenable (pp. 101–102).
  • The article’s final footnote catalogues numeral data from 119 languages across Sumatra, Borneo, Celebes, the Moluccas, the Philippines, Formosa, Micronesia, Melanesia, Polynesia, Madagascar, and Fiji, drawn from the works of Favre, Crawfurd, Wallace, Terrien de la Couperie, Macdonald, Pardo de Tavera, and Tregear (p. 103).

Conclusion

Wilkinson’s definitive takeaway is that the Indonesian numerals furnish “perhaps the best illustration” of the linguistic relationship between the Indonesian and Polynesian races, and that this relationship is best explained by de Quatrefages’ theory of Polynesian migration from the Malay Archipelago rather than by Wallace’s trader-contact hypothesis or by any Semitic origin.

Context

  • The article is a short review piece (5 pages) typical of the JSBRAS in this period, engaging with contemporary philological debates rather than presenting original fieldwork.
  • The primary sources consulted are Pardo de Tavera’s 1894 Manila pamphlet, MacDonald’s 1893 Journal of the Polynesian Society paper, and the comparative numeral data compiled by Favre, Crawfurd, Wallace, Terrien de la Couperie, and Tregear, most held in the Society’s library.

References