AdinaMoshavi 2010 WordOrderintheBiblicalHebrewFiniteClause

Copyright 2010. Eisenbrauns. All rights reserved.
May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable
copyright law.
Account: s9857968
AN: 448598 ; Adina Moshavi.; Word Order in the Biblical Hebrew Finite Clause
EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 8/25/2023 2:20 PM via
Word Order in the
Biblical Hebrew Finite Clause
EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use
00-Moshavi-LSAWS4.indb 1 9/28/10 7:36 AM

Linguistic Studies in Ancient West Semitic
edited by
M. O’Connor† and Cynthia L. Miller

The series Linguistic Studies in Ancient West Semitic is devoted to the ancient
West Semitic languages, including Hebrew, Aramaic, Ugaritic, and their near
congeners. It includes monographs, collections of essays, and text editions in-
formed by the approaches of linguistic science. The material studied will span
from the earliest texts to the rise of Islam.
1. The Verbless Clause in Biblical Hebrew: Linguistic Approaches,

edited by Cynthia L. Miller
2. Phonology and Morphology of Biblical Hebrew: An Introduction,
by Joshua Blau
3. A Manual of Ugaritic, by Pierre Bordreuil and Dennis Pardee
4. Word Order in the Biblical Hebrew Finite Clause: A Syntactic and
Pragmatic Analysis of Preposing, by Adina Moshavi
5. Oath Formulas in Biblical Hebrew, by Blane Conklin

Word Order in the
Biblical Hebrew Finite Clause
A Syntactic and Pragmatic Analysis of Preposing
A DINA MOSHAVI
Winona Lake, Indiana

EISENBRAUNS
2010

Copyright © 2010 Eisenbrauns
All rights reserved.
Printed in the United States of America.
www.eisenbrauns.com
Library of Congress Cataloging-in-Publication Data
Moshavi, A. Mosak (Adina Mosak)

Word order in the biblical Hebrew finite clause : a syntactic and
pragmatic analysis of preposing / Adina Moshavi.
p. cm. — (Linguistic studies in ancient West Semitic ; 4)
Includes bibliographical references and indexes.
ISBN 978-1-57506-191-7 (hardback : alk. paper)
1. Hebrew language—Clauses. 2. Hebrew language—Word
order. 3. Bible. O.T.—Language, style. I. Title.
PJ4717.M67 2010
492.4′5—dc22
2010013367
The paper used in this publication meets the minimum requirements of the American
National Standard for Information Sciences—Permanence of Paper for Printed Library
Materials. ANSI Z39.48-1984.†Ê

To my parents,
Dr. Richard and Orah Mosak
and to my husband,
Shimon
00-Moshavi-LSAWS4.indb v 9/28/10 7:36 AM

00-Moshavi-LSAWS4.indb vi 9/28/10 7:36 AM

Contents
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Technical Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1. The pragmatics of preposing . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. The corpus for the study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3. Overview of the chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Word-Order Markedness in Biblical Hebrew . . . . . . . . . . . . . . . . . . 7
2.1. Basic word order: Typological and generative perspectives . . 7
2.2. Basic word order in Biblical Hebrew . . . . . . . . . . . . . . . . . . 10
2.2.1. The statistical dominance of VSO in Biblical
Hebrew 11
2.2.2. Word order in the basic sentence 13
2.2.3. The generative perspective on basic word order in
Biblical Hebrew 16
2.2.4. Word order and discourse type 17
2.2.5. Word order in the nonverbal and the participial
clause 17
2.3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3. Previous Studies of the Functions of Preposing in
Biblical Hebrew . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.1. The emphasis-centered model . . . . . . . . . . . . . . . . . . . . . . . . 18
3.1.1. Contrast and contrastive structures 19
3.1.2. Circumstantiality, anteriority, and simultaneity 20
3.1.3. Narrative-unit demarcation 24
3.1.4. Attraction 25
3.1.5. Miscellaneous factors 26
3.1.6. Preposing in direct speech 26
3.1.7. Conclusion 26
3.2. The backgrounding and temporal-sequencing models . . . . . 27
vii
00-Moshavi-LSAWS4.indb vii 9/28/10 7:36 AM

viii Contents
3.3. The information-structure model . . . . . . . . . . . . . . . . . . . . . . 31

3.3.1. Introduction to information-structure concepts 32
3.3.1.1. Theme-rheme/topic-comment 32
3.3.1.2. Topicalization 33
3.3.1.3. Focus and focusing 35
3.3.2. Bendavid: Psychological subject and predicate 36
3.3.3. Bandstra 37
3.3.4. Revell 37
3.3.5. Buth: A synthesis of information structure and
foreground-background 38
3.3.6. Gross: A focus-centered model 40
3.3.7. Heimerdinger, van der Merwe, and Shimasaki 41
3.4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4. The Biblical Hebrew Finite Clause and Its Constituents . . . . . . . . . 48
4.1. The clause and the sentence . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2. Finite and nonfinite clauses . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.3. Subordinate and nonsubordinate clauses . . . . . . . . . . . . . . . . 50
4.3.1. The relative clause 51
4.3.2. The adjunct clause 51
4.3.2.1. The conditional adjunct clause 52
4.3.2.2. The causal adjunct clause 53
4.3.2.3. The temporal adjunct clause 53
4.3.2.4. The purpose adjunct clause 54
4.3.3. The content clause 54
4.3.3.1. Complement of verb 55
4.3.3.2. Complement of preposition 55
4.4. Difficult issues in defining subordination . . . . . . . . . . . . . . . 56
4.4.1. The conjunction ‫ו‬: Coordinator and
subordinator? 56
4.4.2. The ‫ ויהי‬construction 57
4.4.3. Direct and indirect speech 59
4.5. Syntactic classification of the constituents of the clause . . . . 61
5. The Syntax of Preposing and Other Word-Order Constructions . . . 64
5.1. Preposing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2. Nonpreposed preverbal constituents . . . . . . . . . . . . . . . . . . . 68
5.2.1. Clausal adverbs 68
5.2.2. Negative particles: ‫ לא‬and ‫טרם‬ 76
5.2.3. Preverbal adjunct clauses 76
5.3. Unmarked preposed constituents . . . . . . . . . . . . . . . . . . . . . . 78
5.3.1. Interrogative pro-forms 78
5.3.2. Clause-initial time adverbs and the demonstrative
adverb ‫ כה‬79
00-Moshavi-LSAWS4.indb viii 9/28/10 7:36 AM

Contents ix
5.4. Complex variants of the preposing construction . . . . . . . . . . 80

5.4.1. Preposing with a focusing adverb 80
5.4.2. Double preposing 81
5.5. Other marked word-order constructions . . . . . . . . . . . . . . . . 81
5.5.1. Left-dislocation 81
5.5.2. The preverbal constituent connected to its clause with
a conjunction 84
5.6. A statistical analysis of word-order constructions
in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.7. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6. Focusing and Topicalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.1. Focusing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6.1.1. Constituent, predicate, and sentence focus 90
6.1.2. Types of givenness: Pragmatic presupposition and
activation 92
6.1.3. Activation in a model of text comprehension 94
6.1.4. Focusing in Biblical Hebrew 96
6.2. Topicalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.2.1. Prince’s and Birner and Ward’s characterization of
topicalization 97
6.2.1.1. Marking a partially ordered set relation
pertaining to the preposed element 98
6.2.1.2. Marking a focus in relation to a given
proposition 99
6.2.2. Topicalization in Biblical Hebrew: A discourse-
connective device 101
7. The Pragmatics of Preposing: A Statistical Analysis . . . . . . . . . . . . 104
7.1. Focusing and topicalization in preposed clauses . . . . . . . . . 104
7.2. Focusing and topicalization in
complement/adjunct-preposed clauses . . . . . . . . . . . . . . 105
7.2.1. The residue: Possible instances of focusing 107
7.2.2. The residue: Fixed expressions with preposed word
order 109
7.3. Topicalization and focusing in subject-preposing
constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
7.3.1. The residue of subject-preposed narrative
clauses 112
7.3.1.1. Anteriority 112
7.3.1.2. Simultaneity 113
7.3.1.3. Background information 114
7.3.1.4. New narrative unit or new scene within the
narrative 115
00-Moshavi-LSAWS4.indb ix 9/28/10 7:36 AM

x Contents
7.3.2. The residue of subject-preposed direct speech

clauses 115
7.3.2.1. Justification 116
7.3.2.2. Affirmation 118
7.3.2.3. Boasting 119
7.4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
8. The Focused Clause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
8.1. Syntactic description of the preposed constituent . . . . . . . . 121
8.2. The activated proposition . . . . . . . . . . . . . . . . . . . . . . . . . . 123
8.2.1. The relation between focused clause and activated
proposition 123
8.2.2. The source of the activated proposition 124
8.2.3. Activation versus pragmatic presupposition in the
focused clause 126
8.3. A taxonomy of focused clauses . . . . . . . . . . . . . . . . . . . . . . 127
8.3.1. Identificational focusing 128
8.3.2. Descriptive focusing 130
8.3.3. Substitutional focusing 131
8.3.4. Additive focusing 133
8.4. Focus of negation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.5. The stylistic use of focusing . . . . . . . . . . . . . . . . . . . . . . . . 140
8.6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
9. The Topicalized Clause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
9.1. Syntactic description of the preposed constituent and its
counterpart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
9.2. The link between the preposed item and its counterpart . . . 146
9.3. The linked segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
9.4. Coherence relations between segments linked by
topicalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
9.4.1. Opposition 155
9.4.2. Similarity 160
9.4.3. Addition 161
9.4.4. Elaboration, summary, and paraphrase 162
9.4.5. Temporal succession 163
9.5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Indexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Index of Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Index of Scripture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
00-Moshavi-LSAWS4.indb x 9/28/10 7:36 AM

List of Tables
Table 1. Word-order constructions in finite nonsubordinate

clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Table 2. Frequency of preposing in preposable finite nonsubordinate
clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Table 3. Relation between discourse register and word order
in preposable finite nonsubordinate clauses in Genesis . . . . . . . . . . . 87
Table 4. Syntactic class of the preposed element in marked
preposed clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Table 5. Syntactic function of the preposed element in marked
preposed clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Table 6. Pragmatic functions of preposed clauses in Genesis . . . . . . . . . . 105
Table 7. Relation between syntactic function of the
preposed constituent and pragmatic function of preposing . . . . . . . . 105
Table 8 . Relation between register and pragmatic function
of preposing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Table 9. Pragmatic functions of complement/adjunct preposed clauses . 106
Table 10. Pragmatic functions of complement/adjunct-preposed
clauses: Narrative versus direct speech . . . . . . . . . . . . . . . . . . . . . . . 107
Table 11. Pragmatic functions of subject-preposed clauses . . . . . . . . . . . 111
Table 12. Pragmatic functions of subject-preposed clauses:
Narrative versus direct speech . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Table 13. Syntactic function of the preposed constituent
in focused clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Table 14. Syntactic category of the preposed constituent in
focused clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Table 15. Syntactic function of the preposed constituent
in topicalized clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Table 16. Syntactic category of the preposed constituent
in topicalized clauses in Genesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
xi
00-Moshavi-LSAWS4.indb xi 9/28/10 7:36 AM

Abbreviations
General
BH Biblical Hebrew
NAB New American Bible
NJB New Jerusalem Bible
NJPSV New Jewish Publication Society Version
NP noun phrase
NRSV New Revised Standard Version
O object
RSV Revised Standard Version
S subject
V verb
VP verb phrase
Reference Works
AfroLing Afroasiatic Linguistics
BDB Brown, Francis; Driver, S. R.; and Briggs, Charles A. A Hebrew and
English Lexicon of the Old Testament. Oxford: Clarendon, 1907
BHS Elliger, K., and Rudolph, W., editors. Biblia Hebraica Stuttgartensia.
Stuttgart: Deutsche Bibelgesellschaft, 1984
BI Balšanut Ivrit
Bib Biblica
BibOr Bibliotheca Orientalis
BN Biblische Notizen
BSOAS Bulletin of the School of Oriental and African Studies
DiscProc Discourse Processes
FolOr Folia Orientalia
GKC Gesenius’ Hebrew Grammar, ed. E. Kautzsch, rev. A. E. Cowley. 2nd ed.
Oxford: Clarendon, 1910
HALOT Köhler, Ludwig and Baumgartner, Walter. The Hebrew and Aramaic
Lexicon of the Old Testament, rev. Walter Baumgartner et al. 1st English
ed. 5 vols. Leiden: Brill, 1994-2000
HAR Hebrew Annual Review
HS Hebrew Studies
HUCA Hebrew Union College Annual
IOS Israel Oriental Studies
JANES Journal of the Ancient Near Eastern Society
xii
00-Moshavi-LSAWS4.indb xii 9/28/10 7:36 AM

Abbreviations xiii
JAOS Journal of the American Oriental Society

JBL Journal of Biblical Literature
JLing Journal of Linguistics
JNSL Journal of Northwest Semitic Langagues
Joüon-Muraoka Joüon, Paul. A Grammar of Biblical Hebrew, trans. and rev.
T. Muraoka. Subsidia Biblica 14. 2 vols. Rome: Pontifical Biblical
Institute, 1991
JPrag Journal of Pragmatics
JQR Jewish Quarterly Review
JSem Journal for Semitics
JSOT Journal for the Study of the Old Testament
JSOTSup Journal for the Study of the Old Testament Supplement Series
JSS Journal of Semitic Studies
JTT Journal of Translation and Textlinguistics
KUSATU Kleine Untersuchungen zur Sprache des Alten Testaments und seiner
Umwelt
Lang Language
LingBer Linguistische Berichte
Leš Lešonénu
LešLaʿam Lešonénu Laʿam
Ling Linguistics
LingInq Linguistic Inquiry
Or Orientalia
OTS Oudtestamentische Studiën
Philos Philosophica
Prag Pragmatics
RB Revue Biblique
SBLSemeiaSt Society of Biblical Literature Semeia Studies
SBLMS Society of Biblical Literature Monograph Series
ScrHier Scripta Hierosolymitana
Sem Semitics
SIL Summer Institute of Linguistics
SIGLA Studies in Generative Linguistic Analysis
SJOT Scandinavian Journal of the Old Testament
SSLL Studies in Semitic Languages and Linguistics
SSN Studia Semitica Neerlandica
TheorLing Theoretical Linguistics
TSL Typological Studies in Language
VT Vetus Testamentum
VTSup Vetus Testamentum Supplements
ZAH Zeitschrift für Althebraistik
ZAW Zeitschrift für die alttestamentliche Wissenschaft
00-Moshavi-LSAWS4.indb xiii 9/28/10 7:36 AM

Technical Notes
The translations of the biblical citations in this work are my own, in consul-
tation primarily with the NJPSV, RSV, and NRSV. An attempt is made to render the
syntactic structure of the Hebrew as literally as possible, for the purposes of
illustrating the syntactic analyses presented in the text. The results are at times
far from idiomatic.
Preposed clauses in the citations are generally marked by underlining when
the citation contains more than one clause. Citations consisting of a single
preposed clause are not underlined.
The symbol * is used to mark an ungrammatical sentence, and # marks a
sentence that is pragmatically unacceptable, being inappropriate in the given
context.
xiv
00-Moshavi-LSAWS4.indb xiv 9/28/10 7:36 AM

Preface
This book has its origins in my 2000 dissertation, The Pragmatics of Word
Order in Biblical Hebrew: A Statistical Analysis, completed at Yeshiva Uni-
versity under the direction of Richard C. Steiner. I submitted the dissertation
to Eisenbrauns that same year, intending to publish it with minor revisions.
Continuing to study pragmatic and syntactic issues related to word order over
the next six years, I worked on what turned out to be, essentially, an entirely
new book on Biblical Hebrew word order. The completed manuscript was sub-
mitted to Eisenbrauns in November 2006 and accepted for publication in May
2007.
My dissertation dealt with the pragmatics of inverted (XV) word order in
declarative nonsubordinate finite clauses in Genesis and showed by means of
multivariate statistical analysis that focus-background structure (including fo-
cusing and topicalization, as defined there) and nonsequentiality are associated
with word order inversion. In preparing the book, I broadened the scope of the
project by including all nonsubordinate finite clauses and analyzed a much
larger corpus, Genesis–2 Kings, with the help of electronic searches for vari-
ous syntactic constructions. The result of the syntactic analysis, presented in
chapter 5, is a taxonomy of marked and unmarked word-order constructions
in the finite clause. I also gave further attention to the fundamental question
whether VSO or SVO is the basic word order in Biblical Hebrew; the conclu-
sions are presented in chapter 2.
On reflection, it became clear that the pragmatic categories used in the dis-
sertation, based as they were to a large degree on discussions in the general
linguistic literature, artificially excluded many biblical clauses that intuitively
appear to belong to the same categories. A key insight was realizing that focus-
ing and topicalization are fundamentally different kinds of pragmatic functions,
not two variations of focus-background structure, as is often claimed (see §6.2
below). In this book, focusing and topicalization have been reconceptualized,
using insights from psycholinguistic research on text comprehension and from
linguistic research on discourse connectives. The newly defined concepts, I
believe, are more intuitively satisfying and represent a significant advance in
the understanding of marked word-order function in Biblical Hebrew and, po-
tentially, in other languages as well.
xv
00-Moshavi-LSAWS4.indb xv 9/28/10 7:36 AM

xvi Preface
In writing this book, I made considerable effort to present linguistic con-

cepts in clear language and keep cumbersome theoretical apparatuses to a
minimum. Multivariate statistical analysis has been omitted on the grounds that
it is unnecessary for the purposes of the analysis here. The discussion of the
literature has been completely reorganized, with the various approaches catego-
rized under three general models and related to the theoretical linguistic theories
that inform them.
An earlier formulation of part of chapter 7 appeared in “The Discourse
Functions of Object/Adverbial-Fronting in Biblical Hebrew,” in Biblical He-
brew in Its Northwest Semitic Setting: Typological and Historical Perspectives
(ed. Steven E. Fassberg and Avi Hurvitz; Publication of the Institute for Ad-
vanced Studies 1; Jerusalem: Magnes / Winona Lake, IN: Eisenbrauns, 2006),
231–45. An abbreviated version of chapter 8 appeared as “Focus Preposing
in Biblical Hebrew” in Leš 71 (2009): 35–55 (in Hebrew), and an abbreviated
version of chapter 9 appeared as “Topicalization in Biblical Hebrew” in Leš 69
(2007): 7–30 (in Hebrew).
The literature on word order in linguistic theory and word order in Bibli-
cal Hebrew, already quite considerable at the time of the completion of the
manuscript in 2006, has continued to grow. Studies appearing in late 2006 and
onward were available too late to be treated in any depth here. An example
is Robert D. Holmstedt’s 2009 study (“Word Order and Information Struc-
ture in Ruth and Jonah: A Generative-Typological Analysis,” JSS 54:111–39),
which presents a more developed pragmatic framework than his earlier stud-
ies of word order (see §§2.2.2–2.2.3 below). Word order in biblical poetry
has received a great deal of attention of late, no doubt stimulated in large part
by Nicholas P. Lunn’s 2006 monograph (Word-Order Variation in Biblical
Hebrew Poetry: Differentiating Pragmatics and Poetics; Paternoster Biblical
Monographs; Milton Keynes: Paternoster). The present work does not address
word order in poetry, but the concepts of focusing and topicalization developed
here can serve as an alternative to the Lambrechtian framework used by Lunn
(see §3.3.7, below).
It is a great pleasure to thank those who helped this project come to comple-
tion. I am deeply grateful to Richard C. Steiner, who has been my teacher and
mentor since I was an undergraduate. As his student, I discovered my love for
Hebrew studies and linguistics. Richard Steiner’s rigorous and innovative ap-
proach to language and textual study has been the inspiration for my research.
He has been a tireless source of support throughout the doctoral years and
beyond. I particularly thank him for his comments on the dissertation and on
the new manuscript.
I would like to thank the other members of my dissertation committee, El-
len F. Prince and Joshua Blau, for their assistance. Ellen Prince provided useful
guidance on linguistic issues, and Joshua Blau offered helpful comments. I am
00-Moshavi-LSAWS4.indb xvi 9/28/10 7:36 AM

Preface xvii
particularly grateful to the editors of this series, the late Michael P. O’Connor
and Cynthia L. Miller, for their detailed and extensive comments on the dis-
sertation. I would like to thank Amy Becker at Eisenbrauns for her careful
copyediting and proofreading of the final manuscript.
I had the opportunity to present some of my work to the international re-
search group on the subject of Biblical Hebrew in its northwest Semitic setting
at the Institute for Advanced Study at Hebrew University in 2001–2. I am
grateful to many members of the research group, including Steven E. Fassberg,
W. Randall Garr, Edward L. Greenstein, Jo Ann Hackett, John Huehnergard,
Avi Hurvitz, Jan Joosten, and Elisha Qimron, for their valuable comments. In
addition, I thank Elitzur Bar-Asher, Yochanan Breuer, Randall Buth, Vincent
DeCaen, Robert D. Holmstedt, Uri Mor, and Yael Ziv for their very helpful
comments on various drafts of my work. I also thank Moshe Bar-Asher and my
colleagues at the Hebrew University of Jerusalem and Bar Ilan University for
their encouragement and support.
Financial assistance came from several sources. I would like to thank the
Bernard Revel Graduate School of Yeshiva University for its generous assis-
tance. I am grateful for a doctoral scholarship from the Memorial Foundation
for Jewish Culture and for a doctoral dissertation fellowship from the National
Foundation for Jewish Culture. I would also like to thank the staff at the Ye-
shiva University libraries for their research assistance.
This book is dedicated to my beloved parents, Richard and Orah Mosak, to
whom I owe everything. They instilled in me a love of learning and encour-
aged me to follow my ambitions. The ideals they imbued in me have guided me
in every area of my life. I also dedicate this work to my husband and soul-mate,
Shimon, who has provided unflagging support for all my endeavors. Without
his help, it would have been unfeasible to immerse myself in academic pursuits
while simultaneously raising five young children. I am also grateful to Shimon
for proofreading and commenting on large portions of the manuscript.
My paternal grandparents, Jacob and Pearl Mosak, have been a source of
inspiration for me in so many ways. I am grateful to them for their love and
support. My late maternal grandparents, Abraham and Shalva Eliezri, were
also a source of much love and encouragement. Their passion for their ideals
made an indelible impression on me and helped to shape the course of my life.
I am grateful to my father- and mother-in-law, Baruch and Ettie Moshavi, for
their love and encouragement. Finally, I thank my children, Elisheva, Chaviva,
Tziona, Gavriel, and Zecharya, for tolerating the many hours I spent complet-
ing this work and, more importantly, for bringing me so much happiness.
00-Moshavi-LSAWS4.indb xvii 9/28/10 7:36 AM

00-Moshavi-FrontMatter.indd vi 9/28/10 2:05 PM

Chapter 1
Introduction
1.1. The pragmatics of preposing

Over the last 40 years, the study of word-order variation has become a
prominent and fruitful field of research. Researchers of linguistic typology
have found that every language permits a variety of word-order constructions,
with subject, verb, and objects occupying varying positions relative to each
other.1 It is frequently possible to classify one of the word orders as the basic
or unmarked order, and the others as marked.2 Converting a sentence from an
unmarked to a marked construction does not affect the semantic content of the
sentence; it does, however, restrict the conditions under which the sentence
could serve as an appropriate uterance. Marked syntactic constructions have
pragmatic meaning, that is, they encode aspects of meaning which are not se-
mantic but concern the relation of an utterance to its context.3 Put another way,
marked constructions provide an instruction to the addressee regarding the in-
terpretation of the utterance in its context. Marked syntactic constructions are
part of a larger group of linguistic forms with a pragmatic contribution to the
meaning of an utterance. The strides that have been made in understanding the
functions of these forms have highlighted the extent of the interaction between
pragmatics and language structure and have shown the importance of studying
language as a communicative system.
This work investigates word order in the finite non-subordinate clause4 in
classical BH. A common marked construction in this type of clause is the pre-
posing construction, in which a subject, object, or adverbial is placed before
the verb. In this work, preposing is formally distinguished from other marked
1. The term typology has several uses in linguistics; the intended meaning here is the
classification of structural language types and the study of language universals, that is, sys-
tematically occurring language patterns (Croft 2003: 1). Greenberg’s (1966a, 1966b) semi-
nal works form the foundation of this field. For a discussion of research on word order uni-
versals and their application to the Semitic language group, see O’Connor (1980: 115–18).
2. On the concept of markedness, see chap. 2, pp. 7ff.
3. For a discussion of pragmatics and its relation to semantics, see Levinson (1983:
1–35).
4. See chap. 4, pp. 48ff., for definitions of these and other syntactic terms relating to the
BH clause.

1.2. The corpus for the study 2
and unmarked constructions, and the distribution of these constructions in BH

is explored. A contextual analysis of a sample of preposed clauses is carried
out in order to determine the pragmatic functions that preposing may express.
My thesis is that the majority of preposed clauses can be classified as one of
two syntactic-pragmatic constructions: focusing or topicalization.
In studying the preposing construction, I hope to further our understanding
of the BH language system in use. I also hope that this study will contribute
indirectly to the cross-linguistic investigation of the relation between linguistic
form and pragmatic meaning.
1.2. The corpus for the study

The corpus for a linguistic study should be of a relatively homogeneous
nature and should be large enough to yield statistically significant results. The
criterion of homogeneity poses a challenge for the study of the Bible, which
contains books of various literary genres written over a long period of time.
The prose and poetic genres in the Bible have been shown to differ in vari-
ous linguistic characteristics. For example, the definite article ‫ה‬, the relative
particle ‫אשר‬, and the object marker ‫ את‬are often lacking in poetry (Freed-
man 1977; Andersen and Forbes 1983). According to O’Connor (1980: 401),
verb gapping occurs in biblical poetry but not in prose.5 More to the point,
poetry appears to exhibit a greater degree of word-order variation than does
poetry. For example, clauses with two constituents preceding the verb are rare
in prose. In poetry, this pattern occurs with greater frequency (Bloch 1946: 39;
O’Connor 1980: 334–35, 342–44).6
The present work is limited to the analysis of word order in prose, because,
as Joosten (2002) writes, “it is a general rule in research in morphosyntax
to take on poetic texts only when the prose rules have been approximately
established,” a rule equally apt with respect to syntactic-pragmatic research.
The exact border between prose and poetry is difficult to determine. Poetry is
usually identified based on an accumulation of poetic criteria, none of which
is exclusive to poetic texts (Freedman 1977: 6; Watson 1984: 46–57). The
identification of isolated poetic segments within a prose text is particularly
hazardous, because prose too may make use of poetic devices such as paral-
lelism and chiasmus.7 In this study, I follow BHS in categorizing passages as
5. On this point, however, see §9.3 n. 19 (p. 152 below). For a comprehensive study of
the syntactic characteristics of biblical poetry, see Sappan (1981). Syntactic studies that are
specifically limited to prose texts include Hoftijzer (1981, 1985), de Regt (1988), and Miller
(2003).
6. On word-order patterns in poetry, see O’Connor (1980: 297–355), Floor (2005),
Rosenbaum (1997), and Lunn (2006).
7. Berlin (1985: 4–16) writes that, though the essential feature of poetry is parallelism,
parallelism is also found in prose. The difference is that, in poetry, parallelism is the con-
structive principle of the text, while in prose it is not used systematically. A similar idea is

3 The corpus for the study 1.2.
prose or poetry, while recognizing that differences of opinion exist regarding

specific passages.
Another source of linguistic heterogeneity in the Bible is the diachronic di-
mension. Scholars distinguish three historical stages of BH: the archaic phase,
exhibited in certain poetic passages; the classical phase, dating from the Pre-
exilic Period; and the late phase, dating from the Exilic and Postexilic Period.8
The classical BH prose corpus consists of the five books of the Pentateuch and
the Former Prophets (Joshua–Kings). Late BH texts, such as Ezra–Nehemiah,
Chronicles, and Esther, betray the influence of Aramaic and other languages
and differ from classical texts in vocabulary, idiomatic constructions, and syn-
tax.9 The present work is restricted to word order in the classical dialect of BH,
that is, word order as manifested in the Genesis–2 Kings corpus.10 Although
source criticism assigns the origins of these books to sources dating from dif-
ferent historical periods, the language of this corpus “presents an astonishing
degree of uniformity,” particularly in the area of syntax ( Joüon-Muraoka §3a).11
The conclusions reached in this study are based on two corpuses, one con-
tained within the other. The syntactic analysis of finite non-subordinate word
order constructions, presented in chap. 5, is based on computerized searches of
the classical BH prose corpus, that is, the prose portions of Genesis–2 Kings.12
In addition, a statistical analysis of the word order constructions in the prose
portions of Genesis (henceforth referred to as the Genesis corpus) was
expressed in Jakobsen (1967). Greenstein (2000) notes that parallelism is often found in di-
rect speech in prose passages in the Bible and suggests that this preserves an ancient literary
convention. Kugel (1981) takes the radical view that the frequent presence of parallelism in
prose texts makes the prose/poetry dichotomy inaccurate and misleading as applied to Bibli-
cal Hebrew. For further discussion of the use of poetic devices in prose, see §8.5.
8. On this classification of the stages of BH, see, e.g., Hurvitz (1972, 1982); Kutscher
(1982: 12); Joüon-Muraoka (§3b); Rendsburg (1991); Steiner (1997: 146).
9. On late BH syntax, see, e.g., Kropat (1909); Hurvitz (1972, 1982); Polzin (1976);
Rooker (1990a, 1990b).
10. This corpus is the basis of linguistic studies such as van der Merwe (1990) and Miller
(2003). It might argue against the linguistic unity of this corpus that the traditional dating
of the Priestly source (P) in source criticism is in the Postexilic Period. Linguistic evidence
shows, however, that this dating of P is mistaken. Hurvitz (1982) finds that the linguistic fea-
tures of the texts assigned to P are compatible with those of classical prose texts, and exclu-
sively late features do not occur at all in P. According to Hurvitz, P predates Ezekiel, which
he views as the transition between classical and late BH. See also Rendsburg (1980). Literary
evidence for the preexilic dating of P is presented in Zevit (1982) and Friedman (1997).
11. Despite an overall picture of uniformity, some degree of variation is present in the
corpus. Variation is not always attributable to diachronic factors and sometimes can be given
a sociological interpretation, as Labov (1972) and others have shown. In addition, as noted
by Miller (2003: 27), a certain degree of variation is present even in a homogeneous speech
community.
12. The searches were carried out using the Bibloi 8.0 software package, published by
Silver Mountain Software.

1.2. The corpus for the study 4
performed (see §5.6). The selection of Genesis for the statistical analysis was
motivated by the fact that this book is a long classical work consisting almost
entirely of prose. Because the overwhelming majority of finite clauses in BH
prose are unmarked, a long text must be analyzed to yield a sufficient number
of marked clauses. The pragmatic analysis of the preposing construction, pre-
sented in chaps. 7–9, is based on a contextual examination of all of the finite
nonsubordinate preposed clauses in the prose portions of Genesis.
The assumption is made here that the present form of Genesis constitutes
a coherent text, regardless of its prehistory. Although Genesis is considered
by source critics to have been composed from more than one source, it is
widely recognized that the book, as it stands, is a unified literary work. Driver
(1913: 8) writes of a “unity of plan” in Genesis that “has long been recognized
by critics.” Skinner (1930: ii, lxv) considers Genesis to be a “complete and
well rounded whole,” despite its composite origin as he sees it. In his view,
the final redactor of Genesis combined several sources in a purposeful design,
giving the work a unity that is “the plan of one particular writer.”13
The text of Genesis is written in two different discourse registers: narrative
and direct-speech quotations.14 Direct-speech quotations in the Bible appear
to be written not in the spoken register actually used in the biblical period but
rather in a literary dialect (Rendsburg 1990: 19). We do not have a record of the
spoken language itself, but it most likely resembled Mishnaic Hebrew, a liter-
ary dialect believed to be closely related to the earlier spoken one.15 Although
direct speech in BH does not represent the spoken register, it is nonetheless
distinct from the narrative register. Both Macdonald (1975) and Mali (1983)
note the greater degree of word order variation in direct speech as compared
to narrative, and Macdonald (1975) notes additional syntactic differences of a
systematic nature.
Polak (1999, 2001, 2003) identifies a variety of styles of quoted speech in
BH, including a conversational style, a formal style, and styles that are com-
binations of the first two. The conversational style resembles spoken language
in having a high number of clauses with at most one verbal complement, a low
number of subordinate clauses, and a low number of long noun phrases. The
formal style contrasts in each of these features, having more than one verbal
13. Other scholars who find unified literary structures in Genesis include Childs (1979:
145–57), Rendsburg (1986), Sarna (1989: xvi), Fokkelman (1991), and Alter (1996: xlii).
14. On the definition of direct speech and its distinction from indirect speech (catego-
rized as narrative), see Miller (2003). The issue is addressed in the present work in §4.4.3.
Heller (2004: 25) also recognizes narrative and direct speech as two primary types in BH
prose but does not provide syntactic or other criteria for distinguishing the two. Heller dis-
tinguishes five types of direct speech, based in large part on Longacre’s (e.g., 1982, 2003)
discourse-type typology; these are narrative, predictive, expository, hortatory, and interroga-
tive discourse.
15. See, e.g., Steiner (1992: 21); Sáenz-Badillos (1993: 112–13).

5 Overview of the chapters 1.3.
complement, many subordinate clauses, and many long noun phrases. Polak
notes (2001: 63; 2003) that some BH narrative is written in a conversational
style, which he argues indicates origin from an oral narrative.16
In this study, I distinguish only two registers, narrative and direct-speech
quotations. Although Polak is undoubtedly right that we should recognize a
variety of direct speech (as well as narrative) registers, the Genesis corpus is
too small for a statistical comparison of all of the varieties. I will show that
even with this oversimplified taxonomy it is possible to identify differences in
the frequency and function of preposing in the two registers.
The text utilized in the present work is the Masoretic Text (MT), as exem-
plified by BHS. It is common practice in statistical studies of BH to base the
analysis on an attested text, normally the MT, rather than a reconstructed one.17
Regardless of the merits of textual criticism in any particular instance, it can be
argued that including emendations in the data introduces undesirable elements
of subjectivity and uncertainty to the statistical analysis. Although the notion
of the MT is admittedly an oversimplification, differences between Masoretic
manuscripts mostly concern vocalization, accentuation, or orthography, and as
such are not relevant to syntax or pragmatics.18 The same is true of the Qere
and Ketib alternations in the Masoretic text of Genesis, with one exception.19
1.3. Overview of the chapters

Chapter 2 discusses word-order markedness and justifies the view that the
basic word order in BH is verb first. A survey of the previous literature on the
pragmatics of preposing is presented in chap. 3. Chapter 4 sets out syntac-
tic definitions necessary for the selection and classification of the data in this
16. See also Polak (1999); he argues there that, in comparison to late BH, which is pri-
marily in a written style, classical BH narrative exhibits more of the informal traits charac-
teristic of spoken language.
17. Hurvitz’s (1982: 19) comments on this procedure are instructive:
This procedure is not followed out of an axiomatic belief in the supremacy of MT, nor
does it imply that it has reached us in exactly the same form in which it left the hands
of the ancient writers. . . . However, at the same time it seems to us that a linguistic
study whose central purpose is to seek facts and avoid conjectures, should base itself
on actual texts—difficult though they might be—rather than depend on reconstructed
texts. These latter are indeed free of difficulties and easy to work with; but we can
never be absolutely certain that they ever existed in reality.
See also Rendsburg (1990: 31–32); Miller (2003: 18).
18. More substantive differences can be found in some medieval manuscripts; e.g., the
word ‫‘ אחר‬after’ in Gen 22:13 has the variant ‫‘ אחד‬one’ in several manuscripts, a variant also
reflected in the versions. The significance of medieval variants is a matter of debate. Many
argue that these variants mostly originate in the medieval period and only coincidentally
reflect ancient variants; see discussion in Tov (1992: 37–39) and Revell (1992: 598).
19. The exception is Gen 30:10 Ketib ‫ בגד‬Qere ‫בא גד‬. The Ketib is probably a preposi-
tional phrase meaning ‘with luck’; the Qere is a complete clause meaning ‘luck has come’.

1.3. Overview of the chapters 6
study. Chapter 5 describes the syntactic varieties of the preposing construction

and distinguishes preposing from other similar word-order constructions. A
statistical analysis of word order in Genesis is also presented. Chapter 6 ex-
plores the concepts of focusing and topicalization and develops definitions to
be applied to the biblical corpus. Chapter 7 examines the various pragmatic
functions of the preposed clauses in Genesis and determines the frequency of
focusing and topicalization. Chapter 8 presents a comprehensive syntactic and
pragmatic analysis of focusing in BH. A comprehensive analysis of topicaliza-
tion is presented in chap. 9. Conclusions are presented in chap. 10.

Chapter 2
Word-Order Markedness
in Biblical Hebrew
The present work proceeds on the principle that verb-subject-object (VSO)

is the basic or unmarked order in BH. Marked constructions in which the sub-
ject, object, or adjunct precedes the verb are termed preposed. I use terms such
as preposing, moving, and the like as a convenient way of describing the corre-
spondence between marked and unmarked constructions, without specifically
implying a transformational analysis. 1
The VSO language group makes up about 10% of the languages in the
world (Carnie and Guilfoyle 2000: 3), including most of the West Semitic lan-
guages, Egyptian, Berber, Celtic, and other languages (O’Connor 1980: 118). 2
Because the classification of BH as a VSO language is not entirely uncon-
troversial, it deserves explicit justification. The present chapter examines the
arguments for and against basic VSO word order in BH. In §2.1, I examine the
concept of basic word order. In §2.2, I examine previous studies of basic word
order in BH. Conclusions are presented in §2.3.
2.1. Basic word order: Typological and generative perspectives

As noted in chap. 1, 3 language typology involves, among other things, the
classification of languages according to their structural types (Croft 2003: 1).
One of the parameters by which languages can be classified is basic word
order. “Basic” is often understood to mean the pragmatically unmarked or
neutral word order. 4 Of the several orders allowed by a particular language,
1. On the theory-neutral use of transformational terminology, see Birner and Ward

(1998: 3); Huddleston and Pullum (2002: 48).
2. Modern Hebrew, however, is an SVO language; see, e.g., Ravid (1977: 23, 38); Ber-
man (1980: 759–60 n. 1; 1997: 323–25).
3. See §1.1 n. 1 (p. 1).
4. The concept of markedness was first developed in the Prague school of linguistics and
referred to the member of a pair of contrasting phonemes that is found in neutral contexts.
Markedness was subsequently extended in a variety of ways to include disparate notions
not exactly comparable to each other. Various criteria are said to be evidence of markedness
in phonology, morphology, and syntax (Greenberg 1966b; Croft 2003: 87). The concept of
markedness is examined in depth in Battistella (1996).

2.1. Basic word order: Typological and generative perspectives 8
usually one order occurs in a wide variety of discourse contexts, whereas the
others have more restricted uses. The word order with a broader contextual dis-
tribution is the unmarked or basic order. Dryer (1995: 112) defines the concept
of pragmatic markedness as follows: “a construction is pragmatically marked
relative to another if the range of contexts in which it is appropriate is a proper
subset of the set of contexts in which the unmarked construction is used.” 5 Be-
cause the unmarked order is pragmatically acceptable in all contexts, the use
of the marked word order is always optional.
Basic word order is sometimes used to mean the statistically dominant order,
the one that is most frequent in spoken or written texts. 6 There is a widespread
assumption that the pragmatically neutral word order is also the most frequent.
According to Greenberg (1966b: 67), textual frequency is the only criterion by
which basic word order can be established. “Statistically dominant” is clearly
a less meaningful definition of basic word order than “pragmatically neutral,”
because frequency is a feature of language use rather than language structure.
In practice, however, researchers usually rely on textual frequency in establish-
ing basic word order, because proving that a particular order is pragmatically
neutral is an extremely involved procedure, requiring the identification and
classification of all discourse contexts in which each word order occurs.
Dryer (1995) notes several potential difficulties with relying on textual
frequency to determine basic word order. Word order frequency may vary in
different text types, raising the question of which text types most accurately
represent the frequencies of the various constructions. Longacre (1995: 333)
states that narrative texts are the most reliable, while Downing (1995: 20)
states that conversational texts are to be preferred. Others claim that conversa-
tion and narrative are both suitable as source data as long as the discourse is
oral, rather than written (Croft 2003: 111–12). The best approach would appear
to be to include a variety of text types in the analysis. A more serious problem
is that it is theoretically possible that a word order might be pragmatically
neutral yet not the most frequently occurring one; however, this appears to be
an atypical situation (Dryer 1995). Dryer concludes that “frequency may be a
useful diagnostic for pragmatic markedness, even if ultimately it is not a defin-
ing characteristic (1995: 116).
5. The concept of pragmatic markedness is a “privative” concept, as defined in O’Connor

(2002). O’Connor writes (2002:32–33) that “where a linguistic phenomenon is associated
with a privative opposition, the unmarked form is acceptable for the whole range of the phe-
nomenon,” and the “marked form is acceptable for a subset of the phenomenon.”
6. For discussion of textual frequency and other criteria used in determining basic word
order, see Brody (1984); Mithun (1992); Dryer (1995). Ziv (1988) examines a variety of cri-
teria for determining basic word order in the register of Modern Hebrew found in children’s
literature.

9 Basic word order: Typological and generative perspectives 2.1.
It is claimed that there are “free word-order” languages that do not have
a basic word order (Thompson 1978; Brody 1984; Comrie 1989: 88; Mithun
1992). In these languages, there is no one order that is statistically dominant
and pragmatically neutral as compared to other orders. In languages of this
sort, the word order in every sentence appears to be pragmatically motivated
(Thompson 1978, Mithun 1992).
A different approach to basic order defines the concept according to the
basic-sentence criterion. Siewierska (1988: 8) writes that “the term ‘basic or-
der’ is typically identified with the order that occurs in stylistically neutral,
independent, indicative clauses with full noun phrase (NP) participants, where
the subject is definite, agentive and human, the object is a definite seman-
tic patient, and the verb represents an action, not a state or event.” 7 Because
sentences satisfying all these requirements are uncommon, the basic-sentence
criterion results in the vast majority of naturally occurring sentences being
removed from consideration in determining unmarked word order. It is very
possible, therefore, that the basic-sentence criterion and the statistical criterion
will yield different unmarked word orders for a particular language.
An important point to be noted is the dependence of the basic-sentence
criterion on the criterion of pragmatic neutrality. Basic sentences may exhibit
more than one word order. To deal with this problem, Siewierska’s criterion
specifies that the unmarked order is the one found in “stylistically neutral” sen-
tences. In effect, then, the basic sentence criterion is essentially the pragmatic
criterion applied to a small subset of sentence types.
The typological conception of basic word order, whether based on prag-
matic neutrality, statistical dominance, basic sentences, or a combination of the
above, pertains to the surface structure of sentences. Generative grammarians
often use “basic word order” in an entirely different sense, meaning the order
that permits the simplest syntactic description of the language. The basic word
order is the structure from which all other structures are considered to be de-
rived. For example, McCawley (1970) argues that English is a VSO language
because this is its underlying word order: SVO sentences are the product of
a transformation inverting subject and verb in the underlying VSO clause. It
should be understood that this argument is irrelevant to the typological char-
acterization of English word order. The underlying word order in a language
is not necessarily the same as the pragmatically neutral / statistically dominant
order.
Subsequent works on word order by generative grammarians distinguish
between the underlying basic word order from a generative perspective and
the pragmatically neutral word order in surface structures. It is widely believed
that all VSO languages have an underlying SVO word order, with the verb-first
7. A similar definition can be found in Mallinson and Blake (1981: 125).

2.2. Basic word order in Biblical Hebrew 10
order produced by movement rules (Carnie and Guilfoyle 2000). This posi-
tion has not broken down the typological distinction between VSO and SVO
languages; on the contrary, current research seeks to explain why the underly-
ing SVO is transformed to VSO in neutral sentences in VSO languages. Both
typological and generative conceptions of basic word order have been invoked
in discussions of BH, as will be shown in §2.2.
2.2. Basic word order in Biblical Hebrew

Biblical scholars have long been aware that finite clauses in BH are most
frequently verb-first. 8 This fact was noted by the 19th-century biblical exe-
gete Malbim (Rabbi Meir Loeb ben Jeḥiel Michael, 1809–79). Malbim (1973:
§111) states that the general rule is that the sentence begins with the verb. In
his commentary on 1 Kgs 20:18, Malbim (1964: 209) explains this rule as de-
riving from the principle that the most important item comes first. The verb is
generally first because it is usually most important. A noun may be preposed
in order to specify something about the noun or in order to express contrast,
contradiction, or exclusion (Malbim 1973: §111). 9
Similar statements regarding word order can be found in the works of schol-
ars such as Ewald (1879: §306b), Müller (1888: §§130–31), König (1897:
§339a–e), Davidson (1901: §105), and GKC (§142a, f). In the view of these
scholars, word order is motivated by emphasis. The first element in every
clause is emphasized, being the newest or most important part of the sentence.
In the ordinary verb-first clause, the emphasis is on the action expressed by the
verb; if a different element, such as the subject, is emphasized, it will precede
the verb. Recent works such as Revell (1989a) and Shimasaki (2002) closely
resemble this approach; in their formulations, the element which is the focus of
attention is the one that occurs first (see §3.3.4 and §3.3.7).
Viewed from the perspective of language typology, the view of BH de-
scribed here comes very close to the modern concept of the free word-order
language, in which the order of every clause is pragmatically motivated. In
modern typology, however, languages are viewed as having free word order
8. The ensuing discussion concerns word order in the nonsubordinate clause. Word order
in subordinate clauses is investigated by de Regt (1991), although his definition of the sub-
ordinate clause differs from the one followed in this work (see §2.3). According to de Regt,
the percentage of verb-first subordinate clauses is even higher than verb-first nonsubordinate
clauses. Peretz (1967: 94) states that word order in the relative clause is usually verb-first, as
it is in the subordinate clause.
9. I am indebted to Richard Steiner for drawing my attention to the Malbim references.
Steiner (1998) notes that the medieval Jewish exegetes are divided on the significance of
word-order variation. Baḥye ben Asher (13th century), for example, argues that word order
can affect meaning (1966: 2:415), while Abraham ibn Ezra (1089–164) states that word or-
der variation is meaningless (commentary to Qoh 5:1).

11 Basic word order in Biblical Hebrew 2.2.
only when they do not exhibit a statistically dominant word order—a situation
not applicable to BH, as we will see.
Most twentieth-century scholars believe that BH does have a basic word
order and that a pragmatic explanation should be sought only for marked word
orders. According to the mainstream view, the basic order is VSO. Exponents
of basic VSO include Brockelmann (1956: §48), Meyer (1972: §91.2), Lam-
bert (1972: §1277), Givón (1977), 10 Bandstra (1982), Muraoka (1985), Waltke
and O’Connor (1990: §8.3), Jongeling (1991), Buth (1995) and de Regt (1991,
2006), among others. Dissenting from this camp and arguing that basic word
order is SVO are several researchers such as Joüon (1947) 11 and Schlesinger
(1953) and, more recently, DeCaen (1995, 1999) and Holmstedt (2002). In
§§2.2.1–2.2.3, I attempt to clarify the points of debate between the VSO and
SVO camps and justify the VSO view. In §2.2.4, I address whether word order
in BH varies according to discourse type, and in §2.2.5, I discuss word order in
the nonverbal and the participial clause.
2.2.1. The statistical dominance of VSO in Biblical Hebrew

Although not every scholar explains how he or she defines basic word or-
der, textual frequency is most frequently cited as evidence for basic VSO word
order. Some scholars, such as Meyer and Lambert, do not use terms such as ba-
sic or unmarked, writing simply that BH word order is “mostly” or “habitually”
verb first. In his revision of Joüon (1947), Muraoka explicitly links markedness
and frequency, writing that “the statistically dominant and unmarked word or-
der in the verbal clause is: Verb-Subject” ( Joüon-Muraoka §155k).
According to Jongeling (1991), a statistical analysis of the book of Ruth
shows that the verb precedes in 87% of finite clauses (including subordinate
ones). In this work, I present a count of nonsubordinate finite clauses in the
10. Givón (1977) states that classical BH had a basic word order of VSO and gradually
shifted to SVO in the Exilic Period. He compares the classical texts of Genesis and 2 Kings
to four books usually held to be late, namely, Esther, Lamentations, Ecclesiastes, and the
Song of Songs. These texts, as Givón admits, differ greatly from the classical texts as well
as from each other in terms of discourse genre and content. Esther is prose narrative, like
Genesis and 2 Kings, while Ecclesiastes belongs to the wisdom-literature genre, and Lam-
entations and the Song of Songs are poetry. As discussed in §1.2, poetic texts often exhibit
a higher frequency of marked word order than prose. Givón’s selection of these four texts as
representative of late BH is particularly puzzling considering the existence of demonstrably
late texts of a much more similar genre to Genesis and Kings, that is, Ezra–Nehemiah and
Chronicles. Contrary to Givón’s conclusion, these books appear to exhibit unmarked VSO
word order, as noted by Buth (1995: 91 n. 11). Additional arguments against Givón’s posi-
tion can be found in Buth (1995: 91).
11. In the revised edition of Joüon by Muraoka ( Joüon-Muraoka §155k), Joüon’s (1947:
§155k) original statement, “L’ordre des mots dans la proposition verbale . . . est normale-
ment: Suject-Verbe” is revised to reflect Muraoka’s support for the standard VSO view (see
§2.2.1.).

book of Genesis, a much larger corpus. The results are similar to Jongeling’s,
with 84% of verbal clauses having nonpreposed word order (see chap. 5). It is
sometimes claimed that figures such as these are misleading because many of
the clauses involved lack an explicit subject and are not really VSO. However,
because BH drops pronominal subjects unless they precede the verb (see §5.1),
one can argue that VO clauses can be classified together with VSO clauses.
The case for this position is particularly strong for consecutive forms, where
the pronoun can only have been dropped from the postverbal position.
In any case, even if only clauses with subjects are included, the evidence
still seems to point to VSO. Jongeling finds that when an explicit subject is
present, the subject precedes the verb in only 20% of the clauses. Hornkohl’s
(2003: 7) M.A. thesis on Genesis reaches a similar conclusion, also including
subordinate clauses in the data: 79% of verbal clauses with an explicit subject
are verb first. 12 The natural conclusion is that the unmarked position for both
subject and object is after the verb.
Some scholars (e.g., Schlesinger 1953; Muraoka 1985: 29; Regt 1991: 160)
are of the opinion that sentences with a waw-consecutive verb form (wayyiqtol/
weqatal) should not be considered in determining basic word order. A reason
given for excluding this type of clause is that the subject necessarily follows
the verb in these clauses (Muraoka 1985: 28). 13 Fleshing this statement out, the
argument would seem to go something like this: the consecutive form is not
just a “positional variant” (Revell 1989b: 32) of the simple form but has se-
mantic or pragmatic significance of its own. A sentence involving this seman-
tic/pragmatic factor (let us call it x) necessarily contains a consecutive form
and, ipso facto, verb-first word order; only clauses lacking this factor exhibit
word-order variation. According to this logic, including consecutive forms in
the data inappropriately skews the results in favor of VSO word order.
What exactly x represents is unclear; but it certainly is not the most obvious
candidate, temporal sequentiality. As will be discussed in §3.2, “consecutive”
forms are not always sequential, nor must a clause relating a sequential event
contain a consecutive form. A simpler explanation of the consecutive/simple-
verb distribution is that verb form is conditioned on word order. The consecu-
tive form is a positional variant of the simple form that is used whenever the
verb immediately follows a conjunction. If any element intervenes between
conjunction and verb, whether a negative particle, clausal adverb, or preposed
constituent, a simple form is selected. As Blau (1993: §60) writes, 14
12. This figure was obtained by combining the figures for verb-first order in narrative
and direct speech, calculated separately by Hornkohl.
13. DeCaen (1995, 1999) and Holmstedt (2002, 2005) argue for the omission of con-
secutive forms on other grounds; for discussion, see §2.2.2.
14. See also Bergsträsser (1962: §6c); Revell (1989b: 3).

The tenses with consecutive waw (wayyqtl, weqtl ) are used whenever the syntac-
tic environment permits the use of waw copulative; otherwise the simple tenses
(qtl, yqtl ) [are used].
It follows from this that eliminating consecutive forms from a text count is
statistically invalid, skewing the results in favor of SVO by eliminating many
VSO clauses from the data. Most clauses in BH start with the conjunction,
particularly in narrative texts, but to a lesser extent in direct speech as well. I
found that in Genesis more than 99% of clauses in narrative have a conjunc-
tion, and 65% in direct speech. The vast majority of VSO clauses (excluding
those containing preverbal particles), then, will necessarily have a consecutive
form and would not be counted.
In his examination of word order, Muraoka (1985: 30–31) excludes consec-
utive forms and restricts his count to a sample of conversational texts, in which
the frequency of SVO word order is significantly higher than in narrative texts.
Although he concludes that VSO order is nonetheless statistically dominant,
this result is not directly relevant to the present study because in Muraoka’s
study all orders in which the verb precedes the subject, including object-first
(OVS) and adjunct-first (AVS) orders, are counted as having “normal” (that is,
relative-VS) order. Hornkohl (2003: 8–10) examines the issue from a different
angle, reasoning that, if BH is in fact an SVO language and the predominance
of VSO is only due to the widespread use of consecutive forms, one would
expect to find SVO as the dominant order in clauses in which consecutive
forms are grammatically constrained from appearing due to the presence of an
initial particle or subordinating conjunction at the head of the clause. Hornkohl
counts all of these clauses in Genesis and shows that, contrary to the SVO hy-
pothesis, VSO is the most frequent word order: 76% of narrative clauses and
58% of direct-speech clauses are verb-first, with an overall frequency of 66%
verb-first word order. 15 It can be concluded that the textual-frequency criterion,
however it is defined, leads to the firm conclusion that basic word order in the
BH finite clause is VSO. 16
2.2.2. Word order in the basic sentence

DeCaen (1995, 1999) has adopted Siewierska’s basic-sentence criterion for
the typological classification of BH, arguing the statistical-dominance criterion
leads to a “naïve and theoretically uninteresting” conception of basic word
order (1999: 118 n. 22). 17 According to DeCaen (1995: 137; 1999: 118 n. 22)
15. I have calculated these percentages using Hornkohl’s raw statistics. The percentages
given by Hornkohl (2003: 8) differ somewhat because he eliminates clauses without an overt
subject.
16. On word order in the nonverbal and participial clause, see §2.2.4.
17. DeCaen’s main interest, however, is in basic word order from the generative per-
spective, as will be discussed in §2.2.3.

the basic-sentence criterion leads to the rejection of VSO as the basic order for
BH. Coordinated clauses are not basic sentences and are excluded from con-
sideration; thus, clauses with consecutive forms, the most common VSO type,
are not considered in determining basic word order. An additional category
that is omitted in the basic-sentence criterion is the modal clause. Citing Nic-
cacci (1987) and Revell (1989b), DeCaen (1995: 24) asserts that clauses with
modal imperfects, jussives, or imperatives have VSO order, whereas clauses
with indicative imperfects have SVO order. 18 Because only indicative clauses
are basic sentences, it follows that SVO is the basic word order. More precisely,
DeCaen’s conclusion is that BH is a verb-second (V2) language, like German,
which requires the verb to be in second place in the clause, whether the first
element is the subject or a different element (DeCaen 1995: 24). 19
DeCaen explains the verb-initial (V1) order characteristic of modal clauses
by noting that the modal category is marked relative to the indicative category.
According to DeCaen (1995: 111–22), the verb form found in wayyiqtol is
actually a modal form; in addition, he posits that the doubling of the initial
prefix represents a complementizer signifying modality (1995: 128–229, 296).
The modal nature of wayyiqtol explains why these forms always have V1 or-
der. Although this interpretation of the consecutive form is certainly open to
challenge, the more fundamental point for present purposes is the connection
drawn between modality and marked word order. Modality is a semantic fea-
ture of the verb that does not vary with discourse context and is independent
of pragmatic neutrality. Although modal verbs are marked in comparison to
indcative verbs, this has nothing to do with whether clauses with modal verbs
are pragmatically marked. In other words, the fact that modal clauses are gen-
erally VSO is perfectly compatible with the VSO basic word-order hypothesis.
What is more critical to the V2 argument is whether indicative clauses are in
fact always or most often SVO.
The latter issue is addressed by Holmstedt (2002). Holmstedt’s (2002: 139)
count of indicative clauses in Genesis (including subordinate clauses but
excluding clauses without overt subjects and excluding wayyiqtol clauses) 20
shows that a slim majority actually have VSO order, with 303 out of 554 clauses
(55%) being VSO. Holmstedt, however, argues that many of these clauses,
such as the subordinate ones, should not be counted because they do not sat-
18. The claim regarding indicative clauses, of course, does not include the most com-
mon indicative clause type, the clause with a consecutive form. For further discussion of this
view, see §3.3.4.
19. According to DeCaen, BH is V1 in subordinate clauses, which typically have verb-
first order.
20. Holmstedt excludes clauses with wayyiqtol due to their “unique morphological char-
acteristics” (2002: 133); see §2.2.3 for his explanation of why clauses with wayyiqtol always
have VSO order.

isfy the basic-sentence criterion. 21 Holmstedt proposes that negative clauses

should be omitted as well because, he asserts, they are closely related to modal
clauses (2005: 147 n. 36). The basic sentence criterion also excludes sentences
that are not stylistically neutral, as Holmstedt points out. But how does one
go about deciding which clauses are neutral, if it is not already known which
order is marked? Holmstedt (2005: 148–49) points out that clauses occurring
in certain contexts, specifically contrastive ones, are likely to be pragmatically
marked; 22 these clauses are therefore not taken into account. Omitting all of
these categories yields a majority of SVO, albeit from a much smaller group of
data (175 clauses in all).
A difficulty with these results lies in the procedure used to weed out the
pragmatically marked clauses. Although contrast is one kind of context likely
to contain a pragmatically marked clause, there are many types of pragmatic
functions that do not involve contrast, including the noncontrastive varieties
of focusing and topicalization (see chaps. 8, pp. 121ff., and 9, pp. 144ff.), as
well as other less-common functions such as marking simultaneity and anteri-
ority or marking the onset of a new narrative unit (see chaps. 3, pp. 18ff., and
7, pp. 104ff.). If clauses appearing to have any of these functions are omitted
as well, the total number of clauses in the data is drastically reduced, leaving
only 65 clauses by my estimate. 23 While it appears that SVO is still the major-
ity word order among the remaining clauses, there is no way to know whether
more clauses need to be omitted.
Holmstedt’s statistics highlight the drastic effect of the basic sentence cri-
terion on text counts. Out of thousands of finite nonsubordinate clauses with
overt subjects in Genesis (including wayyiqtol), only 175 remain according to
Holmstedt’s calculations, and considerably less once more potentially marked
clauses are omitted. Determining the word order of a language on the basis
of such a small sample seems somewhat precarious. More importantly, basic
word order in this approach bears little resemblance to the way the language is
most frequently used.
21. In the following summary I rely on Holmstedt (2005), which is the clearest presenta-
tion of his methodology, although the corpus analyzed there is Proverbs.
22. It can be objected that clauses in a contrastive context need not be marked, even if
BH uses word order to mark contrast: the use of a marked construction is always optional,
because the neutral word order is always available. It is not known how often pragmatically
marked constructions are used in contexts that permit these constructions.
23. The effect is particularly extreme on narrative indicative clauses, where 84 out of the
103 must be omitted. When an additional clause in the data is omitted for technical reasons,
the result is 15 SV clauses out of 18 (83%). In direct speech, omitting the potentially marked
clauses as well as a few found in poetic passages and several others disqualified on syntactic
grounds yields 28 SV clauses out of 47, or 60%.

2.2.3. The generative perspective on basic word order in Biblical Hebrew

DeCaen (1995, 1999) argues that the basic word order in BH from the genera-
tive perspective is SVO. Working within the framework of the widely accepted
generative-grammar theory known as Government-Binding (GB) theory, De-
Caen (1999: 118 n. 2) finds that GB “virtually dictates an underlying SVO for
Hebrew.” 24 This analysis of BH fits the current consensus, mentioned above,
that all languages have SVO basic word order from the generative perspective.
In DeCaen’s view, this underlying SVO order in BH becomes typological V2
on the level of surface structure. Although DeCaen rejects the majority VSO
view in his typological classification of BH, this position is not the logical con-
sequence of assuming SVO underlying word order; rather, it results from his
utilization of the basic sentence criterion. In other words, DeCaen’s findings
regarding the generative analysis of BH do not inherently contradict the major-
ity view that BH belongs typologically to the VSO language group.
Holmstedt’s (2002, 2005) work is set in a “generative-typological” frame-
work, which starts with the generative premise that the underlying structure
in every language is SVO. The main focus of his approach, however, is the
investigation of the basic order of surface structures, that is, in the typological
classification of BH as an SVO language. Holmstedt accounts for the statistical
dominance of VSO by a generative concept known as “triggered inversion.”
He hypothesizes that inversion to VSO occurs when a subordinating or func-
tion particle, negative particle, or interrogative particle precedes the subject
and verb. He further posits that inversion is also triggered by modality, ac-
counting for the majority of modal clauses with VSO order. The VSO order in
wayyiqtol clauses is explained as triggered by a historical “complementizer”
form contained within the consecutive verb and preserved as the doubling of
the verbal prefix. Pragmatically motivated preposing rules are also recognized.
It should be noted that the concept of triggered inversion does not serve as
evidence for typologically basic SVO; rather, it is a mechanism that accounts
for the existence of VSO clauses in Holmstedt’s SVO typological scheme. A
similar mechanism can be invoked by the majority view taking BH to be VSO:
the basic word order of surface structures can be accounted for by a manda-
tory (rather than a triggered) inversion rule to VSO, and SVO clauses can be
explained as the product of pragmatically motivated preposing rules.
24. Buth (1995: 80–81) argues against an underlying SVO, pointing out that SVO lan-
guages such as English typically have OSV as the marked construction, derived by moving
the object to initial position. In BH, OSV word order is almost unheard of; the orders that
appear are SVO and OVS. Although this argument seems highly attractive, the generative
grammarian can account for the discrepancy between English and BH by positing minor dif-
ferences in the underlying SVO structure of the two languages.

17 Conclusion 2.3.
2.2.4. Word order and discourse type

Studies of word order typology generally characterize languages as having a
single basic word order. Longacre (1982), however, has claimed that verb-first
languages such as BH typically exhibit different unmarked orders in certain
discourse types. In BH, he claims, expository discourse exhibits basic SVO
word order. What Longacre actually means by this, as he explains there, is that
nonverbal clauses predominate in expository discourse, and these clauses most
typically have subject-predicate word order, which he takes to be comparable
to SVO (Longacre 1982: 472; 1989a: 111). Thus, it is not actually claimed that
there is a discourse type in which SVO is the unmarked order for finite clauses.
Word order in the nonverbal clause is addressed in the next section.
2.2.5. Word order in the nonverbal and the participial clause

The basic word order in BH nonverbal clauses appears to be subject-
predicate (e.g., Davidson 1901: §103; GKC §141l), in contrast to basic VSO
in finite clauses. Muraoka (1985: 8–9) shows that subject-predicate word or-
der is the most common one, based on a statistical analysis of representative
samples of conversational, narrative, and legal texts. According to Andersen
(1970), the word order depends on whether the clause is “identifying” or “clas-
sifying”: identifying clauses are usually subject-predicate, while classifying
clauses usually have the reverse order. An entirely different approach is taken
by Joosten (1989), who argues that the SP/PS opposition is one of aspect rather
than pragmatic function. Further discussion of nonverbal clause word order
can be found in Rosén (1965); Blau’s (1973b) review of Andersen (1970);
Hoftijzer’s (1973) review essay on Andersen (1970); Revell (1989a, 1999);
Muraoka (1990, 1991); Zewi (1992, 1994); Buth (1999); and DeCaen (1999),
among other studies.
It appears that participial predicates in BH are like nonverbal predicates,
generally occurring after the subject (see, e.g., Andersen 1970; Muraoka 1985;
Joosten 1989; Buth 1999). The present study is restricted to finite clauses, so I
will not explore word order in nonverbal and participial clauses further.
2.3. Conclusion
In conclusion, basic word order in the typological sense is the pragmatically
unmarked order. Basic word order is usually established by the criterion of
statistical dominance. The mainstream view that BH is typologically VSO is
strongly supported by the statistical evidence. BH, like other VSO languages,
has an underlying SVO word order from the generative perspective, but this
fact does not affect the typological classification of the language. Basic word
order in BH does not vary according to discourse genre.

