THE DEVELOPMENT OF THE SYNTAX

OF POST-BIBLICAL HEBREW
 STUDIES IN SEMITIC
LANGUAGES AND LINGUISTICS

EDITED BY

T. MURAOKA AND C.H.M. VERSTEEGH

VOLUME XXIX

THE DEVELOPMENT OF THE SYNTAX

OF POST-BIBLICAL HEBREW

~<q-jt-tG//)~
...., ..<)

7'
o£ r-
\"" I ,...
? ..
~ c,
. I 68) .
 THE DEVELOPMENT OF
THE SYNTAX
OF POST-BIBLICAL HEBREW
BY

CHAIMRABIN

.;:,~ftt-£GI[)~
..., ...,
;.-
<C. r-
..... ,...
? ..
0(' . 16 8') . "

BRILL
LEIDEN . BOSTON' KOLN
2000
 This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Rabin, Chaim.
The development of the syntax of post-biblical Hebrew / by Chaim
Rabin.
p. cm. - (Studies in Semitic languages and linguistics, ISSN
0081-8461 ; 29)
Reprint of the author's thesis (D.Phil.-Christ Church (University
of Oxford)).
Includes bibliographical references and index.
ISBN 9004114335 (cloth: alk. paper)
1. Hebrew language, Medieval-Spain-Syntax-History. 2. Hebrew
language, Medieval-France, Southern-Syntax-History. 1. Tide.
H. Series.
PJ4973.R33 1999
492.4'7'09460902-dc21 99-20 718
CIP

Die Deutsche Bibliothek-CIP-Einheitsaufna1nne

Rabin, Chaim.:
The development of the syntax of post-biblical Hebrew / by Chaim
Rabin. - Leiden ; Boston; Koln: Brill, 1999
(Studies in Semitic languages and linguistics ; 29)
ISBN 90-04-11433-5

ISSN 0081-8461
ISBN 9004 114335

© Copyright 2000 by Koninklijke Brill NV, Leiden, The Netherlands

All rights reserved. No part rif this publication mqy be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.

Authorization to photocopy items for internal or personal

use is granted by Koninklijke Brill provided that
the appropriate fees are paid directly to The Copyright
Clearance Center, 222 Rosewood Drive, Suite 910
Danvers MA 01923, USA.
Fees are subject to change.

PRINTED IN THE NETHERLANDS

 CONTENTS

Preface ........................................................................................ . IX
Foreword by Lewis Glinert ....................................................... . Xlll
Abbreviations ............................................................................. . xv

Introduction
1. Biblical and Mishnaic Hebrew ..................................... . 1.
2. The Early Liturgy ........................................................... . 7
3. The Piyylit ..................................................................... . 11
4. The Midrash .................................................................. .. 19
5. Prose-literature in North-Western Europe ................... . 26
6. The Karaites ................................................................... . 28
7. The Revival of Biblical Hebrew Prose ......................... . 31
8. Spanish Hebrew Poetry ................................................. . 39
9. The First Period of Provenl):al Hebrew Literature ....... . 50
10. The Translators ............................................................... . 56
11. Maimonides ..................................................................... . 63
12. The Second Period of Spanish-Provenl):al Literature ... . 68
13. The Third Period of Spanish Literature ..................... . 76
14. Later Developments of the Hebrew Language ........... . 79

I. General Syntax of the Noun

1. The Article ..................................................................... . 85
2. Concord with Regard to the Article ............................. . 86
3. The Article with Numerals ........................................... . 87
4. Gender of Nouns ........................................................... . 88
5. Concord with Regard to Number ................................ .. 92
6. The Construct State ....................................................... . 93
7. Circumlocution Genitive 94

11. The Pronouns

8. Personal Pronouns 96
9. Demonstrative Pronouns ................................................. . 97
10. Sentence-anaphorics ....................................................... . 101
11. Indefinite Pronouns ......................................................... . 102
VI CONTENTS

Ill. The Nominal Clause

12. The Copula ................................................................ 106
13. The Temporal Copula .............................................. 107
14. The Verb of Existence .............................................. 109
15. Modal Copulae .......................................................... 110

IV. The Verbal Clause

16. The Indefinite Subject ......................... ..................... 113
17. The Agent of the Passive Verb .............. .... .... .......... 113
18. Some Remarks on the Tenses .................................. 115

V. The Object
19. The Direct Object ..................................................... 117
20. The Indirect Object .................................................. 118
2l. The Cognate Object ... .... ........................................... 118
22. The Double Accusative ............................................. 120

VI. The Particles

23. Some Remarks on Adverbs ...................................... 123
24. The Prepositions ........................................................ 125
25. Negation ...................................................................... 127
26. Exceptive Particles .. .... .... .... .................. ............ ......... 129
27. Some Coordinating Conjunctions ............................ 131

VII. Word Order

28. The Verbal Clause with Object 133
29. Adverb and Predicative 134

VIII. The Compound Sentence

30. Extraposition 137

IX. The Verbal Nouns

3l. The Le-Infinitive 141
32. The Subject of the Infinitive Clause 144
33. Gerund and Nomen Verbi 150

X. Substantive Clauses
34. Substantivization of clauses ....................................... 155
35. Clauses as Subjects .................................................... 159
36. Clauses in other Positions ... ................ ........ .............. 165
 CONTENTS vu

XI. Adjective Clauses

37. Asyndetic Relative Clauses ........................................ 171
38. Syndetic Relative Clauses ........ ....... ...................... ..... 172
39. Substantivized Relative Clauses ................................ 174

XII. Conditional Clauses

40. Conditional Clauses 179

Synoptic Table ............................................................................ 183

Bibliography ................................................................................ 188
Index of Passages: Classical Sources .. ............. ......................... 197
 PREFACE

The present thesis is an attempt to discover the general features of

the syntax of the type of Hebrew used by Jews in Southern France
and Spain from the twelfth to the fifteenth century, C.E. in non-
artistic prose, and to establish its relationship to Biblical Hebrew,
Mishnaic Hebrew, and the early mediaeval developments of these two.
The history of Hebrew from the time when it ceased to be spo-
ken was to a large extent a process of re-uniting its two divergent
forms: Biblical Hebrew and Mishnaic Hebrew. In Spanish-Provenc;al
Hebrew a complete synthesis was achieved for the first time. It unites
and harmonizes elements from all preceding stages of literary Hebrew,
and is therefore the most suitable point from which to gain a per-
spective of the development of the Hebrew language during the ear-
lier middle ages. Owing to its large and well-preserved literary
production we are in a better position to know it than we are with
regard to any of the preceding periods, so that our researches into
this form will sometimes help us to elucidate phenomena in other
forms of the language.
No systematic research into the grammar and syntax of Spanish-
Provenc;al prose literature has hitherto been undertaken, although
the poetical language of the 11 th and 12th centuries had attracted
some attention. The study of its philosophical and scientific termi-
nology was initiated by Goldenthal in 1850, but its general vocab-
ulary was neglected until Ben-Yehudah, who discovered in it a rich
source of words that could be made to serve the needs of Modern
Hebrew.
This lack of interest in the linguistic character of mediaeval Hebrew
literature at a time when its historical and literary importance was
fully recognized and much energy devoted to its publication and
analysis appears to be due principally to the nineteenth-century the-
ory that only "living" languages had any value for linguistic research.
Spanish-Provenc;al Hebrew was certainly a "dead", i.e., purely literary,
idiom. It had the added stigma of being "artificial", i.e. of having
made innovations upon the material provided by the older, "living",
forms of the language. This was a crime which even Renan could
not forgive. To-day, however, we have overcome the obsession of
x PREFACE

the reconstructive method in historical linguistics, and also restored

descriptive linguistics to its proper place. We now realize that liter-
ary idioms have some interest for the study of the principles of lan-
guage, in spite of the fact that every one of them is to some extent
both dead and artificial. Indeed, the feeling is growing that mixed
and "dead" languages offer many points of special interest for the
development of language in general.
Spanish-Proven~al Hebrew is in this respect a unique laboratory-
experiment. Its entire development took place in full light of history,
and we know in great detail all the elements that entered into its
make-up. We can, therefore, attempt with some confidence to enter
into the question why some elements were chosen and others rejected,
and also to discover the conditions under which independent evo-
lution of traditional material takes place. It is not claimed that the
present thesis offers a systematic study of these problems. But in a
few instances they have been touched upon, and the results seem
encouraging enough to warrant the continuation of that line of
research once the primary object, the description of the language
and the determination of the provenience of its features is achieved.
I am fully conscious of the fact that this aim is still far away. The
following pages give no more than a study of selected features as a
test of the method employed. It suffers from all the faults of a first
attempt on a very wide subject in which everything is still unknown,
and in which a method has to be created out of the experience
gained in the work itself. The work has not been made any easier
by the lack of systematic modern studies on the syntax of every other
period of Hebrew. Even the syntactical structure of Biblical Hebrew
has yet to be established by more modern methods than those of
Gesenius-Kautzsch and Konig. The only relatively up-to-date stud-
ies are those of Brockelmann in the second volume of his Grundriss
and of Segal in his Di~du~ leshon ha-Mishnah. Both are avowedly
incomplete, but without them the present investigation would have
been impossible. In spite of many efforts, I did not succeed in obtain-
ing volume 11. (Syntax) of M. Szneider's Torath ha-lashon be-hith-
patte&uthah, which appeared recently in Palestine. I could not even
ascertain whether Mr. Szneider has dealt in that book with the period
treated here.
This study restricts itself to syntax because the morphological
differences between the various stages of Hebrew are (at least in so
far as they are discernible in the consonantal script) small enough
 PREFACE Xl

to be negligible for the history of the language. It further deals exclu-

sively with original Hebrew works of a non-artistic character, since
translations offer many special problems that would only confuse the
issue, and poetry and artistic prose employed an idiom basically dis-
tinct from the one we are investigating. An exception must be made
for Shem!ob Falaquera's Meba#esh which, though written in rhymed
prose, uses almost pure Spanish-Proven<;al Hebrew.
The data have been collected from a wide range of works, some
of them critically edited, others in unreliable printed editions or in
manuscripts. It was, of course, not possible to go into questions of
textual criticism in order to eliminate changes of the original texts
made by copyists, printers or editors. The danger of thus recording
printers' errors as features of Spanish-Proven<;al Hebrew is, however,
obviated if the same feature occurs in works by different authors.
Data for which this safeguard was not available have only been
adduced in a very few cases for their particular interest, and the
state of documentation has been pointed out.
 FOREWORD

If ever the phrase 'a man ahead of his time' were aptly to be said,
it might be said of Chaim Rabin, Cowley Lecturer in Post-biblical
Hebrew at the University of Oxford from 1941 to 1956 and Professor
of Hebrew at the Hebrew University of Jerusalem from 1956 to
1985. The present volume, the first published study of Diasporic
Hebrew syntax, was composed mirabile dictu as a doctoral disser-
tation, The Development if the ~ntax if Post-Biblical Hebrew, submitted
to the University of Oxford in 1943. The systematic study of post-
Biblical Hebrew was then in its infancy and Hebrew syntax an almost
unheard-of field of research. It was still commonly believed that
Mishnaic Hebrew was an artificial concoction of rabbinic minds and
that Medieval Hebrew was an unedifying mishmash of Arabic or
Yiddish calques and Paytanic phantasmagoria. The intervening half-
century has continued to pay scant attention to the evolution of
Hebrew as a linguistically authentic and culturally sophisticated expres-
sion of Diaspora life and letters. Our encyclopedias and histories of
Hebrew still have precious little to say about Hebrew in the Sephardic
world and even less about the linguistic contributions of the Ashkenazic
republic of Hebrew letters. What little we do know about key Ash-
kenazic works like Sifer Hasidim is largely due to Chaim Rabin's own
students.
Syntax, too, was for a long time a sideshow in Hebraic acade-
mia. Even as the new spoken and written Hebrew of the Zionist
Yishuv was developing apace and complex syntactic issues were being
decided by mass instinct in the streets of Tel Aviv, the pages of Ivrit
textbooks and Hebrew journals blithely went on paying allegiance
to morphology and phonology, the twin icons of the Medieval gram-
marian's craft-to the total exclusion of syntax, the meat of lan-
guage. "How can this be?" I once asked Chaim Rabin innocently
as we sat over one of our timeless cups of English tea on Jerusalem's
Keren Kayemet St. He gave a mischievous smile: What had been
good enough for Medieval grammarians of Arabic, he said, was
apparently still good enough for the "grammarians" of twentieth cen-
tury Hebrew. Palpably, it was not good enough for him.
It was a nagging enigma to me for the 25 years that I knew him
XIV FOREWORD

that Chaim Rabin's doctoral dissertation remained unknown, even

in the arcane corridors of Jerusalem academe. Why had he not
"worked it up" into a book, in the manner of today's academic
hopefuls? I tried to content myself with the thought that this was a
man of numerous parts, that the syntax of post-Biblical Hebrew was
just on one of the many back burners of his mind. This was indeed
Rabin's second doctoral thesis; the first, Studies in Early Arabic Dialects,
was submitted to London University's School of Oriental Studies in
1939 [and was subsequently revised and published as Ancient West-
Arabian (Taylor's Foreign Press: London, 1951, and translated into
Arabic and published in Kuwait in 1986), to become the classic in
the field.]
Now, alas, Chaim Rabin is no longer with us, and still many of
his dreams and visions are unfulfilled. His 1943 thesis is in many
ways as fresh and challenging today as the day it was submitted. It
is my hesed shel ernet to be able by way of this preface, to introduce
this valuable contribution to Hebrew scholarship. Thanks are also
due to his wife Batya and to Prof. T. Muraoka of Leiden, a former
student of the author, for their careful proofreading and to Mr. Ohad
Cohen for compiling the index.
Let it not be said that Hebrew during two millennia of Diaspora
was petrified, a historical irrelevance. In the study houses and court-
yards of medieval Aragon and Provence, it grew and flourished. Far
more than in its morphology or its lexis, it was in its syntax-over
the conscious horizon of most authors of grammarians-the Hebrew
was able to evolve and adjust. To recognize the vital force of the
written medium in traditional Jewish life and to identifY it above all
in the syntax was an achievement without parallel in Hebrew schol-
arly circles. From today's perspective, too, as fifty years ago, Chaim
Rabin towers over his time.

Lewis Glinert,
Dartmouth College,

6 September, 1999
 ABBREVIATIONS

Abr. Maim. Abraham Maimonides

B., b. bar or ben or ibn; Babylonian Talmud
BH Biblical Hebrew
Brock.* Brockelmann's Grundriss, vo!. ii
Comm. Commentary
Enc. Jud. Encyclopaedia Judaica
f. folio
Falaq. Shemtov b. Falaquera
Gers. Gersonides
Ges.* Gesenius' Hebrew Grammar
JBL Journal if Biblical Literature
Jew. Enc. Jewish Encyclopaedia
JQ.R Jewish Qyarler?J Review
JQ.R (O.S) Jewish Qyarler?J Review (Old series)
Lesh. Leshonenu
m. Mishnah
Maim. Maimonides
MGW] Monatsschrift der Gesellschaflfiir die Wissenschafl des JudentumJ
MH Mishnaic Hebrew
Reck.* Reckendorf's Arabische ~ntax
RE] Revue des etudes juives
S. Sepher
Segal, Seg.* Segal's Mishnaic Hebrew Grammar
SH Spanish-Provem,:al Hebrew
st. stanza
Tg. Targum
ZAW Zeitschrift for altlestamentliche Wissenschafl
ZDMG Zeitschnft der deutschen morgenliindischen Gesellschafl
ZFHB Zeitschrifl flir hebriiische Bibliographie

*Quoted by paragraphs.
 INTRODUCTION

A SKETCH OF THE DEVELOPMENT OF

LITERARY HEBREW

1. Biblical and Mishnaic Hebrew

Es gibt keine Spriinge in der Geschichte des Stils: wenn wir einen RiB
zwischen den Epochen zu sehen glauben, dann ist es an uns, tiefer zu
forschen, damit sich alles erklart. Wenn wir einen starken Wechsel im
Stil sehen, der plotzlich eintritt, eine stilistische Revolution, so ist das
ein Zeichen dafiir (u with umlaut), daB der alte Stil schon lange ver-
wischt und zerftossen war, oder daB ein neuer Stil im stillen gewach-
sen ist: darum wird Altes jah von Neuem verdrangt.
Lipschiitz, Vom lebendigen Hebriiisch, S. 69.

The history of mediaeval Hebrew is the history of a purely literary

language, written by people speaking a variety of languages but never
Hebrew. It shares this character with a number of languages that
continued their written life after having ceased to be spoken: Latin,
Greek, Old Bulgarian, Sanskrit, Pehlevi, and Classical Arabic, to
name only a few of the better-known ones. Mediaeval Hebrew devel-
oped according to the same laws as those languages.
In one respect, however, Hebrew is different from all other "dead"
languages. These are always focused on one classical form of lan-
guage, which is more or less perfectly imitated and used as a meas-
ure by which the quality of style is judged; Hebrew is focused on
two such forms. The various periods in the history of the language
are distinguished mainly by the proportion in which these two forms
entered into the make-up of the style. Moreover, in practically every
period, there existed several styles which differed from each other
in just this one respect.
In later mediaeval times the difference was not always recognized,
e.g. Shabbethai ha-Sofer on Aboth 2.12 (Taylor, ii 144): mt!JQi1 ':J
~iPQi1 1'1t!J';::J i1''1Jt!J, cf. also ~imQi, preface to Mildzlol.
The two basic forms of the Hebrew language were originally used
at different periods of its literary history. Biblical Hebrew, the earlier
form, is the idiom employed in the books of the Canon, composed
between 900 and 150 B.C.E. The later form appears first in oral
2 INTRODUCTION

traditions which date back to the second century B.C.E. but were not
fixed in writing before 200 C.E. I It became the vehicle of a legal,
ethical and narrative literature extending into the 12th century C.E. 2
For convenience, we may call it Mishnaic Hebrew, after the Mishnah,
the oldest work composed in it. 3
The two forms of the language differ from each other in mor-
phology, syntax, vocabulary and style. 4 The majority of these points
of difference cannot be explained by assuming a straightforward
development of forms that existed in Biblical Hebrew. The indu-
bitable similarity of some of these features to Aramaic 5 led A. Geiger 6

1 It is generally assumed now that the Mishnah was written down by Rabbi
Judah himself, although many mediaeval scholars believed that until the 5th cen-
tury C.E. it was only handed down orally, cf. Strack, Einleitung, pp. 15f.
2 See end of chapter IV.
:3 The term "Mishnaic Hebrew" is the usual one in the English-speaking coun-
tries, and in Modern Hebrew works. It has the disadvantage of making it appear
as if only the language of the Mishnah itself was intended (as Biblical Hebrew is
the language of the Bible), but it is preferable to the "Neo-Hebrew" of the Germans,
which ignores all later forms of the language.
4 Of course, they also differed in their phonology. We can only to a very lim-
ited extent reconstruct the history of Biblical Hebrew consonants and vowels. The
consonantal spelling of our Massoretic text is, except for some minor graphical
details identical with that of the epigraphical texts of the seventh century B.C.E.,
and the spelling of the Mishnah is phonologically the same. We do not know at
what point in history this spelling ceased to be phonetical and became purely ety-
mological. The vowel-system of the Massoretes represents the reading pronuncia-
tion of the ninth century C.E. We know that this had undergone profound phonological
changes since the second century B.C.E., when the Septuagint transcriptions were
made. Even the Septuagint, however, was made long after the extinction of spo-
ken Biblical Hebrew, and probably represents merely the Mishnaic pronunciation
of the Biblical text. We do not know whether that tradition preserved the old sounds
to some extent (like the Modern Arabs' reading of the Koran) or replaced them
by the etymologically correspondent Mishnaic ones (as the modern Abyssinians' pro-
nunciation of Ge'ez). The pointing of Mishnah manuscripts, as far as Biblical words
are concerned, is a replica of the Massoretic vocalization. The value of the tradi-
tional pointing and pronunciation of the purely Mishnaic words and forms has not
yet been established. There certainly exists no evidence, at any rate, to justify the
statement of Szneider (Leshonenu iii 18) that the vowel system of Mishnaic Hebrew
was the same as that of Biblical Hebrew.
5 The writers on this subject have usually been content to prove that a word
was to be found in some Aramaic dialect or other, without trying to establish the
influence of one definite Aramaic dialect on Mishnaic Hebrew. The only attempt
to define the provenience of the Aramaic elements is by Avineri, Leshonenu iii 278f,
who gives lists of words found both in the Targumim and in Mishnaic Hebrew. In
view of the artificial character of the Targumic idiom, this does not lead us very
far, quite apart from the probability of a strong influence of Mishnaic Hebrew on
the Jewish Aramaic dialects.
6 Lehrbuch, pp. 22-5. The theory had first been suggested by S. Levisohn, in his
grammatical introduction to the Mishnah-edition of 1812.
 INTRODUCTION 3

to the opinion that Mishnaic Hebrew was not a living dialect at all,
but an artificial jargon, the attempt of Aramaic-speaking scholars to
write Biblical Hebrew, and a most unsuccessful attempt at that.
The authors adhering to Geiger's view assume that Hebrew had
been replaced by Aramaic "at the beginning of the Hellenistic period"
(Brock. i, p. 9) Cf. also Charles, Daniel, p. xlv: "To get in touch
with his countrymen and to bring home to them the ideals for which
they stood, the author of Daniel could not do otherwise than write
in Aramaic. Only through the medium of the vernacular was this
possible, and the vernacular of his day was Aramaic."
As to the 4th century Schaeder, lranische Beitrdge I, pp. 225~6 argues,
quoting Lidzbarski, Ephemeris for semitische Epigraphik, iii. 82 and Cowley,
Aramaic Papyri, 118f., that Aramaic was not the vernacular of the
Elephantine Jews, but that they wrote it with Hebraims, never hav-
ing learnt to write their native Hebrew. 7 Schaeder also points out
relevantly that Palestine was the only place where Imperial Aramaic
became a spoken language. For Hebraims in Biblical Aramaic c(
H.H. Powell, The supposed Hebraisms in the Grammar if Biblical Aramaic,
1907, who denies all Hebraisms.
Geiger's view dominated the whole of the 19th century,8 but it
has been abandoned by most writers today, as it ignores the many
features which cannot satisfactorily be explained by borrowing from
Aramaic. Instead, we have come back to the view expressed for the
first time by S.D. Luzzatto 9 and, developed in great detail by
M. Segal in his various works,lo that Mishnaic Hebrew was the de-
scendant of a sister-dialect of Biblical Hebrew. This they assume to
have existed before the exile in colloquial use only, though it left
some traces in exceptional forms and constructions in the pre-exilic
books of the Canon.
It has been doubted whether Biblical Hebrew was ever spoken in
the form in which we find it in the earlier books. 11 However that

7 Against this C.G. Wagenaar, Dejoodse kolonie van Jeb-fiyene, Proefschr. Groningen
1928, pp. 248-53, the colonists originally spoke Hebrew, but in the 6th century
also Aramaic (with Hebraisms).
8 Brockelmann, in 1908 (Grundriss, i 10), still puts forward the Geiger theory as
an incontestable fact, without even mentioning the other view.
9 Prolegomeni, p. 99.
10 See the list of authorities by Matmon-Cohen, Leshonenu vi 174. Add to these

Bergstrasser, Einfiihrung, p. 46; Cassuto in Enc. Ital. xiii 357.

11 Segal, Diqduq, p. 7, believes that "complicated" usages, such as the consecutive

tenses, the gerund with preposition, or the emphatic cognate object, would not have
been likely to survive long in popular speech, though he grants that they must have
4 INTRODUCTION

may be, it is most probable that even before the Babylonian exile
the actual colloquial speech of the population had become rather
different from the literary language. The few epigraphic documents
of the later kingdom that we possessl 2 were written by professional
scribes and therefore tell us no more about the language of every-
day life than the literary productions incorporated in the Bible. 13
The distance between the literary and the colloquial language was
further increased by the upheavals of the exile and the restoration.
The mixing of people from all parts of the country produced a sub-
standard Koine into which elements from the Northern dialects
entered to some extent. 14 Biblical Hebrew continued as the literary
idiom. During, and immediately after, the exile it was still written
in its purity. From the time of Nehemiah l5 onwards, however, a
marked change came over Hebrew style. While obviously still endeav-
ouring to adhere to the laws of Biblical Hebrew, the authors of Ezra,
Nehemiah, the Chronicles, and many Psalms, succeeded in this only
partially. The result is a laboured, clumsy style which contrasts very
sharply with the easy elegance of early Biblical Hebrew. This is obvi-

existed in it at some time. Matmon-Cohen, Leshonenu vi 172, goes further than that,
and maintains that these, and the complicated vocalic changes in noun-flexion, were
developed only in the literary language, because they are "against the spirit of a
colloquial popular speech". The weakness of these arguments has been exposed by
Szneider, Leshonenu viii 112-122. It is, however, quite possible that Biblical Hebrew
was a court-language, absorbing elements from many dialects and different from each,
like Homeric Greek or Classical Arabic. Its wealth of duplicate forms (cC G.R. Driver,
Problems if the Hebrew Verbal /iystem, p. 105) makes this probable. Furthermore, if the
mixed-language theory is correct (see below note 26), it is not impossible that only
the literary language was as purely Canaanite as our actual Biblical Hebrew, while
the popular language preserved more of the original speech of the Israelites. These
are, however, mere guesses. Our material is quite insufficient to check them.
12 Of all the epigraphic material, the Siloam inscription and the Lachish Letters
are the only texts of sufficient length to permit any linguistic conclusions.
13 Dhorme (Revue Biblique xxxix 62) assumes that, as a written language at least,
Hebrew in its Biblical form, was used by all classes of the people: "l'hebreu ne
restait point confine dans les hautes spheres de la nation".
14 At any rate, the particle she- occurs in an old Northern text, the Song of
Deborah. The Song of Songs, which contains so many Mishnaic features, is con-
sidered by many scholars to be an early product of North Israel (cC Driver, Introduction,
p. 449). Szneider, Leshonemu vi 307, on the contrary, considers Mishnaic Hebrew to
be South:Judaean, brought to Jerusalem by settlers from the South during the exile.
The view of Mishnaic Hebrew as a Koine is also adopted by Segal in his article
in Madda'e ha-Yahaduth i 37. According to him, the dialectal elements are due to
the influx of Northern refugees after the fall of Samaria.
15 Driver, Introduction, p. 505: "The great turning-point in Hebrew style falls in
the age of Nehemiah."
 INTRODUCTION 5

ously due to the fact that these men wrote in an imperfectly acquired
idiom quite different from their ordinary speech. However, no direct
influence of Mishnaic Hebrew is to be discovered in these books. 16
Mishnaic elements appear only in the later books, Esther and Ec-
clesiastes, in which the striving for Biblical diction is hardly more than
a pious wish. The Mishnaic elements in these books presuppose a
colloquial that was practically identical with the language of the
Mishnah itself.
The very last books in Biblical Hebrew, Ben Sira l7 and the Hebrew
parts of Daniel, actually show a return to a somewhat purer Biblical
style. This was achieved by a skilful use of quotations and semi-quo-
tations,18 but the Mishnaic colloquial of the authors breaks through
in almost every sentence. 19
The elevation of Mishnaic Hebrew to the rank of a literary idiom
is bound up with the Pharisaic movement. The middle classes, who
constituted the rank and file of that party, did not possess the nec-
essary education to handle Biblical Hebrew. The possibility of acquir-
ing knowledge through the medium of their colloquial meant for
them emancipation from the tutelage of priesthood and aristocracy.2o
Although Mishnaic Hebrew is thus shown to be an autochthono~s

16 Kropat, o?Jntax, pp. 73-4, Driver, loc. cit.

17 The language of Ben Sira has been analysed, and the Mishnaic elements
pointed out, by Strauss, Sprachliche Studie and Segal, ushonenu vii 100-120.
18 Daniel's prayer, ix 4-20, the only part of the book which does not contain
Mishnaic elements (cr. Charles, p. 227), consists almost entirely of quotations from
earlier texts, cr. Driver, op. cit., p. 507.
19 All the "Aramaisms" in the Hebrew parts of Daniel listed by Charles, p. cviii

are quite current in Mishnaic.

Segal, International Journal qf Apocrypha, 1910, p. 81, suggests that the Hebrew orig-
inals of most apocrypha and pseudepigrapha were written in a similar style; one
"is able to detect, even under the guise of a secondary version, Mishnaic expres-
sions and usages."
This is not the place to discuss the question to what extent, and by what parts
of the population, Aramaic was spoken in the 2nd century B.C.E. The theory adhered
to here excludes the extremist view of Charles (p. xlv) that Aramaic was the only
language understood by the great mass of the people (this is, of course, also the
view underlying Geiger's theory). It presupposes that a certain part at least of the
urban population spoke Mishnaic Hebrew. Otherwise it would be quite unintelli-
gible that Biblical Hebrew-which the author of Daniel could still write in 165
B.C.E.-was abandoned in favour of Mishnaic.
20 If the Zadokite Fragment were really Zadokite and of the first century C.E.
(Charles, Apocrypha and Pseudepigrapha ii 788), it would be an eminent illustration of
the social factor in the language struggle, showing that the Sadducees clung to
Biblical Hebrew long after Mishnaic Hebrew was firmly established.
6 INTRODUCTION

development, it has no doubt absorbed foreign elements to a much

larger degree than Biblical Hebrew. 21 This is due both to the closer
contact with foreign cultures which existed during the post-exilic
period 22 and to the fact that an urban colloquial never has great re-
sistance to borrowings. There are many Greek23 and some Persian 24
words. There is also a number of unquestionable Aramaic loans,25
that betray their origin by their phonology, pattern, or peculiar turn
of root-meaning. These need not have come in from the Aramaic col-
loquial of the 'Am ha-arq, but may derive from contact with extra-
Jewish populations, from the Aramaic administrative language of the
Persian Empire, or may even date from the Babylonian exile.
A good many of the seemingly Aramaic lexical elements, and cer-
tainly most of the apparent Aramaisms of grammar and syntax, are
not due to post-exilic borrowing. Either they belong to the common
West-Semitic stock, and their absence in Biblical Hebrew is an inno-
vation particular to that dialect,26 or they may be remnants of the
language the Israelites spoke before immigrating into Canaan-

21 The number of loan-words in earlier Biblical Hebrew is generally underesti-

mated. Quite apart from the numerous Accadian, Sumerian, Egyptian and Hittite
borrowings, there are probably many from the lost Hurrian dialects of Canaan.
They would account for the high proportion of words without satisfactory ety-
mologies. Then there were of course borrowings from other Canaanite dialects.
22 Cf. Matmon-Cohen, Leshonenu vi 179.
23 Collected by Krauss, Lehnwo"rter.
24 An incomplete list in S. Telagdi, Essai sur la phonologie des emprunts iraniens
en Arameen talmudique, Journal Asiatique ccxxvi (1935) 178-256.
25 Mannes, although including many doubtful cases, could find only 175 Aramaic
roots in Mishnaic Hebrew for half the alphabet. Counting three derivations per
root, we arrived at a maximum of 1000 Aramaic loanwords, i.e., much less than
10 per cent of the Mishnaic vocabulary. If we compare this with the 40 per cent
of Latin and Romance elements in English, we can hardly feel justified in callimg
Mishnaic Hebrew an Aramaic jargon. Cf. also the statistics of Segal, Grammar, pp.
46-53: of the 1,400 verbal roots of Mishnaic, only 25 are definitely Aramaic. This
is perhaps too low. The more ambitious statistics of Avineri, Leshonenu iii 279, are
of doubtful value.
The Aramaic ~:lt~, ~r.:l~ "my father", "my mother" are adduced by Noldeke,
(Semitische Sprachen, p. 26) as proof that the Rabbis spoke Aramaic as their natural
language. On the contrary, it is not infrequent in other languages to call one's par-
ents in public by learned terms. Thus the English public-schoolboy speaks of his
"pater". In Turkish one refers to one's parents only as "pederem" and "validem",
never by the Turkish "ata" and "ana".
26 Renan, Histoire, p. 144, who adds: "11 est it remarquer du reste, que les langues
semitiques different moins dans la bouche du peuple que dans les livres." Cf. also
Segal, Diqduq, p. 10.
 INTRODUCTION 7

whether Aramaic or Amorite. 27 It is reasonable to assume that such

remnants would have lingered on in popular speech, while the edu-
cated classes acquired a pure Canaanite.
By the time Mishnaic Hebrew found its first written expression in
the Mishnah, it was already rapidly disappearing from colloquial use,
its place being taken by Aramaic. It is still far from clear at what
date it finally ceased to be spoken in normallife,28 but it seems to be
certain that by the end of the third century C.E. it was to all practi-
cal purpose a dead language. 29 Like Biblical Hebrew, it enjoyed a
long and vigorous life as a purely literary idiom, and achieved some
of its finest manifestations during the ninth and tenth centuries C.E.
These two basic forms were the only developments of colloquial
Hebrew that attained to literary status. They are also the only two
forms of literary Hebrew that rested directly on a colloquial. Apart
from the short interlude of Paytanic style, the changes which took
place in Hebrew after 200 C.E. were not due to the internal forces
of the language, but either to the interplay of the two basic forms
or to the influence of other languages.

2. The EarlY Liturgy

When Mishnaic Hebrew replaced Biblical Hebrew as a vehicle of

literary composition, the older form of the language did by no means
cease to be an important factor in the linguistic consciousness of the
people. Study and worship brought them into daily contact with it.
Halakhic discussion demanded an understanding of the minutest
details of Biblical idiom. Most important of all, through the liturgical

27 Bauer-Leander, Histomche Grammatik, p. 23; G.R. Driver, Problems qf the Hebrew

Verbal /iystem, chapter x.
28 Noldeke (Semitische Sprachen, p. 25) thinks that Mishnaic Hebrew was originally

a true colloquial, but died out before Maccabeean times, and in its purely literary
use came under Aramaic influence. Ben-Yehudah, Prolegomena 254, thinks that the
final disappearance of Hebrew from daily speech even in scholarly circles took place
shortly after 300 C .E.
M. Zulay, Melila 1944, p. 73, considers it an open question when asked to what
extent(!) Hebrew ceased to be spoken, and even throws up the possibility of Kalir
and Yannai having known Hebrew as a spoken language, or at least drawing upon
creations of a colloquial development close to their own time.
29 I.e. in daily use. On certain occasions (in sermons, between Jews from different
countries, on Sabbath, etc. Hebrew was spoken throughout the Middle Ages.
8 INTRODUCTION

use of the Psalms, the people became accustomed to express their

poetical and religious feelings in the "Holy Tongue". I
When the speakers of Mishnaic Hebrew began to compose prayers
of their own, it was only natural that they should borrow from the
poetry of the Bible those abstract and poetical terms that were lacking
in the colloquial. They also replaced a number of common Mishnaic
words by Biblical ones, because the colloquial words appeared too
prosaic for addressing God, just as in Biblical Hebrew itself many
ideas were expressed by different words in prose and in poetry.
The grammatical and syntactical structure of the early prayers
was, however, purely Mishnaic. 2 Non-Mishnaic features occur with
extreme rarity, and seem to be mainly due to quotation or semi-
quotation. A few of these are:
1. The imperfect to express the present tense, or rather as a
gnomic tense, expressing events that take place throughout eternity.3
That these constructions were felt to be importations from Biblical
style is shown by the fact that they frequently have the archaic end-
ing -fm in the 3rd plural masc.
2. Imperfect with waw consecutive. 4 In the few instance where

1 We do not know if the speakers of Mishnaic Hebrew had any name to dis-
tinguish it from Biblical Hebrew. The designation "Language of the Rabbis" (J1iD'?
CI't:l::ln) appears for the first time in sayings of R. Yol:mnan (ab. 275 C.E.), i.e. after
the language had ceased to be spoken (Bab. 'Aboda Zarah 58b, Bab. f:Iullin 137b,
cf. Segal, Grammar, p. 3.
2 This is so obvious that it would be a vain effort to try to prove it (cf. also the
plain statement of Segal, Grammar, p. 5 top, and Diqduq, par. 3a). It is hard to
understand how Luzzatto could attach such weight to the lexical borrowings as to
maintain that the liturgical language "imita in generale l'Ebraismo biblico, tranne
pochissimi termini che ritenne dall' Ebraismo seriore" (Prolegomeni p. 100, a similar
statement in Elbogen, Gottesdienst, p. 296, perhaps based on Luzzatto.)
M. Szneider, Leshonenu vi 316, puts forward a theory that liturgical Hebrew is a
direct continuation of the idiom of the post-exilic books of the Canon, which was
kept in use as literary language, while Mishnaic Hebrew was the popular speech.
This is open to two grave objections: In the first place, the character of late Biblical
Hebrew is that of a Biblical base with Mishnaic influence in vocabulary and syn-
tax; that of the liturgical style is just the opposite, grammar and syntax being
Mishnaic and only the vocabulary enriched by Biblical elements. Secondly, if there
existed such a literary language (as distinct from a poetical style-variety), it would
be difficult to understand why the very same people who used it laid down their
halakhic and Midrashic utterances in pure colloquial. Such a procedure would be
quite unheard-of in the Orient.
3 Szneider op. cit., p. 317. The majority of imperfects in the liturgy, however,
is not of this type, but ordinary Mishnaic optatives, cr. Segal Grammar, par. 319.
4 E.g. C1it:l'?n1 in i1:li i1:li1~, Singer, p. 39, 10::1'1 in 1J'n1:l~ ml.lJ, ib. p. 43 (quotation
from Ps. cvi. 11), ~~1'1 in i1J1t:l~1 nt:l~, ib. p. 99, iD.lJn1 in C11m ~1i11, ib. 58 (Gaonic?).
 INTRODUCTION 9

this appears, it is quite isolated and does not differ in function from
the ordinary perfect tense.
3. The Biblical infinitive construct instead of the Mishnaic verbal
noun in all its syntactical functions,5 again without semantic differ-
entiation. Very rarely the Biblical construction of the gerund with
the preposition -be instead of a temporal clause is employed. 6
4. The masc. plural always ends in -'im, never -'in. This feature is,
of course, purely graphical, and may be due to the copyists or prayer-
book editors.
5. The compound prepositions and conjunctions are avoided. This
is largely a result of the absence of complicated logical correlations
in the prayers, but there are also deliberate changes from Mishnaic
usage. The most important is the use of'~ "because"7 side by side
with ~ (never ~ ':IElO, etc.). The reason for the adoption of the Biblical
word may have been the desire for a fuller sound, just as i~~ is
often used for the relative -~.
For illustrating the composition and character of the liturgical
vocabulary, it is necessary to analyse a complete piece. The prayer
chosen here is 'n '?~ nO~:I,8 which was composed in Tannaitic or
early Arnoraic times. 9
Apart from the large number of words that are common to both
Mishnaic and Biblical Hebrew, the piece contains quite a few that
occur only in Mishnaic; among them verbs: O:liEl "to sustain", PElOi1
"to do enough"; nouns: m'i~ "creatures", n~~ "praise", 10 t:l'i~~
"limbs"; and particles: ,,?~ (for ,'?) "if", ~'?~ (for t:l~ '~) "except".

5 E.g. T"~ C1,1)1tD n,l)::l "at the time of their crying unto Thee" in 1)'m::l~ nil,l).
Singer p. 43, ]1'~" l::l1tD::l 1)')',1) il)'rnm "and may our eyes be witness of Thy return
to Zion", in the Amidah, ib. p. 51, 1)~1::l1 1)n~~ '1r:ltD "guard our coming and going"
in 1)::l':::ltDil, ib. p. 100.
6 cr. Gesenius, par. 114d. Szneider (op. cit., p. 316) considers this a normal
construction of liturgical Hebrew, but it is so rare that only one example of it
occurs in the whole postbiblical material of the Daily Services (some of Szneider
examples do not belong here at all). One instance from non-liturgical poetry is nr:ltD1
tDtD P"~ tDEl) 1'''~ ~1::l::l "He is glad and rejoices when the soul of one righteous
comes before Him", from an Amoraic dirge, Bab. Mo'ed Qatan 25b. The one
instance in the Daily Service is quite un-Biblical: Til"~ 179 ]1'~" "b~:;J "when it
will be said to Zion, Thy Lord reigneth in 1)')',1) 1~'" Singer, p. 101 (adapted from
Isaiah lix. 7: ... ]1'~" "Q1~). Cr. also below par. 34d.
7 Quite current already in the Amidah.
8 Singer, pp. 125-6.
9 It is mentioned Bab. Berakhoth 5gb and Bab. Taanith 6b.
10 The root n::ltD occurs in Late Biblical Hebrew as an Aramaism, cf. Kautzsch,

Aramaismen, p. 87.
10 INTRODUCTION

Indeed, the liturgical language goes beyond ordinary Mishnaic

Hebrew in its borrowings from Aramaic. The word J?,:;l "to assign,
distribute", is restricted to the Aramaic parts of the Talmud, but
taken over here in a Hebraized form.
Greek and Persian loan-words seem to be entirely absent from
the earlier liturgy. This is quite in keeping with the tendencies of
poetical language everywhere. Purity, i.e. archaism is essential for tra-
ditional poetry.
There is an even larger proportion of words that occur only in Bib-
lical, never in Mishnaic texts. II These, however, are not the usual
words of Biblical prose vocabulary. With some exceptions, like ~~
"nose" (for CC!ln), they are such as are restricted to poetical passages
in the Bible. Such words are: ".l}?~Q "with the exception of", i~~
"to glorify", 12 iJ?~ "god", Jm "to lead" (for Mishnaic Ji1Ji1), ~~J
"exalted",13 ,~ "eternity".
Among the Biblical elements, an important proportion are those
found only very rarely in the Bible, i1P'~ "straits", ri.l}i1 "to fear"14
li.l} "to be equal", occur in the Canon only three times each. The
words C'~niQ "space"15 and mQ~.l}n "fullness of power"16 are hapax
legomena.
It is thus obvious that the liturgical style borrowed from Biblical
Hebrew only for the sake of the more solemn and poetical character
of the Biblical works. Hapax legomena had this character in a special
measure, and were therefore eagerly taken up. Since such words, once
introduced into the liturgy, soon lost their freshness, it became nec-
essary to search the most recondite portions of Biblical vocabulary.
The liturgical style is here the direct forerunner of Piyyut.17 Some
BH words are used in a striking way, e.g. 'J.l)~' i1Q "what is our
power to save ourselves?" (C'Q?'.l}i1 ?::l 1'~i).
On the other hand, owing to this very character, the Biblical borrow-
ings had little effect on the structure and general make-up of the lit-

II I rely on Jastrow's dictionary for Mishnaic occurrences. Some of the words

given here may appear in late Midrashic texts not noted by Jastrow.
12 Only in Isaiah. The word occurs in Amoraic Midrashim, into which it came
perhaps through the prayers.
13 The root ~iDJ is extinct in Mishnaic Hebrew, except in the meaning "to marry".
14 From the liturgy, the word penetrated into the Gaonic Midrash, Pir~e R. Eliezer.
15 The plural form only Hab. i. 6. The singular occurs four times.
16 Ps. lxviii. 36.

17 This disposes of Kenaani's statement, Leshonenu x 175, that the Paytanim were
the first to exploit the Biblical hapax legomena.
 INTRODUCTION 11

urgical style. It is nothing but pure Mishnaic Hebrew enriched by

borrowings from various sources, principally Biblical poetry. There is
no justification for considering it a third dialect of Hebrew, ranking
equally with Biblical and Mishnaic, or even to seek in it the origin of
the mediaeval literary style. 18 It is true that this early liturgical style
entered into the formation of the latter, but their relation is a good
deal more complex, as we shall try to show in the following chapters.
There seems to have existed in early Mishnaic times also a style
which attempted a more direct imitation of narrative Biblical Hebrew.
The only sample we possess of this style is the story of the origin
of Pharisaism in Bab. ~ddushin 66a,19 perhaps a fragment of an
old historical work. The vocabulary and the syntax of that piece is
purely Mishnaic, but a number of Biblical forms occur: consecutive
tenses and an infinitive absolute. There is no consistency in their
use, and they give the impression of serving merely as ornaments.
Interesting is the employment of a verbal noun instead of a gerund
in 1nim:I1 "and when he came back" (for 1irn:I), a hybrid construction
frequent in the later biblicising styles and in Spanish Hebrew. 20 This
style seems not to have survived a long time. 21 Possibly we have here
in fact nothing but the first timid attempts to use Mishnaic Hebrew
in writing, when these Biblical frills were still considered necessary
for giving the style literary respectability.22

3. The Piyyiit

At some point during the Amoraic period, I the semi-poetical style

of the earlier liturgy developed into the peculiar literary form known

18 Szneider, Leshonenu vi 313, 316f, 323 to end of article. If, of course, one is
prepared to consider the Hebrew of authors like Rashi, Maimonides, Judah Hadassi,
and the Tibbonids as one and the same "Literary Hebrew", with only slight styl-
istic differences, then it matters little if the liturgy is also added to the collection.
19 cr. Segal, Grammar, par. 156, Szneider op. cit., p. 319.
20 Similar pieces: I. The story of the Nazir 2. Yeb. XVI,7 end, the story of the
n'p'J1~: Cln'? n-r.:m Clnirn~1 Clnirn~1 1J'i~n n'~ [so in Babylonian Talmud in a
Mishnaic saying].
21 The possibility cannot be excluded that the style remained in use, and that

works like rosippon and Sepher ha-r ashar go back to this tradition rather than to
Saadianic classicist Hebrew. It would be strange, however, that no mention of such
works should have reached us.
22 Just as semi-educated Arabs intersperse their letters with "naJ:iwf" words like
lam (with the perfect!) and qad.
1 In view of the complete absence of any biographical information, no satisfactory
12 INTRODUCTION

as Piyyii!. 2 Basically, the Piyyii! is nothing but an extreme development

of most of the tendencies inherent in the earlier liturgy, replacing
the simplicity of the latter by conscious artistic methods. 3 In this, it
forms by no means an isolated phenomenon, but is parallelled in
many respects by contemporary Byzantine and Syriac4 religious
poetry.5
In their search for unusual and striking ways of expression, the
Paytanim revolutionised the entire Hebrew vocabulary. In their un-
bounded linguistic creativeness they contributed a great deal towards
fitting the Hebrew language for its later literary developments.
Their first means of enlarging the vocabulary consisted in search-
ing Biblical as well as Mishnaic literature for rare words and hapax
legomena. Their utilisation of the recesses of Biblical vocabulary was
a veritable raising of the dead. 6 But these words, by constant use,
became so common that they hardly form a particularly striking fea-
ture of the poems. The use of Biblical vocabulary in this way remained
as a heritage to Spanish poetry, Spanish ornate prose, and the Meli~ah
of the 18th and 19th centuries.

chronology of paytanic literature has yet been established. The date of IS.alir, the
best-known of all is put by various scholars into any century from the 5th to the
12th. Those of his predecessors, Yosi and Yannai, are even more uncertain. The
most probable time, is, however, the sixth to seventh century.
2 The only complete study of this literature is still Zunz's ~nagogale Poesie der
Juden, written in 1855, and republished, with some small additions, by A. Freimann
in 1920. In spite of its rather outdated attitude to Jewish literature as a whole, and
despite the wealth of material that has come to our notice since Zunz's time, it
will be hard even now to excel his treatment. Fortunately for our subject, Zunz
incorporated into this book extensive data on the Paytanic language. His analysis
of the character of the Paytanic innovations in vocabulary still stands, though much
new material has been added. Elbogen's remarks on the language of the Paytanim,
Gottesdienst, pp. 296-99, are a resume of Zunz. The article Cl'~1':li1 nj1:::1n'? by Davidson
in r11ii1'i1 'lli~ i (1926) pp. 187-95 adds nothing new.
The only Paytan whose work has been scientifically published in full is Yannai
('~j' '~1':l, ed. M. Zulay, Berlin 1938). On his language see Zulay, i11'tDi1 1pn'? 11:::1~
n'1::Jlli1 VI (1945) 161-248.
The best selection of Piyyu!im is in Brody and Wiener's Anthology. A complete
catalogue of all known poems is Davidson's monumental n'1::Jlli1 i11'tDi1 1~1~, 4
vols., New York 1924-33, and supplement, Cincinnati 1937-8.
3 Elbogen, op. cit., p. 281.
4 Cf. Duval, Litterature syriaque, p. 27, (speaking of the Syriac poets before 600 C.E.):
"Ils rechercherent les expressions rares ou artificielles, qu'ils affectaient de consid-
erer commes des archaismes propres a donner du relief aux images poetiques."
5 "Demnach kein Werk mit Namen genannte Dichter, zumal der piutischen Gat-
tung, iiber das Jahr 770 hinaufreicht", Zunz, Literatur und Geschichte der Juden, p. 26:
See summary in Schirmann, JQR 44 (1953) 127. Cf. also R. Edelmann, Bestimmung,
Heimat und Alter der synagogalen Poesie, Oriens Christianus III vo!. 7 (1932) 10-31.
6 cr. Kenaani, Lesh. x 27.
 INTRODUCTION 13

The syntax of the piyyutim of Palestinian and South-Italian origin

is strictly Mishnaic in all its principles. Un-Mishnaic biblical construc-
tions, apart from the gerund with prepositions/ are comparatively
rare and often originate in semi-quotations. The necessities of poetical
style, particularly the need for expressing a thought in as few words
as possible, are the cause that certain features of normal syntax
become much more frequent than in prose: e.g., the attributive par-
ticiple, the object without n~, certain emphatic forms of word order.
The main innovation of paytanic syntax, namely, the use of prepo-
sitions as conjunctions without interposing tD,8 is also in keeping with
this urge for brevity. The origin of this construction has not yet been
investigated. 9 Possibly it is an attempt to combine into one the
Mishnaic subordinate clause with final verb and the Biblical gerund
with preposition. Rhythmical reasons cannot have had much weight,
since almost in all instances a gerund construction would have had
the same rhythmic structure (number of syllables and place of stress)
as the final verb with preposition.
The vocabulary of the paytanim is also basically Mishnaic, with an
unlimited amount of Biblical borrowings. Wherever the Piyyiit uses
ordinary words, these are Mishnaic, while the Biblical words are in
the vast majority of cases selected from the more elevated and unusual
elements of the language. It is in this respect the exact opposite of
the language of Spanish poetry, which was founded on the Bible
and only occasionally and grudgingly admitted post-Biblical forma-
tions and roots. ID When Ibn Ezra II accuses the paytanim of having

7 Instances from Yosi b. Yosi, the oldest Paytan, are given by Szneider, ush. iii 27.
8 list in Zunz, pp. 380-82. Instances from the Spaniard Ibn Gabirol in Yellin's
article, ush. vii 243.
9 The use of Kl- before perfect is older than l$.alir, cf. Zunz, p. 121. Cf. on
this -J F1eischer, Kleine Schriften i 387 who compares Ir.:> (cf. M. Wolff, ZDMG 54.9).
In Ethiophic the preposition deara "after" is in poetry used as conj. e.g. Taliba
Taliban vs. 25.
10 Zunz leaves the question rather open, though he seems to imply that the lan-

guage of Piyylit was basically Biblical Hebrew: "The younger forms of Hebrew ...
possessed sufficient vitality to make it impossible and undesirable for the poets to
avoid it, although they wrote on the whole (hauptsachlich) Biblical Hebrew" (p. 118).
As far as vocabulary is concerned the Biblical words are certainly in a majority (as
Romance words are in English). Elbogen (op. cit., p. 296) puts Zunz's implication
in a much less cautious way, by stating that "the Paytanim did not always restrict
themselves to the linguistic material of the Bible", and even goes so far as to claim
that "the entire vocabulary of l$.alir's poetry could easily be traced back to Biblical
roots" (p. 318).
Renan, being a linguist and not under the influence of the Spanish attitude to
poetry, saw matters much more clearly. He goes, however, to the other extreme in
14 INTRODUCTION

admitted post-Biblical words too freely, he is applying~without any

justification~the standards of a completely different linguistic outlook.
His criticism is interesting as showing to what extent the attitude to
Hebrew had changed in the intervening time, largely owing to the
emergence of the revived Biblical style and its Spanish developments.
Already in our analysis of'n ?~ r1QIDJ we saw that the language
of the liturgy sometimes went beyond ordinary Mishnaic prose in its
borrowings from Aramaic. The Piyyii! has continued this tendency l2
and thus contributed a good deal to the syncretism of the two lan-
guages in the later stages of literary Hebrew.
The combined resources of the Biblical, Mishnaic, and Aramaic
vocabularies were, however, soon felt by the paytanim to be insufficient
for the voicing of their religious fervour. They therefore proceeded
to enlarge these by the creation of new words out of existing roots
and patterns. 13
It may seem strange that people should have undertaken to add to
the vocabulary of a dead language. 14
This is, however, a process which we find in other purely liter-
ary languages, such as mediaeval Latin, and Sanskrit. The creations
of the pay!anim, however strange they may seem to us, were con-
ceived in a much more purely Hebrew spirit than those of the
Spaniards. While the latter endeavoured to render expressions of
another language in Hebrew, the Piyyu! does not translate, but works
with associations and suggestions arising out of the Hebrew language
itself. Paytanic thought was guided by the language, not the lan-
guage subjected to the exigencies of a foreign way of thinking. 15

maintaining (Histoire, p. 162) that the Piyyutim were composed "a peu pres dans la
langue de la Mischna". This ignores completely the vast additions to the Mishnaic
basis that have made Paytanic Hebrew different enough to mislead Zunz.
II Cr. note 40 below.
12 List in Zunz, pp. 372-74.
13 Lists of their innovations in Zunz, p. 383; Kenaani, ush. x 25-27; YelJin, ush.,
vii 234-43. Davidson, Gin;:.e Schechter Ill. The dictionaries, including Ben-Yehudah
and Grazovski, are quite incomplete in this field. Grazovski does not even list all
the words found by Zunz. An attempt to achieve completeness is J. Kenaani's j,'?D
~"El'? 'J"~Jii'PJ'p, of which a prospectus appeared in Tel-Aviv 1930.
14 M. Zulay, Melila 1944, pp. 72ff. maintains that much of these innovations is

merely current material from the Palestinian (Hebrew-Aramaic) idiom of the time,
which was different from the Babylonian and is largely lost, owing to the literary
works written in it having passed through Babylonian channels. Cr. with this Gaster's
opinion on Samaritan Hebrew, The Samaritans cr. also Libermann in Clip 'm V. 177,
also Spiegel, Enc. Jud. ix. 819.
15 Of course, these "dead" languages were quite alive in the minds of those who
 INTRODUCTION 15

Moreover, in the mind of those people who conceived Hebrew in

its written form only, roots and patterns had to some extent a sep-
arate existence, and the combining of a pattern with a root was
something of the order of combining two separate words, just as in
English suffixes like -ness, -ize, or prefixes like re- can be attached to
words without restriction. It is remarkable how in literary Arabic
patterns alternate in one and the same root, and new words appear
without the writer ever becoming conscious of them.
In the same way it is not improbable that many of the new cre-
ations of the pay!anim were not intentional at all, but so to say mis-
quotations, lapses of memory. Forms l6 like ~i?i~ "magic", from ~tq~,
'O"'?S? 'joy" from the verb 'O'?.I], r"'?~ "terror" for m~~;l, '?~l "slander"
for m'?':;l"l, ~9j17 "longing" for Cl'El10~, lir, "purity" from l~r, are
words that might have arisen quite accidentally through exigencies
of rhythm or through simple inadvertency. It is certainly no acci-
dent that about half the newly-formed nouns are of the pattern '?S?~, 18
the simplest pattern in the language. Most of these are simplifications
of more complicated nouns. Had these been intentional creations,
with the idea of showing off the poet's powers of innovation,19 we
should expect that unusual patterns would have been preferred.
The nouns of the patterns ii,?'J)~, '?1.1]~, ii,?~~iJ, etc. which Zunz lists
as innovations,20 are no more to be considered as such than the use
of a perfect of a verb of which the imperfect only is to be found in
the Bible. In Mishnaic Hebrew, and therefore also in Pay!anic lan-
guage, these verbal nouns are equivalent to the infinitive,21 and can,
of course, be derived from any verb. The same applies to plurals of
nouns that in the Bible are always singular, and vice versa. 22
Even the striking biliteral nouns from weak roots, as '?iJ "light"

used them. Indeed, the less anxious respect is paid to classical models, the more
we can be sure that the language was really alive. The various periods of classicism
in Mediaeval Latin coincide with epochs of the emergence of nationalism and enthu-
siasm for the vernaculars. "Palestinian Piyyii!, in contrast to the epigonic character
of Spain, is the organic heir of Jewish antiquity." (Spiegel in Enc. Jud. 9.819).
16 Unless specially noted, all the examples in the following are taken from Zunz.
17 Kenaani, Lesh. x 25.
18 Zunz, p. 122; Yellin, Lesh. vii 223, finds that exacdy 50 per cent of the paytanic
words in B. Gabirol are of that pattern.
19 This is the view of Kenaani, op. cit., 24.
20 Op. cit., pp. 387-91, 393-96.
21 Segal, Grammar, par. 21 7.
22 Zunz, pp. 374-77.
16 INTRODUCTION

from '?'?i1, 23 iJ~, i1'J "splendour" from l1JJ, J)'t],i "salvation" from l1rJJ"
t"J1' "beauty" from i1~', need not always be products of a particular
playfulness. For the pay!an, the roots of these words were biliteraF4
and these formations were the natural and correct ones to his mind.
Of course, this does not mean that all the new paytanic nouns
under these classes were results of accident. Many of them were
probably quite deliberate. Moreover, in any other literary period,
when the readiness for innovations was absent, such accidental cre-
ations would soon have been removed by the criticism they aroused.
For the pay!an, even if he noticed them immediately, they were wel-
come additions to the Hebrew vocabulary, and once created, they
were handed on from one generation to another,25 and by their pres-
ence still increased the propensity for analogical slips.
Deliberate invention is probable in a case where a ~al noun is
derived from a verb in another Binyan: n~1 "disputation" from n~1ni1,
iiJt "warning" from ii1ri1, ::l?iJ "insult" from ::l'?l1i1. It will be seen
that these nouns, in contrast to many of the '?.p~-formations, are not
simply doubles of existing words but have a meaning which could
not adequately be expressed by the verbal nouns or any other words. 26
A strong likelihood of intention exists also in the case of rare pat-
terns being chosen, as '?'89 "pardon", 'ptZi "nourishment". But it is
difficult to judge which patterns were considered unusual by the
Pay!an himself, who seems to have restricted himself almost pedan-
tically to the forty most frequent ones. Only with one formative ele-
ment can we definitely assume intention, and that is the -on suffix,
as in 11'~::l "weeping" 11'1' "pain", l1'?"J "greatness" Y These words
were necessary, being abstracts of verbal ideas, and had a rhythmi-
cal and phonetical value for the poet. Indeed, the words in -on were

23'ii: '~ i1~ Job xxxi 26. The word is also used by Rashi (in his commentary
on the verse), Judah Halevi and al-I:Iarizi (cf. Ben-Yehudah, ii 1088).
24 Nathan ben JeI~iel of Rome (11th century) compiled his Talmudic dictionary
11i,l) on a strictly biliteral basis, treating e.g. :m, i1JJ, JJJ, '~JJ all under the head-
ing JJ (cf. Delitzsch, Poesie, p. 152). S. Spiegel, Enc. Jud. viii 819, rightly stresses
the importance of considering the grammatical theories of the time when judging
its language.
25 See the instructive list of fashion-words, Zunz, op. cit., pp. 423 ff

26 In Modern Hebrew the verbal noun has become a real abstract noun and
therefore the ideas of the first two are expressed by the verbal nouns 1J1:;)1 and
i1liJl~. The idea of J7D is expressed by another Paytanic formation, 1i :1'.I) , which
occurs in Genesis Rabba in the meaning of "the state of being insulted".
27 List in Zunz, pp. 397-402. The suffix and the pattern are, of course, Biblical,

cf. Gesenius, par. 85u, Brockelmann, i 389.

 INTRODUCTION 17

by contemporaries considered something peculiar to Piyyiit. In some

verses of the ninth century, obviously a parody on the Paytanic man-
ner, the Romance (or South-Italian Greek) wordsfumo andfurca are
provided with that ending. 28
The same considerations apply to the Paytanic treatment of verbs. 29
The Binyanim contain a stronger element of meaning than the noun-
patterns, and the need for a certain expression may often lead to
the accidental formation of a Binyan by analogy. This is a familiar
phenomenon in Modern Hebrew literature and colloquial. Binyan
innovations are also the only ones that can be found already in the
early liturgy. 30
Even the creation of new roots by incorporating norpheme ele-
ments into the original weak root is not at all surprising. Verbs like
r-r.vni1 "to cause to be dark" from i1El~n,3l ii;1l) "to offer" from i1i1rvn,32
?iT "to shine",33 om "to be fragrant" from Oh'J n'i,34 and 1~l)35 "to
be grieved" from i1'J~n, 36 or nouns like 1~9 "confusion" for i1:l1:lr.:l,
O~O "defeat" for i101:ln,37 are quite in keeping with phenomena in
Hebrew itself38 and the cognate languages. 39 Paytanic Hebrew only
extends these occasional and extraordinary processes until they become
almost grammatical rules.
All this does not detract from the fact that Paytanic Hebrew was
possessed of a strong urge for linguistic experiment and innovation.
The psychological factors of that attitude have not yet been satis-
factorily explained. The riches added to the Hebrew vocabulary by
the Paytanim are so vast that even their cataloguing has so far proved
too much for the lexicographers. Once the peculiar mystical exuber-
ance of the Paytanim, which necessitated the innovations, had ceased,

28 Chronicle qf A~imaa,:;, ed. Salzmann, p. 6, lines 4-5.

29 Lists in Zunz, pp. 41 Off.
30 E.g. :!'jllii "to create evening", Singer, p. 96.
31 Job xi 17.
32 I Sam ix 7.
33 cr. note 21 above.

34 The root of this is, of course, n1J.

3, Kenaani, Lesh. x 26.
36 Root i1J~. Occurs Is xxix 2 and Lam ii 5.

3) Both from Kenaani, op. cit., p. 25.

38 E.g. the Mishnaic ?'nm "to begin" from ii?nn (root ??n), and clin "to bring

an offering" from iir.:lnn.

39 E.g. Arabic takhidha "to take" for ittakhadha (root '-kh-dh) (Wright, Arabic Grammar,

i 77 A), Syriac luniiM. "groaning" from eltanna~ (root '-n-~) (Noldeke, Syriac Grammar,
p. 117, note 2), and many similar cases.
18 INTRODUCTION

not only did the creative process stop, but also the wealth of virtual
synonym, proved altogether too much for the needs of expression,
and most of them fell out of use very rapidly. Even Modern Hebrew
with its insatiable appetite for new words will probably never utilize
the whole of these creations.
The Piyyiit is the special property of the Palestinian orbit: Palestine,
Italy and Greece. 40 At first this included also Spain and N.W.-Europe.
As these two came under the influence of Babylonia, their attitude
to the Paytanic manner changed. In Spain, after it had produced
outstanding Paytanim like Ibn Abitur, Ibn Ghayyath, and Ibn Gabirol,
the Piyyut was abandoned in favour of the new classicist metrical
religious poetry.41 A bitter hostility to Piyyut arose, which found its
spokesman in Moses b. Ezra42 and Abraham b. EzraY Finally all

4{) (But after Saadyah there was also a school of Babylonian Paytanim, cf Davidson,

Ginze Schechter III 147).

41 Cf Zunz, p. 216, Elbogen, p. 339, and below chapter viii.
42 He regretted in his old age the poems which he had written in the Pa},!anic
manner (Shirath Israel, 156f.) Halkin p. 220 top: It is not certain that he means the
Paytanic style.
43 Comm. on Eccl. v I: " ... There are four unpleasant features in the Piyyutim
of R. Eliezer (sic) ha-I$.alir. Most of his poems consist of riddles and parables (fol-
lows detailed discussion of one line) ... Secondly, they are interspersed with Talmudic
vocabulary. It is well known that there are many words in the Talmud that are
not Hebrew. It has been said 'The language of the Bible is one thing and the lan-
guage of the Talmud another' (Bab. f.lullin 137b; the wording of the b. Ezra edi-
tions differs from that of our Talmud text). Then why trouble us with praying in
foreign tongues? ... Why should we not learn from the statutory prayers which are
entirely in elegant Hebrew? Why pray in Median, Persian, Edomite, and Ishmaelite?
Thirdly, even the genuine Hebrew words are employed with serious mistakes ... .
He merely thinks up words in order to impress his hearers with his cleverness ... .
The holy tongue is in the hands of R. Eliezer like a breached city without walls.
He turns masculines into feminines and vice versa. . .. There are in his poems
rhymes so poor that they go a-begging.... Some say: one does not reply to a lion
after his death. That is empty talk. God made all of us. The ancients were shaped
of clay just as we were, and the ear is the sole judge of words .... Fourthly, all
his poems are full of "Mid rash and Aggadah .... one should pray according to the
plain sense of Scripture, without introducing into one's prayers secrets and para-
bles, heretic opinions and ambiguities .... The Gaon R. Saadiah avoided these four
things in his Ba~ashoth, the like of which no one ever composed. They are in
pure Biblical Hebrew and correct grammar, and without any riddles or parables
or Midrash. There was one scholar in France who composed a B*-ashah begin-
ning 'i1 i1J'T~i1 'ir:l~ (author unknown, cf. Zunz Zur Geschichte und literatur der Juden,
p. 223), in which there are many such mistakes I can not mention even one in
thousand of the mistakes these Pa}'!anim make. My opinion is that one should not
introduce their poems into the liturgy but say the statutory prayers only."
The nervous tone of this tirade, and the entirely unjustified distinction between
 INTRODUCTION 19

the works of the older Paytanim disappeared from the prayer books
of the Spanish rite, to be replaced by poems in the Spanish taste. 44
In France and Germany, the external form of the Piyyut was pre-
served and in some cases further developed. But as the Jewries of
these countries began to display their own peculiar character after
the year 1000, the spirit of their Piyylit changed. From an imper-
sonal, didactic, contemplative poetry, it was transformed into an
expression of personal and communal feeling. Its linguistic charac-
ter also altered considerably. All the exuberant wealth of expression
disappeared, and the style became simple and sober. 45 Many of the
particular Paytanic fashion-words and phrases were abandoned. 46
Characteristically, it is the Biblical elements in vocabulary that are
given up, and European Piyylit comes to conform more and more to
the current late Midrashic prose language. 47 Aramaic elements enter,
and whole pieces are written in Talmudic Aramaic. After the end
of the 12th century, Piyyut was drawn into the general cultural
decline of Central European Jewry and finally ceased altogether.
It is interesting that a certain limited amount of interest in it was
revived by the J:Iasidic movement. Today there still exist J:Iasidim
who pride themselves on understanding the "Ma}:lzor words"; and
in some families this esoteric knowledge is handed down from father
to son. In the last few years, Modern Hebrew poets have begun to
take an interest in the resources of the Piyyut. Especially the poems
of Shimshon Melzer show very strongly its influence.

Saadiah and ~alir, are more understandable if one remembers the hard fight of
Ibn Ezra against the European mentality of which Midrash and PiyyU! were the
weakest points. He had to be much more careful where he attacked the literal faith
and the formalism of European Jewry (cf. Bemfeld in Enc. Jud. viii 330). He makes
frequent remarks about Pa)1anic "mistakes" in his commentaries and grammars.
"Saadiah, who is unequalled by any other Pa)1an in his fertility in consciously
invented words." M. Zulay, Melila 1944 p. 74.
44 Elbogen, op. cit. p. 339.
45 Zunz, pp. 125f. Elbogen, p. 331, ascribes the greater simplicity and "correct-
ness" of later European PiyyU! to the spread of grammatical studies after 1050. But
the only grammarian known until 1150 in NW-Europe was Mena\:lem Ben Saruk
who does not condemn PiyyU!. Moreover, the influence of grammarians would have
strengthened the Biblical element, but the opposite was the case.
46 Zunz, pp. 124f.

47 Zunz, p. 188.
20 INTRODUCTION

4. the Midrash

No comprehensive study of the vast field of Midrashic literature has ap-

peared, since Zunz's work, Die Gottesdienstlichen Vortriige der Juden, pub-
lished in 1832.1 Zunz's penetrating analysis enabled him not only to
establish a chronological system which in its main lines has been
generally accepted until now, but also to reconstruct some lost works.
His reconstruction of the Pesi~ta, among others, was fully borne out
by manuscripts subsequently discovered. Yet this book has been much
outdated by the wealth of material published since his time. The
only surveys of the whole literature since Zunz have been brief arti-
cles in the various encyclopaedias, 2 which in the nature of things
cannot provide more than lists of works with a few specimens.
In contrast to his work on the Piyyiit, Zunz did not devote any
space in his Vortriige to the investigation of linguistic features. Where
he employs linguistic criteria for dating works, 3 they almost invari-
ably refer to technical terms and expressions, and not to matters
belonging to the structure of the language. Some of his test-words
have since been shown to occur in much earlier times than he be-
lieved, and thus to be worthless for dating purposes. 4
The whole method established by Zunz of dating Midrashic works
has been called in question by modern scholars, notably Lieberman.
It is probable that the dates assumed now for many Arnoraic and
Gaonic Midrashim will be considerably revised. 5 If it is thus impos-
sible at the present time to undertake any investigation into the devel-
opment of Midrashic language, this task would be an extremely
difficult one even if the dates of the various works were established
beyond doubt. All Midrashim are of a composite character, and very
lengthy researches will be needed to discover to what extent each
Midrash reproduces the language of its sources verbatim or whether
it recasts its material in order to achieve a more homogeneous style.
So far even the basis for that work, scientific editing of all available
Midrashim, has hardly been begun.
It would be of a certain limited usefulness if at least the language

I Re-edited by N. Briill, Frankfurt a.M., 1892.

