VOICE FORM AND FUNCTION Edited by BARBARA FOX University of Colorado, Boulder PAUL J. HOPPER Carnegie Mellon University JOHN BENJAMINS PUBLISHING COMPANY Amsterdam/Philadelphia 1994 Θ The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. Library of Congress Cataloging-in-Publication Data Voice : form and function / edited by Barbara Fox, Paul J. Hopper. p. cm. - (Typological studies in language, ISSN 0167-7373; v. 27) Includes bibliographical references and index. Contents: Active voice and middle diathesis : a cross-linguistic perspective / Manuel Arce- Arenales, Melissa Axelrod, and Barbara A. Fox -- Voice, aspect, and Aktionsart : middle and passive in ancient Greek / Egbert Bakker - A functional typology of antipassives / Ann Cooreman -- Voice : beyond control and affectedness / William Croft — The rise of the Eng­ lish get-passive / T. Givón and Lynne Yang — Passive participles across languages / Martin Haspelmath -- Middle voice, transitivity, and the elaboration of events / Suzanne Kemmer -- On "middle voice" verbs in mandarin / Charles N. Li and Sandra A. Thompson ~ The impli­ cations of ergativity for a Philippine voice system / Marianne Mithun -- A tale of two passives in Irish / Michael Noonan - The Tupi-Guarani inverse / Doris Payne ~ Passives and alterna­ tives in children's narratives in English, Spanish, German, and Turkish / Dan I. Slobin. 1. Grammar, Comparative and general-Voice. I. Fox, Barbara A. II. Hopper, Paul J. III. Series. P281.V57 1993 415~dc20 93-37185 ISBN 90 272 2915 5 (hb.) / 90 272 2916 3 (pb.) (European; alk. paper) CIP ISBN 1-55619-418-8 (hb.) /1-55619-419-6 (pb.) (U.S.; alk. paper) © Copyright 1994 - John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 75577 · 1070 AN Amsterdam · The Netherlands John Benjamins North America · 821 Bethlehem Pike · Philadelphia, PA 19118 · USA Contents Abbreviations vii Introduction ix Active Voice and Middle Diathesis: A Cross-Linguistic Perspective 1 Manuel Arce-Arenales, Melissa Axelrod and Barbara A. Fox Voice, Aspect and Aktionsart: Middle and Passive in Ancient Greek 23 Egbert Bakker A Functional Typology of Antipassives 49 Ann Cooreman Voice: Beyond Control and Affectedness 89 William Croft The Rise of the English GET-Passive 119 T. Givón and Lynne Yang Passive Participles across Languages 151 Martin Haspelmath Middle Voice, Transitivity, and the Elaboration of Events 179 Suzanne Kemmer On "Middle Voice" Verbs in Mandarin 231 Charles N. Li and Sandra A. Thompson The Implications of Ergativity for a Philippine Voice System 247 Marianne Mithun A Tale of Two Passives in Irish 279 Michael Noonan The Tupí-Guaraní Inverse 313 Doris Payne VI CONTENTS Passives and Alternatives in Children's Narratives in English, 341 Spanish, German, and Turkish Dan I. Slobin Index of Languages 365 Index of Subjects 369 Abbreviations A agent EMPH emphatic ABL ablative ERG ergative (case) ABS absolutive (case) EXCL exclusive (first person) ACC accusative ACT active FEM feminine ADJ adjective FUT future ADV-PART adverbial particle GEN genitive AGR agreement 3GEN third person genitive AOR aorist  antipassive IMM immediate ART article IMP imperative ASP aspect(ual marker) IMPERF imperfective IMPERS impersonal BEN benefit IMPF imperfect BENE benefactor INCL inclusive (first person)  common IND indicative CLASS classifier INST instrumental CL.PRO clitic pronoun INTRANS intransitive CNSQ consequential IRR COMP complementizer irrealis(aspect marker) COMPAR comparative LD locative/directional COMPL complement LK linker COMPLE completive LOC locative CONT continuative CONTR contrastive MASC masculine CRS current relevance MID middle marker MVT movement NEG negative morpheme or DAT dative particle DEF definite NOM nominative DEM demonstrative NOMI nominalizer DES desiderative NOMIN nominative DIMIN diminutive NP noun phrase DIR directional NUM PART numerical particle DS directional suffix DU dual O, OBJ object DUR durative DBL oblique (case) DYN dynamic viii ABBREVIATIONS Ρ patient S (intransitive) subject PART particle SG singular PARTCP participle SUBJ subject PASS passive PAST past (tense marker) TG Tupí-Guaraní PERF perfective TOP topic PL plural TRANS transitive POS possessive UNM unmarked PR proper UNSPEC unspecified PRES present (tense marker) V verb Q interrogative VINTR intransitive verb QUOT quotative VTR transitive verb REFL reflexive 1 first person REL relativizer 2 second person REL PART relative particle 3 third person Introduction One of the crucial contributions made by typology in the last 30 years has been the exploration and reanalysis of traditional linguistic categories (e.g. Schachter, 1976; Hopper and Thompson, 1980). Originally based on the grammar of classical languages, these traditional categories have often been found to be inadequate to the task of describing the range of cross-linguistic data explored in typological studies. For example, a typical result of the typological testing of traditional categories is the finding that a particular category fits into a larger functional domain, and that its instantiation as a category in the classical tradition is only one of many possibilities for coding that domain. 