An Outline of Middle Voice in Syriac: Evidences of a Linguistic Category 9781463221270

This study presents a modern linguistic approach to the function of the Syriac et-verbal prefix. Based on a detailed ana

180 30 13MB

English Pages 197 Year 2011

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

An Outline of Middle Voice in Syriac: Evidences of a Linguistic Category
 9781463221270

Citation preview

An Outline of Middle Voice in Syriac

Perspectives on Syriac Linguistics

6

A publication of the International Syriac Language Project (TSLP), this series contains monographs, paper collections, and reference works on Syriac lexicography and related ancient languages.

An Outline of Middle Voice in Syriac

Evidences of a Linguistic Category

Margherita Farina

1 gorgias press 2011

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright© 2011 by Gorgias Press LLC

All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC.

2011

^

^

1

ISBN 978-1-4632-0145-6

ISSN 1935-6900

L i b r a r y of Congress C a t a l o g i n g - i n - P u b l i c a t i o n Data Library of Congress Cataloging-in-Publication

Data

Farina, Margherita. An outline of middle voice in Syriac : evidences of a linguistic category / by Margherita Farina, p. cm. Revised and expanded from the originally presented as the author's Doctoral thesis in 2009. Includes bibliographical references and index. ISBN 978-1-4632-0145-6 (alk. paper) 1. Syriac language--Grammar. I. Title. PJ5421.F37 2011 492'.35--dc23 2011040808 Printed in the United States of America

TABLE OF CONTENTS

List of Tables and Charts Foreword Acknowledgments Abbreviations Note on Transliteration 1 Introduction to the Problem 1.1 Survey of the Linguistic Material 1.2 Preliminary Observations 1.3 Et- Forms in some Modern Syriac Grammars 1.4 Mutawaa and 'Middle' Verbs in Arabic Linguistics

vii xi xiii xv xvii 1 1 8 10 13

2 Middle Voice and some Linguistic Correlations 2.1 Middle Voice 2.2 Aspectual and Actional Correlations 2.3 Unaccusativity 2.4 Formulation of the Working Hypothesis

17 17 23 32 40

3 Et- forms and Inner Passives in Semitic Studies 3.1. A Survey of Ancient Semitic Verbal t- Forms: Status Quaestionis 3.2. West Semitic and Aramaic inner passives

43 43 63

4 Textual Analysis 1: Middle Semantics 4.1 Presentation of Sample 4.2 Methodology of the Analysis 4.3 Voice 4.4 Aspect 4.5 Actionality Charts

69 69 71 73 82 89 98

5 5.1 5.2 5.3 5.4 5.5

Textual Analysis 2: Coreferential Dative 111 Introduction: Syntactic Aspects Ill Coreferential Dative, Dativus Ethicus and Object Marker 113 Coreferential Dative with the Other Verbal Stems 117 First Conclusions: Semantic and Syntactico-Semantic Aspects 120 Other Properties of Coreferential Dative in the Demonstrationes and in Other Syriac Texts 122 5.6 Typological Comparisons 125 5.7 Concluding Remarks 132 6 Conclusions Appendix 1 Diathesis in Medieval Syriac Grammars

135 139 139

v

vi

OUTLINE OF MIDDLE VOICE IN SYRIAC

Appendix 2 145 A. List of Occurrences of the Structure et- Form + / + Coreferential Dative in Aphrahat's Demonstrationes 145 B. List of Occurrences of the Structure et- Form + / + Coreferential Dative in Ephrem's Commentaries on Genesis and Exodus 158 C. Coreferential Dative in the First Seven Acts of Judas Thomas 162 D. Coreferential Dative with et- Stems in the Pesitta Gospels 167 Bibliography

169

LIST OF TABLES AND CHARTS

Table 1. - Syriac verbal system

2

Table 2. - Aspectual categories

24

Table 3. - Actional categories

30

Table 4. - Bakker's event-type classification

37

Table 5. - Semantic structure of stative and eventive verbs

38

Table 6. - Articulation of Syriac reflexive and middle structures

42

Table 7. - Biblical Aramaic verbal system

57

Table 8. - Palestinian Aramaic verbal system

58

Table 9. - Syriac verbal system

59

Table 10. - Voice of the et- stems and inner passives in the BLC

73

Table 11. - Voice of the et- stems and inner passives in the A T (1-7)

74

Table 12. - Voice of the et- stems and inner passives in Aph. Dem. 1

74

Table 13. - Voice of the et- stems and inner passives in Aph. Dem. VIII

74

Table 14. - Voice of the et- stems and inner passives in Aph. Dem. XVI

74

Table 15. - Voice of the et- stems and inner passives in Ephrem's Commentaries 74 Table 16. - Semantic classification of et- stems and inner passives in the BLC ....75 Table 17. - Semantic classification of et- stems and inner passives in the A T (1-7)

76

Table 18. - Semantic classification of et- stems and inner passives in Aph. Dem. 1

76

Table 19. - Semantic classification of et- stems and inner passives in Aph. Dem. VIII

77

Table 20. - Semantic classification of et- stems and inner passives in Aph. Dem. XVI

77

Table 21. - Semantic classification of et- stems and inner passives in Ephrem's Commentaries

78

Table 22. - Semantic categories of Syriac et- stems

79

Table 23. - Aspectual values of the et- stems in the BLC

82

Table 24. - Aspectual values of the inner passives in the BLC

83

Table 25. - Aspectual values of the et- stems in the A T (1-7)

83

vii

OUTLINE OF MIDDLE VOICE IN SYRIAC

Table 26. - Aspectual values of the inner passives in the A T (1-7)

83

Table 27. - Aspectual values of the et- stems in the Aph. Dem. I

83

Table 28. - Aspectual values of the inner passives in Aph. Dem. I

84

Table 29. - Aspectual values of the et- stems in the Aph. Dem. VIII

84

Table 30. - Aspectual values of the inner passives in Aph. Dem. VIII

84

Table 31. - Aspectual values of the et- stems in the Aph. Dem. XVI

84

Table 32. - Aspectual values of the inner passives in Aph. Dem. XVI

84

Table 33. - Aspectual values of the et- stems in Ephrem's Commentaries

85

Table 34. - Aspectual values of the inner passives in Ephrem's Commentaries

85

Table 35. - Actional value et- stems and corresponding bases in the BLC

89

Table 36. - Actional value et- stems and corresponding bases in the A T (1-7) ....89 Table 37. - Actional value et- stems and corresponding bases in Aph. Dem. 1

90

Table 38. - Actional value et- stems and corresponding bases in Aph. Dem. VIII

90

Table 39. - Actional value et- stems and corresponding bases in Aph. Dem. XVI

90

Table 40. - Actional value et- stems and corresponding bases in Ephrem's Commentaries

91

Table 41. - Actional value inner passives and corresponding bases in the BLC ....92 Table 42. - Actional value inner passives and corresponding bases in the A T (1-7)

93

Table 43. - Actional value inner passives and corresponding bases in Aph. Dem. 1

93

Table 44. - Actional value inner passives and corresponding bases in Aph. Dem. VIII

93

Table 45.- Actional value inner passives and corresponding bases in Aph. Dem. XVI

93

Table 46. - Actional value inner passives and corresponding bases Ephrem's Commentaries Table 47. - Combination of Voice and Action in the sample texts

94 96

Chart 1- Semantic distribution of the et- stems and of the inner passives in the BLC

98

Chart 2- Semantic distribution of the et- stems and of the inner passives in the A T (1-7)

98

Chart 3- Semantic distribution of the et- stems and of the inner passives in Aph. Dem. 1

98

LIST OF TABLES AND CHARTS

ix

Chart 4 - Semantic distribution of the et- stems and of the inner passives in Aph. Dem. VIII

99

Chart 5 - Semantic distribution of the et- stems and of the inner passives in Aph. Dem. XVI

99

Chart 6 - Semantic distribution of the et- stems and of the inner passives in Ephrem's Commentaries

99

Chart 7 - Aspect of et- stems and inner passives in the BLC

100

Chart 8 - Aspect of et- stems and inner passives in the A T

101

Chart 9 - Aspect of et- stems and inner passives in Aph. Dem. 1

102

Chart 10 - Aspect of et- stems and inner passives in Aph. Dem. VIII

103

Chart 11 - Aspect of et- stems and inner passives in Aph. Dem. XVI

104

Chart 12 - Aspect of et- stems and inner passives in Ephrem's Commentaries.... 105 Chart 13 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in the BLC

106

Chart 14 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in the A T

107

Chart 15 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in Aph. Dem. I

107

Chart 16 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in Aph. Dem. VIII

108

Chart 17 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in Aph. Dem. XVI

108

Chart 18 - Actional values of the et- stems and inner passives, correlated with those of the corresponding basic stems, in Ephrem's Commentaries

109

FOREWORD

T h e problem of determining the semantic value and the function of the Semitic verbal (-)i- affixes has long been debated throughout the history of Semitic Linguistics. In comparative studies, as well as within the philological traditions of the single languages, several attempts have been made at discovering a unifying meaning, accounting for all the different usages intended either as the original value, in a diachronic perspective, or as an omni-comprehensive semantic category. Although very different from one another, all the approaches to this matter share a kind of 'treasure h u n t ' conception in their search. T h e r e seems to be a quest for a key to the (-)i- affix, a single, most general or, on the contrary, most specific category that will unlock the secret and dissolve the mystery. However, along these lines no fully satisfactory perspective appears to have been provided so far. T h e present work describes the results of a research on the semantics of the Syriac verbal prefix et- and of the et- conjugations. Also the so-called internal passives and the pronominal reflexives are taken into account, to the extent by which they complete and determine the role of the et- conjugations. In my approach to the problem, I did not look for a unique answer or solution, nor do I claim, as a result of my investigation, to have discovered anything as such. Rather, I tried to examine the correlations existing among the numerous syntactic and semantic phenomena in which the Syriac et- prefix appeared to play a role. T h e outcome was a network of parameters, whose interaction defined the borders and internal articulation of the area of influence of this morphological category and of its closest competitors: inner passives and pronominal reflexives. Such a network I refer to as middle voice. T h i s study is subdivided into five sections: •





an introductory part, containing a description of the problem (both synchronically and diachronically), a brief status quaestionis and an overview of the general linguistic and typological theories and instruments that will be used in the course of the work (chapters 1-2-3); an analysis of a sample of early Syriac texts (dating between 3 r and 6 century) and a discussion of its results, also within a frame of typological comparisons (chapters 4-5); a presentation of the conclusions that can be drawn from the analysis (chapter 6);

xi

xii

OUTLINE OF MIDDLE VOICE IN SYRIAC





an appendix containing a digression on the way diathesis and et- forms and inner passives have been described in medieval Syriac grammars (Appendix 1); a second appendix containing some empirical data related to coreferential dative (Appendix 2).

T h i s book is a revision and an extension of my doctoral thesis, which was written during the three years (2005-2008) of Perfezionamento at the Scuola Normale Superiore of Pisa, where it was defended in July 2009. Margherita Farina Florence, July 2011

ACKNOWLEDGMENTS

T h e research the results of which are presented in this book was carried out in the Laboratorio di Lingüistica of the Scuola Nórmale Superiore of Pisa, directed by Pier Marco Bertinetto, Professor of General Linguistics at the Scuola and my advisor. T o him goes all my gratitude for offering me the opportunity of cultivating my scientific interests, improving my knowledge of Semitic languages, also spending long periods in foreign universities, and finally writing this dissertation. I thank Professor Pier Giorgio Borbone, who teaches Biblical Hebrew and Syriac at the University of Pisa, for guiding me during my studies, for giving me the possibility of long, open-minded and fruitful discussions on my research, and for continuously encouraging me to push the boundaries of my investigations out farther and farther, sharing with me his interest and curiosity for Syriac and Eastern culture, from the Syro-Palestinian coast all the way to Central Asia. I would also like to express my gratitude to all the people of the staff of the Laboratorio di Lingüistica of the Scuola Nórmale Superiore, who supported me, both technically and psychologically, through the years of my doctorate. A relevant portion of the research was carried out at the Peshitta Institute at Leiden University. I would like to thank Dr. Wido van Peursen, leader of the T u r gama Project of the Peshitta Institute, for guiding me during my study there, offering me the opportunity to discuss and elaborate my working hypothesis and also for the logistic support I received during my stay. I thank also all the other members of the Turgama Project. Especially, I thank Constantijn Sikkel, scientific programmer of the Turgama Project, who offered me technical support, together with great professionalism and endless patience, without which some of the analysis presented in this book would have not been possible in the short lapse of a lifetime. Indispensable corrections and observations for the improvement and enlargement of my research were given by the board of examiners, during the presentation of my dissertation, as well as during the revision process: I thank Prof. Riccardo Contini, Prof. Holger Gzella and Dr. Alessandro Mengozzi, for their generous remarks and precious bibliographical suggestions. I thank Ouafae Nahli, who teaches Arabic at the University of Pisa, for her indispensable support in enabling me to understand modern Arabic linguistic theories, as well as to read and interpret Arabic classical sources. I thank Beryl Turner, from the Withley College of the University of Melbourne, for correcting the text and giving me precious advises concerning the English: all the mistakes that might be left are entirely due to my responsability. Finally, I thank the Scuola Nórmale Superiore of Pisa, the University of Pisa and all the people (professors, colleagues, friends) whom I have met during my

xiii

xiv

OUTLINE OF MIDDLE VOICE IN SYRIAC

study, for giving me the opportunity to realize my greatest wish and deepest desire: to participate in an international circulation of ideas and perspectives on the human sciences.

ABBREVIATIONS

Alphabetic list of the grammatical glosses used in the illustration of the examples A = active abl. = ablative adj. = adjective adv. = adverb aux. = auxiliary verb C = consonant conj. = conjunction coref. dative = coreferential dative cs. st. = construct state D = intensive/causative stem with redoubling of the second root-consonant (Doppelungsstamm) dat. = dative def. art. = definite article dem. = demonstrative etp. = ethp'el etpa = ethpa'al ettaf. = ettafal G = basic verbal stem (Grundstamm) I = impersonal indc. = indicative ind. art. = indefinite article indp. pron. = independent pronoun inf. = infinitive mood inn. pass. = inner passive intg. = interrogative ipf. = imperfect (or prefix conjugation) ipv. = imperative M = middle marf. = marfu', Arabic nominative case mans. = mansub, Arabic accusative case neg. = negative obj. mrk. = direct object marker (sc. nota accusativi) P = agentless passive P+ = passive with agent expressed part. = particle pass. = passive

xv

xvi

OUTLINE OF MIDDLE VOICE IN SYRIAC

pf. = perfect (or suffix conjugation) prep. = preposition pres. = present tense pret. = preterite, Akkadian prefix conjugation pron. = pronoun pron. suff. = pronominal suffix ptc. = participle refl. = reflexive rel. = relative pronoun / particle (i.e. d-) V = vowel Journals quoted in the bibliography J A O S = T h e Journal of the American Oriental Society J N E S = Journal of Near Eastern Studies K U S A T U = Kleine Untersuchungen zur Sprache des Alten Testaments und Umwelt ZAH = Zeitschrift fur Althebraistik Z D M G = Zeitschrift für die deutschen morgenländische Gesellschaft

N O T E ON T R A N S L I T E R A T I O N

T h e transliteration of the Syriac language, unlike those of other Semitic languages, is not regulated by a unified and shared standard, especially as far as the vowel notation is concerned. T h e attempt at reproducing and respecting the rules of pronunciation often conflicts with the need for preserving the intelligibility of the words to a non-specialist. Moreover, the co-existence of the Eastern and Western traditions of pronunciation, forces the scholar to a choice that will necessarily appear arbitrary and unsatisfactory to a certain extent. T h i s work has linguistic scope and is meant to be accessible not only to the experts of the Syriac language, but also to linguists in general. Therefore, all the quotations from Syriac and from other Semitic languages have been transliterated and annotated with short grammatical glosses. For the same reason, attention was paid to make the root-consonants and the word's morphological structure as recognizable as possible. T h u s , phenomena such as the spirantization of the consonants b, g, d, k, p, t, have not been rendered in the transliteration. As far as the transliteration of the consonants is concerned, the common system of diacritics has been used, with the following exceptions: word-initial alaph is omitted, every time that it is vocalized, and it is thus marked by the corresponding vowel; word-final alaph is omitted: for example, the endings of the masculine singular emphatic state and feminine singular absolute state are transliterated as a, the ending of the masculine plural as e; vowel marked by mater lectionis is indicated by a caret; the spirantization of the consonants b g d k p t, in specific phonetic contexts, is not indicated in the transliteration; consonants that are not pronounced (except for matres lectionis) are always transliterated, in order to preserve relevant morphological elements (such as verbal endings); an underline is used to mark a linea occultans in Syriac writing.

1 See Carl Brockelmann, Syrische Grammatik (Leipzig: Veb Verlag Enzyklopädie, 1965). For a description of Syriac system of diacritical marks and accents, and of its evolution in the centuries, cf. Judah Segal, The diacritical point and the accents in Syriac (London: Cumberlege, 1953).

xvii

xviii

OUTLINE OF MIDDLE VOICE IN SYRIAC

Syriac

Transliteration

alphabet nf

IPA symbol

?

111

b

/b/

g a

/g/ /d/

C7D

h

/h/

O

w

/w/

\

z

/z/

JJ

h

/ii/

t

/tV

y k

M

l

IV

m

Imi

n

Ini

s

/s/

P s

/p/

>

A "p 00

1

/)/

/Y/ /sV /q/

I

q r

JE.

s

/J/

t

k!

J3

/r/

1

INTRODUCTION TO THE PROBLEM

1.1 SURVEY OF T H E LINGUISTIC MATERIAL T h e Syriac verbal system is organized a r o u n d t w o m a i n conjugations (or tenses, in m a n y traditional accounts) characterized by a lexical r o o t associated w i t h a n u m b e r of personal prefixes or suffixes, respectively. F o r this reason, one can also talk a b o u t prefix and suffix c o n j u g a t i o n s (p.c. and s.c.). T h e t w o conjugations differ, at least originally, 3 f r o m t h e aspectual p o i n t of view, b e i n g associated w i t h non-terminative (imperfective) and terminative (perfective) aspects, respectively. Besides t h e s e conjugations, Syriac also has a n u m b e r of forms, or stems, namely m o r p h o l o g i c a l patterns, according t o w h i c h t h e verbal r o o t can be m o d i fied, in order t o m o d i f y its basic lexical m e a n i n g . T h e difference between t h e basic stem and t h e derived ones lies mainly in t h e a r g u m e n t structure (causation, passivization, intransitivization, etc.) and, t o a certain extent, in aspectual representation and actional c o n t e n t . T h i s structure is c o m m o n t o several W e s t e r n Semitic languages f r o m t h e 1 s t m i l l e n n i u m B.C. onwards. Nevertheless, Aramaic (of w h i c h Syriac represents a 2 Cf. e.g. Brockelmann, Kurzgefasste Vergleichende Grammatik, § 168-172, Theodor Nöldeke, Kurzgefasste Syrische Grammatik (Darmstadt: Wissenschaftliche Buchgesellschaft, 1898 [1966]), § 158, Rubens Duval, Traité de grammaire syriaque (Paris: Vieweg, 1881), § 179-183. 3 For a sketch of the evolution of Aspect in Aramaic, cf. David Cohen, L'aspect verbal (Paris: Presses Universitaires de France, 1989). T h e Tense and Aspect system of Syriac is more complex than here indicated. Further indications will be given in ch. 2. Participles play an important role, being the main tool for expressing the Present, but also, in some contexts, expressing an on-going process. Moreover, several compound tenses have been developed, making use of the auxiliary hwâ 'to be' (for which cf., among others, Craig Morrison, " T h e Function of qtal hwa in Classical Syriac Narrative," in Biblical and Oriental Essays in Mamory of William L. Moran (ed. A. Gianto; Roma: Pontificio Istituto Biblico, 2005), 103-131; " T h e Function of qtal hwa in the Acts of Judas Thomas," in Aramaic in its Historical, and Linguistic Setting (ed. H . Gzella and M.L. Folmer; Wiesbaden: Harrassowitz, 2008), 257-285 and Stefan Bombeck, "Das syrische Verbalsystem: hypothetische Sätze und hwä vor Adjectiv oder Partizip," in Tempus und Aspekt in den semitischen Sprachen (ed. R. Nebes; Wiesbaden: Harrassowitz, 1999), 23-36.). In this study the values of some of the compound tenses and of the participles will be analyzed, but only to the extent to which it is relevant for the understanding of the functions of et-forms and inner passives. 4 For the use of the couple terminative/non-terminative to refer to aspectual distinctions, cf. Pier Marco Bertinetto, II dominio tempo-aspettuale. Demarcazioni, intersezioni, contrasti (Torino: Rosenberg and Sellier, 1997) and § 2.2.1 of this book.

1

2

OUTLINE OF MIDDLE VOICE IN SYRIAC

medieval literary5 sub-variant), in the course of time, has developed a remarkable symmetry within its stem-system, which can be considered a peculiarity, in comparison with all other Semitic languages, and which was most productive in Syriac. T h e verbal system of Syriac can be described on the basis of the so-called triliterism. Given a tri-consonantal root, the verbal derivation is realized through morphological operators such as internal apophony (a verbal stem can be represented by a vowel pattern), reduplication of root consonants, and affixation. On the other hand, the inflection is operated by means of prefixes and suffixes. T h e symmetrical stem-system of Syriac, e.g. of a root such as p'l ('to do'), can be sketched as follows (C1C2C3 represent the root-consonants and V a vowel): p'el pa'el (')ap'el sap'el

CiC2VC3 C1VC2C2VC3 'V-QC2VC3 SV-C1C2VC3

etp'el etpa"al ettap'al estap'al

et-CiC 2 VC 3 et C1VC2C2VC3 et-(t)V-CiC 2 VC 3 es-tV-CiC 2 VC 3

Table 1. — Syriac verbal system In the left column of this scheme are listed the basic stem and three derived ones, by means of the following derivational morphemes: • • •

strengthening of the second root-consonant (in the scheme "); causative prefix causative prefix s.

5 On the position of Syriac among the other Aramaic dialects, cf. Eduard Y. Kutscher, "Aramaic," in Linguistics in South West Asia and North Africa (ed. T.A. Sebeok; T h e Hague - Paris: Mouton, 1970), 346-412, Joseph A. Fitzmyer, A Wandering Aramean. Collected Aramaic Essays (Missoula: Scholars Press, 1979). On the origin of Syriac as a classical and literary language, cf. Lucas Van Rompay, "Some preliminary remarks on the origins of Classical Syriac as a standard language," in Semitic and Cushitic Studies (ed. G. Goldenberg and S. Raz; Wiesbaden: Harrassowitz, 1994), 70-89. For the character and literary status of Syriac, also with respect to other Aramaic variants in the late antiquity cf. David G.K. Taylor, "Bilingualism and Diglossia in Late Antique Syria and Mesopotamia," in Bilingualism in ancient society: language- contact and the written text (ed. J.N. Adams, J. Jaeger and S. Swain; Oxford: Oxford University Press, 2002), 298-476. On the linguistic context in which Syriac emerged as a literary language, cf. John F. Healey, "Variety in Early Syriac: T h e Context in Contemporary Aramaic," in Aramaic in its Historical and Linguistic Setting (ed. H. Gzella and M. L. Folmer; Wiesbaden: Harrassowitz, 2008), 221-229. 6 In this stem the prefix alaph Q is assimilated to the -t- of the et- prefix, which results in the latter being redoubled, cf. Duval, Traité, § 198. 7 T h e affix et-, when attached to a verbal root or a verbal stem beginning with a sibilant (s, s, s, z) is infixed between the sibilant and the following consonant, in stead of being prefixed. This rule, common to several Semitic languages, has been referred to as Metathesisregel. On its origin and distribution, cf. Werner Diem, "Die Entwicklung des Derivationsmorphems der t- Stamme in Semitischen," ZDMG, 132, 29-84 and the discussion in chapter 3 of this book.

INTRODUCTION TO THE PROBLEM

3

T h e table also shows how, to every stem in the left column there corresponds a stem in the right column, that is derived f r o m it by means of t h e prefix et-. T h a t is, t h e stems on the right column are characterized by the same morphological element as their counterpart on the left, plus t h e prefix et-. T h e definition of the precise value of t h e prefix et-, of its position within the Syriac verbal system and its relation w i t h its etymological equivalents in other Semitic languages, has been a matter of debate t h r o u g h o u t the history of Syriac grammatical and linguistic studies. 9 If all of the grammarians and linguists agree in indicating a roughly reflexive, reciprocal or passive meaning, they all fail to find a real unified account. Moreover, the exceptions, such as et- forms w i t h active or transitive values, are so numerous to suggest t h e need for f u r t h e r investigation. A n o t h e r morphological category that is correlated with the modification of t h e argument structure and of the actional value of the verb is that of the inner passive. A well known feature of Semitic languages is t h e above-mentioned internal apophony, namely t h e modification of t h e basic lexical meaning of a consonantal root by means of different vowel patterns. In Syriac, t h e productivity of such a p h e n o m e n o n is restricted with respect to more ancient Semitic varieties. One of its surviving traits lies in the inner passives, a n u m b e r of participles related to t h e various verbal stems: qtil, mqattal, maqtal, msaqtal, etc. It is important to notice that this is a morphological label, w i t h only a generic reference to an actual passive meaning of such a form. Finally, a third category that is correlated to reflexivity, and relevant for t h e understanding of t h e role of t h e et- forms, is t h e reflexive pronominal structures. T h e s e are forms constituted by a preposition or by nouns such as napsa 'soul,' qnoma 'person,' yata 'essence,' followed by pronominal suffixes. T h e s e suffixes are coreferential with the subject of t h e governing verb.

8 The prefix s can be described, in a general Semitic perspective, as a causative marker. Nevertheless, in Syriac, it is no longer productive and its functions and origins are not clear. For a complete list of its values in the different verbal roots and on the position of this stem in the Syriac verbal system cf. W. T h . van Peursen, "Inflectional Morpheme or Part of the Lexeme: Some Reflections on the Shaphel in Classical Syriac," in Foundations for Syriac Lexicography III. Colloquia of the International Syriac Language Project (ed. J.W. Dyk and W . T h . van Peursen; Piscataway, NJ: Gorgias Press, 2008), 41-57. 9

An overview of the literature on Semitic t- affix will be given in chapter 3. One could use, for example, the alternative formulation non-active participle, as in Amikam Gai, "The Signifié of the Non-Active Participle in Semitic Languages," ZDMG, 155, 9-23 and Tarsee Li, "Non-Active Participles in the Aramaic of Daniel," Aramaic Studies, 6, 111-136. 11 These nouns have a special status in the language system of Syriac. They have a certain degree of grammaticalization, but their literal meaning can always be restored (as is the case for the cognates of nps throughout the entire Semitic group); cf. in this respect also what is said in Leonid Kulikov, "The reflexive pronouns in Vedic: A diachronic and typo-

4

OUTLINE OF MIDDLE VOICE IN SYRIAC

W h a t follows is an overview of the main possible uses of t h e et- stems and of their apparent competitors: inner passives and reflexive pronominal structures.

logical perspective," Lingua, 117, 1414-1433, esp. 1414-1415 for Vedic reflexives. A very interesting example of the productivity of such an interchange between the grammatical and the lexical function can be found in Ephrem's Commentary on Genesis (ed. Raymond M. Tonneau, Sancti Ephraem Syri in Genesim et in Exodum Commentarii (Louvain: 1955), 17-18). T h e author is explaining the first verse of the book of Genesis, that in Biblical Hebrew is: '¡'""iNn nx 1 DTJUTI riN DTlbN N""Q rPWsn ba-re'sit bärä elohim et-ha-ssämaym w-et hä-äres "in the beginning God created the sky and the earth". Ephrem reports the entire verse, in a sort of Syriac translation-transcription that we reproduce here: k;/ Lxi». b-räsit lam brä alähä yät smayyä w-yät ara. What is interesting here is the word transcribed as yt. In the Hebrew text it corresponds to the direct object marker et (and could have been thus, in principle, translated with the Syriac equivalent /-). In Syriac the word yät exists, with the meaning of'essence, substance.' This word is usually considered to be a cognate of the Hebrew et (for which cf. Meir Max Bravmann, Studies in Semitic Philology (Leiden: Brill, 1977), C. Correll, "Ein neuer Anlauf zur Erklärung der Herkunft der 'Notae Accustativi,'" in Festschrift Ewald Wagner zum 65. Geburtstag. 1 (ed. W. Heinrichs, G. Schoeler and E. Wagner; Beirut, Stuttgart: Orient-Institut der Deutschen Morgenländischen Gesellschaft. In Kommission bei Franz Steiner, 1994), 22-43, David Testen, "Morphological Observations on the stems of the Semitic 'nota accusativi'," Archiv fur Orientforschungen, 44-45, 215-221 and Margherita Farina, "nx e n x in ebraico biblico: nota accusativi e preposizione," Egitto e Vicino Oriente, 18, 245-253) and it was chosen here as the best rendering of the particle. But how is it then explained by Ephrem in his commentary? Does the Syriac word receive grammatical status? This is what Jessie Payne Smith, A compendious Syriac Dictionary. (Oxford: 1903), 198 seems to suggest: "archaism used like Heb. eth as sign of the accusative." Nevertheless, what follows in the Commentary goes in another direction and exploits the ambiguity differently: Tonneau, Sancti Ephraem, 8, 17-18: kj/, ¿»aioo U i , oj^ooio ojoi hänaw den qnömeh da-smayä wa-qnomeh d-ara "and this means the essence of the sky and the essence of the earth." Here a grammatical word from Biblical Hebrew is rendered with a Syriac equivalent that has, nevertheless, prominently a lexical meaning and can be interpreted as a noun, 'essence' (this meaning is quite widespread, for example, in later translations of Greek philosophical and scientific texts). This word is thus interpreted by means of another word (qndmä) that has the same status: it is potentially also a grammatical (pronominal) function, but is here used in its lexical meaning 'essence' (this is confirmed by what follows in the text). For the origin of yät as a particle, cf. the diverging hypothesis of Nöldeke, Kurzgefasste, § 223 and Brockelmann, Syrische Grammatik, § 106c, recently discussed in Aaron D. Rubin, Studies in Semitic Grammaticalization (Winona Lake, Indiana: Eisenbrauns, 2005, 121-122. What has just been observed over Ephrem's interpretation of yät could be taken as an argument in favor of Nöldeke's theory, although the perception of the native speaker does not give information per se about the origin and the linguistic status of a given phenomenon. 12 E.g., nezke napseh 'he will justify himself.' In an analogous manner (in Syriac as well as in many other Semitic languages) also pronominal reciprocals are formed. In Syriac reciprocity is expressed through reduplication of the numeral had 'one' or by using the subst. ahä 'brother,' followed by the pronominal suffix, often under Hebrew influence. Reciprocity and reflexivity are closely related in many world languages, cf. in this respect Suzanne Kemmer, The Middle Voice (Amsterdam - Philadelphia: Benjamins, 1993), 96-108, Gilbert Lazard, L'actance (Paris: Presses Universitaires de France, 1994), 49-52.

INTRODUCTION TO THE PROBLEM

5

Middle-reflexive13 Et-

form

1. G e n 3 8 : 1 4 ( P e s i t t a f 4 w-etkassyat conj. 'and' + III f. sg. pf. etpa 'cover'

b-ardida w-estabbtat prep, 'in, with' + conj. 'and' + III f. sg. pf. 'veil' etpa 'adorn oneself

" S h e c o v e r e d h e r s e l f w i t h a veil a n d a d o r n e d h e r s e l f . " Inn.

pass.

2. Gen 37:23 (Pesitta) Joo) « .-i\ , JiS^'jS; kutina 'robe'

d-pedyata da-Ibis hwa rel. + pi. 'long sleeves' rel. + pass. ptc. m. sg. Ill m. sg. pf. p'el p'el 'wear, be clothed' 'be'

" T h e long-sleeved robe that he was wearing." Refl.

pron.

3. B o o k o f t h e L a w s of t h e C o u n t r i e s 4 6 , 2 3 1 5 tba III m. sg. pf. p'el 'revenge'

napseh 'soul' + pron. suff. III m. sg.

"He revenges h i m s e l f "

13 For the moment, the generic expression middle-reflexive is used, to give a first sketch of the phenomenon. Further in this chapter, the exact meaning of the two terms will be discussed, and the point will be made that for et- forms it is more appropriate to use the label middle, reserving reflexive only for the pronominal structures. 14 T h e text of the Peshitta of Genesis refers throughout to the Leiden Edition: T a e k e Jansma and Marinus Koster (ed.), Peshitta. The Old Testament in Syriac. Part I Fasc. 1. Preface. - Genesis; Exodus (Leiden: Brill, 1977). 15 T h e source text and the page numbers are those of Hendrik Jan W . Drijvers, The Book of the Laws of the Countries. Dialogue on Fate of Bardaisan of Edessa (Assen: Van Gorcum and Comp. N.V., 1965).

6

OUTLINE OF MIDDLE VOICE IN SYRIAC

Impersonal

Et- form 4. Ephrem's Commentaries in Genesis and Exodus 11,5 1 6 JjlX -i^li" etbri

Ihen

'nane

halen

III m. sg. pf. etp.

prep, 'to'/obj.

'clouds'

dem. adj.

'create'

mrk. + pron. I l l f.pl.

lit. "It was created those clouds." > " T h o s e clouds were created."

Inn. Pass. 5. Acts o f Judas T h o m a s VII, 240, 8 1 7 wa-btil

lak

conj. 'and' + pass. ptc. prep, 'to' + pron. m. sg. 'care'

'al

baryätä

prep, 'upon'

pi. 'creature;

II m. sg.

