History of Linguistics, Volume 1: The Eastern Traditions of Linguistics [1 ed.] 0582094895, 9780582094895

Citation preview

LONGMAN LINGUISTICS LIBRARY HISTORY OF LINGUISTICS VOLUME I

LONGMAN LINGUISTICS LIBRARY General editors R. H. Robins, University of London Martin Harris, University of Manchester Geoffrey Horrocks, University of Cambridge

A Short History of Linguistics Third Edition R. H. ROBINS Text and Context Exploration in the Semantics and Pragmatics of Discourse TEUN A. VAN DIJK Introduction to Text Linguistics ROBERT-ALAIN DE BEAUGRANDE AND WOLFGANG ULRICH DRESSLER Psycholinguistics Language, Mind. and World DANNY D. STEINBERG Principles of Pragmatics GEOFFREY N. LEECH Generative Grammar GEOFFREY HORROCKS The English Verb Second Edition F. R. PALMER A History of American English J. L. DILLARD English Historical Syntax Verbal Constructions DAVID DENISON Pidgin and Creole Languages SUZANNE ROMAINE Historical Linguistics Problems and Perspectives EDITED BY C. JONES Aspects in the English Verb Process and Result in Language YISHAI TOBIN The Meaning of Syntax A Study in the Adjectives of English CONNOR FERRIS

General Linguistics An Introductory Survey Fourth Edition R. H. ROBINS A History of English Phonology CHARLES JONES Generative and Non-linear Phonology JACQUES DURAND Modality and the English Modals Second Edition F. R. PALMER Semiotics and Linguistics YISHAI TOBIN Multilingualism in the British Isles I: the Older Mother Tongues and Europe EDITED BY SAFDER ALLADINA AND VIV EDWARDS Multilingualism in the British Isles II: Africa, Asia and the Middle East EDITED BY SAFDER ALLADINA AND VIV EDWARDS Dialects of English Studies in Grammatical Variation EDITED BY PETERTRUDGILLAND J. K. CHAMBERS Introduction to Bilingualism CHARLOTIE HOFFMANN Verb and Noun Number in English A functional explanation WALLIS REID English in Africa JOSEF S. SCHMIED Linguistic Theory The Discourse of Fundamental Works ROBERT DE BEAUGRANDE Aspect in the English Verb Process and Result in Language YISHAI TOBIN History of Linguistics Volume I: The Eastern Traditions of Linguistics Volume II: Classical and Medieval Liguistics GIULIO LEPSCHY

H istoryof Linguistics Volume I: The Eastern Traditions of Linguistics edited by Giulio Lepschy

Routledge

Taylor & Francis Group

LONDON AND NEW YORK

First published 1990 by Longman Group Limited English translation first published 1994 Chapter 5 translated by Emma Sansone Published 2013 by Routledge 2 Park Square, Milton Park, Abingdon, axon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business

Copyright@bySocieteeditriceil Mulino. 1990 English translation Copyright © 1994, Taylor & Francis. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fuliest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury andlor damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. ISBN 13: 978-0-582-09489-5 (pbk)

British Library Cataloguing-in-PublicationData

A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-PublicationData Storia della linguistica. English. History of linguistics / edited by Giulio Lepschy. p. cm. Includes bibliographical references and index. Contents: v. 1. The eastern traditions and linguistics - v. 2. Classical and medieval linguistics. ISBN 0-582-09488-7 1. Linguistics-History. I. Lepschy. Giulio C. II. Title. P61.S7513 1994 93-31169 410'.9-dc20 CIP

Contents Introduction

Vll

Acknowledgements

xv

Noteson the contributors 1

Chinese linguistics Goran Malmqvist 1.1 Introduction 1.2 The period prior to the Qin (-221 BC) 1.3 The periodsofQin and Han (221 BC-220 AD) 1.4 The periodsof Wei, Jin and Nanbeichao(220-581) 1.5 The periodsof Sui, Tang and Song(581-1279) 1.6 The periodsof Yuan and Ming (I260-1644) 1.7 The period of Qing (1644-1912)up to the end of the nineteenthcentury Notes Bibliography

XVI

1

1 1 4 9 II

15 I7 22 23

2 Indian linguistics GeorgeCardona 2.1 Introduction 2.2 Early work on phoneticsand phonology 2·3 Etymology 2-4 Pal)ini 2·5 Pal)ini and later commentators 2.6 Interestin dialects 2·7 Developmentof a philosophyof language 2.8 Summary Notes Bibliography

25

3

61

Linguistics in the ancient Near East Erica Reiner 3. 1 Introduction

25 27 30 33 41 46 48 50 51 58

61

vi

CONTENTS

3.2 Ancient Egyptianlinguistics JanetH. Johnson Notes Bibliography 3.3 Sumerian Miguel Civil Notes Bibliography 3.4 Akkadian Erica Reiner Notes Bibliography 4

5

Hebrew linguistics RaphaelLoewe 4· I Introduction 4.2 Time concepts 4.3 Languageand metaphysics:Torah, Hebrewand Babel 4.4 Late antiquity. Languagesin Palestineand in the Jewishdiaspora 4.5 Jewishlanguage-satellites 4.6 Philological, philosophicaland theologicalinterests 4.7 The cabbalistictransformer 4.8 Renaissance illusions: Haskalahand AufkHirung 4.9 The handmaidenof nineteenth-century nationalism 4.10 The dethronementof tradition 4. II Masterand servant &~ Bibliography Arabic linguistics Henri Fleisch 5.1 Introduction 5.2 ClassicalArabic 5.3 The holy Book 5.4 The 'Grammarians' 5.5 Linguistic analysesof the grammarians 5.6 Greekinfluences &~ Bibliography

Index

63 72 73 76 85 86 87 94 94 95 95 97

97 99 100 I02 I06 I 13 132 133 135 138 142 143 152

164

165 166 169 171

173 179 180 183 185

Introduction

This work originatedin the discussionsheld by a group of advisors for linguistics of the Italian publishers il Mulino. Examining the areasin which new and useful initiatives could be encouraged,it was thought that a large-scalehistory of linguistics would meet a widely felt need, and I was asked to elaboratea plan for such a work. The preparationand completion of the project took about ten years, and the work, written by scholarsfrom different countries, beganappearingin Italian in 1990. This English edition has beenreorganizedinto four volumes.In this introduction I shall say something about the nature and contents of this work, and its placewithin the presentpanoramaoflinguistic historiography. What I had in mind was a history of linguistic thought, rather than an accountof the developmentof linguistic science.In other words, for different societiesand in different periods, I wanted to have a presentationof the prevailing attitudes towards language: its social, cultural, religious and liturgical functions, the prestige attached to different varieties, the cultivation of a standard,the place of language in education, the elaboration of lexical and grammatical descriptions, the knowledge of foreign idioms, the statusof interpretersand translators,and so on. This implies of course a 'view from within', that is, presenting the linguistic interests and assumptionsof individual cultures in their own terms, without trying to transposeand reshapethem into the context of our ideas of what the scientific study of language ought to be. The purpose is an understandingof what certain societies thought about language, rather than an assessmentof their ideason a scaleof scientific progress. The problem is familiar to historiansof scienceand the line I am following does not necessarilyimply a lack of confidencein the