Chapter 3
Previous Studies of the Functions of

Preposing in Biblical Hebrew
This chapter reviews prior studies of the function of preposing in BH. Given
that VSO is the unmarked word order in BH, preposed word orders such as
SVO and OVS can be assumed to have pragmatic functions. The pragmatic
significance of postverbal word order is outside the scope of this study and is
not addressed here. 1
Three general models of preposing can be identified in the literature, each
involving a central concept that motivates all or most preposing: emphasis,
backgrounding, and information structure. The three concepts do not all have
the same scope. Emphasis is not thought to explain all preposed clauses, while
universal scope is sometimes claimed for backgrounding. There is a multitude
of variations of each model, and the approaches of some researchers do not
fit neatly into a single model. Nevertheless, I believe that grouping all of the
approaches into three categories is helpful in making sense of the bewildering
variety of explanations that have been offered for preposing.
3.1. The emphasis-centered model

As discussed in §2.2, according to one school of thought, the first element
in every clause is emphasized, being the newest or most important part of the
sentence (Ewald 1879: §306b; Müller 1888: §§130–31; König 1897: §339a–e;
Davidson 1901: §105; GKC §142a, f; Revell 1989a; Shimasaki 2002). For
these scholars, BH is essentially a free word-order language. A different school
of thought takes VSO as the basic word order, with preposing motivated by
emphasis. Adherents of the latter view include Brockelmann (1956: §48),
Lambert (1972: §1277), Meyer (1972: §91.2), and Muraoka (1985). Thus, Mu-
raoka (1985: 30) writes that in the unmarked VSO word order “neither S nor V
receives special emphasis.”
The concept of emphasis has been criticized as overly subjective and vague.
It is difficult to say whether an emphasis on the preposed element was really
1. Studies discussing postverbal word order in BH include Lode (1984, 1988); Rosen-
baum (1997); Heimerdinger (1999).
18

19 The emphasis-centered model 3.1.
intended or whether the researcher is simply assuming it to exist because of

that element’s position at the head of the clause. In addition, the notion of em-
phasis in and of itself does not explain why and for what purposes the speaker
wishes to emphasize something. Muraoka (1985: i) writes that the term em-
phasis is often used “without much thought being given to precisely what is
meant by the term, nor, more importantly, to the question why the writer or
speaker possibly felt the need for an emphatic form or construction.” 2 Muraoka
attempts to clarify the concept of emphasis, stating that it is a psychological
function that is often accompanied by “intensified emotion” (1985: xiv).
Most scholars who discuss emphasis view it as the dominant but not the only
function of preposing. Muraoka (1985: 32) writes that emphasis is “what seems
to lead to the reverse order in a good number of places, but there remain many
others where one can hardly perceive any emphasis.” For Niccacci (1990: 20),
emphasis (under the name “foregrounding”) constitutes only one of three axes
influencing word order, as explained further in §3.2.
In §§3.1.1–3.1.7, I examine additional functions that have been cited to
explain preposing in the emphasis-centered model. As discussed in these
sections, there is an effort made by certain scholars to link some of these func-
tions, specifically, contrast and circumstantiality, to the concept of emphasis.
3.1.1. Contrast and contrastive structures

Contrast is widely cited as a function of preposing. The exact meaning of
this concept is a difficult problem. Andersen distinguishes between “contrast,”
which compares “the participants in two parallel but in some ways different
activities” (1974: 150), and “antithesis,” which involves not only contrast but
also “contradiction or opposition” (1974: 179). 3 Muraoka (1985: 54, 59) posits
two apparently identical categories, substituting the term juxtaposition for con-
trast. 4 I adopt the terms juxtaposition and antithesis in the ensuing discussion,
with contrast used as an umbrella term encompassing both.
Several scholars note that two distinctive structures are used to achieve con-
trast: “chiastic” structures involving a normal clause followed by a preposed
one and “parallel” structures involving a pair of preposed clauses. 5 The two
patterns are illustrated in (1) and (2), respectively.
2. See also Bandstra (1982: 79), van der Merwe (1989).

3. See also Myhill and Xing (1993), who define contrast as a pair of sentences whose
predicates have opposite meanings and whose contrasted nouns are members of a set. Con-
trast in this sense corresponds to Andersen’s antithesis.
4. The use of preposing for contrast is also discussed in Steiner (1997: 166; 2000: 259–
60).
5. See also Steiner (1997: 166). More than one scholar has noted a resemblance between
contrastive chiastic structures and poetic chiasmus; see, e.g., Sappan (1976); Khan (1988:
89); Muraoka (1985: 36). It should be noted, however, that prose chiasmus is more restricted
in form than the poetic type. Poetry exhibits a wide variety of chiastic patterns (O’Connor

3.1. The emphasis-centered model 20
(1) Gen 4:4–5

‫ ואל קין ואל מנחתו לא שעה‬:‫ אל הבל ואל מנחתו‬′‫וישע י‬
And the LORD paid heed to Abel and his offering; and for Cain and his
offering He paid no heed.
(2) Gen 41:13

:‫ויהי כאשר פתר לנו כן היה אתי השיב על כני ואתו תלה‬
And as he interpreted to us, so it came to pass; me he restored to my of-
fice, and him he hanged.
According to Andersen (1974), the parallel structure is used for juxtaposition,

and the chiastic structure may express antithesis. A separate use of the chiastic
structure according to Andersen is to stress the similarity between two situa-
tions. According to Khan (1988: 88–90), both parallel and chiastic structures
may set up a relation of contrast or similarity.
Contrastive chiastic structures often involve the same verb in both clauses,
as in (1). The connection between repeated verbs and chiasm is discussed by
Driver (1913: 130), Ginsberg (1942: 230), Gordis (1944: 150), Cassuto (1961a:
27; 1961b: 91–92), Ben David (1971: 856), McEvenue (1971:43; 1974), Blau
(1972), Andersen (1974: 46), and Paran (1989). Paran considers chiastic struc-
tures with repeated verbs to be a stylistic device, not necessarily contrastive,
which he calls the “circular inclusio.” 6
There is some disagreement as to whether contrast is a separate category
from emphasis. Ewald (1879: §309a) refers to emphasis and contrast as two
separate functions of preposing, 7 while Davidson (1901: §105) states that em-
phasis usually involves “some degree of antithesis, latent or expressed.” Mu-
raoka (1985: 54) speaks of preposing as expressing emphasis or contrast but
also refers to contrast as “one aspect” of emphasis.
3.1.2. Circumstantiality, anteriority, and simultaneity

Circumstantiality is another concept frequently mentioned alongside empha-
sis in connection with preposing. According to GKC (§156a), the circumstantial
clause concerns “the particular circumstances under which a subject appears as
performing some action, or under which an action . . . is accomplished.” 8 The
circumstantial clause is most frequently a participial or nonverbal clause with
subject-predicate word order: 9
1980: 391–400), while prose chiasmus is restricted to the VSO-SVO/OVS type. On the use
of stylistic devices such as chiasmus in prose, see §8.5.
6. See §8.5 for further discussion of repeated-verb structures.
7. See also Malbim (1973: §§111–14).
8. See also GKC (§§141e, 142a); Ewald (1879: §306c); Müller (1888: §§131, 151);
Driver (1892: §156); Davidson (1901: §150).
9. According to Joosten (2002), there are more than 170 examples of circumstantial claus-
es with a predicative participle in Genesis, Josh 1–10, Judges, 1–2 Samuel, and 1–2 Kings;

(3) Gen 18:1

:‫ באלני ממרא והוא ישב פתח האהל כחם היום‬′‫וירא אליו י‬
And the LORD appeared to him by the oaks of Mamre, and (= while) he
was sitting at the entrance of the tent in the heat of the day.
In the finite circumstantial clause, the subject precedes the verb, mimicking the
subject-predicate order found in the nonfinite circumstantial clause.
Several scholars, including Ewald (1879: §306c), Müller (1888: §§130–
31), and Davidson (1901: §105), link word order in the finite circumstantial
clause to the emphasis principle. Because the circumstantial clause describes a
state rather than an action, the verb is de-emphasized relative to the subject and
hence follows it. S. R. Driver (1892: §160 obs.) and Muraoka (1985: 33), how-
ever, see emphasis and circumstantiality as separate categories. Driver (1892:
§157) and GKC (§141e) link circumstantiality to contrast.
In the classic sense, the circumstantial clause is one that refers to a con-
comitant state or event, that is, an accompanying state or event simultaneous
with the previous clause. The circumstantial clause is viewed as semantically
subordinate to the preceding clause, despite being coordinate on the gram-
matical level. A nonfinite example is (3), above, which can be rendered by a
subordinate clause, ‘while he was sitting’. 10 Oddly, though, it is hard to find
finite clauses of this type. Joosten (2002) points out that a finite clause describ-
ing a concomitant event should contain an imperfect verb; as he shows, there
are hardly any clauses of this type. Although a number of concomitant finite
clauses with perfects are cited in the literature, most of these are not found in
a narrative context but occur in direct speech, following a rhetorical question
or a request: 11
(4) Gen 18:13
:‫למה זה צחקה שׂרה לאמר האף אמנם אלד ואני זקנתי‬
Why did Sarah laugh, saying, “Shall I indeed bear a child, and I am old?”
I will argue in §7.3.2.1 that this type of preposing does not mark circumstan-
tiality but is a special construction that marks the justification for a preceding
utterance. All in all, it would appear that the finite circumstantial clause in the
classic sense is something of a phantom. 12
see there for further citations.

10. See Müller (1888: §151); Driver (1892: §156); Davidson (1901: §137). Eskhult
(1990: 31), who is an advocate of the backgrounding approach (see §3.2), compares the
circumstantial clause to the Arabic ḥal clause, which describes a circumstance under which
a subject performs an action.
11. Other examples of this type cited in the literature are Gen 24:31 and 26:27.
12. See further on in this section (p. 23) for discussion of supposed circumstantial claus-
es that precede rather than follow the main clause.

A number of cited examples of finite circumstantial clauses describe events

anterior to the main clause event; for example, 1 Kgs 1:41 and Judg 16:31.
In order to view examples such as these as circumstantial, it is necessary to
broaden the circumstantiality category to include anterior events as well as
the classical simultaneous type. Considering that finite simultaneous circum-
stantial clauses are hardly to be found, however, this sort of definition seems
artificial. Andersen (1974) and Muraoka (1985) broaden the concept of cir-
cumstantiality even further to include parenthetical comments and clauses that
interrupt the chain of the narrative. This conception comes very close to the
idea of “backgrounding,” as discussed in §3.2, below. Kotzé (1989) finds the
concept of circumstantiality generally wanting, showing that the category is not
consistently understood in the literature and cannot be characterized in terms
of particular syntactic features. He concludes that it is used as a catch-all term.
A more plausible approach is to abandon circumstantiality as a function of
preposing in the finite clause and to view anteriority as the relevant pragmatic
function. The use of preposing to mark anteriority is an unmistakable phenom-
enon, recognized as early as the Jewish exegete Rashi (Rabbi Solomon ben
Isaac, 1040–1105 C.E.). Rashi comments on ‫‘ והאדם ידע את חוה אשתו‬And the
man knew Eve his wife’ (Gen 4:1): 13
‫ שאם כתב‬,‫ וכן ההריון והלדה‬,‫כבר קודם הענין של מעלה קודם שחטא ונטרד מגן עדן‬
‫”וידע אדם“ נשמע שלאחר שנטרד היו לו בנים‬
Already before the above matter, before he sinned and was banished from the
garden of Eden, and so too the pregnancy and birth. For if it had written “ ‫וידע‬
‫ ” אדם‬it would have meant that after he was banished he had children.
Malbim (1973: §117) formulates Rashi’s observation as a general rule: “When

[the text] speaks about a matter that already took place prior to that time, it
always advances the subject ahead of the verb.” Several scholars (e.g., König
1897: §§115–17; GKC §106f; Bergsträsser 1962: §6d) mention anteriority as
one of the uses of the perfect verb form. According to some (Müller 1882:
§152; Driver 1892: §§76 obs., 160 obs.; GKC §142b), it is the preposing of
the subject that conveys anteriority; the anterior clause has a perfect verb as
a result of its SV word order. More recent discussions of preposing to mark
anteriority include the works of Williams (1976: §573), Givón (1977, 1983),
Kutscher (1982: §66), 14 Fox (1983), and Zevit (1998). 15
13. Text from Chavel (1983).

14. Zevit (1998: 13 n. 18) cites Shraga Assif as stating that Kutscher was teaching this
theory to students as early as the late 1960s; in addition, he cites a personal communication
from Menaḥem Z. Kaddari that in the 1940s Neḥama Leibowitz was circulating collections
of examples of this phenomenon discussed by the medieval Jewish exegetes.
15. Zevit (1998: 15) has an unusual understanding of anteriority. He defines anteriority
as including the pluperfect (past perfect) as well as the preperfect, which refers to an event
that had “commenced but not necessarily terminated in the past prior to the beginning of

A separate phenomenon is the use of a pair of preposed clauses to mark

simultaneity. Davidson (1901: §141) and GKC (§141e) note the use of prepos-
ing in clauses such as (5):
(5) Gen 19:23
:‫השמש יצא על הארץ ולוט בא צערה‬
The sun rose upon the earth, and Lot came to Zoar.
Davidson and GKC explain this as an atypical kind of “circumstantial” clause
that precedes the main clause; thus, (5) can be rendered ‘As the sun rose, Lot
came to Zoar’. Given the arguments already presented against circumstantial-
ity as a function of preposing, a different explanation of such examples is desir-
able. A notable feature is the use of preposing in the “main” clause as well as
the “circumstantial” clause, something not seen with the classic circumstantial
clause. In addition, in some cases it is not clear that one event is subordinate
to the other, as in (6):
(6) 1 Sam 9:17

:‫ ענהו הנה האיש אשר אמרתי אליך זה יעצר בעמי‬′‫ושמואל ראה את שאול וי‬
And Samuel saw Saul, and God told him, “Here is the man that I told
you, “This one will govern my people.”
In Joüon’s (1947: §166) view, preposing in these cases marks a temporal

relation of simultaneity between two clauses, rather than a relation of seman-
tic subordination. 16 The structures of both clauses contribute to specifying the
type of simultaneity involved. A sequence of two preposed clauses with per-
fects, such as (5) and (6), indicates two simultaneous instantaneous actions. A
subject-first participial clause followed by a preposed clause with the perfect,
such as (7), indicates a durative action and a simultaneous instantaneous action.
(7) Gen 38:25

‫הוא מוצאת והיא שלחה אל חמיה לאמר לאיש אשר אלה לו אנכי הרה‬
She was being brought out, and she sent to her father-in-law, saying, “By
the man to whom these belong, I am with child.”
When two instantaneous events are concerned, it may be difficult to determine

from the context whether the events are truly simultaneous or one immedi-
ately follows the other. 17 It is plausible, however, that simultaneity and near-
simultaneity constitute a single category from the pragmatic perspective.
another action.” The preperfect category includes many clauses that are more commonly
considered simultaneous.
16. See also Williams (1976: §573).
17. According to GKC (§164), a sequence of perfects refers to a rapid sequence of events,
citing as evidence many of the same verses as does Joüon. See also König (1897: §119).

3.1.3. Narrative-unit demarcation

Driver (1892: 201) notes that subject-first clauses sometimes mark the
“commencement of a new thread” in the narrative:
(8) Gen 16:1

‫ושׂרי אשת אברם לא ילדה לו‬
And Sarai, Abram’s wife, had born him no children.
Similar observations are made by Brockelmann (1956: §48), Jongeling (1980),

Khan (1988: 86–88), and Lambdin (1971: §132), among others. Lambdin and
Khan note that preposed temporal adverbials as well as preposed subjects
occur at the beginning of an episode. Paragraph-marking is a central theme
in Heller (2004)’s study of the function of clause types in BH. According to
Heller (2004: 56), “isolated and independent QATAL clauses consistently mark
paragraph boundaries.” Although most qatal clauses are preposed, Heller does
not make reference to word order in his study.
The kind of narrative unit marked by preposing is a matter of debate. Eskhult
(1990: 33, 40, 50) writes that preposing may mark a new episode or paragraph.
Khan (1988: 86) states that subject-preposed clauses “typically occur at span
boundaries” (for definition of the span, see the next paragraph). The notion of
the boundary is conceived rather broadly by Khan to include circumstantial
or background information, which many would consider to be internal to the
span. For Khan, Lambdin, and Heller, preposing may mark the end as well as
the beginning of a unit; however, evidence for the marking of unit endings is
more tenuous than for the unit-onset marking function. 18
The nature and definition of units such as the episode, the paragraph, and
the span are subject to various interpretations. The episode is a narrative unit,
whereas the paragraph is variously viewed as a literary, thematic, or even a
grammatical unit. 19 Longacre (1979) conceives of the paragraph as a unit ex-
18. See also Mirsky (1977; 1999: 11–36), who states that the ends of verses and larger
thematic units are often marked by change of word order. By this, Mirsky means not neces-
sarily preposing but inversion of whatever word order is exhibited in previous clauses within
the unit. Although in most of Mirsky’s examples the unit ends with a preposed clause, in two
cases the clauses within the unit exhibit preposed order, and the last clause exhibits verb-
first order (Pss 22:24, 23:2–3). Many of his examples are from poetry or nonverbal clauses,
although he does cite several cases involving preposed nonsubordinate finite clauses in the
classical BH prose corpus, including Gen 19:3, 23:15; Deut 17:17; Josh 16:8; 2 Kgs 4:4. It
should be noted that several of these can be alternately explained as topicalized or focused
clauses, as defined in the present work. I am indebted to Uri Mor for the Mirsky reference.
19. Givón (1983: 8) attempts to quantify the description of the paragraph, stating that
the thematic paragraph is linked “in a statistically significant but not absolute fashion” to the
quantitative concept of “topic continuity.” Every noun phrase is considered by Givón to be
a topic. Continuity and discontinuity are coded by a number of means including word order,
morphology, and phonology. Discontinuous topics are generally coded by full noun phrases

hibiting thematic unity and marked by special grammatical structures. Para-

graphs according to Longacre encode notions such as conjoining, temporal
relations, logical relations, elaborative devices, and reportative devices. Exter
Blokland (1995) criticizes Longacre’s model of the paragraph as essentially
literary and semantic and puts forth a different grammatical model based on
syntactic criteria. Khan’s span is a semantic/pragmatic concept, referring to
a textual unit “in which there is some kind of uniformity” (1988: xxxv). He
distinguishes between the topic span, a unit that is “about” a single referent,
and the theme span, which he defines as a unit concerning a single “semantic
domain.”
Any theory of textual units in the Bible should certainly take into account
the traditional Jewish text division, the system of ‫‘ פרשה פתוחה‬open portion’
and ‫‘ פרשה סתומה‬closed portion’. 20 The biblical text is divided into sections,
marked by a space extending to the end of the line (the open portion) or by a
space within the line (the closed portion). The open portion seems to mark a
major division and the closed portion a minor division (Tov 1992: 50). How-
ever, the correspondence between these two division types and narrative units
such as episodes and paragraphs remains to be clarified. The traditional por-
tions mark many kinds of divisions, including divisions between different sto-
ries, between major episodes or scenes in a story, between component sections
of an episode, between paragraphs, and between speeches in a dialogue (Ulrich
2003: 303).
3.1.4. Attraction
Several scholars claim that preposing may result from a process of “attrac-
tion.” According to Malbim (1973: §115), when an object at the end of one
clause is the same as the subject in the next clause, preposed word order may
result:
(9) Gen 25:33–34

‫ ויעקב נתן לעשו לחם ונזיד עדשים‬:‫וימכר את בכרתו ליעקב‬
And he sold his birthright to Jacob. And Jacob gave Esau bread and lentil
stew.
Joüon-Muraoka (§155oa) makes a similar statement regarding a final object

and states that “it is difficult to say whether the resultant chiasmus is by design
or accented independent pronouns, whereas continuous topics are coded by unaccented inde-
pendent pronouns, clitic pronouns, verb agreement, or zero anaphora (1983: 31–32). In verb-
first languages, preposing is also used to code discontinuity (1983: 33). Paragraphs generally
open with a discontinuous topic and generally exhibit continuity of the main topic within the
paragraph (1983: 9). For a study of topic continuity in BH, see Fox (1983).
20. See, e.g., Lundbom (1999: 74); Tov (2001: 50–51); Hoop (2003: 3–4). I am indebted
to Richard Steiner for the Lundbom reference.

or not”; in other words, preposing from attraction may be an unconscious pro-

cess without pragmatic significance. 21 König (1897: §339f) states that similar
syntactic categories are apt to go together; thus when a clause ends with the
subject, the next clause may also start with the subject.
3.1.5. Miscellaneous factors

Muraoka (1985: 34–44, 169) identifies a number of miscellaneous factors
associated with a tendency for subject preposing. 22 Sentences having ‫ איש‬in
the sense ‘someone/everyone’ as the subject 23 or having God as the subject are
frequently preposed. Preposing is also common in sentences at the beginning
of a statement and in sentences occurring in legal texts. Muraoka states fur-
thermore that certain adverbs and adverbial expressions, particularly anaphoric
ones, are normally clause-initial; examples are ‫כן‬, ‫כה‬, ‫כדבר הזה‬, ‫לכן‬, ‫מחר‬, and
‫עתה‬. The implication is that clauses beginning with these words or phrases are
pragmatically unmarked.
3.1.6. Preposing in direct speech

Several scholars note that preposing in direct speech is more frequent than
in narrative and often unexplainable by any of the factors discussed above.
Bloch (1946) notes that speech occupies an intermediate point between nar-
rative and poetry with respect to degree of word order variation. Macdonald
(1975) finds that preposing (“inversion”) is much more common in direct
speech than in narrative, and emphasis, which is largely responsible for pre-
posing in narrative, is far less important in direct speech. He goes so far as to
claim that “in Spoken [Hebrew] Inverted Word Order appears to be the norm”
(1975: 163; capitalization and emphasis original). Mali (1983) finds that direct
speech in the Former Prophets exhibits a greater degree of deviation from VSO
word order than narrative; he asserts, further, that in some syntactic structures
in direct speech the order of verb and subject is free.
3.1.7. Conclusion
Emphasis is widely viewed nowadays as a frustratingly subjective con-
cept, despite Muraoka’s efforts to give the concept a modern linguistic inter-
pretation. Perhaps the main contribution of the emphasis-centered model is
the categorization and description of cases in which preposing is not directly
motivated by emphasis. Many of the functional categories described above,
such as anteriority, simultaneity, and the marking of a new narrative unit, are
significant categories that must be addressed in any comprehensive treatment
21. A more probable explanation of clauses such as these involves the concept of topi-
calization (see §3.3.1.2).
22. These factors are also discussed by Muraoka in his revised version of Joüon’s gram-
mar ( Joüon-Muraoka §155).
23. This was noticed earlier by Bloch (1946: 37), as Muraoka notes.

27 The backgrounding and temporal-sequencing models 3.2.
of word order. As we will see, the achievement of contemporary work on word

order lies not so much in the discovery of new types of preposing but rather
in the formulation of conceptual frameworks that yield a more systematic and
economical account of the types already observed.
3.2. The backgrounding and temporal-sequencing models

A contemporary alternative to the emphasis-centered model is the back-
grounding model, which explains preposing on the basis of the foreground-
background distinction. The temporal-sequencing model is closely related
to the backgrounding model, and is treated together with it in the ensuing
discussion.
According to Grimes (1975), narrative can be divided into “foreground,”
which relates events, and “background,” which presents explanatory material,
introduction of participants, and setting. Hopper (1979: 213) develops this dis-
tinction further in an influential article, defining foreground as “the actual story
line,” or “the parts of the narrative which relate events belonging to the skeletal
structure of the discourse”; background consists of supportive material that
does not itself narrate the main events. 24
Longacre (1994, 2003: 62–118) applies the concepts of foreground and
background to BH, applying these first to verb form and only secondarily to
word order. In narrative, the foreground, which designates sequential events,
is characterized by the wayyiqtol form; qatal designates the background (“off-
line”), which describes secondary, usually nonsequential events. 25 Because
wayyiqtol clauses are always verb-first and qatal clauses are usually preposed,
this is roughly equivalent to a VSO-foreground, SVO-background scheme. 26
24. Hopper (1979) supplies a list of criteria differentiating foreground and background:
e.g., the foreground is frequently characterized by temporal succession, perfective verb
forms (not to be confused with the BH perfect), and subject continuity, while the background
is characterized by simultaneity, imperfective verb forms, and changes of subject. Givón
(1987) argues that the foreground/background distinction is conceptually a graded rather
than a binary distinction.
25. According to Longacre, (1982: 459–60; 2003: 57–58), each discourse type, includ-
ing the narrative, predictive/procedural, hortatory, and expository types, has its own scheme
of marking foreground and background. The link between qatal and backgrounding is char-
acteristic of narrative; in predictive discourse the background is marked by yiqtol (1982: 468;
2003: 106). Longacre’s conception of narrative is not in opposition to direct speech; thus,
direct speech in Longacre’s scheme is also classified as belonging to any of the above types,
including narrative (see also Heller 2004: 23). As mentioned in §2.2.4, Longacre (1982: 472)
views the expository discourse type as having a different basic word order than narrative,
based on the fact that expository discourse consists mainly of nominal clauses with subject-
predicate word order.
26. Longacre, however, views qatal as marking background whether the clause has
verb-first or preposed word order. Qatal with a preposed noun is used to “introduce or feature
a participant or prop” (2003: 71), whereas qatal with verb-first word order is used “to pres-
ent a preliminary action.” Longacre’s theory also distinguishes between different degrees

3.2. The backgrounding and temporal-sequencing models 28
Other scholars with a similar paradigm include Eskhult (1990), Dawson

(1994), Niccacci (e.g., 1990, 1994, 1997), and van Wolde (1997). Niccacci
bases his theory on the linguistic model of H. Weinrich and its application to
BH by W. Schneider. Niccacci (e.g., 1990: 19–21; 1997), distinguishes two
levels of “linguistic attitude”: “narrative” (i.e., foreground) and “commentary”
(i.e., background). 27 Payne’s (1991) analysis also bears a strong similarity to
the backgrounding approach, although he does not use these terms. He ex-
plains preposing as side-stepping the event line for various purposes. Heller
(2004: 431) views foreground and background as important for understanding
the choice between wayyiqtol and other verb forms, but does not address word
order at all. 28 Buth combines the foreground-background framework with an
information-structure approach, as described in §3.3.5.
The nature of the foreground-background distinction is open to various
interpretations. For Longacre, foreground and background relate to the way
in which discourse is structured. Reinhart (1984) understands the concepts
as cognitive in nature, relating to the way information is organized and pro-
cessed by the human mind. 29 Although foreground is usually sequential and
background nonsequential, the foreground/background concepts are inherently
independent from temporal sequencing, because a speaker can choose to repre-
sent sequential events as background.
In the temporal-sequencing model, the somewhat elusive foreground-back-
ground concepts are replaced by the more concrete notions of sequentiality
and nonsequentiality. Givón (1977) finds that the BH imperfect, occurring
mostly in VS clauses, marks temporal continuity (i.e., sequentiality), while the
perfect, generally occurring in SV clauses, marks an anterior or unsequenced
of backgrounding: the “secondary storyline,” which relates secondary, off-line events, is

marked by qatal, whereas more backgrounded activities are marked by participial and non-
verbal clauses, among other constructions (2003: 79). For a critique of Longacre’s theory,
see Heimerdinger (1999); some of Heimerdinger’s criticisms are discussed further in this
section.
27. The “narrative-commentary” axis is only one level of Niccacci’s analysis, which
also includes two other axes, the “foreground-background” axis (actually referring to em-
phasis rather than foreground-background in the previously discussed sense) and a temporal
“linguistic-perspective” axis.
28. Heller’s model is considerably more complex than Longacre’s. According to Heller,
foreground-background is only one of two main functions of non-wayyiqtol clauses; the
other function is marking initial and final paragraph boundaries (see §3.1.3 above). Mul-
tiple non-wayyiqtol clauses are said to mark background information, while isolated non-
wayyiqtol clauses mark paragraph boundaries (2004: 431). As far as wayyiqtol, uninterrupted
wayyiqtol chains in narrative are said to mark main-line events, while isolated wayyiqtol
clauses are claimed to mark backgrounded, rather than foregrounded, information (2004:
430–31). For a review of Heller, see Zewi (2004).
29. A similar approach is adopted Hatav (1985).

29 The backgrounding and temporal-sequencing models 3.2.
event. 30 According to Myhill (1992a; 1992b: 172–77; 1995), BH and other

VSO languages use the unmarked word order for sequential past events and
marked word order for nonsequential events. 31 More recently, Goldfajn (1998)
writes that SV word order signifies an event that is anterior, simultaneous, or
posterior.
There is some disagreement over the intended scope of the backgrounding/
temporal-sequencing models. Longacre would like his backgrounding model
to pertain to all preposed clauses with qatal, and Myhill (1995) makes a simi-
lar claim with regard to the temporal-sequencing model. Other scholars such
as Givón, Buth, and Goldfajn assume that the marking of background/non-
sequentiality is restricted to qatal clauses with a preposed subject; the prepos-
ing of complements or adjuncts is assumed to be motivated by other factors.
The model presented by Longacre, as described above, centers on the as-
sertion that wayyiqtol always marks foreground and qatal background. This
assertion has been challenged by Heimerdinger, who points out that “an evalu-
ative comment, a descriptive detail, a summary, an enumeration,” as well as
“explanatory information such as flashbacks” may all be marked by wayyiq-
tol (1999: 261). 32 Furthermore, qatal clauses are not always backgrounded
(Heimerdinger 1999: 93–98). 33 Cook (2004: 264) asserts that, though fore-
grounding and backgrounding are pragmatically marked by wayyiqtol and
qatal, respectively, there are “any number” of other pragmatic factors that af-
fect the choice of verb form as well.
Givón’s assertion of a clear-cut correspondence between temporal sequenc-
ing and verb form is also open to challenge. Although most consecutive forms
do refer to sequential events, this is not always the case, as pointed out by Buth
(1995: 86–87) and Cook (2004: 257–61). A review of the first 10 chapters of
30. By “imperfect,” Givón means both the consecutive and the simple imperfect. See
also Dempster (1985).
31. Myhill (1995) presents an unconventional description of preposing in clauses with
future time reference: in modal clauses, preposing is used in commands with third-person
animate subjects. In non-modal clauses, preposing marks unilateral good prophecies, neu-
tral prophecies, bad prophecies, unilateral promises, hopes, hostile intentions, bad self-
prophecies, and guesses. Not surprisingly, Myhill (1995: 113) comments: “I cannot say
with absolute certainty that the categories in table 3 [i.e., the nonmodal categories described
above] represent the best possible analysis of these data.”
32. Heimerdinger’s own understanding of the foregrounding concept is different from
the one discussed in this section; a foregrounded element for him is one that is “thrown into
relief by the writer or speaker and is perceived as such by the reader or hearer” (1999: 222).
Thus, clauses in the storyline are not considered foregrounded unless they are specially high-
lighted. According to Heimerdinger, foregrounding in this sense is marked not by specific
clause structures (1999: 239) but by other devices such as repetition and evaluation.
33. Longacre (1989, 1992) addresses this argument by introducing a distinction between
primary and secondary storylines. Perfect clauses carry the secondary storyline, which is
normally background but can be promoted to foreground under certain circumstances.

3.2. The backgrounding and temporal-sequencing models 30
Genesis turns up seven clear instances of wayyiqtol marking a nonsequential

event or state: Gen 2:25 (‫ ;)ויהיו‬3:24 (‫ ;)ויגרש‬5:4 (‫ ;)ויולד‬7:17 (‫ ;)וירבו‬7:23
(‫ ;)וימח‬9:20 (‫ ;)ויחל‬10:30 (‫)ויהי‬. In addition, although Driver (1892: 84–88) is
doubtful that wayyiqtol can indicate the pluperfect (anterior) within a narrative,
Martin (1968–69), Buth (1994a, 1995), Collins (1995), and Cook (2004) cite a
number of convincing examples, including Judg 11:1 (‫ )ויולד‬and 2 Sam 12:27
(‫)וישלח‬. 34
A different way of evaluating the backgrounding and temporal-sequencing
models is examining how they account for the various categories of preposing
discussed in §3.1 above. Preposed clauses traditionally viewed as expressing
emphasis do not fit naturally in either model. The background category in-
cludes anterior clauses as well as parenthetical comments. Pairs of simultane-
ous preposed clauses do not fit the background concept as well, because the
second clause in the pair, if not the first, is usually a “main” event (see §3.1.2).
As far as preposing to mark a new textual unit, this may at times coincide with
backgrounding, as in (8), above, but it is hard to see why this should always be
the case. The backgrounding explanation is particularly inapt for chiastic con-
trastive structures such as (10), as Bailey and Levinsohn (1992) have pointed
out.
(10) Gen 41:51–52

:‫ויקרא יוסף את שם הבכור מנשה כי נשני אלהים את כל עמלי ואת כל בית אבי‬
‫ואת שם השני קרא אפרים כי הפרני אלהים בארץ עניי‬
And Joseph called the name of the firstborn Manasseh, “For God has
made me forget all my hardship and my father’s entire house.” And the
name of the second he called Ephraim, “For God has made me fruitful in
the land of my affliction.”
The backgrounding model assigns the naming of the firstborn child to the fore-
ground and the naming of the second (with qatal ) to the background, although
the second child turns out to be at least as central to the storyline as the first.
The temporal-sequencing model succeeds in accounting for anterior and
simultaneous clauses, as well as many parenthetical clauses. It is also able to
incorporate many contrastive clause pairs, which frequently involve simulta-
neous events. However, the claim that preposing always marks nonsequenti-
34. See also the discussion and further references in Waltke and O’Connor (1990:
§33.2.3). Hatav (1997: 57) has defended the link between wayyiqtol and sequentiality, stat-
ing that her study of wayyiqtol forms in classical Hebrew prose shows that only 6% of 2,445
wayyiqtol clauses are clearly nonsequential (see also Hatav 2000b). Heller (2004: 430) also
believes there is a consistent link between wayyiqtol and sequentiality. According to Cook
(2004: 264), the high degree of correlation between wayyiqtol and sequentiality is due to
the fact that wayyiqtol is a narrative verb and the fact that temporal succession is the natural
order in narrative.

31 The information-structure model 3.3.
ality is easily refutable; many preposed clauses, including some contrastive

clauses such as (10), are clearly sequential. 35
In conclusion, the backgrounding model and temporal-sequencing models
are not applicable to all preposed clauses. The most plausible formulations
of the theory apply exclusively to subject-preposed clauses; thus, object-
preposing and adjunct-preposing are not afforded an explanation. Furthermore,
many subject-preposed clauses do not describe backgrounded or nonsequential
events and cannot be accounted for within these models.
A noteworthy aspect of both models is that they view the function of prepos-
ing as relating to the clause as a whole. The preposed subject is not significant
in and of itself; it is preposed merely in order to create a marked structure. The
perspective is radically different in the model we turn to next, the information-
structure model. In this model, the function of preposing is understood as relat-
ing specifically to the preposed element; it is this element, therefore, that is the
object of attention.
3.3. The information-structure model

A number of scholars have recently adopted an information-structure model
of BH word order. The field of information structure has its origin in work by
H. Weil in the 19th century, was developed further by linguists of the Prague
school, and was subsequently taken in various directions by European scholars
such as Halliday and American scholars such as Chafe. Information structure
concerns the ways in which sentence structures convey a message to the ad-
dressee regarding the interpretation of the sentence in its context. 36
In §3.3.1, I survey information-structure concepts as they are described
in the general linguistic literature. In §§3.3.2–3.3.7, I describe various mod-
els of BH word order that are based on information-structure concepts. 37 The
models are categorized according to the scholar most responsible for their
development.
35. Representative examples from Genesis are Gen 4:18; 10:24, 25, 26–29; 11:27; 14:4;
15:17; 18:7; 19:6, 10.
36. The term “information structure” was introduced by Halliday (1967). Syntactic and
prosodic structures that mark information structure are sometimes known as information-
packaging devices because they organize the informational content of the sentence in a
manner appropriate to the context. For this term, see, e.g., Chafe (1976); Vallduví (1992);
McNally (1998).
37. Information structure has also been applied to the study of Modern Hebrew word
order; see, e.g., Bendavid (1958); Ben-Horin (1976); Rosén (1977: 322–24; 1982); Giora
(1982); Nir and Roeh (1984); Ziv (1988); Glinert (1989: §37); Kuzar (1989, 2002); Azar
(1993).

3.3. The information-structure model 32
3.3.1. Introduction to information-structure concepts

3.3.1.1. Theme-rheme/topic-comment
In work on information structure by linguists from the Prague school, it is
frequently asserted that sentences divide into “theme” and “rheme.” Mathe-
sius states that the theme is “what is given by the context,” “what is being
commented upon”; 38 the new part of the sentence is the rheme. Theme and
rheme are also known as “psychological subject” and “psychological predi-
cate,” respectively. According to the principle of “functional sentence perspec-
tive” (Firbas 1966b, 1992), the normal position for the theme is at the beginning
of the sentence. Sentences beginning with the rheme are marked as emotive.
The terms topic and comment, introduced by Hockett (1958: 201), are generally
taken to be synonymous with theme and rheme. 39 In the ensuing discussion, the
terms are used interchangeably.
There have been many attempts to define rigorously the concept of theme
or topic. Gómez-González (2001) describes three approaches to the concept:
the “aboutness” approach, in which the topic is what the sentence is about; the
“informational” approach, in which the topic is the “given” (i.e., known) part
of the clause, and the “syntactic” approach, in which the topic is the first ele-
ment of the clause. 40 The most widely accepted approach is the “aboutness”
approach. For some, the topic is what the discourse, rather than the sentence, is
about. 41 Discourse topic and sentence topic can be viewed as interrelated; van
Oosten (1985: 23), for example, sees the sentence topic as “a constituent inside
38. See, e.g., Mathesius (1975: 156).
39. Others distinguish between topic-comment and theme-rheme. For Halliday (2004:
65), for example, topic is a particular type of theme.
40. According to Reinhart (1981), a sentence is “about” an entity if it is intended to
increase our knowledge about that entity. The topic is the referential entry under which the
proposition expressed by the sentence is stored in the addressee’s knowledge (1981: 80).
A similar definition can be found in Gundel (1985: 86; 1988: 210). Lambrecht’s (1994:
131) definition of topic is also similar, except that he allows a clause to have more than one
topic (1994: 147). Vallduví (1992) understands topic (which he terms “link”) using a “filing”
metaphor similar to Reinhart’s description: the topic represents the address of the file card
in which the information provided by the sentence is stored (see also Vallduví and Engdahl
1996). The “backward-looking center” of Centering Theory is viewed by some as a model
for topic in the sense of aboutness (Walker, Joshi, and Prince 1998a).
Another variety of the informational approach is Firbas’s (e.g., 1966a, 1966b, 1992)
theory of “communicative dynamism”: the topic is the element that advances the communi-
cation the least. For Givón (1983), topicality is a scalar notion depending on the degree of
continuity with the previous discourse.
For the “syntactic approach,” see, e.g., Halliday (1967). Firbas (1966a) rejects this ap-
proach on the grounds that in emotive sentences the rheme is first, as mentioned above.
41. See, e.g., Keenan and Schieffelin (1976); Van Dijk (1977: 132–40). In Dik’s 1989
version of Functional Grammar, the pragmatic function of Topic refers to the discourse topic,
with four subtypes being distinguished: New Topic, Given Topic, Resumed Topic, and Sub-
Topic (1989: 267). See Bolkestein (1998) for a detailed critique of how the term Topic is used

a sentence which most directly evokes the discourse topic which is relevant in
the current sentence.” 42
The various definitions of topic remain problematic. The most accepted
conception of topic, the notion of “aboutness,” has thus far resisted objective
formulation, despite valiant efforts on the part of many researchers. Gómez-
González (2001: 31) sums up the state of the field as follows: “the intricacies
raised by the numerous and heterogeneous variations of the semantic interpre-
tation have led many scholars to conclude that Theme/Topic in terms of about-
ness cannot be regarded as an objectively identifiable unique category, but as a
clearly intuitive, and therefore subjective concept.”
3.3.1.2. Topicalization
The term topicalization first appears in Ross (1967: 115) as the name for
the English preposing transformation. 43 This transformation moves a noninitial
expression to the head of the clause, yielding a structure such as Cake, I eat
with a spoon. 44 The term topicalization is also used to refer to the structure
produced by the topicalization transformation. It was noticed later by Gundel
(1977: 134–35; 1985: 88 n. 10) and others that the topicalization construction
exhibits two distinct accentuation patterns. 45 The first has the primary accent
on an element other than the preposed constituent, for example, Cake, I eat
with a spoon. 46 A second pattern, restricted to the casual register, has the pri-
mary accent on the preposed element: Cake I asked for, not pie. This type of
sentence marks the preposed constituent as the focus, that is, the new informa-
tion in the clause (see §3.3.1.3). It has become customary to reserve the term
topicalization for the former syntactic-prosodic pattern; the latter pattern is
termed focus-movement or focusing (see §3.3.1.3). 47
in Functional Grammar research. Critical analyses of the concept of discourse topic can be
found in Brown and Yule (1983: 68–124) and Gómez-González (2001: 25–31).
42. Further exploration of the relation between sentence topic and discourse topic can be
found in Goutsos (1997) and in Floor’s (2004) dissertation on BH.
43. Postal (1971: 142) calls this transformation “Y-Movement,” on the grounds that the
resulting sentences are similar to those found in Yiddish. Further discussion of the transfor-
mation and its syntactic constraints can be found in Chomsky (1977).
44. Although Ross’s topicalization includes only preposed NPs, this definition was lat-
er broadened by some to include preposed adverbs and prepositional phrases as well. See
Prince (1986: 210, example 3).
45. See also Creider (1979: 4 n. 1) and Prince (1986: 209); Ben-Horin (1976) describes
corresponding constructions in Modern Hebrew.
46. There is typically a secondary accent on the preposed “cake” as well.
47. Prince distinguishes a third preposing construction in English that she calls
“Yiddish-Movement”; for example, “Can you imagine? Such a rich woman and after all
I’ve done for her, a shirt she gave my Harry when he was bar mitsved.” Yiddish-movement
is syntactically and prosodically identical to focus-movement but is dialectically restricted
with different pragmatic characteristics (see §6.1.2 n. 9).

As their names suggest, topicalization in the narrow sense and focus-

movement appear to have distinct pragmatic functions. This has led to a third
use of the term topicalization, to denote a syntactic-pragmatic concept, that is,
a certain “pragmatically defined type of preposing” (Birner and Ward 1998:
38 n. 9). 48 The latter conception is useful for the cross-linguistic investiga-
tion of topicalization, because other languages may not have prosodic patterns
identical to English, and in some cases (as in BH) prosodic data may not be
available. In other words, topicalization denotes a preposing construction that
has the same pragmatic function as the English topicalization (as opposed to
focus-movement) construction.
What, however, is the pragmatic function of English topicalization? A
widespread theory is that topicalization marks the preposed constituent as the
topic of the sentence. This idea has its origin in the Firbas’s principle of “func-
tional sentence perspective,” referred to above (p. 32). According to functional
sentence perspective, the subject is most frequently the theme and naturally
comes first in the normal clause. When an object or adverbial is the theme, it
is placed in the usual place for the theme at the head of the clause, producing a
topicalized structure. 49 This hypothesis is inapplicable to verb-first languages
such as BH, because the functional-sentence perspective principle does not op-
erate in these languages; 50 in other words; “theme-rheme” is not the operative
organizational principle in the normal BH clause. Because the initial element
in the BH clause is not ordinarily the theme, there is no reason to expect that a
preposed object or adverbial is the theme. In fact, the notion of theme or topic
marking is problematic even for subject-first languages such as English. If the
concept of topic has no clearly defined meaning (see §3.3.1.1), the idea of topi-
calization as topic marking is similarly opaque.
A different theory points to the frequently contrastive contexts in which
topicalized clauses are found, for example, I ate an apple and a pear. I liked
the pear; the apple I detested (Ben-Horin 1976: 194; Creider 1979: 5 n. 2). The
topicalized clause in this example can be said to have the function of contrast-
ing what is said about the apple with what was previously said about the pear.
Givón (2001: 2:225) explains that the contrast in topicalization involves the
breaking of expectations regarding “various members of a group (type, genus)
whose members are expected to display similar behavior, or receive similar
treatment.” The type of contrast involved in topicalization may involve jux-
48. See also Myhill (1985: 181) and Bailey and Levinsohn (1992).
49. See Halliday (1967: 211–23), who terms this function “thematization.” A similar
view can be found in Lambrecht (1994: 147; see also p. 161). For a review of Lambrecht’s
work, including his views on topic, see Ziv (1996b).
50. Verbs are in all but exceptional circumstances part of the rheme, irrespective of the
language involved; hence, unmarked clauses in verb-first languages begin with a component
of the rheme and not the theme.

taposition or antithesis, to use Muraoka’s terms: the contrasted items may be

simply different or sharply antithetical.
Not all topicalized clauses are contrastive, however. A continuative use of
topicalization has also been identified, in which the preposed object refers to
an element mentioned just prior to the clause. Mathesius (1975: 158–59) gives
as an example of this type: In returning he met on the plain of Caraci a scholar
on a bay mule coming from Bologna, and him he questioned about Tuscany.
As we will see in §3.3.3, the contrastive and continuative uses of topicalization
have given rise to some uncertainty as to whether topicalization has the core
function of expressing “discontinuity,” “continuity,” or both.
Some of the most important works on English topicalization have not thus
far had a significant influence on the study of BH. These works, including
Prince (e.g., 1985, 1986, 1988, 1998), Ward and Prince (1991), and Birner and
Ward (1998), are discussed in detail in chap. 6 (pp. 90ff.).
3.3.1.3. Focus and focusing
Gundel (1999) identifies three general uses of the term focus in the linguis-
tic literature: informational, 51 contrastive, and psychological. There seems to
be a fourth use as well, which I term attentional. Focusing is the marking of a
focus by formal means, whether by prosody, syntax, or both.
Informational focus denotes the new information contained in an utterance.
This is the part of the proposition expressed by the sentence that is assumed
by the speaker not to be given, that is, shared by him and the addressee. 52 The
focus supplies the value for an unknown item in the given proposition. In En-
glish, an informational focus may be marked by sentence accent, as in Bill ate
the doughnuts. The accenting of Bill indicates that the proposition “Someone
(X) ate the doughnuts” is given information in the context of utterance. The
focus represents a new value for X; that is, “X = Bill.” 53 Informational focus
answers a wh-question that could be appropriately asked at the time of utter-
ance (Lambrecht 1994: 223; Gundel 1999: 295). In the above example, Bill is
the answer to the contextually appropriate question Who ate the doughnuts?
51. Gundel uses the term semantic focus instead of informational focus.
52. The term focus denoting new information was introduced by Halliday (1967) and
further developed by Chomsky (1970), Jackendoff (1972), Rochemont (1986), and others.
The given information is often termed the “presupposition.” I will not use this term due to the
possibility of confusion with the concept of pragmatic presupposition (see chap. 4, pp. 48ff.).
Focus is related to rheme, if rheme is defined in informational terms (see §3.3.1.1); focus,
however, is usually understood as comprising only part of the rheme.
53. For the sake of convenience, the term focus is used in this work to denote both the
new value for X (“Bill”), as well as the linguistic expression in the sentence that represents
this new value (Bill ). Propositions and the values of the variables they contain are marked by
double quotation marks, and linguistic expressions are in italic type.

Informational focus may be marked by devices other than accenting. One

such device is the cleft construction, as in It’s coffee that I want. As mentioned
above, in the casual register of English, focus may also be marked by prepos-
ing, for example, Coffee I want. In both the cleft and the preposing construc-
tions, the focus is accented in addition to being marked syntactically.
Contrastive focus is a broad concept including, among other things, infor-
mational focuses that are contrary to expectation and contrasted topics. One
definition of contrastive focus is a focus that is selected from a set of possible
candidates (Chafe 1976). For Givón (2001: 225), contrastive focus always
includes some degree of contrary belief on the part of the hearer. The wide
variety of phenomena included here makes it difficult to define the concept
precisely, as noted by Vallduví and Vilkuna (1998: 83). In addition, the term
contrastive focus is confusing, because it includes topics and focuses. Vallduví
and Vilkuna (1998: 85) argue against the merging of focus and contrast, argu-
ing that the two concepts are independent of each other.
Psychological focus is the entity that is currently the center of attention.
This concept involves previously mentioned entities, frequently referred to by
zero or unaccented pronouns. 54 The term psychological focus is somewhat con-
fusing, because this concept is in a sense the opposite of information focus and
is much closer to the notion of topic.
Attentional focus is the entity to which the speaker wishes to direct the at-
tention of the addressee. 55 There are many reasons that a speaker might draw
attention to a particular clause element; thus, attentional focus is a general
concept with a variety of applications. Attentional focus resembles the clas-
sic concept of emphasis and suffers from the same deficiency as that concept,
notably, the nonspecific nature of the definition.
An influential theory incorporating both topic and focus has been developed
by Lambrecht (1994); this theory is discussed in §3.3.7.
The following sections discuss models of word order developed by scholars
of BH that are based on the information-structure concepts discussed above.
3.3.2. Bendavid: Psychological subject and predicate

Bendavid devotes a chapter to BH word order in his work comparing Bibli-
cal and Mishnaic Hebrew (1971: 785–855). Bendavid is clearly influenced by
the Prague school work on information structure, invoking the concepts of
psychological subject and psychological predicate. He classifies word orders
based on the placement of the new information, i.e., the psychological predi-
cate. Bendavid describes BH as having a kind of reverse functional sentence
perspective. In the “calm” sequence, the new information is at the end. An
54. See Gundel, Hedberg, and Zacharski (1993).

55. See, e.g., Taglicht (1984: 1, 7) and Erteschik-Shir (1997: 11; 1998). According to
Taglicht, a focus is an element that is marked for prominence for various pragmatic purposes.

alternative is the “strong” sequence, in which the new information is at the

beginning. The strong sequence is used in answering questions and in order
to express contrast and emphasis. There is also a third order in which the new
information is in the middle.
3.3.3. Bandstra
Bandstra’s dissertation on ‫( כי‬1982: 72–88) appears to be the first work to
present a systematic model of BH preposing based on the concepts of topical-
ization and focusing. 56 Bandstra writes that the topic is the given part of the
clause, “that part of the clause that represents the writer’s thematic choice, that
about which he will say something new” (1982: 74). According to Bandstra, the
topicalization construction marks a topic mentioned in the preceding material
in order to maintain continuity in the text. This category accounts for preposing
traditionally attributed to attraction. Focusing, according to Bandstra, involves
new, unexpected information, or a change in topic (1982: 78). Focusing in this
conception subsumes the traditional “emphasis” and “contrast” categories.
Bandstra’s conception of focus is closest to contrastive focus (see §3.3.1.3):
any preposed constituent that is not continuous with the preceding material is
termed a focus, including contrastive topics. Topicalization is restricted to the
continuative type of preposing (see §3.3.1.2). In a later article (1992), Bandstra
expands the concept of topicalization to include focusing as well. Topicaliza-
tion is described as the placing of new information in the initial slot, which
signals discontinuity or transition; it also is said to include cases in which the
topicalized element is not new, but is fronted in order to provide continuity.
It is not clear how the apparently contradictory categories of continuity and
discontinuity are unified under a single heading.
3.3.4. Revell
Revell (1989a: 2) invokes the concept of attentional focus in his explana-
tion of BH word order, although he does not refer specifically to the linguistic
literature on focusing. As mentioned in §2.2, Revell believes that word order in
every clause, not just in preposed clauses, is pragmatically motivated; in other
words, BH is apparently viewed as a free word-order language. He states that
“the constituent placed first in the clause is that which the author wishes to be
the primary focus of the reader’s attention.” He lists various types of elements
to which the author might choose to draw attention, including a new subject,
the time or place of the action at the beginning of a narrative, information
56. Bandstra was not, however, the first to invoke the concepts of topicalization and
focusing in reference to BH, contrary to van der Merwe (1990: 42). O’Connor (1980: 81,
306–7) discusses the concepts of “focus-marking” and “topicalization” but does not explore
the ramifications for preposing.

significant to the story, the answer to a question, a contrastive element, and an

anaphoric pronoun or adverb that refers to the preceding context.
In Revell (1989b: 21), the principle governing word order is termed “the-
matization” and is defined as the arrangement of the components of the clause
“in the order of their significance for the speaker or narrator.” Revell states
that imperfect verb forms are modal when clause-initial and indicative when
clause-medial; this assertion is adopted by a number of other scholars includ-
ing Cook (2001, 2004), DeCaen (1995), Holmstedt (2002, 2005), Joosten
(1992), and Shulman (1996). It should be noted that Revell in no way intended
his observation to mean that the word-order system as a whole marks the se-
mantic indicative/modal opposition. A position such as this would contradict
Revell’s view that the weqatal form, which serves as a syntactically condi-
tioned variant of the indicative imperfect, is used whenever a speaker wishes
to place an indicative verb in clause-initial position (Revell 1989b: 21). For
Revell, the modal/indicative rule is a secondary consequence of the thematiza-
tion principle: modal verbs tend to be the most significant part of the clause and
are therefore usually clause-initial (Revell 1989b: 21). 57 As Shulman points
out, modals are occasionally clause-medial when the speaker wishes to place
the focus on a different clause component (1996: 241–49). 58
3.3.5. Buth: A synthesis of information structure and foreground-background

In Buth’s (1987) dissertation on Biblical Aramaic and subsequent work on
BH, topic and focus are set within the Functional Grammar framework, devel-
oped by Dik (1980, 1989). 59 In Functional Grammar, language is viewed pri-
marily as an instrument of communication; as a result, “pragmatics is seen as
the all-encompassing framework within which semantics and syntax must be
studied” (Dik 1989: 7). It is posited that all languages have an initial preposed
position (“P1”) that may be used for marking topic and focus. Topic in Func-
tional Grammar denotes what the sentence is about (Dik 1980: 16) or, in a later
version of the theory, what the discourse is about (Dik 1989: 266–77). 60 Topics
that are continuous or resumed are likely to be marked by being placed in P1.
Focus is understood as the most important part of the utterance; this definition
57. Cook’s statement “word order is grammatically relevant in Biblical Hebrew: it dis-
tinguishes between modal and indicative verbs” (2004: 265) is made possible only by his
reinterpretation of weqatal as a modal form. Even with this adjustment, the most common
indicative verb form, wayyiqtol, is excluded from the VX-modal/XV-indicative scheme. On
the relevance of modality to the question of BH basic word order, see §2.2.
58. For examples of XV clauses containing volitive forms in Genesis, see §5.1, p. 65
nn. 2-4.
59. Rosenbaum (1997) has applied the Functional Grammar framework to BH poetry.
60. In an earlier version of the theory (Dik 1980), topicality is defined as what the sen-
tence is about.

includes informational focus as well as contrastive focus (Dik 1989: 277–85). 61

A focus may be marked when it fills in a gap in the addressee’s information, or
is contrastive. Marked topic and focus are denoted in Functional Grammar by
the capitalized terms Topic and Focus.
Buth (1987: 50–51) proposes an additional function, Setting, which “pre-
sents the framework” in which the clause is to be interpreted and usually con-
sists of a spatial or temporal phrase. 62 The implication would appear to be that
temporal phrases can be fronted for no reason other than to provide an orienta-
tion for the clause. In later work, Buth proposes a new definition of Topic that
includes the Setting function. The resulting inclusive category is defined as a
constituent that serves “as a frame of reference for relating a clause to its con-
text” (1994b: 217) and is renamed Contextualizing Constituent (Buth 1994b,
1995, 1999). Buth considers contrastive topics to be simultaneously Topics and
Focuses (1994b: 223).
An innovation of Buth’s is the synthesis of the information-structure and
foreground-background models to yield a comprehensive theory of BH pre-
posing. 63 In his 1987 dissertation, Buth writes that backgrounded clauses
may be marked in Biblical Aramaic by putting a “pseudo-topic” in the pre-
posed position. Buth refines this idea in later works (1994b, 1995). Influenced
by Levinsohn (1990), 64 Buth proposes that wayyiqtol (VSO) clauses mark
continuity-foreground, and X-qatal (SVO/OVS) clauses mark discontinuity-
background. 65 One type of discontinuity concerns a particular constituent in the
clause, which is preposed to mark it as Focus or Contextualizing Constituent. A
different type is temporal discontinuity (i.e., nonsequentiality), which concerns
the clause as a whole. This type of discontinuity may also be marked by pre-
posing, although the discontinuity does not concern the preposed constituent.
61. See also de Jong (1981) and Dik et al. (1981). For critical discussion of Topic and
Focus in Functional Grammar, see, e.g., Siewierska (1991); Bolkestein (1998).
62. On the distinction between Topic and Setting, see Hannay (1991: 146). On Setting in
BH, see also Rosenbaum 1997: 41–44. Winther-Nielsen (1992) cites Gen 1:1 as an example
of preposing to mark Setting in BH.
63. The synthesis of topic and background is most fully explained in Buth’s later ar-
ticles, although it is clearly present as early as his 1987 dissertation.
64. Levinsohn (1990) describes BH topicalization as the marking of discontinuity: the
discontinuity may concern a switch in participants, or a switch in the thematic or temporal
progression of the story. See also Levinsohn’s earlier (1987) discussion of word order and
discontinuity in the book of Acts. The debt to Levinsohn is acknowledged in Buth (1995: 97
n. 14). Levinsohn, however, does not feel that background is a useful concept with respect to
preposing. In Bailey and Levinsohn (1992), the thesis is advanced that all preposing is either
focusing or topicalization.
65. A precedent for this paradigm is Lambdin (1971: §132), for whom we- + verb (VSO)
signifies a “conjunctive-sequential” clause, and waw + nonverb (SVO/OVS) signals a “dis-
junctive” clause. Disjunctive clauses may be contrastive, circumstantial, explanatory, par-
enthetical, terminative, or initial (see also Waltke and O’Connor 1990: §§8.3, 39.21–23).

Buth (1994b: 218) writes, “The marked Contextualizing Constituent (‘Topic’)

may not itself be the point of discontinuity. The clause as a whole is the Dis-
continuity, not the marked term.” A further use of preposing is to mark a new
paragraph or narrative unit (1995: 89–90). Buth suggests that this too can be
viewed as a type of discontinuity, and the preposed element in these cases
can be termed a Contextualizing Constituent (1995: 90): “the contextualizing
constituent . . . can be said to mark the clause for a wide range of relationships
which includes a backgrounding-discontinuity where the discontinuity does
not refer to a switch of a ‘topic,’ but to a higher level of discontinuity.” A
further extension of the contextualizing constituent concept is to include the
preposed constituent that marks a “dramatic pause” within the story (1994b:
226–27). The dramatic pause marks a peak in the story, which is a sequential,
foregrounded event. This too, it is proposed, should be viewed as a subtype of
discontinuity-background (1995: 92).
An apparent inconsistency in Buth’s model regards the claim that Topics are
necessarily discontinuous. As noted above, in Functional Grammar continuous
topics as well as discontinuous topics are thought to be marked by preposing.
Buth’s definition of contextualizing constituent should include continuous top-
ics, because these elements certainly orient the clause to its context. If this is
the case, not all preposing involves discontinuity.
More fundamentally, it is questionable whether Buth is successful in unify-
ing all the different types of discontinuities. By applying the term Contextualiz-
ing Constituent to cases in which the preposed item does not actually function
as Topic or Setting, Buth is calling very different things by the same name. A
similar problem concerns the concept of dramatic pause, which occurs in a
foregrounded clause but is nevertheless included in the background category.
3.3.6. Gross: A focus-centered model

Gross has investigated BH word order in a number of articles (e.g., 1987a,
1988a, 1993a, 1993b, 1994, 2001b, 2004), a monograph on BH word order
(1996) in collaboration with Disse and Michel, and another monograph on
doubly preposed clauses, particularly in poetry (2001a). 66 Disse (1998) also
published his 1996 dissertation on information structure, adopting a similar
theoretical framework.
Gross uses a syntactic framework developed by W. Richter, with some
modifications. He distinguishes a preverbal field (“Vorfeld”) and a main field
(“Hauptfeld”) in the clause (1996: 48, 149). The preverbal field may contain
66. The main conclusions of Gross (1996) are summarized in Gross (1999) and in a re-
view article by van der Merwe (1999c). The monograph on double preposing is examined in
a review article by Bailey (2004). The discussion in this section focuses on the conclusions
of the 1996 monograph, because the pragmatics of double preposing, as well as the functions
of preposing in poetry, are outside the scope of this study (see §§1.2 and 5.4.2).

one or more items. Certain conjunctions, “clause deictics” and “text deictics,”
and clausal adverbs relate to the entire sentence and always stand at the outer
edge of the preverbal field. A number of other syntactic categories are always
or usually in the preverbal field, including interrogative particles and deictic
particles such as ‫שם‬, ‫כן‬, ‫כה‬, ‫זה‬, and ‫אז‬.
Gross (1996: 53–66) discusses the concept of theme (topic) extensively,
but does not make use of it in his pragmatic description of preposing. He uses
topicalization as a purely syntactic term referring to a preposed structure. On
the pragmatic level, preposing is analyzed primarily in terms of focus. In defin-
ing focus, Gross (1996: 66–72) utilizes J. Jacobs’s relational view, in which an
element is considered a focus if that element was chosen from a set of potential
referents. 67 This conception of focus closely resembles Chafe’s (1976) defini-
tion of contrastive focus (see §3.3.1.3 above). A sentence is allowed by Gross
to have multiple focuses.
Gross (1996; 1999: 40–45) describes a number of other circumstances,
aside from focusing, in which preposing occurs. A temporal adjunct may be
preposed without being focused. A nonfocused subject may be preposed at the
beginning of a narrative or quoted speech, 68 in an answer to a discourse-initial
request, 69 in a background description, in “authorial commentary” within a
narrative, or in “supplement” information (e.g., an anterior event). Additional
reasons for preposing are “connection,” in which the preposed item is identical
to an item in the previous clause, and “enumeration,” which is an “on-the-one-
hand/on-the-other” or “neither-nor” construction. 70
Van der Merwe (1999c: 290) points out that because he omits the notion of
topic from his pragmatic analysis of preposing, Gross is unable to provide a
comprehensive account of preposing: “the complexity of his findings . . . and
the number of cases that he has to leave out of consideration . . . indicate that
his notional category of focus can explain only one aspect of his data.” This
criticism is taken into account in Gross’s 2001 monograph, in which the con-
cept of topic is identified as one of the functions of preposing (2001a: 310).
3.3.7. Heimerdinger, van der Merwe, and Shimasaki

Much of the latest research on BH word order is modeled after Lam-
brecht’s (1994) work on information focus. 71 In his revision of his dissertation,
Heimerdinger (1999) develops a theory of preverbal and postverbal word or-
der in which Lambrecht’s model plays an important role. A similar preposing
67. This definition of focus is also adopted by van der Merwe (see §3.3.7). De Regt
(2006: 292) writes similarly that preposing occurs with “restrictive” or “contrastive” focuses,
which are picked “out of a set of possibilities.”
68. See, e.g., 1 Kgs 20:17, 2 Kgs 1:6.
69. See, e.g., Judg 6:18.
70. See, e.g., Deut 7:5, 9:9.
71. For a review of Lambrecht (1994), see Polinsky (1999).

model appears in van der Merwe (1999b) and van der Merwe and Talstra
(2002–3: 83). 72 Floor (2003, 2004) follows Lambrecht and Heimerdinger in
his overall approach. 73 Another work influenced by Lambrecht is Shimasaki’s
(2002) revision of his 1999 dissertation. 74 An additional scholar who applies
Lambrecht’s framework to BH is Bailey (2004), in a review article on Gross
(2001). Lunn has applied Lambrecht’s model to BH poetry in his (2006) revi-
sion of his 2004 dissertation.
Topic for Lambrecht is what the sentence is about. Focus is defined in in-
formational terms as “the element of information whereby the presupposition
and the assertion differ from each other” (1994: 207), 75 in other words, focus is
the information conveyed by the sentence, minus the portion that the addressee
already knows. According to Lambrecht sentences must have a focus, but may
have no topic, one topic, or more than one.
Lambrecht combines focus and topic structures into a single prosodic para-
digm. 76 Most sentences can be categorized as having one of three focus/topic
articulations, each having a characteristic accentuation pattern (the main ac-
cent in the clause is marked by bold type):
(11) Lambrecht’s topic/focus articulations 77

a. Predicate focus = Topic-comment My car broke down.
(Answer to What happened to your car? )
72. Van der Merwe (1999b) cites Heimerdinger as a reference. Van der Merwe’s ap-
proach to BH word has evolved over the years. In van der Merwe (1989, 1991), preposing
is described as marking topic or focus. Like Gross, Van der Merwe defines focus as “a
particular item from a number of possible alternatives” (van der Merwe 1989: 128). Topic is
“that part of a sentence which determines the frame of interpretation” (1989: 128). Topics are
marked by preposing in cases of contrastive topics, new topics, and interruptions of narrative
sequences. In van der Merwe et al. (1999: §47), the functions of preposing are described as
marking focus, marking a new or reactivated topic, and marking a simultaneous or nearly
simultaneous event. Focus is defined there as “the most salient information conveyed by a
particular utterance,” a definition closest to attentional focus.
73. Floor (2003, 2004) elaborates on the category of topic, distinguishing four sub-
categories, and shows that marked word-order constructions contribute to the development
of discourse themes.
74. For a review of Shimasaki, see Holmstedt (2003).
75. Lambrecht (1994: 52) defines the (pragmatic) presupposition of a sentence as “the
set of propositions lexicogrammatically evoked in a sentence which the speaker assumes the
addressee already knows or is ready to take for granted at the time the sentence is uttered”
(this definition is slightly modified later on in the book). Some problems with Lambrecht’s
definition of focus are discussed in Dryer (1996: 517).
76. For Lambrecht, focus/presupposition and topic/comment are closely related notions;
thus, the predicate-focus clause can also be described as having “topic-comment” structure,
where the subject is the topic and the predicate (the focus) is the comment. Sentence-focus
structures are those that have a comment and no topic. Further comments on this paradigm
appear in §6.1.1.
77. See Lambrecht (1994: 222–23).

b. Argument focus = Identificational My car broke down.