2 Heller in the Enc. Jud.; Theodor and Lauterbach in the Jew. Enc.; Strack,
Einleitung, pp. 195-226.
3 pp. 143(a), 235(a), 237, 252(e), 262(b), 274(c).

4 Liebermann in the preface to his edition of Debarim Rabba, p. xxif.

5 Ibid.
 INTRODUCTION 21

of Midrashic literature as a whole, without any attempt at stratification,

were treated from a grammatical and lexicographical point of view. No
separate dictionary of the Midrashim exists, though their vocabulary
has been incorporated into the Talmudic dictionaries and into those
of Ben-Yehudah and Grazovski. As for grammatical features, they have
been mentioned in passing in the various Mishnaic Hebrew gram-
mars, but none of these attempts to treat the features that distinguish
the language of the later Midrashin from that of the Mishnah proper.
In view of the role played by Midrashic Hebrew in forming the
literary language of the Middle Ages,6 the investigation of this stage
of the language, as far as it is possible now, may be considered the
most immediate requirement for the further elucidation of the his-
tory of Post-Biblical Hebrew.
The following remarks can therefore only sum up in the most
general terms the few generally recognised data on the language of
various Midrashic works. One thing which seems to be established
beyond any doubt is the division of this literature into three peri-
ods: the Tannaitic, the Amoraic, and the Gaonic. These three groups
are also clearly distinct in their general linguistic character, although
such characteristics are not yet sufficiently established to determine
the inclusion of a work in any of these three groups.
The Halakhic Midrashim of the Tannaitic period, notably Sifre,
Sifra, and the Mekhiltas, were written down at the time when the
Mishnah took form. Their language is, of course, very close to that
of Mishnah, Tosephta, and Baraithas. In some respects it may even
be more archaic. 7 On the other hand, a work with more pronounced
archaism of style within this group cannot for that reason alone be
considered older than one less archaic. The Seder 'Olam Rabba has
the most archaic style of all earlier Midrashin, 8 but is today placed
in the early Amoraic period. 9
The Amoraic period, from the fourth to the eighth century C.E.,
produced a large number of Midrashic works, of which Genesis
Rabba, Leviticus Rabba and the Pesikta de-Rab Kahana are com-
posed mainly in a very colloquial and idiomatic type of Aramaic. 10

6 See section ix.

7 E.g. the form '~I'~:::l for '~':::l "how", Segal, Grammar, par. 298.
8 Zunz, Vortriige, p. 85.
9 Strack, Einleitung, ch. xxi.
ID Helier in Enc. Jud. i 10 13. Dalman, Grammatik d. judo -pal Aramaisch, p. 23, how-

ever, seems to consider the language of these works artificial and influenced by the
22 INTRODUCTION

Aramaic had become the principal literary language of Near Eastern

Jewry, and dominated halakhic discussion as much as Midrashic lit-
erature. It is the basic idiom of the framework and of the Agadic
parts of both the Babylonian and the Palestinian Talmud. The lit-
erary forms employed were very close to the actual colloquial, as is
proved by their similarity to the contemporary non:Jewish idioms:
Mandaic in the case of the Babylonian dialect, Christian Palestinian
Aramaic in the case of the Galilaean. 11
Yet the use of Hebrew for literary purposes had by no means
ceased. The composition of liturgy and occasional poetical pieces
went on during the Amoraic period, and both Talmuds as well as
the above-mentioned Midrashim contain quite an appreciable amount
of Hebrew passages, quite apart from the pre-Amoraic material,
which is, of course, quoted in Hebrew. It appears that sermons were,
at least in some instances, delivered in HebrewY This would explain
the fact that the sermons of the Yelammedenu-Tanl:lUma I3 cycle are
almost entirely in Hebrew. 14

Targumim, without specifYing in what way it differs from that of the Palestinian
Talmud. What D. probably means is that the Midrashim, with their more ambi-
tious literary character, have drawn to some extent on the vocabulary of Targumic
Aramaic, just as the later Targumim draw on Biblical Aramaic for high-sounding
words (e.g. the word inniDlI: "rebellion" in Tg. Lam. i I, Tg. Cant. vi I, taken
from Ezra iv IS). n:H:::lJ, ~n':::lI':::lI':::lJ in TgJerem xl 5, TgJer. Deut. xxiii 24. Dalman
himself mentions both Pal. Talmud and Midrash as written in pure colloquial on
p. 41, op. cit.
11 Dalman, op. cit., p. 41.
12 In the 4th century C.E., a Babylonian audience was confused by the ambigu-
ity of 1J'?iD C1'O "water kept overnight" or "our water", used by R. Matna, B. Pes.
42a, cf. Segal, Grammar, p. 4, note 2, and Wijnkoop in JQR xv (1903), p. 29.
13 To the editions enumerated in Helier, p. 1010, Strack, ch. xvii, add now the

Debarim Rabba edited by Liebermann, and the passages from the ni1n i1i.l'?n C~l1P'?'
published by Mann, Bible, i, Hebrew part, pp. 270-346, as well as the new Midrash,
ibid. pp. 149-269.
14 Zunz, Vortrage, p. 236, places that part of the cycle which alone existed in print

in his time (Constantinople 1522 and reprints) into the early ninth century and
assumes that it was written in Southern Italy. It would thus belong to the Gaonic
period, so that its Hebrew language would not need any explanation. But today it
is generally assumed to belong to the Amoraic period. The criteria of a linguistic
order which Zunz uses, are either not decisive (cf. the criticism of Liebermann
quoted above) or, in the case of the ample use of foreign words, actually point to
the earlier date.
It is to be investigated whether the use of Hebrew was not somewhat more fre-
quent in Palestine than in Babylonia. E.g. the formula for fon '?1C!1::! is given by
Halakhoth Gedoloth, YJin, p. 134 in Aramaic, by Saadyah (Siddur ~"iDn, p. Y'?P)
first in Arabic then in Aramaic (=j'iD iT1 'in n'i.li~ j1iD'?::! ii.l~ C1~1). In Ye rush ami
Pesal)im 11.2 it is in Hebrew only. (Cf. Side man, Sinai VII.96). The Aramaic is
 INTRODUCTION 23

The Hebrew used in these Midrashim shows clear signs of lin-

guistic degeneration in grammar and syntax as compared with the
Tannaitic period. The language is often barbaric and influenced by
Aramaic idiom. The poverty of the authors's vocabulary is edged
out by a lavish use of foreign words,15 always a sure sign of low styl-
istic culture.
Mter 630 C.E. the Byzantine and Sassanid oppression, which had
largely been responsible for the decline ofJewish intellectual life, came
to an end, but the revival came only about 800, when the economic
security under the early Abbasids and the influence of nascent Arab
science and philosophy combined to bring about a recovery ofJewish
literature. One of the various branches of literature that benefitted
from this was the Midrash. A very large number of Midrashic works
was composed between 800 and 1200, the most important among
them Pes~ta Rabbathi, Tanna debe Eliyahu, Pirke R. Eliezer, Midrash
Tehillim, the Rabboth on the Megilloth, and Midrash Abkir. They
are of a much more literary character than the earlier Midrashim,
each with a definite plan and style of its own. Although they largely
use older material, they rewrite it so as to fit it into their own frame.
Their principal difference from the Amoraic Midrashim lies, however,
in their exclusive use of Hebrew and in the particular type of Hebrew
they employ.
By 800 Aramaic had probably been ousted by Arabic from collo-
quial use among the urban Jewish masses. It lingered on for some
time, in Babylonia perhaps longer than in Palestine. The weekly por-
tion and other Biblical passages were still translated into Targumic
Aramaic-which had, of course for a long time been a purely liter-
ary dialect, utterly different from the colloquial, particularly in Baby-
lonia. Some Targums were only written down at this time, such as
the Targum on Canticles with its Arabic names for the stones of the
Ephod. 16 Natronai Gaon (859-69) was the first to write responsa in
ArabicY Such a step would only have been taken under very great
pressure, and indeed, Aramaic continued to be employed in halakhic

the same as in HalakllOth Gedoloth; '?1C!l::J'? i1'::J ~jl)" ~'?1 '~miD'::J ~::l'~' ~"r:ln '?::l
~'~l)::l '}i171 '~miD'r:I.
15 Zunz, loc. cit., note (e).
16 Tg. Cant. v 14; 'aqzq, ai}mar, barqiin za'aftiin, kubtz jahar, tabiiz, jirilzq.;; murawwaq.
In the Yemenite text, ed. Melamed, these names have all, with the exception of
the first, been replaced by the Hebrew ones.
17 Mendelsohn in Jew. Enc. ix 190.
24 INTRODUCTION

literature throughout the Gaonic period. Not before Saadiah (d. 942)
was the Bible translated into Arabic. The fact that this was done,
and that Saadiah employed Arabic for his great work on the Jewish
religion, appears to prove that large masses of Jews did not know
either Hebrew or Aramaic sufficiendy well to read fluendy, yet in
the 11 th century still Hai Gaon (d. 1038) states that the country
folk, both Jewish and Gentile, spoke Aramaic quite fluendy.18 A story
told by Moses b. Ezra implies that Hai himself was most at home
in that language. 19 The same Hai wrote halakhic treatises in Arabic.
In Saadyah's time people had already to be admonished to use
Aramaic in ritual formulas, see the quotation above at note 14.
The immediate benefit of the disappearance of Aramaic went not
to Arabic but to Hebrew. The reasons for this will be discussed in
the next chapter. It may suffice here to mention that in the field of
Massora, too, Aramaic gave way to Hebrew about this time. 20 From
the Midrash it vanishes altogether. Stories that are told in Aramaic
in the older collections are translated into Hebrew in the Gaonic
works. 21
The Hebrew employed in the latter is also much purer than that
of the preceding period. It is a synthesis between the pure Mishnaic
Hebrew of the Tannaitic Midrashim, with their absence of foreign
words,22 and the liturgical language. They take over the Biblical
words of the latter, as well as typical turns of phrase. 23 Pesil.cta
Rabbathi, especially, has a poetical diction strongly reminiscent of
the Piyyu!im. 24

18 Hakedem, ii 82, quoted by Baron, iii 87. However, Maimonides, in the preface

to the MishTl£h Torah, states that in his time (he says: Cl'J'~Jii '1:l':::l!) people in Babylonia
did not understand the language of the Talmud without having been taught.
19 Shirath Israel, p. 104: Saadiah appears to him in a dream to give him the

"Nabataean" equivalent of an Arabic (or Hebrew?) word in one of his writings.

20 Levias in Jew. Enc. viii 365. From the ninth century onwards, too, Hebrew
begins to appear again with regularity on epigraphic monuments, cf. Cassuto, Enc.
Ita!. xiii 357.
21 There seem to have even been endeavours to substitute Hebrew for Aramaic
in legal formulas, cf. the Hebrew n'r.:l'?iI.J'i' ii:::l'n~ found in various copies, published
by Berliner (i' '?J) P'P) (1893); Gulak, miOiI.Jii i~'~, no. 31, pp. 35~6, Sidemann,
Sinai vii 98.
22 Midrash Abkir is particularly remarkable for its purism, cf. Strack, ch. xvii.,
par. 6 (e). Foreign words found in the older sources are ~ not always correctly ~
rendered by Hebrew ones, Krauss, Lehnwbrter, p. xxvi.
23 Zunz, Vortriige, p. 113, note (c) and p. 313, note (a).
24 Zunz, p. 313, note (a); Waxman, History, i 142. It is still to be investigated whether
 INrRODUCTION 25

There is no doubt that the purism of these works is connected

with the revival of interest in Biblical Hebrew, which will be discussed
in the next chapter. It is a rather remarkable proof of the general
linguistic conservatism of the Jews that the revived Biblical style never
invaded Midrashic literature proper,25 and that even the influence
of Piyyutic language is very small. The main reason for this is, how-
ever, the fact that the whole of Midrashic writing26 belonged to the
Palestinian orbit: Palestine, Italy, Greece and North-Western Europe.
Here, where Mishnaic Hebrew had once been a real colloquial, the
language was still alive and could not be replaced by an artificial crea-
tion of philologyY We shall see that the Biblical revival never really
took root within the Palestinian orbit. The new ideas of purism and
artistic style, which were current in Babylonian Jewry, were applied
in Palestine to the later form of Hebrew. It is impossible to decide
with any certainty whether to see in this an influence of the Babylonian
renaissance or a direct effect of Arabic culture on Palestinian Jewry.
However, the first seems more probable, as Syrian and Palestinian
Arab civilization never reached the heights of the culture of Baghdad,
and especially philological development was centered in Iraq, and it
also appears that the contact between Jews and Moslems was less
close in Palestine than in Babylonia.
The European countries certainly contributed to the literature of
the third period, though it is as yet not clear to what extent. 28 These
countries produced the only Midrash works of which we know the
authors: Bereshith Rabbathi by Moses ha-Darshan of Narbonne

any definitely paytanic words occur in these Midrashim. Zunz, !iJnag. Paesie, p. 381,
gives an example of Piyyulic syntax (~ before a finite verb) from the Pesi~ta Rabbathi.
25 The Sepher Hayashar and similar works can hardly be reckoned among Midrashic
literature, in spite of Zunz having included them in his Vartrage. If anything, they
are an attempt to break away from the Midrashic tradition in form and spirit as
well as in language.
26 I.e. Midrash as a literary genre. Aggadah and aggadic exegesis, of course, were
common heritage of the Palestinian and the Babylonian developments of Judaism.
Indeed as Philo proves, the aggadic outlook was typical of Judaism everywhere.
The only group who ever rebelled against it were the Spanish rationalists.
27 Delitzsch (Paifsie, pp. 145-157) was the first to distinguish clearly between the
Palestinian and the Babylonian cultural circle. He describes the attitude of the for-
mer to the Hebrew language as "a heritage which was at the disposal of the nation,
a national capital that could increase without limit" (p. 146).
28 Zunz connects Pesi~ta Rabbathi with Greece, Midrash Tehillim with Italy (pp.
244, 268). Numeri Rabba, which draws on Moses Darshan (Zunz, p. 259(d)), may
be of French or German origin.
26 INTRODUCTION

(11th century)29 and the Midrash Le~aQ Tob by Tobias ben Eliezer. 3o
More important still are the systematic Midrash collections which
are the specific contribution of European Jewry to this literature: 3l Y~t
Shime'oni,32 Yal~ut Makhiri 33 and Yal~ut Talmud Torah. 34 These,
in arranging the material in a strict exegetical order, form the tran-
sition from Midrash to the European type of Bible commentary.

5. Prose-literature in North- Western Europe

In two places did J ewries belonging to the Palestinian orbit come

under the influence of Babylonia: Spain and North-Western Europe,
i.e. Northern France, England, and Western Germany. The Provence
originally belonged to this area as well, but during the first half of the
twelfth century passed into the orbit of Spanish:Jewish civilization. '
European prose literature begins about the year 1000 C.E., with
Moses ha-Darshan of Narbonne, midrashic scholar and originator of
the midrashic Bible-commentary, and Gershom b. Judah of Mainz
960-1028, Talmudist and halakhic decisor. 2 During the next cen-
turies European literature proceeded on these two paths: BiblicaP
and Talmudic exegesis. In the second half of the 12th century, a

29 The Prague MS of this name, edited by C. Albeck, Jerusalem 1940, is only

an extract of Moses' work, and it is still to be investigated how far it reproduces
his language. The Commentary on Midrash Rabba ascribed to M. in Cowley's
Bodleian Catalogue (where our author is confused with Moses ben Joshua Vidal of
Narbonne, d. 1362, the Maimonides-commentator), however, does not exist at all.
The Catalogue reference is to two pages of Midrashic material in the Wilna 1878
ed. of Midrash Rabba, pp. 376-7, which may be a fragment of the Bereshith
Rabbathi. Epstein, r11'J1r.lip i. ch. xi., ascribes to Moses the authorship of Midrash
Tadshe.
30 Lived in Castoria in Bulgaria (Buber in his introduction to the Le~aQ. Tob,
Wilna 1880, p. 21). He mentions dates from 1090 to 1106 C.E.
31 Only the Midrash Hagadol was composed in Yemen (not before the 13th cen-

tury, cf. Helier, Enc. Jud. i. 1033.

32 Composed in the first half of the 13th century by Simeon ha-Darshan of
Francfort (Zunz, op. cit., p. 300). The Bodleian MS of 1307 C.E. (Neubauer-Cowley,
no. 2637) refers to the author as deceased.
33 By Makhir b. Abba-Mari, who lived perhaps in Provence, according to Poznanski
in the 14th century (cf. Seligsohn in Jew. Enc. viii 246).
34 Composed in 1331 by R. I:Ianan'el the Sicilian on the basis of 234 works.
Extracts ed. by Mann, Bible, i 270.
1 Dubnow, Weltgeschichte, iv 142.
2 Moses Darshan, teacher of Nathan of Rome, knew Arabic, cf. Rashi on Prov.
v 19.
3 A list of Bible commentators in Ginsburger, HUCA vii p. 439.
 INTRODUCTION 27

few works were written on other subjects: contemporary martyrologies,4

the travelogue of Pethal)iah of Regensburg (travelled 1175-90),5 and
the ethical sentences ofJudah I:Iasid (1150-1217).6
The language of this literature was that of the later Midrashim,
on which the European Jews depended for their lighter reading.
They took this language over as they found it, and wrote it with
fair skill during the 11 th century. The responsa of the early French
and Rhenish Rabbis 7 are in pure Midrashic Hebrew, much freer
from Aramaisms than the Mishnaic Hebrew written by the con-
temporary Babylonian scholars.8 The language of the commentators
betrays more influence of the Amoraic Midrashim. The style of the
remains of Moses ha-Darshan's9 and Menal)em b. I:Ielbo'slo works
is very much like that of Rashi, in whose commentary most of the
quotations are to be found, and it may be that Rashi rewrote his
sources, so that we would have no means of discovering anything
about the language of these early commentators. I I
The most fertile of these earlier writers was Solomon b. Isaac
(Rashi, 1040-1105). He was also an outstanding Paytan. 12 In his
Bible commentaries and in his responsa he uses a vigorous and lucid
Midrashic Hebrew. His usage has been carefully investigated by
J. Avinery,13 revealing to what extent he has integrated the language
of his sources. The large number of neologisms (some of which may
derive from his predecessors) shows that Hebrew was for him as
much a living language as for the Paytanim.

4 Hebriiische Berichte iiber die JudenverfOlgungen wiihrend der Kreuzziige, ed. A. Neubauer
and M. Stem, Berlin, 1892. nDn C''?iD1i" n~i,"1 IJ:::liDl4i mi'IJ i~O, 10i::lil.
5 Ed. J. Griinhut, Frankfurt a. M. 1905. Berakhya Naqdan (ab. 1200), ''?iDO
C''?.ll1iD, ed. Habermann, Jerusalem 1946; C'J::Il4i m:::l, Lapidarium (MS Bodl.).
6 C'i'On i~O, ed. J. Wistinetzki, Berlin 1891-3.
7 i'm'?1 n~i," 'O:::ln m::l1iDn, ed. J. Mueller, Vienna 1881.
B Cf. Dubnow, iv 148, who calls their style "cultured" (gepflegt).

9 A collection of quotations published by A. Epstein, Vienna 189!. Now H. Albeck

(ed.), 'n::li n'iDl4ii::l iDiiO (jerusalem, 1940).
10 A collection of quotations published by S. Poznanski, Warsaw 1904. e£ also
Bemstein: 1~'~ i1n::l '"iDi, Hadoar, Rashi number 23, Tebeth 1940 (Heb. d. 480).
II Moses Darshan was available to Rashi in writing only (?), c£ Rashi on Prav.
v. 19 i"ilO 'i 'i::Ji::J 'n'l4ii. The quotation there contains a reference to Arabic! Cf.
on the connection Albeck in 'n::Ji n'iDl4ii::J iDiiO p. 34.
12 Zunz, !iJnagogale Poesie, p. 18!.
13 '"iDi '?:::l'il. The first volume, all that has been published so far, contains a dic-

tionary of Rashi's innovations and a selective list of his conjunctions, prepositions,

and other words of interest.
28 INTRODUCTION

There must have been a fairly large literature in this type of He-
brew. It may have varied according to the countries in which it was
used. The Hebrew written in the Provence at Rashi's time seems to
have been more Mishnaic than Midrashic, if we assume that Abraham
bar f.liyya's style 14 corresponds closely to the current style of his
own time.
Rashi wrote his commentary on the Talmud in the Aramaic lan-
guage of that work itself, and his example was followed by the
Tosafists of the suceeding generations.
In the 12th century we notice a distinct decadence of Hebrew
style. It was probably due to the disorganisation ofJewish community
life and education by the Crusade massacres and other forms of
oppression. The style of the Crusade chronicles (written about 1150)
is clumsy and monotonous. PethaJ:llah's Hebrew 15 is harsh, long-winded,
and tortuous. 16 The style of the !:l'i'On i~O is not better.
In the 13th and 14th centuries European literature declined even
further. It produced occasionally works of a non-Talmudic charac-
ter, but these were few and far between. The polemics of Moses
Tako against Spanish rationalism 17 in the 13th century, and the anti-
Christian I'n~j of Yomtob Lippmann of Miihlhausen 18 may be men-
tioned here. The only non-halakhic literature that flourished were
the works on ethics and the pious way of life. 19

6. The Karaites

In the second half of the last century, it was fashionable in Jewish

scholarship to attribute every advance in Judaism during the ninth
and tenth centuries either to the Karaites themselves or the reaction
against them in Rabbanite circles. Among other things, it was believed
that their insistence on the study of the Bible 1 had forced the Rabbanite

14 See below chapter ix.

15 The report was probably not written by himself. A note in the text points to
it having been composed or redacted by Judah I:Iasid, the author of the Cl'Ton 1£)0.
It should be interesting to compare the usage of these two books.
16 Griinhut in the preface p. v.

17 Cl'On :m:l, published in 10nJ 1,"1~ vo!. iii.

18 First half of the 15th century. Printed Numberg 1644.

19 More than 30 works between 1050 and 1490, cf. Zunz, -?,ur Geschichte und

LiteratuT, i. 122-57.
I Summed up in Anan's dictum "search thoroughly in the Law" (~n"11~:l 1tZl'£)n
1'£)tZl).
 INTRODUCTION 29

Jews on their part to take up again the occupation with the half-
forgotten Sacred Books, and thus brought about the revival of Mas-
sorah and grammar. They were credited with a particularly strong
love of the Hebrew language, so much so that it was implied that
the revival of Biblical Hebrew prose, described in the next chapter
was more or less due to them. 2
The utilisation of Biblical vocabulary in the Piyyiit shows that the
"searching in the Law", at least from a linguistic point of view, had
been practised long before the advent of Karaism. If we review the
early literary productions of the sect, we can discover no particular
predilection for Hebrew, and no trace of any desire to return to the
Biblical form of the language. Anan's work has come down to us in
ordinary Babylonian Aramaic, 3 and there is no reason to assume
that this was not the original form.4 Aramaic was the idiom employed
by Karaite writers in their short "pre-Arabic" period. 5 We possess
indeed fragments of a Book of Precepts (Cl"J,n mtLll1 '0) by one
Nissi ben Noal:t, who is alleged to have lived in the 8th century,
and which is written in rhymed prose in an archaizing, piyyutic style.
In the preface 6 he states that he has composed it "in clear and ele-
gant language, in the words of the Hebrews and not in the speech
of Assyrians and Aramaeans, which is the shameful language of the
diaspora, for the sake of which the Hebrews have forgotten their
own tongue, and write their wisdom and thoughts in a barbaric lan-
guage/ so that they make mistakes in reading and are confused8 in
their exegesis9 and deviate from the literal meaning."lo However,
Frankl, lion internal evidence, has declared the work to be a plagiarism

2 The Massorete family of Ben Asher is claimed as Karaites, cf. P. Kahle, 17ze
Cairo Geni::;ah (Oxford 1959), p. 105 against M. Zucker, Tarbi::; xxvii (1957) 70, and
A. Dotan, Sinai 20 (1957) 280-312, 350-62.
3 m1~0i1 iOJO, ed. Schechter, Documents rif Jewish Sectaries ii, Cambridge 1910.
4 For the source of Anan's saying, see Harkavy, C'JiD' CJ C'iDin, (1958), I, 7
p. 33, cf. Zucker, Tarbi::; 27 (1958) 74 n. 71.
5 Karpeles, Gesch. d. jiid. Iiteratur, p. 411.
6 Pinsker, rWJ11:lip 'Cl1p'? pp. 37ff.
7 C'J'?ll piD'? (Is. xxxii 4), used by later writers to translate the Arabic lisiin al-
'qjam.
8 C'J10J, i.e. the Arabic ta~qyyari1.
9 C'J1inOJ. The word, meaning originally the interpretation of a dream, is not
used for "exegesis" before the 11th century.
10 Cltq~. This word occurs only in the Aramaic portions of the Talmud. In Hebrew
texts it does not occur before the II th century.
11 Ersch und Gruber, Realencyclopaedie, xxxiii 14, note 23; cf. Steinschneider Arab.
Literatur, p. 75.
30 INTRODUCTION

from Judah Hadassi (12th century), and stylistic evidence goes to

support his opinion. Rhymed prose was not used in Hebrew before
the tenth century, and the vocabulary, as well as the sentiments, of
this piece all point to a late age.
The Karaites seem to have taken to writing Arabic considerably
earlier than the Rabbanites. 12 The reason for this was evidently their
desire to reach the great mass of the people with their propaganda.
The earliest Karaite Arabic writer whom we know is Benjamin
Nihawandi (8-9th century), the real founder of Karaism as we know
it. 13 He wrote all his polemic and exegetic works in Arabic, only a
collection of laws by him14 is in Hebrew. His Hebrew style is ordi-
nary Mishnaic, remarkable only by a tendency to avoid technical
terms used by the Rabbanites. 15 Arabic is the sole medium of Karaite
polemics until the twelfth century. Then the newly developed cen-
tre in Constantinople, where Arabic was unknown, necessitated the
translation of the literature of the sect into Hebrew. The situation
is parallel to that existing in the Provence at the same period and
possibly the origin of the Constantinople school of translators is not
quite unconnected with the translating activity going on in the west.
The first Karaite to compose an original work in Hebrew was
Judah Hadassi, who wrote in 1148. His "'~~i1 '?~iV~ is written in a
rhymed prose that is a clumsy adaptation of the language of the
prayer-book, including the piyyiit. "Hadassi's diction has nearly all
the shortcomings of :&alir's language without any of its beauties".16
Indeed Hadassi's Hebrew gives ample proof that there existed no
Karaite tradition of Hebrew writing, and he had to create a style
out of nothing. In view of this, it is certainly characteristic of the
attitude of his sect to the Hebrew language that he made no attempt
to get back to pure Biblical Hebrew.

12 Steinschneider, Arabische Literatur, p. xxi, finds in Karaite literature "an objec-

tion to Arabic in principle". Surely such an objection would have restrained Nihawandi
from using Arabic a hundred years before Saadiah.
13 Nemoy, Karaite Anthology, p. 21: flourished in the 2nd quarter of the 9th cent.
(i.e. ca. 825-50!).
Best on him (ace. to Nemoy) Poznaitski in Yebr. Enz. V 483-90.
14 rO'J:l n~tvo, ed. by Firkovich at the end of Aaron b. Joseph's t:l'J'JEl 1n:lO,

Eupatoria 1834.
15 Mann, Texts & Studies 11, 12-13, shows that all Nihawandi's works were in
Hebrew, called by Nemoy, Karaite Anthology 22, "fluent & idiomatic". His style shows
distinct influence from DSS!
16 Frank!, MGl1J xxxi 11. Cr. also Skoss in Enc. Jud. vii 774, who calls his style
"forced and primitive".
 INfRODUCTION 31

The new Karaite Hebrew writing came, however, soon under the
influence of the Spanish-Provenc;al style. The model is unmistake-
able in the diction of Aaron b. Joseph (1250-1320),17 of whom we
know that he was conversant with contemporary Rabbinic litera-
ture. 18 He even takes over such typical features as the prefixed m.

7. The Revival if Biblical Hebrew Prose

In the latter half of the ninth century, a new Hebrew prose style
came into use. This was based on Biblical grammar and syntax, and
imitated the style of the Bible to the extent of employing the Massoretic
accents and verse divisions. I At the same time, it absorbed many
elements from Mishnaic and Midrashic Hebrew, as well as from the
innovations of the Paytanim.2
Until the discovery of the Genizah the only known samples of this
style were two fragments from the works of Saadiah Gaon (d. 942):
part of the introduction to his dictionary 1ii.:l~i1 iElO and five lines
from the text of his '1'?.:li1 iElO.3 Now we possess further fragments
from the latter work, 4 and from two other books of Saadiah, the
polemic against the heretic, I:Jiwi of Balkh, 5 and that against the
Palestinian Ben Me'ir. 6 Apart from Saadiah's writings, there is a col-
lection of heretical arguments 7 and a "scroll" of 10 12 C.E. dealing

17 Of his Bible commentary, the parts on the Pentateuch, the First Prophets, and
Isaiah, were printed at Eupatoria, 1834-5.
18 Simchoni in Enc. Jud. i 51 f.

I Unfortunately the editions of these texts reproduce only the pointing, but not

the accents. A study of the principles followed by the authors in accenting their
texts would perhaps throw some interesting light on the semantic value of Biblical
accentuation schemes.
2 Saadiah's writings show a clear development from Paytanic style (Palestinian?)
to pure Biblical in '1'Jil '0. Did he learn this in Iraq?
3 Both published by Harkavy, Studien und Mittheilungen V. I.
4 Ed. Schechter, JQ,R xiv 41-47 and Chapira, RE] lxix 1-14.
5 Ed. by Davidson, New York 1915, by Poznanski, Warsaw 1916, and by
Wertheimer, Jerusalem 1931. The name should probably be pronounced l:Iayyu.ye
('1'0) or l:Iiyyu.ye, i.e. l:Iayyim or l:Iiyya with the Persian diminutive ending as in
~ibiiya ("-i.~).
6 Cf. Malter, Saadia Gaon, pp. 409-19, and JQ,R iii (1912) 500-509. Not ~iD~
''?iDr.:I, which is a Piyyut-but in fact, so is the reply to l:Iiwi! Was Saadia's inno-
vation that he used Piyyut for polemical purposes-but then he uses a more BH
style in his polemic piyyutim than in his liturgical ones.
7 "Bible difficulties by aJew", ed. Schechter, JQ,R xiii (1901) 345-74, cf. Malter,
op. cit., pp. 386f.
32 INTRODUCTION

with the sufferings of the Egyptian community.8 Saadiah9 tells us

of a similar "scroll" concerning the persecutions in Kairouan,lo
which was written in the same style as his own "':1i1 ""E)O and pro-
vided with vowels and accents. A reminiscence of another work of
this type may be preserved in Hadassi's obscure reference to a ""E)O
...,'~' P omQ. II
If the list of extant works is not very impressive,12 it yet seems to
point to the existence of a much larger quantity and of a consider-
able public who read this style but were unfamiliar with the lan-
guage and technique of halakhic discussion. This can be the only
reason why Saadiah carried on his controversies in this idiom rather
than in Aramaic or in Mishnaic Hebrew, both of which were much
better equipped with terminology to deal with the matters under dis-
cussion. Saadiah stresses in his Egron l3 that it was intended to assist
in the composition of Hebrew prose and poetry. 14 The lost stylistic
chapters of his "'Ji1 ""E)O, too, were to give assistance of this kind.
We may, therefore, assume that there was a fairly widespread
movement and an established tradition of writing a Hebrew prose
of a Biblical type already before Saadiah's time and extending at
least into the first half of the 11 th century.15 This is the first con-
scious revival of any form of Hebrew, and therefore of great impor-
tance for elucidating the phenomena connected with later undertakings
of this kind. It also had an interest because it is the first beginning
of a development that led through Spanish Hebrew poetry and ornate
prose to the Haskalah revival and thus became a factor in the for-
mation of Modern Hebrew.
The scantiness of material makes it, however, very difficult to
investigate the language of this literature. Apart from a short char-
acterization by Bacher,16 the only study so far undertaken is one of
Saadiah's syntaxY

8Ed. by Mann, Jews in Egypt, ii 30-37.

9Harkavy, op. cit., p. 151, line 19.
ID Cf. Poznaitski, Harkavy-Festschrift, p. 176 and MaIter, JQR iii 489f.
11 1El~i1 "~iU~, Alphabet, 224, letters C)-El.
12 It is stilI an unsolved problem in what language I:Iiwi's original Book of two
hundred questions was composed. Davidson in his introduction (pp. 2 If.) adduces argu-
ments for assuming it to have been in Arabic, but these are by no means decisive.
13 Harkavy, p. 55.

14 The words employed are r1J1~nr::l and nn1nr::l.

15 Did the movement arise in Persia? Cf. I:Iiwi, Nihawandi!

16 Jew. Enc. x 583, based on the fragments ed. by Harkavy and Schechter.
17 Rabin, "Saadya Gaon's Hebrew prose style", in &sf!Ys on Samfya, Manchester 1943.
 INTRODUCTION 33

Although Saadiah's language endeavours to achieve a close approx-

imation to Biblical Hebrew, it does not imitate any definite Biblical
style, but rather constructs a completely new idiom out of the ele-
ments of Biblical Hebrew, with a preference for the rare and unusual
and for the features of poetical style, which are here imported into
prose. One cannot fail to notice the similarity of this procedure with
that of the Pay!anim, only that here the grammatical structure, too,
is taken from Biblical Hebrew. The close relationship of this prose
style to Piyyu! is further shown by the large percentage of post-
Biblical and purely pay!anic elements not only in vocabulary but
also in grammar and syntax. Thus Saadiah, while always employ-
ing gerundial constructions instead of dependent clauses,18 often has
the Mishnaic verbal noun instead of the correct Biblical gerund
forms, as in lrI:J''?iT in':JiDiT ~'? "he did not let him forget his having
gone" (for in:J,?).19 He also has the Mishnaic -~O with infinitive 20
instead of F~ with gerund, e.g. i''?~ J.)iiD'?Q ~~'? iin~ JiOJ ~'? "their
heart did not shrink back from appealing to Him."21 He also exhibits
some remarkable Arabisms, such as the use we- in a consecutive
sense, equal to the Arabicfil' al-jawiib,22 e.g. nniDJ ~'? iJ'Ji nQiD ~'? iTQ'?
"why did He not guard him, so that his offspring might not perish?",23
or the imperfect instead of the participle after the auxiliary verb,
e.g., iDiJ~ '?:J fii.IJn iT~ iTn"iT iT'iT iiD~ "by which thou wast wont to
terrorize every mortal."24 With these he already foreshadows later
Spanish usage. Moreover, we find in these passages tendencies which,
without being Arabisms, are closely parallel to features found in
Spanish-Provenc;al authors. Thus 'itn is used impersonally: '?:J P
iDiEl'? 'itn iJ'i~i' ni'?''?.IJ "in this manner it is proper to explain all
the actions of our Creator".25 The le-infinitive is employed with a
subject of its own: '?i~n~i ii~Q~ iTEl iiQ.IJ'? i'~nn ':J :"J~i "and you
still want him to stay here in trouble and distress?"26 iD' is employed
as copula: ~'iiiQ ~'~i iD' iT'?~ '?:J imiDJ.)~ "although He does all this,
many are rebelling" Y He even shares the typical tendency of Spanish

18 As already the early liturgical style and the PiyyU!.

19 /jiwi, Polemic, stanza 3.
20 Segal, Grammar, par. 346.
21 /jiwi, stanza 48.
22 Wright ii, par. 15d, Brockelmann, Grundriss ii, par. 302 i.
23 /jiwi, stanza 7.
24 RE] Ixviii 3, line 4.
25 /jiwi 64.
26 /jiwi 12.
27 /jiwi 13.
34 INTRODUCTION

Hebrew to place the object or adverb before the predicate, e.g.

l:l'ipE)~ ,In:JtoIi !:Il1i1 '?11 ..• r''?~ '?:I "every orator ... we appoint over
the people".28
In his vocabulary, Saadiah is equally catholic. He employs freely
Mishnaic words and even phrases, such as j~',? 1'? i1'i1 "you should
have said,"29 j~'m '?p "all the more SO".30 Indeed, he included the
Mishnaic vocabulary in his Egron;31 wishing his readers, of course, to
employ these words. He also uses many paytanic formations: i1~iE)J
"diaspora"32. '::1'?-'toIii'::1 "the thoughts of my heart", 33 ni'O' "original
form" (~i),34 i1n~E) for n'~E) "thou spakest".35 He also invented his
own words, e.g. j~toIi'? n'ClID "thou hast spoken foolishly", 36 j'il1i1 "to
withhold"37 n1:1i1 "speech". 38
The close contact of this prose style with Piyyii! is shown not only
by the use of paytanic stanza-forms in the ljiwi-Polemic 39 but also by
the fact that Saadiah in the preface of the Egron40 holds up the early
Pay!anim, Yosi ben Yosi, Yannai, ~alir, Joshua, and Phineas as
models of composition. 41 The excellence of later poets is measured
in terms of their similarity to the ancients, just as an Arab literary
critic would use the pre-Islamic poets as a standard.
Thus it is clear that the new prose style was not intended to be
merely a slavish imitation of the Bible. There are some important
deviations from Biblical prose-usageY Yet it is obviously much closer

28 JO",R xiv 45, line 22. This is, however, found already in the liturgy, cf. the
treatment of this feature below, p. 134.
29 /jiwi 7.
30 /jiwi 13 and 27.

31 Harkavy, op. cit., pp. 29 and 69.

32 Harkavy, p. 54, line 10.
33 lb. 57, line 3.
34 lb. 57, line 13.

35 /jiwi, stanza 23, cf. Zunz, Synag. Poesie, p. 380.

36 /jiwi, stanza 31.
37 /jiwi 11. The word is employed in Spanish-Proven~al texts and by Karaite
writers, cf. Ben-Yehudah, 1hesauTUS, pp. 43-48.
38 Harkavy, op. cit., p. 55, line I. Used in this sense by Moses Chiquitilla in his

translation of I:!ayyiij's /juriij aI-un.

39 The four-line stanza is the most usual form of the older Selii).oth, cf. Zunz,

Synag. POifsie, p. 91.

40 Harkavy, p. 5 I.

41 A number of piyyiitim by Phineas ha-Kohen have been published recendy by

A. Mamorstein, Ha-Zopheh, vols. v-vi, and M. Zulay, Mitteilungen des Forschungsinstituts

for hebrdische Dichtung, pp. 150-74.
42 The consecutive tenses, for example, are used infrequendy and irregularly. The
 INTRODUCTION 35

to Biblical Hebrew than any PiyyU!. It must have been the result of
a deliberate revision of the latter.
At present one can do no more than offer some suggestions as to
the causes of this revival. The first outstanding fact is that it belongs
entirely to the Babylonian orbit of Jewish civilization. 43 Here neither
Piyyu! nor Midrash had ever found a home. The literary language
of its Jewry had been purely Aramaic, and there was no tradition
of Mishnaic Hebrew writing as in Palestine. The movement was,
therefore, not so much one of substituting Biblical for Mishnaic
Hebrew as the introduction of Hebrew into literary prose.
One reason for this renaissance may have been the claims made
by the Moslems for the holiness and universality of Arabic. These
claims caused the Jews to assert the value of their own ancient lan-
guage. This would of course be done best by stressing its oldest and
most revered form, Biblical Hebrew. 44
It may also be the successes of Arabic philology and the great
popularity of the newly developed ornate prose which spurred Jewish
writers to undertake the "purification" of their own language and to
prove that Hebrew was capable of the same elegance (~~, mn~)45
as Arabic. 46 It is perhaps not insignificant that some, at any rate, of
our texts come from that interesting circle of sceptics, who, with
their leanings towards Zoroastrian and Mu'tazilite theories, seem to
have been an assimilationist movement.
Their exaggerated purism may have come out of their opposition
to talmudic Judaism, just as the purism of the Haskalah writers was
an expression of their dislike of mediaeval Judaism. 47
Finally, it is possible to assume a purely linguistic reason for the
origin of this style. The Palestinian Piyyu! seems to have penetrated

, ... 'i1', construction is not employed by Saadiah. Both are quite correctly used in
the Egyptian scroll, however.
43 Egypt appears to have been under the influence of both orbits. Only when
contact with Babylonia had been broken off by the advent of Fa!imid rule, did
Egypt turn entirely to the Palestinian centre.
44 Baron, Social and religious history, i 353.
45 Saadiah in Harkavy, op. cit., p. 45, line 6: "a book from which they may
learn elegant modes of expression".
46 Roth, Short History, p. 170.
47 It is possible that the "Zadokite" fragment belongs to the same time. Its lan-

guage is rather too anxiously Biblical for Mishnaic times. One should expect some-
thing more like Ben Sira. Its purism would be accounted for by the same anti-Talmudic
attitude. However, at the present stage of our knowledge, linguistic arguments can
bear but little weight in dating any book.
36 INTRODUCTION

into Babylonia only at that time,48 perhaps together with the puristic
Midrash style, and as a linguistic tendency often gains momentum by
being transferred from one area to another, the principle of Biblical
borrowing was brought to its logical conclusion on the new soil.
This view is reinforced by the fact that at approximately the same
time paytanic features began also to enter the prose style of Palestine
to a much larger degree than in the late Midrashim. In the letters 49
and responsa50 of the Palestinian Gaonim of the tenth and eleventh
centuries, we find a style which, while grammatically and syntacti-
cally purely Mishnaic, is in vocabulary not too dissimilar from the
Babylonian Renaissance style.
As an example we may quote the responsum of Solomon ben
Judah Gaon (1027-51).51 The legal discussion is in Aramaic, but the
introductory part, containing the greetings and the statement of facts,
is in a high-flown, biblicizing Hebrew strongly reminiscent of the
Prayer-book. The short fragment contains, among others, the fol-
lowing paytanic creations: '?J1' "ability", 1':Jni1 "to cause to come
together",52 11n1J "force, compulsion", i1Jmn "doubt", n11j~Q "bun-
dle".53 It also shows points of contact with late Midrashic language,
as 1'J .!l~Qni1 "come between".54 Again, we know this epistolary litera-
ture only from scanty remains preserved mainly through the Genizah,
and we can for the present recognize no more than its outline and
general tendencies.
Its home was Palestine and Egypt, but it seems to have spread from
there soon after its beginnings to Italy. In Southern Italy was com-
posed in the 11 th century the only prose work in pure PiyyD.! style
that we possess, the Chronicle of Al).imaa~.55 There we find Paytanic

48 Saadiah, the Egyptian, appears to have been the first Paytan in Babylonia.
For a bibliography of his liturgical poetry, see Malter, Saadia Gaon, pp. 329-39.
"Saadiah ... who talks in his liturgical compositions both the most flowing and the
most heavy language, here a worshipper and there a Pay!an, but never really a
poet" (Zunz, Literaturgesch, p. 93).
49 Many in the second volume of Mann's Jews in Egypt. Cf. ib. p. 9: "We see
how the Pay!anic phraseology pervaded the style of ordinary private letters."
50 For publications up to 1933, see list in Assaf, Gaonica, p. 90.
51 lb. 90-95.

52 i1i':Jn~ in Job xvi 4 means "to speak elegandy".

53 cr. Saadiah's n1'J~r::l, Zunz, fiynag. Poesie, p. 407.
54 Cant. Rabba, 1. 66, on Cant. iv I.
55 Ed. Neubauer, Mediaeval Jewish Chronicles, ii 111-32, and Salzman, New York
1924.
 INTRODUCTION 37

peculiarities like biliteral verb-forms, prepOSItIOns as conjunctions,

construct noun-forms instead of absolute ones,55 and others. The voca-
bulary is Paytanic throughout.
However, some time before Al:limaa~, another type of classicist
prose style had arisen in ltaly.57 This style modelled itself even more
closely upon the Bible than that of Saadiah, but instead of the poet-
ical, it took the historical books as its guide. It achieves thus a much
simpler narrative style. On the other hand, it abandons the imita-
tion of Biblical verse structure and produces a flowing historical nar-
rative of a European type. European influence is also umistakable
in its romantic and heroic approach to Jewish history. The admix-
ture of Mishnaic and Paytanic elements is so small that it can be
assumed to be unintentionaP8
We possess only historical (or rather pseudo-historical) works in
this style: 59 Yosippon, which was composed in Italy about the middle
of the lOth century, 50 and the Sepher ha-Yashar, of the same date and
origin. 51 Probably from the same century or a little later dates the
mUQ ~tU Cl'Q'i1 '"1::1i. 52
When metrical poetry was first introduced in the East,53 probably
in the beginning of the tenth century, it employed a style which hel9
the balance between Saadiah's prose and genuine Piyyur.54 This style

56 ntDi~:::l ii'? iO~l (for iitD~'?) ntD~'? ~iP "he called the woman and spoke to her
the following words", Salzman, p. 4; ''?':::ltD C1P"'? iO., ''''O:::l 10tD "his name was
Basilios, he arose to make paths crooked", ib., p. 6. Cf. n(lj~ "5l~ Ps. lviii 9. A list
of such cases in the Bible: Sperber, JBL lxii 216ff.
57 It is possible that the same development had taken place in the East during

the 70 years between Saacliah's death and the "Egyptian scroll". The language of
the latter is very close to that of the Italian historical works, but keeps up the con-
vention of accents and verses.
58 A list of such elements in Yosippon, Zunz, Vortrdge, p. 148 (c). The philo-
sophical terms ib. note (d) are certainly later than the date of the book, and must
have come in through later copyists.
59 It is possible, but improbable, that the use of Biblical Hebrew for historical

works derives from a tradition which was quite independent of the revival discussed
in this section, cf. I1, note 21.
60 Cassuto in Enc. Jud. ix 425, Zunz, Vortrdge, pp. 151f.
61 Ochser in Jew. Enc. xii 589. Zunz, op. cit., p. 156, decides for Spain and the

12th century.
62 Zunz, op. cit., p. 145. It is older than Exodus Rabba.
63 The view that Dunash ben Labra! invented the Hebrew system of metre
(Bacher in Jew. Enc. v 11) is now recognized to be a legend (cf. Wilensky in Enc.
Jud. vi 117). Whether the Jerusalem Karaite of the 9th century Meborakh ben
Nathan invented it (Gross, Menachem ben Saruk, p. 16, and Pinsker, nl'Jl0ip 'Olp'?
pp. 64, 139) is, however, rather doubtful, too.
64 Many examples in Mann's Jews in Egypt, ii. For example, a poem of about
38 INTRODUCTION

was also employed by Dunash ben Labrat (ca. 910-80), who intro-
duced metrical poetry into Spain. 65 It was only there that it received
its peculiar classicist character, as we shall see in the next chapter.
Saadianic prose was represented in Spain by Menahem ben Saru~
(ca. 9lO-70). His m:Jnr:l 66 and the letter to the King of the Khazars,
which he composed for I:Iisdai ibn Shaprut,67 are written in a prose
simpler and more Biblical than Saadiah's, yet containing a good deal
of Paytanic material. The very fact that MenaJ:tem composed his dic-
tionary points to the existence of a Hebrew-writing public in Spain.
We find in the next few generations private letters written in Hebrew. 68
But they and the ornate-prose compositions of Samuel ha-Nagid
(982-1055) are already in the new classicist Spanish style. The Spanish
and Proven<;al ornate-prose works of the 13th century are written in
a language that is derived from Spanish poetry, and thus only indi-
rectly connected with the Saadianic style.
In Christian Spain, however, Saadianic prose survived much longer.
Towards the middle of the 12th century, Abraham b. Da'ud of
Toledo (1110-1180) wrote his three historical works 69 in a much
watered-down prose of the paytanic classicist type,70 with a strong
influence of late Mishnaic Hebrew, such as his contemporaries Abra-

the year 1000 (ib. 11-13) contains Pay~anic innovations like p~~ "signing", PI 'joy",
P~?ll, notp id., n:~.!1 "reply", ';n~ "arrived", n:;J9 "the noble one" for "Israel" and
constructions like 1nijn::) ... ?:lilll 'JJ "when the Bene A. were put to flight". The
word '1JO is also employed by Yannai and ~alir. Habermann, Leshonenu iv (1932)
186, investigates its meaning and from the occurrences quoted concludes that it
means "important, great" and in the plural "many". It is, of course, a Hebraized
form of the Aramaic 'JO "great". For Eastern poetry cf. J. Werfel: ?iV 1'C!l1'El1 1'i'iV
plllJ 'IIIn Ji, Sinai I (1938) 592-676.
6.\ A poem by him in Mann, op. cit., ii 21-3 contains the following Mishnaic
words: nl''?JiO "pearls", OiJ "to learn", nfl "to entide" ~J[nl 1J?111 "the world to
come". Pananic innovations: n?iVn "to send", '1?1l "exalted", 1l~1pO ~C!l1JO "per-
fumed with stacte (Exodus xxx 34) and cassia", 10iJ "scholar", '01 "bloody" (Mann's
emendation is unnecessary and syntactically difficult), and others. Other poems by
Dunash were edited by D. Kahana, Warsaw 1904.
66 Ed. Filipowski, London 1854.
67 Text with most 'il1::l-editions. The authorship is fixed by the acrostich of the
dedicatory poem, cf. Gross, Menachem, p. 44, Brutzkus in Enc. Jud. v 348f.
68 E.g. the letters of Judah ha-Levi, in his Drwan (ed. Brody 1894) vo!. i, pp.
207-25.
69 I. History of the Israelite kings in the period of the second temple; 2. History of
Rome; 3. History of talmudic studies (n?Jpn iElO). All three printed Mantua 1516
and often. No. 3 was critically edited by Neubauer, Mediaeval Jewish Chronicles, i 47-84.
70 Elbogen, Guttmann-Festschrift, p. 203, calls his language "elevated and beau-
tiful, modelled on the Bible without becoming heavy or euphuistic".
 INTRODUCTION 39

ham bar !:Iiyya and Abraham ibn Ezra were using. There are also
many Arabisms. The choice of this type of Hebrew is obviously due
to his model and source, Yosippon. Where he quotes Yosippon, how-
ever, he rewrites him in his own style. 71
B. Dii'iid was the last author in Spain to employ this type of He-
brew. About the same time the last books in paytanic Hebrew were
composed in Italy by Menal).em ben Solomon, who wrote about
1140.72 His style is more Piyyutic than that of Menal).em ben Saruq.
It abounds in Paytanic words and heaping of attributes aud verbs
in the paytanic manner. 73 But when he wrote his philological work,
Ibn Ezra's activity in Italy had already begun. As Menal).em's anti-
quated Saru1.cian grammar was replaced by the more modern sys-
tem of !:Iayyiij, so was his tenth century prose style by the new
idiom created by the Spanish exiles.
In N.W.-Europe, the revived Biblical style seems to have found
no echo. Menal).em's work was known and highly esteemed there 74
and the controversy between his and Dunash's pupils still occupied
the minds of European Jewry when both their systems had long been
superseded in Spain. Jacob ben Meir Tarn of Rameru (1100-1171)
still thought it worth while to write a work in defence of Menal).em
against Dunash. But Menal).em was only esteemed as a help to exe-
gesis, not as a guide to Hebrew composition.

8. Spanish Hebrew Poetry

The Hebrew language was cultivated in Moslem Spain between 1000

and 1200 C.E. almost exclusively in the one literary sphere of met-
rical poetry. Although this genre had originated within the paytanic-
classicist literature of Babylonia and Egypt, and used its language,
it underwent a profound transformation very soon after its introduc-
tion to Spain. It became increasingly Biblical and narrowly classicist,
until it not only was quite unlike its parent idiom, but its representatives
began to urge a literary campaign against everything connected with

71 Elbogen, loc. cit.

72 :11C!l '?:liD iDiiQ on the Pentateuch, and ln~ p~, a grammar and dictionary. C[
Bacher in Graetz-Jubelschrift pp. 84-115.
73 Bacher, op. cit., p. 110.
74 Ginsburger, HUCA vii 443, Bacher, Jew. Enc. viii 470, Dubnow, Weltgeschichte,

iv 230.
40 INTRODUCTION

the Pay!anic manner of writing the Hebrew language. To understand

these changes, we must consider the peculiar situation of Hebrew
in Spain and the influences bearing upon the linguistic attitude of
its Jews.
Serious Hebrew prose ceased in Spain shortly after the year 1000,
when the controversy between the disciples of Mena}:lem and Dunash
died down. All the rich intellectual activity of Spanish Jewry in its
Golden Age used Arabic as its medium.
The reasons for this are obvious. The sciences and ideas that
formed the material of this great creative effort were imported. They
were not the property of the Arabs, but had been taken over by
them from the Greeks. It was, however, in an Arab linguistic garb
that they reached the Jews. These attained to the new ideas only
by virtue of their more or less complete Arabicization. I Both for the
acquisition of science and for success in public life, a complete Arabic
education was indispensable. The wide course of studies required for
it left little leisure for Hebrew studies. A knowledge of Hebrew,
though hard to obtain, was of little use in worldly life, and in that
age of prosperity was only sought after by those who were wealthy
enough to have the necessary leisure or idealistic enough to dispense
with success. It is therefore little wonder that the standard of Hebrew
among the Jewish intelligentsia was low, as we learn from the inces-
sant plaints of the grammarians. 2 Even in the East, Saadiah had felt
impelled to translate the Bible into Arabic 3 and to compose in that

1 f.lisdai b. Shaprut, in the middle of the tenth century, is said to have obtained
his high office on account of his beautiful Arabic handwriting. Even if this is only
a legend, it shows that there was nothing strange in the idea of a Jew spending
years on learning Arabic calligraphy. Various errors in the Oxford MS of Moses
b. Ezra's Kitiib al-mudhiikara suggest that the author wrote his work in Arabic char-
acters (Shirath Israel, Introduction pp. 26f). Among Eastern Jews, the use of the
Arabic alphabet for writing Arabic was extemely rare. Likewise, a Jewish Arabic
poet like Abu Sahl al-Isra'rIr would have been impossible in the East.
2 Dukes, Ehrensaulen p. 23; Sachs, Religiose Poesie p. 221; Steinschneider, Arab.
Literatur p. xxxiv; Friedlander, in Moses ben Maimon i 423. In the 13th century still,
Mordecai KimJ.!i writes: "in our time most of our sons and daughters speak Romance
and Arabic and Greek, and every language under the sun but the vast majority
cannot speak a word of Hebrew. How then can we expect women to understand
words used in the Gemara?" (Neubauer in RE] xii 82).
3 Saadiah's translation forms only part of his commentary! Saadiah never seems
to have translated the whole Bible. Indeed, perhaps not even the whole Pentateuch,
cf. the list of translators from which the final edition was put together in Mann,
Jews in Egypt, II, 310 (wrongly taken by Mann as commentators). See also Mas'udi,
Tanbfh p. 112f. On early Jewish-Arabic Bible translations, cf. Gehmann in the JBL
44, 327.
 INTRODUCTION 41

language his Kztiib al-amiiniit wal-i'tiqiidiit,4 which was destined for the
general public. In Spain, Abo Sa'i"d Faraj Ibn I:Iisdai in the 12th
century "was among those who translated the most important works
of Hebrew literature into Arabic",5 presumably for the use of Jews
who took an interest in that literature. Yosippon was translated into
Arabic (thence into Ethiopic), cf. Wellhausen, Der arabische Josippus.
The demand for the Arabic Bible was so great that Saadiah's trans-
lation was not sufficient and improved versions had to be made. 6
Not only the ignorance of prospective readers prevented the use
of Hebrew for scientific and philosophical literature. Arabic possessed
a wide range of terminology and syntactical adaptability to deal with
these subjects. The revived Biblical prose of Saadiah's and Menal:tem's
time, on the other hand, was still an undeveloped idiom, with all the
disadvantages of artificiality and deliberate purism. One feels in
Saadiah's Hebrew style the difficulties he must have experienced in
expressing his comparatively simple thoughts without violating the
rules of Biblical syntax. Had the Hebrew used in Spain at the out-
set been of the late Midrashic type, like that of N.W.-Europe, they
might have found it more fitted for the complicated lines of thinking
imposed by the new learning. 7 However, even then it is doubtful
whether the Spanish writers would have gone to the trouble of adapt-
ing Hebrew when the whole immense wealth of Arabic was at hand.
It also appears that throughout the earlier Middle Ages the lan-
guage of prose literature was not viewed under any nationalist aspect,
quite in contrast to the language of poetry. Prose was still too new,
and too difficult, for most peoples, and even those who already pos-
sessed a prose literature submitted quite willingly to the claims of a
language of superior culture. s

4 mDi1 mJ1r.:l~ i5:l0 in the Hebrew translation.

S Shirath Israel p. 76.
6 The Talmudist Isaac b. Ghayyath translated Ecclesiastes (Steinschneider, Arabische
Literatur p. 136). A list of Jews who translated the Bible into Arabic is in Mas'ildf,
Tanhfh, ed. De Goeje, p. 112f.
7 Terminology would, of couse, have been lacking in this case, too, but short-
comings of vocabulary are in such cases easily overcome, as the beginnings of
Hebrew scientific prose show.
8 The only exception to this were the Arabs. This was possible because of the
great diversity of languages with which Arabic came in contact, and the fact that
all of them were just then on the decline. We see in Himyarite Yemen and in the
Nabataean state what happened when Arabic was confronted with another lan-
guage without having the backing of religion and a supra-national state.
42 INTRODUCTION

We therefore find that not only the minor scientific writers of the
11 th and 12th centuries employed Arabic, but even the great mas-
ters of Hebrew poetry and ornate prose: Solomon b. Gabirol, Ba}:tya
ibn Pa~udah, Moses ibn Ezra, Judah ha-Levi, and Maimonides. The
last-named closed the line of great writers who addressed Jews in
Arabic. We shall be able to observe in his own productions the tran-
sition to the Hebrew age of Jewish literature.
It is necessary here to mention the real and apparent exceptions
to this rule. The Talmudists, who were few and unimportant in these
centuries, seem to have continued to employ in their more techni-
cal works the traditional Gaonic mixture of Mishnaic Hebrew and
Aramaic. Isaac b. Ghayyath 9 of Lucena (d. 1089) wrote various trea-
tises in that style. These consist mainly of quotations from Talmudic
and Gaonic literature; the compiler's connecting and explanatory
matter imitates closely the language of the sources.lO Ibn Ghayyath
also translated into Arabic and we shall meet him as a Paytan.
Isaac b. Baruch Albalia (1035-94) is credited with a !:l'~:l1i nElp,
probably a commentary to the Talmud, and with a i1:llli1 iElO on
the calendar, as well as other works. Both are lost, and we can make
but guesses as to their language. 11 The only specimen we possess of
his responsa is in Arabic. 12
It is possible that more halakhic works in Hebrew existed. They
were written for specialists, and allow no conclusions as to the lan-
guage of the general public, any more than the books of !:l'iDrm pub-
lished in our days. Maimonides stated in 1170 that the Talmudic
style was unknown to most of his contemporaries,13 and the same
was certainly true in earlier times, when the Babylonian Gemara
had been but recently imported.
The Introduction to the Talmud ascribed to Samuel ha-Nagid (982-
1035),14 is written in a Hebrew very much like that of post-Tibbonid
literature. If it were genuine, we would have to revise all our ideas
of the development of Spanish Hebrew. However, the authenticity

9 This is the only correct fonn of the name, which its bearer translated into
Hebrew as l)'iD1r.l "the helper" (Arabic ghiitha "to help").
10 Lucena is south of Cordoba, and was still Muslim in 1148.
II Steinschneider (Arab. Literatur, p. 131) thinks it was Arabic.

12 Horowitz, Halachische Schriflen der Geonim, ii 35. The question is in Hebrew, the
reply in Arabic.
13 r"'~r.li1 ;:;)0, ed. Bloch, p. 2. Cr. note 2 above.
14 Printed in Talmud editions at the end of Berakhoth.
 INfRODUCTION 43

of the text was doubted already by Steinschneider,15 who believed it

to be a translation of part of an Arabic work by that author. Harkavyl6
suspected that the work was not by ha-Nagid, but by Samuel b.
I:Iophni, the last Gaon of Sura (died lO34). When Cowley discov-
ered the first leaf of the Arabic original,17 this supposition was proved
to be correct. Harkavy, and following him Epstein,18 thought that it
was Samuel ha-Nagid who translated the work into Hebrew. As,
however, we know nothing about any translating activity by the
Nagid, the considerations of style seem to exclude such an assump-
tion. The work was probably translated in the 13th century, and the
Nagid's name attached to it merely by a confusion. It may be that
he really wrote some work of this kind, as we know him to have
been the author of halakhic treatises in Aramaic. 19
Benjamin of Tudela, who travelled in 1165-73, and wrote the
account of his journeys in Hebrew,20 cannot be considered as a rep-
resentative of Spanish Jewry. His home, Navarra, had never been
under Arab rule. Although he seems to have known some Arabic,
his outlook and language was entirely that of a Southern-French Jew.
The only Spaniard who during the twelfth century wrote original
works of a serious nature in Hebrew was Abraham b. Da'ud, whose
works and style have been discussed in the preceding section. 21 By
his use of Saadianic, not Spanish ornate style, as well as by his
dependence on Yosippon, B. Da'ud is shown to have followed a tra-
dition different from that of Moslem Spain. Indeed, he does so
already by his choice of subject, for we find no other historian in
Spain either before him or after him until the last years of Spanish
Jewry. Like other scholars of Christian Spain, B. Da'ud participated
in both Spanish and European Jewish culture. He wrote his philo-
sophical work, "The sublime dogma" (al-'aq'ida ar-rq['i'a, m1Q~ i1Qi) ,
in Arabic, as it was of interest only to Spanish Jews, but for his
historical works he chose the language in which the historical books

15 Catalogus Bodleianae, p. 2471, ef. Baron, i 340.

16 Studien und Mittheilungen, iii 4.
17 In a Genizah fragment, Bodleian MS. Heb. d 62, fol. 98, published Harkavy-
Festsehrift pp. 161ff.
18 Harkavy-Festsehrift pp. 168f.
19 He wrote a ~m':::lj ~n:::l"i1, ef. Marmorstein, Moses Maimonides: Anglo-Jewish
papers, p. 165.
20 Critical edition by Adler, London 1907.
21 Apparently born in Moslem Spain, and martyred at Toledo.
44 INTRODUCTION

of the preceding centuries had been composed. As the main pur-

pose of his work was to counteract Karaite propaganda, it is possi-
ble that he wrote Hebrew in order to reach also Jewish circles that
knew no Arabic. Finally, it is not unlikely that he was influenced by
the new Hebrew movement then beginning in the Provence.
Translating into Hebrew seems to have been rare. Moses b.
Chiquitilla (Gi~atilia, ab. 1060 C.E.) translated two treatises of J:Iayyiij
on grammar,22 but these seem to have remained quite unknown in
the Hebrew-reading countries. His fellow Paytan, Isaac b. Reuben
of Barcelona (born 1043), translated Hai Gaon's Kztab at-bait wash-
shira' (1:lQQi11 npQi1 '0).23 While B. Chiquitilla employed a slightly
simplified ornate style, the language of Isaac b. Reuben's translation
was determined by his subject-matter. It is the Aramaicized halakhic
Hebrew of the Responsa-literature of both Spain and N.W.-Europe.
These few and comparatively small works constitute all the seri-
ous Hebrew prose composed on Spanish soil before 1200, at a time
when hundreds of books on a variety of subjects were written in
Arabic. 24 The same people, however, who thus used Arabic in prose,
showed great interest in Biblical Hebrew, both as an object of philol-
ogy and as the almost exclusive vehicle of poetical expression.
This is not the place to comment on the achievements of Hebrew
philology in Spain. What is important for us is the fact that it occu-
pied itself solely with the grammar and vocabulary of the Bible. It
considered Mishnaic Hebrew only in so far as it offered additional
confirmation of Biblical facts.25 This stress on the classical language
was quite in keeping with the practice of the Arab grammarians,
who considered nothing but pre-Islamic poetry and the Koran. From

22 Ed. ].W. Nutt, London 1870. The translation was made for the "learned youth
Isaac, the son of Solomon the Prince", perhaps a FrenchJew. Cf. Steinschn. Uebers.,
p. 916.
23 Printed Vienna 1800. Note that Catalonia had only for a short time been
under Moslem rule, and was more closely connected with the Provence than with
the rest of Spain. The Jews there were the most suitable mediators between Spain
and Europe.
24 For Jewish-Arabic literature, see the series of articles by Steinschneider "Intro-
duction to the Arabic literature of the Jews" JQR ix (1897) to xiii (1901), the same
author's Arabische Literatur der Juden (1902), and the article by Zobel in the Enc. Jud.
iii 53-89.
25 Both Ibn Janal,l and Parl,lon record occasionally Mishnaic occurrences of the
rarer Biblical words. Cf. Parl,lon p. xxii: "Whatever in the language of Mishnah or
Talmud resembles Biblical Hebrew, I shall include in my dictionary."
 INTRODUCTION 45

the Arab grammarians the Jewish philologists also learned to reproach

the poets for any deviation from the standards they proclaimed. 26
An Arabic heritage was also the exclusive attention of the philolo-
gists to poetry. Prose did not matter; their own works were written
in Arabic.
While the pupils of Dunash about 1000 C.E. had still quoted Yannai
and ~alir with approval, 27 later grammarians turned their sharpest
criticism against them. After the grammarians came the aesthetics
and literary critics, as Moses b. Ezra,28 and Abraham b. Ezra. 29
The criticism of the philologists was all the more likely to have
an effect as metrical secular poetry was itself an imitation of Arabic
models. At first only the form had been adopted. The productions
of the earlier Egyptian poets were still called by the traditional name
of Piyyiit,3o but in Spain this name disappears. Only i'tD is employed,
thus emphasizing the break with tradition and the return to Biblical
ideals. 31 Arabic forms, Arabic themes, and Arabic ways of descrip-
tion and expression pervade every piece of Spanish Hebrew poetry.
Its most original and happy phrases are often no more than remi-
niscences of Arabic poetry, and even the words themselves some-
times take their connotation from some Arabic word of similar sound. 32
Since the Hebrew poets were intent on imitating only the best in
Arabic poetry, they followed to the letter the aesthetic theories of the
Arabic literary critics,34 and produced poetry more correct in manner
and language than anything the Arabs ever knew.

26 Zunz, Synag. Poesie p. 216.

27 m:ntDnil '0, ed. Stem, p. 37.
28 Shirath Israel, p. 147: "Because sometimes one experiences difficulties in com-
posing poetry, some of our masters of language have made things easier for them-
selves, and formed new nouns not found in the Bible by analogy on the ordinary
nominal patterns, just as one would derive forms from verbs, because the latter
obey the laws of analogy throughout. However, the full connotation of a noun, in
Hebrew as in Arabic, can only be known from usage (sima')."
29 See section Ill, note 40.
30 An obituary poem has the superscription Cl'n~il 'JtD... ,?., 01'E) (Mann, JfWS

in Egypt, ii 79).
31 Perhaps the choice of this word, in preference to its Biblical synonyms, is not
quite unconnected with its similarity to the Arabic shier. That may be why we always
find i'tD, never the more appropriate ili'tD for a single poem.
32 Many instructive examples are given by Goldziher JQR xiv 719-36. e£ also
Kaufmann, ZDMG lii 307 on Arabisms in B. Gabirol's Diwan.
33 Faithfully explained by Moses b. Ezra, Shirath Israel, chapter viii. On p. 157,
the author expresses his regret at being able to include only twenty poetical devices
46 INTRODUCTION

It was for the Hebrew poet much easier to keep to the strictest
purism because his language was completely divorced both from
prose, with its needs of intelligibility, and from the obtrusion of a
steadily changing colloquial. There was no popular colloquial poetry
to challenge the literary standards, such as in Arabic Spain produced
Ibn Guzman and the Muwashsha}:l. Hebrew poetry lived in a rarified
atmosphere of literati and their highly-educated patrons where taste
and conformity to rules were more applauded than depth of feeling.
The reason for the employment of Hebrew in poetry was without
doubt the desire to have something specifically Jewish which would
equal the achievements of Arabic literature in the field of which it
was proudest. We find that just in the same way, the Persian national
revival began with poetry. As Rudagi and FirdausI were intent in
their language to keep as close to pre-Islamic models as possible,
and wrote in a style that was probably much purer and freer from
Arabisms than their normal speech, so the Hebrew poets kept anxiously
to the most ancient form of Hebrew.
The result was, at least theoretically, a language that made no
innovations on Biblical material. 34 Actually, this ideal was never
achieved. The language of Spanish poetry was a child of Saadianic
Hebrew, and however much it endeavoured to deny its origin, it
could never fully do so without endangering its own existence. It
inherited from the Piyyu! the preference for hapax legomena35 and
the exaggerated use of rare grammatical forms, especially where these
were needed for metre and rhyme. 36 No Spanish poet is quite free
from words and usages that were first created by the Pay!anim. 37 It

"for we cannot ImItate the Moslems in the rest of these artifices, as our language
is not suited to them. For each deVice, I shall bring one example from Arabic
poetry and compare it with a verse in our Bible, so that we may not be despised,
and that it might not be said that we do not know these laws, and that Arabic is
the only language possessing pithy expressions and striking turns of phrase, while
Hebrew has none of these .... "
34 Shirat~ Israel, loc. cit.: "You may use all words you can find in the Bible, but

do not introduce any in your poetry by analogy only. Go with the language, and
stop where it stops. Imitate it, but do not create anything new in it. Follow it but
do not go ahead of it."
35 Kenaani, Leshonenu x 175.
36 Shortened forms of imperfects, imperatives, and participles of tertiae y6dh,

Zunz, ~nag. Poesie, p. 480. Infinitive absolute after the finite verb, ib., pp. 482f.
The later European poets who used metrical form did not employ these licences,
cf. ibid. p. 219.
37 Cf. K. Albrecht, "Zum Lexikon und zur Grammatik des Neuhebraischen"
(ZAW xix 135-55) on post-Biblical material in Moses b. Ezra's "Tarshish" (The
 INTRODUCTION 47

was impossible to cut oneself off completely from the development

of the last 1500 years, but the will to do so existed. It is more notice-
able, as might be expected, in the minor poets than in the really
great ones, who were not afraid of creating their own language. The
worse the poet, the more his work becomes a simple mosaic of
Biblical quotations. 38
The classicist standards were established in secular poetry by Samuel
b. Nagdelah ha-Nagid of Cordova (982-lO55), who was himself an
outstanding philologist. 39 The imitation of Biblical standards in his
Diwan was outwardly marked by its division into "sons" of the Psalms,
Proverbs and Ecclesiastes. 4o Though the ideas and images are largely
drawn from Arabic belles lettres,4! he succeeded remarkably well in
clothing these in pure, almost Biblical, Hebrew. 42
Both through his poetical fame 43 and his influence as a patron of
the arts, the Nagid set an example for all contemporary and future
poets. The classicist principle immediately took hold of secular poetry
and was never abandoned. It was not so easily introduced into

provenience of that material is not investigated). No study exists on Pananic ele-

ment in Spanish poetry. Kenaani's article "Linguistic influence of the early Pananim
on the Spanish poets", Lesh. x 173-82, in spite of its promising title, deals only
with the Piyyft!im of Solomon b. Gabirol, cr. below note 43.
With regard to Mishnaic and Midrashic vocabulary, there is a general attitude
of uneasy toleration. It would in many cases have been quite impossible to man-
age without a Mishnaic word, and the Spaniards never went to the lengths of some
Haskalah writers, like Professor HaIevy, who would rather form a new word than
accept a post-Biblical one. Moses b. Ezra (Kokovtsov p. 220, Shirath Israel, p. 60)
says: "One is quite entitled to seek assistance from the language of the Mishnah.
Even if some of it violates the rules of grammar, it is genuine Hebrew. We can
rely on its compilers, for they lived in a period that was close to the full life of the
language. It would be most desirable if one could explain their irregular forms
according to the rules of grammar. Ibn Jana\:l and other grammarians tried to do
so, each one according to his ability." For a list of post-Biblical words, etc. in the
poems of Samuel ha-Nagid, see D. Yellin in Madda'e ha-Yahaduth i (Jerusalem 1926)
157. A similar list for al-l;Iarizi's translation of the Maqamas of al-l;Iariri, Schirmann,
pp. 91-4.
38 cr. Delitzsch, Poesie p. 146: "The scientific Babylonian-Spanish school treated
the language of the Bible as an unalterable established thing, the classical model
of a dead language, which one is only allowed to imitate, never to develop the
exis.te~t sacred treasure. The style of the Spaniards is a mosaic, like Ciceronian
Latm.
39 Poznanski, Les ouvrages linguistiques de Samuel ha-Nagid, RE] lvii 253, lviii 183.
4{) Diwan, ed. D.S. Sassoon, Oxford 1934.