Vlll

INTRODUCTION

possibility of obtainingreliable findings in linguistic study. Whether 'science'is the most suitable term for what linguistics does, is of courseopen to debate:seeChomsky(1969, 56) and Graffi (I99Ib). In any case,the authorsof different chapterswere also encouraged not to enlarge on methodological problems posed by linguistic historiography but to concentrateinstead on the presentationof facts, on the interpretationof texts which they consideredrelevant and interesting for the periods and cultures they were discussing. But, after obtaining the agreementof my authors concerningthe generalaims, I left them completelyfree to organizetheir chapters as they thought fit, without trying to impose a unifying format, becauseI thought that differences of perspectiveand treatment would contribute to highlighting specific features of individual traditions. For instance,the chapterson Indian and Arabic concentrate on the elaborationof grammaticalanalysis,in both casesan important contribution of thesecivilizations, whereasthe Hebrew chapterplacesmore stresson the theological, mystical and philosophicalcontextof Jewishreflectionson language. The content of the four volumes of the English edition is as follows: Volume I includes the ancient traditions, each of which develops in a manner which is, from a viewpoint both cultural and chronological,largely independent,apartfrom obviousconnections, like thosebetweenArabic and Hebrewthought in the Middle Ages. (The Graeco-Roman tradition, which is the basis of those reflections on languagewhich we presentchronologically in Volumes II-IV, appearsat the beginning of the secondvolume.) Chapter I, by G6ran Malmqvist, of Stockholm University, describes the developmentof Chineselinguistics, analysing the relevant lexicographical and grammatical works, and throwing light on the particular shapeimposed on phonological analysis by the logographic nature of Chinesescript. Chapter 2, by George Cardona, of the University of Pennsylvania, presents the Indian grammaticaltradition, its cultural and religiousimplications,and particularly the contribution of pa1).ini, illustrating its systematiccharacter and its attentionto detail. Chapter3, supervisedby Erica Reiner of the University of Chicago, presentsand interprets the documents which bear witness to linguistic interests and knowledge in the civilizations of the ancient near East; the chapter is divided into three sections, devoted to Ancient Egyptian (by Janet Johnson), Sumerian (by Miguel Civil), and Akkadian (by Erica Reiner). Chapter 4, by Raphael Loewe, of the University of London, examinesthe place of languagewithin the Hebrew tradition, from the Biblical period, throughthe Talmudists,the mystics,the enlightenment,down to the rebirth of Hebrew as an everydaylanguage,

INTRODUCTION

ix

payingparticularattentionto the philosophicalandcultural implications of thesetrends. Chapter5, by the late Henri Fleisch, of Saint JosephUniversity in Beirut, probably the last essayto flow from the pen of this eminentscholar,dealssyntheticallywith the original system of grammatical analysis elaboratedby the great Arabic civilization of the Middle Ages. The following volumes presentthe main stagesof the European tradition, in their chronologicalsuccession.They include two chapters each. In Volume II, the first chapter, by Peter Matthews, of CambridgeUniversity, deals with classical linguistics and offers a reading of the main texts of the Graeco-Romanworld which elaborate the grammatical categorieson which we still base our analysis of language.The second chapter, by Edoardo Vineis, of Bologna University, and (for the philosophy of language) by Alfonso Maieru, of Rome University, presentsa detaileddiscussion of language study from the end of the sixth to the end of the fourteenth century, not limited to the late medieval period which hasreceivedmost attentionin recentyears(with particularreference to Modistic philosophy),but extendingto the lessfrequentlystudied early Middle Ages. In Volume III, the first chapter, by Mirko Tavoni, of Pisa University, covers the fifteenth and sixteenth centuries, and is completed by two sections, one on Slavia Romana, by M. D. Gandolfo,and one on Slavia Orthodoxa,by S. Toscano;the bibliography of this chapter takes advantageof the great Renaissance Linguistic Archive set up by M. Tavoni at the Istituto di Studi Rinascimentaliin Ferrara.The secondchapter,by RaffaeleSimone, of Rome University, offers a helpful map of the varied terrain constituted by seventeenth-and eighteenth-centuryculture, in which many of the roots are found from which the great plant of later comparativelinguistics derives its nourishment; R. Simone, who is a professional linguist, keeps in mind the philosophical perspectivethat is particularly relevantfor this period. In Volume IV, the first chapter,by Anna Morpurgo Davies, of Oxford University, examinesthe flowering of historical and comparativelinguistics in the nineteenthcentury, stressingin particular someaspectswhich traditional presentations,focusing on the Neogrammarians,sometimesleave in the shadow, like the interest in typological classifications, and the importance for comparative philology of the newly constitutedGermanuniversity systemin the first threedecadesof the century.The secondchapter,by the editor of this work, offers a synthesis of the main developmentsin twentieth-centurylinguistics, extendingfrom the progressof comparative studies to linguistic theory, philosophy of language,and

