Professional Documents
Culture Documents
BOSTON
2003
CHAN-14-schiff.qxd 17/12/2002 14:05 Page iii
This book is printed on acid-free paper.
Library of Congress Cataloging-in-Publication Data
Semitic papyrology in context / edited by Lawrence A. Schiffman.
p. cm. (Culture and history of the ancient Near East ; v. 14)
Papers given at a conference on March 5 7, 2000 at New York University.
Includes bibliographical references and index.
ISBN 9004128859 (hc.)
1. Manuscripts, Semitic (Papyri) Congresses. 2. Paleography, Semitic Congresses. I.
Schiffman, Lawrence H. II. Series.
PJ3091 .S45 2003
492'.0411 dc2l 2002032271
Die Deutsche Bibliothek CIP-Einheitsaufnahme
Semitic Papyrology in Context: A Climate of Creativity, Papers from a New York
University conference marking the retirement of Baruch A. Levine /
ed. by Lawrence H. Schiffman -- Leiden ; Boston : Brill, 2003
(Culture and History of the Ancient Near East ; Vol. 14)
ISBN 90 04 12885 9
ISSN 1566-2055
ISBN 90 04 12885 9
Copyright 2003 by Koninklijke Brill NV, Leiden, The Netherlands
All rights reserved. No part of this publication may be reproduced, translated, stored
in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.
Authorization to photocopy items for internal or personal use is granted
by Koninklijke Brill provided that the appropriate fees are paid directly
to The Copyright Clearance Center, 222 Rosewood Drive,Suite 910
Danvers, MA 01923, USA.
Fees are subject to change.
printed in the netherlands
CHAN-14-schiff.qxd 17/12/2002 14:05 Page iv
CONTENTS
Preface ........................................................................................ vii
Abbreviations .............................................................................. ix
Ancient Egyptian ScriptsLiterary, Sacred, and Profane
Oorrx Gorrr+, Jn. ................................................................ 1
The Samaria Papyri and the Babylonio-Aramean Symbiosis
Dotor.s M. Gnorr ................................................................ 23
Elephantine and the Bible
Brz.rrr Pon+rx .................................................................... 51
The Corpus of the Qumran Papyri
Ev.xtrr To\ ........................................................................ 85
The Roman Census in the Papyri from the Judaean Desert
and the Egyptian kat okan pograf
H.xx.n M. Co++ox .............................................................. 105
The Language of Power: Latin in the Inscriptions of
Iudaea/Syria Palaestina
Wrnxrn Eck .......................................................................... 123
Oral Establishment of Dowry in Jewish and Roman Law:
hrymab ynqnh yrbd and dotis dictio
R.xox K.+zorr .................................................................... 145
Witnesses and Signatures in the Hebrew and Aramaic
Documents from the Bar Kokhba Caves
L.vnrxcr H. Scnirrv.x ...................................................... 165
The Roman Near East: The View from Below
Fn.xk E. Pr+rns .................................................................... 187
The Decipherment and Edition of the Petra Papyri:
Preliminary Observations
Ltrvio Korxrx (In Collaboration with R.Ch.
Caldwell, R.W. Daniel, and T. Gagos) .............................. 201
An Early Arabic Legal Papyrus
Grorrnrv Kn.x .................................................................... 227
The Voice of the Jewish Poor in the Cairo Genizah
M.nk R. Conrx .................................................................... 239
v
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page v
vi cox+rx+s
Epilogue
B.ntcn A. Lr\ixr ................................................................ 257
Index of Ancient Sources Cited ................................................ 275
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page vi
vii
PREFACE
On 57 March 2000, a conference was held at New York University,
sponsored by the Skirball Department of Hebrew and Judaic Studies
as part of the Edelman Lectures, entitled A Climate of Creativity:
Semitic Papyrology in Context. The conference was planned to
mark the retirement of Professor Baruch A. Levine after thirty years
at New York University. Levine had been honored previously by a
Festschrift,
1
and so it was decided upon his retirement to organize a
conference that would revolve around the work he was then engaged
in, and which is now published, the edition of the Naal ever
papyri rst discovered by Yigael Yadin.