(Answer to What broke down? )
c. Sentence focus = Event reporting My car broke down.
(Answer to What happened? )
Predicate focus is the unmarked articulation, and argument and sentence focus
the pragmatically marked types. In predicate focus, the focus is “projected”
from a constituent within the verb phrase to include the entire verb phrase. 78
Thus, although only the word down is accented in (11a), the focus is the verb
phrase broke down. The predicate-focus structure is also a topic-comment
structure: the subject is the topic and the verb phrase the comment. In the
argument-focus articulation, the accent is on the focused argument; in (11b),
the focus is the subject my car. In sentence focus, the entire sentence is the fo-
cus and simultaneously the comment: sentence-focus sentences have no topic.
Sentence focus in intransitive clauses is marked by an accent on the subject, as
in (11c); 79 the focus is projected from the subject to include the entire sentence.
The sentence-focus articulation, as in (11c), is identical to the argument-focus
articulation with a focused subject, as in (11b): in both of these articulations,
the accent is on the subject.
According to Lambrecht (1994: 138–40), the sentence-focus category in-
cludes two types of sentences: “presentational” and “event-reporting” sen-
tences. Presentational sentences indicate the existence of a referent (e.g., There
was a man), or the appearance of a new referent in the discourse (e.g., John
arrived). Event-reporting sentences announce an event concerning a referent
new to the discourse (Lambrecht 1994: 14), (e.g., My car broke down.) Lam-
brecht explains that both of these types are “thetic judgments,” that is, sen-
tences that simply recognize a fact, as opposed to the ordinary sentence, which
involves both recognizing a subject and recognizing what is expressed about
the subject by the predicate. 80
The intriguing identity between argument- and sentence-focus articulations 81
in Lambrecht’s paradigm leads Heimerdinger, van der Merwe and Shimasaki
78. On focus projection, see Halliday (1967: 208); Chomsky (1970: 92–93); Ladd (1980:
74); Selkirk (1995: 554); Gussenhoven (1999).
79. On this sentence pattern, see Schmerling (1976: 22).
80. See Kuroda (1972). Kuno (1972) calls sentence-focus sentences “neutral descrip-
tions.”
81. It should be noted that argument-focus (with a focused subject) and sentence-focus
articulations are identical only for intransitive clauses. In transitive sentence-focus sentenc-
es, both subject and complement are accented: The children went to school. See, e.g., Selkirk
(1995: 556) and Lambrecht (1994: 121, example 4.2d). This accentuation pattern is identical
with the predicate-focus articulation, rather than the argument-focus articulation. However,
in a later work Lambrecht (2000: 620–21) argues that transitive thetic sentences with a lexi-
cal subject and object should be excluded from the sentence-focus category, thereby preserv-
ing the identity between subject- and sentence-focus articulations.

to hypothesize a comparable phenomenon in BH: “Although we do not have

access to the accentuation patterns of BH, I will hypothesize below that also
in BH one and the same construction is used in instances of argument and
sentence focus, viz. the fronting of nonverbal constituents in verbal clauses”
(van der Merwe and Talstra 2002–3: 77). This leads to the following model:
unmarked word order, which is verb-first, is typically used for predicate fo-
cus, while preposing is used to mark argument and sentence focus. It is noted
that not every verb-first clause has predicate focus; the clause may have an
argument-focus structure that is marked only by accenting (van der Merwe and
Talstra 2002–3: 80).
It is clear that argument and sentence focus, as defined by Lambrecht, are
not sufficient to account for all BH preposing. Backgrounded/nonsequential
clauses, for one thing, would not seem to fit either category. Van der Merwe
and Talstra (2002–3: 83) address this issue by broadening the sentence-focus
category to include background clauses. Other preposing types not included in
the paradigm, as noted by van der Merwe and Talstra, include preposed topics
that are compared, contrasted, or listed (e.g., Deut 4:7–8, 4:13–14, 28:67). 82
Heimerdinger’s model is a combination of Lambrecht’s paradigm and a no-
tion he calls the “dominant focal element.” The dominant focal element is the
part of the focus which is “the informationally pivotal element of the assertion”
(1999: 167–68) and is determined by a combination of syntactic and pragmatic
factors. Clauses with predicate focus are usually verb-first, unless the choice of
the dominant focal element is surprising in some respect (1999: 201–6, 212); in
these cases, the dominant focal element is preposed. Another cause of prepos-
ing in the predicate-focus clause is the contrastive-topic construction (1999:
183–85). Argument-focus clauses are always preposed, as are sentence-focus
clauses. Like van der Merwe and Talstra, Heimerdinger widens the sentence-
focus category, this time to include narrative flashbacks.
Shimasaki (2002), who models his theory loosely after Lambrecht, asserts
that the focus in BH is always at the head of the clause. BH is thus apparently
seen as a free word-order language, in which word order is always pragmati-
cally determined (see §2.2). Shimasaki takes an attentional approach to focus,
defining it as a constituent that is marked as a “prominent piece of information”
(2002: 42). All preposing can be explained as argument or clause (= sentence)
focus. Clause focus for Shimasaki is a clause “marked for a high-cost cogni-
tive effort” (2002: 145). Clause-focus clauses may have “inter-clausal level
implications” (2002: 150), including exclamation/proclamation, contrasting
the whole proposition, and circumstantiality. The concept of circumstantial-
ity according to Shimasaki “covers a wide range of connotations,” including
82. It should be noted Lambrecht did not intend his paradigm to cover these pragmatic
sentence types. He also recognizes a separate “contrastive topic” accentual pattern; that is,
The children went to school, and the parents went to bed (1994: 124, 291–95).

anteriority, simultaneity, parenthetical or explanatory information, condition,

concession, and cause. Alternatively, a clause with clause-focus structure may
have a “text-unit level implication” (2002: 163) such as textual-unit onset or
closure, topic announcing (which he terms “topicalization”), and topic shifting.
According to Shimasaki, clauses with clause-focus structure are typically non-
sequential (2002: 148).
The proper definition of the sentence-focus category is an important issue
for word-order models based on Lambrecht. Sentence focus in Lambrecht’s
definition, being restricted to event-reporting and presentational clauses,
clearly does not apply to many preposed clauses. 83 Van der Merwe and Tals-
tra’s proposal to include backgrounded clauses in the sentence-focus category
is problematic, given that a backgrounded clause in an English narrative would
not normally be uttered with the articulation used for event reporting. Even
more problematic is Shimasaki’s catch-all conception of sentence-focus as in-
cluding a wide variety of pragmatic functions.
An even more fundamental issue, however, is whether the link between
preposing and sentence focus, in the strict sense of the term, is valid for classi-
cal BH prose in the first place. Lambrecht (2000: 638) notes the use of subject
fronting for sentence focus in some verb-initial languages, although he states
that it is more typical in these languages for the subject to be placed in the post-
verbal object position. 84 The evidence that preposing marks sentence-focus in
BH is underwhelming. According to more than one scholar, event-reporting
sentences in BH typically have VS, not SV, word order ( Joüon-Muraoka
§155nd; Qimron 1998). 85 As far as presentational clauses, there is a preposed
example of an existential clause in Job (1:1): ‫‘ איש היה בארץ עוץ‬There was
a man in the land of Uz’. In the classical BH prose corpus, however, this
existential structure is extremely rare. 86 The usual formula for an existential
clause involves a nonpreposed ‫ ויהי‬form, as shown in (12): 87
83. Lambrecht’s characterization of sentence focus in a later work (2000: 620–21) is

even narrower, excluding all transitive clauses except for those with pronominal objects (see
p. 43 n. 81 above).
84. In SVO languages, preposing is not used to mark subject focus; instead, sentence-
focus sentences may be given VS order (Lambrecht 2000: 634).
85. Qimron provides a list of examples, including Gen 27:35, 31:1, 42:29–30, 45:16;
Judg 16:23, 16:24; 1 Sam 10:2, 14:29; 2 Sam 3:23, 12:18, 15:10; 1 Kgs 21:13, 14; 2 Kgs 8:7,
9:13. Some of these (Gen 27:35, 2 Sam 3:23, 2 Kgs 8:7) can be categorized as presentational
clauses. Genesis contains two counterexamples to Qimron’s observation: the direct-speech
clauses in Gen 37:20 and 37:33 are event-reporting clauses with SV word order.
86. Two examples that I found in the classical BH prose corpus are 2 Sam 12:1 (in a
direct-speech quotation) and 2 Kgs 7:3.
87. Additional examples from the classical BH prose corpus are Num 9:6; Judg 17:1;
1 Sam 1:1, 9:1; and 2 Sam 21:20.

3.4. Conclusion 46
(12) Judg 13:2

‫ויהי איש אחד מצרעה ממשפחת הדני‬
There was a man from Zorah from the tribe of the Danites.
The other type of presentational clause, as mentioned above, indicates the

appearance of a new referent; in BH, this kind of clause would presumably
contain the verb ‫ בוא‬or a synonym. It has already been remarked by Joüon-
Muraoka (§155nd n. 2) that clauses with ‫ בא‬typically have VS, not SV, order.
A statistical examination of all of the finite nonsubordinate clauses in the clas-
sical BH prose corpus containing the verb ‫ בוא‬in the Qal and having a presenta-
tional meaning confirms this observation: a total of 36 out of 44 (82%) clauses
have VS word order. 88 Although this figure does not prove that preposing does
not mark sentence-focus in the eight preposed clauses, 89 the small number of
clauses involved does not provide convincing evidence that sentence-focus, as
opposed to some other pragmatic category, is the function of preposing in these
clauses. In any case, it is evident that presentational preposed clauses are quite
rare in classical BH prose.
3.4. Conclusion
As the the emphasis-centered model has largely fallen out of favor, back-
grounding/temporal-sequencing and information-structure models dominate
the field of contemporary research on BH word order. In the backgrounding
model and the temporal-sequencing variation of this model, preposing is seen
as marking a characteristic relating to the clause as a whole. These models fall
short of a global paradigm for word order but succeed in explaining some types
of subject-preposing, including parenthetical remarks, anterior clauses, and (in
the case of the temporal-sequencing model) simultaneous clause pairs. The use
of preposing to mark a new narrative unit cannot necessarily be incorporated
in the background concept; however, this function has in common with back-
grounding functions the characteristic that it pertains to the clause as a whole
rather than to the preposed constituent.
In the information-structure approach, the function of preposing is seen as
relating specifically to the preposed constituent. Focusing and topicalization
together account for all of the clauses formerly considered emphatic, contras-
88. SV clauses falling into alternate pragmatic categories already identified, such as si-
multaneous clauses, parenthetical clauses, or topicalization, were eliminated from the count,
because the category of sentence focus is not needed to explain the marked word order in
these cases. The remaining SV clauses are Gen 46:31, 47:1, 48:2; Josh 2:2; Judg 13:6; 1 Sam
23:27; 1 Kgs 13:1; and 2 Kgs 4:42.
89. As explained in §2.1, marking a pragmatic category by preposing is always a choice
made by the speaker/writer; thus, even if preposing could be used to mark sentence-focus,
this does not mean that every sentence with a sentence-focus interpretation, or even the ma-
jority of these sentences, would necessarily be preposed.

47 Conclusion 3.4.
tive, or due to attraction. The division of labor between focusing and topical-
ization is a matter of debate among Hebraists, mirroring a variety of opinions
found in the general linguistic literature. Scholars disagree as to whether it is
informational, contrastive, or attentional focus that is relevant to preposing. An
additional point of disagreement is whether topicalization marks continuity,
discontinuity, or both.
Like the backgrounding/temporal-sequencing models, the information-
structure model cannot serve as a comprehensive framework for BH word
order. Preposing to mark anteriority or simultaneity does not fit either the topi-
calization or the focusing category, nor does preposing clauses marking the be-
ginning of a narrative unit. Although an attempt has been made to integrate the
backgrounding and information-structure models in a global “discontinuity”
paradigm, the two models seem to concern two fundamentally different types
of pragmatic functions. It is likely that both information-structure functions
and backgrounding-type functions are involved in the preposing phenomenon.
It is not clear, however, which type of function is widespread in the biblical
corpus.
The present study explores the significance of information-structure func-
tions for preposing in BH. The concepts of focusing and topicalization are
clarified and redefined so that they provide insights into when and why prepos-
ing occurs. A sample of preposed clauses is examined to determine whether
information-structure functions are statistically dominant or whether functions
that relate to the clause as a whole, such as simultaneity and anteriority, are
the dominant kind. In addition, differences between preposing in narrative and
direct speech are explored. In subsequent chapters, focused and topicalized
clauses are analyzed in detail from the syntactic and the pragmatic perspectives.

Chapter 4
The Biblical Hebrew Finite Clause

and Its Constituents
This chapter defines syntactic categories relating to the BH verbal clause

and its constituents. A structural approach, in which syntactic categories are
defined using formal rather than semantic criteria is followed wherever pos-
sible. Unfortunately, linguistic categories do not always have neat boundar-
ies even when defined in structural terms (Huddleston and Pullum 2002: 90).
A category may have central members that possess all of the characteristics
typically associated with the category, as well as peripheral members that have
only some of these characteristics. Trying to decide which category to assign to
a peripheral item is at times a fruitless and artificial endeavor. My present pur-
pose, however, is not to write a BH grammar but simply to delineate criteria for
selecting and categorizing the clauses in the Genesis corpus. The definitions
set out here, therefore, are designed to minimize the number of cases viewed
as indeterminate.
As noted in chap. 1, the present work analyzes word order in the non-
subordinate finite clause. The terms clause, finite, and nonsubordinate are all
given explicit definition here. In §4.1, I discuss the distinction between sen-
tence and clause. In §4.2, I define finite and nonfinite clauses. In §4.3, I address
the subordinate clause and its various semantic subtypes. Certain problems in
distinguishing subordination from coordination, including the syntactic analy-
sis of direct and indirect speech, are addressed in §4.4. The constituents of the
clause and their syntactic classification are discussed in §4.5. I concentrate on
issues relevant to the Genesis corpus and do not attempt to provide a compre-
hensive description of the BH finite clause.
4.1. The clause and the sentence

A clause minimally includes a predicate. In the verbal clause, the predicate
is the verb; in the nonverbal clause, the predicate is nonverbal and may be, for
example, a noun phrase, prepositional phrase, or subordinate clause. Clauses
are either simple, containing a single predication and no subordinate clauses, or
48

49 The clause and the sentence 4.1.
complex, containing a main predication and one or more subordinate clauses.

Both simple and complex clauses are included in the present study.
The sentence is a syntactic unit consisting of a single clause or a group
of coordinated clauses. Although the idea of the sentence is intuitively clear,
the formal definition of this concept continues to elude linguists. In spoken
language, prosody is used to indicate which clauses go together to constitute
a unit, and punctuation serves the same purpose in written language. Yet it is
not clear that units so delineated always constitute a syntactic sentence. 1 For
example, clauses that most likely constitute two separate sentences may none-
theless appear in a single orthographic sentence, joined by a semicolon or dash:
(13) I saw Marcia yesterday; she looked upset.
In Biblical Hebrew, the quest for the sentence is probably an exercise in

futility. The researcher trying to define the sentence in Biblical Hebrew must
grapple with texts that appear to be one interminably long sentence, because al-
most every clause in narrative begins with the coordinator ‫ו‬. Although the tra-
ditional verse division groups clauses into units, there is no particular reason to
think that verse division corresponds to syntactic sentence boundaries. Given
this situation, I ignore the coordination of clauses to form sentences and treat
each clause, whether simple or coordinated, separately. The only coordination
recognized here is the coordination of subordinate clauses, which generally
involves a clear end boundary (see §4.3).
Determining the boundary of the individual clause is also problematic. Be-
cause verbs in BH inflect for number, gender, and person, clauses need not
contain an explicit subject. If compound verbs are recognized, one faces the
problem of deciding whether two successive verbs with the same implicit sub-
ject constitute two separate clauses or a single clause. In this study, I make the
simplifying assumption that each verb belongs to a separate clause. 2
1. Huddleston and Pullum (2002: 1729) write that the question whether an orthographic
unit marked by punctuation is a syntactic sentence “may have no determinate answer.”
2. One case that cannot be analyzed in this way is ‫ואת כל חילם ואת כל טפם ואת נשיהם‬
‫‘ שבו ויבזו ואת כל אשר בבית‬And all of their wealth and all of their children and their wives
they took captive and took as booty, and everything in the house’ (Gen 34:29). The two verbs
are preceded by a sequence of three coordinated noun phrases and followed by a fourth. The
first and last of the noun phrases, ‫ ואת כל חילם‬and ‫ואת כל אשר בבית‬, serve as a discontinuous
coordinate complement of the second verb, ‫ויבזו‬. The intermediate coordinated phrases ‫ואת‬
‫ כל טפם ואת נשיהם‬relate to the first verb ‫שבו‬. The sense of the sentence is, roughly, ‘they
took captive and took as booty respectively their children and wives, and all of their wealth
and everything in the house’. The interwoven structure of the Hebrew construction makes
it necessary to consider it a single clause with a compound verb phrase rather than two co-
ordinated clauses.

4.2. Finite and nonfinite clauses 50
4.2. Finite and nonfinite clauses

The finite clause is a clause with a finite verb form as predicate. The BH
finite forms are listed in (14).
(14) Finite verb forms

Perfect ‫כָּתַ ב‬
Perfect consecutive ‫ְוכָתַ ב‬
Imperfect ‫ִכתֹּב‬ ְ‫י‬
Imperfect consecutive ‫ִכתֹּב‬ ְ ‫ַויּ‬
Jussive ‫יִבֶן‬
Cohortative ‫ִכ ְתּבָה‬ְ‫נ‬
Imperative ‫ְכּתֹב‬
Jussive and imperfect forms are distinct only in Hiphil verbs, in certain weak
verbs (‫ו‬″‫ ע‬and ‫י‬″‫ )ע‬in Qal, and in ‫ה‬″‫ ל‬verbs in all binyanim.
The nonfinite forms are the infinitive construct (‫)ל ְכתֹּב‬, ִ infinitive absolute
(‫)כָּתוֹב‬, and participle (‫)כּוֹתֵ ב‬. Certain forms are ambiguous in theory, capable
of being interpreted as a participle or a perfect (e.g., ‫קָם‬, ‫) ָבּא‬, but these can usu-
ally be disambiguated in context based on the meaning of the clause in context.
4.3. Subordinate and nonsubordinate clauses

The subordinate clause functions as a constituent of the superordinate
clause that contains it. Clauses that are not subordinate to other clauses are
termed independent or nonsubordinate clauses. 3
Finite subordinate clauses may or may not be marked as such by a subordi-
nator. The most common subordinators are ‫ אשר‬and ‫כי‬. The conditional sub-
ordinate clause has its own subordinators, as described in §4.3.2.1. Infinitive
clauses are always subordinate and are not marked by subordinators. When
several subordinate clauses are coordinated, the subordinator is frequently
omitted in all but the first clause:
(15) Deut 22:7

:‫שלח תשלח את האם ואת הבנים תקח לך למען ייטב לך והארכת ימים‬
You shall let the mother go, and the young you may take for yourself, so
that it may go well with you, and [so that] you may live long.
Although not explicitly marked as such, ‫ והארכת ימים‬is subordinate because it

is coordinated with the preceding subordinate clause, ‫למען ייטב לך‬.
I now discuss the various types of subordinate clauses, including relative
clauses, adjunct clauses, and content clauses. The discussion is limited to the
3. In this study, I use the term nonsubordinate, rather than independent, because the
latter is sometimes used to refer to a clause that is not coordinated with other clauses. A
nonsubordinate clause, in contrast, may be coordinated with another clause.

51 Subordinate and nonsubordinate clauses 4.3.
finite subordinate clause, because nonfinite subordinate clauses are in any case
excluded from the data.
4.3.1. The relative clause

The relative clause modifies a noun phrase and contains an explicit or im-
plicit anaphoric pronoun referring to that noun phrase. Finite relative clauses
in classical BH prose are generally marked by ‫אשר‬:
(16) Gen 18:8

‫ויקח חמאה וחלב ובן הבקר אשר עשה‬
And he took curds and milk and the calf that he had prepared
A relative clause may occur without an antecedent noun phrase:
(17) Gen 43:16

‫ויאמר לאשר על ביתו‬
And he said to [the person] who was in charge of his house
4.3.2. The adjunct clause

The adjunct clause is a subordinate clause that functions as an adjunct in
the superordinate clause (see §4.5 for a definition of the adjunct). The adjunct
clause category as understood in the present study is narrower than in some
other treatments, because many structures traditionally considered to be ad-
junct clauses are treated here as prepositional phrases. In the traditional ap-
proach, words such as ‫‘ אחרי‬after’ and ‫‘ עד‬until’ are defined as prepositions
when their complements are noun phrases but as subordinators when their
complements are clauses. Thus, ‫ עד‬in ‫‘ עד היום הזה‬until this day’ (Exod 10:6)
is a preposition, whereas in ‫‘ עד יצמח זקנכם‬until your beards grow’ (2 Sam
10:5) ‫ עד‬is a subordinator, and the entire structure is an adjunct clause.
According to a prominent school of thought in modern linguistics, however,
prepositions can take clauses as complements. Hebraists adopting this ap-
proach include Waltke and O’Connor (1990: §§38.3–38.7) and Steiner (1997:
169). According to this approach, both ‫ עד היום הזה‬and ‫ עד יצמח זקנכם‬are
prepositional phrases. 4 In the present study, particles that govern noun phrases
and clauses are considered prepositions in both cases. The clause governed by
a preposition is considered a content clause, as discussed in §4.3.3.2.
4. See, e.g., Huddleston and Pullum (2002: 598–601). I differ from Huddleston and Pul-
lum in regarding particles as prepositions only when they have a complement; when they
occur alone, I retain the traditional classification as adverbs. Thus, ‫ אחר‬is a preposition in the
phrase ‫‘ אחר הדברים האלה‬after these things’ but an adverb in the clause ‫‘ ואחר ילדה בת‬and
afterwards she bore a daughter’ (Gen 30:21).

4.3. Subordinate and nonsubordinate clauses 52
A finite adjunct clause may express the semantic categories of time, cause,
condition, or purpose, as discussed in §§4.3.2.1–4.3.2.4. The various semantic
types can in some cases be distinguished by their characteristic subordinators.
4.3.2.1. The conditional adjunct clause
Conditional clauses are marked by the subordinators ‫ אם‬and ‫כי‬. Hypotheti-
cal conditionals are marked by the subordinator ‫לו‬, and negative hypotheticals
by ‫לולי‬. An example of each subordinator is shown in (18)–(21).
(18) ‫אם‬
:‫אם אמצא בסדם חמשים צדיקם בתוך העיר ונשׂאתי לכל המקום בעבורם‬
If I find in Sodom fifty righteous ones within the city, I will forgive
the whole place for their sake. (Gen 18:26)
(19) ‫כי‬
‫וכי יגח שור את איש או את אשה ומת סקול יסקל השור ולא יאכל את בשׂרו‬
If an ox gores a man or a woman and s/he dies, the ox shall be stoned,
and its flesh shall not be eaten. (Exod 21:28)
(20) ‫לו‬
‫ להמיתנו לא לקח מידנו עלה ומנחה ולא הראנו את כל אלה וכעת‬′‫לו חפץ י‬
‫לא השמיענו כזאת‬
If the LORD had meant to kill us, he would not have accepted a burnt
offering and a grain offering at our hands, and would not have shown us
all these things, and he would not now have announced to us such things
as these. (Judg 13:23)
(21) ‫לולא‬
‫לולי אלהי אבי אלהי אברהם ופחד יצחק היה לי כי עתה ריקם שלחתני‬
If the God of my father, the God of Abraham and the Fear of Isaac,
had not been with me, you would have sent me away empty-handed.
(Gen 31:42)
In oaths, ‫ אם‬may be used without an explicit apodosis:
(22) Gen 14:22–23 5

‫ אם מחוט ועד שרוך נעל ואם‬:‫ אל עליון קנה שמים וארץ‬′‫הרימתי ידי אל י‬
‫אקח מכל אשר לך‬
I swear to the LORD, God Most High, Maker of heaven and earth, if I
take a thread or a sandal-strap or anything that is yours [may I be
punished].
5. Additional examples from the Genesis corpus are Gen 21:23, 26:29, 31:52, and 42:15.

According to Richard Steiner (personal communication) “an oath may be for-

mulated as a conditional and a self-imprecation is understood as the apodosis:
‘if I do such and such (may God do such and such to me as a punishment)’.” 6
The full formula with the apodosis can be found in verses such as 1 Sam 3:17
and 2 Kgs 6:31.
In addition to its use as a conditional subordinator, ‫ לו‬serves as an optative
adverb expressing a wish (GKC §151e; Joüon-Muraoka §163): 7
(23) Gen 17:18
:‫לו ישמעאל יחיה לפניך‬
O that Ishmael might live before you!
Optative ‫ לו‬could theoretically be explained as a subordinator marking a con-
ditional clause with a missing apodosis, by analogy to oaths with ‫אם‬. 8 Because
the full formula with the apodosis does not occur in the case of ‫לו‬, however, it
is preferable to view optative ‫ לו‬as an adverb from a synchronic perspective,
whether or not it derives historically from the conditional subordinator.
4.3.2.2. The causal adjunct clause
Finite clauses expressing cause are marked in Genesis by the subordinator
‫כי‬:
(24) Gen 2:5
:‫ אלהים על הארץ‬′‫וכל עשב השדה טרם יצמח כי לא המטיר י‬
And no grasses of the field had yet sprouted because the LORD God had
not caused it to rain upon the earth.
4.3.2.3. The temporal adjunct clause

Temporal finite clauses are marked by ‫ כי‬or, rarely, by ‫אם‬, as shown by the
parallel verses in (25) and (26) ( Joüon-Muraoka §166p).
(25) Deut 12:20
‫ אלהיך את גבולך כאשר דבר לך ואמרת אכלה בשר כי תאוה‬′‫כי ירחיב י‬
‫נפשך לאכל בשר בכל אות נפשך תאכל בשר‬
When the LORD your God enlarges your border, as He promised you,
and you say, “I shall eat meat,” because you have the urge to eat
meat, you may eat meat whenever you desire.
6. According to a different view, ‫ אם‬in oaths should be viewed from a synchronic per-
spective as a negative asseverative particle meaning ‘certainly not’ (see, e.g., GKC §150;
Joüon-Muraoka §166).
7. See also Gen 30:34.
8. In GKC (§151e n. 1), this is suggested as a historical explanation of the optative
use. In Huehnergard’s (1983: 575) view, optative ‫ לו‬does not derive from the conditional;
rather, ‫ לו‬was historically a particle used to mark hypothetical propositions, including those
expressed by unreal conditionals and optative clauses.

4.3. Subordinate and nonsubordinate clauses 54
(26) Deut 19:8

‫ אלהיך את גבלך כאשר נשבע לאבתיך ונתן לך את כל הארץ אשר‬′‫ואם ירחיב י‬
:‫דבר לתת לאבתיך‬
And when the LORD your God enlarges your border, as he has sworn to
your fathers, and gives you all the land that he promised to give to your
fathers
4.3.2.4. The purpose adjunct clause

Finite adjunct clauses expressing negative purpose are marked by ‫‘ פן‬lest’:
(27) Gen 19:15

:‫קום קח את אשתך ואת שתי בנתיך הנמצאת פן תספה בעון העיר‬
Arise, take your wife and your two daughters who are here, lest you be
consumed because of the iniquity of the city.
Positive purpose adjuncts are usually expressed by content clauses governed by

the prepositions ‫‘ בעבור‬for the sake of’ or ‫‘ למען‬for the sake of’ (see §4.3.3.3).
4.3.3. The content clause

Content clauses are the default type of subordinate clause, lacking special
features found in other types and being less differentiated syntactically from
main clauses (Huddleston and Pullum 2002: 950). 9 Content clauses are marked
by ‫ כי‬and ‫אשר‬.
It should be noted that ‫ כי‬is not always a subordinator: it also used as an
adverb marking the apodosis of a conditional clause, in the combination ‫כי עתה‬
( Joüon-Muraoka §167s):
(28) Gen 43:10

‫כי לולא התמהמהנו כי עתה שבנו זה פעמים‬
For if we had not delayed, we could have already returned twice.
Another use of ‫ כי‬is as an adverb meaning ‘but, rather’ (Blau 1993: §104): 10
9. Content clauses are also known as complement clauses.

10. The words ‫ כי עתה‬and ‫‘ כי‬rather’ are categorized as clausal adverbs; for further
discussion, see §5.2.1. Many also recognize an asseverative adverb ‫‘ כי‬surely, indeed’ (Mui-
lenberg 1961; Schoors 1981: 160; Muraoka 1985: 160–61; Waltke and O’Connor 1990:
§39.3.4e; HALOT s.v. ‫ ;כי‬Follingstad 2001). Bandstra (1982) argues against the existence of
asseverative ‫כי‬. Claassen (1983) and Aejmelaeus (1986) show that many supposed instances
of the asseverative ‫ כי‬involve the causal particle, which functions on various levels in the
discourse. According to Aejmelaeus the asseverative use is the most probable explanation
of ‫ כי‬in oaths (e.g., Gen 42:16) and in cases in which ‫ כי‬occurs in the middle of the clause,
(e.g., Gen 18:20). Both of these examples from Genesis are in nonverbal clauses. A probable
instance of asseverative ‫ כי‬in Genesis in a verbal clause is Gen 22:17.

(29) Gen 17:15

‫שרי אשתך לא תקרא את שמה שרי כי שרה שמה‬
Sarai your wife, you shall not call her Sarai; rather, her name shall be
Sarah.
Content clauses have a variety of syntactic functions, as described in

§§4.3.3.1–4.3.3.2.
4.3.3.1. Complement of verb
Finite content clauses are frequently the complements of verbs 11 and as such
are usually marked by ‫כי‬:
(30) Gen 30:1

‫ותרא רחל כי לא ילדה ליעקב‬
And Rachel saw that she had borne Jacob no children.
Less frequently, they are marked by ‫אשר‬: 12

(31) 1 Sam 18:15
‫וירא שאול אשר הוא משכיל מאד‬
And Saul saw that he was very successful
4.3.3.2. Complement of preposition

A finite content clause may be the complement of a preposition. Preposi-
tions governing content clauses include ‫‘ אחרי‬after’, ‫‘ כ‬when’ or comparative
‘as’, ‫‘ עד‬until’, ‫‘ מאז‬from the time’, ‫‘ למען‬for the sake of’, ‫‘ בעבור‬for the sake
of’, and ‫‘ יען‬because of’. Some of these require the content clause to be marked
by a subordinator such as ‫ אשר‬or ‫כי‬. Some prepositions, including ‫ יען‬and
‫בעבור‬, allow the content clause to be marked or unmarked:
(32) Num 11:20

‫ אשר בקרבכם‬′‫יען כי מאסתם את י‬
Because you have rejected the LORD who is among you
(33) Num 20:12

‫יען לא האמנתם בי להקדישני לעיני בני ישראל‬
Because you did not trust me, to sanctify me in the eyes of the people
of Israel
11. A content clause that is the complement of a verb is also known as an object clause
or an objective complement clause. A content clause can also be the subject of a verb, but
finite examples in BH are rare. On verbal complementation and, in particular, indirect speech
complements, see Miller (2003: 95–129).
12. Other examples (from Joüon-Muraoka §157c) are Exod 11:7, Deut 1:31, 1 Kgs 22:16.

4.4. Difficult issues in defining subordination 56
A prepositional phrase containing a content clause may serve as an adjunct

expressing time, purpose, cause, comparison, or another semantic concept.
Some common types are:
(34) Temporal adjunct
:‫וישבת עמו ימים אחדים עד אשר תשוב חמת אחיך‬
And stay with him a while, until your brother’s fury subsides (Gen
27:44)
(35) Purpose adjunct

‫אמרי נא אחתי את למען ייטב לי בעבורך‬
Please say that you are my sister so that it may go well with me because
of you. (Gen 12:13)
(36) Causal adjunct

‫על כן היתה חברון לכלב בן יפנה הקנזי לנחלה עד היום הזה יען אשר מלא‬
:‫ אלהי ישׂראל‬′‫אחרי י‬
So Hebron became the inheritance of Caleb the son of Jephunneh the
Kenizzite to this day, because he wholeheartedly followed the LORD,
the God of Israel. (Josh 14:14)
(37) Comparative adjunct

‫וכאשר יענו אתו כן ירבה וכן יפרץ‬
And as they oppressed them, so they multiplied. (Exod 1:12)
4.4. Difficult issues in defining subordination

In this section, we discuss certain problematic issues relating to subordina-
tion: the seemingly subordinate use of the conjunction ‫ו‬, the syntax of the / ‫והיה‬
‫ ויהי‬construction, and the syntax of direct and indirect speech.
4.4.1. The conjunction ‫ו‬: Coordinator and subordinator?

A coordinated clause with ‫ ו‬may occasionally substitute for a subordinated
clause. Such a clause may be semantically equivalent to a conditional clause
( Joüon-Muraoka §167b), a purpose clause ( Joüon-Muraoka §168b), 13 a com-
plement clause (Steiner 1997: 169), or a relative clause (Steiner 1997: 169). An
example of each type is shown in (38)–(41).
(38) Conditional
:‫ועזב את אביו ומת‬
And (= if) he leaves his father, he will die (Gen 44:22)
13. In this case, a volitive verb form is usual.

57 Difficult issues in defining subordination 4.4.
(39) Purpose
‫והביאה לי ואכלה‬
And bring it to me, and (= so) I may eat it (Gen 27:4)
(40) Complement
‫ואם ידעת ויש בם אנשי חיל‬
And if you know and (= that) there are capable men among them
(Gen 47:6)
(41) Relative
‫הבה נבנה לנו עיר ומגדל וראשו בשמים‬
Let us build a city and a tower and its (= whose) head is in the heavens
(Gen 11:4)
Scholars are divided as to the proper syntactic classification of clauses with

a “subordinate” ‫ו‬. Van der Merwe et al. (1999: §40.8) recognize a separate use
of ‫ ו‬as a subordinating conjunction. Waltke and O’Connor (1990: §38.1h), in
contrast, write that “the system expressed in the text may skew the unexpressed
semantic system”; that is, the formally nonsubordinate ‫ ו‬clause may be used to
express a logically subordinate idea. Steiner (1997: 168) takes an intermediate
view between these positions, stating that “the boundary between coordination
and subordination in BH is not as sharp as in English.” He views ‫ ו‬as a univer-
sal connector that can be used to connect coordinated or subordinated clauses.
Because no formal means of distinguishing “subordinate” from coordinate ‫ו‬
are available, the two types are not regarded as distinct grammatical categories
in the present work. Steiner’s fuzzy-boundary solution is undesirable for pres-
ent purposes because it greatly complicates the classification of the data.
4.4.2. The ‫ ויהי‬construction

The syntactic analysis of the ‫והיה‬/‫ ויהי‬construction is a contentious point in
BH syntax. Several varieties of this construction are shown in (42)–(44).
(42) Gen 4:8

:‫ויהי בהיותם בשׂדה ויקם קין אל הבל אחיו ויהרגהו‬
And it came to pass when they were in the field, and Cain rose up against
his brother Abel, and he killed him.
(43) Gen 8:13

‫ויהי באחת ושש מאות שנה בראשון באחד לחדש חרבו המים מעל הארץ‬
And it came to pass in the six hundred and first year, in the first month,
on the first of the month, the waters dried up from the earth.

(44) Gen 4:14

:‫והיה כל מצאי יהרגני‬
And it will come to pass; anyone who meets me may kill me.
In (42), ‫ ויהי‬is followed by a temporal phrase and then by a finite clause begin-
ning with a coordinator. Example (43) is similar except that the finite clause
lacks the coordinator. In (44), ‫ ויהי‬is followed immediately by a finite clause,
without a coordinator.
In a common analysis of these constructions, the finite clause is taken to be
a subordinate clause functioning as the subject of ‫ויהי‬. 14 Example (42), in this
view, is best rendered ‘And it came to pass when they were in the field [that]
Cain rose up against his brother’. 15 Alternatively, ‫ ויהי‬can be considered a sub-
jectless predicate and the finite clause nonsubordinate. The latter analysis is
supported by the fact that finite clauses following ‫ ויהי‬are never marked with a
subordinator such as ‫ אשר‬and frequently begin with the coordinator ‫ו‬. In keep-
ing with my preference for syntactic classification based on formal criteria, I
adopt the latter approach.
The subjectless interpretation of ‫ ויהי‬constructions has the additional advan-
tage of avoiding the problem of determining the end-boundary of the purported
subject clause. For example, in Gen 4:8, above, the subject clause would pre-
sumably include the two coordinated clauses describing the event that hap-
pened while they were in the field: ‘And Cain rose up against his brother Abel
and killed him’. In some instances an argument could be made that the entire
following narrative unit should be considered to be one long coordinated sub-
ject clause, as in (45):
(45) Gen 22:1

‫ויהי אחר הדברים האלה והאלהים נסה את אברהם ויאמר אליו אברהם‬
And it came to pass after these things, and God tested Abraham, and He
said to him, “Abraham!”
14. See, e.g., BDB §1961.2; Longacre (1989: 67); Niccacci (1990: 160). For a survey
of these and other approaches to the syntactic and pragmatic analysis of ‫ויהי‬, see van der
Merwe (1999a). According to van der Merwe et al. (1999: §44.5), ‫ ויהי‬is a discourse marker
that “anchors an event, state of affairs, scene, episode or narrative to the time line.” In this
analysis, ‫ ויהי‬presumably does not function as a verb.
15. This analysis assumes that the temporal phrase serves as an adjunct in the matrix
clause. Another possibility is that the temporal phrase belongs to the following finite clause,
as suggested in Steiner (1979: 148). The verse would then be translated ‘And it came to pass
that when they were in the field Cain rose up’. In this interpretation, ‫בהיותם בשדה ויקם קין‬
is viewed as a clause with a preverbal adjunct connected to its clause by a conjunction (see
§5.5.2). The case for taking the temporal phrase as belonging to the following finite clause
is stronger in cases such as (43), where the finite clause does not start with a conjunction
(Richard Steiner, personal communication).

59 Difficult issues in defining subordination 4.4.
4.4.3. Direct and indirect speech

BH exhibits two types of reported speech, direct and indirect speech (or
direct and indirect discourse). 16 The two types can be distinguished based on
their deixis, that is, the way in which elements in the utterance such as personal
and demonstrative pronouns, tense, and temporal and spatial adverbs relate to
the spatiotemporal context of the act of utterance (Lyons 1977: 2:636; Levin-
son 1983: 54). In direct speech, deictic elements are interpreted in relation to
the context of the quoted utterance: I and you refer to the reported speaker and
his addressee, rather than to the reporting speaker and addressee. In indirect
speech, the deictic elements are interpreted in relation to the context in which
the speech is reported. First- and second-person pronouns refer to the reporting
speaker and his addressee, while the quoted speaker and addressee are referred
to by third-person pronouns. The difference is illustrated by (46) and (47):
(46) Gen 29:25

‫ויאמר אל לבן מה זאת עשׂית לי‬
And he said to Laban, “What is this you have done to me?”
(47) Gen 29:12

‫ויגד יעקב לרחל כי אחי אביה הוא‬
And Jacob told Rachel that he was her father’s kinsman
The direct speech clause in (46) refers to the speaker by a first-person enclitic
pronoun, ‫‘ לי‬to me’. The addressee is referred to implicitly by means of a verb
inflected for second person. The indirect speech clause in (47), in contrast, re-
fers to the speaker by the third-person independent pronoun ‫‘ הוא‬he’ and to his
addressee by a third-person enclitic pronoun, ‫‘ אביה‬her father’.
A complex relationship exists between quotations and speech that is re-
ported. Researchers have found that direct speech and indirect quotations
never exactly reproduce the reported utterance (Tannen 1986). Quoters are
generally unable to remember precisely what has been said and often delib-
eratively shorten or otherwise alter utterances in quoting them. Some direct
speech quotations relate not to an actual utterance but to an utterance that a
person could have said or did not say: Why didn’t you tell me, “Don’t go to the
beach today!’” Other quotations represent an internal speech: I said to myself,
“I must go to sleep right now.” Tannen concludes that all quoted speech is “con-
structed dialogue,” whether the quotation relates to a real or fictitious event. 17
16. For a comprehensive discussion of the two types of reported speech and their syntac-
tic and pragmatic characteristics, see Miller (2003).
17. The constructed nature of quotations is reflected in requotations in the Bible. Mali
(1983) and Savran (1988) show that when one biblical character’s speech is requoted by
another, the second quotation nearly always exhibits changes such as shortening, lengthen-
ing, or paraphrase.

The syntactic status of indirect speech in BH is fairly clear: the indirect

speech quotation serves as the complement of the verb in the reporting clause.
Like other complement clauses, indirect speech is usually marked by a subor-
dinator such as ‫ כי‬and must consist of a complete clause or clauses rather than
a sentence fragment (Miller 2003: 76–77). 18
Direct speech clauses in BH exhibit signs of syntactic independence. 19 Di-
rect speech is not introduced by the subordinator ‫ כי‬20 and occasionally lacks
a reporting clause. 21 Unlike indirect speech, direct speech may contain voca-
tives and exclamations and may consist of a sentence fragment (Miller 2003:
74–75). 22
Some direct speech clauses, however, have traits suggesting they are sub-
ordinated. For example, the direct speech clause is often obligatory in the re-
porting clause (Miller 2003: 74–75, 199–220), suggesting that it is in fact a
complement clause. It has been claimed that ‫ לאמר‬functions as a subordinator
introducing a certain type of direct speech clause (Miller 2003: 163–212); if
this is so, citations following ‫ לאמר‬are subordinated. We may conclude that
direct speech clauses vary in their degree of subordination to or independence
from the reporting clause. 23
18. Miller (2003: 120) notes that when an indirect speech quotation is embedded in
direct speech, the subordinator may be omitted; e.g., Gen 12:13.
19. Partee (1973) takes the view that direct speech is syntactically independent in En-
glish. See also the discussion in Munro (1982) and Li (1986).
20. It has been claimed that direct speech quotations, like other complement clauses,
may be introduced by ‫( כי‬the so-called kî recitativum); see, e.g., GKC (§157b); Williams
(1976: §452); Joüon-Muraoka (§157c); Goldenberg (1991: 79–96). According to this view,
‫ כי‬at the beginning of a direct speech citation belongs to the reporting clause rather than to
the quoted utterance. An example is ‫( ויאמר כי את שבע כבשת תקח מידי‬Gen 21:30), which
is understood to mean something like “And he said that ‘These seven ewes you shall take
from my hand.’ ” Some other supposed instances of this phenomenon are Gen 29:32; Exod
3:12, 4:25; 1 Kgs 21:6; 2 Kgs 8:13. Other scholars strongly object to the kî recitativum
theory, arguing that ‫ כי‬introducing direct speech is best interpreted as belonging to the quoted
utterance; see Esh (1957); Schoors (1981: 256–59); Aejmelaeus (1986); Bandstra (1982:
165–66); Miller (2003: 103–15). A look at the conversation preceding the quoted utterance
reveals in almost all cases that ‫ כי‬is one of the following: a causal subordinator (“because”),
a temporal subordinator (“when”) (see §4.3.2.3), or a conjunction (“rather”) (see §4.3.3). For
example, in Gen 21:30, Abraham’s utterance is a response to Abimelek’s preceding question,
“What are these seven ewes. . . ?” Rather than respond to the literal meaning of this question,
Abraham answers the implied question “Why did you put aside these seven ewes. . . ?” His
response is properly rendered, “And he said, ‘Because these seven ewes you will take from
my hand.’ ” Schoors (1981: 258–59) concludes that “the kî recitativum, as a specific syntac-
tic category, should be deleted from grammars and dictionaries.”
21. An example is 2 Kgs 10:15.
22. An example is Gen 42:7.
23. In a similar vein, Quirk et al. (1985: 1023) write that in English “there is a gradient
from direct speech that is clearly independent to direct speech that is clearly integrated into
the clause structure.”

61 Syntactic classification of the constituents of the clause 4.5.
Hatav (2000a) claims that there is a third type of reported speech, “free
direct discourse,” which is marked by ‫לאמר‬. Unlike direct speech, which pur-
ports to be an exact rendition of the original utterance, free direct discourse
represents “more or less” what the original speaker intended to say and in
some cases does not report speech at all. 24 Miller (2003: 199), however, as-
serts that quotations prefaced by ‫ לאמר‬are direct speech, although they may
be condensed, hypothetical, or fabricated (Miller 2003: 351). 25 Miller (2003:
412–18) presents a number of arguments against Hatav’s view, pointing out
the implausibility of considering divine legislation introduced by ‫ לאמר‬to have
been “flagged by the narrator/writer as not necessarily presented accurately”
(2003: 417). Considering Tannen’s (1986) finding that reported speech is never
an exact representation of the original utterance, it would seem that the ‫לאמר‬
quotation simply represent the end of the direct speech spectrum in terms of its
distance from the quoted utterances.
Direct speech quotations can be analyzed on two levels: the level of the re-
ported utterance and the level of the reporting clause (Miller 2003: 200–201).
A direct speech quotation functions as a nonsubordinate clause on the level of
the reported utterance, whether or not it is subordinated to the reporting clause.
Due to the uncertainty regarding their syntactic status on the level of the report-
ing utterance, direct speech clauses are analyzed in this study exclusively on
the level of the reported utterance.
4.5. Syntactic classification of the constituents of the clause

Clauses are made up of constituents, groups of words that form a syntactic
unit. The main constituents participating in the structure of the verbal clause
are subjects, predicates (i.e., verbs), complements, and adjuncts. A difficult
question that is not resolved here is whether BH predicates and complements
participate directly in clause structure or combine to form a verb phrase. On the
surface, BH does not seem to involve a VP, because the subject regularly inter-
venes between the verb and its complements. However, as discussed in chap. 2
(pp. 9–10), according to generative grammar, VSO languages have a VP con-
stituent in their underlying structure, with the position of the verb resulting
from some sort of movement. Despite the theoretical importance of the VP
issue, it is not particularly relevant to the surface description of BH preposing.
In the present work, complements and adjuncts are referring to as clause-level
constituents, without intending to reject the possibility of an underlying VP.
24. See also Follingstad (2001: 453–555), who considers quotations introduced by ‫לאמר‬
as “semi-direct discourse.” The word ‫ לאמר‬according to Follingstad indicates that the origi-
nal citation “has been paraphrased, summarized, reinterpreted—i.e., recontextualized—by
the reporting speaker (or narrator), typically with regards to another context” (2001: 546).
25. Even direct speech quotations without ‫ לאמר‬do not always relate to actual speech
situations, as both Miller (2003: 290–96) and Hatav (2000a: 15) note.

4.5. Syntactic classification of the constituents of the clause 62
A clause-level constituent can be classified according to its syntactic class

(e.g., noun phrase, prepositional phrase, verb) as well as its syntactic function
(subject, verb, complement, or adjunct.) The subject is the clause constituent
that triggers verb agreement: ‫‘ ויכתב האיש‬and the man wrote’ versus ‫ויכתבו‬
‫‘ האנשים‬and the men wrote’. The subject is generally a noun phrase but is on
rare occasions a nonfinite clause. 26 Subject pronouns take the form of indepen-
dent personal pronouns, e.g., ‫‘ אני‬I’, ‫‘ אתה‬you’.
Complements are for the most part clearly distinguishable from subjects.
Unlike subjects, complements do not trigger verb agreement. Again unlike
subjects, complements may consist of a prepositional phrase headed by ‫את‬
(although in rare instances ‫ את‬governs what appears to be the subject of the
clause). 27 Complement pronouns may take the form of enclitic pronouns at-
tached to the verb (‫‘ ראיתיו‬I saw him’).
The distinction between complement and adjuncts is much less straight-
forward. Semantically, complements generally correspond to arguments of
the semantic predicate, while adjuncts express circumstantial concepts such
as cause, time, and purpose. A structural distinction can be made between the
most central members of the two categories. 28 The prototypical complement is
a noun phrase, whereas the prototypical adjunct is an adverb, adverb phrase,
or prepositional phrase. The prototypical complement is a required clause el-
ement, while the prototypical adjunct is optional. Prototypical complements
may be advanced to subject when the verb is changed to a passive binyan (Piel
to Pual, Hiphil to Hophal); adjuncts are not likewise advanced.
Unfortunately, many peripheral types of adjuncts and complements are dif-
ficult to categorize. Some complements are optional, resembling adjuncts.
Some adjuncts are noun phrases, resembling complements. Another difficult
case concerns verbs of motion and location, which may require a constitu-
ent expressing source, goal, or location. These constituents are considered by
some to be complements because they are required and are restricted to certain
classes of verbs; others consider these a separate class of “required adjuncts.”
26. See, e.g., Gen 4:26.
27. The particle ‫ את‬may mark what one would expect to be the subject in passive con-
structions; see also, e.g., Gen 4:18, 21:5, and 27:42. According to Steiner (1997: 160), in con-
structions containing a certain type of medio-passive verb, the object of the corresponding
active sentence regularly remains an object in the passive construction and is not advanced
to subject; thus, this is not really a case of ‫ את‬governing the subject. In these constructions,
the verb is usually in the third person singular and the object retains the accusative marker.
In rare cases, ‫ את‬governs the subject of an intransitive verb; e.g., 1 Sam 17:34, Judg 20:44,
2 Kgs 6:5. Studies of the so-called nominative ‫ את‬are cited in Waltke and O’Connor (1990:
§10.3b).
28. On the complement/adjunct distinction in English, see Huddleston (1984: 177–
225); Matthews (1981: 121–45); Quirk et al. (1985: 723–38); Andrews (1985: 62–153). On
complements versus adjuncts in Biblical and Modern Hebrew, see, e.g., Azar (1972); Blau
(1973a); Lerner (1975); Ben-Asher (1973: 54–71); Muraoka (1979).

63 Syntactic classification of the constituents of the clause 4.5.
In addition to the theoretical controversies regarding the definitions of com-

plement and adjunct, formidable difficulties are involved in distinguishing the
two types in practice in BH. The optionality of a constituent is a central criterion
in the classification. Yet deciding whether a constituent is optional frequently
requires an in-depth investigation into the meaning or meanings of the clause
verb and the various constructions associated with each meaning. Attempting
such an investigation is not practical when analyzing a large textual corpus.
Due to the theoretical and practical difficulties involved, I do not systemati-
cally differentiate between complements and adjuncts in the present study.

Chapter 5
The Syntax of Preposing and Other

Word-Order Constructions
A preposed clause in BH generally corresponds to a semantically equiva-

lent verb-first clause; that is, a systematic correspondence exists between the
marked and unmarked constructions. In this chapter, I describe the syntactic
varieties of preposing and identify the unmarked constructions to which they
correspond. In addition, I distinguish preposing from several similar construc-
tions that are not the subject of the present work.
In §5.1, I describe the basic preposing construction and show that almost
any verb-first clause in BH can be altered to produce a grammatical and seman-
tically equivalent preposed clause. Although verb-first word order is unmarked
in BH, not every preverbal constituent is considered to be preposed. In §5.2,
I discuss constituents that normally precede the verb and are not preposed as
such. In some cases, an element is obligatorily preposed, producing an un-
marked preposed structure. Structures of this type are described in §5.3.
Complex variants of the preposing construction, including preposing with
a focus adverb and double preposing, are discussed in §5.4. In §5.5, I describe
marked word-order constructions that resemble preposing but are syntactically
distinct from it. In §5.6, I present statistical data regarding the distribution of
the various word-order constructions in Genesis. Conclusions are summarized
in §5.7.
The syntactic generalizations in this chapter apply to the classical BH prose
corpus, that is, the prose portions of Genesis through Kings. The book of Gen-
esis was analyzed in its entirety and the conclusions tested with respect to the
rest of the corpus by means of computerized searches, as noted in chap. 1.
5.1. Preposing
Virtually every verb-first construction can be converted to a preposed one,
although this usually requires alterations beyond a simple word-order change.
If the unmarked clause contains a consecutive verb form, this is replaced with a
simple perfect or imperfect in the preposed construction. The contrast between
the verb forms in unmarked and preposed clauses is illustrated by (48), which
64

65 Preposing 5.1.
has two coordinated clauses, the first unmarked with a consecutive verb and
the second preposed with a perfect:
(48) Gen 1:5

‫ויקרא אלהים לאור יום ולחשך קרא לילה‬
And God called the light Day, and the darkness he called Night.
Even clauses consisting of a bare finite verbal form can be converted to a

preposed structure. This is accomplished by retaining the normally dropped
personal pronoun and preposing it, as in (49): 1
(49) Gen 21:24

:‫אנכי אשבע‬
I swear.
Clauses containing volitive verb forms can undergo preposing as well, as

shown in the following examples:
(50) Imperative 2
‫והרכש קח לך‬
And the possessions take for yourself (Gen 14:21)
(51) Jussive 3
‫ועינכם אל תחס על כליכם‬
And never mind about your belongings (Gen 45:20)
(52) Cohortative 4
‫ואני והנער נלכה עד כה‬
And I and the lad shall go over there (Gen 22:5)
It is even possible to prepose the ordinarily dropped subject of an imperative:
(53) Imperative 5
‫ואתם האסרו‬
And you be (= remain) confined (Gen 42:16)
1. Preposed personal pronouns are quite common; representative examples from the
Genesis corpus include 10:8, 10:9, 14:23, 15:15, and 33:3.
2. Preposing with imperatives is not uncommon; examples from Genesis include 6:21;
8:17; 14:21; 19:12, 17; 20:13, 15; 23:6, 15; 24:60; 31:16, 32; 42:16, 18, 19, 33 (2×); 43:11,
12, 13; 44:17; 45:17, 19; 47:6.
3. Examples from Genesis include 1:22, 37:27, 44:33, and 45:20.
4. Examples from Genesis include 22:5 and 33:14.
5. Other representative examples are Gen 24:60, 42:19, and 44:17.

5.1. Preposing 66
One construction that cannot be preposed is the ‫והיה‬/‫‘( ויהי‬and it came/will

come to pass’) clause. Because there is no implicit pronominal subject in this
structure, a personal pronoun cannot stand in front of the verb. Thus, (54) for
example, cannot be changed to ‫הוא היה בנסעם מקדם‬.
(54) Gen 11:2

‫ויהי בנסעם מקדם‬
And it came to pass when they traveled from the East
Although one might expect the temporal phrase following ‫ ויהי‬clause to be

preposable, this does not in fact occur; that is, (54) cannot be altered to ‫בנסעם‬
‫מקדם היה‬. It seems that ‫ היה‬has the meaning ‘come to pass’ only in the waw-
consecutive form (BDB §1961). Note that when the verb has copular meaning,
preposing is perfectly appropriate, as in (55).
(55) Gen 29:17

:‫ורחל היתה יפת תאר ויפת מראה‬
And Rachel was beautiful in form and beautiful in appearance.
The syntactic function of a preposed constituent may be subject, comple-

ment, or adjunct:
(56) Subject
‫וכוש ילד את נמרד‬
And Cush fathered Nimrod (Gen 10:8)
(57) Complement
‫ואת הצפר לא בתר‬
And the bird he did not cut in two (Gen 15:10)
(58) Adjunct
‫בעוד שלשת ימים ישא פרעה את ראשך מעליך‬
In another three days Pharaoh will lift up your head from you (Gen
40:19)
The syntactic category of the preposed constituent may be noun phrase, prepo-
sitional phrase, adverb, or adjective:
(59) Noun phrase

:‫ואביו שמר את הדבר‬
And his father kept the matter in mind. (Gen 37:11)

67 Preposing 5.1.
(60) Prepositional phrase

‫ולחשך קרא לילה‬
And the darkness He called Night (Gen 1:5)
(61) Adverb
:‫שמה קבר אברהם ושרה אשתו‬
There Abraham and his wife Sarah were buried (Gen 25:10)
(62) Adjective
‫תמים היה בדרתיו‬
Blameless was He in his generation (Gen 6:9)
Preposing is normally performed on an entire clause-level constituent.
When the constituent involved is a compound phrase, there are several prepos-
ing options. In most cases the entire phrase is preposed, even when it is a long
coordinated chain: 6
(63) Gen 47:1
‫אבי ואחי וצאנם ובקרם וכל אשר להם באו מארץ כנען‬
My father and my brothers, and their flocks and herds and all that
is theirs, have come from the land of Canaan
An alternative is to prepose the first part of the compound phrase and postpose
the remainder to the end of the clause. 7 The preposed element may or may
not be resumed by a pronoun in the postposed phrase. The resumptive type is
shown in (64) and (65) and the nonresumptive type in (66) and (67). 8
(64) Gen 44:3
:‫והאנשים שלחו המה וחמריהם‬
And the men were sent away, they and their asses
(65) Gen 17:9

:‫ואתה את בריתי תשמר אתה וזרעך אחריך לדרתם‬
And you, keep my covenant, you and your offspring after you
throughout their generations.
6. Additional examples of compound noun phrases preposed in their entirety in the Gen-
esis corpus include Gen 9:2; 20:5; 22:5; 27:37; 31:38, 42; 35:11; 46:32; and 47:5.
7. Postposing of part of a compound phrase is common in verb-first clauses as well,
particularly when the compound constituent is the subject. Representative examples from
the Genesis corpus include Gen 6:18; 33:6; 35:6; and 50:14, 22. For discussion of postponed
subjects with resumptive pronouns, see Muraoka (1985: 62–63); Waltke and O’Connor
(1990: §16.3.2c); Joüon-Muraoka §146c; Naudé (1999). Naudé (1999: 97) states that clause-
final phrases such as ‫ המה וחמריהם‬occupy the same position as right-dislocated constituents.
8. See also Gen 24:38, 43:15.