x

INTRODUCTION

the investigationof languageuse in different areas,from literature to social communication.The Italian edition also includesa chapter on the history of Italian linguistics and dialectology, by Paola Beninca, of Padua University, which has been omitted from the English edition. Let us briefly look at the presentstate of linguistic historiography. Over the last three decadesthere has been a considerable revival of interestin this field. In 1974 Konrad Koerner, a German scholar teaching in Canada,founded the journal Historiographia Linguistica, which has becomea forum for internationaldiscussion on the history of linguistics; in 1978 he was the organizer of the first internationalcongresson the history of linguistics (the meetings continueat three-yearlyintervals). There are also severalother associationsdevoted to the history of linguistics, like the Societe d'histoire et d'epistemologiedes sciencesdu langage, created in 1978, presidedover by S. Auroux, which publishesa Bulletin and the journal Histoire Epistemologie Langage (1979-); the Henry Sweet Society for the History of Linguistic Ideas,founded in 1984 on the initiative of Vivian Salmon,which organizesregularmeetings and publishes a Newsletter; the North American Association for the History of the LanguageSciences,foundedin 1987. The interest in the history of linguistics over this period is also indicatedby the publication of monographs and collections of studies, such as Hymes (1974), Parret (1976), Grotsch (1982), Schmitter (1982), Chevalierand Encreve(1984), Bynon and Palmer(1986), Formigari and Lo Piparo (1988), and by the two tomes devoted to the historiographyoflinguistics by Sebeok(1975)· A separatestudy would be necessaryto examinethe main available histories of linguistics, from the great works of the nineteenth century devoted to Classical linguistics by Steinthai (1863), to Oriental philology by Benfey (1869), and to Germanicphilology by R. von Raumer(1870). Subsequentstudieswitness the triumph of Neogrammariancomparativephilology, from Delbriick (1880) to Meillet (1903) to Pedersen(1924). Sincethe 1960snumeroushistorical presentationshave been published, from the large and wellinformed work by Tagliavini (1963), lacking however in historical perspectiveand theoreticalinsight, to the two-volume contribution by Mounin (1967-72), the neat and well-balancedbook by Robins (1967), the acute and comprehensivesynthesisby Law (1990), the disappointing attempts by Malmberg (1991) and Itkonen (1991)· Large-scaleworks have also startedappearingin the last few years, under the direction of Auroux (n.d.) and of Schmitter (1987). There is no spaceto mention in this context the great many studies devoted,since the 1960s,to individual periodsand problemsin the

INTRODUCTION

Xl

history of linguistics, although they often provide the detailed groundwork that makes possible overall synthetic assessments (sometitles will be quotedin the list of references). What is the place of our history of linguistics against this background?To me it seems to be placed in an advantageous position, in the middle ground between the concise, one-author profiles on the one hand, and the extended,multi-authored,multivolume series on the other. Comparedto the former, it has the greater richness of detail, which is made possible by the larger space available and the higher degree of reliability that derives from the authority of contributors who are specialists in the individual areas; comparedto the latter, it is more compact and coherentin perspectiveand basic assumptions,and it can not only be consultedfor single questions,or studiedin individual sections, but also be readin its entirety. If I were askedto presenta schematicprecis of the main features which I seeas inspiring this work and characterizingits realization, I would list the following points: A perspectivedirected towardsunderstandingthe past, rather than dealing with present-dayconcerns.The aim is to reconstruct and illustrate different epochs and traditions within their own context and on the basisof their own values,rather than of their appeal to present-daypreoccupations;to highlight their linguistic interests,ratherthan our own. 2. This is a history oflinguistic thought,of interestsand attitudes toward language.Thesemayor may not find a place within the elaborationof a 'scientific' study of language(howeverwe may want to define it), but in any caseI feel that an account of the preoccupationswith linguistic mattersin different societies proves to be an interesting and worthwhile object of historical investigation. 3. I have consideredit essential,in my choice of authors for individual chapters,that they should be specialists,able to analyse the relevant texts in the original languages,and to presentthem to a lay readership.I aimed at obtaining not an account of what is known, derived from current literature, but a series of original contributions based on first-hand study of the primary sources. 4. From what precedesit is clear that this work is prevalently concernedwith a historical and philological study of ideas, texts from the past, rather than with methodological and theoreticalproblemsposedby historiography.It is an 'extroverted', rather than an 'introverted' history, dealing with the I.

xii

INTRODUCTION

facts it analyses,ratherthan with the theoreticaland ideological assumptionswhich lie behind the work of the historian. This obviously does not imply that methodologicalquestions are not a legitimate object of study; but I believe that it is possibleto offer useful contributionson the history of linguistics, without dealing in the first instancewith the theory of historiography. 5. One of the main linguists of our times, Yakov Malkiel, observedsomeyearsago (Malkiel and Langdon 1969) that to produce good work on the history of linguistics it is not enoughto be a linguist: one has also to be a historian, and to fulfil the expectationsnormally raised by a historical essay. Here, of course, one can only observethat the ability to set some episodesof the history of linguistics within their social and cultural context, is the exception rather than the rule (examplesthat come to mind are thoseof Dionisotti (1967a,b, 1972) or Timpanaro(1963, 1965, 1972, 1973)). What I had in mind for this history of linguistics was the more modestaim of providing information aboutideason language,in different periodsand societies,which are not easily (and in somecases not at all) accessibleelsewhere. I know from direct experiencethat linguists feel the need for a work of this kind, and I hopeit may also appealto readerswho are interested to know how people, at different times and within different cultural traditions,havelooked at one of the mostessential and challenging featuresof our common humanity - that is, language. G. C. Lepschy References From Locke to Saussure. Essays on the Study of LanguageandIntellectualHistory, University of MinnesotaPress,Minneapolis. AMIROVA, T. A., OL'KHOVIKOV, B. A. and ROZDESTVENSKIJ, JU. v. (1975) Oterki po istorii lingvistiki, Nauka, Moskva (German tf. Abriss der Geschichte der Sprachwissenschaft,Veb Bibliographisches Institut, Leipzig 1980). AUROUX, s. (ed.) (n.d.) Histoire des idees linguistiques. Tome I: La naissance des metalangages.En Orient et en Occident, Mardaga, Liege/ Bruxelles. AUROUX, s. (1992) Histoire des idees linguistiques. Tome 2: Le developpementde la grammaireoccidentale,Mardaga,Liege. AARSLEFF, H. (1982)

