Word order in discourse Ed. by Pamela Downing and Michael Noonan (review)

REVIEWS185 English, by the often critical audience evaluation that follows the sermon, by a convention that sermons avoid political topics, and by a general Nukulaelae distrust of those who claim or assert authority. This last is a feature of many small-scale egalitarian societies in Polynesia and Micronesia—authority and group consensus are necessary for certain social undertakings, but they are frequently opposed and undermined by headstrong individual and family action. In many of these societies, excluding women from public speaking roles is one way in which hierarchy is produced in otherwise egalitarian social organization (see Lederman 1991 on the Papua New Guinea Mende and Kulick 1992 on the Papua New Guinea Gapun for similar accounts ). B contextualizes and widens his specific account of Nukulaelae literacy events and practices through connected analyses of gender and emotion. As he argues in the concluding chapter (169-87), reading and writing are often directly involved in the reproduction ofinequality, in Western and non-Western contexts, and gender is often implicated in such processes—through determining the extent to which individuals are given access to literacy and through ideological judgments of gendered literacy in which women's literacy, like women's work, is rendered invisible or devalued. The male-composed sermon is one obvious way in which such hierarchizing processes operate among the Nukulaelae. There are other ways, however, in which gender and literacy are intertwined—for example, gender is 'indexed' in letter writing. A significant function of letter writing as an institution is that letters 'centralize emotions and ways of experiencing emotions that are symptoms of vulnerability ... to which women [are seen to] have privileged access' (183). While men and women are avid letter writers and readers, letters foreground alofa, 'feelings of love and empathy', and they highlight the vaaivai 'vulnerability' of the sender. All Nukulaelae have and express such feelings upon addressing separation from kinspeople, but—no surprises here—feeling empathy and vulnerability are 'womanly' . Men are allowed to take on these feminine traits in letter writing, but within limits. The 'emotional economy' of letters is regulated by and in interaction with a more basic regulation of gender. It is a measure of the worth ofLiteracy, emotion, andauthority that it can develop substantial anthropological accounts of such matters while telling us a great deal about the linguistic situation of Nukulaelae society, including its century-long engagement with letters. REFERENCES Besnier, ????. 1988. The linguistic relationships of spoken and written Nukulaelae registers. Language 64.707-36. Kulick, Don. 1992. Anger, gender, language shift, and the politics of revelation in a Papua New Guinea Village. Pragmatics 2.281-96. Lederman, Rena. 1984. Who speaks here? Formality and the politics of gender in Mendi, Highland Papua New Guinea. Dangerous words: Language and politics in the Pacific, ed. by Donald Brenneis and Fred Meyers, 85-107. New York: New York University Press. Department of Anthropology College of Arts & Sciences Social Science 263 State University of New York, Albany Albany, New York 12222 Word order in discourse. Ed. by Pamela Downing and Michael Noonan. Amsterdam & Philadelphia: John Benjamins, 1995. Pp. vi, 596. Reviewed by Jeanette K. Gundel, University ofMinnesota* This volume contains selected papers from the symposiumWord Order in Discourse, held at the University ofWisconsin, Milwaukee, April 12-14, 1991. There are eighteen chapters, including an * In preparing this review, I have benefited considerably from discussions with Ann Mulkern. 186LANGUAGE, VOLUME 74, NUMBER 1 (1998) introduction by Pamela Downing which situates the papers in a general overview of word order research, its findings, and its methodological problems. The papers are strongly functionalist and empiricist in their orientation. Most use a quantitative text distribution approach to investigate discourse factors influencing word order in a single language or in languages of a particular word order type. Five studies examine conditions under which noun phrases occur preverbally in languages with verb-first order. Susanna Cumming ('Agent position in the Sejarah Melayu', 51-84) finds that preverbal agents in the classical Malay text she examines occur only in clauses with preverbal patients. She proposes that they serve to 'reorient the reader to participants which are given but in a potentially new relationship, thereby initiating a text structure boundary'. Michael Darnell ('Preverbal nomin...