"And you are worried about the creatures." Passive 18

Et- form 6. A T VI, 2 3 1 , 8 ^o Iboioo w-netmle conj. 'and' + I pi. ipf. etp. 'fill'

men prep, 'from/by'

tarbiteh 'nourishment' + pron. suff. Ill m. sg.

" W e will be filled by his nourishment."

16

From, now on, text and references according to Tonneau, Sancti Ephraem.

From now on (AT), text and references according to William Wright, Apocryphal Acts of the Apostles, I. The Syriac texts (London: Williams and Norgate, 1871). 17

The value of the term 'passive,' for Semitic languages in general and for Syriac in particular, will be discussed in detail in ch. 3. For the moment we indicate, with this label, a verbal form of which the subject is undergoing the action described, by intervention of an agent, that can be expressed in the sentence, introduced by a preposition (such as men, h~, I-, byäd...).

INTRODUCTION TO T H E PROBLEM

Inn.

7

pass.

7. A T V, 214, 18 iSoi" j ^ . OlA^O menneh ilid ant prep, 'from/by' + pron. Ill ptc. pass. m. sg. p'el pron. II m. sg. m. sg. 'generate' " F r o m h i m y o u were g e n e r a t e d . " A c t i v e intransitive Et-

form

8. Ex 3:5 (Pesitta) 1 9 jzhoLL JJ Id 'not'

tetqarrab l-harkd II m. sg. ipf. etpa prep, 'to' + adv. 'here' 'approach'

" D o n o t c o m e close." Inn.

pass.

9. A T I, 182, 11 w-eskah conj. 'and' + III m. sg . of. 'find'

ennon pron. Ill m. pi.

kad adv. 'when'

yatibin ptc. pass. pi. p'el 'sit'

"And he found t h e m sitting." A c t i v e transitive Et-

form

10. E p h r e m ' s c o m m e n t a r i e s o n Genesis a n d E x o d u s , 60,14 ooo) ^-ojiioo v po).-'i\\ JJ, d-la rel. part. + neg.

'aryawata 'lions'

l-abayhon prep, 'to'/obj. mrk. + 'forests' + pron. Ill m. pi.

metdakkrin ptc. etpa. m. pi. 'remember'

hwaw Ill m. pi. pf. 'be'

" T h e lions d o n o t r e m e m b e r t h e i r forests." 19

T h e text of the Peshitta of the book of Exodus refers throughout to the Leiden Edition (1977): Jansma and Köster, Peshitta. The Old Testament in Syriac.

8

OUTLINE OF MIDDLE VOICE IN SYRIAC

Inn. pass. 11. A T II, 193, 18 JjO.JjO Joo) y^/o w-ahid

hwä

qeryona

conj. 'and' + ptc. pass. m.

Ill m. sg. pf. 'to be'

'candle'

sg. 'hold, carry'

"And he was holding a candle." 1.2 PRELIMINARY OBSERVATIONS T h e examples presented in the previous paragraph show, on the one hand, that a certain variety of meanings can be expressed by the same morphological element. On the other hand, analogous meanings seem to be expressed by different forms. But what is the property that binds together all these different meanings? Are all those forms mere equivalents, or is it possible to individuate different specializations for them? In order to try to give an answer to these questions I will reconsider the distribution of et- forms, inner passives and reflexive markers in the light of the concept of middle voice. T h e leading principle of my analysis is based on the following statement formulated by Kemmer: Recurring instances of different meanings being expressed by the same formal or structural means is an indication that the meanings in question are related. Furthermore, the more direct the semantic relationship between two meanings, the more likely they are to be subsumed under a single form of expression, both within and across languages.20 T h e situation of Syriac is not isolated among the Semitic languages. On the contrary, a morphological connection between intransitive/stative, reflexive and passive is characteristic more or less of the entire language family. In his monograph on Semitic diathesis, Retso investigates the historical origin of two verbal categories proper to West Semitic dialects: the causative conjugations of the type yuqtil-h/'aqtal and the inner passive of the type yuqtal-qutil/qatil. As the subtitle indicates, the book contains mainly a comparative morphological study. Nevertheless, in the introductory section, as well as in various parts of the account, a number of observations are found that are relevant for the present study. Retso starts his analysis of the various ways of expressing the passive in Semitic languages with the following remark: "all morphological devices in Semitic

20

Kemmer, The Middle Voice, 4. Jan Retso, Diathesis in the Semitic languages: a comparative morphological study (Leiden: Brill, 1989). For a critical discussion of the material presented by Retso cf. ch. 3. 21

INTRODUCTION TO THE PROBLEM

9

occurring in PC 22 have other functions as well." Trying to understand the reason for such polyfunctionality, the scholar indicates an important, although very general, correlation between intransitive, stative and passive verbs: In Semitic at least intransitive marking is widely spread for marking passive subjects (derivative verbs with t- affix). T h e reason for this is not difficult to understand: one characteristic of the verb in a PC is intransitivization [...]. In Semitic there is another category that plays a crucial role in the PC, viz. the stative [...]. It is clear that the reason for this must be a semantic feature which the subject of a stative verb shares with the subject in a PC: non-agentivity. 23

According to Retso, thus, the expression of passive is realized in Semitic languages by means of an analogical extension of two properties: argument reduction, on the syntactico-semantic level, and non-agentivity, on the semantic one. As will be shown in the course of this study, the analogical extension of semantic and syntactic properties, on the basis of metonymic and metaphoric associations, plays a very important role in the distribution of forms and functions in Syriac middle voice. Another important contribution given by Retso is a discussion about the nature of the passive construction in Semitic languages that he analyses (among which is also Aramaic). A well known feature of Semitic languages is the fact that the agent of a passive construction is very seldom expressed. For this reason, passivization has been viewed in these languages mainly as an agent-demotion device. Nevertheless, in many languages of this family, and certainly in all of them from the first millennium onwards, there is the syntactic possibility of expressing the agent of a passive construction by means of prepositional phrases. This second type of construction is considered as a topicalization of the object. Retso proposes the following reconstruction: T h e PC, originally perhaps a device for handling the absence of agent, thus gives the languages a means for grammaticalization of topicalization. T h u s the PC in Semitic should probably be seen as two different constructions, (1) deletion or absence of the semantic subject (= the "agent") for different reasons, (2) topicalization of the object. 25

22 23

Sc. passive construction. Retso, Diathesis, 6.

24

Cf. also Retso, Diathesis, 28: "It is however essential to make the distinction between syntactic and lexico-semantic categories. It is perhaps reasonable to agree that the term stative is apt to design a lexico-semantic category found in all languages. [...] T h e other terms involved here, viz. the transitive, intransitive, active and passive, should likewise be seen as primarily indicating syntactic categories connected with the phenomena of valency and diathesis." 25

Retso, Diathesis, 3.

10

OUTLINE OF MIDDLE VOICE IN SYRIAC

As we will see in ch. 4, t h e overt expression of t h e agent is a possible criterion for classifying different constructions involving t h e same verbal form, at least to a certain extent. If t h e presence of t h e agent is certainly not indispensable for a construction to be passive, we can consider it as an important ingredient for establishing a scale of passivity and a scale of thematic relevance, among the various passive structures. Individuating formal and semantic criteria for classifying the different uses of t h e same verbal form is important, in order to identify the prototypes around which t h e verbal system is organized, and on t h e basis of which all analogical extension of forms and functions is operated. In t h e following chapter, definitions will be given of reflexive and middle, that will constitute t h e guidelines for my textual analysis. 1.3 E T - FORMS IN SOME M O D E R N SYRIAC GRAMMARS

Let us briefly review what is said about t h e verbal forms with et- prefix in some of t h e main Western grammars of classical Syriac. Duval does not treat the meaning of et- forms separately. Nevertheless, reference to it is made in several instances. T h e first instance is in § 184: Ainsi a disparu du verbe le passif interne, dont les participes et quelques noms d'action ont seuls conservé des vestiges; l'araméen lui a substitué les réfléchis. 26

In ch. 5 Des verbes dérivés, § 190-191 is said: Les dérivés du verbe simple ou radical sont au nombre de sept: trois actif pa'el aph'el et saph'el; quatre réfléchis, formant le passif du verbe simple et des trois actifs dérivés, dont ils se distinguent par les voyelles et la particule des réfléchis i i [et]. [...] Le passif du verbe simple ^¡¡ol/ [etkteb]. [...] Ethp'el a encore quelquefois le sens réfléchi,

jtapaitéo|iai.

[est'el] se faire prier, refuser

28

T w o terms are used here to refer to the et- forms: réfléchi and passif. N o special indication is given about t h e way such labels should be understood. However, it seems that the first term is used to indicate t h e morphological category in general and makes reference to an original function, whereas t h e second defines t h e current semantic value of t h e form. In order to understand what exactly is meant by sens réfléchi, one can look at t h e examples given by Duval in § 193 for etpa'ah se rendre intelligent, comprendre, s'étonner, admirer, se rendre compte, réfléchir etc.

26

Duval, Traité, 176.

27

An analogous observation is found in Duval, Traité,

28

Duval, Traité, 178.

§ 193 for etpa'al.

INTRODUCTION TO THE PROBLEM

11

Rather than real reflexive actions (transitive events in which the agent and the patient are coreferential), they describe events experienced by the subject. In Nöldeke the stems with et- prefix are referred to as Reflexiv of the respective basic or derived stem. As far as the meaning is concerned, the following remark is added: "Die Reflexiva haben meist Passivbedeutung gewonnen." 29 Brockelmann gives the following general rule: "zu jedem dieser Stämme bildet es ein Reflexiv, jetzt, meist mit passiver Bedeutung, mit dem Praefix U." Nöldeke's and Brockelmann's observations, although more synthetic, are along the same lines as those found in Duval: the et- prefix had an original reflexive meaning (whatever is meant by this term) and has subsequently developed a passive one, that has taken over the primary function. A slightly more detailed discussion can be found in Muraoka: T h e semantic or functional opposition between these six patterns is still a matter of debate. T h e three Eth-prefixed patterns are partly reflexive, passive or ingressive [...]. T h e passive voice is partly indicated by the ¿í/j-patterns. 31

Muraoka also gives some interesting information about the relation existing between the et- participle and the inner passive: Where an inner passive participle is attested side by side with an external, ethprefixed one, the former stresses a result, the latter a process.

Such an observation introduces aspectual and actional values of the form, that will be extensively dealt with in the present study. What is important, for the moment, is to notice how the remarks mentioned so far, although all correct in principle, turn out to be rather vague and not completely clearly related to one another. Moreover, they often apply only to a subset of the et- forms, from which they try to derive a general/generic rule. Let us now focus on the reflexive value that is attributed to the et- forms. If one looks at the examples mentioned above, that Duval33 gives for such a usage, one is struck by the fact that very few of them actually correspond to what is generally meant by the term reflexive (see infra). Moreover, all the authors that were quoted seem to distinguish between a formal label reflexive (referring to some kind of morphological correspondence between the et- stems and the basic ones), and

29

Noldeke, Kurzgefasste, 101-102, § 159.

30

Brockelmann, Syrische Grammatik, 82.

31

Takamitsu Muraoka, Classical Syriac. A Grammar with a Chrestomathy. Second, Revised Edition (Wiesbaden: Harrassowitz, 2005), 41-42. 32 Muraoka, Classical Syriac, 42. 33 Brockelmann and Noldeke do not give examples related to meaning, but only forms relevant for the understanding of the paradigm. All the examples given by Muraoka, Classical Syriac are passive.

OUTLINE OF MIDDLE VOICE IN SYRIAC

12

t h e semantic value, t h a t is rather located in t h e area of t h e passive. 34 A s m e n t i o n e d in § 1.1, Syriac has a n o t h e r way of expressing reflexivity strictu senso, by means of specific p r o n o u n s or p r o n o m i n a l structures. Observing some of t h e examples given in t h e various grammars for these categories, we see h o w they qualify as reflexive structures m u c h better t h a n t h e etforms: loot -¿at «¿aio£. [la-qnûmeh lân^oo

[dhî mdabbrâ

âmar

napsâh wa-mtaksâ

hwâ] "sprach zu sich", l-yâtâh]

¿»isu Jbpo -5),

"welche sich selbst f ü h r t u n d

ordnet"; 3 5 oiiz >ol Jj5 [nâsê 'am yäteh] "streitet m i t sich selbst"; 3 6 otAsu ¿oL. [hnaq napseh] "il s'étrangla",

^ o i o [qnûmy ettel] "je m e donnerai." 3 7

A s regards t h e definition of reflexivity t h a t was current at t h e t i m e of t h e Syriac grammars here taken into account, this is w h a t L e o n a r d Bloomfield says of reflexivity, in a m o m e n t n o t exceedingly distant in t i m e : In English we say he washed him when actor and goal are not identical, but he washed himself (a reflexive form) when they are the same person. 38 O n t h e one h a n d , one could simply assume t h a t t h e terminology related t o r e flexivity was extended t o t h e et- stems in a sort of compromise, as t h e closest linguistic category at h a n d . O n t h e o t h e r h a n d , this may n o t appear to be t h e m o s t obvious solution, nor does it actually follow a generalized t r e n d in t h e grammatical descriptions of its time. 3 9 A s far as medieval Syriac grammars are concerned, t w o m a i n approaches t o t h e description of t h e et- stems can be observed: either it is indicated as t h e means for expressing passivity {m'abdânûtâ), or is it simply considered f r o m t h e m o r p h o 34

Cf. supra Duval: "quatre réfléchis, formant le passif du verbe simple et des trois actifs dérivés"; Nöldeke: "Die Reflexiva haben meist Passivhedeutung gewonnen;" Brockelmann: "zu jedem dieser Stämme bildet es ein Reflexiv, jetzt, meist mit passiver Bedeutung." 35 Nöldeke, Kurzgefasste, 168-169. Brockelmann, Syrische Grammatik, 108. 37 Duval, Traite, 296. 38 Leonard Bloomfield, Language (Chicago: Chicago University Press, 1933). 39 Cursorily, it can be observed that in Georg Curtius, Griechische Schulgrammatik (Prag: Tempsky, 1875 ), one of the most popular didactic grammars of Classical Greek of its time, the terms Reflexivpronomen and reflexiven gehrauch occur only twice (p. 75, § 210, referring to pronominal reflexives). In connection with verbal morphology, only terms related to middle and passive are found, and the term Medium, with reference to verbal diathesis, is found 14 times. On the diffusion of Curtius' grammar and on its role in the elaboration of 19 century linguistic terminology for verbal semantics, cf. Mario Squartini, "Metalinguaggio traduzione e il 'raffreddamento' della terminologia," in Per le vie del mondo (ed. P. De Gennaro; Torino: Trauben, 2009, 258-266 which was expanded in Squartini's communication to the congress of the Società di Linguistica Italiana, Verona 26-28 September 2009: "Tradurre il metalinguaggio prima e dopo i Neogrammatici: il caso della terminologia tempo-aspettuale." Cf. e.g. the translation of the Greek Téchne Grammatiké by Joseph Huzaya, or the quaestiones by Elias of Tirhan (11 cent.).

INTRODUCTION TO THE PROBLEM

13

logical point of view, as a formal counterpart of the other basic and derived stems. 41 No third alternative is given and no reference is made to reflexivity. 1.4 MUTÂWA'A

AND 'MIDDLE' VERBS IN ARABIC LINGUISTICS

On a purely speculative basis, one could look also at another concept, differently connected with verbal diathesis and belonging to the Arabic linguistic tradition: that of mutâwa'a. T h e word mutâwi', used for the first time by Sïbawayhi in his Kitâb, literally means 'submissive, obedient' 43 and refers to the verbal stems built with ta- and nprefixes and with -t- infix. According to Fleisch, who gives a thorough explanation of the usage of the term by Arab grammarians, mutâwi' has the following functions: •



to underline the 'aspectual' resultative value of certain verbal forms. In the sentence ^ [kassara-hu fa-takassara] "il le brisa et il fut en pièces", the ta- stem expresses the result of the action of the corresponding intensive stem, with a meaning that Fleisch calls résultatif. to describe, in general, the meaning of some derived stems, as compared with their corresponding 'basic' ones: tafa'ala, tafaala, infa'ala, ifta'ala. In this case Fleisch quotes as-Sïràfï "le sens de notre dire ^ j l k . [mutâwi'hi] est que le complément d'objet direct revient à celui qui est en jouissance de ce que désirait l'agent." In other words, the subject of, e.g., tafa'ala is coreferential with the direct object o f f a ' a l a .

What is important, in order to understand the value of the term mutâwi', is that the two functions described here are not independent from one another, but, on the contrary, make reference to the same idea: the subject of the verb allows for an action to take place, the consequences of which it happens to undergo. In the Western scholarly tradition, the second component of the meaning of mutâwi' has often been interpreted as the Arabic equivalent of the Western reflexive. At the same time, in all the main descriptions of the Arabic language, from the 41

As it is the case, for example, in Bar Zu'bî (beginning of 13 th cent.) and in Barhebraeus (13 cent.), who considers the et- stems as schemes, morphological constructions. Cf. Margherita Farina, "Diathesis and Middle Voice in the Syriac Ancient Grammatical Tradition: T h e Translations and Adaptations of the Téchne Grammatiké and the Arabic Model "Aramaic Studies, 6, 175-193 for a detailed study, and Appendix 1 in this book for a short resume. 42

Ch. 446. Or, according to Henri Fleisch, Traité de philologie arabe. Pronoms, morphologie verbale, particules. (Beyrouth: Dar el-Machreq, 1979), 305 "celui qui correspond à, s'accorde 43

V

avec . 44 45

Fleisch, Traité de philologie, 181. Sarh ms. Le Caire, Ve Partie, folio 92.1.7, cf. Fleisch, Traité de philologie, 305.

14

OUTLINE OF MIDDLE VOICE IN SYRIAC

19 th century up to today, the forms with ta- prefix (V and VI forms, respectively) are described as the reflexive of the II and III forms. T h i s correspondence was observed and analyzed by Larcher, who noticed a tendency in the modern Western scholars to identify the concept of mutâwa'a with reflexivity. However, in spite of a certain semantic contiguity (and, sometimes, even an overlap), the two linguistic categories stem from very different perspectives, as clearly emerges if one compares Bloomfield's above-mentioned definition of reflexivity with what Fleish says about the mutâwa'a. Particularly significant, in this respect, is the following remark by Joiion (1930): la nuance du «moyen» grec [...] est, par contre, parfaitement sensible dans de nombreux verbes sémitiques aux conjugaisons «réfléchies», particulièrement en arabe à la conjugaison ifta'ala.47

According to Larcher, the term reflexive is used in modern western Arabic grammar in a rather specific sense, as a label, to refer to the morphological correlation existing between the verbal stems: la réflexivité des grammairiens arabisants, est d'abord une notion morphologique: les formes VIII, V, VI et X sont dites réfléchies de I, II, III et IV comme formes augmentées en [t] de celles-ci, même si sous l'influence de la mutâwa'a des grammairiens arabes, la notion de réflexivité a été étendue à la VII par rapport à la I. La mutâwa'a des grammairiens arabes est d'abord une notion sémanî 48 tique r[...J.

Coming back to Syriac, it is possible to notice a certain analogy between the way the derived et- stems are described by Brockelmann, Duval and Nôldeke and the way the derived (-)t(a)~ stems are described in modern Arabic grammar, according to Larcher. Rather than suggesting an influence of one field of study upon another (which could not be proved), I would like to draw attention to the similarity in perspectives on the connection between reflexivity and -t- affixes that Syriac and

46

Pierre Larcher, Le système verbal de l'arabe classique (Aix-en-Provence: Publications de l'Université de Provence, 2003). Cf. the following remark on p. 69: "La seconde corrélation établie par les grammairiens arabes est la mutâwa'a. Celle-ci est souvent confondue avec la grande corrélation établie par les grammairiens arabisants, celle de réflexivité." 47 Paul Joùon, "Sémantique des verbes statifs de la forme qatila (qatel) en arabe, hébreu, arameéen," Mélanges de l'Université Saint-Joseph, 15, 3-32.: 4. T h e passage, in its broader context, refers mainly to the so-called stative conjugations, namely specific vowel patterns associated, in many Semitic languages, with intransitive meanings. However, a few pages after, Joiion states: "En sémitique, cette nuance du réfléchi, c'est dans les formes réfléchies qu'on la trouve, et parfois très légèrement exprimée, comme dans beaucoup de verbes «moyens»." (p. 29). 48

Larcher, Le système verbal, 70.

INTRODUCTION TO THE PROBLEM

15

Arabic Western studies show. The aim of this digression is, in fact, to show how, in the history of Syriac studies, the attribution of a precise function to the et- prefix and its connection with reflexivity, was not a straightforward and rudimentary identification process. On the contrary, it was the result of a (rather arbitrary) compromise and an intermingling of different suggestions and perspectives on the language, of which it is important to be aware, in handling labels and grammatical terminology. I quoted Joiion above, on the possibility of a middle nuance of the Arabic tforms. The broader context of that statement is a study over the West Semitic 49 qatila / qatel verbal patterns, namely verbal patterns that display, in a relevant proportion, an intransitive-stative meaning. Among these verbs are, for example, the Arabic sami'a 'to hear,' 'amila 'to work,' sariba 'to drink,' labisa 'to wear'; the Hebrew lâbes 'to wear,' kâbed 'to be heavy,' sâken 'to dwell'; and the Syriac Ibes 'to wear,' dhel 'to fear' qreb 'to come near' etc. Joiion analyzes the semantic value of these verbs in various Semitic languages, trying to individuate a common feature. His conclusion is that all the qatila verbs express either a state or an event that is conceived from 'less active' up to definitely passive. The key of this explanation is in a scalar conception of the semantic categories: "Mais la nuance d'«activité» varie beaucoup d'un verbe à l'autre. Il y a assurément moins d'«activité» dans dormir que dans courir, dans recevoir que dans i

„50

prendre... Although Joiion thinks that these verbs should not be assimilated to the Greek middle verbs, their semantics are in fact closely related to that of middle voice, as examined by Kemmer, and should certainly be analyzed as such. As far as Syriac is concerned, this class is no longer productive and moreover has been made opaque by phonetic changes, therefore in this study it will not be specifically dealt with. Nevertheless, one more observation is necessary, as regards the qatila verbs. In modern Arabic grammar they are sometimes referred to as a middle pattern. This term is meant to indicate a verbal class which is not purely 'active,' like the qatala verbs, nor clearly 'stative,' like the qatula verbs, but shares semantic properties with both classes and is thus to be considered as being intermediate. on sera tenté de comprendre le dédoublement d e f a ' i l a en un verbe d'action et en un verbe d'état, certes à partir d'un unique résultatif, mais par l'intermédiaire de la syntaxe: fa' ila est interprété comme verbe d'action quand il est transitif direct et comme verbe d'état quand il est intransitif, l'emploi comme transitif indirect représentant le même état "intermédiaire" sur le plan syntaxique que l'interprétation comme "moyen" sur le plan sémantique. 5 2

49

Arabie in primis,

Hebrew and Aramaic to a minor extent.

50

Joùon, "Sémantique des verbes statifs," 5.

52

Larcher, Le système

Cf. e.g. Larcher, Le système verbal,

verbal, 24.

24.

16

OUTLINE OF MIDDLE VOICE IN SYRIAC

Therefore, although closely related to middle semantics and also to the original Greek concept of middle diathesis, this terminology does not fully correspond to the modern definition of middle. T h e Arabic qatila verbs (as well as their Semitic cognates) do not represent middle voice tout court, but, rather, are included in it.

53

For a different and more comprehensive overview on Arabic middle verbs, cf. Zouhair Maalej, "Middle Verbs," in Encyclopedia of the Arabic Language and Linguistics (ed. C.H.M. Versteegh; Leiden: Brill, 2008), 224-22.

2

MIDDLE VOICE AND SOME LINGUISTIC

CORRELATIONS

2 . 1 MIDDLE VOICE T h e i m p o r t a n c e of t h e category of middle voice was u n d e r l i n e d in t h e previous chapter. L e t us n o w give some preliminary definitions, in order t o clarify t h e p e r spective of this study. A s regards voice, I refer to t h e following definition by Shibatani (1993), in t h e I n t r o d u c t i o n to t h e b o o k : Voice is to be understood as a mechanism that selects a grammatically prominent syntactic constituent - subject - from the underlying semantic functions (case and thematic roles) of a clause.54 T h e r e f o r e , in this view, voice is a p h e n o m e n o n related t o t h e syntaxsemantics interface and is correlated t o w h a t is considered t o be t h e grammatical subject in a given language. T h e use of t h e t e r m middle, in this context, derives f r o m t h e ancient G r e e k grammatical tradition, in w h i c h it designated a class of verbs t h e m e a n i n g of w h i c h could be b o t h active and passive, d e p e n d i n g o n t h e circumstances. A classic e x a m ple is f o u n d in t h e chapter o n verb ( j t e p i prjjiaxoc;) in t h e Techne Grammatike, controversially a t t r i b u t e d t o Dionysios T h r a x :

54 Masayoshi Shibatani (ed.), Passive and Voice (Amsterdam - Philadelphia: Benjamins, 1988), 3. 55 T h e way this definition is phrased works essentially for nominative-accusative language type, to which Syriac belongs. A definition along the same lines, but focusing rather on the relational aspect of diathesis (in this case), is given in Retso, Diathesis, 622: "Diathesis can be defined as the syntactic relationship between the verbal core of a sentence and its nominal constituents, i.e. the verbal predicate and the parts of speech directly relating to it, mainly subject and object. T h e relationship encompasses both semantic and morphosyntactic categories." 56 For the problem of the attribution of the Téchne, see Vincenzo Di Benedetto, "Dionisio Trace e la Techne a lui attribuita," Annali della Scuola Normale Superiore di Pisa. Lettere, Storia e Filosofìa. Serie 2, 27, 169-210; "Dionisio Trace e la Techne a lui attribuita," Annali della Scuola Normale Superiore di Pisa. Lettere, Storia e Filosofìa. Serie 2, 28, 87-118; "La Techne spuria," Annali della Scuola Normale Superiore di Pisa. Classe di Lettere e Filosofìa. Serie 3, 3, 797-814; Vivien Law and Ineke Sluiter (eds.), Dionysius Thrax and the Techne Grammatike (Miinster: 1995).

17

18

OUTLINE OF MIDDLE VOICE IN SYRIAC

SiaBeoeig e lai tpelg evepyeia, JtàBog, iieaótrig' evepyeia [lèv otov TUJtta), JtàBog 8è otov TUJTtoum, [leaotrig, 8è f| Jtotè [lèv èvépyeiav jtotè 8è JtàBog J t a p i a t a a a , otov jtéjtriya 8iécp8opa, èjtoirià|iriv, ¿Ypm|)à|iTiv. Diatheses are three: active, passive, middle. Active is like TDJtTCO (I hit), passive is like TDJtTO|iai (I am hit), middle, then, is what is sometimes expressing active and sometimes passive, like JteJtT]ya (I was fastened), 8iecp8opa (I destroyed), èjtOLTlà|xr|V (I was made), èypm|)à|lTlV (I was written). 57

T h e origin of this grammatical category (^£OÓTT](;) is debated. It could be related to the stoic linguistic doctrines. From this Greek definition it is clear that, originally, middle was meant as a semantic category, rather than a morphological one. It would be interesting to understand how and when it has become, at least in Greek, the label for an inflectional category: the verbal - ^ a i conjugation. 59 Unfortunately, this goes beyond the scope of the present study.

57 For a detailed analysis of this passage, also with respect to its Syriac translation attributed to Joseph Huzaya (6 cent.), cf. Adalbertus Merx, "Historia artis grammaticae apud Syros," Abhandlungen fur die Kunde des Morgenlandes, 9; Riccardo Contini, "Considerazioni interlinguistiche sull'adattamento siriaco della Techné Grammatiké di Dionisio Trace," in La diffusione dell'eredità classica nell'età tardoantica e medievale - Il Romanzo di Alessandro e altri scritti (ed. R.B. Finazzi and A. Valvo; Alessandria: Edizioni dell'Orso, 1998), 95-111; Farina, "Diathesis and Middle Voice." T h e Greek text follows Gustav Uhlig, "Dionysii Thracis ars grammatica: qualem exemplaria vetustissima exhibent subscriptis discrepantiis et testimoniis quae in codicibus recentioribus scholiis erotematis apud alios scriptores interpretem armenium reperiuntur," in Grammatici Graeci; Partis prima, vol. 1. (ed. Leipzig: Teubner, 1883). T h e last two examples, which were translated as 'I was made' and 'I was written,' can actually also receive a middle interpretation (I made for myself,' 'I wrote for myself), according to the definition of middle just given in the Greek text. 58

As it emerges from some scholia to the Téchne, cf. Alfred Hilgard, "Scholia in Dionysii Thracis Artem Grammaticam," in Grammatici Graeci I, 3 (ed. Leipzig: Teubner, 1901), in some stoic linguistic doctrines, a correlation was established between the system of the noun genders and the verbal conjugations. As for the genders, a tripartite scheme was found, with a masculine, a feminine and a neuter (Lat. ne-uter, Gr. Ou8etepOV 'none of the two,' which was neither masculine nor feminine), and an analogous division was established for conjugations. What was neither active nor passive, had been indicated by some scholars as Ou8etepOV, by others as |iecroTT] diathesis. This account had the only aim of giving a short and general sketch of the origin of the term middle. T h e situation, as regards ancient Greek grammatical theories, is far more complicated and a great simplification has been undertaken here, for sake of brevity. For further details on the topic see, among others, Frédérique Ildefonse, La naissance de la grammaire dans l'antiquité grecque (Paris: Vrin, 1997), Jean Lallot, La grammaire de Denys le Thrace (Paris: CNRS, 1989). As regards the correlation between grammatical genders and verbal conjugations, cf. Pierangiolo Berrettoni, La logica del genere (Pisa: Edizioni PLUS, 2002). 59 For a thorough analysis of this passage, see the commentary by Lallot, La grammaire, 166-167 and also Farina, "Diathesis and Middle Voice" for further observation and a

19

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

Two terms have been used so far in connection with middle: voice and diathesis. According to Arce-Arenals, Axelrod and Fox (1994), it is only correct to speak about middle diathesis, as opposed to active and passive voice. Middle diathesis is a syntactic and morphological way a language has of signaling the affectedness of an active subject, and is therefore a sub-product of active voice. In modern grammar and linguistics, the term middle voice is nevertheless quite widespread and has retained both semantic and morphological values, indicating a specific verbal conjugation, as well as a set of semantic values that are usually associated with such a conjugation. This clearly shows, for example, in the definition of middle voice given by the Routledge Dictionary of Linguistics: Verbal voice contrasting with active and passive which is found in Sanskrit and classical Greek. T h e middle voice is semantically similar to reflexive constructions in that it describes an action which is performed by the subject for his/her own benefit or in which the subject affects itself: Grk louo (act.) 'I wash' vs. louomai 'I wash myself.' T h e r e is also a middle construction without an agent subject: didasko (act.) 'I teach' vs. diddskomai (mid.) 'I have myself taught,' which is similar to passive in meaning. Many Indo-European languages developed passive from middle-voice forms (see the typological summary of Kemmer 1993). 60

In this work middle voice is considered to be a semantic category, that has various manifestations in the Syriac language, which are related to a range of syntactic phenomena and to some morphological elements which are more appropriately definable as middle diathesis markers. But what is exactly meant by middle semantics? In general terms, one could say that middle voice expresses the subjectaffectedness, a certain degree of involvement of the grammatical subject in the process described by the verb. This definition has been more or less adopted by all the scholars who have dealt with middle voice. However, it is still quite vague, and can give rise to perplexities. Moreover, as will be shown in the course of this study, it is limited to one aspect of middle voice, but does not include a number of important semantic and syntactic features. 2.1.1 Kemmer's Middle Voice Among the contemporary studies on middle voice, one of the most exhaustive and well known is certainly Kemmer's. This work is fundamental for a number of theoretical and methodological assumptions, as well as for the frame of semantic categories that it provides for the analysis of middle systems in world languages. Refer-

comparison with the Syriac translation. 60

Hadumod Bussmann, Routledge Dictionary Routledge, 1996), 306. 61

of Language and Linguistics

(London:

Cf. John Lyons, Introduction to Theoretical Linguistics (Cambridge: Cambridge University Press, 1969), 373, quoted in Kemmer, The Middle Voice, 1.

20

OUTLINE OF MIDDLE VOICE IN SYRIAC

ence to it will be often made in the present study. For this reason it is useful to briefly recall some of its salient points. T h e scholar conducts a systematic analysis of the middle and reflexive patterns and of the related semantic values in a number of different languages, belonging to various families: Indo-European, Bantu, Turkic, Austronesian etc. Semitic languages are not represented. An initial very important point made by Kemmer is that it is necessary to distinguish reflexivity stricto sensu from middle voice. T h e two phenomena are closely related both from a semantic and (often) morphological point of view,62 but need nevertheless to be considered separately. More precisely, there are a number of semantic values that are proper to middle voice and that cannot be accounted for by means of reflexivity. A reflexive event can be defined as a transitive two-participant event in which the agent happens to be coreferential with the patient. T h e difference between this type of situation and a middle one (an often intransitive event, indicating a certain degree of subject-affectedness) emerges clearly from the definition of prototypical twoparticipant event, that Kemmer derives from Givon and from Hopper and rr~telic]

[-stative]

[-durative]

[+telic]

[-stative]

accomplishment gradual completion verbs

102

achievements Table 3. — Actional categories

T h e first feature [idurative] regards the fact that a verb can describe a process in which the starting point and the endpoint coincide (such as 'to sneeze'), or a process that lasts longer in time (such as 'to walk'). T h e second feature [itelic] indicates the presence or absence of an endpoint that needs to be reached for the action described by the verb to be really performed. T h u s 'to walk' is [-telic], because if one says 'John was walking, but then he met a friend and he sat on a bench,' it can still be said that John has walked. On the other hand 'to write a letter' is [+telic] and if one says 'John was writing a letter, when suddenly the phone rang,' it cannot be said that John has actually written the letter. In other words, when a telic verb is conjugated in an imperfective Tense, no hypothesis can be made about the actual conclusion of the event. Finally, the feature [istative] regards the compatibility with continuous Tenses and the Imperative. A stative verb,

101

Zeno Vendler, Linguistics in Philosophy (Ithaca, London: Cornell University Press,

1967). 102

I have derived this category from Pier Marco Bertinetto and Mario Squartini, "An attempt at defining the class of'Gradual Completion Verbs'," in Temporal Reference, Aspect and Actionality I: Semantic and Syntactic Perspectives (ed. P.M. Bertinetto, V. Bianchi, J. Higginbotham and M . Squartini; T o r i n o : Rosenberg and Sellier, 1995). It refers to verbs that indicate a gradual approach to a goal, that is more or less determined and fixed, according to the pragmatic context. T h e s e verbs, as shown by Bertinetto and Squartini "An attempt" are a subclass of accomplishments, although they have a hybrid behavior with respect to a number of features. T h i s category proves to be particularly appropriate for a number of derived stems from stative bases indicating a quality or a condition: the derived stems sometimes seem to indicate that the subject gradually assumes such a condition. Moreover this seems to apply also to some active causative verbs.