INTRODUCTION

Xlll

BENFEY, T. (1869) Geschichte der Sprachwissenschaftund orientalischen Philologie in Deutschland, seit dem Anfange des 19. Jahrhunderts mit einemRiickblick aufdie friiheren Zeiten, Cotta, Miinchen. BORST, A. (1957-63)Der Turmbau von Babel. Geschichteder Meinungeniiber Ursprung und Vielfalt der Sprachenund Volker, Hiersemann,Stuttgart. BRINCAT, G. (1986) La linguistica prestrutturale,Zanichelli, Bologna. BYNON, T. and PALMER, F. R. (eds) (1986) Studies in the History of WesternLinguistics. In Honour of R. H. Robins, CambridgeUniversity Press,Cambridge. CHEVALIER, J.-C. and ENCREVE, P. (eds) (1984) Vers une histoire sociale de la linguistique(LangueFran9aise,63), Larousse,Paris. CHOMSKY, N. (1966) Cartesian Linguistics. A Chapter in the History of Rationalist Thought,Harperand Row, New York. CHOMSKY, N. (1969) Linguistics and Philosophy, in Languageand Philosophy. A Symposium,edited by S. Hook, New York University Press, New York, 51-94. DELBRijcK, B. (1880) Einleitung in das Sprachstudium.Ein Beitrag zur Geschichteund Methodik der vergleichendenSprachforschung,Breitkopf und Hartel, Leipzig. DIONISOTTI, c. (1967a)La lingua italiana da Veneziaall' Europa.II Fortunio e la filologia umanistica.Niccolo Liburnio e la letteraturacortigiana,in Rinascimentoeuropeo e Rinascimentoveneziano,edited by V. Branca, Sansoni,Firenze,1-46. DIONISOTTI, c. (1967b) Geografia e storia della letteratura italiana, Einaudi, Torino. DIONISOTTI, c. (r972) A Year's Work in the Seventies.The Presidential Address of the Modern Humanities Research Association delivered at University College, London, on 7 January 1972, reprinted from The Modern LanguageReview,Vol. 67, NO.4. FORMIGARI, L. and LO PIPARO, F. (eds) (1988) Prospettive di storia della linguistica, Editori Riuniti, Roma. GRAFFI, G. (r99Ia) La sintassi tra Ottocento e Novencento,il Mulino, Bologna. GRAFF!, G. (199Ib) Concetti 'ingenui' e concetti 'teorici' in sintassi,Lingua e Stile, 26, 347-63. GROTSCH, K. (1982) Sprachwissenschaftsgeschichtsschreibung. Ein Beitrag zur Kritik und zur historischenun methodologischenSelbstvergewisserung der Disziplin, Kiimmerle Verlag, G6ppingen. HYMES, D. (ed.) (1974) Studies in the History of Linguistics, Indiana University Press,Bloomington. ITKONEN, E. (1991) Universal History of Linguistics,Benjamins,Amsterdam. KOERNER, K. (1978) Towards a Historiography of Linguistics. Selected Essays,Benjamins,Amsterdam. KOERNER, K. (1989) Practicing Linguistic Historiography. SelectedEssays, Benjamins,Amsterdam. LAW, v. (1990) Languageand Its Students:the History of Linguistics, in An Encyclopedia of Language, edited by N. E. Collinge, Routledge, London/NewYork, 784-842.

xiv LEPSCHY, G.

Bologna.

INTRODUCTION

(ed.) (1990) Storia della linguistica, vol.

I

and 2, il Mulino,

and LANGDON, M. (1969) History and Histories of Linguistics, RomancePhilology, 22, 530-74. MALMBERG, B. (1991) Histoire de la linguistique de Sumer a Saussure, PressesUniversitairesde France,Paris. MEILLET, A. (1903) Introduction a{'etude comparativedes languesindoeuropeennes,Hachette,Paris(eighth edition, 1937). MOUNIN, G. (1967-72) Histoire de la linguistique, 2 vols, PressesUniversitairesde France,Paris. OL'KHOVIKOV, B. A. (1985) Teorija jazyka i vid grammaticeskogoopisanija v istorii jazykoznanija. Stanovleniei evolucija kanona grammaticeskogo opisanija v Evrope,Nauka,Moskva. PARRET, H. (ed.) (1976) History of Linguistic Thought and Contemporary Linguistics,de Gruyter, Berlin. PEDERSEN,H. (1924) Sprogvidenskaben i det Nittende Aarhundrede.Metoder og Resultater,GyldendalskeBoghandel,K0benhavn(English tr. Linguistic Sciencein the NineteenthCentury, Harvard University Press,Cambridge, Massachusetts1931; and with the new title The Discovery of Language.Linguistic Sciencein the NineteenthCentury, Indiana University Press,Bloomington 1962). RAUMER, R. VON (1870) Geschichteder germanischenPhilologie vorzugsweisein Deutschland,Oldenbourg,Miinchen. ROBINS, R. H. (1967) A Short History of Linguistics, Longman, London (third edition, 1990). SALMON, v. (1979) The Studyof Languagein Seventeenth-Century England, Benjamins,Amsterdam. SCHMITTER, P. (1982) Untersuchungenzur Historiographie der Linguistik. Struktur-Methodik- TheoretischeFundierung,Narr, Tiibingen. SCHMITTER, P. (ed.) (1987) Zur Theorie und Methode der Geschichtsschreibung der Linguistik. Analysen und Reflexionen (Geschichte der Sprachtheorie,I), Narr, Tiibingen. SEBEOK, T. (ed.) (I975) Historiography of Linguistics (Current Trends in Linguistics, 13), Mouton, The Hague. STEINTHAL, H. (1863) Geschichteder Sprachwissenschaft bei den Griechen und Romern mit besondererRiicksicht auf die Logik, Diimmler, Berlin (secondedition, 189~I). TAGLIAVINI, c. (1963) Panorama di storia della linguistica, Patron, Bologna. TIMPANARO, s. (1963) La genesi del metodo del Lachmann, Le Monnier, Firenze(new edition, Liviana, Padova1985). TIMPANARO, s. (1965) Classicismo e illuminismo nell'Ottocento italiano, Nistri Lischi, Pisa(second,enlargededition, 1969). TIMPANARO, S. (1972) Friedrich Schlegele gli inizi della linguistica indoeuropeain Germania,Critica Storica, 9, 72 - 105. TIMPANARO, s. (1973) II contrastotra i fratelli Schlegele Franz Bopp sulla strutturae la genesidelle lingue indoeuropee,Critica Storica, 10, 53--90 .

MALKIEL, Y.

Acknowledgements

In the preparationof the Italian edition I availed myself of the advice of many friends and colleagues.For Volume I, I should like to thank in particularZyg Baranski,Verina Jones,Anna Morpurgo Davies, Joanna Weinberg (Introduction); Michael Halliday and Michael Loewe (Chapter I); Anna Morpurgo Davies (Chapter2); Arnaldo Momigliano (Chapter 3); Ada Rapoport and Joanna Weinberg (Chapter4); Bernard Lewis (Chapter5); for Volume II, Tullio De Mauro (Chapter2); for Volume III: GiuseppeDell'Agata (ChapterI), Tullio De Mauro (Chapter2). I am grateful to Emma Sansone,who has sensitively and helpfully performedthe difficult job of translating into English chapters originally in other languages(Volume I, Chapter 5; Volume II, Chapter 2; Volume III, Chapters I and 2; Volume IV, Chapter 2). I should also like to thank the series'editors for their advice and commentsduring the preparationof the English edition. The English version introduces some updatingsand improvementson the previousItalian edition. In particular Volume I, Chapter4 includesa more detailed discussion of Hebrew grammaticalideas.For help and advice concerning the history of Arabic linguistics (Volume I, Chapter 5) I am grateful to ProfessorE. Ullendorff and ProfessorA. F. L. Beeston.