2
As was characteristic of
Levine, this work had led him into far reaching, but related areas
of Semitic and Greek papyrology, and it was thought that a con-
ference on this subject would be an appropriate way of honoring
him on the occasion of his retirement.
The conference was meant to accomplish two major goals: First,
we wanted to advance the cross-cultural aspect of research in this
eld, emphasizing specically the relation of the various collections
of Semitic papyri one to another and to the better known Greek
material. Second, we wanted to encourage and foster personal rela-
tionships and scholarly cooperation between scholars working on
these diverse but in many ways related materials. The days we spent
together more than justied the eort, as readers will see when they
read the various contributions to this volume. A summary of the
results of the conference and some personal reections by Baruch
Levine close the volume, and indicate the extent to which his own
work in this eld helped to form the conference agenda.
I wish to thank the contributors to this volume for their cooper-
ation in the preparation of their presentations and papers. Andrew
1
R. Chazan, W.W. Hallo, L.H. Schiman (eds), Ki Baruch Hu: Ancient Near Eastern,
Biblical and Judaic Studies in Honor of Baruch A. Levine (Winona Lake, IN: Eisenbrauns,
1999).
2
Y. Yadin, J.C. Greeneld, A. Yardeni, and B.A. Levine, The Documents from the
Bar Kokhba Period in the Cave of Letters: Hebrew, Aramaic and Nabatean-Aramaic Papyri
( JDS 3; Jerusalem: Israel Exploration Society; Institute of Archaeology, Hebrew
University; Shrine of the Book, Israel Museum, 2002).
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page vii
viii rnrr.cr
Gross, Research Assistant at New York University, prepared the vol-
ume for publication and compiled the index of sources with his usual
combination of both technical and scholarly acumen. Hans van der
Meij and Patricia Radder of Brill Academic Publishers were kind
enough to publish the volume, and in so doing to show once again
Brills commitment to advancing scholarship on the ancient world
and its complex cultural interactions.
Lawrence H. Schiman
3 June 2002
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page viii
ix
ABBREVIATIONS
For full bibliographic information on papyrus publications, the reader
is often referred to J.F. Oates, R.S. Bagnall and others, Checklist of
Editions of Greek, Latin, Demotic and Coptic Papyri, Ostraca and Tablets
(BASPSup 9; Oakville, CN: American Society of Papyrologists, 2001
5
);
and in permanently updated form:
<http://scriptorium.lib.duke.edu/papyrus/texts/clist.html>.
AASOR Annual of the American Schools of Oriental Research
ABAW Abhandlungen der Bayerischen Akademie der Wissen-
schaften
ABD D.N. Freedman (ed.), Anchor Bible Dictionary (New York:
Doubleday, 1992).
ADAJ Annual of the Department of Antiquities of Jordan
AE LAnne pigraphique
AfO Archiv fr Orientforschung
AGJU Arbeiten zur Geschichte des antiken Judentums und des
Urchristentums
AIPHOS Annuaire de lInstitut de philologie er dhistoire orientales et slaves
ALAD G. Khan, Arabic Legal and Administrative Documents in the
Cambridge Genizah Collections (CUL Genizah series 10; Cam-
bridge: Cambridge University Press, 1993).
ANRW Aufstieg und Niedergang der rmischen Welt
APBH W. Diem, Arabische Briefe auf Papyrus und Papier aus der
Heidelberger Papyrus-Sammlung: Textband (Wiesbaden: Otto
Harrasowitz, 1991).
APK G. Khan, Bills, Letters and Deeds: Arabic Papyri of the 7th11th
Centuries (London: Nour Foundation in association with
Azimuth Editions and Oxford University Press, 1993).
APEL A. Grohmann, Arabic Papyri in the Egyptian Library (Cairo:
Egyptian Library Press, 193474).
ARU J. Kohler and A. Ungnad, Assyrische Rechtsurkunden in
Umschrift und Uebersetzung nebst einem Index der Personen-
namen und Rechtserlauterungen (Leipzig: E. Pfeier, 1913).