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

31

such as 'to be tall,' cannot be used in a continuous Tense: *'John is being tall,' nor can somebody be ordered *'be tall!' More subtle distinctions within the various classes can be found in Bertinetto (1986) - where a thorough description of the properties of each class is given and will be occasionally used in the course of this study. As observed above for Tense and Aspect, the categories Aspect and Actionality also interact constantly with each other, but are nevertheless distinct and operate at different levels within a language. Aspectual distinctions relate to the morphological specifications of the various languages, whereas Actionality is to be conceived mainly as a lexical property of the individual verbal stems. T h e syntactic context or the aspectual value of the Tense in which a verb is used can often have an important influence on its position with respect to Actionality. This fact is crucial for the analysis of Syriac verbs, because of the great influence (although maybe not general prominence) of Aspect in the use of the Tenses. Let us, for example, consider the verb Ibes 'to wear, to put on,' in the following sentences: 18. Gen 38:19 (Pesitta version) OjlQ^Oj/

»^ K

-

w-lebsat 'and' + III f. sg. pf. 'put on, wear'

r, Ibuse 'garments'

armlutah 'widowhood' + pron. suff. Ill f. sg.

"And she put on the clothes of her widowhood." 19. Gen 37:23 (Pesitta version) 103

,

Bertinetto, Tempo, Aspetto.

104 In the words of Rundgren, "Das altsyrische Verbalsystem," 54-55: "Die sog. natürlichen Aktionsarten stellen objektiv gegebene Grössen dar, die unter verschiedenen Gesichtspunkten beschrieben werden können. [...] Im Unterschied stellen die Aspekte eine subjektive Kategorie dar, die als eine polare, privative Opposition funktioniert". Some authors do not distinguish clearly between Aspect and Actionality and use the first label for both linguistic features. T h i s is the case, for example, in Timberlake's description of Aspect: T h e scholar considers the latter as "in part a property of the lexical semantics of verbs" (Timberlake, "Aspect, tense, mood," 286), referring to what we have considered as the actional value of the verbs. However, the possibility of a variation in perspective and in the description of a given event (which constitutes Aspect stricto sensu), forces Timberlake to add: "Still, it has to be stated that the lexical aspect of verbs is not fixed [...] Any given lexical item thus presents a characteristic, though not rigid, view of the flow of situations over time, as uniform or changing or liminal" (Timberlake "Aspect, tense, mood," 286). W i t h respect to the Syriac verbal system, a discussion of the importance of a distinction between the three categories of Aspekt, Tempus and Aktionsart in found in Rundgren "Das altsyrische Verbalsystem," 52-57.

32

OUTLINE OF MIDDLE VOICE IN SYRIAC

Joo) >.~|\) lii^'jS; kutina 'tunic'

d-pedyata

da-Ibis hwa

rel. part. + pi. 'long rel. part. + pass, ptc. 105 sleeves'

+ III m. sg. pf. 'be'

"The tunic with long sleeves that he was wearing." In 18 the verb Ibes is used in the perfect tense, with Aoristic value. T h e verb is there [+durative], [+telic] and [-stative], thus an accomplishment. In 19 on the other hand, the verb is used in the compound tense qtil hwa, with imperfective continuous value. T h e verb appears here to be [+durative], [-telic] and [+stative], thus a state. Situations like the one just analyzed are frequent in Syriac and the different forms of the verb, together with the Tenses, constantly interact with Actionality. Such an interaction will be the main concern of the textual analysis presented in chapter 4 of this book. T h e relevance of the interplay of the three categories of Tense, Aspect and Actionality (rather than the prominence of one of the three upon the others) for an understanding of the verbal system of Semitic languages was efficaciously underlined by Zaborski: Semantic aspect and tense (and Aktionsart) coexist and cooperate in every language [...] and the only reasonable question is whether and how they are expressed, first of all morphologically, then syntactically and lexically and finally how they work in an interplay of morphology, syntax and lexicon.106

In this study particular attention will be devoted to the aspectual and actional value of the forms with et- prefix and of the inner passives. 2 . 3 UNACCUSATIVITY

In traditional grammar and linguistics, verbs are usually divided into intransitive, which only have one argument (the grammatical subject of the clause) and transitive, which have at least two arguments and prototypically express an action performed by an agent on a patient. Intransitive verbs can be either unaccusative (when they express a state, or an event on which the subject has low control) or unergative (when they express an 105 'p ass _ p t C - ' ( a s w e y a s ' a c t - p t C - ' u s e ( J above) only refers to the morphological pattern (in this case qtil), without any implication of voice. In this case, for example, the participle has active-stative meaning. 106

Zaborski, "On the interplay," 871. See, among others, David Perlmutter, "Impersonal passives and the unaccusative hypothesis," in Proceedings of the fourth Annual Meeting of the Berkeley Linguistics Society (ed. J. Jaeger et al.; Berkeley: Berkeley Linguistic Society, University of California, 1978), 157189 and William Foley, "A typology of information packaging in the clause," in Language Typology and Syntactic Description (ed. T . Shopen; Cambridge: Cambridge University Press, 2007), 362-446.

M I D D L E VOICE AND SOME LINGUISTIC C O R R E L A T I O N S

33

action). For example, in 'John walks,' the English verb 'to walk' is intransitive and unergative, the grammatical subject has control over the action he performs. Conversely, in 'the vase broke' the grammatical subject undergoes the event described by the verb; the verb 'to break' (used intransitively) is unaccusative and its grammatical subject can be considered an object in the deep structure of the sentence. Unaccusativity can surface in different ways in languages. In Italian, for example, unaccusative intransitive verbs receive a different auxiliary (essere 'to be') from unergative intransitives (which take avere 'to have'). Moreover, the unergative verbs have the same auxiliary received by active transitive verbs (such as mangiare, 'to eat'), while unaccusatives take the same auxiliary which is used to express the passive. This proximity to the passive shows how the subject of the unaccusative verb is considered as undergoing an action or a process, just like the subject of passive verbs. Both subjects are objects at a deeper level of analysis. T h e situation of Italian can be exemplified as follows: Italian partire 'to leave' vs. camminare 'to walk' a - Stefania ieri sera adv. yesterday'

evening

e

partita

III sg. pres.

past part. f. sg.

aux 'be'

'leave'

"Yesterday evening Stefania left." P-

Stefania

ieri

sera

ha

camminato

adv. 'yesterday'

evening

III sg. pres.

past part. m. sg.

aux 'have'

'walk'

"Yesterday evening Stefania walked."108 Italian mangiare 'to eat' a

'-

la

mela

è

mangiata

da

def. art.

'apple'

III sg. pres.

passive part. f.

prep, 'by'

aux. 'be'

sg. 'eat'

Stefania

"The apple is eaten by Stefania." P'-

Stefania

ha

mangiato

la

mela

III sg. pres. aux

past part. m. sg.

def. art.

'apple'

'have'

'eat'

"Stefania has eaten the apple."

108

W i t h a Terminative Aoristic interpretation.

34

OUTLINE OF MIDDLE VOICE IN SYRIAC

A n o t h e r counterpart of the correlation here exemplified is that unaccusative and unergative predicates behave differently in participial constructions. 1 0 9 T h u s , a " and (3 " are possible, while y is ungrammatical in Italian: a

"- Mangiata

la

passive part. def. art. f. sg.

mela,

Stefania

'apple'

torrid

a

cantare

III sg. simple past prep, 'to'

f. sg. 'eat'

inf. 'sing'

'go back'

"[Having] eaten t h e apple, Stefania went back singing." P"-

Partita

Stefania,

spensero

tutte

le

luci

past part. f. sg.

ILL pi. simple past

'all'

def. art. f. pi.

'lights'

'leave'

'turn off

" W h e n Stefania left [lit. left Stefania] they t u r n e d off all t h e lights." "/"•

Camminata

Stefania,

spensero

past part. f. sg.

ILL pi. simple past

'walk'

'turn off

tutte 'all'

le

luci

def. art. f. pi.

'lights'

** "Walked Stefania, they turned off all t h e lights." F r o m these examples it can be concluded that t h e grammatical subject of an unaccusative verb shares, in its deep structure, some properties w i t h the subject of a passive construction: b o t h of t h e m are considered as patients. O n the other hand the subject of an unergative verb shares s o m e ^ r o p e r t i e s w i t h t h e subject of a t r a n sitive verb: b o t h are considered to be agents. Unaccusativity is related to a n u m b e r of syntactical properties of t h e verb and its arguments, a m o n g which is t h e possibility of using a differential object marker (when available in a given language) to introduce the subject of unaccusative verbs, in order to underline its low volitionality and low control of the action, as is t h e case for t h e following Biblical Hebrew sentence:

109 Cf. Martin Haspelmath, "Passive Participles across Languages," in Voice. Form and Function, (ed. B.A. Fox, P.J. Hopper; Amsterdam - Philadelphia: Benjamins, 1994), 151177, esp. 157. This article will be discussed in detail in chapter 4. 110 The concepts of patient and agent are here viewed as internally articulated categories.

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

35

2 Kings 6:5

••»on bx Vaj bran nni mipn Vao -man wa-yahi

ha-ehad

mappil

ha-qqorah

w3-et

and +III m. sg. ipf.

art.det.+num.

part. att. caus.

det. art. + 'rafter'

and + dir.

narr. 'to be'

one

'to fall'

ha-bbarzel

napal

el

det. art. + 'iron axe'

III m. sg. pf. 'to

prep, 'towards'

obj. mrk.

ha-mmayim det. art. + 'water'

fall'

"And one was cutting down a rafter and the iron axe fell into the water." In this example the particle et, the nota accusativi, which is normally used in Biblical Hebrew to introduce a determined direct object, precedes instead the subject of the verb. T h e verb considered here, napal 'to fall' is unaccusative its subject is not prototypically agentive and has low control over the event (i.e. is an object in the deep structure of the sentence). T h i s fact has been stressed by the use of the particle et, the direct object marker. 2.3.1 Unaccusativity, Aspect, Actionality and Middle Voice In many ancient Indo-European languages such as Latin, Greek, Vedic, Hittite there is a verbal diathesis, usually called 'middle,' which covers several different values, including reflexive, reciprocal, passive and many others. It is not easy to determine its original and core function, but in the historical languages the middle diathesis is clearly opposed to the active one, so that, for the majority of verbs, to every active form and meaning, a middle one corresponds, with either reflexive or passive value. In all the languages mentioned above there is also a number of verbs, the so-called media tantum, that do not allow for an active counterpart. In the same way, there are also activa tantum, namely verbs that have only active diathesis. Many scholars in the course of time have argued that the original core of middle verbs was represented by these media tantum, but they did not agree on the 111

Other examples of this use of et are found in Nm. 5,10; Dt. 20, 8; Js. 17,11; Jd. 20, 44; IS. 20, 13; 2S. 11, 5; IK. 18, 35; Jr. 3,9; Ez. 35, 10. For a semantic analysis of the use of et as patient marker, cf. Jacob Hoftijzer, "Remarks regarding the Use of the Particle 'et in Classical Hebrew," Oudtestamentische Studien, 14, 1-99; William Randall Garr, "Affectedness, Aspect and Biblical 'et," ZAH, 4/2, 119-134; Pablo Kirtchuk, "/'et/ ou ne pas /'et/, l'actant Y en hébreu et au-delà," Actances, 7, 91-133; Michael Malessa, "Différentielle Objektmarkierung im klassischen Hebräisch," KUSATU, 1, 133-156. Finally, Margherita Farina, "La marca differenziale dell'oggetto e la particella 'et in ebraico biblico, prime prospezioni," Quaderni del Laboratorio di Linguistica della Scuola Normale Superiore di Pisa, 5 for synthetic re-examination of the problem and a discussion of the literature on the matter. 112

ml

For a list of the main values of middle diathesis see Kemmer, The Middle Voice", Romano Lazzeroni, "La diatesi come categoria linguistica: studio sul medio indoeuropeo," Studi e saggi linguistici, 30, 1-22. 113 Such as Berthold Delbrück, Vergleichende Syntax der indogermanischen Sprachen (Strassburg: Karl Trübner, 1897); Emile Benveniste, "Actif et moyen dans le verbe," in

36

OUTLINE OF MIDDLE VOICE IN SYRIAC

function and values of Indo-European middle endings. Benveniste (1966) also understood that the ancient distribution of verbal endings in Indo-European was lexical, so that every verbal root could have only one series of endings (either active or middle). Only over time was a symmetry developed in the verbal systems of the individual languages, so that each verb could have both active and middle diathesis. Bakker (1994) examines the correlation between ancient Greek middle diathesis, Aspect and Aktionsart. T h e scholar considers middle diathesis as an encoding of subject affectedness, which, in its turn, is put within the frame of scalar transitivity, as defined by Hopper and Thompson (1980). Bakker identifies the degree of transitivity of an event as determined by the following semantic parameters: Volitionality, Agency and Causation (in this order). Aspectual opposition is considered by Bakker in terms of the opposition between aorist and imperfective verbal bases. By Aktionsart Bakker essentially means the opposition durative vs. punctual. Bakker shows that Aktionsart interacts with Aspect in determining whether a given verb is basic in the aorist and derived by means of suffixation in the imperfect (with punctual roots) or vice versa (in durative roots). On the other hand, the selection of aspectual morphology, such as the aorist markers -sa- and -the-, depends also upon the degree of transitivity of the event, namely upon the distribution of the above-mentioned parameters. Bakker has examined a number of ancient Greek verbs according to a series of event-types individuated through the transitivity continuum, classifying them on the basis of the presence/absence of some parameters: the possibility of middle morphology, the affectedness of the subject, and the possibility of the aorist morphemes -sa- and -the-. I reproduce here the table in which the results are summarized: Event-type 1-participant events Objective states Physical processes (E)motion, cognition +/- volitionality Inherent reciprocals Affectedness and agency 2-participant events

Middle

causatives Patient-directed

Inherent aff. of s.115 -sa-

-the-

-

-

-

-/+

-

-

-

+

+

-

+

+

+

-

+

+

+

+

+

+

+

+

-

+

+

+

-

+

-

+

+

+

+/-

+

+

+

+

-

+

Problèmes de linguistique générale (Paris: Gallimard, 1966), 68-175. 114 Egbert J. Bakker, "Voice, Aspect and Aktionsart," in Voice. Form and Function, (ed. B.A. Fox, P.J. Hopper; Amsterdam - Philadelphia: Benjamins, 1994), 23-47. 115 "Inherent Affectedness of the subject."

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

activity 'Kill'

-

-

-

37

-

Table 4. - Bakker's event-type classification

T h e event-types are very close to the categories individuated by Kemmer as relevant for the selection of middle voice (cf. 2.1.2). T h e conclusion is that "middle voice may occur in all but the event-types that are the lowest and the highest in transitivity." T h e -sa- morpheme is associated with the events that are higher in volitionality, while the -the- one "occurs where affectedness takes the aspectual value of inherent duration." 117 T h e analysis shows that both aspectual morphology and diathesis vary according to the same parameters, which are connected with the inherent semantics of the events. What has not been developed by Bakker is the internal articulation of Aktionsart as a category, which the author considers exclusively in terms of +/- duration and which therefore does not appear clearly in the whole of the transitivity continuum. Once telicity and stativity are introduced into the equation, the dynamics of the correlation with diathesis are more clearly explained. In Lazzeroni (1990) and Lazzeroni (2004) some interesting hypotheses are made about the original distribution and function of Indo-European middle verbal endings. He analyses the list of Indo-European media tantum given by Delbrück (1897), among which are verbs meaning: to be born, to die, to follow, to walk, to own, to lay down, to seat, to enjoy/benefit, to meditate, to abandon, to lose, to be/get warm, to be sad, to feel ashamed, to feel pain etc.

All these verbs have a stative or eventive value. Nevertheless, these values of the media tantum do not explain, by themselves, all the values that middle voice conveys in historical languages. T h e solution proposed by Lazzeroni (1990) is to conceive original middle voice as a scalar category, built around a prototype.

116

Sc. 1-participant events and verbs of the type 'to kill'. Bakker, "Voice, Aspect," 42.

117

Bakker, "Voice, Aspect," 44.

118

A striking anticipation of this idea, with a considerable systematization, although only briefly sketched, can be found in Joiion. H e draws a parallel between ancient IndoEuropean and modern Romance middle conjugations on the one hand and Semitic socalled stative vowel patterns on the other. In order to justify the extension of a given pattern to verbs that apparently express an activity, Joiion observes: "la nuance d'«activité» varie beaucoup d'un verbe à l'autre. Il y a assurément moins d'«activité» dans dormir que dans courir, dans recevoir que dans prendre, dans entendre que dans écouter [...] et l'on conçoit fort bien que telle langue [se. Arabie] ait pu exprimer morphologiquement cette difference sémantique.": Joùon, "Sémantique des verbes statifs," 5.

38

OUTLINE OF MIDDLE VOICE IN SYRIAC

It can be said that the original value of Indo-European middle verbs was stative, if stative and eventive verbs can be seen as part of a single prototypical category defined by some features: agentivity Statives Eventive s Agentive statives

processuality +

+

Table 5. - Semantic structure of stative and eventive verbs

Another important feature of some media tantum is the so-called inner diathesis: they sometimes describe a process that involves the subject and refers to it (as in 'to wash' or 'to wash one's hands'). T h e reflexive and reciprocal values are also related to this feature. Finally, passive value is secondarily derived from both subjective and unagentive values. An original Indo-European passive construction is not reconstructed, but the passive is always a single language development: "La categoria del medio si è, dunque, costituita in seguito all'estensione metonimica delle proprietà del prototipo." An important contribution in the same direction is given by Romagno. An analysis of the main categories, in which Indo-European media tantum can be classified, shows that the hypothesis of a stative prototype is not completely adequate. A number of motion verbs is in fact included in the original core of middle verbs (such as Greek oikhomai 'to leave,' néomai 'to come back,' e'rchomai 'to go,' and Sanskrit gahate 'to plunge' nasate 'to come back,' variate 'to turn' etc.). Examining the semantic and syntactical properties of media tantum, Romagno concludes that the prototype of ancient Indo-European middle verbs was more likely represented by intransitive unaccusative verbs. Unaccusativity must here be considered as a scalar category characterized by a number of features, among which the notion of state was prominent, but not unique. Also low control and low volitionality of the subject must have had an important role in the structure of the ancient media tantum class. On the one hand some verbs of motion may also have the notion of state in their logical structure (for example telic verbs of motion such as 'arrive' or 'come back' can be considered from the point of view of the state resulting from the event they describe) and subsequently belong to the category of media tantum. On the other hand, as was mentioned above, the subject of unaccusative verbs shares some properties with the direct object of a transitive clause and can be considered as an object in the deep structure. This fact accounts for the extension of middle endings to reflexive and passive values, expressing an action that the subject either performs on itself or undergoes. 119

Romano Lazzeroni, "La diatesi come categoria," 14.

120

Domenica Romagno, "Diatesi indoeuropea e verbi di movimento greci: alcune considerazioni sull'intransitività," Archivio Glottologico Italiano, 87, 163-174.

121

T h e same can be said also for all the other values of middle voice, such as reciprocal, benefactive etc.

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

39

In Lazzeroni122 further steps are made in the comprehension of the original structure of the Indo-European verbal system. T h e author focuses on the Vedic verbal suffix -ya, originally combined only with active endings, to express intransitive, reflexive, reciprocal and passive meaning, covering a semantic area analogous to the one originally covered by middle endings. Moreover, middle endings and -ya derivatives seem to be in complementary distribution to express the same semantic values: a verb could have either middle endings (medium tantum) or form a -ya derivative (in which case it was an activum tantum). Lazzeroni then concludes that the original value of the -ya suffix was to give to the activa tantum a medial value, namely an anticausative value, in a time when middle endings were lexically selected. Of course, this suffix could be added only to transitive verbs, but not to unergative verbs. However, the Vedic verbal system also had some affixes to express causative meaning: -aya, -«-. T h e range of application of those affixes follows the same criteria individuated fo the -ya suffix. T h e -n- infix, denoting high transitivity and high telicity, is limited to a class of verbs that express a change in state. As noticed by Lazzeroni, this class also systematically uses the -ya suffix with anticausative function. All the other verbal classes can express causation either lexically (pure transitive verbs) or by means of the suffix -aya. T h e distribution of verbal diathesis in Vedic, as described by Lazzeroni can be described as follows: two cores of activa and media tantum lexically selecting different series of endings. A core of verbs denoting change in state, that use the affixes -ya and -n- to express the anticausative and the causative respectively. A series of intransitive (unaccusative) and not highly transitive verbs using the -aya suffix to express causation. From ancient Greek, Vedic and Latin data, it can be concluded that the lexical selection of active and middle endings, in ancient Indo-European, was a surface morphological display of split intransitivity. There were, moreover, a number of morphological devices to modify the argumental structure of the predicate, according to its original semantic value. In the development of the individual languages, the selection and extension of the various morphological elements has followed the pattern of metonymic association along the scale of prototypical categories.

122

Romano Lazzeroni, "Inaccusatività indoeuropea e alternanza causativa vedica," Archivio Glottologico Italiano, 89, 139-164. 123

Anticausative is a process represented as spontaneous or internally caused, independent from an external cause. For example 'the house burns' vs. 'John burns the house,' or 'the vase breaks' vs. 'John breaks the vase.' Cf. Haspelmath, Transitivity Alternations; "More on the Inchoative-Causative," and also § 2.1.1 n. 14 in this chapter.

40

OUTLINE OF MIDDLE VOICE IN SYRIAC

2 . 4 FORMULATION OF THE WORKING HYPOTHESIS

As mentioned in § 2.1.2, Kemmer does not include any Semitic language. If one tries to apply the definitions of middle and reflexive markers given above to Syriac, some problems arise. T h e reflexive marker is defined as follows: A reflexive marker is a productive grammatical device that is used obligatorily to mark direct reflexive contexts in at least the third person. 124 [...] Faltz found no languages in his sample of 30 which had more than one reflexive marker [...]. If this limitation is not an absolute universal, it is certainly a universal tendency, and thus I will refer to a reflexive marker in a given language as the reflexive marker for that language. 126

Having to point at 'the reflexive marker' in Syriac one would probably pause for hesitation. In the case of Syriac (but the same could be said also, for example of Biblical Hebrew) one can at best talk about a specific reflexive construction, a pattern that associates a preposition or a noun referring to 'soul,' 'body,' 'person,' etc. to a coreferential pronominal suffix. Moreover, Syriac can hardly be classified as either a one-form or two-form language type. T h e r e is not a single morpheme or structure expressing both reflexivity and middle, but there is also not a dichotomy of forms and functions. As shown in § 1.1, there are at least three structures concurring in the expression of reflexive and middle values. Nevertheless, I believe that applying Kemmer's distinction between pure reflexivity and middle semantics can still be fruitful, as well as making use of the semantic classes individuated by the scholar, provided that some other linguistic categories and factors are icluded in the analysis. T h e existence of a connection between the Syriac verbal et- patterns and the semantics of middle voice, as they have been defined by Kemmer, has already been pointed out in Mengozzi, in a study on the semantics of some Neo-Aramaic constructions with a 3rd singular feminine 'empty' object. T h e author also clearly differentiates the middle structures of the type 'ethzi "he appeared" and the

124

Kemmer, The Middle Voice, 47.

Leonard Faltz, Reflexivization: A study in universal syntax (New York: Garland Publishing, 1985). 126 Kemmer, The Middle Voice, 47. 127 Alessandro Mengozzi, "Middle Markers: Neo-Aramaic and Italian verbal forms with a 3rd singular feminine pronominal object," in Atti del XII Incontro italiano di linguistica afroasiatica (Ragusa 6-9 giugno 2005) (ed. M. Cassarino and M. Moriggi; Soveria Mannelli: Rubbettino, 2007), 105-116. 128 "...the stems with prefix 'et- attested in earlier stages of Aramaic (cf., e.g., Classical Syriac 'elp'el, 'elpa"el, 'ettafel), which all convey a reflexive-middle voice meaning": Mengozzi, "Middle Markers," 105.

MIDDLE VOICE AND SOME LINGUISTIC CORRELATIONS

41

reflexive groper, expressed by pronominal forms, such as hawwi napseh "he showed himself." 29 Moreover, Mengozzi's work, although specifically focused on a NeoAramaic phenomenon, rather than on the Classical Syriac et- stems, offers an approach to middle voice that is interesting from the methodological point of view. Middle voice is considered, in fact, to be an underlying semantic connection between a number of structures, emerging at the syntactical level in some idiomatic expressions. As I tried to show in the preceding paragraph, stativity and unaccusativity are the two features that some scholars have considered as being prototypical for the Ancient Indo-European middle voice. T h e existence of a scalar continuum between the two features explains the use of a unique series of endings to convey both values. Nevertheless one can also imagine to group such features differently and formally distinguish, for example, between a state or permanent property and an action undergone or one upon which the subject has low control. This is the structure that I propose to use in the description of Syriac middle voice. Combining the results of a number of general linguistic studies in the field of middle voice, such as those by Kemmer, Lazzeroni, and Romagno, also referring to more general works on phenomena that are closely related to it^ such as Perlmutter, Bossong, Bertinetto, Bertinetto and Squartini, Cennamo, I will attempt to sketch a distribution of forms and functions for the Syriac verbal system. More specifically, I will argue that key concepts for the understanding of the value of the et- forms are : a)

b)

129 130

unaccusativity, according to the formulation given by Perlmutter, and also to its applications to various Indo-European languages by Kulikov, Lazzeroni, Romagno; the scalar articulation of semantic categories such as transitivity, as introduced by Hopper and Thompson.

Mengozzi, "Middle Markers," 105.

In chapter 5 of this book an analogous analysis will be provided for the syntactical phenomenon of the coreferential dative. 131 An analogous connection between unaccusativity on the one hand and Arabic middle verbs and i-stems on the other is described in Maalej, "Middle Verbs," 224-225. 132 ml Kemmer, The Middle Voice", Lazzeroni, "Inaccusatività indoeuropea;" Romagno, "Diatesi indoeuropea;" Perlmutter, "Impersonal passives;" Bossong, "Le marquage différentiel;" Bertinetto, Tempo, aspetto; Bertinetto and Squartini, "An attempt at defining;" Michela Cennamo, "Late Latin pleonastic reflexives and the unaccusative hypothesis," Transactions of the Philological Society, 97, 103-150. 133 Leonid Kulikov, "Passive and middle in Indo-European," in Passivization and Typology - Form and Function (ed. W . L . Abraham, Larisa; Amsterdam - Philadelphia: Benjamins, 2006), 62-81. 134 Hopper and T h o m p s o n , "Transitivity."

42

OUTLINE OF MIDDLE VOICE IN SYRIAC

O n the basis of t h e above mentioned scientific works and of a detailed analysis of a sample of early Syriac texts, I will a t t e m p t to reconstruct t h e following organization of morphological and syntactical material, w i t h respect to middle voice and reflexivity: reflexive (and reciprocal) (prep.) + pron. suff. (co-referentiality of agent and n o u n 'body' / 'soul' + pron. suff. (depending on t h e degree of emphasis required patient) by the context) inner passive middle - stative [- process] middle - unaccusative external affixation: et- stems [- control] [- volition] [- agentivity] [+ affectedness] Table 6. - Articulation of Syriac reflexive and middle structures T h e functions indicated in t h e scheme correspond to t h e semantic values that are prototypical for each structure. T h e r e is a greater distance between the first class on t h e one hand and t h e other two classes on the other hand. Pure reflexive structures are semantically differentiated from the middle ones in their emphatic (and not always homogeneously grammaticalized) character and in their role of markers of co-referentiality between two different participants of a transitive event. T h e two forms that are classed as middle (inner passives and et- stems), instead, b o t h belong to a macro-category, that of unaccusatives, defined by some basic semantic features: affectedness of the subject and lack of control, volition, and agentivity. W i t h i n this macro-category, moreover, a sub-class is individuated. T h e one represented by the inner passives is also endowed w i t h t h e feature of lack of process and, therefore, of stativity. As it will be shown in t h e course of this study, such values undergo various kinds of drifts and extensions, by means of metonymic and metaphoric associations, giving rise to t h e variety of meanings that are traditionally attributed to etstems and inner passives.

135

On the criteria for the selection of the sample, see chapter 4.

3

E T - FORMS AND INNER PASSIVES IN SEMITIC

STUDIES

3 . 1 . A SURVEY OF ANCIENT SEMITIC VERBAL T - F O R M S : S T A T U S QUAESTIONIS

T h e verbal prefix et- is the Syriac variant of an affix that is widespread among most Semitic languages, from its earliest attestation up to today. Both the morphological context and the semantic value of this affix vary throughout Semitic as well as within the individual languages, and have been a matter of debate in the history of linguistic studies. For this reason, it may be useful to briefly run through some relevant problems and hypothetical reconstructions, in order to set the Syriac etforms in a broader comparative context. T h e basic function commonly indicated in traditional studies for the Semitic (-)i- affix is that of reflexive marker. In addition, it is said that this morpheme can express reciprocal and passive meaning. As we saw in chapter 2, these labels are generic and in some cases extemporaneous, and they do not descend from a systematic semantic classification, but they are, rather, the result of a progressive calcification. In different terms, one could say that, throughout the Semitic family, an affix (-)i- is widespread, which is used, in several instances, as a de-transitivization device, from different verbal stems. However, reducing the verbal valency is not the only function of such a morpheme (or class of morphemes), nor has (-)i- the monopoly of this function. On the contrary, the prefix n-, some vowel patterns and pronominal structures also compete with (-)f-, to a different extent and with different balances in the verbal systems of the individual languages. T h e (-)i- morpheme is either prefixed to the verbal root (in which case it is usually preceded or followed by a vowel, because most Semitic languages tendentially avoid the sequence CC at the beginning of a word), or infixed after the

136 q . • y a j m y Q v ô n > " T h e Pragmatics of De-Transitive Voice: Functional and T y p o logical Aspects of Inversion," in Voice and Inversion (ed. T . Givôn; Amsterdam - Philadelphia: Benjamins, 1994), 3-46; R . M . W . Dixon and Alexandra Aikhenvald, "A Typology of Argument-Determined Constructions," in Essays on language function and language type: dedicated to T. Givôn (ed. J. Bybee, J. Haiman and S.A. Thompson; Amsterdam - Philadelphia: Benjamins, 1997), 71-113. 137 Cf. R. Lee Humphreys, "Valency changing alternations," in Concise Encyclopedia Grammatical Categories (ed. K. Brown and J. Miller; Oxford: Elsevier, 1999), 391-400.

43

of

44

OUTLINE OF MIDDLE VOICE IN SYRIAC

first root-consonant. When the first root-consonant is a dental, the (-)f- affix, if unvocalized, is assimilated. T h i s affix is not uniformly represented in the Semitic languages. In Akkadian and Ugaritic it is widespread and can be used with almost all verbal stems (except for N, a stem built through an n- prefix). In Classical Arabic three main f-stems are found: one is built upon the basic stem and the others upon derived ones. In Aramaic the (-)f-morpheme is widely productive and has progressively specialized in the expression of passive meaning, together with the progressive recession of the inner passive and the disappearance of the «-stems (see infra). T h u s , a system of oppositions has developed, in which a f-passive corresponds to every verbal stem. In Western Neo-Aramaic the i-forms have been preserved, whereas in Eastern Neo-Aramaic, through the influence of the Iranic adstratum, i-forms have been almost completely lost, at least as a productive derivational process. T h e y are nevertheless^reserved in participles employed in the construction of the periphrastic passive. Let us now examine a few examples of the uses of i-forms in some classical Semitic languages: 1. Biblical Hebrew - IK 20:22 p t n n n

f ?

iV

n o w i

wa-yyo'mer l-o conj. 'and' + III m. sg. prep, 'to' + impf. conv. 'to say' pron.III m. sg.

lek hithazzaq ipv. qal III m. ipv. hitp. Ill m. sg. sg. 'to go' 'to be strong'

"And he said to him: strengthen yourself!" qal (basic stein/ 4 0 - hazaq 'to be strong;' pi'el (D stem) - hizzaq 'to make strong;' hithpa'el (tD stem) - hithazzeq 'to strengthen oneself.'