Notes on the contributors

(to all four volumes)

GeorgeCardonareceiveda BA from New York University and an MA and PhD in Indo-EuropeanLinguistics from Yale University. He has spent several years studying texts of traditional sastras with paI).], p [ph], m [m], f[f]; d [g], t [th], n [n], I [I]; z [g~], c [tsh], S [s]; zh [g~], ch [t~h], sh [~], r [~]; j [gc], q [tch], x [c]; g [~], k [P], h [x]; Pinyin finals: i [i] (after the initials z, c, s the vowel i is realizedas a vocalized[?]; after the initials zh, ch, sh, r the vowel i is realizedas a vocalized[?]), a [a], e [Y-], o ["], ai [ail, ei rei], ao [au], ou [ou], an [an], en [;)n], ang [a1)], eng [A1)], ong [U1)], er [;)']; ia rial, ie riel, iao [iau], iu [iu] (with tones I or 2), [iou] (with tones 3 or 4), for the tones see p. 9, ian [im], in [in], iang [ia1)],

23

NOTES

ing [iI)], iong [iuI)]; u [u], ua [uo], uo [U::l], uai [uai], ui lui] (with tones 1 or 2), [uei] (with tones 3 or 4), uan [uan], un [un], uang [UOI)], ueng [UAI)]; U (after the initials j, q, x pinyin omits the trema above the vowel u) [y], ue [YE], uan ryan], un [yn]. 1. Fengsu tongyi ('A comprehensiveaccount of customsand mores'), by Ying Shao(secondcenturyBC). 2. Hsun Tzu (Xun Zi) (1963) Basic Writings, trans. by B. Watson,Columbia University Press,New York and London, 144. 3. Examplesof the six typesof graphsin the Shuowenjie zi in their ancient and modernform: Pictographs Ideographs Compoundideographs Phoneticcompounds

, El

El

~

+

-

IlJJ ,wI

If

jJ

.E-

sA ;~

5PJ

'sun' 'moon' 'child' 'one' 'three' 'bright' The Yangzi river' The Yellow river'

In thesegraphsthe common left-hand elementis a classifier signifying 'water'. The right-hand element in the first graph reads gong (Old Chinesekung) and servesto indicate the pronunciationof the graph in which it is included, which was Old Chinesekung, modernjiang. The phonetic in the second graph is ke (Old Chinesek'a), approximating the pronunciationof the graph in which it is included, he (Old Chinese g'a).

*"

*Loan characters usedto denotethe homophoneverb 'to come'. Derivative graphs ~ ~ ~

~

'wheat'. 'high age'

~W'

4. Teng, Ssu-Yii (1968) Family Instructions for the Yen Clan (Yen-shih chia-hsun)by Yen Chih-t'ui. An AnnotatedTranslation with Introduction (Monographiesdu T'oungPao,4), E. J. Brill, Leiden. 5. JosephNeedham(1961) Scienceand Civilisation in China, Vol. I, Cambridge University Press,Cambridge,34. 6. 111 nan; jfa nuo. 7. ~ yu; M! yong.

Bibliography c. (1954) A Linguistic Study of the Shih Ming (HarvardYenching Institute Studies, II), Harvard University Press,Cambridge, Mass. COBLIN, w. S. (1972) An Introductory Study of Textual and Linguistic Problemsin Erh-ya (PhD Thesis,University of Washington),University Microfilms, Ann Arbor, Michigan. BODMAN, N.

24

CHINESE LINGUISTICS

(1940) Grammata Serica, Bulletin of the Museum of Far EasternAntiquities, 12, 1-471. KARLGREN, B. (1954) Compendiumof Phoneticsin Ancient and Archaic Chinese,Bulletin of the Museumof Far EasternAntiquities,26, 211-367. LO, CH'ANG-P'Ei (1944) Indian Influence on the Study of ChinesePhonology, Sino-IndianStudies,I, 115-27. STIMSON, H. (1962) Phonology of the Chung-yuan yin-yun, Tsinghua Journal of ChineseStudies,New Series,3, 114-59. TENG, SSU-YU and BIGGERSTAFF, K. (1950) An AnnotatedBibliography of SelectedChineseReferenceWorks, HarvardUniversity Press,Cambridge, Mass. THERN, K. L. (1966) Post/aceof the Shuo-wenChieh-tzu, the First ComprehensiveChineseDictionary (Wisconsin China Series, I), Departmentof EastAsian Languagesand Literature, University of Wisconsin. KARLGREN, B.

2

Indian linguistics George Cardona

2.1 Introduction *

Ancient and medieval Indian scholarsproducedworks of linguistic interest concerning Sanskrit, the high language (samskftii viik 'refined, ritually pure speech') acceptedby an early Indo-Aryan elite as the most correct speechform, and related Middle Indic vernaculars(called priikfta), as well as Dravidian languages.In the following, I shall concentrateon the works that deal with Sanskrit. I shall considerin particular some of the major concernsof Indian grammarians- principally within the pa1).inian traditions - and others with respect to topics of linguistic importance. These are justifiably consideredthe most original and importantcontributions early Indian scholarshavemadeto the studyoflanguage,and pa1).ini is generallyacknowledgedas a grammarianof the first rank. 2.1.1 Linguistic work connectedwith the Vedas Early Indian thoughtconcerninglanguagewas intimately connected with the sacredtexts called the Vedas, India's most ancient literary documents.From early on thesewere associatedwith six ancillaries (vediinga [nom. pI. nt. vediingiimJ, 'limbs of the Veda') intended to help in the proper maintenanceand application of these texts. Three of the vedangasconcern areasthat in modern times would come within the purview of linguistics: sik~a ('phonetics'),vyakara1).a (,grammar'), and nirukta (,etymology'). Works in the first areadeal with the productionof soundsand their properpronunciation for various recitational traditions. Such phonetic matters are treatedalso in works called priitisiikhya. In addition, thesedeal in greatdetail with the relationsbetweencontinuouslyrecited versions

* Seebeginningof Notes.