AS Assyriological Studies
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page ix
x .nnnr\i.+ioxs
ASAW Abhandlungen der Schsischen Akademie der Wissen-
schaften
BASOR Bulletin of the American Schools of Oriental Research
BASPSup Bulletin of the American Society of Papyrologists: Sup-
plement
BAR Biblical Archeological Review
BE The Babylonian Expedition of the University of Penn-
sylvania
BGU Aegyptische Urkunden aus den Kniglichen (later Staatlichen)
Museen zu Berlin, Griechische Urkunden (Berlin: Weidmann,
18951912) [see Checklist].
Bib Biblica
BIDR Bullettino dellIstituto di Diritto romano
BM British Museum
BN Biblische Notizen
BRM A.T. Clay, Babylonian Records in the Library of J. Pierpont
Morgan (New York, 191223).
CAD The Assyrian Dictionary of the Oriental Institute of the University
of Chicago (Chicago: Oriental Institute, 1956).
Camb J.N. Strassmaier, Inschriften von Cambyses, Knig von Babylon
(529521 v. Chr.) von den Thontafeln des Britischen Museums
(Leipzig: Eduard Pfeier, 1890).
CdE Chronique dgypte, Bulletin priodique de la Fondation gyp-
tologique Reine lisabeth
ChLA Chartae Latinae Antiquiores
CIL Corpus Inscriptionum Latinarum
CJ Codex Justinianus
COS W.W. Hallo and K.L. Younger (eds), The Context of
Scripture (Leiden: Brill, 19972002).
CPA A. Grohmann and R.G. Khoury, Chrestomathie de papyro-
logie arabe: documents relatifs la vie prive, sociale et adminis-
trative dans les premiers siecles islamiques (HO. Erste Abteilung,
Nahe und der Mittlere Osten. Erganzungsband II, 2;
Leiden/New York/Kln: Brill, 1993).
CPL (= C.Pap.Lat.) R. Cavenaile (ed.), Corpus Papyrorum Latinarum
(Wiesbaden, 1958) [see Checklist].
CRAI Compte Rendu de lAcadmie des inscriptions et belle-lettres
CTN Cuneiform Texts from Nimrud
CUL Cambridge University Library
DAWW Denkschriften der Kaiserlichen Akademie der Wissen-
schaften in Wien, Phil. Hist. Klasse. Vienna
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page x
.nnnr\i.+ioxs xi
DC The Drower Collection of Mandean manuscripts in the
Bodleian Library, Oxford
DE Discussions in Egyptology
DJD Discoveries in the Judaean Desert
DJPA M. Sokolo, Dictionary Of Jewish Palestinian Aramaic (Ramat-
Gan: Bar Ilan University Press, 1990).
DNP H. Cancik and H. Schneider (eds), Der neue Pauly: Enzy-
klopdie der Antike: Altertum (Stuttgart: J.B. Metzler, 1996).
Drower & E.S. Drower and R. Macuch, A Mandaic Dictionary (Oxford:
Macuch Clarendon Press, 1963).
ENA Elkan Adler Collection at the Jewish Theological Seminary
of America
ErIsr Eretz Israel
FIRA Fontes Iuris Romani Antejustiniani [see Checklist].
GAG W. von Soden, Grundriss der akkadischen Grammatik (Analecta
orientalia 33; Rome: Ponticium Institutum Biblicum,
1995
3
).
GM Gttinger Miszellen: Beitrge zur gyptologische Diskussion.
Gttingen
GOF Gttinger Orientforschungen
HALAT L. Koehler, W. Baumgartner, and J.J. Stamm, Hebrisches
und aramisches Lexikon zum Alten Testament (Leiden: Brill,
196795).
HO Handbuch der Orientalistik
HSM Harvard Semitic Monographs
IEJ Israel Exploration Journal
IFAO Papyrus grecs de lInstitut Franais dArchologie Orientale. Cairo.
(Institut Franais dArchologie Orientale du Caire. Bib-
liothque dtude) [cf. P.IFAO in Checklist]
IG Inscriptiones Graecae
ILS H. Dessau, Inscriptiones Latinae Selectae (Berlin: Weidmann,
18921916).