138

Cf. the sketch in § 1.1. Cf. Otto Jastrow, "The Neo-Aramaic Languages," in The Semitic Languages (ed. R. Hetzron; New York - London: Routledge, 1997), 334-377, esp. 341; Fabrizio Pennacchietti, in Tipologie della convergenza linguistica - Atti del convegno della Società Italiana di Glottologia, Bergamo, 17-19 dicembre 1987 (ed. V. Orioles; Pisa: Giardini editore, 1988), 93-107. 140 t e r m ^ m e a n j n g "light", indicates the basic stem in the Hebrew grammatical tradition. 141 Intensive/causative stem with redoubling of the second root-consonant. 139

E T - FORMS AND INNER PASSIVES IN SEMITIC STUDIES

2. Classical Arabic jojj-l tanassafa III m. sg. pf. V form 'to dry oneself

bi-lawbi-hi prep, 'in, with' + 'garment' + pron. suff. Ill m.sg.

l-gadldi det. art.+ adj. 'new'

" H e dried himself with his new garment." fa'ala - nasafa 'to soak, to become soaked;' fa'ala (D stem) - nassafa 'to dry;' tafa'ala (tD stem, mutawi' of the D) - tanassafa 'to dry oneself.' 3. Akkadian 1 4 3 samnam acc. m. sg. 'oil'

iptasas-ma III m. sg. pret. 'to anoint' (+ ventive part.)

" H e anointed himself with oil." G stem - pasasu 'anoint;' Gt stem - pissusum (< *pitsusum) 'to anoint oneself.' 4. Classical Arabic

ga'ala la-kumu l-nuguma li-tahtadu III m. sg. pf. 'ere- prep, 'to' + pron. suff. det. art. + 'stars' prep. 'to'+ II m. pi. ate' II m. pi. ipf. VIII 'to guide' " H e created t h e stars for you to be guided." fa'ala - hada 'to guide;' ifta'ala - ihtada 'to be (well) guided.' 5. Biblical H e b r e w - Gen 42:1

i*nnn noV lammah conj. 'why'

titrau II m. pi. ipf. hitp. 'to see'

" W h y do you look at each other?" 142

For the concept of mutawaa cf. § 1.2. T h e example is taken from John Huehnergard, A Grammar of Akkadian (Wi Lake, Indiana: Eisenbrauns, 2005), 397.

46

OUTLINE OF MIDDLE VOICE IN SYRIAC

qal - raah 'to see;' pi'el - not attested; hithpael - hithraeh 'to look at each other;' nifal - nir'ah 'to be seen, to let oneself be seen.' 3.1.1 History o f the (-)f- Affix: some Hypotheses o f Reconstruction T h e affix (-)i- has different forms and positions in the various Semitic dialects. It m u s t be added that variation occurs also within t h e individual languages, depending on w h e t h e r t h e m o r p h e m e attaches to t h e basic stem or a derived one. W h i l e Biblical H e b r e w has preserved only one prefixed i-form, built u p o n t h e D stem, Akkadian has several i-forms, all constructed by infixation. By contrast Classical Arabic has b o t h prefixed and infixed i-forms, giving rise to different verbal stems. T h e problems related to t h e origin and history of t h e (-)i- affix involve basically two topics: a. t h e spreading and extension of use in t h e various dialects; b. the position of t h e m o r p h e m e (prefixed or infixed). Diem discusses b o t h topics thoroughly, and his conclusions are broadly shared by scholars. D i e m analyses the situation of the following languages: A r a maic, Hebrew, Phoenician, Classical Arabic, Ugaritic, Early South Arabian, E t h i o pic, Akkadian and M o d e r n South Arabian. Aramaic, H e b r e w and Phoenician are characterized by the so-called Metathesisregel, a process in which, in roots beginning with a sibilant, t h e (-)i- affix is infixed after t h e first root-consonant. Otherwise it is always prefixed. T h i s p h e n o m e n o n is usually explained as being due to the general tendency of Semitic languages to avoid the sequence dental-sibilant. In t h e remaining languages (except for Ethiopic), there is a complementary distribution of the forms respectively with

144 In passages such as Jud. 20:15,17; 21:9; Num. 1:47; 2:33; 26:62; IK. 20:27 forms of a i-stem are found which do not display, according to the masoretic punctuation, the redoubling of the second root-consonant, which is expected for i-stems derived from the intensive pi'el (or qittel) ones. Such forms have received different explanations, cf. Bruce K. Waltke and Michael P. O'Connor, An introduction to biblical Hebrew syntax (Winona Lake, Indiana: Eisenbrauns, 1990), § 26 for a description of the problem. See § 3.1.6 in this book for further discussion. 145

Such as tafa"ala, tafä'ala, ifta'ala, istaf ala. Werner Diem, "Die Entwicklung des Derivationsmorphems der t- Stämme in Semitischen," ZDMG, 132, 29-84. 147 Some exceptions are however attested, cf. the form ytsm' "es sei nicht gehört", found in the Sfire Stele (I, 29, cf. Stanislav Segert, Altaramäische Grammatik, mit Bibliographie, Chrestomathie, und Glossar (Leipzig: VEB Verlag Enzyklopädie, 1986), 256). 146

148

Cf. Joseph H. Greenberg, "The Patterning of Root Morphemes in Semitic," in On Language. Selected Writings ofJoseph H. Greenberg (ed. K. Denning and S. Kemmer; Stanford: Stanford University Press, 1950 [1990]), 365-387.

E T - FORMS AND INNER PASSIVES IN SEMITIC STUDIES

47

prefixed and infixed (-)i-: the infix is selected by t h e basic stem, t h e prefix by all 149

other stems. Most scholars have considered t h e metathesis - which in Aramaic, Hebrew (and Phoenician) is limited t o a subset of t h e lexical roots - as t h e source of the infixed -t- forms in the various languages. Protosemitic would have formed derived stems only by means of prefixes («-, s-, h-, £-, '-); in the roots beginning with a sibilant, the t- m o r p h e m e would have undergone a metathesis. Subsequently this p h e n o m e n o n would have been extended to all other roots, creating t h e stem with infixed -£-. As noticed by Diem, this reconstruction is not completely satisfactory, mainly because it does not explain the presence of b o t h infixed and prefixed (-)i- in several languages. According to Diem, t h e distribution of t h e m o r p h e m e in t h e various dialects is related to other phonetic changes. T h e basic elements of his reconstruction are the alternation of (-)i- and ta- and t h e alternation of prefix and infix forms. D i e m observes that, while -t- can be b o t h prefixed and infixed, ta- can only be prefixed to t h e verbal root. T h e r e f o r e , he supposes that t h e (-)i- affix derives f r o m a reduction of an original ta- in particular contexts, i.e. after a CV structure, as in *yataqatil>*yatqatil. In roots beginning w i t h a sibilant, however, this reduction would have generated a dental-sibilant sequence, which would subsequently have undergone m e tathesis. Finally, in some dialects, the Metathesisregel would have been extended to all roots, regardless of t h e nature of t h e first root-consonant. In Stempel a significant objection to Diem's reconstruction is formulated. T h e Arabic VIII form, with infixed -t- (corresponding to t h e Akkadian structure) seems to be older t h a n t h e ones with prefixed ta-. In the first one, in fact, the i m perfect is vocalized according to t h e rules of apophony (iqtatala - yaqtatilu), while in t h e others it simply follows t h e vocalization of the perfect (taqattala - yataqat-

149

This type of metathesis is regarded as a "Pan-Semitic feature" in Takamitsu Muraoka and Bezalel Porten, A Grammar of Egyptian Aramaic (Leiden: Brill, 1998), 25-26. However, Margaretha Folmer, "Metathesis in Jewish Aramaic: a so-called "Pan-Semitic Feature" reconsidered," in Hamlet on a Hill (ed. M. Baasten, T . Muraoka and W . T h . Van Peursen; Leuven: Peeters, 2003), 223-243 correctly suggested such a perspective be modified. With respect to Jewish Epigraphical Aramaic, she notices: "One of the remarkable features within these Aramaic texts is that metathesis in the Itp. does not take place when the initial radical is one of the sibilants just mentioned. This is true both for the Bar Kosiba letters and for the contemporary legal documents. In the Bar Kosiba letters the following relevant forms are found: SK 1.7 htskw (suffix conj. - a scribal error for htskh.w)', SK 1.10 ytskh. (prefix conj.); SK4.4 htsdr (an imperative)." (p. 234). 150

Reinhard S. Stempel, Abriss einer historischen Grammatik der semitischen Sprachen (Frankfurt am Main: Peter Lang, 1999). 151

Pace Jan Retsö, "Diathesis," in Encyclopedia of the Arabic Language and Linguistics (ed. C.H.M. Versteegh; Leiden: Brill, 2006), 622-626, esp. 624.

48

OUTLINE OF MIDDLE VOICE IN SYRIAC

T o this objection a few more may be added. T h e Metathesis-Theory presupposes, for Protosemitic verbs, a highly regular derivational system, in which all derived stems were built by means of prefixes. A sort of de-regularization would then have occurred, with the creation of an infixed -t- stem. Another problem is the hypothesis that the infix forms derive from the generalization of a process, that in the beginning was (and in some languages, such as Hebrew, which remained) conditioned by the presence of a sibilant, as first root-consonant. Although they are quite frequent, their number does not justify the dragging of the entire lexicon. Lipinski seems to reconstruct an original position of the (-)i-affix after the first root-consonant, and not outside the root. 3.1.2 Other De-Transitivizing Morphemes and Structures As mentioned above, Semitic languages use different strategies to operate detransitivization. Besides the (-)i- affix, they use other affixes and various nominal periphrases. a. T h e n-prefix the {~)t~, displays an array of meanings, basically related to intransitivity, but ranging from fully passive to rather reflexive-like events. It attaches to the basic stem. In Akkadian it is also vocalized according to the vocalization only of the basic stem, which does not occur for the other derived stems. Moreover, the reflexive meaning is rare in Akkadian. It has to be noticed that Semitic languages tend to avoid the expression of the agent, in a passive structure. Therefore (as we will see in detail in ch. 4), the distinction between reflexive and passive meaning is often unclear. Let us now examine some examples of the use of the «-prefix in Hebrew: 6. Biblical Hebrew - Ruth 1:17 -QpN Dun rnoN T n o n ba-aser prep, 'in' + rel. pron.

tamuti II f. sg. ipf. 'to die'

amut w3-säm eqqäber (*e-n-qaber) I sg. ipf. conj. 'and' + adv. 'there' I sg. ipf. nifal 'to bury' 'to die'

"Where you will die I will die, and there I will be buried." qal - qäbar 'to bury;' nifal - niqbar ' t o be b u r i e d . '

152

Edward Lipinski, Semitic Languages: Outline of a Comparative Peeters, 1997), 395-402.

Grammar

(Louvain:

E T - FORMS AND INNER PASSIVES IN SEMITIC STUDIES

49

7. Biblical Hebrew - 2S 20:10 m m

xb KDDJN

wa-'amasa conj.'and'+Amasa

lo conj. 'not'

nismar Ill m. sg. pf. nifal 'to watch'

ba-hereb prep.'in, with'+ 'sword'

"And Amasa did not protect himself from the sword..." qal - samar 'to w a t c h , t o k e e p ; ' nifal - nismar 'to be o n one's g u a r d . '

T h e «-prefix is also used, in Post-Biblical Hebrew, in nitpa'al which the n- is prefixed to the derived stem with prefixed £-.

forms, 154 in

8. Mishnaic Hebrew - 'Erubin V i, 53a 156 D T I

D r n i n

n D " p ™

bne pi. cs. st. 'son'

D J I ^ V by

i r a p r u y

m i r r

m

yahudah Judah

se-hiqpidu al rel. + III m. pi. pf. hif*il prep, 'on' 'to care for' bson-am nitqayyamah torat-am ba-yad-am 'language' + pron. Ill m. Ill f. sg .pf. nitpa'al 'law' + suff. Ill m. pi. prep, 'in' + 'hand' + pron. pi. 'to stand' suff. Ill m. pi.

"The Judeans who cared for their language, their law was preserved in their hands."

153 preposition b- introduces here the object from which Amasa negleted to protect himself. Cf. Francis Brown, S.R. Driver and Briggs Charles A., A Hebrew and English lexicon of the Old Testament: with an app. containing the biblical Aramaic: based on the lexicon of William Gesenius as transi, by Edward Robinson (Oxford: Clarendon, 1952), 1037: "by reason of the sword, i.e. against it." 154 Cf., e.g. for Rabbinic Hebrew, Miguel Pérez Fernández and John Elwolde (transi.), An Introductory Grammar of Rabbinic Hebrew (Leiden: Brill, 1997), 95-96: "In the perfect, the Hitpa'el is replaced by a Nitpa'al conjugation [...]. In some instances a Nittafal form is probably attested [...]." And, further on: " N i t p a ' a l . T h i s binyan is a fusion of the Nifal and Hitpa'eF. For a periodization of the Hebrew language, cf. Angel Sáenz-Badillos and John Elwolde (transi.), A history of the Hebrew language (Cambridge: Cambridge University Press, 1993).

A few forms that are considered as probable nitpa'al or nitpa'el patterns are also found in Biblical Hebrew: Dt 21:8; Ez 23:48; Prv. 27:15. Cf. Paul Joiion, Grammaire de l'hébreu biblique (Roma: Pontificio Istituto Biblico, 1965 ), § 59f. On the vocalization cf. Pérez Fernández and Elwolde, An Introductory, 96: "Its vocalization as Nitpa'el is an incorrect composite form of copyists and publishers - the Sefardi and Yemenite traditions have retained the original pronunciation in a." 155

T e x t according to Israel W . Slotki, '"Erubin. transi, [from the Hebrew] into English with notes, glossary and indices," in Hebrew-English edition of the Babylonian Talmud. Seder Mo'ed (ed. I. Epstein; London: Soncino Press, 1983).

50

OUTLINE OF MIDDLE VOICE IN SYRIAC

qal - qum 'to stand, to stand up;' hithpael - hitqayyem and nitpaal- nitqayyam 'to be established, to be preserved.' 157

Siebesma analyses H e b r e w nifal, also in its relationship with the other passive and reflexive stems. H e concludes that it is not possible to formally distinguish between passive and reflexive functions, in Biblical Hebrew. T h i s is due to t h e tendency to leave t h e agent unexpressed: "One may ask t o what extent Biblical H e brew allows for t h e distinction reflexive/passive for the ni." H e notices that, in t h e Biblical T e x t , nifal forms to be translated w i t h a passive are m u c h more freq u e n t than the ones to be translated with a reflexive. T h i s result contrasts with the assumption of Joiion (1965), that t h e original meaning of t h e nifal is reflexive of t h e basic stem, and that only in some cases can it have a passive meaning. 1 5 9 b. Periphrastic

structures

Several Semitic dialects have nominal periphrasis that convey reflexive meaning. 1 6 0 T h e y are usually built with nouns indicating 'body,' body parts, 'soul,' followed by a pronominal suffix referring to the subject. For example: • • •

Akk. raman 'body': ramas-su ipattar 'he will redeem himself;' Classical Arabic nafiun 'soul': qatala nafia-hu 'he killed himself;' Misnaic and Medieval Hebrew 'esem 'bone:' wa-tippol asma-h min ha-ssus 'and she t h r e w herself f r o m t h e horse.'

157 Pieter Anne Siebesma, The function of the niph'al in Biblical Hebrew (Assen / Maastricht: Van Gorcum, 1991). Slotki, '"Erubin," 34. Joiion, Grammaire, § 51c. 160 See also § 1.1 n.10 for Syriac. 161 On the use of lexemes of such a semantic area in intensive structures in the world languages, cf. Edith A. Moravcsik, "Some crosslinguistic generalizations about the intensifier constructions," in Papers from the eighth regional meeting of the Chicago Linguistic Society (ed. P.E.A. Peranteau; Chicago: Chicago Linguistic Society, 1972), 271-277; on intensive and emphatic reflexives see also Ekkehard König and Peter Siemund, "Intensifiers and reflexives - a typological perspective," in Reflexives: Forms and Functions (ed. Z. Frajzyngier and T . Curl; Amsterdam - Philadelphia: Benjamins, 2000), 41-74.; Ekkehard König and Volker Gast, "Focused assertion of identity: A typology of intensifiers," Linguistic Typology, 10, 223-276. 162

Morris Epstein, Tales of Sendebar: an edition and translation of the Hebrew version of the Seven sages: based on unpublished manuscripts (Philadelphia: T h e Jewish Publication Society of America, 1967). 1. 525. Cf. Pérez Fernández and Elwolde, An Introductory, 45. T h e structure 'esem + pronoun with reflexive meaning is often preceded by the nota accusativi: JJUH 1QSJ) n x D'UTO DTK pN en ädäm mesím et 'asmo räsä' "a man cannot make himself guilty."

E T - FORMS AND INNER PASSIVES IN SEMITIC STUDIES

51

T h i s type of reflexive formation is quite common, even outside the Semitic group: both Vedic and Sanskrit, for example, reflexivity is expressed by nominal expressions with a semantic close to those described above. Beside the nominal element sva-y-am 'self,' Sanskrit uses the oblique cases of atman- 'soul,' and Vedic the oblique cases of tanu- 'body.' T h e semantic value of such reflexive structures often implies a certain degree of emphasis and intensification, in order to stress the fact that the event or action described is performed by the subject on itself, or in order to avoid ambiguity. c. Akkadian -tan- infix T h e Akkadian -tan- is a morpheme that has no parallel in any other Semitic language and therefore is commonly considered as an Akkadian innovation, as is the case also for the so-called perfect, a -t- infix with tempo-aspectual value (see infra § 3.1.7). For example, astanappar 'I sent you regularly,' aktanarrabakkum 'I continuously pray you.' As far as the etymology of this stem is concerned, according to Kurylowicz (1972) the presence of a nasal element in this morpheme is due to a reinterpretation of a second geminated root-consonant. In Akkadian a CC sequence can be interpreted as C+C or as n+C. A D stem uparris has a Dt reflexive uptarris in the preterite tense, uptarras in the present. Analysing -rr- as \njoioo -ojo^/J

I

ayk

qurbana

d-ahuh

d-metqabbal

hwa

conj. 'as'

'offer'

rel. part. + 'brother' +

rel part. +

III m. sg.

suff. pron III m. sg. ptc.etpa"al m. sg. pf. 'to be' 'received'

"As the gift of his brother that had been accepted." As regards the inner passives, in the compound tense with hwa, the generic value of resultative perfect is modified into the expression of an on-going state contemporary with the narrated events. Thus, in example 8 above, the presence of the auxiliary indicates that the state, the position of the master is described as concomitant to the main scene of Judas partaking in a feast. This is the case for all

267

nal sin.

Here reference is made to God's repeated act of creating men's lives, after the origi-

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

89

intransitive verbs, such as Ibes 'to wear, to put on': Aph. Dem. VIII, 369, 11 Ibis hwa "he was wearing." 4.5 ACTIONALITY As regards the actional value of the et- forms, the majority of the verbs in all the texts examined are achievements. T h e tables below reproduce the values in context, for the various texts examined, with respect to those of the corresponding basic stems. actional value etbases 107 achievement 83 18 12 accomplishment 1 2 gradual completion achievem./accompl. 1 18 achievem./grad. compl. 9 14 achievem./activity 36 42 activity 58 28 state state/achievem. 4 state/activity 9 Table 35. - Actional value of et- stems and corresponding bases in BLC actional value achievement accomplishment gradual completion achievem./accompl. achievem./grad. compl. achievem./state achievem./activity activity state state/activity/achievem. activity/state activity/accompl. not attested

et187 18 1 -

3 -

37 63 -

1 -

bases 139 3 -

10 9 8 5 56 28 21 15 4 12

Table 36. - Actional value of et- stems and corresponding bases in A T (1-7)

OUTLINE OF MIDDLE VOICE IN SYRIAC

actional value achievement accomplishment achievem. /activity achievem./grad. compl. achievem./accompl. activity state state/activity state/activity/achievem. activity/accompl./achiev. not attested

et84 4 -

3 19 -

bases 68 2 1 6 1 12 14 1 1 1 3

Table 37. - Actional value o f et- stems and corresponding bases in Aph. Dem. I

actional value achievement accomplishment gradual completion achievem./activity achievem./grad. compl. achievem./accompl. activity state activity/accompl./achiev. state/activity/achievem. state/activity not attested

et68 5 1 -

13 13 -

bases 37 11 1 1 2 4 14 3 1 3 7 16

Table 38. - Actional value o f et- stems and corresponding bases in Aph. Dem. VIII

actional value achievement achievem./activity accompl./achievem. activity state

et38 -

8 7

bases 33 2 1 11 6

Table 39. - Actional value o f et- stems and corresponding bases in Aph. Dem. XVI

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

actional value achievement accomplishment gradual completion achievem./accompl. achievem./state achievem./activity activity/grad. com pi. activity state state/activity/achievem. state/activity not attested

et-

240 2

1 1 1 -

19 25 -

91

bases 165 2 1 37 6 9 1 31 16 3 6 12

Table 40. - Actional value of et- stems and corresponding bases in Ephrem's Commentaries

In order to investigate the function of the et- prefix, with respect to Actionality, a comparison is necessary with the actional value of the corresponding basic stems. For a less fragmentary overview one can also consult the charts on pp. 115118 below. T h e ratio between the values of the et- stems and the corresponding bases is not homogeneous in the various texts. In BLC and in A T , most of the verbs encountered are achievements already in the basic stem. On the other hand, in the Demonstrationes and the Commentaries, there seem to be more achievements among the et- forms than among the corresponding bases. An important class of verbs that is relevant for establishing the behavior of cisterns with respect to Actionality are those that in the charts have been indicated as ambiguous. One immediately notices that such a category is quite well represented among the basic stems, but absent among the et- forms. T h e verbs that have been classified as ambiguous are those which admit more than one meaning and therefore more than one actional value in their basic stem. Among them is the verb hza, which, in the p'el, means "to see, to watch, to notice, to consider," and can be therefore a state, an activity or an achievement, but in the corresponding et- stem means "to appear, to be seen" and is thus telic and non-durative. Another example is the verb sma "to hear, to listen," a state or an activity, which in the et- stem means "to be heard" or "to obey," in both cases telic, achievement and accomplishment, respectively. Finally, from a different semantic area, the verb nuh "to cease," achievement, but also "to be at rest," state, which in the et- stem means "to take rest, to relieve oneself," in the sense of passing from a state to another (thus, also, "to die"), achievement. One can safely say that, when there is an ambiguity in the value of the basic stem, the et- prefix turns it into an achievement. One of the possible reasons for the apparent equivalence of actional values between basic and et- stems lies in the fact that, among the et- forms, there is a certain quantity of etpa'al stems: these are derived from the intensive-causative stem pa"el, which has been considered as the basic reference point for a comparison of

92

OUTLINE OF MIDDLE VOICE IN SYRIAC

the actional values. However, causatives generally have a higher probability of being telic and thus rank among achievements (and accomplishments). Another phenomenon that may be observed in the charts is the increasednumber of states in the et- stems, as compared with the basic stem, and the corresponding diminishment of the activities. This corresponds to the relevant amount of perfects found among the et- stems, some of which may also express a resulting state. T h e situation is much clearer for the inner passive participles. Here the predominance of the states is almost absolute. Nevertheless, activities are also fairly represented. T h e tables below summarize the results of my analysis. actional value achievement accomplishment achievem./activity achievem./ accompl. activity state state/achievem. state/activity

inn. pass. 17 3 -

1 49 6 -

bases 44 3 2 2 10 13 -

2

Table 41. — Actional value of inner passives and corresponding bases in BLC

268 js case, for example, of the verb dmar, attested in the etpa'al in A T IV, 209, 16, and in Aph. Dem. I, 40, 18. This verb in the p'el means 'to tremble,' in the pa"el 'to make marvelous, to make tremble in amazement.' T h e etpa'al describes the consequence of the pa'el, thus meaning 'to wonder, to be amazed,' as an achievement, a change of state.

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

actional value achievement accomplishment gradual completion achievem./actitvity achievem./grad. compl. achievem./ accompl. achievem./state activity/accompl. activity state state/activity state/accompl./ achiev. not attested

inn. pass. 21 4 -

-

10 69 3 -

bases 29 7 1 6 3 2 6 5 14 26 4 1 2

- Actional value of inner passives and corresponding b

actional value achievement activity state

inn. pass. -

25

bases 10 8 7

Table 43. - Actional valu of inner passives and corresponding bases in Aph. Dem. I

actional value achievement accomplishment achievem./state activity state state/activity not attested

inn. pass. 7 -

28 -

bases 10 3 3 11 1 2 5

Table 44. - Actional value of inner passives and corresponding bases in Aph. Dem. VIII

actional value achievement activity state

inn. pass. 1 -

7

bases 3 5 -

Table 4 5 . - Actional value of inner passives and corresponding bases in Aph. Dem. XVI

94

OUTLINE OF MIDDLE VOICE IN SYRIAC

inn. pass. 8

actional value achievement achievem./activity achievem./state activity state not attested

bases 18 2 1 8 2 6

-

1 28 -

Table 46. - Actional value of inner passives and corresponding bases Ephrem's Commentaries

It may be interesting to examine some examples of this category. 12. A T I, 176, 10 "And I saw it [sc. a hand]..."

kad

grir

lah

kalba

adv. 'when'

pass. ptc. m. sg.

prep, 'to'/dir.obj.

'dog'

'drag'

marker + pron. suff. Ill f. sg.

"...while a dog was dragging it." Here the inner passive participle grir expresses an activity and has also active transitive value. This verb belongs to a semantic category to which also verbs such as ehad and sqil (attested in the sample texts) also belong, both meaning "to take, to lift" or "to carry," the participles of which have active value. In the case of ehad and sqil, the passive participle means "holding," and therefore also "carrying."269 However, a passive meaning of such participles is also possible.270 These verbs are signaled as a semantic sub-group in some grammars , and their behavior has also been discussed by Goldenberg: T h e active meaning of some passive participles of the simple stem (whatever may be its origin) is in Syriac a special feature of certain verbs, mostly transitive, that signify "take, hold, carry" or the like. 272

269

Cf. also the following example: "One often comes across what is passive in form only, but active in meaning: e.g. L k 14.2 Lx> lot» „ 'a man who has collected [knîs hwâ]water,' i.e. hydropic; M k 14.13 Lx>, Vu, 'a man carrying [da-sqîï\ a water container (i.e. having picked up...and carrying):" Muraoka, Classical Syriac, 67. Cf. also Joiion, "Sémantique des verbes statifs," 25. 270 a m J 3 j v a ] e n c e 0 f v o i c e i s n o t uncommon in Syriac, another example was given in § 4.3 for the etp'el of the verb hzâ. Nôldeke, Kurzgefasste, § 280. 272

Goldenberg, "Aramaic Perfects," 118.

T E X T U A L ANALYSIS 1 : MIDDLE SEMANTICS

95

A possible explanation of the phenomenon can be seen in the origin of the inner passives from stative adjectives. Since the participle originally indicates a more or less permanent property of the subject, actions such as holding and carrying may also be seen in this perspective. Moreover, such events are also included in the semantic area that Kemmer defines as 'indirect middle,' and which encompasses all the actions through which the subject achieves the possession of something. Finally, the participle in I, 176, 10 expresses an ongoing action and has progressive aspectual value. It is interesting to notice that, from the verb gar, an etp'el is also available. Therefore, this example seems to contradict Muraoka's statement about the correlation between the passive participle and et- stem participles from the same roots, quoted in § 3.2.3. We saw in the previous paragraph that, in the compound tense with the verb 'to be,' the inner passive expresses a concomitant state. When the passive participle is formed from a transitive root, however, the value of the compound does not seem to be so much that of an on-going state, even though it results from a previous event. On the contrary the forms seem to have, rather, a Terminative value. Let us consider the following passage: 15. Aph. Dem. VIII, 381, 17 "A great favour was done to us..." ^nro Joo) psiq hwa sakran pass. ptc. m. sg. 'to III m. sg. pf. 'to 'hope' + pron. cut off be' suff. I pi.

"...our hope had been cut off, but there exists another one for us." In this case the semantics of the root, both in terms of transitivity in general and in terms of actionality (the verb is an achievement) has an influence on the aspectual value of the verbal compound. An analogous influence of transitivity and actionality on the aspectual value of this compound is also found in the structures of the type ktib It This structure c o r r e s p o n d s grosso modo t o t h e r o m a n c e perfects o f t h e k i n d habeo lectum librum.

It

is a possessive structure of the type mihi est, in which the event, expressed by the inner passive participle, is presented as a possession of the agent. In Aph. Dem. VIII, 368, 6 we find "Everyone will be rewarded in his body whatever..."

273 274

Kemmer, The Middle Voice, 74 ff.

T h i s structure will have fundamental developments in the modern Neo-Aramaic dialects. For a possible reconstruction of the origin cf. Eduard Y. Kutscher, "Two 'passive' constructions in Aramaic in the light of Persian," in Hebrew and Aramaic Studies (ed. B . - H . Zeev, A. Dotan and G. Sarfatti; Jerusalem: T h e Magness Press - T h e Hebrew University, 1977), 70-89. For a description of the distribution and variants in the Neo-Aramaic dialects and for their correlation with some Iranian equivalents cf. Fabrizio Pennacchietti, "Verbo neo-aramaico." Finally, for the parallel Iranian mana kartam type, cf. Geoffrey Haig, Alignment Change in Iranian Languages (Berlin - New York: Walter de Gruyter, 2008).

96

O U T L I N E OF M I D D L E VOICE IN SYRIAC

1 6 .

Joo)

)CL,JJ0

d-men

^ O j

qdim

'bid

hwa

leh

rel. part. + prep. adv. 'before' pass. ptc. m. sg. 'to III m. sg. pf. 'to 'from'

do'

be

prep, 'to' + pron. suff. ILL m. sg.

"...he has done before." In these structures the compound has a telic value and is most often an achievement, unlike the simple passive participle that, as was shown, is usually a state. 4.6 Voice, Aspect and Actionality Crossing the results of the analysis of Voice, Aspect and Actionality, the following conclusions can be drawn: Et- stems convey a spectrum of meanings going from self preservation, body care and indirect middle actions, more immediately connected with proper reflexivity, up to agentless passive and passive with agent, covering all the semantic areas that are indicated by Kemmer as being specific to middle voice. T h e most represented combination is the one of the agentless passive, with Terminative aspectual value (aorist or perfect in equal measure) and with actional value of achievement or, more seldom, accomplishment. Overall, the combination terminative - achievement appears to be the most frequent. States and Activities are basically concentrated among the following semantic categories: body actions, positionals, emotion middle, cognition middle, perception middle, and spontaneous events (these are, however, more frequently terminative and achievement, like agentless passives). These categories are also the ones among which the Non-Terminative aspects are more numerous (although, in minor measure, they are found through the entire spectrum),and which are almost always related with activities and states. T h e distribution of the more frequent values for each category can be summarized in the following table: 4

Terminative

achievement (accomplishment)

passive + agent

Terminative

achievement (accomplishment)

16

Terminative

achievement (accomplishment)

self preservation, 8,

Terminative

achievement (accomplishment)

Non-Terminative

activity/state

10, 11, 12, 13, 14, 15 self preservation, 8, 10, 11, 12, 13, 14, 15 T a b l e 47. - Combination of Voice and Action in the sample texts

Inner passives form a much more homogeneous category. From the aspectual point of view they are almost always resultative perfects. Their Aktionsart is in the

T E X T U A L ANALYSIS 1: MIDDLE SEMANTICS

97

great majority of cases state. T h e data for Syriac are completely in keeping with Haspelmath's description of passive participles.

275

Haspelmath, "Passive Participles."