26

INDIAN LINGUISTICS

(safnhitiipiithiih) of Vedas and the analysed texts (padapiithiilJ)

associatedwith them. The proceduresfor positing the various padapathasshow that the authorsof theseanalysedtexts followed definite principles wherebydivisions were madeat particularplaces - not only between syntactic words and membersof compounds but also at certain morph boundarieswithin syntacticwords - and particularwords were left unanalysed.Moreover, padapathascame to be viewed as the sourcesfrom which sarilhitapathaswere derived, so that phonologicalrules of derivation, togetherwith phonological classifications, came under consideration. VyakaraI).a serves to accountfor the relations amongcomponentsof sentences(viikyiini 'utterances')through derivational systems.Nirukta concernsitself with the etymological explanation (nirvacanam) of particular words in Vedic texts. 2.1.2 Pal,lini and his successors The culmination of early linguistic thought in India is represented in the grammar,called the ('group of eight chapters'), of the renowned grammarianpal).ini, who cannot reasonablybe consideredto have lived later than the early fourth century Be. pal).ini's work not only surpassedand eclipsed works of earlier grammarians,upon whose work he drew, but also became the object of intense scrutiny. The earliest author whose discussions are available is Katyayana (c. third century BC), who takes up problematic issues concerning particular rules (siitra, nom. pI. siitriilJi) in his varttikas (see note I), and Patafijali (mid-second century BC), whose Mahabha~ya(,great commentary') not only incorporatesKatyayana'svarttikas - which are paraphrasedand discussed- but also includesindependentconsiderationsof certain problems. Traditionally, varttikas are viewed as intended to consider what is said, what is left unsaid,and what is possibly wrongly said in sutras.1 There can be no reasonabledoubt that commentatorialworks on pal).ini's grammarwere producedbefore Katyayana, and commentariescontinued to be producedlater, right up to the present day. These commentariesdiffer among themselvesnot only as concernsparticular interpretationsacceptedand defendedbut also in the way they organizethe sourcetext. Somefollow the order of the sutras as found in the The most famous and widely used commentaryof this type is the Kasikavrtti (c. seventh century AD) of Jayaditya and Vamana, which is based on a conftatedversionof the including additionsand modificationssuggestedin varttikas of Katyayanaand acceptedearlier by Candragominin his own grammar, the Candravyakaral).a. Some

A~tadhyaYI

A~tadhyayl.

A~tadhyaYI,

EARLY WORK ON PHONETICS AND PHONOLOGY

27

A~tiid­

commentaries,on the other hand, reorder the sutrasof the hyiiYl into thematicgroups: rules concerningterminology (sainjiiiisiitriifJi),metarules sandhi rules, sutras that apply in the derivation of nominal forms, and so on. The most widely used text of this kind is the Siddhiintakaumudlof Bhatto(sixteenthcentury).

(paribhii~i si triifJi),

jidlk~ita

2.1.3 Non-Piil.linian works Although grammaticalworks associatedwith Piil)ini are justly the most well known, there are other grammaticaltreatisesand commentaries. One such grammar, Candragomin's,is renowned not only becauseit reflects modifications in Piil)ini's statementsbut also because,accordingto tradition, Candragominis consideredto have revived the study of Piil)inian grammarafter this had ceased to be handeddown traditionally from teacherto pupil. Other wellknown grammarsare the Kiitantra and a whole seriesof works by Jainas, among them the Jainendravyiikaral)aand Hemacandra's Siddhahemasabdiinusiisana. Some of these works, such as the Kiitantra, are methodologically interesting in that they revived proceduresfound in priitisiikhya descriptions but abandonedby Piil)ini and other grammarians. The discussionsfound in commentatorialworks are not restricted to circles of grammariansalone - whether of the Piil)inian 'school' or within any of the various non-Piil)inian traditions - but extend to thinkers in areassuch as the exegesisof ritual texts (mlmiiinsii), poetics, different schools of logic, and others, including Buddhist thinkers. It is not without reasonthat Patafijali says that grammar is the principal one among the six vediitigas (pradhiinain ca vyiikarafJam: Abhyankar1.1, line 19).

~a(sv

aizge~u

2.2 Early work on phonetics and phonology

Speechproduction Early priitisiikhya texts such as the Rgvedapriitisiikhya and the Taittirlyapriitisiikhya deal with the relations between padapiitha and sarilhitiipiitha and also include descriptionsof speechproduction which show considerablesophistication.Attention is paid not only to places (sthiiniini) where particular sounds are produced but also to the types of articulation involved - whether full occlusion slight occlusion and so on - as well as to how the air streamis modified as it is emitted. Different varieties of air stream emitted, dependenton the status of the glottis, are also distinguishedand associatedwith distinct groupsof sounds. A brief summaryof the major points madein the Taittirlyapriiti2.2.1

(sP':~tam),

(dusP':~tam),

28

INDIAN LINGUISTICS

sakhya concerningspeechproduction and phonologically distinct sets of soundswill serve to illustrate.2 Three varieties of air stream are recognized,dependingon whether the glottis (kalJfha) is close (sml1V1:te kalJfhe), open [kalJfhe]) or midway (madhye) between these states; these are respectively called niida, sviisa, and hakiira. 3 They are the basic matter of articulatedsounds nada for vowels (svara) and voiced consonants (gho!javat), hakara for h and voiced aspirated stops (caturtha), svasa for voiceless consonants(agho!ja).4 The same work also distinguishesbetweenthe gesturesmade in producing vowels and consonantsat placesof production (sthiina). In producing vowels, a speaker brings an active articulator (karalJam) into proximity (upasainharati)with a place of production(sthiinam); for producing other sounds, one brings an articulator into contact (spar.sayati 'causes to touch') at a place of production.5 For example, in producing soundsof the group c ch j jh Ii (cavarge), one lets the blade of the tongue (jihviimadhyena 'middle of the tongue') make contactat the palate(tiilau), but for producing i- vowels and e one brings the tongue blade Uihviimadhyam) into proximity with the palate.6

(viv~te

(varlJaprak~tayal;):

(prak~ti)