IstMitt Istanbuler Mitteilungen
JAC Jahrbuch fr Antike und Christentum
JAOS Journal of the American Oriental Society
Jastrow M. Jastrow, A Dictionary of the Targumim, the Talmud Babli
and Yerushalmi, and the Midrashic Literature (New York: G.P.
Putnam, 1903).
JBL Journal of Biblical Literature
JCS Journal of Cuneiform Studies
JDS Judean Desert Studies
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page xi
xii .nnnr\i.+ioxs
JEA Journal of Egyptian Archaeology
JJP Journal of Juristic Papyrology
JNES Journal of Near Eastern Studies
JQR Jewish Quarterly Review
JRA Journal of Roman Archaeology
JRASup Journal of Roman Archaeology Supplement Series
JRAS Journal of the Royal Asiatic Society
JRS Journal of Roman Studies
JSOTSup Journal for the Study of the Old Testament Supplement
Series
JSJ Journal for the Study of Judaism
JSS Journal of Semitic Studies
KAJ E. Ebeling, Keilschrifttexte aus Assur juristischen inhalts (Leipzig:
J.C. Hinrichs, 1927).
KT Kleine gyptische Texte
KTU M. Dietrich, O. Loretz, and J. Sanmartin, Keilalphabetischen
Texte aus Ugarit (Alte Orient und Altes Testament 24/1;
Neukirchen-Vluyn: Neukirchener Verlag, 1976). 2nd
enlarged edition: The Cuneiform Alphabetic Texts from Ugarit,
Ras Ibn Hani and Other Places (KTU) (Abhandlungen zur
Literatur Alt-Syrien-Palastinas und Mesopotamiens 8;
Mnster: Ugarit-Verlag, 1995).
Le Leshonenu
LexSyr C. Brockelmann, Lexicon Syriacum (Halle: Max Niemeyer,
1928
2
; reprint, Hildesheim: G. Olms, 1995).
LingAeg Lingua Aegyptia
LSJ H.G. Liddell, R. Scott, and H.S. Jones, A Greek-English
Lexicon (Oxford: Clarendon Press, 1996
9
).
MAOG Mitteilungen der Altorientalishen Gesellschaft
MDAIK Mitteilungen des Deutschen Archologischen Instituts: Abteilung
Kairo
MEF Y. Rib, Marchands dtoffes du Fayyoum au III
e
/IX
e
sicle
dapres leurs archives (actes et lettres) (Cairo: Institut francais
darcheologie orientale, 1985).
MRS Mission de Ras Shamra
Mur (= P. Murabba't) Wadi Murabba'at Papyri [see Checklist].
Nbk J.N. Strassmaier, Inschriften von Nabuchodonosor, Knig von
Babylon (604561 v. Chr.) von den Thontafeln des Britischen
Museums (Leipzig: Eduard Pfeier, 1889).
Nbn J.N. Strassmaier, Inschriften von Nabonidus, Knig von Babylon
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page xii
.nnnr\i.+ioxs xiii
(555538 v. Chr.) von den Thontafeln des Britischen Museums
(Leipzig: Eduard Pfeier, 1889).
NJPS New Jewish Publication Society translation
NovT Novum Testamentum
NRV M. San Nicol and A. Ungnad, Neubabylonische Rechts-
und Verwaltungsurkunden (Leipzig: J.C. Hinrichs, 192935).
NTOA Novum Testamentum et Orbis Antiquus
OEANE E.M. Meyers (ed.), The Oxford Encyclopedia of Archaeology in
the Near East (New York: Oxford University Press, 1997).
Or Orientalia (NS)
PAAJR Proceedings of the American Academy of Jewish Research
PIASH Proceedings of the Israel Academy of Sciences and
Humanities
P.Cair.Isid. The Archive of Aurelius Isidorus in the Egyptian Museum,
Cairo, and the University of Michigan [see Checklist].
P.Flor. Papiri greco-egizii, Papiri Fiorentini [see Checklist].
P.Hever (= Xev/Se) The Seiyl Collection [see Checklist].
P.Lond. Greek Papyri in the British Museum [see Checklist].
P.Mich. Michigan Papyri [see Checklist].
P.Oxy The Oxyrhynchus Papyri [see Checklist].