'Voice' is one of these categories. Well-known among the Sanskrit, Greek, and Latin grammarians as one of the possible paradigmatic sets of the verb — Active, Middle, and Passive — the term was readily extended, when the time came, to the vernacular languages that came under their con­ trol. Active, middle, and passive were then accepted as universal categories: for example, Bloomfield, explicating the focus system of Tagalog in his 1917 grammar, unhesitatingly postulated an analysis based on multiple "passives". The importance of voice within typological approaches has become especially apparent since the further examination of the closely related questions of subject- and objecthood (e.g. Li 1976). The investigations into other voice-related questions such as Transitivity (Hopper and Thompson 1980) and Ergativity (Comrie 1977; Dixon 1982) have further assured a continued interest in the nature of passives and other such phenomena. The study of voice also has clear connections to what has come to be called 'in­ formation flow' (Chafe 1987), with its origins in Prague School syntax, and to such unfamiliar phenomena as 'inverse forms' that first came to the attention of linguists through Native American languages. Efforts to cap­ ture the complexity of these phenomena, both functionally and formally, have resulted in a sophisticated understanding of the nature of voice in par- χ INTRODUCTION ticular, and of form-function relationships in general. The title of the present volume reflects the concern of all of the con­ tributions with formulating, both for specific languages and across lan­ guages, the multiple relationships between form and function in the domain of grammatical voice. While each of the papers focuses on a slightly differ­ ent facet of form-function correlations, several recurrent questions emerge out of the individual perspectives. Perhaps the most basic question addressed in these chapters is how to identify a clause-type in a given language as an instance of given voice cate­ gory. That is, given a set of categories for the linguist to choose from, and some possible constructions in a specific language, what criteria should the linguist use to classify a construction as an instance of a voice category? Each of the papers confronts this issue at least implicitly, and some of them address it explicitly: Noonan addresses possible mismatches between form- based and function-based definitions for passive voice; Li and Thompson examine a candidate for middle voice in Mandarin Chinese, Kemmer and Arce, Axelrod, and Fox explore definitions for middle phenomena cross- linguistically, Mithun takes another look at voice phenomena in a Philip­ pine language, and Cooreman discusses her decision to start with a form- based definition of antipassive in order to find functional correlates. Croft provides a framework within which all voice phenomena can be situated. Having identified a construction as an instance of a given voice catego­ ry, we still must find ways to understand and describe the function of a given voice construction in a specific language. Bakker discusses the func­ tions of middle voice in Ancient Greek, Cooreman looks at the functions of antipassives, Givón and Yang describe the changes in function of get con­ structions in English, Haspelmath presents intriguing data on passive par­ ticiples cross-linguistically, Kemmer explores the functions of middle voice, Noonan examines the functions of two passive-like constructions in Irish, and Payne argues for the function of inverse in Tupi-Guarani. The next level is one of scope: what is the range of voice phenomena in the languages of the world? In addition to the traditional categories active, passive, and middle voices, Payne argues for including the notion of an inverse construction, a transitive clause which is nonetheless patient-cen­ tered. Arce et al. suggest that certain constructions in specific languages which are currently treated as instances of either middle or passive voice ought to be instead treated as middle diathesis, a construction-type which exhibits active voice marking but is also marked for an affected subject. INTRODUCTION xi Noonan discusses an impersonal construction in Irish, and Cooreman pre­ sents a typological study of antipassives. The possibility that form-based definitions of voice categories may lead to different classifications than function-based definitions leads several of the authors to delve deeper into the relationships between form and func­ tion. Bakker examines the mutually constraining relationship between Aktionsart and overt voice marking. Croft explores the cognitive models which appear to underlie voice systems and discusses the voice categories associated with specific deviations from the central Idealized Cognitive Model. Kemmer adopts a similar framework for exploring middle voice phenomena. Many of the other papers discuss this issue in passing (e.g. Cooreman, Haspelmath, Li and Thompson, Mithun, Noonan). The relationships between form and function extend beyond simple voice in some of the cases studied. In particular, verbal aspect turns out to be closely associated with the functions of voice. The papers by Bakker, Arce et ai. and Cooreman treat this association in some depth; others men­ tion it in passing (e.g. Noonan and Payne). The importance of historical evolution in understanding the organiza­ tion of voice systems cannot be overlooked in a functional approach to voice. Givón and Yang's paper takes a primarily historical perspective on one voice construction, the so-called get passive in English. Kemmer, Li and Thompson, and Arce et al. touch on possible historical sources for the phenomena they observe. The construction-types explored in these papers are: passive, middle, impersonal, inverse, and anti-passive, as these are variously defined by the authors. Passive is understood for the most part by the authors as a syntactic construction which performs the following functions (Givón, 1981): - topicalizes the patient - defocuses the agent - stativizes the event Arce et al., Givón and Yang, Mithun, Noonan, and Slobin explore the use of constructions that have been called passive in diverse languages. Arce et al., while focusing for the most part on what they have called middle dia­ thesis, look briefly at the periphrastic passive in Spanish and English, com­ paring these passive constructions to what they claim to be active-voice/ middle-diathesis counterparts: se constructions in Spanish and get construc-