2.2.2 Phonological classificationsand operations A distinction is made between phonetics and phonology. Terms like svara, 'vowel', vyalijana, 'consonant',gho!javat, 'voiced consonant', agho!ja, 'voiceless consonant' are used with reference to membersof phonologicalclasses,so that in treatiseswhich observe the distinction strictly gho!javat is not used of vowels, although these are obviously voiced. The phonological classesin question are pertinent to phonological rules of conversion, of the type a ----> b, whereby forms in sarilhitapathasare derived from forms in padapathas.For example,the sarilhitapathaof Rgveda 8.IOO.Ioa: yad vadanty avicetanJni, 2.I8-4b: J caturbhir J !jaqbhir huyamiinal; contain yad, 'when', before vJg (nom. sg.) 'speech', which is followed by vadanty'uttering' (nom. sg. fern. agreeingwith vJg), and !jaq- 'six' before the instrumentalplural ending,-bhir. Th~ corresponding passagesof the padapathaare: yilt/viik/vadanti/ avi-cetanJni/, J/ catul;-bhil;/J/!jaq-bhil;/ huyamiinal;, with pauses not only between syntacticallyseparatewords such as yilt and viik but also between the ending bhil; and precedingstems. A rule of the Rgvedapratisakhya 7 provides that voicelessunaspirated stops (prathamiil;) changeto corresponding(sviin 'own') voiced unaspirated stops if !hey are follo\\;:ed b~ voiced (gho!javatpariil;): yat viik ----> yad viik, viik ----> viig vadantl, !jat-bhil; ----> !jaqbhil;. Similarly, a rule of the Taittirlyaprati-

wig

(t~tTyiin)

c~msonan!s

vadanti

EARLY WORK ON PHONETICSAND PHONOLOGY

29

sakhya8 provides that a voiceless unaspiratedstop (prathamal;) changesto a voiced unaspiratedstop (tr:tryam) of the same class (savargryam)ifit is followed by a vowel or a voicedconsonant(svara).

gho~avatparal;

Criteria for recognizingseparateconstituentsin padapiithas As notedearlier, the continuouslyrecitedtext of a Veda is generally treatedas derivedfrom a correspondinganalysedtext. The Rgvedapratisakhya9 explicitly states that the continuous recitation in close junction (samhitii) has for its source the separatelyrecited padas(padaprakl:til;) of the padapatha,and it goes on to say that the former (sii) is derived from the latter through bringing padafinal sounds (padiintiin) into contact with the initial sounds of padas (padiidibhil;) without any interval (kiiliivyaviiyena).lo As I have also pointed out, the padaswhich the padapatharecognizes as separateunits are not only syntacticwords, including clitics, but also endingssuchas -bhil; and stemsthat precedethem. Such divisions are basedon phonologicalcriteria. Thus, one has final -0 instead of -al; before such endings (e.g. tamobhil;: tamal; [nom.-acc.sg.] 'darkness'),just as -0 is found insteadof -al; at a full word boundarywherea voicedconsonantfollows; and -n-stemssuch as rajan- 'king' have forms without -n- not only at full word boundaries(nom. sg. rajii) but also before an ending like -bhil; (instr. pI. rajabhil;). Accordingly, composersofpadapathas (padakiiriil;) such as Sakalya,who composedthe padapathato the Rgveda, treat such endings as precededby a pada boundary. On similar grounds,certainderivationalaffixes arealso treatedas precededby a pada boundaryunder the appropriatecircumstances.For example, go-mat 'rich in cattle' (acc. sg. masc.gb-mantam)is split as shown, but the suffix of {irjasvantal; (nom. pI. masc.)'full of strength'is not so split: if -vantal; were precededby a pada boundary,one should have -0 insteadof -as before -v-. Again, correspondingto 'acts like a bull' of the samhitapatha,the padapathahas recognizing that a derivate with the denominative suffix -yacorrespondsto a syntacticcomplexthat includesa nominativeform of the stem iva acarati. Of course,componentsof compoundsare split, for comparable reasons;e.g. praja-patil;. In addition, padakarassplit items into two componentpadasa.:t most, and in doing so they recognizea hierarchy. Thu,s, pra-jii 'creature, pcajii-vat 'rich in progeny', prajii-patil;. 'Prajapati', prajiipati-gfhitayii 'seized by Prajapati'(instr. sg. fem.) are split as shown. Moreover, the authors of padapathasshow evidence of letting semanticconsiderationsalso influence their analysis. For example, 2.2.3

VI:~iiyate vf~a-yate,

vf~an-: vf~iiyate = vf~i

pro~eny',

30

INDIAN LINGUISTICS

althoughrtvij- 'ritual officiant, priest' is obviously divisible into rtu'appropriatetime, season'and ij, an agent noun derived from yaj 'venerate,performa sacrificial rite', the componentsof forms suchas rtvijam (acc. sg.) are not separatedby pausein the padapiitha.This reflects a possible ambiguity in the precise phrase to which the compoundis consideredto correspondandthe fact that rtvij- cameto have a meaning ('priest') not associatedwith a fixed etymological segmentation.In its commentaryonPiit:lini 3.2.59,wherertvij- is given asa ready-madederivatewith a zero-affix kvin (= v with markersk, i, n; seesection24I) the Kiisikiivrtti gives different paraphrases of the compound,including onessuch that this correspondsto phrasesin which yajati 'performsa rite, venerates'is construedwith a locative (rtau yajati ' ... at the appropriatetime, season')or an accusative (rtum yajati); it also notes that this is used as a term with a fixed meaning not necessarilyassociatedwith its etymological analysis. of Rgveda3.24.I, the padapatha Again, correspondingto has dustaraIJ instead of duIJtaraIJ. The same verse occurs in the Yajurveda,as Vajasaneyisamhitii9.37. The padapathacorresponding to this shows that du~(arah is a compoundby giving dustara iti dustaraIJ, but it does not separatethe pre-verb from the following verbal derivative by meansof a pause,so that it does not give duIJ-taraIJ. In the Vajasaneyipratisakhya,11 moreover, is given as one of a series of compoundsin which there is no separate pronunciation (niivagrahaIJ) of componentsdivided by pause,and commentatorsexplain that this reflects a possibleambiguity concerningthe derivation of the verbal noun, from different verbs tf. Centuries earlier than these commentators,Bhartrhari, in the autocommentaryon verse 2. 13 of his Viikyapadiya, remarks that dustaraIJ is given in the padapiitha without pause between componentsbecausethe second element of the compound could derive from either of two verbs.12 The authors of padapathasthus clearly followed certain principles in setting up their analysedtexts. They were also clearly aware 13 One of of grammaticalderivations and phonological processes. thesepadakaras,moreover,antedatesPat:lini, who refers to Sakalya, the author of the padapathaassociatedwith the Rgvedasamhita (seesection2-4.3).

du~(araIJ

du~(araIJ du~(araIJ

2.3 Etymology 2.3.1 Introduction Vedic texts,especiallyBriihmat:lasand Arat:lyakas,aboundin etymological explanationsmeant to demonstratethe essentialproperties on accountof which entities are called by the namesthey bear. For