5.2. Nonpreposed preverbal constituents 68
(66) Gen 26:26

:‫ואבימלך הלך אליו מגרר ואחזת מרעהו ופיכל שר צבאו‬
And Abimelech went to him from Gerar, and Ahuzzath his adviser
and Phicol the commander of his army.
(67) Gen 44:2

‫ואת גביעי גביע הכסף תשׂים בפי אמתחת הקטן ואת כסף שברו‬
And my cup, the silver cup, put in the mouth of the sack of the young-
est, and his money for the grain
Preposing together with postposing occasionally occurs in clauses containing a

noun phrase modified by a prepositional phrase. An example is (68), where the
phrase ‫ את האנשים אשר פתח הבית‬is preposed, and its modifying phrase ‫מקטן‬
‫ ועד גדול‬is postponed to the end of the clause. 9
(68) Gen 19:11
‫ואת האנשים אשר פתח הבית הכו בסנורים מקטן ועד גדול‬
And the men who were at the entrance of the house they struck with
blindness, from small to great
5.2. Nonpreposed preverbal constituents

Some constituents stand before the verb in the unmarked clause and are not
considered preposed in that position. These include clausal adverbs (§5.2.1),
negative particles (§5.2.2), and certain adjunct clauses (§5.2.3).
5.2.1. Clausal adverbs

Although the unmarked position of most adjuncts is postverbal, clausal ad-
verbs 10 such as ‫אולי‬, ‫לכן‬, ‫הנה‬, and others always occur in clause-initial posi-
9. See also Gen 19:4, where the posponed modifying prepositional phrase is followed
by an appositive phrase: ‫ואנשי העיר אנשי סדם נסבו על הבית מנער ועד זקן כל העם מקצה‬
‘And the men of the city, the men of Sodom, surrounded the house, both young and old,
all the people to the last man’.
10. Clausal adverbs are also known as sentence adverbs. Blau (1977) discusses sentence
adverbials that are separated from the rest of the clause by a conjunction (see §5.5.2) but does
not discuss the sentence adverb (i.e., a one-word adverbial) specifically. The term clausal
adverb appears in Waltke and O’Connor (1990: §39.3.2–5) but is defined more broadly to
include negative and restrictive particles, which have very different syntactic characteristics
than the ones discussed here. According to Gross (1996:138–40), a variety of adverbs and
particles that function on the clause or text level, such as ‫עתה‬, ‫הנה‬, and ‫לכן‬, stand at the
margin of the preverbal field, before preposed constituents (see §3.3.6). According to Gross,
‫ )ו(הנה‬is a “Satzdeiktikon,” ‫ )ו(עתה‬is a “Textdeiktikon,” and ‫ לכן‬is a conjunctional adverb
(1996: 129–31). Van der Merwe et al. (2002: §41.3) classify as modal adverbs some words
I consider clausal adverbs, such as ‫‘ אך‬surely’. Others, such as ‫הנה‬, are classified there as

69 Nonpreposed preverbal constituents 5.2.
tion. The clausal-adverb class in BH can be defined in a manner similar to the

definition of the parallel category in English. Clausal adverbs in English are
syntactically and prosodically detached from the sentence and most frequently
occur in initial position. 11 The syntactic detachedness of clausal adverbs in
English is demonstrated by the fact they cannot be the focus of a cleft sen-
tence and cannot be the focus of negation or interrogation (Quirk et al. 1985:
612, 631; Huddleston and Pullum 2002: 666–68). From the semantic perspec-
tive, clausal adverbs have a semantic scope that extends over the entire clause.
Quirk et al. introduce the terms disjunct and conjunct to describe two types of
clausal adverbs. 12 Although disjuncts and conjuncts may not be syntactically
distinct, the terms are useful in understanding the different types of pragmatic
functions that clausal adverbs may serve. Disjunct adverbs express the speak-
er’s attitude to the content of the clause or the manner in which it is expressed;
examples in English include certainly, perhaps, and fortunately. Conjunct ad-
verbs signal relations between clauses or larger discourse segments; examples
include therefore, similarly, and however. 13
In BH, the group of adverbs that correspond semantically to English clausal
adverbs constitutes a distinct group on syntactic grounds, occurring exclusively
in initial position 14 and being syntactically detached from the clause. Although
the detachedness of clausal adverbs in BH cannot be established by the syntac-
tic tests referred to above (i.e., the possibility of focus of a cleft construction
and focus of negation), evidence on this point can be inferred from the position
of clausal adverbs relative to other preverbal elements, as discussed further on
in this section. The following is a list of the clausal adverbs occurring in finite
“discourse markers” that “comment on the content of a sentence and/or sentences from a
meta-level” (§44). On discourse markers, see n. 13 below.
11. Quirk et al. (1985: 615). English clausal adverbs can also occur anywhere within the
clause, set off by commas. Huddleston and Pullum (2002: 1350–62) view clausal adverbs
as belonging to the supplement category, a category that also includes appositional phrases
and interpolations.
12. Disjuncts and conjuncts may be phrases as well as single-word adverbs. An example
of a disjunct phrase in BH is ‫‘ ולישמעאל שמעתיך‬As for Ishmael, I have heard you’ (Gen
17:20). The disjunct and conjunct classes are distinct from the adjunct class (Quirk et al.
1985).
13. Quirk et al. (1985: 627–28, 642). Conjunct adverbs belong to a larger pragmatic
category known as discourse markers or discourse connectives. Discourse connectives are
linguistic elements that mark relations between clauses and may be adverbs, coordinating or
subordinating conjunctions, interjections, and perhaps even clauses. A tutorial overview of
the literature on discourse markers, accompanied by an extensive bibliography, can be found
in Schourup (1999). For further discussion, see chap. 6.
14. Some BH adverbs with semantic scope over the clause have freer positioning within
the clause, for example, ‫‘ אפוא‬therefore’. It would seem that these belong to a different syn-
tactic class from the one described here.

clauses in the Genesis corpus. 15 Examples (69)–(73) are clausal adverbs of the
disjunct type.
(69) ‫אולי‬
‫אולי יחסרון חמשים הצדיקם חמשה‬
Perhaps the fifty righteous lack five. (Gen 18:28)
(70) ‫ הנה‬16
‫הנה נתתי לכם את כל עשב זרע זרע אשר על פני כל הארץ‬
Behold, I give you every seed-bearing plant that is upon the face of all
the earth (Gen 1:29)
(71) ‫ אך‬17
‫אך טרף טרף‬
Surely he has been torn to pieces (Gen 44:28)
Clausal adverbs of the conjunct type are shown in (72)–(83).
(72) ‫( הן‬expressing justification) 18

‫הן גרשת אתי היום מעל פני האדמה‬
Behold, you have driven me this day away from the soil (Gen 4:14)
(73) ‫( הלא‬expressing justification) 19

‫הלא הוא אמר לי אחתי הוא‬
Surely he said to me, “She is my sister.” (Gen 20:5)
15. Some of the adverbs mentioned have other uses as well, as mentioned in the notes
below.
16. The word ‫ הנה‬is a presentative with many shades of use, as discussed in, e.g., Labu-
schagne (1973); McCarthy (1980); Muraoka (1985: 137–40); Kogut (1986); Waltke and
O’Connor (1990: §40.2.1); Zatelli (1994). Often, ‫ הנה‬is a conjunct rather than a disjunct,
marking the justification for a following utterance (see §7.3.2.1 n. 21, p. 117). Zewi (1996)
states that ‫ הנה‬following the speech verb ‫ אמר‬often functions as a kind of conjunction intro-
ducing direct speech, on the analogy of Arabic inna. In such cases, in her view, ‫ הנה‬belongs
to the reporting clause rather than to the quoted utterance, like the supposed kî recitativum
(see §4.4.3 n. 20, p. 60).
17. The word ‫ אך‬is also a focusing adverb. On preposing with the focusing adverb ‫אך‬,
see §5.4.1.
18. The word ‫ הן‬has some of the same functions as ‫ ;הנה‬see Labuschagne (1973); Mu-
raoka (1985); Zewi (1997); Garr (2004).
19. The clausal adverb ‫ הלא‬is viewed here as distinct from the combination of the inter-
rogative particle ‫ ה‬and the negative particle ‫לא‬. Steiner (1979: 7) writes that ‫ הלא‬can be used
idiomatically to introduce an assertion and in this case is not synchronically analyzable as ‫ה‬
‫ לא‬+. Historically, the clausal adverb may have a separate origin from the interrogative; see
Brown (1987); Sivan and Schniedewind (1993); Ben-Ḥayyim (2000: 320); alternatively, the

(74) ‫אולם‬
‫ואולם אחיו הקטן יגדל ממנו‬
Nevertheless, his younger brother shall be greater than he (Gen 48:19)
(75) ‫( עתה‬expressing conclusion or consequence) 20

‫לגור בארץ באנו כי אין מרעה לצאן אשר לעבדיך כי כבד הרעב בארץ כנען‬
:‫ועתה ישבו נא עבדיך בארץ גשן‬
To sojourn in the land we have come; for there is no pasture for your
servants’ flocks, for the famine is severe in the land of Canaan; and now,
pray, let your servants dwell in the land of Goshen. (Gen 47:4)
(76) ‫( על כן‬expressing consequence)

‫ויאמרו איש אל אחיו אבל אשמים אנחנו על אחינו אשר ראינו צרת נפשו‬
:‫בהתחננו אלינו ולא שמענו על כן באה אלינו הצרה הזאת‬
Then they said to one another, “Truly we are guilty concerning our
brother, in that we saw his distress, when he pleaded with us and we
would not listen; therefore this distress has come upon us.” (Gen 42:21)
clausal adverb may have developed from the use of ‫ לא‬+ ‫ ה‬in negative rhetorical questions.
Syntactic evidence for the clausal adverb includes the occurrence of ‫ הלא‬in front of a pre-
posed element, as in ‫‘ הלוא נכריות נחשבנו לו‬Surely as foreigners he regards us’ (Gen 31:15;
appears as example (89) below). Nonidiomatic interrogative ‫ הלא‬is composed of two sepa-
rable particles (Richard Steiner, personal communication). As described in §§5.2.2 and 5.3.1
below, when a noun phrase or prepositional phrase precedes the verb, ‫ ה‬occurs in front of the
preposed phrase and ‫ לא‬immediately precedes the verb; see, e.g., Gen 18:25; 2 Sam 19:22;
Job 11:2. Thus, if ‫ הלוא‬in Gen 31:15 were a combination of the interrogative and the nega-
tive, one would expect the order to be ‫הנכריות לא נחשבנו לו‬. Additional syntactic evidence
for the clausal adverb is the occurrence of ‫ הלא‬preceding a conditional clause (e.g., Gen 4:7,
1 Sam 15:17) and preceding a left-dislocated element (e.g., Num 23:12, Judg 11:24). Clausal
adverbs occur in these structural positions (see below, p. 73 in this section, and §5.2.3), while
interrogative ‫ה‬, as well as the negative particle, does not. Thus, the position of clausal adverb
‫ הלא‬in ‫‘ הלוא אם תיטיב שאת‬Surely, if you do right, there is uplift’ (Gen 4:7) contrasts with the
position of interrogative ‫ ה‬in ‫‘ אם יגע טמא נפש בכל אלה היטמא‬If someone defiled by a corpse
touches any of these, will it be defiled?’ (Hag 2:13; see also Job 14:4). For further discus-
sion, see Moshavi (2007c, forthcoming). Brongers (1981) contains a useful discussion of the
asseverative use of ‫הלא‬, although he does not distinguish it formally from the interrogative.
The word ‫ הלא‬can serve as a disjunct or a conjunct and strongly resembles ‫ הנה‬in all of its
uses. In its occurrences in Genesis, it is typically a conjunct, marking the justification for an
adjacent utterance. Although the clausal adverb ‫ הלא‬is usually understood as an asseverative
(‘surely’), I argue in a forthcoming article (Moshavi forthcoming) that it is best seen as a
presentative, like ‫הנה‬.
20. The clausal adverb ‫עתה‬, as opposed to the temporal adverb ‫‘ עתה‬now’, is preceded
by ‫ ;ו‬see Waltke and O’Connor (1990: §39.3.4f).

(77) ‫( לכן‬expressing consequence)

:‫לכן ישכב עמך הלילה תחת דודאי בנך‬
Then he may lie with you tonight in return for your son’s mandrakes.
(Gen 30:15)
(78) ‫( כי‬after negative clause) 21

:‫ויאמר לא כי צחקת‬
And he said, “No, rather, you did laugh.” (Gen 18:15)
(79) ‫( כי אם‬after negative clause) 22

‫לא יירשך זה כי אם אשר יצא ממעיך הוא יירשך‬
This one shall not be your heir; rather, your own issue shall be your heir
(Gen 15:4)
(80) ‫( כי עתה‬introducing apodosis after hypothetical conditional) 23

:‫כי לולא התמהמהנו כי עתה שבנו זה פעמים‬
For if we had not delayed, we could have already returned twice. (Gen
43:10)
(81) ‫( אם לא‬after negative clause)

‫ אם לא אל בית אבי‬:‫לא תקח אשה לבני מבנות הכנעני אשר אנכי ישב בארצו‬
:‫תלך ואל משפחתי ולקחת אשה לבני‬
You shall not take a wife for my son from the daughters of the Canaan-
ites, in whose land I dwell; rather to my father’s house you shall go and
to my family, and take a wife for my son. (Gen 24:37–38)
(82) ‫ גם‬24
‫לא מצאתיה וגם אנשי המקום אמרו לא היתה בזה קדשה‬
I have not found her; and furthermore, the people of the place said,
“There has been no prostitute here.” (Gen 38:22)
21. See also, e.g., Gen 19:2, Josh 5:14, 2 Sam 20:21. Both ‫ כי‬and ‫( כי אם‬79) are discussed
further in the context of the focused clause in §8.3.3. On asseverative ‫כי‬, see §4.3.3 n. 10
(p. 54).
22. See also, e.g., Gen 35:10, Lev 21:14, Num 10:30, 1 Sam 8:19. A slightly different
usage is found in Gen 40:14, which is more similar in meaning to the clausal adverb ‫ ;רק‬see
example (83).
23. The conjunct is not represented in the translation due to the lack of an appropriate
English equivalent. See also, e.g., Gen 31:42, Num 22:29, 1 Sam 14:30.
24. The conjunct ‫ גם‬differs semantically from the more common focus adverb ‫‘ גם‬also’
in having scope over the whole clause. The function of the conjunct is to add one utterance to
another (‘furthermore’, ‘moreover’). See also, e.g., Exod 3:9, 6:2–5; 1 Sam 4:17. The word

(83) ‫ רק‬25
‫ואם לא תאבה האשה ללכת אחריך ונקית משבעתי זאת רק את בני לא תשב‬
‫שמה‬
And if the woman is not willing to follow you, then you will be free from
this oath of mine; but do not take my son back there (Gen 24:8)
When a clausal adverb occurs in a clause with a detached element, such as a

left-dislocated element or an element connected to the clause with a conjunc-
tion (see §5.5.2), the adverb precedes the detached element, as shown in (84)
and (85). 26 This pattern clearly indicates that the clausal adverb itself occupies
a detached position.
(84) 2 Kgs 1:6, 16

‫לכן המטה אשר עלית שם לא תרד ממנה‬
Therefore the bed upon which you have alighted, you shall not come
down from it
(85) 2 Kgs 17:36

‫ אשר העלה אתכם מארץ מצרים בכח גדול ובזרוע נטויה אתו תיראו‬′‫כי אם את י‬
Rather, the LORD who brought you out of the land of Egypt with great
power and with an outstretched arm, him you shall worship
Further evidence for the detachedness of clausal adverbs is their position rela-
tive to a preposed subject, complement, or ordinary adjunct. The clausal ad-
verb always precedes the preposed element. 27 The following examples include
nearly all the clausal adverbs listed above.
‫ גם‬as a conjunct may precede the disjunct ‫הנה‬, as in the nonverbal clauses in Gen 38:24 and
Exod 4:14. According to Labuschagne (1966), ‫ גם‬is at times a purely “emphatic” adverb,
lacking an additive meaning. Muraoka (1985: 146), however, feels that “the particle gam
almost always retains its additive force.” In his comprehensive study of the particle, van
der Merwe (1990: 198) comes to a similar conclusion, writing that “Gam almost always has
an additive, inclusive or at least a connective connotation.” On preposing with the focusing
adverb ‫גם‬, see §5.4.1.
25. See also Gen 19:8, 1 Kgs 11:13. The conjunct ‫ רק‬differs semantically from the fo-
cus adverb ‫‘ רק‬only’ in having scope over the whole clause. In addition, the conjunct has a
contrastive meaning (‘but, however’) rather than a restrictive one. On the conjunct use, see
Kogut (1996: 204). According to BDB and Muraoka (1985: 131), ‫ רק‬is an asseverative par-
ticle in several cases; e.g., Gen 20:11.
26. For examples of a clausal adverb preceding an element separated from its clause by
a conjunction, see examples (131) and (132). In the nonverbal clause, however, a dislocated
element sometimes precedes a clausal adverb, for example: ‫ויתר דברי שלמה וכל אשר עשׂה‬
‫‘ וחכמתו הלוא הם כתבים על ספר דברי שלמה‬And the other events of Solomon’s reign, and all
his actions . . . surely they are recorded in the book of the Annals of Solomon’ (1 Kgs 11:41).
27. This result agrees with Gross’s (1996: 138–40) finding that certain adverbs stand at
the margin of the preverbal field, before preposed constituents (see n. 10, p. 68 above).

(86) Exod 32:34 28

‫הנה מלאכי ילך לפניך‬
Behold my angel will go before you
(87) Exod 9:16 29

‫ואולם בעבור זאת העמדתיך‬
But for this purpose I have spared you
(88) Gen 3:22 30

‫הן האדם היה כאחד ממנו‬
Behold, the man has become like one of us
(89) Gen 31:15

‫הלוא נכריות נחשבנו לו כי מכרנו‬
Surely as foreigners he regards us, for he has sold us
(90) 1 Sam 28:18

:‫ היום הזה‬′‫על כן הדבר הזה עשה לך י‬
Therefore this thing the LORD has done to you this day.
(91) 1 Sam 28:2

:‫לכן שמר לראשי אשימך כל הימים‬
Therefore my bodyguard I will make you for life.
(92) Gen 19:2

:‫לא כי ברחוב נלין‬
No, rather, in the street we will spend the night.
(93) Gen 21:26

:‫וגם אתה לא הגדת לי וגם אנכי לא שמעתי בלתי היום‬
And furthermore, you did not tell me. And furthermore, I have not
heard of it until today.
28. The word ‫ הנה‬followed by a preposed element is a particularly common construc-

tion, occurring also in Exod 32:34; Num 22:5, 32; Josh 2:2, 24:27; Judg 7:13, 14:16, 19:22;
1 Sam 8:5, 24:10, 28:9; 2 Sam 1:6, 13:35, 19:38; 2 Sam 24:17; 1 Kgs 8:27; 2 Kgs 10:9,
19:11.
29. See also Gen 48:19 (example [74], p. 71 above).
30. See also, e.g., Gen 27:37, 39:8, 44:8; Exod 6:12; Lev 10:19.

(94) Gen 24:8

:‫רק את בני לא תשב שמה‬
But my son do not take back there.
(95) Gen 35:10

‫כי אם ישראל יהיה שמך‬
Rather, Israel shall be your name
(96) Gen 24:38

‫אם לא אל בית אבי תלך ואל משפחתי‬
Rather, to my father’s house you shall go and to my family
(97) Gen 48:19

‫ואולם אחיו הקטן יגדל ממנו‬
Nevertheless his younger brother shall be greater than he
The fact that the clausal adverb always precedes the other preverbal item sup-
ports the conclusion that the clausal adverb is not preposed. If both items were
preposed, we would expect to find variation in the relative order of the two
constituents. Furthermore, although double preposing does occur in classical
BH prose, it is much rarer than single preposing (see §5.4.2). 31 Additional con-
firmation of the detached analysis is a case in Genesis of a clausal adverb fol-
lowed by double preposing:
(98) Gen 4:15

‫לכן כל הרג קין שבעתים יקם‬
Therefore any one who kills Cain, sevenfold vengeance shall be taken
[on him]
Because triple preposing is practically unheard of in classical prose, this con-

struction provides additional support for the position that the clausal adverb is
detached rather than preposed.
In conclusion, clausal adverbs stand in a position more peripheral than ei-
ther preposed or even left-dislocated elements. A clause that begins with a
clausal adverb and continues with the verb is not considered to be preposed.
Because the preverbal position of the clausal adverb is syntactically obligatory,
clausal adverb-verb is the unmarked word order.
31. My count in the Genesis corpus identified 21 occurrences (see table 1, p. 85 below).
Single preposing is far more common; as shown in table 1, there are 402 occurrences of
single preposing in the Genesis corpus.

5.2.2. Negative particles: ‫ לא‬and ‫טרם‬

The negative particle ‫ לא‬precedes the verb whenever clausal negation, rather
than subclausal negation, is involved. 32 The negative temporal adverb ‫‘ טרם‬not
yet’ also precedes the verb in its unmarked position. 33 These negative particles
are tightly bound to the verb. As a result, when clauses with ‫ לא‬or ‫ טרם‬undergo
preposing, the preposed element precedes both negative particle and verb:
(99) Gen 16:1 34

‫ושרי אשת אברם לא ילדה לו‬
And Sarai, Abram’s wife, had not borne him children
(100) Gen 24:45 35

‫אני טרם אכלה לדבר אל לבי‬
I had not yet finished speaking in my heart
In rare instances the preposed item intervenes between the negative particle
and the verb: 36
(101) Gen 45:8

‫ועתה לא אתם שלחתם אתי הנה‬
And now, it was not you who sent me here, but God
The construction exemplified in (101) has a special function that is explored

in chap. 8.
5.2.3. Preverbal adjunct clauses

Despite the unmarked postverbal position of adjuncts in general, most types
of adjunct clauses regularly occur at the head of the main clause and are there-
fore unmarked in the preverbal position. 37 Preverbal adjunct clauses occupy a
32. In clausal negation, the whole clause is treated as negative, whereas in subclausal
negation, a word or phrase is negated (Huddleston and Pullum 2002: 789).
33. In some cases (none of which occur in Genesis), ‫ טרם‬is not an adverb but a preposi-
tion governing a finite clause, as in ‫‘ וישא העם את בצקו טרם יחמץ‬And the people took their
dough before it was leavened’ (Exod 12:34). See Waltke and O’Connor (1990: §31.6.3c);
HALOT s.v. ‫טרם‬.
34. See also, e.g., Gen 20:4 and 21:26.
35. See also, e.g., Josh 2:8 and 1 Sam 3:7.
36. Goldenberg (1971, 1977, 1998a) views this construction as an “imperfectly-transformed
cleft sentence.” The completely transformed cleft would have a relative particle, yielding
something like ‫לא אתם אשר שלחתם אותי הנה‬. Goldenberg cites parallels to this construction
in Mishnaic Hebrew and various dialects of Aramaic. Given the marginality in BH of cleft
constructions and negative clefts (if the latter exist at all), it is doubtful whether the ‫ לא‬+
preposing structure is related to clefting.
37. An exception is the causal ‫ כי‬clause, which is normally at the end of the clause. A
preverbal causal ‫ כי‬clause would be marked; no clauses of this sort, however, are to be found
among the finite clauses in Genesis.

position in the clause that is intermediate between the preposed position and
the clausal-adverb position. Adjunct clauses are more detached from the clause
than preposed constituents. This can be seen from the fact that the adjunct
clause is regularly connected to the rest of the clause by a conjunction, some-
thing that does not ordinarily happen with preposed constituents: 38
(102) Gen 34:17

‫ואם לא תשמעו אלינו להמול ולקחנו את בתנו‬
And if you will not listen to us to be circumcised, we will take our
daughter
(103) Gen 32:18–19
:‫כי יפגשך עשו אחי ושאלך לאמר למי אתה ואנה תלך ולמי אלה לפניך‬
‫ואמרת‬
When Esau my brother meets you, and asks you, “To whom do you
belong, and where are you going, and whose are these ahead of
you?” you shall say
A preverbal adjunct clause may be followed by a preposed constituent. In

(104), for example, a conditional clause is followed by a preposed subject.
(104) Gen 42:19

‫אם כנים אתם אחיכם אחד יאסר בבית משמרכם‬
If you are honest men, one of your brothers shall be confined in your
place of detention
As compared to the clausal adverb, the adjunct clause is more closely con-
nected to the clause. When a clause contains both a clausal adverb and a pre-
verbal adjunct clause, the clausal adverb precedes the adjunct clause: 39
(105) Gen 24:49
‫ועתה אם ישכם עשים חסד ואמת את אדני הגידו לי‬
And now, if you will deal loyally and truly with my master, tell me
(106) Judg 21:21
‫והנה אם יצאו בנות שילו לחול במחלות ויצאתם מן הכרמים‬
And behold, if the daughters of Shiloh come out to dance in the
dances, come out of the vineyards
38. A conjunction sometimes intervenes between an initial element, usually a time ad-
junct, and the verb; see §5.5.2 for further discussion. I consider this to be a separate construc-
tion from preposing.
39. A similar observation is made in Blau (1977: 9) with regard to the relative order
of sentence adverbial and conditional clause. See also, e.g., Exod 19:5, 32:32, 33:13; Num
22:34; Judg 9:16; 1 Sam 20:29.

5.3. Unmarked preposed constituents 78
5.3. Unmarked preposed constituents

Certain grammatical forms, including interrogative pro-forms and a number
of time adverbs, are either constrained to or generally appear before the verb. 40
The preverbal position is the unmarked position for these forms. In §§5.3.1 and
5.3.2 it is shown that, unlike clausal adverbs, these forms occupy a preposed
position in the clause.
5.3.1. Interrogative pro-forms
Interrogative pro-forms occurring in finite nonsubordinate clauses in Gen-
esis are listed in (107).
(107) Interrogative pro-forms
(‫זאת‬/‫מה )זה‬ ‘what’ 41
‫מי‬ ‘who’
‫מתי‬ ‘when’
‫אנה‬ ‘to where’
‫אי מזה‬ ‘from where’
‫מאין‬ ‘from where’
(‫למה )זה‬ ‘why’
‫מדוע‬ ‘why’
‫איך‬ ‘how’
That interrogative forms stand in the same position as preposed constituents
can be seen from the fact that preposing does not ordinarily occur in clauses
with an interrogative. 42 When interrogative forms occur in left-dislocated con-
structions, they occur after the dislocated element. Example (108) is an example
of a dislocated element followed by an interrogative pronoun.
(108) Gen 31:43
:‫ולבנתי מה אעשה לאלה היום או לבניהן אשר ילדו‬
And to my daughters, what can I do to these today, or to their children
whom they have borne?
40. Gross (1996: 105–6) lists various forms that are obligatorily or preferably preposed;
these include interrogative particles and a variety of deictic particles including demonstrative
pronouns and spatial and temporal adverbs (see §3.3.6). His list includes a number of words
that are not preposed in the majority of cases, such as ‫‘ מחר‬tomorrow’ and ‫‘ היום‬today’. In the
present work, the category “unmarked preposed forms” is restricted to forms that are always
or nearly always preposed.
41. The particle ‫ מה‬is also used as an exclamatory particle, as in ‫‘ מה נורא המקום הזה‬How
awesome is this place!’ (Gen 28:17); see also, e.g., Gen 38:29. Exclamatory particles such as
‫מה‬, like their homonymous interrogative counterparts, are always preposed. Interrogative and
exclamatory ‫ מה‬must be distinguished from the indefinite pronoun ‫מה‬, as in ‫וראיתי מה והגדתי‬
‫‘ לך‬And if I learn anything, I will tell you’ (1 Sam 19:3). Unlike the interrogative pronoun,
the indefinite pronoun need not be preposed. See also, e.g., 2 Sam 18:22, 18:29.
42. One exception is Gen 27:37, in which the interrogative is preceded by a prepositional
phrase and an adverb: ‫‘ ולכה אפוא מה אעשה בני‬And for you, then, what can I do, my son?’
On ‫אפוא‬, see n. 14 (p. 69).

79 Unmarked preposed constituents 5.3.
Although it stands at the head of the clause, interrogative ‫ ה‬is not preposed. 43
Unlike the interrogative pro-forms, interrogative ‫ ה‬can occur with a preposed
element. In these cases, interrogative ‫ ה‬precedes the preposed element: 44
(109) Gen 18:25

:‫השפט כל הארץ לא יעשה משפט‬
Shall the Judge of all the earth not deal justly?
(110) Gen 34:31

:‫הכזונה יעשה את אחותנו‬
Should he treat our sister as a harlot?
5.3.2. Clause-initial time adverbs and the demonstrative adverb ‫כה‬

Although time adjuncts are normally postverbal, 45 certain time adverbs, as
well as the demonstrative adverb ‫כה‬, normally or exclusively occur before the
verb. 46 The time adverbs occurring in Genesis are listed in (111).
(111) Clause-initial time and ‫כה‬
‫‘ אחר‬afterward’ 47
‫‘ אחרי כן‬afterward’ 48
‫‘ אז‬then’ 49
43. Interrogative ‫ ה‬rarely appears with a coordinating conjunction in the classical prose
corpus. An apparent exception, ‫‘ והלוא עמך שם צדוק ואביתר הכהנים‬And surely Zadok and
Abiathar the priests will be with you there’ (2 Sam 15:35), is not actually the interrogative
‫הלוא‬. The word ‫ הלוא‬in this clause is the asseverative clausal adverb (see n. 19, pp. 70–71
above). This interpretation is followed by the NJPSV and NRSV, which take the sentence as
declarative and omit the adverb. (The RSV, however, which generally does not recognize the
clausal adverb ‫הלא‬, translates the verse with a rhetorical interrogative, ‘is it not?’).
44. See also, e.g., Gen 3:11, 17:17; Num 11:22; Judg 14:15; 2 Sam 19:22. In Deut 4:32
and 34, ‫ ה‬appears after ‫או‬.
45. The normal postverbal position for time adjuncts in general is confirmed by a survey
of a group of representative time adjuncts in the classical BH prose corpus. All of the occur-
rences of the following words and phrases were checked: ‫‘ מחר‬tomorrow’; ‫‘ היום‬today’; ‫ביום‬
‘on the . . . day (of)’; ‫‘ בחדש‬in the . . . month (of)’; ‫בשנת‬/‫‘ בשנה‬in the . . . year (of)’; all tem-
poral infinitive phrases involving the prepositions ‫ב‬, ‫כ‬, or ‫אחרי‬. The incidence of preposing
of these expressions is 27% overall. This result indicates clearly that the normal position for
time adjuncts is postverbal, as it is for adjuncts in general. The time adverbs listed in (111)
are exceptional in normally preceding the verb.
46. A similar observation is made by Gross (1996: 105–6).
47. Temporal ‫ אחר‬is always clause-initial in the classical BH prose corpus
48. The phrase ‫ אחרי כן‬is almost always clause-initial, except in ‫ ויהי‬clauses, which al-
ways begin with the verb (see §5.1), and in Josh 10:26, 1 Sam 24:8, and 2 Sam 21:14.
49. The word ‫ אז‬is always clause-initial in verbal clauses in the classical prose corpus. In
the nonverbal clause, it occurs between the subject and predicate, as in Gen 12:6 and 13:7.

5.4. Complex variants of the preposing construction 80
‫עתה‬ ‘now’ 50
‫כה‬ ‘thus’ 51
Just like interrogative pro-forms, the adverbs in (111) do not occur in clauses
with an additional preverbal constituent. The natural conclusion is that the ad-
verbs stand in the preposed position and are not detached like clausal adverbs. 52
5.4. Complex variants of the preposing construction
5.4.1. Preposing with a focusing adverb

A constituent that is the focus of a focusing adverb such as ‫‘ גם‬also’, ‫אף‬
‘also’, ‫‘ רק‬only’, or ‫‘ אך‬only’ may undergo preposing. 53 Several different struc-
tures are possible, corresponding to different unmarked constructions. 54 There
are at least two different means of forming a postverbal phrase with a focus
particle: 55
(112) Gen 19:35

‫ותשקין גם בלילה ההוא את אביהן יין‬
And they made their father drink wine that night also
(113) Gen 22:20

:‫הנה ילדה מלכה גם הוא בנים לנחור אחיך‬
Behold, Milcah, she also has born children to your brother Nahor:
In (112), ‫ גם‬precedes the focused expression, whereas in (113) ‫ גם‬focuses a

personal pronoun that is coreferential with the immediately preceding expres-
sion, ‫מלכה‬. A structure such as (112) can be converted to a preposed structure
by moving the ‫ גם‬phrase to a preverbal position:
(114) Gen 13:5

:‫וגם ללוט ההלך את אברם היה צאן ובקר ואהלים‬
And also Lot, who went with Abram, had flocks and herds and tents
50. The word ‫( עתה‬without ‫ו‬, as a time adjunct rather than a clausal adverb) is preverbal
in the vast majority of cases. The exceptions are Num 22:38; Judg 8:2, 8:6, 9:38, 11:7; 1 Sam
9:12, 17:29; 1 Kgs 12:4, 21:7.
51. The demonstrative adverb ‫‘ כה‬thus’ is almost always preverbal in the classical BH
prose corpus, with the exceptions of Exod 5:15 and Num 22:30.
52. In Deut 7:5, ‫ כה‬stands after an initial clausal adverb.
53. The focusing uses of ‫ גם‬and ‫ רק‬are distinct from their uses as conjuncts, as noted
above in §5.2.1 (pp. 72–73).
54. For a comprehensive study of constructions with ‫גם‬, see van der Merwe (1990).
55. In a rare type that does not occur in Genesis, ‫ גם‬precedes a cataphoric pronoun that
refers to a following appositive lexical expression, for example, ‫ויעשו גם הם חרטמי מצרים‬
‫‘ בלהטיהם כן‬And they also, the Egyptian magicians, did the same with their spells’ (Exod
7:11).

81 Other marked word-order constructions 5.5.
The structure exemplified in (113) may be preposed in one of two ways. First,
both the lexical expression and the following ‫ גם‬phrase may be preposed to-
gether. An alternative is to prepose the lexical expression, with the ‫ גם‬phrase
remaining in its original position. The two options are illustrated by the close-
to-minimal pair in (115) and (116). 56
(115) Gen 4:26

‫ולשת גם הוא ילד בן‬
And to Seth, [to] him also a son was born
(116) Gen 10:21

‫ולשם ילד גם הוא‬
And to Shem were born [sons], [to] him also
5.4.2. Double preposing

An unusual variant of preposing is the double preposing construction. In
this construction, two clause constituents precede the verb: 57
(117) Gen 6:16
‫ופתח התבה בצדה תשים‬
And the entrance of the ark in its side you shall put
(118) Gen 48:5

:‫אפרים ומנשה כראובן ושמעון יהיו לי‬
Ephraim and Manasseh, like Reuben and Simeon, shall be mine.
5.5. Other marked word-order constructions

In this section, I discuss marked word-order constructions that resemble
preposing but are syntactically distinct from it. These include left-dislocation
and the clause in which an initial element is separated from the clause by a
conjunction.
5.5.1. Left-dislocation
Left-dislocation, traditionally known by Hebraists as casus pendens, bears a
close resemblance to preposing but has an entirely different syntactic structure.
In both constructions, a constituent stands before the verb. In left-dislocation,
56. Representative occurrences of the first pattern from the Genesis corpus include Gen
4:22 and 19:38. A representative example of the second pattern is Gen 4:4.
57. A discussion of the functions of this construction can be found in Disse (1998:
1:190–99). Gross’s (2001a) monograph is devoted to the syntactic and pragmatic analysis of
double preposing in BH poetry and prose.

5.5. Other marked word-order constructions 82
the initial element is resumed by a coreferential pronoun within the body of

the clause; 58 no resumption of this sort is found in the preposed clause. The
resumptive element in left-dislocation is usually a pronoun or pro-adverb but
may also be an anaphoric lexical expression such as ‫‘ הנפש ההיא‬that person’;
see example (124). In the examples that follow, the left-dislocated element and
the resumptive element are bold.
(119) Gen 26:15

‫וכל הבארת אשר חפרו עבדי אביו בימי אברהם אביו סתמום פלשתים‬
And all the wells which his father’s servants had dug in the days of
Abraham his father, the Philistines had stopped them up
(120) Gen 3:12

‫האשה אשר נתתה עמדי הוא נתנה לי מן העץ‬
The woman whom you gave to be with me, she gave me from the tree
In (119), the direct object is dislocated, followed later by an enclitic resumptive

pronoun. In (120), the subject is dislocated, followed by a resumptive personal
pronoun. 59
When a phrase headed by ‫ את‬or another preposition is dislocated, the prepo-
sition may be retained, as in (121), or dropped, as in (122).
(121) Gen 35:12

‫ואת הארץ אשר נתתי לאברהם וליצחק לך אתננה‬
The land which I gave to Abraham and Isaac, to you I will give it
(122) Gen 28:13

:‫הארץ אשר אתה שכב עליה לך אתננה ולזרעך‬
The land upon which you are lying, to you I will give it and to your
descendants.
The detachment of the left-dislocated element as compared to the pre-

posed element is evident from several facts. First, the clause following the
58. The term left-dislocation originates from work in transformational-generative gram-

mar (Ross 1967). As applied to BH, of course, left-dislocation is a misnomer, because the
NP standing at the head of the clause is to the right of the clause, not the left. Left-disloca-
tion is termed “isolation of the natural subject” by Reckendorf (1895–98) and Bravmann
(1953); in Waltke and O’Connor (1990: §4.7), it is called the nominative absolute. On left-
dislocation in BH, see further, e.g., Driver (1892: §§193–201); GKC (§143); Brockelmann
(1956: §123a–h); Lambert (1972: §1282); Bloch (1986); Gross (1986, 1987b, 1988b); Khan
(1988); Naudé (1990); Joüon-Muraoka (§156); Zewi (1999). For Gross (1987b: 187–90), the
pendens construction need not involve resumption.
59. Other representative examples from the Genesis corpus include Gen 15:4, 24:7, and
44:17.

83 Other marked word-order constructions 5.5.
left-dislocated element is syntactically complete without it, which is not the

case with the preposed clause. Second, as shown in (122), a preposition is
often omitted from the dislocated element, something that does not occur in
preposed elements. These two characteristics suggest that the left-dislocated
element does not have a syntactic function (e.g., subject or complement) with
respect to the clause — hence the term casus pendens, ‘suspended case’. The
preposed constituent, in contrast, has the same syntactic function that it would
have in its unmarked position.
A third mark of detachment is that the left-dislocated element is often con-
nected to the clause with a conjunction, like the preverbal adjunct clause (see
§5.2.3). The left-dislocated element may also be preceded by a second con-
junction. In (123), a single conjunction occurs between the dislocated element
and the clause; in (124), two conjunctions are present. 60
(123) Josh 15:16
:‫אשר יכה את קרית ספר ולכדה ונתתי לו את עכסה בתי לאשה‬
Whoever attacks Kiriath-sepher and takes it, I will give Achsah my
daughter to him as wife.
(124) Gen 17:14

‫וערל זכר אשר לא ימול את בשר ערלתו ונכרתה הנפש ההוא מעמיה‬
Any uncircumcised male who is not circumcised in the flesh of his
foreskin, that person shall be cut off from his people
Although it is detached, the left-dislocated constituent is not entirely exter-

nal to the clause. As already mentioned in §5.2.1, a clausal adverb precedes a
left-dislocated constituent. In (125), for example, the dislocated element stands
after the conjunct ‫לכן‬. 61
(125) 2 Kgs 1:6, 16

‫לכן המטה אשר עלית שם לא תרד ממנה‬
Therefore the bed upon which you have alighted, you shall not come
down from it
The resumptive element in the dislocation construction may appear in the

normal postverbal position, as in (119), or may be preposed, as in (120). Re-
sumptive subject pronouns are generally preposed. 62
60. See also, e.g., Lev 26:36; 1 Kgs 12:17, 15:13.

61. This appears as (84), p. 73 above. See also 2 Kgs 17:36, example (85), p. 73 above.
62. See example (120), p. 82; additional representative examples include Gen 14:24,
15:4, 24:7, 44:17; Exod 12:16; Deut 1:30, 38, 39.

5.5. Other marked word-order constructions 84
5.5.2. The preverbal constituent connected to its clause with a conjunction

In this unusual construction, a normally postverbal constituent is placed in
preverbal position and connected to the rest of the clause with a conjunction: 63
(126) Gen 22:4

‫ביום השלישי וישא אברהם את עיניו‬
On the third day, Abraham lifted up his eyes
(127) Gen 27:34

‫כשמע עשו את דברי אביו ויצעק צעקה גדלה ומרה עד מאד‬
When Esau heard his father’s words, he cried out with an exceedingly
great and bitter cry
The intervening conjunction makes this construction resemble left-dislo-

cation more than preposing. As with the left-dislocated element, a preverbal
constituent followed by a conjunction may also be preceded by one:
(128) Exod 32:34

:‫וביום פקדי ופקדתי עליהם חטאתם‬
And on the day when I make an accounting, I will bring them to ac-
count for their sins.
Many cases of preverbal constituents followed by conjunctions involve

temporal adjuncts, as in the examples above. In some cases, however, the pre-
verbal constituent is the subject: 64
(129) Gen 44:9
‫אשר ימצא אתו מעבדיך ומת‬
Whichever of your servants it is found with, [he] shall die
(130) 1 Sam 17:20

:‫והחיל היצא אל המערכה והרעו במלחמה‬
And the army that was going forth to the battle line, [they] shouted
the war cry.
63. Adjunct clauses that are regularly preverbal, as discussed in §5.2.3, are not included
in this category. As noted there, an intervening conjunction is quite common in clauses of
this type. Blau (1959, 1977) discusses the sentence adverbial separated from its clause by a
conjunction in BH and other Semitic languages. In his view, the adverbial typically serves
as the psychological subject of the clause, that is, the part that is known from the context.
Steiner (1979: 9) rejects this pragmatic interpretation in his review of Blau (1977), stating
“it seems to me that sentence-initial temporal adverbials are used in BH to introduce a new
temporal frame of reference—not to refer to an old one.” Driver’s (1892: §§123–29) discus-
sion of casus pendens includes many examples of clauses with a preverbal constituent con-
nected to its clause with a conjunction. The corresponding Arabic construction is discussed
in Bravmann (1953: 17–21), Blau (1977), and Kinberg (1985).
64. Additional examples are cited in Steiner (1979: 9).

85 A statistical analysis of word-order constructions in Genesis 5.6.
Table 1. Word-order constructions in finite

nonsubordinate clauses in Genesis.
Number Percentage
65
1. Nonpreposed 2,901 84.0
2. Marked preposing 402 11.6
66
3. Unmarked preposing 71 2.1
4. Double preposing 21 0.6
5. Preposing with focus adverb 26 0.8
67
6. Other marked constructions 33 1.0
Total 3,454
Like the left-dislocated constituent, the initial constituent connected to its

clause by a conjunction comes after a clausal adverb:
(131) 1 Sam 25:27

‫ועתה הברכה הזאת אשר הביא שפחתך לאדני ונתנה לנערים המתהלכים‬
:‫ברגלי אדני‬
And now this present that your maidservant has brought to my lord,
let [it] be given to the young men who follow my lord.
(132) Num 22:11 68

‫הנה העם היצא ממצרים ויכס את עין הארץ‬
Behold, the people that has come out of Egypt, [it] covers the face of
the earth
5.6. A statistical analysis of word-order constructions in Genesis

The Genesis corpus contains 3,454 nonsubordinate finite clauses, ac-
cording to the definitions established in chap. 4. Table 1 69 categorizes these
clauses according to their word-order construction, based on the definitions
set forth in this chapter. Table 1 confirms that the nonpreposed construction is
65. This category includes clauses that are not preposed and do not have a marked con-
struction such as left-dislocation. A clausal adverb, negative adverb or particle, or adjunct
clause may precede the verb.
66. This category includes clauses with obligatorily preverbal elements.
67. This category includes left-dislocations and clauses with a constituent (other than an
adjunct clause) connected to the clause by a conjunction.
68. Although this segment is often rendered as two conjoined clauses (e.g., NRSV: “A
people has come out of Egypt and has spread over the face of the earth”), this interpretation
requires ignoring the definite articles in ‫ העם‬and ‫היצא‬.
69. In this and other tables, the percentages do not add up to precisely 100% because of
rounding.

5.6. A statistical analysis of word-order constructions in Genesis 86
the unmarked construction, constituting 84% of all the finite nonsubordinate

clauses in Genesis. Preposing is by far the most common marked word-order
construction, with a frequency of about 12%. Unmarked preposing constitutes
2% of the clauses, and double preposing and preposing with a focus adverb
occur in less than 1% of the clauses each. The other marked constructions
together account for 1%.
In the chapters that follow, I explore the pragmatic function of marked
preposing, as represented in category 2 in table 1. My purpose is to identify
the speaker’s/writer’s motivation in choosing the more unusual preposed con-
struction over the corresponding unmarked construction. Unmarked prepos-
ing, as represented in category 3, has no special pragmatic function, because a
nonpreposed option is not available for these clauses. Clauses exhibiting un-
marked preposing, therefore, are not addressed further. From this point on, the
term preposing will be used to denote exclusively the marked type of prepos-
ing. The constructions represented in categories 4 and 5 are certainly worthy
of pragmatic investigation, but there are too few occurrences of these types in
Genesis to allow for a reliable analysis. 70 Left-dislocations and clauses with a
constituent connected to the clause by a conjunction (both included in category
6) are also sparsely attested in Genesis. As such, they are not addressed further
in this work, despite their undeniable interest from the pragmatic perspective.
Table 2 shows the frequency of preposing as compared to unmarked word
order. In this table, the number of preposed clauses is compared with the num-
ber of clauses that are unmarked but are hypothetically preposable. In the first
row, the nonpreposed clauses (category 1 in table 1) are tallied, subtracting the
nonpreposable ‫והיה‬/‫ ויהי‬clauses, of which there are 72. The next row tallies the
marked preposed clauses from category (2) in table 1. Table 2 shows that pre-
posing occurs in 12.4% of the clauses that theoretically allow this construction.
As discussed in §3.1.6, several researchers have claimed that direct speech
exhibits a higher frequency of preposing than does narrative. Table 3, which
breaks down the results in table 2 according to discourse register, confirms
this claim. The frequency of preposing in narrative is only 7.6%, whereas the
frequency in direct speech is considerably higher, at 21.7%. The chi-square test
shows that the difference between the two frequencies is statistically signifi-
cant. 71 Despite the higher frequency of preposing in direct speech, however, it
70. Buth (1999: 88) contends that in double fronting the first element is a topic and the
second a focus (see §3.3.5 for further discussion). Double preposing in Modern Hebrew
is discussed in Ziv (1996a: 179), who states that the principle governing sentences of this
sort is that “the initial fronted constituent must be anchored in the discourse, if any fronted
entity is.”
71. The chi-square test is used to determine whether the difference between two popula-
tion proportions is statistically significant. The chi-square value is associated with a prob-
ability, p. According to standard statistical methodology, p must be less than .05 to be con-
sidered significant; in this case, p is less than or equal to .001.

87 Conclusion 5.7.
Table 2. Frequency of preposing in preposable finite

nonsubordinate clauses in Genesis.
Word order Number Percentage

Nonpreposed 2,829 87.6
Preposed 402 12.4
Total 3,231
Table 3. Relation between discourse register and word order

in preposable finite nonsubordinate clauses in Genesis.
Discourse register Nonpreposed Preposed Total

Narrative No. 1,964 162 2,126
% 92.4 7.6
Direct speech No. 865 240 1,105

% 78.3 21.7
Total No. 2,829 402 3,231

% 87.6 12.4
Chi-square = 131.682
p ¯ .001
is clear that preposing is still the marked order in this register, since the major-
ity of direct-speech clauses (78.3%) are not preposed.
Table 4 (p. 88) categorizes preposed clauses according to the syntactic
class of the preposed element. Almost all preposed elements are noun phrases
(64.9%) or prepositional phrases (32.8%). 72
Table 5 (p. 88) shows that preposed clauses are about evenly divided be-
tween subject preposing and complement/adverbial preposing.
5.7. Conclusion
The preposed constituent must be carefully distinguished from other ele-
ments that precede the verb. Clausal adverbs and most types of adjunct clauses
are unmarked and non-preposed in the preverbal position. There are two types
of preposing: the optional, marked type and the syntactically obligatory and
hence unmarked type (e.g., the preposing of interrogative pro-forms and cer-
tain time adverbs).
72. Noun phrases governed by ‫ את‬are considered prepositional phrases.

5.7. Conclusion 88
Table 4. Syntactic class of the preposed element in

marked preposed clauses.
Class Number Percentage
Noun phrase 261 64.9
Prepositional phrase 132 32.8
73
Adverb 3 0.7
74
Adjective 2 0.5
Headless relative clause75 3 0.7
76
Other 1 0.2
Total 402
Table 5. Syntactic function of the preposed element in

marked preposed clauses.
Class Number Percentage

Complement or adverbial 204 50.7
Subject 196 48.8
77
Ambiguous 2 0.5
Total 402
The various types of preverbal constituents discussed in this chapter differ

in terms of their degree of detachment from the clause. The most detached is
the clausal adverb, which stands as a rule before all other constituents except
for an initial conjunction. In the next place are the preverbal adjunct clause, the
73. These are 18:5 (‫)כן‬, 25:10 (‫)שמה‬, and 31:42 (‫)ריקם‬.
74. These are Gen 6:9 (‫ )תמים‬and 27:33 (‫)ברוך‬.
75. These are Gen 34:11, 41:28, and 41:55.
76. The preposed element in Gen 47:9 is a compound phrase consisting of a coordinated
noun and adjective, ‫‘ מעט ורעים‬few and bad’ (see HALOT s.v. ‫)מעט‬.
77. The ambiguous cases are Gen 36:13, 14. Both clauses contain the copular verb ‫היה‬,
a demonstrative pronoun and a definite noun phrase. Because both the demonstrative and
the noun phrase are definite, it is unclear which is the subject and which the predicate. Ac-
cording to Dyk and Talstra’s (1999) rules for determining subject and predicate in nominal
clauses, however, there is a definitive answer in these cases: in a nominal clause containing a
demonstrative pronoun and a definite noun phrase, the demonstrative pronoun is the subject;
see also Lowery (1999).

89 Conclusion 5.7.
left-dislocated constituent, and the element separated from the clause by a con-
junction. There are insufficient data to distinguish between the degrees of de-
tachedness of these three types. Next is the preposed element, whether marked
or unmarked. Most integrated is the negative particle, which comes after a
preposed constituent and is nearly always immediately adjacent to the verb.
Of all of the marked word-order constructions, preposing is by far the most
common. In the following chapters, I examine the pragmatic functions of
preposing.

Chapter 6
Focusing and Topicalization
This chapter explores the pragmatic concepts of focusing and topicalization

and applies these concepts to BH preposing. The approach here draws on previ-
ous research on information structure as well as on psycholinguistic research.
I will show that, although both focusing and topicalization provide an instruc-
tion regarding the interpretation of the clause in its context, the two types differ
regarding the context involved. Focusing signals a relation between the clause
and the context of the addressee’s attention state, whereas topicalization sig-
nals a relation between the clause and the linguistic context that accompanies
it. As such, topicalization functions in a manner similar to discourse connec-
tives that signal a pragmatic relation between two sentences or text segments.
Focusing is discussed in §6.1 and topicalization in §6.2.
6.1. Focusing
As noted in §3.3.1.3, there appear to be four distinct conceptions of focus:
informational focus, contrastive focus, psychological focus, and attentional fo-
cus. Psychological focus is very similar to the concept of topic, as mentioned
in §3.3.1.3. The shortcomings of the contrastive and attentional focus concepts
have already been discussed: contrastive focus blurs the distinction between
topic and focus, and attentional focus is essentially a modernized version of
emphasis. 1 Informational focus, therefore, appears to be the most useful con-
ception of focus for explaining preposing. Informational focus is the part of the
proposition expressed by the sentence that is assumed by the speaker/writer to
be new, rather than given. The ensuing sections examine the nature of informa-
tional focusing in more detail.
6.1.1. Constituent, predicate, and sentence focus

As mentioned in §3.3.7, Lambrecht has advanced the idea that informational
focus may be an argument, a predicate, or an entire sentence. For purposes
of clarity, I substitute the term constituent focus for argument focus, because
Lambrecht’s “argument focus” may be any clause constituent (other than the
predicate) and is not necessarily an argument of the verb. In this section, it is
1. The concept of attentional focus may be useful in understanding a “stylistic” use of

focusing, discussed in §8.5.
90

91 Focusing 6.1.
argued that the concepts of predicate focus and sentence focus relate to given
information of an entirely different type from constituent focus, and that the
concept of informational focus, strictly speaking, applies only to the constitu-
ent type.
As discussed in §3.3.1.3, informational focus identifies an unknown ele-
ment in the given proposition, answering an implicit wh-question. An example
of constituent focus is the sentence Bill ate the doughnuts, with the primary
accent on Bill. This sentence expresses a semantic proposition, represented
informally as “Bill ate the doughnuts.” The accenting of Bill indicates that
the proposition “x ate the doughnuts” is given information in the context of
the utterance 2 and reflects an assumption that Who ate the doughnuts? is a
contextually appropriate question. The given proposition is directly related to
the semantic content of the sentence and is derived by replacing the focus Bill
with a variable.
Given this understanding of informational focus, one would expect predi-
cate and sentence focus to reflect given propositions in which the variable x
represents the predicate and the entire sentence, respectively. Thus, Lambrecht
(1994: 223) writes that a predicate-focus sentence such as My car broke down
could be used appropriately following the question What happened to your
car?, reflecting the given proposition “My car x ”; the sentence-focus sentence
such as My car broke down is said to be appropriately used following a ques-
tion such as What happened? Sentence-focus, then, apparently reflects the
given proposition “x happened.”
As Lambrecht’s further discussion makes clear, however, this description
of predicate and sentence focus is imprecise. Predicate-focus sentences can
also be uttered in the absence of a given proposition in which x represents the
predicate. According to Lambrecht, the only proposition that can be assumed
to be given with respect to a predicate-focus sentence such as My car broke
down is a proposition to the effect that “the speaker’s car is pragmatically
available as a topic for discussion”; in other words, the given information is the
proposition “the speaker’s car is a topic for this utterance” (Lambrecht 1994:
226). This proposition is not directly related to the content of the sentence, as
in constituent-focus sentences; furthermore, the focus broke down does not
2. Propositions of the form “x ate the doughnuts” are open propositions, that is, proposi-
tions containing a variable (Prince 1985, 1986). Dryer (1996: 512) points out that, techni-
cally speaking, open propositions cannot be given, in the sense of believed or known, “since
they are not the type of thing that can be true or false” (see §6.1.2 on various conceptions
of givenness). It would be more correct to say that the given proposition is the “existential
closure of an open proposition” (Dryer 1996: 512); i.e., “there exists an x such that x ate the
doughnuts.” In the present study, given propositions are represented as open propositions,
with the understanding that the open proposition is shorthand for the existential closure of
that proposition.

6.1. Focusing 92
identify a variable in this proposition. The given proposition is best character-

ized as a metalinguistic proposition regarding the topic of the sentence. 3
The essential difference between predicate- and sentence-focus sentences,
in Lambrecht’s view, is that the latter do not have a topic. A topic is lacking
in the sentence-focus sentence because it informs the addressee of an event
rather than relating information about the subject (Lambrecht 1994: 124). In
other words, my car is the topic of predicate-focus My car broke down but
not of sentence-focus My car broke down. Sentence-focus sentences need not
involve the given proposition “x happened” and can be uttered out of the blue
(Lambrecht 1994: 124). Neither predicate-focus sentences nor sentence-focus
sentences, then, need involve a given proposition with an unknown.
It can be concluded that the only type of focus that fits the strict definition
of informational focus is constituent focus. Only constituent focus identifies an
unknown in a given proposition and answers a wh-question. Both predicate-
focus and sentence-focus sentences do not involve a piece of given informa-
tion closely related to the content of the sentence. Furthermore, the difference
between predicate and sentence focus has nothing to do with focus but con-
cerns the presence or absence of a topic. Using the term informational focus to
include predicate and sentence focus, therefore, obscures the very significant
differences between constituent focus and the other two types.
Constituent focus is the only kind of focus marked by preposing in the ca-
sual English register (see §3.3.1.3), and, I contend, is the only type of focus
marked by preposing in BH. Convincing evidence is lacking for the marking of
sentence focus by preposing in BH, as discussed earlier in §3.3.7. Predicate fo-
cus is the pragmatically unmarked articulation (see §3.3.7) and, as such, would
be expected to correspond to verb-first word order in BH.
In the following sections, I examine the concept of constituent focus (hence-
forth referred to simply as “focus”) in more detail. The following sections ex-
plore the concept of givenness as it relates to focus.
6.1.2. Types of givenness: Pragmatic presupposition and activation

Two types of givenness can be distinguished with regard to informational
focus: pragmatic presupposition and activation. 4 Pragmatically presupposed
information is information that is part of the common ground of speaker and
addressee, that is, information that the speaker assumes to be known or be-
3. In Lambrecht and Michaelis (1998), the two types of given information are clearly
distinguished. Given information involving an open proposition is termed a “knowledge
presupposition,” and given information relating to sentence topic is termed a “topicality
presupposition.”
4. Although presupposition and activation are often discussed in relation to discourse
entities represented by noun phrases, the concepts are equally applicable to the propositions
represented by sentences (Dryer 1996: 483).

93 Focusing 6.1.
lieved by him and his addressee. 5 Activated information is information that is

assumed by the speaker to be in the consciousness of the addressee at the time
of the utterance. 6 Pragmatic presupposition and activation correspond roughly
to the cognitive states of long-term and short-term (or working) memory. 7
Long-term memory stores all of a person’s knowledge and beliefs and has a
very large capacity. Short-term memory, on the other hand, has room for only
a very small amount of information. Long-term memory is relatively stable,
whereas short-term memory is in a state of flux, with items entering and leav-
ing on a regular basis.
The relation between presupposition and activation is a subject that has not
received adequate treatment in much of the linguistic literature. Activated enti-
ties are widely held to be a subset of presupposed entities. 8 Dryer (1996), how-
ever, shows that pragmatic presupposition and activation are logically distinct
categories; that is, an entity can be activated but not pragmatically presupposed.
Dryer (1996: 483) offers the proposition “the earth is flat” as an example of a
proposition that may be in a person’s consciousness (once it is mentioned by
someone else), despite that person’s lack of belief in the proposition.
When we talk about a construction marking a focus with respect to a given
proposition, it must be clarified whether we mean that this proposition is pre-
supposed or activated. In Prince’s (1986: 209–10) view, focusing constructions
5. See, e.g., Karttunen (1974) and Stalnaker (1974). Pragmatic presupposition is not
to be confused with semantic presupposition, which is an entailment of the sentence that is
constant under negation. Clark and Haviland’s (1977) “given” information is pragmatically
presupposed information. Prince’s (1981) term for pragmatic presupposition is assumed fa-
miliarity. Lambrecht (1994: 44) uses the term pragmatic presupposition somewhat differ-
ently to refer to mentally represented propositions that may or may not be believed (see n. 8).
6. For the term activation, see, for example, Gundel (1985); Chafe (1987); Gundel, Hed-
berg, and Zacharski (1993: 278); Lambrecht (1994); Kintsch (1998). The concept originates
in Chafe’s seminal article (1976: 30), in which given information is defined as “the knowl-
edge which the speaker assumes to be in the consciousness of the addressee at the time of the
utterance. In Prince (1981), the concept is termed “saliency.”
7. On short-term and long-term memory, see, e.g., Kintsch (1998: 215–46). Kintsch’s
theory of memory and its role in comprehension is described in further detail in §6.1.3.
8. Prince (1985: 70), for example, writes that English topicalization marks a proposi-
tion “as representing ‘shared knowledge’, more particularly, Chafe-given [i.e., activated]
knowledge” (text in brackets is mine). In another article (1986: 209), she states that focus-
presupposition constructions mark a proposition as “salient shared knowledge”; that is, pre-
supposed and activated. See Dryer (1996: 511–12) for further discussion of Prince’s views.
In the “givenness hierarchy” described by Gundel, Hedberg, and Zacharski (1993), activated
entities are a subset of presupposed entities; however, this hierarchy does not specifically ad-
dress propositions. A different taxonomy of givenness is Prince’s (1981) “familiarity” scale.
Despite its superficial resemblance to the givenness hierarchy referred to above, Prince’s
scale relates exclusively to states of pragmatic presupposition. Lambrecht (1994) also views
activated propositions as a subset of presupposed ones, because presupposed propositions for
him are those that are “mentally represented” (see n. 5 above); all activated propositions are
thus presupposed by definition.

6.1. Focusing 94
relate to a proposition that is both presupposed and activated. 9 Dryer (1996),

however, demonstrates that different focusing constructions relate to different
kinds of given information: accenting relates specifically to activated infor-
mation (whether presupposed or not), while clefting relates specifies to pre-
supposed information (whether activated or not). The pragmatic difference
between accenting and clefting is illustrated by (133) and (134):
(133) A: Did anyone see Mary?

B: John saw her.
(134) A: Did anyone see Mary?

B: #It was John who saw her.
As Dryer notes, A’s question activates the proposition “x saw Mary” but does
not presuppose it, because the point of the question is precisely whether this
proposition is true. In this context, John can be marked as a focus by accenting,
as in (133), but cannot be marked as focus by a cleft construction, as shown by
the inappropriateness of (134). 10
6.1.3. Activation in a model of text comprehension
What kind of information do speakers/writers assume to be activated for
their addressees, and how does this information become activated, that is,
placed in short-term memory? I first consider written-speech contexts. It is
proposed here that activated information for the reader of a text is information
either contained in or inferrable from the previous sentence. Support for this
claim can be gathered from current research on the process of text compre-
hension. A broad body of psycholinguistic research supports the theory that
readers understand texts by constructing a coherent mental representation of
the text (Singer 1990; Sanders and Spooren 2001). Short-term and long-term
memory play key roles in building the mental representation. In the ensuing
discussion, activation is examined in the light of a well-known comprehen-
sion model, Kintsch’s “construction-integration” model (1998). The model is
supported by a large number of laboratory experiments by Kintsch and others.
9. Prince (1986: 210) has stated that Yiddish-movement (see §3.3.1.2, p. 33 n. 47) dif-
fers from the aforementioned constructions in involving presupposed but not necessarily
activated information (see also Prince 1988: 512–15). In later work (1992, 1999), however,
she states that the given information in Yiddish-movement may be merely plausibly infer-
able and not necessarily presupposed.
10. Similarly, B may answer Nobody saw her, but not It was nobody who saw her.
The first sentence, in which the focus is marked by accent alone, is acceptable because it
marks the proposition “Somebody saw him” as activated. The cleft sentence is unacceptable
because it presupposes the proposition “Somebody saw him,” while its content contradicts
this presupposition. For additional proofs that focusing by accent does not involve presup-
position, see Dryer (1996).

95 Focusing 6.1.
The relevant aspect of Kintsch’s theory for present purposes is the concept
of working memory, which includes short-term memory as well as a section of
long-term memory. Long-term memory, storing knowledge, beliefs, and ex-
perience, is a network of propositions, a large and intricately interconnected
structure. Short-term working memory, in contrast, is an extremely small buf-
fer containing up to four or perhaps seven “chunks” of information—about
the amount of information contained in a single sentence (Kintsch 1998: 217,
411). Strictly speaking, the only information that is activated is the proposition
currently in short-term working memory. This is not the only memory avail-
able to the reader, however. The reader also has near-instant access to all of the
propositions in long-term memory that are directly linked to the proposition
in the short-term memory. These easily retrievable propositions constitute the
long-term working memory. Propositions in long-term working memory can be
easily activated and placed in short-term memory and are known as accessible
propositions. 11
The essence of the comprehension process is described by Kintsch (1998:
93) as follows: “We comprehend a text, understand something, by building a
mental model. To do so we must form connections between things that were
previously disparate: the ideas expressed in the text and relevant prior knowl-
edge.” A reader builds a mental representation of the text in the form of a
network of propositions derived from the text and stores it in long-term mem-
ory. As the reader proceeds through a text, a proposition is constructed corre-
sponding to each sentence and stored in short-term working memory. After the
sentence processing is completed, the proposition representing the sentence is
copied to long-term memory and linked to the textual representation already
stored there. Furthermore, additional propositions, drawn from the reader’s
knowledge and experience, are added to the representation and linked to the
sentence representation. These propositions include, for example, bridging in-
ferences regarding referring expressions in the text, inferences about causal
connections between sentences in the text, and elaborative inferences that fill
in details unspecified in the text (Kintsch 1998: 188–99). 12
11. This term is from Dryer (1996: 481), although Dryer does not connect accessibility
to long-term working memory. According to Dryer, accessible information is inferable from
activated information and is therefore more easily activated than unrelated information. Ac-
cessible information is similar but not identical to Chafe’s (1987; 1994: 29) “semi-active”
information. Accessibility also has a different meaning in Ariel’s Accessibility Theory (for a
recent summary and discussion, see Ariel 2001). For Ariel, accessibility is a scalar property
applicable to activated and nonactivated entities and refers to the ease of retrieval of an item
from memory.
12. Higher-level propositions are also inserted into the text representation to represent
the macrostructure of the text, that is, its main ideas and themes. On the various types of
inferences made during text comprehension, see Singer (1990: 167–89); Kintsch (1998:
193–98); Graesser et al. (2001). A controversy exists over which types of inferences are

6.1. Focusing 96
As the reader moves on to the next sentence, the central information con-
tained in the previous sentence is retained in short-term memory in order to aid
in comprehending the next sentence. All of the propositions in the text repre-
sentation directly linked to the information in short-term memory, including
inferred information, are in long-term working memory and can be quickly and
easily converted to activated information. In short, when reading a sentence,
the gist of the previous sentence is activated and information inferred from or
directly linked to the previous sentence is accessible.
In Kintsch’s model, activated and accessible propositions have much in
common, both being part of working memory. Moreover, it is hard to draw
a line between activated and accessible information, because information in
long-term working memory can quickly and easily be activated and transferred
to short-term memory. It is plausible, therefore, that both activated and acces-
sible propositions can serve as given information in connection with a focusing
construction. In fact, it can be argued that a focusing construction relating to
an accessible proposition reflects an assumption that this proposition will have
been activated by the reader. From this point, on I will use activated to refer
to information in short-term memory as well as accessible information that is
treated by the writer as activated.
The comprehension model as outlined above relates to text comprehension
and does not include many aspects of spoken-language comprehension. In spo-
ken discourse, information assumed to be in the consciousness of the addressee
is not restricted to propositions derived from the preceding discourse but also
includes propositions pertaining to the identities of speaker and addressee,
their emotions and motivations, as well as elements of the real-world environ-
ment in which the dialogue takes place. All of this information cannot possibly
fit into the small short-term memory buffer. Kintsch (1998: 411) suggests that
this additional information should be thought of as located in permanently acti-
vated nodes in long-term memory. Focusing constructions in spoken discourse,
then, may relate to activated parts of the addressee’s mental representation of
the preceding utterances, as well as additional information of the permanently
activated type.
6.1.4. Focusing in Biblical Hebrew

Preposing in BH can serve as a focusing construction, as will be demon-
strated by the statistical analysis in the next chapter. Left-dislocation may also
have a focusing function, as noted by Khan (1988: 93–94). 13 Because the prag-
matic analysis in the present work is restricted to the preposing construction,
typically generated during text comprehension and which only later during retrieval. Readers
appear to vary greatly in the degree of inference that they perform during the comprehension
process.
13. See, e.g., Gen 15:4, 44:17; Deut 1:36, 38, 39; 1 Kgs 22:14; 2 Kgs 17:36.

97 Topicalization 6.2.
the term focusing is used as shorthand for focusing by preposing; similarly, a

“focused clause” is one in which the focus has been marked by preposing.
Does BH focusing relate to activated or presupposed information? An ex-
amination of focusing in Genesis shows that activation is the relevant type of
givenness for this construction. In narrative, focused clauses relate in every
case to a proposition contained in or derived from the previous clause; in other
words, the given proposition is assumed to be in short-term or long-term work-
ing memory. In direct speech, focused clauses relate to information derived
from the previous dialogue or to information pertaining to the addressee’s per-
sonal details, motivations, and perceptions of the environment. The activated
proposition associated with a focused clause need not be presupposed, i.e.,
believed or known by the hearer. A full examination of focusing in BH is pre-
sented in chap. 8.
6.2. Topicalization
As described in §3.3.1.2, the term topicalization at first denoted a syntactic
concept, later shifting to a syntactic-prosodic one; yet another use of the term
is to denote a syntactic-pragmatic concept. The latter conception of topicaliza-
tion is the most useful for investigating BH, a language for which we have no
prosodic data. 14 It is important to keep in mind, of course, that correspond-
ing structures in different languages may not have exactly the same pragmatic
properties. Thus, the function of BH topicalization may be subtly different
from topicalization in other languages. In order to determine whether topical-
ization exists in BH, however, we need to understand exactly what the function
of topicalization is.
As discussed in §3.3.1.2, the linguistic literature cited in studies of BH is
somewhat unclear on this crucial point, with topicalization variously charac-
terized as marking discontinuity, continuity, or both. There are, however, other
studies of topicalization that have not been fully taken advantage of by schol-
ars of BH. These are addressed in §6.2.1. In §6.2.2, I present a new character-
ization of BH topicalization based on prior research on topicalization and on
discourse connectives.
6.2.1. Prince’s and Birner and Ward’s characterization of topicalization

Prince and, subsequently, Birner and Ward have presented a theory regard-
ing the function of topicalization that does not involve the concept of topic
and is equally applicable to verb-first and subject-first languages. 15 Although
their description of topicalization is open to challenge in some respects, it
14. On the cross-linguistic investigation of topicalization, see, e.g., Myhill (1985, 1992b)
and Givón (2001).
15. See Prince (e.g., 1985, 1986, 1988, 1998); Ward and Prince (1991); Birner and Ward
(1998).

