The problemAlthough language families such as Indo-European and Sino-Tibetan have more speakers, the Austronesian language family is the world's secondlargest in number of languages, with well over 900 members (Ruhlen 1987).These languages have a wide geographical distribution in Taiwan, the Philippines, Malaysia, Indonesia, portions of mainland Southeast Asia, Madagascar, and a variety of Pacific nations located within the broad geographical regions of Mclanesia, Micronesia and Polynesia.As with any large and widely distributed collection of languages or cultures, there is a great deal of linguistic and cultural diversity among the Austronesianspeaking peoples.Since they reflect a common ancestral tradition, the differences among these various languages and cultures must have arisen through an accumulation of gradual linguistic and cultural changes over many generations.It is the task of the comparative linguist to reconstruct as much as can be recovered of the language which existed before these changes.Because language and culture are interconnected in sometimes surprising and unexpected ways, the reconstruction of particular linguistic features may under favourable circumstances also entail the reconstruction of interconnected features of culture.Somewhat over a decade ago I argued on the basis of comparative linguistic evidence that early Austronesian (AN) society had ancestororiented kin groups (descent groups), matrilateral cross-cousin marriage (hereafter simply 'matrilateral cousin marriage'), dual divisions and some * ACKNOWLEDGEMENT I am indebted to Alice Dewey, Gregory Forth, Rodney Needham
Stored only when both papers are in the corpus — 97 citations exist upstream.