ETYMOLOGY

31

example,puru/ia ('man') used in the senseof 'life breath' (priilJalJ) is explained as follows: The life breath resides (sete 'lies') in the body (puri), so that, being so (purisayam santam [acc. sg.]), it is referred to (iicak/iate '[the gods] call') puru/ia, in an oblique manner (parok/ielJa 'beyond direct observation');for the gods (deviilJ) are fond of what is not directly observable(parok/iapriyii iva) and dislike what is obvious(pratyak/iadvi/ialJ).14 2.3.2 Yaska's Nirukta

Etymology came to be used as a tool in Vedic exegesis.An early collection of words subject to such analysisis the Nighal).tu (nom. pI. nighalJtavalJ), which Yaska took as a basis for etymological discussionsin his work called Nirukta. In this work, Yaska also explains the statusof nirukta relative to grammar(vyiikaralJa) and the proceduresto be followed in giving etymologicalexplanations. He remarksthat without this (idam antarelJa) etymologicalscience there is no proper understandingof the meaningsto be understood in Vedic mantras (mantre/iv arthapraptyayo na vidyate).15 Moreover, one who doesnot comprehendthe meaning(artham apratiyatalJ [gen. sg.]) of a given item cannothave an absoluteunderstanding of the proper accentand grammaticalformation of the term in question(niityantam svarasamskiirodddalJ),since theseare dependent on the meanings of terms; for example, a derivate may be accented in two different manners, depending on its meaning. Hence, this science of etymology, which is the abode of other sciences(vidyiisthiinam), is a complement to grammar (vyiikaralJasya kiirtsnyam) while it also serves its own purpose (sviirthasiidhakamca).16

2.3.3 Grammar presupposedin Nirukta

Clearly, Yaska assumesthat etymologicalexplanationpresupposes grammar.Thus, for example,he recognizesfour classescomparable to what Westernerscall parts of speech:nominal forms (niiman), verb forms (iikhyiita) , preverbs(upasarga), and particles (nipiita). Yaskagoeson to note what are said to be the distinctivecharacteristics of nominal and verbal forms. A verb form (iikhyiitam) has an activity as its principal meaning (bhiivapradhiinam), but nominal forms (niimiini) have substantialbeingsas their principal meanings (sattvapradhiiniini). Moreover, he remarks, in cases where both nominals and verbals signify actions, one uses a verb form to signify (iikhyiiteniica/i{e 'expresseswith a verb form') an act in sequence,in progress (piirviiparTbhiitam bhiivam), as when one says vrajati, 'is going', pacati, 'is cooking'; and nominal forms like the action nouns vrajyii, 'going', paktilJ, 'cooking' are used to

32

INDIAN LINGUISTICS

signify an action as a being, an embodiedentity (miirtam sattvabhiitam sattvaniimabhih) encompassingan act from beginning to end (upakramaprabhrty apavargaparyantam),but not viewed as sequential.17 2.3.4 Views of etymologistsand grammarianscontrasted Yaska contraststhe views of etymologistsand grammarians:The convention of etymologistsis that all substantives(niimiini 'nominals') arise from verbs (niimiiny iikhyiitajiini), and this is accepted also by the grammarianSakatayana;but other grammarians hold that not all substantivesderive from verbs (na sarviilJlti), and this view is acceptedby the etymologist Gargya.18 Adherentsof these opposing views also present arguments. Two of the arguments given against the position that all substantivesderive from verbs are as follows. Any entity that carried out the action (yah kasca tat karma kuryiit) which servesas a basis for somethingbeing designatedby a particular substantivederived from a verb denoting that action would be referred to by that term (sarvam tat sattvam tathiicakfor example, if a horse is called asva becauseit covers a distanceon a path (adhviinam asnute), anything that performs the action denoted by the verb asnute should equally be called asva, but this is not SO.19 In addition, things would be referred to by means of derivates with the formations which are appropriate, accordingto grammaticalrules of derivation, to their being derived from the verbs in question and such as would be immediately one would refer to understood;accordingly,insteadof using the entity in question by means of the derivate purisaya,20 and insteadof asva one would use a derivatein which as, 'reach', is followed by the generalagentivesuffix tr. 21

~Tran);

a~tr-,

puru~a

2.3.5 The etymologists'standmaintained Adherents of the etymologists' position defend their stand. The first objection noted is met with an appealto generalusage:we see (pasyiimah)that of many peoplewho perform the sameact (samiinakarmalJiim), some receive a particular appellation (niimadheyaby which they are specifically referred to pratilambham but others do not for example, is usedonly of somepeoplewho hew, not of anyonewho engagesin this act.22 Against the second objection, etymologists argue that there are indeed other instancesof derivateswith primary suffixes (krtah) that have restricted usage (alpaprayogiih)23 and whose meaningsare neverthelessclearly understood. Yaska stands firmly by the etymologists' tenet: it is fine to

eke~iim) (naike~iim);

tak~an-'carpenter'

accent

33

follow the derivations of grammariansin the case of words like karta (nom. sg. of kar-tr- 'one who does, makes, an agent'), for which accentand derivationalformation (svarasamskarau)in terms of basesand affixes are clearly capable(samarthau)of association with the meanings,being linked harmoniouslywith a modification of a verb (vikarelJanvitau). Even if such obvious derivation is not available, however, one should explain words in terms of acts signified by verbs, and in no instanceshould one not give such an etymological explanation(na tv eva na nirbruyat), giving priority always to the meanings(arthanityalJ) of terms being explained.24 2.3.6 Yaska and pa.,ini Although Yaska doesindeedpresupposegrammaticaldescriptions, the available evidence is not sufficient to determine whether he precededor succeededpaI).ini. From the point of view of methodology, however, this is not a crucial issue, since such early works as Sakalya'spadapathato the Rgveda already give clear evidenceof sophisticationregardinggrammaticalanalysis. 2.4 PalJini

24I Introduction The culmination of early phonological and grammatical work in ancient India is to be seenin the work of India's most outstanding and famous grammarian of all time, paI).ini. He follows earlier traditions in that he describesSanskrit by meansof a derivational systemwherein final utterancesare obtainedby performing operations on posited strings (see section 2-4-4.2). paI).ini also adopts a typical grammarian'sstancewith respectto the etymologicalcontroversy (see section 2.3-4): he does indeed provide for the derivation of such obviously analysableitems as kar-tr- (