6.2. Topicalization 98
provides significant insights relevant to BH. Its most important achievement

is to provide a framework that unites the continuous and discontinuous uses
of topicalization.
Prince and Birner and Ward see topicalization as having a double function.
In §§6.2.1.1 and 6.2.1.2, I present a discussion and critical analysis of the two
proposed functions.
6.2.1.1. Marking a partially ordered set relation pertaining to the
preposed element
One function of topicalization according to Prince is to mark a relation be-
tween the preposed item and an item in the prior context. The preposed item
may be an entity already evoked in the discourse, or else in a “salient [i.e.,
activated] set relation to something already evoked” (Prince 1985: 73; words
in brackets are mine). 16 The former type is what has been called continuous
topicalization, in which the preposed element is identical to an element men-
tioned earlier. The latter type, in which the preposed element is related to an
earlier element, frequently standing in contrast with it, is what has been called
discontinuous topicalization.
In later work by Ward and Prince (1991: 173), Prince’s description of topi-
calization is reformulated to incorporate the idea of the “partially ordered set
relation.” 17 The partially ordered set relation, or “poset,” includes a broad va-
riety of logical relations, e.g., “A is identical to / a member of / subset of /
type of / part of B,” or “A and B are members of / subsets of / types of / parts
of C.” Two items belonging to a set stand in a poset relation to each other:
{husband, wife}, {father, son}, {bread, wine}. “Functional dependency” rela-
tions, such as “A has a B” (e.g., a book has a cover), are not poset relations
and, it is claimed, are not characteristic of topicalization. 18 According to Ward
and Prince (1991), topicalization marks the preposed item as standing in a
salient partially ordered set relation to an entity or entities already evoked in
the discourse. 19
16. Levinsohn (1990: 22) cites this definition, attributing it erroneously to Andrews
(1985: 78). Levinsohn’s understanding of topicalization is much broader than Prince’s, in-
cluding the use of preposing to mark all kinds of discontinuities, whether relating to the
preposed constituent or to the thematic progression of the story.
17. The logical definition of this concept is highly technical and difficult to understand
intuitively; according to Ward and Prince (1991:173), partially ordered sets “are defined by
a partial ordering R on some set of referents, b, such that, for all b1, b2, and b3 that are ele-
ments of b, R is either reflexive, transitive, and antisymmetric… or, alternatively, irreflexive,
transitive, and asymmetric.” The linguistic application of the partially ordered set originates
in Hirschberg (1991).
18. See Ward and Prince (1991: 175–76); Prince (1998: 8).
19. See also Prince (1998).

In Birner and Ward’s (1998) modification of Prince’s theory, 20 the descrip-

tion of topicalization is similar, except that it specifies that the entity to which
the preposed item is related need not have been explicitly evoked in the prior
context; it may be merely inferable from the prior context. The partially or-
dered set relation must be “contextually licensed” (Birner and Ward 1998: 20);
that is, it must be directly derivable or inferrable from the context.
The partially ordered set relation involved in topicalization may be highly
idiosyncratic and restricted to the particular discourse context; in Birner and
Ward’s (1998: 234) words, it may be “entirely ad-hoc.” This point is illustrated
by (135):
(135) Philadelphia Inquirer, 8/31/1983 (Birner and Ward 1998: 234).

It is nearly 8 and Ellerbee, back from dinner, tidies up some details in her
narrow, windowless third floor office before moving down the hall to the
studio. Her son and his friend pore over computer workbooks in a small
area outside her office. Joshua and his 14-year-old sister, Vanessa, take
turns spending Friday nights at the studio. “Sleep they can catch up on,”
Ellerbee says, “Mom they can’t.”
Example (135) contains paired topicalized clauses, with sleep and Mom as the
respective preposed constituents. Mom is obviously meant to be related to the
preceding sleep, despite the lack of any natural relation between the two con-
cepts outside the given context. Birner and Ward (1998: 234) explain that these
two items belong to an ad-hoc partially ordered set, {things children need}.
This example and others like it raise the question whether the elaborate logical
apparatus of the poset is really necessary for the description of topicalization.
The possibility of ad-hoc posets means, in essence, that any items that con-
stitute a pair in a given context qualify as a poset. 21 It can be argued that this
loose conception of the relation between the linked items obviates the need for
partially ordered set concept. The point is that the linked items in topicalization
are contextually related in a way that is recognizable by the addressee.
6.2.1.2. Marking a focus in relation to a given proposition
Prince, as well as Birner and Ward, argue that in addition to marking a
relation pertaining to the preposed constituent, topicalization simultaneously
marks a different constituent as focus. The focus in a topicalized clause re-
mains in its original position and is marked by accent alone. Thus in Prince’s
view, a sentence of the form Cake I like to eat with a spoon would only be
appropriate in a context in which “I like to eat cake with an x ” is an activated
20. Birner and Ward (1998) have extended and modified Prince’s theory of topicaliza-
tion to explain a wide variety of marked word-order constructions in English and other lan-
guages; my comments on their work are restricted to its ramifications for preposing.
21. I am indebted to Yael Ziv for this insight (personal communication).

proposition, and “spoon” is the new value for x. 22 Birner and Ward (1998)
agree with Prince that most types of topicalization, except for those involving
preposed locatives, mark a focus.
Gregory and Michaelis (2001: 1673 n. 8), raise doubts regarding the focus-
marking function of topicalization. They note that the final accent in Cake, I
like to eat with a spoon need not represent constituent focus but can just as
well be interpreted as the unmarked articulation, in which the accent falls on
the last lexical expression in the clause. 23 Thus, “I like to eat cake with an x ” is
not necessarily given information with respect to the clause. 24
An examination of some of the examples of topicalization cited by Birner
and Ward (1998) confirms that topicalization need not involve constituent fo-
cus. In examples (136) and (137), it is unlikely that the (presumably) accented
constituent is interpretable as the focus.
(136) S. Pintzuk to G. Ward in conversation about Ph.D. exam

(Birner and Ward 1998: 22)
G: So, how’d it go?
S: The historical question I had some problems with, but I think
it’s ok. The descriptive I just wrote a whole lot.
(137) M. Schultz to G. Ward in conversation (Birner and Ward 1998: 222)

G: How could you take an exam with all those students pestering
you?
M: It wasn’t easy, but they were all done by three-thirty, and I had
until five. The hard part I left for the end.
In the underlined topicalized clause in (136), the accented constituent is pre-

sumably problems. Yet replacing this constituent with a variable yields a prop-
osition that is neither activated nor inferrable: “The historical question I had
some x (or had x) with.” Similarly, in (137) end is presumably accented, but
“The hard part I left for x ” cannot be reasonably be considered an activated
proposition in this context.
22. In Prince (1985), the given proposition is arrived at by replacing the accented constit-
uent with a variable and by replacing the preposed constituent with the set containing it, that
is, “I like to eat {types of foods} with an x.” In Prince (1998: 293) the procedure is described
somewhat differently: the preposed constituent is not replaced by the set containing it but
is accompanied by an explicit representation of the set relation; that is, “I like to eat cake, a
member of the set {types of food}, with an x.” It is unclear to me how the latter formulation
differs in its propositional content from the simpler version, “I like to eat cake with an x.”
23. This is what Lambrecht calls the “predicate-focus” articulation, in which the focus is
projected from the accented constituent over the entire verb phrase (see §3.3.7).
24. It can be further noted that, under the predicate-focus interpretation, “I do x ” cannot
be assumed to be given information either; see §6.1.1.

6.2.2. Topicalization in Biblical Hebrew: A discourse-connective device

In this section, I present a new characterization of the function of topical-
ization as it is manifested in BH. The description here is based in part on the
insights of Prince and Birner and Ward but also draws on a different area of
pragmatic research―the function of conjunct adverbials, also known as dis-
course markers or discourse connectives (see §5.2.1, p. 69 n. 13).
A problematic aspect of both Prince’s and Birner and Ward’s descriptions
of topicalization is that they do not specify where the linked item or the item
from which it is inferred is located in the previous discourse. It can be argued
that without a specification of this sort their descriptions are too broad to be
useful for cross-linguistic comparison: since most clause elements are related
at least indirectly to something mentioned at some earlier point (especially in
long narrative texts), nearly any preposed clause can by viewed as topicalized
under these descriptions. In most of the cited examples in Birner and Ward,
however, the source of the linked item is the immediately preceding context. It
is unlikely that this is a coincidence.
As a working hypothesis, then, it is posited that topicalization indicates a
contextual relation between the preposed constituent and another element in
the immediately preceding context. An examination of the Genesis corpus, as
described in the following chapters, reveals a large number of preposed clauses
that satisfy this description, thus justifying the topicalization category for this
language. There is also, however, a group of BH clauses in which the clause
containing the preposed entity and the clause containing the item to which it
is related are not immediately adjacent to each other. This type is addressed
further in this section in the discussion of discourse connectives.
In another group of BH clauses, the preposed element is related to an ele-
ment in the immediately following context, rather than the preceding context.
In fact, a topicalized clause may occur at the beginning of a discourse, where
the only possible relation must concern the following clause (see §9.3). It is
probable that forward-pointing topicalization occurs in English as well, al-
though this cannot be conclusively demonstrated from the examples cited by
Birner and Ward.
As will be shown in chap. 9, ad-hoc links between entities bearing no natu-
ral relation are not uncommon in BH topicalization; the only requirement is
that the linked entities bear a contextual relation which is perceptible by the
addressee. As in English, BH topicalization does not necessarily involve a con-
stituent focus.
Marking a relation between entities, however, is only one functional aspect
of BH topicalization and, mostly likely, of English topicalization as well. Con-
sider again example (135), repeated in (138):
(138) “Sleep they can catch up on, Mom they can’t.”

The ad-hoc contextual relation between sleep and Mom is hardly significant
in and of itself. Rather, this relation serves to indicate a larger-scale relation
between the clauses—in this case, contrast. In most cases of BH topicalization,
a specific relation, usually contrast or similarity, exists between the topicalized
clause and the clause containing the other linked entity. When the two linked
items are identical, the relation between clauses is often one of explanation or
restatement. Occasionally, the relation is one of temporal sequence.
In marking relations between clauses, BH topicalization bears a functional
resemblance to discourse connectives, which are words or phrases that mark
the relation between two adjacent sentences or text segments (Schiffrin 1987:
31; Fraser 1990, 1996; Schourup 1999). 25 Discourse connectives and the rela-
tions they signal have been studied by researchers in pragmatics as well as by
psycholinguists. Relations between text segments, including causal, contras-
tive, and concessive relations, are frequently referred to as coherence rela-
tions (Sanders 1997; Sanders and Noordman 2000; Knott 2001; Sanders and
Spooren 2001). Coherence relations are often left to the reader to infer and not
explicitly signaled on the formal level. Alternatively, coherence relations may
be signaled by discourse connectives such as so, but, however, and similarly.
In the context of psycholinguistic theories of comprehension, discourse con-
nectives are said to facilitate the building of a coherent text representation by
guiding the reader in correctly linking the representations of the two segments
(Sanders and Noordman 2000; Sanders and Spooren 2001).
Coherence relations are generally divided into two types: those that con-
cern the content of the two segments, and those that concern the speaker’s
reasoning or the speech acts accomplished by the segments. These types are
sometimes known as semantic and pragmatic relations, respectively (Van Dijk
1979: 449); 26 alternative terms are external and internal relations (Halliday
and Hasan 1976: 240), and ideational and rhetorical relations (Redeker 1990).
Sweetser (1990: 21) divides the pragmatic relation category into epistemic (re-
lating to the speaker’s beliefs or reasoning) and speech act relations. Following
Sweetser, I will refer to three types of coherence relations: content, epistemic,
and speech-act relations. 27
The difference between the various types of relations is illustrated in (139).
As the examples show, a single discourse connective can often signal more
than one type of relation.
25. Many discourse connectives are conjunct adverbs or adverbials (see §5.2.1).
26. Van Dijk’s use of the term pragmatic is narrower than the one adopted in the present
work. I am using this term to include all aspects of the relations between clauses and their
contexts, including the so-called semantic relations.
27. The question of whether epistemic and speech-act relations should be subsumed un-
der a single category does not concern us here; for an attempt to settle this issue, see Knott
(2001). On the relations signaled by discourse connectives in Modern Hebrew, see Abadi
(1988).

(139) Semantic, epistemic, and speech-act relations (Knott 2001)

a. Bill was starving, so he had a sandwich.
b. Bill had five sandwiches, so he must have been starving.
c. Bill is starving. So why isn’t he eating?
In all three examples, the discourse connective so marks a cause-and-effect re-

lation between the clauses, but in each case the relation is of a different type. In
(139a), the discourse connective marks a relation between the contents of the
two clauses. In (139b), so marks a relation concerning the speaker’s reasoning:
the clause he must have been starving represents a conclusion derived from the
premise “Bill had five sandwiches.” The sequence can be paraphrased as Bill
had five sandwiches. It may be concluded that he must have been starving. In
(139c), so marks a relation between the fact that Bill is starving and the speech
act of questioning. The sequence can be paraphrased as Bill is starving. There-
fore I ask, “Why isn’t he eating?”
It is proposed here that topicalization is a kind of generalized discourse
connective. By marking a link between the preposed item and a second item,
topicalization cues the reader to the coherence relation between the segments
containing these items, 28 a relation which specifically concerns the pair of
linked items. The exact nature of the coherence relation is not specified, but is
left to the reader’s inference. In marking either forward-pointing or backward-
pointing relations, topicalization resembles some BH conjuncts that may point
either forward or backward. An example is ‫הלא‬, which marks the justification
for a preceding or a following clause (see §7.3.2.1, p. 117 n. 21).
Viewing topicalization as a discourse-connective device explains why in
some cases in BH the linked items are not in adjacent clauses. As mentioned
above, discourse connectives can link adjacent sentences or adjacent text seg-
ments of longer length. In cases in which the linked items in BH topicalization
are not in adjacent clauses, the items nonetheless stand at the head of adjacent
text segments.
A full examination of topicalization in BH is presented in chap. 9.
28. Blakemore (1987: 86) notes that discourse connectives do not always link segments
in a single discourse, but may relate to a different speaker’s utterance, as in (1), or even to
the extratextual situation, as in (2):
(1) A: You take the first turning on the left.
B: So I don’t go past the hospital. (Blakemore 1987: 85)
(2) [Having seen someone arrive home laden with parcels]
A: So you’ve spent all your money. (Blakemore 1987: 86)
Some of the examples of topicalization cited by Birner and Ward (1998) are similar to (1),
linking the preposed item to an item in a different speaker’s utterance, as in (3).
(3) A: You know this album?
B: This song I know. (Birner and Ward 1998: 44)
This type of topicalization does not appear to occur in the Genesis corpus, but occurs in
Jer 1:14, where the preposed ‫‘ מצפון‬from the north’ in God’s utterance is linked to ‫‘ צפונה‬the
north’ in Jeremiah’s immediately preceding utterance.

Chapter 7
The Pragmatics of Preposing:

A Statistical Analysis
In this chapter, the pragmatic function of preposing in the 401 marked pre-
posed clauses in Genesis (see §3.6) is analyzed, with the aim of establish-
ing the dominance of the information-structure concepts of focusing and
topicalization.
7.1. Focusing and topicalization in preposed clauses

Table 6 shows the frequency of information-structure functions in preposed
clauses in Genesis. Topicalization is shown to be a widespread function of
preposing, occurring in 42.5% of all preposed clauses. Focusing is consid-
erably less common, accounting for 12.4%. Including the seven clauses that
are ambiguous—either focused or topicalized—the two information-structure
functions account for 56.6% of preposing in Genesis.
A closer examination of the data reveals that the frequency of information-
structure functions varies according to the syntactic function of the preposed
constituent (subject vs. complement/adjunct) 1 and according to the register of
the clause (narrative vs. direct speech). The figures are displayed in tables 7 and
8, respectively. 2 Table 7 shows that information-structure functions constitute
the majority of complement/adjunct-preposed clauses, 68.6% of the total, but
a minority of subject-preposed clauses, only 43.6%. The chi-square test shows
that the difference between the two frequencies is statistically significant.
Table 8 shows that information-structure functions account for 66.7% of
narrative clauses and a smaller portion, 50%, of direct-speech clauses. Again,
the chi-square test shows that the difference between the two groups is statisti-
cally significant. In §§7.2 and 7.3, I examine complement/adjunct preposing
and subject preposing separately. In each section, the functions of the preposed
clauses in narrative is compared to their functions in direct speech.
1. As explained in §4.5, I do not systematically differentiate between complements and

adjuncts, due to practical and theoretical difficulties in doing so.
2. As noted in §5.6 n. 77 (p. 88), in two of the 402 clauses, the syntactic function of
the preposed constituent cannot be definitively determined. These clauses are omitted from
table 7.
104

105 Focusing and topicalization in preposed clauses 7.2.
Table 6. Pragmatic functions of preposed clauses in Genesis.
Number Percentage
Focusing 50 12.4
Topicalization 171 42.5
4
Focusing/topicalization 7 1.7
Residue 174 43.3
Total 402
Table 7. Relation between syntactic function of the preposed constituent

and pragmatic function of preposing.
Information
Syntactic function structure Other Total
Subject No. 86 110 196
% 43.9 56.1
Complement/adjunct No. 140 64 204
% 68.6 31.4
Total No. 226 174 400
% 56.5 43.5
chi-square = 24.914
p ¯ .001
7.2. Focusing and topicalization in

complement/adjunct-preposed clauses 3
Table 7 shows that information-structure functions constitute 68% percent
of all clauses with a preposed complement/adjunct. As mentioned in §3.3.5, it
has been suggested that preposed time adjuncts may be preposed for no reason
other than to provide a setting for the clause. I checked various representa-
tive time-adjunct types, examining all of the preposed occurrences in finite
nonsubordinate clauses in the classical BH prose corpus. The expressions
checked were ‫‘ מחר‬tomorrow’, ‫‘ היום‬today’, ‫‘ ביום‬on the . . . day’, ‫‘ בחדש‬on
the . . . month’, ‫בשנת‬/‫‘ בשנה‬in the . . . year’; and temporal infinitive phrases
3. This category includes clauses that allow either a focused or a topicalized interpreta-
tion; i.e., Gen 8:17, 9:2, 18:7, 34:26, 34:29, 43:15 (‫)בידכם‬, and 46:7.

7.2. Focusing and topicalization in preposed clauses 106
Table 8. Relation between register and

pragmatic function of preposing.
Information
Register structure Other Total
Narrative No. 108 54 162
% 66.7 33.3
Direct speech No. 120 120 240
% 50.0 50.0
Total No. 228 174 402
% 56.7 43.3
chi-square = 10.944
p ¯ .001
Table 9. Pragmatic functions of complement/adjunct

preposed clauses.
Pragmatic function Number Percentage

Focusing 42 20.5
Focusing/topicalization 5 2.5
Residue 64 31.4
Total 204
involving the prepositions ‫כ‬, ‫ב‬, or ‫אחרי‬. Of the 123 clauses containing pre-
posed expressions of these types, a total of 72% have an information-structure
function. When this figure is compared to the 68% of complement/adjunct-
preposed clauses that have an information-structure function, it can be seen
that information-structure functions are at least as dominant in time-adjunct-
preposed clauses as in complement/adjunct-preposed clauses in general. These
results indicate that there is no reason to consider time-adjunct preposing to be
functionally distinct from complement/adjunct preposing in general.
Table 9 shows the distribution of pragmatic functions in complement/ad-
junct preposed clauses. The table shows that topicalization is over twice as
common as focusing in complement/adjunct preposed clauses, 45.6%, as com-
pared to 20.5%.

Table 10. Pragmatic functions of complement/adjunct-preposed clauses:

Narrative vs. direct speech.
Focus./
Register Focus. Topic. Topic. Residue Total
Narrative No. 7 47 4 7 65
% 10.8 72.3 6.2 10.8
Direct speech No. 35 46 1 57 139
% 25.2 33.1 0.7 41.0
Total No. 42 93 5 64 204
% 20.6 45.6 2.5 31.4
In table 10, the functions of complement/adjunct preposing are tallied

separately for narrative and direct speech. 4 Several interesting facts may be
observed. First, the frequencies of focusing and topicalization constructions
differ sharply in narrative and direct speech. In narrative, focusing (10.8%) is
much less common than topicalization (72.3%). In direct speech, the frequen-
cies of the two functions are more similar: 25.2% focusing and 33.1% topical-
ization. Comparing the size of the residue group in the two registers, we see
that almost all the residue clauses are direct speech clauses. Of the 64 clauses
in the residue column, 57 (89%) are in direct speech and only 7 (11%) are in
narrative. Looked at another way, only 7 out of 65 narrative clauses (11%) are
residue, whereas 57 out of 139 direct-speech clauses (41%) are residue.
It may be concluded from these figures that focusing and topicalization are
the predominant and perhaps the only significant functions of complement/
adjunct preposing in narrative. In direct speech, though focusing and topical-
ization still together account for the majority (58%) of complement/adjunct
preposing, other factors, presently unclear, are obviously at work as well. In
the ensuing discussion, I offer some tentative remarks regarding the residue,
without attempting to give a systematic account.
7.2.1. The residue: Possible instances of focusing

A number of the clauses in the direct-speech residue may actually be in-
tended to be focused. In the present study, clauses are classified as focused
only if examination of the context makes it clear that a proposition relating
4. The chi-square text was not performed on this table. The chi-square test is not consid-
ered accurate for this table due to the presence of two cells with frequencies of less than 5.

7.2. Focusing and topicalization in preposed clauses 108
to this focus is activated or accessible for the addressee. This procedure may
not identify all of the cases in which the speaker/writer is treating a particular
proposition as activated. The preposed clauses in (140) and (141), for example,
may have been intended to be focused, although they are not classified as such
due to lack of evidence of an activated proposition.
(140) Gen 29:25

‫ויהי בבקר והנה הוא לאה ויאמר אל לבן מה זאת עשׂית לי הלא ברחל עבדתי‬
:‫עמך ולמה רמיתני‬
And it came to pass in the morning, and behold, it was Leah; and Jacob
said to Laban, “What have you done to me? Surely for Rachel I served
you! And why have you deceived me?” 5
(141) Gen 24:50

‫ יצא הדבר‬′‫ויען לבן ובתואל ויאמרו מי‬
And Laban and Bethuel answered, and said, “From the LORD the thing
has come.”
In (140), it is hard to determine whether “I worked for you for x ” is accessible
by inference from the preceding clause, ‫‘ מה זאת עשית לי‬What have you done to
me?’ Nevertheless, it is certainly plausible that Jacob assumes that this proposi-
tion is accessible. Such an explanation is plausible for (141) as well.
A different phenomenon, known as accommodation, involves a speaker
treating a proposition as presupposed or activated, even though he knows this
is unlikely. An example given by Dryer is a speaker who says “Watch out, the
dog will bite you,” even though he knows the addressee is unaware of any dog
nearby. By acting as if the existence of the dog is presupposed, the speaker in-
duces the addressee to adopt this presupposition. 6 The sentence serves as short-
hand for “There is a dog nearby, and he is likely to bite you.” Accommodation
may be at work in clauses such as (142). 7
(142) Gen 12:7

‫וירא י‘ אל אברם ויאמר לזרעך אתן את הארץ הזאת‬
And the LORD appeared to Abram, and said, “To your offspring I will
give this land.”
5. The translation follows NJPSV in giving ‫ הלא‬an asseverative rather than an interrogative
interpretation (see §5.2.1, p. 70 n. 19 for the distinction between the two). In this clause, ‫הלא‬
marks the justification for the preceding rhetorical question, as explained further in §7.3.2.1.
6. Chafe (1976: 34) calls this a “quasi-given” state, where the speaker engages in a pre-
tense that conscious information is involved. On accommodation, see Lewis (1979); Dryer
(1996).
7. Similar instances are Gen 15:18–21, 24:7, and 31:15. See also Gen 21:6, where the
speaker is addressing a nonspecific audience or herself.

In this verse God speaks as if “I shall give this land to x ” is an activated propo-
sition, although there is no reason to assume that Abram was thinking about
this proposition.
7.2.2. The residue: Fixed expressions with preposed word order

In some of the clauses in the direct-speech residue, preposing seems to be
due to the presence of a fixed expression with characteristically preposed word
order. This type of preposing does not have any pragmatic function at all. An
example is the expression ‫בי נשבעתי‬:
(143) Gen 22:16

‫ויאמר בי נשבעתי נאם י‘ כי יען אשר עשית את הדבר הזה‬
By myself I swear, says the LORD, that because you have done this
This expression occurs three times elsewhere, each time with preposed ‫בי‬. 8 Al-
though these other occurrences are not in the classical prose corpus, the striking
resemblance to the word order in (143) seems to indicate a fixed expression.
Another recurring expression that is usually preposed is ‫ כדבר הזה‬or ‫כדברים‬
‫ האלה‬with the meaning ‘such and such’, as used in direct speech: 9
(144) Gen 32:20

‫ויצו גם את השני גם את השלישי גם את כל ההלכים אחרי העדרים לאמר כדבר‬
‫הזה תדברון אל עשו במצאכם אתו‬:
He also instructed the second and the third and all who followed the
droves, saying, “Such and such you shall say to Esau when you meet
him.”
(145) Gen 39:19

‫ויהי כשמע אדניו את דברי אשתו אשר דברה אליו לאמר כדברים האלה עשה‬
‫לי עבדך‬
When his master heard the words which his wife spoke to him, saying
“Such and such your servant did to me”
Another characteristically preposed formula is /‫ אעשה‬. . . ‫צוה‬/‫כל אשר אמר‬

‫‘ נעשה‬whatever x says/commands, I/we shall do’: 10
8. These are Isa 45:23; Jer 22:5, 49:13. The similar expression ‫ נשבעתי‬′‫‘ בי‬by the LORD
I swear’ occurs in 2 Sam 19:8.
9. Additional examples include 1 Sam 18:24 and 2 Sam 17:6. In 2 Sam 14:3, the equiv-
alent expression is not preposed.
10. Additional examples include Exod 24:3, 7; Josh 1:16. A similar formula has a left-
dislocated structure instead of a preposed clause (Num 22:20, 23:26). Folmer (1995: 569–72)
shows that in Aramaic texts of the Achaemenid period, phrases containing general relative
clauses of the form ‫ זי‬. . . ‫‘ כל‬every x that’ are usually preposed; this is also true of similar
types such as ‫‘ מה זי‬whatever’ and ‫‘ מן זי‬whoever’. (I am indebted to Richard Steiner for the

7.3. Topicalization and focusing in subject-preposing constructions 110
(146) Gen 31:16

:‫ועתה כל אשר אמר אלהים אליך עשה‬
And now, whatever God has said to you, do.
(147) Exod 19:8

‫ויענו כל העם יחדו ויאמרו כל אשר דבר י‘ נעשה‬
And all the people answered together and said, “Whatever the LORD has
spoken we will do.”
(148) 2 Kgs 10:5

‫עבדיך אנחנו וכל אשר תאמר אלינו נעשה‬
We are your servants, and whatever you say to us we will do.
A similar formula occurs in (149)–(151), with the head of the relative clause
omitted.
(149) Gen 41:55

:‫לכו אל יוסף אשר יאמר לכם תעשו‬
Go to Joseph; whatever he says to you, you shall do.
(150) Gen 34:11

:‫אמצא חן בעיניכם ואשר תאמרו אלי אתן‬
Let me find favor in your eyes, and whatever you say to me I will give.
(151) 2 Sam 18:4

‫ויאמר אליהם המלך אשר ייטב בעיניכם אעשה‬
The king said to them, “Whatever you think best I will do.”
7.3. Topicalization and focusing in subject-preposing constructions

Table 7 (p. 105) showed that focusing and topicalization account for only
43.9% of subject-preposed clauses. As discussed in chap. 3 (pp. 18–47), sub-
ject preposing appears to have a number of functions that relate to the clause as
a whole, rather than to the preposed subject in particular; these include marking
anteriority or a pair of simultaneous events (§3.1.2), marking a new narrative
unit/scene (§3.1.3), or marking background information (§3.2). It is sometimes
difficult to distinguish between the use of preposing for topicalization and the
other functions. When a preposed subject is the same as an item in the pre-
ceding clause, for example, the clause fits the criterion for topicalization. 11 In
Folmer reference.) In BH, however, ‫ כל אשר‬phrases are not typically preposed. It is only in
the formula exemplified by examples (146)–(148) that ‫ כל אשר‬is typically found in initial
position.
11. In a case of this sort, the relation between linked items is one of identity. See §9.2
for further discussion.

111 Topicalization and focusing in subject-preposing constructions 7.3.
Table 11. Pragmatic functions of subject-preposed clauses.
Pragmatic function Number Percentage

Focusing 9 4.6
Focusing/Topicalization 2 1.0
Residue 110 56.1
Total 196
these cases, it is necessary to decide whether the preposing is for the purpose
of linking the identical entities and the clauses containing them or whether the
purpose is to mark a different function such as anteriority or simultaneity.
Table 11 shows the distribution of information-structure functions in
subject-preposed clauses. Cases clearly intended to have a function relating to
the clause as a whole are counted as residue even if the sentence could tech-
nically be taken as topicalization. If the data in table 11 are compared to the
comparable data for complement/adjunct-preposed clauses in table 9 (p. 106),
it can be seen that focusing is much less common in subject-preposed clauses.
Only 4.6% of subject-preposed clauses are focused, as compared to 20.5%
of complement/adjunct-preposed clauses. The frequency of topicalization in
both groups is more similar: 38.3% of subject-preposed clauses and 45.6% of
complement/adjunct-preposed clauses are topicalized. Turning to the residue,
it can be seen that the residue of subject-preposed clauses is much larger than
the residue of complement/adjunct clauses, at 56.1% as compared to 31.4%.
In table 12 (p. 112), the frequencies of information-structure functions
in subject-preposed clauses are tabulated separately for narrative and direct
speech. Although some differences may be noted between the two registers,
the differences are probably not statistically significant. 12 Topicalization is
much more common than focusing in both registers, although the frequency is
somewhat higher in narrative, constituting 47.4% of the total, as compared to
only 30.7% in direct speech. In both registers, focusing is a rare phenomenon,
occurring in 3.2% of narrative clauses and 5.0% of direct-speech clauses. In-
terestingly, the residue of subject preposing constitutes nearly 50% or more of
the clauses in both registers. This contrasts with the residue of complement/
adjunct preposing in narrative, which is quite small (see table 10, p. 107). Al-
though the sizes of the residues in narrative and direct speech are comparable,
12. The chi-square test cannot be performed on this table because one of the cells has a
frequency of 0.

Table 12. Pragmatic functions of subject-preposed clauses:

Narrative versus direct speech.
Focus./
Register Focus. Topic. Topic. Residue Total
Narrative No. 3 45 0 47 95
% 3.2 47.4 0 49.5
Direct Speech No. 5 31 2 63 101
% 5.0 30.7 2.0 62.4
Total No. 8 76 2 110 196
% 4.1 38.8 1.0 56.1
the functional distribution of the residue clauses differs considerably in the two
registers, as I discuss next.
7.3.1. The residue of subject-preposed narrative clauses

About three-quarters of the subject-preposed clauses in the narrative resi-
due (34 clauses) appear to have a function relating to the clause as a whole.
The function of preposing in the remaining third of the residue is not clear. In
§§7.3.1.1–7.3.1.4, I list of the functions that were identified, which are mark-
ing anteriority, simultaneity, background information, or the beginning of a
new narrative unit or a scene within the narrative. Because these have already
been discussed in detail in chap. 3, I merely cite examples of each type from
the Genesis corpus and elsewhere.
The examples cited include, wherever possible, an instance with a preposed
lexical expression and an instance with a preposed pronoun. The pronomi-
nal examples serve as evidence that functions relating to the entire clause are
marked by preposing the subject, as opposed to a complement or adjunct: in
most of the pronominal examples, a complement or adjunct is present that
could have been preposed, but instead the normally dropped subject pronoun
is retained in order to produce the desired subject-verb order.
7.3.1.1. Anteriority
Preposing to mark anteriority is shown in (152) and (153). The anterior
preposed clause can often be rendered with the English past perfect, as in these
examples. 13
13. Additional examples from the Genesis corpus include Gen 20:4 and 31:19. Addi-
tional representative examples from the classical BH prose corpus include Josh 2:6; 1 Sam
9:15, 30:1.

(152) Gen 31:33–34

‫ויבא לבן באהל יעקב ובאהל לאה ובאהל שתי האמהת ולא מצא ויצא מאהל‬
‫ ורחל לקחה את התרפים ותשמם בכר הגמל ותשב עליהם‬:‫לאה ויבא באהל רחל‬
And Laban went into Jacob’s tent, and into Leah’s tent, and into the
tent of the two maidservants, and he did not find them. And he came
out of Leah’s tent, and he entered Rachel’s tent. And Rachel had taken
the household gods, and she put them in the camel cushion, and she sat
upon them.
(153) Josh 2:4–6

‫ ויהי השער לסגור בחשך‬:‫ותאמר כן באו אלי האנשים ולא ידעתי מאין המה‬
:‫והאנשים יצאו לא ידעתי אנה הלכו האנשים רדפו מהר אחריהם כי תשיגום‬
:‫והיא העלתם הגגה ותטמנם בפשתי העץ הערכות לה על הגג‬
And she said, “True, the men came to me, and I did not know where they
were from. And the gate was [about] to be closed, at dark, and the men
went out; I do not know where the men went; pursue them quickly, for
you will overtake them.” And she had brought them up to the roof, and
she hid them under the stalks of flax which she had laid out on the roof.
7.3.1.2. Simultaneity
Preposing can be used to mark a simultaneous event. In several instances,
the preposed clause is simultaneous with an immediately preceding nonverbal
or participial clause: 14
(154) Gen 38:25

‫הוא מוצאת והיא שלחה אל חמיה לאמר לאיש אשר אלה לו אנכי הרה‬
She was being brought out, and she sent word to her father-in-law, say-
ing “By the man to whom these belong, I am with child.”
In (155), a triplet of preposed finite clauses denotes three simultaneous or

nearly simultaneous events: 15
14. Appears in chap. 3 as (7) (p. 23). Additional examples from the Genesis corpus in-
clude 29:9, 38:25, and 44:3. Additional representative examples from the classical BH prose
corpus include Judg 18:3; 1 Sam 9:11; 1 Kgs 20:39.
15. As discussed in §3.1.2, it seems likely that simultaneity and near-simultaneity con-
stitute a single pragmatic category. Pairs of finite preposed simultaneous clauses occur, for
example, in Josh 2:8, Judg 18:22, 1 Sam 9:5, and 2 Sam 2:24.

(155) Gen 19:23–24

‫ המטיר על סדם ועל עמרה גפרית ואש‬′‫ וי‬:‫השמש יצא על הארץ ולוט בא צערה‬
:‫ מן השמים‬′‫מאת י‬
The sun rose on the earth, and Lot came to Zoar. And the LORD rained
upon Sodom and on Gemorrah sulfur and fire from the LORD from heaven.
In two cases, a preposed clause follows a ‫ ויהי‬clause with which it is simul-

taneous: 16
(156) Gen 15:12

‫ויהי השמש לבוא ותרדמה נפלה על אברם‬
And the sun was setting, and a deep sleep fell upon Abraham.
7.3.1.3. Background information

A preposed clause may provide background information that is pertinent to
the narrative as it is unfolding. This function partially overlaps with the ante-
riority category, as anterior events often constitute background information.
Background information, however, may be a current state or event: 17
(157) Gen 48:9–10
‫ויאמר יוסף אל אביו בני הם אשר נתן לי אלהים בזה ויאמר קחם נא אלי‬
‫ ועיני ישראל כבדו מזקן לא יוכל לראות ויגש אתם אליו וישק להם‬:‫ואברכם‬
: ‫ויחבק להם‬
And Joseph said to his father, “They are my sons, whom God has given
me here.” And he said, “Bring them to me, please, and I will bless them.”
And Israel’s eyes were dim with age; he could not see. And Joseph
brought them near him; and he kissed them and embraced them.
(158) Gen 42:22–23

‫ויען ראובן אתם לאמר הלוא אמרתי אליכם לאמר אל תחטאו בילד ולא שמעתם‬
:‫ והם לא ידעו כי שמע יוסף כי המליץ בינתם‬:‫וגם דמו הנה נדרש‬
And Reuben answered them, “Did I not tell you, ‘Do not sin against the
boy?’ And you did not listen. And behold, there comes a reckoning for
his blood.” And they did not know that Joseph understood, for there was
an interpreter between them.
16. The other example is Gen 22:1. In a different pattern, the ‫ ויהי‬clause is a bare verbal
form without a temporal adverbial, followed by two preposed simultaneous clauses (Gen
15:17 and 27:30, ‫)ויהי אך‬. See §4.4.2 for the syntactic analysis of ‫ ויהי‬clauses adopted in the
present work.
17. An additional example is Gen 37:3.

7.3.1.4. New narrative unit or new scene within the narrative

In several instances, preposing marks the beginning of a new narrative unit:
(159) Gen 16:1

:‫ושרי אשת אברם לא ילדה לו ולה שפחה מצרית ושמה הגר‬
And Sarai, Abram’s wife, had not borne him children. And she had an
Egyptian maidservant, and her name was Hagar.
(160) Gen 39:1

‫ויוסף הורד מצרימה ויקנהו פוטיפר סריס פרעה שר הטבחים איש מצרי מיד‬
:‫הישמעאלים אשר הורדהו שמה‬
And Joseph was taken down to Egypt. And Potiphar, an officer of Pha-
raoh, the captain of the guard, an Egyptian, bought him from the Ishma-
elites who had brought him down there.
Both (159) and (160) occur at the beginning of a ‫‘ פרשה סתומה‬closed portion’
(see §3.1.3). 18
Somewhat more frequently, subject-preposing marks the clause as begin-
ning a new scene within a narrative unit. The new scene is characterized by
new participants and/or a new setting: 19
(161) Gen 24:61–62

‫ותקם רבקה ונערתיה ותרכבנה על הגמלים ותלכנה אחרי האיש ויקח העבד את‬
‫ ויצחק בא מבוא באר לחי ראי‬:‫רבקה וילך‬
And Rebekah and her maids arose, and they mounted the camels and fol-
lowed the man; and the servant took Rebekah, and he went his way. And
Isaac had come from Beer-lahai-roi
(162) Gen 26:25–26

: ‫ ויט שם אהלו ויכרו שם עבדי יצחק באר‬′‫ויבן שם מזבח ויקרא בשם י‬
:‫ואבימלך הלך אליו מגרר ואחזת מרעהו ופיכל שר צבאו‬
And he built an altar there, and he called upon the name of the LORD,
and he pitched his tent there. And Isaac’s servants dug a well there. And
Abimelech went to him from Gerar, and Ahuzzath his adviser and Phicol
the commander of his army.
7.3.2. The residue of subject-preposed direct speech clauses

The residue of subject-preposing in direct speech is more difficult to cat-
egorize than the narrative residue. The functions described above were not
18. Additional examples include Gen 3:1, 4:1, and 21:1. The latter two clauses are at the
beginning of a ‫פרשה סתומה‬.
19. Additional examples include Gen 13:14, 18:17, 34:5 (‫)ויעקב‬, and 37:36.

found in this group. Only 11 clauses (17.5%) in the residue of subject-preposed

direct speech clauses fall into plausible pragmatic categories, as discussed in
§§7.3.2.1–7.3.2.3.
7.3.2.1. Justification
A specialized use of preposing is found in several clauses traditionally
considered to be circumstantial clauses. As discussed in §3.1.2 (pp. 20–21),
although narrative texts exhibit ample instances of nonverbal and participial
circumstantial clauses, finite circumstantial clauses are scarce in these texts.
The examples cited in the literature are mostly found in direct speech and typi-
cally follow a directive or a rhetorical question. The distinctive environment of
these preposed clauses points to a different interpretation of this marked con-
struction. I propose that preposing in these verses has the pragmatic function
of marking the justification for the preceding utterance:
(163) Gen 24:56

‫ הצליח דרכי‬′‫אל תאחרו אתי וי‬
Do not delay me. And (= for) the LORD has made my journey successful.
In (163), the preposed clause ‫ וי‘ הצליח דרכי‬marks the justification for the pre-
ceding directive ‫אל תאחרו אתי‬. 20 The coordinated phrases may be idiomatically
rendered ‘Do not delay me, for the LORD has prospered my way’.
In examples (164)–(166), the justification provided by the preposed clause
relates to the implication of a rhetorical question. Rhetorical yes-no questions
imply an assertion with the opposite polarity of the question (Pope 1972: 46).
For example, “Didn’t I just tell you that yesterday?” implies the assertion “I
told you that yesterday.” Question-word rhetorical questions frequently imply
that the answer is a null set; for example, “Who could pass that test” implies
“Nobody could pass that test” (Pope 1972: 59). Examples of preposing to jus-
tify rhetorical questions include:
(164) Gen 24:31

: ‫למה תעמד בחוץ ואנכי פניתי הבית ומקום לגמלים‬
Why do you stand outside? And (= for) I have prepared the house and a
place for the camels.
(165) Gen 18:13

:‫למה זה צחקה שרה לאמר האף אמנם אלד ואני זקנתי‬
Why did Sarah laugh, and say, “Shall I indeed bear a child? And (= for)
I am old.”
20. See also Gen 26:27 and 32:12. Examples found in the classical BH prose corpus
include Judg 21:7; 1 Sam 15:6, 16:1, 28:16.

(166) Gen 18:17–18

‫ ואברהם היו יהיה לגוי גדול ועצום‬:‫המכסה אני מאברהם אשר אני עשה‬
Shall I hide from Abraham what I am about to do? And (= for) Abraham
shall become a great and mighty nation
In (164), the rhetorical question ‫‘ למה תעמד בחוץ‬Why do you stand outside?’
implies the assertion “There is no reason for you to stand outside.” The subse-
quent preposed clause, ‫‘ ואנכי פניתי הבית‬I have prepared the house’, supplies a
justification for this implication. In (165), the question ‫‘ האף אמנם אלד‬Shall I
indeed bear a child?’ implies “I will not bear a child.” The preposed clause, ‫ואני‬
‫‘ זקנתי‬I am old’, supports this prediction. The relation between the rhetorical
question and preposed clause in (166) is comparable.
Preposing for justification is an alternative to marking with the clausal ad-
verb ‫( הלא‬see §5.2.1). As in the case of preposing, utterances justified by ‫הלא‬
are usually directives or rhetorical questions (Moshavi 2007b): 21
(167) Gen 19:20

‫הנה נא העיר הזאת קרבה לנוס שמה והוא מצער אמלטה נא שמה הלא מצער‬
‫הוא‬
Behold, that city is near enough to flee to, and it is a little place. Let me
escape there. Surely it is a little place
(168) Gen 44:4–5

‫ הלוא זה אשר ישתה אדני בו‬:‫למה שלמתם רעה תחת טובה‬
Why have you repaid good with evil? Surely this is the one from which
my master drinks
The similarity between justificational clauses marked by preposing and jus-

tificational clauses marked by ‫ הלא‬is illustrated by the close-to-minimal pair
in (169) and (170):
21. See also, e.g., Gen 19:20, 44:5; Judg 6:14, 11:7 (appears as example 170); 1 Sam
17:29; 2 Sam 11:20; 2 Kgs 6:32. The interpretation of justificational ‫ הלא‬as a clausal adverb
is supported by its interchangeability with ‫הנה‬. Representative examples of ‫ הנה‬justifying a
directive include Gen 16:2; 27:2–3, 6–8; 42:2; for ‫ הנה‬justifying a rhetorical question, see,
e.g., Gen 25:32; 1 Sam 20:2, 28:9. Unlike ‫הנה‬, which almost always precedes the justified
utterance, ‫ הלא‬may precede or follow. When a justificational ‫ הלא‬clause contains a preposed
element, ‫ הלא‬occurs in front of the preposed element, in the characteristic position of the
clausal adverb (e.g., Gen 20:5, 29:25, 31:15; Josh 22:20; Judg 4:14, 11:7; 1 Kgs 1:13). For
further discussion on distinguishing the clausal adverb from interrogative ‫הלא‬, see Moshavi
(forthcoming).

(169) Gen 26:27

:‫מדוע באתם אלי ואתם שנאתם אתי ותשלחוני מאתכם‬
Why have you come to me? And (= for) you hate me and sent me away
from you.
(170) Judg 11:7

‫הלא אתם שנאתם אותי ותגרשוני מבית אבי ומדוע באתם אלי עתה כאשר צר‬
:‫לכם‬
Surely you hate me, and you drove me out of my father’s house. Why
have you come to me now when you are in trouble?
The justification in (169) is the segment ‫;ואתם שנאתם אתי ותשלחוני מאתכם‬
in (170), it is the very similar ‫הלא אתם שנאתם אותי ותרגשוני מבית אבי‬. The
supported rhetorical question in both cases is ‫מדוע באתם אלי‬. In (170), the
justificational segment is marked by ‫( הלא‬and by preposing), whereas in (169)
it is marked by preposing alone. One difference between marking justification
by ‫ הלא‬as opposed to preposing is that ‫ הלא‬may precede or follow the justified
utterance, 22 but the preposed clause always follows.
Example (169) illustrates that preposing may mark a justification that is a
segment longer than a clause. The justification in this example includes the
clause following the preposed one, ‫ותרגשוני מבית אבי‬. In such cases, only the
first clause in the segment is preposed.
7.3.2.2. Affirmation
In several cases, a preposed first-person pronoun occurs in a promise to
carry out a certain action. 23 The exact pragmatic function of preposing in this
group is difficult to characterize; to say that the pronoun is “emphatic” is hardly
sufficient. Nevertheless, the pattern exemplified by these clauses is undoubt-
edly significant. It could be that this category and the category of boasting, de-
scribed in §7.3.2.3 (p. 119), are subtypes of a broader category the parameters
of which are currently unclear.
(171) Gen 21:24

:‫ויאמר אברהם אנכי אשבע‬
And Abraham said, “I swear.”
(172) Gen 47:30

:‫ויאמר אנכי אעשה כדברך‬
And he said, “I will do as you have said.”
22. For ‫ הלא‬preceding, see, e.g., Gen 37:13; Judg 15:2, 11; 1 Sam 9:21, 26:15.
23. See Waltke and O’Connor 1990: §16.3.2b. Additional examples are Gen 38:17 and
50:21. Additional clear examples from the classical BH prose corpus are Exod 8:24; Judg
6:18, 17:10; 1 Kgs 2:18, 21:7; 2 Kgs 6:3.

119 Conclusion 7.4.
7.3.2.3. Boasting
Joüon-Muraoka (§146a) note that preposed first-person pronouns occur in
boasts, as in the well-known instances in the Mesha inscription. Genesis con-
tains one clause of this type:
(173) Gen 14:23

:‫ולא תאמר אני העשרתי את אברם‬
And you shall not say, “I made Abram rich.”
An additional example from the classical BH prose corpus is shown in (174). 24
(174) 2 Sam 1:16

:′‫ויאמר אליו דוד דמך על ראשך כי פיך ענה בך לאמר אנכי מתתי את משיח י‬
“And David said to him, “Your blood be on your own head! Your own
mouth testified against you, saying, ‘I put the LORD’s anointed to death.’”
7.4. Conclusion
Statistical analysis of the pragmatic functions of preposing in the Genesis
corpus reveals a complex picture in which information-structure functions play
the leading roles, accompanied by various other functions, not all of which
can be currently identified. Focusing and topicalization account for 56.6% of
all preposed clauses, with topicalization being over three times as common as
focusing. The actual numbers of focused clauses may be somewhat higher,
because some plausible cases were excluded due to lack of textual evidence of
activation. In comparison to information-structure functions, which constitute
the majority of preposed clauses, functions relating to the clause as a whole
(i.e., marking anteriority, simultaneity, background information, the start of a
new scene or episode, justification, affirmation, and boasting) together consti-
tute only 11.2% of preposed clauses over all. Preposing in fixed expressions
accounts for an additional 1.5%. The function of preposing in the remaining
30.6% of clauses is currently unclear; nearly all of these (83.7%) are in direct
speech.
Complement/adjunct preposing and subject-preposing exhibit distinct func-
tional patterns; these patterns also vary according to discourse register. Com-
plement/adjunct preposing in narrative is almost exclusively for the purpose
of focusing or topicalization. In direct speech, however, information-structure
functions constitute only three-fifths of the preposed-complement/adjunct
total. Much of the residue of complement/adjunct preposed clauses in direct
speech is hard to characterize, but in some of the cases preposing appears to
reflect a fixed expression and does not have a pragmatic function. Almost half
24. An additional example is 1 Kgs 1:5.

7.4. Conclusion 120
of subject-preposed clauses have information-structure functions; the number

is a bit higher in narrative and lower in direct speech. Although the relative
sizes of the subject-preposed residues are comparable for the two discourse
registers, most of the clauses in the narrative residue have a function relating
to the clause as a whole, while most of the residue of direct speech lacks a
specific explanation. A new function of subject-preposing in direct speech was
identified: the marking of the justification for a preceding request or rhetorical
question.
The impression that word order in direct speech is “freer” than in narrative
can be explained as the result of a combination of two factors. First, as noted
in §5.6 (table 3, p. 87), preposing is more frequent in direct speech than in nar-
rative. Second, the proportion of preposed clauses with no clearly identifiable
pragmatic function or other explanation is higher in direct speech. While in
narrative only 12.3% of the preposed clauses have no clear function, in direct
speech the percentage is 42.9%. Put another way, 83.6% of the unexplained
preposed clauses are in direct speech.
In the following chapters, I discuss focused clauses and topicalized clauses
in more detail, examining them from a syntactic and pragmatic perspective.

Chapter 8
The Focused Clause
This chapter examines BH preposed clauses with a focusing function

(henceforth referred to as focused clauses). The analysis is based on the 50
focused clauses in Genesis and selected clauses from elsewhere in the classical
prose BH corpus. In §8.1, I describe the preposed constituent in the focused
clause. In §8.2, I examine the activated proposition to which the focused clause
relates. In §8.3, I set out a taxonomy of focused clauses based on how the
clause is to be interpreted with respect to the relevant activated information.
In §8.4, I discuss a special type of focusing known as focus of negation. In
§8.5, I discuss the relevance of focusing to a particular stylistic device. Conclu-
sions are presented in §8.6. In the citations in this chapter, focused clauses are
marked by underlining and focuses by bold type.
8.1. Syntactic description of the preposed constituent

The preposed constituent in the focused clause may be a subject, comple-
ment, or adjunct:
(175) Subject
:‫בלעדי אלהים יענה את שלום פרעה‬
Not I. God will give Pharaoh a favorable answer. (Gen 41:16)
(176) Complement
‫ את צאנם ואת בקרם‬:‫בני יעקב באו על החללים ויבזו העיר אשר טמאו אחותם‬
:‫ואת חמריהם ואת אשר בעיר ואת אשר בשדה לקחו‬
And the sons of Jacob came upon the slain, and plundered the city, be-
cause they had defiled their sister. Their flocks and their herds and
their asses, and whatever was in the city and whatever was in the
field they took. (Gen 34:27–28)
(177) Adjunct
:‫כל רמש אשר הוא חי לכם יהיה לאכלה כירק עשב נתתי לכם את כל‬
Every moving thing that lives, for you shall be for food. Like the green
grasses, I give you everything. (Gen 9:3)
121

8.1. Syntactic description of the preposed constituent 122
Table 13. Syntactic function of the preposed constituent in

focused clauses in Genesis.
Function Number Percentage

Subject 8 16
Complement/adjunct 42 84
Total 50
Table 14. Syntactic category of the preposed constituent

in focused clauses in Genesis.
Category Number Percentage

Noun phrase 29 58
Prepositional phrase 21 42
Total 50
Focusing is primarily associated with complement/adjunct preposing. Table 13

shows that 84% of focused clauses have a preposed complement or adjunct.
Table 14 shows that 58% of focused clauses have a preposed noun phrase, and
42% have a preposed prepositional phrase. Other syntactic categories, such as
adverbs or adjectives, do not occur as preposed constituents in focused clauses
in the Genesis corpus.
Although the preposed constituent in a focused clause is commonly referred
to as the focus of the clause, this is not entirely accurate. Because preposing is
normally performed on entire clause-level constituents, material that belongs
to the activated proposition is often preposed along with the focus. This phe-
nomenon is particularly common when the preposed item is a prepositional
phrase. In these cases, it is usually the noun phrase governed by the preposition
that is the focus:
(178) Gen 19:2

‫ויאמר הנה נא אדני סורו נא אל בית עבדכם ולינו ורחצו רגליכם והשכמתם וה־‬
:‫לכתם לדרככם ויאמרו לא כי ברחוב נלין‬
And he said, “My lords, please turn aside to your servant’s house and
spend the night, and wash your feet. And [= then] you may rise early
and go on your way.” And they said, “No. In the square we will spend
the night.”
The preposed constituent is ‫‘ ברחוב‬in the street’, but the focus is ‫‘ הרחוב‬the
street’. The focused clause relates to the proposition “We will spend the night
in x,” activated by Laban’s invitation, ‫סורו נא אל בית עבדכם ולינו‬. The clause

123 The activated proposition 8.2.
asserts “x = the street.” The preposition ‫ ב‬is necessarily preposed along with its
noun phrase, despite being extraneous to the focus.
In the focused clauses cited in this chapter, the preposed element is bolded
in its entirety, whether or not it is identical with the focus.
8.2. The activated proposition

As noted in §6.1.4, focusing in BH relates the focus to an activated proposi-
tion. In §8.2.1, I describe the formal relation between the focused clause and
the relevant activated proposition. In §8.2.2, I discuss the means by which the
activated proposition acquires this cognitive status. In §8.2.3, I discuss the is-
sue of activation versus pragmatic presupposition in the focused clause.
8.2.1. The relation between focused clause and activated proposition

In focused clauses that express statements, the relevant activated proposi-
tion consists of the proposition expressed by the clause, with a variable “x”
substituted for the focus. In (179), for example, the activated proposition is
“You did x to me,” with a variable substituted for the complement ‫מעשים אשר‬
‫לא יעשו‬.
(179) Gen 20:9
‫מה עשית לנו ומה חטאתי לך כי הבאת עלי ועל ממלכתי חטאה גדלה מעשים‬
:‫אשר לא יעשו עשית עמדי‬
What have you done to us, and how have I sinned against you, that you
have brought such great guilt on me and my kingdom? Things that
ought not to be done you have done to me.
When the focused clause is a question, request, or command, its relation to
the activated proposition is more indirect, because these illocutionary types do
not themselves express propositions. The activated proposition in these types
of clause is the one expressed by the statement most resembling the clause:
(180) Gen 23:3–6
‫ גר ותושב אנכי עמכם תנו‬:‫ויקם אברהם מעל פני מתו וידבר אל בני חת לאמר‬
:‫ ויענו בני חת את אברהם לאמר לו‬:‫לי אחזת קבר עמכם ואקברה מתי מלפני‬
‫שמענו אדני נשיא אלהים אתה בתוכנו במבחר קברינו קבר את מתך‬
And Abraham rose up from before his dead, and spoke to the Hittites,
saying, “I am a resident alien among you; give me a burying place
among you, and [= so that] I may bury my dead from before me.” And
the Hittites answered Abraham, saying to him, “Hear us, my lord; you
are a prince of God among us. In the choicest of our burial places bury
your dead.”
The activated proposition associated with the underlined focused clause is
“You will bury your dead in x.”

8.2. The activated proposition 124
8.2.2. The source of the activated proposition

The given information in BH focusing relates to an activated proposition,
that is, a proposition assumed to be in short-term or long-term working mem-
ory (see §§6.1.2–6.1.3). Activated information for the reader of a written text
includes propositions derived directly from the previous clause; elements in the
mental representation of the text that are directly linked to the previous clause
may be activated as well. The mental representation incorporates inferences
based on the reader’s knowledge and experience; thus information not explic-
itly referred to in the text may nonetheless constitute activated information.
In (181), the activated proposition “I will require the life of man from x ” is
derived directly from the previous clause, with the addition of an unspecified
argument:
(181) Gen 9:5

‫ואך את דמכם לנפשתיכם אדרש מיד כל חיה אדרשנו‬
But for your own life-blood I will require a reckoning. Of every beast
I will require it
The first clause, ‫ואך את דמכם לנפשתיכם אדרש‬, activates the proposition “I will
require your life-blood of x ”; the source argument is not present in the surface
structure of the clause. The focused clause supplies the value “every beast”
for x.
In (182), the activated proposition “He will father x ” is derived by entail-
ment from the preceding clauses. 1
(182) Gen 17:20
‫הנה ברכתי אתו והפריתי אתו והרביתי אתו במאד מאד שנים עשר נשיאם‬
‫יוליד‬
Behold, I bless him and I will make him fruitful and exceedingly numer-
ous. Twelve princes he will father
In (183), the activated proposition is inferred from the preceding clause:
(183) Gen 24:37–38

:‫וישבעני אדני לאמר לא תקח אשה לבני מבנות הכנעני אשר אנכי ישב בארצו‬
:‫אם לא אל בית אבי תלך ואל משפחתי ולקחת אשה לבני‬
And my master made me swear, saying, “You shall not take a wife for
my son from the daughters of the Canaanites, in whose land I dwell. But
to my father’s house and to my family you shall go, and take a wife
for my son.”
1. See also, e.g., Gen 35:11.

125 The activated proposition 8.2.
Abraham’s command to the servant not to take a wife for Isaac from the Ca-
naanites implies that he must go elsewhere to find a wife for Isaac; that is, “You
will go to x to take a wife for Isaac.” The focused clause identifies “my father’s
house” as the value for x.
In spoken discourse, activated information includes textually derived in-
formation in the mental representation of the preceding dialogue, as well as
extra-textual information pertaining to the speaker’s and addressee’s personal
details, their motivations, and their perceptions of the environment. Direct
speech in the Bible may be treated like spoken discourse for the purposes of
analyzing focusing. Although direct speech in the Bible is not an exact ren-
dition of spoken discourse, as discussed in §§1.2 and 4.4.3, it nevertheless
represents the quoted citation from the perspectives of the quoted speaker and
addressee. Focusing in direct speech relates to information activated for the
quoted addressee, rather than for the reader of the citation. Either textual or
extratextual sources of information, therefore, may be relevant in analyzing
direct-speech focusing.
An example of extra-textual activation is shown in (184):
(184) Gen 30:16

‫ויבא יעקב מן השדה בערב ותצא לאה לקראתו ותאמר אלי תבוא כי שכר שכ־‬
‫רתיך בדודאי בני‬
And Jacob came from the field in the evening, and Leah went out to meet
him, and she said, “To me you will come, for I have hired you with my
son’s mandrakes.”
Because Leah knows that Jacob is heading for Rachel’s tent, she takes the
proposition “Jacob will go to x ” as information activated by the extratextual
situation. Her statement, ‫אלי תבוא‬, replaces “Rachel,” the value for x activated
for Jacob, with a new value, “Leah.” 2
When a person comes to see someone else, the proposition “I have come” is
treated as activated extratextual information. An adjunct describing the purpose
of the visit can be marked as a focus in relation to this activated proposition: 3
(185) Gen 47:2–4

‫ ויאמר פרעה אל אחיו מה‬:‫ומקצה אחיו לקח חמשה אנשים ויצגם לפני פרעה‬
‫ ויאמרו אל‬:‫מעשיכם ויאמרו אל פרעה רעה צאן עבדיך גם אנחנו גם אבותינו‬
‫פרעה לגור בארץ באנו‬
And from among his brothers he took five men and presented them to
Pharaoh. And Pharaoh said to his brothers, “What is your occupation?”
2. This clause is of the substitutional type, where the focus replaces a previously acti-
vated value for x (see §8.3.3).
3. An additional example of this type is 1 Sam 16:5.

8.2. The activated proposition 126
And they said to Pharaoh, “Your servants are shepherds, as our fathers
also were.” And they said to Pharaoh, “To sojourn in the land we have
come.”
(186) Judg 15:11–12
‫וירדו שלשת אלפים איש מיהודה אל סעיף סלע עיטם ויאמרו לשמשון הלא‬
‫ידעת כי משלים בנו פלשתים ומה זאת עשית לנו ויאמר להם כאשר עשו לי כן‬
‫ ויאמרו לו לאסרך ירדנו לתתך ביד פלשתים‬:‫עשיתי להם‬
And three thousand men of Judah went down to the cleft of the rock
of Etam, and said to Samson, “Surely you know that the Philistines are
rulers over us. What have you done to us?” And he said to them, “As
they did to me, so have I done to them.” And they said to him, “To bind
you we have come down, so that we may give you into the hands of the
Philistines.”
8.2.3. Activation versus pragmatic presupposition in the focused clause

Propositions that are activated are generally presupposed as well, as (187)
illustrates:
(187) Judg 1:1–2
‫ ויאמר‬:‫ לאמר מי יעלה לנו אל הכנעני בתחלה להלחם בו‬′‫וישאלו בני ישראל בי‬
‫ יהודה יעלה‬′‫י‬
And the Israelites inquired of the LORD, saying “Who shall go up first
for us against the Canaanites to fight against them?” And the LORD said,
“Judah shall go up.”
The question ‫ מי יעלה לנו‬both presupposes and activates the proposition “x will
go up with us first against the Canaanites.” That is, the question reflects the
Israelites’ assumption that both they and God believe the truth of this proposi-
tion. By uttering the question, the people activate this proposition for the ad-
dressee, God. God answers with a focused clause supplying a value for the x :
‫‘ יהודה יעלה‬Judah shall go up [first against the Canaanites]’.
Despite the general association between presupposition and activation, pre-
supposition is not a necessary condition for focusing, as shown by the follow-
ing examples:
(188) Gen 23:3–6 4

‫ גר ותושב אנכי עמכם תנו‬:‫ויקם אברהם מעל פני מתו וידבר אל בני חת לאמר‬
:‫ ויענו בני חת את אברהם לאמר לו‬:‫לי אחזת קבר עמכם ואקברה מתי מלפני‬
‫שמענו אדני נשיא אלהים אתה בתוכנו במבחר קברינו קבר את מתך‬
And Abraham rose up from before his dead, and spoke to the Hittites,
saying, “I am a resident alien among you; give me a burying place site
among you, and [= so that] I may bury my dead from before me.” And
4. Appears above as (180) (p. 123).