41 Shirath Israel, p. 66.

42 Bemstein, Tarbiz; xi 298.

43 Moses b. Ezra praises him more lavishly than any other poet, Shirath Israel,
pp. 65-7.
48 INI'RODUCTION

religious poetry, which was dependent on the approval of worship-

pers accustomed to the Paytanic manner. In liturgical poetry we can
almost watch the transition taking place.
Piyyut had already found worthy representatives on Spanish soil.
Joseph b. Abitur, who died about 10 10, produced more than a hun-
dred poems still preserved. 44 He wrote of course pure Paytanic He-
brew, although a certain prominence of Biblical quotations is noticeable
even in his work. 45 Solomon b. Gabirol, the contemporary of Samuel
ha-Nagid, used the purest classicist style in his secular poetry, but
in his Piyyut he was entirely traditional and in no way different from
~alir. 46 He employed two completely different languages, because it
was still unthinkable to apply the standards of Arabic literary taste
to sacred poetry.47
Isaac b. Ghayyath, in the next generation, does already mark a
significant change. In his hundreds of Piyyutim, he never employs
metre,48 but they are constructed around Biblical quotations to an
extent never known before. His language, too, abandons entirely the
Paytanic pattern and becomes as purely Biblical as that of secular
poetry.49 We can understand why Moses b. Ezra calls him "a well
of elegant style".50 The new features of his style seem to be due to
the influence of the Nagid, which is also noticeable in other respects. 51
His contemporary, Isaac b. Reuben, too, shows a marked tendency
to employ Biblical material, and a similar purism in language. 52
Around these two a definite movement can be discovered towards
establishing Biblical Hebrew in liturgical poetry.53
This movement seems to have met with some opposition. Moses
b. Ezra makes some cryptic remarks about adversaries of Samuel

44 Elbogen, Enc. Jud. viii 309-12.

45 Bemstein, op. cit., p. 297.
46 A list of his lexical and grammatical peculiarities, Yellin, Leshonenu vii 234-43.
47 Yellin, op. cit., 219.
48 Bemstein, op. cit., p. 295.
49 Ibid., 297f.

50 Shirath /sr{JJ!l, 72: "He knew the secrets of the Hebrew language as well as hav-
ing command of the Aramaic ... He wrote more ethical, liturgical, laudatory and
commemorative poems than anyone before him, but he did not compose many
metrical poems, because his knowledge in the Arab sciences was small."
51 Sachs, Relig. Poesie, 262, note 3.

52 Schirmann, Enc. Jud. viii 543. "He chose the best out of the Prophets and
incorporated it into the prayers." (Shirath /sr{JJ!l, p. 73.)
53 Bemstein, op. cit., p. 299 calls it "a popular movement" (n"1:::l'~ il.ll1ln).
 INTRODUCTION 49

ha-Nagid,54 and tells us that Abu Arnr b. Sahl (d. 1124) carried on
a poetical war against "the group of men who attacked the ornate
style and condemned the products of the poets". 55
The new school, however, was in the end victorious. It conquered
popular taste so completely that in the end all the older Piyyutim
disappeared from the prayer books of the Spanish rite and were
replaced by metrical compositions. By the year 1100 already, poets
in Narbonne and Rome used metre and the Spanish linguistic man-
ner for religious compositions,56 i.e. Spanish poetry come together
with Spanish philosophy, only it spread more rapidly and farther.
By 1150 these had been introduced to North-Western Europe, where
Jacob b. Meir Tarn was the first poet to use metre. 57 The purism
of the language of Spain, however, never penetrated to France and
Germany. The poets of these countries were too familiar with Midrash
and Talmud to deprive themselves of the linguistic treasures of
Mishnaic Hebrew.
As Piyyut led to the revival of Biblical Hebrew prose in the age
of Saadiah, so Spanish poetry produced a new ornate prose. This,
which was thus a descendant of Piyyut in the third generation, was
at first distinguished by a purism that went even beyond that of the
poets. It seems to have had its origin in the rhymed prose used in
the opening lines, and sometimes in the main part, of private letters,
in the 11 th century. The first work in rhymed prose was the 15:)0
':li~::Jnn58 of Solomon b. Sa~bel, a contemporary of Judah ha-Levi.
The great age of rhymed-prose literature was, however, the thir-
teenth century, when quite a number of authors employed it. Catalonia
appears to have been the centre of this literary movement, though
its most outstanding representatives, Judah al-I:Iarizi (1170-1230) and
Jedaiah ha-Penini Bedershi (i.e. of Beziers, 1270-1340) lived outside
it. In its later stages, this literature absorbed a great deal of features
from the contemporary Hebrew scientific and philosophical prose.

54 Shirath Israel, p. 67.

55 Ibid., p. 74. Cr. also the statement to this effect (without reference) in Zunz,
~n. Poesie, p. 220.
56 Zunz, op. cit., p. 219.
57 Ibid., p. 248.
58 Ed. Schor in Hechaluz iii 154ff. The t:l"~r:l l1r:l~ ~j m~nr:l by the Karaite
Moses b. Abraham Dar'! in Egypt was considered to be older than B. Saqbel, but
the editor of the work, Davidson (mi;;';; 'l1ir:l ii 297) has shown that Dar'! lived
after Judah ha-Levi, perhaps only in the 13th century.
50 INTRODUCTION

The Maqamas of Shemtob Falaquera (or Palaquiera, died after 1290)59

are written in a language that is only slightly different from the prose
of his time.
This was inevitable, as the Maqamists largely dealt with the prob-
lems of contemporary serious literature, and had to make use of its
vocabulary and manner, if only to be intelligible to their readers.
On the other hand, we notice in many prose authors of the 13th
and 14th centuries some influence of the aesthetic principles and the
phraseology of the Maqama. Most authors inserted some rhymed
prose in the opening lines of their books and thus had to be famil-
iar with this style.

9. The First Period if Provenfal Hebrew Literature

Jewish-Arabic literature in Spain had been going on for nearly two
hundred years, and some of its greatest works had seen the light of
the world, when North-West EuropeanJewry obtained its first glimpse
of this new world through the efforts of two men: Abraham Bar
I:Iiyya and Abraham Ibn Ezra. Both were emigrants from Spain;
both possessed a thorough Arabic and Jewish education and felt the
urge to communicate their superior civilization to the Jews of Europe.
Apart from that, the two were utter contrasts in their methods and
in their language.
Abraham Bar I:Iiyya, who died before the year 1136, was a native
of Barcelona. The Jewish, as the Christian, culture of Catalonia
formed a bridge between Spain and France. Abraham probably
enjoyed both an Arabic education of the normal Spanish type and
a European Jewish one, which gave him a command of Talmud
and Midrash and a mastery of their language. His literary activities
took place in Southern France, perhaps Marseilles.' He wrote on a
variety of subjects: religious philosophy, mathematics, astronomy and
astrology, and Messianic mysticism.
Since the Proven<;al and French Jews, for whose instruction he
composed his books, understood no Arabic, he had to write his books
in Hebrew. 2 Since he departed from the custom of his native coun-

59 Malter, JQ,R (NS) i (1910) 151-8l.

IGuttmann in Enc. Jud. i 430-37, where also a list of his works.
2 Ibid., p. 431, quoting Steinschneider. The opinion, sometimes given in popu-
lar works, that he wrote Arabic, seems to originate in a supposition of Neubauer,
 INTRODUCTION 51

try only for the sake of his readers, he did not employ the classicist
style which was the only one then written in Spain, but the type of
Hebrew to which the Jews of France and Germany were accustomed:
simplified late Midrashic Hebrew, i.e. the language of the Gaonic
Midrashim, but largely deprived of its idiomatic and picturesque
expressIOns.
The choice of a type of Mishnaic Hebrew by Bar I:Iiyya must be
considered the most important event in the history of the language
from the cessation of spoken Hebrew until its revival in the nine-
teenth century. Mishnaic Hebrew became the linguistic vehicle of
the most productive intellectual movement of the Jewish Middle Ages,
and through the needs of that movement was developed into a highly
adaptable instrument of thought. Once so transformed, it was des-
tined to remain the basis of all subsequent evolution of Hebrew.
Even the Haskalah reaction in favour of a return to Biblical Hebrew
in the 18th and 19th century was doomed to failure. But the adop-
tion of Mishnaic Hebrew by Spanish writers also meant that it now
absorbed the characteristics of the classicist style in its peculiar Spanish
variety. It came under the influence of the grammarians' purism and
of the sense of style that Arabic education had inculcated into the
intellectuals of Spain. The result was a language in which Biblical
and Mishnaic idiom complemented each other and which was fur-
ther enriched by the close contact with Arabic.
The basis of Bar I:Iiyya's style must have been the Hebrew current
in his time in the Provence. As we have no indubitably authentic
texts in this, we cannot discover to what extent Bar I:Iiyya simply
accepted the current style, and how far he altered it by a closer con-
formity with the style of the Mishna proper. His style is different
from the Northern French idiom of Rashi, many of whose peculiar-
ities are entirely absent from Bar I:Iiyya's prose. Particularly his range
of conjunctions differs markedly from Rashi's. He is altogether more
Mishnaic than the latter. Apart from that, however, Bar I:Iiyya's style
is very plain and refrains from anything unusual and experimental.
It is obvious that for him the language was a means to an end, and

Catalogue qf Hebrew MSS in the Bodleian, i 467, who discovered some MSS of ]1'Ji1
iD:lJi1 varying rather strongly from the printed text, and attributed the differences
to their representing another translation from the Arabic. Final judgment on this
can only be given when those differences are investigated in detail. The language
of the iD:lJi1 j1'Ji1 is identical with that of the mathematical works, which certainly
were written originally in Hebrew. It is quite unlike any translators' style.
52 INTRODUCTION

that he aimed only at giving a clear and correct expression to his ideas.
In general Bar J:Iiyya took over Midrashic Hebrew as he found
it and wrote it with an easy fluency. Yet the language of his edu-
cation, Arabic, subtly pervades his Hebrew style. He is very far from
the Arabic syntax in Hebrew words of the thirteenth-century authors.
One feels that he thinks in Arabic and mentally translates into
Hebrew, but into idiomatic Hebrew. His deviations from Hebrew
idiom belong to the higher regions of syntax: word-order, use of
prepositions, structure of dependent clauses; all matters on which
grammatical guidance was in his time entirely absent3 and which
varied even in the Mishnaic and Midrashic texts available to him
to such an extent as to produce an uncertainty as to what was cor-
rect usage. Since the syntactical apparatus of Arabic was so much
richer than that of Hebrew it naturally produced a subconscious feel-
ing of restriction O'iD'?i1 i~'p in the Tibbonid phrase). Often he
betrays by some awkward turn of phrase that the Hebrew con-
struction he employed did not express adequately what he wanted
to say, and that an entirely different-the Arabic-construction was
at the back of his mind. In general he avoids definite Arabisms, yet
in a large number of cases these came in, by inadvertence or other-
wise. Compared with the style of even the following generation,
to say nothing of the thirteenth century, his Hebrew is distinctly
archaic in character. He exhibits many of the typical Spanish-Hebrew
features only in an embryonic state, and has some Midrashic Hebrew
features that disappeared later on.
There is also a noticeable influence of the Spanish poetry and
ornate prose, both in his choice of words and in his word-order. 4
Abraham ibn Ezra (l092-1167)5 was perhaps one of the most
interesting personalities of the Middle Ages. He was born in Christian
Spain, at Toledo or Tudela6 and enjoyed a good Arabic and Jewish

3 Both because the Spanish grammarians paid hardly any attention to syntax and
because no one had yet investigated Mishnaic Hebrew.
4 Examples for the features of Bar I:Iiyya's syntax, as well as that of all subse-
quent authors, will be found in Part Two.
S A choronological list of his woks in Levy, Reconstrnction, p. xviii., a fuller list

without dates by Bernfeld, Enc. Jud., viii 334ff.

6 Bernfeld, op. cit., 327. The Abu Is\:laq ibn Ezra, whom Moses ibn Ezra mentions
(Shirath Israel, p. 75), side by side with Judah ha-Levi, as his own contemporary
(Moses died after 1138), and whom he calls "the terror of theologians (if Halpel."'s
t:l'i~il:l is correct) and rhetoricists", mayor may not be our Abraham. He is there
represented as born in Toledo and living in Cordova, and nothing is said about
his travels.
 INTRODUCTION 53

education at Lucena in Moslem Spain. Driven by his unsteady spirit,

he wandered across Jewish Europe as a missionary of the philosoph-
ical Judaism of his homeland, engaging in a bitter fight with obscu-
rantism and traditionalism. His numerous works were written in an
offhand manner, and he sometimes repeated writing the same book
in different localities.
To the accomplishments of Bar I:Iiyya he added those of a gram-
marian and a poet. 7 Neither his works on grammar, nor his treatises
on philosophy or mathematics, however, were either original or thor-
ough. They were by-products of his mind, written at the request of
friends or patrons. The great work which occupied him all his life
was his Bible-commentary. In it he utilized all his encyclopaedic
knowledge in one vast edifice of Weltanschauung.
Like Bar I:Iiyya, he had to employ Hebrew as the sole means of
communication with his European public. 8 But his difference from
the thorough, pedestrian Bar I:Iiyya is also mirrored in his language.
He was a philosopher of language and a connoisseur of style, and
he put his theories into practice. His Hebrew still allows us to dis-
cern the Mishnaic basis in its tense system and dependent clauses,
but that foundation is overlaid with elements taken from Biblical
Hebrew, Saadianic style, poetry, Arabic, and perhaps Ibero-Romance,
with the addition of some features that are the fruit of Ibn Ezra's
grammatical speculations. It is a gallery of linguistic curiosities per-
meated with his curious personality, a style severely his own and
never fully achieved after him.
Ibn Ezra was acquainted with the w.orks of his older contemporary
Abraham bar l:Iiyya,9 but it seems that his own style was not influenced
by the latter. These two men may be said to have been the joint
creators of Mediaeval Scientific Hebrew. lO Ibn Ezra's pupils, Ibn
Parl:lOn and the ~ml:Iis, followed his principle of enriching Mishnaic
Hebrew by classicist elements, but with much more moderation than

7 D. Rosin, Die Reime und Gedichte des R. Abraham ibn E::;ra, Breslau 1885-94.
8 Bemfeld, op. cit., 329, recognises that the need for being understood was the
only reason that led him to write Hebrew. There is nothing in his views on Judaism
to warrant the view that in writing Hebrew he followed some definite educational
policy.
9 ct Friedlander, the Commentary qf Ibn E::.ra on Isaiah, i, p. xi, and Steinschneider,
Zeitschrifl for Mathematik und Physik, xiii 11, note 20.
10 Correct accordingly the popular view, expressed by Roth (Short History, p. 171):
"Ibn Ezra ... may indeed be said to have created Hebrew prose as a medium for
scientific purposes".
54 INTRODUCTION

their master. Although there is something of Ibn Ezra's experimen-

tal spirit in practically every Spanish-Provenc;:al writer, it was on the
whole the more sober approach of Bar I:Iiyya that prevailed.
The activity of these two had aroused the interest of European
Jewry in the ideas of Spain. The demand was satisfied by a group
of exiles from Spain, refugees of the Almohade persecutions of 1148,
who provided further information in Hebrew, partly in the shape of
original summaries of Spanish science and scholarship, partly by
translating the standard works, both Jewish and non:Jewish, from
Arabic into Hebrew.
The original presentation of Spanish science in Hebrew was, in
the generation immediately following Ibn Ezra, entirely restricted to
philology and Biblical exegesis. These two sciences had been treated
in Spain only in a disconnected way unsuited for popularisation.
Thus, while a number of specialized grammatical studies had appeared,
no complete grammar of Hebrew existed in Arab Spain. 11 No com-
plete commentary to the Bible had been written since Ibn Chiquitilla. 12
The great advances in exegesis made since his time had not yet been
incorporated into a complete commentary. Ibn Ezra had provided
both, and the other Spanish exiles followed in his wake. For phi-
losophy and natural sciences, on the other hand, it was possible to
utilise the existing Arabic works.
The principal workers in the field of original composition were
the I:>..im}:li 13 family in N arbonne. J oseph I}..im}:li (1105-70) had still
been educated in Spain. His Hebrew bears strong marks of the
influence of Spanish ornate prose style and resembles in character,
though not in details, that of Ibn Ezra. It can be studied only in
the prefaces to his two grammatical works 111:Jr i~O and ',?Ji1 i~O.
The grammatical text of these books is very laconic (as they were
probably intended to be learnt by heart) and unsuitable for syntac-
tical analysis. The apologetic treatise Sepher ha-Berfth ascribed to him l4
is rather different in its language from the above-named works, and
makes the impression of a much later date.
More important was the work of his two sons, Moses (d. 1190)

11 Ibn Pari:lOn in the preface of his dictionary, p. xxii.

12 Steinschneider, Arab. Literatur, p. 138.
13 On the pronunciation, J>..iml)i or JS.aml)i, cr. Felsenthal, Kohut Memorial Volume,
pp. 127~138.
14 Printed in i1:l1n nr.ln'?r.I, Constantinople 1710, f. 18bff. Cr. Suler, Enc. Jud., ix
1243.
 INTRODUCTION 55

and David (1160-1235). The former wrote two short grammars,

(n.v'i1 ''?'~iD l'?i1r~ and ~,~ '?~iD) and commentaries on some Biblical
books. The latter worked in the same fields, and became the best-
known grammarian and commentator of the Middle Ages. Both these
authors wrote a Hebrew of the same type of Bar I:Iiyya, plain, sim-
ple late Midrashic style without attempts at linguistic brilliance. The
Arabisms and independent new tendencies, which in Bar I:Iiyya are
still in their beginning, have in their style become settled features of
the language, as the anteposition of i1r and the impersonal con-
structions. Other features typical of the two older writers, such as
the infinitivus cum nominativo, are in course of disappearing.
Solomon Ibn Parl:lOn was a younger contemporary ofJoseph .J>.imlP.
Born at Calatayud in Aragon, he came from the same environment
as Bar I:Iiyya. He emigrated to Salerno where in 1160 he completed
his dictionary l".vi1 ni~nr~ in order to give the Jews of Southern
Italy, who until then had possessed only the dictionary of Mena}:lem,
access to the achievements of Spanish philology, Biblical exegesis,
medicine (to which he gives the queer name of grammatica) and
natural sciences. 15 Following Spanish tradition, he admitted only
Biblical words, but gave also instances from Mishnah, Targum and
Talmud where these words occurred. His Hebrew is in vocabulary
and syntax similar to that of Bar I:Iiyya, but his loose sentence-
construction betrays an influence of ornate prose.1 6
Compared with post-Tibbonid Hebrew, the language of the writ-
ers of this second and third generation is free from glaring Arabisms.
Almost all the constructions they used could be justified by exam-
ples from Biblical and Mishnaic or Midrashic texts. But the spirit of
Arabic subtly pervades everything they write. Constructions that are
unusual or exceptional in the earlier styles appear in their writings
very frequently. In most cases these will be found to be the ones
paralleled in Arabic syntax. Arabic influences above all the idiomatic
parts of the language. The following extract from Judah b. Tibbon's
instructions to his son illustrates well the uncertainty these authors
felt about their own language: 17

15 All this in his own preface, p. xxiif, c£ Bacher, Jew. Enc., ix 526.
16 "Although Ibn Parl.lOn introduces a few Aramaic phrases to satisfy the taste
of his (Italian) readers, the language of his lexicon, with its pure Hebraisms and
the fluency and precision of his style, betrays the influence of his teacher, Ibn Ezra" ,
(Bacher, ibid.).
17 Testament, ed. Steinschneider, p. 7.
56 INTRODUCTION

And beware of errors in the choice of words,18 in word-formation l9

and in grammar, and with regard to masculine and feminine nouns,
for often the habits of the foreign language lead to mistakes in this
respect. . . . And 20 be careful with conjunctions and prepositions, how
to use them and how to apply them to verbs. If you are in doubt
about anything, and you have no book at hand to look it up, then
avoid it. . . . And do not employ formations or words which are not
in current usage,21 even if these are correct according to analogy, for
the unusuaP2 is not agreeable to nature.

The authors who came from Aragon and Catalonia, and those who
lived in Provence, spoke in their daily life Romance languages. The
influence of these on their Hebrew is, however, hardly noticeable,
and may not have been direct at all, but by way of colloquial Arabic.
The features coming under this head are the use of 0[1 with [I' and
the preference for the infinitivus cum nominativo. Romance influence
is indicated for the latter particularly by the fact that it disappears
gradually in the later, more arabicised, language.
The development of Hebrew initiated by this school of writers was
not allowed to reach its logical conclusions. Although they served as
a model to later authors, the style of the latter was even more deeply
influenced by the language of the translations from Arabic, which
began at the same time as the original compositions treated in this
chapter.

10. The Translators

Soon after the year 1000 C.E. the "unbridgeable gulf dividing the
way of learning in the Provenc;:al or Franco-German schools from
the methods used in the East and Spain"l began to make itself felt.
The separation between Spanish and European Jewry became much
more definite than the linguistic barrier would ever have warranted. 2

18 jiiD'? = Arabic lugha "lexicography".

19 Cl'J'J::l = Arabic tarn] "derivation of nouns and conjugation of verbs".
20 Ibid., p. 8.
21 m1r, literally "strange (ghanba) formations or words".
22 '1::1J, i.e., the Arabic munkar.
Marmorstein in Moses ben Maimon: Anglo-Jewish Essays, p. 171.
I

In fact it is doubtful if such a linguistic barrier ever existed. Both in Italy and
2
in the Provence, there were enough Jews who knew some Arabic, and the difference
between Spanish or Catalan and Provenc;:al was no barrier to understanding between
the Christian populations of these countries (Provenc;:al was even used as the language
 INTRODUCTION 57

Mter the enthusiastic reception Mena1:tem ben Saruq's philological

work found in Europe, and the veneration in which it was held
there, the improved system of his successor I:Iayyuj was completely
ignored there, in spite of the fact that almost immediately after its
appearance it was translated by Moses b. Chiquitilla. Even Abraham
b. Ezra's activity did at first not succeed in popularizing it. Although
he had translated I:Iayyuj's treatises a second time about the year
1140 while in Italy,3 and moreover had made I:Iayyuj's system the
basis of his Bible-commentaries, Mena1:tem ben Solomon wrote in
that country in 1143 his 1n~ p~, in which he mentions as only a
philological authority Mena1:tem b. Saru~, and even him without ever
mentioning his name. 4 And still in 1160, Solomon b. Par1:ton, com-
ing to the important community of Salerno, found I:Iayyuj's name
unknown there, and B. Saru~ as the only authority.s
Metrical poetry penetrated into the European countries without
any difficulty,6 which proves that contact was close enough. It was
also quite possible to obtain translators, as is proved by the work of
Moses Chiquitilla and Isaac b. Reuben. Apart from the three trans-
lations made by them, no others were forthcoming. The only con-
clusion we can draw from that is that works such as the Fons Vitae
of Ibn Gabirol, the Duties if the Heart by Ba1:tya b. Pa~udah, and the
Khazar of Judah ha-Levi did not provoke any curiosity among the
N.W.-European Jews.
The change of attitude after 1150 was due to two factors: the con-
scious efforts of Bar I:Iiyya7 and Ibn Ezra, and the closer contact with

of lyric poetry all over Spain). R. Moses ha-Darshan of Narbonne (10-11) is quoted
by Rashi on Prov. v 19 as comparing a Hebrew word with Arabic ('::Jill l1iV'?).
3 Steinschneider, Uehers., p. 916. This is said to have been his first prose work
(Friedlander, The Comm. on Isaiah, i, p. xxi). The translation was edited by Ewald
and Dukes, Stuttgart 1844.
4 He refers to him as C'r111 'J10ip "men of ancient times", and that 150 years

after Mena\:lem's death! Bacher (Graetz-Festschrift, p. 96) suggests that the work
was deliberately written as a protest against the new philological theories.
5 In the preface to his dictionary, p. xxii.
6 See last section, towards the end. It is a riddle how metre in liturgical poetry
could be accepted so easily and early in Europe, while in Spain it had such a hard
struggle for recognition. Perhaps it was to the European mind merely another
Piyyu!ic device, while the Spanish conservatives resisted it because they were aware
of its Arab origin and secular associations.
7 Bar I:Iiyya worked for some time with Plato of Tivoli as an interpreter of Arabic
into Romance. He did, however, not translate any Arabic work into Hebrew (cr.
Steinschneider Uehers., pp. 502, 529, 532, 97If).
58 INI'RODUCTION

Spanish culture in the Provence owing to the immigration of Spanish

intellectual refugees after 1148. Ibn Ezra himself translated not only the
above-mentioned treatises of f.layylij, but also Avicenna's philosophical
novel Ifayy ibn Yaq?iin, the anonymous Principles if the astronomical tables
if al-Khwiiri;:.m'i, az-Zanatl's Geomancy, and perhaps also the Masii'il
(De Interpretationibus) of Masha'allah b. Athrf. 8
The refugee scholars began their work of translation almost imme-
diately after their arrival in the Provence. Probably they found inter-
est in SpanishJudaism already aroused. 9 Between 1148 and his death
about 1160, Joseph I>iml:J.i translated Bal:J.ya b. Pa~udah's Duties if
the heart. His work was, however, not exact enough, and was soon
superseded by the more literal rendering of Judah b. Tibbon. 10
Joseph's translation of the Selection if Pearls attributed to GabiroP 1
was even less literal. It was rather a new work based upon the Arabic
original.
Systematic and scientific translation was inaugurated by Judah b.
Tibbon (1120-90) "the Father of Translators" and his son Samuel
(1150-1230), who between them translated some 15 larger works,
both of Moslem and Jewish origin. They laid a firm foundation for
the introduction of Arabic:Jewish science and philosophy into the
Provence. Both specialised almost entirely in translation. Judah's only
original Hebrew composition were his admonitions to his son 0i~1~),
and of Samuel we have two short works of a partly philosophical,
partly exegetical, character (C1'Qi1 11p' iQ~Q, n'?i1p tD1i'~). To these
must be added the short prefaces which they prefixed to some of
their translations.
In their Hebrew writings, both Tibbonids used the current Proven<;al-
Spanish style of Bar f.liyya and the younger I>iml:J.i. But in their
translations they employed a style that follows slavishly every turn
of the Arabic phrase~and attempts to imitate the very etymologi-

8 cr. Steinschneider op. cit., pp. 286, 572, 857, 916. "Ibn Ezra was the first
real translator from Arabic" (ibid., p. 502).
9 By a strange slip, Renan (Histoire, p. 164) makes out that the translations were
made for the Spanish refugees, who were forgetting their Arabic.
10 Suler, Enc. Jud., ix 1243. The fragments still extant of Joseph's work were
edited by Jellinek as an appendix to Benjacob's edition of the Tibbonid translation
of the ni:::l:::l'?i1 ni:::l1n, Leipzig, 1846. Suler seems to be wrong, cr. Steinschneider,
Hebraische Uebersetzungen, 373, who quotes Tibbon's own preface, by which ISimchi
started only after Tibbon had done the first tractate, but Tibbon did the rest of the
work after Kimhi had finished.
I1 iDipi1 SpiD,·ed. H. Gollancz, Oxford, 1919.
 INfRODUCTION 59

cal similarities of the Arabic words. There resulted a Hebrew which

is only intelligible to the uninitiated by the process of mental re-
translation into Arabic. The idiomatic and syntactical features of the
Hebrew language (in any of its forms) are constantly violated and
often constructions result which in any other Hebrew would bear a
meaning quite different from that intended by the translator.
This is quite in keeping with the ancient and mediaeval ideas of
translation. The task of the translator, especially of religious or philo-
sophical and scientific books, was not to produce a readable text,
but to render as closely as possible the wording of the original. 12
Translations were revised, not in order to improve their style, but
to make them more similar to the original, even if this meant dimin-
ishing their intelligibility. The instances of the Aquila translation, or
of the successive revisions of the Syriac Bible, are well known. The
unidiomatic character of the Tibbonid translations is fully equalled
by the Syriac and subsequent Arabic translations of Greek works.
True, some of the masters of the Hebrew language raised their
voice against this barbaric procedure. Maimonides wrote to Ibn
Tibbon the following lines, which are a classical statement of a trans-
lator's duties: 13
If anyone, in translating from one language into another, undertakes
to render each word in the one by one single word in the other, and
at the same time to keep to the original order of words and sentence-
construction, he will find this most difficult, and his translation will in
the end be ambigious and clumsy.'4
A translator must first of all understand the meaning of a passage,
and then relate it clearly in such a manner that this meaning becomes
intelligible in the second language. It is impossible for him to avoid
altering the order of elements l5 or rendering one word by several. ...
This is the method I:Iunain B. IsJ:1aq applied in translating the works
of Galen, and likewise his son, IsJ:1aq, in translating Aristotle, and thus
it comes that all their translations are very clear.

12 In the programmatic preface to his first translation, Ibn Pa~udah's, m:l1n

m~~"i1, Judah states his intention of translating more literally than his predecessors.
13 !lobe;:;, ii 27.

14 niD~1iDr.n np~10r.:l. The words could also be translated "doubtful and faulty",

but they obviously refer to the intelligibility of the Hebrew text, not its relation to
the original.
15 'n~'1 Cl1ip" i.e., the Arabic technical phrase taqtfim wa-ta'kkir "irregular word-

order".
60 INTRODUCTION

This, and similar admonitions,16 however, were of no avail. The

mediaeval mind did not see in the translation a work of art in its
own language, but demanded only one thing of it: faithfulness; and
there is little doubt that by sacrificing the idiomatic character of the
second language greater faithfulness could be achievedY Nor was
the stylistic taste of the Proven<;al Jews sufficiently developed to rebel
against the barbarisms of the translators, as did the more cultured
Spaniards. On the contrary (as with the authors of the Amoraic
Midrashim) the weird and unintelligible phrases may have held a
strange fascination for them, and as we shall see, they were only too
eager to imitate them.
Another reason for the low quality of the style in our translations
may lie in the fact that translation soon became an industry, car-
ried on by dozens of specialists. 18 The demand for production was
so great that there was no time to strive for perfection. Where the
Tibbonids had failed through being pioneers, their weaker succes-

16 Maimonides' words were copied by several translators in their prefaces, though

they did not keep to them. There seems to have been a tradition of such precepts
for the translator. Cf. the advice of Judah b. Bolat (16th century, cf. Steinschneider,
Obersetzungen, p. 29) in his i,"P "";', fol. lOb: "He who intends to translate must
realise that a general knowledge of the two languages is not sufficient. He must
make a profound study to discover which expressions and constructions are com-
mon to both, so that they can be translated without any trouble, and which are
peculiar to one of them. In the case of these latter, the translator must conform to
the usage of the language into which he translates, not of the original language.
Otherwise he will only produce confusion, and his work will be in vain." Translation
into Hebrew had largely ceased in B. Bolafs day, and he probably merely sum-
marises some older source.
17 "Faithfulness is often carried to slavish imitation, at the expense of the idiomatic
character of the language-the epigones see their ideal in letter-worship, everywhere
the same phenomenon!" (Steinschneider, Uebersetzungen, p. xix.) In the second half
of the 13th century, Solomon b. Ayyiib, the third translator of the n11,"Qil i:;JO, dis-
tinguishes between an earlier school of translators, who endeavoured to produce a
readable Hebrew, and the later school, whose only aim is a literal, mechanical ren-
dering without regard to Hebrew style (Steinschneider, op. cit., p. 928, the preface
is not in Peritz's edition of the translation). An interesting excuse is given by Joseph
b. Isaac al-Fawwal (cf. Steinschneider, p. 923), in his preface to the translation of
Maimonides' commentary on Mo'ed (in the Talmud editions in the appendix to
Shabbath): "The earlier translators might in many cases have expressed themselves
more clearly and precisely if their style and language had not been attracted by
the language from which they translated. This seems to be a natural failing of all
translators. Now I myself have experienced this: my thought, grammar, and style
have been affected by the Arabic. This is the reason why I have so often erred
with regard to particles and other matters of usage".
18 Most of the minor translators worked to order, although they never mention
in their colophons any fee received for the work, as the copyists often do (Stein-
schneider, op. cit., p. xvi).
 INTRODUCTION 61

sors were content to ImItate their failures. Possibly a large amount

of this work was carried on with the help of vocabularies, similar
to the well known Greek-Syriac word-lists. In any event, translation
tended to become stereotyped, not only with regard to the choice
of words, but also in the grammatical and syntactical sphere. Even
where the Hebrew phrase was not an exact replica of the Arabic
one, certain Hebrew constructions came with time to be accepted
as the conventional renderings of this or that Arabic construction,
and were sometimes employed even when the Arabic construction
bore quite a different sense. 19
The only exception to the general rule was Judah al-I:Iarizi (1170~
1230), who treated translation as an art and achieved in it some
artistic triumphs. He was also the only mediaeval Jew who trans-
lated from the Arabic a work of real literary merit, and for the sake
of its artistic value, not of its instruction or entertainment. 2o He chose
the most difficult of all: the Maqamas of al-I:Iarlr1. 21 Not satisfied
with this lour de force, I:Iarizi undertook to produce an equally in-
genuous series of linguistic acrobatics based on the resources of the
Hebrew language. 22 He applied the same sense of artistry to his
translations of scientific works.23 He explained his principles of transla-
tion in the rhymed-prose preface to his version of Maimonides' Mishna-
Commentary, from which we give here some extracts: 24
In most passages I translate word for word, but first of all I endeav-
our to express the meaning adequately .... And when I find in the
Arabic text a refractory word ... my thoughts pursue it until they hunt
it down 25 ... But if a clumsy26 word chances to come into my Hebrew

19 One of the most important desiderata for the textual criticism of the transla-
tions would be a collection of the syntactical devices by which various Arabic con-
structions are rendered. Such an undertaking should also be a useful contribution
to the much neglected field of Middle Arabic syntax.
20 The translations under the heading of "entertainment" are all of international
popular tales, as Kalilah wa-Dimnah and the Alexander-romance. The more refined
Spanish-Arabic Adab-works were not translated.
21 "~'n'~ mi~nO ed. Chenery, 1872. This work was undertaken at the request
of Spanish patrons (Steinschneider, op. cit., p. 851), who certainly were quite well
able to understand Arabic. It is also hardly a translation in our sense, but substi-
tutes and changes the original most freely.
22 'J10:Jnn, critical ed. by Lagarde, 1883.
23 See the list by Brody, Enc. Jud., v 317f.
24 In the editions of the Talmud, supplement to Berakhoth.
25 C1'i~iDi1 iD, a witty adaptation of Josh. vii 5.
26 i1iDP, i.e. the Arabic thaqfl. "Difficult" gives no sense.
62 INTRODUCTION

text, then I change it for another, and refrain from using it27 ••• And
whenever I want to translate an Arabic word, I search for three or
four different Hebrew words, and choose the most fitting one ... And
I pick the sweetest words out of the Holy Tongue ... so that the words
of my text may appeal to him who hears them .... 28
The important feature for us is not in the high artistic aspirations
of al-I:Jarizi, but in the fact that he attempted to translate into Spanish
ornate style. This reaction against current practice is typical for the
Spaniard I:Jarizi. 29 It is, so-to-say, the last stand of the stylistic ideals
of Spanish:Jewish culture before the rising tide of Midrashic Hebrew.
The attempt was unsuccessful, both in its execution and in its
reception. I:Jarizi was forced to incorporate into philosophical trans-
lations a good deal of Mishnaic elements and of the new creations
of the Proven<;al translators, yet his translation of the Cl':J1::lJ i1...,10,
while it makes much better reading than that of Samuel Ibn Tibbon,30
is in fact less clear and the philosophical thought in it more diffi-
cult to follow. The translation was sharply criticized not only by
I:Jarizi's rival translator, but even by his own friends,3l and, after a
short period of popularity, was almost completely forgotten. 32 He
found no imitators; all subsequent translators followed the Tibbonid
tradition.
The volume of translating activity was immense. 33 Within a rela-
tively short time, practically every Arabic work of any importance

27 ilji'Otli i1i':::lJr.l1 "remove it from being a queen", I Kings xv 13.

28 The same artistic view of translation had been expressed nearly a hundred
years earlier by Moses b. Ezra (Kokovtzov, Vostochinia Zametki, p. 216; Shirath Israel,
pp. 55f.): "Sometimes certain nouns or verbs are found in one language, but not
in the other. In such cases the translator is compelled to employ metaphorical
expressions that circumscribe them or designations that give their meaning approx-
imately, without being identical. Then the beautiful effect intended by them and
the whole interplay of meaning is irretrievably lost. It is impossible for the nouns
and the particles and other devices of syntax to be the same in every language ....
But once the translator has understood clearly the general meaning of a passage,
he must not be too pedantic with regard to the wording of the original, or too
penible with rendering it, for what one wants in the sciences is their sense." In
Moses b. Ezra's day, such a discussion was of course largely academic, and could
refer only to the comparatively small body of translation from Hebrew in Arabic.
29 He originated from Granada or Toledo (Brody, op. cit., p. 312).
30 The two can conveniendy be compared in Klatzkin's Anthology, pp. 63-103, where
they are printed in parallel columns. The translation of M. Lepin printed there,
pp. 71-2 (Zolkiew 1829), was not made from the Arabic but from B. Tibbon's text.
31 cr. Brody, op. cit., p. 318.

32 It was not printed before 1851, cr. Steinschn., Arab. Literature, p. 207.
33 Steinschn., Arab. Literature, p. xx., Uebersetzungen, p. xxii.
 INTRODUCTION 63

in philosophy and medicine had been translated into Hebrew, as

well as a good number of books on natural history, as also a number
of popular tales and ethical sentences. It was possible for a Jew with-
out any knowledge of Arabic to acquire the complete range of Arab
education in these subjects. About 1300, Gershom b. Solomon of
Arles composed an encyclopaedia (1:l'~tDi1 1.litD) , without any direct
reference to Arabic sources, only out of translated materia1. 34 True,
this work is full of mistakes, but this was not the fault of Gershom's
sources. The important philosopher, Levi b. Gershom (Gersonides,
1288-1344) of Bagnols in Provence, also seems to have known no
Arabic, and derived his profound knowledge of philosophy from
translated texts. 35
This is not the place for treating the lexical creations of the trans-
lators. Although no complete study exists, there are at least some
collections. 36 The importance of these words in the general history
of Hebrew has been much over-estimated. They were not, as the
Paytanic creations, a sign of the formative powers of the language,
and have left its structure untouched.

11. Maimonides

Moses ben Maimon (1135-1204) marks an important turning-point

in the annals of Hebrew as a literary language, pardy owing to his
own efforts, and even more so through the literary warfare that broke
out over his work.
He spent his early youth, until the year 1148, when he fled Spain
during the Almohade terrorism, in Cordova, the heart of Arab Spain
and of Arabic-speaking Spanish Jewry. On the other hand, his youth
was witness of the beginnings of an original Hebrew literature in the
Spanish cultural tradition. His talmudic education was more thor-
ough than was usual among the contemporary intelligentsia, and this,
too, may have determined his linguistic oudook. He was utterly

34 Steinschneider, Ueberset;:;ungen, p. 9.
35 e[ Adlerblum, A study qf Gersonides, p. 32.
36 J. Goldenthal, Grund;:;iige und Beitrage ;:;u einem sprachvergleichenden rabbinisch-philosophi-
schen Wiirterbuche, Denkschriften der kais. Akad. der Wiss. in Wien, i (1850) 419-53;
Steinschneider, Ueberset;:;ungen, pp. 1036-45; H. Kroner, :(,ur Terminologie der arab.
Medi;:;in und ;:;u ihrem ;:;eitgenossischen hebriiischen Ausdrucke; J. Klatzkin, Cl'nm:lil '~i~
Cl"~iOi?'~il vols. i-iii, Berlin 1928-32.
64 INTRODUCTION

strange to the poeticising, euphuistic, philological tradition of Hebrew

writing which was typical of Spanish Jews, including even an inde-
pendent spirit like Ibn Ezra. We possess not a single piece of poetry
by Maimonides, surely a singular instance in the world of scholar-
poets that was Spain.
In Fez, where he had settled with his family, the young Maimonides
made his literary debut in Arabic with the so-called irJiDi1 nij~ "The
treatise on apostasy". I His first large work, the Mishna-commentary
(1168), written in Fostat, also followed the custom by employing the
Arabic language. Not long after, he addressed a letter to the Jews
of Yemen. 2 To this he prefixed, again in accordance with custom,
a Hebrew part containing greetings and general matter, but he
changes over to Arabic on coming to the subject of his letter, intro-
ducing the change with the following words: 3
And as to the other matters about which you asked in your letter, I
prefer to reply to these in Arabic, so that they can be easily under-
stood by everyone, men, women and children, since my reply to one
of your questions deserves to be carefully studied by every member of
your communities.

The Hebrew in this letter is Spanish ornate style of the less poeti-
cal kind. It actually contains many elements belonging to the newly
created Proven<;al Spanish Hebrew prose. But there is no conscious
attempt to break away from the ornate-style tradition.
About the year 1170, however, Maimonides began to work on his
i1i1n i1JiDr.:l (also called i1prn i') which he completed in 1180. The
purpose and plan of this book were a daring innovation; so, too,
was its language. It was the first major work in Hebrew composed
in an Arabic-speaking country after the year 1000.
It would lead too far here to attempt an investigation into the
philosophical and religious considerations that led Maimonides to
choose Hebrew and not Arabic for this work which was addressed
to the large masses of these countries. 4 That he pursued with this a

I Kobe;:. ii 12a-15b, under the name CltDi1 tD",p ir::l~r::l.

2 Kobe;:. ii I a-7b.
3 Ibid., fol. 2b.
4 As against the "Guide", which was destined for the intellectuals and formed
the crowning completion of his philosophical system for those who were able to
understand it, Maimonides intended the Mishneh Torah as a popular work, addressed
to "all men". Its purpose was, by a simple and uncontroversial exposition of tra-
ditional Judaism in an attractive form, to combat the two dangers besetting the
 INTRODUCTION 65

definite educational policy is shown by his reply to a correspondent

who had asked for permission to have the Mishneh Torah translated
into Arabic: 5
It is worth your while to learn enough Hebrew to understand our
book, for it is easy to understand, and the necessary knowledge can
be acquired within a short time. After having practised on one part,
you will be able to understand the whole work. But under no cir-
cumstances will I agree to have it translated into Arabic, for then all
its benefits (1mD'liJ) should be lost. On the contrary, I intend now (?)
to have the Mishna-commentary and the Book of Precepts translated
into Hebrew.

People did indeed study Hebrew in order to enable them to read

the Mishneh Torah. Not long after Maimonides' death, Tanl:lUm
Yerushalmi composed a Hebrew-Arabic dictionary on the Mishneh
Torah,6 thus proving that there existed a need for instruction and
assistance in understanding the text.
Not only the choice of Hebrew was an important departure from
custom, but even more so the type of Hebrew which Maimonides
employed. 7 Again, we possess his own statement: 8

common people, Karaism and irreligion (Cf. Schwarz, in Der Mischneh Thorah,
p. 75; Strauss in Essays on Maimonides, p. 90f.; Marmorstein in Moses Maimonides:
Anglo-Jewish Essays, pp. 169f.). There is some reason to believe that the reintroduc-
tion of Hebrew into religious literature was part of Maimonides' plan for this strug-
gle for the soul of the people.
S Letter to joseph b. jabir, Kobe;; ii 15b-16b.

6 Cf. Goldziher, Studien tiber Tanchum jeruschalmi, Diss. Leipzig, 1870, and
Bacher, Aus dem Wbrterbuche Tanchum Jeruschalmi's, Strassburg, 1903. The book itself,
still unpublished, exists in several MSS in the Bodleian Library.
7 The vocabulary-innovations and some stylistic features are treated by Bacher
in Aus dem Wbrterbuche Tanchum Jeruschalmi's, pp. 117-146, and in Moses ben Maimon.
ii. 280ff. Schwarz, Der Mischneh- Thora, pp. 74-94, considers Maimonides' style mainly
from an aesthetic point of view. See on all this Sideman in Sinai vi. 429-433.
Attempts to hebraise the technical language of the Halakha were made in two
books: I) 1~, m:I"i1 (ed. Schlossberg, Versailles 1886) and 2) ,'i1rr.l ed. I.M. Freimann
(Sid.). On the other hand, Maimonides used a number of Aramaic words, proba-
bly mostly those current in contemporary Hebrew, cf. such as, cf. ~r.l"P ~:l"O, cf.
Sideman, Sinai vii 10 I.
8 m1~r.li1 '0, ed. Bloch, p. 2; ed. Peritz, p. 2 of text. Cf. also the letter to
R. PinJ:!as ha-Dayyan, flobe;;, no. 140: "The method I had adopted was to write it
in the manner (1'1) of the Mishnah & the language of the Mishnah". Also cf. in the
introduction to the commentary on the Mishnah: R. judah was l1iD'?:l t:l1~i1 '?:lr.l
J'?mr.l1 l1iD'? n~ iD1pn. The importance which Maimonides attached to the linguistic
character of the Mishneh Torah was so great that at the time when he was only
working out the general plan (cf. pp. 7-8 in Peritz's text), he wrote his Book of
Precepts in Arabic in order to prevent it being considered part of the Mishneh
Torah itself (statement of Moses b. Tibbon, quoted by Peritz, p. i.). This proves at
66 INTRODUCTION

I thought it best not to write It m the language of the Bible, since

that holy tongue is insufficient for us to-day to express clearly in it
legal ideas, nor did I compose it in the language of the Talmud, as
only very few among our people understand it to-day, and many words
in it are unusual and difficult even for Talmudic experts. But I wrote
it in the language of the Mishnah so that it might be easy for the
greatest possible number of people.

By "language of the Bible", Maimonides probably means the Spanish

ornate style derived from Biblical Hebrew, not an independent rever-
sion to imitation of some Biblical book or other. In the same way,
"language of the Mishna" implies for him merely what we should
call Mishnaic Hebrew in its widest sense, and that in its particular
late Midrashic form is employed by his Proven~al contemporaries.
In fact, Maimonides' language in the Mishneh Torah cannot be con-
sidered an imitation of the style of the Mishnah. 9 The structure of
the sentences and the arrangement of thoughts in a passage is quite
that of Spanish-Proven~al scientific literature. A number of Biblical
Hebrew constructions occur, notably the gerund instead of depend-
ent clauses. ID Other constructions are typical of contemporary style,
like the infinitive with separate subject, and the addition of ~tv to
tv', as well as other Arabicisms. II
On the other hand, the style of the Mishneh Torah is distin-
guished from that of the Spanish-Proven~al writers by a consider-
ably smaller number of Arabisms and late Midrashic features. While
reproducing neither Bible nor Mishnah, it has comparatively few fea-
tures that are not, at least occasionally, to be found in these sources.
A particularly interesting feature is the avoidance of all Aramaic
phraseology. The technical terms of Halakha are largely replaced by
Hebrew words created for the purpose.1 2
We may well ask whether one should see in this a conscious ten-
dency to purism. Although the question can hardly yet be definitely

least two things: That the language of the work was an important part of its general
plan, and that Maimonides himself considered it an unusual undertaking to write
such a book in Hebrew.
9 Bacher, Tanchum Jeruschalmi, p. 118.
10 Biblical influence is also noticeable in the vocabulary, er. Bacher, op. cit.,

p. 118. Another typical feature is the use of Biblical phrases as part of sentences
(Mosaic style), cf. the instances by Bacher, Moses h. Maimon, ii 291-94.
11 Bacher, Tanchum Jeruschalmi, pp. 121 and 123. Cr. also Friedlander, Moses ben
Maimon, i 424.
12 Schwarz, Der Mischneh-7horah, pp. 80ff.
 INTRODUCTION 67

answered, it may be pointed out here to what extent the author was
likely to be influenced by the language of his sources, especially after
the prolonged and penetrating study he had devoted to the Mishna
itself He does not insist on pure Hebrew in any of the passages
quoted is which he recommends the use of Hebrew, nor does he
exhibit any particular purism in those of his letters and responsa
which we may assume to have been composed by him in Hebrew.
Neither is his Arabic style any more purist or classicist than that of
other scientific writers of his period. I3 If Maimonides had held purist
views with respect to Mishnaic Hebrew, we should expect a reflexion
of these in the style of his closest pupils, especially his son Abraham.
The latter, however, writes a quite usual, strongly arabicized Hebrew.
Maimonides made an important step forward by recognizing the
value of Mishnaic Hebrew also in theory as equal to that of Biblical
Hebrew,I4 in contrast to the opinion current among Spanish philol-
ogists, who admitted its value only in so far as it confirmed or
clarified Biblical usage. I5 One could hardly expect him to have realised
also, without the help of any philological reference works such as
existed for Biblical Hebrew, the laws and structure of the Mishnaic
language. We may, therefore, assume that the archaism of the style
of the Mishneh Torah was purely accidental. It was not consciously
imitated, and appears altogether to have had little influence on the
further development of Hebrew prose.
The language of the Mishneh Torah was certainly not recognised
as pure Mishnaic Hebrew by the more inimical of his critics. Abraham
ben David of Posquiers accuses Maimonides of talking to the peo-
ple in a foreign tongue I6 and deviating from the language of the
sages, and of employing Hebrew idioms with meanings that are not
proper to them. I7
Much more, however, than Maimonides influenced the history of
Hebrew directly by his Mishneh Torah, he did so indirectly by his
Arabic work, the Guide if the Perplexed, (dalalat al-ba'ir'in, Hebrew i111Q
Cl'J1JJ). The controversy which was started by this book was the start-
ing point of Mediaeval Hebrew literature in its proper sense.

13 FriedHinder, Sprachgebrauch, p. xiv.; Moses ben Maimon, i 426[

14 C[ also the quotations in Bacher, Die Bibelexegese Maimuni's, pp. 165[
15 Thus Ibn jana\:!, Par\:!on, and David ISim\:!i sometimes give examples from the
Mishnah for Biblical words, but do not mention any specifically Mishnaic words.
16 'n~ l1iD'? Does he mean the Arabisms?
17 In various passages of his mJiDi1, quoted by Bacher, Moses ben Maimon, ii 281.
68 INTRODUCTION

12. The Second Period if Spanish-Provenfal Literature

In the age of the earlier Spanish emigrant scholars, from Bar J:Iiyya
to Samuel ibn Tibbon, the communities of the Provence had been
merely passive recipients of the new learning which came from
Arabicized Spain. The Jews of Northern France and Germany had
stood aside; they were still suffering from the devastations of the
Crusades and perhaps also too much occupied with their own
Talmudic studies.
The Second Period, from 1200 to 1400 C.E., is marked by two
features: the complete cultural amalgamation of Spanish and Provenc;al
Jewry, and the gradual elimination of Moslem Spain. The principal
centres were Barcelona and Narbonne. Both were in constant touch
with each other, and each had a number of minor, yet important,
communities grouped round it.
The whole literary development of this second period was deter-
mined by the series of events with which it began, the controversy
about the Guide if the Perplexed. l Through it, the communities of the
two countries were brought together for the first time on a matter
of common interest. The two warring parties were evenly distrib-
uted in both countries, and both contributed in an equal measure
to the polemics. Hebrew, as the only common medium of inter-
course, was the language of all the controversy. When the latter
came to a temporary standstill with the Barcelona ban of 1305, it
left as perhaps its one definite result the permanent establishment of
Midrashic Hebrew as the literary language of all Jewry.2
The first two writers of our period still used Arabic for their more
important works; Joseph b. c~nln (1160-1226? and Abraham b.

I I was unfortunately unable to consult the only monograph on this subject:

J. Sarachek, TIe history qf the anti-Maimonidean controver~, New York, 1932. A collec-
tion of letters connected with the earlier stages of the struggle is to be found in
the third part of the flobe;:;.
2 Renan, Histoire, p. 158: "The history of post-Biblical Hebrew falls into two
sharply-distinct periods. During the first, from the close of the Canon to the 12th
century C.E., Hebrew was still written, but rarely and at large intervals .... In the
second period, from the 12th century to the present day, Hebrew became again
the literary language of the Jews". To be quite correct, this statement should be
qualified by interposing a period of transition, from 800 to 1200 C.E.
3 Steinschneider, Arab. Literatur, pp. 230f. It is doubtful which, if any, of his works
were composed in Hebrew. It is probable for his '10'?m 1Ii1:l0 (ed. Breslau, 1871) and
his commentary on Aboth, ed. Bacher, Berlin, 1910 (cf. Helier, Enc. Jud., ii 36).
 INTRODUCTION 69

Maimon (1186-1237).4 After them, no major Jewish work was pro-

duced in Arabic, either in Spain or outside it, although the literary
use of Arabic never ceased entirely among Spanish Jews. Steinschneider
mentions nineteen Jewish-Arabic writers in that country between
1200 and 1492. Most of them wrote on astronomy, mathematics, or
medicine, all subjects of common interest to Jews and Moslems. In
the latter part of the 15th century, however, Saadiah b. Maimon
Ibn Danan of Granada still found it worth while to compose in
Arabic not only a Hebrew grammar and dictionary (arj-rjarufi fi
'l-lugha), but also a chronology of the Kings of Judah and Israel. 5 As
the same writer was even more active in Hebrew,6 it is possible that
his Arabic writings were destined for the use of readers in North
Africa.
Apart from these few exceptions, all the literature of the 13th and
14th centuries was in Hebrew. The works of this period did perhaps
never equal the great Arabic philosophical treatises of the 12th century,
but in quantity at least it was considerably more fertile than the pre-
ceding age. The reason for the abandonment of Arabic in literary
use cannot be sought in the fact that most of Spain was gradually
coming under Christian rule. The majority of the Jews had not SPQ-
ken Arabic under the Moslems any more than the native Moslem
(Aljamiado) population. 7 Arabic had been the literary language of
the intellectuals. As the above-mentioned authors, and the many
translators, prove, these kept up a practical knowledge of the lan-
guage. The people spoke Romance dialects and did not know Hebrew
any better than Arabic. 8 A certain amount of popular literature in

4 His main Hebrew production was the letter about the Guide to Judah b. Gershom
(flobe;;, iii 5Off. also published by A. Dubno, Wilna, 1821, under the tide ';1 nion'?o.
His monumental Kifiyat al-'iibidin was in Arabic, and only small parts of it were
translated into Hebrew (cf. Steinschneider, Uebers., p. 907; Arab. LiteratuT, p. 221).
One chapter was published by S. Rosenblatt, Baltimore, 1937-8, under the title
"High ways to perfection".
5 Steinschneider, Arab. LiteratuT, p. 172. I was unable to consult the monograph
by Blumgrund, S::.aadja ibn Danon, elele es milvei.
6 See Section 13, note 9.
7 In the tenth century already, a distinction was made in Moslem Spain between
Arabic-speaking and Romance-speaking Jews (cf. Mezan, Enc. Jud., x 557). It is
most unlikely that the Spanish Jews, who until the present day have preserved the
language of the country that expelled them in 1492, should have given up Arabic
so rapidly if it had really been their colloquial speech at the time of the conquests.
8 "In Spain ... an inconceivable degree of ignorance of Jewish matters prevailed
after the end of the fourteenth century ... a great many adult Jews could not even
read Hebrew" (Jacobs in Jew. Enc., xi 499).
70 INTRODUCTION

Judaeo-Spanish existed,9 and even the Bible had to be translated into

Castilian in the 13th century. IQ Had the abandonment of Arabic
been due mainly to the change-over to Spanish, or to political rea-
sons, we should expect a beginning of a Jewish literature· in Castilian
or Catalan. The first independent work on aJewish subject in Spanish
was, however, not written until more than 70 years after the expul-
sion, 1564 when Moses Almosnino's Regimiento de la vida appeared in
Salonica. II
The real causes of the change of language were in purely inter-
nal Jewish developments. Beginning with the Almohade persecutions
in 1148, a shift of the cultural centre had set in, which was practi-
cally completed by 1250. 12 The principal scene of literary activity
was now in countries where Arabic had never ruled. The Proven<;al
Jews, who had always considered Hebrew as their literary language,
played an increasingly important part in literature, and the need to
reach this wide reading public forced Hebrew even upon writers
whose educational background was Arabic. The Maimonidean con-
troversy itself, with its many journeys and correspondences between
Spain and the Provence, probably had itself a considerable influence
in reintroducing Hebrew into Spain.
Even more important was the reaction against Arab humanism
during the 13th century. Increasing stress was laid on the national
values and upon the mystical character of the Hebrew language.
Instead of the Spanish Jews spreading Arab ideas in N.W.-Europe,
two great intellectual movements were introduced from Europe into
Spain. Intensive Talmudic studies in the French and German man-
ner took root in Spain through the efforts of Moses b. Nal).man
(Nal).manides, 1194~ 1270), Solomon b. Adret (l235~ 1310), and Nissim
b. Reuben Gerondi (d. ca. 1380)Y All these three lived in Barcelona.
That centre of Talmudic studies was outside the former Moslem
Spain, and had been in close contact with Europe already in the

9E.g. the Coplas de Yo<;:ef (ed. J. GonzaJez Uubera, Cambridge 1935).

10 Cf. Jacobs, Jew. Enc., xi 489. This translation was, however, never printed
before the middle of the 16th century (Pentateuch, Constantinople, 1547; Bible,
Ferrara, 1553).
11 All works printed before were translations from the Hebrew (cf. Kayserling in

Jew. Enc., vii 324).

12 Cf. Dubnow, op. cit., iv 385.
13 Howdezki, Enc. Jud., vii 299.
 INTRODUCTION 71

eleventh century, as the instances of Isaac b. Reuben and Abraham

bar J:Iiyya show.
Halakhic writings had their own technical language, a blend of
Mishnaic Hebrew and Talmudic Aramaic, and this had necessarily
to be employed by the Spanish Talmudists. Their halakhic treatises
and responsa were intended for a small circle of specialists, and
affected the language of general philosophical and ethical literature
only to a small degree. Whatever increase of Talmudic elements in
the vocabulary of this period there may be found should be ascribed
rather to the more widespread occupation with Talmud and Midrash
among philosophical writers. One can certainly not observe any
influence of Talmudic Aramaic syntax on the language of this time.
Samuel b. ?:arzah, for example, who wrote at the very end of our
period, quotes in his 'El1' ""~~ (1369 C.E.)14 many Talmudic passages
without his style showing any signs of Mishnaic or Aramaic influence.
On the contrary, he often paraphrases these passages into Spanish
Hebrew.
Into Halakha itself, Spanish Hebrew was introduced by Menal).em
b. Zeral). (d. 1385), whose l""i1"~ is a curious mixture of halakhic
and ethical disquisitions,15 and of Halakhic and Spanish Hebrew.
When the Talmudists wrote for the general public they employed
the ordinary style of their contemporaries. Both Nal).manides in his
popular half ethical, half halakhic treatises and in his Bible com-
mentary, and Nissim Gerondi in his Sermons l6 used quite a regular
Spanish-Provenyal style.
The Kabbalah, too, had been introduced from the Provence l7 and
found its home in Leon, the heart of Christian Spain. In contrast
to the halakhic literature, however, kabbalistic works employed the
language of Spanish-Provenc;:al literature, and even the kabbalistic
technical vocabulary was largely based on the terminology created
by the philosophical writers. The use of Aramaic in the Zohar by
Moses de Lean (1250-1305) was a deliberate deviation from usage

14 Bodleian MS. Seld Arch. A65, Neubauer 1296 and Mich. 141, Neubauer 1297.
15 " ••• Some room is given to moral and dogmatic disquisitions, but they are
only such as to illustrate the general degeneration of speculative theology in his
time". Dubnow, v 266.
16 Constantinople 1530. Gerondi's authorship was doubted for a long time, but
is now definitely established, c[ Rosenmann, Schwarz-Festschrift, p. 491, note I.
17 Abraham b. Isaac of Narbonne (1110-1179) and Abraham b. David of Posquiers
(d. 1199), the contemporaries and opponents of Maimonides, had been students of
Kabbalah (Lichtenstein, Cl'i1.lJil] p. 104; Bialoblocki, Enc. Jud., i 457).
72 INTRODUCTION

in order to give the work the authority of a high age. IS In his other
works the same author employed ordinary Spanish Hebrew.
When the writers in Spain began to write Hebrew, they did so, as
we have seen, through the influence and for the benefit of the Pro-
ven<;:al Jews. It was therefore natural that they also took over the
particular form of Hebrew in use in the Provence, that is, late
Midrashic Hebrew as it had been developed by the Spanish exiles
in the 12th century, and especially by the translators. We do not
know if there was any reaction against this in Spain, where the
ornate style had been current hitherto. The attempt of al-I:Iarizi to
employ ornate style for philosophical translations rather suggests that
such a reaction existed.
From the Hebrew of the first period that of the second period is
distinguished above all by the much more profound influence of
Arabic upon its syntax. The authors derived their education mainly
from the translations of the Tibbonid school, and imitated their lan-
guage quite deliberately. Thus the Kabbalist Isaac La!if of Toledo
(1220-1290) tells us that he modelled his own style on that of the
translations. 19 The writers of the first period had created their own
style from elements of the various older forms of the language, and
Arabisms had been more or less accidental, and were always restricted
by the sense of style derived from the older models. To the authors
of our period, the Arabic constructions of the translations came in
a Hebrew guise, as organic elements of the only style in which
scientific discussion was possible. Their resistance to such construc-
tions was therefore much weakened, and became progressively weaker
as a body of original Hebrew literature in this language grew up.
The authors of the 14th century, who were further removed from
direct Arabic influence, exhibit a more thoroughly Arabicized syn-
tax than those of the 13th, who still knew Arabic and were aware
of the foreign character of these constructions.

18 The language of the Zohar is a Western Aramaic imitating the idiom of the
Palestinian Talmud and the Amoraic Midrashim. Dubnow, Weltgesch., v 152, thinks
that the work is based mainly on lost Aramaic Midrashim, whose language Moses
imitated in those passages which did not derive directly from his sources. Cf. also
the discussion of the authorship of the book by G. Sholem, nlii1'i1 'llir.:l, i 16-29.
19 At the end of the preface to his L:l'r.:ltDi1 1lltD (cf. Steinschneider, Uebers., p. xix.
The text, published in Hashai:Jar, vo!. ii, was not accessible to me). The depen-
dence of the language of this period on the translations was stressed already by
Goldenthal, GrundZiige und Beitrdge, p. 421.
 INTRODUCTION 73

One of the most strongly Arabicized styles is that of Levi b.

Gershom (Gersonides, 1288-1344), who probably knew no Arabic
at all. His Hebrew, both in his popular Bible-commentary and in
his highly technical philosophical works, such as 'j1 n1r.ln'?r.l, reads
often as if it was Arabic transposed into Hebrew words. The Arabic
syntactical features are completely integrated. One can hardly speak
of an Arabic influence any more: what we find here is rather a
fusion of the two languages. Gersonides himself was probably no
more conscious of the composite character of his language than we
are of the varied origins of English words. The result is a fluent and
lucid style. 20 The extreme rapidity with which he composed his works 21
indicates that this form of Hebrew was natural for him and his time,
not the result of reflection of the individual writer.
In general, the experimental character of the Hebrew of the first
period is entirely absent. Apart from normal individual variations of
style, the scientific writers use all the same language. 22 They were
fully aware of its individuality. Alone among all the forms of the
Hebrew language, it even was distinguished by a contemporary name,
m1:ln 11tv'? "the Astronomers' Language".23 It was a traditional liter-
ary idiom which was taken over by every writer as it was.
Being without the backing and control of a closely related collo-
quial idiom, Spanish-Proven~al Hebrew was less narrowly stabilised
than a modern literary language. It possessed a large number of
duplicate forms and constructions. If one comes to it with a knowl-
edge of other forms of the Hebrew language, it is easy to pick out
features belonging to all of these, and often duplicates of which the
one is, say, Biblical and the other Mishnaic. One gains therefore the
impression that the style was thrown together ad hoc from those var-
ious elements. This is, however, not true. The composite aspect of
the language is due to its origins,24 but in the 13th century was not

20 Cr. Husik, Enc. Jud., xi 330.

21 cr. Adlerblum, A stutfy of Gersonides, pp. 29f.
22 The statement of Luzzatto about mediaeval Hebrew (Prolegomeni, p. 103): "tanti
Ebraismi, quanti sono gli autori", is due to the fact that Luzzatto does not distin-
guish between periods, but only between styles (ib., pp. 103-4).
23 Steinschneider, Arab. Literatur, p. xlii.

24 Hebrew was in all its stages (with perhaps the single exception of early Mishnaic)
a highly composite language, owing both to the impact of other languages upon it,
and to the continuity of its literature, which led to the various periods influencing
each other much more than is usual elsewhere. Perhaps its inclination to style mix-
ture was conditioned by the fact that already Biblical Hebrew was so composite
and rich in duplicates.
74 INTRODUCTION

felt any more. The duplicates exist in an almost equal mixture in every
author.25
Another feature which gives to the "astronomers' language" a curi-
ously unhomogeneous appearance is its habit of introducing Biblical
quotations as parts of sentences, and of falling on occasion into ornate
prose. This is, of course, merely due to the literary taste of the time,
and is paralleled very closely by Arabic. It is still to be investigated
to what extent a direct influence of the aesthetic standards of ornate
prose can be traced in the purely scientific style.
Biblicizing ornate prose was the style considered suitable for pri-
vate letters,26 and for the opening paragraphs of books and chap-
ters. This age, too, saw the rise of a rhymed-prose literature in
imitation of the Arabic Maqama. The rhymed-prose compositions of
Solomon b. Sa~bel and Joseph b. Zabara in the twelfth century had
been formed on the model of the older Arabic sq/-compositions. The
Maqama-form, first used in Arabic literature by AJ:!mad b. al-I:Iusain
al-Hamadhanl (969-1007 C.E.), was introduced into Hebrew proba-
bly by J oseph b. (~nin (1160-1226).27 It found its greatest expo-
nent in al-I:Iarizi (1170-1230).28 During the thirteenth century, it
developed into a large and popular literature. 29 The jocular warfare
that was carried on by various authors 30 presupposes a large and

25 Some examples are: the use of imperfect and participle to express the present
tense; the gerund with preposition and the dependent clause for adverbial expres-
sions of time; the equivalence of nomen verbi and gerund; the many synonyms in
prepositions and conjunctions (this is also a feature of Arabic); iD and iiD~ as rela-
tive particles.
26 Thus Nissim Gerondi, who wrote his Sermons in "astronomers' language",
used pure ornate style in his letters (ed. S. Assaf, Horeh, iii 96-100).
27 cr. Kiein, Enc. Jud., ii 36; Schirmann, Die arah. Uehers. der Maqamen des Hariri,
p. 112. Nothing of his Maqamas is preserved; we know of them only by a men-
tion in al-l:Iarizi.
28 'J1a::lnn, ed. Lagarle, Gottingen, 1883. Verses from additional Maqamas were
published by Davidson, Rabinowitz-Festschrift, pp. 83-10 I, and by Bernstein, Horeh,
i 179-187. l:Iarizi's authorship of these verses is not quite certain, but Bernstein
(op. cit., p. 181), adduces weighty arguments in its favour.
29 The names listed by Schirmann, op. cit., pp. 111-32, are probably only a
small part of what did exist. Indeed, about many of them very little is known. The
poor preservation of this literature is due to its largely ephemeral and personal
character.
30 Isaac of Catalonia (ca. 1210) states in the preface to his C1'iDJ nil.ll (Bodleian
MS, Heb. f. 10, Neubauer 2768) that he wrote his work in order to counteract the
dangerous impression made by Judah b. Shabbethai's C1'iDJil ~J1iD ili1il' nma "The
offering of Judah the Misogynist" (published in Eliezer Ashkenazi's anthology C1.1l~
C1'Jpl, Frankfurt a.M., 1854, fols. 1-12).
 INTRODUCTION 75

appreciative public. It found also worthy representatives in the

Provence.
The last rhymed-prose writer in the Provence was also the last
Hebrew author on French soil, Jacob ?:arphathi, who composed his
JP.lJ' mJ~IDr.:l in 1384. 31 In Northern Spain, the production of Maqamas
seems to have ceased not much later, with Vidal Benveniste's n~''Ir.:l
itJ'" '5:).lJ. 32
Poetry, too, still existed, but its great time had passed with the
twelfth century. The Castilian courtier Todro b. Judah Abulafia
(1247-1306) produced a Diwan which equalled in size, though not
in quality, those of the most famous earlier poets. 33 Other poets were
Meshullam de Fiera in the Provence (13th century), his son Solomon
in Saragossa (died 1317), and Solomon Bonfid (1380-1450).34 The
greatest poet in the Spanish manner during this period was Immanuel
b. Solomon of Rome (1270-1330).
The Spanish prose style entered Italy in the 13th century, per-
haps through the literary activity ofJacob Anatoli, the translator and
ethical writer from the Provence (d. ca. 1260). The ethical treatises
nnr.:li1 m'l.lJr.:l by Je}:liel b. Jel.wthiel 'Anaw of Rome (1278?5 betray
in their vocabulary a strong influence of the "astronomers' language",
but their syntax is still purely late Midrashic. His contemporary,
Hillel b. Samuel of Verona (1220-95),36 however, uses a language
quite indistinguishable from that of the Spanish authors.37
The last writers belonging to this period were, in Spain, Samuel

31 Cf. Renan-Neubauer, Hist. lit. de la France, xxxi 710. The third chapter was
published by Kaufmann in RE] xxx 56-64.
32 Schirmann, op. cit., p. 126.