CHARTS Voice

4

12

Chart 1- Semantic distribution of et- stems and inner passives in BLC



m 2

4

4

12

¡ H I III

567

8

10

11

13

14

16

16

Chart 2- Semantic distribution of et- stems and inner passives in A T (17)

4

12

13

16

Chart 3- Semantic distribution of et- stems and of passives in Aph. Dem. I

98

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

4

11

13

16

Chart 4 - Semantic distribution of et- stems and inner passives in Aph. Dem. VIII

4

16

Chart 5 - Semantic distribution of et- stems and inner passives in Aph. Dem. XVI

I 2

4

7

15

Chart 6 - Semantic distribution of et- stems and inner passives in Ephrem's Commentaries

100

OUTLINE OF MIDDLE VOICE IN SYRIAC

Aspect Legend: ao: aoristic at: attitudinal co: continuous fr: frequentative gn: gnomic ha: habitual: in: incohative

it: iterative nt: non-terminative pf: perfect pr: progressive te: terminative un: undetermined

et-stems

BLC

inner passives

Chart 7 - Aspect of et- stems and inner passives in B L C

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

et-stems

AT

inner p a s s i v e s

Chart 8 - Aspect of et- stems and inner passives in A T

102

OUTLINE OF MIDDLE VOICE IN SYRIAC

fr it te nt un ha at gn

co pr ao pf et-stems

AD1

inner passives

Chart 9 - Aspect of et- stems and inner passives in A p h . D e m . I

T E X T U A L ANALYSIS 1: MIDDLE SEMANTICS

et-stems

AD8

inner passives

Chart 10 - Aspect of et- stems and inner passives in Aph. Dem. VIII

OUTLINE OF MIDDLE VOICE IN SYRIAC

fr it te nt un ha at gn co pr ao pf et-stems

AD16

inner p a s s i v e s

Chart 11 - Aspect of et- stems and inner passives in A p h . D e m . XVI

T E X T U A L ANALYSIS 1 : MIDDLE SEMANTICS

fr it te nt un ha at gn I co pr

I 1

| I

ao pf

et-stems

EC

inner p a s s i v e s

Chart 12 - Aspect of et- stems and inner passives in Ephrem's Commentaries

105

106

OUTLINE OF MIDDLE VOICE IN SYRIAC

Aktionsart Legend: ac: accomplishment ah: achievement at: activity gc: gradual completion na: not attested st: state

gc na ac 1

1 st at

13,,=:,

1/'

' et-stems

ah BLC

inner passives

Chart 13 - Actional values of et- stems and inner passives (light grey), correlated w i t h those of the corresponding basic stems (dark grey), in BLC

TEXTUAL ANALYSIS 1 : MIDDLE SEMANTICS

GC

na 1

ac

1

St at ah

I?:

et-stems

AT

inner passives

Chart 14 - Actional values of et- stems and inner passives (light grey), correlated with those of the corresponding basic stems (dark grey), in AT

gc na r

ac st

-

at ah

1 et-stems

AD1

inner passives

Chart 15 - Actional values of et- stems and inner passives (light grey), correlated with those of the corresponding basic stems (dark grey), in Aph. Dem. I

107

108

OUTLINE OF MIDDLE VOICE IN SYRIAC

gc na ac St at ah

1 et-stems

AD8

inner passives

Chart 16 - Actional values of et- stems and inner passives (light grey), correlated with those of the corresponding basic stems (dark grey), in Aph. Dem. VIII

gc na ac st at ah

1

et-stems AD16 inner passives Chart 17 - Actional values of et- stems and inner passives (light grey), correlated with those of the corresponding basic stems (dark grey), in Aph. Dem. XVI

T E X T U A L ANALYSIS 1 : MIDDLE SEMANTICS

gc na ac St at ah

I--; • • et-stems

EC

inner passives

Chart 18 - Actional values of et- stems and inner passives (light grey), correlated with those of the corresponding basic stems (dark grey), in Ephrem's Commentaries

109

5

TEXTUAL ANALYSIS 2: COREFERENTIAL DATIVE

5.1 INTRODUCTION: SYNTACTIC ASPECTS

T h i s chapter describes some syntactical and semantic features of the coreferential dative, a phenomenon closely related to middle voice and, as I will try to show, to unaccusativity. T h e aim of this analysis is to establish a semantic connection between the verbal forms with an et- prefix on the one hand, and other verbal forms which, although not displaying a specific morphological mark, can be, for various reasons, classified as unaccusatives and semantically middle. Such a connection can be highlighted by examining a syntactical property that all these verbal forms share: coreferential dative. In a number of classical Syriac texts a structure is found in which a verbal form is immediately followed by the preposition /- attached to a suffix pronoun, coreferential with the subject of the verb. T h i s structure is well represented in many classical Syriac works, but so far little attention has been devoted to it in linguistic research, and hardly any in the main reference grammars of the Syriac language. T h e verbal form can be either a basic or a derived stem, active or with et- prefix. In order to give an account of this phenomenon, attention will be paid first to its variant with et- stems, which is represented with particular frequency in Aphrahat's Demonstrationes.

276 rpj^ m a j n reference studies on this topic are Jan Joosten, "The Function of the socalled Dativus Ethicus in Classical Syriac," Orientalia, 58, 473-492 and Riccardo Contini, "Considerazioni sul presunto 'dativo etico' in Aramaico pre-cristiano," in Études sémitiques et samaritaines^offertes à Jean Margain (ed. C.-B. Amphoux, A. Frey and U. Schattner-Rieser; Lausanne: Editions du Zèbre, 1998a) 83-94, who gives an extensive review of the current bibliography on the subject. The same phenomenon has been noticed for Biblical Hebrew, and secondarily for Syriac, by Muraoka, who uses for it the term centripetal dative (cf. T a kamitsu Muraoka, "On the so-called Dativus Ethicus in Hebrew," The Journal of Theological Studies, New Series, 29, 495-498). More recently, Gideon Goldenberg, "Actants and Diathesis, Directions of Transitivity & c.: Some Satzgestaltungen and Their Background in Semitic and Elsewhere," in From Beyond the Mediterranean. Akten des 7. internationalen Semitohamitistenskongresses (ed. R. Voigt; Aachen: Shaker, 2007), 283-296 has included this structure in a concise but incisive linguistic study on transitivity and diathesis. 277 Aphrahat's Demonstrationes contain 77,222 words (counted according to Syriac spelling). T h e et- forms followed by a coreferential dative are 91, cf. Appendix 2 for a complete list.

Ill

112

OUTLINE OF MIDDLE VOICE IN SYRIAC

As regards voice, the verb in t h e structure examined here has passive value and t h e cluster /- + pronominal suffix does not express a dative, b u t anaphorically recalls t h e grammatical subject, namely t h e patient of t h e action described by t h e verb. T h u s , for example, in D e m . II 52, 10-11 278 is found:

kad

mta

adv.

Ill m. sg.

conj. 'time' + pron.

ger

zabneh

'when'

pf. 'to ar-

'for'

suff. Ill m.

rive'

d-namosa rel part. + 'law'

ettawsap

leh

III m. sg. pf. prep, dat./obj. etp. 'to add'

sg.

mark. + pron. suff. Ill m. sg.

"(For) w h e n t h e time of the law arrived, it [i.e. t h e law] was added." T h e structure ettawsap leh is composed of the verbal form ettawsap and of the syntagm leh, in which t h e pronominal suffix refers to namosa, t h e logical subject of t h e verb and t h e patient of the action described by it. It is important to notice that this sentence is at the beginning of a new section of t h e text and no other possible referent for leh is available. O n e could consider this as an impersonal structure, in which leh simply indicates t h e direct object of t h e verb. 279 However, t h e range of structures found in t h e Demonstrationes is more complex. In D e m . II 56, 2 t h e following is found: 2. )ooi^>/, oti.;j w-etmalkat conj. 'and' + III

JlOj^ii.

ioYsDl/o

lah

'abduta

'al

zar'eh

prep, dat./obj.

'slavery'

prep, 'on,

'seed, off-

rel part. +

spring' +

'Abraham'

f. sg. pf. etp. 'to mark. + pron. promise'

suff. Ill f. sg

to'

d-abraham

pron. suff. Ill m. sg.

"And slavery was promised to the offspring of Abraham." 2 8 0 Here t h e et- form etmalkat is a 3 r d ps. f. sg., refers t h e n o u n 'abduta 'slavery,' coreferential with t h e pronominal suffix in lah, placed immediately after the verb. T h e agreement in the feminine gender clearly shows that no impersonal structure is involved here. Likewise, in D e m . II 57, 13-14:

278 r p ^ Syriac text and the Latin translation, correspond to the ones provided by Parisot, "Aphrahatis Demonstrationes." T h e passages are quoted according to the column and line numbers in that edition. 279 For the use of the preposition /- as a direct object marker, see infra § 5.2. 280 Also in Dem. II 53, 23; 60, 10; 60, 14; 84, 20.

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

3.

113

Vj o)K.lhar> vpo»\ a ^ ^ l / Lœovn ^ ^ ^O}Qa*JLO« hûsâhaivhy

gêr

m. pl. 'use' + pron. suff. Ill

conj.

d-nâmôsâ

m. sg.

'for'

rel. part. + 'law'

etbattalw

l-hôn

III m. pl. pf.

prep. dat./obj. mark. +

etpa. 'to

pron. suff. III m. pl.

cease'

b-me'tîteh

d-mahyânan

prep, 'in, with' + 'arrival' + rel. part. + 'life-giving' + pron. suff. III m. sg.

pron. suff. I pl.

"For t h e uses of t h e law ceased w i t h t h e arrival of our Savior." 281 In this example t h e a g r e e m e n t at t h e 3 r d ps. m . pi. of t h e verb, t h e subject and t h e suffix p r o n o u n , proves t h a t this is n o t an impersonal structure. 5 . 2 ( P R E F E R E N T I A L DATIVE , DATIVUS ETHICUS AND OBJECT MARKER 5.2.1 T h e Classical Syriac Grammars N o n e of t h e traditional grammars of t h e Syriac language explicitly m e n t i o n s this kind of construction. Nevertheless, traces of it can be f o u n d in some marginal o b i

>282

servations in connection w i t h t h e so-called 'dativus ethicus.' following remark, in Duval: 2 8 3

L e t us consider t h e

Le pronom suffixe avec lâmadh se met souvent après certains verbes, comme pronom réfléchi, mais sans ajouter beaucoup au sens du verbe; on peut comparer nos verbes neutres, s'en aller, s'enfuir, se mourir; tels sont notamment ezal leh il s'en alla, 'rab leh - il s'enfuit, npaq leh - il sortit, nhet leh - il descendit, qam leh il se leva, etâ leh - il vint, mît leh - il mourut, gha leh - il voulut, stâ leh - il était insensé, îteh leh - il était assis, hwa leh - il exista, shar leh - il pensa, estakah leh - il fut trouvé, shaq leh - il laissa (après lui) etc.

281

Similarly Dem. II 57, 25. On the use of the term dativus ethicus cf. Contini, "Considerazioni sul presunto." It is important to distinguish the uses discussed here from the dativus ethicus proper. T h e latter structure is a dative that accompanies the event described by the verb in a given utterance, but without displaying any coreferentiality, either with the object or with the subject of the verb. T h e function of the dative is in this case to underline an involvement of the speaker and of the interlocutor in the event, e.g. the English "I'll rhyme you so eight years together" (Shakespeare, As You Like It, iii. 2.) or the Latin "At tibi repente paucis post diebus [...] venit ad me Caninius" (Cic. Fam. IX.2.1). 282

283

Duval, Traité, § 305c. Cf. also, along the same lines but less detailed, Nöldeke, Kurzgefasste, § 224: "Die Präposition mit reflexivem Personalpronomen steht oft bei einem Verbum, ohne dessen Bedeutung wesentlich zu modificieren (Dativus Ethicus) [...] sehr oft bei Verben der Bewegung [...] und so grade bei Afr. [sc. Aphrahat] oft neben Passiven u.s.w."

114

OUTLINE OF MIDDLE VOICE IN SYRIAC

A b o u t t h i s passage some observations are o p p o r t u n e . First of all, as regards t h e verbs listed by Duval, it is n o t e w o r t h y t h a t all of t h e m b u t one ("estakâh leh - il f u t trouvé") are active in f o r m , t h u s t h e y do n o t seem t o b e l o n g t o t h e same typology of t h o s e involved in t h e structure used by A p h r a h a t . Moreover, it is possible to p o i n t t o some features c o m m o n to t h e verbs listed above, f r o m a semantic p o i n t of view: t h e y are all verbs expressing m o t i o n , existence or e m o t i o n . Observations similar t o Duval's, b u t m o r e systematic, are also f o u n d in Nowicki's accurate study of E p h r e m ' s syntax. In t h e section of his w o r k in w h i c h prepositions are analyzed, a paragraph is devoted t o t h e dativus ethicus. T h e scholar lists h e r e grosso modo t h e same examples given by Duval (with specific reference t o E p h r e m ' s prose), b u t organized in semantic categories: Verben der Bewegung der Wahrnehmung der Gemütsbewegung die einen Mangel bezeichnen, die einen Zustand ausdrücken. O n c e again, only active verbs are listed, and n o c o n s t r u c t i o n w i t h an et- stem is m e n t i o n e d . T h i s usage of t h e dative has also been recognized for Biblical H e b r e w by J o ü o n , 2 8 w h o prefers t h e definition of dativus commodi:287 Le / du dativus commodi est employé d'une façon très particulière avec le pronom de la même personne que celle du verbe. On obtient ainsi, surtout avec les verbes intransitifs (particulièrement avec les verbes du mouvement et leurs contraires) une nuance réfléchie indirecte, qui peut équivaloir à peu près à certaine nuance d'une forme verbale réfléchie (p. ex. nifal). A n o t h e r use of t h e preposition /- t h a t is useful to recall, and w h i c h was m e n t i o n e d in t h e previous section, is t h a t of object m a r k e r , w i t h p r o n o m i n a l objects and, less frequently, w i t h n o u n s . T h e presence of a particle t h a t introduces t h e 284

Paul Nowicki, Beiträge zur Syntax inAphrems Memren (Freiburg: 1916).

285

Nowicki, Beiträge, 22-23.

Joüon, Grammaire. Joüon, Grammaire, § 133d). In Paul Joüon and Takamitsu Muraoka, A Grammar of Biblical Hebrew (Revised edition) (Roma: Pontificio Istituto Biblico, 2006), § 133 is enriched with new bibliographical references on the topic, such as the already mentioned Muraoka, "On the so-called." 287

288

Cf. Joüon and Muraoka, A Grammar, 77-79. For direct object markers in some classical Semitic languages, see Rubin, Studies in Semitic, 91-110. In the fifth chapter of his study, Rubin also offers a typological comparison with the direct object markers in Romance languages and in Hindi (110-112) and analyzes the grammaticalization process that underlies the shift from 'dative' to 'accusative' preposition (112-115). As regards Aramaic, Rubin individuates the reason for the rise of direct object markers in "the loss of inherited accusative case." (112). Moreover, the distribution of the direct object markers in the earliest Aramaic dialect shows that they were at first used only in combination with animate objects, and progressively extended also to inanimate (or less animate) direct objects (112). T h u s the original role of these markers was to stress the 'irregular' syntactic status of objects, for lexemes that were considered as prototypically subjects. T h i s also explains why the dative preposition is often used in this function: it is the marker of the indirect object, the one that appears most frequently with animate entities (114).

115

T E X T U A L ANALYSIS 2 : C O R E F E E N T I A L D A T I V E

direct object of a verb, under particular syntactic and semantic circumstances 289 is common to several Semitic languages, from the 1 st millennium B.C. onwards. In this function, among others, Aramaic, Hebrew and some Arabic dialects have also adopted the dative preposition /-. As we will see in the course of this chapter, the role of /- as object marker is important for understanding the semantic relevance of coreferential dative. For the moment, let us examine only one example of this use: 4. Jn. 1,43 (Pesitta version): "And another day Jesus wanted to depart from Galilee..." w-eskah

l-pilipos

conj. 'and' + III m.

prep, dat./obj. mark. +

sg. pf. p'el 'to

pron. suff. Ill m. s. +

find'

'Philip'

"...and he found Philip." 5.2.2 A First Key to Interpretation: Aphrahat's Quotations from the Scriptures Let us now examine again some examples of the structure found in Aphrahat. In two passages it occurs in quotations from the New Testament, but each time with a certain degree of modification of the text, that is indeed related to the insertion of the group verb + /- + pron. suff. Thus, Dem. II 52, 11 (see ex. 1. above) is related to Gal. 3:19: 5 .Otrer.) ) } I) t-M. Krev.^^^O IgOOYH mana hakil namosa conj. 'why'

conj. 'therefore'

'law'

mettul

mastyanuta

ettawsap

conj. 'because

'transgression'

III m. sg. pf.

of

etp. 'to add'

"Why then the law? Because of transgression it was added." T h e original passage shows a regular construction, in which the noun namosa is the subject of a verb in an et- stem, with passive value: ettawsap. In Aphrahat's rephrasing, however, the verb is followed by leh, referring to the subject of the verb, assigning to it the semantic role of patient. A hybrid construction arises, in between a real passive and an impersonal followed by a direct object. Dem. Ill 116, 24-26 presents a construction that seems to be analogous to the ones thus far analyzed:

A s far as Syriac is concerned, Nöldeke, Kurzgefasste, § 288 observes that " D a s / dient sonst facultativ zur Bezeichnung des determinierten Objects. Die Determination ist stärker, wenn das dem Subst. entsprechende Objectsuffix am Verbum hinzukommt." 289

290 r p j ^

s a m e

voo»;, UAo cA

kind of rephrasing occurs also with the verb npal, in D e m . V 201, 1-2 po

[w-kad npal leh klilä d-rishon] 'cum autem cecidit corona capitis eorum,'

cf. with citation from L a m . 5:16 in 201, 7-8.

116

OUTLINE OF MIDDLE VOICE IN SYRIAC

wa-b-dina da-sba la-mdän ettdin w-ba-kyälä conj. 'and' + prep. rel. part. + III m. sg. prep, 'to' + inf. 'to III m. sg. pf. conj. and' + prep. 'in, with' + 'judgpf. 'to want' judge' ettaf. 'in, with' + ment' 'to judge' 'measure' da-ba rel. part. + III m. sg. pf. 'to seek, desire'

la-mkalu prep, 'to' + inf. 'to measure'

ettkil III m. sg. pf. ettaf. 'to measure'

leh w-ayk prep, dat./obj. conj. 'and' + conj. mark. + pron. 'like, as' suff. Ill m. sg.

d-etthassab l-me'bad et'bed leh rel. part. + III m. prep, 'to' + inf. 'to do' III m. sg. pf. etp. prep, dat./obj. sg. pf. etpa. 'to 'to do' mark. + pron. think, reckon' suff. Ill m. sg.

"With the judgment with which he wanted to judge he was judged, and with the measure with which he wanted to measure he was measured, and as he planned to do it was done to him." In this passage Aphrahat is rephrasing Mt. 7:2:291

b-dina prep, 'in, with' + 'judgment'

ger conj.

d-daynin rel. part. + part. m. pi. 'to judge'

w-ba-kyalta da-mkilin conj. 'and' + prep. rel. part. + inf. 'in, with' + 'meas- part. m. pi. afel ure' 'to measure'

anton tetdinun92 pron. ind. II II m. pi. pf. ettaf. m. pi. 'to judge'

anton pron. ind. II m. pi.

mettkil l-kon part. m. pi. ettaf. 'to prep, dat./obj. mesure' mark. + pron. suff. II m. pi.

"With the judgment that (with which) you judged you will be judged and with the measure that (with which) you measured it will be measured to you." In this case the model of the Gospel suggests a different and more complex interpretation for the construction used by Aphrahat. T h e author rephrases the passage in Mt. 7:2 according to the context, the 2 n ps. pi. is transformed into a 3r ps. sg. T h e text of the Gospel, moreover, is a literal translation from the Greek ev a)i

yap

Kpin/m

icpivete

icpiSrioeTe,

icai

ev

a)i

nitpan

^tetpeiTe

^leTpriSrioeTai i)^iiv.293 therefore the cluster l-kon renders a real dative here. T h e 291

Partially corresponding also to Mk. 4,24 and Lk. 6, 38. 292 r p j ^ Curetonian version of the Gospel has here the variant ^oW antòn], with participial construction. 293

[metdinin

Eberhard Nestle, Erwin Nestle, Barbara Aland and Kurt Aland, Novum Testamentum Graece (Stuttgart: Deutsche Bibelgesellschaft, 2001). Lit. "In that judgement in which you judge you will be judged, and that measure in which you measure, it will be measured to you." T h e last portion of the verse meaning 'you will receive'. Gr. [letpeco TL TLVL = 'to

117

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

verb kul, in Syriac, is not normally followed b j on the contrary, this construction seems to be triggered by the Greek model 94. An interesting chain of structures is thus generated in the passage of Demonstration III: a direct transitive ettdin is followed by an ambiguous structure ettkil leh (in which leh can be meant either as dative or as a nota accusativi) and by a final real dative et'bed leh, that summarizes and generalizes the two preceding statements, and is not in Mt. 7:2. 295 5.3 (PREFERENTIAL DATIVE WITH THE OTHER VERBAL STEMS

5.3.1 Demonstrationes Another series of passages deserves attention. In a number of occurrences a 'dative' (/- + pron. suff.) coreferential with the subject of the verb follows an active verb, instead of an et- form. Such a structure corresponds, therefore, to the typology described by Duval and Joiion. Dem. II 60, 13-14:

men

zabnâ

gêr

prep, 'from'

'time'

conj.

d-etyahbat

hdatâ

rei. part. + III f. sg.

'new'

pf. etp. 'to give'

betlat

lâh

III f. sg. pf. 'to prep, dat./obj. mark. cease'

'tîqtâ 'old, ancient'

+ pron. suff. Ill f. s

g

"From the time the new [Testament] was given, the old one ceased." T h e verb btel 'to cease' is followed by lâh, referring to the subject 'tîqtâ 'old one' (namely dyatîqî 'Testament,' feminine).

give portion of something to someone.' 294 For the translation techniques in early Syriac texts see Sebastian Brock, "Greek into Syriac and Syriac into Greek," Journal of the Syriac Academy, 3, 406-422; "Aspects of translation technique in antiquity," Greek Roman and Byzantine Studies, 20, 69-87; "Towards a history of translation technique," in Ille Symposium Syriacum, 1980. Les contacts du monde syriaque avec les autres cultures (ed. R. Lavenant, S. J.; Roma: 1983), 1-14. For the Gospel quotations in Aphrahat see Tjitze Baarda, The Gospel quotations ofAphrahat, the Persian sage (Amsterdam: 1975). 295 A deeper examination would be necessary, from both the Greek and Syriac point of view. This passage poses interesting problems as regards the biblical sources used by Aphrahat, the translation and quotation methods and the influence of the language of the Greek original on early Syriac.

118

O U T L I N E OF M I D D L E VOICE IN SYRIAC

w-en

nestkah

conj. 'and' + conj. 'if III m. sg. ipf. etp. 'to

ne kid

b-'arqawhy

d-bayta

'deceit, guile'

prep, 'in' + 'beam,

rel. part. +

find'

nepel III m. sg. pf. 'fall'

framework'

leh

kulleh

prep, dat./obj. mark. + 'all' + pron. suff. pron. suff. Ill m. sg.

'house'

benyana 'building'

Ill m. sg.

"If deceit was found in the framework of the house, the entire building would fall down." 10. napla

lah

kullah

haymanuta

part. f. sg. 'fall'

prep, dat./obj.

'all' + pron. suff. Ill

'faith'

mark. + pron.

f. sg.

suff. Ill f. sg

"All faith collapses." Dem. II 84, 17-18 and 84, 20: the verb npal 'to fall' is followed by /- + pron. referring to the subject of the verb, in one passage in the masculine gender, in the other in the feminine.

5.3.2 The Acts of Judas Thomas In the first seven Acts of Thomas thirty-seven occurrences of the coreferential dative are found. Of those, only five cases are in combination with et- stems. Examining the aspectual value of the remaining thirty-two cases with non-etstems, one finds that twenty-five of t h e m have aoristic value, three of t h e m are perfects, three progressives and one has gnomic value. One observes, therefore, a clear predominance of the aoristic aspect. T h i s feature emerges clearly from the comparison of the following passages: 11. A T I, 182, 11 ^-JliK. jO voV ••."I« /

eskah

ennon

kad

yatibin

III m. sf. pf. 'find'

pron. II pi.

adv. 'when'

ptc. pass. m. pi. 'sit'

"He found t h e m sitting."

296

Namely: I, 177, 17 estahlap leh 'it had changed/was t r a n s f o r m e d ; ' IV, 210, 2 mestqla lah mennyi 'will be taken away f r o m m e ' ; VI, 220, 11 etkses leh men ewkaristiya 'he was rejected f r o m t h e Eucharist;' VII, 244, 2 la mestle leh 'he has n o t been forced;' VII, 248, 5 mestren Ihen 'are loosened.' Cf. A p p e n d i x 2 for a list of all t h e coreferential datives f o u n d in t h e first seven Acts.

119

T E X T U A L ANALYSIS 2 : COREFEENTIAL DATIVE

12. A T I, 178, 2 Q|\~io\ w-kad

¡^ h vi\« Jlyoj p o zamarta

conj. 'and + adv. 'when'

selmat

'song'

itbat

III f. sg. pf. 'cease, end'

lah

l-qubleh

Ill f. sg. prep, dat./obj. prep, 'in front of + pf. 'sit' mark. + pron. suff. pron. suff. Ill m. Ill f. sg sg.

"And when the song finished she went and sat next to him." In these two passages the same verb iteb 'to sit' is found. T h i s verb, just like the English 'to sit,' can indicate both a state and an achievement, according to its syntactic context. In the first sentence the verb is a 'passive' participle of the basic stem and it has the progressive aspect 'to be seated.' In the second sentence, instead, the verb is in the suffix conjugation of the basic stem, followed by the coreferential dative, and it has aoristic value: it indicates the moment in which the subject (a dancer) takes her place next to Judas. In 11, thus, the presence of the coreferential dative gives an aoristic nuance to the verbal root, as the adverb 'down' does in the English language with the verb 'to sit down.' Another clear example of the aoristic value given by the coreferential dative is the following: 13. A T I, 193, 19 vpojjojaj Joo) otnSio w-pkah

leh

conj. 'and' + III m. sg. pf. prep, dat./obj. mark 'to grow pale' + pron. suff. Ill m. s

nürhon 'light' + pron. suff. Ill m. sg.

g-

"and their light became pale." T h i s can be compared with the following occurrence of the same verb, also in the basic stem of the suffix conjugation, in Aphrahat: 14. Dem. XIV 577,21 I. A vt lo*n2> : Ijojaj ^JL*. yOOUioQ^o wa-byawmayhon

[...] pekhat

melha

conj. 'and' + prep, 'in' + 'days' + pron. suff. Ill m. pi.

III f. sg. pf. 'grow pale, lose flavour'

'salt'

"And in their days [sc. of the evil rulers] the salt lost flavor." Here the form pekhat is a Perfect from the aspectual point of view, and the coreferential dative is absent. T h i s feature of coreferential dative has already been pointed out by Joosten. 297 T h e scholar notices, for example, that "when a stative verb is accompanied by the DE [sc. dativus ethicus] it never means 'to be -' but

297

Joosten, " T h e Function."

120

OUTLINE OF MIDDLE VOICE IN SYRIAC

invariably 'to become -'." 298 Joosten analyses and classifies the occurrences of such a structure in a corpus of early Syriac texts consisting ofthe Pesitta translation of the New Testament, Aphrahat's first seven Demonstrationes and the Acts of Thomas. His conclusion is that "the D E defines the content of the preceding verb-form as an entering into the state associated with the verb."299 Further remarks on this theory will be presented in § 5.7.1. For the moment it is interesting to notice that the number of et- forms followed by a coreferential dative found in the Acts of Thomas is strikingly smaller than that of those found in Aphrahat. As we will see, this is also true for the other Syriac texts examined (with variable proportion). T h e combination of coreferential dative with the et- stems seems to be, therefore, a peculiar characteristic of Aphrahat's style. 5 . 4 FIRST CONCLUSIONS: SEMANTIC AND SYNTACTICO-SEMANTIC ASPECTS So far it has been shown how the structures with et- stems and those with active forms followed by the coreferential dative are fundamentally analogous, both from the formal and syntactical point of view. T h e two structures share the following features: •

T h e verb is immediately followed by the preposition /- (generally used to express a dative or to introduce the direct object of a verb) and by a pronominal suffix.



T h e suffix pronoun is coreferential with the subject of the verb, usually already mentioned before, and very close to it, in the sentence (in 8 and 9, however, the suffix pronoun precedes the grammatical subject).

It is also possible to establish a semantic link between the structures analyzed so far, that all seem to be expressions of the same linguistic feature. All the verbal forms with et- prefixes in the passages so far analyzed are passive in meaning, according to both the semantic value of the verbal root and to the syntactical context. T h e subject of the verb plays the role of a patient. T h e other verbs examined here share the same feature: they are highly unagentive verbs, their subject undergoes the action described, or has very low control on it. Verbs such as 'cease,' 'fall' belong, in fact, to the category of unaccusatives, being characterized by low volitionality and by the fact that their subject is an object (or a patient) in the deeper structure of the sentence. This last property is

298

Joosten, " T h e Function," 474.

299

Joosten, " T h e Function," 475.

For a complete list of all the et- forms + coreferential datives occurring in the Demonstrationes cf. Appendix 2. 300

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

121

morphologically expressed in the construction analyzed in this chapter, in the cluster /- + pron. s u f f , in which a pronoun referring to the subject is introduced by the particle that normally precedes the direct object, in order to stress the split between (prototypical) syntactic and semantic roles. Unaccusativity can receive different morphological and syntactical expressions in the various languages, and can emerge more or less systematically (or even not emerge at all). In the text analyzed here it is possible to identify an epiphenomenon of the link existing between medio-passive morphological structures (verbal forms with et- prefix) and unaccusative verbal roots. 15. Aph. Dem. I 8,12-14: ^ ^ O u JJ j j o j ' . . VI ^ J D O : Imisoo JJ w-benyâneh

L.O-,-

men

conj. 'and' + 'building' + prep, 'from, pron. suffi Ill m. sg.

w

^OO

:

JJ JJS^

n)i.i->n

ga[le



mettzi

'wave' pi.

neg. part, 'not'

part. m. sg. etp. 'to



metnkê

w-men

quake'

w-men

ruhe

conj. 'and' + prep, 'from

'wind' pi.

mahsûlê

la

napel

'storm' pi.

neg. part,

part. m. sg. att.

'not'

'to fall'

neg. part, 'not' part. m. sg. etp. 'to conj. 'and' + prep, 'from damage, injure'

by'

"And the building is not shaken by the waves, is not damaged by the wind and does not fall because of the storm." T h i s is a series of parallel structures, used by the author to describe analogous situations. T h e first two sentences contain a participle of an et- form with a passive value (mettzi, metnke), preceded by the negative particle la and by the agent, regularly expressed through the preposition men 'from, by' (men galle, men ruhe). T h e third sentence consists of a participle active in form (napel), preceded by la and by the same group men + noun (men mahsule), indicating the agent of an action that the grammatical subject undergoes, even though expressed by an active verb. T h e verb npal is therefore treated, in this passage, in the same way as the forms with an et- prefix, from both the syntactic and semantic point of view .

301 An analogous phenomenon was noticed for Biblical Hebrew in § 2.3.1 of this book. In that case reference was made to the particle et; however, the same use is found also for the Hebrew preposition /- (later biblical texts and Rabbinic language, cf. Joiion, Grammaire, § 1251), e.g. ICh. 3,2 Dbunxb ,U^t7U;n ha-slisi h-'absalom "the third (is) Absalom," 2Ch. 7,21 DUP vbp ""QJ> b^h h-hal-'ober 'alayw yissom "and every one passing by it will be astonished...". Finally, cf. in this respect also a marginal, but interesting, observation in Muraoka, Classical Syriac, 79: "Only one of two direct objects may be marked by the preposition Lamadh, and the object so marked is mostly the grammatical subject of the underlying active voice clause." 302

Another interpretation is also possible. The participles, could simply have middle

122

OUTLINE OF MIDDLE VOICE IN SYRIAC

T h e final step of this analysis is the specification of the link between the forms with an et- prefix and the more general category of the verbs that can be followed by a dativus ethicus, according to the descriptions found in Duval, Nowicki and Joiion. For this purpose it is only necessary to recall the semantic classification made by Nowicki already mentioned above: verbs of motion, perception, affection, expressing an absence and state. At least five of these classes can easily be considered as part of the unaccusative category. 5.5 O T H E R PROPERTIES OF COREFERENTIAL DATIVE IN T H E DEMONSTRATIONES AND IN O T H E R SYRIAC T E X T S

5.5.1 The Demonstrationes A n overview of the structure in the Demonstrationes enlightens some other interesting features. First of all, as already shown in ex. 6, the phenomenon is not limited to the 3 r person, but is also attested for the other persons, even if less frequently. T h u s , in Dem. IV 173, 23 we find the 2 n ps. m. sg. ^O ^ ^ ^ ^ u [tetbattal lak men sluta] "you shall cease from prayer", and in Dem. VII 332, 3-4 the 1 s t ps. sg. ^

JJiW

-j3tiLo W IW

W [d-dalma

end d-la-'hrane

akrzet

end qnumy

estle It],

rendered by Parisot as "ne, cum ceteris praedicaverim, ipse reprobus efficiar." As far as voice is concerned, a minor (but relevant) number of passages shows a middle or reflexive value, rather than a passive one. T h u s , for example, in Dem. VI 293, 12 l a ^ i v » [mettamra lah ruha] "spiritus absconditur" and Dem. VIII 364, 24-25 [w-'al d-methabbal leh pagra] "de corporis autem corruptione" (lit. "of the body corrupting") are middle in meaning, rather than passive; Dem. X I 469, 14 voo>\ a ^ s o i / L^>o [wa-bnaya ethpekw Ihon] "filii [...] se converterunt", has middle value, but the presence o f the coreferential dative makes it closer to the reflexive pole. Finally, it is sometimes the case that the very same construction is repeated within a few lines, with and without the coreferential dative, as in Dem. VI 293,1214:

unagentive value, rather than passive value, and the preposition men (usually indicating an agent) could simply introduce the causer of the event, always with low volitionality. Similarly, et- form participles and active participles of the verb npal receive the same syntactical treatment and are considered semantically analogous.

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

123

16. lâh

rûhâ

ptc. f. sg. etp. 'to

mettamra

prep, dat./obj.

'wind,

hide'

mark. + pron. suff.

spirit'

napsânâytâ 'living, animal'

am prep, 'with'

Ill f. sg.

pagra

w-regstâ w-regsta

mestaqla

lâh

menneh

'body'

and + 'perception'

ptc. f. sg.

prep, dat./obj.

perp. 'from' +

etp. 'to lift,

mark. + pron.

pron. suffi Ill

take away'

suff. Ill f. sg.

m. sg.

" T h e animal spirit is hidden with the body and sensation is taken away from it [sc. 'the body']." and in D e m . VI 293, 23-24: 17 .ou^o JJjoiSJL^o JiSJL^o

IjYi^too IK .1« on L,o;o

w-ruha

napsanayta

and + 'wind, spirit'

'living, animal'

mettamra hide'

ba-kyanah prep, 'in, with' + 'nature'

lah

ptc. f. sg. etp. 'to prep, dat./obj. mark. + pron. suff. Ill f. sg.

w-regsta

mestaqla

and + 'perception'

ptc. f. sg. etp. 'to

prep, 'from' + pron.

mennah

lift, take away'

suff. Ill f. sg.

" T h e animal spirit is hidden in its nature and sensation is taken away from it [sc. 'the spirit']." In 11. w-regsta mestaqla lah menneh, a coreferential dative is found; a few lines later, instead, t h e very same concept w-regsta mestaqla mennah is formulated w i t h out t h e dative.