127 A taxonomy of focused clauses 8.3.
the Hittites answered Abraham, saying to him, “Hear us, my lord; you
are a prince of God among us. In the choicest of our burial places bury
your dead.”
(189) Gen 43:11–13

‫ויאמר אלהם ישראל אביהם אם כן אפוא זאת עשו קחו מזמרת הארץ בכליכם‬
‫ וכסף‬:‫והורידו לאיש מנחה מעט צרי ומעט דבש נכאת ולט בטנים ושקדים‬
‫משנה קחו בידכם ואת הכסף המושב בפי אמתחתיכם תשיבו בידכם אולי משגה‬
:‫ ואת אחיכם קחו וקומו שובו אל האיש‬:‫הוא‬
And their father Israel said to them, “If it is so, then do this: take some
of the choice fruits of the land in your bags, and carry down to the man
a present, a little balm and a little honey, gum, resin, pistachio nuts, and
almonds. And double the money take with you; and the money that was
returned in the mouth of your sack bring back with you; perhaps it was
an oversight. And your brother take, and go again to the man.
In (188), the focused clause ‫ במבחר קברינו קבר את מתך‬relates to Abraham’s

earlier request to the children of Het to give him a burial place. Abraham’s
request activates the proposition “Abram will bury his dead in x,” but does not
presuppose it: in fact, Abram implies that he will have no burial place for Sarah
if the children of Het do not accede to his request. In (189), the focused clause
relates to the proposition activated by Jacob’s previous command, i.e., “You
will take x.” The value for x supplied in the focused clause is added to the items
Jacob has already specified (on additive focus, see §8.3.4.) Again, the proposi-
tion “You will take x ” is not presupposed. In general, focusing in commands
and requests involves activation but not presupposition. 5
8.3. A taxonomy of focused clauses

The focus in a focused clause supplies a value for a variable x in the acti-
vated proposition; in other words, a focused clause containing a focus f asserts
the identifying proposition “x = f ” with respect to an activated proposition p. In
some cases, a value for x (let us call it v ) is already activated for the addressee
at the time that he processes the clause. In this case, an identifying proposition
of the form “x = v ” is activated for the addressee. In §§8.3.1–8.3.4, I propose
a taxonomy of focused clauses based on the relation between the new proposi-
tion “x = f ” and the proposition “x = v,” if a proposition of the latter type is
activated. Four types of focusing are distinguished: identificational, descrip-
tive, substitutive, and additive. 6
5. For other examples of focusing in commands or requests, see, e.g., Gen 20:15, 44:2,
43:12, and 47:6.
6. A different taxonomy is found in Dik (1989: 282), where six types are distin-
guished: completive, parallel, replacing, expanding, restricting, and selecting; see also Dik
et al. (1981). This taxonomy is followed in Rosenbaum (1997) and Heimerdinger (1999).

8.3. A taxonomy of focused clauses 128
8.3.1. Identificational focusing

In identificational focusing, the value of x is unknown to the addressee at
the time of clause processing, that is, there is no currently activated proposition
“x = v.” Identification is the most common type of focusing, constituting about
three-fifths of the focused clauses in Genesis. 7 The classic case of identifica-
tional focus, as cited in the linguistic literature, is the answer to a previously
posed question. An example from BH is (187) above (p. 126). 8 This type of
focusing is quite rare in BH. The only remotely similar example in Genesis is
(190):
(190) Gen 20:9 9

‫מה עשית לנו ומה חטאתי לך כי הבאת עלי ועל ממלכתי חטאה גדלה מעשים‬
‫אשר לא יעשו עשית עמדי‬:
What have you done to us? And how have I sinned against you, that
you have brought such great guilt on me and my kingdom? Things that
ought not to be done you have done to me.
The focused clause echoes the rhetorical question posed just before by the
same speaker, ‫‘ מה עשית לנו‬What have you done to us?’ Although on the sur-
face this looks like identificational focus, an interpretation such as this misses
the implication of the rhetorical question. 10 In contrast to the genuine question,
an RQ is not a request for information but an implicit assertion (Ilie 1994: 38,
45; Schmidt-Radefeldt (1977: 376–77). The assertion implied by Abimelek’s
question is “You have done a terrible thing.” 11 Because the focus, “things that
Completive focus corresponds to my identificational focus, replacing focus is the same as

substitutional focus, and expanding focus is the equivalent of my additive focus. Restricting
focus, in which a compound value for x is reduced to a simple one, and selecting focus, in
which one of a number of alternative values for x is selected, are not recognized as separate
categories in my scheme. The former is included in the substitutive category and the latter
in the identificational category. Parallel focus refers to contrasted items in a pair of clauses,
as in John and Bill came to see me. John was nice, but Bill was rather boring. John and Bill
are not informational focuses, according to the definition of focusing set out in §§3.3.1.3 and
6.1; thus, there is no need here for the parallel category. As mentioned in §3.3.5, Dik defines
focus in terms of importance rather than in terms of new and given information.
7. Representative examples of identificational focused clauses from the classical BH
prose corpus include Num 9:3; Deut 24:19, 20, 21; Judg 1:2, 2:29; 1 Sam 1:27.
8. See also Judg 6:29.
9. Appears as (179) above (p. 123).
10. I am indebted to Yochanan Breuer for pointing this out.
11. Question-word rhetorical questions frequently imply that the answer is the null set;
e.g., “Who could pass that test” implies “Nobody could pass that test” (Pope 1972: 59). Al-
though the question ‫“ מה עשית‬what have I done” usually has the interpretation “I have done
nothing” in the Bible (see, e.g., Num 22:28; 1 Sam 20:1, 26:18, 29:8), the question “What
have you done” usually implies “You have done a terrible thing” (see e.g., Gen 3:13, 4:10,
12:18, 26:10, 29:25, 31:26).

ought not to be done,” is merely a reformulation of the value for x that has
already been activated (“a terrible thing”), this is actually descriptive focus
(see §8.3.2).
More commonly, identificational focusing answers a question that could ap-
propriately be asked by the hearer/reader, not one that has actually been asked: 12
(191) Gen 34:27–28 13

‫ את צאנם ואת בקרם‬:‫בני יעקב באו על החללים ויבזו העיר אשר טמאו אחותם‬
:‫ואת חמריהם ואת אשר בעיר ואת אשר בשדה לקחו‬
And the sons of Jacob came upon the slain, and plundered the city, be-
cause their sister had been defiled. Their flocks and their herds, their
asses, and whatever was in the city and in the field they took.
(192) Gen 9:5 14

‫ואך את דמכם לנפשתיכם אדרש מיד כל חיה אדרשנו‬
But for your life-blood I will surely require a reckoning. Of every beast
I will require it.
In (191), the previous clause ‫‘ ויבזו העיר‬they plundered the city’ entails the
proposition “they took x.” This proposition can appropriately be the subject
of a question: “What did they take?” The focused clause supplies the answer:
. . . ‫‘ את צאנם ואת בקרם‬Their flocks, their cattle’, and so on. Similarly, in (192),
the reader may ask subsequent to reading the first clause, “Of whom will God
require the blood?” The preposed clause indicates the answer: ‫‘ מכל חיה‬Of
every beast’. 15
When the focus in an identifying clause is an adjunct, the focus supplies
a previously unspecified detail regarding the circumstances of the event de-
scribed in an activated proposition. Consider (193) and (194): 16
(193) Gen 23:11

‫השדה נתתי לך והמערה אשר בו לך נתתיה לעיני בני עמי נתתיה לך‬
I give you the field, and I give you the cave that is in it. In the presence
of my people I give it to you.
12. See also, e.g., Gen 17:20 (‫ )שנים עשר נשיאים‬and 35:11.
15. An additional value for x is supplied in the continuation of the verse; the latter clause
is an example of additive focus (see §8.3.4.)
16. See also, e.g., Gen 7:20, 9:3 (‫)כירק עשב‬, 20:5 (‫)בתם לבבי‬, and 41:11. In the latter
verse, the preposed element is an adjunct phrase containing the word ‫איש‬: ‫איש כפתרון חלמו‬
‫‘ חלמנו‬Each man according to the interpretation of his dream we dreamed.’ The word
‫ איש‬cannot be the subject of the clause, as is evident from the mismatch in person and num-
ber between ‫ איש‬and the verb (additional examples are Gen 41:12 and 49:28).

(194) Gen 17:10–12

:‫זאת בריתי אשר תשמרו ביני וביניכם ובין זרעך אחריך המול לכם כל זכר‬
‫ ובן שמנת ימים‬:‫ונמלתם את בשר ערלתכם והיה לאות ברית ביני וביניכם‬
‫ימול לכם כל זכר לדרתיכם יליד בית ומקנת כסף מכל בן נכר אשר לא מזרעך‬
:‫הוא‬
This is my covenant, which you shall keep, between me and you and
your offspring after you: every male among you shall be circumcised.
You shall circumcise the flesh of your foreskins, and it shall be a sign of
the covenant between me and you. At the age of eight days every male
among you shall be circumcised throughout your generations, whether
the slave born in your house, or bought with your money from any for-
eigner who is not of your offspring.
In (193), the clause preceding the focused one activates the proposition “I give
the cave to you.” Although this proposition does not appear to contain a vari-
able x corresponding to the adjunct “in the presence of my people,” a variable
is present in covert form. Every proposition admits an unspecified number of
adjuncts regarding the circumstances of the event (e.g., time, place, manner, and
cause). The proposition “I give the field to you” is thus more fully represented
as “I give the field to you at time x, in place y, in manner z, for reason a,” and
so on. The phrase ‫‘ לעיני בני עמי‬In the eyes of my people’ supplies the value for
an implicit adjunct variable in the activated proposition. In (194), similarly, the
focus ‫‘ בן שמנת ימים‬at eight days old’ supplies the value for an implicit adjunct
in the proposition activated by v. 10, “Every male will be circumcised.”
8.3.2. Descriptive focusing

Descriptive focusing presents an alternate or more explicit description for
an already identified variable. In descriptive focusing, a proposition of the form
“x = v ” is activated for the hearer/reader at the time of processing the clause,
such that v and the new value, f, are referentially identical. In other words, the
focused clause asserts the proposition “x = f,” where “f = v.” Clauses of the
descriptive-focus type do not answer a potential question, because the hearer/
reader already has an activated value for x.
Descriptive focusing appears to be rare in BH. The Genesis corpus contains
only three examples, two of which are shown in (195) and (196). The third is
(190), p. 128 above. The expressions representing v and f are bold.
(195) Gen 6:17

‫ואני הנני מביא את המבול מים על הארץ לשחת כל בשר אשר בו רוח חיים‬
:‫מתחת השמים כל אשר בארץ יגוע‬
And I, behold, I will bring a flood—waters upon the earth—to destroy
all flesh in which is breath of life from under the sky. Everything
that is on the earth shall perish.

(196) Gen 7:21–22

‫ויגוע כל בשר הרמש על הארץ בעוף ובבהמה ובחיה ובכל השרץ השרץ‬
‫ כל אשר נשמת רוח חיים באפיו מכל אשר בחרבה‬:‫על הארץ וכל האדם‬
:‫מתו‬
And all flesh that crawls upon the earth, of the birds, and of the
cattle, and of the beasts, and of all the swarming creatures that
swarm upon the earth, and every human being, perished. Everything
in whose nostrils was the breath of life, from all that are on the dry
land, died.
In (195), the proposition “x will die” is an activated proposition, inferred from

the first clause. A value for x is already supplied by this first clause: “all flesh
in which is breath of life.” The focus of the preposed clause, ‫כל אשר בארץ‬
‘everything that is on the earth’, denotes the same set of referents as the former
value but is a new way of describing this set. The focus in (196), ‫כל אשר נשמת‬
‫‘ רוח חיים באפיו‬everything in whose nostrils was the breath of life’, bears a
similar relation to the expression in the previous verse, ‘all flesh that crawls
upon the earth’.
8.3.3. Substitutional focusing

Substitutional focus replaces an old value for x with a new one. In this type
of focus, a value v has previously been activated for x, and the new value, f, is
not referentially identical to v. Substitutional focus carries with it the implica-
tion that the previous value for x is to be rejected; that is, the assertion “x = f ”
implies the negation of the activated proposition “x = v.”
In many cases of substitutional focus, the proposition “x = v ” has been ne-
gated by the speaker immediately prior to the utterance of the focused clause. 17
The substitutional focused clause in this environment answers a potential ques-
tion that could be asked: if x is not v, then what is x? Focused clauses that
replace previously negated values for x are frequently introduced by one of the
conjuncts ‫כי אם‬, ‫כי‬, or ‫אם לא‬, each in the sense ‘rather’. The adverb signifies
that the new value is to be substituted for the previously rejected one. Exam-
ples are shown in (197) and (198), with v and f marked in bold. 18
17. The negation of the proposition “x = v ” does not remove this proposition from the
hearer’s consciousness, because the negated proposition remains in short-term memory; it
does, of course, strip it of presupposed status if it previously had this status.
18. Representative examples with ‫ כי‬from the classical BH prose corpus include 1 Sam
8:7 and 10:19. Representative examples with ‫ כי אם‬are Num 10:30 and 2 Kgs 17:40. This
phrase ‫ אם לא‬in this sense occurs in Gen 24:38. On the elliptical ‫ כי‬clause, see §8.4.

(197) Gen 19:2 19

‫ויאמר הנה נא אדני סורו נא אל בית עבדכם ולינו ורחצו רגליכם והשכמתם‬
:‫והלכתם לדרככם ויאמרו לא כי ברחוב נלין‬
And he said, “My lords, please turn aside to your servant’s house and
spend the night, and wash your feet. And you may rise early and go on
your way.” They said, “No. In the square we will spend the night.”
(198) Gen 35:10

‫ויאמר לו אלהים שמך יעקב לא יקרא שמך עוד יעקב כי אם ישראל יהיה שמך‬
And God said to him, “Your name Jacob, your name will no longer be
called Jacob. Rather, Israel will be your name.”
Lot’s request in (197) activates the proposition “You will spend the night at
x” and, along with it, the proposition “x = my house.” The three men reject
this value for x with their response, ‫‘ לא‬No’, and then specify “the street” as
the substitute. In (198), God rejects the value “Jacob” for x in the proposition
“Your name will be x.” “Israel” is then substituted for the rejected value.
In other cases, the focused clause follows immediately after the explicit or
implicit rejection of the old value, with no introductory conjunct: 20
(199) Gen 41:16 21

:‫ויען יוסף את פרעה לאמר בלעדי אלהים יענה את שלום פרעה‬
Joseph answered Phraoh, saying, Not I. “God will give Pharaoh a favor-
able answer.”
(200) Gen 42:7–9

‫ ויכר יוסף את אחיו‬:‫ויאמר אלהם מאין באתם ויאמרו מארץ כנען לשבר אכל‬
‫ ויזכר יוסף את החלמות אשר חלם להם ויאמר אלהם מרגלים‬:‫והם לא הכרהו‬
:‫אתם לראות את ערות הארץ באתם‬
And he said to them, “Where do you come from?” And they said, “From
the land of Canaan, to buy food.” And Joseph recognized his brothers,
and they did not recognize him. And Joseph remembered the dreams
which he had dreamed of them; and he said to them, “You are spies. To
see the nakedness of the land you have come.”
In (199), the activated proposition is “x will give Pharaoh a favorable answer,”

with the associated proposition “x = Joseph.” Both propositions can be in-
ferred indirectly from Pharaoh’s statement in the previous verse, ‫ואני שמעתי‬
‫‘ עליך לאמר תשמע חלום לפתר אתו‬I have heard about you that you can interpret
20. See also, e.g., Judg 8:23.

dreams’. This statement serves as an indirect directive to Joseph to interpret

the dream and, presumably, to supply a favorable interpretation. Joseph first
negates the proposition “x = Joseph” with ‫‘ בלעדי‬Not I’, that is, “it is not I who
will give you a favorable interpretation." A focused clause follows, substitut-
ing “God” as the correct value for x. In (200), Joseph’s statement ‫מרגלים אתם‬
‘you are spies’ implicitly rejects the brothers’ earlier assertion that they came
with the purpose of buying food. In other words, the proposition “x = buy
food” is rejected with respect to the proposition “We have come in order to
x.” The focused clause substitutes a different value for x, “see the nakedness
of the land.”
In some cases of substitutional focus, the old value v has not been pre-
viously rejected, and the question What is x? is not appropriate. In example
(184), repeated here as (201), the focused clause is substitutional, despite the
fact that it is conversation-initial. As discussed above, the speaker substitutes
“Leah” for the previously activated value, “Rachel.”
(201) Gen 30:16

‫ויבא יעקב מן השדה בערב ותצא לאה לקראתו ותאמר אלי תבוא כי שכר שכ־‬
‫רתיך בדודאי בני‬
And Jacob came from the field in the evening, and Leah went out to meet
him, and she said, “To me you will come, for I have hired you with my
son’s mandrakes.”
8.3.4. Additive focusing

Additive focusing resembles substitutional focus in that there is a previ-
ously activated value for x that is not referentially identical to the focus f. Ad-
ditive focus, however, does not imply the negation of the proposition “x = v ”
but implies that f is to be added to v. The result is a new identifying proposition
in which the two values form a compound expression: “x = v + f.” Additive
focus does not answer a potential question, because the addressee has no way
of knowing that another value is to be added to v. An example of additive focus
is (202):
(202) Gen 35:12

‫ואת הארץ אשר נתתי לאברהם וליצחק לך אתננה ולזרעך אחריך אתן את‬
:‫הארץ‬
The land that I gave to Abraham and Isaac I give to you. And to your
offspring after you I give the land.
The first clause activates the proposition “I give the land to x ” and the as-
sociated proposition “x = you.” The focused clause adds a new value “your
offspring after you,” yielding the proposition “x = you and your offspring after

you.” Additive focusing differs from other types of focus in that it generally
begins with the conjunction ‫ ;ו‬22 the other types are almost always asyndetic.
Additional examples of additive focusing appear in (203)–(205). 23
(203) Gen 9:5 24

‫ואך את דמכם לנפשתיכם אדרש מיד כל חיה אדרשנו ומיד האדם מיד איש‬
:‫אחיו אדרש את נפש האדם‬
But for your life-blood I will require a reckoning. Of every beast I will
require it. And of man, of every man for that of his fellow man I will
require the life of man.
(204) Gen 8:18–19

‫ כל החיה כל הרמש וכל העוף כל‬:‫ויצא נח ובניו ואשתו ונשי בניו אתו‬
:‫רומש על הארץ למשפחתיהם יצאו מן התבה‬
And Noah and his sons and his wife and his sons’ wives with him
came out. Every beast, every creeping thing, and every bird, every-
thing that creeps upon the earth, by families came out of the ark.
(205) Gen 43:11–13 25

‫ויאמר אלהם ישראל אביהם אם כן אפוא זאת עשו קחו מזמרת הארץ בכליכם‬
‫ וכסף‬:‫והורידו לאיש מנחה מעט צרי ומעט דבש נכאת ולט בטנים ושקדים‬
‫ ואת אחיכם קחו‬. . . ‫משנה קחו בידכם‬
And their father Israel said to them, “If it is so, do this: take some of
the choice products of the land in your baggage, and carry them down
as a gift for the man—some balm and some honey, gum, resin, pista-
22. Gen 8:19 is an exception to this generalization.

23. Additional examples are Gen 17:27, 34:25–26, and 44:2. Gen 34:25–26 is an excep-
tional case of additive focusing, in that f is included in v rather than added to it: ‫ויהרגו כל זכר‬
‫‘ ואת חמור ואת שכם בנו הרגו לפי חרב‬And they killed all the males. And they killed Hamor
and his son Shechem by the sword’. Some examples of additive focus from the classical BH
prose corpus are Exod 36:10; Lev 26:29; Num 6:3; Deut 17:17; 1 Kgs 6:22.
24. My interpretation of this verse follows the cantillation marks, in accordance with
NJPSV. Some translations, such as RSV, consider ‫‘ ומיד האדם‬and of man’ to be appended to the
end of the previous (also focused) clause, yielding ‘Of every beast I will require it, and of
man’. The second focused clause would then be ‫‘ מיד איש אחיו אדרש את נפש האדם‬Of every
man for his fellow man I will require the life of a man’. According to this clause division,
the first focused clause relates to the proposition “I will require it [i.e., the life of man] of x,”
and asserts “x = every beast and man.” The second focused clause asserts “x = every man
for his fellow man.” This value is a subset of the set denoted by the previous focus and is a
different way of describing that set. The focus in this interpretation bears characteristics of
additive and descriptive focusing.
25. Appears on p. 127 above as (189).

chio nuts, and almonds. And double the money take with you. . . . And
your brother take.
The last clause, ‫ואת אחיכם קחו‬, adds their brother to the other items the broth-
ers are to take. The previous clause, ‫וכסף משנה קחו בידכם‬, is an additive-focus
clause as well, assuming that the expression ‫ קחו בידכם‬is not meant to be taken
literally and the money was to be placed in the brothers’ bags along with the
gifts.
Additive focusing is functionally similar to the focusing adverb ‫גם‬. For
example, the preposing construction in (202) may be replaced by or supple-
mented with ‫גם‬: 26
(206) Paraphrase of Gen 35:12 with ‫גם‬
‫ואת הארץ אשר נתתי לאברהם וליצחק לך אתננה ונתתי את הארץ גם לזרעך‬
:‫אחריך‬
The land that I gave to Abraham and Isaac, to you I give it. And I give
the land also to your offspring after you.
Additive focusing by preposing and focusing by ‫ גם‬are not pragmatically

identical: the two devices involve different kinds of given propositions. Focus-
ing by preposing, as we have seen, relates to an activated but not necessar-
ily presupposed proposition. The focusing adverb, in contrast, may relate to
a proposition that is presupposed but not activated. In (207), for example, the
relevant proposition was introduced much earlier in the discourse and is no
longer activated in the consciousness of the reader.
(207) Gen 27:30–31

‫ויהי כאשר כלה יצחק לברך את יעקב ויהי אך יצא יצא יעקב מאת פני יצחק‬
‫ ויעש גם הוא מטעמים‬:‫אביו ועשו אחיו בא מצידו‬
And it came to pass when Isaac finished blessing Jacob; and it came to
pass—Jacob had scarcely gone out from the presence of Isaac his father,
and Esau his brother came in from his hunting. And he also had prepared
delicacies.
The clause ‫ ויעש גם הוא מטעמים‬asserts “x = Esau” with respect to the presup-
posed proposition “x prepared delicacies,” with the implication that the value
“Esau” is added to an earlier value, “Rebekah.” The information that Rebekah
prepared delicacies is presented much earlier in the story, in v. 14. The rela-
tively large distance between the ‫ גם‬clause and the source of the given propo-
sition is apparently no barrier to the use of the focusing adverb. The use of
preposing to mark additive focus in this context, in contrast, would be inappro-
priate and unclear, because the presupposed proposition is no longer activated.
26. As discussed in §5.4.1, focus adverbs can also be combined with preposing.

8.4. Focus of negation 136
8.4. Focus of negation

There are two ways in which negation and focusing can operate together in
a clause, as illustrated in (208) and (209):
(208) A: Is there anyone who didn’t finish the assignment?

B: Mary didn’t finish it.
(209) A: Did anyone finish the assignment? 27

B: I don’t know. I know Mary didn’t finish it.
Mary is marked as the focus by accenting in (208) and (209), both of which
involve clause-level negation. Nevertheless, the two clauses have different
pragmatic interpretations. The focused clause in (208) relates to the activated
negative proposition, “x did not finish the assignment,” and asserts “x = Mary.”
In (209), the activated proposition is the affirmative “x finished the assign-
ment,” and the focused clause asserts “x is not Mary.” 28 The negation here is
external to the activated proposition. This type of focus is termed “focus of
negation.” 29
As seen above, ordinary focusing with negation and focus of negation are
not distinguishable by pitch accent in English; in both types, the focus is ac-
cented. The cleft construction, however, does distinguish the two types:
(210) A: Who did not finish the assignment?

B: It was Mary who didn’t finish it.
(211) A: Who finished the assignment?

B: I don’t know. I know it wasn’t Mary who finished it.
Example (210) is the cleft equivalent of the ordinary focusing structure with
negation in (208), and (211) is the cleft equivalent of the focus-of-negation
structure in (209). In the cleft versions, the given propositions must be prag-
matically presupposed and are not necessarily activated (although in these ex-
amples they are). The cleft in (211) would be inappropriate as an answer to the
question in (209), because there the proposition “x finished the assignment” is
not presupposed (i.e., the speaker does not know if anyone finished).
In BH, ordinary focusing with negation and focus of negation are repre-
sented by two distinct forms of preposing, in a manner reminiscent of English
clefting. An example of ordinary focusing with negation is the additive clause
in (212). 30
27. The example is adapted from Dryer (1996: 490).
28. Note that in both cases the proposition is activated but not presupposed, because the
speaker does not know whether there is any such x. See further the next paragraph.
29. Huddleston and Pullum (2002: 796). See also Jackendoff (1972: 255), where the
phenomenon is termed “association of negation with focus,” and Horn (1989: 515–16).
30. No examples of this type were found in the Genesis corpus.

137 Focus of negation 8.4.
(212) Num 6:3

‫מיין ושכר יזיר חמץ יין וחמץ שכר לא ישתה וכל משרת ענבים לא ישתה‬
From wine and intoxicants he shall separate himself. Vinegar made
from wine or intoxicants he shall not drink. And all juice of grapes he
shall not drink
The activated presupposition in this case is “He will not drink x.” The focused
clause adds “any juice of grapes” to the items already identified as values for x.
Focus of negation is rare in BH but unmistakable, due to its unique syntactic
structure. As opposed to all other instances of clause-level negation, where the
negative particle and the verb form an inseparable unit, the focus-of-negation
structure places the negative particle before the preposed focus. An example
is (213).
(213) Gen 45:5–8
‫ועתה אל תעצבו ואל יחר בעיניכם כי מכרתם אתי הנה כי למחיה שלחני אלהים‬
‫ כי זה שנתים הרעב בקרב הארץ ועוד חמש שנים אשר אין חריש‬:‫לפניכם‬
‫ וישלחני אלהים לפניכם לשום לכם שארית בארץ ולהחיות לכם לפליטה‬:‫וקציר‬
‫ ועתה לא אתם שלחתם אתי הנה כי האלהים‬:‫גדלה‬
And now do not be distressed, or angry with yourselves that you sold
me here, for it was to preserve life that God sent me before you. For it
is two years that the famine has been in the land; and there are yet five
years in which there will be neither plowing nor harvest. And God sent
me before you to preserve for you a remnant on earth, and to keep you
alive in a great deliverance. And now, it was not you who sent me here;
rather God [sent me here].
The proposition involved in the underlined focused clause is “x sent me here,”
activated by the previous clause, ‫‘ וישלחני אלהים לפניכם‬and God sent me be-
fore you’. By assigning responsibility for his exile to Egypt to God, Joseph
implicitly rejects the brothers’ culpability. The following focused clause ex-
plicitly specifies that the brothers are not ultimately responsible for what hap-
pened. To this end, Joseph utilizes a focus-of-negation clause that asserts “x is
not you” with respect to the affirmative proposition “x sent me here.” 31
31. In this analysis the focus-of-negation structure is understood as involving clause-
level negation, like the English accented-focus construction “I know Mary didn’t finish it”
(example 209, p. 136 above). An alternate analysis is presented in Snyman and Naudé (2003);
Snyman (2004). Utilizing the framework of the Minimalist Program, they argue that the BH
focus-of-negation construction involves “constituent negation” rather clause-level negation.
The concept of constituent negation is different from the word-level negation exhibited in,
e.g., ‫‘ הם קנאוני בלא אל‬They incensed me with non-gods’ (Deut 32:21). Thus, it is not argued
that Gen 45:8 should be rendered as ‘Non-you sent me here’ but rather as ‘Not-you sent me
here’. See also Waltke and O’Connor (1990: §39.3.2) on clausal versus “item” (i.e., constitu-
ent) negation. A variation of Snyman and Naudé’s analysis is proposed by the editors (see
n. 33 below).

8.4. Focus of negation 138
The focus-of-negation construction in BH is often assumed to be the equiv-

alent of the English focus-of-negation cleft, above, It is not x that y. This is
true, however, only when the relevant activated proposition happens to be pre-
supposed as well. In (213), the given proposition is both activated and presup-
posed, and a cleft translation is perfectly appropriate. In other instances of BH
focus of negation, however, the activated proposition is not presupposed. Here
the assertion “x is not f ” does not imply that there is a correct value for x ; the
implication of the focused clause may in fact be that there is no value at all for
which the activated proposition is true. An English cleft rendering is odd and/
or misleading for focused clauses of this type, as shown by (214).
(214) Num 16:28–29

‫ שלחני לעשות את כל המעשים האלה כי לא‬′‫ויאמר משה בזאת תדעון כי י‬
′‫ אם כמות כל האדם ימתון אלה ופקדת כל האדם יפקד עליהם לא י‬:‫מלבי‬
:‫שלחני‬
And Moses said, “By this you shall know that the LORD has sent me to do
all these things, [and] that it has not been of my own devising. If these
men die as all people do, and if their fate be the fate of all people, the
LORD did not send me.”
The focused clause in (214), ‫ שלחני‬′‫לא י‬, asserts “x is not the LORD” with
respect to the activated proposition “x sent me.” The activated proposition is
derived from Moses’ previous statement, ‫ שלחני‬′‫‘ בזאת תדעון כי י‬by this you
shall know that the Lord has sent me’. At first glance, the focused clause has
an appropriate cleft rendering: ‘It is not the LORD who sent me’. This cleft,
however, implies that someone other than God sent Moses. If God did not send
Moses, however, Moses must have fabricated his mission, as Moses implies in
the previous verse: ‫‘ כי לא מלבי‬it has not been of my own devising’. The intent
of the conditional in v. 29 is “If these men die a normal death then the LORD did
not send me, that is to say, no one sent me.” It is clear, then, that “x sent me” is
activated but not presupposed information. The cleft rendering inappropriately
converts the activated presupposition into a presupposed one and blocks the
intended implication that no one sent Moses.
An additional example of this type is (215).
(215) Num 16:15

‫ אל תפן אל מנחתם לא חמור אחד מהם נשאתי‬′‫ויחר למשה מאד ויאמר אל י‬
And Moses was very angry, and he said to the LORD, “Do not pay heed to
their offering. I have not taken one ass from them”
The focused clause ‫ לא חמור אחד מהם נשאתי‬asserts “x is not one ass” with
respect to the activated proposition “I took x from them.” 32 The inappropriate-
32. The previous context does not make it clear how the proposition “I took x ” has been
activated. Although this sort of clause would not be counted as focused according to the strict

139 Focus of negation 8.4.
ness of the cleft rendering here is obvious: “It is not a single ass that I took”
implies that the speaker is guilty of a different theft, whereas Moses’ actual
intent is “I did not take (even) a single ass—in fact, I didn’t take anything!” 33
Focus-of-negation clauses involving presupposition are often followed
by an elliptical clause beginning with the clausal adverb ‫‘ כי‬rather’ or ‫כי אם‬
‘rather’. 34 The second clause functions as a substitutional-focus clause, supply-
ing the correct value for x in the activated proposition:
(216) Deut 5:3

‫ את הברית הזאת כי אתנו אנחנו אלה פה היום כלנו‬′‫לא את אבתינו כרת י‬
‫חיים‬
It is not with our fathers that the LORD made this covenant; rather, with
us, we who are here today, all of us alive, [He made this covenant].
The focus-of-negation clause in (216) relates to the proposition “The LORD

made this covenant with x ,” activated by Moses’ utterance in the previous
verse: ‫ אלהינו כרת עמנו ברית בחרב‬′‫‘ י‬The LORD our God made a covenant with
us at Horeb’. The activated proposition is also presupposed, because the ad-
dressees presumably accept Moses’ assertion that a covenant took place. The
focus-of-negation clause asserts “x is not our fathers” and is followed by an
elliptical clause ‫‘ כי אתנו‬Rather, with us, we who are here today, all of us alive,
[He made this covenant]’. 35 The ‫ כי‬clause provides the correct value for x, “us.”
criteria I followed in classifying the Genesis corpus, it is included in the discussion here
for purposes of illustration. Because the ‫ לא‬+ preposed constituent structure has virtually
no other uses besides focus-of-negation, it is reasonable to assume that focus of negation is
involved in (215) as well. Furthermore, it is intuitively obvious that Moses intends the clause
to be focused. This is plausibly viewed as a case of accommodation (§7.2.1, pp. 108–109),
where in his agitation Moses treats the proposition as activated even though this is not neces-
sarily the case for his addressee.
33. The editors suggest that one can explain the difference between focus-of-negation
clauses that can be rendered by clefts and those that cannot by means of the clause-negation/
constituent-negation distinction. Examples such as (213) and (216) below, where the cleft
is an appropriate rendering, involve constituent negation, whereas examples that cannot be
rendered by clefts, such as (214) and (215), involve clause-level negation. In contrast, I am
claiming that the same level of negation is involved in both cases and that the difference
between the two groups of clauses is solely due to the type of givenness involved.
34. See also (213), above. For an example with ‫כי אם‬, see (217) below. The particles ‫כי‬
and ‫ כי אם‬seem largely interchangeable in this usage. Follingstad, however, distinguishes
between the two, claiming that ‫ כי אם‬has scope over a constituent and marks it for replace-
ment (i.e., substitutional) focus (2001: 563–66), whereas ‫ כי‬has scope over the entire clause,
marking it for “assertive” focus (2001: 569–83).
35. Although the (‫ כי )אם‬clause following a focus-of-negation clause usually undergoes
ellipsis, the full version is occasionally attested, e.g., Lev 21:14 (Richard Steiner, personal
communication). The editors argue that ‫ כי‬following a focus-of-negation clause does not
start a new clause. Rather, ‫ כי‬is the continuation of the focused constituent. In Gen 45:8
(example (213), p. 137 above), it is claimed, there is a discontinuous two-part subject, ‫לא‬
‫‘ אתם כי האלהים‬not you but God’. This interpretation is dependent on a constituent-negation

8.5. The stylistic use of focusing 140
Example (217) is a somewhat complex case of a focus-of-negation clause

followed by a ‫ כי אם‬clause:
(217) Gen 32:28–29

‫ ויאמר לא יעקב יאמר עוד שמך כי אם‬:‫ויאמר אליו מה שמך ויאמר יעקב‬
‫ישראל‬
And he said to him, “What is your name?” And he said, “Jacob.” And he
said to him, “Your name shall no longer be called Jacob; rather, [your
name will be called] Israel.”
In this example it seems that the adverb ‫ עוד‬is external to the activated proposi-
tion, along with the negative; in other words, the activated proposition is “Your
name will be x,” inferred from Jacob’s prior assertion that his name is Jacob,
and the clause asserts “x will no longer be Jacob.” The focused clause is fol-
lowed by an elliptical ‫ כי אם‬clause identifying the correct value for x, “Israel.”
8.5. The stylistic use of focusing

Focusing may be used to achieve a stylistic effect. An example is the “re-
peating verb” structure discussed by Blau (1972) and Paran (1989), among
others, in which a verb is repeated over the course of two or more adjacent
clauses. 36 The second half of the structure typically exhibits preposing and a
perfect verb form. Examples are shown in (218)–(220).
(218) Lev 23:11

‫ לרצנכם ממחרת השבת יניפנו הכהן‬′‫והניף את העמר לפני י‬
He shall elevate the sheaf before the LORD for acceptance in your behalf;
on the day after the sabbath the priest shall elevate it.
(219) Lev 4:12

:‫ושרף אתו על עצים באש על שפך הדשן ישרף‬
And he shall burn it on a fire of wood; on the ash heap it shall be burned.
analysis (see n. 31 above). If clause-level negation is involved, ‫ לא‬is not part of the subject;
and the two-part subject would have to be ‫אתם כי האלהים‬, which is impossible from both the
syntactic and the semantic perspective. Against the two-part subject analysis, as the editors
themselves note, is the fact that the verb in Gen 45:8 agrees with the negated ‫אתם‬, rather than
with ‫האלהים‬. A further argument is the occurrence of ‫ כי אם‬after clauses that clearly involve
clause-level negation and hence exclude the possibility of a two-part subject, e.g., ‫ויאמר לא‬
‫‘ עכרתי את ישראל כי אם אתה ובית אביך‬I did not bring trouble on Israel; rather you and you
father’s house [brought trouble on Israel]’ (1 Kgs 18:18; see also Num 26:33, 1 Sam 2:15,
1 Kgs 22:18). Because ‫ כי אם‬introduces an elliptical clause in these verses, it seems most prob-
able that ‫ כי אם‬following a focus-of-negation clause introduces an elliptical clause there too.
36. See §3.1.1 for additional references. As mentioned in §3.1.1, Paran terms this struc-
ture the “circular inclusio.” According to Paran (1989: 49), the circular inclusio is character-
istic of the Priestly style and appears more than 100 times in the Priestly source in legal and
cultic contexts alone. For further references, see Paran (1989: 49–50).

141 The stylistic use of focusing 8.5.
(220) Gen 41:12 37

‫ויפתר לנו את חלמתינו איש כחלמו פתר‬
And he interpreted our dreams to us; each man according to his dream
he interpreted
As Blau points out, in each cases the double clause structure can be condensed
into a single clause. Example (218) could have been expressed as ‫והניף את‬
‫ לרצנכם ממחרת השבת‬′‫‘ העמר לפני י‬he shall elevate the sheaf before the LORD
for acceptance in your behalf on the day after the Sabbath’, and (219) could
have been expressed as ‫‘ ויפתר לנו את חלומתינו איש כחלומו‬and he interpreted
our dreams to us, each man according to his dream’.
In the light of the above discussion of focusing, it can be seen that the pre-
posed clause in the repeating-verb structure is a focused clause. The second
sentence is preposed not simply to create an inclusio but in order to mark the
preposed element as the focus in relation to the proposition expressed by the
first clause. All three clauses involve identificational focus. The examples are
repeated in (221)–(223), with the focused clauses underlined and focuses bold.
(221) Lev 23:11

‫ לרצנכם ממחרת השבת יניפנו הכהן‬′‫והניף את העמר לפני י‬
(222) Gen 41:12
‫ויפתר לנו את חלמתינו איש כחלמו פתר‬
(223) Lev 4:12
:‫ושרף אתו על עצים באש על שפך הדשן ישרף‬
On the level of the clause, stylistic focusing is indistinguishable from ordinary

focusing. What makes these structures “stylistic” is that the writer has inten-
tionally set up the first clause as a lead-in to the focused clause, when the same
idea could have been expressed more economically by a single clause.
A particularly elaborate example of focusing for stylistic effect is (224),
cited in Blau (1972: 235): 38
(224) Gen 1:27
‫ויברא אלהים את האדם בצלמו בצלם אלהים ברא אתו זכר ונקבה ברא אתם‬:
And God created man in His image; in the image of God He created
him; male and female he created them.
37. Paran (1989: 78–79) views this example as exceptional in that it is not from the
Priestly source. He also views its style as different from the one usually employed in the
Priestly source.
38. Although this verse is often taken to be poetry, the appearance of a stylistic device
such as the repeated-verb structure does not necessarily justify the classification of the verse
as poetry, as discussed further on in this section.

8.5. The stylistic use of focusing 142
In this three-part structure the initial clause, ‫ויברא אלהים את האדם בצלמו‬, pro-
vides the activated proposition for the following focused clause, ‫בצם אלהים‬
‫ברא אתו‬. This is technically speaking descriptive focus, although the old value
v and the new value f are semantically identical. The two values differ only
with regard to the form of the expressions involved: where v contains an en-
clitic pronoun, f substitutes the lexical expression ‫אלהים‬. Next comes an iden-
tificational focus clause, ‫זכר ונקבה ברא אתם‬. The verb ‫‘ ברא‬created’ recurs in
each of the three clauses in the structure, all of which can be condensed into a
single clause, ‫‘ ויברא אלהים את האדם בצלמו זכר ונקבה‬And God created man in
His image, male and female’.
Although on the level of the clause there is nothing exceptional about stylis-
tic focusing, it is worth investigating why in these cases the speaker has chosen
to use a repeated-verb structure with focusing, rather than a simpler mode of
expression. It is plausible that in many cases the repeated-verb structure has
a specific pragmatic function. 39 Pragmatics operates on the discourse level as
well as the clause level, motivating the speaker both to express an idea as a
particular combination of clauses and to structure those clauses in particular
ways. A plausible hypothesis for some of the examples of stylistic focusing is
that they draw the listener’s attention to a constituent that might be otherwise
overlooked or insufficiently appreciated. Bendavid (1971: 857) states that the
repeated verb structure has the purpose of giving “weight to each piece of new
information individually”; a similar view is expressed in Paran (1989). 40 In
(223), for example, the identificational focus ‫ על שפך הדשן‬is brought to atten-
tion to stress that it is here and nowhere else that the offering is to be burnt.
The concept of attentional focus (see §3.3.1.3) may prove to be useful in un-
derstanding this type of function. In other cases, as in (224), focusing may be
motivated more by esthetics than by pragmatics.
A separate issue is whether “stylistic” focusing serves as an indication that
a segment of text is poetry rather than prose. Example (224), for example, is
considered poetry in many Bible translations, including NRSV, NAB, and NJB.
Although investigation of this issue is beyond the scope of this work, it is
worthwhile to note that “poetic” devices are widespread even in ordinary lan-
guage use. As Jakobsen (1967: 302) writes, “Any attempt to reduce the sphere
of poetic function to poetry or to confine poetry to poetic function would be a
39. Stylistic devices often have pragmatic functions, although in some cases the func-
tion is purely esthetic. Watson (1984: 32–34) states that poetic devices may have functions
related to structure or esthetics, among other functions. Structural functions are pragmatic,
whereas esthetic functions are not. According to Wimsatt (1967: 369 n. 31) the meaning of
stylistic devices may be “emotive,” function that may or may not be included in pragmatics.
40. Paran suggests that the device is also intended to facilitate memorization of the text
(1989: ix). According to Blau (1972: 237), the device reflects an oral style that was originally
used to express an afterthought. He also suggests that the preposed element serves as the
“psychological predicate” of the previous sentence.

143 Conclusion 8.6.
delusive oversimplification.” 41 Devices such as parallelism and repetition are

not exclusive to poetic or even elevated discourse, but can be found in every-
day language as well. 42 It should not be surprising, therefore, to find devices
such as stylistic focusing in prose texts.
8.6. Conclusion
The activated proposition relating to a focused clause may be derived from
the textual or extratextual context. Activation frequently goes hand-in-hand
with presupposition, but there are many examples of focused clauses in which
presupposition is not involved. Focusing in commands and requests generally
does not involve presupposition.
Focused clauses may be identificational, descriptive, substitutional, or ad-
ditive, depending on the relation the focus bears to previously activated values
for x. Substitutional clauses that replace previously negated values for x may
be introduced by the conjuncts ‫כי‬, ‫כי אם‬, or ‫אם לא‬. Additive focusing and the
focusing adverb ‫ גם‬are in some ways similar, but the two devices involve dif-
ferent types of given information. The focusing adverb relates to presupposed
but not necessarily activated information, whereas additive focusing by prepar-
ing relates to activated but not necessarily presupposed information.
Focus of negation is a special type of focusing with a distinct syntactic
form. Like ordinary focus, focus of negation also relates to an activated but
not necessarily presupposed proposition. The English cleft is an appropriate
rendering only for focus-of-negation constructions involving presupposition.
Focus-of-negation clauses involving presupposition may be followed by an
elliptical ‫ כי‬or ‫ כי אם‬clause supplying the correct value for x.
There are stylistic uses of focusing in which a repetitive structure is inten-
tionally set up containing a focused clause. The purpose of a structure of this
sort may be to set up the focused constituent as a focus of attention.
41. I am indebted to Michael O’Connor for bringing this reference to my attention. See
also Berlin (1985: 3–5).
42. According to Muilenberg (1953), repetition is a characteristic stylistic feature of
biblical narrative as well as poetry. See also Holes (1995), who shows that repetition and
parallelism can be found in the everyday speech of certain communities of nonliterate Arabic
speakers, and Johnstone (1991), who discusses the use of repetition and parallelism in differ-
ent types of nonpoetic Arabic discourse. See also the discussion in §1.2.

Chapter 9
The Topicalized Clause
In this chapter, I describe BH topicalization, based on the 171 occurrences

in Genesis and selected additional examples from the classical BH prose cor-
pus. Topicalization highlights a link between the preposed item and a second
item in an adjacent text segment, cueing the addressee to a coherence relation
between the linked segments. Because segments linked by topicalization are
most frequently single clauses, I refer to them as clauses, except when larger
segments are concerned. The item linked to the preposed item is termed the
“counterpart” of the preposed item.
In §9.1, I examine the preposed constituent and its counterpart from a syn-
tactic perspective. In §9.2, I examine the nature of the link between the pre-
posed item and counterpart. In §9.3, I examine the structure of the segments
linked by topicalization. In §9.4, I describe the coherence relations obtaining
between the linked segments. In the citations in this chapter, preposed clauses
are marked by underlining, and the preposed items and their counterparts are
bold.
9.1. Syntactic description of the preposed constituent

and its counterpart
The preposed constituent in the topicalized clause may be a subject, com-
plement, or adjunct:
(225) Subject
:‫ויקרא לו לבן יגר שהדותא ויעקב קרא לו גלעד‬
Laban called it Jegar-sahadutha; and Jacob called it Galeed. (Gen 31:47)
(226) Complement
:‫את בנתם נקח לנו לנשים ואת בנתינו נתן להם‬
We will take their daughters for ourselves as wives, and our daughters
we will give to them. (Gen 34:21)
144

145 Syntactic description of the preposed constituent 9.1.
Table 15. Syntactic function of the preposed constituent in

topicalized clauses in Genesis.
Function Number Percentage
Subject 76 44.4
Complement/adjunct 93 54.4
Unclear 2 1.2
Total 171
Table 16. Syntactic category of the preposed constituent in

topicalized clauses in Genesis
Category Number Percentage
Noun phrase 101 59.1
Prepositional phrase 68 39.8
Infinitive clause 1 0.6
Adverb 1 0.6
Total 171
(227) Adjunct
‫ויחפש בגדול החל ובקטן כלה‬
And he searched: with the eldest he began, and with the youngest he
ended. (Gen 44:12)
Table 15 shows the frequency of the various syntactic functions of the pre-
posed constituents in topicalized clauses from Genesis. Complement/adjunct-
preposed topicalization is somewhat more common than the subject-preposed
type. As shown in table 16, the preposed constituent is most frequently a noun
phrase but may also be a prepositional phrase or, more rarely, an infinitive
clause or adverb. About 60% of preposed constituents are noun phrases.
In contrast to the preposed item, the counterpart of the preposed item need
not be present in the surface structure of the text. A subject counterpart in (228)
and a complement counterpart in (229) have been deleted from the surface
structure. 1 The deleted items are indicated in brackets.
(228) Gen 35:18

:‫ותקרא ]רחל[ שמו בן אוני ואביו קרא לו בנימין‬
And [Rachel] called his name Ben-oni, and his father called his name
Benjamin.
1. See also, e.g., Gen 22:5 and 24:53.

9.2. The link between the preposed item and its counterpart 146
(229) Gen 20:15–16

‫ ולשרה אמר הנה‬:‫ויאמר אבימלך ]לאברהם[ הנה ארצי לפניך בטוב בעיניך שב‬
‫נתתי אלף כסף לאחיך‬
And Abimelech said [to Abraham], “Behold, my land is before you;
dwell where it pleases you.” And to Sarah he said, “Behold, I give your
brother a thousand pieces of silver”
In exceptional cases, the counterpart of the preposed item is not a phrase,

but an entire clause, or even a group of coordinated clauses. In (230), for ex-
ample, the word ‫זאת‬, used as a pro-clause, has as its counterpart the several
clauses following, marked in bold.2
(230) Gen 42:33–34
‫בזאת אדע כי כנים אתם אחיכם האחד הניחו אתי ואת רעבון בתיכם קחו‬
‫ והביאו את אחיכם הקטן אלי‬:‫ולכו‬
By this I shall know that you are honest men: one of your brothers
leave with me, and for the famine of your households take [some-
thing], and go. And bring your youngest brother to me.
9.2. The link between the preposed item and its counterpart
The link between the preposed item and its counterpart generally pertains
to the referents of the linked items, rather than to the expressions themselves. 3
Because preposing is normally performed on entire clause-level constituents,
the preposed constituent may contain material that is irrelevant to the link with
the counterpart. Prepositions are extraneous to the link:
(231) Gen 10:10–11

‫ מן הארץ ההוא‬:‫ותהי ראשית ממלכתו בבל וארך ואכד וכלנה בארץ שנער‬
‫יצא אשור‬
And the beginning of his kingdom were Babel, Erech, and Accad and
Calneh 4 in the land of Shinar. From that land Ashur went forth 5
2. The topicalized clause in (230) is forward-linking (see §9.3), with the preposed item
and its counterpart linked by a relation of identity. The counterpart itself contains two items
linked by topicalization: ‫‘ אחיכם האחד‬one of your brothers’ is linked to ‫את רעבון בתיכם‬
‘[something for] the famine of your households’. The former is to be left behind and the
latter taken back to Canaan. Additional examples of clauses or groups of clauses serving as
counterparts are Gen 42:15–16, 18–20; 43:11–13; 45:17–18.
3. In Gen 48:11, however, the link concerns the meaning of the preposed infinitive
clause ‫‘ ראה פניך‬seeing your face’ and the meaning of the following verse.
4. The NRSV translates ‘all of them’ instead of ‘Calneh’, emending the vocalization to
‫ ֻכּלָּנָה‬.
5. The NRSV translates ‘From that land he went into Assyria’, taking Ashur as the object
rather than the subject of the clause.

147 The link between the preposed item and its counterpart 9.2.
In (231), ‫‘ מן‬from’ is extraneous to the link of identity obtaining between the

referent of ‫ הארץ ההוא‬and the referent of the preceding ‫ארץ שנער‬. 6
A noun that is part of an idiomatic expression may also be extraneous to
the link:
(232) Gen 43:9

‫אנכי אערבנו מידי תבקשנו‬
I will be surety for him; from my hand you shall require him.
Preposed ‫ מידי‬in the second clause has ‫ אנכי‬in the preceding clause as its coun-
terpart. The word ‫ ידי‬is part of the idiomatic expression ‫‘ בקש מיד‬hold re-
sponsible’ and has no independent meaning and hence no reference. The link
obtains between the referents of ‫ אנכי‬and the coreferential enclitic pronoun in
the phrase ‫מידי‬. A more idiomatic rendering of the verse would be ‘I will be
surety for him; you may hold me responsible’, with the linked items marked
in bold.
An examination of topicalization in Genesis and elsewhere reveals a wide
variety in the links obtaining between the linked items. The common denomi-
nator in all cases is a contextual relation between the referents of the preposed
item and counterpart; that is, the referents function as a pair in the discourse
context. I refer to the referents of the linked items as A and B, where A is the
item found in the first segment and B is in the second segment.
In most instances of topicalization, A and B have been mentioned just be-
fore the linked segments in a way that establishes them as a pair. Examples are
shown in (233) and (234). 7
(233) Gen 1:4–5
‫ ויקרא‬:‫וירא אלהים את האור כי טוב ויבדל אלהים בין האור ובין החשך‬
‫אלהים לאור יום ולחשך קרא לילה‬
And God saw that the light was good; and God separated between the
light and the darkness. And God called the light Day, and the darkness
He called Night.
(234) Gen 29:16–17

‫ ועיני לאה רכות ורחל‬:‫וללבן שתי בנות שם הגדלה לאה ושם הקטנה רחל‬
:‫היתה יפת תאר ויפת מראה‬
And Laban had two daughters; the name of the older was Leah, and the
name of the younger was Rachel. And Leah’s eyes were weak, and Ra-
chel was beautiful in form and beautiful in appearance.
6. See also, e.g., Gen 9:19, 10:32, 12:16, and 41:44.

7. Additional representative examples include Gen 3:16; 4:2, 5; 12:16; 13:12; 31:47;
36:4; and 40:21–22.

9.2. The link between the preposed item and its counterpart 148
In (233), light and darkness are mentioned together in the preceding verse;
similarly, in (234) Rachel and Leah are named just before as Laban’s two
daughters. In these examples, as in many others, A and B bear a logical rela-
tion that automatically establishes a contextual link: light and darkness are
opposites, and Rachel and Leah are members of the set of Laban’s children.
In some cases, neither A nor B has been previously mentioned, but they nev-
ertheless bear a logical relation to each other. An example is (235), in which
the linked items, the pure animals and the impure animals, are mentioned for
this first time. Since these constitute subsets of the set of all animals, a link
between them is automatically established. 8
(235) Gen 7:2

‫מכל הבהמה הטהורה תקח לך שבעה שבעה איש ואשתו ומן הבהמה אשר‬
:‫לא טהרה הוא ]תקח לך[ שנים איש ואשתו‬
Of every clean animal you shall take for yourself seven pairs, a male
and its mate; and of every animal that is not clean [you shall take] two,
a male and its mate.
Logical relations are not necessary for linkage by topicalization. In many

cases, the link is ad hoc, restricted to the particular context: 9
(236) Gen 18:33
‫ כאשר כלה לדבר אל אברהם ואברהם שב למקמו‬′‫וילך י‬:
And the LORD departed when He had finished speaking to Abraham; and
Abraham returned to his place
Abraham and God are not a natural pair, but they are linked by the prior context
by virtue of being mutually engaged in conversation. Another ad-hoc pair is
shown in (237).
(237) Gen 47:19

‫למה נמות לעיניך גם אנחנו גם אדמתנו קנה אתנו ואת אדמתנו בלחם ונהיה‬
‫אנחנו ואדמתנו עבדים לפרעה ותן זרע ו]אנחנו[ נחיה ולא נמות והאדמה לא‬
:‫תשם‬
Why should we die before your eyes, both we and our land? Buy us and
our land in exchange for food, and we and our land will be slaves to
Pharaoh; and give us seed and (= that) [we] may live, and not die, and
the land may not become desolate.”
8. See also Gen 19:3, where the preposed ‫‘ מצות‬unleavened bread’ bears a part-whole
relation to ‫‘ משתה‬a feast’ in the previous clause.
9. The idea of the ad hoc link is discussed in Birner and Ward (1998: 234). Note that
these scholars use the term link somewhat differently to refer to the referent of the preposed
constituent rather than to a relation between that referent and another item (1998: 20).

149 The link between the preposed item and its counterpart 9.2.
The topicalized clause ‫ והאדמה לא תשם‬is linked to the previous two clauses,
‫ונחיה ולא נמות‬. The preposed element ‫ האדמה‬is linked to the implicit subject
pronoun in the preceding segment, ‫( אנחנו‬i.e., the people). A contextual link
between the land and the people was established previously in the verse, where
the people propose to sell themselves and their land to Pharaoh. The pair /‫אנחנו‬
‫ גויתנו‬and ‫ אדמתנו‬occur in coordinated or appositive phrases no less than three
times in the preceding text (shown above in bold), making the link impossible
to miss.
In some cases a previously nonexistent ad hoc relation is established by the
linked clauses themselves:
(238) Gen 27:15–16

[‫ותקח רבקה את בגדי עשו בנה הגדל החמדת אשר אתה בבית ותלבש ]אותם‬
:‫ ואת ערת גדיי העזים הלבישה על ידיו ועל חלקת צואריו‬:‫את יעקב בנה הקטן‬
Then Rebekah took the best clothes of Esau her older son, which were
with her in the house, and she put [them] on Jacob her younger son; and
the skins of the kids she put upon his hands and upon the smooth part
of his neck.
In this example, A is the referent of the omitted object ‫ אותם‬in the first under-
lined clause, that is, Esau’s clothes mentioned in the previous verse. B is the
goat skins. It is hard to think of a natural relation obtaining between Esau’s
clothes and the goat skins, outside of this particular context. Nevertheless, be-
cause the clauses referring to these items describe similar actions performed
with the two items—both are used to clothe Jacob—a link is established be-
tween the two. Other unnatural pairs of this sort include, among others: the
wood of a wagon and the cows pulling the wagon (1 Sam 6:14), the head of
the dead Philistine and his weapons (1 Sam 17:54), and a household idol and
a goat’s hair quilt (1 Sam 19:13). 10 The link between each of these pairs is
perfectly comprehensible in its context: the wood and the cows are both used
in preparing a sacrifice, the Philistine’s head and his weapons are retrieved
from the battlefield as war trophies, and the idol and quilt are used to simu-
late the form of David’s body lying under the covers. Idiosyncratic examples
like these make it clear that logical relations like partially ordered set relations
(see §6.2.1.1), even if involved in some instances of topicalization, are beside
the point.
A notable type of link is between two items with identical reference. The
first item is often a lexical expression and the second a coreferential expres-
sion containing a personal pronoun (independent or enclitic), as in (239), or a
demonstrative pronoun, as in (240). 11
10. The last example bears a striking relation to (238).

11. For additional examples, see, e.g., Gen 2:10; 10:11; 25:10; 36:13; and 42:15.

9.3. The linked segments 150
(239) Gen 4:20

:‫ותלד עדה את יבל הוא היה אבי ישב אהל ומקנה‬
And Adah bore Jabal; he was the father of those who dwell in tents and
amidst cattle.
(240) Gen 22:20–23

‫ את עוץ בכרו‬:‫ויגד לאברהם לאמר הנה ילדה מלכה גם הוא בנים לנחור אחיך‬
‫ ואת כשד ואת חזו ואת פלדש ואת‬:‫ואת בוז אחיו ואת קמואל אבי ארם‬
‫ ובתואל ילד את רבקה שמנה אלה ילדה מלכה לנחור אחי‬:‫ידלף ואת בתואל‬
:‫אברהם‬
And it was told to Abraham, saying, “Behold, Milcah also has born chil-
dren to your brother Nahor: Uz the first-born, and Buz his brother,
and Kemuel the father of Aram, and Chesed, and Hazo, and Pildash,
and Jidlaph, and Bethuel.” And Bethuel fathered Rebekah. These eight
Milcah bore to Nahor, Abraham’s brother.
The lexical expression may also be repeated in place of a pronoun, as in (241). 12

In many cases both items are pronouns, as in (242). 13
(241) Gen 36:1–3

‫ עשו לקח את נשיו מבנות כנען את עדה בת‬:‫ואלה תלדות עשו הוא אדום‬
‫ ואת בשמת בת ישמעאל‬:‫אילון החתי ואת אהליבמה בת ענה בת צבעון החוי‬
:‫אחות נביות‬
These are the descendents of Esau, that is, Edom. Esau took his wives
from the Canaanites: Adah the daughter of Elon the Hittite, Oholibamah
the daughter of Anah the daughter of Zibeon the Hivite, and Basemath,
the daughter of Ishmael, the sister of Nebaioth.
(242) Gen 9:19

:‫שלשה אלה בני נח ומאלה נפצה כל הארץ‬
These three were the sons of Noah; and from these the whole world
branched out.
9.3. The linked segments

In the simplest and most common situation, topicalization links two adja-
cent clauses. The second clause is generally topicalized, pointing back to the
previous one. This is termed backward-linking topicalization. The preceding
clause most often has normal word order: 14
12. Additional examples include Gen 1:27; 7:19; 8:5; 9:23; 11:27; 40:13, 19; and 41:50.
13. Additional examples include Gen 9:19, 10:32, 31:39, 36:14, 41:40, 43:9, and 46:4.
14. These are the “chiastic” structures discussed by Andersen and Khan (see §3.1.1).
Additional representative examples include Gen 1:5, 1:10, 4:2, 12:12, 18:33, 24:53, 25:5–6,
25:33–34, 31:47, 33:16–17, 35:18, 37:11, 40:21–22, and 41:51–52.

151 The linked segments 9.3.
(243) Gen 4:4–5

‫ ואל קין ואל מנחתו לא שעה‬:‫ אל הבל ואל מנחתו‬′‫וישע י‬
And the LORD paid heed to Abel and his offering, and to Cain and his
offering He paid no heed.
(244) Gen 37:11

:‫ויקנאו בו אחיו ואביו שמר את הדבר‬
And his brothers were jealous of him, and his father kept the matter
in mind.
Backward-linking topicalizations are almost always coordinated with the previ-

ous clause with the conjunction ‫ו‬, as in (243) and (244). When the linked items
have identical reference, the topicalized clause is often asyndetic, as in (245):
(245) Gen 40:12–13

‫ בעוד שלשת ימים ישא פרעה את ראשך‬:‫שלשת השרגים שלשת ימים הם‬
‫והשיבך על כנך‬
The three branches are three days. In another three days Pharaoh will lift
up your head and return you to your post.
Forward-linking topicalization points ahead to a subsequent clause, which

usually undergoes topicalization as well.15 A particularly clear-cut example of
forward-linking topicalization is the underlined clause in (246), which occurs
in the absence of any preceding discourse:
(246) Gen 2:16–17

‫ ומעץ הדעת טוב ורע‬:‫ מכל עץ הגן אכל תאכל‬:‫ אלהים על האדם לאמר‬′‫ויצו י‬
:‫לא תאכל ממנו כי ביום אכלך ממנו מות תמות‬
And the LORD God commanded the man, saying, “From every tree of
the garden you may eat; and from the tree of the knowledge of good
and evil, you shall not eat from it, for on the day that you eat from it you
shall die.”
The initial underlined clause marks a relation between the preposed ‫מכל עץ הגן‬
and the expression in the following clause, ‫מעץ הדעת טוב ורע‬.
Forward-linking topicalization is almost always asyndetic. 16 The syndesis/
asyndesis opposition between backward- and forward-linking topicalization
15. These are Andersen’s (1974) and Khan’s (1988) “parallel” structures (see §3.1.1).
An exception to this rule is Gen 45:17, where a forward-linking topicalization is followed by
a clause with normal word order.
16. One exception is Gen 50:20, where the waw is attached to the first topicalized clause,
rather than the second, as would be expected: ‫ואתם חשבתם עלי רעה אלהים חשבה לטבה למען‬
:‫‘ עשה כיום הזה להחית עם רב‬And you intended me harm, God intended it for good, to bring
it about that many people should be kept alive, as they are today’.

guides the addressee in correctly identifying the linked segment. If the topi-
calized clause is syndetic, the linked segment precedes. If it is asyndetic, the
linked segment lies ahead. Additional examples of forward-linking topicaliza-
tion are shown in (247) and (248). 17
(247) Gen 41:13

:‫ויהי כאשר פתר לנו כן היה אתי השיב על כני ואתו תלה‬
And as he interpreted to us, so it came to pass: me he restored to my of-
fice, and him he hanged.
(248) Gen 13:11–12

:‫ אברם ישב בארץ כנען ולוט ישב בערי הככר‬:‫ויפרדו איש מעל אחיו‬
And they separated from each other. Abram dwelt in the land of Canaan,
and Lot dwelt among the cities of the plain.
When the linked clause following a forward-linking topicalization contains

material identical to the former clause, the repetitious material may be gapped:
(249) Gen 7:2 18

‫מכל הבהמה הטהורה תקח לך שבעה שבעה איש ואשתו ומן הבהמה אשר‬
:‫לא טהרה הוא ]תקח לך[ שנים איש ואשתו‬
Of every clean animal you shall take for yourself seven pairs, a male
and its mate; and of every animal that is not clean [you shall take] two,
a male and its mate.
The counterpart of the preposed item is at the head of the gapped clause, and
the verb, along with other material identical to that in the topicalized clause,
has been omitted. The gapped material is shown in brackets. 19
In several instances, a forward-linking topicalization is followed by left-
dislocation: 20
17. See also, e.g., Gen 3:2; 14:4; 31:38, 40; 34:9, 21; 42:19, 33; and 44:12, as well as
examples (249)–(250).
19. A similar example is Gen 31:40. This verse shows that verb gapping does occur in
BH prose, contrary to O’Connor’s (1980: 401) denial of this possibility, discussed above in
§1.2. Additional examples of verb-gapping in prose can be found in (254) below. See Kugel
(1981: 321–22) for some additional examples.
20. An additional example is Gen 3:2. In Exod 1:22, a left-dislocation is followed by a
topicalized clause.

153 The linked segments 9.3.
(250) Gen 2:16 21

‫ ומעץ הדעת טוב ורע‬.‫ מכל עץ הגן אכל תאכל‬:‫ אלהים על האדם לאמר‬′‫ויצו י‬
shall die.”
The linked items in this example are ‫ מכל עץ הגן‬and ‫מעץ הדעת טוב ורע‬. The
first item is preposed, and the second is dislocated, with a resumptive enclitic
pronoun.
When the segments linked by topicalization are larger than a single clause,
the preposed item and its counterpart occur in the first clause of their respective
segments. Large discourse segments are often involved when the linked items
are time adverbials. The segments include all of the clauses referring to events
occurring at the designated time:
(251) Gen 8:13–14

‫ויהי באחת ושש מאות שנה בראשון באחד לחדש חרבו המים מעל הארץ‬
‫ ובחדש השני בשבעה‬:‫ויסר נח את מכסה התבה וירא והנה חרבו פני האדמה‬
:‫ועשרים יום לחדש יבשה הארץ‬
And it came to pass in the six hundred and first year, in the first
[month], on the first day of the month, the waters dried up from the
earth. And Noah removed the covering of the ark, and he looked, and
behold, the surface of the ground was dry. And in the second month, on
the twenty-seventh day of the month, the earth was dry.
The first segment encompasses all of the clauses in v. 13, and the second seg-
ment consists of the topicalized clause in v. 14.
Although topicalization most frequently links two segments, chains of
linked segments also occur: 22
(252) Gen 3:14 –19

‫ אל‬. . . .‫ אלהים אל הנחש כי עשית זאת ארור אתה מכל הבהמה‬′‫ויאמר י‬
‫ ולאדם אמר כי שמעת לקול‬. . . .‫האשה אמר הרבה ארבה עצבונך והרנך‬
‫אשתך‬
And the LORD God said to the serpent, “Because you did this, you are
more cursed than all beasts. . . .” To the woman he said, “I will greatly
increase your pangs in childbearing. . . .” And to Adam he said, “Be-
cause you listened to the voice of your wife”
21. Appears above as (246).
22. See also Gen 42:36, where the first two segments are nonverbal clauses and the third
a topicalized clause.