33 r11i'nm Cl'?tDt:li1 p, ed. D. Yellin, Jerusalem 1932-4. It has 800 poetical pieces,

some of them of considerable length. A large section is devoted to acrobatics, such

as poems that can be read backwards, etc., in the taste of the Maqamists. These
artistries were also popular in Christian Spanish literature of the time, cf. the col-
lection by A. Aguilar y Tejera Las poesias mas extravagantes de la lengua castellana,
Madrid, 1920.
34 Solomon b. Meshullam de Fiera lived ca. 1340-1420. His Diwan, ed. by

S. Bernstein, part i, New York 1942, exhibits a remarkable linguistic virtuosity and
command of exceptional Biblical usages. Its language contains perhaps no post-
Biblical elements at all, in any case much fewer than that of any earlier poet.
35 Printed Cremona 1556.

36 tD=:lJi1 '?,t:lJi1 '0, ed. S. Halberstam, Lyck 1874.

37 Owing to war-conditions, the anti-rationalist work Cl't:li1 :m~ by the German,

Moses Tako (printed in it:lnJ i,"'~ vo!. iii) was not accessible to me. The style of
this work would show whether the Spanish Hebrew penetrated also to Germany.
76 INTRODUCTION

b. Zaq;ah of Valencia (ab. 1370?8 and the above-mentioned Nissim

b. Reuben Gerondi (died ca. 1380); in the Provence the compiler
Isaac de Lates (wrote 1372).39 The fourteenth century was under the
shadow of the disasters that had befallen the Jews of Spain and
France: the expulsion from the crown lands of France, 1306-1315, and
the massacres in Spain in 1328 and 1355. Mter the middle of the
century, the older generation of scholars gradually died out, and the
break-up of the educational system during those years of terror made
itself felt 40 in the ignorance and lack of interest of those who had in
the meantime grown up. A definite break in the continuity of liter-
ary tradition of more than twenty years took place, and those who
took up the thread again after 1400 were of a different character.

13. The Third Period if Spanish Literature

The fifteenth century was for the Jews an intellectual interregnum.

Persecutions and internal exhaustion had put an end to the civiliza-
tion they had developed in the Middle Ages. French Jewry had been
utterly destroyed, the East was after the Mongol invasions rapidly
falling into stagnation. German Talmudism, as far as it managed to
survive expulsions, massacres and financial exploitation, became more
and more narrow-minded and uncreative. The new centre in Eastern
Europe was developing an approach that was strange and hostile to
the spirit of the Spanish Hebrew civilization.
The only countries where the tradition of the Middle Ages still
existed were Italy and Spain. Spanish Jewry was in a process of
decadence. The recurrent persecutions disturbed the continuity of its
spiritual life, and the periods of comparative prosperity only threw
up the difficult problems of assimilationism and the interference of
apostates into the affairs of the community. Spanish culture was as

38 The name is spelled with many variations i1~i~, n~i~, i1iiO, i1Ji~, i1i~. His
'El1' ??::lO (cC note 14 of this section) is still unpublished. c"n i1pO, super-commentary
on B. Ezra to the Bible, was printed at Mantua, 1559. The only monograph on
him, by T. Reinach, Revue d'anthropologie iv (1889), was not accessible to me.
39 His theological work, iElO n'ip (Bodleian MS. Mich. 602 Neubauer 1298) is
still unpublished.
40 Gersonides complained somewhere that he could not go on with his writings

"on account of the calamities of the times which interfered with clear thinking." In
a place like Avignon, there was in his time some difficulty in obtaining copies of
the Talmud and even the Bible! (Adlerblum, A stu4J qf Gersonides, p. 27).
 INTRODUCTION 77

yet primitive, and the Moslem civilization of Granada too decadent

to fertilize Jewish thought.
This period witnesses the gradual weakening of the mediaeval
Spanish tradition and the transportation of its last remnants first to
Italy and later to the East. At its end comes the final destruction of
Spanish Jewry. This event may be said to be the beginning of the
modern period of Judaism.
During the first part of the century, such literary activity as there
existed was devoted to a summing-up of the developments of the
second period. There were the scholars who attempted to achieve a
final solution of the problem of faith and philosophy: J:Iisdai Crescas
of Barcelona (1340-1410), Joseph Albo of Tortosa (1370-1444),
Simon ben Zema}:l Duran of Mallorca (1361-1444), J:Iayyim b. J udah
Ibn Musa of Salamanca (1380-1460), I and Zera}:liah b. Isaac ha-
Levi 1"?iD (ab. 1415).2 The last-named is interesting as being one of
the last mediaeval Jewish translators from the Arabic, the last to
translate a philosophical work. It is characteristic of him and his
time that the work which he sought thus to make accessible to the
Jewish reader was none other than the Discomfiture if the Philosophers
(Tahiifot al-Jaliisifa) of al-GhazzalI} the work which did so much to
eliminate philosophy from the Islamic world.
Apart from these there was the grammarian, astronomer, apolo-
gist and historian Profiat Duran Ephodi in Catalonia,4 the moralist
Solomon Al <ammi,5 and the Kabbalist Moses Botarello (wrote in
1409). 6 These authors as a group constitute a transition between the
14th and the late 15th century. They were still following the lin-
guistic traditions of the second period, but signs of disruption are
noticeable. The Arabic element is much weakened, and the Talmudic
studies of the writers begin to influence their language. It is at once
simpler and more experimental than that of the preceding age. Mter

I A letter, and specimens from his apologetic work nO'i' pO were edited by
D. Kaufmann in "o'?n n'~, ii 110-25.
2 A sermon by him was published by Schor in f,'?nil vii 96-10 1. Cf. Rosenmann,
Schwarz-Festschrift, p. 494.
3 Cf. Steinschneider, Uebersetzungen, p. 328.
4 "Ein typischer Vertreter des Epigonentums in der Spanisch-jiidischen Literatur"
(Suler, Enc. Jud., vi 124).
5 iO'Oil n;j~, ed. A. Jellinek, Leipzig, 1854.
6 His commentary on the ili'~' i:lO was printed in the edition Mantua 1562.
The name is spelled '?'iO'~ in the heading on the first page, but ''?'iO'~ in the
opening line of the text itself.
78 INTRODUCTION

the uniformity of the latter, we come back to a picture of extreme

diversity. Each author has his own special mixture. Albo writes a
simple, but rather pleasing, Spanish-Provenc;:al style, but Crescas'
Hebrew is harsh. His sentence-construction is so loose as to become
often hard to follow. One can almost feel the disintegration of the
syntax of the thirteenth-century Hebrew. I:Iayyim b. Miisa's letter is
written in the normal mosaic of Bible-verses. He uses the same style
in his nD'i' pD, as far as can be judged from the specimens. This may
be due either to that treatise having been in letter form, or to a
confusion of styles symptomatic of the weakening of literary tradition.
Mter this group, there was again an interval of about 30 years.
The writers of the end of the 15th century heralded the beginning
of the "modern" period. The most outstanding among them was
Isaac Abarbanel (or Abravanel, 1437-1508). He was still closest to
the Spanish tradition, and this is also visible in his language, which
is a simplified form of the Spanish idiom. A strange anachronism is
Saadiah b. Maimon Danan (d. about 1500), who in the last days of
Moslem Granada produced in Arabic a number of works quite in
keeping with the best Arab tradition, though on Jewish subjects. 7 In
his Hebrew writing he also followed the manner of the 13th cen-
tury. He wrote tracts on philosophical and theological questions,S
and poems in the best tradition of Arabic love poetry, so sensual
that they provoked the disapproval of the staid Abraham Gavison. 9
The typical product of this age are the histories. It is indeed strange
that in spite of the Arab example, the Jews of Spain produced no
historian until the time of their distress. Writing of contemporary
history began with Profiat Duran's lost I11iDiDiT 1'i:lr, and at the turn
of the century we find the two great historians, Solomon b. Verga
with his m'iT' c;):JiD. lO and Abraham Zaccuto (1450-1510) with his
rom' i~O.11 The Hebrew of both these works is harsh and clumsy,
a hybrid between European Rabbinic Hebrew and Spanish Hebrew.

7See Section Ill, note 5.

8Some in the MS. Bod!. Or. 108 (Neubauer 2233).
9 In his commentary on Proverbs, i1n:::ltDii iQ.Il, Livorno 1748, f. 131 b.
10 Ed. Wiener, Hanover 1855. Although the work was based on the materials

by earlier authors, it was Solomon's credit to have brought it into the chronolog-
ical form. Cr. F. Baer, Untersuchungen iiber Qye/fen und Komposition des Schebet Jehuda,
Berlin, 1923, and '''1J) '?.Il n1tDin nn.llii, Tarbiz, vi 152-79.
11 Ed. Filipowski, London, 1857.
 INTRODUCTION 79

The expulsion of 1492 dispersed the Spanish Jews over Europe,

Asia and Mrica. When they began again to produce literature, it
was largely in Judaeo-Spanish. Their Hebrew literary compositions
were restricted to Talmudic studies and religious poetry.12 The tradition
of the "astronomers' language" was preserved in a degenerated form
in the occasional moralist writings and in private letters until the
present time, when it is gradually giving way to Modern Hebrew.

14. Later Developments if the Hebrew Language

The civilization of the Jewish Ghetto from the 16th to the 19th cen-
tury was based entirely on the study of the Talmud. Mter the short-
lived existence of the Palestinian centre in the 16th century, where
there were at least some echoes of the Spanish mentality and lan-
guage, I literary production was limited to Germany and Poland. The
latter country, where in 1210 no scholars could be found to serve
as Rabbis for the new communities,2 had developed a number of
flourishing Yeshiboth during the early 15th century. The study of
the Talmud, the only literary possession of Polish Jewry, was devel-
oped into a fine art. Especially since Jacob Polak (died 1530) had
devised the new method of the I:Iillu~, 3 new possibilities were opened
for an almost unlimited amount of literary activity.
To the books of O"t01in were later on, under the influence of
I:Iasidism, added moral treatises (i010 'iElO) and a certain number
of stories (m'tO.!/o). The greater part of this more popular literature
was, however, written in Yiddish. 4 This was the real language of the
people, while the Hebrew knowledge of the common man rarely
went beyond the understanding of the prayer-book, the Pentateuch,
and the easier portions of the Mishnah.

12 Cf. the lists in Zunz, Syn. Poesie, pp. 356-58.

1 Joseph IS.-aro (1488-1575), who was born in Toledo and wrote in Constantinople
and Palestine, employed even in his legal work 1".11 In'?i/J a language that was basi-
cally Spanish-Proven~al.
2 Cf. Dubnow, Weltgeschichte, v 220.
3 I.e., the search for contradictions between various passages of the Talmud or
its commentators and codifiers, and their resolution by logical methods as an intel-
lectual exercise, which was carried on at regular intervals in the Yeshiboth.
4 The term 'Judaeo-German" should be reserved for the Literary German in
Hebrew characters that was employed from Mendelssohn onwards by many edu-
cated writers.
80 INTRODUCTION

The Hebrew employed by the Talmudists of this time was a devel-

opment of the halakhic style: Mishnaic Hebrew with a strong admix-
ture of Aramaic elements. The intrusion of the latter often reached
such dimensions that one can almost speak of a fusion· of the two
languages. The difference between the vocabularies of the two disap-
peared entirely. Aramaic words were used side by side and instead
of the Hebrew ones, be it in their Aramaic or in a thinly Hebraized
form. Even the morphological elements of the two languages were
used promiscuously.5
The complete ignorance of grammar and absence of any aesthetic
approach to life or language soon led to an almost unbelievable dis-
integration of the morphological and syntactical structure of Hebrew
in the writings of the Talmudists and moralists. 7 Distinctions of gen-
der, conjugation, sentence construction, were completely obliterated.
The vulgar Yiddish tongue penetrated to a considerable extent into
the non-halakhic writings, especially the stories, until these were some-
times nothing but Yiddish idioms with Hebrew words. In spite of its
utter lawlessness, the resulting style often has a certain simplicity and
directness that holds the attention. 8
The movement of enlightenment in Germany and Lithuania during
the 18th century broke with the linguistic tradition of the early mod-
ern age. In their search for a more cultured form of Hebrew for their
prose writings, the earlier Maskilim fell back on the Hebrew of Spain.
All the circle around Mendelssohn composed their more serious works
in a language that was a fairly close approximation to it. 9 Thus did
Mendelssohn (1729-86) in his contributions to the i'~::J, in his i'~
i1::J'm'? and in his commentary on Maimonides l 1":Ji1i1 m'?~,1O N.H.
Wessely (1725-1805) in his .l)l '?'?i1~, Solomon Dubno (1738-1813)

5 cr. E.M. Lipschiitz, lJi1EliD, i 20.

6 Even the very elementary grammatical explanations of Rashi were treated in
the schools as an impenetrable mystery, understood by neither teacher nor pupil.
Cr. the amusing description by Avineri, '"iD' '?:l'il, i 34.
7 "Mediaeval Hebrew ranks among the most slovenly of languages" (Bergstrasser,
Eirifiihrung, p. 47). This language was caricatured by Joseph Perl in his r"r:l~ il'?Jr:l.
Vienna 1819.
B In recent years the style of these stories has been imitated with great success
by SJ. Agnon.
9 Cf. Szneider, Leshonenu, vii 56ff.
10 Lipschiitz, Vom lebendigen Hebriiisch, pp. 11-12 states clearly that Mendelssohn
and Wessely wrote "rabbinical" Hebrew. The same view is held by ADad Ha-am
in his ilnnElOl lliD'?il, APD i 18.9.
 INTRODUCTION 81

III the preface to the 11~:1 on Genesis, Isaac Satanow (1732-1805)

in his n1iDi1 IElO,11 Judah Ben-Zeeb (1764-1811) in his 11iV? i1D?n
'1:11), and in the next generation Abraham Ber Lebensohn (1789-1878)
in his 1:l'?D Ipn and his commentary on Ben-Zeeb's grammar.
This choice of language was quite natural. The writings of the
Spanish rationalists, particularly the Guide-the "Bible of the Mas-
kilim"-were a constant source of inspiration to them. Here were
thoughts similar to their own, expressed in an idiom that seemed
perfectly suited to the expression of pilosophical argument.
For poetry and artistic prose, the early Maskilim also imitated the
Spanish example in employing Biblical Hebrew, or rather its Spanish
ornate-prose development. This again was the most obvious and eas-
iest way to take. The tradition was, however, challenged in 1785,
when Isaac Satanow published an edition of the Selil).oth in which
he, in full recognition of the facts, explained many words by refer-
ence to Mishnaic Hebrew. He drew upon himself the wrath of a
critic,12 who, in accordance with the Spanish tradition, deprecated
the introduction of such elements into poetry. In his famous reply,
published under the name of his son,13 Satanow defended Mishnaic

r
11 Professor J. Klausner, Madda'e ha- ahaduth, i I 64f, assigns to Satanow a most
important place in the history of Modern Hebrew, asserting that it was he who re-
introduced Mishnaic Hebrew into literary style. "The m1r.li1 iElO is composed in a
language that is practically identical with that of the Mishnah, with but a certain
leaning towards the idiom of the Tibbonids, such as was imposed by the subject-
matter; religious problems, morals, and ethics" (ib., p. 165). It may seem pre-
sumptuous to contradict in this respect the creator of the history of Modern Hebrew
literature, but even a superficial glance at the books quoted here should convince
one that they are not composed in Biblical Hebrew. Nor did Satanow in the work
mentioned by Klausner employ anything like the style of the Mishnah. His lan-
guage is modelled very closely upon that of the 13th and 14th century writers (thus
only indirectly on the Tibbonid idiom). He employs almost all the post-Mishnaic
features that are discussed in the following chapters: The proclitic zeh, the con-
junction ki, the various impersonal clause-predicates, the imperfect for the present
tense, etc. His vocabulary and phraseology are most similar to that of Gersonides.
In all these respects he is in full agreement with his contemporaries.
12 According to the note at the end of Satanow's reply, the reviewer's name was
Tobias Gutmann of Piotrk6w. Klausner, op. cit., p. 164, states that the review
appeared in the Me' asseph, but no communication under that name, or on the sub-
ject of Satanow's Seli~oth, is to be found in the volumes 1786-7 and 1787-8 of that
periodical, apart from a favourable notice signed "-:1-' (1786-7, p. 48). It is quite
impossible that Satanow should have directed his wrath against this review. In the
History if Modern Hebrew literature, i 151, Klausner just refers to "a reviewer" with-
out specifYing any place. If the original review escaped even the diligent researches
of Prof. Klausner, the suspicion arises that such a review never existed. Perhaps
Satanow invented it merely in order to air his views on Mishnaic Hebrew.
13 Ha-Me'asseph 1787-8, pp. 82-95, under the title i1r.l"tu J)r.ltu (and not, as Klausner,
82 INTRODUCTION

Hebrew as an organic branch of the Hebrew language, and mam-

tained that it was fit to be used for poetry.14
Perhaps it was Satanow's spirited plea for Mishnaic Hebrew,15 or
it may have been the direct influence of the Beth ha-Midrash, that
moved his townsman and younger contemporary Mendel Lepin
(1749-1826) to write his translations in an imitation of Mishnaic
Hebrew proper. He even "translated" the Moreh of Maimonides from
Samuel Ibn Tibbon's Proven~al translator's language into this neo-
Mishnaic idiom. 16 Actually, Lepin's language borrowed from the
Mishnah only its basic vocabulary and a number of characteristic
idioms. Syntactically it was a blend between late Midrashic, Spanish-
Hebrew and European languages. The introduction of Mishnaic and
Midrashic elements into prose was, however, an important principle
which was taken up again by S. Abramowitz (1837-1917) and
Ch.N. Bialik, and through them became a standing feature of pre-
sent-day literary Hebrew.
In the later course of the development of Haskalah literature, the
standards of the Biblical ornate prose spread to an even greater
extent over scientific writing and other fields. Since Abraham Mapu
(1808-67) it became the normal idiom of narrative prose. Finally,
post-Biblical Hebrew was entirely driven out from Modern Hebrew
literature. Typical of this time was the attitude of Prof. Joseph Halevy
(1827-1917), who advocated the creation of new words from Biblical
material as preferable to the admission of Mishnaic words into the
language. 17
A reaction in favour of Mishnaic-Midrashic Hebrew came with

loc. cit., states, i1~''?Qi11 j1iD'?i1 ':l'iQ. This is the column heading which recurs in
every number of that volume.). The article is written in a curious mixture of Mishnaic
and Spanish Hebrew with high-flown poetical ornate style.
14 Cf. Szneider, op. cit., p. 59. The first poet who introduced Mishnaic Hebrew

into his poems on an equal footing with Biblical, was Ch.N. Bialik (1873-1934).
15 This is the opinion of Klausner, op. cit., p. 167.
16 Zolkiew 1829. The title page announces that the work is "translated into the
language of the Mishnah" (cr. also section 10, note 30). In one of his letters, Lepin
refers to the language of his translations as "The language commonly used by
Talmudic scholars" (cr. Klausner, op. cit., p. 174).
17 Mai:Jberet, Recueil de compositions Mbrafques de J.H., Paris:Jerusalem, 1894, pp.
4-8: Cl',i1ClQi1 i1Pi~ j'Q''? Cl'in~ Cl',:li (appeared originally in Ha-Maggid, v 93-4).
Lipschiitz, Vom lebendigen Hebriiisch, p. 13, does not know the reason for that "purism"
i.e. the spread of ornate style into prose. He suspects Spanish influence. Writers
who did not use ornate prose (? Lipschiitz): S.D. Luzzatto, Krochmall, I. Roll, and
E. Zweifel.
 INTRODUCTION 83

S. Abramowitz (pen-name Mendele Mocher Sepherim) (1837-1917).17

Soon after him A. Ginsberg (Al:tad Ha'am, 1856-1927) began to
reintroduce elements of Spanish Hebrew into his language, and his
disciple E. Ben-Yehudah went even further than he in admitting typ-
ical Spanish-Proven<;al features. 18 Both were, however, far from actu-
ally writing that type of Hebrew. The basis of their syntax was
European, and the Spanish elements found a place in it only in so
far as they did not clash with European syntactical principles.
With the revival of spoken Hebrew through Ben-Yehudah's efforts,
and the establishment of a Hebrew daily press in Palestine, the exper-
imental stage of Modern Hebrew came to an end. The present-day
literary language is becoming settled in its syntax. 19 Spanish-Proven<;al
Hebrew is entering into the making of its syntactical structure only
to a very small extent, but in its vocabulary it has a predominant
position.

17 Abramowitz's followers S. Benzion and Bialik. Other stream: J.M. Pines -

Z. Jawetz - David Yellin. Third (most important, so Lipschutz) Katzenelenson -
Kantor, Ben Hillel Hakohen - Frischmann (grows out of Meliza).
IR There were some authors who deliberately imitated "astronomers' language",
as J. Goldenthal, in the preface to his edition of the Hebrew translation of Averroes'
commentary on the Rhetorics, Leipzig, 1842; HJ. Mathews in the preface toJoseph
I)..imJ:!i's ','?Jil ,:lO, Berlin 1887; and other editors of mediaeval Hebrew works. Such
attempts belong, however, hardly to Modern Hebrew literature.
19 Some teachers and makers of school books are attempting to bring up the
younger generation on Biblical Hebrew. Professor J. Klausner is compaigning for
the recognition of Mishnaic Hebrew as the standard idiom (Spanish Hebrew has
not yet found a propagandist). Meanwhile the literary language is slowly finding a
way of fitting European syntactical categories into the morphological material pro-
vided by the two older stages of Hebrew. The colloquial is developing its own syn-
tactical devices, often of a rather striking nature.
It is hardly necessary to stress that this sketch does in no way claim to do jus-
tice to the extremely complicated history of Modern literary Hebrew.
 CHAPTER ONE

GENERAL SYNTAX OF THE NOUN

1. The Article

a. The functions of the definite article are the same as in BH. I The
generic use is found in BH (Ges., l26n) and in MH (Segal, 374
iv-v). It is not more frequent in SH than in the two other dialects.
The usage of Arabic, demanding the article whenever an entirely un-
defined object or quantity is referred to (Reek., 106.8a), is not followed
in SH, and instances of generic terms without the definite article are
frequent.
h. Only the generic article after partitive 11:l appears to be due to
Arabic influence. It is, however, not frequent: '?:m~ ... rr:nniT 1r.l '?;:!,~
n'n'?'niT 1r.l' '?iiniT 1r.l C!lDr.l "one may take vinegar ... or some mus-
tard and assa foetida" (Maim., De'oth, iv 8); n'r.lr.l i'::l'iiT 1r.l ID' "some
talk brings death" Falaq., MebaMesh, 19a), iT::liiT::l' ~'i::liiT 1r.l iT::liiT::l
~'Jr.l1'iT 1r.l "in many matters and on many occasions" (Gers., De'oth, £ 8a).

c. There seem to be no traces of a tendency to employ in~ "one" as

an indefinite article, such as is found on rare occasion in the Mishnah,
e.g., in~ 1PO,'?J ''? np pi' m'J~ ,'? np "buy for me a bundle of veg-
etables, buy for me a white loaf" (Demai, vi 12), in the later forms
of Midrashic Hebrew, and in N.W.-European Hebrew, e.g. i'::l
in~ ~'C!lO''? "in the hand of a robber" (Rashi on I Sam. xix 17). The
only instance of which I know occurs in the ~'r.l~DiT iElO, wrongly
ascribed to Ibn Ezra (ed. Grossberg, London 1901, p. 6): ,'? i'JiT
~'J'r.lJiTiT ''?'iJr.l in~ l'r.lJiT "I was told by a nobleman, of the great
noblemen".
d. In MH already, some set expressions are never provided with the
article, e.g. iTi,n i'r.l'?n "the study of the Torah", ~'r.l '?.v::l "one with
a blemish" (Seg., 378). The conditions under which this takes place
have not been properly investigated. Segal's contention that they "are
considered definite in themselves" might be allowed for the above
cases of construct-compounds-although it would still have to be dis-
covered why just these few cases of constructs differ from all the
86 CHAPTER ONE

others. It can certainly not be applied to the noun-adjective groups,

such as ?n:1 1i1:J (Sanh. ii l, contrasted with l?Qi1 ib., 2) "The High
Priest", :J'~ Cl" "a festival".
In SH, the construction is extended: i1'i1 1m~:JtD m?:1 tD~i "the
exilarch of that generation" (B. Daud, flabbalah, p. 63), i1i,n ?lJ in'
:In::l:JtD "in addition to the Written Law" (J. NaQmias, Aboth Comm.,
f. la).
e. The inverse case, i.e. construct-compounds that must always have
the article, also exists in SH, both in cases where these are taken
over from MH (cf. Seg., loc. cit.) and in additional ones. There are
even instances where an article is prefixed to such a perpetually
determined expression' i1i'~'i1 C1?1lJi1 "the created world" (B. I:Iiyya,
Hegyon, f. 5), ltDQi1 m'm~i1 "the semivowels" (B. Ezra, Yesod Mora,
p. 18). Both cases may be due to copyist's errors. In the Eastern
European Hebrew of the 17th-18th century, however, such con-
structions became very frequent, and forms like nOJ:Ji1 n':Ji1 "the
Synagogue", i1i,ni1 iElOi1 "the Torah scroll" found their way into
epigraphy.
f. In Gersonides, the determinate form of C1"n ?lJ:J "animal" is always
C1"n ?lJ:Ji1. This irregular construction is occasionally found in all
later forms of Hebrew, e.g., the inscription i1ii1 nm:J n':Ji1 in a
Jerusalem Sephardic synagogue, and Modern colloquial forms like
?:J1~ i,ni1 "the dining-room". It is, however, exceptional in all of
them. There may be other cases of this kind in SH, too.

2. Concord with Regard to the Article

a. The normal construction is in SH, as in BH and MH, deter-

mined by the rule that either both adjective and noun have the arti-
cle or neither of them. However, instances of the article being prefixed
to the adjective only, occur in BH since the earliest times (Ges.,
l26v, Driver, Tenses, 209, full list in Kbnig, 334n-r. Kropat, p. 46,
believes all the instances to be constructs not recognized by the Mas-
soretes). It also occurs in early Phoenician ('lJ:Ji~i1 Q" "the fourth
day").
In MH the construction has become more stabilised and serves

1 Some instances of this occur in BH, where they are probably due to corrup-
tion of the text (cf. Ges., l27f).
 GENERAL SYNTAX OF THE NOUN 87

with some regularity to distinguish the specialising adjective from the

descriptive one, e.g. P?ii ?1E) "white beans" (as distinct from other
kinds), ii:::J1~ii tDm "a good appetite", and so especially with words
denoting provenience: '~1'~ii r~1n "Idumaean vinegar", 'P?~'~ii 1"
"Italian wine" (cf. the list in Segal, ]Q,R xx 665ff, Grammar, 376).1
h. The same phenomenon occurs also in post-classical and in col-
loquial Arabic (Brock., 132c). In SH, its application was enlarged to
cover newly-formed technical phrases similar to those in which it
was found in the Mishnah: F?l)ii '~l)~ ?~ "to the upper station"
(B. I;Iiyya, Hegyon, f. 1); 'j1j':::Jii p?n "the middle part" (ib., f. 11);
D'P"~? m:lii i1~ "the light stored up for the righteous" (id., Megillah,
p. 22); tD'iE)~ii 1m~ "the separating sphere" (id., clbbur, p. 10); ?tD~
':I1~'Pii "the ancient adage (al-Fakhir, J;(obe?-, iii 2a); ... D'~nii D'?~~~
D'iPii O'?~~~ "hot food ... cold food" (Maim., Dloth, iv 8); niD1:::Jii
Cl'l)iii Cl'i:::J'~ ii'ii' "shame comes from evil things" (Falaquera,
MebaMesh, f. 20); 'iD1j~ii r~ ?l) "upon the human race" (Gerson.,
Dloth, f. 6b). As these examples show, the MH distinction between
specialising and descriptive attribute has been lost again, and these
forms are quite unorganic. Some of them may again be due to copy-
ists, but their very frequency shows that such constructions were
employed deliberately. Perhaps the intention of the authors was to
unifY these technical phrases by adopting for them a construction
that was associated with set phrases.
c. The inverse case, the article before the noun only, is rarely, if
ever, found in normal prose, although it exists in BH (K6nig, 334m, s)
and MH (Segal 184, ]Q,R xx 668). The only instance I have comes
from the ornate prose of B. Daud: '?:::J:::J '1~?nii n~ 1~'O Dii "they
completed the Babylonian Talmud" (J;(abbalah, p. 59).

3. The Article with Numerals

a. When the numeral precedes its noun, and the latter is defined,
various constructions are possible in SH. The normal BH construc-
tion, with the numeral in the construct, occurs: mi:mii mi1~ii niD1?iD
"The three above-mentioned shapes" (B. I;Iiyya, Hegyon, f. 4), niD1?iD~

I Szneider (Leshonenu iii 26) makes two mis-statements with regard to this con-
struction: that it was the common one in MH, and that it was not taken over into
mediaeval literary Hebrew.
88 CHAPTER ONE

i1?tl:i1 I:J'J'ai1 "from these three kinds" (Maim., (lobe?:" ii 25). For some
reason, however, it must have been felt awkward. Probably it was
the usage of Arabic, where the numbers from 11 to 99 take their
noun in the accusative, and those from 3 to lOin an improper
annexation, so that the article always comes to stand before the
numeral: ath-thaliithatu rijiilin or ath-thaliithatu 'r-rijiili.
h. In fact the first of the two Arabic constructions' is the most fre-
quent one in SH: ,??i1 I:J'::JJ,J 'JtDi1 "these two stars" (B. I:Iiyya,
Megillah, p. 116), i1?tl:i1 l:J'p?n i1l7::J..,tl:i1 "these four parts" (id., 'Ibbur,
p. 7), r1lJ1tD? i1tD,?tDi1 "the three ways of expression" (B. Ezra, Saphah
Berurah, f. 2b), I:J'J?a i1l1::JtDi1 "the seven long vowels" (id., Za~oth,
( 2b); l:J'tD..,tD i1l7::J..,tl:i1 "the four elements" (ie. Yesod Mora, p. 2), 1:J'5:l?tl:i1
1:J'::J,m "the two thousand dinars" (B. Daud, (labbalah, p. 70), ,ntl:
nltD:J..,i1 tDani1a "one of the five senses" (Maim., (lobe?:" ii 25), i1?tI:
1:J'?175:l 'JtDi1 "these two actions" (Falaq., Meba~~esh, ( 17a).

c. The third construction; article before the numeral as well as the

noun, seems to be a speciality of Ibn Ezra's: nI'nltl:i1 I:J'JtDi1 "the two
letters" (Saphah Berurah, f. 18a), l:J'i1I::J:Ji1 1:J'?:J?:Ji1 I:J'JtDi1 "the two upper
spheres" (Ha-Shem, f. 2a), i15:ltDi1 nI'nltI: i1tD,?tDi1 "the three labial con-
sonants" (Yesod Mora, p. 18). It occurs only once in BH (Ex. xxviii 10).
B. Ezra's preference for it may be due to some grammatical theory
of his.2

4. Gender qf Nouns

a. One of the most characteristic features of SH (and the subse-

quent developments of Hebrew)' is its laxity with regard to the gen-
der of substantives. This is a phenomenon that pervades all MSS of
all writers. Some of it may be ascribed to the copyists, but these
were on the whole careful enough in other matters. If they err so

I It occurs twice in BH (Num. xvi 35, Josh. iv 4).

2 I could not use Herner's ~ntax der Zahlwiirter im Alten Testament, Lund, 1893.
The grammars, and Konig's article "Zur Syntax, etc." AJSIL xviii, do not deal
with this matter.
I I.e. until the second period of the Haskalah movement. An early Haskalah

writer like Isaac Satanow was still very careless about gender. It is interesting to
speculate whether, but for the Haskalah reaction, the category of gender would not
have disappeared from Hebrew altogether.
 GENERAL SYNTAX OF THE NOUN 89

persistently in gender, this would in itself be significant. However,

textual variants concerned with gender are not more frequent than
others, and mostly the-from our point of view-incorrect gender is
common to all MSS of a work.
From various indications we know that the correct handling of
gender was considered one of the greatest difficulties in writing
Hebrew. Thus Judah Tibbon singles out gender as one of the points
of which one has to take special care (Testament, p. 7). His son Samuel
admits (Preface to the Moreh Nebukhim translation) that gender has
been a frequent source of mistakes with him.
h. Since the gender of inanimate things has in the historical lan-
guages that possess it lost all rational character it may ever have
possessed, it is always one of the first categories to suffer upon any
change of language, whether in an individual or a group. Among
Semitic languages, Ethiopic is the classic example. In Ge'ez the con-
cord forms still exist, but are regularly applied only to words for
animate beings. All other words can quite indifferently be construed
masculine or feminine. 2 In the modern dialects, all gender differences
have disappeared where inanimate objects are concerned. The rea-
son for this is probably to be sought in the fact that Ethiopic was
more profoundly influenced by non-Semitic languages than any other
Semitic dialect.
c. A weakening of the feeling for gender can be observed already
in Saadiah's ornate prose. In lfiwi Polemic, stanza I, we find the
word r.l1
"tree" construed as a feminine to suit the rhyme. There is
no justification for this in any BH or MH text. Although no other
inconsistencies with regard to gender occur in the existent fragments
of Saadiah, the willingness to sacrifice gender to rhyme is of some
significance.
d. The general tendency in SH is to make syntactical gender depend
on the external appearance of the word. This means that all femi-
nine nouns without any feminine ending are more often than not
constructed as masculines. Furthermore only -ah is recognised as a
real feminine ending, while the other so-called feminine endings,
-ith, -ath, -eth, and -uth tend to be treated as masculines. With -uth,
especially, the masculine is practically the rule. It is possible that in

2 Brock, vo!. i, par. 227B(

90 CHAPTER ONE

this last case the masculine gender goes back to late Mishnaic Hebrew.3
If this is correct, then one may further assume that the gender of
words with the other endings mentioned was influenced by those
with -uth.
Examples: 1:l':J'Q~Qi1 n.u1 ,i1i'1 "and this is the opinion of the believ-
ers" (B. I:Iiyya, Hegyon, f. 4); mi1 m'?:JJ "in this exile" (id., Megillah,
p. I); na~i1 1~JmD 1.u "until the truth becomes clear" (id., Schwarz-
Festschrift, p. 25); 1'iD~1i1 n:li1 "the first group" (id., Hegyon, f. lIb);
t']":li1 miDJ1 i1'i1" "and the doubling of the Kaph is ... " (J. I)..iml).i,
Zikkaron, p. 3); 1:l1p ':JJ '?:lQ '?"J na~i1 "truth is greater than all the
men of old" (id., Sepher Ha-Galuy, p. 2); 'O~'Q m:JQ'~i1 .u1" ':J'~iD 'Q
"he who does not know a craft despises it" (ibid., p. 3); '?'1J m.u~
"a great mistake" (Maim., ~kkum i I); ~'i1i1 n''?:lm "that purpose"
(B. Abbas, f. 4b); iDp'JQi1 n.u1i1 "the required knowledge" (M. Tibbon,
Canticle, p. 5); prm '?'1J n,Q'Qn "a great and intense heat" (Falaq.,
Meba~~esh, 13a); 'm~'~Q i1iDp' "it is difficult to find it" (ibid., 14b);
J'~ ~'i1iD niD'JJ '1Q~ "they say about modesty that it is a good thing"
(ibid. 20a); ~~Q:J ~'i1 I:l~ iDEm m1~iDi1 p.uQ 1'pn,? "to inquire concern-
ing the permanence of the soul, if it is real" (Gersonides; Mil&amoth,
f. 4a); J1 np,,?nQ "much dispute" (ibid., f. 26a); 1'J m:l'?nm i1'i1'iD
l:l'iD:J~i1 'JrQ "so that there is some difference between men's charac-
ters" (ibid., f. 28b); ~.u'Q ~'? m.u~ i1.u~ "he made not a small mis-
take" (ibid., f. 46b); nI:J~"iDi1 m "this overlordship" (id., Pent. Comm.,
f. 12b); niD1i1 mJ "in this net" (ibid., f. 48b); l'1~i1 m'?Ji1 mJ "in
this long exile" (id., De'oth, f. 9b); iDp'JQi1 n''?:lm i1i'1 "and that is the
desired purpose" (Gerondi, Sermons, f. 47b); J1 nQ,.u1n mJ 'iD.u'? i1'i1
"Esau might have had weighty complaints about this" (ibid., 30b);
i1n'~ iDJ" mQ'Qni1 "the heat dried it up" (B. Zaqah, Mikhlol, f. 5b);
Jn1 ,,? i1iD.u:J 1iD~ n',m "the angle that is wide" (Zeral).iah, He-lfalu;:"
iii 97).

3 In the Mishnah, nouns with this ending are feminine, except for two cases in

Aboth v: Cl"'!)' ~:::J m,j "exile comes to the world" (Mish. 9, in all Taylor's MSS.);
,:::J '"n Cl':::Jiii m:l1 "the Divine grace of the many now depends on him" (Mish. 18
in Taylor's MSS. B., C., S., some others have ii"'n). As both words are frequendy
feminine in other parts of the Mishnah, the exceptional gender may point to some
dialectic peculiarity in the tradition of this chapter. In Rashi, -uth nouns are mostly
feminine, sometimes masculine, e.g. m'n ii:::J ~ii'tD "that life might be in her" (Gen.
i 20); rii~'" l:liiJ m:l'inii "the sharp edge bends" (Josh. v 2). In Talm. Aram., as
in Syriac (Noldeke, 76a) words in -iithii are always feminine, but these words that
come into Arabic are almost all masculine: tiibiit (tebh6thii) "ark"; malakiit (malkhiithii)
"kingdom"; but ~aliit feminine (~etothii) "prayer"; Mniit masculine and Mniit feminine
(biiniithii) "wine-shop". Perhaps these words were masculine in some Aramaic dialects.
 GENERAL SYNTAX OF THE NOUN 91

e. Similarly, masculine nouns with feminine endings are sometimes

treated as feminines: il:l1i~ ili~Pi11 ili~p il::Jn~il il'?''?il p "thus the
long night becomes short and the short one long" (B. J:Iiyya, 'Ibbur,
p. 7); n'liJClil m~il m~1il "sorrow is (the principal cause of) natural
death" (Falaq., Musar, p. 52).
f. This applies frequently to plurals in -oth of masculine nouns: .. .
nmm m'm Cl'J::Ji::Jil m::J'?i1r~ "the courses of the stars are guided .... "
(B. J:Iiyya, Megillah, p. 114, sing. l'?i1r~); mnJi~Qi m~' mn'Q1 "beauti-
ful and excellent images" (id., Hegyon, f. 12b, sing. 1i'Q1); m'?'?::J ':I~
m'?i1:1 "two important rules" (ibid., ( 13a, sing. '?'?::J); m'?iliil mJJ'?il
m~pi11 "the uncircumcised and obdurate hearts" (ibid., 16a);4 m'?rQil
Cl'?iliil mQi~ '?li mCl'?i~ "the constellations govern the nations of the
world" (B. I:Iiyya, Schwarz-Festschr., p. 2S); mi~lii~ mJ~'nn 1liQ'? "that
his doubts 5 might be allayed" (B. 'Abbas, f. 5a), ni1:lmQ mi~pil 'n~
"the two extremes are opposed" (Falaq., Meba~~esh, ( 20b, sing. il~P).

g. A special case of this are the loanwords from Talmudic Aramaic.

While the Aramaic words taken over into Mishnaic Hebrew lost the
aleph emphaticum, those newly adopted into European and SH keep
the aleph, but are construed as feminines: 6 ~P'~Oil . .. n~ri1 ~iJOil
n~ri1 "this opinion ... this doubt" (= MH p;JO" B. J:Iiyya, Megillah,
p. Ill), n~ri1 ~'~iPil "this problem" (B. Ezra, Yesod Mora, p. 4); il'iln
~P'~O "there exists doubt" (J. ~m}:!i, Ha-Galuy, p. 1); mJi mp'~o
"many doubts" (B. 'Abbas, ( 9a); ~iJOil n~r'? "for this opinion"
(Gers., Mil~amoth, f. 42b).
h. In other cases there appears to be no reason at all for the change
of syntactical gender: ~iiJil n~ Cl'i'::JQi Cl"n il'?~il mii~il "these forms
are alive and acknowledge God" (Maim., Yesode ha-Torah, i S); ~i'?~
mil 'ipli ili~li "the thirteen articles of faith" (B. 'Abbas, f. 9b);
m~~Q:I ri~J Clil i~~ Cl'~iJil i~~ "the other bodies found on earth
(B. J:Iiyya, Hegyon, f. 1); ~'? Cl~ i1iOQi '?JJiQ il'?~il Cl'iJ1il nliJil il'il
"if the emergence of these things were not fixed and orderly" (Gers.,
De'oth, f. Sa). Perhaps these really are copyists' errors.

4 ::J" is feminine in Prov. xii 25.

5 This meaning is late Mishnaic (PirJ:-e R. Eliezer, beg. of ch. xx).
b In Rashi they are not only feminine in construction n~r ~'JiO "this problem"
(Yoma l3a); ir ilP'ElO "this doubt" ('Arakh. Bb); ili::JO ilm~ "that opinion" (Niddah
7b) but are also inflected as feminines in -ah: in'rDiP "his question" (Be?ah lOb).
::J"il ni::JO "an unfounded opinion" (Baba Bathra l2b). ilP'ElO occurs already in a
letter of the Palestinian Gaon Solomon b. Judah (Assaf, Gaonica, p. 94).
92 CHAPTER ONE

i. These outright violations of the gender rules are, however, rare,

and even in those cases where gender and form contrast with each
other, the correct gender is very frequently used. In spite of the
statement of Judah Ibn Tibbon quoted above/ the influence of the
corresponding Arabic words is hardly a noticeable factor in the irreg-
ularities of gender. There was simply a certain amout of wavering,
which favoured analogical formations. Probably it was the general
spread of grammatical knowledge which prevented complete law-
lessness in this respect, such as existed in Ethiopic, and in early
Modern Rabbinic Hebrew.

5. Concord with Regard to Number

a. The usage in SH is the same as that in BH and MH: predicate

and attribute agree in number with the governing noun. The Arabic
grammatical rule, demanding singular predicates and attributes in
many cases although the governing word is in the plural, appears
to have had no influence on the original Hebrew writings. In some
cases there are irregularities, e.g. ~'~~ ~i' 'i::J1 iir r~ "these are not
the words of a God-fearing man" (B. J:Iiyya, Megillah, p. 49); ~1ni1 '~'
~i1'? i1p ~1i1 1J'? "the days of heat for us are cold for them" (id.,
'Ibbur, p. 7). These are probably due to inadvertence on the part of
the author, who may have thought of the Arabic qaul (sg.) in the
first case, and of some word like "summer" in the second.
h. Bar J:Iiyya is altogether careless about number. He also uses some-
times plurals by constructio ad sensum, although singular is required,
e.g.: 1'? ~'im ~1~i1 '~' iElO~ "the number of a man's days are decided
for him by fate" (Megillah, p. 111), ~mlic!)::J 1liC!) m~ 1n~ '?~ "every
one of them were mistaken in the same way" (Meshi~ah, p. 3).
c. It is quite another thing when collective nouns are construed with
the plural, a procedure that is common not only in the earlier stages
of Hebrew (Ges. 148b-f; Sega1, 448), but also in other Semitic lan-
guages (Brock., 51 b). Examples: ~'li11' ~'?1lii1 '?~ "all the world know"
(J. Tibbon, Testament, p. 11), 1'?1~ 1~ ~'?1lii1 ::J1i 1::J ~'li1c!)~ i::J1i1 ~1i11
"and that is the thing in which almost all! the world are mistaken"

7 See above p. 56.

I The idiom is an Arabism: aktharu 'd-dunyii au kulluhii, lit. "most of the world, if
not all of it".
 GENERAL SYNTAX OF THE NOUN 93

(Maim., (lobe;:;, ii 25), i10i1 1mn ':J 1?'~' ~?1 1?'l71' ~? ilD~ 1mni1 "van-
ity that do not bring any benefit or help, for they are vanity" (ibid.).
The following example is interesting in showing how SH authors
derived their justification for such constructions from BH models:
1O:J !:l'J~ 1? 1'i1' llDn~ !:l':J?1i1i1 !:ll7i1 i~1D mmm~ m1m ?:J1
!:ll7i1 !:ln1~ "if one follows the practices of the other 'people that walk
in darkness', one will have sons like those people" (Maim., De'oth,
v 5); the quotation is from Is. ix 1).

6. The Construct State

a. One nomen regens cannot depend in SH on more than one

nomen rectum any more than in BH (Ges., 128a) or MH (Seg., 382).1
Circumlocution must be employed: 1'n11~n1 1':I1Jl7n1 !:l?1l7i1 'i1i1ii1 ?:J
!:l'l7ii1 "all the evil thoughts and pleasures and lusts of the world"
(B. I:Iiyya, Hegyon, f. 17a) where !:l'l7ii1 applies equally to all three
regentia; 1'P0l71 1'n1~1Cl1 !:l?1l7i1 nmJl7n !:l1l7JO ~?ID "whom the pleasures,
attractions and occupations of the world did not distract" (G. Tibbon,
Ecclesiastes, Comm., f. 16a).
h. On the other hand, several regentia can govern one nomen rec-
tum. This is also possible in BH, especially in its latest forms (Kropat,
p. 55, Konig, 276b) and in MH, e.g., !:l'i~~1 !:l'~?n iClpi1 "the burn-
ing of fat and limbs" (Ber. i I); i1lDom m)~i~ '01?lDn "payments of
the fourfold and fivefold value" (Sanh. i I). But while in the earlier
forms of the language the words were always "such as naturally went
together or were more or less synonyma" (Konig, loco cit.) and the
sequence was always cumulative, it can be distributive in SH and
broken by particles: ~ipOi1 m10~ ~? m'~... n?p?1p0i1 m10~i1 n1~1
'10?ni1 ~?1 mIDOi1 ~?' "and that corrupt dogma is neither the creed
of the Bible, nor of the Mishnah, nor of the Talmud" (Abr. Maim.,
(lobe;:;, iii 17a). This seems to be under the influence of Arabic where
all kinds of words can come between the recta (Reck., 171,3). 2
c. The epexegetic genitive to convey an attributive idea occurs in
BH, mainly in poetical passages (Ges., 128r), and in Arabic poetry

1 Such cases occur in the Karaite Paytanic prose of Hadassi: lt1i1i'P 'i'i.l'?m 'iDJtIi
"the men and scholars of Kairouan" (Eshkol ha-Kopher, Alphabet, 224 ~).
2 They are rare in early Arabic, but become increasingly frequent in post-clas-
sical prose (Fleischer, Kt. Schr., i 624).
94 CHAPTER ONE

(e.g. ~iilibu ayyiim'i "my righteous life", c( Reck., 73.2). It also appears
occasionally in SH, where it seems, however, to have lost its poeti-
cal flavour: C1i1':Jii .llii:::J n:J~~ "to walk in their evil ways" (B. 'Abbas,
f. 7b); rmJip~i1 '~ino nnEl' "it explains the difficult passages" (Falaq.,
Meba#esh, ( 15b);3 C1itlii1 n'~~' C1i1:::J iiDtIi C1'iD.ll~i1 iiDi' 'n~~niD liOn:::J
C1~i.lli1 i1r:::J "please, teach me the righteous deeds by which a man is
successful in this world" (ibid., 18a). The plural of the retrospective
pronoun in the last example shows that such constructions were felt
to be fully equivalent to noun plus adjective.

7. Circumlocution Genitive

a. Circumlocution by le- is frequent in BH (Ges., 129), and occurs

occasionally in MH (Seg., 384, fuller list in ]Q.R xx 725f). In spite
of the fact that it is also a common feature of Arabic (Reck., 135.4),
it is very rare in SH. The only example I have is from Gersonides;
C1'intlin~i1 n~p~ '.ll':n n.lli 1t1i:J:::Ji "and in this connection, there exists
a fourth opinion, of some recent scholars" (Milbamoth, f. 4b).
h. Circumlocution by means of ~iD (Seg., 385-6) is very frequent,
and appears to have been fully equivalent to the construct state: It
exists both with nominal rectum: n.lli ~iD C1'i:::Ji "words of good
sense" (Maim., I;lobe::" ii 26); TiElO ?iD ni:Jr~i1 "the list of your books"
(j. Tibbon, Testament, p. 4), and with pronominal rectum: l li.ll:::J
lii.lli1 1~i~iD "in its proper entry in the 'Arukh" (Ibn Parl).on, xxii);
i1~iD 1i~itlii1 tliii1iD i1n'iJ "her body, which is her palace" (B. Ezra,
Yesod Mora, p. 16); C1:J~iD ni~tliiDi1 'piipi ~:J "all the details of your
questions" (Maim., I;lobe::" ii 26).
c. ~iD with suffixes is also used substantivally (Seg., 406), e.g., i':::JtIi
"if both his father
i':::JtIi ~iD 1:J intlii i:::Ji ~iD ntli n'J~ tliiD~:::J C1'tliiDiJ i:::Jii
and his teacher are carrying burdens, he must ease first that of his
teacher and afterwards that of his father (Maim., Talmud Torah, vi).

3 This is an imitation of BH constructions like t:n:ltl: 'p'?n "smooth stones"

(I Sam. xvii 40); with the article ltl:;;n 'i'll;; "the young sheep" (Jer. xlix 20).
I The construction with pronouns goes back to the frequent BH '? iilltl: (Seg.,
409; Kbnig, 28i, 0). '?ill with pronoun-suffix occurs in Ben Sira xiii 5, but is still
comparatively rare in the Mishnah, where only 40 instances can be found (cr. Segal,
JQR xx 732f).
 GENERAL SYNTAX OF THE NOUN 95

d. The circumlocution of the genitive with anticipation (Seg., 387-8)

is extremely rare. As in MH, the nomen rectum must not have the
article. This goes so far that even the article of stereotyped expres-
sions is dropped mpo '?ID in:lID 'nOipnID 'n~i "and after I have first
of all expressed praise of God" (B. I:Iiyya, Megillah, p. l), although
~ipon in the sense of "God" always has the article in MH and the
liturgy.
 CHAPTER TWO

THE PRONOUNS

8. Personal Pronouns

a. With regard to the lexical forms of the personal pronouns, SH

agrees with BH against MH in employing i:Jn:Jtol: "we", cntol: "ye"
(masc.) and cn "they" (masc.) for i:Jtol:, 1ntol:, 1n (Segal, 68, 70). It seems
probable that in this it did not go back to BH of its own accord,
but followed the usage of the liturgy (cf. Segal, ]QR xx 655, note 3).
For the first pers. sg. ':Jtol: is employed almost exclusively. Here SH
agrees with MH, but also with late BH (Kropat, p. 75) and the
liturgy.
h. As in late BH (Kropat, p. 1) and MH (Seg., 404), the personal
pronoun is sometimes expressed with the perfect or imperfect "to
give added emphasis to the statement as a whole" (Kropat): 'ntol:::liD~i
iY"I''?O m'iQ'? ':Jtol: "and when I came to the city of Salerno" (B. Par-
}:lon, p. xxii); 1ElOn i1r ':Jtol: 'i1i::l'n ..• 1',l]~n ':Jtol:i "and I, the young
one, ... I composed this book" (J. ~ml:Ii, Ha-Galuy, p. 2).
c. It is especially frequent when some phrase or clause intervenes
between subject and predicate, and there is need to remind the
reader to which subject the predicate belongs. It is then repeated
both before finite tenses and the present participle: Ci'?iD ,'? 1i1i:Jn 'n'?tol:
Ci'?iD cn'? rtol:iD C',l]iD1n '?,l] 1m tol:in "my Lord, who gives me peace,
He decrees upon the wicked ones that they should have no peace"
(B. I:Iiyya, Hegyon, f. 12b); 1iDinn ntol: O'Eltol:Q tol:in r1tol:n '?,l] r'ElQ 1itol:niD~
i1:JQQ "when the light spreads over the earth, it dislodges the dark-
ness from it" (id., Megillah, p. 27); cn'? 11"1' tol:in nQ~n en'? 1miD CiDm
nQ'?iD Cn1i~iDQ "and the Lord, who gave them wisdom, may make
their reward complete" (B. Ezra, Yesod M-Ora, p. 9); 1i,l]QiD ':JEl'? iiQ'?
... iiQ'?'? i1rm CJ1~ '?~ iiQ'? tol:'? cm 1~El:JiD i,l] p'i~n "and they
studied with Simeon the Just, until he died, but they had not learnt
enough, so they went on studying" (J. Na}:lmias, Aboth Comm., f. 3b;
perhaps this should rather be taken as a circumstantial clause).
d. The ethical dative (Ges., 119i; Seg., 361 iii) is not very frequent,
e.g. tol:'iD:Jn '?~tol: i'? 1'?' "let him go to the prince" (Maim., Talm. Torah,
 THE PRONOUNS 97

vii 10). Its use is extended to a nominal clause in the following

instance: 1n~ iD1~i1 l'? im "and this is another difference for you"
(B. I:Iiyya, Hegyon, f. 7). This instance seems, however, to be quite
isolated, and the ethical dative becomes very rare with later authors.
e. In some cases, anaphoric personal pronouns precede the nouns
to which they refer: J,P.Ii' np''? i1'i1 ~'? nI'1.1ii1 '?J 1'O'~ .li1" i1'i1 ,,?'~
,n' nI'n~ 'niD 'J'J~ "if he had known all the forbidden degrees of
relationship, our father Jacob would not have married two sisters at
the same time" (B. Ezra, Yesod Mora, p. 6); nlJ,C!l nI'D 1i1''?.liJD "D'?J
"when he learns from those who practise them, good qualities"
(B. Abbas, f. 4b). Instances of this occur in early Arabic (cf. Reek.,
139,6) but not in Middle Arabic prose. The origin of these con-
structions seems to be accidental, perhaps also due to the influence
of European languages, where such constructions are nothing unusual.

9. Demonstrative Pronouns

a. m "this", fern. ,r, n~r, pI. ,,?~, i1'?~

This pronoun is constructed in two ways: either the noun pre-
cedes and the pronoun agrees with it in determination (mi1 iD'~i1 or
m iD'~). This is the common method of BH and MH (Segal, 411).
It is the only one employed by Mena}:lem and the one most fre-
quently found in the works of B. I:Iiyya, as well as in Maimonides'
Mishneh Torah. Instances occur in all writers, though very rarely in
those of the second period. But Gersonides, who employs consis-
tently the other construction, has occasionally phrases like i1ri1 1~0i1
"this book" (Mil~amoth, f. 48b). In the sermons of the Talmudist
Nissim Gerondi, suffixed zeh is almost as frequent as prefixed.
h. The typical construction of SH is, however, that in which indeter-
minate zeh precedes the determinate noun: iD'~i1 m It is still rare in
B. I:Iiyya: 1'.lii1 m '?.Ii "for this sin" (Megillah, p. 25); nlDJni1 ,,?~ "JJ
"the honour in which these sciences are held" (Meshi~ah, p. 2); nlJ'?D
Cl'iDJ~i1 ,,?~ "from the hearts of these men" (Megillah, p. Ill).
With Ibn Ezra, prefixed zeh is already the rule: nlD'PDi1 i1'?~ '1~'O
"the scribes of these localities" (Saphah Berurah, f. 5a); i1.1i':I'i1 n~r "this
effort" (Yesod Mora, p. 1). His contemporary B. Daud also employs
mostly prefixed zeh, although he writes ornate prose: 1m0i1 '"i1'i1 m
''?iD "this Jew, my scribe" (flabbalah, p. 72); '?"m OJi1 m "this great
miracle" (ibid., p. 54).
98 CHAPTER TWO

In the writings of the ~mDis and Tibbonids, the prefixed position

IS the normal one.
Maimonides, though generally employing enclitic zeh in his Mishneh
Torah, has the prefixed form in his letters, e.g., IJ'...,:l,i1 i1'?~ '?J ...,n~'
"and after all these events" (Teshuboth, p. lix); ~:li1 IJ'?,.lJi1! IJ'?,.lJi1 m:l
"in this world and the world to come" Cflobe?, ii 26).
It would be idle to quote here instances from authors of the sec-
ond period.
c. There are some cases of the prefixed zeh in BH (Ges., l26d, note
2; not mentioned by Konig). These are often emended or explained
away, but may actually represent the older usage (Brock., 40e).! In
MH it probably did not exist. 2 It appears in European texts from
the 11 th century. There are some isolated instances of it in Josippon
(e.g. ID'PrJi1 m col. 64), but the textual tradition of that work is too
uncertain to warrant their originality.
It is already fairly frequent in ADimaa~, e.g., IJ','oni1 i1'?~ 'rJ':l "in
the days of these pious men" (Chronicle, p. 3); i1.IJIDi1 ,r:l' r1.IJi1 ,r:l "at
this time and in this hour" (ibid., p. 8, in rhyme).
In Rashi, it occurs occasionally: P.IJi1 mJ "in this fashion" ('Ab.
Zarah 40b); 1rJri1 m l,m "within this time" (Baba Mqi'a, 99b).
It also is frequent in the Crusade Chronicles of 1140: ...",i1 m "this
generation", IJ'rJ' 'jlD i'?~:l "in these two days" (Kreuzziige, p. 2), and
seems to be the rule in the Sepher /fasidim (ab. 1200), e.g. IJiPrJ in'1~'?
IJID ...,mOi1 mID "to that place where this merchant is" (no. 826). 3 In

I It appears that with regard to the position of the demonstratives the Semitic
languages fall into two groups marked by the two extreme cases: Assyrian with con-
sistent posterior position, Ethiopic with just as consistent anterior position. Arabic
follows Ethiopic, except in proper names and nouns with possessive suffixes (Reek.,
149.1). In the oldest Aramaic texts, the demonstratives normally follow the noun,
but in Biblical Aramaic and the Elephantine Papyri, it occasionally precedes (Rowley,
Aramaic if the O. T, p. 105). In Syriac and Mandaic, both constructions occur with
equal frequency, in the Neo-Syriac of Urmia it always precedes, but in Torani it
can also follow (Brock., 40d). Hebrew seems to have gone through the opposite
development, from prefixed to posterior position. It is interesting that MH agrees
here, as so often, with Phoenician, where the demo pron. always follows (Harris,
Phoen. Gramm., pp. 53-4).
2 The cases of prefixed zeh enumerated in Seg. 411 (to which DiMu~, 79 adds
the reading of the Pal. Talm. [Pes. vi I] i1r~?llj i1~?i1 11 [as against i1r~?llnj 11 i1~?i1
of Bab. Pes. 66a]) are almost certainly to be explained as asyndetic relative clauses
dependent on m (as Seg. himself does in par. 477, but in Di~du~, loc. cit. he thinks
this improbable). n1~r:l1C!l 1?~ ?ll (Nazir, vii 2) is to be translated "for which cases
of uncleanness ... ?" (er. Segal, 419 end).
3 But in stereotyped phrases, zeh follows, e.g. mi1 Cl?111i1. zeh is very rare in the
SI:I, 1n1~ being the commonly employed demonstrative.
 THE PRONOUNS 99

the East, we find prefixed zeh in a fragment from a letter of the

Yemenite communities to the Gaon Samuel b. Eli (1164-94 pub!.
by Harkavy, ZFHB ii 126) and in the commentaries of the Constan-
tinople Karaite Aaron ben J oseph (1250-1320).
On the other hand, prefixed zeh is entirely absent from Spanish
poetry or ornate prose before 1200. We may, therefore, assume that
this construction arose somewhere in the Palestinian orbit, be it in
Europe or in the East, but more probably in the latter, in or before
the 10th century. Its origin was most probably not in an imitation
of the exceptional Biblical instances, but either through the influence
of the Jewish Aramaic prefixed ri
(Dalman 17; Margolis 9.1), or
through analogy with 1rmt In either case, however, the constant use
of the definite article with the noun is not satisfactorily explained.
Its great popularity in SH may be due to its similarity with the con-
struction of Arabic hiidhii. 4
d. The wavering between the two constructions also produced, it
appears, the strange hybrid form in which determinate zeh nii1 :l1n:J1
IQ1~ "and this verse says" (B. I:Iiyya, Hegyon, 13a) follows an inde-
terminate noun. This may be a copyist's mistake, but the same con-
struction recurs in Tobia ben Eliezer (ca. 1100, see Introduction,
p. 48), e.g. mn p10~ ~tll 1tllliQ (Le~a~ Tob on Gen. i 1), where, how-
ever, the normal MH m p10~ seems to be more frequent.
e. SH goes further than Arabic in employing prefixed zeh even in
contexts where hiidhii would have to come after the noun: with nouns
defined by a genitive and with proper names without the article
(Wright, ii 277c). E.g. '''1n~ n1'n1~:1 ~~~ ni1n' 'I m 1:1'i ~~ "This
R. Judah dealt only with the semi-vowels" (B. Par}:lon, p. xxii); nr1
i1:ltllQn n~ 1:1'n tllnpn 1J:11 "and this holy Rabbi of ours composed
the "Mishnah" (B. Daud, J:labbalah, p. 57); n.!]in "J'!] n~~ "these
people devoid of sense" (Abr. Maim., J:lobe?, iii 18a); nQ~tll 'I m 'l:1i
"the words of this R. Solomon" (ibid., 19a).

4 The prefixed z:.eh is one of the few SH features that have penetrated into pre-

sent-day literary Hebrew. It is-apparently only since recent times-employed where

a ::.eh following the noun break up an attributive sequence, e.g. 11n':::l ':::l1lli1 iEl:J:::l
ri~i1 'j:::l '?iD C1:m C11'i1 1ll i1:::l C1"1iiDiD n''?:J'?:Ji1-n'ni:::lni1 i1i1~i1 1r:::l, "especially in
the Arab village, in that social and economic complex in which until to-day the
majority of the population dwell" (A. Boneh, i1'?:J'?:Ji11 ri~i1, 1938, p. 7); C1'jEl i1~:J
iiD1~i1 ~ll~ m'? iiD1~'? 1'i1 "how many aspects happiness has, that little happiness of
mankind!" (J. Zar/:li, C1'El1~i1 ii1, 1940, p. 43); n'j1:J~:::l i1iD~i1 n~r1 "and that woman
in the motorcar" (ibid., p. 95).
100 CHAPTER TWO

f. There is in all post-Mishnaic forms of Hebrew very litde difference

between the various demonstrative pronouns. In some styles (e.g. the
N.W.-European of the c','on i~O) there is a strong tendency to
employ only one, mosdy 1ni~ etc., in all contexts. In the SH texts,
practically all the EH and MH demonstratives are employed, with-
out clear notional differentiation. 5
g. The MH demonstrative ~1i1 without the article (Seg., 73) is found
in SH only in the stereotyped phrase C1':l 1:l (Mishnah Shabb. i 4,
and often), e.g. Ci1'in~ C1':l 1::1 C'~i:lJi1 C'OiDi1 t"J1.11 "the birds of heaven,
that were created after them on the same day" (B. I:Iiyya, Hegyon,
f. 7).6 The attempt of Saadiah to revive this pronoun in rm1iD'?:l C:l
"in those languages" (lfiwi Polemic, st. 38) met with no success.
h. U,]r'?i1 (Seg., 74) is very rarely used. E.g. i1pmmi1 '?1.l1iDO:l ~1:l'?
m'?i1 i~i1 "to enter this narrow path of translation" (S. Tibbon, pref-
ace to the Moreh).
i. ~1i1i1 generally refers to something which has just been mentioned
and corresponds to the Arabic dhiilika. It is in this parallel to the
EH use of this word (Ges., 136a), but since all other pronouns can
stand in the same position, its notional range is not the same. It is
fully synonymous with 1ni~, but rarer. Examples: 'i1' C'i1'?~ iO~'1
C'P tl;1i1i1 i1~i1 i1'i1 'i1' n'?O::1 iOtl;Oii nn~iD 11no ... '1~ "and God said:
Let there be light. . . since he began the utterance with the words
'Let there be', that light is permanent" (E. I:Iiyya, Hegyon, f. 6); ':J
~1iii1 ~.l10::1 'J:l ,'? nJii C:JO.l1 '10.l1'? ,'? iD' ~.l10 'J~ "for I have little
time to stay with you, let me enjoy that little, my son" (J. Tibbon,
Testament, p. 11); nio:Jn ~1iiii iDi::lOii 1::1 tI;~01 ... C'iDi::lOii n~p 1iO~ p1
ni:li "and so said some commentators ... and such a commentator
found in it many sciences" (S. Tibbon, Eec!., f. 15b).
~1iiii follows the noun, but it may come between the noun and
its attribute: C":l:JJii Ciiii C'J1'~'? "To those honoured gendemen"
(Maim., (lobe;::, ii 27); 'ii'?~ii ~1iii1 ,:l,'? "for that divine utterance"
(Gerondi, Sermons, f. 30b).

5 It seems almost as if only a colloquial speech, with requirements of rapid intel-

ligibility, could maintain a distinction between demonstrative pronouns. The notional
of these words, even in colloquials, is the reason for their tendency to cumulative
composition and their great variation from one Semitic language to the other.
6 C1':::l 1:::l is frequent in Rashi, e.g. Gen. xv 15.
 1HE PRONOUNS 101

k. 1mtl: precedes the noun, which may indifferently have the article
or be without it. Its use is exactly the same as in MH (Seg., 417),
but the form without pronominal suffix (Kropat, 2ff; Konig, 270f;
Segal, loc. cit.) is not found in SH. Examples: 1mtl: P'EJO' tI:'? '~ tI:~01
C1t1: 'j:::J'? iEJOii "He found that book would not fulfil all the require-
ments of the public" (B. Pari:lOn, p. xxii); cmtl: m'o'?:::J Cii'O' '?~ 11:::JtI:
C'iEJO "all their days were wasted in studying these books" (Maim.,
.(lobe;?:" ii 25), mm 110ii'? iEJOii 1mtl: i:::J'n tI:'? "he did not write that
book for the common people" (Abr. Maim., .(lobe;?:" iii 17b).
1. In substantivised form, only m appears. 1mtl: seems to occur as
substantive in translations only, to render the Arabic dhiilika, e.g.
1:::J'?:::J iiO~nii tl:1ii Imtl:O n''?~m "the purpose of those others is purely
theoretical" (Averroes' Summary if the Republic, translated by Samuel
b. J udah of Marseilles, f. 1).

