This study focuses on the Na people of northern Yunnan and their unusual kinship system. Cai Hua argues that with the Na "social anthropology is faced, for the first time in the history of the knowledge of humankind, with a society that is without fathers and husbands".

Marriage is the exception rather than the rule in this matrilineal society, in which men "visit" women's houses. While this is not completely new information - there have been many studies published in Chinese, particularly since the 1960s - it is the first substantial piece of work published in English in recent times and is based on extensive fieldwork with the Na and a study of their language.

Yunnan province is in the far southwest of China on the borders of Burma, Laos and Vietnam. It is ethnically the most complex region of China, and many of the minority communities are related to groups in southeast Asia or India. The Na, who live in the area around Yongding close to the border between Yunnan and Sichuan, are part of the ethnic group classified as Naxi (Nakhi) in the official Chinese list of national minorities. The Na themselves argue that they are completely separate and, according to Cai, claim that their language and that of the Naxi who live further south in the Lijiang region, are not mutually intelligible, although both belong to the Tibeto-Burman language group. The Naga of Assam are close relatives.

Cai is professor of social anthropology at Beijing University and, judging by conversations I had with academics in China last autumn, is highly respected there. He studied anthropology at the Universite de Paris X, Nanterre, in the Levi-Strauss school, and his book carries an endorsement from the master.

ÁñÁ«ÊÓƵ

The information on Na kinship was collected during two short field trips in 1985 and 1986 and a year-long stay between 1988 and 1989, when Cai worked in the village of Wenquan, which had the highest percentage of people who practised the "visit" system.

The body of the book details the Na kinship system, including the "furtive visit", the "conspicuous visit", cohabitation in various degrees of permanence, and marriage (although the author claims that this is rare). Cai has gone to great trouble to construct the complex lineages that the Na system produces. Almost all the data are derived from conversations with Na informants, some of whom were initially reluctant to discuss their sexual lives after similar inquiries in the 1960s prompted the Chinese authorities to try to "civilise" them with a monogamy campaign.

ÁñÁ«ÊÓƵ

It would have been useful to know how these conversations took place. Were they conducted solely in the Na language - very much a foreign language to a Chinese researcher - or was a Chinese-speaking intermediary involved? The reliability of the data depends so much on these conversations. An index would have been useful, and there are some errors in the spelling of Chinese words in Pinyin , but that apart, this is a valuable study that is likely to stimulate discussion on the nature of kinship and marriage in east and southeast Asia.

Michael Dillon is senior lecturer in modern Chinese history, University of Durham.