P.Ryl. Catalogue of the Greek and Latin Papyri in the John
Rylands Library, Manchester [see Checklist].
P.Tebt. The Tebtunis Papyri [see Checklist].
P.Yadin (= P.Babatha = 5/6ev) Naal ever papyri [see
Checklist].
PRU C.F.-A. Schaeer and J. Nougayrol, Le Palais royal dUgarit
(Paris: Impr. nationale, 195565).
P-S R. Payne-Smith, A Compendious Syriac dictionary: founded upon
the Thesaurus Syriacus of R. Payne Smith (Oxford: Clarendon
Press, 1903; reprint, Winona Lake, IN: Eisenbrauns, 1998).
PSI Papiri greci e latini (Pubblicazioni della Societ Italiana
per la ricerca dei papiri greci e latini in Egitto; Florence:
E. Ariani, 191229) [see Checklist].
RA Revue dAssyriologie
RB Revue Biblique
Rd Revue dgyptologie
RE J.J. Herzog (ed.), Realencyklopdie fr protestantische Theologie
und Kirche (Leipzig: J.C. Hinrichs, 18961913).
RIDA Revue internationale des droits de lantiquit
RSO Rivista degli studi orientali
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page xiii
xiv .nnnr\i.+ioxs
SAOC Studies in Ancient Oriental Civilizations
SAPKC G. Khan, Arabic Papyri: Selected Material from the Khalili Collec-
tion (London; New York: Nour Foundation in association
with Azimuth Editions and Oxford University Press, 1992).
SB Sammelbuch griechischer Urkunden aus Aegypten [see Checklist].
SBLSymS Society of Biblical Literature Symposium Series
SBLWAW Society of Biblical Literature Writings from the Ancient
World
SCI Scripta Classica Israelica
SEG Supplementum epigraphicum graecum
SJLA Studies in Jewish Law in Antiquity
STDJ Studies on the Texts of the Desert of Judah
TAD B. Porten and A. Yardeni, Textbook of Aramaic Documents
from Ancient Egypt ( Jerusalem: Hebrew University, Dept.
of the History of the Jewish People, 198999).
Tal A. Tal, A Dictionary of Samaritan Aramaic (HO; Leiden:
Brill, 2000).
TCL Textes cuniformes du Louvre
ThWAT G.J. Botterweck and H. Ringgren, Theologisches Wrterbuch
zum Alten Testament (Stuttgart: W. Kohlhammer 1970).
TLOT E. Jenni (ed.), with assistance from C. Westmann, Theo-
logical Lexicon of the Old Testament (trans. M.E. Biddle;
Peabody, MA: Hendrickson, 1997).
TS Taylor Schechter Collection at Cambridge University
TSAJ Texte und Studien zum antiken Judentum
VAB Vorderasiatische Bibliothek
VAS Vorderasiatische Schriftdenkmler
VAT Vorderasiatische Tontafelsammlung
Wb. A. Erman und H. Grapo (eds), Wrterbuch der aegyptischen
Sprache (Berlin: Akademie-Verlag, 1982
4
).
W.Chr. (= Chrest.Wilck.) L. Mitteis and U. Wilcken, Grundzge
und Chrestomathie der Papyruskunde (Leipzig-Berlin: B.G.
Teubner, 1912) [see Checklist].
WDSP Wadi Daliyeh Samaritan Papyri
WZKM Wiener Zeitschrift fr die Kunde des Morganlandes
YClS Yale Classical Studies
YOS Yale Oriental Series
ZDPV Zeitschrift des deutschen Palstina-Vereins
ZPE Zeitschrift fr Papyrologie und Epigrak
SHIFFMAN_prelims_v-xiv 12/11/02 9:05 AM Page xiv
ANCIENT EGYPTIAN SCRIPTSLITERARY,
SACRED, AND PROFANE
Ogden Goelet, Jr.
Introduction: The Egyptian Systems of Writing
Since Egypt represents the geographical as well as the chronologi-
cal margins of this conference, so to speak, if I were simply to pre-
sent a well-illustrated sketch of Egyptian writing systems, many in
this audience would probably nd that alone rather informative.