10. Sentence-anaphorics

a. The MH sentence-anaphorics, or pronouns referring to the "neuter",

1~ and pi are rarely employed in SH. Examples: 1j~i~1ii 1~'? tI:?
iin,I] "not this is what we need now" (Maim., .(lobe;?:" ii 25); 1j'?:::JP p
1j'm:::Jtl:o "this we received as a tradition from our forefathers" (B. Ezra,
;:.a~oth, f. 6a).

h. The BH fern. ntl:f (Ges., 122q, end) is even rarer: ntl:f 'intl:1 "and
thereafter" (B. I:Jiyya, Hegyon, f. 17a). tI:'ii (Ges., loco cit.) seems not
to occur at all in SH.
c. The normal sentence-anaphoric is m, corresponding to Arabic
hiidhii or dhiilika (Reek., 191 ):2 1jOO tI:';:m m iitl:itD l1no "since he per-

1 Segal (Wduk, 289) quotes cases of clearly pronominal use of these two words
under the heading of "adverbs of manner and degree". Nor does their pronomi-
nal character (which is also pronounced in Modem Hebrew) seem to be mentioned
in any other grammar or dictionary. The following examples from the Mishnah
should give proof that these words are indeed pronouns: ~1n l~ m'?m "0 "This
is the order of inheritance" (Baba Bathra, viii 2); m:l'? l~ '?~ "why all this" (Menal}oth
x 3); 'n'?~iD l~ ~'? "not this I asked" (Taanith iii 8); l~::l 1:l'~1' !:l"::ln ''? iD' "some
of my friends want it" (Sanh. vii 10), pr:l mn~ "less than this" (Meg. i 3); 11~nl p'?
n'?'nnr:l "that is what he intended at first" (Temurah v 3); P nlr:l !.;i.l) "for this rea-
son" (Baba Bathra, v 9). Both are frequent in Rashi in this function (cf. Avineri,
Hekhal, pp. 260 and 265).
2 It is to be investigated whether this feature was due to Arabic influence, or
goes back to late Mishnaic Hebrew. If it originated in Spain-Provence, Romance
102 CHAPTER TWO

ceived this, he concluded ... " (B. I:Iiyya, Schwarz-Festschr, p. 2S);

C1':1i C1'J'Jl7 'l7 m~ri1 i1Q:Ji1:1 ~~1Q i1n~ m 11J:J1 "and the like of it you
find in this science with regard to many matters" (id., Megillah,
p. 113); pn' ~, i1r1 i"1':1 :1n:Ji1, '1~i i1,Qi1tD 1:1tDn "they' thought the
word should be spelled with a Yod, but that is out of the question"
(B. Ezra, Saphah Berurah, f. 7a); m ':J 'ni:Jrm "and I mention all
this" (id., Yesod Mora, p. 6); m n'1r :11tDnn '~1 "and do not imagine
anything else" (B. 'Abbas, f. 3b); m C1'l7i1' C1i~ 'J:11 "and people know
it" (J. Tibbon, Testament, p. 10); m 'J5:lQ1 "and therefore" (Abr. Maim.,
!lobe;::., iii 17b); m ,:J nl7i' "to know all this" (Falaq., MebaMesh,
f. 13a); C1'Ql75:l' m i1~iJtD 1Q:J "as we sometimes see it" (Gers., Milba-
moth, f. 42b); m 'l7 i1i1'1 "and this is proved by ... " (id., Pent.
Comm., f. 12a); ... tD n1~' i1r1 "and this a sign that" (ibid., f. SOb);
m' "therefore" (frequent in second period writers, esp. Gers.); m:1 i1~i
"he means by this" (Gerondi, Sermons, f. 47b).
d. The Arabic anticipatory sentence-pronoun, the rjam'ir ash-sha'n
(Reek., lS4.5) seems not to have penetrated into original Hebrew
texts. There is a case of anticipatory pronoun, however, which is
quite independent of Arabic usage; the pronominal suffix after the
negative copula, if this precedes its subject at some distance: '5:l'
... tD il7 i:1i FQii1 n1:J:1 '5:l1J 1J'~tD "because nothing takes shape in
the imaginative faculty, unless ... " (Sepher ha-ca;:;amim, ascr. to Ibn
Ezra, f. 5); C1"ni1 'l7:11 nQ1~i1 'n'1r 11r'J 1J'~ "nothing partakes of nour-
ishment except plants and animals" (Falaq., MebaMcesh, f. 12a). The
possibility cannot be excluded, however, that these are nothing but
instances of confusion between r~ and 1J'~, cf. below, par. 25.

11. Indifinite Pronouns

a. There is a notable tendency in SH to make good the lack of spe-

cialised and unambiguous indefinite pronouns in the earlier forms of
Hebrew (Seg., 434). This urge explains the borrowing of the typical
Arabic indefinite suffixed -mii (Reek., 150, Brock., 149k) in the authors
of the second period: i1n'?:Jn ?~ i1Q C1i~ l7'J' ~, "no man whatever

influence may have played some part, as in all Romance languages sentences are
of the masculine gender, and sentence-anaphorics masculine. Talmudic-Aramaic in-
fluence is excluded, as in that dialect sentences are mostly feminine (cr Schlesinger, 35).
 THE PRONOUNS 103

will grasp its final meaning" (B. 'Abbas, f. 4b); 'Jm,n m'~DD i1D m~D
"any injunction of our law" (ibid., 12a); fit1;i1D i1D CnpD:::J "in any
place on earth" (Gers., De'oth, ( 6b); i1D 1El,tI;:::J "in some way or other"
(id., Mil&amoth, 28b, and often); In Gersonides it is, against Arabic
usage, also employed with the plural: i1D Cl'iYTD Cl'pninD Cl'IDJtI; n~p
"Some men refrain from some things" (ibid., f. 8b).'
h. The late MH2 Cl'1D "any" is employed in the sense and function
of Arabic ayyuh: m.lJ~ ,tI; i1:IJID Cl'1D Cli1'? i1t1;i' Cltl;, "and if he sees them
make any mistake or error" (B. J:Iiyya, Megillah, p. 3); Cl'pO.lJnDi1
i1:JtI;'?D Cl11D:::J "those who practise any craft" (ibid., p. 114); i11D.lJ tI;'?,
,'? 'i1t1;ii1 tI;'?lD Clitl; Cl'ID'? m'?ID'? i'lD "and he did not compose a poem
to send to any person without letting me see it" (J. Tibbon, Testament,
p. 9); lD~.lJ:::J lD,n'D Cl'ID:::J ID'Jin Cltl; "if you feel any pain in your own
body" (ibid., p. 12).
c. '?:J is employed in the sense of "any" in negative clauses only, as
in BH (Ges., 152b):3 'mJ~p 'JElD Cli1D iElO '?:J i:::JtI;' tI;'?lD "that no book
of their number should be lost because of its small size" (D. I>iml;i
on Hosea, p. 9).
In contrast to both BH and MH, it is also used substantivally:
Cli1'? '?:J 1'tI; Cli1'~Elm Cli1'iDnD '?:JD "they haven't anything left of all
their treasures and precious possessions" (Falaq., Meba~~esh, ( lla).
d. The plural indefinite "some" is expressed, as in MH,4 by ID' or
•.. ID ID', literally "there are some who": Cl'iD,tI; Cli1D ID' "some of
them say" (Maim., flobe~, ii 25); lD'i'El Clm'~'? 'lDi'ElID Cl'J'lDtI;ii1D ID'

I The indefinite -mah was taken over into Modem Hebrew, where it serves to

form certain stereotyped expressions e.g., i1r.:l-i:::li "something", i1r.:l-l'1ir.:l:::l "to some
extent".
2 Neither the indefinite nor the much more common negative 01iD is recorded
by Segal. Indefinite 01iD is found in TanJ:!uma, Pi~ude iii.: i:::li 01iD i:::l1i ~1i1iD:l1
"and when he says anything". Its origin may be from the early MH use of the word
to mean "a minute quantity", e.g., ,l)il l'1:::l:liD 01iD ••• fiiD 01iD "a minute quantity
of vermin, of semen" (Bab. Middah 43b). It is very frequent in Rashi, cf. Avineri,
Hekhal, p. 338. The negative 01iD is also absent from Talmudic and early MH texts.
There is no doubt that both uses of the word came into SH from N.W.-European
Hebrew, but the etymology and history of this word requires further investigation.
3 Its use is much wider in MH (Seg., 435).
4 E.g., r'?'nJr.:l1 r'?ml iD' "some both inherit and bequeath" (Baba Bathra viii I).
i1'?'l 1l'~iD ~J1iD iD'1 i1'?1J ~J1iD iD' "some enemies are exiled, some are not" (Makkoth
ii 3), and the frequent O'ir.:l'~ iD' "some say", 1ir.:l~iD iD' "some said". (Segal does
not record this usage).
104 CHAPTER TWO

"some ancient scholars give an explanation for their form" (J. ~m1:ti,
Zikkaron, p. 2).
Under the influence of Arabic ba'tj, which can refer to one as well
as to several individuals (Reek., 91.2), [... iD] iD' also is applied in
the sense of "one who" "n'i1 15:l0 '~1P' 15:l0 1:::l'niD C':l'~:Ii10 iD" "and
one of the Geonim wrote a book and called it The Book of Oneness"
(B. Ezra, resod Mora, p. 3).

e. The most common expression for the plural indefinite is the MH

n~po (Seg., 437v), more frequently in SH reduced to n~p: 1pl1i1
cn~po '~1iD i1:~1 ~'i1 "the main point is that some people have seen
it" (J. ~m1:ti, Galuy, p. 1); r1IiD110i1 n~po ~iD5:l "the literal meaning
of some Midrashic passages" (Abr. Maim., (lobe;::., iii 18a); nl1' '5:l,
C'p'p'o n~p "according to the view of some grammarians" (B. Ezra,
Ha-Shem, f. 2b); "1'5:l'0' :::l,nJi1 '1:::l' n~p m:::li1 ':l'0 ':liD "two ways to
understand some scriptural expressions and stories" (M. Tibbon,
Cant., p. 5); C'iD15:l0i1 n~p "o~ p' "and so said some commentators"
(id., Eccl., f. 15b); C'1n~n0i1 n~p' '11':::l1 nl1' "a fourth opinion, held
by some of the recent writers" (Gers., Milbamoth, f. 4b); 10 n~p
c"ni1 '~1~ C':l'iD~1i1 "some of the ancients enjoyed a long life" (id.,
De'oth, f. 6b); ,"r1 n~p iD1':::l i1r '10~ 1:::l~' "and some of our Rabbis
have said this in a Midrashic form" (id., Pent. Comm., f. 12b).5
f. n~p became so much identified with ba'tj that in the authors of
the second period it is, like the latter, also used in reciprocal expres-
sions (Reek., 9l.4): cn~p ,~ cn~po l5:li1m, r1I,'O'i1 11:::l~ "it is the
nature of the elements to change into each other" (B. ?:aq;a, Mikhlol,
f. 23a). It imitates the construction of ba'tj to such an extent as to
drop the suffix pronoun of the second n~p: cn~p 1'0i1i1 ':l':l11 1'0'
n~p:::l "to regulate the affairs of the common people with each other"
(B. 'Abbas, f. 6a = Arab. ba'tjahum bi-ba'tjin); r1I1'~i1 ,,~ ,:::l'P 10,ni1
n~p r1Il1~O~:::l cn~p "matter received these shapes by transition from
one to the other" (Gers., Pent. Comm., f. 12b); n~p cn~p C'1r110 C'1:::l'
"words that contradict each other" (ibid.), but cn~p' cn~p 'P'l1' ~"
"and should not crowd each other out" (ibid., f. 80b). The sentence:

5 The word, which occurs in late BH (Neh. vii 70, Dan. i 2), is a loan from the
Aramaic n~p-1Q (Elephantine texts xv 3, xxxvii 4; Dan. ii 42). The simple n~p for
n~PQ is again a N.W.-European word, cr. Cl'Q; n~p m1iD 1j1'iD "that they should be
worth some money" (Rashi on Baba Me~ica 47a) and the other examples from
Rashi in Avineri, op. cit., p. 326.
 THE PRONOUNS 105

n~p'? ~n~p i1'?.lJii1 n'iD~'i1 liDQi1~ "as cause and effect are connected
with each other" is particularly instructive, as only two objects are
involved, and it is quite impossible to take n~p in its original sense
of "part of". Older authors employ only the MH m ... m (Seg.,
433), e.g., m ~.lJ m m":::Ji1 mmi1 "the behaviour of people towards
each other" (IS.iml).i on Psalm i 3).
g. On paronomastic expressions for the indefinite ct below, par. 16b.
 CHAPTER THREE

THE NOMINAL CLAUSE

12. TIe Copula

a. From early times, all Semitic languages have provided for a pos-
sibility to give the implied relationship between subject and predi-
cate an analytic expression by means of a copula. I This is in SH,
as in BH (Ges., l4lg) and MH (Segal, 405), homonymous with the
personal pronoun of the third person.
h. Notionally, the nominal predication with copula appears to be
in no way distinguished from that without copula. There is no
justification in the examples to assume that it emphasizes the sub-
ject (Schlesinger) or the predicate. The reason for putting the cop-
ula into a sentence may in many cases have been the rhythm of the
sentence, a factor that is impossible for us to assess, owing to our
ignorance of SH pronunciation. It may also have served to lend
weight to the sentence, and therefore is more frequent the less col-
loquial the style. It occurs a good deal more frequently in late BH
than in the texts written during the colloquial life of BH (Ewald,
297b). Its comparative frequency in SH may thus be accounted for
by the rather slow and explicit character of its style.
c. The most frequent order of words in the nominal clause with
copula is subject-copula-predicate: FD'?i1 l'ir.:l O'tv"Elr.:l 0i1 o':n O'r.:lllc:l
"many accents are to be explained on grammatical grounds" (B. Ezra,
Yesod Mora, p. 1); 0i1'? 0"'0r.:l 0i1 Oi~ 'JJ 'tvllr.:l '?J "all actions of men
are of their own free will" (Maim., .(lobe?:., ii 26) i1r.:l'?tv " i1r "Ji
P'" '?Ji1 0i1 O'?,J "the words of this R. Solomon are all nonsense"
(Abr. Maim., .(lobe?:., iii 19a). This is the only possible word order

1 Brock., 52c, note; Kbnig, 338c; Albrecht, ZA W viii 250ff maintain that the ~'i1
is a pure copula (Kbnig calls it "Koinzidenzpunkt") as against Driver, Tenses, 193ff;
Ges., l4lg; Schlesinger, p. 10; Segal, DiMu~, 333, who consider the subject to be
in extraposition. Even if Driver's view, in spite of its many difficulties, is admitted
as an explanation for the origin of the construction, it has no validity for late forms
of the language, such as SH.
 THE NOMINAL CLAUSE 107

when the predicate is determinate: i1'?'jDi1 '?~ Ti~i1 ~'i1 'JtDi1 '?'?::li11
n~ri1 "and the second principle is the one required for the purpose
of this scroll" (B. I:Iiyya, Megillah, p. 6).
d. The order predicate-copula-subject is rather uncommon: ~'i1 ~'n
. .. tD1ii'? r~D tD'~ '?::l "every intelligent man is bound to investi-
gate ... " (B. I:Iiyya, Hegyon, f. 1). It must, of course, be employed
when the predicate is an interrogative adverb: i1~i1~i1 ~'i1 i~'::1 "What
is love like?" (J. Na}:lmias, Aboth Comm., f. 3a).
e. More frequent is the order subject-predicate-copula: '::1 '!)i' ~'?
... ~'tDi1'? mi1 C1'?''!)~ ~'i1 Ti~ C1i~i1 "he did not know that man must
give back in this world ... " (B. I:Iiyya, Megillah, p. 50); r~tD i~i '?::1
i1D::1ni1 P'!)D ~'i1 ~~" :")'0 ,,? "everything that has no limit is outside
the province of science" (ibid., p. 10); ~'i1 lJ'~i '?~i1i1 mtD C1'DiD
"they imagine that this nonsense is Thy will" (Maim., 'Akkum, i 1).2
f. The following two instances present some interesting phenomena
of analogy: ~'i1 i1i'D i1J'~' i1i'D '::1 1"ji1~ po.!) ~'? "he did not study
logic, for it is a doubtful accomplishment" (J. ~m}:li, S. ha-Galuy,
p. 2); ~'i1 i1''?~ li'~i1tD i1D::1n '::1 i'~'.!)i1 1D •.. i'D'? "learn some cal-
endar-theory, for it is a science of which one is in need" (J. Tibbon,
Testament, p. 11). Here the anaphoric personal pronoun has moved
into a position normally reserved for the copula in SH. Furthermore,
although the copula normally agrees in gender with the subject
(Brock., 53a, note), and in this case the anaphoric pronoun should
certainly be of the same gender as its referent, it has in these cases
been attracted into the gender of the intervening predicate. 3 The
same has happened in ~'i1 n'D::10i1 i1::1~'?D ~n::1 '?::1 "every writing is
a conventional skill" (Saadiah Dan6n, MS. Bodl. Or. 108, f. la).

13. TIe Temporal Copula

a. The verb i1'i1 is employed to give a nominal clause a time range.

The construction does not differ from that in the older forms of the

2 The order copula-subject-predicate, recorded by Albrecht, ZA W viii 251, does

not exist. In the example given there, i1~" cn:lim Cl,.,':>r:l lIi'i1 ..,iLllli (Ecc!. vii 26),
lIi'i1 is an extraposed subject. If it were a copula, the masc. would be required. (For
another objection, cf. Konig, 338e, note.)
3 The similarity of the two instances, both of which come from scientifically edited
texts, seems to exclude the possibility of scribal error.
108 CHAPTER THREE

language, except that it is in our texts used with much greater reg-
ularity than in BH, where the time-notion is often left to be sup-
plied from the context. 1
h. The copula with participle, which in MH had a durative or a
modal sense ("wont to", "likely to", cf. Segal, 324-7), had lost this
completely already in late Midrashic Hebrew, and the compound
tenses become simply equivalent to the perfect and imperfect, as in
the later forms of Eastern Aramaic. The causes for this may be the
inherent analytical tendencies of the language, reinforced by Aramaic
influence and in the case of SH by Middle Arabic, where the difference
between simple tenses and those compounded of kiina (yakunu) and
the imperfect practically disappears. Examples of compound tenses
of this kind abound in all writers. E.g. '?'El~ T,ElO ?~r.l m?!:'nm
cm'r.ltD i1r.l, Cli1 i1r.l m~l? i1~'1 n"i1 ~? "you ignored all your books,
you did not even want to see what they were and what their titles
were" (J. Tibbon, Testament, p. 4); ~'Ji1 mJ!:'? Cl"'~r.l ,J"i1 ••• tD 'JElr.l,
tDElJi1 mJ!:' Cl"J ItDJm "and because of. .. we were commended to
castigate our body and flesh on the day of castigation of the soul"
(B. J:liyya, Hegyon, 15a).
c. The subconscious influence of the Arabic construction with kiina
shows itself in the treatment of a coordinated participle in a nega-
tive clause: i1r.ltDJi1 nC!l'tDEl 1.,0 !:'.,,' ~?' i1r.l~ni1 11" r::lr.l i1'i1 ~?tD "who
did not understand the method of science, so as to grasp the proper
sequence of the disembodying of the soul" (B. J:liyya, Hegyon, f. 5).
The !:'.,,' is here dependent on the i1'i1, and connected with rJr.l by
a waw of apodosis (cf. below, par. 28c), nevertheless it is, in accord-
ance with Arabic usage, separated from it by the repeated negation.
d. The employment of finite verb-forms instead of participles depend-
ent on i1'i1 is an innovation of Saadianic prose style: i1n"i1 i1'i1 ItD~
tD'J~ ?~ Y'I!:'n i1J "by which you were want to terrorize every mor-
tal" (Saadia, Sepher ha-Galuy. RE] lxviii, p. 3, line 4). It occurs even
in the very Biblicizing rhymed prose of al-J:larizi (Translation of
J:larlri's Maqiimas, ii 6, ii 15, xxiv 10). In SH it is rather rare: ,'i1'tD
'J'r.l~" ,.," mr.l'~i1 ?!:' ~Ji1 1,!:',Eli1 Cl'~'ltD F~ ?~ltD' "that Israel might,

I jj'jj is used much more regularly in the later stages of pre-exilic BH, i.e., in D

and P, as compared with JE (er. Albrecht, ZA W viii 252, note I). On omission of
jj'jj demanded by the sense, cf. K6nig, 3261).
 THE NOMINAL CLAUSE 109

when they see the requital come upon the nations, thank and believe"
(B. I:Iiyya, Megillah, p. 64); ?:l nipni 'nipn l]T 16 :n iT'iT'iD iiDEl~ '~
~:liT C?'l]:l 'iT 'i:ll] "it is impossible that a Rabbi should not be aware
of his own hope, and that of all servants of God, for the world to
come" (Abr. Maim., .(lobe;;:" iii 19a). This is, of course, identical with
the Arabic construction after kiina (Wright, ii 21B). Yet its rarity in
writers of the second period (I have so far found no instance in
Gersonides) suggests that in the above examples it was a result of
inadvertence rather than of Arabic influence. 2
e. Gersonides employs iT'iT with the perfect for the pluperfect, in
imitation of Arabic (Wright, ii Se): ,JrJrJ 'l]JrJ i:l:l •.• C'i'~':liT m iTr?'
iElOiT iTr 'i:li nJ:liT "the commentaries. .. had prevented us from
understanding this book" (Comm. on Job, preface), and frequently.

14. The Verb if Existence

a. No examples are needed for the use of iT'iT in an absolute sense
"to exist". In this function it also forms a present, iTjiT, as already
in late BR (Eccl. ii 22, Neh. vi 6) and MR (Segal, 212).
h. The Niph'al of ~~rJ serves in the same meaning already in BR
(e.g., Gen. xlvii 14), and is used in SR, especially in philosophical
texts as the most explicit term. It has also produced the derivatives
ni~'~rJ "existence", I (C'-) ni~~rJJiT "existing things", 2 "~rJ "an object".

c. The likewise old-inherited word iD' "exists, there is", and its neg-
ative r~, are employed in SR in the same way, but often with the
addition of CiD "there" iTi'rJ~ CiD r~ ni~~rJJ CiD r~ C~ "if there are
no things in existence, there can be no commanding" (B. I:Iiyya,
Hegyon, f. 2); 1rJr CiD r~ ?J?JnrJ iT'iT'iD =-]1?'n CiD r~ C~ "if there is no
change taking place, there is no time" (id., Megillah, p. 8); CiD iD' ?:l~
~i':liT Cl] "~rJ in~ i:li "but there is one thing co-existing with God"
(Maim., .(lobe;;:" ii 2S); in~ m?~ CiD iD'iD "that there exists another
God" (id., Yesode ha- Torah, i 6); iTr :l:l,:l ~?~ m?~ CiD r~ "there is no

2 This is supported by a similar occurrence in Rashi: i''':1i'' ir.lip i'Jip i~~r.:IJ

"thus its horns are mentioned before its feet" (on Psalm !xix 32), with the modal
copula ~~r.:IJ (par. ISd).
I First found in Ibn Gabirol's ni:l"r.:I in:l, line 22.
2 First (?) occurrence in Bar I:Iiyya, Hegyon.
110 CHAPTER THREE

God but this star" (id., 'Akkum i 1). Cf. also Bacher, Aus d. Wiirterbuche
etc., p. 12l.
This expansion of OlD does not seem to occur in late MH or N.W.-
European texts. It is found both in Arabic, written by Jews as well
as Moslems (cf. Friedlander, Sprachgebrauch des Maimonides, p. xiii) and
in early Romance. 3 Although it is quite possible that Hebrew devel-
oped this idiom on its own, borrowing is very probable.

15. Modal Copulae

a. Although n'n can also have the meaning of "to become" it was
already in BH times felt too weak for expressing modified existence
in addition to all its other functions. As n'n itself had at some point
changed its original meaning "to fall" for that of "to be", I so through-
out the history of the language ever new verbs became modal cop-
ulae and gradually lost their meaning until finally they became mere
synonyms of n'n and were in their turn replaced by new verbs.
The full treatment of these verbs is in the province of lexicogra-
phy and style. But as SH lexicograply is non-existent, and in order
to illustrate the syntactic nature of the modal copulae, a selection of
terms is given here.
h. The Niph cal of n'n serves in the meaning of "to become" already
in BH (Deut. xxvii 9). In SH, it seems to have become a simple
copula, serving as present of n'n ~al: ?lJ nmi1J nI'i1J O':l:l1:ln nI:l?i1i.~
cmp on? ...,Oi.:l ""ID~ ...,ion "the courses of the stars are guided accord-
ing to the order with which their Creator endowed them" (B. I:Jiyya,
Megillah, p. 114).
c. :l11D (BH. Is. xxix 17) is very popular with the SH writers, prob-
ably owing to its corresponding to Arabic 'iida "to become" (Reek.,
56): OID:l 1D...,11D n:l1D pn~' i"1' "the Yod of 'Yi~~a~) becomes in that
name an integral part of the root" (B. Ezra, S. Ha-Shem, f. 3a); ~""EJn
??:l :lID "the particular case turns thus out to be a general rule"
(ibid., f. 3b); P,...,11D:l O"i.:ln nm? :lID P1...,11D:l i"i'n nJ:ln ...,1:llJ:l "because

3 The idiom (French "il y a", etc.) is well known. Its earliest occurrence inJudaeo-
Spanish is probably in the "Coplas de YOyef" (st. 53): r,1I;
Clill ill' i1,"~ i1l111; '11; ":111;.
 THE NOMINAL CLAUSE III

the Yod is vocalized with a Shuru~, the Mem becomes vocalized

with a Shuru~" (J. ~ml:ti, s. ha-Galuy, p. 12); "the water did not
spread over the earth like water poured on top of other water IJ':liViV
1n~ ~'J I:lnQ "which becomes one single entity with the latter"
(B. Zar~ah, Mikhlol, f. 23a).
This word is used in the rhymed prose of al-I:Iarizi almost with-
out exception to translate the Arabic ~iira (cf. Schirmann, Maqamen
des Hariri, p. 11). It is to be investigated whether I:Iarizi was influenced
in this respect by SH or followed an ornate-prose usage.
d. A speciality of Hebrew are the verbs of consequential existence,
i.e., existence dependent on the truth of some argument that pre-
cedes the statement. The principal verb to express this is the Niph 'al
of ~~Q. Its function as modal copula is possibly foreshadowed by
Esth. ii 23; ~~9" ...,:l1n iVp:l" "the matter was investigated and turned
out (to be true)". It is very frequent in MH, e.g. 1:l'1Q, ':J'~Q:J "thus
we learn ... " (Siphre, Deut. par. 210); r..."n-p pn ~~Q:J "therefore
the son is free" (~id. iii 13). 2
Examples: n'.l/Q ,:1 " mrm 1:l1~n '~ m':ln n~~Q:J' "thus the abil-
ity of man to speak is found to give him every advantage" (B. I:Iiyya,
Hegyon, f. 12b); n:ln c:l~iVQ:1 n;J07 ntp-PO ""iJ~ ~~Q:J m, "and thus the
phrase 'a window thou shalt make to the ark' (Gen. vi 16) is con-
structed on the same pattern as the other instance" (B. Ezra, Za~oth,
f. 3a); niVQ m,n 1:l',c:l:lQ l:l'iV~c:ln n'~ '~~Q:J "therefore these fools are
denying the validity of the law of Moses" (Maim., $obez, ii 26).
e. Maimonides employs ~i:l in the same sense: np~oo inpmm ~i:ln
...,ni':l nn:liiVQi "his translation will then be most adequate and excel-
lent" ($obeZ, ii 27). Such a use of the verb is not to be found any-
where in older literature. It is simply a translation of the Arabic jii' a
"to become" (Wright, ii, p. I03B), a word of which Maimonides is
fond in his Arabic writings (cf. FriedHinder, Sprachgebrauch, p. 21).
f. An interesting innovation, without parallel in Arabic, is ~'nnn "to
be sometimes":3 1:l'~"":J m I:l'Q'.l/:J ni~'nnQ ''''n~ ni'ni~ "the letters "n~

I The Talm. Aram. '1i1 is always modal (Schlesinger, p. 14).

2 Full list of Mishnah instances in Kosovski, pp. I 160b-1161c, cf. also Ben-Yehudah
s.v. Not recorded by Segal. The word is current in Rashi, cf. also par. 13, note 2.
3 This is, of course, a loan from Arabic ikhtalaJa "to differ". The Arabic word is
not used as a modal copula, according to the dictionaries (but cf. ikhtalaJa ilii "to
go repeatedly to a place, to visit a place one after another").
112 CHAPTER THREE

are sometimes vowels, sometimes consonants" (B. Ezra, lp~oth, ( 4b).

This creation of Ibn Ezra's seems, however, not to have penetrated
into common usage.
g. Similarly isolated is the modal copula use of i~tD:l "still to be":4
IJ.I.1c!) rln? ~'i1i1 tDi5:l~i1 ?.1.1 i~tDJ "that commentator has still to give a
reason ... " (S. Tibbon, Eccl. Comm., f. 15b). The last two cases
show with what case verbs could be adapted for the copula function.
There is little doubt that SH inherited this tendency from Arabic.

4 I.e., the Arabic baqiya, so used by Maimonides (Friedlander, Sprachgebrauch,

p. 9 and in the colloquials.
 CHAPTER FOUR

THE VERBAL CLAUSE

16. The Indifinite Subject

a. The indefinite person is mostly expressed by the third pI. masc.,

as in BH (Ges., 144f) and MH (Segal, 439). The use of the third
sing. is less frequent, but also occurs, particularly in Ibn Ezra and
Maimonides.
h. Likewise, the indefinite person can be expressed by a parono-
mastic participle, again in accordance with BH (Ges., 144e) and MH
(Seg., 438): ?tI;1iD ?tI;iD Ctl; "if anyone asks" (B. Ezra, Saphah Berurah,
f. 4b); iQ1t1; iQtI; Ctl; "if one says" (J. ~ml:ti, s. ha-Galuy, p. 1). The
purely syntactic origin of this notionless subject is well illustrated by
its occurrence in negative contexts, where by definition it does not
correspond to any reality, e.g.: 'i::l1 Pii? PQ ?;:'1' tI;? "no one will
be able to understand what I say" (B. Ezra, ?,a~oth f. 7b); cf. already
in BH "that thou bring not blood upon thy house, 1JQQ ?ElJii ?El' ';:'
should anyone fall from thence" (Deut. xxii 8).1

17. The Agent if the Passive Verb

a. The original character of the Semitic passive was that of an
intransitive, related only to its grammatical subject. It was essential
that the originator of the action was not named (Brock., 73a). But
already in BH (Ges., 121f) and MH (Segal, 361) it had become pos-
sible to introduce the agent by prepositions indicating either an instru-
mental or a spatial (ablative) relationship. This is even more common
in SH, which in this respect contrasts sharply with Arabic. In the
latter language the indication of the agent is in all periods rather
uncommon (Reck., ~nt. Verh., p. 43).

I There are instances of more direct negative construction in BH, but in these

the participle is possibly to be taken as full noun, e.g., Is. xvi 10: 1iiil 1ii' ~"
"the vintner shall not tread".
114 CHAPTER FOUR

h. The agent of the passive is in SH introduced by· the following

prepositions:
le-, the most frequent one in BH (Konig, 102b; Giesebrecht, Priipos.
Lamed, pp. 63f), is ,comparatively rare: imi1i1 iDEl:J" niD:J:I:J ~'i1 C~ "if
it is conquered by the intellectual soul" (B. I:Iiyya, Hegyon, f. lIb);
1" i1',:JP i1'i1n ':J'''El i1iiD "the field of so-and-so will be bought by
you" (id., Schwarz-Festschr., p. 28); ,,, ri:liD:J r~' i:liD:J ':J'~ "he is
not employed and others may not be employed by him" (Maim.,
Talmud Torah, vii 5).
c. mm IS very rare III this function in BHl and in MH.2 In SH it
is frequent, perhaps owing to the use of min in Arabic for this pur-
pose (Reck., 137.la), or under the influence of Aramaic (cf. Schlesinger,
par. 83 end; Dalman, p. 228): 1i0" 'i:JiO io'no ':J'~ "he becomes
not chastized by the words of him who reproaches him" (B. I:Iiyya,
Hegyon f. 13a); c'pono iD:J.lI:J ':J'~ "he is not punished by God" (ibid.,
f. 13b); C':JC!lPi1 10, c'''nJi1 10 i:J':lO i1'i1n "you will be honoured by
great and small alike (J. Tibbon, Testament, p. 7); ,:JOO C'io"m r~
"one is not taught by him" (Maim., Talmud Torah, iv 1); C'Jm'Oi1
,JOO "those who are led by him" (Gers., Dloth f. 6b); i1":Jpi1 nmiDi1
,:JOO c'miD'O:J "the care of God for those who are specially guided
by Him"3 (ibid., f. 8b); n"iDi110 imiD i10 "that which has been decreed
by God" (ibid., f. 9a); m",ro Ci1:J iO~:JiD i10 "what has been said
about them by others than he" (Crescas, Or Adonai f. 4a).
d. The instrumental expression for the passive agent by means of
be- is rare in BH (Konig, 106; Brock., 73b) and seems to be no
more frequent in SH. E.g., miDiiOi1 1m~ C!liDEl:J i1nEln'iD '0 "if any- ":1
one is misguided by the literal meaning of those Midrashic passages"
(Abr. Maim., flobe;::, iii 18a).
e. 'i' ".lI is in MH used only to mean "by the agency of", as in
"~prn' 'i' ".lI iDilElO "explained through the agency of Ezekiel" (Tamid
iii 7; the actual agent is, of course, God). In B. Daud, it introduces
a real passive agent: C'C!liD n:l"o 'f' ".lI Jim "he was put to death

I Molin, Om prepositionen min, p. 39, finds only three occurrences of such a min

in BH (Ps. xxxvii 23, Job xxiv I, Ecc!. xii, 11). K6nig, 107, adds another instance.
It is hardly possible to take this use of the preposition with Molin as a special case
of "min causale". It is, of course, simply local min.
2 E.g. 010iT 1i.l O','?1JiT '?~ "all those sired by a horse" (Kil'aim viii 4).
3 This is a case of a passive from a verb with indirect government (~ or '?lJ).
 THE VERBAL CLAUSE 115

by the queen of Shinim" Cflabbalah, p. 51). It does not occur in SH

writers proper. 4

18. Some Remarks on the Tenses

a. The tenses and their connotation from the point of view of time,
aspect, etc., are not in the province of syntax, but of morphology.'
No systematic collection of material on this subject has been under-
taken for this thesis. It may, however, be useful to make here a few
observations on the SH tenses, based on the general impression of
the texts.
h. The BH and MH tense systems are thoroughly different from
each other. The first is a scheme of two aspects, with a differentia-
tion between the first verb in a passage and those that follow up in
the same aspect. It is further complicated by the intrusion of other
categories.
MH had a hybrid between a three-tense system and a two-aspect
system. The categories of the latter (durative and punctual) were not
the same as those of BH (accomplished and incomplete). Added to
this was a rudimentary mood distinction.
c. In late Midrashic and N.W.-European Hebrew, the MH system
was simplified by the lapse of the aspect category, the compound
(durative) tenses becoming synonymous with the simple (punctual)
ones (cf par. 13b).
d. In SH, the late Midrashic three-tense system was disturbed by the
influence of the two other systems, preserved through poetry and ornate
prose. 2 The balance of the tense system was, however, completely
upset by the impact of Arabic. The principles of the tense-system of

4 This meaning is not mentioned by Ben-Yehudah or Avineri. It is, however,

very common in Modern Hebrew, and requires further investigation.
I The inclusion of tense-connotations in syntax is a· remnant of the system of
Miklosich, who held that syntax represented the study of the meanings of gram-
matical forms, as against morphology, which dealt with forms only. Its retention in
modern books on syntax is a concession to the reader (see Reek., p. iii.). It is partly
justified by the fact that tense-meanings (as case-functions) play an important part
in syntax, and are often not semantically, but only syntactically, definable.
2 It seems as if Spanish poetry showed a tendency towards a temporal tense sys-
tem with BH material: perfect for the past tense, imperfect for present and future.
If correct, this would be identical with the system current in modern poetry.
116 CHAPTER FOUR

that language are as yet rather imperfectly known, 3 owing mainly to

its extreme complication. The basis is one of aspect, rather similar
to the BH (but without the secondary-verb category), overlaid with
a series of living and fossile categories of various nature. Apart from
that, certain frequent constructions (conditional clause, idhii) neutralise
the aspect or time category. This and several other factors bring
it about that in non-narrative prose it does not matter in about 50
per cent of all cases which tense is used.
e. One can almost watch in SH authors the gradual disruption of
the MH system. In the authors of the first period it was modified
above all by a tendency to employ the imperfect instead of the par-
ticipial present tense. Consecutive tenses did not intrude. In the sec-
ond period, usage settled more and more into something akin to the
Arabic system. A cursory examination of Gersonides, for example,
produces no cases of tenses that could not have been the same in
an Arabic text. This is, of course, due to the influence of the trans-
lations, where verbs seem always to have been rendered in the same
tense as in the original. It is to be expected, however, that the usage
of these authors did differ from the Arabic in many of the subtler
points. Perhaps there was also some influence of the Romance ver-
naculars spoken by the authors.

3 Cr. the multiplicity of meanings attributed to both tenses in Reek., parr. 7

and 8.
 CHAPTER FIVE

THE OBJECT

19. the Direct Ollject

a. The nota accusativi ntoll before determinate object, which appears

with fair regularity in BH (Wilson, Hebraica, vi 139fT, Konig, 264ff)
and even more so in MH (Se gal, 351) is very rarely employed in
SH. This is a development of N.W.-European Hebrew. l Already
in Al).imaa?, eth is very rare. In Rashi it appears sporadically, and
in the Sepher lfasidim it is almost entirely lacking. In SH, the N.W.-
European tendency may have been reinforced by the absence of
such a particle in Arabic.
eth is occasionally found in all writers, perhaps most frequently in
those under strong MH influence as Maimonides in the Mishneh
Torah. But there is no regularity in its use. One feels that it had no
function whatsoever.
h. eth is not even employed where its insertion would help to clar-
ify the construction, i.e., where the object precedes the subject: 2
1:l'?'O~ ~?tolI 1'i:Ji 1J':J' tolI?rv 'El ?lJ ~tolI "although a thousand fools may
not understand his words" (Abr. Maim., [lobe?:., iii l7b); iQtolI P10Eli1 i1r
r10i'? i1o'?rv "this verse Solomon wrote to indicate ... " (M. Tibbon,
Cant. Comm., p. 5). Cf. examples in par. 29b and d.
c. There is no trace in SH of le- as a nota accusativi, although this
is found in BH (Ges., 117n),3 including Ben Sira (Strauss, p. 70),

I Perhaps it originated from the liturgical style, where eth with determinate object

is considerably less frequent than in BH prose. The prayers imitate in this the poet-
ical style of BH. The omission of eth would thus ultimately be a poetical feature
which penetrated into prose.
2 A list of such cases in BH prose, Brock., 212. Examples from European Hebrew:
~~o ~" ij~i1 "the mule he did not find" (Al;Iimaa~, Chronicle, p. 3); jtD~~ '~ 'j'tDlli11
n1JO'? "and the tenth one cannot count" (Rashi, Gen. viii 5).
3 Kropat 35 (Brock., 211d) considers this le- an Aramaism, but Giesebrecht, Prap.
Lamed, pp. 82f, argues that it is genuinely Hebrew (i.e. in fact Proto-Semitic, as it
occurs in all Semitic dialects). Giesebrecht admits, however, that its later extension
in use was due to the influence of Aramaic.
118 CHAPTER FIVE

and MH (Sega1, 168), and occurs, somewhat rarely, in Arabic (Reck.,

135.3b).4
d. Contrary to the usage of BH, MH, and Arabic, one object can
be dependent on several verbs: 1:J'i1'?~ lii:J1 i':J:J 1:J "by it we recognize
and know our Lord" (J. ~ml).i, S. ha-Galuy, p. 1): 1lii'1 1a'p ~'?
i1'1m rm:m :J"',n pli "they did not observe or know the 613 pre-
cepts of the Law (B. 'Abbas, f. 1Oa); '~:ltDa '~:J'?1 '1atD'? i1a'p~1 'nli:JtD:J
i1Pi~ "I have sworn and shall carry out to preserve and explain the
righteous sentences" (J. Nal).mias, Aboth Comm., f. Ob). This is a
heritage from the liturgical style, where long series of verbs with one
object are common, e.g., Q'tD'ipm Q'~"lia1 Q'i~:la1 Q'n:JtDa1 Q':J':Ja1
... QtD n~ Q':J''?aa1 "they bless and praise, glorify and reverence,
sanctify and bring homage to the name ... " (Singer, p. 38).

20. the Indirect Object

The prepositions by which a noun can be governed by a verb are

a concern of lexicography. It may, however, be of some interest for
understanding the nature of Arabic influence on SH to give a few
instances of deviations from BH and MH usage from authors not
strongly influenced by the translations:
Qi1'?li '1pn? "to inquire into them" (B. I:Iiyya, Megillah, p. 1): BH
accus., Arabic babatha calii.
Q'p'?n i1tD'?tD '?li Q'p'?nm "are divided into three parts" (ibid. p. 3):
MH ... '? Q'p'?1n (Parah iii 10), Arab. qasama Calii.
i1m~ Q'J'naa "they wait for it" (ibid.): MH ... '? rnai1 Arab. inta;;ara
c. accus.
'?li O:J:J:J "enter upon, join" (J. ~ml).i, S. ha-Galuy, p. 12): MH
'?~ O:J:J:J, Arabic dakhala calii.
in~ '?:J 1''?li '?1:J' "everyone can do it" (Falaq., Meba#esh, f. 18b):
BH accus., Arabic qadara calii.

21. the Cognate Object

a. The cognate object specified by an attribute is extremely fre-

quent, since it makes up for the lack of adverbs derived from adjec-
tives. While in this respect SH does not differ from the older dialects,

4 This construction, is, however, quite alive in the Hebrew of the Karaite Aaron
 THE OBJECT 119

it has an innovation in so far as for all Binyanim except the ~al it

mostly employs a standardized verbal noun of the form pi'ill, what-
ever the verbal noun peculiar to the Binyan in question may be:
Niph'al: 'ill' p'J', l:'jlJii ~tI; npJ'J ii"~ "a form that is immedi-
ately inherent in the body" (B. f.liyya, Hegyon, 5).
Pi'el and Pu'al: 'J".lJ ""0 iir ~:J l~ "00 'JJii "I arrange all this
for you in a systematic (?) manner" (B. 'Abbas, [, 6a).
Hiph'il and Hoph'al: lP,no, chill p,m ,J','J 'J'ii prmii "this is
fully and firmly established" (B. f.liyya, Megillah, p. 10).
Hithpa'el: "~J 'J'tI;ill "J'n "~tI; "Jnm "they became indissolubly
attached to it" (id., Hegyon, f. 3a); ,,"J ntl;rii iitl;"iiiiO 'J~ ''In',
... ii~' "and it becomes quite clear to us from this admission ... "
(id., Megillah, p. 10).
h. The cognate object without any attribute is on the contrary rather
infrequent. SH follows here BH (Ges.,117q, K6nig, 329c-f) and
earlier Arabic (Reek., 48.3a). E.g., ill"'~ Clm'~~ 'ill"~ill Cl'J'illtl;'ii 10 ill"
"some ancient scholars have given an explanation for their form"
(J. ~m1;i, Zikkaron, p. 2).
c. It is, however, frequent in a peculiar construction, in which the
cognate object appears as part of the major phrase, the verb on
which it is dependent being attached to it as a relative clause. The
cognate object thus crystallized out of the clause represents the ver-
bal idea in a syntactically nominal form and is equivalent to a sub-
stantivised clause introduced by she- or ki. In European languages it
would best be rendered by some phrase like "the fact that".
Examples: iin'Oii ,~ iiO'J J'Jr ~.lJJ ntl; ~tI;illill 'n~tI;ill "the fact that
he inquired of the Baal Zebub was the cause of his death" (B. f.liyya,
Megillah, p. Ill): Cln~po ,tI;'ill ii'~l 'P.lJii "the main thing is that
some people have actually seen it" (J. ~m1;i, S. ha-Galuy, p. 1); tI;'ii
nC!l'J ii'iiill 1J18C!lJ ,~~ iiJ'C!l~ ,r ClJ 'O,tI; ii'ii "he used to say 'this, too,
is for the good', because he had faith ... " (B. f.liyya, Hegyon, f. l6b);
Cl"'~:Jii n~tI;illO "'iiriiill ii'iirtl;ii ~.lJ ".lJ "pn~ "rm "we shall inquire
later into the fact that we were warned against consulting the Chal-
daeans" (id., Schwarz-Festschr., p. 26); 'nltI;Jill nI~J illtl;, ntl;Jill 'J~O
,tI;J'ill ii'iiill '''ii "because the Exilarch of that generation was his
enemy" (B. Daud, 1labbalah, p. 63).

b. Joseph: '?~'tl),'? 1j'm' 'tI)~ '?i1pi1 'jpr "the elders who lead Israel" (MiMar
Yesharim on Cant. i 10): i11J::In? ?~'tI)' 'p'J" P "thus does Israel nurture wisdom"
(ibid.)".
120 CHAPTER FIVE

d. This extraposition of a cognate object is something quite different

from the cases where a paronomastic verb is attached as a relative
clause to a noun, and where the action, etc., described by the major
sentence applies to the idea of the noun, not to the verbal idea de-
scribed by the relative clause. Thus: m:m 1i11~ nn~ '?~ itD~ iii'~Ji1I
!:J'J,tD !:J'J'O '?~ np'?m ii'JtDii '?.l) "the superiority which each possesses
over the other can be divided into various kinds" (B. I:Iiyya, Hegyon
f. 11) could not possibly be translated "the fact that each possesses
superiority: ... ". Similarly: mO'~ii '?.l) !:JP'J ~'ii itD~ iiOpJii ~'ii m
"this is the revenge he takes upon the nations" (B. I:Iiyya, Megillah,
p. 86); ~tDntD 'n~tDno ':J~ "according to the idea he had" (B. Ezra,
Yesod Mora, p. 6). Although there may be cases in which both expla-
nations are possible, it seems that the abstract cognate object forms
a definite semantic category.
e. From the instances collected by Reckendorf (Paronomasie, pp. 159f)
it appears that neither in BH nor in any other of the older Semitic
languages did the concrete construction develop into the purely
abstract one under (c). It is possible that the shift of meaning came
into the construction within the N.W.-European development of
Hebrew. It occurs in Rashi, e.g. iiJO~ii '?.l) m~r'? !:Ji~~'? ii~tDn ii"~Pii
,~ rO~iitD "God reckoned it as a meritorious act for Abram, because
of the fact that he believed in Him" (Gen. xv 5).

22. The Double Accusative

a. When a verb is followed by two objects, the relation between the

three terms may be of three different kinds, without there being, in
Hebrew, any formal criteria to distinguish them. The kind of rela-
tion that obtains can only be determined by the sense of the verb.
Both may be direcdy governed by the verb, e.g. 'J~ n~ !:Ji~ iO'?'
ii'pJ mJm~ "let everyone teach his son a cleanly craft" (~idd. iv
14).1 One of them may be a cognate object (see examples above).
In either of these two cases the syntactic principles are the same as
if every object was singly attached to the verb. The case that inter-
ests us here is the one in which the two objects stand in a pred-
icative relationship to each other, the Latin accusative cum infinitivo.

I The second accusative may be one of specification, as 1Elll 'i1~"r.:l' "he shall fill
it with dust" ('U4in ii 10).
 THE OBJECT 121

The real object of the verb is then neither of the two nouns, but
the rudimentary clause constituted by them. In fact, such a double
accusative could always without change of meaning be replaced by
a dependent clause. 2
h. In SH the range of verbs that can take such a double accusative
appears to be smaller than in BH and MH, where it could be used
after any verb signifying transformation or production, as "to build"
(I Kings xviii 32); "to bake" (Ex. xii 39); "to tear" (I Kings xi 30),
"to write" (Gigin ix 8). It is mainly employed after verbs meaning
"to consider", and rarely after i1tDll "to make into", and lnJ "to give
as". Though I have no examples for this, they undoubtedly also
occur after verbs of sense-perception.
c. The second, or predicative, object can be a substantive or adjec-
tive noun, or a participle: i1m~ lli"'i1 ?::l ?ll i1::l.,n n~ri1 mJm~i1 ~~~m
"thus you find the practice of this art to be a duty for anyone who
is skilled in it" (B. J:Iiyya, Schwarz-Festschr., p. 31); 1~ ~~.,~ i1n~
.,n1'~' nll::l C1i~? C1'1m C1"ni1 ::l.,n::li1 "you find in Scripture the span
of life to have been decreed for each man at the time of his being
created" (B. J:Iiyya, Megillah, p. Ill); C1"p~i1 i1Jll~? C1'::l'1~ C1'~tDi1 ~~~m
.,,::l? "thus you find the heavens to be in need of the help of God
alone" (ibid., p. 114); ::l.,t!) mi1 C1"'i1 ::l.,n::li1 i1~1 ~?., "and Scripture did
not consider this day to be good" (ibid., p. 23); ?::l::l C1::ln Tn'~1
i1::l~?~ "I perceive you to be skilled in every art" (Falag., Meba~~esh,
f. 11 a); m'n~~ C1i1i1 m~'pi1i1 n'JJ "we presume those premises to be
true" (Crescas, Or Adonai, f. 4a); CI"n::l~ CI'n'iDll ~? "1 did not pro-
duce them in a scattered form". (B. Parl).on, p. xxii); m::lr i1? lmJ
r~.,mn ~::li1 C1?.,lli1 "He gives to her the privileges of the world to
come as a consolation (B. J:Iiyya, Hegyon, f. lib).
d. When it is desired to negate the predication itself (usually the
negation is attached to the governing verb) this is done by inflected
r~. Thus the double accusative is transformed into an asyndetic nom-
inal object clause: mi1 Ji1J~i1 mmJ i1J'~ yPi1 ?ll i11'pni1 ~~.,~ i1n~ "you
find that the inquiry about the end of the world is not conducted
after this pattern" (B. J:Iiyya, Megillah, p. 2).

2 The distinction between the double accusative with predicative relationship and
the verb with two accusatives is not sufficiently clear in Ges., 117h and ii, and
Segal, Di~du~ 384, but er. Kbnig, 327y.
122 CHAPTER FIVE

e. In Bar J:liyya already, the copula is also found in positive phrases

of this kind: '~mlD :l:l1:l m':lnm nm ~'i1 ... 1'?1l i1i'llDi1 m~i1 1~~D
... Clll "they found the sign for it to be the conjunction of Saturn
with ... " (Megillah, p. 116). It is not clear to what extent the use of
the copula in this construction was current in SH. It occurs in Arabic
(Reek., 141.4, Brock., 209d, note 1).
f. The final development of the double accusative is the admission
of a finite verb instead of the nominal predicate, the accusativus cum
verbo finito (Reek., !$ynt. Verh., pp. 515f, where its difference from
the circumstantial clause is proved). It seems that, of all Semitic lan-
guages, Arabic alone took this step (Reek., 189; Brock., 342), and
SH borrowed it from there in the same way as Neo-Syriac did
(Noldeke, Neusyr. Cr., 180).
g. The verb can be in the imperfect: 1~~11D' Cl'i'D?m :l1' i1~1' i1n~
1~~D' ~?1 ,:JO IDp:l? "you see most students go about searching for
a book without finding it" (J. Tibbon, Testament, p. 4).
More frequently, however, it is in the perfect: ,'ni1 :l1n:li1 'n'~'
':li:l ~~11D? Cl':l'? "I saw Scripture having permitted the multitude
to explore this matter" (B. J:liyya, Megillah, p. 2); C1"1:l'j Cl'~1' 1:l~
mi1 plli1 ?ll 1'pn ... i1D:ln:l "we see great scholars having inquired
into this matter" (ibid.); ID'~ ':l:l Clj Cli~ ':l:l ClJ 1'D~ Cl~'n ~?i1 "do
you not see that they said 'sons of Adam' as well as 'sons of man'"
(B. Ezra, Saphah Berurah, f. 9a; the ed. has i1~'n); 1:l?i1 Cl':l11D~'i1 'n'~'
Cl':l11D Cl':l'i:l i1lDn':J:l "I saw that the ancients had adopted various
methods in interpreting it" (M. Tibbon, Cant. Comm., p.5).3
It is in the use of the perfect that the real advantage of this con-
struction lies, as it permits to indicate that the action of the depend-
ent verb took place before that of the principal verb. In the Semitic
languages, which do not distinguish time in non-finite verbal forms,
this was the only way to do so.
h. In the older dialects, the double accusative can be resolved by
inserting le- (BH, Ges., l19t) before the predicative. This is rare in
SH, perhaps because the construction was not possible in Arabic.
E.g., 'p:l Cli? C1'Dii1 1? 1:llDnJ "the blood was counted for him as the
spilling of innocent blood" (D . .£5im}:li on Hosea, p. 14).

3 I have so far not found any instance of this construction in the works of
Gersonides. Should it really have lapsed in the second period?
 CHAPTER SIX

THE PARTICLES

23. Some Remarks on Adverbs

a. Adverbs as a class offer no special syntactical interest in SH. They

are a poorly developed category, owing to the lack of a regular pro-
cess for forming adverbs out of adjectives. l
Besides reviving all the BH adverbs that were lost in MH (Segal,
294), SH possessed quite a number of new adverbs not found in the
earlier dialects, but these are of interest for lexicography only. Here
it may be useful, however, to treat briefly of three adverbs that have
shifted their syntactical and stylistic function to some extent.
h. i:J~, a late BH borrowing from Aramaic (occurs only in Eccl.).
If it ever meant "already, a long time ago" its meaning was already
much weakened in the Eccl. passages (especially ix 6, 7), and even
more so in its frequent MH use, 2 where it merely served to under-
line the completed aspect of the perfect tense.
It corresponds thus to the Arabic perfect-particle qad (Wright, ii
p. 3C). It is frequently employed in this sense in SH: i1D nl.li' i:J~
'n'?:JOiD "you know well what I have suffered" (J. Tibbon, Testament
p. I 0); 1~' 1~ 1'r1~'?EljD '~i i:J~iD ''? 'iD~ "they told me they had
seen such-and-such of his miracles" (Maim., (lobe!:" ii 26); i'~' '?'~iDDi1
'?~O i1'i1 i:J~iD 'jElD '?'O~i1 "the enlightened person recognizes a fool,
because he himself once was an ignoramus" (Falaq., Musar, p. 50).
With i1'i1, it denotes the pluperfect (cf. par. 13e).
c. Through its identification with qad, it comes to be used by authors
of the second period also before the imperfect and in nominal clauses3
in the sense of "sometimes, possibly" (Wright, i 286c; Reek., 155.3).

I Only Modem Colloquial Hebrew has made an attempt to solve this problem

by the use of 1;;1~~ "in a manner" with the adjective, e.g., or.mo lEl'~~ ".,~ i~'i
"he spoke to him politely", for O'O'J~.
2 E.g., ~'imo i1",I) i~:I "Sennacherib has come up ... " (Yadayim iv 4). Other
exx. 'Erub. iv 3, 'Arakh. viii 7, etc.
3 ~'i1 i~:I i1'i1iV i10 (Ecc!. iii 15) does not, of course, bear this sense of kebaT. A
similar use Gigin iv 5.
124 CHAPTER SIX

This is extremely frequent in Gersonides' wntmgs, e.g. "whoever

takes another's possession unrighteously, takes the life of its owner"
,n::l1n ii'?Jn ~'?iD 'i~ ,m1ii' 1~~tD '?"1 "i.e. he is quite likely to kill
him so that his vileness should not be discovered" (Comm. on Prov.
i 18); .li1ii 1~ 1~iDii'? ii~.li ,jr1' 1~~ "sometimes they advise how to
keep away from evil" (Dloth ( 7b); ... iD 1iD::l~ 1~~ "it is sometimes
possible that ... " (ibid., f. 6b). The last instance is already an
autonomous development of Hebrew, since in Arabic qad can not
be used in nominal clauses.
d. mii is used much more frequently in SH than in BH. The word
itself is a revival in SH, and in MH it had been replaced by '1ii or
1ii (Segal, 305 end), and its regained life is probably not quite uncon-
nected with the existence of Arabic inna. Indeed, its function is often
reduced to that of the Arabic word in introducing a nominal clause,
especially at the beginning of an utterance (Reck., 65.1), e.g. '?iD~ii mii
~'ii "a proverb is ... " (Gers. on Prov., i 1). In contrast to the Arabic
word, mii is just as frequently employed to introduce verbal clauses,
e.g. CJ.li~ii 1'~rii mm "and he mentioned the reason" (B. Ezra, Yesod
Mora, p. 8).
e. Interesting is the tendency to make ii:lii also take over the role
of the MH '1ii as particle introducing the apodosis of a conditional
clause: 4 1::!.li '?,l),::l 'Jii mn::l ii'ii C1~ "if it· has a PathaJ:! in the second
syllable, then it is a perfect" (B. Ezra, Zaboth, ( 9a); "there are many
strange and unintelligible commandments" CJi~ii CJi~iD' ~'? CJ~ mm
iii,n ~'?~ 1~iD' mii CJ~.li~ .lii'iD i.li "and if one would not observe
them unless one knows their reason, then one would remain with-
out the Law" (id., Yesod Mora, p. 13).
f. In BH, r~ serves in the same function, as particle to introduce
the apodosis of conditional clauses (Brock., 465b). It was revived in
Saadianic prose, and at the same time its use was extended to the
apodoses of temporal clauses: 5 mn' r~ p 'J'miD.li~ "when we do so,
He will show mercy to us" (Saadiah, Egron, verse 20). In the Hebrew
of Rashi, the particle is currently employed for introducing the prin-

4 Examples: "if he saw a koi and said ... " C'1'IJ C'?,:l '1iT "then in each case
he is a Nazir" (Nazir v 7); '?C!l:l 'JiDiT '1iT liT~ in~ i:l~ "if one of them is lost, then
the second is void" (Ginin ix 5). er. also Ginin vii 2, Baba Me? x 2, Baba Bathra
v 2, Nega'im xi 9, etc.
5 There are, however, in BH cases closely approaching this use, although a con-
ditional tinge always remains. E.g., 2 Sam. v 24, 2 Kings v 3.
 THE PARTICLES 125

cipa1 clause after temporal clauses beginning with in~?, iiD~:J, -lD:J,
-iD F:J, -iD, etc. (cf. Avineri, Hekhal, p. 199). From N.W.-European
Hebrew, it came into SH. Possibly its use is restricted to Ibn Ezra.
E.g., ,'J'l1 'in~ ,~'? i1n' r~ ... p'iD~ 'rJ1'ii~ "when one is in a pub-
lic place ... then one's heart will go after one's eyes" (Yesod Mora,
p. 9): CI1~ii rJ1Q' r~ ii":Iii 1Q ... n1iii n~r n~~~' "and when this
spirit ... leaves the body, a man dies" (ibid., p. 11).
g. Typical is also the use of apodotic r~ with a purely Mishnaic
conditional construction: 6 ,,? ~'C!l r~ .. ; iii'~l1 i~'Q 'Q~l1 l1J,Qii ':J "if
one refrains from a sin ... then it is good for him" (Yesod Mora,
p. 11). So in the Sepher Ifasidim: ~C!l'n r~ ii'?":1 iD'iD iiJC!lp np,'?ii "if
one marries a minor when there is a grown-up girl available, then
one commits a sin" (par. 1145).

24. The Prepositions

a. When the same preposition is applied to two coordinated nouns,

it is in SH mostly placed only before the first. This occurs already
fairly frequently in late BH (Konig, 276e, 31ge; Kropat, p. 43).1 The
regularity of the feature in SH is doubtlessly due to the fact that it
is the Arabic usage (Reck., 173.2).2
Examples: C1'~'C!l C1'iDl1Q, iiQ:Jn~ "with wisdom and good works"
(B. I:Iiyya, Hegyon, f. 5); C1J'~i' C1rJ1iDiQ "of their own right and will"
(id., Megillah, p. 112); C1nl1'piD' C1'~:J':Jii rJ1'?11~ "at the time of the stars
rising and setting" (ibid., p. 114); ii'?'Dii1 m'?'l1:J mJi n-n::m "many
commandments, such as burnt-offerings and circumcision" (B. Ezra,
Yesod Mora, p. 8); C1'~'?ElJii "~n~' "iDl1Q~ CI1~ii p'~n'iD iil1iD~ "when
one considers His deeds and marvellous creatures" (Maim. Yesode
ha- Torah, ii 2); iDJ'l1ii1 i:JiD~ rQ~ii "believed in reward and punish-
ment (B. 'Abbas, f. 5b).
h. No attempt shall be made here to discuss the idiomatic uses of
the prepositions in SH as compared with other forms of Hebrew.
These are largely a lexicographical matter of great complication, and

6 r~ does not occur in the Mishnah at all.

I I have no data for MH. Segal does not mention this feature, which is also
ignored in Ges.
2 Reckendorf does not point out that this construction is almost exclusively
employed.
126 CHAPTER SIX

with comparatively little bearing on syntax. Uses of syntactical impor-

tance are discussed in other chapters. It is also impossible to deal
with them satisfactorily at present, when the study of MH preposi-
tions outside the Mishnah itself has not yet been undertaken, to say
nothing of N.W.-European Hebrew.
c. The importance of the latter for the understanding of SH prepo-
sitions may be illustrated on one example. "after" is employed for
denoting a multiple series of units in i'i in~ i'i "generation after
generation" (J. Nal).mias, Aboth Comm., f. la). This is against BH
usage 3 and does not seem to occur in the Mishnah. 4 It occurs, how-
ever, in Saadiah in the same phrase (Egron, v 8), and is frequent in
Rashi (cf. Avineri, p. 201). It is therefore unnecessary to assume here
direct influence of Arabic in SH, although the Arabic idiom may
have played a role in Saadiah's choice of the phrase.
d. In the first period at least, direct borrowing Ooan-translation) from
Arabic is comparatively rare, though the influence of that language
can be felt in the frequency of some exceptional or secondary BH
usages, such as the partitive min or the causative le-. Closer investi-
gation may well show that the emphasis of these constructions had
changed, too. There are, however, some definite borrowings in func-
tion, e.g. the explicative min (cf. Wright, ii 137) Cl'OiD 10: Cl'~~O,Jil '?:J
Clil',J':::JiD ilO' ri~' "all existent things: heaven and earth and what is
between them" (Maim., Yesode ha- Torah, i 1), or the min meaning
"with respect to": m'?rOil 10 1i0.t10 Cl'pO :::J:J,:J, :::J:J,:J '?:J:::J n.t1i'? "to
know of every planet its position with relation to the constellations"
(B. f.liyya, Schwarz-Festschr., p. 29).
e. SH revived all the BH prepositions that were lost in MH (Segal,
300):5 n~ "with"; '?'?.J:::J "for the sake of" (e.g. B. f.liyya, Megillah,
p. 29); 'n'?:::J "without" (with suffixes: Cln'?:::J, D. ~ml).i, Mikhlol, f. la;
Falaq., Meba~~esh, f. 19b); i':::J.t1:::J "for" (e.g., B. Ezra, Yesod Mora,
p. 1); i.t1:::J "for" (e.g. Falaq., Musar, p. 49); n~o "from" (e.g. Abr.

3 BH employs for this 'al (Is. xxxii 10) or le- (Is. xxviii 10), cf. Budie, Prap. 'al,
pp. 39f, Giesebrecht, Prapos. Lamed, p. 30.
4 In a phrase like ~j 'n~ ~j "a divorce-document where one has been sent
already" (Yebamoth, v I) it has its original meaning, and only two terms are implied.
5 Some are added here to Segal's list, which is not quite exhaustive.
 THE PARTICLES 127

Maim., $obe;:;, iii 19a); 'El:l "according to" (e.g. B. I:Iiyya, Hegyon,
( l5a), and others.
f. There are also a few newly-formed prepositions, such as TI' ';l]
"by way of", t:l"pO "before" (Maim., $obe;:;, ii 27), ,~o "because of"
(B. I:Iiyya, Hegyon, ( 5), T"'O "with regard to" (frequent).
g. In the second period, the whole system of prepositions changed
under the influence of Arabic. A host of terms was introduced through
the translations, mostly etymological renderings of Arabic, such as
n';,r::l "without" (bi-ghairi); ,~ ';l] "by way of, by" ('alii jihati), ml]~O~::l
"by means of" (biwiisi{ati), etc. These were then in turn given new
senses and functions by the Hebrew writers.

25. Negation

a. The system of negations with verbs is basically that of MH: ~';

with the finite tenses, r~ with the present tense-( cf. Segal, 471-6).
The use of ~'; with the present participle, which is fairly frequent
in BH (Ges., l52d) and MH (Segal, 340), seems to be entirely absent
from SH.
h. ~'; is, however, used for r~ when two nominal predicates are
coordinated, in accordance with MH usage (Segal, 339ii): ~~m ~';
t:li1'';l] "o~ ~';, t:li1::l "it does not apply to them and is not said about
them" (B. I:Iiyya, Megillah, p. 7); i1C~0'; ~';, i1';l]0'; ~'; nnm nl],m
n'l]~1:I~ ~'i1 p' "the sound of the Pathal). is neither high nor low,
but medium" (B. Ezra, ?.a~oth, f. lb). So also when the second term
has some other negative: FD'; p' ,::l, ~'; t:liDi1 "a name is not a
thing, only a word" (B. Ezra, Yesod Mora, p. 4). This feature is, how-
ever, rare in SH, being mostly superseded by the construction men-
tioned in the following paragraph.
c. An Arabism that entered the language at the very beginning of
SH prose is the habit of continuing every negation in a coordinated
clause or term by ~'; (Wright, ii 303D).
Examples: 10r ~';, n~ t:l'iD... t:l'~~mi1 m'~' 'm'; r~ "before the
creation of existing things, there was neither point of time nor length
of time" (B. I:Iiyya, Hegyon, ( 2); n1:l ~';, miD' i1'; r~ "it has neither
a right nor the force to ... " (ibid., ( 5); l]OiDJ ';'p ~';, ,,::l', r~ "no
word or sound is heard" (ibid.) 'J'J::l ~';, nr ';l] Ji11J t:l';'l]i1 no, r~
128 CHAPTER SIX

l:l ?.lI iO':J "the course of the world does not run thus, nor IS ItS
structure founded on this" (id., Megillah, p. 50); ~?1 n1~i~n n~ ii10 1:J'~
... Cn1~ p'?1n "he cannot measure or divide lands ... " (id., Meshi~ah,
p. 2); nnr n1?0 ~?1 C'ir C':J':J::J ~'::Jn ?~ "do not employ unusual verb-
forms or words" (J. Tibbon, Testament, p. 8); 'n?::J cn... c'p?nn
C'E)P10 ~?1 C'?::J:m "the varieties .... are unlimited and unclassifiable"
(Gers. on Job, Preface). ~? seems, however, to be employed in this
way also in N.W.-European Hebrew, e.g., n1:JP? fE)n ':J'~iD .lIi ?.lI::J
npi~ nn? ~?1 iE)O "a wicked husband who does not want to buy a
book, nor to give alms' (S. Ijasidim, par. 669). r?.lIO r~ n101?n 'i::Ji
C','i10 ~?1 "things one sees in dreams do not add anything, nor
detract anything" (ibid. par. 1138). No cases of this kind are noted
for Rashi by Avinery, nor have I found any myself. Further inves-
tigation is necessary, but it is quite likely that this, like so many
other apparent Arabisms, will turn out to be a heritage from N.W.-
European Hebrew. It may have developed out of the MH ~? r~
~?1 . .. construction by dropping the first ~?
This construction seems to have been avoided by Ibn Ezra and
Maimonides. It is the rule in the authors of the second period.
d. The coordinating ~? seems to have been contaminated with r~
in the following two instances from the Megillah of Bar J:Iiyya, in
which uninflected r~ stands where one would expect the inflected
form: in~ ji1:JO ?.lI ni01.l1 r~1 ~1jn ?~ np::Ji:J ni'~ "shape is inherent
in the body and cannot exist in any other way (p. 11); iiD~ c':J':J.lIn
C?1.l1? c':Jno r~ 1~ ni1~ Cn? r~ "those things that have no shape or
have no effect upon (?) the world". In the first case, i1:J'~ should be
expected; in the second sentence, the second r~ should be C:J'~. It
is possible that B. J:Iiyya wanted at first to write ~?, and corrected
himself, but not completely, so that a hybrid form arose. l
e. Uninflected r~ is also employed when the subject of a nominal
clause is preceded by ~?~ etc., "except" in an exceptive sentence in
which the general term is not expressed (istithnif muforragh, Wright,
ii 336A);2 Cnnp?1 pi ni1m iDi1E)0 r~ "only 'and ye shall take' is
explicitly said in the Law" (B. Ezra, Yesod Mora, p. 10); ?'t!l? '1~i r~

I It must be admitted that this explanation may not be quite convincing. Perhaps

further instances will help to clarifY the construction.

2 SH agrees in this with MH, e.g. ~.,~... ]':::l ]'~ "There is no difference
between ... except ... " (Megillah i 4ff).
 THE PARTICLES 129

ilD:l' cn, 'Oi:J tol:,amlD 'a tol:,tol: OiiEl:l "only he whose belly is full of
bread and meat is fit to walk in the orchard" (Maim., Yesode ha-
Torah, iv 13); 1J1.lJ:l 'lD:Jjn Ctol: ':J n:l'lDn:l :l'n rtol: "only he who has
fallen in sin is liable to repentance" (B. I:Iiyya, Hegyon, f. 8). This is,
of course, only an extended application of the MH rule which
demands uninflected rtol: before the subject, inflected rtol: when the
particle follows the subject (Segal, Di~du~ 334 end), e.g. C'tol:intol: ,jtol: rtol:
"we are not responsible" (Demai iii 5), but ,j'atol:a ,j'tol: tol:,m "and he
does not deem him trustworthy" (ibid. vii 1).
The above rule is, however, violated in the instance: 'n"r 1,r'j ,j'tol:
c"nn '.lJ:l, na'~n "only plants and animals take nourishment" (Falaq.,
Meba#esh, f. 12a). It may be that usage changed in the second period:
I have no data to confirm or disprove it. In any event, the editor
of the work is not above the suspicion of having altered the text
according to his own grammatical ideas, or the MSS, may have
altered the original wording.
f. 1'tol: is mostly uninflected with clause-predicates, see above par. (e).
g. In accordance with MH usage,3 an otiose she- is inserted before
negative infinitives and prepositional phrases: ni'a.lJ " ID' Cip,an
Cip,a:l tol:,tol: ni'a.lJ " rtol: imtol:am ,mtol:a:l tol:'lD tol:a'p' "the earlier has
a definite existence, if not accompanied by the later, but the later
has no existence except if accompanied by the earlier" (B. I:Iiyya,
Megillah, p. 9); m,ran mi.lJ:J tol:'lD ntol:':ljn n'nnlD 1':Jj "it is proper that
prophecy should be not according to the evidence of the constella-
tions" (id., Schwarz-Festschr., p. 33); p m'n' ... ,mtol: llD,a intol: i:li
p m'n, tol:'ID' "one thing induces him. . . to be thus and not to be
thus" (Maim., (lobe;:, ii 26); C',.lJ' C,:Jtol:, tol:'lD Citol:' "tol:i "it is bet-
ter for one not to eat them at all" (id. De'oth, iv 9).