5.5.2 Other Syriac Texts As already discussed in § 5.1.2, instances of t h e coreferential dative have been described in traditional linguistic literature, under the label dativus ethicus, w i t h reference to various Syriac texts. Duval lists a n u m b e r of verbs that can appear with such a construction; Nowicki, semantically analyses t h e p h e n o m e n o n only for t h e writings of E p h r e m . Nevertheless, no specific attention has ever been devoted to the structures in which an et- form is involved.

A Pesitta

Gospels

I will now consider t h e status of t h e phrase et- form + /- + pron. suff. in the Pesittâ version of the Gospels and in Ephrem's Commentaries on Genesis and Exodus.

303

Duval,

304

Nowicki, Beiträge.

Traité.

124

OUTLINE OF MIDDLE VOICE IN SYRIAC

As far as t h e four Gospels are concerned, t h e only two examples of et- form + coreferential dative are: M t . 13:15 and M k . 10:14. 1 8 .

JJO) h a ^ J

OI^A.

et'abi III m. sg. pf. etp. 'harden, thicken'

leh

ger

lebbeh

d-'amma

prep, dat./obj. conj. 'thus' 'heart' + pron. suff. mark. + pron.

rel. + 'people'

Ill m. sg.

hänä dem. pron ./adj. 'this'

suff. Ill m. sg.

" T h e heart of this people has become insensitive." M t . 13:15 is a quotation from Is. 6:10 (also found in Acts 2 8 : 2 7 ) . T h e q u o t a tion is almost literal and, as far as t h e consonantal text is concerned, t h e two passages are identical. Nevertheless, in t h e Gospel there has been an adaptation to the context, so that the imperative et'bay (cons, 't'by) in Isaiah has been substituted by a perfect et'abi (cons, 't'by) in Matthew. T h i s can be easily verified by looking at t h e H e b r e w text of Isaiah:

19. n m Dj?n ±> hasmen

leb

ha-'am

ha-zzeh

II m. sg. ipv. hifil 'harden, cs. state 'heart' art. det. + 'peo - art. det. + dem. thicken'

pie'

'this'

"Make the heart of the people become insensitive. T h i s slight modification has great consequences for the interpretation of t h e rest of t h e passage. In Isaiah t h e dative leh is the direct object of an imperative, t h e subject of which is external to t h e sentence (Isaiah, ordered by God). In Matthew, however, leh is a coreferential dative, referred t o leb 'heart,' t h e subject of t h e perfect et'abi 'it has become harder.' It is also interesting to consider the corresponding Greek text of t h e two passages. B o t h t h e LXX and in t h e quotation in M t . 06 have: eJtaxiWQr] Y&P 11 i c a p S i a ToiS XaoiS TOIJTOU,307 w i t h a III sg. of t h e indicative aorist passive of J t a x w o ^ i a i 'to be made fatter.' T h e distinction between t h e two versions, found in t h e Syriac texts of Isaiah and M a t t h e w respectively (and preserved by all the m o d e r n translations) is witnessed also by t h e Vulgate .

305

Cf. Alfred Rahlfs, Septuaginta id est Vetus Testamentum graece iuxta 70 interpretes (Stuttgart: Württembergische Bibelanstalt, 1971 ). 306 Cf. Nestle et al., Novum Testamentum. 307 heart of this people has become thicker." 308

Is. 6:10 excaeca cor populi huius "blind the heart of this people" vs. incrassatum est enim cor populi huius "the heart of this people has become thicker" B. Fischer, R. Gryson and Robert Weber, Biblia sacra iuxta Vulgatam versionem adiuvantibus B. Fischer ... [et al] recensuit et brevi apparatu critico instruxit Robertus Weber (Stuttgart: Deutsche Bibelgesellschaft, 1994).

125

T E X T U A L ANALYSIS 2 : COREFEENTIAL DATIVE

Instances of coreferential dative are also found in the Gospels in connection with 'basic' stems. Some examples are: Mt. 2: 20, 3:2, 4:12, 4:17; Mk. 4:28; Lk. 8:49, 8:53, 10:34, 10:39, 15, 15. Particularly interesting is Mk. 4:28: 20. ar'a

Lis^o ^ Kji" ger

maytya

Ih

l-pi'ra

'earth'

conj. 'thus'

ptc. f. sg. afel 'come'

prep, dat./obj. mark. + pron. suff. Ill sg.

prep, dat./obj. mark. + 'fruit'

" T h e earth will bring forth fruit." According to the main available editions, such as Pusey 309 the cluster Ih is vocalized as leh, proleptically referred to l-pi'ra. However, the Greek original and a number of translations seem to suggest a different vocalization and interpretation. T h e Greek has in fact: OOITOUXXTTI fj yfl icapjrocpopei, rendered by the Vulgate as ultro enim terra fructificat , which points, for the Syriac text, in the direction of a dative coreferential with 'earth,' rather than with 'fruit,' thus to be vocalized as lah {ar'a is feminine), instead of leh. B Ephrem's Commentaries to Genesis and Exodus T h e other texts that have been checked for the structure et- form + /- + ]won. suff., within this analysis, are Ephrem's Commentaries on Genesis and Exodus. In this work the coreferential dative with et- forms is quite well represented, even if it is not as frequent as in the Demonstrationes: Ephrem's Commentaries contain 33,231 words (according to Syriac spelling) and 36 et- stems followed by coreferential dative. 5 . 6 TYPOLOGICAL COMPARISONS

5.6.1 Coreferential Dative in some Arabic Dialects Lentin 3 1 3 presents a number of particular uses of prepositional datives in some Arabic dialects of the so-called Bilad al-Sam, a dialectal area including Syria, Lebanon, Palestine and Trans-Jordan. Among these uses a form of coreferential dative is also found, involving the preposition 1(a)- 'to' (cognate of the Syriac /-) and sharing several syntactic and semantic features with the Syriac phenomenon discussed

309

Pusey,

310

Fischer et al. Biblia sacra. King James: 'For the earth bringeth forth fruit of herself.'

Tetraeuangelium.

311

T h e text used for the analysis is the edition by Tonneau, Sancti Cf. Appendix 2 for a complete list. Cf. n. 3 for Aphrahat's Lentin, "Datif éthique."

Ephraem.

Demonstrationes.

126

OUTLINE OF MIDDLE VOICE IN SYRIAC

in this chapter. T h e occurrences presented by Lentin come both from modern speech and from Middle-Arabic texts. T h e Arabic structure closely parallels the Syriac one, as is shown in the following example: 21. wa-anam-ll conj. 'and' + I sg. I ipf. 'to sleep' + prep. dat. + pron. suff. I sg.

laylah 'ni ght'

'indakum prep, 'by' + pron. suff. II m. pi. 1

"And I will sleep one night at; your place.

»315

22. ma conj. 'not'

flk prep, 'in' + pron suff. II sg.

t3nt3Z3r-lak II m. sg. VIII ipf. 'to wait' + prep. dat. + pron. suff. II sg.

swayye? adv. 'a bit, a little'

"Can't you wait a little?"316 In 21. the verb nama 'to sieep is conjugated m tne nrst person singular or tne imperfect (or prefix conjugation) and is followed by the dative preposition 1(a)- + a suffix pronoun, agreeing with the verb. In 22. the verbal form tantazar is a stem with -t- infix, from a root meaning 'to wait,' accompanied by 1(a)- + a suffix pronoun, coreferential with the subject of the verb. Although Syriac and the Arabic dialects discussed by Lentin are cognate languages, also belonging to approximately the same geographical area, the correspondences related to the coreferential dative will be examined here from a typological perspective only. However, in principle, a diachronic and family-internal analysis could have also been undertaken and would probably be very interesting and fruitful, especially if extended, for example, to all the Semitic languages of the Levantine area.

314

For Middle Arabic cf. a classical study by Joshua Blau, The Emergence and Linguistic Background of Judaeo-Arabic. A study of the Origins of Middle Arabic (Jerusalem: 1981 [1999]), but also the recent observations by Jonathan Owens, A Linguistic History of Arabic (Oxford: Oxford University Press, 2006). 315

Lentin, "Datif éthique," 112. Lentin, "Datif éthique," 112.

317

Cf., e.g., the remark by Michel Feghali, Syntaxe des parlers arabes actuels du liban (Paris: Imprimerie Nationale - P.Geuthner, 1928), 299-300 about some examples of coreferential dative, quoted by Lentin: "«probablement sous l'influence de la langue syriaque»: qàm «se lever», 'raq «s'enfuir» [...]" Lentin, "Datif éthique," 111.

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

127

Lentin 318 discusses in detail the features and the function of the coreferential dative 319 in the above-mentioned Arabic dialects. Lentin immediately gives a semantic interpretation of the phenomenon he is going to analyze: Par contraste avec la construction simple, sans pronom coréférent, son emploi permet de réitérer, par un moyen syntaxique, la mise en évidence de la relation déjà exprimée par les moyens morphosyntaxiques habituels - du sujet du verbe avec le procès (où l'état); cette relation avec l'action qu'il accomplit (ou avec le procès dont il est le siège) lui étant ainsi en quelque sorte réattribuée, au moyen d'une préposition précisément attributive, le sujet apparaît de ce fait comme lié, associé de façon encore plus étroite à ces procès ou à cet état.

T h u s , according to Lentin, the coreferential dative is a syntactic means of expressing a particular semantic correlation between the verb and its subject, a correlation of the type often underlined for middle voice, an involvement of the subject in the process or state indicated by the verb. After presenting a rich list of examples, organized according to the verbal form with which the coreferential datives occur, Lentin clusters such verbs into 13 semantic classes, which can be described, approximately, as verbs of speech, of translational motion (both transitive and intransitive), of perception, of ingestion, of receiving, of waiting, of resting and sleeping, of using, of cognition and of biological events. It is easy to recognize (even if they are cut somewhat differently) the same classes as those individuated by Kemmer for middle voice (cf. ch. 2 and 4), and also those listed by Nowicki with respect to the Syriac dativus ethicus (cf. § 5.2.1). Indeed, Lentin connects the Arabic data with Kemmer's categories and with what has been observed by Contini on the coreferential dative in PreChristian Aramaic, and Aramaic in general. According to Lentin, the coreferential dative has to be considered as one of the morphological markers that in Arabic can express middle voice, together with vowel patterns and the 10 stem (pattern sta-CCaC).323 These three devices are not synchronic with one another, but rather have been productive at different

318

Lentin Lentin, "Datif éthique," § 5. Coreferential dative is set and defined by Lentin within a system of oppositions, regarding the reference of the pronoun to which the dative is attached: "«le datif éthique», où le pronom personnel n'est pas non plus coréférent au sujet du verbe, mais sans référer pour autant à un objet qu'il affecterait ou impliquerait [...], le datif «réfléchi», «coréférentiel», où au contraire il lui est coréférent" Lentin, "Datif éthique," 101. 320

Lentin, "Datif éthique," 111. Lentin, "Datif éthique," 115-116. 322 For the role of this semantic category with respect to transitivity and middle voice cf. e.g. Mengistu Amberber, Verb classes and transitivity in Amharic (Muenchen: LINCOM EUROPA, 2002) and Jouon, "Sémantique des verbes statifs," 21. 321

323

See also Contini, "Considerazioni sul presunto," discussed infra, § 5.7.2.

128

OUTLINE OF MIDDLE VOICE IN SYRIAC

stages of language development, and also in different registers. T h e coreferential dative appears to be t h e most recent and to belong to a rather popular register. Of particular interest, with respect to t h e Syriac situation, is Lentin's remark about the marginality of forms which accrue more t h a n one middle mark: here t h e coreferential dative co-occurs with a middle stem only w h e n this stem has been lexicalized, and is, therefore, no longer part of a system of morphological oppositions. T h i s seems to partly contrast with t h e situation of Syriac, where we found an abundance of coreferential datives in combination with et- stems, which are by and large still fully functional, as opposed to the non-et- stems and, as we saw above, which tend to rank on the passive border of middle voice.

5.6.2 Latin Pleonastic Reflexives Cennamo 3 2 5 analyses a Latin p h e n o m e n o n displaying important affinities w i t h t h e one so far described for Syriac. It is t h e so-called pleonastic reflexive: namely the use of reflexive p r o n o u n s se/sibi beyond their regular function of direct/indirect object markers, with a certain (variable) degree of grammaticalization (in Cennamo's terms: Non-canonical/Pleonastic se/sibi): 23. a. Quid igitur sibi volt pater?27 b. Animus nescit quid

24. Ille solemniter sibi

vs.

velit,328

ambulabat.329

T h e p h e n o m e n o n , as shown by t h e examples above, is already present in archaic Latin. However it fully develops in t h e later stages of t h e language, w h e n t h e grammaticalization process concludes and t h e p r o n o u n becomes semantically c o m pletely unmotivated. 3 3 0 C e n n a m o demonstrates h o w in late Latin t h e pleonastic

324 Lentin, "Datif éthique," 118. This seems to be the case not only for the morphological categories indicated by the scholar as characteristic of middle voice, but also for other verbal stems built with the affix -t~, at least in some of the instances quoted by Lentin (cf. e.g. the one in 22. above). On the origin and meaning of the lexicalized VIII stems in Arabic, cf. Stempel, Abriss, 134 and also the brief but incisive discussion in Zaborski, "On the interplay," 873. 325

Cennamo, "Late Latin pleonastic." All the examples of this section are taken from Cennamo, "Late Latin pleonastic." 327 "What does the father then want?" Ter. Andr. 375 328 s Q u j Jogs n o t know what it wants" Enn. Trag. 199. 329 w a g w a ] k j n g solemnly" Formul. Tur. 30. 330 Cennamo "Late Latin pleonastic", 124: "Already from its earlier occurrences (dating back to the 3rd century A.D.) [...] and even more so later on, sibi appears also fully pleonastic, that is, it is no longer weakly motivated [...], but occurs in fully grammaticalized

129

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

reflexive intervenes in a reorganization of the verbal system. By means of a thorough analysis of syntactic structures and verbal categories involved in these kinds of expressions, making use of the categories and the conceptual framework elaborated by the so-called Syntactico-Semantic

Approach,

the author shows that in late

Latin, between the 4 and 5 centuries A.D., pleonastic reflexives se/sibi become markers of split intransitivity. T h e structures examined can be classified according to the semantic properties of subject and verb: T w o parameters appear to be crucial in the pattern of active-like syntax we are arguing for: the nature of the surface subject, i.e., its Control (Animacy, Agentivity...) and inherent lexical aspect, in particular the notion of Telicity, although also such features as the Dynamic/Static.

Moreover, these structures are distributed along a gradient, exemplified in Cennamo. High unaccusatives are characterized by the presence of all parameters but Control, as with change of state verbs, having the features telic, dynamic, concrete, Control. Low Unaccusatitvity on the other hand is characterized by the absence of the above-mentioned parameters [...]. High unergatives, on the other hand, denote atelic, dynamic, concrete situation with an Actor subject.

T h e correlation with telicity, as we will see in § 5.7.1, emerges also from Joosten's (1989) observations on the Syriac coreferential dative. 5.6.3 A Possible Origin of the Syriac and the Latin Phenomena? T h e comparison between the Latin phenomenon studied by Cennamo and the use of the coreferential dative in Syriac can be further developed, revealing other analogies. If one observes the examples given for Non-canonical/Pleonastic

se/sibi, one no-

tices that many of these are characterized by a certain degree of emphasis, they are emphatically marked, compared to their equivalents without reflexive pronouns. This is the case, for example, with the suo sibi forms:

forms." T h e scholar gives a number of examples, for one of which Syriac can give a further confirmation. It is 37a. on p. 125, taken from Vulg. Act. 28,16 (Fischer et. al., Biblia sacra): permissum est Paulo manere sibimet cum custodiente se milite. T h e corresponding Greek of sibimet is here Kxx9' e a u t o v "by himself, according to his will", rendered in Syriac just with Bj, W [ayka da-zba] "where he wanted." 331

Cf. Cennamo, "Late Latin pleonastic" for a detailed bibliography, and particularly Robert D. Van Valin, "Semantic Parameters of Split Intransitivity," Language, 66, 221-260. 332 t e r m j n ( Jj c a tes the phenomenon in which intransitives are morphologically or syntactically subdivided into unergatives and unaccusatives. 333

Cennamo, "Late Latin pleonastic," 135.

334

Cennamo, "Late Latin pleonastic," 136, tab. 2.

130

OUTLINE OF MIDDLE VOICE IN SYRIAC

25. ita nunc ignorans suo sibi servit patri.

335

26. suo sibi suco vivunt (sc. cochleae). 3 3 6 27. suo sibi gladio hunc

iugulo.

337

In all these passages the pleonastic reflexive is used to underline the exceptional character of the event described. T h e same can be said for the cluster sibi + adj.: 338

28. sed solus sibi immotus Atreus constat atque ultro deos terret

minantes.

And also for the equivalents of per se: 29. nam scorpio sibi ipse pulcherrimum

medicamentum

est.m

Cennamo makes no particular reference to phenomena of emphasis that are not relevant for such analysis. Nevertheless, if compared with what is found in Aphrahat, the Latin examples can contribute to building a hypothesis for the possible origins of coreferential dative in Syriac. As mentioned before, some examples of the coreferential dative in Aphrahat's Demonstrationes, are found in^cjuotations of biblical passages, modified by the author according to the context . Such passages can be used for comparison, in order to individuate another function of the coreferential dative. Dem. II 52, 10-11 is based on Gal. 3:19, but the biblical passage is rephrased with the addition of the coreferential dative. T h e same is true for Dem V 20 L, 1-2, based on Lam 5:16. This time both the rephrasing and the literal quotation from the Bible are found in Aphrahat's text, in close proximity.

335

"So now, without knowing it, he is serving his own father" Plaut. Capt., Prl., 50.

336

"[Snails] live out of their own juice" Plaut. Capt. 1,1,13.

337

"I slaughter him with his own sword" T e r . Ad., 957. Cf. Cennamo, "Late Latin pleonastic," 118-119. 338 "But Atreus, firm, on his own, stands and appals the threatening gods" Sen. Thy., 703-4. Cennamo, "Late Latin pleonastic," 120. 339 "In fact, the scorpion is itself an excellent medicine against itself' Cels. 5, 27, 5. Cennamo, "Late Latin pleonastic," 120.

T h e problem of biblical quotations in the early Syriac texts is quite complex and still a matter of debate, both as regards translation techniques and the kind of text that the authors had at their disposal. For Aphrahat, cf. Baarda, The Gospel. 341

T h e Pesitta text of Lam. 5:16 is actually simply

UAo

[npal klild d-risan].

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

131

It is evident, from the examples given, that the quotations by Aphrahat differ from the original biblical text because of the presence of the coreferential dative . Let us now consider the context of such quotations. Aphrahat's Demonstrationes are a doctrinal text. T h e y are set in the framework of a long epistle, as a response to an epistola interrogator is. T h e author, addressing his interlocutor with the second person singular, explains some essential points of Christian faith and provides a clue for the understanding and interpretation of the Scriptures. Be the epistola interrogatoris a literary fiction or not, the strong rhetorical character and the persuasive purpose of the text are evident. T h e Demonstrationes are also referred to as Homiliae (cf., e.g., the title of the edition by Wright 1869). Dem. II 52,11 is part of a section in which Aphrahat discusses the statement Iustis non est lex posita sed impiis (I T i m . 1:9). T h e law, therefore, has not always existed, but was introduced by God because of the transgression of mankind. T h r o u g h a series of biblical examples, the author aims at proving that si iustitia apud homines permansisset, opus lege non fuisset. Therefore, the sentence in II 52,11 is not just the description of an obvious fact, but the emphatic stating of an interpretation that Aphrahat wants the reader to share. T h u s the text goes on: Legem enim additionem esse [Scriptura] demonstravit. T h e Syriac text of Gal. 3:19 is a faithful reproduction of the Greek original. In Dem. V 201 the mechanism of the quotations is more complicated, but can be neglected for the purpose of this analysis. At lines 1-2, the cluster npal leh is used to introduce a sudden event, interrupting a static situation previously described: the king of Cherubs rises over his people, sitting at his feet, in his shade. T h e function of the coreferential dative is here to emphasize an unexpected event. In the light of what has been exposed so far, it is possible to hypothesize, both for the Syriac and Latin coreferential dative, an original emphatic value, that has been lost in time, through the increasing grammaticalization of the pronoun. A variable degree of emphasis and markedness is still present in such structures, that often keep a rhetorical function. Nevertheless they progressively reach the status of standard construction, at least when combined with some verbal categories.

342 js case _ The coreferential dative is in some instances already a j wa y g present in the biblical source, be it in the Old or the New Testament. In Dem. XXI 957, 19,23 the coreferential dative [mitw Ihon] 'they died' is already in the original text of the Pesitta of Ex. 4:19. 343 For the various names given to Aphrahat's work see Baarda, The Gospel, 9-10, Jacobus Forget, De Vita et Scriptis Aphraatis, Sapientis Persae (Louvain: 1882), 131. 344 Dem. II, 47. 345 Dem. II 52, 12. 346 For the emphatic origin of the markers of coreferentiality, cf. also what Kemmer, The Middle Voice, 47 says about the origin of reflexive markers. The author also admits the possibility of intermediate stages in the grammaticalization of the markers, in which the emphatic function is preserved to a certain extent.

132

OUTLINE OF MIDDLE VOICE IN SYRIAC

Another observation can be made, regarding the origin of the Syriac coreferential dative. According to Kemmer (1993), the origin of middle markers is often to be found in the extension of reflexive markers. From the expression of coreferentiality stricto sensu, between Agent and Patient of a transitive action, the marker is progressively extended to other domains that relate to middle voice, rather than to reflexivity. As was shown in chapters 1 and 3, Syriac has a system of reflexive markers based on the use of pronominal suffixes, connected with prepositions determined by the governing verb. T h e preposition /- is widely used in Syriac, as a marker of direct and indirect objects and as such is one of the most frequent markers of reflexivity. On the basis of this observation, it is thus possible to see the coreferential dative also as an original reflexive marker that has progressively extended its scope towards middle voice and has been later grammaticalized in a new function of middle voice marker. T h e group l-+pron. suff., per se, is still a marker of reflexivity, but, in the specific syntactico-semantic context discussed in this section, it has developed into a middle voice marker. 5 . 7 CONCLUDING REMARKS

5.7.1 Coreferential Dative and Aspect In § 5.3.2 reference was made to the study on dativus ethicus by Joosten. 347 T h e author notices that the structure is mainly associated with verbs indicating a state (or expressing an action eventually resulting in a state) and believes that the function of the DE is precisely that of signalling 'entering into a state' (p. 482). He also gives a list of all the verbs occurring in this kind of structure, organized in categories (pp. 476-478). All the et- forms are listed together, under the label of 'passives,' but without any attention to their quantity and specific relevance. Moreover, the author gives important data about the distribution of the coreferential dative with respect to the various verbal tenses. It appears that there is an incompatibility of the structure with compound tenses constructed with the participle, whereas it is diffused with all the other narrative and perfective tenses. Joosten concludes: "This indicates that the use of the D E is incompatible with the notion of durativity." Here a brief terminological digression is required. In his article Joosten speaks of Aktionsart and Phasenaktionsart, concepts that, as was shown in chapter 2, belong to the inherent semantics of the various verbal roots. However, the value that Joosten attributes to the coreferential dative rather belongs to the aspectual domain. T h e verbs followed by a coreferential dative seem to be mainly aoristic. T h i s distinction is important, because it also allows us to individuate two different levels at

347 348

Joosten, "The Function." Joosten, "The Function," 89.

TEXTUAL ANALYSIS 2: COREFEENTIAL DATIVE

133

which coreferential dative can be examined: a) the semantics of the verbs with which it occurs; b) the properties of the sentences in which it can be found. T h e first level relates to the semantic categories by which coreferential dative can be found, according to Joosten: statives, intransitives indicating change of state, motion verbs, and et- forms. As was shown earlier, all these categories can also be defined as unaccusative. T o the second level belongs instead the remark that the coreferential dative is not found with participial compounds indicating an ongoing process, but rather with 'narrative' and 'perfective' verbal tenses. Is it possible to specify the connection between the two levels? According to Joosten the underlying property binding all examples of coreferential dative is that they express "entering into a state." In terms of Actionality, one can say that the verbs followed by a coreferential dative are telic,349 It was also shown in the previous chapters and here in § 5.6.1 that telicity is also one of the features of unaccusatives. As we saw in chapter 2, telicity has tendentially a greater compatibility with the perfective Aspect and perfective tenses, although it can in principle be combined also with the imperfective ones. 350 This explains why, as we saw above in § 5.3.2, in the Acts of Thomas, the majority of coreferential datives occurring with non-et- stems turn out to have aoristic value. And this also explains Joosten's findings about the distribution of the phenomenon with respect to the verbal tenses. 5.7.2 Coreferential Dative and the Aramaic Background of the New Testament Black, 351 in a study on the Aramaic background of the New Testament devotes a paragraph to the possible traces of the dativus ethicus in the N T Greek. Here the following remark is found: T h e place of reflexives in the Semitic languages is largely taken by the dativus ethicus, a very common construction in Aramaic. It is not unknown in Greek, where it is usually classed as an extension of the dativus commodi or incommodi, but it is rare as compared with Aramaic usage, where it is a means of stressing the action of the subject and corresponds more to the Greek Middle than to the dativus commodi?52

Black appears fully conscious of the connection between the dativus ethicus and the values traditionally associated with middle voice. He does not develop his ob349

Cf. in this respect, Joosten's observation that "With regard to the term Phasenaktionsart, it is necessary to specify one more point. T h e phase of action on which the DE focuses attention is never viewed as a process, but as a limit, a terminus" Qoosten, "The Function," 487). 350

In this respect, Joosten's statement that "the use of the DE is incompatible with the notion of durativity" is rather drastic (cf. Joosten, "The Function," 489). Matthew Black, An Aramaic approach to the Gospel and Acts (Oxford: Clarendon Press, 1967). 352

Black, An Aramaic, 102.

134

OUTLINE OF MIDDLE VOICE IN SYRIAC

servation any further (it being beyond his scope), but he gives a list of interesting examples of datives in the Greek text that are better explained as a literal rendering of Aramaic dativi ethici, thus proving he is fully aware of the precise usages and function of such structures. 5.7.3 Contini's Analysis of Coreferential Dative as a Middle Voice Marker T h e functions of the coreferential dative have been described and analyzed already by Contini (1998). In the first part of the study, the scholar gives a broad overview of the phenomenon in various Semitic (and also non-Semitic) languages, including references to the major studies on the matter. T h e scholar clearly shows the difference existing between the proper dativus ethicus and the coreferential dative. Contini also underlines the fact that all the various values and usages of such a dative are better understood if connected to the category of middle voice: nonostante la coincidenza formale di questi pronomi con i corrispondenti riflessivi, proprio nella combinazione con verbi intransitivi il vero carattere di questi sintagmi si rivela consistere nell'espressione di una più intensa partecipazione del soggetto all'azione, ovvero della sfera semantica della diatesi media. 353

Contini's idea is that the coreferential dative serves as a middle voice marker "in the expression of a more intense participation of the subject in the action", in the absence of a precise morphological category in the verbal system. As we saw before, this idea is shared also by Lentin, with respect to the coreferential dative in Levantine Arabic. T h e main object of Contini's study is Pre-Christian Aramaic, therefore the Syriac phenomenon is only briefly examined in the preliminary overview. Nevertheless, the perspective that such a study offers on the different Aramaic dialects (and also on parallel phenomena in Biblical Hebrew) gives a very important insight into the underlying semantic processes that surface to different extents in the various languages.

353

"In spite of the formal overlap between these two pronouns and the corresponding reflexives, just as in the combination with intransitive verbs, the real character of these syntagms proves to consist in the expression of a more intense participation of the subject in the action, namely of the semantic area of middle diathesis." Contini, "Considerazioni sul presunto," 86. 354

Contini, "Considerazioni sul presunto," 87.

6

CONCLUSIONS

In this book a number of phenomena of the Syriac language system have been analyzed, which are correlated to the semantic categories of middle voice and unaccusativity. T h e aim was to describe the net of semantic, syntactic and morphological phenomena that, operating in different domains of the Syriac linguistic system, turn out to play a part in the expression of middle semantics. Because the topic had not been studied per se, so far, my work was to a great extent merely a survey, and would require deeper and broader investigation. T h e term outline, chosen for the title, expresses first of all such a limitation. However, this word also says something about the method that I used in my research, and about the way the results are organized and described. A number of phenomena were examined, whose linguistic reason and interconnection can be viewed as evidence of the productivity of a certain semantic category within the Syriac verbal system. I tried to let middle voice emerge in its extension and in its internal articulation, through its manifestations in different domains of the syntax and of the semantics. In chapter 1 heterogeneous morphological and syntactic elements were presented, which, according to traditional Syriac grammar, may all be used to express meanings such as reflexive, passive and reciprocal. Such categories are: the verbal et- prefix, inner passives, and pronominal reflexive structures. It was also noticed how the majority of examples, that grammars present for the reflexive meaning of verbal et- forms, do not fit into a rigorous definition of reflexivity, but rather relate to the expression of the affectedness of the subject and to what, in other languages, is usually called middle voice. In chapter 2 the term and the semantic category of middle voice have been introduced and analyzed. Particular attention has been devoted to Kemmer's (1993) study of the internal semantic articulation of the categories and of the parameters, on the basis of which reflexivity and middle voice needed to be distinguished. Reflexivity was thus defined as a two-participant event, in which the agent and the patient happen to be coreferential, whereas middle voice is represented prototypically by a one-participant event in which the grammatical subject is affected by the process described. In this chapter unaccusativity was also introduced, a semantic feature of a class of verbs, determined basically by the following set of parameters: affectedness of the subject, low control on the process, low agentivity, low volitionality, and telic Aktionsart. Unaccusativity has been individuated, by a number of scholars, as the main semantic feature of Indo-European middle voice. In the light of all the preceding observations, the following plan was made for the study of Syriac middle voice:

135

136

OUTLINE OF MIDDLE VOICE IN SYRIAC

1 distinguishing between proper reflexivity and reflexive structures of middle voice; 2 checking for the semantic categories that Kemmer has individuated for middle voice, within some specific Syriac verbal classes; 3 verifying whether unaccusativity could play a role in the distribution of forms and functions of the morphological elements and syntactic structures presented in chapter 1. More specifically, the following distribution of forms and functions was hypothesized: reflexive (and reciprocal) (prep.) + pron. suff. (co-referentiality of agent and noun 'body' / 'soul' + pron. suff. (depending on the degree of emphasis required patient) by the context) inner passive middle - stative [- process] middle - unaccusative external affixation: et- stems [- control] [- volition] [- agentivity] [+ affectedness] In chapter 3 a brief overview was given of the Semitic parallels of the Syriac verbal morphemes involved in this study. T h e result was that also in the other languages taken into account, especially as far as the -t- verbal affix is concerned, there was a certain degree of vagueness in the correlation of forms and functions. However, reference is constantly made to the domains of reflexivity and passivity. Moreover, it emerged that, with respect to both the i-affix and the inner passive, a number of aspectual and actional values were pointed out by several scholars, as concurring with the expression of voice. Chapters 4 and 5 present a textual analysis of a sample of early Syriac texts, dating from between the 3r and the 6 century A.D. In chapter 4 the results are presented of a classification of et- stems and inner passives, found in the sample texts, according to Kemmer's semantic categories, aspect and Aktionsart. All Kemmer's classes appeared to be represented among Syriac verbs, although to a variable extent. Moreover, there seems to be a preference for Terminative aspect and for telic Aktionsart, as far as et- forms are concerned, while inner passives turned out to be mainly resultative perfects and states. As regards Aktionsart, it is important to recall that each one of the four main categories that have been used in this study (achievement, accomplishment, activity, state) is determined on the basis of three parameters [itelic], [idurative], [istate]. Therefore, when we say that inner passives are in almost all occurrences classified as states, we are saying something more than what is already common knowledge about the stative value of the qtil pattern. With respect to the hypothesis formulated in chapter 2, the following points were confirmed:

137

CONCLUSIONS

• • •

the connection of passive internal vowel patterns with stativity; the connection of the et- prefix with telicity; the connection of both morphological categories with the semantic area of middle, as defined by Kemmer.

In chapter 5 a particular phenomenon of Syriac was examined, which is strictly connected with middle voice: coreferential dative. With this term a syntactic compound is indicated, in which a verbal form is immediately followed by the preposition /- (expressing dative) and a pronominal suffix coreferential with the subject of the verb. This phenomenon had already been studied by Joosten for Syriac and by Contini for Aramaic in general, and for Pre-Christian Aramaic in more detail. T h e two scholars have underlined the following features of the phenomenon, respectively: • •

the preference for the expression of aoristic incohative aspect; the correlation with the expression of semantic affectedness of the subject.

In this study attention was devoted, on the one hand, to a more specific connection between et- forms and coreferential dative (especially observed in Aphrahat's Demonstrationes and in Ephrem's Commentaries) and, on the other hand, to a strong connection between the semantic classes of verbs that can be followed by a coreferential dative also in their basic stems and unaccusatives. Finally, a typological comparison was established between the Syriac phenomenon and the analogous coreferential dative in Levantine Arabic dialects, described byJ Lentin and the 357

Latin pleonastic dative, that was studied by Cennamo. T o sum up, with respect to the hypothesis of chapter 2, in this chapter a correlation is established between verbal et- forms and a syntactical epiphenomenon of unaccusativity. Appendix 1 presents an overview of some documents of medieval grammatical tradition, with respect to diathesis. T h e material there presented constitutes a brief resumé of a previous study. 358 Diathesis, translated in Syriac with the terms qyàmè 'positions,' or aynàywàtà 'qualities' is described from different perspectives in Syriac grammars, according to whether they are of Hellenizing or Arabicizing inspiration. Only in the oldest grammatical document at our disposal, Huzaya's translation of the Téchne Grammatiké, is found reference of the existence of a middle diathesis. As already noticed by Contini, Huzaya gives a purely semantic definition of such a diathesis, without indicating any morphological element as being a characteristic 355

Kemmer, Hoe Middle Voice.