(253) Gen 14:4–6

‫ ובארבע‬:‫שתים עשרה שנה עבדו את כדרלעמר ושלש עשרה שנה מרדו‬
‫עשרה שנה בא כדרלעמר והמלכים אשר אתו ויכו את רפאים בעשתרת קרנים‬
‫ ואת החרי בהררם שעיר עד איל‬:‫ואת הזוזים בהם ואת האימים בשוה קריתים‬
:‫פארן אשר על המדבר‬
Twelve years they served Chedorlaomer, and in the thirteenth year
they rebelled. And in the fourteenth year Chedorlaomer came, and the
kings who were with him, and they defeated the Rephaim at Ashteroth-
karnaim, and the Zuzim at Ham, and the Emim at Shaveh-kiriathaim,
and the Horites in their hill country of Seir as far as El-paran, which is
by the wilderness.
Each of these examples contains three linked segments. In (252), a normal

clause is followed by two topicalized clauses, each of which includes a direct-
speech citation. 23 Example (253) contains three segments, the first two consist-
ing of one clause each and the last probably extending to the end of v. 6.
A particularly long chain from the classical BH prose corpus is shown in
(254); only part of the chain appears in the citation.
(254) Num 7:12–83 (excerpt)

‫ וקרבנו‬:‫ויהי המקריב ביום הראשון את קרבנו נחשון בן עמינדב למטה יהודה‬
‫קערת כסף אחת שלשים ומאה משקלה מזרק אחד כסף שבעים שקל בשקל‬
‫ כף אחת עשרה זהב מלאה‬:‫הקדש שניהם מלאים סלת בלולה בשמן למנחה‬
‫ שעיר עזים אחד‬:‫ פר אחד בן בקר איל אחד כבש אחד בן שנתו לעלה‬:‫קטרת‬
‫ ולזבח השלמים בקר שנים אילם חמשה עתודים חמשה כבשים בני‬:‫לחטאת‬
‫ ביום השני הקריב נתנאל בן צוער נשיא‬:‫שנה חמשה זה קרבן נחשון בן עמינדב‬
‫ ביום השלישי ]הקריב[ נשיא לבני זבולן אליאב בן חלן‬. . . .‫יששכר‬
The one who offered his offering on the first day was Nahshon the son
of Amminadab, of the tribe of Judah; and his offering was one silver
plate whose weight was a hundred and thirty shekels, one silver basin of
seventy shekels, according to the shekel of the sanctuary, both of them
full of fine flour mixed with oil for a meal offering; one golden ladle of
ten shekels, filled with incense; one bull of the herd, one ram, one lamb
in its first year, for a burnt offering; one goat for a sin offering; and
for the sacrifice of well being, two oxen, five rams, five he-goats, and
five yearling lambs. This was the offering of Nahshon the son of Am-
minadab. On the second day Nethanel the son of Zuar, the leader of Is-
sachar, made an offering. . . . On the third day the leader of the children
of Zebulun, Eliab son of Helon [made an offering]
23. As discussed in §4.4.3, the syntactic status of direct-speech citations is uncertain,

making it hard to determine how many nonsubordinate clauses there are in each segment.

155 Coherence relations between segments linked by topicalization 9.4.
The complete chain contains 12 segments. The first clause in each segment
contains a time adverbial: ‫ביום השני‬, ‫ ביום הראשון‬and so on. The first segment
opens with a clause with normal word order, and the second segment begins
with a backward-linking topicalized clause. The initial clause in the rest of the
segments has a preposed time adverbial and a gapped verb; for example, ‫ביום‬
‫‘ השלישי ]הקריב[ נשיא לבני זבולן אליאב בן חלן‬On the third day the leader of the
children of Zebulun, Eliab son of Helon [made an offering]’ (v. 24). 24
9.4. Coherence relations between segments linked by topicalization

Topicalization marks a coherence relation obtaining between the linked
clauses or segments, a relation concerning the pair of linked items. The spe-
cific relation involved in each case is left to the reader’s inference. Coherence
relations are categorized as content, epistemic, or speech-act relations, as de-
scribed in §6.2.2.
In most cases of topicalization the clauses or segments stand in a relation
of opposition or similarity. The categorization of these coherence relations is a
matter of debate. Halliday and Hasan (1976: 242) view opposition and similar-
ity as pragmatic (i.e, epistemic) relations, whereas Sanders et al. (1992) catego-
rize them as semantic (i.e., content). In my view, opposition and similarity are
epistemic relations, relating to speaker reasoning: the opposition or similarity
between the segments exists in the mind of the speaker rather than in the real
world. The distinction between opposition and similarity is to some degree sub-
jective and is not fully determined by the content of the sentences (see §9.4.2).
In some cases, the coherence relation involved in topicalization is of the
speech-act type. A relation of similarity concerning speech acts is termed addi-
tion. Other speech-act relations associated with topicalization are elaboration,
summary, and paraphrase. A type rare in the Genesis corpus but amply attested
elsewhere is the content relation of temporal succession. 25 It is likely that other
coherence relations are possible as well but do not occur in the corpus.
The coherence relations associated with topicalization are explored in more
detail in §§9.4.1–9.4.5.
9.4.1. Opposition
Segments linked by topicalization most frequently stand in a relation of op-
position: the linked items in the segments are marked in order to highlight the
difference between what is said about them. The opposition between the two
segments might be expressed in English by the conjunct adverbial in contrast;
the closest equivalent in BH is ‫אולם‬, although use of this conjunct is much less
24. Another long chain involving fronted time adverbials is found in the passage de-
scribing the offerings for the Sabbath and festivals (Num 28:9–29:38).
25. In a few cases, it is difficult to identify a specific coherence relation; these are Gen
2:10; 7:18–19; 10:11; 19:6, 10–11; 25:10; and 46:32.

9.4. Coherence relations between segments linked by topicalization 156
common than topicalization. Opposition is a binary relation; thus, this relation

is to be found in pairs of linked segments but not in topicalization chains.
Oppositional segments may consist of a normal word-order clause followed
by a backward-linking topicalization or a forward-linking topicalization fol-
lowed by a backward-linking one. The linked clauses often exhibit parallel
syntactic and lexical structure, as in (255) and (256). The linked clauses in
these examples have an identical syntactic structure and involve the same verb.
The opposition is between the postverbal constituents in both clauses. 26
(255) Gen 41:51–52

:‫ויקרא יוסף את שם הבכור מנשה כי נשני אלהים את כל עמלי ואת כל בית אבי‬
:‫ואת שם השני קרא אפרים כי הפרני אלהים בארץ עניי‬
And Joseph called the name of the first-born Manasseh, “For God has
made me forget all my hardship and my father’s entire house.” And the
name of the second he called Ephraim, “For God has made me fruitful
in the land of my affliction.”
(256) Gen 4:2

:‫ויהי הבל רעה צאן וקין היה עבד אדמה‬
And Abel was a keeper of sheep, and Cain was a tiller of the soil.
When oppositional segments exhibit parallel structure, it is often possible

to interpret the last constituent in the second clause as a focus. In (254), for
example, we are told that Joseph had two sons and named the first Menasseh.
This activates by inference the proposition that he also named the second son,
that is, “He named the second x.” The topicalized clause supplies the value for
x, “Ephraim.” Similarly, in (256) “a tiller of the soil” is identified as the value
for x with respect to the indirectly activated proposition “Cain was x.” Al-
though “Ephraim” and “a tiller of the soil” satisfy the pragmatic conditions for
focus, they are not marked as such on the syntactic level, because they stand in
their usual position in the clause.
Topicalized clauses such as (255) and (256) are indistinguishable structur-
ally and semantically from others that do not have a focus. An example of the
latter is (257).
(257) Gen 35:18 27

:‫ותקרא ]רחל[ שמו בן אוני ואביו קרא לו בנימין‬
And [Rachel] called his name Ben-oni; and his father called his name
Benjamin.
26. Additional representative examples include Gen 1:10, 2:16–17, 4:5, 7:2, 12:12,
13:12, 14:21, 15:10, 25:5–6, and 50:20.

The two clauses in this example are syntactically and semantically parallel,
just as in the examples above. Here, however, there is no part of the topicalized
clause that can be interpreted as a focus, because the fact that Jacob gave Ben-
jamin a different name is unexpected information that is not inferrable from
the preceding clause.
When the linked segments do not exhibit parallel structure, as is often the
case, there is no possibility of an unmarked focus. The opposition is inferred
from the content of the two clauses: 28
(258) Gen 42:3–4

‫ ואת בנימין אחי יוסף לא שלח‬:‫וירדו אחי יוסף עשרה לשבר בר ממצרים‬
‫יעקב את אחיו כי אמר פן יקראנו אסון‬
And ten of Joseph’s brothers went down to buy grain in Egypt. And
Benjamin, Joseph’s brother, Jacob did not send with his brothers, for
he said, “Lest disaster befall him.”
(259) Gen 12:15–16

‫ ולאברם היטיב בעבורה‬:‫ותקח האשה בית פרעה‬
And the woman was taken into Pharaoh’s house. And with Abram it
went well because of her.
In (258), an opposition is drawn between the 10 brothers’ going down to Egypt

and Benjamin’s remaining in Canaan. The topicalized clause does not state
straightforwardly that Benjamin did not go down but relates this fact indi-
rectly by stating that Jacob did not send him. In (259), an opposition is drawn
between the tragic event of Sarai’s abduction to the palace and the benefits
accruing to Abram as a result.
An interesting variety of opposition is found in cases in which the linked
referents refer to a set and a member of that set, respectively. The first clause
makes a statement regarding set A, and the second clause makes a directly
contradictory statement regarding member B. In order to reconcile the two
statements, the addressee is forced to conclude that the expression apparently
designating set A actually refers to all members of A except for B. Examples
include (260) and (261).
(260) Gen 2:16 29

‫ מכל עץ הגן אכל תאכל ומעץ הדעת טוב ורע‬:‫ אלהים על האדם לאמר‬′‫ויצו י‬
28. Additional representative examples include Gen 6:6–8, 17:20–21, 29:17, 33:14, 48:11,
and 50:24.
29. Appears as (246) and (250) above (p. 151 and p. 153, respectively).

shall die.”
(261) Gen 41:54

‫ותחלינה שבע שני הרעב לבוא כאשר אמר יוסף ויהי רעב בכל הארצות ובכל‬
:‫ארץ מצרים היה לחם‬
And the seven years of famine began to come, as Joseph had said. And
there was famine in all of the lands, and in all the land of Egypt there
was bread.
In (260), man is given permission to eat from every tree in the garden. The
subsequent clause makes it clear that this means every tree except for the Tree
of Knowledge. Similarly, in (261), ‫ כל הארצות‬actually means all of the lands
except for the land of Egypt. 30 Recognizing the link between the items in the
two segments and the coherence relation between the segments is critical for
the proper understanding of the first segment. By explicitly marking the links,
topicalization plays an important role in guiding the addressee to the intended
semantic interpretation.
Another interesting type of oppositional topicalization involves an a for-
tiori argument (‫‘ קל וחמר‬the argument from the minor to the major’). In this
type of argument, a speaker argues for a proposition p by asserting or implying
a weaker version of p, from which the stronger version naturally follows: “if
the ‘minor’ has this or that property then the ‘major’ must undoubtedly have
it” ( Jacobs 1972: 221). An example is (262).
(262) Gen 44:8
‫הן כסף אשר מצאנו בפי אמתחתינו השיבנו אליך מארץ כנען ואיך נגנב מבית‬
‫אדניך כסף או זהב‬
Behold, the money which we found in the mouth of our sacks, we
brought back to you from the land of Canaan, and how could we have
stolen from your master’s house silver or gold?
The linked items are ‫ כסף אשר מצאנו בפי אמתחתנו‬and ‫כסף או זהב‬. The first seg-
ment is a forward-linking topicalized clause with the clausal adverb ‫הן‬, 31 and
30. A slightly different type is ‫לכלם נתן לאיש חלפות שמלת ולבנימן נתן שלש מאות כסף‬
‫‘ וחמש חלפת שמלת‬To each of them he gave a change of clothing, and to Benjamin he gave
300 pieces of silver and five changes of clothing’ (Gen 45:22). The topicalized clause does
not contradict the literal meaning of the previous one, because Benjamin, like his brothers,
received a change of clothing (following most translations, which take ‫ חלפות‬as referring
to a single set of clothing.) Instead, the contradiction is between the implication of the first
clause that the brothers (including Benjamin) received only a change of clothing and the fol-
lowing assertion that Benjamin received a large sum of money in addition to five changes
of clothing.
31. On the use of ‫ הן‬in a fortiori arguments, see Blau (1993: §103.2); Garr (2004: 332).

the second segment is a rhetorical question. The rhetorical question implies the
assertion “We did not steal gold or silver from Joseph’s house.” Topicalization
highlights the “minor/major” relation between the linked items: “the money
which we found in the mouth of our sack” is money that the brothers could
have been justified in keeping, whereas they had no right to “silver or gold”
from Joseph’s house. “If we returned the money that we could have kept for
ourselves,” the brothers say, “we certainly would not steal Joseph’s money.”
The use of topicalization in a fortiori arguments occurs elsewhere in the
classical BH prose corpus as well, as shown in the following examples.
(263) Exod 6:12

‫הן בני ישראל לא שמעו אלי ואיך ישמעני פרעה‬
Behold, the Israelites have not listened to me, and how should Pharaoh
listen to me?
(264) 2 Kgs 10:4

:‫הנה שני המלכים לא עמדו לפניו ואיך נעמד אנחנו‬
Behold, the two kings could not withstand him, and how can we with-
stand [him]?
(265) Judg 14:16 32

‫הנה לאבי ולאמי לא הגדתי ולך אגיד‬
Behold, I have not told my father or my mother, and you I should tell?
(266) 1 Kgs 8:27

:‫הנה השמים ושמי השמים לא יכלכלוך אף כי הבית הזה אשר בניתי‬
Behold, heaven and the highest heaven cannot contain you; how much
less this house which I have built.
As in (262) above (p. 158), the arguments in these examples start with a
topicalized clause beginning with ‫ הן‬or ‫הנה‬, followed by a rhetorical question.
In (265), the rhetorical question is topicalized as well. In (266), the topicalized
clause is followed by an elliptical clause introduced with ‫‘ אף כי‬how much
more so’. 33 Everything in the elliptical clause has been omitted except for the
linked item. In each case, a forward-linking topicalization highlights a “minor/
major” relation between the linked items and an oppositional relation between
the segments.
32. The rhetorical question here is unmarked, lacking an interrogative particle or adverb.
33. For additional examples (not necessarily involving topicalization of a finite clause)
of the ‫ אף כי‬. . . ‫הנה‬/‫ הן‬sequence, see 1 Sam 23:3 and 2 Sam 16:11. The phrase ‫ אף כי‬occurs
without preceding ‫הנה‬/‫ הן‬in 1 Sam 14:29–30; Prov 15:11; 19:7, 10. See Labuschagne (1973:
7) and Stec (1987: 479–80).

9.4.2. Similarity
Segments linked by topicalization frequently stand in a relation of similar-
ity, a relation that might be expressed in English by the conjunct similarly or
likewise. The distinction between opposition and similarity is not completely
determined by the content of the clauses but depends on whether the speaker
views what is said about the linked items as different or similar. As such, simi-
larity and opposition are epistemic rather than content relations. The distinc-
tion between opposition and similarity is at times very subtle, and some similar
segments are apt to be taken at first glance as oppositional. Unlike the relation
of opposition, the similarity relation may obtain in a chain of linked segments.
A relation of similarity is obvious when closely related things are stated
about each of the linked items. A chain of topicalized segments of this type
from the classical BH prose corpus is shown in (267):
(267) Deut 7:5
‫כי אם כה תעשו להם מזבחתיהם תתצו ומצבתם תשברו ואשירהם תגדעון‬
‫ופסיליהם תשרפון באש‬
But thus shall you do to them: their altars you shall break down, and
their pillars you shall smash, and their sacred posts you shall cut down,
and their graven images you shall burn with fire.
The verbs in the linked clauses denote various types of destruction. The series
of topicalized clauses highlights the common action that is to be performed on
all of the cultic objects mentioned in the clauses. 34
Paradoxically, a relation of similarity may exist even when what is stated
about the linked items involves antonyms or near-antonyms:
(268) Gen 46:4

‫אנכי ארד עמך מצרימה ואנכי אעלך גם עלה‬
I will go down with you to Egypt, and I will also bring you up.
Although the verbs ‫ ארד‬and ‫ אעלך‬are near-antonyms from a semantic perspec-

tive, the clauses in their entirety are more similar than oppositional: the point is
that God will be with Jacob both at the start and the conclusion of the exile to
Egypt. A similar example is (269):
(269) Gen 34:9 35

:‫והתחתנו אתנו בנתיכם תתנו לנו ואת בנתינו תקחו לכם‬
Intermarry with us. Your daughters give to us, and our daughters take
for yourselves.
34. Other representative examples of chains with a relation of similarity from the classi-
cal BH prose corpus include 1 Sam 8:11–17; 2 Sam 12:3; 2 Kgs 3:19, 23:11–13. See also the
pairs of linked segments in Josh 11:6, 9.
35. See also Gen 34:16, Deut 7:3.

Although ‫ תתנו‬and ‫ תקחו‬are antonymous, the topicalized clauses are similar in

that they describe two ways in which intermarrying may take place. As such,
they serve as an elaboration of the introductory clause, ‫והתחתנו אתנו‬.
In several cases, segments that appear to be oppositional at first glance are
revealed on further reflection to be of the similarity type: 36
(270) Gen 18:33 37

:‫ כאשר כלה לדבר אל אברהם ואברהם שב למקמו‬′‫וילך י‬
And the LORD departed when he had finished speaking to Abraham; and
[similarly] Abraham returned to his place
In this verse, the similarity between God’s and Abram’s departure is high-
lighted, rather than an opposition between their different destinations or be-
tween God’s going and Abram’s returning. In rendering the verse, similarly
yields a more suitable result than in contrast: ‘And God left when he finished
speaking to Abram. Similarly/in contrast Abram returned to his place’.
Topicalization chains in genealogy passages involve the similarity relation.
This is a frequent type of topicalization in Genesis: 38
(271) Gen 4:18
‫ויולד לחנוך את עירד ועירד ילד את מחויאל ומחייאל ילד את מתושאל‬
:‫ומתושאל ילד את למך‬
And Irad was born to Enoch; and Irad fathered Mehujael, and Mehu-
jael fathered Methushael, and Methushael fathered Lamech.
The point here is the continuation of the chain of lineage, rather than an op-
position between the names. 39
9.4.3. Addition
The additive relation pertains to speech acts rather than speaker reasoning
and can be paraphrased as “I say in addition.” In English, conjuncts such as
furthermore and moreover mark an additive relation, and in BH the conjunct
‫ גם‬may be used for this purpose, as an alternative to or in combination with
36. See also, e.g., Gen 1:20, 22; and 45:14. A number of cases are ambiguous, admitting
either oppositional or similarity interpretations; these include Gen 24:53, 25:33–34, 31:40,
32:1–2, and 40:21–22.
38. Additional representative examples of topicalized clauses in genealogy passages in-
clude Gen 10:8, 13–14, 15–18, 24 (2×), 25, 26–29; 11:12, 14, 27 (2×).
39. In this example and others like it, there is an ambiguity regarding the counterparts of
the preposed items. In the second clause, for example, Irad may be linked with the previous
mention of this individual in the preceding clause (bold). Alternatively, Irad may be linked
with his father, Enoch.

topicalization. Examples of topicalization with the additive relation include: 40
(272) Gen 42:36

‫ויאמר אלהם יעקב אביהם אתי שכלתם יוסף איננו ושמעון איננו ואת בנימן‬
‫תקחו‬
And Jacob their father said to them, “You have bereaved me: Joseph is
no more, and Simeon is no more, and Benjamin you would take.”
(273) Gen 44:9 41

:‫אשר ימצא אתו מעבדיך ומת וגם אנחנו נהיה לאדני לעבדים‬
Whichever of your servants it is found with, he shall die, and further-
more, we will be my lord’s slaves.
In (272), Jacob justifies his assertion that he has been and will be bereaved by
the brothers with a three-part proof. Two nonverbal clauses are followed by a
topicalized clause. In (273), the brothers specify two penalties they will incur
if the thief is among them. The first clause takes the form of a left-dislocation,
whereas the second is a topicalized clause containing the conjunct ‫גם‬. 42
9.4.4. Elaboration, summary, and paraphrase
Elaboration, summary, and paraphrase are coherence relations that concern
speech acts. There are no linguistic markers that explicitly signal the elabora-
tion relation in English or in BH; the summary and paraphrase relations can be
marked in English by to sum up and that is to say, respectively. Topicalization
involving elaboration, summary, or paraphrase generally involves segments
with identical linked items.
In one pattern of the elaborative type, a forward-linking topicalization with
a preposed demonstrative pronoun is followed by a segment consisting of sev-
eral clauses: 43
(274) Gen 45:17–18
‫ויאמר פרעה אל יוסף אמר אל אחיך זאת עשו טענו את בעירכם ולכו באו‬
‫ וקחו את אביכם ואת בתיכם ובאו אלי‬:‫ארצה כנען‬
And Pharaoh said to Joseph, “Say to your brothers, ‘Thus you shall do:
load up your beasts and come to the land of Canaan. And take your
father and your households, and come to me.’”
40. Clauses linked by an additive relation usually resemble each other in content, mak-
ing it hard at times to determine whether similarity or addition is intended. Ambiguous ex-
amples include Gen 31:38, 39:8, and 41:40.
41. The word ‫ גם‬should be interpreted here as a conjunct rather than a focusing adverb
(see §5.2). See also Gen 21:26.
42. An additional example with ‫ גם‬is Gen 44:9.
43. Additional examples include Gen 42:15–16, 18–20, 33–34; 43:11–13; and 45:19.

In a second pattern of elaboration, an initial segment is followed by a backward-

linking topicalization with a preposed lexical expression: 44
(275) Gen 40:12–13 45

‫ בעוד שלשת ימים ישא פרעה את ראשך‬:‫שלשת השרגים שלשת ימים הם‬
‫והשיבך על כנך‬
Then Joseph said to him, “The three branches are three days. In another
three days Pharaoh will lift up your head and return you to your post.”
A summary relation can be observed in (276), which has a backward-linking

topicalized clause with a preposed demonstrative pronoun. 46
(276) Gen 22:20–23 47
‫ את עוץ בכרו‬:‫ויגד לאברהם לאמר הנה ילדה מלכה גם הוא בנים לנחור אחיך‬
‫ ואת כשד ואת חזו ואת פלדש ואת‬:‫ואת בוז אחיו ואת קמואל אבי ארם‬
‫ ובתואל ילד את רבקה שמנה אלה ילדה מלכה לנחור אחי‬:‫ידלף ואת בתואל‬
:‫אברהם‬
And it was told to Abraham, “Behold, Milcah also has born children to
your brother Nahor: Uz the first-born, and Buz his brother, and Kem-
uel the father of Aram, and Chesed, and Hazo, and Pildash, and Jid-
laph, and Bethuel.” And Bethuel fathered Rebekah. These eight Milcah
bore to Nahor, Abraham’s brother.
In (277), the linked segments are related by a paraphrase relation. 48
(277) Gen 43:9 49

‫אנכי אערבנו מידי תבקשנו‬
I will be surety for him; from my hand you shall require him.
9.4.5. Temporal succession

Temporal succession is a relation that pertains to the contents of the two
segments and involves segments with linked temporal adjuncts. Although
Genesis contains only examples of this coherence relation, the type occurs
frequently in the larger classical BH prose corpus. The linked segments range
from single clauses to large units; long topicalization chains also occur. Ex-
amples of single-clause segments follow:
44. Additional examples include Gen 8:5, 11:27, 36:1–3, and 40:18–19.
46. An additional example is Gen 36:13.
48. An additional example is the very similar Gen 31:39.

(278) Exod 23:12 50

‫ששת ימים תעשה מעשיך וביום השביעי תשבת‬
Six days you shall do your work, and on the seventh day you shall rest.
(279) 2 Sam 11:12

‫ויאמר דוד אל אוריה שב בזה גם היום ומחר אשלחך‬
Then David said to Uriah, “Remain here today also, and tomorrow I
will send you off.”
(280) 1 Kgs 6:37–38

‫ ובשנה האחת עשרה בירח בול הוא‬:‫ בירח זו‬′‫בשנה הרביעית יסד בית י‬
‫החדש השמיני כלה הבית לכל דבריו ולכל משפטו משפטיו‬
In the fourth year the foundation of the house of the LORD was laid, in
the month of Ziv. And in the eleventh year, in the month of Bul, that
is the eighth month, the house was finished in all its parts, and accord-
ing to all its specifications.
An example involving larger segments is (281). 51 Each segment describes

a sequence of events that begins at the time specified by the linked adjunct at
its head. The first segment extends from v. 3 to v. 7, and the second segment
probably extends from v. 8 to v. 21.
(281) 2 Kgs 25:3–8

‫ ולא היה לחם לעם הארץ ותבקע העיר‬52 ‫בתשעה לחדש ויחזק הרעב בעיר‬
‫וכל אנשי המלחמה הלילה דרך שער בין החמתים אשר על גן המלך וכשדים על‬
‫ ובחדש החמישי בשבעה לחדש היא שנת‬. . . :‫העיר סביב וילך דרך הערבה‬
‫תשע עשרה שנה למלך נבכדנאצר מלך בבל בא נבוזראדן רב טבחים עבד‬
‫ ואת בית המלך ואת כל בתי ירושלם ואת‬′‫ וישרף את בית י‬:‫מלך בבל ירושלם‬
. . . .‫כל בית גדול שרף באש‬
On the ninth day [in the fourth month] 53 and the famine was severe
in the city and there was no food for the people of the land. And the city
was breached; all the soldiers [fled] by night through the gate between
50. It is common for the first-time adjunct to express duration and the second-time ad-
junct to express a point in time immediately following the specified time span, as in this
example. Additional representative examples from the classical BH prose corpus include
Exod 22:29; 31:17; 35:2; Lev 12:2–3; 15:28–29; 19:23–24; Deut 15:12; Josh 6:3–4; 1 Kgs
8:65–66; 2 Kgs 11:3–4.
51. See also (253) and (254) (p. 154), above. Other examples include the synchronistic
verses in Kings in which preposed time adverbials mark long segments describing a king’s
reign, for example, ‫‘ ובשנת עשרים לירבעם מלך ישראל מלך אסא מלך יהודה‬And in the twentieth
year of Jeroboam the king of Israel, Asa the king of Judah became king’ (1 Kgs 15:9).
52. The temporal adverbial is connected to its clause by a conjunction (see §5.5.2).
53. Following the standard translation, the phrase “in the fourth month” is filled in from
the parallel passage in Jer 52:6.

165 Conclusion 9.5.
the double walls, which is by the king’s garden, and the Chaldeans were
all around the city. And they went in the direction of the Arabah. . . . And
in the fifth month, on the seventh day of the month (that was the
nineteenth year of King Nebuchadnezzar, king of Babylon) Nebu-
zaradan, the captain of the bodyguard, a servant of the king of Babylon,
came to Jerusalem. And he burned the house of the LORD, and the king’s
house and all the houses of Jerusalem; and every great house he burned
down.
The events in the first segment do not necessarily all precede the events in the
second; that is, it is not clear that Zedekiah reached Babylon before the seventh
day of the fifth month. The segments are nevertheless temporally successive
in that the first sequence begins prior to the inception of the second sequence.
9.5. Conclusion
Topicalization functions as a discourse-connective device, signaling that the
marked clause bears a coherence relation to an adjacent discourse segment, ei-
ther preceding or following. Topicalization marks the preposed item as linked
to an item in the first clause of the adjacent segment; the linked items bear a
contextual relation which may be of a logical or ad hoc nature.
The specific coherence relation between the linked segments is not speci-
fied by topicalization. By examining the segments in the light of the linked
items, the addressee infers the intended relation between the clauses. The most
common relations are opposition and similarity, both of which are epistemic
relations. Opposition always obtains between a pair of linked segments, while
similarity may involve a topicalization chain. Oppositional segments may or
may not exhibit parallel semantic and syntactic structure. In some cases, the
relation between the linked items is crucial to the semantic interpretation of the
two segments, as in the case of literally contradictory segments. Oppositional
segments also occur in a fortiori arguments, highlighting the “major/minor”
relation crucial to the argument.
Topicalization may involve a speech act relation such as addition, elabora-
tion, summary, or paraphrase. The last three types are characteristic of topi-
calization involving identical linked items. When linked time adjuncts are
involved, the coherence relation is the content relation of temporal succession.
This type may involve a chain of long discourse segments, each describing a
series of sequential events.
The topicalized clause, at least in the Genesis corpus, always occurs in the
context of a second segment to which it is related. In contrast, the focused
clause can occur as an isolated clause, relating to information activated by the
extratextual situation.

9.5. Conclusion 166
Topicalization does not mark a constituent as focus. Although in a subset

of oppositional topicalized clauses, it is possible to view the final constituent
as a focus from a pragmatic perspective, this constituent is not syntactically
marked in any way.

Chapter 10
Conclusion
The purpose of this study was to investigate BH preposing from a syntactic

and pragmatic perspective, using the Genesis corpus as the basis of a statisti-
cal analysis. To this end, syntactic categories were defined, making possible
the delineation and classification of clause units. Preposing was distinguished
from other similar syntactic constructions and the preposed constituent dis-
tinguished from other preverbal elements such as clausal adverbs, negative
particles, and left-dislocated elements. The various preverbal elements were
ranked in terms of their degree of detachment from the clause. Marked prepos-
ing was distinguished from syntactically obligatory preposing, which lacks a
pragmatic function.
The thesis was advanced that the majority of instances of preposing can be
classified as either focusing or topicalization. In order to substantiate this the-
sis it was necessary to develop a clear and precise understanding of these two
concepts, which have been understood in a variety of ways in the linguistic lit-
erature and in the literature on BH word order. Focusing was defined based on
the concept of informational focus and was put in the context of a current psy-
cholinguistic theory of text comprehension. The distinction between activation
and presupposition was shown to be crucial for the proper understanding of BH
focusing. Theories on the English topicalization construction were explored in
order to glean insights regarding BH topicalization. A new conception of BH
topicalization as a discourse-connective device was presented. It was shown
that the pragmatic functions of focusing and topicalization are of two different
types. Both address the interpretation of the clause in its context, but the rele-
vant context in the case of focusing is information in the hearer’s conscious-
ness, and the context in the case of topicalization is an adjacent text segment.
A statistical analysis of Genesis was presented, demonstrating that focusing
and topicalization account for about 57% of the preposed clauses in the cor-
pus; this figure is higher (67%) in narrative clauses and lower (50%) in direct-
speech clauses. The actual figures may be slightly higher, because it is likely
that some clauses in the residue were intended as focused clauses, despite the
lack of contextual evidence for an appropriate activated proposition.
In contrast to the relatively high frequency of information-structure func-
tions, functions relating to the clause as a whole, such as marking simultaneity,
167

Conclusion 168
anteriority, background information, the beginning of a narrative unit, justifica-

tion, affirmation, or boasting, together characterize a little more than one-tenth
of the clauses. The explanation for the preposing in 31% of clauses is currently
unclear; the overwhelming majority of the unexplained clauses are in direct
speech. Thus, functions discussed in this work provide a relatively complete
account of preposing in narrative and account for preposing in direct speech to
a lesser degree.
The statistical analysis revealed significant differences between preposing
patterns in direct speech and in narrative. The impression that word order in di-
rect speech is “freer” than in narrative is due to two factors: first, direct speech
has a significantly higher incidence of preposing than narrative, and, second,
direct speech has a higher proportion of preposed clauses the pragmatic func-
tion of which is unclear.
It was shown that preposing for focus in BH relates to an activated but not
necessarily presupposed proposition. Despite the fact that BH focused clauses
are often rendered with a cleft sentence in English, the English constructions
most equivalent to BH focusing are focusing by accent marking and preposing
(in the casual register), both of which relate to activated propositions. The pre-
supposition/activation distinction clarifies several important points regarding
BH preposing for focus, including the fact that yes-no questions and commands,
which often do not involve presupposition, may be focused. The distinction
also helps to explain how additive focusing by preposing differs from focusing
by ‫גם‬, an adverb that can relate to a presupposed but non-activated proposi-
tion. It also explains why some focus-of-negation clauses cannot be appropri-
ately translated with an English negative cleft sentence: because clefts relate
to presupposed information, a cleft is an appropriate translation for focus-of-
negation only when the relevant activated proposition happens to be presup-
posed as well.
As a discourse-connective device, topicalization functions in a manner
similar to the clausal adverb of the conjunct type (see §5.2.1). Like conjuncts,
topicalization can link single clauses or larger adjacent segments. Like the
conjunct ‫הלוא‬, topicalization can create a link to a preceding or a following
segment. Both topicalization and conjuncts are used to indicate a coherence
relation between discourse segments, such as opposition, addition, elabora-
tion, and summarizing. Nevertheless, there are important differences between
the two devices. First, conjuncts are usually specific to a particular coherence
relation; for example, ‫ אולם‬indicates contrast and ‫ גם‬an additive relation on the
speech-act level (see §5.2.1, p. 72 n. 24, and §9.4.3). Topicalization, however,
marks a range of coherence relations. A feature unique to topicalization is the
link it highlights or creates between two items and the connection between this
link and the coherence relation obtaining between the clauses. For example,
while the conjunct ‫ גם‬can be used to connect clauses with an overall thematic

169 Conclusion
similarity but no specific parallel elements, the similarity in additive topical-

ization directly concerns the pair of linked items.
By highlighting a relation such as opposition or similarity, topicalization
gives the hearer/reader a nuanced appreciation of the contribution of the clause
to its context. At times, understanding the coherence relation signaled by topi-
calization is essential for the correct interpretation of the linked clauses. In
seemingly self-contradictory units in which the linked items are a set and a
member of that set, topicalization assists the reader in inferring which propo-
sition is actually expressed by the first clause. In the a fortiori argument, the
hearer/reader’s acceptance of the argument hinges on the recognition of a “mi-
nor/major” relation between the items linked by topicalization.
Although the present research has underlined the importance of information-
structure functions for the understanding of preposing, the fact that a third of
the clauses remain unexplained means that the picture is far from complete.
Future research is needed to illuminate the function of preposing in clauses
that are not focused or topicalized and do not have any of the clause-level
functions discussed here. Other avenues of research unexplored here include
the syntactic and pragmatic analysis of preposing in subordinate clauses and
the investigation of the pragmatics of marked constructions such as double
preposing, preposing with focus adverbs, and left-dislocation. Like preposing,
left-dislocation can be used for focusing and topicalization. An issue worthy
of investigation is whether Polak’s (1999, 2001, 2003) distinction of various
direct speech and narrative styles (see §1.2) sheds further light on the statistical
distribution of the various functional types of preposing. An important topic
explored by Lunn (2006) and worthy of continued investigation is the degree
to which pragmatic functions such as focusing and topicalization play a role in
the word order of biblical poetry.
The concepts of focusing and topicalization as developed in this work lead
to a richer understanding of the importance of preposing for the pragmatic in-
terpretation of the sentence. Focusing and topicalization are conceived here as
syntactic-pragmatic categories that have equivalents in English and other lan-
guages. The findings of this study, therefore, are potentially relevant to marked
constructions in other languages as well. As such, this study plays a part in the
cross-linguistic investigation of marked forms with a pragmatic contribution
to sentence meaning.

References
Abadi, Adina
1988 ‫[ תחביר השיח של העברית החדשה‬Discourse syntax of contemporary Hebrew].
Jerusalem: Magnes.
Aejmelaeus, Anneli
1986 Function and interpretation of ‫ כי‬in Biblical Hebrew. JBL 105: 193–209.
Alter, Robert
1996 Genesis: Translation and commentary. New York: Norton.
Andersen, Francis I.
1970 The Hebrew verbless clause in the Pentateuch. Journal of Biblical Literature
Monograph Series 14. Nashville: Abingdon.
1974 The sentence in Biblical Hebrew. Janua Linguarum Series Practica 231. The
Hague: Mouton.
Andersen, Francis I., and Forbes, A. Dean
1983 “Prose particle” counts of the Hebrew Bible. Pp. 165–83 in The Word of the
Lord shall go forth: Essays in honor of David Noel Freedman in celebration
of his sixtieth birthday, ed. Carol L. Meyers and M. O’Connor. American
Schools of Oriental Research Special Volume Series 1. Winona Lake, IN:
Eisebrauns.
Andrews, Avery
1985 The major functions of the noun phrase. Pp. 62–154 in Clause structure,
vol. 1 of Shopen 1985.
Ariel, Mira
2001 Accessibility theory: An overview. Pp. 29–87 in Sanders, Schilperoord, and
Spooren 2001.
Azar, Moshe
1972 ‫ עיון תחבירי וסמנטי‬:‫ מלות הדרכה והיחידות המילוניות של הפועל‬,‫סימני הצרכה‬
‫[ בפעלים בעלי משלימים‬A syntactic and semantic approach to verbs governing
prepositional phrases]. Leš 36: 220–27, 282–86.
1993 ‫זממנו‬-‫[ לקראת הבנת המשפט הממוקד בעברית בת‬Towards an understanding of
the focused clause in contemporary Hebrew]. Pp. 87–99 in vol. 1 of ‫העברית‬
‫תרבותיים בעקבות כנס‬-‫ קובץ מחקרים על הלשון בהקשריה החברתיים‬:‫שפה חיה‬
‫ן‬′′‫[ אורנים תש‬Hebrew: A living language. Studies on the language in social
and cultural contexts], ed. Uzi Arnon, Rina Ben-Shahar, and Gideon Toury.
Haifa: University of Haifa Press.
Bailey, Nicholas A.
2004 A second look at double preverbal constituents. Review of Gross (2001). HS
45: 253–76.
170

171 References
Bailey, Nicholas A., and Levinsohn, Stephen H.

1992 The function of preverbal elements in independent clauses in the Hebrew nar-
rative of Genesis. JTT 5: 179–207.
Bandstra, Barry L
1982 The syntax of particle KY in Biblical Hebrew and Ugaritic. Ph.D. diss. Yale
University.
1992 Word order and emphasis in Biblical Hebrew narrative: Syntactic observa-
tions on Genesis 22 from a discourse perspective. Pp. 109–23 in Bodine
1992.
Battistella, Edwin L.
1996 The logic of markedness. New York: Oxford University Press.
Ben-Asher, Mordechai
1973 ‫פועל‬-‫[ להבחנה בין מושא עקיף ותיאור‬On the distinction between indirect object
and adverbial]. Pp. 54–71 in ‫[ עיונים בתחביר העברית החדשה‬Studies in modern
Hebrew syntax]. Tel Aviv: Hakibbutz Hameuhad.
Bendavid, Abba
1958 ‫[ כיצד סדר נושא ונשוא‬On the order of subject and predicate]. LešLaʿam 9:
67–75, 97–107, 149–65.
1971 ‫[ לשון מקרא ולשון חכמים‬Biblical Hebrew and Mishnaic Hebrew], vol 2: ‫דקדוק‬
‫[ ותרגילי סגנון‬Grammar and style]. Tel Aviv: Dvir.
Ben-Ḥayyim, Zeʾev
2000 A grammar of Samaritan Hebrew: Based on the recitation of the Law in com-
parison with the Tiberian and other Jewish traditions, rev. and trans. Abra-
ham Tal. Jerusalem: Magnes.
Ben-Horin, Gad
1976 Aspects of syntactic preposing in spoken Hebrew. Pp. 193–207 in Studies in
Modern Hebrew syntax and semantics: The transformational generative ap-
proach, ed. Peter Cole. Amsterdam: North-Holland.
Bergen, Robert D., ed.
1994 Biblical Hebrew and discourse linguistics. Dallas: SIL.
Bergsträsser, Gotthelf
1962 Hebräische Grammatik, part 2: Verbum, ed. E. Kautzch. Hildesheim: Olms.
Berlin, Adele
1985 The dynamics of Biblical parallelism. Bloomington: Indiana University
Press.
Berman, Ruth A.
1980 The case of an (S)VO language: Subjectless constructions in modern Hebrew.
Lang 56: 759–76.
1997 Modern Hebrew. Pp. 312–33 in The Semitic languages, ed. Robert Hetzron.
Routledge Language Family Descriptions. London: Routledge.
Birner, Betty J., and Gregory Ward
1998 Information status and noncanonical word order in English. Studies in Lan-
guage Companion Series 40. Amsterdam: Benjamins.

References 172
Blakemore, Diane
1987 Semantic constraints on relevance. Oxford: Blackwell.
Blau, Joshua
1959 Adverbia als psychologische un grammatische Subjekte/Prädikate im Bibel-
hebräisch. VT 9: 130–37.
1972 ‫[ על חזרת הנשוא במקרא‬On the repetition of the predicate in the Bible]. Pp.
234–40 in ‫ מחקרים לזכרו של י‘ ליוור‬:‫[ המקרא ותולדות ישראל‬Bible and Jewish
history: Studies in Bible and Jewish history dedicated to the memory of Ja-
cob Leiver], ed. Benjamin Uffenheimer. Tel Aviv: Tel Aviv University Press.
Repr., pp. 124–30 in ‫[ עיונים בבלשנות עברית‬Studies in Hebrew linguistics] by
Joshua Blau. Jerusalem: Magnes, 1996.
1973a ‫[ כיצד להבחין בהוראה בין מושא עקיף לתיאור‬How to distinguish between indi-
rect objects and adverbials in teaching]. Leš 37: 202–4.
1973b ‫[ על מבנה המשפט השמני בתורה‬On the structure of the nominal sentence in the
Bible]. Review essay on Andersen 1970. Leš 37: 69–74.
1977 An adverbial construction in Hebrew and Arabic: Sentence adverbials in
frontal position separated from the rest of the sentence. Proceedings of the
Israel Academy of Sciences and Humanities 6/1. Jerusalem: Israel Academy
of Sciences and Humanities.
1993 A grammar of biblical Hebrew. 2nd ed. Porta Linguarum Orientalium n.s. 12.
Wiesbaden: Harrassowitz.
Bloch, Alfred
1946 Vers und Sprache im altarabischen: Metrische und syntaktische Untersuc-
hungen. Acta Tropica Supplementa 5. Basel: Recht & Gesellschaft.
Bloch, Ariel A.
1986 Studies in Arabic syntax and semantics. Wiesbaden: Harrassowitz.
Bodine, Walter R., ed.
1992 Linguistics and Biblical Hebrew. Winona Lake, IN: Eisenbrauns.
1995 Discourse analysis of biblical literature: What it is and what it offers.
SBLSemeiaSt. Atlanta: Scholars Press.
Bolkestein, A. Machtelt
1998 What to do with topic and focus? Evaluating pragmatic information. Pp. 193–
214 in Functional grammar and verbal interaction, ed. Mike Hannay and
A. Machtelt Bolkestein. Studies in Language Companion Series 44. Amster-
dam: Benjamins.
Bosch, Peter, and van der Sandt, Rob, eds.
1999 Focus: Linguistic, cognitive, and computational perspectives. Studies in Nat-
ural Language Processing. Cambridge: Cambridge University Press.
Bravmann, M. M.
1953 Studies in Arabic and general syntax. Cairo: Institut Francais d’Archeologie
Orientale.
Brockelmann, Carl
1956 Hebräische Syntax. Neukirchen-Vluyn: Moers.
Brody, Jill
1984 Some problems with the concept of basic word order. Ling 22: 711–36.

173 References
Brongers, Hendrik A.
1981 Some remarks on the biblical particle “halō.” Pp. 177–89 in Remembering all
the way: A collection of Old Testament studies published on the occasion of
the fortieth anniversary of the Oudtestamentisch Werkgezelschap in Neder-
land, ed. B. Albrekton. OTS 21. Leiden: Brill.
Brown, Gillian, and George Yule
1983 Discourse analysis. Cambridge Textbooks in Linguistics. Cambridge: Cam-
bridge University Press.
Brown, Michael L.
1987 “Is it not?” or “indeed!”: HL in Northwest Semitic. Maarav 4: 201–19.
Buth, Randall
1987 Word order in Aramaic from the perspectives of functional grammar and
discourse analysis. Ph.D. diss. University of California.
1990 Word order differences between narrative and non-narrative material in Bibli-
cal Hebrew. Pp. 9–16 in vol. 1 of Proceedings of the Tenth World Congress
of Jewish Studies, Jerusalem August 16–24 1989. Jerusalem: World Union of
Jewish Studies.
1992 Topic and focus in Hebrew poetry: Psalm 51. Pp. 83–96 in Language in con-
text: Essays for Robert E. Longacre, ed. Shin Ja J. Hwand and William R.
Merrifield. SIL and the University of Texas at Arlington Publications in Lin-
guistics 107. Arlington: SIL and the University of Texas at Arlington.
1994a Methodological collision between source criticism and discourse analysis.
Pp. 138–54 in Bergen 1994.
1994b Contextualizing constituent as topic, non-sequential background and dra-
matic pause: Hebrew and Aramaic evidence. Pp. 215–31 in Function and
expression in functional grammar, ed. Elisabeth Engberg-Pedersen, Lisbeth
F. Jakobsen, and Lone S. Rasmussen. Functional Grammar Series 16. Berlin:
de Gruyter.
1995 Functional Grammar, Hebrew and Aramaic: An integrated, textlinguistic ap-
proach to syntax. Pp. 77–102 in Bodine 1995.
1999 Word order in the verbless clause: A generative-functional approach. Pp. 79–
108 in Miller 1999.
Carnie, Andrew, and Guilfoyle, Eithne, eds.
2000 The syntax of the verb initial languages. Oxford Studies in Comparative Syn-
tax. New York: Oxford University Press.
Cassuto, U.
1961a A commentary on the book of Genesis, part 1: From Adam to Noah, trans.
Israel Abrahams. Publications of the Perry Foundation for Biblical Research
in the Hebrew University of Jerusalem. Jerusalem: Magnes.
1961b The documentary hypothesis and the composition of the Pentateuch: Eight
lectures, trans. Israel Abrahams. Publications of the Perry Foundation for
Biblical Research in the Hebrew University of Jerusalem. Jerusalem: Magnes.
Chafe, Wallace
1976 Givenness, contrastiveness, definiteness, subjects and topics. Pp. 27–55 in Li
1976.

References 174
1987 Cognitive constraints on information flow. Pp. 21–51 in Coherence and

grounding in discourse: Outcome of a symposium, Eugene, Oregon, June
1984, ed. Russell S. Tomlin. TSL 11. Amsterdam: Benjamins.
1994 Discourse, consciousness and time: The flow and displacement of conscious
experience in speaking and writing. Chicago: University of Chicago Press.
Chatman, Seymour, and Samuel R. Levin, eds.
1967 Essays on the language of literature. Boston: Houghton.
Chavel, Chaim Dov, ed.
1966 ‫ באור על התורה‬:‫[ רבנו בחיי‬Rabbeinu Baḥye: Commentary on the Pentateuch].
3 vols. Jerusalem: Mossad Harav Kook.
1983 ‫ עם‬,‫ כתב־יד אוכספורד ומהדורת ברלינר‬,‫י על התורה על פי דפוס ראשון‬′′‫פרוש רש‬
‫ הערות וביאורים‬,‫ מקבילות‬,‫ ציוני מקורות‬,‫ שינויי נוסחאות‬,‫[ מבוא‬Rashi’s commen-
tary on the Pentateuch]. Jerusalem: Mossad Harav Kook.
Childs, Brevard S.
1979 Introduction to the Old Testament as Scripture. Philadelphia: Fortress.
Chomsky, Noam
1970 Deep structure, surface structure and semantic interpretation. Pp. 52–91 in
Studies in general and Oriental linguistics: Presented to Shiro Hattori on
the occasion of his sixtieth birthday, ed. Roman Jakobson and Shigeo Kawa-
moto. Tokyo: TEC. Repr., pp. 62–119 in Studies on semantics in generative
grammar, ed. Noam Chomsky. Janua Linguarum Series Minor 107. Paris:
Mouton, 1972.
1977 On Wh-Movement. Pp. 71–132 in Formal syntax, ed. Peter W. Culicover,
Thomas Wasow, and Adrian Akmajian. New York: Academic Press.
Claassen, Walter T.
1983 Speaker-orientated functions of kî in Biblical Hebrew. JNSL 11: 29–46.
Clark, Herbert H., and Haviland, Susan E.
1977 Comprehension and the given-new contract. Pp. 1–40 in Discourse produc-
tion and comprehension, ed. Roy O. Freedle. Discourse Processes: Advances
in Research and Theory 1. Norwood, NJ: Ablex.
Collins, John C.
1995 The wayyiqtol as “pluperfect”: When and why. Tyndale Bulletin 46: 117–40.
Comrie, Bernard
1989 Language universals and linguistic typology: Syntax and morphology. 2nd
ed. Oxford: Blackwell.
Cook, John A.
2001 The Hebrew verb: A grammaticalization approach. ZAH 14: 117–43.
2004 The semantics of verbal pragmatics: Clarifying the roles of wayyiqtol and
weqatal in Biblical Hebrew prose. JSS 49: 247–73.
Coulmas, Florian, ed.
1986 Direct and indirect speech. Trends in Linguistics Studies and Monographs
31. Berlin: de Gruyter.
Creider, Chet A.
1979 On the explanation of transformations. Pp. 3–21 in Givón 1979.

175 References
Croft, William
2003 Typology and universals. 2nd ed. Cambridge Textbooks in Linguistics. Cam-
bridge: Cambridge University Press.
Culicover, Peter W. and Louise McNally, eds.
1998 The limits of syntax. Syntax and Semantics 29. San Diego: Academic Press.
Davidson, A. B.
1901 Hebrew syntax. 3rd ed. Edinburgh: T. & T. Clark.
Davis, Steven, ed.
1991 Pragmatics: A reader. Oxford: Oxford University Press.
Dawson, David A.
1994 Text-linguistics and Biblical Hebrew. JSOTSup 177. Sheffield: Sheffield Aca-
demic Press.
DeCaen, Vincent
1995 On the placement and interpretation of the verb in Standard Biblical Hebrew
prose. Ph.D. diss. University of Toronto.
1999 A unified analysis of verbal and verbless clauses within Government-Binding
theory. Pp. 109–31 in Miller 1999.
Dempster, Stephen G.
1985 Linguistic features of Hebrew narrative: a discourse analysis of narrative
from the classical period. Ph.D. diss. University of Toronto.
Dijk, Teun A. van
1977 Text and context: explorations in the semantics and the pragmatics of dis-
course. Longman Linguistics Library 21. London: Longman.
1979 Pragmatic connectives. JPrag 3: 447–56.
Dik, Simon C.
1980 Studies in functional grammar. London: Academic Press.
1989 The theory of functional grammar, part 1: The structure of the clause. Func-
tional Grammar Series 9. Dordrecht: Foris.
Dik, Simon; Hoffman, Maria; de Jong, Jan R.; Djiang, Sie I.; Stroomer, Harry; and
de Vries, Lourens
1981 On the typology of focus phenomena. Pp. 41–74 in Hoekstra et al. 1981.
Disse, Andreas
1998 Informationsstruktur im biblischen Hebräisch: Sprachwissenschaftliche
Grundlagen und exegetische Konsequenzen einer Korpusuntersuchung zu
den Buchern Deuteronomium, Richter und 2 Könige. 2 vols. Arbeiten zu Text
und Sprache im Alten Testament 56. St. Ottilien: EOS.
Downing, Pamela
1995 Word order in discourse: By way of introduction. Pp. 1–28 in Downing and
Noonan 1995.
Downing, Pamela, and Noonan, Michael, eds.
1995 Word order in discourse. TSL 30. Amsterdam: Benjamins.
Driver, S. R.
1892 A treatise on the use of the tenses in Hebrew and some other syntactical
questions. 3rd ed. London: Oxford University Press. Repr., Biblical Resource
Series. Grand Rapids: Eerdmans, 1998.

References 176
1913 An introduction to the literature of the Old Testament. 9th ed. Edinburgh:
T. & T. Clark.
Dryer, Matthew S.
1995 Frequency and pragmatically unmarked word order. Pp. 105–35 in Downing
and Noonan 1995.
1996 Focus, pragmatic presupposition and activated propositions. JPrag 26: 475–
523.
Dyk, Janet W., and Talstra, Eep
1999 Paradigmatic and syntagmatic features in identifying subject and predicate in
nominal clauses. Pp. 133–85 in Miller 1999.
Erteschik-Shir, Nomi
1997 The dynamics of focus structure. Cambridge Studies in Linguistics 84. New
York: Cambridge University Press.
1998 The syntax-focus structure interface. Pp. 211–40 in Culicover and McNally
1998.
Esh, Shaul
1957 ‫פתיחה לפני דיבור ישיר בעברית‬-‫[ על מלות‬On particles introducing direct speech
in Hebrew]. Leš 22: 48–53.
Eskhult, Mats
1990 Studies in verbal aspect and narrative technique in Biblical Hebrew prose.
Studia Semitica Upsaliensia 12. Uppsala: Almqvist & Wiksell.
Exter Blokland, A. F. den
1995 In search of text syntax: Towards a syntactic text-segmentation model for
Biblical Hebrew. Applicatio 14. Amsterdam: VU University Press.
Ewald, Heinrich
1879 Syntax of the Hebrew language of the Old Testament, trans. James Kennedy.
Edinburgh: T. & T. Clark.
Firbas, Jan
1966a On defining the theme in functional sentence analysis. Travaux linguistiques
de Prague 1: 267–80.
1966b Non-thematic subjects in contemporary English. Travaux linguistiques de
Prague 2: 239–56.
1992 Functional sentence perspective in written and spoken communication. Stud-
ies in English Language. Cambridge: Cambridge University Press.
Floor, Sebastian J.
2003 From word order to theme in Biblical Hebrew narrative: Some perspectives
from information structure. JSem 12: 197–236.
2004 From information structure, topic and focus, to theme in Biblical Hebrew
narrative. Ph.D. diss. University of Stellenbosch.
2005 Poetic fronting in a wisdom poetry text: The information structure of Prov-
erbs 7. JNSL 31: 23-58.
Fokkelman, J. P.
1991 Narrative art in Genesis: Specimens of stylistic and structural analysis. 2nd
ed. Biblical Seminar 12. Sheffield: JSOT Press.

177 References
Follingstad, Carl M.
1995 Hinnēh and focus function: With application to Tyap. JTT 7: 1–24.
2001 Deictic viewpoint in Biblical Hebrew text: A syntagmatic and paradigmatic
analysis of the particle ‫כי‬. Dallas: SIL.
Folmer, M. L.
1995 The Aramaic language in the Achaemenid period: A study in linguistic varia-
tion. Orientalia Lovaniensia Analecta 68. Leuven: Peeters.
Fox, Andrew
1983 Topic continuity in early Biblical Hebrew. Pp. 215–54 in Givón 1983.
Fraser, Bruce
1990 An approach to discourse markers. JPrag 14: 383–95.
1996 Pragmatic markers. Prag 6: 167–90.
Freedman, David N.
1977 Pottery, poetry and prophecy: An essay on Biblical poetry” JBL 96: 5–26.
Friedman, Richard Elliott
1997 Who wrote the Bible? 2nd ed. New York: Harper & Row.
Garr, W. Randall
2004 ‫הן‬. RB 111: 321–44.
Ginsberg, H. L.
1942 Aramaic Studies Today. JAOS 62: 229–38.
Giora, Rachel
1982 ‫פרגמטי של משפטים‬-‫ נתוח פונקציונליסטי‬:‫סדר המלים במשפט ויחסו אל הטקסט‬
‫[ ממוקדים‬The function of topicalization at the sentence and discourse level].
Pp. 301–39 in ‫ קובץ מאמרים‬:‫[ עיונים בחקר השיח‬Studies in discourse analysis],
ed. Shoshana Blum-Kulka, Yishai Tobin, and Raphael Nir. Jerusalem: He-
brew University Press.
Givón, Talmy
1977 The drift from VSO to SVO in Biblical Hebrew: The pragmatics of tense-
aspect. Pp. 181–254 in Mechanisms of syntactic change, ed. Charles N. Li.
Austin: University of Texas Press.
1979 Discourse and syntax. Syntax and Semantics 12. New York: Academic Press.
1983 Topic continuity in discourse: A quantitative cross-language study. TSL 3.
Amsterdam: Benjamins.
1987 Beyond foreground and background. Pp. 175–88 in Coherence and ground-
ing in discourse: Outcome of a symposium, Eugene, Oregon, June 1984, ed.
Russell S. Tomlin. TSL 11. Amsterdam: Benjamins.
2001 Syntax: An introduction. Rev. ed. 2 vols. Amsterdam: Benjamins.
Glinert, Lewis
1989 The grammar of Modern Hebrew. Cambridge: Cambridge University Press.
Goldenberg, Gideon
1971 Tautological infinitive. IOS 1: 36–85. Repr., pp. 66–115 in Goldenberg 1998b.
1977 Imperfectly-transformed cleft sentences. Pp. 127–33 in vol. 1 of Proceedings
of the Sixth World Congress of Jewish Studies. Jerusalem: World Union of
Jewish Studies. Repr., pp. 116–22 in Goldenberg 1998b.
1991 On direct speech and the Hebrew Bible. Pp. 79–96 in Jongeling et al. 1991.

References 178
1998a On verbal structure and the Hebrew verb. Pp. 148–96 in Goldenberg 1998b.
English translation of ‫על תורת הפועל והפועל העברי‬. Language Studies 1
(1985): 295–348.
1998b Studies in Semitic linguistics: Selected writings. Jerusalem: Magnes.
Goldfajn, Tal
1998 Word order and time in Biblical Hebrew narrative. Oxford Theological
Monographs. Oxford: Clarendon.
Gómez-González, María A.
2001 The theme-topic interface: Evidence from English. Pragmatics and Beyond
n.s. 71. Philadelphia: Benjamins.
Gordis, Reuben (Robert)
1944 ‫[ על מבנה השירה העברית הקדמה‬On the structure of ancient biblical poetry].
Sefer Hashanah: The American Hebrew Year Book 7: 136–59.
Goutsos, Dionysis
1997 Modeling discourse topic: Sequential relations and strategies in expository
text. Advances in Discourse Processes 59. Norwood: Ablex.
Graesser, Arthur C.; Wiemer-Hastings, Peter; and Wiemer-Hastings, Katja
2001 Constructing inferences and relations during text comprehension. Pp. 249–71
in Text representation: Linguistic and psycholinguistic aspects, ed. Ted Sand-
ers, Joost Schilperoord, and Wilbert Spooren. Human Cognitive Processing
8. Philadelphia: Benjamins.
Greenberg, Joseph H.
1966a Some universals of grammar with particular reference to the order of mean-
ingful elements. Pp. 73–113 in Universals of language. Cambridge: MIT
Press.
1966b Language universals, with special reference to feature hierarchies. Janua
Linguarum Series Minor 59. The Hague: Mouton.
Greenstein, Eliezer (Ed)
2000 ‫[ הדיבור הישיר וצורת התקבלת‬Direct discourse and parallelism]. Pp. 33–40 in
‫ מנחות ידידות והוקרה לאוריאל סימון‬:‫[ עיוני מקרא ופרשנות כרך ה‬Studies in Bible
and exegesis, vol. 5: Presented to Uriel Simon], ed. Moshe Garsiel, Shmuel
Vargon, Amos Frisch, and Jacob Kugel. Ramat Gan: Bar Ilan University
Press.
Gregory, Michelle L., and Michaelis, Laura A.
2001 Topicalization and left-dislocation: A functional opposition revisited. JPrag
33: 1665–706.
Grimes, Joseph E.
1975 The thread of discourse. Janua Linguarum Series Minor 207. The Hague:
Mouton.
Gross, Walter
1986 Zum problem der Satzgrenzen im Hebräischen: Beobachtungen an Pendens-
konstruktionen. BN 35: 50–72.
1987a Zur Syntagmen-Folge im hebräischen Verbalsatz: Die Stellung des Subjekts
in Dtn 1–15. BN 40: 63–96.

179 References
1987b Die Pendenskonstruktion im biblischen Hebraïsch: Studien zum althebrae-

ischen Satz 1. Arbeiten zu Text und Sprache im Alten Testament 27. Mün-
chener Universitätsschriften. St. Ottilien: EOS.
1988a Der Einfluss der Pronominalisierung auf die Syntagmen-Folge im he-
bräischen Verbalsatz, untersucht an Dtn 1–25. BN 43: 49–69.
1988b Satzgrenzen bei Pendenskonstruktionen: Der Pendenssatz. Pp. 249–58 in
Text and Context: Old Testament and Semitic Studies for F.C. Fensham, ed.
W. Claassen. JSOTSup 48. Sheffield: JSOT Press.
1993a Die Position des Subjekts im Hebräischen Verbalsatz, untersucht an den asyn-
detischen ersten Redesätzen in Gen, Ex 1–19, Jos–2 Kön. ZAH 6: 170–87.
1993b Das Vorfeld als strukturell eigenständiger Bereich des hebräischen Verbal-
satzes. Pp. 1–24 in Syntax und Text: Beiträge zur 22. Internationalen Öku-
menischen Hebräisch-Dozenten-Konferenz 1993 in Bamberg, ed. Hubert
Irsigler. Münchener Universitätsschriften. Arbeiten zu Text und Sprache im
Alten Testament 40. St. Ottilien: EOS.
1994 Zur syntaktischen Struktur des Vorfeldes im hebräischen Verbalsatz. ZAH 7:
203–14.
1996 Die Satzteilfolge im Verbalsatz alttestamentlicher Prosa: Untersucht an den
Buchern Dtn, Ri und 2Kon. Forschungen zum alten Testament 17. Tübingen:
Mohr Seibeck.
1999 Is there really a compound nominal clause in Biblical Hebrew? Pp. 19–49 in
Miller 1999.
2001a Doppelt besetztes Vorfeld: Syntaktische, pragmatische und ubersetzung-
stechnische Studien zum althebraischen Verbsaltz. Beihefte zur Zeitschrift
fur die alttestamentliche Wissenschaft 305. Berlin: de Gruyter.
2001b Die Stellung der Zeitangabe in Sätzen mit zwei oder mehr nominalen/pro-
nominalen Satzteilen vor dem Verbum finitum in alttestamentlicher Poesie.
Pp. 35–50 in Sachverhalt und Zeitbezug: Semitistische und alttestamentliche
Studien. Adolf Denz zum 65. Geburtstag, ed. Rüdiger Bartelmus and Norbert
Nebes. Wiesbaden: Harrassowitz.
2004 Satzteilfolge—Aufmerksamkeitsleitung—Gedankenführung. KUSATU 5: 33–
66.
Gundel, Jeanette K.
1977 The role of topic and comment in linguistic theory. Bloomington: Indiana
University Linguistics Club.
1985 Shared knowledge and topicality. JPrag 9: 83–107.
1988 Universals of topic-comment structure. Pp. 209–39 in Studies in syntactic ty-
pology, ed. Michael Hammond, Edith Moravcsik, and Jessica Wirth. TSL 17.
1999 On different kinds of focus. Pp. 293–305 in Bosch and Sandt 1999.
Gundel, Jeanette K.; Hedberg, Nancy; and Zacharski, Ron
1993 Cognitive status and the form of referring expressions in discourse. Lang 69:
274–307.
Gussenhoven, Carlos
1999 On the limits of focus projection in English. Pp. 43–55 in Bosch and Sandt
1999.