26. Exceptive Particles

a. The normal exceptive construction in SH is that in which the

general term is not mentioned (cf. par. 26e). There are, however,
cases of the more common Arabic construction with introduction of

3 E.g. 11~' ~'iD 1Jjjj •.• 11~' 1Jjjj "they used to say the blessing ... they used not

to say the blessing" (Megillah iv 1); 015) '~P' ~'iD m~ 'll... 015) '~P' m~ 'll
"in order to receive a reward ... not in order to receive a reward" (Aboth i 3).
130 CHAPTER SIX

the general term (istithna' mutt~il): tI:?tI: 1?1~ 1t1: D?117ir J1..., 1J D'171~
D'1'n' D'iD:JtI: "almost all people are mistaken in this, except a few"
(Maim., 1lobe?:" ii 25). BH would in such a case employ 1J?0, e.g.
Gen. xlvi 26.
h. The most commonly employed exceptive particle is the MH tI:?tI:
(Segal, 506). The construction of this particle is assimilated to the
Arabic illa. Thus it can take a complete verbal clause (cf. Reck.,
262.6): 1?'iD~0 ir'ir mtl: tI:?tI: ••• ...,J1 miD17? ?1~' l?Oir ir'ir tI:? "the
king could do nothing but that his brother put obstacles in his way"
(B. Daud, 1labbalah, p. 72). Similarly with DtI: '~ in N.W.-European
Hebrew: pir 1mtl: DtI: '~ r...,tl:ir '017 1'ir 1'ntl: ?~1 "and all his brothers
were illiterate, except that one son" (s. /fasidim, par. 669).
c. B. I:Iiyya follows N.W.-European Hebrew (Avineri, Hekhal, 259)
in employing in addition to this the BH DtI: '~ (Ges., 163); 1'...,J1 1'tI:
D'?Jir DtI: '~ "his words are mere nonsense" (Megillah, p. 50; and
instance in par. 26e). In a hybrid typical for SH, he combines the
BH particle with MH supplementary particle 1J?J (Segal, 507): tI:?
1J?J ?tI:...,iD' DtI: '~ ""~17ir 10 1~'P' "only the Jews alone will wake up
from the dust" (Megillah, p. 49).
d. Ibn Ezra revived the BH P"" (Kbnig, 392fV 11iD?ir 10 171:J tI:? 1:Jn:JtI:
iD11pir '''''~O:::J :::J1n:Ji1 pi "we know of the Hebrew language only that
which is written down in the Bible" (Saphah Berurah, f. 4b), and often.
With istithna' mutta.Jil and 1:::J?J: T~""1 ?:JJ m1t1: D'OiD tI:..." '~ lJ ':JtI: 1711'
1J?J JtI: 11J~ ...,:::J1J p"" "I know that you are God-fearing in all your
ways, with the only exception of honour to the father" (J. Tibbon,
Testament, p. 10). The inconvenience of this construction is that it
gives rise to ambiguity with the positive p"" "only". E.g., the sentence
tl:Jir D?117? J1~ 1? ir'ir'iD irO p"" irrir D?117? iDpJ'iD ?'~iDO? 11~:J 1'tI: "it
is not right for the intelligent man to strive in this world except for
what will be good for him in the world to come" (B. Ezra, Yesod Mora,
p. 16) could, very much against the intention of the author, be trans-
lated also "it is not right to strive only for such things as are good
for the world to come". This ambiguity was probably the reason
that the use of this particle was not continued by later writers.

I The particle does not seem to have been used in N.W.-European Hebrew in

this function.
 THE PARTICLES 131

e. In the second period, also 'n'?,r, corresponding to the Arabic ghaira,

is used (cf the instance from Falaquera above par. 26e).

27. Some Coordinating Conjunctions

a. As with other particles, SH uses the MH and the BH conjunc-

tions together. While it has made a great deal of innovations in the
sphere of subordinating conjunctions (see below under adverbial
clauses), it is fairly conservative with regard to the coordinating ones.
h. In Arabic, waw copulativum is omitted between adjectives refer-
ring to the same noun. This is occasionally found in SH writers,
too: n'?El~ n'?EliV i1:l~p n~":J ~'niV lJi" "he knows that he is a puny,
lowly, and dull creature" (Maim., Yesode ha-Torah ii. 2); "o'n m
ro~n prnn "this strong and firm fundament" (B. 'Abbas, f 9a).

c. Waw is sometimes equivalent to the Arabic fa) al-jawab "so that"

(Wright, ii 30f, Brockelmann, 302i). This occurs already in Saadiah's
Hebrew: nniV:J ~'? ,:J':J, "OiV ~'? no'? "why did He not guard him, so
that his offspring would not have perished?" (lfiwi Polemic, stanza 7).
SH examples: ~":Jn 1'? n'?o',. . . 1'? :"j'?niV no '?:l :I'iVn'? ':J:J n~'n Cl~
':J~ Cl:l 1'? n'?o~, "if you want, my son, to regain all you have missed ...
so that God may forgive you, and I too" (J. Tibbon, Testament, p. 9);
n:J,n i'OEl" ~lJO n1:Jn''?.. ,m':l n'nn ~'? "his intention cannot be
to enjoy a little so that he should lose a great deal" Maim., De'oth,
v 12). These constructions may be unconscious Arabisms, but they
may also be in imitation of BH as understood by the grammarians
of the period. Thus D. ~m1:Ii found a fii) al-jawiib in 0'0'" :IiV"
(Psalm vii 6), which he understood as "so that he might overtake
and tread down".
d. ,~ "or" is used in the same manner on the lines of the Arabic
au with the subjunctive (Wright,,ii 33B) to mean "or else". It is then
equal to the MH ,~'? Cl~' (Segal, 489): '?~'iV' O1:J~" Cl"':llJ i'OlJ'iV:l'
':J:I,n' ,~ n"m n1"o~n n1'~on '?:l0 nn~ '?lJ ,,:JlJ'? "if an idolator
forces aJew to transgress anyone of the commandments of the Law,
or else he will kill him ... " (Maim., Yesode ha-Torah, vI).
e. For "also", the BH m is used more frequently than its MH equiv-
alent :"j~. By one of the hybrid processes typical of SH, it takes the
MH peculiarity of being prefixed to the subject of the coordinated
132 CHAPTER SIX

clause even if its force applies to another part of the sentence.' E.g.,
n'niD~ iiO~ ~in !:lJ ':::l iir.l~i "it has been said that drinking, too, is
forbidden on that day", (B. Ezra, Yesod Mora, p. 7); iiO~ ~in !:lJ mi1l
n~'ni~ "it is also forbidden to wash on that day" (ibid.) "if the
camel-nature caused that one animal to fly in the air" nr.liiJ ~'n !:lJ
p m'n~ rr.ln ~:::l~ "then it would cause the whole species, too, to
have the same characteristic" (al-Fakhir, {(obe?:" iii 2a). By BH usage,
the gam could in these instances apply only to the word it precedes,
so that the first quotation would mean "on that day, too, drinking
is forbidden (i.e. as well as on other days)", obviously against the
author's intention.
f. Following a rare BH use (Gen. xxx 8, Psalms xxxvii 25, etc.), gam
is employed by first-period writers as an ordinary copulative con-
junction equivalent to we-, e.g., !:lJ !:l'i:::lrJn !:l'i~in n~~ ~:::l ii~.ll~
!:l'in~ "for all the reasons mentioned and others (B. Ezra, Yesod Mora,
p. 20); pnp,n 'i5:)O i'r.l~ m 'ir.l~n~ !:l'POi.ll m "they were busy with
the Talmud and also studied grammatical works" (J. ~mJ::ti, s. ha-
Galuy, p. 1).
g. 'i.lli is used like !:lJ and t"]~ as in MH:2 !:l'~~.llr.liD'n 'i5:)O~ 'P~ n'n 'i.lli
!:lJiiD~~i "he was also learned in Arabic language and literature"
(B. Daud, {(abbalah, p. 71). Out of it, a new conjunction was devel-
oped-iD 'i.lli "furthermore": i1InniDi1I iJtI;~r.liD 'i.lli "further we find
'wehishta~awah'" (B. Ezra, Za~oth, f. 8a); n~'nn !:l,tI; ~iD in.ll' r~iD'r.l !:In
... n~i'Jn n~i~n !:lniD 'i.lli "they firstly give man pleasure, and also
are the great gift ... " (Maim., Yesode ha- Torah, iv 13). As in all other
cojunctions the she- in it can be replaced by the BH ki: ~:::liJ ':::l 'i.lli
ir.ltl;~ "and further we can say ... " (B. Ezra, S. ha-Shem, f. 3a).

I E.g., n'J'J i1~i ~li1 ~~ "he also saw a skull" (Aboth i 6). Also in Shebiit ii 2,
iv 2, Sotah ix 6, I:Iallah ii 5, Sukkah iii 7, Nedarim viii 6, etc. In Job vii 11, CJ
is used in a similar way to apply to the predicate, while preceding the subject, but
there it is not coordinating, but intensive. Perhaps it foreshadows the MH usage,
as ~~ was originally an emphatic non-coordinating particle, and remained so in
1"El~ "even if".
2 E.g. Shebiit viii 10, 'Erub ii 5, RH ii 8, Taanith iii 6, etc. In the Hebrew of the
Palestinian Talmud, simple ill) is used in this function (Segal, Wdu~, 309). --m ill)l
does not occur in the Mishnah, and is not mentioned by Avineri as occurring in
Rashi.
 CHAPTER SEVEN

WORD ORDER

28. the Verbal Clause with Oldect

a. The normal non-emphatic order in SH is either Subject-Verb-

Object or Verb-Subject-Object. The relative position of subject and
verb seems to be irrelevant.
This word order is very often changed for reasons of emphasis.
There are no arrangements that are peculiar to SH as against BH
and MH, as the order of words in those dialects was, if possible,
even more sensitive to demands of emphasis than the SH one. Certain
schemes appear to be more frequent in SH, though, than in the
older dialects.
h. The order Verb-Object-Subject is in BH restricted almost entirely
to poetical language (cf. Ges., 142f).1 In SH it is rather freque~t
and there seems to have been a tendency for it to become the distinc-
tive feature of dependent clauses, a tendency that is also noticeable
in Modern Hebrew: l:I'OJni1 .,n:J'nJ .,J.".,i1I .,'?.l.1 .,J.,,'iiri1l .,J.,o,p I:IJO~"
"J'J~? 'iD~ "and indeed there pointed it out to us before and taught
us its method the scholars who lived before us" (D. ~ml).i, Miklzlol,
f. 1a); C1rn?:l:l C1'~:lIVi1 C1':l:l "''11i1 C1'~:lIVi1 mlVl1 .,'?:l1V ',ntl; "after the
ten tribes had gone into exile, there begat sons the tribes in their
exile" (D. ~ml).i, Hosea Comm., p. 16); .,',:l, "J':J' ~?iD '~ ?.l.1 =-J~
I:I'?'OJ =-J?~ "although there would not understand his words a thou-
sand fools" (Abr. Maim., (lobe~, iii I 7b); 1:I'~i1 I:Im~ i1iD.l.1'iD T'~ r~
l:I?iDi1 "it is not for the perfect man to do these things" (Falaq.,
Meba~~esh, f. 18b). This order is also frequent in the paytanic prose
of the A&imaa~ Chronicle, e.g. 1:Ii1'J':J l:I'iDJ~i1 i1:J"O .,iD.l.1., "and there had
a quarrel the people amongst themselves" (p. 5, in rhyme).
c. The normal word-order is, however, quite common in dependent
clauses, e.g. i10ni1 mp? I:I,.,J iiri1 j'J.l.1i1 'iD~J "as this thing causes

1 Cf. the instances given by Ges., the only one from a prose passage is Num.
v 23. The order occurs occasionally in SH (Segal, ~du~, 395.2) and in other
Semitic languages (Brock., 26ge).
134 CHAPTER SEVEN

eclipses of the sun" (B. I:Iiyya, Megi1lah, p. 57); 1rJJ '?~il itD~ in~.,
i1r.~:Jn tDP~'? l~'?~ "and since God has given you the desire for wisdom"
(Falaq., Meba#esh, f. l8a).
d. Object-Verb-Subject is another scheme that is very rare in BH
(Brock., 269h; Konig, 339m; Kropat, p. 59) and does not seem to
occur in MH in enunciative sentences at all (Segal, Di~du~, 395.2).
Its comparative frequency in SH is perhaps connected with the fact
that it is common in Arabic (Reck., 7l.1). It occurs, however, not
infrequently in liturgical style, e.g. "Jmil~ il~i il~il~ "with abound-
ing love Thou hast loved us" (Singer, p. 39). Mostly, however, this
position is, in the liturgy, due to chiasmus. In SH, where chiasm us
is rare, it occurs in ordinary sentences. Examples: il'?~il t1"1r.l:Jnil 'ntD
~il''?.l) ~'ip.,n "'ilI ,?"r "J't1"1~i .,n~'tD "our Rabbis esteemed these two
sciences highly and inquired into them" (B. I:Iiyya, Schwarz-Festschr.,
p. 30); nr.li'? ilr.l'?tD ir.l~ p.,OElil i1r "this verse Solomon made in order
to indicate" (M. Tibbon, Cant. Comm., p. 5); ~tDn il'?'Jr.lil ':J .l).,i'
tDi.,Pil mi~ ilr.l'?tD "it is well known that Solomon conceived this scroll
under divine inspiration" (ibid., p. 6).
e. Konig (loc. cit.) points out that the emphasis is on the object in
this scheme. This may be true for BH,2 but is hardly supported by
the above quotations. The emphasis in them is rather one of the
whole statement. If anything, it is the subject-verb complex that is
emphasized. This is quite in keeping with the tendency of SH to
put the emphasized parts of the sentence at its end.
f. The material collected does not show clearly what happens if two
coordinated objects are affected by these inversions. In the one
instance at my disposal, the second object is separated from the first
and placed into what would be its normal non-emphatic position:
~n~.,~r.l" ~i~ 'J~ ~ilr.l .,Jil' ~'? ':J "for men do not enjoy them and
their benefits" (D. J>iml)i on Psalm i 4).

29. Adverb and Predicative

a. The normal position for these two elements (which are not prop-
erly distinguished in SH) is after the verb. In BH they very rarely

2 It is definitely true for the Arabic examples in Reck.

 WORD ORDER 135

precede the verb, and in that case they mosdy stand at the begin-
ning of the clause (e.g. 1 Sam. xx 8, Is. xxiii 12, Jer. xx 6, xxxii 5,
Micah i 10, Job xix 23). Only twice, in poetry, they stand between
subject and verb: plD' !:l'1.j' =,1n'l 1'11:::lr "Zebulun shall dwell at the
shore of the seas" (Gen. xlix 13) and i11.ji1' !:l''I'In:J :::l~11.j'l ':::l'l "my
heart shall sound for Moab like pipes" (Jer. xlviii 36). Once the
adverb precedes a participial predicate in BH 'I'~1.j ','1.j r~ "there
is no one to save from my hand" (Deut. xxxii 39 = Is. xliii 13), and
twice before infinitive: 1'li1 1':J":::l 1:::l~ ~'I "they did not wish to go
in His ways" (Is. xlii 24) and :::l~'m'l 1Q.!.l r~ "none is able to with-
stand Thee" (2 Chr. xx 6, cf. Kropat, p. 60). All the latter exam-
ples, too, occur in poetical passages.
In ordinary MH, this scheme does not seem to be found at all. 1
h. This rare poetical word-order is, however, extremely frequent in
liturgical style, e.g. m'p 'll'l 1m11.j~1 1m:J'I1.j "His kingdom and His
faith endure forever" (Singer, p. 42); i1"1D 1:Jll 1'1 'I~'ID' ':J:::l1 i11D1.j
"Moses and the children of Israel sang a song unto Thee" (ibid.,
p. 43); i1'i1n 'I:J'I 'Elll:J 'IDEl:J "let my soul be unto all as the dust"
(ibid., p. 54).
c. It is just as frequent in N.W.-European Hebrew. Thus three
instances occur in Rashi on Gen. 1.1 alone: ~:::l n"1i1'1 '1.j~n !:l~1
1~':::l:J i1'1'nn 1'1~1D "and if you say, he wants to indicate that, if they had
been created first ... " and i1n'i1 'm !:l'1.ji1 n~":::l ~'p1.ji1 i1'1'j ~'I "Scrip-
ture does not reveal when the creation of the water took place". It is
also common in the Sepher Ijasidim, e.g..l.l1r n:::l:JtD ':::l~1.j 1j1~'1.j ',i1
"then he wastes semen with full intent" (par. 1144); m'l i1~':J !:l~
!:l"1.j1~ :::l1t!l:::l ,m' !:li11D "if they consider that they will give a better
judgment" (par. 1378).
d. It came into SH not only via N.W.-European Hebrew, but also
through the Saadianic prose style: !:l'~'p:Ji1 i1'1m1 p'ln 'I~ i1n':JEl i1n~ p
"thus thou hast turned to those called share and heritage" (Saadiah,
ljiwi Polemic, st. 40); !:l',j1:::l 1n1:::l ID' "some betray His faith (ibid., st.
13); !:l"PEl1.j 1:Jm~ !:llli1 'Ill ..• f''I1.j 'I:J "every orator. . . we appoint

1 A case like Cl'~n:::lj iElO~ TiU.lli.l '?~1 "and all thy deeds are written in the book"
(Aboth ii I), in a passage of elevated and sententious diction, belongs practically to
the liturgical style. In 1i.lni' Cl'i.liUil 1i.l1 "and from Heaven they will have mercy
upon her" (b. Yebamoth 12b), the words "from Heaven" are construed as if they
were a noun, being merely a circumlocution for "God".
136 CHAPTER SEVEN

over the people" (id. JQ,R xiv. 45, line 22); 'J'il~~ in~ ilril 1'iD~~ "our
Lord elected this language" (Menal).em b. SaruJ..<., Mabbereth, p. 1).
e. In SH, this construction is almost as common as the "normal"
one. 2 The emphasis it implies is often very slight. Examples: 1:lJ'~
"JEl~ l:J'~~OJ l:JmJ~ "they are not bodily existent before Him" (B. I:Iiyya,
Hegyon, f. 4); l:J'O'il ~J mm.\] in~ n'J~n ~.\] .\]'piil ni'~ "the shape
of the sky remains all the time in the same state (ibid., f. 1); "when
they see the sun eclipsed" l:J'inElO' l:J'~il'O i'O l:Jil "they immediately
begin to hope and fear" (id., Megillah, p. 113); ~'il Ti~O 'J~ l:J"iliDJ
i~pO l:Jil~ "when the day is long with us, it is short with them" (id.,
'Ibbur, p. 7);3 il~iJ m 'JiDil ~'iDil ilJil "the second Sheva seems to be
quiescent" (B. Ezra, Zabot, f. 2b); 'O'pO~ 'm'il~ l:Ji~il 'm~ iliD'\]' ~~nil
~P'iD "sin is committed by man while he stays in one place" (I5iml).i
on Psalm i 1); il'il l:J''\]'~il 10 'o~.\]~ iD'J~ "Enos himself was one of
those who erred" (Maim., 'Akkum, i 1); mo'~ ilril l:J~''\]il P r~ 'J
"J'iD l:J~''\] ~iD ,.\]~~~ n'iDOil "for between this world and the Messianic
age there is no difference as far as the physical structure of the world
is concerned" (Abr. Maim., (lobe?, iii ISb) ~J n'iD~i~ iliD'\]O p.\]~
l:J'i'.\]O ilr~ ilr l:J'~'nJil "with regard to the creation, all scripture pas-
sages confirm each other" (al-Fakhir, (lobe?, iii la); l:Jm'o~~ l:J'n~~o
mnJ 'niD~ l:J'p~m il'\]iil "I found them to be divided, with regard to
their evil beliefs, into two groups" (B. 'Abbas, f. 11 a); ni:Jil~ m~
'~JiD l:Jir' "thereby his intelligence will perforce become purified"
(ibid., f. 5a); '\]'~p iiD'.\]il ~J~J~ i'~Jil ~J'J 'J il~i "convince your-
self that the star of honour is fixed to the firmament of riches"
(Falaq., Mebaf!,f!,esh, f. 4b); ilO~iD mn~~ilil 'niDO iln'il ilO~ "in what
respect it is the more perfect of the two virtues" (ibid., f. lOb);4 lJ
iO~ il~'nn~ ~'il "so he said first ... " (Gerondi, Sermons, f. 61 a).

2 SH is in this respect quite unaffected by Arabic, where this position of adverb

and predicative was very rare, if it occurred at all (er. Reek., 71.3).
3 This example shows clearly that the feature has nothing to do with chiasmus
(Ges., p. 352, note I). Chiasmus is altogether rare in SH.
4 In the Meb~esh the emphatic position is virtually the normal one. Exigencies
of rhyme are not a sufficient explanation as it would have been just as easy to
rhyme with other word-order. It is the emphatic and elevated diction of the Maqama
that caused the writer to prefer it.
 CHAPTER EIGHT

THE COMPOUND SENTENCE

30. Extraposition

a. All West-Semitic languages have gready developed the facility of

making a sentence, as it were, a predicate to one of its own ele-
ments.! The difference between this and the ordinary construction
was by some scholars believed to be a particular emphasis on the
extraposed word (K6nig, 3411, Brock., 271a, Ges., 143c; Segal, 211).
This is, however, in many cases quite impossible, as e.g., in the quo-
tations below in which a m "this", referring to something just men-
tioned, is thus extraposed. Emphasis, moreover, was effectively achieved
by altering the normal order of words without extraposition, as seen
in the preceding chapter. Nodel, Der zusammengesetzte Satz, par. 20a,
shows that in MH extraposition regularly takes place under certain
purely syntactical conditions, without any reference to emphasis.
It seems preferable to avoid seeking any difference of meaning and
to treat extraposition with S.R. Driver (Hebrew Tenses, p. 196) as "one

I In Assyrian, this construction has no importance at all (Brock., 271 a end), but
hardly for the reason given by Brockelmann: that in that language the subject comes
in any event in the beginning of the sentence. This would have been an added
reason for extraposition in order to give prominence to other parts of the sentence.
It is preferable to assume that extraposition as a normal grammatical process in
Semitic developed after the separation of Accadian, or perhaps died out there
through the influence of the non-Semitic substrate. The situation is complicated by
the fact that extraposition is common in Egyptian, where it follows the same prin-
ciples as in Semitic (Gardiner, Egyptian grammar, 146-8). Extraposition also occurs
in modem European languages as a strongly affective device mainly restricted to
poetry and oratory, e.g. "The rain, it raineth every day"; "Zionism-what is that
to me?" (cf. Jespersen, Ana[ytic ~ntax, 12.1). As a regular linguistic process it occurs
only in colloquial Italian (personal information from Prof. A. Momigliano), e.g., "i
piccoli debiti bisogna pagarli". (Colloquial German has developed a device with the
same effect, but different through employing a special set of pronouns and keep-
ing the original case, e.g. "den Miiller, den kann ich nicht leiden"). In the Western
European languages, the only commonly employed form of extraposition is that
introduced by "as to ... ", French "quant it ... ", etc., a process which Arabic has
paralleled by its ammii, Modem Hebrew by its -'? iiD~. Modem Hebrew shows its
affinity to European ways of thinking by having lost the whole process of extrapo-
sition beyond the degree to which it is usual in the latter.
138 CHAPTER EIGHT

of the commonest and most characteristic artifices of which Hebrew

avails itself for the purpose of avoiding an unwieldy sentence. 2 This
would also account for the phenomenon that extraposition is more
frequent in the later stages of the Semitic languages, and compara-
tively rare in colloquial speech. 3 It is a natural product of the ten-
dency existing in Semitic of dissolving the sentences into a series of
self-contained units connected with each other by a system of antic-
ipatory and retrospective pronominal elements.
h. While it must be kept in mind that the extraposed element is
really the complex which serves as predicate, it is for practical pur-
poses easier to consider the process of extraposition from the point
of view of the function in that complex of the pronoun referring
back to the major subject. As a shorthand description, we can say
that the subject is extraposed from its original position in that func-
tion. If we thus classify the elements that can be extraposed, we find
that all parts of the sentence are affected by this process, except the
predicate (the latter can only be extraposed by the device described
above, par. 22c).
c. The subject (as in BH, cf. K6nig, 341 a), though rather rarely:
tI;'i1 i1ii~ Wtl; O'::l:Ji:Ji1 ni:J'?i10 mi~ '?lJ 'iOlJ'? I:Jni:J i1'i1 tI;'? "they were
not able to discover the shape of the courses of the planets~what
shape it was" (B. I:Iiyya, Megillah, 113). 4
d. The direct object (for BH, cf. K6nig, 341 c): 1:l':J'i~ iJtI; 1'tI; m ':J
::l'? '?lJ ini'?lJi1'? "for this~we must not let it enter our minds" (B. I:Iiyya,
Hegyon, ( 3); l:l'OiD::l i1'i1 iiDtI; iDn'ni1 initl; I:liJ fitl;::l mi1 iDn'ni1 "this
event on earth~there caused it the event that happened in the sky"
(id., Megillah, p. 114); l:li'mi1'? ::l'n i1ntl; ::liplJ' itl;iD '?::ltl; "but other
Jews~you are obliged to warn them" (id. Hegyon, f. l3b); iOiOi1 i1ri
1i::ln '1 inii1' 'i . •. initl; i:m "and this admonition. . . R. Judah b.
Tibbon composed it" (J. Tibbon, Testament, p. 1); initl; i1iDlJ' tl;C!lni1
C!lpiiD impO::l ini'i1::l l:l,tI;i1 "sin~man does it while he stays in one

2 Reckendorf, who in his !fyntaktische Verhdltnisse still attempted to account for the
phenomenon by emphasis, changed to Driver's view in his Arabische !fyntax.
3 It may, indeed, be argued with some justification that it was the possibility of
resolving any difficulties by extraposition, that prevented the development of a more
complex subordinative structure in the Semitic languages. Modern Hebrew, which
gave up extraposition, is now forced to develop its subordinate clauses.
4 This is a classical example of the great advantages of extraposition. To express
this idea in any other way would have required a paragraph.
 THE COMPOUND SENTENCE 139

place (D. ~m}:li on Psalm i 1); i1tZi?iDr.l in~:::l 1iiOni1 ii1:J'iD' '?.liiEl '?:l i1:Ji1
Cl'JEl "every action-shortcomings affect it in one of three ways"
(Zera}:liah, Hechalut;::., vii 96); i1i~:::l'? i1~ii 'JJi1 n~r:::l i1:::l'Oi1 "the reason
for this-I wish to explain it" (ibid.).
e. An adverb: i1r ~'?:::l i1r ~~r.l' ~'? in' Cli1'JiDi "the two together-one
is not found without the other" (Maim., -!lobe?:, ii 25); "they have
weapons" 1C!lP C!lnr.l i'?'El~ iJ'? r~ iJ'?i "but to us-to us there belongs
not even a small needle" (B. Verga, Shebet Jehudah, p. 4).
f. With adverbs formed with a preposition, it is more usual to extra-
pose only the nominal kernel (as in BH, cf. K6nig, 341f.): ... i1:::lii1
i''?.li i:::li~ i1r 1i:l:l "much of this kind-I shall speak about it ... "
(B. Par}:lon, Mabbereth, p. xxii) '?:::l~ i1r.lipn Cli1'? i1'i1n Clmr.l:::l Cl'P'i~i1
i1r.lipn Cli1'? i1'i1n ~'? Cl'.liiDii1 "the righteous-they will have a resurrection
after their death, but the wicked-they shall have no resurrection"
(D. ~m}:li on Psalm i 5).
g. The same applies to the indirect object: "he divides them into
three classes" Cli1''?.li ni:::li:l n'r.li1:::li1 iDmi1 Cl'?i:li "and all of them-the
animalic soul rules them" (B. J:Iiyya, Hegyon, f. 11); 'JElr.l i:::l 'mr.l~i1 i1r
i1'?:::lpi1 "this-I believe in it because of tradition" (Maim., -!lobe?:,
ii 25).
h. Only the nomen rectum of a construct can be extraposed (as in
BH, K6nig, 34Ig): iDiipi1 mir.l Cli':::l i1'?:::lp Cli1 iiD~ ,?"r iJ-ni:::l1 "our
teachers who-in their possession there is a tradition from the divine
spirit" (B. J:Iiyya, Megillah, p. 14); '?i:l~'? iiDElJ i1i~nn ~'? n.lii iD'~ '?:l1
iJp'r'iD i:::li "and every sensible man-his soul will not lust to eat a
thing that will harm him" (B. Ezra, Yesod Mora, p. 10); Cl':l'iOi1 Clm~ '?:li
Clip.li ~ii1 in~ 1'?'~ "and all those branches-one tree is their ori-
gin" (Maim., -!lobe?:, ii 24).
i. The regens cannot be extraposed without its nomen rectum: '?:::l~
i''?~ i1:lr ~'? i1r.l:lni1 pp "but the acquisition of science-he did not
achieve it" (B. J:Iiyya, Megillah, p. 5); Cli~i1 '?:l:::l i'J.li r~ mnQi1 '?:::lpr.l
"the recipient of the gift-there is no one among men as mod-
imr.l:l
est as he" (J. Na}:lmias, Aboth Comm., f. la). Sometimes a whole
construct chain is extraposed: Cli1iD m'J.lir.l i~i:::lr.l iDmn 'p'?i1 i~iDiD 'El'?
Cl'iom "because the remainder of the parts of the soul-it is clear
from their functions that they are transient" (Gers., Milbamoth, f. 4a).
k. Sometimes the elements that are extraposed together make up
considerable groups: ~'~ir.l i1'i1 ~ii1 ••. imJr.li~ iniin m'i1'? i1~iiD '0 '?:li
140 CHAPTER EIGlIT

moo t:lil'?,l] "and whoever wanted his learning to be his livelihood-

he spent his wealth upon them" (B. Daud, (Cabbalah, p. 72); n1,l]c;)i1i
il'?,l] O::lnJil ~1il t:l'~il "0 ~~'tD "and the mistakes that come out from
a man's hand-he is the one that is blamed for them" (J. Tibbon,
Testament, p. 7); 1n1Jn1J,l] ,l]1,'1 '1?J il"~Pil ~1iltD ln1Jil "he who gave it,
namely the Holy one blessed be He-His modesty is well-known"
(J. Na}:lmias, Aboth Comm., p. la); il"pnil ,~o iltDpn 1~i' mtD il01
1~ "and that which behaves in this manner-investigation of it is
very difficult" (Gers., Mill;amoth, f. 48b); ?"ri t:l1~iP 1? il01'il ?~1 m
il~il~ ?tD 1'i10' "this and all that is like it--our sages called it sufferings
of love" (ibid., ( 8b).
1. The element is sometimes extraposed not merely out of a clause,
but out of a complex sentence with several dependent clauses, and
may thus stand at a great distance from the pronoun representing
it: n10~n~ P0,l]ni10 ... 1i1i,l]JO ~? 1?0 P 1?0 il'iltD '::l ?,l] ~~ ~1il 'iiltD
"for he-although he was a prince, the son of a king, they did not
prevent him from studying the sciences" (S. Tibbon, Eccl.-Comm.,
( 16a); i1iil'tD ilO '~i~il in10ilO 1~ il'il'tD itD::l~ 'n?~ 1ion n1P'? t:l'~il
1? po 1JOO "man-because of the tenuity of his matter it is impos-
sible that there should be in him enough material surplus to protect
him" (Gers., De'oth, f. 7a).
Ill.The highest degree of complexity is achieved in those rare cases
where two distinct elements are extraposed from one clause, and are
put into relation to each other quite apart from the relationship their
representative pronouns have within the minor sentence. E.g. lOril
1JOO 'i::lil? t:l'?1~' t:lJ'~ t:l?1~ t:l?1,l]il 'i1~' t:l'~1i~? "time-with regard
to all the creatures of the world-they are not able to exist apart
from it" (B. I:Iiyya, Megillah, pp. 7-8). The subject and the indirect
object have both been extraposed in order to establish between them
a special relationship that could not be expressed within the frame
of the minor sentence.
 CHAPTER NINE

THE VERBAL NOUNS

3 1. TIe Le-irifinitive

a. Alone of all Semitic languages, Hebrew had in every historical

period of its development two classes of verbal nouns. Their for-
mation, and the distribution of syntactical functions between them,
changed, but their number was always two. When one became obso-
lete, it was replaced by another form and when through some cause
three verbal nouns existed, the resulting confusion soon reduced the
effective number of categories to two.
h. In BH, we have at the outset the absolute infinitive and the con-
struct infinitive. Already in the earliest books, a construct infinitive
with le- prefix began to appear as a non-distinctive variant of the
plain construct infinitive (Ges., 45fg).1
At some stage in the development leading up to MH, the le-
infinitive also took over most of the functions of the infinitive absolute.
In literary MH, the infinitive (always with le-, cf. Segal, 344) is con-
trasted with the nomen verbi (Segal, 217), which served for the old
infinitive construct in its more definitively substantival functions.
The later forms of Hebrew, from the liturgical style onwards,
revived the infinitive construct, but used it as a non-distinctive vari-
ant of both the le- infinitive and the nomen verbi.
c. This is still the situation in SH. The normal infinitive in all ver-
bal functions is the le- infinitive, which we may henceforward refer
to as the infinitive. Arabic, which possessed no infinitive but only a
verbal noun doing service for all Hebrew infinitival forms and as a
common abstract noun, seems to have made little impression in this
respect. On the contrary, the infinitive slightly extended its functions
in SH beyond those it possessed in H.

I The le-, which originally only preceded the final adverbial infinitive, became,

just as the English "to", an integral part of the infinitive in all its functions.
142 CHAPTER NINE

d. The infinitive can be employed in all syntactical functions in

which a noun can stand:
As subject (in BH cf. Brock., 25g; Kahan, Verbal-nominale Doppelnatur,
p. 29; in MH in the construction rWil? ,EJ'O "he will surely be" cf.
Segal, 350): 1iDnil lt1:~? lt1:iDj "it remains to explain the darkness"
(B. I:Iiyya, Megillah, p. 23), nw"tI: 'niDO 1~P l:liD m'il? pn' tI:'I "it is not
possible for a noun to have less than two letters" (B. Ezra, S. ha-Shem
f. la). For further examples, see under clause-predicates (par. 36).
e. As predicate (rare in BH, cf. Brock., 91 b; Kahan, p. 33, in MH
after 11', e.g. m.l)iD 'J~ "O.l)? 1:l':l?O 'j~ 11' "it is the habit of princes
to rise at the third hour" Ber. i 2, also Pe'ah viii 4; m. Sh. iv 10;
Erub. viii 5, etc.): iliD.l)O? ntl:~'I ,m:l "its potentialities are such as for it
to become reality" (B. I:Iiyya, Megillah, p. 15); m'il? 1~Pil nnEJil '~'1
iltl:lj m "1ntl: "most occurrences of PathaJ:! are such as there to be
after it a quiescent consonant" (B. Ezra, Za~oth f. 6a); "I asked the
greatest men of our generation what its meaning was" I:ln~'iDm
r'~1 mp~ mn,? "but their answer consisted in giving evasive argu-
ments"2 (B. Ezra, Yesod Mora, p. 6); 1iDOj m'il? l:l,tI: 'liD 'ntl:'1~ 11'
,'.l)1 lntl: "iD.l)O~' "m.l)'~ "it is the nature of man to be influenced
in his opinions and actions by his associates" (Maim. De'oth, vi 1).
f. As object (in BH, cf. Kahan, p. 31): mtl:l? 'jOO .l)j,O n"il "you
did not allow me to see" (J. Tibbon, Testament, p. 9); CinO:Jn iUnil tI:'I
,j? 1:l'.l)"'O n"il? I:lnJ'~' "their wisdom and knowledge did not suffice
to tell us ... " (B. I:Iiyya, Hegyon, f. 5); ilO:ln iDP~? 1~?~ 1nJ 'Itl:il "God
has put it into your heart to seek wisdom" (Falaq., Meba#esh, f. IBa).
For further examples, see par. 33.
g. As predicative part of a double accusative: rWil? iDl'il i1r 'OiD
tl:n:lOOtl:?' mtl:'I "they made this Derash to be a sign and a mnemonic
device" (B. Ezra, Saphah Berurah, f. 5b).3
h. As adverb with a final implication (Ges., l14g): ntl: t"]'PO tI:'iliD
1?'~Jil?' 10'nnil? ,'nnn n'tI:~mil "which surrounds the things that are
beneath it in order to shut them off and to limit them" (B. I:Iiyya,
Hegyon, f. 4); il?J.l)il 'p ?.l) ml'il? 'p 1nl'~ "its shape is a straight

2 Literally "to push away with a broken reed". This is a blend between the com-
mon Mishnaic mp~ i1n1 "give an evasive reply" (e.g. b. I:Iullin 27b, also tDP~ i1n1)
and the Biblical expression r'~' mp "a broken reed" (Is. xlii 3).
3 The infinitive in this example can, however, also be explained as a final one.
 THE VERBAL NOUNS 143

line, to symbolize the radius of the circle" (B. Ezra, ;:,a~oth, f. 1b);
n~~ nnp? I:ln'Ji11 l?n "he went and gave them up in order to marry"
(J. Tibbon, Testament, p. 10).
In inverted word-order, there seems to be a tendency to make the
adverbial infinitive follow immediately upon the verb, e.g., ""'17n? ...,o~
1:l'~'~Jn 1n~ nr ?17 "there said, to draw attention to this, the master
of all prophets" (B. 'Abbas, f. 4a).
i. In BH already (Ges., 1140) there are many instances in which a
le- infinitive stands in a loose connection to the predicate, so vague
that it might almost be treated as an apposition to the sentence as
a whole rather than to any of its parts. It covers a variety of notional
relations, as motive, circumstance and closer definition. 4 This func-
tion of the infinitive appears to have been lost in MH but was revived
in N.W.-European Hebrew, e.g. n...,:>o~ ?,5:l'? n...,~o I:l,' ~'n~ !:l1~ ?17
mp'J'n~ "for it is a day for misfortune because of croup affecting
little children on that day" (Rashi on Gen. i 14).
k. In SH, this construction was very frequent, although it had no
parallel in Arabic or Romance. E.g., 1'~?n mp'i I:lno:>n~ I:l·...,n~ ~.,
I:ln'mp?nO' I:l'J'J~n n17i? "there are others whose speciality is gram-
mar, i.e., to know the forms and their classifications" (B. Ezra, Yesod
Mora, p. 1);5 i~:>Jn I:l~n in'? nn~o,~o n~ipno i1J'~ "she does not
cleanse herself of her impurity by declaring the unity of God" (ibid.,
p. 5); mn~?' ?':>~? nrn 1:l?'17n m~n l...,i ,·...,,17JO I:li~n iO?' n?'nno
mo~?, "at first man learns in his youth the lustful ways of this world,
namely to eat and drink and rejoice" (D. ~mJ:li on Psalm i 1); ...,iV5:ltl:
f5:lm 1'~"" l? nn? I:l'o~n 10 lnr17' I:l~ ""~'nn~ "perhaps you will find
the right way, if God helps you by giving you good will and energy"
(J. Tibbon, Testament, p. 5); ·...,~i '?~~ ...,o~? ...,nOJ ?~ 1n17i f'5:lP' i'O
no:>n "at once the headstrong man jumps to conclusions, saying that
these are words of wisdom" (Maim., (lobe?:;, ii 25); I:l·n...,~non I:l•..."...,~
I:ln':>...,i 17'''''~ n:>?? I:ln'?17 "cursed be those who join them by walk-
ing in their evil ways" (B. 'Abbas, f. 7b).

4 A similar use is sometimes made in English of abstract nouns, e.g.: "her face
was very pale, a greyish pallor" (cf. Jespersen, Ana!Jtic Syntax, 20.4).
5 It is impossible to consider the infinitive in this and the following quotations
an attribute, and the whole as a case of infinitive in nominal function. If the infinitive
were replaced by a noun, the construction would be quite different; at least a 'J'~,
IQ'" would have to be inserted.
144 CHAPTER NINE

1. Infinitive as attribute to a noun: CiDil ntl: ::l'iltl:? 1'::lQ? nnEl 'il'iD ',J
"that they might be a help for the intelligent to love God" (Maim.,
resode ha- Tora, ii 2); 1::l'? iD"Pil 'iDJtI: Ji1JQ i1JJ "thus is the custom
of holy men to speak" (B. Ezra, Saphah Berurah, f. 9a). Perhaps also:
mil '?,niT tl:El1? nlJ ,? In1j "he gives him strength to overcome this
illness" (B. Hiyya, Hegyon, f. 11), but in this last instance the infinitive
may be taken as final.
In. The nominal force of the infinitive is not sufficient for it to be
governed by prepositions. That is the special function of the nomen
verbi. The only exception to this in MH is the infinitive with min
after verbs of preventing, e.g. tI:'::lil?Q ltljQj "they refrained from bring-
ing" (I:Iallah i 7, cf. Segal, 346), corresponding to BH inf. constr. as
in n,?Q 'il 'j'~ll "the Lord has restrained me from bearing" (Gen.
xvi 2). This feature of MH penetrated into Saadianic prose, e.g.,
"Jt?Q niDC!lj "thou hast omitted to mention" (lfiwi Polemic, st. 19)
and "?tI: ll'iD?Q C::l? "ntl: j'Oj tI:? "their heart did not shrink from
appealing to Him" (ibid. st. 48). It was also preserved in Spanish
ornate-prose style: ,ElO?Q C'll.:l' 'n::liD' "ElO?Q C',::lJ '" "my hands
are too clumsy to count, and my praises too weak to tell" in a let-
ter of Judah ha-Levi (Diwan, ed. Brody, i 209).
n. It is also frequently used in SH, e.g. ,n'J'il?Q 'ilrlQ ilntl: "you are
warned against reproaching him" (B. I:Iiyya, Hegyon, f. 13a); ' l '
C'jil::l nJ?QQ ,? nI'il?Q COtl:QiD "so that He despised them too much
for letting them be His theocracy" (Abr. Maim., (lobe;:, f. iii 17b);
lC!liDElJ Inn'QtI: nm?Q '? il?'?n "God forbid that their real meaning
should be the literal one" (ibid., f. 19b). But some difficulty seems
to have been experienced with this idiomatic construction, and there
was a tendency to replace the infinitive in it by the BH gerund:
mil '::l'il ?tI: ll'jilQ iD?m, iDiD,n cn" "their spirit is too limp and weak
to attain this purpose" (B. I:Iiyya, Megillah, p. 13); POllnilQ 'i1IlljQ tI:?
nlQJn::l "he was not prevented from occupying himself with the sci-
ences" (S. Tibbon, Eccl.-Comm., f. 16a).

32. The Subject if the lrifinitive Clause

a. In the majority of cases in BH and throughout in MH, the sub-

ject or agent of the action expressed by the infinitive is not expressly
indicated, but it is left to the hearer to discover which part of the
 THE VERBAL NOUNS 145

main clause is subject of the infinitive. Thus in m?tv? i'tv iTtv.t1 ~?

Cli~ Clitv? "he did not compose a poem to send it to anyone"
(J. Tibbon, Testament, p. 9), the subject of the principal clause is also
subject of the infinitive. In rm~i? 'JDD .t1JiD n"iT "you did not let me
see" (ibid.), the subject of the infinitive is equivalent to the pronoun
included in the adverbial phrase.
h. In BH already, there existed also the possibility to indicate the
subject of the infinitive, if it was different from any item of the prin-
cipal clause, within the infinitive clause itself. In the case of the
infinitive construct without le-, this was mosdy done by attaching the
subject as a nomen rectum. With the le- infinitive, which was not a
noun, this was not possible, and the subject was therefore placed
after it like the subject of a finite verb, a construction which we may
call infinitivus cum nominativo. E.g., iJ? ?~ l?DiT Clitv? "for the king
to take it to heart" (2 Sam. xix 20); n~iiT iTDtv OJ? "for the mur-
derer to flee thereto" (Num. xxxv 6). By analogy, the construction
may have been transferred also to infinitives without le-, which are
in the Massoretic text pointed according to the absolute form. I The
same construction is quite current in Assyrian (Brock., 87f) and prob-
ably proto-Semitic (Noldeke, Zur Grammatik, p. 74, note 3).
c. In MH this construction did not exist; the need for it was avoided
by employing dependent clauses with final verbs. It is also not found
in the liturgical style of MH.
Its revival seems to date from Saadianic prose style,2 e.g., "there-
fore the tree of life was withheld from him Cl'?'O~ ?~? iOiD im'iT?

I Cr. Ges., ll5g; Brock., 87d; SeJlin, Verbal-nominate Doppelnatur, p. 81. The view
here proposed differs from that of the above authors in connecting each syntactic
construction with a definite morphological form. The cases in which the vocaliza-
tion or position of the infinitive construct without le- points to the subject being in
the nominative may also be explained otherwise. Konig, 230f, doubts the correct-
ness of the Massoretic vocalization in these words; Ges. loco cit., suggests that they
may be constructs, the absolute form having been preserved for phonetic or seman-
tic reasons.
2 Its resumption may partly have been due to Aramaic influence, since the con-
struction occurs in Syriac (Noldeke, Syr. Grammar, p. 226), Babylonian Talmudic
Aramaic (cr. Schlesinger, p. 202), and, which is perhaps here of special importance,
in late Targumic Aramaic, e.g. 'O.ll~ c;)''?iLl 1i1'~ '1i1D? i1D'?iLl ~:I'?D'? i1~1~J~ 'D~n~
rC;)~iLl "it was predicted to King Solomon that he would be ruler over the ten tribes"
(T. Cant. viii 12). The use of the infinitive here is remarkable, since in general
Targumic Aramaic replaces BB infinitive clauses by dependent di- clauses (cf.
Stevenson, Grammar, p. 53).
146 CHAPTER NINE

"for it to be a warning to all fools" (lfiwi Polemic, st. 1); 1'~nn ':J :"j~1
'?1~n:11 i1~0:J nEl 1'0.ll'? "and you still want him to stay here in trou-
ble and distress?" (ibid., st. 12).
Its use was continued in N.W.-European Hebrew, beginning with
Yosippon: "therefore Sanballat asked Alexander for permission to build
a temple on the mount of Gerizim" 1Jnn I:lID 1n:J m'n'? "for his son-
in-law to be priest there". It is quite common in Rashi, e.g., r~
1'El:J i1:l1D 1:l'01D I:lID m'n'? 11D.lI 1i' "it is not Esau's manner for the
name of God to be current on his lips" (Gen. xxvii 21); n1:J 1:J nn'?
.lI01D:J1 i':J:lO 1'?1p nm'? "to give him strength for his voice to be
strongly audible" (Ex. xix 19).
d. In SH, the infinitivus cum nominativo was employed with much
greater frequency than in either SH or N.W.-European Hebrew as
an alternative to the dependent clause whenever the subject of the
secondary action was not referred to in the principal clause. E.g.,
n'?1D001 n:J'?oo 1'? m'n'?.. I:l'~n n:Jr "man was granted that king-
ship and rule should be his portion" (B. I:Iiyya, Hegyon, f. 1); ni1~n
nm~ i':JO :"j1:li1 '1D1no 'n~ m'n'? .. n'?1:J' m'~ "it is not possible for
abstract form that any of the five senses should perceive it ... " (ibid.,
f. 2); ni1n I:ln'? 1mn'? 1:l"1~i 1:l'?1.l1n ':J:J m ~'? "the denizens of the
world were not worthy for the Law to be given to them" (id., Megillah,
p. 24); ':llDn '?,l) n1,l),n ':lIDO 'n~ n~:l'? .. n1:J 1:li1:l r~ "there is no
power in them for one of the two opinions to prevail over the other"
(ibid., p. 113); 1:J'?:J niiEl:J n":lpn m~'~o n1'n'? 1:J:li nlDo IDP':J ':J
1:l'~~o:Jn i~IDO "for Moses wished for the existence of God to be a
distinct factor in his mind from that of other beings" (Maim., Yesode
ha- Torah, i 10). ri~n i1:J~0 n~oo 1:l:J'?no nm'? n1ii11 lD~n 1i' "the
nature of fire and air is for their movement to be upwards, away
from the centre of the earth (ibid., iv 2); mro n1'n'? 10~.lI '?':Ji' ~'?
I:lno "let him not accustom himself for his food to be of them" (id.,
De'oth, iv 9); i:J1D n1p1:J'nn ,o'?o np''? m',on :lmo n'n "if it is the cus-
tom of the land for the elementary teacher to receive wages ... "
(id., Talmud Torah, i 7); m1:J01~ 1nim nm'? n~ilD '0 "whoever wished
for his learning to be his livelihood" (B. Daud, fCabbalah, p. 72); 1nm
I:l'm~no I:l.lln m'n'?.. 1:l':J'0'? mpo "when he made concessions to
heretics ... so that the people should be afflicted" (al-Fakhir, fCobez,
iii 2a); 1~IDEl:J 1nn'0~ m'n'?o ,'? n'?''?n "God forbid that their true
meaning should be the same as their literal sense" (Abr. Maim.,
fCobez, iii 19b); mlDn C1i~n I:lnilD'n:J :l'lDn'? ~'n ... ni1nn m'm n''?:Jn
 THE VERBAL NOUNS 147

~ii:li1 "the purpose of the giving of the laws is for man to grasp by
their guidance the nature of God (B. 'Abbas f. 3b); C1'i'i1rrJ C1'rJ~ni1
i1'PJ mJrJ~ C1i~ iirJ'?'? "the sages advise that one should learn a clean
craft" (Falaq., Meba~~esh, ( 12b); C1'rJi1 nnn i1'p'?n nm'? ri~i1 .ll:lCl "the
nature of the earth is for parts of it to be under water" (B. {:aq:ah,
Mikhlol, f. 23a).
e. The popularity of this construction went so far that for its sake
elements were taken into the infinitive clause which did not really
belong to it, e.g. piiiD:l C1:'rJi1 m'i1'? :liD piiiiD:l i"i'i1 ni:li1 ii:l.ll:l
"because of the Yod being pronounced with a shuru~ it comes to
the Mem also being pronounced with a Shuru~ (J. ~mi:li, S. ha-
Galuy, p. 12). Here the original subject of :liD is C1"rJi1, but in order
to bring a subject into the infinitive clause, the verb is construed
impersonally. A similar case is the Falaquera quotation in the pre-
ceding section (last instance but one).
f. In early SH, the infinitivus cum nominativ0 3 construction can also
be applied when the subject of the infinitive clause is already men-
tioned in the principal clause, and therefore need only be referred
to by a pronoun. This is not found in BH, Saadianic prose,4 or
N.W.-European Hebrew. Examples: C1m'i1'? n.llii1 '?iP'iD '?.ll C1i1 C1"i~i
C1'ii:m C1'ir:JJ C1iiprJ "they are seen, on closer investigation, to have
originally been scattered and isolated" (B. J-:Iiyya, Hegyon, ( 2); C1iiJ
i1rJiPrJ n~ nEl'?nrJ i1nm'? t']iJi1 '?~ np:liJi1 i1ii~i1 '?.ll "it causes the form
inherent in the body to change its place" (ibid., ( 4); miDi-lnJ ~'?
on~'i::J::J '?'.lliO ini'ii'? O'?i.ll::J i::Ji'? "He did not permit anything in
the world to assist in their creation" (id. Megillah, p. 51); .ll'Ji1 ~'?
lnJ'l' n:ln~ im'i1'? in'?.llrJ "his rank was not high enough for him to be
like his friend J onathan" (B. Ezra, Saphah Berurah, ( 1Ob); niiDrJ t']"'?~i1

3 Exception might be taken to the application of the term nominative to the

suffix-pronouns employed in these constructions. As it is, however, quite out of the
question to construe the subjects in the earlier instances as being syntactically in
the genitive, and the identity of the two constructions cannot be doubted, we must
admit the possibility of the suffix-pronouns expressing a virtual nominative in SH. Fur-
thermore, the le- infinitive in SH is non-nominal, and can therefore not be con-
sidered regens of a construct unless one wishes to deprive these terms of any
syntactical reality. In any event, it must be kept in mind that terms like nomina-
tive, genitive, etc., in a language devoid of nominal fiexion are merely convenient
shorthand expressions of strictly limited value.
4 The quotation from the Ijiwi Polemic stanza 14 (above under c) is not a case
in point. The pronoun there refers to a word in a preceding sentence, not to any
term of the principal clause.
148 CHAPTER NINE

i~r iJ1rJ 1'n' lrJ'o 1rWi1'? Cl'rJ,l]El'? "the Aleph serves sometimes to be
the prefix of the first sing. masc." (ibid., f. 17b) m1'i1'? 1"m Ji1JrJ
pi'nJ "the normal thing for the Taw is for it to be vocalized with
a l:Iire~"; D'niiDrJ orWi1'? ~~1rJJ D''?Pi1 i1nJ'? 11~Ji1 "it is proper to
select the easily articulated consonants for them to be grammatical
elements" (id., Yesod Mora, 17). 5
g. In writers of the second period, this construction is rare. I have
no instances of it from Maimonides, Ibn 'Abbas, or Falaquera, who
are so fond of the infinitivus cum nominativo in general. Strangely
enough, it turns up again, though sporadically, at the very end of
the second period, e.g., mnEl 1rWi1'? 01~'rJi1 1~ '?,l] n"'iDi1 1i1Jrl]'iD ~'?
mJJ1 "not that God would leave him, out of contempt, to become
lowly and despicable" (Gers., De'oth f. 9b); 1m~ii1'? ~o~ iD1prJi1 ~iPJ
1J i1'?,l]iD 'rJ D~1,l] '?,l] "the Temple is called a throne to indicate the
might of Him who dwells on it" (B. ~ar?:ah, Mikhlol, f. 8b).
h. In Arabic, the nominative subject with the infinitive exists in
grammatical theory, but hardly ever in actual usage (cf. Reck., 101;
Brock., 87a). It seems never to occur in Arabic prose at all. On the
other hand, a very similar usage occurs in all earlier Western Romance
dialects, and is still quite current in modem Spanish and Portuguese,
e.g., "mas liviano trabajo es pasar un mamello por el ojo de una
aguja que entrat unrico en el reino de Dios" (cf. Lerch, Hist. ftanz.
!iYntax, ii 152ff'), The further development mentioned in section (f)
does not seem to occur in Romance. 6
i. The history of this construction offers various problems. It cer-
tainly came in from N.W.-European Hebrew, and maintains itself at

5 I cannot account for the pronoun attached to the infinitive in the sentence:
~1tDj ~tD1jil 1nl'il? i:Jiil l:liln' l1tD'?:J "in actual usage it is the other way round, the
subject becoming predicate" (E. Ezra, S. ha-Shem, f. 8b). The subject of the infinitive
clause is ~tD1jil. Unless the pronoun is an attempt to reproduce the Arabic context-
pronoun (tjamfT ashsha'n, which, however, would hardly be applicable in that sen-
tence) it is strongly suspect of being a textual error.
6 Professor W J. Entwistle was kind enough to give me the following informa-
tion: "The infinitive with personal pronoun occurs in Spanish and Catalan only
when there is need to call attention to another subject. The strictly logical subject
may remain the same, as 'Las dificultades, al verlas yo par primera vez, me pare-
cian insuperables'. In Portuguese the same effect is obtained by the personal con-
jugation of the infinitive. Ptg. vermos = Sp. ver nosotros. In Brazilian Portuguese
the correct use of the personal infinitive is not always maintained, and there I
believe one might find a personal infinitive without change of subject."
 THE VERBAL NOUNS 149

first against the influence of Arabic. It is possible, though not nec-

essary, that the existence of the parallel construction in Romance
increased its popularity with the older authors. If, however, the
Romance factor had been dominant, we should have expected its
use to grow during the second period. Instead, the frequency of its
occurrence diminished, a fact which we must ascribe to the influence
of Arabic through the translations. 7 We find thus the paradoxical
situation that Romance influence was greater during the "Arabic-
speaking" period, but Arabic influence during the "Romance-speaking"
period. The truth is, of course, that neither "period" had any real-
ity as such, at least as far as the Hebrew language was concerned.
The real situation was a good deal more complicated.
k. The origin of the extension of our construction to the case under
(f) is unexplained. Even if it was invented by B. J:Iiyya, no reason
seems to exist for it, especially in view of the general linguistic con-
servatism of that author (c( above p. 51). Also, it would in this case
have to be assumed that B. Ezra took over this feature from him,
while in general no direct linguistic influence of the former on the
latter can be discovered. There are two possible explanations, both
rather uncertain:
(1) That the construction had been thus developed in the Hebrew
of the Provence and Catalonia at the time of B. J:Iiyya, without,
however, having penetrated into the Northern-French Hebrew used
by Rashi.
(2) That the extension had taken place in the Catalan colloquial
of the eleventh century and was introduced into Hebrew by both
authors indepedently.
1. A particularly interesting extension of the infinitivus cum nomi-
nativo is its transference to the impersonal construction. As we find
in BH an impersonal passive with accusative object (Ges., 121ab) so
we find an impersonal passive infinitive with an accusative object in
ii"m n~ ,J 1n:Jii? "~, ii'ii iirii O"ii "this day was fit for the Torah
to be given on it" (B. J:Iiyya Megillah, p. 24), literally "ut detur in
eo legem". Similarly p n1iD.I.lii? ii1~ l?r:lii "the king commanded for
this to be done" (B. Daud, flabbalah, p. 51).

7 It occurs, however, also in translations, e.g., in the "O'?r1il ~'::lO wrongly ascribed
to Samuel ha-Nagid (c£ above p. 43): r10',PO il::l'?il ilr1m'? ilO'P'?, il::l'?ilil prn'? "to
strengthen and establish the halakhah so that it becomes an established halakhah".
It is not clear what Arabic construction ilr11'il'? translates: hardly an infinitival one.
150 CHAPTER NINE

33. Gerund and Nomen Verbi

a. SH uses both the BH gerund (i.e. infinitive construct without

le-) and the MH nomen verbi, without any semantic or functional
differentiation. As the functions of the old gerund were wider than
those of the nomen verbi, it was inevitable that the latter should in
SH be applied in constructions in which it could never have stood
in MH, simply because these did not exist in normal MH. This can-
not be considered an extension of the functions of the nomen verbi
as such, since the whole syntactical category did not exist in SH
except as an alternative morphological form to the gerund. The
gerund, on the other hand, is not identical with the BH infinitive
construct, many of the functions of the latter having been taken over
by the le- infinitive.
h. The nomen verbi has also functions to itself: (1) the purely nom-
inal function in which it is really nothing but an abstract noun
derived from a verb, and often turns into a more or less concrete
term, such as 1i~'~ "comment", n1tD'n "guidance", nl1~nnn "con-
junction of heavenly bodies" etc.; (2) its use as cognate object described
above, par. 22a, the nominal character of which is proved by the
fact that in it the nomen verbi is able to take an attribute. In either
case it is not a morphological but an etymological category.
c. The gerund or nomen verbi is very rarely employed as subject
or object. In these positions the old construct infinitive is displaced
normally either by the le- infinitive or by the MH dependent she-
(ki-) clause, and the use of gerunds appears to be a conscious archaism,
perhaps due to the influence of ornate style. It occurs thus as sub-
ject: n"on 10 Ion r"'J'~ 10'0 1'wn" pn' ~" it is impossible for the
imperfect-prefixes to be missing in this word" (B. Ezra, Za~oth, f. 3a);
As object: n'~~ 'o~n ~'1~.li 'JtD~ n'JtDn ~.li i1JitD~1n nI'n" no'., "he
imagined the former to be with respect to the latter 'like two young
roes that are twins' (Cant. iv 5)" (al-Fakhir, flobe(:., iii 2a). A second
object: ... ni1~nnn nm ~'n i'''.li n.,'.lion nI~n i~~O "they found the
sign bearing witness to it to be the co~unction (of Saturn and Jupiter
in one section of the Zodiac)" (B. I:Iiyya, Megillah, p. 116).
d. While in these few instances the gerund or nomen verbi appeared
as an alternative for the le- infinitive, there are cases in which it
could not be replaced by the latter: ... ~10~ nr ".li nprnn ~n'~li
 THE VERBAL NOUNS 151

"their strongest proof for this is that they say ... " (S. Tibbon, Eccl.
Comm. f. lSb); lJ:J n~ n?1i'J ... i1:li "remember that she brought
up your son" (]. Tibbon, Testament, p. 11); n~:ln 'n~ lr11'n iD1ii "seek
that you become a brother of wisdom" (ibid., 13). So already in
Saadiah: 1n:l'?n 1n':liDn ~? "he did not let him forget that he had
gone" (ljiwi Polemic, stanza 3).
This is a normal construction in Arabic, and indeed the third quo-
tation seems to be a translation from that language, while Di~~ in
the first is but an imitation of the Arabic qauluhum. It is again a
proof of the resistance of the SH sYM-tactical structure (as distinct
from idiomatic detail) to Arabic influence that SH managed to develop
its own distinctive categories.
e. The most frequent function of the gerund in SH is after prepo-
sitions. The BH infinitive construct in this position l had been replaced
in MH by dependent clauses, introduced by the same or equivalent
prepositions plus she- (cf. Segal, 344). It is still almost non-existent
in the early liturgy (cf. above p. 19, note 6). In Piyyut it became
again a regular linguistic feature. E.g., n5:l' p5:liD n~?~:J "when the
prosperity (?) of the beautiful one became full"; n:J1iDn l:l ?:l lr11'n p"
"because you are so noble" (both from the I:Iinah n~?~:J i~ by I>.alir,
Brody and Winer, p. 44, cf. also the instances from Yosi b. Yosi,
collected by Szneider, Leshonenu, iii 27). From there it was taken into
the possible liturgical pieces of the Gaonic period, e.g. Dn~~:J D'n~iD
D~1:J:J D'iDiD1 "they rejoice while they go forth and are glad when
they return" in 11i~ ?~ (Singer, p. 129); 1i~,l):J nm 1iD~i:J m~5:ln ?'?:l
'J'O in ?,l) TJ5:l? "a diadem of glory didst Thou place upon his head,
when he stood before Thee on Mount Sinai" in niD~ n~iD' (Singer,
p. 138). It is also common in Saadiah's prose in the paytanic prose
of Al).imaa~, in the biblicizing historical style, and in metrical poetry
and the ornate prose derived from the latter.

1 The use of the gerund instead of a dependent clause is characteristic of late

Biblical Hebrew. The Chronicles replaces almost throughout the dependent clauses
of its sources by gerunds (cf. Kropat, fiyntax, p. 73, examples ibid., pp. 68-70). This
is perhaps an instance of over-correctness: because the colloquial of the author of
Chronicles did not possess the construction, it was considered a hall-mark of liter-
ary elegance, and applied wherever possible. For this reason, too, the gerund and
the consecutive tenses were chosen to give the historical piece (b. ~dd. 66a) liter-
ary respectability, cf. above p. 11. In Babylonian Talmudic Aramaic, too, infinitive
clauses are used in preference to dependent clauses wherever there is no danger of
ambiguity (cf. Schlesinger, p. 196).
152 CHAPTER NINE

f. On the other hand it does not seem to have penetrated to any

appreciable extent into late Midrashic Hebrew and its N.W.-European
offshoot. Its use in SH from the very beginnings of that idiom must
therefore be due to the influence of the ornate prose style, perhaps
reinforced by the frequency of the same construction in Arabic.
g. In using the gerund proper with prepositions SH continues BH
usage, though the prepositions employed are often post-Biblical: I:lJ'~
i11D.!i~ 'i''? in~~ in~ ~'?~ iJi l:l'i':J~ "they do not perceive a thing
except after it has come into being" (B. !:Iiyya, Megillah, p. 114);
i1~'?1D in.!iiJn nI'i1 iiJ.!iJ I:l'?in ~iPJ "it is called !:Iolam because its
sound is full" (B. Ezra, Za~oth, f. la); i1ii'?n inl'?.!ii1~ '?:JIDi1 nli~ilD
"they guard the mind from getting rusty" (B. 'Abbas, f. 6a); C'JiJi1
ip.!ii1 I:li~'? I:lit!l ... "who build ... before they have learnt the essen-
tials" (ibid., f. 8a); I:l'Jit!li1 l:l'ID.!i~i1 I:lnllD.!iJ l:l'n~1D 1:li1 "they are pleased
because they do good deeds" (Falaq., MebaMesh, f. l8b); 1:l1Di1 nJID
nI'?i.!iEli1 i'?~ i~:J nllD.!i~ "God ceased from doing the like of these
actions" (Gers., DeCoth, f. 9b); i1'?JJi1 nJi:J n''?:Jn'? i.!i'Ji1:J "when he arrives
at knowing the final intention of that which is revealed" (M. Tibbon,
Cant. Comm., p. 5); ri~i1 ni,?~i} in~'? i~iJJ C':J~'?~i1 "the angels were
created after the draining of the earth" (B. Z:aq;ah, Mikhlol, f. 5b);
1:l':JElIDJi1 1:l'~i1 t!lIDElm:J i1''?.!i it!lIDElm ~'? "it did not spread on it like
water poured out" (ibid., f. 23a); nr.:l~ i1~iPi1i1 n~r nI'i1 in~ "since
this promise is true" (Zeral:tiah, He-lfalu?:" vii 96).
h. This is specially frequent with the causative le-. Though causative
use of that preposition is common in MH (cf. Giesebrecht, Priipos.
Lamed, p. 85), it does not occur there with infinitive. Its frequent
use in SH, both with nouns and with gerunds, is probably due to
the influence of the Arabic liim at-taCITI (Wright, ii l50D):2 m iiJ.!iJ
in:Ji.!i~ i~:J iniJn~ nI'i1'? '.!i~~~ I:l'JIDi1 iElO~ i1'i1 "therefore the number
two is the middle, because its double is equal to its square" (B. Ezra,
S. ha-Shem, f. 9b); iOi~i1 pp I:l'J'JPi1 inJ~ ':J 'n.!ii'? "because I know
that the choicest of possessions is the possession of worldly wisdom"
(Falaq., Musar, p. 47); cmr.!i i.!iIDJ r~ ':J iJlDn'? I:lElii" 1:l''?ElIDi1 m"1D '~
"he who cheats and oppresses the poor, because he thinks that they
have no assistance from the law" (Gers., De'oth, f. 9a).

2 The causative function seems to have been considered, by some Spanish gram-
marians at least, the essential function (a~l) of ie-. Thus B. Ezra says (Yesod Mora,
p. 17): "The name of the letter Lamed is derived from the root 1i.l" "to learn"
because its essential function is to indicate why a thing exists (i~1i1 i1'i1 i1i.l")".
 THE VERBAL NOUNS 153

i. We find the nomen verbi used in the same manner already in

the early Baraitha b. :&idd. 66a: 1mrn:::l "when he came back". SH,
in making the nomen verbi the equivalent of the gerund in this func-
tion, hardly followed an older tradition, however. The same situation,
i.e., the existence of three verbal nouns where only two categories
could be utilized, produced the same results. Examples: ~'? mm :::l1i'p:::l
C'01nO C'i:::l1i1 1'i1' "when the time is brought near, these things will
not be ambiguous any more" (B. I:Iiyya, Megillah, p. 2); C'n:::lniDO
m10 '?1I cm'Oll:::l "they boast of having discovered its secret" (B. I:Iiyya,
Schwarz-Festschr., p. 29); rE)",?~ ':liD nn:::lnm i1:::l1l:::l miDOi1 ion "the
short vowel is elided because the two Alephs meet" (B. Ezra, Saphah
Berurah, f. 3a); i1i':::lll n'iDlIO 10~1I 1I:l10i1 "if one refrains from com-
mitting a sin" (id., Yesod Mora, p. 11); 1n~':::l ':JE)'? "before he comes"
(Maim., (lobe;::., ii 26); m:J c:Jn'? 1:::l m'~o'? 'mom 'nnOiD "I was very
pleased and gratified3 that such a learned man should have a son"
(ibid., f. 27); C'?1~':::l'? '?~:::l' ~'? "He does not cease to exist because
they cease to exist" (id., Yesode ha-Torah, i 2); 1m~ 1I1n Ci1'? l110''?:::l
"by studying them you will know Him" (J. Tibbon, Testament, p. 12);
m~'~Oi1 n:JiDOi1'? 'ni:Ji1i1 li1~i1 "the essential necessity to maintain
the existence of the world" (Abr. Maim., (lobe;::., iii l8b); C'i:::l1 i1n'~o
n1'?:Jn0i11 nm1:::lm 'n'?:::lO i1'?1I0 'E)'?:J "(they) utter blasphemy without
considering and thinking" (B. 'Abbas, f. lla); 1m'pn m~ 1'? ipn
m''?:JiDi1 m'?1I0:::l C:Jni1 nn'?~i1 "before investigating the happiness of
the wise, he investigated the intellectual virtues" (Falaq., Meba~~esh,
f. lla); "the Torah tells us the relationship of God to the creation
of the world" 1:::l ~1i1iD 1E)1~i1 m ?1I 1'on 1n1~ 1n"'POiD ~'i11 "namely,
that he preserves it forever in the state in which it is" (Gers., De'oth,
f. 9b); C'?111i1 1iD11'n nll:::l ?lIE)iD m?111E)i1 "the deeds He accomplished
when He created the world" (ibid.); i'1~:::l 1m:JE)i1m:::l r1~':Ji1 n1l1m
"the movement of the spark while it revolves in the air" (Zeral).iah,
He-lfalu;::., vii 97).
j. The equivalence of gerund and nomen verbi is proved by the
instances in which they are used in parallelism to each other, e.g.
Cnll'piD1 C':::l:J1:Ji1 m'?lI:::l cn:J~'?O:::l 10':JiD nm'o:J "experiences they have
made in their craft with regard to the rising and setting of the plan-
ets" (B. I:Iiyya, Megillah, p. 114); '?:Ji1 m'fllO :::l1~ ~1I10i1 nnp "taking
a little is better than giving up all" (Falaq., Meba~~esh, f. 19b).