356

Lentin, "Datif éthique."

357

Cennamo, "Late Latin pleonastic,"

358

Farina, "Diathesis and Middle."

359

Contini, "Considerazioni interlinguistiche."

138

OUTLINE OF MIDDLE VOICE IN SYRIAC

of middle. All the later Syriac grammars only refer to two diatheses, an active and a passive one. Moreover, the concept itself of diathesis seems to oscillate between a semantic and a morphological one. Appendix 2 presents a list of all the occurrences of coreferential dative in combination with et- stems (and, for some texts, also with basic stems) that are found in the corpus I have examined. In this book a number of phenomena and problems have been mentioned that would deserve further investigation, in a broader sample of Syriac texts. Among the desiderata, I would indicate, at least: • • • • •

an analysis of the aspectual and actional behavior of the opposite pole of the verbal system: intensives and causatives; an examination of the differences in meaning between the various etstems of a single verbal root; a more selective analysis of the differences between the various verbal 'tenses'; the extension of the corpus of Syriac texts to translations from Hebrew and Greek and to later texts; a comparison, as far as Aphrahat's Demonstrationes are concerned, with the extant Middle Arabic translation.

360 On which Butrus Yousif and S. Khalil Samir, "La version arabe de la troisième démonstration d'Aphrahat (Sur le jeûne) - Texte établi par Kh. Samir. Etude et traduction de P. Yousif," in Actes du deuxième congrès internationsl d'études arabes chrétiennes (ed. S. Khalil Samir; Rome: Pontificium Institutum Studiorum Orientalium, 1986), 31-66.

APPENDIX 1

DIATHESIS IN MEDIEVAL SYRIAC GRAMMARS

This Appendix provides a brief account 361 of the way diathesis was treated by medieval Syriac grammarians. T h e interest of this matter is, of course, historical, but also paradigmatic in nature. On the one hand, in fact, following the development of concepts and terminology that are still in use in linguistic studies can provide a better understanding of their internal articulation. On the other hand, investigating the perception that native speakers and ancient writers had of a given language can offer a precious insight into a number of phenomena that have hitherto eluded complete understanding. Finally, it is important to bear in mind that our knowledge of the Syriac language (as well as of most classical languages) is the endpoint of a chain of interdependent accounts. A number of ideas we have about the structure of a given language often ultimately depend on perspectives on the language, that originated in a given historical period, under specific cultural influences of which we have to be aware. T h e earliest phases of Syriac linguistic tradition are related, as is often the case, to orthoepy and to a masoretic annotation of the Scriptures. A record of the correct pronunciation and vocalization of the words depends, of course, on the knowledge of the grammar and morpho-syntax of the language, apart from the mere tradition. This phase mirrors, of course, the Jewish masoretic tradition. However, from the 5 -6 century A.D., the study of Greek theology and philosophy was already triggering the first translations of Greek texts into Syriac 361

I have discussed this topic more extensively in Farina, "Diathesis and Middle

Voice." For a history of the Syriac grammatical tradition, cf. Merx, "Historia artis grammaticae," Contini, "Considerazioni interlinguistiche;" " T h e role of linguistics in Syrian society," in History of the Language Sciences (ed. S. Auroux, E.F.K. Koerner, H.-J. Niederehe and C.H.M. Versteegh; Berlin - New York: Mouton de Gruyter, 2000), 341-344; "Storia della scienza siriaca. III. Le scienze del linguaggio," in Storia della scienza: vol. IV: Medioevo — Rinascimento (ed. Roma: Istituto dell'Enciclopedia Italiana, 2001), 26-36, 68-69; Rafael Talmon, "Foreign influence in the Syriac grammatical tradition," in History of the Language Sciences (ed. S. Auroux, E.F.K. Koerner, H.-J. Niederehe and C.H.M. Versteegh; Berlin New York: Mouton de Gruyter, 2000), 337-341; John Watt, "Grammar, Rhetoric and Enkyklios Paideia in Syriac," ZDMG, 143, 45-71. 363 Cf. Brock, "Greek into Syriac;" Watt, "Grammar, Rhetoric;" Henri HugonnardRoche, La logique d'Aristote du grec au syriaque: études sur la trasmission des textes de

139

140

OUTLINE OF MIDDLE VOICE IN SYRIAC

and, therefore, the necessity of also creating a code in order to translate grammatical and linguistic categories. In this context there emerges a first stream of Syriac grammatical tradition that is largely based on a Greek model. Subsequently, after the development of Arabic linguistic thinking and the great systematizations by authors such as Sibawayhi and Zamahsari, also Syriac grammar was progressively influenced. Elias of Tirhan's work, which is explicitly and fully based on the Arabic model, dates back to the 11 century. T h e Ktaba dsemhe (Book of Explanations), a grammar written by Barhebraeus in the 13 century, is considered to be the final synthesis of the Greek and Arabic models applied to the Syriac language. T h e first complete description of Syriac that we have is a translation of the Greek Techne Grammatike, made by Joseph Huzaya in the 6 century A.D. T h i s text is particularly interesting, from the perspective of the present study, because, to the best of our knowledge, it is the only text throughout the entire Syriac tradition that makes reference to a middle verbal conjugation. Following the Techne, in fact, Huzaya mentions and describes three diatheses (qyame, lit. "positions") of the verb: the active (ma'bdanuta lit. "action, fact"), the passive (hasa lit. "passion") and the middle one (mes'ayuta lit. "midst, midde"). T h e middle diathesis is defined semantically (as it is the case for its Greek counterpart), as "that which is expressed sometimes by the active and sometimes by the passive." Of particular interest are the examples that Huzaya gives for the mes'ayuta: "I travelled (rbet), I went (helket), I sat (yetbet), I ran (rehtet), I passed {'ebret)." These verbs, unlike what happens in the examples of the active and passive categories, do not correspond at all, in meaning, to those of the Greek model (jteJtr|YCX SiecpQopa, ejroiria^iriv, eYpa^a^iriv). It may well be a coincidence, but four of these five examples are motion verbs, and all of them belong, from the semantic point of view, to the class of unaccusative verbs. Of course, I do not suggest that Huzaya was by any means aware of such a semantic feature, which is an arbitrary cognitive category of our time. However, the examples that he chose for the mes'ayuta must have appeared to him as bound, in their meaning, by a common underlying property, relevant from the point of view of diathesis. Huzaya's possible intuition remains, nevertheless, isolated within Syriac grammatical tradition, and all the other authors, whose work has reached us, only recognize an active and a passive diathesis in the Syriac verbal system. Alternatively, one can also reverse the perspective, and investigate how the etstems and the inner passives are described by ancient Syriac grammarians and how these two structures are fitted into the verbal system. l'Organon et leur interprétation philosophique (Paris: Vrin, 2004). 364 For the corresponding Greek passage see pp. 17-18 of this book. Cf. above § 2.1, Lallot, La grammaire, 166-167; Contini, "Considerazioni interlinguistiche;" Farina, "Diathesis and Middle Voice," 178-180. Merx, "Historia artis grammaticae," 60*.

APPENDIX 1

141

First of all, as noticed in § 1.3, in Syriac grammar there is not a specific term, or class, that encompasses all the non-active derived stems, such as the Western 'reflexive' or the Arabic mutawaa. T h e r e are two different ways in which the etstems and the inner passives are addressed: a semantic and a morphological one. Leaving aside the middle, which, as we saw, only includes active stems, in Huzaya's account the passive voice is exemplified by a series of et- stems: "metmhe na (I am hit), met'bed '_ná (I am done), metkteb na (I am written)." No further explanation is given with respect to the meaning of the forms. A clear semantic interpretation of the function of the et- prefix is given by Elias of Tirhan, whose work, as we saw, is the first Syriac grammar constucted according to the Arabic model: When we mention something effected by means of a verb meaning an action, we distinguish what is effected and that is enough, we do not need a different vowel of the noun, as is the use in Arabic. So for example we say that etmahhi nuh ("Noah was beaten"): a transformation of the verb is enough for us to indicate that Noah [is the object] of the wounds. This [happens] also when we speak about actions such as etkteb ktaba ("the book/script was written") and esta'rat 369 sa'ürüta ("the action was done") and etyahbat zküta ("the victory was given").

Bar Z u b i ( 1 2 - 1 3 1 cent), considers, on the contrary, the et- stems as the passive (hasüsütá) counterpart of the active diathesis imabdánütá). T w o lists of verbal forms are juxtaposed. T h e first one consists of basic and non-et- derived stems, the second one of the corresponding et- stems, from the same roots quoted in the first list (plus a few new forms): "the passive is like metbzez (spoiled), mettdin (judged), metbne (built) ..." Although no specific distinction is made, between the semantic and the morphological level, one gets the impression that Bar Zu bi is here mainly concerned with the morphological correlation between non-et- and cisterns. In Barhebraeus' Ktaba d-semhe (13 cent.) the distinction between active and passive diathesis is described by means of different vowel patterns. T h e inner pas-

367 f o r m s a r e e v e n m o r e interesting because, they have all been inserted in participial constructions, but each could have been replaced by an inner passive participle. Grammatîqî arabî'âyt in Elias' words. For the text of Elias' grammar cf. Friedrich Bäthgen, Syrische Grammatik des Mar Elias von Tirhan (Leipzig: 1880); Theodor Nöldeke, "review of F. Baethgen, Syrische Grammatik des Mar Elias von Tirhan," Göttingische gelehrte Anzeigen, 1, 721-734. 369 Bäthgen, Syrische Grammatik, 8-9*. 370 Bar Zu bî, like most grammarians after Huzaya, uses for diathesis the term aynâyûtâ 'quality.' 371 Georges Bohas, "Les accidents du verbe dans la grammaire de Bar Zu'bî ou: une adaptation de la Technê," Langues et littératures du monde arabe, 4, 54-86.

142

OUTLINE OF MIDDLE VOICE IN SYRIAC

sive is thus considered as a prototype for t h e semantic role of patient. In t h e section of the grammar devoted to t h e affixes (nqipwata) of t h e verb we read: T h e fifth is quality (aynayuta), active and passive, such as mamlek ('being king') with rbasa lamad, on the one hand, and mamlak ('being made king') with ptaha lamad on the other hand. 373

T h e et- stems are also treated in the list of t h e nqipwata, but in this case a m o n g t h e schemes (eskim), namely the possible variations of the verbal structure, patterns, we would say: T h e sixth is scheme, simple, composed, "overcomposed", such as hza ('he saw') ethzi ('he appeared') ethzawzi ('he distinguished himself). 3 7 4

Here again a morphological interpretation of t h e et- stems emerges, this time even clearer than in Bar Zu'bi. T h e et- stems are no longer associated with the expression of diathesis, which is, at least in this passage, left to t h e inner passive participles. A final interesting aspect of t h e treatment of diathesis in t h e medieval Syriac grammars examined here regards t h e concept of passivity that they reveal. In this book we often referred to t h e fact that Semitic languages categorize and express passivity in a different way from Western languages. T h e y tend not to express t h e agent overtly and they do not always operate a clear-cut distinction between a real passive and a tollerativum or a spontaneous event experienced by t h e subject. In this respect, a few remarks found in some of the texts examined here give an outline of two different perspectives. T h e first one is witnessed in t h e above-mentioned grammar by Elias of T i r h a n . T h e grammar has the structure of a series of questions and answers subdivided into chapters. As mentioned earlier, this grammar is based on t h e Arabic model and follows t h e Arabic categorization of language. Elias does not address t h e problem of diathesis directly, but is rather concerned with t h e distinction between agent and patient in a transitive sentence. 375 T h e question related to the use of et-

372

On the meaning of this term in Syriac grammar cf. Axel Moberg, Buch der Strahlen: die grössere Grammatik des Barhehräus / Ubersetz. nach einem kritisch berichtigten Texte mit textkrit. Apparat und einem Anhang: Zur Terminologie von Axel Moberg (Leipzig: Harassowitz, 1907): *67-*68. 373

Axel Moberg, Le livre des splendeurs. La grande grammaire de Grégoire Barhebraeus. Texte syriaqe édité d'après les manuscrits avec une introduction et des notes par Axel Moberg (London - Oxford - Paris - Leipzig: 1922), 90. Moberg, Le livre des splendeurs, 90. T h u s one of his questions is "by means of which signs do we indicate the agent ('bûdâ) that is the performer {saura) in the Syriac language? T h e Arabic language has in fact in grammatical terms signs for the one who effects and the one who undergoes, namely fail and mafül/' to which he answers: "First is the essence (yâtâ) i.e. the natural possibility (mkânûtâ) of the agent and of the patient (met'abdânâ). Second by means of letters that are called article (masryâtâ / sâryâtâ), like b-, d-, /-. Although properly we indicate the Patient

APPENDIX 1

143

stems is phrased as follows: "From which signs and marks is recognized a name that is patient whose agent was not metioned together with it?" The answer, already quoted above, is: when we mention something effected by means of a verb meaning an action, we distinguish what is effected and that is enough, we do not need a different vowel of the noun, as is the use in Arabic. So for example we say that etmahhi nub "Noah was beaten", a transformation of the verb is enough for us to indicate that Noah [is the object] of the wounds 3 7 7

It is clear, from this example, that the expression of an agent (whatever its syntactical status may be) is considered as superfluous in a passive construction. A remark by Barhebraeus in the Ktaba d-semhe (in the above mentioned section on the affixes of the verb) goes in a rather different direction. In order to clarify what a passive diathesis is, Barhebraeus quotes the following example from the Scriptures: Gospel: 378 "and when he heard that Archelaus was made king ( m a m l a k ) of Judaea", namely by Cesar ( m e n qesar).

In this case, the author considers it necessary to add an overt mention of the agent men qesar to the original quote from the Gospel, in order to explain the passive value of the afel participle mamlak.

with /-. T h i r d from the order and the precedence of the nouns." (Bäthgen, Syrische Grammatik: 4*). T h e same idea lies behind the definitions of agent and patient given by Barhebraeus in the Ktäbä d-semhe: "And the agent {'abüdä) is the noun in front of which is the verb, and what is like it, such as in 'Paul preaches' (akrez pawlüs) and 'Paul, his disciple is an announcer' ( p a w l ü s msahrän talmideh). [...] And the patient ( h a s ü s ä ) is the one on which the verbal action falls ( m e l a t saürüta), such as in 'our Lord struck the Egyptians' (mha märyä l-mezräye) and 'he went to the land of the Gadarenes' (w-eta l-aträ d-gädräye)," Moberg, Le livre des splendeurs, 36). 376 377

Bäthgen, Syrische

Grammatik,

8-9*.

Bäthgen, Syrische

Grammatik,

8-9*.

M t 2, 22. 379

Moberg, Le livre des splendeurs,

90.

APPENDIX 2

A . LIST OF OCCURRENCES OF THE STRUCTURE T-FORM + L + COREFERENTIAL DATIVE IN APHRAHAT'S DEMONSTRATIONS 3 8 0 Dem. I 5, 6-7 0*1>0

v> Jio,

OO) OO)

hü haw meddem d-kalè metnseb leh menneh et qui habet, si quecumque rogantem se volit prohibere, auferetur ab ipso quidquid ipse dare recusat 8, 1 0 - 1 1

Islo

0)\

lloiYI.Q|\ ^iol/ v*^ JLli", bo

ma d-'nas gèr etqarrab 1-haymänütä ettsìm leh 'al ki'pä quando enim quis accedit ad fìdem, constituitur super petram Dem. II 52, 10-11 .°imoll(' Lcdclìoj, ot±Z2)

l^o

kad m t i gèr zabneh d-namòsa ettawsap leh... quando autem legis advenit tempus, addita est... 53, 17-19 )ooip/, Q)~|\~>

i,jLLl [ . . . ] ]jO) JJSYÜD )ooi^>JJ O)\ ItWli", o o o i ^ >

oi^>

beh gèr b-haw zabnä d-et'amrat leh 1-abraham melti hadè [...] bhaymänütä ettsìr leh b-lebbeh d-abraham eo enim ipso tempore quo Abrahae factus est sermo iste: proles erit tibi, per fìdem formatus est [sermo] 381 in corde Abrahae

380 381

The Latin translation is the one found in Parisot, "Aphrahatis Demonstrationes." Also Parisot interprets leh as refering to melta.

145

146

OUTLINE OF MIDDLE VOICE IN SYRIAC

53, 23-25 looip/, Q)~l\~i

[...]

.s/o

w-äp melti [...] etbatnat läh b-lebbeh d-abraham sermo [...] intimo Abrahae corde conceptus est 56, 2-3 )ooip/, ou^j

Jloj^ii.

t*a\x>L/o

w-etmalkat läh 'abdütä 'al zar'eh d-abrähäm promissa fuit servitus Abrahae posteris 56, 21-22 ^JJJOQO,

l

l.^x-. IcOOYlì

estayyak leh namòsa wa-nbìyè 'al halèn trèn pùqdanìn lex autem et prophetae in duobus illis mandatis comprehensi sunt 57, 13-14 ^i...

o)K.lbaz> vpo)\

Lcoovi^ ^ ^

husahawhy gèr d-namòsa etbattalw 1-hòn b-me'tyeh d-mahyanan finis enim legis evacuatus est per adventum vivificatoris nostri 57, 25 - 60, 1 lt.\» «

If^r^ \Ooò>.

u0)0;ÌD/, Jjji. ^oo

w-men edam d-asrawhy et'esarw 1-hòn 'ad'ìdè b-sislätä ex quo eum vinxerunt, festivitates eorum vinculis ligatae sunt 60, 10-11

Jis^ojjo Lois,,

iiAsl/o

w-etmalyat läh diyatìqì qadmäytä ba-'hräytä textamentum prius novissimo est impletum 77, 9-10 Jo)\JJ lvi\\

)Q*TO ^^OQj

d-badmut sta yawmìn sìm leh 'alma 1-alähä mundum, sicut intra sex dies a Deo constitutus fuit 80, 7-8 OUOQS ^O nettsìr leh men pùmeh proprio irretiatur ab ore

APPENDIX 2

80, 26 - 81, 1-2 1«..,\ OO) )c\h**./

•^

^^-^jO OJKHA.

JJj ^^JOO

w - m e t t u l d - l a sbaq la-knateh qalíl d - l e h saggí estabeq leh estlem h ü 1-dahseh parum conservo non dimisisset is cui m u l t u m condonatum fuerat, satellitibus ipse traditus est 88, 16-18

...» Oltoo I...» "O, boj'o) ^O

Ji£D, ^o

man d-sané gér had men hadamé da-msíha metpsah leh men kolleh pagra qui autem u n u m de membris Christi odit, a toto corpore divelletur 88, 18-19 loi^j

^o

-O)Q«JJ Jato, ^200

w - m a n d-sané l - a h u h y metpres leh men bnaya d-alaha et qui fratrem odit, a filiis Dei separabitur D e m III 100, 10-12 ba.j

-¡joIí" lis« \i\, ^00

j ^o )Oj, boo

w - m a d - s a m men da-l-mea'kal w men da-l-mesta etqri leh sayama quamdiu cibo potuque abstinet, ieiunare dicitur 116, 25

w-ba-kyala da-b'a la-mkalü ettkil leh mensura qua metiri intendebat demensum est ei D e m IV 141, 5 - 7 o>\ ; W o>\

Jt-Ar> v>

Oj\ IM./ jO )oo)p/ 3 / 0

w-ap abraham kad asar leh alaha mülkana dabra metiled leh e m i r leh... A b r a h a m etiam, quando firmavit ei Deus de filio ipsi bascituro promissionem, dixit ei... 156, 4 - 6 JJoo/ 1>oqo,

¿dKco/o : l;nro

o^oo2> o j i a d o ^^Ljj o o l —^jO

w-sali tub daní'él wa-slóteh puma daryawata sakrat. w-estakar leh puma aküla Daniel etiam oravit, eiusque oratio leonum ora praeclusit; obturatum est os voracissimum

148

OUTLINE OF MIDDLE VOICE IN SYRIAC

168, 11-12 l..~i,V> )0,j0

La^>iSJL!0

JiOO)

hàkanà ethassab b-tar'ìtak da-slòtàk mestabqa lah qdam madbha T i b i autem animo fìnge, orationem tuam ante altare relictam esse 172, 10-12 on^Sj ^o

.a^Kcolo

l l o ^ j ^ÌÌO ¿Ò>.

*jLaS.j o vt\ » vt\ ¡^V

w-'ad b a e ' ant la-msalamu slòtàk tetpallat lah mennak sbuta da-nyaha w testaggap lak men sebyàneh wanyàhe d-allaha Nam priusquam orationem compleveris, res ilia quae Deum alleviarci tibi excidet, ispoque opere quod Deo placet defraudaberis 173, 23

tetbattal lak men siòta (noli) orationem omittere (lit. 'cease from prayer') 176, 3 ^vjOll JJo w - l a tetqatta' lak neque obtorpescas

Dem V 196, 24-25

estqel leh kolleh ablatae sunt omnes [scil. 'divitiae' = ' u t a r ] 196, 26 - 197, 1 I1Q~- 1 ^ou^o

^ nt i o

w-testqel làh menhòn malkutà regnum auferretur ab eis 209, 24 JJjiS, llo-i\v>;

^A-li"

ethlet leh b-'eqara d - m a l k u t à d-parzla radici regni ferrei sese commiscuisse 216, 20-21 JJjiS,

^=.11/

ettabar leh nìra d-parzla i u g u m ferreum [...] confractum est 217, 1

ettrìm leh sibi exaltavit 217, 2 .Si" vOOl^ Q^yoli' etmretw lhòn ap gepawhy avulsae sunt etiam eius alae 232, 8 - 1 0 loìi^j PIVI. fl\ .^¡JOÌÌOOJ

l°H..\

apen gèr 1-hanpè plah barnasa. barsa'teh d-metqreb la-qyameh d-alaha etharrar leh quamvis enim servierit quis gentilibus, ubi primum foederi Dei appropinquat, extemplo fit liber

150

OUTLINE OF MIDDLE VOICE IN SYRIAC

Dem VI 248, 9-10 Ijl^JO, JisL.^0 ^o

l l j o ^ l o L;qjo

^O

man d-rahem qurye w-te'gurta metglez leh men mdinta d-qaddise qui rura et mercatum amat prohibebitur a civitate Sanctorum 252, 20 JJ, d-la nethayyab leh ne forte superetur 252, 22 0)t!0 Ol^i bsj JJ

I^oOiOO;

man d-metqat'a leh la zabe beh mareh qui taedio indulget, non sibi complacet in eo Dominus eius 253, 10-11 *j00JSJL)0

^JQJLJ ^O

man d-saqel nira d-qaddise neteb w-nestuq leh qui iugum portat sanctorum, sedeat et taceat 256, 26 o)lo;~i«~>

^jiSJLi' (Oj/o

w-adam estdel leh b-sabruteh Adam imperitia sua deceptus est 261, 1

w-nethseb leh gayara et adulter existimetur 280, 22

metb'e leh mennan nobis desumptum est 293, 12

mettamra lah ruha napsanayta spiritus animalis absconditur

APPENDIX 2

293, 13-14 OIÜO

JIOÎSjlûd IJSJL^JO

w-regstâ mestaqlâ läh menneh quod sensu destituitur (lit. 'and feeling is taken away from it') 293, 23

w-rûhâ napsânâytâ mettamrâ lâh b-kyânâh animalis autem spiritus in natura sua sepelitur 296, 26 - 2 9 7 , 1 Ik.l.YI»

L.Oj

w-metbal'â lâh rûhâ napsânâyta b-rûhâ smayânâytâ Spiritus animalis a Spiritu caelesti absorbebitur 304, 5-6 fjli^O j-liSJ JJj

•ÇfOI

haydên dhel ganâbâ w-'âreq d-lâ netthed wnettsîd leh pertimescit latro et aufugit, ne arripiatur et comprehendatur 308, 11-12

mettul d-metbal'â lâh napsânâytâ b-rûhânâytâ quod enim animale est absorberetur a spiritalis 309, 8-9 11. VI« P,/ OO), lljOj

~l\ vQJO) yOOli

w-hâylên aylên d-methalpîn lhôn henôn lâbsîn zûrtâ d-haw âdâm smayânâ qui immutabuntur, formam induent Adae caelestis

Dem VII 316, 4-5 OtTûjOQX mettbar leh 'ûqsâh tune stimuls ille confringetur 320, 3

âp zakkâyê nethasbôn lhôn hayyâbê innocentes in culpa esse iudicentur

151

152

OUTLINE OF MIDDLE VOICE IN SYRIAC

321, 5 - 6 Li'^OO

Jlnro vOO»\

-.K v>r>

w-methasbìn lhòn saklè w-bìsè w-maràhè amantesque et improbi et imprudentes reputabuntur 329, 4 J,0)

^olìoo

w-nettkel leh 'al hadè et ille in hoc confisus fuerit 329, 16

w-nettkel leh'al hadè et super hoc confisa 332, 3-4 JJiS-CDi" uJOQLO Jjl IjtOi" Jj'i-X, Jjl b A . » d - d a l m a ena d-la-hrànè akrzet ena qnòmy estlè lì ne, cum ceteris praedicaverim, ipse reprobus efficiar 340, 12 Ji°lJL )V!.\ . ; «°ll ^ U o Ma^oì

K.Ì ]o), ^

" W l JJ

la tetkel lak d - h a ìt asyuta w-tetel napsak la-sma sapla noli ergo prae nimia fiducia medicinae attendere, ut tibi ipsi infamia afferras 352, 3

tetkel lak tibi ipsi n i m i u m confidere 352, 5 OO)

l2>iÈO

netrapè leh haw d-bala' ille qui percussus fuerit animo frangatur 352, 7 )CL.;lìo JJ,

d - l a nettrìm w-nethayyab leh non autem superbire, et reum inde fieri

APPENDIX 2

Dem VIII 364, 24-25

w-'al d-methabbal leh pagra de corporis autem corruptione (lit. 'of the body corrupting') 3 6 8 , 14

U ^ J lj'oi£D ^

OÌSJLi" OO) JiOO) ^O

w-en hakana hw estkhan lan sàhdè daggalè si autem ita res se habuerit, invenimur falsi testes 368, 24

Jl.VI«

-i-Dli

etqri leh smayana caeleste dicetur 368, 25

lii.;/ o)i.-i~>

-ioli

etqri leh bakyaneh ar'ana secundim naturam suam vocabitur terrestre 369, 14-15

liì^liV llfOj metbala lah zurta ar'anayta ba-smayanta forma terrestri a caelesti absorpta 369, 15-16

Jl.VI«

luoìoo

metqrè leh pagra smayana corpus caeleste praedicabitur 3 6 9 , 17

La^V

luobo

metqrè leh ar'anaya terrestre nuncupabitur 400, 4

IIqjI^)

LO^JOO, IYI\\ Jjoi^i

b-hana 'alma d-metalqa lah taybuta b-ki'nuta in m u n d o isto in quo iustitia gratiam absumit

154

OUTLINE OF MIDDLE VOICE IN SYRIAC

400, 10 o)l 11/

ettaqnat lah antta the woman was formed

383

etpallagw

lhon atrawtä

"per loca sua dispersi sunt" T o n n e a u , Sancti

Ephraem.

APPENDIX 2

46, 2 VOOLAI.

I^LI"

etgzar leh 'layhon it was decreed against them 3 8 4 48, 20 jlITSII' etebe's leh he got angry 48, 21 (idem 2x) 49, 5

netqassas leh he will be reckoned 51, 25

estkahw lhon they are found 52, 27 Uli tetqare läh she will be called 55, 2 vOO|i ^¡JOli" etqariw lhon they are called 55, 8

nett'iq leh he will be grieved

384

Impersonal with leh functioning as an empty subject.

159

160

OUTLINE OF MIDDLE VOICE IN SYRIAC

56, 13 ¿1^ ¡SJL^iSJLi' estagsat lah she is troubled 58, 23 ^OO)^ ^..^K rev. mestarhin lhon they were destroyed 63, 22

mestakhan lhen they are found 68, 22

esta'al leh he abstained 74, 29

ettrim leh menneh alaha God is taken away from him 76, 16

tet'iq lah she will be grieved 76, 23 h CI*°>L~»K V» metkasse leh he concealed himself 103, 8

estkah leh he was found 103, 18

et'si leh he was forced 110, 22 « « ol/ etqassas leh he is declared older

162

OUTLINE OF MIDDLE VOICE IN SYRIAC

1 1 2 , 20

estawtap leh lewi l-sem'on Levi joined Simeon 112, 28

etpset leh he spread out 116, 25

etbaddarw lhön they are dissolved Exodus 122, 13

ethattan leh he connected himself 146, 2 vooii a » ; ! / etr'emw lhön they got angry C . ( P R E F E R E N T I A L D A T I V E IN T H E F I R S T SEVEN A C T S OF J U D A S T H O M A S W i t h et- s t e m s I, 177, 17 -°>\ ..K . ! estahlap leh it had changed/was transformed IV, 210, 2 . .1 Vt

JXojbv • V

mestqlä läh menní will be taken away from me

APPENDIX

VI, 220,

11-12

L^-cotooi" ^o

. mrml/

etkses leh men ewkaristiya he was rebuked by the Eucharist VII, 244, 2

U ; ^

JJ

la mestle leh he has not been forced VII, 248, 5

mestren lhen are loosened

With other verbal stems I, 1 7 3 , 16

steq leh he was silent I, 1 7 4 , 11

Loot I o q j o

L,o;

qamat hwat lah ruha the wind had risen/was steady I, 1 7 8 , 2

itbat lah she sat I, 1 8 0 , 9

lis—j nehtat lak you descended I, 180, 21

nepqet hwayt lak you had gone out

2

164

OUTLINE OF MIDDLE VOICE IN SYRIAC

I, 181, 3 V^S)

Joo)

iteb hwa leh maran and our Lord sat down I, 183, 1-2 • O.'.ji/ . .ivt menni etrahhaq leh from me it moved away I, 184, 12 loo) hzaq hwa leh he had gone, departed II, 188, 7-8 nt sqil leh he takes for himself II, 189, 7-8 JV l l i ma et 'na ly I am dying II, 190, 14 Joo) ;ol twar hwa leh he was astonished

II, 191, 7

tezban lak you will buy yourself

385

Wright, Apocryphal Acts, II: "it was moved far from me" (impersonal).

APPENDIX 2

165

II, 194, 14 and II 195, 8 are borderline between coreferential dative and real

dativus

II, 193, 8 IfOO^.

)QjO

qam leh ihuda and Judas stood up II, 193, 19 vpojjaj

loo) otn2>

pkah hwa leh nurhon and became pale their light II, 193, 20 vOOti^O

/

etkassi leh menhon he concealed himself from t h e m

II, 194, 14-15

nehsar lah she will want (for herself) II, 195, 8

sbu Ikon receive (for yourselves) 3 8 6 III, 203, 10

nebtel leh it comes to an end III, 204, 7 ou^o

hood

tahme lak menneh you will turn away from h i m

386

ethicus.

166

O U T L I N E OF M I D D L E V O I C E IN S Y R I A C

III, 2 0 4 , 7 - 8

^

m

te'te lak you will go VII, 2 3 4 , 10,

vOO)\ OjOJX 'raqw lhon they fled VII, 2 3 4 , 12

^OlAi yOOli olI etw lhon 'alayhen they arrived/ they came upon them VII, 2 3 6 , 1 2 - 1 3

vQolo^ ^O

Jj^ji"

ezal 'na li men lewatkon I will go away from you VII, 2 3 6 , 2 1

sbeq leh it goes away VII, 2 3 8 , 1 5 - 1 6

-OO) -JJ l'i hway lhen they got tired VII, 2 3 8 , 16

uooi ..vin qamy hway lhen they stood still VII, 2 3 9 , 1

Joo) -to kri hwa leh was becoming short

APPENDIX 2

167

VII, 2 4 5 , 1

mity lhen they died VII, 245, 4-5

etayt lak (3x) you have come VII, 249, 18

steq leh he went silent D . ( P R E F E R E N T I A L D A T I V E W I T H ET- S T E M S IN T H E P E S I T T A G O S P E L S Mt 13, 15 Jjoi bai., OI^A.

.