References 180
Halliday, M. A. K.
1967 Notes on transitivity and theme in English: Part 2. JLing 3:199–244.
2004 An introduction to functional grammar. 3rd ed. London: Arnold.
Halliday, M. A. K., and Hasan, Ruqaiya
1976 Cohesion in English. English Language Series 9. London: Longman.
Hannay, Mike
1991 Pragmatic function assignment and word order variation in a functional
grammar of English. JPrag 16: 131–55.
Hatav, Galia
1985 Criteria for identifying the foreground. TheorLing 12: 265–73.
1997 The semantics of aspect and modality: Evidence from English and Biblical
Hebrew. Studies in Language Companion Series 34. Amsterdam: Benjamins.
2000a (Free) direct discourse in Biblical Hebrew. HS 41: 7–30.
2000b ‫[ תנועת הזמן בסיפור המקראי‬Movement of time in biblical narrative]. BI 47:
63–84.
Heimerdinger, Jean-Marc
1999 Topic, focus and foreground in ancient Hebrew narratives. JSOTSup 295.
Sheffield: Sheffield Academic Press.
Heller, Roy L.
2004 Narrative structure and discourse constellations: An analysis of clause func-
tion in Biblical Hebrew prose. Harvard Semitic Studies 55. Winona Lake, IN:
Eisenbrauns.
Hirschberg, Julia L.
1991 A theory of scalar implicature. Outstanding Dissertations in Linguistics. New
York: Garland.
Hockett, Charles F.
1958 A course in modern linguistics. New York: Macmillan.
Hoekstra, Teun; van der Hulst, Harry; and Moortgat, Michael, eds.
1981 Perspectives on Functional Grammar. Dordrecht: Foris.
Hoftijzer, Jacob
1973 The nominal clause reconsidered. Review essay on Andersen 1970. VT 23:
446–510.
1981 A search for method: A study in the syntactic use of the h-locale in Classical
Hebrew. SSLL 12. Leiden: Brill.
1985 The function and use of the imperfect forms with nun paragogicum in Classi-
cal Hebrew. SSN 21. Assen: Van Gorcum.
Holes, Clive
1995 The structure and function of parallelism and repetition in spoken Arabic: A
sociolinguistic study. JSS 40: 57–81.
Holmstedt, Robert D.
2002 The relative clause in Biblical Hebrew: A linguistic analysis. Ph.D. diss. Uni-
versity of Wisconsin.
2003 Adjusting our focus. Review essay on Shimasaki 2002. HS 44: 203–15.
2005 Word order in the book of Proverbs. Pp. 135–54 in Seeking out the wisdom of
the ancients: Essays offered to honor Michael V. Fox on the occasion of his

181 References
sixty-fifth birthday, ed. K. G. Friebel, D. R. Magary, and R. L. Troxel. Winona

Lake, IN: Eisenbrauns.
Hoop, Raymond de
2003 Genesis 49 revisited: The poetic structure of Jacob’s Testament and the an-
cient Versions. Pp. 1–32 in Korpel and Oesch 2003.
Hopper, Paul J.
1979 Aspect and foregrounding in discourse. Pp. 213–41 in Givón 1979.
Horn, Laurence R.
1989 A natural history of negation. Chicago : University of Chicago Press.
Hornkohl, Aaron
2003 The Pragmatics of the X + Verb Structure in the Hebrew of Genesis. M.A.
thesis. Hebrew University.
Huddleston, Rodney D.
1984 Introduction to the grammar of English. Cambridge Textbooks in Linguistics.
Cambridge: Cambridge University Press.
Huddleston, Rodney D., and Pullum, Geoffrey K.
2002 The Cambridge grammar of the English language. Cambridge: Cambridge
University Press.
Huehnergard, John
1985 Asseverative *la and hypothetical *lu/law in Semitic. JAOS 103: 569–93.
Hurvitz, Avi
1972 ‫ לתולדות לשון המקרא בימי בית שני‬:‫[ בין לשון ללשון‬The transition period in
Biblical Hebrew: A study in post-exilic Hebrew and its implications for the
dating of Psalms]. Jerusalem: Bialik Insititute.
1982 A linguistic study of the relationship between the priestly source and the book
of Ezekiel: A new approach to an old problem. Cahiers de la Revue Biblique
20. Paris: Gabalda.
Ilie, Cornelia
1994 What else can I tell you? A pragmatic study of English rhetorical questions as
discursive and argumentative acts. Stockholm Studies in English 82. Stock-
holm: Almqvist & Wiksell.
Jackendoff, Ray
1972 Semantic interpretation in generative grammar. Studies in Linguistics Series
2. Cambridge: MIT Press.
Jacobs, Louis
1972 The qal va-ḥomer argument in the Old Testament. BSOAS 35: 221–27.
Jakobsen, Roman
1967 Linguistics and poetics. Pp. 296–322 in Chatman and Levin 1967.
Johnstone, Barbara
1991 Repetition in Arabic discourse: Paradigms, syntagms, and the ecology of lan-
guage. Pragmatics and Beyond n.s. 18. Amsterdam: Benjamins.
Jong, Jan de
1981 On the treatment of focus phenomena in Functional Grammar. Pp. 89–115 in
Hoekstra et al. 1981.
Jongeling, Bastiaan
1980 Some remarks on the beginning of Genesis I, 2. FolOr 21: 27–32.

References 182
Jongeling, K.
1991 On the VSO character of Hebrew. Pp. 103–11 in Jongeling et al. 1991.
Jongeling, K.; Murre-Van den Berg, H. L.; and van Rompay, L., eds.
1991 Studies in Hebrew and Aramaic syntax: Presented to Professsor J. Hoftijzer
on the occasion of his sixty-fifth birthday. SSLL 17. Leiden: Brill.
Joosten, Jan
1989 The predicative participle in Biblical Hebrew. ZAH 2: 128–59.
1992 Biblical Hebrew “weqātāl” and Syriac “hwā qātel” expressing repetition in
the past. ZAH 5: 1–14.
2002 Do the finite verbal forms in Biblical Hebrew express aspect? JANES 29:
49–70.
Joüon, Paul
1947 Grammaire de l’hébreu biblique. 2nd ed. Rome: Pontifical Biblical Institute.
1991 A grammar of Biblical Hebrew, trans. and rev. T. Muraoka. 2 vols. Subsidia
Biblica 14. Rome: Pontifical Biblical Institute.
Karttunen, Lauri
1974 Presupposition and linguistic context. TheorLing 1: 181–94. Repr., pp. 406–
15 in Davis 1991.
Keenan, Elinor Ochs, and Schieffelin, Bambi B.
1976 Topic as a discourse notion. Pp. 337–84 in Li 1976.
Khan, Geoffrey
1988 Studies in Semitic syntax. London Oriental Series 38. Oxford: Oxford Univer-
sity Press.
Kinberg, Naphtali
1985 Adverbial clauses as topics in Arabic: Adverbial clauses in frontal position
separated from their main clauses. Jerusalem Studies in Arabic and Islam 6:
353–416.
Kintsch, Walter
1998 Comprehension: A paradigm for cognition. Cambridge: Cambridge Univer-
sity Press.
Knott, Alistair
2001 Semantic and pragmatic relations and their intended effects. Pp. 127–51 in
Sanders et al. 2001.
Kogut, Simcha
1986 On the meaning and syntactical status of ‫ הנה‬in Biblical Hebrew. Pp. 133–54
in Studies in Bible 1986, ed. Sara Japhet. Scripta Hierosolymitana 31. Jerusa-
lem: Magnes.
1996 ‫[ טעמי ’רק‘ למיעוט – כיצד‬The excluding biblical ‫רק‬: Its syntactical usages as
reflected in its accentuation]. Leš 59: 203–6.
König, Friedrich Eduard
1897 Historisch-kritischen Lehrgebaudes des hebräischen, part 3: Historisch-
comparative Syntax der hebräischen Sprache. Leipzig: Hinrichs.
Korpel, Marjo C. A., and Oesch, Josef M., eds.
2003 Unit delimitation in Biblical Hebrew and Northwest Semitic. Pericope: Scrip-
ture as Written and Read in Antiquity 4. Assen: Van Gorcum.

183 References
Kotzé, Robert J.
1989 The circumstantial sentence—a catch-them-all term? A study in sentence re-
lationships in 1 Samuel 1–12. JNSL 15: 109–26.
Kropat, Arno
1909 Die Syntax des Autors der Chronik verglichen mit der seiner Quellen: Ein
Beitrag zur historischen Syntax des Hebraeischen. Beihefte zur Zeitschrift
für die alttestamentliche Wissenschaft 16. Giessen: Alfred Topelmann.
Kugel, James
1981 The idea of biblical poetry: Parallelism and its history. New Haven, CT: Yale
University Press.
Kuno, Susumu
1972 Functional sentence perspective: A case study from Japanese and English.
LingInq 3: 269–320.
Kuroda, S.-Y.
1972 The categorical and the thetic judgment: Evidence from Japanese syntax.
Foundations of Language 9: 153–85.
Kutscher, Edward Yechezkel
1982 A history of the Hebrew language, ed. Raphael Kutscher. Jerusalem: Magnes.
Kuzar, Ron
1989 ‫[ מבנה המסר של המשפט בעברית ישראלית‬Message structure of the sentence in
Israeli-Hebrew]. Ph.D. diss. Hebrew University.
2002 ‫[ תבנית החג” ם הפשוטה בלשון המיוצגת כמדוברת‬The simple impersonal con-
struction in texts represented as colloquial Hebrew]. Pp. 329–52 in ‫מדברים‬
‫ לחקר הלשון המדוברת והשונות הלשונית בישראל‬:‫[ עברית‬Speaking Hebrew:
Studies in the spoken language and in linguistic variation in Israel], ed.
Shlomo Izre'el and Margalit Mendelson. Teudah 18. Tel Aviv: Tel Aviv Uni-
versity Press.
Labov, William
1972 Language in the inner city: Studies in the Black English vernacular. Philadel-
phia: University of Pennsylvania Press.
Labuschagne, Casper J.
1966 The emphasizing particle “gam” and its connotations. Pp. 193–203 in Stu-
dia Biblica et Semitica: Theodoro Christiano Vriezen qui munere professoris
theologiae per XXV annos functus est, ab amicis, collegis, discipulis dedi-
cata. Wageningen: Veenman.
1973 The particles ‫ הן‬and ‫הנה‬. Pp. 1–14 in Syntax and meaning: Studies in Hebrew
syntax and biblical exegesis, ed. C. J. Labuschagne et al. OTS 18. Leiden:
Brill.
Ladd, D. Robert
1980 The structure of intonational meaning: Evidence from English. Bloomington:
Indiana University Press.
Lambdin, Thomas O.
1971 Introduction to Biblical Hebrew. New York: Scribners.

References 184
Lambert, Mayer
1972 Traité de grammaire hébraïque, ed. Gérard E. Weil. 2nd ed. Publications de
l’Institut de Recherche et d’Histoire des Textes: Section Biblique et Massoré-
tique: Collection Massorah Series 3. Hildesheim: Gerstenberg.
Lambrecht, Knud
1994 Information structure and sentence form: Topic, focus, and the mental repre-
sentations of discourse referents. Cambridge Studies in Linguistics 71. Cam-
bridge: Cambridge University Press.
2000 When subjects behave like objects: An analysis of the merging of S and O
in sentence-focus constructions across languages. Studies in Language 24:
611–82.
Lambrecht, Knud, and Michaelis, Laura A.
1998 Sentence accent in information questions: Default and projection. Linguistics
and Philosophy 21: 477–544.
Lerner, Yoel
1975 ‫[ עיון מחודש בשאלת ההבחנה בין מושא לתיאור‬A reexamination of the question
of the distinction between complement and adverbial]. Leš 40: 148–51.
Levinsohn, Stephen H.
1987 Textual connections in Acts. SBLMS 31. Atlanta: Scholars Press.
1990 Unmarked and marked instances of topicalization in Hebrew. Work Papers of
SIL, University of North Dakota Session 34: 21–33.
Levinson, Stephen C.
1983 Pragmatics. Cambridge Textbooks in Linguistics. Cambridge: Cambridge
University Press.
Lewis, David
1979 Scorekeeping in a language game. Journal of Philosophical Language 8:
339–59. Repr., pp. 416–27 in Davis 1991.
Li, Charles
1976 Subject and topic. New York: Academic Press.
1986 Direct speech and indirect speech: A functional study. Pp. 29–45 in Coulmas
1986.
Li, Charles N., and Thompson, Sandra A.
1976 Subject and topic: A new typology of language. Pp. 459–89 in Li 1976.
Lode, Lars
1984 Postverbal word order in Biblical Hebrew: Structure and function. Semitics 9:
113–64.
1988 Postverbal word order in Biblical Hebrew: Structure and function part 2. Se-
mitics 10: 24–39.
Longacre, Robert E.
1979 The paragraph as a grammatical unit. Pp. 115–34 in Givón 1979.
1982 Discourse typology in relation to language typology. Pp. 457–84 in Text
processing: Text analysis and generation, text typology and attribution. Pro-
ceedings of Nobel symposium 51, ed. Sture Allen. Data Linguistica 16. Stock-
holm: Almqvist & Wiksell.
1989 Two hypotheses regarding text generation and analysis. Discourse Processes
12: 413–60.

185 References
1992 The analysis of preverbal nouns in Biblical Hebrew narrative: Some overrid-
ing concerns. JTT 5: 208–24.
1994 Weqatal forms in Biblical Hebrew prose. Pp. 50–98 in Bergen 1994.
1995 Left shifts in strongly VSO languages. Pp. 331–54 in Downing and Noonan
1995.
2003 Joseph: A story of divine providence. A text theoretical and textlinguistic
analysis of Genesis 37 and 39–48. 2nd ed. Winona Lake, IN: Eisenbrauns.
Lowery, Kirk E.
1999 Relative definiteness and the verbless clause. Pp. 251–72 in Miller 1999.
Lundbom, Jack R.
1999 Jeremiah 1–20: A new translation with introduction and commentary. An-
chor Bible 21a. New York: Doubleday.
Lunn, Nicholas P.
2006 Word-order variation in Biblical Hebrew poetry: Differentiating pragmatics
and poetics. Paternoster Biblical Monographs. Milton Keynes: Paternoster.
Lyons, John
1977 Semantics. 2 vols. Cambridge: Cambridge University Press.
Macdonald, J.
1975 Some distinctive characteristics of Israelite spoken Hebrew. BibOr 32: 162–
75.
Malbim, Meir Loeb ben Jeḥiel Michael
1964 ‫[ לב מלכים‬Heart of Kings]. In ‫ מקראות גדולות עם פירוש מלבים‬:‫נביאים וכתובים‬
[The Prophets and Writings], vol. 3: ‫[ מלכים‬Kings]. Jerusalem. Based on the
1897–99 Lublin ed.
1973 ‫[ אילת השחר‬Hind of the dawn]. In ,‫חמשה חומשי תורה עם תרגום אונקלוס‬
‫ בעל הטורים והתורה והמצוה‬,‫ שפתי חכמים‬,‫[ פירש“י‬The Pentateuch], vol. 3:
‫[ ויקרא‬Leviticus]. Jerusalem. Based on the 1880 Warsaw ed.
Mali, Uziel
1983 ‫[ לשון השיחה בנביאים ראשונים‬The language of conversation in the former
prophets]. Ph.D. diss. Hebrew University.
Mallinson, Graham, and Blake, Barry J.
1981 Language typology: Cross-linguistic studies in syntax. North-Holland Lin-
guistic Series 46. New York: North-Holland.
Martin, W. J.
1968–69 “Dischronologized” narrative in the Old Testament. Pp. 179–86 in Con-
gress volume: Rome 1968. VTSup 17. Leiden: Brill.
Mathesius, Vilem
1975 A functional analysis of present day English on a general linguistic basis, ed.
Josef Vachek. Janua Linguarum Series Practica 208. The Hague: Mouton.
Matthews, P. H.
1981 Syntax. Cambridge Textbooks in Linguistics. Cambridge: Cambridge Univer-
sity Press.
McCarthy, Dennis J.
1980 The uses of w ehinnēh in Biblical Hebrew. Bib 61: 330–42.
McCawley, James D.
1970 English as a VSO language. Lang 46: 286–99.

References 186
McEvenue, Sean E.
1971 The narrative style of the priestly writer. Analecta Biblica 50. Rome: Pontifi-
cal Biblical Institute.
1974 The style of a building instruction. Sem 4:1–9.
McNally, Louise
1998 On the linguistic encoding of information packaging instructions. Pp. 161–83
in Culicover and McNally 1998.
Merwe, Christo H. J. van der
1989 The vague term “emphasis.” JSem 1: 118–32.
1990 The old Hebrew particle ‘gam’: A syntactic-semantic description of gam in
Gn–2Kg. Philosophische Fakultat. Arbeiten zu Text und Sprache im Alten
Testament 34. Munchener Universitatsschriften. St. Ottilien: EOS.
1991 The function of word order in Old Hebrew, with special reference to cases
where a syntagmeme precedes a verb in Joshua. JNSL 17: 129–44.
1999a The elusive biblical Hebrew term “vayehi”: A perspective in terms of its syn-
tax, semantics, and pragmatics in 1 Samuel. HS 40: 83–14.
1999b Explaining fronting in Biblical Hebrew. JNSL 25: 173–86.
1999c Towards a better understanding of Biblical Hebrew word order. Review essay
of Gross 1996. JNSL 25: 277–300.
Merwe, Christo H. J. van der, and Talstra, Eep
2002–3 Biblical Hebrew word order: The interface of information structure and for-
mal features. ZAH 15–16: 68–107.
Merwe, Christo H. J. van der; Naude, Jackie A.; and Kroeze, Jan H.
1999 A Biblical Hebrew reference grammar. Biblical Languages: Hebrew 3. Shef-
field: Sheffield Academic Press.
Meyer, Rudolph, ed.
1972 Hebräische Grammatik, vol 3: Satzlehre. 3rd ed. Sammlung Göschen 5765.
Berlin: de Gruyter.
Miller, Cynthia L.
2003 The representation of speech in Biblical Hebrew narrative: A linguistic anal-
ysis. Harvard Semitic Monographs 55. Atlanta: Scholars Press.
Miller, Cynthia L., ed.
1999 The verbless clause in Biblical Hebrew: Linguistic approaches. Linguistic
Studies in Ancient West Semitic 1. Winona Lake, IN: Eisenbrauns.
Mirsky, Aharon
1977 Stylistic device for conclusion in Hebrew. Sem 5: 9–23.
1999 ‫[ סגנון עברי‬Hebrew style]. Jerusalem: Bialik Institute.
Mithun, Marianne
1992 Is basic word order universal? Pp. 15–61 in Payne 1992.
Moshavi, Adina
2006 The discourse functions of object/adverbial-fronting in Biblical Hebrew. Pp.
231–45 in Biblical Hebrew in its northwest Semitic setting: Typological and
historical perspectives, ed. Steven E. Fassberg and Avi Hurvitz. Publication
of the Institute for Advanced Studies 1. Jerusalem: Magnes / Winona Lake,
IN: Eisenbrauns.

187 References
2007a ‫[ הטופיקליציה עברית של המקרא‬Topicalization in Biblical Hebrew]. Leš 69:

7–30.
2007b ‫ הלא‬as a discourse marker of justification in Biblical Hebrew. HS 48: 171–86.
2007c Syntactic evidence for a clausal adverb ‫ הלא‬in Biblical Hebrew. JNSL 33:
51–63.
2009 ‫[ הקדמת הפוקוס בעברית מקראית‬Focus preposing in Biblical Hebrew]. Leš 71:
35–55.
forthcoming Rhetorical question or assertion? The pragmatics of ‫ הלא‬in Biblical
Hebrew. JANES 32.
Muilenberg, James
1953 A study in Hebrew rhetoric: Repetition and style. Pp. 97–111 in Congress
volume: Copenhagen, 1953. VTSup 1. Leiden: Brill.
1961 The linguistic and rhetorical usages of the particle ‫ כי‬in the Old Testament.
HUCA 32: 135–60.
Müller, August
1888 Outlines of Hebrew syntax, ed. and trans. James Robertson. Glasgow: Macle-
hose. [Translation of Part 3 of Hebräische Schulgrammatik. Halle: Max Nie-
meyer. 1878]
Munro, Pamela
1982 On the transitivity of “say” verbs. Pp. 301–18 in Studies in transitivity, ed.
Paul J. Hopper and Sandra A. Thompson. Syntax and Semantics 15. New
York: Academic Press.
Muraoka, T.
1979 On verb complementation in Biblical Hebrew. VT 29: 425–35.
1985 Emphatic words and structures in biblical Hebrew. Jerusalem: Magnes.
1990 ‫[ הפסוק השמני בלשון המקרא המאוחרת ובלשון חז“ל‬The nominal clause in late
Biblical Hebrew and Mishnaic Hebrew]. Language Studies 4: 219–52.
1991 The Biblical Hebrew nominal clause with a prepositional phrase. Pp. 143–51
in Jongeling et al. 1991.
Myhill, John
1985 Pragmatic and categorial correlates of VS word order. Lingua 66: 177–200.
1992a Word order and temporal sequencing. Pp. 265–78 in Payne 1992.
1992b Typological discourse analysis: Quantitative approaches to the study of lin-
guistic function. Oxford: Blackwell.
1995 Non-emphatic fronting in Biblical Hebrew. TheorLing 21/2–3: 93–144.
Myhill, John, and Xing, Zhiqun
1993 The discourse functions of Patient fronting: A comparative study of Biblical
Hebrew and Chinese. Ling 31: 25–57.
Naudé, Jacobus A.
1990 A syntactic analysis of dislocations in Biblical Hebrew. JNSL 16: 115–30.
1999 Syntactic aspects of co-ordinate subjects with independent personal pro-
nouns. JNSL 25: 75–99.
Niccacci, Alviero
1987 A neglected point of Hebrew syntax: Yiqtol and position in the sentence.
Liber Annus 37: 7–19.

References 188
1990 The syntax of the verb in Classical Hebrew prose, trans. W. G. E. Watson.
JSOTSup 86. Sheffield: JSOT Press.
1994 The Stele of Mesha and the Bible: Verbal system and narrativity. Or 63: 226–
48.
1997 Basic facts and theory of the Biblical Hebrew verb system in prose. Pp. 167–
202 in Wolde 1997.
Nir, Raphal, and Roeh, Itzhak
1984 ‫פתיחה של פריטי חדשות ברדיו‬-‫[ היבטים רטוריים של מיקוד במשפטי‬Rhetorical
aspects of topicalization in opening sentences of radio news items]. Balšanut
Ivrit Ḥapašit 22: 25–45.
O’Connor, M.
1980 Hebrew verse structure. Winona Lake, IN: Eisenbrauns.
2002 Discourse linguistics and the study of Biblical Hebrew. Pp. 17–42 in Con-
gress volume: Basel 2001, ed. A. Lemaire. VTSup 92. Leiden: Brill.
Oosten, Jeanne van
1985 The nature of subjects, topics and agents: A cognitive explanation. Bloom-
ington: Indiana University Linguistics Club.
Paran, Meir
1989 ‫ מבנים‬,‫ שימושי לשון‬,‫ דגמים‬:‫[ דרכי הסגנון הכוהני בתורה‬Forms of the Priestly
style in the Pentateuch: Patterns, linguistic usages, syntactic structures].
Publication of the Perry Foundation for Bbiblical Research in the Hebrew
University of Jerusalem. Jerusalem: Magnes.
Partee, Barbara H.
1973 The syntax and semantics of quotation. Pp. 410–18 in A Festschrift for Mor-
ris Halle, ed. Stephen R. Anderson and Paul Kiparsky. New York: Holt, Rine-
hart, & Winston.
Payne, Doris L., ed.
1992 Pragmatics of word order flexibility. TSL 22. Amsterdam: Benjamins.
Payne, Geoffrey
1991 Functional sentence perspective: Theme in Biblical Hebrew. JSOT 1: 62–82.
Peretz, Yitzhak
1967 ‫[ משפט הזיקה בעברית לכל תקופותיה‬The relative clause]. Dvir: Tel Aviv.
Polak, Frank H.
1999 The oral and the written: Syntax, stylistics and the development of biblical
prose narrative. JANES 26: 59–105.
2001 The style of the dialogue in biblical prose narrative. JANES 28: 53–95.
2003 Style is more than the person: Sociolinguistics, literary culture and the dis-
tinction between written and oral narrative. Pp. 38–103 in Young 2003.
Polinsky, Maria
1999 Review essay on Lambrecht 1994. Lang 75: 567–82.
Polzin, Robert
1976 Late Biblical Hebrew: Toward an historical typology of Biblical Hebrew
prose. Harvard Semitic Monographs 12. Missoula, MT: Scholars Press
Pope, Emily N.
1972 Questions and answers in English. Ph.D. diss. Massachusets Institute of
Technology.

189 References
Postal, Paul M.
1971 Cross-over phenomena. Transatlantic Series in Linguistics. New York: Holt,
Rinehart, & Winston.
Prince, Ellen F.
1981 Toward a taxonomy of given-new information. Pp. 223–55 in Radical prag-
matics, ed. Peter Cole. New York: Academic Press.
1985 Fancy syntax and “shared knowledge.” JPrag 9: 65–81.
1986 On the syntactic marking of presupposed open propositions. Pp. 208–22 in
Papers from the parasession on pragmatics and grammatical theory, at the
twenty-second regional meeting, Chicago Linguistic Society, ed. Anne M.
Farley, Peter T. Farley, and Karl-Erik McCullough. Chicago: Chicago Lin-
guistic Society.
1988 On pragmatic change: The borrowing of discourse functions. JPrag 12: 505–
18.
1992 The ZPG letter: Subjects, definiteness and information-status. Pp. 295–325 in
Discourse description: Diverse linguistic analyses of a fund-raising text, ed.
William C. Mann and Sandra A. Thompson. Pragmatics and Beyond n.s. 16.
1998 On the limits of syntax, with reference to left-dislocation and topicalization.
Pp. 281–302 in Culicover and McNally 1998.
Qimron, Elisha
1998 ‫[ כתובות עבריות חדשות וייחודי לשונן‬New Hebrew inscriptions: Their linguis-
tic contribution]. Leš 61: 181–85.
Quirk, Randolph; Greenbaum, Sidney; Leech, Geoffrey; and Svartvik, Jan
1985 A comprehensive grammar of the English language. London: Longman.
Ravid, Dorit
1977 ‫[ מספר היבטים של בעית סדר המרכיבים בעברית ישראלית מודרנית‬Several aspects
of the problem of the order of elements in general Israeli Hebrew]. Balšanut
Ivrit Ḥapašit 11: 1–45.
Reckendorf, H.
1895–98 Die syntaktischen Verhaltnisse des Arabischen. 2 vols. Leiden: Brill.
Redeker, Gisela
1990 Ideational and pragmatic markers of discourse structure. JPrag 14: 367–81.
Regt, L. J. de
1988 A parametric model for syntactic studies of a textual corpus, demonstrated on
the Hebrew of Deuteronomy 1–30. SSN 24. Assen: Van Gorcum.
1991 Word order in different clause types in Deuteronomy 1–30. Pp. 152–71 in
Jongeling, Murre-van den Berg, and van Rompay 1991.
2006 Hebrew syntactic inversions and their literary equivalence in English: Robert
Alter’s translations of Genesis and 1 and 2 Samuel. JSOT 30: 287–314.
Reinhart, Tanya
1981 Pragmatics and linguistics: An analysis of sentence topics. Phil 27: 53–94.
Repr., Bloomington: Indiana University Linguistics Club, 1982.
1984 Principles of gestalt perception in the temporal organization of narrative
texts. Ling 22: 779–809.

References 190
Rendsburg, Gary
1980 Late Biblical Hebrew and the date of P. JANES 12: 65–80.
1986 The redaction of Genesis. Winona Lake, IN: Eisenbrauns.
1990 Diglossia in ancient Hebrew. American Oriental Series 72. New Haven, CT:
American Oriental Society.
1991 The strata of Biblical Hebrew. JNSL 17: 81–99.
Revell, E. J.
1989a The conditioning of word order in verbless clauses in Biblical Hebrew. JSS
34: 1–24.
1989b The system of the verb in standard Biblical prose. HUCA 60: 1–37.
1992 Masoretic Text. Pp. 598–99 in vol. 4 of The Anchor Bible dictionary, ed.
David N. Freedman et al. New York: Doubleday.
1999 Thematic continuity and the conditioning of word order in verbless clauses.
Pp. 297–319 in Miller 1999.
Rochemont, Michael S.
1986 Focus in generative grammar. SIGLA 4. Amsterdam: Benjamins.
Rooker, Mark F.
1990a Ezekiel and the typology of Biblical Hebrew. HAR 12: 133–55.
1990b Biblical Hebrew in transition: The language of the book of Ezekiel. JSOTSup
90. Sheffield: JSOT Press.
Rosén, Haim B.
1965 ‫פועל בעברית המקראית‬-‫[ על משפטים חסרי‬On some types of verbless sentences
in Biblical Hebrew]. Pp. 167–73 in Third World Congress of Jewish studies:
Jerusalem, 25th July–1st August 1961. Jerusalem: World Union of Jewish
Studies.
1977 Contemporary Hebrew. Trends in Linguistics: State-of-the-art Reports 11.
The Hague: Mouton.
1982 ‫[ אספקטים בחקר סדר חלקי המשפט בעברית הישראלית הכתובה‬Aspects in the
study of sentence order in literary Israeli Hebrew]. Pp. 43–49 in Proceedings
of the eighth world congress of Jewish studies, Jerusalem, August 16–21,
1981, vol. 4: Division D: The Hebrew language and languages of the Jews;
Jewish folklore and art. Jerusalem: World Union of Jewish Studies.
Rosenbaum, Michael
1997 Word-order variation in Isaiah 40–55: A functional perspective. SSN 35.
Assen: Van Gorcum.
Ross, John R.
1967 Constraints on variables in syntax. Ph.D. diss. Massachusets Institute of
Technology. Repr., Bloomington: Indiana University Linguistics Club, 1968.
Sáenz-Badillos, Angel
1993 A history of the Hebrew language, trans. John Elwolde. Cambridge: Cam-
bridge University Press.
Sanders, Ted J.
1997 Semantic and pragmatic sources of coherence: On the categorization of co-
herence relations in context. DiscProc 24: 119–47.

191 References
Sanders, Ted J., and Noordman, Leo G.

2000 The role of coherence relations and their linguistic markers in text processing.
DiscProc 29: 37–60.
Sanders, Ted; Schilperoord, Joost; and Spooren, Wilbert, eds.
2001 Text representation: Linguistic and psycholinguistic aspects. Human Cogni-
tive Processing 8. Amsterdam: Benjamins.
Sanders, Ted, and Spooren, Wilbert
2001 Text representation as an interface between language and its users. Pp. 1–25
in Sanders et al. 2001.
Sanders, Ted J; Spooren, Wilbert P.; and Noordman, Leo G.
1992 Toward a taxonomy of coherence relations. DiscProc 15: 1–35.
Sappan, Raphael
1976 ‫[ הכיאזם בשירת המקרא‬Chiasmus in Biblical poetry]. Beit Mikra 21: 534–39.
1981 ‫[ הייחוד התחבירי של לשון השירה המקראית‬The typical features of the syntax of
biblical poetry in its classical period]. Jerusalem: Kiryat Sefer.
Sarna, Nahum M.
1989 Genesis: The traditional Hebrew text with the new JPS translation. JPS Torah
Commentary. Philadelphia: Jewish Publication Society.
Savran, George W.
1988 Telling and retelling: Quotation in Biblical narrative. Indiana Studies in Bib-
lical Literature. Bloomington: Indiana University Press.
Schiffrin, Deborah
1987 Discourse Markers. Studies in Interactional Sociolinguistics 5. Cambridge:
Cambridge University Press.
Schlesinger, K.
1953 Zur Wortfolge im hebräischen Verbalsatz. VT 3: 381–90.
Schmerling, Susan F.
1976 Aspects of English sentence. Austin: University of Texas Press.
Schmidt-Radefeldt, Jürgen
1977 On so-called “rhetorical” questions. JPrag 1: 376–77.
Schoors, Antoon
1981 The particle “kî.” Pp. 240–76 in Remembering all the way: A collection of
Old Testament studies published on the occasion of the fortieth anniversary
of the Oudtestamentisch Werkgezelschap in Nederland, ed. B. Albrektson.
OTS 21. Leiden: Brill.
Schourup, Lawrence
1999 Discourse markers: Tutorial overview. Lingua 107: 227–65.
Selkirk, Elisabeth O.
1995 Sentence prosody: Intonation, stress and phrasing. Pp. 550–69 in The hand-
book of phonological theory, ed. John A. Goldsmith. Blackwell Handbooks
in Linguistics 1. Cambridge: Blackwell.
Shimasaki, Katsuomi
2002 Focus structure in Biblical Hebrew: A study of word order and information
structure. Bethesda, MD: CDL.

References 192
Shopen, Timothy
1985 Language typology and syntactic description. 3 vols. Cambridge: Cambridge
University Press.
Shulman, Ahouva
1996 The use of modal verb forms in Biblical Hebrew prose. Ph.D. diss. University
of Toronto.
Siewierska, Anna
1988 Word order rules. Croom Helm Linguistics Series. London: Helm.
1991 Functional grammar. Linguistic Theory Guides. London: Routledge.
Singer, Murray
1990 Psychology of language: An introduction to sentence and discourse pro-
cesses. Hillsdale: Erlbaum.
Sivan, Daniel, and Schniedewind, William
1993 Letting your “yes” be “no” in ancient Israel: A study of the asseverative ‫לא‬
and ‫הלא‬. JSS 38: 209–26.
Skinner, John
1930 A critical and exegetical commentary on Genesis. 2nd ed. International Criti-
cal Commentary. Edinburgh: T. & T. Clark.
Snyman, F. P. J.
2004 The scope of negative lōʾ in Biblical Hebrew. Acta Academica Supplemen-
tum 3. Blomfontein: Acta Academica.
Snyman, F. P. J., and Naudé, J. A.
2003 Sentence and constituent-negation in Biblical Hebrew. JSem 12: 237–67.
Stalnaker, Robert C.
1974 Pragmatic presuppositions. Pp. 197–213 in Semantics and philosophy, ed.
Milton K. Munitz and Peter K. Unger. New York: New York University Press.
Repr., pp. 471–81 in Davis 1991.
Stec, D. M.
1987 The use of hēn in conditional sentences. VT 37: 478–86.
Steiner, Richard C.
1979 Review of Blau 1977. AfroLing 6: 5–10.
1992 A colloquialism in Jer. 5:13 from the ancestor of Mishnaic Hebrew. JSS 37:
11–26.
1997 Ancient Hebrew. Pp. 145–73 in The Semitic Languages, ed. Robert Hetzron.
Routledge Language Family Descriptions. London: Routledge.
1998 Saadia vs. Rashi: On the shift from meaning-maximalism to meaning-mini-
malism in medieval biblical lexicology. JQR 88: 213–58.
2000 Does the Biblical Hebrew conjunction -‫ ו‬have many meanings, one meaning,
or no meaning at all? JBL 119: 249–67.
Sweetser, Eve
1990 From etymology to pragmatics: Metaphorical and cultural aspects of seman-
tic structure. Cambridge Studies in Linguistics 54. Cambridge: Cambridge
University Press.
Taglicht, Josef
1984 Message and emphasis: On focus and scope in English. English Language
Series 15. London: Longman.

193 References
Tannen, Deborah
1986 Introducting constructed dialogue in Greek and American conversational and
literary narrative. Pp. 311–32 in Coulmas 1986.
Thompson, Sandra A.
1978 Modern English from a typological point of view: Some implications of the
function of word order. LingBer 54: 19–35.
Tov, Emanuel
1992 Textual criticism of the Hebrew Bible. Minneapolis: Fortress. [Revised and
enlarged edition of ‫ פרקי מבוא‬:‫ביקורת נוסח המקרא‬. The Biblical Encyclopedia
Library 4. Jerusalem: Bialik Institute, 1989]
Ulrich, Eugene
2003 Impressions and intuition: Sense divisions in ancient manuscripts of Isaiah.
Pp. 279–307 in Korpel and Oesch 2003.
Vallduví, Enric
1992 The informational component. Outstanding Dissertations in Linguistics. New
York: Garland.
Vallduví, Enric, and Engdahl, Elisabet
1996 The linguistic realization of information packaging. Ling 34: 459–519.
Vallduví, Enric, and Vilkuna, Maria
1998 On rheme and contrast. Pp. 79–108 in Culicover and McNally 1998.
Walker, Marilyn A.; Joshi, Aravind K.; and Prince, Ellen F.
1998a Centering in naturally occurring discourse: An overview. Pp. 1–28 in Walker,
Joshi, and Prince 1998b.
1998b Centering Theory in Discourse. Oxford: Clarendon.
Waltke, Bruce K., and O’Connor, M.
1990 An introduction to Biblical Hebrew syntax. Winona Lake, IN: Eisenbrauns.
Ward, Gregory L., and Prince, Ellen F.
1991 On the topicalization of indefinite NPs. JPrag 16: 167–77.
Watson, Wilfred G. E.
1984 Classical Hebrew poetry: A guide to its techniques. JSOTSup 26. Sheffield:
JSOT Press.
Williams, Ronald J.
1976 Hebrew syntax: An outline. 2nd ed. Toronto: University of Toronto Press.
Wimsatt, W. K.
1967 Style as meaning. Pp. 362–73 in Chatman and Levin 1967.
Winther-Nielsen, Nicolai
1992 “In the beginning” of Biblical Hebrew discourse: Genesis 1:1 and the fronted
time expression. Pp. 67–80 in Language in context: Essays for Robert E.
Longacre, ed. Shin ja J. Hwang and William R. Merrifield. SIL and the Uni-
versity of Texas at Arlington Publications in Linguistics 107. Arlington: SIL
and the University of Texas at Arlington.
Wolde, Ellen van
1997a Linguistic motivation and Biblical exegesis. Pp. 21–50 in Wolde 1997b.
1997b Narrative syntax and the Hebrew Bible: Papers of the Tilburg conference
1996. Biblical Interpretation Series 29. Leiden: Brill.

References 194
Young, Ian, ed.

2003 Biblical Hebrew: Studies in chronology and typology. JSOTSup 369. New
York: Continuum.
Zatelli, Ida
1994 Analysis of lexemes from a conversational prose text: “hnh” as signal of a
performative utterance in 1 Sam. 25:41 ZAH 7: 5–11.
Zevit, Ziony
1982 Converging lines of evidence bearing on the date of P. ZAW 94: 481–511.
1998 The anterior construction in classical Hebrew. SBLMS 50. Atlanta: Scholars
Press.
Zewi, Tamar
1992 ‫ההסבות התחביריות הכרוכות במבנה הפונקציונלי של המשפט בעברית מקראית‬
[Syntactical modifications reflecting the functional structure of the sentence
in Biblical Hebrew]. Ph.D. diss. Hebrew University.
1994 The nominal sentence in Biblical Hebrew. Pp. 145–67 in Semitic and Cushitic
Studies, ed. Gideon Goldenberg and Shlomo Raz. Weisbaden: Harrassowitz.
1996 The particles ‫ הנה‬and ‫ והנה‬in Biblical Hebrew. HS 37: 21–37.
1997 On similar syntactical roles of inūma in El Amarna and ‫והנה‬, ‫הנה‬, and ‫ הן‬in
Biblical Hebrew. JANES 25: 71–86.
1999 Interrupted syntactical structures in Biblical Hebrew. ZAH 12: 83–95.
2004 Review of Heller 2004. HS 45: 298–302.
Ziv, Yael
1988 Word order in children’s literature: FSP and markedness. Pp. 123–44 in The
Prague School and its legacy in linguistics, literature, semiotics, folklore, and
the arts, ed. Yishai Tobin. Linguistic and Literary Studies in Eastern Europe
27. Amsterdam: Benjamins.
1996a Discourse grounding: A constraint on preposing. Pp. 168–93 in IATL 3: The
proceedings of the eleventh annual Conference, Tel Aviv University 1995 and
of the Workshop on Discourse, the Technion 1996, ed. Edit Doron and Shuly
Wintner. N.p.
1996b Review of Lambrecht 1994. JPrag 26: 699–710.

Index of Authors
Abadi, A. 102 Brown, G. 33

Aejmelaeus, A. 54, 60 Brown, M. L. 70
Alter, R. 4 Buth, R. xvii, 11, 16, 17, 28, 29, 30, 38,
Andersen, F. I. 2, 17, 19, 20, 22, 151 39, 40, 86
Andrews, A. 62, 98
Ariel, M. 95 Carnie, A. 7, 10
Assif, S. 22 Cassuto, U. 20
Azar, M. 31, 62 Chafe, W. 31, 36, 41, 93, 95, 108
Chavel, C. D. 22
Bailey, N. A. 30, 34, 39, 40, 42 Childs, B. S. 4
Bandstra, B. L. 11, 19, 37, 54, 60 Chomsky, N. 33, 35, 43
Bar-Asher, E. xvii Claassen, W. T. 54
Bar-Asher, M. xvii Clark, H. H. 93
Battistella, E. L. 7 Collins, J. C. 30
Ben Asher, B. 10 Comrie, B. 9
Ben-Asher, M. 62 Cook, J. A. 29, 30, 38
Bendavid, A. 20, 31, 36, 142 Creider, C. A. 33, 34
Ben-Ḥayyim, Z. 70 Croft, W. 1, 7, 8
Ben-Horin, G. 31, 33, 34
Bergsträsser, G. 12, 22 Davidson, A. B. 10, 17, 18, 20, 21, 23
Berlin, A. 2, 143 Dawson, D. A. 28
Berman, R. A. 7 DeCaen, V. xvii, 11, 12, 13, 14, 16,
Birner, B. J. 34, 35, 97, 98, 99, 100, 17, 38
101, 103, 148 Dempster, S. G. 29
Blake, B. J. 9 Dijk, T. A. van 32, 102
Blakemore, D. 103 Dik, S. C. 32, 38, 39, 127, 128
Blau, J. xvi, 12, 17, 20, 54, 62, 68, 77, Disse, A. 40, 81
84, 140, 141, 142, 158 Downing, P. 8
Bloch, A. 2, 26 Driver, S. R. 4, 20, 21, 22, 24, 30, 82,
Bloch, A. A. 82 84
Bolkestein, A. M. 32, 39 Dryer, M. S. 8, 42, 91, 92, 93, 94, 95,
Bravmann, M. M. 82, 84 108, 136
Breuer, Y. xvii, 128 Dyk, J. W. 88
Brockelmann, C. 11, 18, 24, 82
Brody, J. 8, 9 Engdahl, E. 32
Brongers, H. A. 71 Erteschik-Shir, N. 36
195

Index of Authors 196
Esh, S. 60 Heimerdinger, J.-M. 18, 28, 29, 41, 42,

Eskhult, M. 21, 24, 28 43, 44, 127
Exter Blokland, A. F. den 25 Heller, R. L. 4, 24, 27, 28, 30
Ewald, H. 10, 18, 20, 21 Hirschberg, J. L. 98
Hockett, C. F. 32
Fassberg, S. E. xvii Hoftijzer, J. 2, 17
Firbas, J. 32, 34 Holes, C. 143
Floor, S. J. 2, 33, 42 Holmstedt, R. D. xvi, xvii, 11, 12, 14,
Fokkelman, J. P. 4 15, 16, 38, 42
Follingstad, C. M. 54, 61, 139 Hoop, R. de 25
Forbes, A. D. 2 Hopper, P. J. 27
Fox, A. 22, 25 Horn, L. R. 136
Fraser, B. 102 Hornkohl, A. 12, 13
Freedman, D. N. 2 Huddleston, R. D. 7, 48, 49, 51, 54, 62,
Friedman, R. E. 3 69, 76, 136
Huehnergard, J. xvii, 53
Garr, W. R. xvii, 70, 158 Hurvitz, A. xvii, 3, 5
Ginsberg, H. L. 20
Giora, R. 31 Ibn Ezra, A. 10
Givón, T. 11, 22, 24, 27, 28, 29, 32, 34, Ilie, C. 128
36, 97
Glinert, L. 31 Jackendoff, R. 35, 136
Goldenberg, G. 60, 76 Jacobs, L. 158
Goldfajn, T. 29 Jakobsen, R. 3, 142
Gómez-González, M. A. 32, 33 Johnstone, B. 143
Gordis, R. 20 Jong, J. de 39
Goutsos, D. 33 Jongeling, B. 24
Graesser, A. C. 95 Jongeling, K. 11, 12
Greenberg, J. H. 1, 7, 8 Joosten, J. xvii, 2, 17, 20, 21, 38
Greenstein, E. L. xvii, 3 Joshi, A. K. 32
Gregory, M. L. 100 Joüon, P. 3, 11, 23, 25, 26, 45, 46, 53,
Grimes, J. E. 27 54, 55, 56, 60, 67, 82, 119
Gross, W. 40, 41, 42, 68, 73, 78, 79,
81, 82 Kaddari, M. Z. 22
Guilfoyle, E. 7, 10 Karttunen, L. 93
Gundel, J. K. 32, 33, 35, 36, 93 Keenan, E. O. 32
Gussenhoven, C. 43 Khan, G. 19, 20, 24, 25, 82, 96, 150,
151
Hackett, J. A. xvii Kinberg, N. 84
Halliday, M. A. K. 31, 32, 34, 35, 43, Kintsch, W. 93, 94, 95, 96
102, 155 Knott, A. 102, 103
Hannay, M. 39 Kogut, S. 70, 73
Hasan, R. 102, 155 König, F. E. 10, 18, 22, 23, 26
Hatav, G. 28, 30, 61 Kotzé, R. J. 22
Haviland, S. E. 93 Kropat, A. 3
Hedberg, N. 36, 93 Kugel, J. 3, 152

197 Index of Authors
Kuno, S. 43 Moshavi, A. 71, 117

Kuroda, S. Y. 43 Muilenberg, J. 54, 143
Kutscher, E. Y. 3, 22 Müller, A. 10, 18, 20, 21, 22
Kuzar, R. 31 Munro, P. 60
Labov, W. 3 Muraoka, T. 3, 11, 12, 13, 17, 18, 19,
Labuschagne, C. J. 70, 73, 159 20, 21, 22, 25, 26, 35, 45, 46,
Ladd, D. R. 43 53, 54, 55, 56, 60, 62, 67, 70, 73,
Lambdin, T. O. 24, 39 82, 119
Lambert, M. 11, 18, 82 Myhill, J. 19, 29, 34, 97
Lambrecht, K. 32, 34, 35, 36, 41, 42, Naudé, J. A. 67, 82, 137
43, 44, 45, 90, 91, 92, 93, 100 Niccacci, A. 19, 28, 58
Leibowitz, N. 22
Lerner, Y. 62 Nir, R. 31
Levinsohn, S. H. 30, 34, 39, 98 Noordman, L. G. 102
Levinson, S. C. 1, 59
Lewis, D. 108 O’Connor, M. P. 1, 2, 7, 8, 11, 19, 30,
Li, C. N. 60 37, 39, 51, 54, 57, 62, 67, 68, 70,
Lode, L. 18 71, 76, 82, 118, 137, 143, 152
Longacre, R. E. 4, 8, 24, 25, 27, 28, Oosten, J. van 32
29, 58
Lowery, K. E. 88 Paran, M. 20, 140, 141, 142
Lundbom, J. R. 25 Partee, B. H. 60
Lunn, N. P. xvi, 2, 42, 169 Payne, G. 28
Lyons, J. 59 Peretz, Y. 10
Polak, F. H. 4, 5, 169
Macdonald, J. 4, 26 Polinsky, M. 41
Malbim, M. L. ben J. M. 10, 20, 22, 25 Polzin, R. 3
Mali, U. 4, 26, 59 Pope, E. N. 116, 128
Mallinson, G. 9 Postal, P. M. 33
Martin, W. J. 30 Prince, E. F. xvi, 32, 33, 35, 91, 93, 94,
Mathesius, V. 32, 35 97, 98, 99, 100, 101
Matthews, P. H. 62 Pullum, G. K. 7, 48, 49, 51, 54, 69, 136
McCarthy, D. J. 70
McCawley, J. D. 9 Qimron, E. xvii, 45
McEvenue, S. E. 20 Quirk, R. 60, 62, 69
McNally, L. 31
Merwe, C. H. J. van der 3, 19, 37, 40, Rashi 22
41, 42, 43, 44, 45, 57, 58, 68, Ravid, D. 7
73, 80 Reckendorf, H. 82
Meyer, R. 11, 18 Redeker, G. 102
Michaelis, L. A. 92, 100 Regt, L. J. de 2, 10, 11, 12, 41
Miller, C. L. xvii, 2, 3, 4, 5, 55, 59, Reinhart, T. 28, 32
60, 61 Rendsburg, G. 3, 4, 5
Mirsky, A. 24 Revell, E. J. 5, 10, 12, 14, 17, 18, 37, 38
Mithun, M. 8, 9 Richter, W. 40
Mor, U. xvii, 24 Rochemont, M. S. 35

Index of Authors 198
Roeh, I. 31 Taglicht, J. 36
Rooker, M. F. 3 Talstra, E. 42, 44, 45, 88
Rosenbaum, M. 2, 18, 38, 39, 127 Tannen, D. 59, 61
Rosén, H. B. 17, 31 Thompson, S. A. 9
Ross, J. R. 33, 82 Tov, E. 5, 25
Sáenz-Badillos, A. 4 Ulrich, E. 25
Sanders, T. J. 94, 102, 155
Sappan, R. 2, 19 Vallduví, E. 31, 32, 36
Sarna, N. M. 4 Vilkuna, M. 36
Savran, G. W. 59
Schieffelin, B. B. 32 Walker, M. A. 32
Schiffrin, D. 102 Waltke, B. K. 11, 30, 39, 51, 54, 57, 62,
Schlesinger, K. 11, 12 67, 68, 70, 71, 76, 82, 118, 137
Schmerling, S. F. 43 Ward, G. L. 34, 35, 97, 98, 99, 100,
Schmidt-Radefeldt, J. 128 101, 103, 148
Schneider, W. 28 Watson, W. G. E. 2, 142
Schniedewind, W. 70 Weil, H. 31
Schoors, A. 54, 60 Weinrich, H. 28
Schourup, L. 102 Williams, R. J. 22, 23, 60
Selkirk, E. O. 43 Wimsatt, W. K. 142
Shimasaki, K. 10, 18, 41, 42, 43, 44, 45 Winther-Nielsen, N. 39
Shulman, A. 38 Wolde, E. J. van 28
Siewierska, A. 9, 13, 39
Singer, M. 94, 95 Xing, Z. 19
Sivan, D. 70
Skinner, J. 4 Yule, G. 33
Snyman, F. P. J. 137
Spooren, W. P. 94, 102 Zacharski, R. 36, 93
Stalnaker, R. C. 93 Zatelli, I. 70
Stec, D. M. 159 Zevit, Z. 3, 22
Steiner, R. C. xv, xvi, 3, 4, 10, 19, 25, Zewi, T. 17, 28, 70, 82
51, 53, 56, 57, 58, 62, 70, 71, 84, Ziv, Y. xvii, 8, 31, 34, 86, 99
109, 139
Sweetser, E. 102

Index of Scripture
Genesis Genesis (cont.) Genesis (cont.)

1:1 39 5:4 30 10:30 30
1:4 147 6:3 118 10:32 147, 150
1:5 65, 67, 150 6:6–8 157 11:2 66
1:10 150, 156 6:9 67, 88 11:4 57
1:20 161 6:16 81 11:12 161
1:22 65, 161 6:17 130 11:27 31, 150, 161,
1:27 141, 150 6:18 67 163
1:29 70 6:21 65 12:6 79
2:5 53 7:2 148, 152, 156 12:7 108
2:10 149, 155 7:17 30 12:12 150, 156
2:16 153, 157 7:18–19 155 12:13 56, 60
2:16–17 151, 156 7:19 150 12:15–16 157
2:25 30 7:20 129 12:16 147
3:1 115 7:21–22 131 12:18 128
3:2 152 7:23 30 13:5 80
3:11 79 8:5 150, 163 13:7 79
3:12 82 8:13 57 13:11–12 152
3:13 128 8:13–14 153 13:12 147, 156
3:14 153 8:17 65, 105 13:14 115
3:16 147 8:18–19 134 13:42 67
3:22 74 8:19 134 14:4 31, 152
3:24 30 9:2 67, 105 14:4–6 154
4:1 22, 75, 115 9:3 121, 129 14:21 65, 156
4:2 147, 150, 156 9:5 124, 129, 134 14:22–23 52
4:4 81 9:19 147, 150 14:23 65, 119
4:4–5 20, 151 9:20 30 14:24 83
4:5 147, 156 9:23 150 15:4 72, 82, 83, 96
4:7 71 10–11 155 15:10 66, 156
4:8 57, 58 10:8 65, 66, 161 15:12 114
4:10 128 10:9 65 15:15 65
4:14 58, 70 10:10–11 146 15:17 31, 114
4:18 31, 62, 161 10:11 149, 155 15:18–21 108
4:20 150 10:21 81 16:1 24, 76, 115
4:22 81 10:24 31 16:2 117
4:26 62, 81 10:25 31 16:14 88
4:42 46 10:26–29 31 17:9 67
199

Index of Scripture 200
Genesis (cont.) Genesis (cont.) Genesis (cont.)

17:10–12 130 21:23 52 27:30–31 135
17:14 83 21:24 65, 118 27:33 88
17:15 55 21:26 74, 76, 162 27:34 84
17:17 79 21:30 60 27:35 45
17:18 53 22:1 58, 114 27:37 67, 74, 78
17:20 69, 124, 129 22:4 84 27:42 62
17:20–21 157 22:5 65, 67, 145 27:44 56
17:27 134 22:13 5 28:13 82
18:1 21 22:16 109 28:17 78
18:5 88 22:17 54 29:9 113
18:7 31, 105 22:20 80 29:12 59
18:8 51 22:20–23 150, 163 29:16–17 147
18:13 21, 116 23:3–6 123, 126 29:17 66, 157
18:15 72 23:6 65 29:25 59, 108, 117,
18:17 115 23:11 129 128
18:17–18 117 23:15 24, 65 29:32 60
18:20 54 24:7 82, 83, 108 30:1 55
18:25 71, 79 24:8 73, 75 30:10 5
18:26 52 24:31 21, 116 30:15 72
18:28 70 24:37–38 72, 124 30:16 125, 133
18:33 148, 150, 161 24:38 67, 131 30:21 51
19:2 72, 74, 122, 132 24:45 76 30:34 53
19:3 24, 148 24:49 77 31:1 45
19:4 68 24:50 108 31:15 71, 74, 108,
19:6 31, 155 24:53 145, 150, 161 117
19:8 73 24:56 116 31:16 65, 110
19:11 68 24:60 65 31:19 112
19:12 65 24:61–62 115 31:26 128
19:15 54 25:5–6 150, 156 31:32 65
19:17 65 25:10 67, 88, 149, 31:33–34 113
19:20 117 155 31:38 67, 152, 162
19:23 23 25:32 117 31:39 150, 163
19:23–24 114 25:33–34 25, 150, 31:40 152, 161
19:35 80 161 31:42 52, 72, 88
19:38 81 26:10 128 31:43 78
20:4 76, 112 26:15 82 31:47 144, 147, 150
20:5 67, 70, 117, 129 26:25–26 115 31:52 52
20:9 123, 128 26:26 68 32:1–2 161
20:11 73 26:27 21, 116, 118 32:12 116
20:13 65 26:29 52 32:18–19 77
20:15 65, 127 27:2–3 117 32:20 109
20:15–16 146 27:4 57 32:28–29 140
21:1 115 27:6–8 117 33:3 65
21:5 62 27:15–16 149 33:6 67
21:6 108 27:30 114 33:14 65, 157

201 Index of Scripture
Genesis (cont.) Genesis (cont.) Genesis (cont.)

33:16–17 150 41:12 129, 141 44:28 70
34:5 115 41:13 20, 152 44:33 65
34:9 152, 160 41:16 121, 132 45:5–8 137
34:11 88, 110 41:28 88 45:8 76, 137, 139,
34:16 160 41:40 150, 162 140
34:17 77 41:44 147 45:14 161
34:21 144, 152 41:50 150 45:16 45
34:25–26 134 41:51–52 30, 150, 45:17 65, 151
34:26 105 156 45:17–18 146, 162
34:27–28 121, 129 41:54 158 45:19 65, 162
34:29 49, 105 41:55 88, 110 45:20 65
34:31 79 42:2 117 45:22 158
35:6 67 42:3–4 157 46:4 150, 160
35:10 72, 75, 132 42:7 60 46:7 105
35:11 67, 124, 129 42:7–9 132 46:31 46
35:12 82, 133, 135 42:15 52, 149 46:32 67, 155
35:18 145, 150, 156 42:15–16 146, 162 47:1 46, 67
36:1–3 163 42:16 54, 65 47:2–4 125
36:4 147 42:18 65 47:4 71
36:13 88, 149, 163 42:18–20 146, 162 47:5 67
36:14 150 42:19 65, 77, 152 47:6 57, 65, 127
36:16–17 150 42:21 71 47:9 88
37:3 114 42:22–23 114 47:19 148
37:11 66, 150, 151 42:29–30 45 47:30 118
37:13 118 42:33 65, 152 48:2 46
37:20 45 42:33–34 146, 162 48:5 81
37:27 65 42:36 153, 162 48:9–10 114
37:33 45 43:9 147, 150, 163 48:11 146, 157
37:36 115 43:10 54, 72 48:19 71, 74, 75
38:17 118 43:11 65 49:28 129
38:22 72 43:11–13 127, 134, 50:14 67
38:24 73 146, 162 50:20 151, 156
38:25 23, 113 43:12 65, 127 50:21 118
38:29 78 43:13 65 50:22 67
39:1 115 43:15 67, 105 50:24 157
39:8 74, 162 43:16 51
39:19 109 44:2 68, 127, 134 Exodus
40:12–13 151, 163 44:3 67, 113 1:12 56
40:13 150 44:4–5 117 1:22 152
40:14 72 44:5 117 3:9 72
40:18–19 163 44:8 74, 158 3:12 60
40:19 66, 150 44:9 84, 162 4:14 73
40:21–22 147, 150, 44:12 145, 152 4:25 60
161 44:17 65, 82, 83, 96 5:15 80
41:11 129 44:22 56 6:2–5 72

Exodus (cont.) Numbers (cont.) Joshua (cont.)

6:12 74, 159 22:20 109 14:14 56
7:11 80 22:28 128 15:16 83
8:24 118 22:29 72 16:8 24
9:16 74 22:30 80 22:20 117
10:6 51 22:34 77 24:27 74
11:7 55 22:38 80
12:16 83 23:12 71 Judges
12:34 76 23:26 109 1:1–2 126
19:5 77 26:33 140 1:2 128
19:8 110 28:9–29:38 155 1:16 119
21:28 52 2:29 128
22:29 164 Deuteronomy 4:14 117
23:12 164 1:30 83 6:14 117
24:3 109 1:31 55 6:18 41, 118
31:17 164 1:36 96 6:29 128
32:32 77 4:7–8 44 7:13 74
32:34 74, 84 4:13–14 44 8:2 80
33:13 77 4:32 79 8:6 80
35:2 164 4:34 79 8:23 132
36:10 134 5:3 139 9:16 77
7:3 160 9:38 80
Leviticus 7:5 41, 80, 160 11:1 30
4:12 140, 141 9:9 41 11:7 80, 117, 118
10:19 74 12:20 53 11:24 71
12:2–3 164 15:12 164 13:2 46
15:28–29 164 17:17 24, 134 13:6 46
19:23–24 164 19:8 54 13:23 52
21:14 72, 139 22:7 50 14:15 79
23:11 140, 141 24:19 128 14:16 74, 159
26:29 134 24:20 128 15:2 118
26:36 83 24:21 128 15:11 118
28:67 44 15:11–12 126
Numbers 32:21 137 16:23 45
6:3 134, 137 16:24 45
7:12–83 154 Joshua 16:31 22
9:3 128 1:16 109 17:1 45
9:6 45 2:2 46, 74 17:10 118
10:30 72, 131 2:4–6 113 18:3 113
11:20 55, 117 2:6 112 18:22 113
11:22 79 2:8 76, 113 19:22 74
16:15 138 5:14 72 20:44 62
16:28–29 138 6:3–4 164 21:7 116
20:12 55 10:26 79 21:21 77
22:5 74 11:6 160
22:11 85 11:9 160

203 Index of Scripture
1 Samuel 1 Samuel (cont.) 1 Kings (cont.)

1:1 45 28:2 74 12:17 83
1:27 128 28:9 74, 117 13:1 46
2:15 140 28:16 116 15:9 164
3:7 76 28:18 74 15:13 83
3:17 53 29:8 128 18:18 140
4:17 72 30:1 112 20:17 41
6:14 149 20:18 10
8:5 74 2 Samuel 20:39 113
8:7 131 1:6 74 21:6 60
8:11–17 160 2:24 113 21:7 80, 118
8:19 72 3:23 45 21:13 45
9:1 45 10:5 51 21:14 45
9:5 113 11:12 164 22:14 96
9:11 113 12:1 45 22:16 55
9:12 80 12:3 160 22:18 140
9:15 112 12:18 45
9:17 23 12:27 30 2 Kings
9:21 118 13:35 74 1:6 41, 73, 83
10:2 45 14:3 109 2:18 118
10:19 131 15:10 45 3:19 160
14:29 45 15:35 79 4:4 24
14:29–30 159 16:11 159 6:5 62
14:30 72 17:6 109 6:31 53
15:6 116 18:4 110 6:32 117
15:17 71 18:22 78 7:3 45
16:1 116 18:29 78 8:7 45
16:5 125 19:8 109 8:13 60
17:20 84 19:22 71, 79 9:13 45
17:29 80, 117 19:38 74 10:4 159
17:34 62 20:21 72 10:5 110
17:54 149 21:14 79 10:9 74
18:15 55 21:20 45 10:15 60
18:24 109 24:17 74 11:3–4 164
19:3 78 17:36 73, 83, 96
19:13 149 1 Kings 17:40 131
20:1 128 1:5 119 19:11 74
20:2 117 1:13 117 23:11 160
20:29 77 1:41 22 25:3–8 164
23:3 159 6:22 134
23:27 46 6:37–38 164 Job
24:8 79 8:27 74, 159 1:1 45
24:10 74 8:65–66 164 11:2 71
25:27 85 11:13 73 14:4 71
26:15 118 11:41 73
26:18 128 12:4 80

Psalms Qohelet Jeremiah

22:24 24 5:1 10 1:14 103
23:2–3 24 19:7 159 22:5 109
49:13 109
Proverbs Isaiah 52:6 164
15:11 159 45:23 109
19:10 159 Haggai
2:13 71

AdinaMoshavi 2010 WordOrderintheBiblicalHebrewFiniteClause

Uploaded by

Copyright:

Available Formats

AdinaMoshavi 2010 WordOrderintheBiblicalHebrewFiniteClause

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AdinaMoshavi 2010 WordOrderintheBiblicalHebrewFiniteClause

Uploaded by

Copyright:

Available Formats

Copyright 2010. Eisenbrauns. All rights reserved.

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 1 9/28/10 7:36 AM

M. O’Connor† and Cynthia L. Miller

1. The Verbless Clause in Biblical Hebrew: Linguistic Approaches,

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 2 9/28/10 7:36 AM

Winona Lake, Indiana

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 3 9/28/10 7:36 AM

Library of Congress Cataloging-in-Publication Data

Moshavi, A. Mosak (Adina Mosak)

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 4 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb v 9/28/10 7:36 AM

00-Moshavi-LSAWS4.indb vi 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb vii 9/28/10 7:36 AM

3.3. The information-structure model . . . . . . . . . . . . . . . . . . . . . . 31

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb viii 9/28/10 7:36 AM

5.4. Complex variants of the preposing construction . . . . . . . . . . 80

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb ix 9/28/10 7:36 AM

7.3.2. The residue of subject-preposed direct speech

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb x 9/28/10 7:36 AM

Table 1. Word-order constructions in finite nonsubordinate

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xi 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xii 9/28/10 7:36 AM

JAOS Journal of the American Oriental Society

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xiii 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xiv 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xv 9/28/10 7:36 AM

In writing this book, I made considerable eﬀort to present linguistic con-

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xvi 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb xvii 9/28/10 7:36 AM

00-Moshavi-FrontMatter.indd vi 9/28/10 2:05 PM

1.1. The pragmatics of preposing

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 1 9/28/10 7:36 AM

and unmarked constructions, and the distribution of these constructions in BH

1.2. The corpus for the study

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 2 9/28/10 7:36 AM

prose or poetry, while recognizing that diﬀerences of opinion exist regarding

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 3 9/28/10 7:36 AM

EBSCOhost - printed on 8/25/2023 2:20 PM via . All use subject to https://www.ebsco.com/terms-of-use

00-Moshavi-LSAWS4.indb 4 9/28/10 7:36 AM

1.3. Overview of the chapters