3 iTr:lr1 takes this meaning "to be pleased with" from its Arabic synonym 'ajaba.
154 CHAPTER NINE

k. Owing to the use of the nomen verbi as abstract noun some con-
fusion results with non-verbal abstract nouns, and these can be
employed instead of the gerund (as already in BH, cf. Ges. 45c-e).
E.g., ?1D~~i1 iOi1 1n~1~::J i1::J11Dm lii i1~i~ 'JIDi1 ::J1n~i1 "the next verse
shows the meaning of repentance by commanding: remove the ob-
stacle ... " (B. I:Iiyya, Hegyon, f. 8). Thus even abstracts, the full nom-
inal character of which is established by their syntactical function
within the sentence, govern as if they were infinitives: ?1iJ mJi~~ m
i1?lln' n1::J~::J "this is a complete denial of His glory" (B. I:Iiyya,
Megillah, p. 10); m j'Jll::J 'm~ Tm?pi1~ i~O~ i1~ "what shall I tell of
the times you treated me with contempt 4 in this matter" (J. Tibbon,
Testament, p. 9); 1m~ 1Jn::Ji1~ ?i1J? "because we love Him so gready"
(J. Nai).mias, Aboth Comm. f. 3a).
1. It seems that the opposite case could also occur, the gerund being
employed in positions where one would expect an abstract noun, or
a nomen verbi functioning as such: 1'm1~~ i1~tq::J1 ?~i1 n~i'::J p::JinlD
"that you may cling to the fear of God and to the keeping of His
commandments" (J. Tibbon, Testament, p. 9). It is, however, possible
to take the word i1~1D as nomen verbi of the Pieel, and thus to
remove the difficulty.
Dl. As occasionally in BH5 and always in Arabic (which possesses
no passive infinitive), the verbal nouns of the passive conjugations
are sometimes replaced by the active ones: ?~ilD'? i1i1n i1Jn'J mi1 ~1'::J
~i1'?ll i1i'~i1 ~i1? i1nJ'nJ::J1... "on that day the Law was given to
Israel ... and in being given to them it shone upon them" (B. I:Iiyya,
Megillah, p. 27); i11D~ i'::J i1i1m nn 'J:l? "before the Law was given
to Moses" (B. Ezra, Yesod Mora, p. 8). Perhaps these should be treated
as impersonal constructions parallel to the impersonal use of the pas-
sive le- infinitive (above par. 33 1).
This construction was by no means the only one in use. The pas-
sive verbal nouns occur quite frequendy, e.g. ~'i1?~ m?i1 m ?ll ::Jn~i1::J
"when the name of God is inscribed upon this tablet" (B. Ezra, Yesod
Mora, p. 16).

4 It seems that this is an imitation of the Arabic verbal noun with feminine end-
ing to indicate a single event (ism al-marrah, Wright, i 122D).
5 cr. a list of instances in Koch, Der semitische lrifinitiv, p. 67. Koch argues that
this is the original Semitic construction, the passive infinitives being a late innovation.
 CHAPTER TEN

SUBSTANTIVE CLAUSES

34. Substantivization if Clauses

a. Asyndetic substantive clauses are very rare already in BR (Brock.,

332; 339.a), and apart from oratio recta do not exist in MR (Segal,
DiMu~, 422).
In SR, too, only oratio recta is construed asyndetically. This is,
however, an exceptional syntactical process, as really the structure
of the sentence is broken, and the speech part is no more depend-
ent on the principal clause than a parenthesis.
h. Real asyndetic substantive clauses occur, however, when a depend-
ent clause is coordinated with an infinitive clause: n~r ?::l 'Ill 'J~ ?i::l'
Cli1::l ipini Cli11.:lll ~~itD1.:l i1'i1~i Cl':lii1 Clll niJ1.:li1? "nevertheless I can
count myself among the many, and wander about with them and
investigate as they do" (B. I:Iiyya, Megillah, p. 2); Cli~ ?::l? 'i~ii1 '::l
'~iiJ r1.:l~'i lli'i ?~i11.:l i1~i' i? ni'i1? i1?'nnJ "the befitting thing for
every man is first of all for him to have fear of God and that he
should know and believe with certainty" (B. 'Abbas, f. 9a). These
clauses are constructed as if the first clauses, to which they are coor-
dinated, were a dependent finite clause. In such a case there would
be no need to repeat the conjunction. We have here merely cases
of anacoluth. Its occurrence in two writers as different as B. I:Iiyya
and B. 'Abbas shows, however, that there was a definite tendency
towards it in SR. 1

1 A similar situation produced the same effects in bab. Talmudic Aramaic, where
asyndetic dependent clauses, though otherwise non-existent, appear occasionally in
coordination with infinitive clauses (cf. Schlesinger, p. 20 I); e.g., '~ 1i1"i~J "1Jn~:::l
1n~i iD 1i1"i~J" lit;lJ1 p~i "by letting their husbands study in the house of study
and that they wait up for their husbands until they come home" (bab. Berakhoth
17a). Such cases occur already in BH (Is. xiii 9, Amos viii 6) and Biblical Aramaic
(Ezra iv 21). Cr. also the ;::,adokite fragment p. 2 line 21: ~"1 I:m~n n~ ~n1tDD~ ~i1'tDD
n1~r.l n~ 1ir.ltD "because they did their own will and kept not the commandments
of their creator". The Zadokite fragment is altogether fond of this construction, which
also occurs 3.11, 6.16, 7.1, 7.3, etc.
156 CHAPTER TEN

c. Asyndetic, as in the older dialects, are also the dependent questions:

"pn'? "to investigate whether it has an end" (B. I:Iiyya,
i1'?~n ,,? tzl'i1
Megillah, p. 10); ~'i1 i1"~ 1r'~ O'J~'~i1 n'~'?i1Ci m'~ '?l) "Cil)'? "to dis-
cover the shape of the courses of the planets-what shape it is (ibid.,
p. 113); O'~~'?Cii1 '~'JJ 'm'~ '?"r, nl)1 'mn~ 'J~ "I have already
indicated the opinion of the Rabbis as to when the angels were cre-
ated" (B. ~af?;ah, Mildzlol, ( 7b).
d. Syndetic substantive clauses are introduced either by the MH she-
or by the BH ki, rarely by the BH asher. These three are entirely
equivalent, and even in compound adverbial conjunctions can be
substituted for each other. No examples need be given here, as there
will be ample illustration in the following paragraphs.
e. The revival of ki as a conjunction for introducing dependent
clauses dates from the early liturgy. This uses side by side she-, e.g.
'i1 ~'i1 i1n~iV 1'? 'Jm~ 0'1'Ci "we acknowledge unto Thee that Thou
are the Lord" (Singer, p. 51) and ki, e.g. O'iVl)Ci ,JJ r~ '~ 'Jl)1' "we
know that we have no good works of our own" in o,n, ~'i11 (Singer,
p. 59). Kz in the sense of "that" is, however, rather rare, especially
if compared with the innumerable instances of the same word with
the meaning "because".
In SH, too, Kz is probably rarer than she-. Some authors, like Ibn
Ezra, prefer the first, others, such as Maimonides in the Mishneh
Torah, use exclusively the latter.
f. Infinitive clauses are fully equivalent to dependent clauses. It is
quite impossible to give any reasons for the choice of the one or
the other construction in a given case. To what extent they were
felt to be the same is proved by the phenomenon treated under (b)
above. In the following the two are in some cases for convenience
considered together.
g. A favourite device of SH is the extraposition of some member of
a subordinate clause into the principal clause, where it stands in the
syntactical position which would have been occupied by the clause.
The reason for using this construction seems to have been a tendency,
visible throughout SH, to avoid clauses as subjects or objects. The
same tendency produced the extraposition of a cognate object dis-
cussed above, par. 22c, which is really nothing but a special case of
the feature illustrated here. An added advantage in the case of depend-
ent questions was that the extraposed word could be governed by a
 SUBSTANTIVE CLAUSES 157

preposition, which would have been impossible with the clause. E.g.
nQ~n O'JniJ O~ ,'nEm '?,I] -npn'? 'J'~i mn "we now think best to inves-
tigate whether its symptoms give true indications" (Crescas, Or Adonai,
f. 4a).

h. Other examples: ,::J ~~911iD nn~ n::J'c:!::J o'~n 'iiD~' ~'? "one does
not praise a man happy because one good quality is found in him"2
(D. J>iml:li on Psalm i 1); n'?,n n'n'iD ~mi ,QJ niD'::J' nm r~ 'J "noth-
ing is as disgraceful and embarrassing as that a doctor should fall ill"
(B. Tibbon, Testament, p. 10); ~~QJ ~'n !::J~ iDEl:Ji1 nii~iDn pt1Q i'pm "we
shall inquire as to whether the survival of the soul is something real"
(Gers., Mil~amoth, f. 4a); n"'iD'? niQ,n'? 'piDn' ... 'i'~'iDJ on 'J ",1]'
"furthermore, when they realise. . . they long to become similar to
God" (id., DeCoth, f. 9b). In the last instance, the extraposition of on
(a word that is quite unnecessary within the subordinate clause) helps
to avoid heaping of conjunctions.
i. This has, of course, nothing to do with the cases in which a mem-
ber of the subordinate clause is extraposed, but remains dependent
on the conjunction introducing it. E.g., ".:JniD 0',1],,' ,'n ,,?~ on 'J
,'OJJ "::J~' "if they had known that the gentile~his fortune would
be lost" (B. I:Iiyya, Schwarz-Festschr., p. 29);3 n,l]'r::J 'n~n 'J iQ~"
'n~'~i n'nn "he said that the one~his cure would be by perspira-
tion" (Falaq., Meba#esh, f. l3a). These offer no special syntactical
problems beyond those of extraposition in general.
k. The same tendency to have a word rather than a clause as mem-
ber of the principal clause has also led to phenomena correspond-
ing to the "lesser subject" of European languages (e.g. "it is nice that
you have come"). A pronoun representing the whole clause is employed,
and the clause placed in apposition to it. zeh can be used as such
a "pseudo-substantive", e.g. m~'Q ~'? o~ O'JiD nQJ ni::J'iDm 'in~miD nrl
''?,nn i::J,Q "that the replies are a few years late is, if not because
of my worries, then through my illness" (Maim., Teshuboth, ed.
Freimann, p. Ix); ,::J'?::J r,l]n n'~iQ'? n'?Jm r.llJ 1ni~ O'~'i 'J~iD nrl
~'n "that we see the sky as something blue is merely an optical illu-
sion" (id. Yesode ha- Torah, iii 3).

2 This cannot be translated "for one good quality that is found in him", for then
the verb should be n~~r.ll
3 This sentence contains both kinds of extraposition treated here.
158 CHAPTER TEN

1. ~'i1 is also employed for this purpose: n~p rJ'tD ?':JtDO? ~'i1 J'~'
C1i1'iJi "it is good for the educated person that he understands
some of their statements" (B. Ezra, Yesod Mora, p. 1); ':J ~'i1 ?p iJi ~?
C1'o:Jn 'iJi i~OJ 'D~O:l "it is not a small matter that the words of
the Rabbis have lost all their authority" (Letter of the Yemenites to
Samuel b. Eli, ZFHB ii 126); J'~ ~'i1 ':J i1~.ll 1ii C1'm C1i1? C1'~iO
n10tD?, 1'00 ~PD~? "they suggest to them by way of advice that it is
good to gather wealth and enjoy life" (D. ~m}:li on Psalm i 1); i1i?
n1~~0:li1 i~tDJ ~'?tD i1'i1~tD "~i ~'i1 "therefore it is befitting that I
should be lord over the other existent things" (Gers., Pent. Comm.,
f. 5 7b); i1~?::l:l ,J i1n:ltDi1i1 i1'i1ntD "~i ~'i1tD i~'JO ~'i1 ':Jii i1itD i10, "if
anything is one of this nature, it is clear that it is befitting that it
should enjoy special providence"4 (id., De'oth f. 9b).
Dl. This ~'i1 can, of course, be explained as copula, but in view of
the existence of the parallel i1i, and its unusual position in front of
the predicate, it seems at least probable that it should be considered
a true pronoun. It would be a special case of the sentence-anaphoric
(above, par. lOc). Its emergence in SH may be due to the influence
of the Romance languages, where such pseudo-substantives were
employed from early times. Modern colloquial Hebrew, under the
pressure of European speech-habits, is developing i1i again as a lesser
subject, e.g. n~::JiD ::J'~ i1i "it is good that you have come".
n. It is rare to find subordinate clauses of a complex structure. The
range of subordinating devices was, as in Arabic, too simple to main-
tain periods. This may be the reason why occasionally in such sub-
ordinated periods the subordinating conjunction is repeated before
the major dependent clause if the minor one (i.e., the one subordinated
to the dependent clause) intervenes. E.g.: i1'i1 C1~ ':J i':Jri1? 1i'~ r~
i1'i1' ?'iJ fOPi1 5':J C1?.ll:l n:l ~'OJ ~"i1i1 "it is superfluous to point out
that if the He at the end is quiescent and vocalic, that then the ~me~
is a long vowel" (B. Ezra, Za~oth, f. I b); ':J ipi10i1 'o:Jn 'iO~ p'
,? iJnni1? ?intD'tD Ti~~ C1i~i1 'n1~ ~~O'tD:J "likewise the philoso-
phers have said that when a man finds such a one, that he must
needs endeavour to become attached to him" (Falaq., Meba~~esh,
f. 21a).

4 The second tli1il may either be a lesser subject as well as the first, or it may
be an anaphoric pronoun referring to the implied antecedent of the relative clause.
5 The ed. has Cltli 'J the second time, too. This is obviously wrong.
 SUBSTANTIVE CLAUSES 159

35. Clauses as Subject

a. Clauses as subject in ordinary sentences are rather infrequent,

e.g. i1:Jii:J ~i:JO m'~ •.. l:l'i1:JO I:li~ 'J:J iiZ.l~1 "and if people think ...
that is not a legitimate opinion" (B. I:Iiyya, Megillah, p. 23); Ti':J i1t;.lI'
I:li~ t;:J i'i1ri1t; :::J'n i1n~ ':J mo "the conclusion is that you are obliged
to warn everybody" (id., Hegyon, f. l7b). The use of iiZ.l~ to intro-
duce a subject clause in the first quotation is in accordance with the
late Biblical usage (Brock., 397), and indicates that the author felt
the construction to be an archaism. Normally it is avoided in SH
by the devices discussed in par. 22c and 35 k-l.
h. Subject-clauses are, however, very frequent in SH with a special
class of predicates, which, for want of a better term, we may here
refer to as clause-predicates. The words serving as predicates in this
construction were originally adjectives, verbs in the third sing. masc.,
or adverbs, and are often still employed as such in SH apart from
their impersonal use as clause-predicates.
c. These predicates always precede their subject-clauses, and are
inflexible. Semantically, the complex is very close to an ordinary sen-
tence with the clause-predicate as adverb. In translation it will often
be of advantage to change the construction in this way.
The subject-clause is either finite and introduced by she-, ki, or
asher, or an infinitive clause. The latter has no subject when an
indefinite subject is implied, otherwise the infinitivus cum nomina-
tivo is employed.
d. This construction occurs already in BH, especially with :::J1t;l "it is
good, it is better", e.g., l:l:Jn~ ipn' ':J :::J1t;li1 "is it better that He
should search you out?" (Job xiii 9); 1i:::Jt; 1:li~i1 nm :::J1t;l-~t; "it is
not good for man to be alone" (Gen. ii 18), also Gen. xxix 19, 2
Sam. xviii 3, Ruth ii 22. :J1t;l has already the typical feature of being
accompanied by an ethical dative; I:l'i;m-n~ i:J.lI 1Jt; :::J1t;l "it would
have been better for us to serve the Egyptians" (Ex. xiv 12). Similarly
with 11:J~ "it is right": p miZ.l.llt; 11:JJ ~t; "it is not right to do thus"
(Ex. viii 22). An indication that these were felt to be in some way
different from ordinary predicative adjectives may be found in the
peculiar method of negation by means of ~t;.
e. In MH, the construction was also applied to a small number of
words, such as .lI1i'[:J] "it is common knowledge", Ti~ "it is necessary",
160 CHAPTER TEN

but was still far from becoming a general syntactical category. The
wide extension of this construction seems to belong to late Midrashic
Hebrew, and took perhaps place under the influence of Arabic and
Aramaic, in both of which it is much more frequently used than in
MH.! In N.W.-European Hebrew and in SH, many words that in
the earlier dialects could only be used in personal constructions
became clause-predicates. The exact lexicographical range of the con-
struction is still unknown. It will be difficult to discover, because of
the wide variations in range between different authors.
f. Although the subject-clause construction was used by Saadiah, its
use seems to have gone back in poetry and ornate prose. This may
explain the reluctance to employ the impersonal constructions we
sometimes notice with Abraham bar J:Iiyya.
This construction, once firmly established in the language by SH
usage, remained widely used in all later forms of Hebrew. It is full
alive and still extending in Modern Hebrew, both in its literary and
colloquial forms.
g. The gradual development of the category is illustrated by the his-
tory of the most frequent of all clause-predicates, "~i, a term of a
vague meaning expressing most connotations of the English "should"
or "might". This is a purely MH word, 2 occurring in the Bible only
once: i1'? nn'? n;'~iiJ n1i.!iJi1 "the maidens which were meet to be

I In Talmudic Aramaic, the infinitive (with or without le-) appears as subject after
t"j'i., "it is better", ~"i~ ni1~ "it is customary", Ti~, '.,:::1'0 "it is necessary", etc.
(cf. Schlesinger, p. 196). In Arabic syndetic subordinate clauses are used after words
like khairun "it is better", kha/igun, jadfrun "it is meet", sawa'un "it is immaterial", la
ba'sa "it does not matter" (cf. examples in Brock., 395a, Reck., 194, where, how-
ever, these are not differentiated from the~comparatively rare~clauses as subject
of regular sentences).
2 In Jewish Aramaic, n!IJ is used in the same manner (Dan. iii 19, Targ. Lev.
v 10, Targ. Job xv 11, in the Palestinian Targum 'on; not in Syriac), cf. also the
Palmyrenian 'rnn~ "it seemed good" (Cook, Glossary, p. 51). The meaning of the
root "to consider good, worthy, etc." seems to be common Semitic, cf. also the
Arabic ra'a "to think fit", and huwa ar'a bi- "he is more apt to ... ". The word
re'uyyoth is hardly a passive participle of n~i, as it would then be the only instance
in BH of such a participle with short u (cf. Ges., 75v). It is probably a qatul-adjec-
tive (Ges., 84h; Brock., i 120). Perhaps the Aramaic forms are loan-translations
from Hebrew, dating from a time when '1~i had come to be considered a par-
ticiple passive.
But cf. the analogous double sense of the Syriac Paqba, paqqab 1. "suitable",
2. "better" for esp. paqqab wa leh l-gavra haw ellu /ii eliled. (Mark 14.21 Pesh.), an
exact translation of the Mishnah phrase. Altogether these forms, both nouns, are
the nearest thing to '1~i I know.
 SUBSTANTIVE CLAUSES 161

given her" (Est. ii 9, cf. Kcmig, 236b).3 In MH it is employed very

frequently, with the meanings 1. "Fit, legally fit" (Yeb. vii 6, BB viii
5, Hor. i 1, etc.), 2. "Able to, apt to" (Zeb. xi 3, Ter. vi 6, 'Orlah
iii 7, etc.), 3. "Likely to" (Bekh. v 4, Pe'ah v 1, Sot. iv 3, etc.), 4.
"Suitable for" (Zeb. ix 1, Shab. xv 2, Yeb. xi 7, etc.), 5. "Proper
for" (Ned. xi 4, B~. i 4, San. iv 5, etc.), 6. "Sufficiently good for,
worthy of" (Pes. vi 5, San. viii 4). It is never used in impersonal
construction, except once in our Mishnah text in a passage that is
almost certainly corrupted. 4
h. The impersonal construction appears in the Gaonic Midrash Pir~e
R. Eliezer. iitv~ rmp'? "~i 1'?' "and you it is meet to call woman"
(ch. xii). It is also employed thus by Saadiah (IJiwi Polemic, stanza
64): tvi~'? "~i 'j'i~" m'?''?.l1 '?:J p "thus it is meet to explain all the
deeds of our Creator". In Mishnaic, these would have been ii"~i n~
iitv~ 1'? nnp'? and tvi~'? m"~i. Perhaps Saadiah merely imitated
the usage of the corresponding word in Arabic, khalfq, which is fre-
quently employed as clause predicate by Saadiah's older contempo-
rary, I:Junain b. Is}:laq (d. 910) to render the Greek isiis.
i. B. I:Jiyya knew the impersonal construction of the word, and occa-
sionally employs it so: ... D1'~ ,'tv.l10 '?'nn'tv t::Ji~'? ,,? "~i r~ "it is
not right for a man to begin his business on the day when ... "
(Megillah, p. 23). Yet he appears to have been rather reluctant to
employ it, and often produces awkward constructions in order to
construe "~i personally: m:Jo '?:JO iiitii'? t::J"'~i n.l1i 'tvj~ '?:J "and
all sensible men ought to beware of any danger" (Schwarz-Festschr.,
p. 24); ii~ p'rnii'?, iim~ i'O'?'? ii"~i t::J'~:J':Jii nO:Jn "one ought to study
astrology and pay attention to it" (ibid., p. 28). The meaning of the
word in these two quotations is quite un-Mishnaic, and does not fit
the construction. There are cases where the struggle of the author
to avoid impersonal "~i leads to even stranger results: t::J"'~i '?~~

3 Ibn Ezra's only comment on the passage is: .,"ri 'i::l;::l .Il1;' j1iD"il "this expres-

sion is well known from rabbinic literature".

4 l:Iagigah ii I, our Mishnah text has O"1.1l" ~::l ~., 1"~:l 1" '1~i (second time:
~::l ~"iD). 1"~:l makes no sense, and should probably be 1"~. The passage then trans-
lates smoothly: "it were better for him if he had not come into the world". The
Bab. Talmud has, however, '1r1; instead of '1~i. Levi explains '1r1i = '1::'i "it would
have been more desirable"; Rashi derives from ilr1'i "he was lenient", and trans-
lates: "it would be a mercy for him if". The very difficulty of explaining the hapax-
legomenon '1ni proves it to be the original word. '1~i was substituted in violation
of Mishnaic usage.
162 CHAPTER TEN

nlili1'?1 r QtI;i1'? "one should believe and acknowledge" (Hegyon, f. 5)

with the indefinite person, or: '?J Cli1:J '?'nm'? ClJI~'?iD 'Q' tl;1i1 'ItI;i
i1iD.liQ "it is fit for the days of their dominance to begin on them all
kinds of business" (Megillah, p. 23). Here the-to the author~nat­
ural construction obtruded itself so far as to substitute the singular
of 'ItI;i for the plural one would have expected. Often the con-
struction is balanced in such a way that it is impossible to decide
whether the 'ItI;i is personal or impersonal, e.g.: nlJr i1:JiQ I:J n~I:Ji1
i1Iin tI;'?Q I:J'? nI'i1'? 'ItI;il i1I~QI "he who trusts in Him acquires much
merit and grace, and is worthy (or it is meet) that his heart should
be filled with joy" (Megillah, p. 1).
j. A similar tendency to avoid the impersonal construction appears
to exist in the style of B. 'Abbas, more than a hundred years later:
'?tI;i1Q i1t1;i' I'? rWi1'? i1'?'nn:J Clitl; '?J'? 'ItI;ii1 "the proper thing for
every man is first of all for him to have fear of God" (f. 9b); nlJIJ
iD'tI; '?J r QtI;i1'? nI'ltI;ii1 nmJJi1 i1ilm "the true meanings of the Law,
which are fit for every man to believe" (f. 11 a).
k. In all other writers, the impersonal construction is used almost
exclusively: CliDi1 I'? lnJiD i1Q '5JJ In'iD 'ItI;iiD Cl.li~i1 "the reason is that
he should give in proportion to that which God has given him"
(B. Ezra, Yesod Mora, p. 17); '?J '?.Ii I:J i:Ji'? 'ItI;iiD lli.li tl;iPQi1 i1J'i~
tI;"n':J tI;"5J,?tI; "the Bible requires a systematic dictionary, in which one
should speak of the whole Alphabet" (B. Parl:lOn, ~rukh, p. xxii);
IJn'tI;'? Cl'i1 i'rni1'? 'ItI;i "it would be right to restore the sea to its pri-
mordial state" (Maim., Teshuboth, ed. Freimann, p. lix); ClJii tI;'i1 Ir
i1:J l'?''? 'ItI;iiD Cl'QJn 'i'Q'?n '?iD "this is the way of the learned in
which one should walk" (id., Talmud Torah, vii 17); IJ:JI'iD 'ItI;i lJ
Cl'tI;':JJi1 'i:Ji "thus the words of the prophets should be understood"
(Abr. Maim., {lobe?:, iii 19a); Cl'p'?n 'J'? i1ri1 iQtI;Qi1 p'?miD 'ItI;i "we
must divide this statement into three parts" (Crescas, Or Adonai, f. 4a);
'.Ii:J~i1 iiOi1 '5J'? nI'i1'? 'ItI;i P "this should be so according to the
system of nature" (B. ~ar~ah, Mikhlol, f. 23a).
1. While in all the above examples 'ItI;i has the connotation of "must,
should", it was by some authors also used in the more Mishnaic
sense of "able". In this sense it occurs in one and the same text
both in the MH personal construction: r:Ji1'? 'ItI;i i1'i1'iD i.li Cli1:J P'Q.lii1
Cl'QJni1 nlilO "to study them deeply until he is able to understand
the hidden meanings of the Rabbis" (S. Tibbon, Eccl. Comm., f. l6a);
 SUBSTANTIVE CLAUSES 163

and with the SH impersonal construction: "to collect in it a great

deal of wisdom" 1iO~'? o~n Oi~'? '1~i il'il ~'?tD... lii::l "so much
so that one wise man would not have been able to compose it"
(ibid., f. 15b).
Dl. Instead of introducing the subject of the dependent verb into
the dependent clause (whether as subject of a finite verb or by the
infinitivus cum nominativo device), it can also be attached to the
clause predicate by means of an ethical dative, and, if necessary, be
taken up within the dependent clause. This we have already with
the BH ::l1~: ptD!ln '~ l'? :n~il "it is better for thee that thou shouldst
oppress?" (Job x 3, and cf. the quotation above under d). It is fre-
quent in MH. In SH it occurs very frequently with "~i and similar
words, e.g. 11tD'? '?~ 1'tD'?o p'n!lil'? 1'Ji11 '1~i '~i1::l l'? "you are surely
a fit and suitable person to translate from one language into another"
(Maim., (lobe?, ii 26); '?'nn'tD Oi~'? ,,? '1~i r~ "it is not right for
man to begin ... " (B. J:Iiyya, Megillah, p. 23); 1'?~~'? ~'?iD Oi~'? '1~i
"one should not eat them" (Maim., De'oth, iv 9); i'pn'? Oi~'? ,,? :n~
1''?!I "man should inquire into it" (B. J:Iiyya, Megillah, p. I).

n. The negative used with clause-predicates is in BH ~'? (cf. above

under d), in MH r~ e.g., 1'iD~i no i1r'~ !In' r~ "it is not known
which one died first" (Yeb. vii 4); iO''? Ti~ r~ "it is needless to
say" (Bab. Ber 14a), and the frequent iiDEl~ '~ "it is not possible"
(c£ Segal, Di~du~, p. 79 note 2). Similarly in N.W.-European Hebrew:
~'::l1::li!l::l O'iDOiDO 'WiD liD,n'?, i'~'? il~J r~ "it is not right that light
and darkness should be mixed" (Rashi on Gen. i 4).
SH follows the tradition of MH:5 n!li '?iD O'i::li n':Jil'? Oi~'? "~i r~
"one should not ignore sensible opinions" (Maim., (lobe?, ii 26); r~
O'?tDil Oi~il om~ iliD!I'iD Ti~ "the perfect man must not do these"
(Falaq., MebaMesh, f. 20a); ~ipOil no~no P'i ,nm'? '?'~iDO'? 1'~:J r~
"it is not right for an educated person to be devoid of Bible knowl-
edge" (B. Ezra, resod Mora, p. 2); ~i'niD '1~i r~ "you must not be
afraid" (Gers., Pent. Comm., f. 47b).
o. Most clause-predicates are original adjectives. Adverbs are com-
paratively rare. Thus !I'i'::l "it is well known" is Mishnaic, e.g. !I'i'::l

5 Modem colloquial Hebrew has gone back to ~., while literary Hebrew avoids
~., (the use of which with adjectives is considered a colloquialism) and rather awk-
wardly helps itself with the inflected form 'j'~.
164 CHAPTER TEN

"it is well known that both of them are wont

in~ Cl'l'::J r~mitl Clii'JitIitI
to kill on the same day" (J:lul. v 3), also Bab. RH 20b, Bab. Sukkah
49b. It is employed in this form only by B. J:liyya: ~~QJ 'niiitl .!I'li'::J
"it is clear that living things exist" (Megillah, p. 9). Later writers prefer
the MH alternative .!I'li' (Dem. vi 11, Ter. iii 1) as more in keeping
with the general trend, e.g.; itli'lPii nn::J iiQ'?itI ::Jitln ii'?':lQiiitl .!I'li' "it
is well known that Solomon composed the scroll under divine inspi-
ration" (M. Tibbon, Cant. Comm., p. 6).
p. The only verb appearing in this role is pn' "it is possible". In
BH this is an ordinary verb form, meaning "to be straight" (4 times
in Ezekiel). It does not occur in MH, and seems to have received
its new syntactical function and new meaning somewhere in late
Midrashic. It is quite current in N.W.-European Hebrew, e.g., T~
iQ~' i::J.!Iiiitl pn' "how is it possible that a worshipper should say"
(Sepher IJasidim, par. 547). Its first appearance in SH is with Ibn Ezra:
Cl'?.!I;) m 'l'in~ nm'? pn' "it is possible for it to be followed by a
vocalic quiescent" (Saphah Berurah, f. 1b). The negative of this is, of
course, ~'?, e.g., n'l'n'l~ 'nitlQ 1~P Clitl n'l'ii'? pn' ~'? "it is not possible
for a noun to be shorter than two letters" (S. ha-Shem, f. la).
q. It is at present impossible to aim at an exhaustive list of the words
that can be used as clause-predicates. Almost every author has his
own favourites. It is interesting to observe, however, that those words
are sometimes strengthened by adding a synonym to a clause-predicate
in common use and that these additions tend then to be used instead
of the original words. Thus we find in the 13th century in B. 'Abbas:
Cl'Q:Jn ':Jii i'lQ'?'? ''l~i'l ::J'nQ "it is obligatory and proper to learn
the ways of wise men" (f. 7b). In the middle of the 14th century,
B. Z,aqah can already say: nnn ri~ii 'iinitl Cl'Qiii ri~ii .!I::J~ '::J:J ::J''lnQ
Cl'Qii "according to the nature of earth and water it would be proper
that the earth should be underneath the water" (Mikhlol, f. 23a).
r. Of some interest, though by no means an innovation, is the use
of the verb of existence as clause predicate in the sense of "must".
Thus itI' is used in BH: 1'? i::Ji'? itI'ii "is there any need to speak
for thee?" (2 Kings iv 13); r~ appears more frequendy in this con-
struction, mosdy in post-exilic texts: Psalm Ix 6, Est. iv 2, viii 8,
2 Chr. v 11 (cf. Ges., 114 1).6 Possibly the origin of this usage is an

6 Ges. mentions a similar use of t6 with the ininitive (Amos vi 10, I Chr. xv
 SUBSTANTIVE CLAUSES 165

ellipsis, starting from phrases like ~'~ii? r~ iii1iDr1 ':J "for there is
no gift to bring" (1 Sam. ix 7).7 It appears in Ben Sira (x 23; xxxix
21). Perhaps by accident it does not occur in the Mishnah, though
it is employed in Tannaitic Midrash, e.g.... 'i~' ?J) ~'iDii? r~ "one
cannot reply to Him ... " (Mekhilta, Beshallal). ii 6) .... ? iD' seems
to occur only in Gaonic Midrashim, however, e.g., Deut. Rabba ii,
in the sense of "to be able" (cf. Ben-Yehudah, p. 2170).
They are common in N.W.-European Hebrew, e.g. l:I'J)01iD 1:J"ii 1?~
rO~ii? r~ "if we had heard it, it would not have been possible to
believe" (Rashi on Is. liii I); 'ir1iD ~?~ 1? r~ "there is nothing for
you except to go down" (id., 1 Sam. xxi 12). SH examples: 1? iD'
r1J)'? "you must know" (B. Zaq;ah, Mikhlol, f. 5b); 'in~ iiiiii? r~
riDJ)O~ P"~ii "one must not doubt Him Who is just in His deeds"
(B. I:Iiyya, Hegyon, f. 13b); ... iD iD'~ ~? ?J) r11?J)ii? r~ "it should not be
allowed to enter anyone's mind that ... " (B. Zaq:ah, Mikhlol, f. 5b).
s. The past tense of this construction, ii'ii with the ethical dative
(instead of the BH ?J) ii'ii) occurs once in the Mishnah: 'r1"ii 1?~
ii'iDii r1~ '? r1i:JOiD:J l?iD r1~ J)iElii? 1? ii'ii 1? ~'n "if I had owed
you anything, you should have collected your money when you sold
me the field" (Ketubh. xiii 8).8 It is frequent in texts of the Arnoraic
period, both in the meaning "must", e.g. 1i'~ ?"Elnii? 1" ii'ii "you
should have prayed on the way" (B. Ber. 3a); and "could": ii'ii iir
iDiEl? 1? "this he could explain" (B. BB 71 a). In Rashi it is very fre-
quent, cf. the examples in Avineri, p. 229. It occurs in Saadiah's
Hebrew: iO~? 1? ii'ii "you should have said" (lfiwi Polemic, stanza
7), and is current in N.W.-Europe, e.g., ::J1m? 1? ii'ii "he should
have written" (Rashi on Gen. i 5); 1i::J? 1:I:J? ii'ii ?10r1~O ~?ii "should
you not have said the benediction yesterday?" (Sepher lfasidim, par.
548). In SH: i':Jrii? 1? ii'ii "he should have mentioned" (B. I:Iiyya,
Megillah, p. 4); r11iDJ)? 1? ii'ii "you should have done it" (J. Tibbon,
Testament, p. 11).

2), but the instances quoted are negatived jussive infinitives, the ~'? being employed
in its ordinary adverbial function (cr. K6nig, 399z).
7 Or it may be proto-Semitic, cr. the Arabic lii an "it must not be that ... ", which
in its contraction fan became the negation of the future tense (cr. Brock., 395b).
8 Perhaps in this passage the phrase has still its literal sense "you had (the means)
to make yourself paid". The construction is also employed in bab. Talmudic Aramaic
in the frequent idiom ir:l'r:l'? ii''? '1ii "he might have argued". Since it is not found in
Syriac it is impossible to say whether this new meaning of the phrase arose first in
Jewish Aramaic or in the MH of the Amoraic period.
166 CHAPTER TEN

36. Clauses in Other Positions

a. Clauses as predicates of nominal clauses are rare in Arabic and

Syriac (Brock., 400) and do not seem to occur at all in BH.l In MH
it only stands in the rather difficult adversative construction with ~ii11
iV (Segal, Di~du~, 436), e.g., i~iVJiV ~ii11 no i~ iO~ "he answered 'it
is dead', whereas it was lamed" (Sheb. viii 2). In SH, they occur,
though infreqently, and are introduced in the same way as subject-
clauses: iinniV'i Cl'~i 1iiV~~ iiO~iV ,.lli11 "and the proof is that one
says in the plural iinniV'i" (B. Ezra, <,a~oth, f. 8a); m lEli1 ~'i1 ~~~
riin 'J~ CnI~ C'iVn ~'i1iV "but it is the opposite in that it makes them
free" (Gers. De'oth, f. 7a). In the latter quotation, the she- may also
be taken to mean "because".
Predicate-clauses are occasionally employed in sentences having as
their subject the sentence-anaphoric ~ii1, a usage that probably devel-
oped out of a similar construction with substantivized relative clauses.
Cf. below par. 40f.
h. Object-clauses are so frequent, and have been so amply illustrated
in former chapters, that there is no need to give further examples
here. They are introduced by she- and ki, as other substantive clauses,
and in their syntax are identical with the syndetic object-clauses of
BH (Ges., 157) and MH (Segal, Di/:Ldu/:L, 435). As in the latter, object
clauses stand also after other than verba cordis, in positions where
BH would have an infinitive, e.g. i'i1TO i1'i1'iV ii~~ nOiiJ i10ni1 m'iT
C~i.ll~ "the action of the sun in shining causes the light to become
bright in the world" (B. I:Iiyya, Megillah, p. 9).
c. Clauses dependent on preposltlOns replace to an even greater
extent than in MH (Segal, 302, 482) the adverbial clauses introduced
by conjunctions proper. The BH plain conjunctions, most of which
were lost in MH, occur occasionally, but, it seems, always with a
poetical undertone, e.g. i10i' nI~~~ i~ r~ ,.ll C'J~ '''J~ ~~ ii10'J
"God tried him sorely with the absence of his sons abroad until he
had no equal among fathers" (J. Tibbon, Testament, p. 3); i10Jn r~

I The instances of predicative clauses in BH which Konig (383b) quotes (e.g.

Gen. xli 27; Ps. x 4; Ecc!. vii 12) are all asyndetic, and their syntactical character
not satisfactorily explained. Still, there may be one or two examples of this in the
Old Testament. For Talmudic Aramaic, Schlesinger asserts the existence of such
clauses (p. 228), but in all the other paragraphs referred to there fails to adduce
any real instance.
 SUBSTANTIVE CLAUSES 167

pp '?:l Oi~ii n'n' iiQ.I] 1.1]' ii:l~'?Qii '?~ T'.I]n "no wisdom equals a
craft, for by the latter man gains everything to be possessed" (Falaq.,
Mebaf:o,~esh, f. 11 a); 11'n~ii 1:l mp' :lp.l] 11iD~'ii mQ'?iDii n1JP'? "to acquire
the former perfection because he acquires by it the latter one"
(B. 'Abbas, f. 7b). Even the Mishnaic she- "because" is rare.
d. Perhaps owing to its being reinforced by the parallel use of M;na
in Arabic, the Biblical n.l]:l "at the time when" followed by a syn-
de tic genitive clause (Job vi 17, 2 Chr. xx 22, Ben Sira xi 19), con-
tinues to be employed, e.g. O'Q ~':lii'? 1J1~' ii'ii n.l]:l "when He wished
to bring water ... " (B. I:Iiyya, Megillah, p. 112). The BH n.l] (Jer.
xlix 8, 1 31, Ez. xxvii 34) is also frequently used. The MH equiv-
alent 1Qr is assimilated to n.l] in these constructions: 1:l'? n~ n'Oii 1Qr
1:l "when he incited his heart in him" (J. Tibbon, Testament, p. 3). 2

e. The only conjunction not containing she- which is commonly used

is ':l "because". This BH word was revived already in the liturgy as
an equivalent to causal she-, e.g. iin~ n'?101 '?mQ ':l "for Thou dost
pardon and forgive" (Singer, p. 46); 1J'1p In.l]1iD''? ':l "because we
wait for Thy salvation" (ibid., p. 49). It does not seem to have been
employed in N.W.-European Hebrew, however, and in SH constitutes
a loan from Spanish ornate prose. Examples abound in all writers,
e.g. on, '?:l mnnE:lQ n':l ':l "for in His hand are the keys of every
womb" (B. Ezra on Psalm cxxvii 3); pn' ~'? ':l "for it is not possi-
ble ... " (id., Zaboth, f. 7b); P 'Q~'? 11:lJ ~'? ':l "for it is not right to
say this" (B. I:Iiyya, Hegyon, f. 8); Oi~ '?:l'? '1~'ii ':l "for the proper
thing for everyone is ... " (B. 'Abbas, f. 9b); 1iin'niDn ?'O:lii n?1~ ':l
"because the foolishness of the fool brings destruction upon him"
(Gers. on Prov. xvii 12).
f. By analogy, she-, etc., is suffixed to some conjunctions which in
BH governed a clause directly. Thus we find iD CiC!l:3 '?.I] ':li~iD O'C!l
O'Q.I]C!lii '?.I] ':li~... m'm~ii "before I speak about the letters. . . I
shall treat of the accents" (B. Ezra, Saphah Be171rah, f. 4b).

2 Already in the responsa of the Spanish Talmudist al-Fasf (1013-1103), JD1'? is

employed with a finite verb, in imitation of the BH nll'?, cf., Ben-Yehudah, s.v.
3 In authors of the second period, r::liI:;) is also employed with the gerund (while
in BH only '!:;)::J or '!:;)D could be constructed in that manner), e.g.,nn'?~n 1ni'pn Cli!:;)
Cl:lnn "before inquiring into the happiness of the wise man" (Falaq., Meba#esh,
f. II a); Cl~1::J Cli!:;) "before they come" (Gers., De'oth, f. Ba).
168 CHAPTER TEN

g. To this group seems to belong also iD '1ll ?J "as long as", a deri-
vation from '1ll ?J in Job xxvii 3, which was taken by the Jewish
commentators as a conjunction: 4 '1rJ'?~ ~, 1rJr "rJn'iD '1ll ?J "as
long as he persists for a considerable period in his studies" (B. 'Abbas,
f. 4b).
h. New composite conjunctions are formed by employing BH, MH,
and newly-created SH prepositions before substantive clauses. They
are a concern of the lexicon rather than of a syntax, as are the new
functions attributed to old-inherited composite conjunctions. An
example may be cited: iD '~rJ "because" (cf. in Arabic min jihati an,
employed by Ibn Khaldun, cf. Dozy, Supplement, ii 787): m? 1iD1n' ~?
ni1m m1J nrJ i1i'~~ 1ll" ~?iD '~rJ "they are not keen on this be-
cause they do not know clearly what the intention of the Law is"
(B. 'Abbas, f. lOa).
i. All these composite conjunctions can be employed with iD, iiD~ or
'J indifferendy. Thus in~ and its secondary forms ('in~, in~rJ, in~?),
which in BH is always construed with iiD~, has all three particles in
SH: mrJ1~n 'rJJn 'i~' 1:Ji~'O iiD~ in~ "after we have recounted the
opinions of gentile scholars" (B. I:Iiyya, Megillah, p. 14); P ~1niD in~
"since it is thus" (S. Tibbon, Eccl., l6b); n~iDnrJ nJ~?rJ ?J~ 'J in~rJ
OJ1nrJ nn~ "since you are an expert in every handicraft" (Falaq.,
Mebal),f!,esh, f. 11 b).
j. To the BH iiD~ i1~ll~ is formed not only a variant with she-: i1~ll~
'n~ ~~1rJrJ oniD "because they are articulated in one place" (J. J>.iml).i,
Zikkaron, p. 4), but also one with ki: i~'n l~nn' 11iD?~ 'J i1~ll~ "be-
cause in actual usage the situation is reversed" (S. ha-Shem, f. 8b).
The BH iiD~ 1llrJ? is supplemented by a ki form: nrJJn OiD 'J 1llrJ?
... ?ll irJ~J "because the term wisdom is applied to ... " (Falaq.,
MebaMesh, f. lla). To iiD~ 'J~rJ is added a 'J 'J~rJ: 11'Jn~ pOll ~?
~'n i11rJ i1J'~1 n'rJ 'J 'J~rJ "he did not occupy himself with logic,
because it is a doubtful accomplishment" (J. J>.iml).i, S. ha-Galuy, p. 2).
In this respect SH merely continues a tendency already existent
in BH (Brock., 416), and it may well be pure accident that our Bible
text does not provide instances of the above prepositions introduc-
ing ki- clauses.

4 Neither Rashi nor B. Ezra comments on the verse. Mqudath :?:iyyon renders
the phrase by iD 101 .,~ "as long as".
 SUBSTANTIVE CLAUSES 169

k. From this adverbial use of conjunctional clauses governed by

prepositions must be distinguished the cases in which a syndetic
clause is indirectly governed by a verb, or by a preposition in its
proper function. E.g.: 1:l'.!I,i1 1:l''?:J~QJ m'?pi1 'iD~Q 1i1 T''?n '?:J "all
your illnesses derive from the fact that you were careless about harm-
ful food" (J. Tibbon, Testament, p. lO); l:liDiniD '?~ I:lm"Q .!I'm ~'?
i1piiDni1 n~r I:li1J "their perfection does not go so far that this desire
should be instilled into them" (Gers., De'oth, f. Bb). 1:l1~' 'iD~J '?'nm
1:li1'? 'JJnQ "he begins with the fact that their lusts get the better of
them" (B. I:Iiyya, Hegyon, f. 14a).
1. As, however, in the adverbial constructions the substantival char-
acter of the clause was no longer felt, and the she-, etc., was a mere
appendage to the conjunction similar to the French que, a need seems
to have existed to differentiate the more properly substantival clauses
after cojunctions. A means for that was found in the substantivized
relative clauses (see below, par. 40). Substantival and relative clauses
being rather similar in construction, there was no great difficulty in
substituting the one for the other. Of course, the process was not as
conscious as it is described here, but must be imagined to have arisen
from a subconscious identification, or in other words, false analogy.
Possibly the Arabic mii ma~darryya as equivalent of an (Reck., 192.2),
played some role in favouring this development. Examples: lm'?:JOQ m
TJ'.!IJ l:l:Jn n"i1iD i1QQi "this comes from your foolishness and from
the fact that you were clever in your own eyes" (J. Tibbon, Testament,
p. 9); 1:l'~iOi'?'~i1 i~'?nni1 'J:JiD i1Q'? "owing to the fact that the philoso-
phers differed" (Gers., Mil~amot, f. 4a); m'iO' l:li1iD i1QJ nniO'i1 n'?ir
"except the elements in so far as they are elements" (Zeral).iah, He-
Haluf:;., vii 96).
:m. Once established in such cases, the i1Q also penetrated into the
adverbial constructions, and can be, though rarely, inserted into any
composite conjunction whatsoever. E.g.: ... OiQ'Ji1 mJ iJ'iD'iD i1Q ,~o
"because by this set of rules they achieve ... " (Gers., De'oth, f. 9b);
l:l'?i.!lJ i1iD'~ i1.!1'i1iD i1~'iD i1Q 'J~Q "because he saw that evil spread
in the world" (id., Mil~amoth, f. 57b). It is not quite clear, on this
scanty evidence, whether there was really a tendency for confusing
the constructions, or whether the more substantial construction was
chosen by Gersonides in order to import some specific shade of
meaning to the sentence.
170 CHAPTER TEN

n. Although not strictly a substantive clause, the /fal-clause may be

mentioned here with the other adverbial clauses. While in BH (Brock.,
321) and MH (Segal, Di~du~, 419) asyndetic /fa I-clauses still occur,
though less frequently than those introduced by we-, the syndetic
type is the only one in use in SH. The circumstantial clause had
been preserved in N.W.-European Hebrew, e.g., Cl~tD'? 1'r~ ~'~~~
nim~ e'po'lJ cm i:JtDtD' "he provides food for the tribe of 1ssachar,
while the latter occupy themselves with the Law" (Rashi on Gen.
xlix 13), and it is possibly there that the we- type was generalized.
o. Examples: 'n~ ~~,~~ cm m~ in~ m~ i~''? ,'~'?nn '?lJ ntDp "it is
difficult for the pupil to pronounce two sounds after each other when
they are articulated in the same place" (B. Ezra, Saphah Berurah,
f. 43b); 'm~ e'J':JnD eJ'~' li,n '?lJ l'?n'? e'~~n '?:J "all those who
come to adopt a way of life without fully intending to do so" (B.
I:Jiyya, Hegyon, f. l4b); ~lJi ~'m ,,~''?n ptD '?:J, r",~n n'Jp e,~'? ~'?
en'?'? "it is impossible for a man to acquire virtues, let alone to
study, while he is hungry for bread" (B. 'Abbas, f. 4b); itD~~ '~
n'?,n ~'m i~' r~'tD "he cannot possibly understand anything while
he is ill" (Maim., De'oth, iv 1); r~n ~'? ~'m 'm~n ,'~'?nn i~~' ~'?
"the pupil must not say 'I have understood' when in fact he has not
understood it" (id., Talmud Torah, iv 4); ~'m ~in ni'~ '?~P' ~'? i~~n
i~~ "wool will not take the shape of a sword as long as it remains
wool" (Zeral:tiah, He-/falur:;, vii 96).
 CHAPTER ELEVEN

ADJECTIVE CLAUSES

37. A.ryndetic Relative Clauses

a. Already in BH, asyndetic relative clauses occur almost without

exception after indeterminate antecedent (Ges., l55d; Baumann, Hebr.
Relativsiitze, p. 6; Strauss, Ben Sira, p. 70). In MH, where asyndetic
clauses are very rare, they also seem to occur only after indetermi-
nate antecedents (cf. Segal, 477). As this agreed also with the rules
of Arabic grammar, I asyndetic clauses are in SH found only after
indeterminate and generically determinate nouns.
h. Examples with indeterminate antecedents: ~~ C1'1Ji i1Q~nJ tD'
C1tD11':l 1.IJi' "there are matters in the Talmud of which they do not
know the meaning" (B. Ezra, Yesod Mora, p. 2); n'-'1Qj m'~1 F.lJiT iT~1Q
1i1iT 'tDJ~ '1Ji 1tD'nJ' "theoretical consideration provides decisive argu-
ments which disprove the statement of the people of India" (id., S.
ha-Shem, f. l2a); iT'c:l1:l ~.IJ n'jtDiTQ 11~p C1'J1tD C1'J1iJ iTtD11':lJ 1J~iT
"they adopted various methods for explaining it, which fell short of
accounting for points of detail" (M. Tibbon, Cant. Comm., p. 5);
iT01JQ ~:l11 iTJ iT'iT iT~1ij iTJ'iQ "a great city in which there was an
expert physician" (Falaq., Meba#esh, f. 12b); 1QiPQ:J riT' n1J:Ji C1'J:J i'~i'
"begets sons and daughters that will take his place" (id., Wikkuab,
p. 82); .. C1iTJ C1'1P' n1~~QJiT .IJ:Jc:lQ n1:J1 n1J.lJc:l 1~~9' 1:JJ "there are
many arguments from the nature of the existing world by which one
may confirm ... " (Gers., Milbamoth, f. 48b); 11JJ iT'iT iT.lJ1Q C1~'~' ~~
C1iT'~.IJ ~1J~ "it will not save them from an evil that was about to
befall them" (id., De'oth, f. 8a).
c. As in BH (Baumann, op. cit., p. 16) and in Arabic (Reck., 200.2;
Brock., 255b), an antecedent with generic article may also be fol-
lowed by an asyndetic relative clause. This is frequent in the sec-
ond period: iT'J~iT 'J1~ ~J 1pn' ~J1niT J1J ~1iT "he is like a captain
who looks after the needs of a ship" (Falaq., Meba~~esh, f. l3b); i~Q

1 There the rule is not always observed in practice, cf. Reek., 200.3.
172 CHAPTER ELEVEN

C'!]'O 1~tD" C'1il P11:l' prnil m1il "by the strong wind that crum-
bles mountains and splinters rocks" (Gers., De'oth, f. 6b); C:I il~T 1~:J
••• Ci1~ '~.l' C'Pi'~il mQ"nil rJ!]Q P "this is also seen from the true
dreams in which one sees ... " (ibid., f. 8a).
d. I have found no cases of asyndetic relative clauses with deter-
minate antecedent. Although these occur in BH (K6nig, 380cd,
Driver, Tenses, p. 537 note 30), it appears that Arabic grammatical
theory has exercised a decisive influence on this point. 2
e. For asyndetic clauses with participles see below par. 38c.

38. Syndetic Relative Clauses

a. Of the two particles, she- and asher, introducing relative clauses

with finite verbs, SH prefers she-, but in contrast to MH uses also
asher. In this respect, it follows the liturgical language (cf. Segal, 77).
As asher is hardly to be found in N.W.-European Hebrew, we may
assume that it came into SH from Spanish ornate prose. It occurs
in all writers from Bar J:Iiyya onwards.
h. There are no cases with ha- as relative particle in verbal relative
clauses (Ges., I 38ik). On the other hand, ha- is employed regularly
before nominal and participial clauses of which the subject is identical
with the antecedent. I The fact that it can be employed after inde-
terminate antecedent, as in MH (Segal, 376a) proves that it is different
from the attributive use of the participle: ,~ np~iJil il1'~ '!] C1':1
... ilm'il' :"J':Iil "it is the cause for a form which is inherent in a
body to be ... " (B. J:Iiyya, Hegyon, f. 4); nn:J1n CilQ in~ ,:J n':J'il'
" il"~1il "to accord to every one of them an admonition that is proper

2 It remains to be investigated whether the inverse case: syndetic clause after

indeterminate antecedent, occurs in SH. It was frequent in BH (cf. Baumann,
op. cit., pp. l4f) and virtually the rule in MH.
1 The existence of a relative ha- is admitted by most grammarians (Ges., l38i;
Brock., 366d; Konig 52 and 61 is not quite clear on this point) only where such
clauses contain finite verbs. The ha- with the participle is considered an ordinary
article (cf. also Sellin, Doppelnatur, p. 30). This mayor may not be true for BH. It
is hardly true for Arabic, where the al- in phrases like al-maqtiilu abiihu "whose father
has been killed" cannot be anything but a relative pronoun (cf., Reek., !iJnt. Verh.,
pp. 596f). In MH this ha- is already in important respects different from the arti-
cle (cf. Segal, 376, who describes it as "virtually equivalent to a relative clause";
Strauss, Ben Sira, p. 70).
 ADJECTIVE CLAUSES 173

for him" (ibid., f. l4a). An interesting illustration for the need that
was felt for this introductory particle is in D. ~ml).i on Psalm i 3,
where the Cl'r.PJ?El ?.lJ ?1ntD f.lJ:J "like a tree that is planted on rivers of
water" of the text is rendered in the commentary by ... ?.lJ ?1ntDiT f.lJ:J. 2
c. We find, however, rare instances of such participles without ha-.
It is doubtful whether these were intended to be attributive participles
or asyndetic relative clauses. E.g. Cl'?J?JiT Cl'JtDiT mi:JnQ m'1pJ 'ntD? fQi
"a symbol for two points connecting the two spheres" (B. Ezra,
S. ha-Shem, f. 2a); iT?1'J n?.lJ1n 1J? iT~':JQ iTQ'PiT ~':J~ "I shall give an
introduction which teaches us an important point" (Zeral).iah, He-
Ijatui:" vii 96). Cf. also below par 39d.
d. There seems to be an incipient tendency to construct she- clauses
on the pattern of those introduced by ha-, as in: itD~ mi1~iT?:J
mQQ m:J1~n "all the shapes that are derived from it" (B. I:Iiyya,
Hegyon, f. 3). It does not appear to have spread much.
e. In the majority of cases the retrospective subject-pronoun IS In-
serted, as in ~iPQiT l1tD?? iTQ1' ~1iTtD iTQ ?:J1 "and everything that re-
sembles the language of the Bible" (B. Parl).on, 'Aruldz, p. xxii); Cl:J? n:JtDnQ
iT.lJi? ~'iTtD "the thought of their heart, which is for evil" (D. ~ml).i
on Ps. v 10); 'Q1i l1tD?:J tD~ ~1iTtD l1i~~Q "Me!a!ron, which means
fire in Latin" (Gers., on Prov., i 8).3 SH is in this respect more con-
servative than N.W.-European Hebrew, where the retrospective pro-
noun can be inserted or omitted at will, e.g. 1? P1':J ~1iTtD Cl'~ "a
man who is trusted by him" (s. Ijasidim, 1099); liT? 1"1~itD lm~ "those
who are suitable for them" (ibid., 1109). Nor was it influenced in
this respect by contemporary Arabic prose, in which the omission
of the restrospective subject pronoun is extremely frequent (cf. Reck.,
Synt. Verh., p. 528).
f. The relative clause always follows immediately upon its antecedent,
even where a lengthy clause breaks the continuity of the major sen-
tence. E.g.: 1:J?iT iTQiT 1mn ':J 1?'~' ~?1 1?'.lJ1' ~? itD~ 1mm 'in~1 "and

2 An imitation of Arabic constructions with an impersonal passive participle of

the type al-maghrfftbu 'alaihi (Reek., 206), is found in m.:lr:l pnllir:lil jiiD'?il "the lan-
guage from which a translation is made"; i''?~ pnllir:lil jiiD'?il "the language into
which a translation is made" (Samuel Tibbon, in his preface to the Moreh transla-
tion). It seems, however, not to have been currently employed.
3 Contrary to Arabic usage, relative clauses in SH can be not only qualifYing

but also descriptive, as in this case, which in Arabic would have to be rendered
by means of a f:lal-clause.
174 CHAPTER ELEVEN

after the vain theories that do not help or save, because they are
idle, they have gone" (Maim., -!lobe?, ii 25); ilQli t-l;'iliD miil nt-l;i nt-l;~JI
Oit-l;il mo' it-l; il'IJil 10 i'It-l;'? "and when there goes out this spirit,
which resembles a whiff of air, from the body, then man dies"
(B. Ezra, Yesod Mora, p. 11).
g. The relative particle is placed only before the first of two coordi-
nated relative clauses. This is done not only when the two clauses have
the same rection with regard to the retrospective pronoun (as in the
first quotation of the preceding section), but also when the retrospec-
tive pronoun has a different function in each: milOI il'l~ I'? iD'iD 'It-l;il
O'?I.l1 ':JJ'? "the light, in which a form is inherent and which benefits
the inhabitants of the world" (B. I:Iiyya, Megillah, p. 16); O"5:l0il
On'?li n':Jm OilJ p0.l1nniD "the books with which you occupy yourself,
and others than which you leave alone" (Maim., -!lobe?, ii 27).

39. Substantivized Relative Clauses

a. As in BH (Ges., 138e; Brock., 368d) and in contrast to MH

(Segal, 422) relative clauses can in SH be substantivized without any
change in construction: Oil''?.l1 O'5:lil.l1 O'O:ln O:J't-l;iD "those who are
not clever are more numerous than they" (B. I:Iiyya, Megillah, p. 2);
.l1ilJ nJ:lJ iiU~ il'il ~" i1iU.l1r.li1 '''I'? "but for its emerging into reality,
that which existed potentially would not have been known" (ibid.,
p. 9); I:JOO '?~:Jm'? nJ:l I'? rt-l; 'iDt-l;1 ':J.l1i1 IPiD.l1J "by oppressing the
poor man and him who has no power to save himself from him"
(D. ~ml:J.i on Ps. xl).
h. These may be dependent on prepositions: 'JIJ r:::n~' 'iU~r.l int-l; rt-l;
... 'It-l;, Oil''?.l1 "no one of those whose passions rule them is wor-
thy ... " (B. I:Iiyya, Hegyon, f. 14a); iD' 'iDt-l;'? iliD.l1n t-l;'? ml~OJ nil~ rt-l;
rJil'? J'? I'? "there is no trouble involved in the negative com-
mandments for him who has sense to understand" (B. Ezra, Yesod
Mora, p. 10); 1':J5:l'? 'iDt-l;'? iDl,'5:l i1iil PI05:lil "this verse is an explanation
to that which (or: the one which) precedes it" (D. J5..iml:J.i on Ps. i 6);
m.l1'ilO Oil'? t-l;IJ'iDJ Ci1'iD.l10 '?.l1 Oip5:l'iD t-l;'? "not that He punishes
them for their deeds by the evils that befall them" (Gers., De'oth,
f. 8a; cf. below under rn).

c. Ha-clauses: OilJ 'i1i:Jil O'nJiDO Oil "they praise him who is careful
about them" (B. I:Iiyya, Hegyon, f. 16b); O':J.l1:J O:J't-l;iDl O'O'?iD O':J.l1:Jil
 ADJECTIVE CLAUSES 175

~'?~:l "those who respond are hale, and those who do not respond
perish" (id., Megillah, p. 115); ~~" ~'ntD :ltD,nn nl'~ p ?l' "therefore
he who thinks it transitive is mistaken" (B. Ezra, Saphah Berurah,
f. 24a); ]i10 in~? nO'i ':J'~ rll':li~O i:l,nOn ~~9' "thus that which is
combined from the four of them does not resemble any of them"
(Maim., Yesode ha-Torah, iv 2).
d. It appears that the asyndetic relative participles described above
(par. 38c) could also be substantivized. One might, however, prefer
to take the participles in the following quotations as substantivized
attributive participles: 'rm~p:l ni,n iO'? tD'tD ':J~O "for there are some
who learn the Law in their youth" (B. I:Iiyya, Hegyon, f. 17a, equivalent
to ... iO'? ~'ntD '0 tD'); ... niO?O ~'n itD~ n:l ro~o, ni,m iO'?
"one who studies the Law and believes in it, and who teaches it ... "
(ibid.).
e. When a substantivized relative clause is employed as a participle,
the subject may be the sentence-anaphoric ~'n (which in SH can be
employed only in this construction): ':J'n1:li no~ itD~ ~'n "this is
what our Rabbis said" (B. I:Iiyya, Megillah, p. 20); iO'~ ~':l:JntD ~'n
"that is what the Prophet means by saying" (Maim., Yesode ha- Torah,
i 4); 'niO~tD no ~'m "and that is what I meant by saying" (Gers.,
on Job xxi 34), the last with pseudo-antecedent, cf. below under g).
This construction occurs in N.W.-European Hebrew, e.g. iO~tD ~'n'
ntDo "and that is what Moses meant by saying" (Rashi on Gen. xlix
13) and must have been taken over from there. It may, however,
be found in older types of MH, though it does not appear in the
Mishnah.!
f. A further development of this is the employment of ~'n as sub-
ject to a predicative substantive clause: no,o, "i~ ~i~ ':J:l ,:JOO ,np'
,ni,n ,:JOO nO?'tD ~'m "people take his fruit and good example,
i.e., they learn from him his knowledge of the Law" (D. ~ml:ti on
Ps. i 3). This looks like a rendering of the Arabic dhiilika anna or
huwa anna (Reck., 195). Possibly the extension of the construction to

I This construction is certainly not connected with the Talmudic Aramaic use of
~1i1 as copula before a substantivized relative clause employed as predicate as in:
,., 'lJ:::l'~i ~1i1 ~ii1 "that is just what is doubtful to me". Such sentences have in
every case a subject preceding the ~1i1 (cf. Schlesinger, p. 221). The same idiom as
in SH occurs in Arabic in phrases of the type huwa 'lladhf gala "that is what he
said".
176 CHAPTER ELEVEN

conjunctional clauses took place under the influence of the similar-

ity between Arabic huwa lladhl and huwa anna.
g. Much more frequendy, MH usage (Segal, 422) is followed and
the substantivized clause given a pseudo-antecedent ("correlative",
Brock., 377ft'). This may be, as normally in MH, the interrogative
pronoun: 2 ~1l1 1'; tD'tD 'rJ tDP~'; "to search for one who had a fowl"
(B. Ezra, Yesod Mora, p. 6); i1~ 1'rJ~rJ 1j'~tD 'rJ'; i11rJ';rJ "he teaches it
to one who does not believe in it" (B. I:Iiyya, Hegyon, f. l7a); 1n1~
n1~r ~~'; 'jj'1'tD 'rJ n1~r ~~'; 1'1' "may our Rock be forbearing to
him who is forbearing towards me" (J. ~ml:ti, s. ha-Galuy, p. 2);
1n~i1~~ nt;)~ntD 'rJ~ ••• fll'ni1 "take counsel with one in whose affection
you have confidence" (J. Tibbon, Testament, p. 6). With plural pred-
icate by constructio ad sensum: ';t;)~'; 1';1ntD'tD 'rJ~ i1~'i1 i1'P' m~
n"'tDi1 n'rJ "the like of this often happens to those who endeavour
to frustrate the decisions of God" (Gers., De'oth, f. 9b).
h. With non-personal subject implied, the neutric interrogative is,
of course, employed: n1tD, ~1i1tD i1rJ n11~rJ~ 10'j~i1tD tD' "some include
among commandments things that are merely permitted" (B. Ezra,
Yesod Mora, p. 6); i1tDlItD i1rJ ';~ i1tDlI 1'1i1 1rJ ')usdy he did all he has
done" (B. Parl:ton, 'Arukh, p. xxii); 1'j'tDrJ 1l111' ~'; 1mrJ lI11' ~';tD i1rJ
"a thing of which the nature is unknown-its accidental features are
perforce unknown" (Gers., Mil~amoth, f. 4a); i1rJ~ n'i1'?~i1 i1J1~i1 cr'?tDi1'?
1'1~lI~ Cl~'~tD "to achieve the divine intention concerning the pur-
pose for which He created them" (id., De'oth, f. 9b); i11prJ~tD i1rJ 1np';
Cl~lI~tD i1rJ Cl1prJ~ "they took the features of the accident for the fea-
tures of the essence" (Zeral:tiah, He-ljaluz, p. 96). A rather unusual
constructio ad sensum, giving the substantivized clause the function
of a plural because it represents a plural noun, is found in: 1mlljrJ ~';
Cl':I1jllm1 Cl'lI1:1l1:1i1 Clm mrJ CllljrJ'; Cl'~';rJ 'j~ l'1tD i1rJ n1rJ~nJ pOllmrJ
"there did not prevent him from studying the sciences those things
which usually prevent princes therefrom, namely lusts and pleasures"
(S. Tibbon, Eccl.-Comm., f. 16a).
For relative clauses of this type employed as substantive clauses,
see above par. 36, n. 1.

2 It occurs already in BH, though rarely, e.g., Ex. xxxii 33, 2 Sam xx 11, and
twice in Ecc!. (cf. Brock., 377a). It is found in all Aramaic dialects.
 ADJECTIVE CLAUSES 177

i. Already in MH, a certain confusion is to be noticed between these

substantivized relative clauses preceded by an interrogative pronoun
and the dependent questions (cf. Segal, DiMu/c, 425). Indeed it is
found even among the few cases where the construction is employed
as a Mishnaism in Eccl. 1'1n~ i1'i1'1D i10:l m~1~ "to see what shall
be after him" (Eccl. iii 22). The confusion continues in N.W.-European
Hebrew: 11'JIDi11D i1J1r 1r'~ Cl'~'1E)i1 ,~ Cl'1'JO "the debauchees tell me
which courtesan they have hired" (S. /fasidim, 80). It is remarkable
that such relative clauses standing for dependent questions are so
rare in SH. E.g., l~ i11plD i10 n'~1 "you have seen what happened
to you" (J. Tibbon, Testament, p. 9); 'n~:lOID i10 nJ)1' "you know what
I suffered" (ibid., p. 10).
k. As in the Aramaic dialects (Brock., 378bc) the interrogative pro-
nouns as pseudo-antecedents became weakened with time and were
replaced by demonstratives. zeh is used in this way in MH: i11Dl7' i10
~~O ~~ID i1r "what should he do that did not find" (Sanh. iii 8), and
rarely also in N.W.-European Hebrew: :l1~ 1n1' ~1i11D i1r~ In:l 1n
"give your daughter to the one that is better" (S. /fasidim, 1142). It
is not very frequent in SH: i1'1:l1D plD1ni1 i1rO :l1~ ~~ 11D' n:l np1~i1
"he who marries a Jewess is better than the one who falls in love
with a captive woman" (B. Ezra, Yesod Mora, p. 6); 'J~E) ~l7 ~1i11D i1r1
~1:l' ~~ Cl'O "and the one that stands on rivers of water shall not
wither" (D. ~m}:li on Ps. i 3).3
1. As the pronoun for the nearer object, so that for the farther object
can serve as pseudo-antecedent from MH onwards. In its only occur-
rence in the Mishnah it has still the full demonstrative meaning:
i11rl7i1 nnE) ~l71D 1m~~ Cli1~ r~:l "they come to those that are at the
gate of the courtyard" (Sanh. xi 2). In a Tosephta-passage it is already
used as pure pseudo-antecedent: 11J1J 1m~ r11p ~~i1 n':l 1'i11D 1rI1~~
•.. ~11P i1'i1 i111i1' '1 "him whom the Hillelites call a glutton,
R. Judah used to call ... " (Tosephtah, ed. Zuckermandel, Niddah,
ix 19), and it is frequent in Midrashic literature of the Amoraic
period (cf. Ben-Yehudah, p. 441).
In N.W.-European Hebrew: 1:J~1D 1m~J "like those we have" (Rashi
on Shabbath 84a); Fn:l:l 1'i11D Cln1~ "those who were in their houses"

:3 In Modern Colloquial Hebrew, i1t is mostly employed as pseudo-antecedent,

probably under the influence of European languages (German "deIjenige welcher",
Yiddish "der was").
178 CHAPTER ELEVEN

(Solomon b. Simon in Kreuzziige, p. 2); l'1i1 n~ '~'iD Dn1~ "people

who will give false judgment" (s. /fasidim, 1379); 11jJ~iD i1r "the one
opposite you" (ibid., 1142). It is, in fact, the most frequent word for
this purpose in that dialect.
In SH, it is comparatively rare, and seems to be lacking completely
in some authors: '?~l'rJiD' n1~1~J "i1iD Dn1~ "people who were in
Moslem countries" (J. J>iml;i, S. ha-Galuy, p. 3); D'10nmiD Dn1~ 'rJ~
1rJ''? "like those who try to appear pious by saying" (Gerondi, Sennons,
( 23b).
Ill. The peculiar Arabic man . .. min or mii ... min construction replac-
ing attributive relative clauses (Reek., 211) seems to have been intro-
duced into Hebrew first by the translators, and is frequent in authors
of the second period. It was adapted to Hebrew idiom by substitut-
ing iD i1rJ for the Arabic relative mii.4 E.g. ,JJ l"ji1iD i1rJ ... n'tI;1 1J~
i1'?'"Di1 1rJ P Dj ,'1ntl; "you have seen what honour fell to our lot
after that" (J. Tibbon, Testament, p. 4); ~1'Ji1 lJ'niD i1rJJ ... 11"JiDrJ 'J~
'1'J~rJ "I adjure you by the respect for me that God had com-
manded you" (ibid., p. 10); D'1J1i1rJ mi1 1~0i1 ''?'?~iD i1rJ m'nnJ ... p1~
D"J".lii1 "Chapter ... completing the theoretical matters contained in
this book" (Gers., Milbamoth, ( 48b); 'J'rJ~i1'? ,'?'ni11 ,''?.li ,'?1jiD i1rJ
i11'pm n~rJ m~i1 nl'ji1rJ "the fact that they were brought up in the
habit to believe that the truth can be discovered by this type of
inquiry" (ibid. 4a).