I1

et'bi leh ger lebeh d-'ama hana the heart of this people has become insensitive M k 10, 14

eteb'es leh he got angry

BIBLIOGRAPHY

Amberber, Mengistu. "Quirky Alternations of Transitivity: T h e Case of Ingestive Predicates." Pages 1-19 in Language universals and variation Edited by Mengistu Amberber and Peter Collins. Westport, C T : Praeger, 2002. Arce-Arenals, Manuel, Melissa Axelrod and Barbara A. Fox. "Active Voice and Middle Diathesis. A Cross-Linguistic Perspective." Pages 1-21 in Voice. Form and Function. Edited by Barbara A. Fox and Paul J. Hopper. Amsterdam Philadelphia: Benjamins, 1994. Baarda, Tjitze. The Gospel quotations of Aphrahat, the Persian sage. Amsterdam: 1975. Bakker, Dirk. "Review of Jerome Lund, 'The Book of the Laws of the Countries. A Dialogue on Free Will Versus Fate. A Key-Word-in-Context Concordance'." Aramaic Studies 5 (2007): 273-286. Bakker, Egbert J. "Voice, Aspect and Aktionsart." Pages 23-47 in Voice. Form and Function. Edited by Barbara A. Fox and Paul J. Hopper. Amsterdam Philadelphia: Benjamins, 1994. Bäthgen, Friedrich. Syrische Grammatik des Mar Elias von Tirhan. Leipzig: 1880. Benveniste, Emile. "Actif et moyen dans le verbe." Pages 68-175 in Problèmes de linguistique générale. Edited by Emile Benveniste. Paris: Gallimard, 1966. Berrettoni, Pierangiolo. La logica del genere. Pisa: Edizioni PLUS, 2002. Bertinetto, Pier Marco. Tempo, aspetto e azione nel verbo italiano. Il sistema dell'indicativo. Firenze: Accademia della Crusca, 1986. - Il dominio tempo-aspettuale. Demarcazioni, intersezioni, contrasti. Torino: Rosenberg & Sellier, 1997. Bertinetto, Pier Marco and Mario Squartini. "An attempt at defining the class of 'Gradual Completion Verbs'." Pages 11-26 in Temporal Reference, Aspect and Actionality I: Semantic and Syntactic Perspectives. Edited by Pier Marco Bertinetto, Valentina Bianchi, James Higginbotham and Mario Squartini. Torino: Rosenberg & Sellier, 1995. Beyer, Klaus. The Aramaic Language. Göttingen: Vandenhoek & Ruprecht, 1986. BHS Biblia Hebraica Stuttgartensia. Edited by Karl Elliger and Wilhelm Rudolph. Deutsche Bibelgesellschaft, 1997 (Fünfte, verbesserte Auflage). Black, Matthew. An Aramaic approach to the Gospel and Acts. Oxford: Clarendon Press, 1967. Blau, Joshua. The Emergence and Linguistic Background ofJudaeo-Arabic. A study of the Origins of Middle Arabic. Jerusalem: Ben Zvi Institure, 1999. Bloomfield, Leonard. Language. Chicago: Chicago University Press, 1933. Bohas, Georges. "Les accidents du verbe dans la grammaire de Bar Zu'bì ou: une adaptation de la Technè." Langues et littératures du monde arabe 4 (1998): 54-86.

169

170

OUTLINE OF MIDDLE VOICE IN SYRIAC

Bohas, Georges, Jean-Patrick Guillaume and Djamel Eddin Kouloughli. The Arabic linguistic tradition. London - New York: Routledge, 1990. Bombeck, Stefan. "Das syrische Verbalsystem: hypothetische Sätze und hwä vor Adjectiv oder Partizip ." Pages 23-36 in Tempus und Aspekt in den semitischen Sprachen. Edited by Robert Nebes. Wiesbaden: Harrassowitz, 1999. Bossong, Georg. "Le marquage différentiel de l'objet dans les langues d'Europe." Pages 193-258 in Actance et Valence dans les Langues de l'Europe. Edited by Jack Feuillet. Berlin - New York: Mouton de Gruyter, 1998. Bravmann, Meir Max. Studies in Semitic Philology. Leiden: Brill, 1977. Broch, Jens Peter (ed.). Al-Mufassal, opus de re grammatica arabicum, auctore Abu'lKâsim Mahmûd bin 'Omar Zamahsario, ad fidem codicum manu scriptorum edidit - . Ed. altera, denuo recognita, adnotationibus criticis aliisque aucta. Christianiae: Libraria T . P. Mallingii 1879. Brock, Sebastian. "Greek into Syriac and Syriac into Greek." Journal of the Syriac Academy 3 (1977): 406-422. - "Aspects of translation technique in antiquity." Greek Roman and Byzantine Studies 20 (1979): 69-87. - Syriac studies: a classified bibliography (1960-1990). Jounieh: Université SaintEsprit de Kaslik, 1996. Brockelmann, Carl. Syrische Grammatik. Leipzig: Veb Verlag Enzyklopädie, 1965. Brown, Francis, S.R. Driver and Briggs Charles A. A Hebrew and English lexicon of the Old Testament: with an app. containing the biblical Aramaic: based on the lexicon of William Gesenius as transi, by Edward Robinson Oxford: Clarendon, 1952. Bussmann, Hadumod. Routledge dictionary of language and linguistics. London: Routledge, 1996. Cennamo, Michela. "Late Latin pleonastic reflexives and the unaccusative hypothesis." Transactions of the Philological Society 97 (1998): 103-150. Cohen, David. L'aspect verbal. Paris: Presses Universitaires de France, 1989. Comrie, Bernard. Aspect. Cambridge: Cambridge University Press, 1976. Contini, Riccardo. "Considerazioni sul presunto 'dativo etico' in Aramaico precristiano." Pages 83-94 in Études sémitiques et samaritaines offertes à Jean Margain. Edited by Christian-Bernard Amphoux, Albert Frey and Ursula Schattner-Rieser. Lausanne: Editions du Zèbre, 1998. - "Considerazioni interlinguistiche sull'adattamento siriaco della Techné Grammatiké di Dionisio Trace." Pages 95-111 in La diffusione dell'eredità classica nell'età tardoantica e medievale - Il Romanzo di Alessandro e altri scritti. Edited by Rosa B. Finazzi and Alfredo Valvo. Alessandria: Edizioni dell'Orso, 1998. - " T h e role of linguistics in Syrian society." Pages 341-344 in History of the Language Sciences. Edited by Sylvain Auroux, E.F.K. Koerner, Hans-Josef Niederehe and Cornelis H.M. Versteegh. Berlin - New York: Mouton de Gruyter, 2000. - "Storia della scienza siriaca. III. Le scienze del linguaggio." Pages 26-36, 68-69 in Storia della scienza: vol. TV: Medioevo - Rinascimento. Roma: Istituto dell'Enciclopedia Italiana, 2001.

BIBLIOGRAPHY

171

Correli, C. "Ein neuer Anlauf zur Erklärung der Herkunft der 'Notae Accustativi'." Pages 22-43 in Festschrift Ewald Wagner zum 65. Geburtstag. 1. Edited by Wolfhart Heinrichs, Gregor Schoeler and Ewald Wagner. Beirut, Stuttgart: Orient-Institut der Deutschen Morgenländischen Gesellschaft. In Kommission bei Franz Steiner Verlag, 1994. Costaz, Louis. Dictionnaire Syriaque-Français / Syriac-English dictionary / Qämüs Siryänl-Arabl. Beyrouth: Imprimerie Catholique, 1963. Cowley, Arthur E. Aramaic Papyri oft the Fifth Century B.C. Oxford: 1923. Curtius, Georg. Griechische Schulgrammatik Prag: Tempsky, 18759. De Medicina - A. Cornelii Celsi, De Medicina, Libri Octo. Edited by Charles Daremberg. Leipzig: Teubner, 1891. Delbrück, Berthold. Vergleichende Syntax der indogermanischen Sprachen. Strassburg: Karl Trübner, 1897. Derenbourg, Hartwig (ed.^. Le Livre de Sîbawayhi, traité de grammaire arabe par Sîbouya dit Sîbawayhi, I—II. Paris: Imprimerie Nationale, 1881. Di Benedetto, Vincenzo. "Dionisio Trace e la Techne a lui attribuita." Annali della Scuola Normale Superiore di Pisa. Lettere, Storia e Filosofia. Serie 2 21 (1958): 169-210. - "Dionisio Trace e la Techne a lui attribuita." Annali della Scuola Normale Superiore di Pisa. Lettere, Storia e Filosofia. Serie 2 28 (1959): 87-118. - "La Techne spuria." Annali della Scuola Normale Superiore di Pisa. Classe di Lettere e Filosofia. Serie 3 3 (1973): 797-814. Diem, Werner. "Die Nominalform fu al." ZDMG 120 (1970): 43-68. - "Die Entwicklung des Derivationsmorphems der t - Stämme in Semitischen." ZDMG 132 (1982): 29-84. Dixon, R.M.W, and Alexandra Aikhenvald. "A Typology of ArgumentDetermined Constructions." Pages 71-113 in Essays on language function and language type : dedicated to T. Givón. Edited by Joan Bybee, John Haiman and Sandra A. Thompson. Amsterdam: Benjamins, 1997. Dombrowski, B.W.W. "Some remarks on the Hebrew hithpa'el and inversative - t in the Semitic languages." JNES21 (1962): 220-223. Donner, Herbert and Wolfgang Röllig. Kanaanäische und aramäische Inschriften. Wiesbaden: Harrassowitz, 1966-69. Drijvers, Hendrik Jan Willem. The Book oft the Laws of the Countries. Dialogue on Fate ofBardaisan ofEdessa. Assen: Van Gorcum & Comp. N.V., 1965. Duval, Rubens. Traité de grammaire syriaque. Paris: Vieweg, 1881. - La littérature Syriaque. Paris: Lecoffre, 1907. Enn. Trag. - The Tragedies of Ennius. Edited by H.D. Jocelyn. Cambridge: Cambridge Univeristy Press, 1967. Epistulae ad Familiares - M. Tulli Ciceronis Epistulae ad familiares. Edited by David Roy Shackleton Bailey. Stuttgart: Teubner, 1988. Epstein, Morris. Tales oft Sendebar: an edition and translation of the Hebrew version of the Seven sages: based on unpublished manuscripts. Philadelphia: T h e Jewish Publication Society of America, 1967. Faltz, Leonard. Reflexivization: A study in universal syntax. New York: Garland Publishing, 1985.

172

OUTLINE OF MIDDLE VOICE IN SYRIAC

Farina, Margherita, "ns e n s in ebraico biblico: nota accusativi e preposizione." Egitto e Vicino Oriente 18 (2005): 245-253. - "La marca differenziale dell'oggetto e la particella 'et in ebraico biblico, prime prospezioni." Quaderni del Laboratorio di Linguistica della Scuola Normale Superiore di Pisa 5 (2005). - "A survey on Semitic verbal t-forms: status quaestionis." Quaderni del Laboratorio di Linguistica della Scuola Normale Superiore di Pisa 6 (2006). - "Diathesis and Middle Voice in the Syriac Ancient Grammatical Tradition: T h e Translations and Adaptations of the Téchne Grammatiké and the Arabic M o d e l . " Aramaic Studies 6 (2008): 175-193. - "Review of Li, Tarsee (2008). Non-Active Participles in the Aramaic of Daniel. Aramaic Studies, 6(1), 111-136." Quaderni del Laboratorio di Linguistica della Scuola Normale Superiore di Pisa 7 (2008). Feghali, Michel. Syntaxe des parlers arabes actuels du Liban. Paris: Imprimerie Nationale - P. Geuthner, 1928. Fici Giusti, Francesca. "Diathèse et voix marquée dans les langues d'Europe." Pages 347-389 in Actance et Valence dans les Langues de l'Europe. Edited by Jack Feuillet. Berlin - New York: Mouton de Gruyter, 1998. Fischer, Bonifatius and Robertus Weber. Biblia sacra iuxta Vulgatam versionem adiuvantibus B. Fischer ... [et al] recensuit et brevi apparatu critico instruxit Robertus Weber. Stuttgart: Deutsche Bibelgesellschaft, 1994. Fitzmyer, Joseph A. A Wandering Aramean. Collected Aramaic Essays. Missoula: Scholars Press, 1979. Fleisch, Henri. Traité de philologie arabe. I. Préliminaires, phonétique, morphologie nominale. II. Pronoms, morphologie verbale, particules. Beyrouth: Dar el-Machreq, 1979. Foley, William. "A typology of information packaging in the clause." Pages 362446 in Language Typology and Syntactic Description. Edited by T i m o t h y Shopen. Cambridge: Cambridge University Press, 2007. Folmer, Margaretha. "Metathesis in Jewish Aramaic: a so-called "Pan-Semitic Feature" reconsidered." Pages 223-243 in Hamlet on a Hill. Edited by Martin Baasten, Takamitsu Muraoka and W. T h . van Peursen. Leuven: Peeters, 2003. Formul. Tur. - Pages 33-40 in Merowingische und Karolingische Formulare. Edited by J. Pirson Heidelberg: Winters, 1913. Gai, Amikam. " T h e Signifié of the Non-Active Participle in Semitic Languages." ZDMG 155 (2005): 9-23. Garr, William Randall. Dialect Geography of Syria - Palestine, 1000-586 B.C.E. Winona Lake, Indiana: Eisenbrauns, 1984. - "Affectedness, Aspect and Biblical et." ZAH4/2 (1991): 119-134. Gass, Erasmus. Die Moabiter - Geschichte und Kultur eines ostjordanischen Volkes im 1. Jahrtausend v. Chr. Wiesbaden: Harrassowitz, 2009. Givón, Talmy. Syntax: A functional-typological introduction. Amsterdam Philadelphia: Benjamins, 1984. - " T h e pragmatics of de-transitive voice: Functional and typological aspects of inversion." Pages 3-46 in Voice and Inversion. Edited by Talmy Givón. Amsterdam - Philadelphia: Benjamins, 1994.

BIBLIOGRAPHY

173

Goldenberg, Gideon. "Aramaic Perfects." Israel Oriental Studies 12 (1992): 113-137. - "Actants and Diathesis, Directions of Transitivity & c.: Some Satzgestaltungen and Their Background in Semitic and Elsewhere." Pages 283-296 in From Beyond the Mediterranean. Akten des 7. internationalen Semitohamitistenskongresses. Edited by Rainer Voigt. Aachen: Shaker, 2007. Gottlieb, Hans and Erling Hammershaimb The Old Testament in Syriac according to the Peshitta version. The Book of Kings. Leiden: Brill, 1976. Greenberg, Joseph H. " T h e Patterning of Root Morphemes in Semitic." Pages 365-387 in On Language. Selected Writings of Joseph H. Greenberg. Edited by Keith Denning and Suzanne Kemmer. Stanford: Stanford University Press, 1950. Gzella, Holger. Tempus, Aspekt und Modalität im Reichsaramäischen. Wiesbaden: Harrassowitz, 2004. Haig, Geoffrey. Alignment Change in Iranian Languages. Berlin - New York: Walter de Gruyter, 2008. Haspelmath, Martin. Transitivity Alternations of the Anticausative Type. Köln: Institut fur Sprachwissenschaft, Universität zu Köln, 1987. - "More on the Inchoative-Causative Verb Alternation." Pages 87-120 in Causatives and Transitivity. Edited by Bernard Comrie and Maria Polinsky. Amsterdam Philadelphia: Benjamins, 1993. - "Passive Participles across Languages." Pages 151-177 in Voice. Form and Function. Edited by Barbara A. Fox and Paul J. Hopper. Amsterdam Philadelphia: Benjamins, 1994. Healey, John F. "Variety in Early Syriac: T h e Context in Contemporary Aramaic." Pages 221-229 in Aramaic in its Historical and Linguistic Setting. Edited by Holger Gzella and Margaretha L. Folmer. Wiesbaden: Harrassowitz, 2008. Hilgard, Alfred. "Scholia in Dionysii Thracis Artem Grammaticam." in Grammatici Graeci I, 3. Leipzig: Teubner, 1901. Hoftijzer, Jacob. "Remarks regarding the Use of the Particle 'et in Classical Hebrew." Oudtestamentische Studien 14 (1965): 1-99. - "Interpretation and Grammar." Pages 173-302 in Aramaic texts from Deir Alia. Edited by Jacob Hoftijzer and G. van der Kooij. Leiden: Brill, 1976. Hopper, Paul J. and Sandra A. Thompson. "Transitivity in grammar and discourse." Language 56 (1980): 251-299. Huehnergard, John. A Grammar of Akkadian. Winona Lake, Indiana: Eisenbrauns, 2005. Hug, Volker. Altaramaische Grammatik der Texte des 7. und 6. Jh. s. v. Chr. Heidelberg: Heidelberger Orientverlag, 1993. Hugonnard-Roche, Henri. La logique dAristote du grec au syriaque : études sur la trasmission des textes de l'Organon et leur interprétation philosophique. Paris: Vrin, 2004. Humphreys, R. Lee. "Valency changing alternations." Pages 391-400 in Concise Encyclopedia of Grammatical Categories. Edited by Keith Brown and Jim Miller. Oxford: Elsevier, 1999. Ildefonse, Frédérique. La naissance de la grammaire dans l'antiquité grecque. Paris: Vrin, 1997.

174

OUTLINE OF MIDDLE VOICE IN SYRIAC

Jansma, Taeke and Marinus Koster Peshitta. The Old Testament in Syriac. Part I Fase. 1. Preface. - Genesis; Exodus. Leiden: Brill, 1977. Jastrow, Otto. " T h e Neo-Aramaic Languages." Pages 334-377 in The Semitic Languages. Edited by Robert Hetzron. New York - London: Routledge, 1997. Jenni, Ernst. Die hebräischen Präpositionen. Bandi: Die Präposotion Beth. Stuttgart Berlin - Köln: Verlag W. Kohlhammer, 1992. Joosten, Jan. " T h e Function of the so-called Dativus Ethicus in Classical Syriac." Orientalia 58 (1989): 473-492. Joüon, Paul. "Sémantique des verbes statifs de la forme qatila (qatel) en arabe, hébreu, arameéen." Mélanges de l'Université Saint-Joseph 15 (1930): 3-32. - Grammaire de l'hébreu biblique. Roma (Deuxième édition): Pontificio Istituto Biblico, 1965. Joüon, Paul and Takamitsu Muraoka. A Grammar of Biblical Hebrew (Revised edition). Roma: Pontificio Istituto Biblico, 2006. Kemmer, Suzanne. The Middle Voice. Amsterdam - Philadelphia: Benjamins, 1993. Kienast, Burkhart. Historische Semitische Sprachwissenshaft. Wiesbaden: Harrassowitz, 2001. Kiraz, George. Comparative Edition of the Syriac Gospels. Leiden: Brill, 1996. Kirtchuk, Pablo, "/'et/ ou ne pas /'et/, l'actant Y en hébreu et au-delà." Actances 7 (1993): 91-133. Klijn, Albertus F. J. The Acts of Thomas. Leiden: Brill, 2003. König, Ekkehard and Volker Gast. "Focused assertion of identity: A typology of intensifiers." Linguistic Typology 10 (2006): 223-276. König, Ekkehard and Peter Siemund. "Intensifiers and reflexives - a typological perspective." Pages 41-74 in Reflexives: Forms and Functions. Edited by Zygmunt Frajzyngier and Tracy Curl. Amsterdam - Philadelphia: Benjamins, 2000. Kulikov, Leonid. " T h e reflexive pronouns in Vedic: A diachronic and typological perspective." Lingua 117 (2007): 1412-1433. Kurylowicz, Jerzy. Studies in Semitic Grammar and Metrics. Wroclaw: 1972. Kutscher, Eduard Y. "Aramaic." Pages 346-412 in Linguistics in South West Asia and North Africa. Edited by T h o m a s A. Sebeok. T h e Hague - Paris: Mouton, 1970. Lallot, Jean. La grammaire de Denys le Thrace. Paris: CNRS, 1989. Larcher, Pierre. Le système verbal de l'arabe classique. Aix-en-Provence: Publications de l'Université de Provence, 2003. Law, Vivien and Ineke Sluiter, Dionysius Thrax and the Technë grammatikë, Münster: Nodus Publikationen, 1995. Lazard, Gilbert. L'actance. Paris: Presses Universitaires de France, 1994. Lazzeroni, Romano. "La diatesi come categoria linguistica: studio sul medio indoeuropeo." Studi e saggi linguistici 30 (1990): 1-22. - "Inaccusatività indoeuropea e alternanza causativa vedica." Archivio Glottologico Italiano 89 (2004): 139-164. Lentin, Jérôme. "Datif éthique, datif coréférentiel et voix moyenne dans les dialectes arabes du Bilâd al-Sâm et quelques problèmes connexes." Cahiers de Linguistique de l'INALCO (2003-2005) (2008): 99-130. Li, Tarsee. "Non-Active Participles in the Aramaic of Daniel." Aramaic Studies 6 (2008): 111-136.

BIBLIOGRAPHY

175

Liddell, Henry G. and Robert Scott. A Greek-English lexicon. Oxford: Oxford University Press, 1978. Liebermann, Stephen J. "Review o f R e t s ö 1989: Diathesis in Semitic Languages: A Comparative Morphological Study.".//IOS 111 (1991): 650-653. Lipinski, Edward. Semitic Languages: Outline of a Comparative Grammar. Louvain: Peeters, 1997. Lund, Jerome. The Book of the Laws of the Countries: a Dialogue on Free Will Versus Fate: a Key-Word-in-Context Concordance. Piscataway, NJ: Gorgias Press, 2007. Lyons, John. Introduction to theoretical linguistics. Cambridge: Cambridge University Press, 1969. Maalej, Zouhair. "Middle Verbs." Pages 224-227 in Encyclopedia of the Arabic Language and Linguistics. Edited by Cornells H. M. Versteegh. Leiden: Brill, 2008. Macdonell, Arthur Anthony. A Vedic grammar for students. London: Oxford University Press, 1916. Malessa, Michael. "Différentielle Objektmarkierung im klassischen Hebräisch." KUSATU 1 (2000): 133-156. Mengozzi, Alessandro. "Middle Markers: Neo-Aramaic and Italian verbal forms with a 3rd singular feminine pronominal object." Pages 105-116 mAtti del XII Incontro italiano di linguistica afroasiatica (Ragusa 6-9 giugno 2005). Edited by Mirella Cassarino and Marco Moriggi. Soveria Mannelli: Rubbettino, 2007. Merx, Adalbertus. "Historia artis grammaticae apud Syros." Abhandlungen fur die Kunde des Morgenlandes 9 (1889). Moberg, Axel. Buch der Strahlen : die grössere Grammatik des Barhebräus / Ubersetz. nach einem kritisch berichtigten Texte mit textkrit. Apparat und einem Anhang: Zur Terminologie von Axel Moberg. Leipzig: Harassowitz, 1907. - Le livre des splendeurs. La grande grammaire de Grégoire Barhebraeus. Texte syriaque édité d'après les manuscrits avec une introduction et des notes par Axel Moberg. London, Oxford, Paris, Leipzig: 1922. Moravcsik, Edith A. "Some crosslinguistic generalizations about the intensifier constructions." Pages 271-277 in Papers from the eighth regional meeting of the Chicago Linguistic Society. Edited by Paul Peranteau et al. Chicago: Chicago Linguistic Society, 1972. Morrison, Craig. " T h e Function of qtal hwa in Classical Syriac Narrative." Pages 103-131 in Biblical and Oriental Essays in Mamory of William L. Moran. Edited by Augustinus Gianto. Roma: Pontificio Istituto Biblico, 2005. - " T h e Function of qtal hwä in the Acts of Judas Thomas." Pages 257-285 in Aramaic in its Historical, and Linguistic Setting. Edited by Holger Gzella and Margaretha L. Folmer. Wiesbaden: Harrassowitz, 2008. Muraoka, Takamitsu. "On the so-called Dativus Ethicus in Hebrew." The Journal of Theological Studies, New Series 29 (1978): 495-498. - Classical Syriac. A Grammar with a Chrestomathy. Second, Revised Edition. Wiesbaden: Harrassowitz, 2005. Muraoka, Takamitsu and Bezalel Porten. A Grammar of Egyptian Aramaic. Leiden: Brill, 1998.

176

OUTLINE OF MIDDLE VOICE IN SYRIAC

Nestle, Eberhard, Erwin Nestle, Barbara Aland and Kurt Aland. Novum Testamentum Graece. Stuttgart: Deutsche Bibelgesellschaft, 2001. Nöldeke, Theodor. "Review of F. Baethgen, Syrische Grammatik des Mar Elias von Tirhan." Göttingische gelehrte Anzeigen 1 (1880): 721-734. - Kurzgefaßte Syrische Grammatik. Darmstadt: Wissenschaftliche Buchgesellschaft, 1898 [1966], Nowicki, Paul. Beiträge zur Syntax inAphrems Memren. Freiburg: 1916. Pazzini, Massimo. Grammatica siriaca. Jerusalem: Franciscan Printing Press, 1999. Parisot, Jean. Aphrahatis Demonstrationes 7, II. Paris: Firmin-Didot, 1894. Payne Smith, Jessie. A Compendious Syriac Dictionary. Oxford: 1903. Payne Smith, Robert. Thesaurus Syriacus, 7-77. Oxford: 1879-1901. Pennacchietti, Fabrizio. "Verbo neo-aramaico e verbo neo-iranico." Pages 93-107 in Tipologie della convergenza linguistica - Atti del convegno della Società Italiana di Glottologia, Bergamo, 17-19 dicembre 1987. Edited by Vincenzo Orioles. Pisa: Giardini editore, 1988. Pérez Fernández, Miguel and John Elwolde (transi). An Introductory Grammar of Rabbinic Hebrew. Leiden: Brill, 1997. Perlmutter, David. "Impersonal passives and the unaccusative hypothesis." Pages 157-189 in Proceedings of the fourth Annual Meeting of the Berkeley Linguistics Society. Edited by J. Jaeger et al. Berkeley: Berkeley Linguistic Society, University of California, 1978. Peursen, van W. T h . "Inflectional Morpheme or Part of the Lexeme: Some Reflections on the Shaphel in Classical Syriac." Pages 41-57 in Foundations for Syriac Lexicography III. Colloquia of the International Syriac Language Project. Edited by Janet Dyk and W. T h . van Peursen. Piscataway, NJ: Gorgias Press, 2008. Phillips, George Syriac Grammar. The Third Edition. Cambridge - London: 1866. Plaut. Aul. - Plaute. Comédies. Amphitruo-Asinaria-Aulularia. Tome I. Texte établi et trad, par A. Ernout. Paris: Les Belles Lettres - 9 e tirage revu, corr. et augm. par J. C. Dumont, 1989. Plaut. Captivi - Plaute. Comédies. Bacchides-Captivi-Casina. Tome II. Texte établi et trad, par A. Ernout. Paris: Les Belles Lettres, 1943. Plaut. Trinummus - Plaute. Comédies: Trinummus. Truculentus. Vidularia. Fragmenta. Tome VII. Texte établi et traduit par Alfred Ernout. Paris: Les Belles Lettres, 1961. Pusey, Philip E. Tetraeuangelium Sanctum juxta simplicem Syrorum versionem. Oxford: 1901. Rahlfs, Alfred. Septuaginta id est Vêtus Testamentum graece iuxta 70 interpretes (9th ed.). Stuttgart: Württembergische Bibelanstalt, 1971. Rainey, Anson. "Mesha' and Syntax." Pages 287-306 in The Land That I Will Show You: essays on the history and archaeology of the ancient Near East in honour of J. Maxwell Miller. Edited by John A. Dearman and Matt P. Graham. Sheffield: Sheffield Academic Press, 2001. Reichenbach, Hans. Elements of Symbolic Logic. London [New York] : Macmillan [The Free Press], 1947 [1966], Retsö, Jan. Diathesis in the Semitic languages: a comparative morphological study. Leiden: Brill, 1989.

BIBLIOGRAPHY

177

Retsö, Jan. "Diathesis." Pages 622-626 in Encyclopedia of the Arabic Language and Linguistics. Edited by Cornelis H. M. Versteegh. Leiden: Brill, 2006. Romagno, Domenica. "Diatesi indoeuropea e verbi di movimento greci: alcune considerazioni sull'intransitività." Archivio Glottologico Italiano 87 (2002): 163174. Rosen, Carol G. " T h e Interface between Semantic Roles and Initial Grammatical Relations." Pages 38-77 in Studies in Relational Grammar II. Edited by David Perlmutter and Carol G. Rosen. Chicago - London: Chicago UP, 1984. Rosenthal, Frank. A Grammar of Biblical Aramaic. Wiesbaden: Harrassowitz, 1961. Rubin, Aaron D. Studies in Semitic Grammaticalization. Winona Lake, Indiana: Eisenbrauns, 2005. Rundgren, Fritihof. "Das altsyrische Verbalsystem. Vom Aspekt zum Tempus." Acta Societatis Linguisticae Uppsaliensis 11 (1960): 49-75. Sàenz-Badillos, Angel. Historia de la lengua hebrea. Sabadell: AUSA, 1988. Sàenz-Badillos, Angel and John Elwolde (transl.). A history of the Hebrew language. Cambridge: Cambridge University Press, 1993. Segal, Judah B. The diacritical point and the accents in Syriac. London: Cumberlege, 1953. Segert, Stanislav. Altaramäische Grammatik, mit Bibliographie, Chrestomathie, und Glossar. Leipzig: VEB Verlag Enzyklopädie, 1986. Seneca, Thyestes - L. Annaei Seneca, Tragoediae. Edited by Otto Zwierlein. Oxford : OUP, 19883. Shibatani, M. and P. Pardeshi. " T h e causative continuum." Pages 85-126 in The Grammar of Causation and Interpersonal Manipulation. Edited by Masayoshi Shibatani. Amsterdam - Philadelphia: Benjamnins, 2001. Siebesma, Pieter Anne. The function of the niph'al in Biblical Hebrew. Assen Maastricht: Van Gorcum, 1991. Slotki, Israel W. '"Erubin. transl. [from the Hebrew] into English with notes, glossary and indices." In Hebrew-English edition of the Babylonian Talmud. Seder Mo'ed. Edited by I. Epstein. London: Soncino Press, 1983. Sorace, Antonella. "Gradients in Auxiliary Selection with Intransitive Verbs." Language 76 (2000): 859-890. Speiser, Ephraim A. " T h e Durative Hithpael: A tan-Form J AOS 75 (1955): 118121. Squartini, Mario. "Metalinguaggio traduzione e il 'raffreddamento' della terminologia." Pages 258-266 in Per le vie del mondo. Edited by Piero De Gennaro. Torino: Trauben, 2009. Steiner, Gerd. "Die sog. tan-Stämme des akkadischen Verbums und ihre semitischen Grundlagen." ZDMG 131 (1981): 9-27. Stempel, Reinhard S. Abriss einer historischen Grammatik der semitischen Sprachen. Frankfurt am Main: Peter Lang, 1999. Stevenson, William B. A Grammar of Palestinian Jewish Aramaic. Oxford: Clarendon, 1924. Streck, Michael. "Das «Perfekt» iptaras im Albabylonischen der HammurapiBriefe." Pages 101-126 in Tempus und Aspekt in den semitischen Sprachen. Edited by Robert Nebes. Wiesbaden: Harrassowitz, 1999.

178

OUTLINE OF MIDDLE VOICE IN SYRIAC

Talmon, Rafael. "Foreign influence in the Syriac grammatical tradition." Pages 337-341 in History of the Language Sciences. Edited by Sylvain Auroux, E. F. K. Koerner, Hans-Josef Niederehe and Cornells H. M. Versteegh. Berlin, New York: Mouton de Gruyter, 2000. Taylor, David G.K. "Bilingualism and Diglossia in Late Antique Syria and Mesopotamia." Pages 298-476 in Bilingualism in ancient society: language-contact and the written text Edited by J. N. Adams, J. Jaeger and S. Swain. Oxford: Oxford University Press, 2002. Terentius, Adelphoe; Andria- Publi Terenti Afri, Comoediae. Edited by Robert Kauer and Wallace M. Lindsay. Oxford : OUP, 1926. Testen, David. "Morphological Observations on the stems of the Semitic 'nota accusativi'."Archivfür Orieniforschungen 44-45 (1998): 215-221. Thackston, Wheeler M. Introduction to Syriac. Bethesda, Maryland: IBEX Publishers, 1999. Timberlake, Alan. "Aspect, tense, mood." Pages 280-333 in Language typology and syntactic description. Vol. Ill: Grammatical categories and the lexicon. Edited by T i m o t h y Shopen. Cambridge: Cambridge University Press, 1997. Tonneau, Raymond M. Sancti Ephraem Syri in Genesim et in Exodum Commentarii. Louvain: 1955. Tropper, Joseph. "Althebräisches und semitisches Aspektsystem." Zeitschrift für Althebraistik 11 (1998): 153-190. Tsereteli, Konstantin. Grammatica generale dell'aramaico. Torino: Zamorani, 1995. Uhlig, Gustav and Richard Schneider. Grammatici Graeci I-IV. Leipzig: Teubner, 1878 f. Van Rompay, Lucas. "Some preliminary remarks on the origins of Classical Syriac as a standard language." Pages 70-89 in Semitic and Cushitic Studies. Edited by Gideon Goldenberg and Shlomo Raz. Wiesbaden: Harrassowitz, 1994. Van Valin, Robert D. "Semantic Parameters of Split Intransitivity." Language 66 (1990): 221-260. Vendler, Zeno. Linguistics in Philosophy. Ithaca, London: Cornell University Press, 1967. Voigt, Rainer M. "Die tan-Stämme und das System der Verbalformen im A k k a d i s c h e n . " Z D M G 137 (1987): 246-265. Waltke, Bruce K. and Michael P. O' Connor. An introduction to biblical Hebrew syntax. Winona Lake, Indiana: Eisenbrauns, 1990. Watt, John. "Grammar, Rhetoric and Enkyklios Paideia in Syriac." ZDMG 143 (1993): 45-71. Wright, William. A Grammar of the Arabic Language, translated from the German of Caspari and edited with numerous additions and corrections, 3rd ed. revised by W. Robertson Smith and M. J. de Goeje. Cambridge: Cambridge University Press, 1859 [1977], - Apocryphal Acts of the Apostles, I. The Syriac texts. II. The English translation. London: Williams and Norgate, 1871. Yousif, Butrus and S. Khalil Samir. "La version arabe de la troisième démonstration d'Aphrahat (Sur le jeûne) - Texte établi par Kh. Samir. Etude et traduction de P. Yousif." Pages 31-66 in Actes du deuxième congrès internationsl

BIBLIOGRAPHY

179

d'études arabes chrétiennes Edited by Samir Khalil Samir. Rome: Pontificium Institutum Studiorum Orientalium, 1986. Zaborski, Andrzej. "On the interplay of Tense, Aspect and Aktionsart in Semitic Languages." Pages 869-876 in Sprich doch mit deinen Knechten aramäisch, wir verstehen es: 60 Beiträge zur Semitistik: Festschrift fur Otto Jastrow zum 60. Geburtstag. Edited by W. Arnold and H. Bobzin. Wiesbaden: Harrassowitz, 2000.