4 Besides SH, only Nabataean has this construction, no doubt also by borrow-
ing from the Arabic. Just as in SH, a pseudo-antecedent is employed: i'?'n' 'i na
j'iJi la nJi 1El'?n'? "all male children that will be born to this Khalaf" (Brock.,
378a).
 CHAPTER TWELVE

CONDITIONAL CLAUSES

40. Conditional Clauses

a. Asyndetic conditional clauses (Ges., 159b Segal, 484) are rare in

SH, except in the more Mishnaic style of Maimondes' Mishneh Torah.
Even in that book they are much less common than in MH, e.g.
1'? ''?mQi!l 1lJ C1i!lQ rr 1J'~ i1J1i!ln i1i!llJ1 ni:J i1J r~i!l i1i!llJ n1~Q '?lJ C11~ 1JlJ
"If a man transgresses a positive commandment which is not pun-
ishable by extirpation and repents, then his sin is forgiven even before
he moves from his place" (Teshubah, i 4).
h. The only type of asyndetic construction which is common in SH
is the one in which the protasis is a substantivized relative clause,
as in MH (Segal, 442) and Arabic (Reek., 257.2). As in MH, the
subject of the apodosis need in such cases not be identical with that
of the protasis: 1i11m ~'i1 J'? 1'? ill' C1~ 'i1 ni1nJ 1'Qn m1i1i1 "if one
always ponders over the Law, it will show him the way, if he has any
sense" (B. Ezra, Yesod Mora, p. 13); i1'i11 ... C1'Q'i1 1~i!lJ 1i!l5)J mlJQi1
'?Ji1 '1J1 '?:JQ 1J1i!l'?1 1'5) 1Qi!l ~1i1i!l ..• 1n'JlJn "if one fasts on other
days, then let his fast consist in refraining from idle talk" (B. I:Iiyya,
Hegyon, f. 15a); l'?Qi1 1JJ:J C1'1J1 1J1Qi!l 'Q 1Q:J n'QQ 11J1i1 IQ ill'
"some talk is deadly, as when one speaks against the king" (Falaq.,
MebaMesh f. 19a); C1i1''?lJ 01lJ:J' ~'? C1'1'Q'?m 1J'Ji1 ~'?1 1Q'?i!l J1i1 "if a
teacher teaches something, and the pupils do not understand, let
him not be angry with them" (Maim., Talmud Torah, iv 4).
c. The conjunction introducing the protasis of syndetic relative clauses
is C1~ for real conditions, as in all other types of Hebrew. For hypo-
thetical conditions, either the BH 1'? (Ges., 159 1) or the MH 1'?~
(Segal, 490, also Eccl. vi 6, Esth. vii 4). Very frequently, however,
C1~ is used in hypothetical conditions as well as in real ones. This
was common in BH (list of cases K6nig, 390t), I but in MH the two
kinds of clause were sharply distinguished, and C1~ employed only

1 Similarly in Talm. Aramaic (Schlesinger, p. 270), in Syriac, and always in

180 CHAPTER TWELVE

for real conditions (cf. Nodel, Der ;::usammengeset;::te Sat;::, 43d). The BH
confusion was reintroduced in the European writings of the Gaonic
period, e.g. r1~iJ'; ';1:1' i1'i1 1i1J ~i1P ~1i11D '~ ';:1 i'Oi1 ';17 1Jn'J C~
C';117 "if they were given in their proper order, anyone who read
them could create a world" (Midrash Tehillim on Ps. iii 1, ed. Buber
1891; the print Warsaw 1865 has '';~';~). It is frequent in N.W.-
European Hebrew, especially in Rashi (cf. Avinery, p. 204), and was
taken over into SH from there.
d. Examples: ';e!)J m';r~i1 m:1 i1'i1 ~'; C~ C';117'; ~J ';1J~i1 i1'i1 ~'; "the
Flood would not have come into the world if the power of the con-
stellations had not been annulled first" (B. I:Iiyya, Schwarz-Festschr.,
p. 28); rp'rJ i'O'; C':1'i~ 1'i1 ~'; C'P"~ ';~ilD' ';:1 1'i1 C~ "if all Jews
were righteous, they would not require the penal and civil laws"
(B. Ezra, Yesod Mora, p. 3); P 1~1D i1'i1 ~'; 'iJ17 i1'i1 C~ "if he had been
a Hebrew, he would not have been called by that name" (J. ~mI:ti,
Zikkaron, p. 4); Ci':1~ r1"i1 ~'; C'in~ ,'J cm~ i1~1i r1"i1 C~ "if you
had seen them in other people's hands, you would not have recognized
them" (J. Tibbon, Testament, p. 4); i1i1r1i1 i1';'171i1 i1~ P iJ'i1 i1'i1 C~
1D'~i1 m'; . .. "if it were so, what good would the Law ... be to that
man?" (Maim., flobe?, ii 26); i'10~1 ';J:l1~ Ci1i1 C'iJ'i1 r117:1i1 i1'i1 ~'; C~
C~1J Cie!) i117','i1 Ci1J i1'i1r11D Pr1' ~'; "if the emergence of those things
were not orderly and according to a plan, it would not be possible
for them to be known before they happen" (Gers., De'oth, f. 8a).
e. The negative of 1'; is, as in MH (Segal, DiMu~, 454), employed
mainly with a single noun following in the meaning of "were it not
for".2 The particle used is, however, not the MH 1';1';~ or '';~';~,
but the BH '';1';, which has this construction only once in the Bible:
'r~ . .. 1J'; i1'i11D 'i1 '';1'; "had it not been for the Lord who was on
our side" (Ps. cxxiv 1, 2) in a passage that betrays strong Mishnaic
influence. 3

Mandaic (Brock., 428b). The hypothetical character is in these, as in SH (below

under 2), expressed by the tense. In colloquial Arabic lau can, on the contrary, be
employed with real conditions (Brock., 426b).
2 ,.".,~ is not followed by a nominal clause, as Segal states, but the ~., contained
in the conjunction together with the following noun constitutes an existential one-
member sentence (Brock., 434a), in which the noun stands in the nominative, as
in the Arabic laula with the nominative, e.g. laulii anta "were it not for thee" (cf.
Reek., 264.2).
3 Further instances may exist in Is. i 9, Ps. cvi 23, cxix 92. In each of these the
 CONDITIONAL CLAUSES 181

Examples: Cl'pnr.:l "W1n n'n ~, ni'~n "" "Were it not for form,
the hyle would not continue its existence" (B. J:liyya, Hegyon, f. 2);
n':lpn "" ... i1r Cl'l1'" ':J"n ~, "we would not have known this ...
were it not for the tradition" (B. Ezra, <p~oth, f. 6b); nO:J~ 'iD:J~ ',?"
':J'n'~ m,n n':l~ i:l~... n"':In "had it not been for the men of
the Great Assembly, the Law of our Lord would have been lost"
(id., Yesod Mora, p. 9); nr.:l n"n ~'nn n"':In ni'r.:liDn "" "had it not
been for that thorough care, you would have died" (J. Tibbon,
Testament, p. 10).
f. Cl~' is employed in the sense of "although, even if" (as in Arabic
wa-in): nr.:l 'Em Cl~' n'n1 nr.:l'p Cl'~n nr.:liD:J "the soul of man remains
alive, although his body dies" (B. J:liyya, Megillah, p. 58). In this
function also iD Cl~' could be employed: 4 li'~r.:l ~'niD Cl~' pr.:ln nr.:l~n
Cl"l1n "mathematics, although it is a wordly necessity ... " (B. J:liyya,
Meshi~ah, p. 2); for further examples from the second period (with-
out she-) see Ben-Yehudah Thesaurus, p. 263. Although Cl~ could be
employed in the sense of "although" in BH (Num. xxii 18; I Kings
xiii 8, etc.,), the SH idiom seems to be a direct loan from the Arabic.
g. Nominal clauses in the protasis (Ges., 159p) are extremely rare.
Interesting in this connection is the following instance: n'?r.:ln Cl~
1~P nn~ n'n' "11'r.:l n'n Cl~' 1~P fr.:lP:l n'n' l1i'r.:l "if the word has
ultimate stress, it will have a Zere, but if it has penultimate stress,
there will be a Segol" (B. Ezra, <.a~oth, f. 6a). The first sentence has
a nominal protasis, but in the coordinated second sentence the author
changes to the more convenient verbal clause. The reason for this
preference may lie in Arabic speech-habit, where a nominal clause
after in is impossible (Reck., 255.2).
h. Not enough material has been collected on the tense system of
the conditional complex. The usage of authors seems to vary con-
siderably. B. Ezra prefers perfect in the protasis, imperfect in the
apodosis: Cl'l1:J m iD:I,n nnn 'r.:l'iD' 1'i:ln m'm~r.:l 1'iD~in m~n n'n Cl~

words following on ,."., may be construed either as nominal clauses (nominal clause
after ,."., is found in Ps. xciv I 7) or as nouns in the nominative with attributive
relative clauses (in the last case with apposition). The Targum chooses the latter
way in each case, but this proves nothing, since the linguistic background of the
translator was MH. In view of the fact that in all other cases ,."., is followed by
a full clause, it appears preferable to assume that in these verses, too, nominal
clauses are intended.
4 Cr. the Talm. Aramaic ... , '~ "if" (Schlesinger, p. 273).
182 CHAPTER 1WELVE

"if the first radical is a laryngal, the Dagesh is replaced by a long

vowel" (Saphah Berurah, f. 7a). B. ~iyya, on the other hand, has either
both clauses in the perfect or both in the imperfect: rQ~Q i1"~Pi1 i1'i1 C~
i1ri1 C'nll~ ,Q'l) C'~i1 i1'i1 . . . 1'QQi1 m~i1 n~ "if God would reinforce
the digestive faculty, man would remain forever in this world" (Megillah,
p. 50); 1Q '1ElJ iD~i1 i1'i1' ... n1~':1 i1~ ,n'nQ 1iD~ i1QiDJi1 i1'i1n C~
n'Qi1~i1 "if the individualizing soul in it became stronger, the fire
would separate itself from the beasts (id. Hegyon, f. 11). In hypo-
thetical conditions, the compound tenses (cf. par. 13b) are frequent,
cf. examples above under (d).
i. Normally the protasis precedes the apodosis, but occasionally it
precedes, seemingly without any syntactical reasons (cf. the rules for
MH in Nodel, op. cit., 4lc). E.g., the first quotation under (d) and:
i1~ 1~iDJ C~ ,,'?~ ml)'JQ ,'i1 ~'? mQ'~l) mn'?~i1 ,,? 'l)'Ji1 "he achieved
great successes which he would not have obtained had he remained
there" (Gers., De'oth, f. 8b). The position is in both cases due to sty-
listic considerations.
k. SH either developed independently or, more probably, borrowed
from Arabic the possibility of having only a very loose connection
between protasis and apodosis, so that the truth of the apodosis could
not possibly be taken as dependent on the truth of the protasis
("Verschiebung", cf. Reek., 261.1). These are frequent in Arabic,
and also occur in Syriac and Mandaic, but never in the older Hebrew
dialects (Brock., 430). Examples: pn' ~'? m i1'i1 'n~~ 1n~ '~ C'1m~i11
"and if anyone says that this happened after he had come out (he
is wrong, because) that is not possible" (B. Ezra, Yesod Mora, p. 9);
1'~J 'J'~ m'?:Ii1 Pl) '?l) m iD1ElQi11 "if anyone explains this as referring
to the exile, that is not correct" (D. J>imJ:Ii, Hosea Comm., p. 17).
 SYNOPTIC TABLE

(The dates indicated are those of death. Many of these are only approximate. Dates of particular uncertainty are marked
by a ° before the name of the person.)

C.E. The East Spain Provence N.W.-Europe Italy

940 Saadiah Gaon

950
960
970 Mena}:lem b. Saru~
980 Dunash b. Labrat Donnolo ~
o
990 ~
1000 Cl
1010 oB. Chiquitilla
oB. Abitur ~
1020 °Moses Darshan ~
1030 I:Iai Gaon Gershom b. Judah
1040
1050 Samuel Nagid
B. Gabirol o A\:limaa~ b. Pal tiel
1060
1070
1080 B. Ghayyath
1090 Isaac b. Reuben °Mena}:lem I:Ielbo
Isaac Albalia
co
w
(Table cont.)
..,.co
C.E. The East Spain Provence N.W.-Europe Italy

1100 Nathan b. YeJ:!iel (111,1))

1110 Rashi
1120
1120
1130 Moses b. Ezra Abraham b. Crusade Chroniclers
1140 Solomon b. S~bel I:Iiyya JeraJ:!meel
1150 MenaJ:!em b. Solomon
1160 'judah Hadassi Abraham b. Ezra 00

1170 Joseph ~mJ:!i Jacob Tarn; Rashbam °Solomon Parhon ~

0
1180 Abraham b. Da'ud
°Benjamin of Tudela ~
("]
1190 Moses ~mJ:!i
J udah Tibbon ~
1200 PethaJ:!iah ~
Maimonides
1210 J udah I:Iasid
1220 Josef b. '~nin
°Ibn Zabara
1230 al-I:Iarizi
°Isaac of Catalonia Samuel Tibbon
David ~mJ:!i
1240 Abraham Maimon Eleazar Worms
Meir Abulafia
(Table cont.)

C.E. The East Spain Provence N.W.-Europe Italy

1250 oB. Shabbethai

°Abraham b. I:Iisdai °Abraham
Bedershi
1260 'judah b. 'Abbas 'joseph Ezobi
1270 Tanl:mm Jerush. Nal)manides 'jacob Anatoli
°Moses of Salerno
1280 Moses Tibbon
Isaac b. La!:if [/0
1290 'jeJ:!iel b. Je~utiel
Abraham Abulafia ~
0
°Shemtob Falaguera °Moses Tako °ZeraJ:!iah Ben >tl
...,
Hillel of Verona 0
1300 °Todros Abulafia
Josef b. Chiquitilla ~
t"'
Moses de Leon t"fj

1310 Solomon Adret (Rashba) °Gershom b. Sol.

°Isaac Aboab
°Ibn Sahula Levi b. Abraham
°Estori ha-ParJ:!i
1320 Aaron b. Jos. °Isaac Israeli
Asher b. JeJ:!iel
MenaJ:!em Recanati
Benjamin b. Judah
1330 °Abba Mari °Immanuel Romi
J>.alonymos b. J>.alon (Xl
(Jl
(Table cont.) co
O'l
C.E. The East Spain Provence N.W.-Europe Italy

1340 Isaac I:Ialawa Joseph Kaspi

°Isaac I:Ielo
Jedaiah Bedershi
Gersonides
1350 ':Joseph NaJ:!mias
°Isaac Pulqar
1360 °Moses b. Judah Moses N arboni
1370 Aaron b. Elijah en
1380 Nissim Gerondi °Isaac de Lates ~
0
°Samuel Zaqah '1:1
MenaJ:!em b. ZeraJ:! ::l
()
1390 ':Jacob Zarphati --l
1400 ~
1410 I:Iisdai Crescas ~
1420 ZeraJ:!iah Halevi
°Solomon Alammi
°Prophiat Duran
Moses Botarel
1430
1440 Joseph Albo
Simon Duran
1450 Yomtob Miihlhausen
(Table cont.)

C.E. The East Spain Provence N.W.-Europe Italy

1460 °I:Iayyim b. Musa Moses da Rieti

1470
1480
1490 Shemtob b. Shemtob Messer Leon
1500 Saadiah b. Danon
Isaac Abarbanel
1510 Abraham Zaccuto
1520 °Solomon b. Verga
~
o
::J
8
~
~

00
'-l
 BIBLIOGRAPHY

(a) Sources

Abraham Maimonides, Letter on the Moreh Nebukhim, in the Jlobe;;. Teshuboth ha-Rambam,
ed. Lichtenberg, Leipzig 1859, part iii, fols. 15a-2Ia.
B. 'Abbas (Judah b. Samuel), Ya'ir Nethib, Bodleian MS., Neubauer 1280.
i\I:limaa~, The Chronicle if Ahimaaz, ed. M. Salzman. Columbia University Oriental Studies,
no. xviii. New York 1924.
B. Da'ud (Abraham), Sepher ha-Jlabbalah, ed. A Neubauer in Mediaeval Jewish Chronicles
i (Oxford 1887), pp. 47-84.
B. Ezra (Abraham), Saphah Berurah, ed. G.H. Lippmann, Fiirth 1839.
Sepher ?,aI,loth, ed. G.H. Lippmann, Fiirth 1827.
Sepher ha-Shem, ed. G.H. Lippmann, Fiirth 1834.
Yesod Mora, ed. S. Waxman, Jerusalem 1938.
, Commentaries on the Bible, quoted from the Warsaw Mif:rra'oth Gedoloth.
B. I:Iiyya (Abraham), Megillath ha-megalleh, ed. A Poznanski and M. Guttmann, Berlin
1924.
Hegyon ha-nephesh, ed. B. Freimann, Leipzig 1860.
Ijibbur ha-meshi~ah weha-tishboreth, ed. M. Guttmann, Berlin 1912-13.
Sepher ha-'ibbur, ed. H. Filipowski, London 1851.
, Letter on astrology, ed. AZ. Schwarz in the Festschrift fUr Adolf Schwarz,
Berlin 1917, pp. 23-36.
B. Parl).on, MaI,lbereth he-'aiukh, ed. S.G. Stern, Pressburg 1844.
B. Verga (Solomon b. Judah), iii1ii' O~iD, ed. M. Wiener, Hanover 1855-6.
B. Z:ar~ah (Samuel), Mikhlol Yophi, Bodleian MS., Neubauer 1296.
Crescas (I:Iisdai), Or Adonai, Vienna 1859.
David l}.iml).i, Hosea-Commentary, ed. H. Cohen, Columbia University Oriental Studies, no.
xx., New York, 1929.
- - , Commentary on Psalms r-xli, ed. S.M. Schiller-Szinessy, Cambridge 1883.
- - , Mikhlol, ed. I. Rittenberg, Lyck 1842.
al-Fakhir (or al-Fakhkhar, Judah b. Joseph), Letter in Jlobe;;. Teshuboth ha-Rambam, 111
Ib-5b.
Falaq. (Shemtob b. Falaquera or Palquiera), Sepher ha-Meba#esh, ed. M. Tama, The
Hague 1779.
- - , Iggereth ha-wikkual,l asher ben ha-torah weha-~okhmah, Prague 1610.
- - , Iggereth ha-musar, ed. AM. Habermann, Jlobe;;. tal yad no. xi., Jerusalem 1936.
Gerondi (Nissim b. Reuben), Sermons, Venice 1596.
Gersonides (Levi b. Gershom), Mil~amoth Adonai, Riva di Trento 1560.
, Sepher ha-de'oth weha-middoth, ed. S. Mahariah, Warsaw 1865.
- - , Pentateuch-Commentary, Venice 1547.
- - , Commentary on Prav. and Job, in Mi~ra'oth Gedoloth, Warsaw.
I:Iayyim b. Musa, Letter to his son, ed. D. Kaufmann, Beth Talmud ii (1882) 117-25.
Joseph l}.iml).i, Sepher Zikkaron, ed. W. Bacher, Berlin 1888.
- - , Sepher ha-Galuy, ed. HJ. Mathews, Berlin 1887.
Joseph Nal).mias, Commentary on Aboth, ed. M.L. Bamberger, Berlin 1907.
Judah b. Tibbon, Testament, ed. M. Steinschneider, Berlin 1852.
Kreuzziige, Hebriiische Berichte uber die Judenveifolgungen wiihrend der Kreu;::zuge, ed.
A. Neubauer and M. Stern, Berlin 1892.
 BIBLIOGRAPHY 189

Maimonides (Moses b. Maimon), Mishneh Torah, Berlin 1880 (quoted by title of

section, chapter and paragraph, without title of work).
~-, Letters, in Jlobe;;. etc., Leipzig 1859, pt. ii.
~-, Responsa, ed. AH. Freimann, Jerusalem 1934.
Moses b. Tibbon, Commentary on Canticles, Lyck 1874.
Rashi (Solomon b. Isaac) Commentary on the Pentateuch, ed. A Berliner, Frankfurt a.M.
1905.
~-, Commentaries on Bible and Bab. Talmud quoted from Avinery, Hekhal Rashi
(cf. bibliography "b").
Saadiah (b. Joseph, Gaon), Fragments from the Egron and the Sepher ha-Galuy, ed.
A Harkavy, Studien und Mittheilungen v., St. Petersburg 1891.
~-, Fragments from Sepher ha-Galuy, ed. S. Schechter, "Saadyana", JQR (O.S.)
xiv (1901) 37-63.
~-, Fragments inedits du Sifer Haggaloui, ed. B. Chapira, RE] Ixviii (1914) 1-14.
~-, Polemic against f:liwi al-Balkhi, ed. j. Davidson, New York 1915.
Saadiah Danon (or Danan), Treatises, Bodleian MS, Neubauer 2233.
Samuel b. Tibbon, Commentary on Ecclesiastes, Bodlcian MS., Neubauer 132.
~-, Preface to the translation of Maimonides' Guide qf the Perplexed, quoted from
the print Berlin 1875.
Sepher f:lasidim (by Judah b. Samuel l:Iasid), ed. by j. Wistinetzki, Berlin 1891-3.
Singer, S., Authorised dairy prqyer book, annotated edition by j. Abrahams, London 1914.

(b) Special Bibliography

Works dealing with the history and grammar qf post-Mishnaic Hebrew.
Albrecht, K., Zum Lexikon und zur Grammatik des Neuhebraischen. ZA W xix
(1899) 135-55, 310-28).
(A fairly complete list of deviations from Biblical Hebrew in vocabulary and grammar
found in the Tarshish, a collection of poetry by Moses Ibn Ezra. The provenience
of these features is not investigated, and no distinction made between Mishnaic
or Paytanic material and innovations of the Spanish poets or Ibn Ezra himself.)
Avinery, j., '''iDi '?:J'i1. Encyclopaedia in five volumes, containing alphabetically all
that Rashi created in the field of language and exegesis. Vo!. i: New words and
expressions. Tel-Aviv 1940.
(The "dictionary of usages", pp. 197-350, contains a large number of remarks
on prepositions, conjunctions, adverbs, and pronouns, with careful documenta-
tions from Biblical and Talmudic literature.)
Bacher, W., Ueber den sprachlichen Charakter des Maimuni'schen Mischne-Tora.
In: Aus dem Wiirterbuche Tauchum Jeruschalmi's, Strassburg 1903.
~-, Zum sprachlichen Charakter des Mischne Thora. In: Moses ben Maimon, ed.
Bacher, etc., Leipzig 1914, ii 280 ff.
(Deal mainly with vocabulary and idiom, without distinguishing between the
innovations of Maimonides' predecessors and those of his own.)
Bergstrasser, G., Eirifiihrung in die semitischen Sprachen, Munich 1928.
(Short note on pp. 46 and 47 states that mediaeval Hebrew, which was one
of the most slovenly of all languages, was much improved by the efforts of the
Spanish period.)
Cassuto, U., Article "Epoca postbiblica (Neoebraica)" in the Enciclopedia Italiana xiii
(1932) 357-8.
Goldenthal, j., Grundziige und Beitrage zu einem sprachvergleichenden rabbiniseh-
philosophise hen Worterbuche, in: Denkschriften der kais. Akad. der Wiss. in Wien, i
(1850) 419-53.
190 BIBLIOGRAPHY

Kenaani, j., i''?Pi1 '?iD miD'? i~1~O, Leshonenu x (1938) 21-29.

- - , iiElO 'ii1iDO '?ll Cl'Oi1Pi1 Cl'jt!l'E1i1 '?iD n'j1iD'?i1 ClnllEliDi1, Leshonenu x (1938)
173-82.
- - , t!l1'E1'? 'j1'~jii1pj1P ]1'?0, Prospectus, Tel-Aviv 1930.
Kroner, H., Zur Terminologie der arabischen Medizin und zu ihrem zeitgeniissischen hebrai-
schen Ausdrucke, Berlin 1921.
Landau, j., Geist und Sprache der Hebraer nach dem zweyten T empelbau, Prague 1822.
(A very superficial account of Mishnaic Hebrew.)
Levias, D., Article "Hebrew language" in the Jewish Encyclopaedia vi (1904) 306-10.
(On mediaeval Hebrew pp. 308-309.)
Lipschiitz, E.M., l1iD'?i1 m'?~iD", in Sephathenu i (Odessa 1917) 17--42.
(pp. 17-25 on the influence of Aramaic on mediaeval Hebrew.)
- - , Vom lebendigen Hebraisch. Berlin 1920.
Luzzatto, S.D., Prolegomeni ad una grammatica regionata della lingua ebraica, Padova 1836.
(p. 103: "Mediaeval Hebrew ... does not belong to the province of the study
of the Hebrew language, but to the critical study of rabbinical writings". Divides
Rabbinical Hebrew into six distinct styles that all coexisted in time. No mention
of any historical development.)
Metmann, L., Die hebraische Sprache. Ihre geschichtliche und lexikalische Entwicklung
seit Abschluss des Kanons. Berne Diss. , Jerusalem (1904).
(Grammatical and syntactical matters are hardly touched upon, except for
Modern Hebrew. On mediaeval Hebrew pp. 40-63, of which pp. 50-63 a list
of words from Piyyur, Poetry, Original Hebrew literature of all types and peri-
ods, and translations, with many inexactitudes. Metmann claims that the gener-
ally employed popular style (in which works?) was that of the historical books of
the Bible, which however approached more closely the style of the Mishnah than
that of the Bible. The exegetes (Ibn Ezra, Simeon ~ara and B. Helbo!) used a
pure Biblical style which avoids every Mishnaic expression. The third style was
that of the translators and philosophers, distinguished by idioms that were nei-
ther Biblical nor Mishnaic, but Arabic.)
Rabin, C., Saadya Gaon's Hebrew prose style, in Essays on Saa4Jia, ed. E. Rosenthal,
Manchester 1943.
(A list of post-Biblical usages in the /jiwi Polemic, the Egron and the Sepher ha-
Galuy, with parallels from Spanish-Proven~al Hebrew.)
Renan, E., Histoire generate des langues semitiques, 5th ed., Paris 1878.
On mediaeval Hebrew pp. 164-7. A number of pertinent remarks on the char-
acter of the language. Renan objects to the creation of new words in the Spanish
period. Nevertheless, he hardly deserved the bitter attack of M. Griinbaum, "Renan
iiber die spateren Formen der hebraischen Sprache", in: Semitic Studies in Memory
qf A. Kohut, Berlin 1897, pp. 226-34.)
Schwarz, A., Der Mischneh Thorah, Karlsruhe 1905.
(On pp. 74-94 the language of Maimonides is considered, mainly from an aes-
thetic point of view.)
Sideman A.I., Cl":JOi'? i1i1n nJiDO j1jjO, Sinai VI. 428-38 VII. 96-101.
Schirmann, j., Die hebraische Uebersetzung der Maqamen des Hariri. Schriflen der
Geschichte for die Wissenschrift des Judentums, no. 37, Frankfurt a/M. 1930.
(Deals with Arabisms and post-Biblical elements in al-Harizi's style.)
Steinschneider, M., Die hebraischen Uebersetzungen des Mittelalters, Berlin 1893.
(Many remarks on linguistic matters, and valuable lexicographical collections
in the index of Hebrew words, pp. 1036-45.)
Szneider, M., n'miElOi1 n'i:Jlli1 l1iD'?i1, Leshonenu vi (1934) 1933-5, 301-26, vii (1935)
52-73.
(Advances the theory that the liturgical language was a direct development of
Late Biblical Hebrew, and that Piyyiit. N.W.-European Hebrew, Spanish-Proven~al
Hebrew, and the Karaite idiom all derived from it.)
 BIBLIOGRAPHY 191

- - , ilr11nnElr1i1:J rilhil n1ln, vo!. i, Wilna 1923, vo!. ii, Tel-Aviv 1939.
(See above in preface, p. vii).
- - , (O'OPOlQ) lliD?il iDli.JiD n1ln ;:JiD '?p ,ilr11nnElr1i1:J lliD'?il n-lln, 1939 il:J'?'l. The
~1:Ji.J, pp. 1-38, deals with history, particularly with the origins of MH (MH was
brought in by immigrants from Edom during the first exile; the il'?l,j nOl:J 'iDl~
wrote in BH the post-exilic books-which were for the 1:J''?':JiDi.J-and in MH the
oral law for the people. Reaffirms theory of n'r11'ElO lliD'?, fights against the 1'?i1i.J
iD,n (pp. 33-5) (for preferring MH). The syntax deals with the supposed J1iD'?
n'r11'ElO, not with modern Hebrew, no examples from modern authors.
Torczyner, H., Article "Hebraische Sprache" in the Encyclopaedia Judaica, vii 1031-66.
(Devotes to mediaeval Hebrew columns 1061-2, only a few general remarks.)
Yellin, D., lO'El 'In:J '?,,':JJ 1:J i1i.J'?iD, Leshonenu vii (1935) 219-45.
(Some additional material on the paytanic language.)
Wijnkoop, J.D., The Neo-Hebraic language and its literature, JQ.R O.S. xv (1902)
23-55.
(Points out the importance of the study of post-Biblical Hebrew for the under-
standing of the Bible. Stresses mainly similarities between Biblical, Mishnaic, and
mediaeval Hebrew, and gives a short sketch of Mishnaic Hebrew from this point
of view. Post-Mishnaic features are mentioned only in passing.)
Zulay, M., l:J'lO'Elil J1iD'? '?iD ilr11i.J''?, Melila, Manchester, 1944, pp. 69-80.
, "j' 'Ol'El:J lliD'? 'll'~: Mitteilungen des Instituts for EifOrschung hebriiischer Dichlung vi.
- - , in J1iD'? 'l'lll, ed. H. Yalon, vols. J"iDn-YiDn.
Zunz, L., Die synagogale Poesie der Juden des Mittelalters, Berlin 1855-59.
(Cr. above p. 11, note 2)

(c) General Bibliography

This bibliography includes all works from which one or more quotations occur in
the preceding. It does not include books or articles that are only named as titles,
or articles in the Jewish Encyclopedia and the Encyclopaedia Judaica. See also
Special Bibliography (b).*
Aaron b. Joseph, 1:J"iD' ,n:Ji.J, Eupatoria 1834.
Abraham Gavison, i1n:JiDi1 'i.Jll, Livorno 1748.
Adlerblum, N.H., A stu4J if Gersonides in his proper perspective, New York 1926.
Albrecht, K., Die Wortstellung im hebraischem Nominalsatz, ZAW vii (1887) 218-23,
viii (1888) 248-63.
Assaf, S., Gaonica, Jerusalem 1936.
- - , '?"1 'iJ1i'j Cl'Ol " n~i.J 1:J':Jn:Ji.J illl:J'~, Horeb iii (1936) 93-100.
Avinery, I., n":Jllil '?ll n'i.J'~il nllEliDil, Leshonenu iii (1931) 273-90.
Bacher, W., Die Bibelexegese Moses Maimuni's, Strassburg 1897.
- - , Der "PrUfstein" des Menachem b. Salomo, in the Graetz-Jubelschrifl, Breslau
1887, pp. 94-115.
Baron, S., Social and religious history cif the Jews, 3 vols., New York 1937.
Baumann, F.V., Hebriiische Relativsiitze, Diss. Leipzig 1894.
Bauer, H. and Leander, P.*, Historische Grammatik der hebriiischen Sprache des Alten
Testamentes, vo!. i, Halle 1918-22.
Ben-Yehudah, E., Thesaurus totius hebraitatis, vo!. i-ix, Jerusalem 1910-30.
- - , Prolegomena, Jerusalem 1940.
Bergstrasser, G., Das hebraische Prafix iD, ZAW xxix (1909) 40-56.
Bernstein, S., 'l"n'?~ ililil' ",? r11iDin "r11i.J~pi.J", Horeb i (1934) 179-87.

* Note: Books marked * are quoted by paragraphs.

192 BIBLIOGRAPHY

, n~'j '1 pn~' 'I' !:l'iV,n !:l't:l1'El, Tarbiz xi (1939/40) 295-325.

Boneh, A., n':l':ln1 r,~n, Tel-Aviv 1938.
Brockelmann, C.*, Grundriss der vergleichenden Grammatik der semitischen Sprachen, 2 vols.,
Berlin 1898-1908.
(All quotations without volume-number refer to volume ii.)
Brody, H. and Wiener, M., Anthologia Hebraica, Leipzig 1922.
Budie, O.R.M., Die hebriiische Priiposition 'ai, Halle 1882.
Charles, R.H., "The Apocrypha and Pseudepigrapha of the Old Testament, 2 vols., Oxford 1913.
---, A critical and exegetical commentary on the book of Daniel, Oxford 1929.
Coplas de Yoc,:ef, ed. j. Gonzalez Ilubera, Cambridge 1935.
Cowley, A.E., A concise catalogue of the Hebrew printed books in the Bodleian Library, Oxford
1929.
Cowley, E., 'jEln p '~1aiV
Petersburg 1908, pp. 160-3.
'I' '1a,n ,~ ,,~ ,j,a ,~, in Harkavy-Festschrift, St.

Dalman, G., Grammatik des jiidisch-paliistinischen Aramiiisch, 2nd ed., Leipzig 1905.
Davidson, j., Saadia's Polemic against Ijiwi al-Balkhi, New York 1915.
, nn::Jna ,Eloa ",iV, A.Z. Rabbinowitz Anniversary volume, Tel-Aviv 1924,
pp. 83-101.
- - , !:l't:l1'Eln nmn" Madda'e ha-Yahaduth i (Jerusalem 1926) 187-95.
Delitzsch, F., Zur Geschichte der jiidischen Poiisie vom Abschluss der heiligen SchriJten alten
Bundes bis auj die neueste Zeit, Leipzig 1836.
Dhorme, P., L'ancien hebreu dans la vie courante, Revue Biblique xxxix (1930) 62-73.
Dotan, A., ?,~,p n'n 'iV~-p !:lja~n, Sinai 20 (1957) 280-312, 350-62.
Dozy, R., Supplement aux dictionnaires arabes, Leiden 1881.
Driver, G.R., Problems of the Hebrew verbal system, Edinburgh 1936.
Driver, S.R., Introduction to the literature of the Old Testament, 8th ed., Edinburgh 1909.
- - , A treatise on the use of the tenses in Hebrew, 3rd ed., Oxford 1892.
Dubnow, S.M., Weltgeschichte des jiidischen Volkes, German translation, 10 vols., Berlin
1927.
Dukes, L., Ehrensiiulen und Denksteine zu einem kiirifiigen Pantheon hebriiischer Dichter und
Dichtungen, Vienna 1837.
Duval, R., La litterature syriaque. [Anciennes litteratures chretiennes ii], Paris 1900.
Elbogen, j., Der jiidische Gottesdienst in seiner geschichtlichen Entwicklung, Leipzig 1913.
- - , Abraham b. Daud als Geschichtschreiber, in j. Guttmann-Festschrift, Leipzig
1915, pp. 186-205.
Epstein, A., ,::J::J m::J'iV'1 !:l'j1~jn m,1p' m,1pa, in Harkavy-Festschrift, St. Petersburg
1908, pp. 164-74.
- - , !:l"1n'n m'j1a'pa, vo!. i, Vienna 1887.
Ewald, H., Ausfiihrliches Lehrbuch der hebriiischen Sprache, 8th ed., Gottingen 1870.
Frankl, P., Karaische Studien, neue Folge, MonatsschriJt der Geschichte}iir die WissenschaJt
des Judentums xxxi (1882) 1-13, 72-85, 268-75.
- - , Article "Karaer" in Ersch und Gruber's Realencyclopaedie, Leipzig 1818-40, Section
ii, vo!. xxxiii.
Friedlander, j., Der Sprachgebrauch des Maimonides, ein lexikalischer und grammatischer Beitrag
zur Kenntniss des Mittelarabischen, pt. i, Frankfurt a.M. 1902.
- - , Die arabische Sprache des Maimonides, in Moses ben Maimon i, Leipzig 1908,
pp. 421-38.
Gardiner, A.H.*, Egyptian grammar, Oxford 1927.
Geiger, A., Lehr- und Lesebuch zur Sprache der Mishhna, Breslau 1845.
Gesenius' Hebrew grammar as edited by E. Kautzsch*, second English edition by A.E.
Cowley, Oxford 1910.
Gesenius, W., Hebriiisches und Aramiiisches Handwiirterbuch, bearbeitet von F. Buhl, 16th
ed., Leipzig 1915.
Giesebrecht, F., Die hebriiische Priiposition Lamed, Halle 1876.
 BIBLIOGRAPHY 193

Ginsburger, M., L'exegese biblique des juifs d'Allemagne au moyen age, HUCA vii
(1930) 439-56.
Goldziher,J., Bemerkungen zur neuhebraischen Poesie,JQR (O.S.) xiv (1902) 719-36.
Gross, S., Menahem ben Saruq, ein Beitrag zur Geschichte tier hebrdischen GrammatiJc, Breslau 1872.
Hadassi (Judah), i5:l::lil '?::liD~, Eupatoria 1836.
Harkavy, A., Studien und Mittheilungen aus der Kaiserlichen offentlichen Bibliothek
zu St. Petersburg v. I: Leben und Werke des Saa4Jah Gaon, St. Petersburg 189 I.
- - , Fragment einer Apologie des Maimonidischen Cl'11ail 11'n11 ia~a, Zeitschrififiir
hebrtiische Bibliographie ii (I897) 125-8.
Harris, Z.S., A grammar if the Phoenician language. American Oriental Series, no. 8,
New Haven 1936.
Horowitz, C., Halachische Schriften der Geonim, pt. 2, Frankfurt a.M. 1881.
Jastrow, M., Dictionary if the Targumim, etc., Authorized ed., New York 1925.
Jespersen, 0.*, Ana!J!tic 5iJntax, London 1937.
Josippon, Ad fidem Mantuae editionis ed. denuo D. Giinzburg, Berdischev 1896- I 9 I 3.
Judah b. Bolat, i;;p '?'?::l, Constantinople (1530).
Kahan, J., Ueber die verbal-nominale Doppelnatur der hebraischen Participien und
Infinitive, Diss. Leipzig 1889.
Karpeles, G., Geschichte der jiidischen Literatur, 2 vols., Berlin 1886.
Kautzsch, E., Aramaismen im Alten Testament, i. Lexikalischer Teil, Halle 1902.
Klatzkin, J., 11'i:::lJ)il il'5:l101'?'5:lil '?iD il'J1'?1m~, Berlin 1926.
Klausner, J., iliDinil mi5:l0:::l miDail pJO '?iD 1m'?iD'?11iDil'?, Madda'e ha-Yahaduth i (1926)
163-78.
- - , iliDinil 11'i:::lJ)il mi5:l0il '?iD il'i1~Oil, vo!. i, Jerusalem 1930.
Koch, A., Der semitische lrifinitiv, eine sprachwissenschojiliche Untersuchung, Stuttgart 1874.
Koenig, E., Historisch-comparative hebrdische 5iJntax, Leipzig 1897.
Kosovsky, HJ., miDail ]liD'? i;;1~, 2 vols., Jerusalem 1914-27.
Krauss, S., Griechische und lateinische Lehnwiirter im Talmud, Midrasch und Targum, Berlin
1898-9.
Kropat, E., Die 5iJntax des Autors der Chronik mit der seiner Qy.ellen verglichen. Beihefte zur
ZAW no. 16, Giessen 1909.
Lerch, E., Historische jranZiisiche 5iJntax. Leipzig 1925-34.
Levy, L., Reconstruction des Commentars Ibn Ezra's zu den ersten-Propheten, Diss.
Heidelberg 1903.
Lichtenstein, Z., ii5:l0 iD1iJ iJ) :)iJ)a:::J 11nil'il 'T::lia 11n'a;;a '?~iiD' 'a' 'i:::Ji:::J Cl'i1J)iD,
Tel-Aviv, 1939.
Liebermann, S., ed., ~:::li Cl'i:::li iDiia, Jerusalem 1940.
Malter, H., Saadia Studies, JQR iii (1913) 487-509.
- - , Saadia Gaon, his lift and works, Philadelphia 192 I.
Mann, J., TIe Jews in Egypt and Palestine under the Fatimid Caliphs, 2 vols., Oxford
1920-22.
- - , TIe Bible as read and preached in the old 5iJnagogue, vo!. i, Cincinnati 1940.
Mannes, S., Ueber den Einfluss des Aramaischen auf den Wortschatz der Misnah,
pt. i, ~-a, Diss. Erlangen 1897.
Margolis, M.L., A manual if the Aramaic language if the Babylonian Talmud. Clavis lin-
guarum semiticarum pars iii, Munich 19 I O.
Marmorstein, A., The place of Maimonides' Mishneh Torah in the history and
development of the Halachah, in Moses Maimonides: Anglo-Jewish papers, ed. J. Epstein
London 1935, pp. 159-74.
Matmon-Cohen, J., '?:::l:::l m'?J imli 11'i:::lJ)il, Leshonenu vi (1934) 172-88, vii (I 935)
136-44, 257-65, 349-55, viii (1936) 15-22, 123-30, ix (1937) 65-75.
Molin, C., On prepositionen min i bibelhebraiskan, Diss. Upsala 1893.
Moses b. Ezra, Kitab al-mudhakara wal-mu\:iaqara, ch. i-iv, ed. by P.K. Kokovtsov
in Vostochnrya Zametki, St. Petersburg 1895, pp. 193-200.
194 BIBLIOGRAPHY

, ?~iiD' ni'iD, trans!' by B.-Z. Halpern, Leipzig 1924.

Moses b. Maimon, i11~r:li1 i~O, ed. M. Bloch, Paris 1888.
Moses b. Maimon, Das Buch der Gesetze ... nebst der hebraischen Uebersetzung
des Schelomoh benJoseph ibn Ajub (sic.), ed. M. Peritz, pt. i, Diss. Leipzig 1881.
Neubauer, A, Catalogue qf the Hebrew manuscripts in the Bodleian Library, Oxford 1886.
- - , Documents in edits xviii: Nouvelles collections de consultations casuistiques,
RE] xii (1886) 80-94.
Nbldeke, T., Die semitischen Sprachen, 2nd ed., Leipzig 1899.
--*, Compendious .'iYriac Grammar, London 1904.
- - , Grammatik der neusyrischen Sprache am Urmia-See und in Kurdistan, Leipzig 1868.
- - , Zur Grammatik des classischen Arabisch. Denkschriften der K. Akad. d. Wiss. in
Wien, Phi!.-hist. C!. Vo!. xlv, Abh. 2, Vienna 1897.
Nodel, j.*, Der ;;.usammengesetzte Sat;;. im Neuhebraischen, Frankfurt a.M. 1928.
Pinsker, S., i11'J1r:l'P 'Cl1P?, Vienna 1860.
Pir~e R. Eliezer, ed. j.E. Stand, Lemberg 1870.
Poznanski, SA., 1~1i'p 'iDJ~, in Harkavy Festschrift, St. Petersburg 1908, pp. 175-220.
Reckendorf, H.*, Die syntaktischen Verhaltnisse des Arabischen, Leiden 1895-8.
- - , Arabische .'iYntax, Heidelberg 1921.
- - , Ueber Paronomasie in den semitischen Sprachen, Giessen 1909.
Renan, H. and Neubauer, A, Les rabbins fran-;ais du 14me siecle. Histoire litteraire
de la France, vols. xxvii (Paris 1877) and xxxi (1893).
Rosenmann, M., Das Lehrhaus des R. Nissim Gerondi in Barcelona, Schwarz-
Festschrift, Berlin 1917, pp. 489-98.
Roth, C., A short history qf the Jewish people, London 1936.
Rowley, H.H., The Aramaic qf the Old Testament, Oxford 1929.
Sachs, H., Die Partikeln der Mischna, Diss. Giessen 1897.
Sachs, M., Die religiose Poesie der Juden in Spanien, Berlin 1845.
Schechter, S., Geniza Specimens. The oldest collection of Bible difficulties by aJew,
J(LR (O.S.) xiii (1901) 345-74.
Schirmann, j., Die hebraische Uebersetzung der Maqamen des Hariri. Schri.ften der
Geschichte flir die Wissenschafl des Judentums, no. 37, Frankfurt a.M. 1930.
Schlesinger, M., .'iYntax der aramaischen Sprache des babylonischen Talmuds. Verbffentlichungen
der A Kohut-Stiftung, vo!. I, Leipzig 1928.
Segal, M., Misnaic Hebrew and its relation to Biblical Hebrew and to Aramaic,
J(LR (O.S.) xx (1909) 647-737.
- - , Hebrew in the period of the Second Temple, International Journal qf Apocrypha,
1910, 79-82.
- - , i1'nn?1i11 i1~~1r:l miDr:li1 riD?, Madda'e ha-Yahaduth i (1926) 30-44.
--*, A grammar qf Mishnaic Hebrew, Oxford 1927.
--*, i1JiDr:li1 PiD? pnp', Tel-Aviv 1936.
- - , ~i'O P ?iD miD?, Leshonenu vii (1935) 100-120.
Sellin, E., Die verbal-nominale Doppelnatur der hebraischen Participien und Infinitive,
Diss. Leipzig 1889.
Sepher ha-ca~amim (wrongly ascribed to Abraham ibn Ezra), ed., M. Grossberg,
London 1901.
Steinschneider, M., Catalogus librorum hebraeorum in bibliotheca bodleiana, Berlin 1852-60.
- - , Die arabische Literatur der Juden, Frankfurt a.M. 1902.
Stern, S.G., '1~1 11i~p '1 i1"i1 p1iO '1 ::lp'" P C1nJr:l '1'r:l?n i11::l1iDn i~O, Vienna 1870.
Stevenson, W.B., A grammar qf Palestinian Jewish Aramaic, Oxford 1924.
Strack, H.L., Einleitung in Talmud und Midraf, 5th ed., Munich, 1930.
Strauss, D., Sprachliche Studien in den hebraischen Sirachfragmenten, Diss. Zurich
1900.
Strauss, L., The literary character of the Guide for the Perplexed, in Essays on Maimonides,
ed. S.W. Baron, New York 1941, pp. 37-91.
 BIBLIOGRAPHY 195

Szneider, M., '~iPDil plipi'? 'nliDDil plipiil on" Leshonenu iii (1931) 15-28.
- - , ?t1i~liD ]liD'? n'~ipDil ]liD'?il i1r1'i1il, Leshonenu viii (1936) 112-22.
Targum on Canticles, ed. R.H. Melamed, JQR x (N.S.) 377-410, xi 1-20, Xli
57-Ill.
Taylor, C., Sqyings rif the }ewish Fathers, Cambridge 1877-1900.
Tobia b. Eliezer, ~l0 np'? iDiiD, ed. S. Buber, vo!. i, Wilna 1880.
Tosephta, ed. M.S. Zuckermandel, Trier 1882.
Waxman, M., A history rif Jewish literature, vo!. i, New York 1930.
Wilson, A.M., The particle n~ in Hebrew, Hebraica vi (1890) 139-150, 212-224.
Wright, W., A grammar rif the Arabic language, 3rd ed., 2 vols., Cambridge 1896-8.
Zadokite Fragment, ed. S. Schechter, Documents ofJewish Sectaries, vo!. i, Cambridge
1910.
ZarJ:ii, J., C1'Ell,"il iil, novel, Jerusalem 1940.
Zucker, M., ?''''?iDD ~iD~" 0l'Elil n~ ll~j il'i.110 ~i ~n:::l 'D ijJ Tarbiz 27 (1958) 61-82.
Zunz, L., Die gottesdienstlichen Vortriige der Juden, Berlin 1832.
- - , Zur Geschichte und Literatur der Juden, vo!. i, Berlin 1845.
 INDEX OF PASSAGES*

Bible

Genesis 2.18 159 29.17 110

6.16 III 32.4 29 (7)
16.2 144 32.10 126 (3)
29.19 159 39.2 17 (36)
30.18 132 42.3 142 (2)
41.27 166 42.24 135
46.26 130 43.13 135
47.14 109 Jeremiah 20.6 135
49.13 135 32.5 135
Exodus 8.22 159 48.36 135
12.39 121 49.8 167
14.12 159 49.20 94 (3)
28.10 88 50.31 167
30.34 38 (65) Amos 6.10 164 (6)
32.33 176 8.6 155 (I)
Numbers 5.23 133 Micha 1.10 135
16.35 88 (I) Habakkuk 1.6 10
22.18 181 Psalms 7.6 131
35.6 145 37.23 114 (I)
Deuteronomy 22.8 113 37.25 132
27.9 110 50.4 166
32.39 135 58.9 37 (56)
Joshua 4.4 88 (I) 60.6 164
7.5 61 (25) 68.36 10
I Samuel 9.7 17, 165 94.17 181
17.40 94 (30) 106.11 8 (4)
20.8 135 106.23 180
2 Samuel 5.24 124 (5) 119.92 180
18.3 159 124.1 180
19.20 145 124.2 180
20.11 176 Proverbs 7.25 91 (4)
I Kings 4.13 164 Job 6.17 167
8.8 181 7.11 132 (I)
11.30 121 10.3 163
15.13 62 (27) 11.17 17
18.32 121 13.9 159
2 Kings 5.3 124 (5) 16.4 36 (52)
Isaiah 1.9 180 19.23 135
9.1 93 24.1 114 (I)
13.9 155 (I) 27.3 168
16.10 113 (I) 31.26 16 (23)
23.12 135 Canticles 4.5 150
28.10 126 (3) Ruth 2.22 159

* Numbers (enclosed) in brackets refers to footnotes.

198 INDEX

Lamentations 2.5 17 (36) 8.8 164

Ecclesiastes 2.22 109 Daniel 1.2 104 (5)
3.15 123 (3) 2.42 104 (5)
3.22 178 3.19 160
6.6 179 Ezra 4.15 22 (10)
7.12 166 4.21 155 (I)
7.26 107 (2) Nehemiah 6.6 109
9.67 123 7.70 104 (5)
12.11 114 (I) I Chronicles 15.2 164 (6)
Esther 2.9 160 2 Chronicles 5.11 164
4.2 164 20.6 135
7.4 179 20.22 167

Ben Sira

10.23 165 39.21 165

I !.I 9 167

Mishnah

Aboth 1.3 129 (3) 7.4 163

1.6 132 (I) 7.6 160
2.1 135 11. 7 160
2.12 I 16.7 II (20)
5.9 90 (3) Yadayim 4.4 123 (2)
5.18 90 (3) Kil'aim 8.4 114 (2)
Baba Batrah 5.2 124 (4) Ketubboth 13.8 165
5.9 101 (I) Megillah 1.3 101 (I)
8.1 103 (4) 1.4 128 (2)
8.2 IQ] (I) 4.1 129 (3)
8.5 160 Makkoth 2.3 103 (4)
Baba Mqi'a 10.2 124 (4) Menal,Ioth 10.3 101 (I)
Baba J>.ama 1.4 160 Nega'im 11.9 124 (4)
Bekhoroth 5.4 160 Nedarim 8.6 132 (I)
Berakhoth !.I 93 11.4 160
1.2 142 Nazir 5.7 124 (4)
Ginin 4.5 123 7.2 98 (2)
7.2 124 (4) So!ah 4.3 160
9.5 124 (4) 9.6 132 (2)
9.8 121 Sukkah 3.7 132 (I)
Demai 3.5 129 Sanhedrin l.l 93
6.11 164 2.1 86
6.12 85 2.2 86
7.1 129 3.8 177
Horayoth 1.1 160 4.5 160
Zebal,Iim 9.1 160 7.10 101 (I)
11.3 160 8.4 160
J:Iagigah 2.1 161 11.2 177
J:Iullin 5.3 164 'U4in 2.10 120 (I)
J:Iallah 1.7 144 'Erubin 2.5 132 (I)
2.5 132 (I) 4.3 123
Yebamoth 5.1 126 (4) 8.5 142
 INDEX 199

'Arakhin 8.7 123 (3) 4.2 132 (I)

'Orlah 3.7 160 4.10 142
Pe'ah 5.1 160 8.10 132 (2)
8.4 142 Shabbath 1.4 lOO
Pesal).im 6.1 98 (2) 15.2 160
6.5 160 Temurah 5.3 101 (I)
Parah 3.10 118 Tamid 3.7 114
~ddushin 3.13 I11 Ta'anith 3.6 132 (2)
4.14 12 3.8 101 (I)
Rosh Hashanah 2.8 132 (2) Terumoth 3.1 164
Shebi'ith 2.2 132 (I) 6.6 160

Talmud Babli

Baba Batra 71a 165 Mo'ed IS-a!an 25b 9 (6)

Berakoth 3a 165 Middah 43b 103 (2)
14a 163 Sukkah 49b 164
17a 155 'Aboda Zarah 58b 8 (I)
59b 9 (9) Pesal).im 42a 22 (12)
I:Iullin 27b 142 (2) 66a 98 (2)
137b 8 (I), ~ddushin 66a 11, 153
18 (43) Rosh Hashnah 20b 164
Yebamoth 12b 135 (I) Ta'anith 6b 9 (9)

Targumim

Deuteronomy 23.24 22 (10) Canticles 6.1 22 (10)

Jeremiah 40.5 22 (10) Lamentations 1.1 22 (10)

Commentators

B. Ezra Abraham 1.4 134

1.5 139
Commentary on the Bible:
1.6 174
Psalms 127.3 167
5.10 173
Ecclesiastes 5.1 18 (43)
Rashi
Gersonides (Levi b. Gershom)
Commentary on the Bible:
Proverbs 1.1 124 Genesis 1.4 163
1.8 173 1.5 165
1.18 124 1.14 143
17.12 167 1.20 90 (3)
Job 21.34 175 8.5 117 (2)
15.15 100 (6)
"im~i David 27.21 146
Commentary on Psalms: 49.13 170, 175
Psalms 1.1 136, 139, Exodus 19.19 146
143, 157, Joshua 5.2 90 (3)
158 1 Samuel 19.19 85
1.3 105, 173, 21.12 165
175,177 Isaiah 53.1 165
200 INDEX

Psalms 69.32 109 (2) 99b 98

Proverbs 5.19 26 (2); Be~ah lOb 91 (6)
27 (11) Yoma 13a 91 (6)
Job 31.26 16 (23) Niddah 7b 91 (6)
'Aboda Zara 40b 98
Commentary on Talmud Babli: 'Arakhin 8b 91 (6)
Baba Batra 12b 91 (6) Shabbath 84a 177
Baba Mqi'a 47a 104 (5)

Medieval Texts

A~imaa?, 70 88
(The chronicle 71 132
of Al:Iimaa~) 3 98, 117 72 97, 130,
5 133 140, 146
6 17 (28) B. E:;.ra Abraham,
8 98 Yesod Mora 106, 126,
151 143, 158
Botarallo Moses, 2 88,163,171
Commentary 3 104, 180
on the 4 91, 127
Sepher 5 143
Yqirah 77 (6) 6 97, 102,
B. 'Abbas 142, 176,
(B. Samuel 177
Judah), Ya'ir 8 124, 125,
Nethib 3b 102, 147 154
4a 143 9 96, 125,
4b 103, 168, 181, 182
170 10 128, 139
5a 136 II 125, 153,
5b 125 174
6a 119, 152 13 124, 179
7b 143, 164, 16 94, 130,
167 154
8a 152 17 148, 152,
9a 131, 155 162
9b 162, 167 18 86,88
lOa 118, 168 20 132
Ila 136, 153, Saphah Berurah Ib 164
162 2b 88
12a 103 3a 153
B. Baruch Isaac 4b 113, 130,
Albalia, 167
~uppat 5a 97
Rokhlim 42 5b 142
B. Da'ud 7a 102, 182
Abraham, 9a 122, 144
Sepher lOb 147
ha-~abbalah 22 99 17b 148
51 115, 149 18a 88
59 87 24a 174
63 86, 119 43b 170
 INDEX 201

Sepher 19a 85, 179

ha-'A+amim 5 102 19b 126, 153
Sepher 20 87
ha-?:ai:loth la 152 20a 90, 163
Ib 127, 158 20b 91
2b 88, 136 21a 158
3a Ill, 150 Wikkuai:l 82 171
4b 112 B. lfryya
6a 101, 142, Abraham,
166, 181 Hegyon
6b 181 ha-nephesh 87, 106,
7b 113, 167 136, 146
8a 132 2 109, 127,
9a 124 146, 147,
Sepher 181
ha-Shem la 142, 164 3 138,173
2a 88, 173 3a 119
2b 104 4 87,90,
3a 110, 132 136, 142,
3b 110 147, 172
8b 148 (5),168 5 86, 108,
9b 152 118, 125,
12a 170 127, 142
B. E::.ra Moses, 6 lOO
Shirat Israel 26 40 (I) 7 97, lOO
55 62 (28) 8 129, 154,
60 47 (37) 167
65-7 47 (43) 11 87, 120,
66 47 (41) 139, 144,
67 49 (54) 182
72 48 (50) lib 90, 114,
73 48 (52) 121
74 49 (55) 12b 91, 96, III
76 41 (5) 13a 91, 99,
104 24 (19) 114, 144
147 45 (28) 13b 114,138,
157 45 (3) 165
46 (34) 14a 169, 173
B. Falaquera 14b 170
Shemtob, 15a 108, 179
Sepher 16a 91
ha-Mebai,cl,<:.esh 4b 136 16b 119,174
lOb 136 17a 93, 101,
lla 103, 121, 175, 176
153,167,168 17b 159
lib 168 51 (2),
12a 102, 129 109 (2)
12b 147, 170 I:Iibbur
13a 90, 102, 157 ha-Meshii:lah
13b 171 weha-tishboreth 2 97, 128
14b 90 3 92
17a 88 Megillath
18a 94, 134, 142 ha-Megalleh 90,95, 118,
18b 118, 133, 152 162, 163
202 INDEX

2 121, 122, B. Judah Ibn

153, 155, Musa if
174, 181 Salamanca,
3 103, 118 Magen
4 165 wa-Roma\:l 77 (1), 78
5 128, 139 B. Maimon
6 106 Abraham, Kifayat
7 127, 140 al-'abidin 69 (4)
8 109 140 B. Par~on,
9 129, 164, Ma\:lbereth
165, 174 he-'arukh 22 94, 101,
10 107, 119, 121, 139,
154, 156 162, 173,
11 128 176
13 144 55
14 139, 168 B. Sakbel
15 142 Solomon,
16 174 Sepher
20 175 Ta\:lkemony 49
22 87 B. Samuel if
23 121, 142, Verona Hillel,
159, 161, Sepher
162, 163 Tagmuley
24 146, 149 ha-Nephesh 75 (35)
25 97 B. Saruf:;
27 96, 154 Men~em,
29 126 Ma\:lbereth 136
49 92, 130 B. Solomon
50 107,130,182 Gershom,
51 147 Sha'ar
57 134 ha-Shamaim 63, 72 (19)
58 181 B. Solomon
64 109 Menaly.em,
86 120 'Eben Bo\:lan 57
I 11 91, 92, 97, B. Tibbon Joseph,
119,121 She~el
112 125, 167 ha-J>.odesh 58 (11)
113 102, 136, B. Tibbon Judah,
138, 146, Tibbonid
156 translation of
114 91, 103, the f.lobot
110, 121, ha-Lebabot 58, 59 (12)
125, 138, Testament I 138
152, 153 3 166, 167
115 175 4 94, 108,
116 88, 122, 150 122, 178,
Sepher 180
ha-'Ibbur 7 88,91, 92, 6 176
136 7 89,114,140
10 87 8 128
42 9 103, 131,
B. Jalf.uthiel Jely.iel, 142, 145,
Ma'alot 154, 169,
ha-Middot 75 177
 INDEX 203

10 102, 123, 9b 90, 148,

143, 157, 152, 153,
169, 177, 157, 158,
178, 181 169, 176
II 92, 107, MilJ:iamoth
151, 165 'Adonai 4a 90, 139,
12 103, 153 157, 169,
B. Tibbon Moses, 176, 178
Canticles 4b 94, 104
Commentary 5 117,122, 8b 103, 140
171 26a 90
B. Tibbon Samuel, 28b 90, 108
Ecclesiastes 42b 91, 102
Commentary 15b lOO, 112, 46b 90
163 48b 97, 140,
16a 162, 176 171, 178
16b 168 57b 169
Ma'amar 73
Yi~awu Pentateuch
ha-Maim 58 Commentary 12b 90, 104
Perush Kohelet 58 47b 163
B. Verga Solomon, 48b 90
She be! Yehudah 78 80b 104
B. Zar;:.ah, Mikhlol f:lay Gaon,
Yophi 71, 76 (37) Kitab al bai'
B. Zerah Mena~em, washira 44
~edah Laderekh 71 Isaac if Catalonia,
Crescas (f:lisdai), 'Ezrath Nashim 74 (30)
'Or 'Adonai 4a 114, 121, Josippon, Zeh
157, 162 ha-Mil,<dash 64 98
Gavison Abraham, Maimonides
'Omer (B. Maimon
ha-ShikheJ:ia 78 Moses),
Gerondi 'Akkum l.l 90, 107,
(B. Reuben 110, 136
Nissim), De'oth 4.1 170
Sermons 23b 178 4.8 85, 87
30b 90, lOO 4.9 129, 146,
47b 90, 102 163
61a 136 5.5 93
Gersonides 5.12 13/
(B. Gershom 6.1 142
Levi), De'oth 6b 87, 103, ~obe~
104, 114, Teshuboth
124,172 ha-Rambam I 4 179
7a 140, 165 II la-7b 64 (2)
7b 124 II 2b 64 (3)
8a 85, 91, II 12a-15b 64 (I)
167 (3), II 15b-16b 65 (5)
171, 172, II 24 139
173, 180 II 25 88,93,101,
8b 114,169, 103, 109,
182 130, 139,
9a 114,152 143, 174
204 INDEX

II 26 94, 98,106, 3 33, 150

111,123, 7 33, 34,
129, 153, 131, 165
163, 180 12 33, 146
II 27 100, Ill, 13 33, 34, 135
127, 174 14 147
III 68 (1) 19 144
III la 136 23 34
III 2a 87, 132, 27 34
146, 150 31 34
III 17b 101, 102, 38 100
117,133, 40 135
144 48 33, 144
III 18a 99, 104, 114 64 33, 161
III 18b 136, 153 Sepher 'Emunot
III 19a 106, 109, wede'oth 41
127, 162 Sepher ha-Yashar 25 (24), 37
III 19b 144, 146 Sepher ha-Zohar 71, 72 (18)
III 179 93 Sepher I:Iasidim
III 199 99 (by Judah B.
Mishneh Tora 24(4), 64, Samuel IfasidJ 80 177
65, 66, 547 164
67, 117 548 165
Sepher 669 128, 130
ha-Mi~voth 42 (13), 826 98
60 (17) 1099 173
Talmud Torah 1.7 146 1109 173
4.1 114 1138 128
4.4 170, 179 1142 177, 178
5.1 94 1144 135
7.5 114 1145 125
7.10 96 1378 135
7.17 162 1379 178
Yesode ha-torah 1.1 126 28, 117
1.2 153 Solomon
1.4 175 al-'ammi,
1.10 146 'Igeret
2.2 125, 131, ha-Musar 77 (5)
144 :tlim~i David,
3.3 157 Mikhlol la 126, 134
4.2 146, 174 5b 90, 152, 165
4.13 129, 132 7b 156
5.1 131 8b 148
7 132 23a 104, Ill,
Meli~at 'Opher 147, 152,
weDina 75 162, 164
Nahmias J. Aboth 1
Commentary 3a 154 Kimhi Joseph,
3b 96 Sepher ha-Berith 54
Saadiah Gaon, Sepher ha-Galuy 91, 104,
Egron 31, 34, 113,118,
124, 126 119, 132
Polemic against 2 90, 96, 107,
I:Iiwi al Balkhi 89, 146 168, 176
 INDEX 205

3 90, 178 flim~i Moses,

12 111,118, Mahalakh
147 Shebiley Da'at 55
31, 32, 54 Sekhel Tob 55
Sepher Zaccuto Abraham,
ha-Zikkaron 2 104,119 Sepher Yul}asin 78
3 90 Zarphthi Jacob,
4 168, 180 Mishkenoth Ya'a~ob 75
54
 STUDIES IN SEMITIC
LANGUAGES AND LINGUISTICS

3. Corre, AD. 1he Daughter qf MY People. Arabic and Hebrew Paraphrases of

Jeremiah 8.13-9.23. 1971. ISBN 90 04 02552 9
5. Grand'Henry, J. Les parlers arabes de la region du Mziib (Sahara algbien).
1976. ISBN 90 04 04533 3
6. Bravmann, M.M. Studies in Semitic Philology. 1977. ISBN 90 04 047433
8. Fenech, E. Contemporary Journalistic Maltese. An Analytical and Compara-
tive Study. 1978. ISBN 90 04 05756 0
9. Hospers, J.H. (ed.). General Linguistics and the Teaching qf Dead Hamito-
Semitic Languages. Proceedings of the Symposium held in Groningen, 7th-
8th November 1975, on the occasion of the 50th Anniversary of the
Institute of Semitic Studies and Near Eastern Archaeology of the State Uni-
versity at Groningen. 1978. ISBN 90 04 05806 0
12. Hoftijzer, J. A Search for Method. A Study in the Syntactic Use of the H-
locale in Classical Hebrew. With the collaboration of H.R. van der Laan
and N.P. de Koo. 1981. ISBN 90 04 06257 2
13. Murtonen, A Hebrew in its West Semitic Setting. A Comparative Survey of
Non-Masoretic Hebrew Dialects and Traditions. Part I. A Comparative
Lexicon.
Section A Proper Names. 1986. ISBN 90 04 07245 4
Section Ba. Root .$ystem: Hebrew Material. 1988. ISBN 90 04 08064 3
Section Bb. Root .$ystem: Comparative Material and Discussion. Sections C,
D and E: Numerals under 100, Pronouns, Particles. 1989.
ISBN 90 04 08899 7
14. Retso, J. Diathesis in the Semitic Languages. A Comparative Morphological
Study. 1989. ISBN 90 04 08818 0
15. Rouchdy, A Nubians and the Nubian Language in Contemporary Egypt. A Case
of Cultural and Linguistic Contact. 1991. ISBN 90 04 09197 1
16. Murtonen, A Hebrew in its West Semitic Setting. A Comparative Survey of
Non-Masoretic Hebrew Dialects and Traditions. Part 2. Phonetics and
Phonology. Part 3. Morpho,ryntactics. 1990. ISBN 90 04 09309 5
17. Jongeling K., H.L. Murre-van den Berg & L. van Rompay (eds.). Studies
in Hebrew and Aramaic .$yntax. Presented to Professor J. Hoftijzer on the
Occasion of his Sixty-Fifth Birthday. 1991. ISBN 90 04 09520 9
18. Cadora, F J. Bedouin, Village, and Urban Arabic. An Ecolinguistic Study.
1992. ISBN 90 04 09627 2
19. Versteegh, C.H.M. Arabic Grammar and OJtr'iinic Exegesis in Early Islam.
1993. ISBN 90 04 09845 3
20. Humbert, G. Les voies de la transmission du Kitiib de Szbawqyhi. 1995. ISBN
900409918 2
21. Mifsud, M. Loan Verbs in Maltese. A Descriptive and Comparative Study.
1995. ISBN 90 04 10091 1
22. Joosten, J. The .$yriac Language qf the Peshitta and Old .$yriac Versions qf Mat-
thew. Syntactic Structure, Inner-Syriac Developments and Translation
Technique. 1996. ISBN 90 04 10036 9
23. Bernards, M. Changing Traditions. Al-Mubarrad's Refutation of SIbawayh
 and the Subsequent Reception of the Kitab. 1997. ISBN 90 04 10595 6
24. Belnap, R.K. and N. Haeri. Structuralist Studies in Arabic Linguistics. Charles
A. Ferguson's Papers, 1954-1994. 1997. ISBN 90 04 10511 5
25. Talmon R. Arabic Grammar in its Formative Age. Kitab al-'4Jn and its Attribu-
tion to tIalTl b. Alpnad. 1997. ISBN 90 04 10812 2
26. Testen, D.D. Parallels in Semitic Linguistics. The Development of Arabic ~­
and Related Semitic Particles. 1998. ISBN 90 04 109730
27. Bolozky, S. Measuring Productivity in Word Formation. The Case of Israeli
Hebrew. 1999. ISBN 90 04 11252 9
28. Ermers, R. Arabic Grammars if Turkic. The Arabic Linguistic Model Applied to
Foreign Languages & Translation if 'Abii !byyiin al-'Andalusl's Kitab al-1.drak
li-Lisan al-'Atrak. 1999. ISBN 90 04 113061
29. Rabin, Ch. The Development if the !iJntax if Post-Biblical Hebrew. 1999.
ISBN 90 04 114335