Although I shall concentrate on those scripts found on papyri, I shall
rst provide a brief overview of the Egyptian writing system as a
whole. At the same time there is another theme I wish to develop
today that rearms Egypts distinctive character, namely the close
connection that always existed between Egyptian writing and mon-
umentalitythe connection between papyrus and hard copy in its
truest and most literal meaninginscriptions on the stone surfaces
of stelae, tombs, and temples. Stone inscriptions may seem at rst
to take one far aeld from a conference on papyrology, but it is fair
to say that some parts of Egyptian writing systems were informed
by an intimate working relationship between papyrus and monumen-
tal writing. This integration of script and monument, inherent in the
strongly decorative character of the Egyptians scripts in all their
forms, was, in turn, intimately connected to their desire to make
texts and representations closely cohere as they tried to recreate the
cosmos in tombs and temples. I hope to show one aspect of this,
namely how books could act as substitutes for tombs as well as the
reverse of the situationhow a tomb could become a book.
The Terminology of Egyptian Script Forms
One of the most interesting and accurate accounts in classical liter-
ature of the Egyptian writing system comes from the early Christian
author, Clement of Alexandria, in a passage describing the training
of the literate Egyptians of the second century cr: . . . the men of
1
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 1
2 oorrx oorrr+, n.
1
Clement of Alexandria: Stromata V 4, 20, as quoted by R. Parkinson and
S. Quirke, Papyrus (The Egyptian Bookshelf; London: The British Museum Press,
1995) 29.
2
Illustrations of this object (BM EA 24) are far too common to list, but two
recent publications by the British Museum are particularly useful. For a discussion
of the discovery of the Stone, its role in the decipherment of the hieroglyphs, as
well as its subsequent treatment and conservation, see R. Parkinson, Cracking Codes:
The Rosetta Stone and Decipherment (Berkeley: University of California Press, 1999)
1245, esp. Fig. 3 on p. 18, and Color Plate 1; a line-drawing restoration of the
entire object appears in Fig. 8 on p. 26. S. Quirke and C. Andrews have produced
a brief study which presents a full-size facsimile and line drawing of the object
along with a translation of all three texts on the stela; see S. Quirke and C. Andrews,
The Rosetta Stone: Facsimile Drawing with an Introduction and Translation (New York and
London: British Museum Press, 1988).
3
The remark appears in line 14, the nal line of the preserved hieroglyphic text,
see Parkinson, Cracking Codes, 54, for an enlarged representation of this part of the
inscription, supplemented by a transliteration and translation. The Demotic version
(line 32) is similar, except that the term for the Greek language is translated as
the script of the Ionians. The Greek parallel (line 54) has tow te erow ka gxvroiw
learning among the Egyptians learnt rst of all that method of writ-
ing called the epistolographic, and second the hieratic, which the
sacred scribes use, and then, last of all, the hieroglyphic.
1
The evi-
dence is strong that this situation was essentially true throughout
Egyptian history. Actually, Clement might have taken his narrative
one step further and added Greek to this picture, because, from
about the middle of the Ptolemaic Dynasty onwards, many educated
Egyptians knew how to write Greek as well. I would like to exam-
ine Clements statement a bit more in light of what prevailed at the
time when the Rosetta Stone was written, close to four centuries
earlier.
The Rosetta Stone,
2
which should really be called the Memphis
Decree after the place where the document was probably rst com-
posed and published, is often described as a trilingual stela. This
is not quite accurate. First of all, no version of the text could be
properly described as an accurate translation of the others. Secondly,
the stela presents roughly the same text written in three dierent
scripts, but not in three languages. There are really only two unre-
lated tongues here, Greek and Egyptian, but the latter occurs in two
dialects, each in its own script. This distinction is made explicitly
in the last line of the hieroglyphic text where it commands that
copies of the decree are to be made on a hard stone stela in the
writing of the gods words (mdw nr), in the writing of letters, and
in the script of the Aegeans.
3
In fact, the scripts on the stela were
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 2
.xcirx+ rovr+i.x scnir+s 3
inscribed in that same order that Decree requiredthe hieroglyphs
on top;
4
in the middle was demotic, the script which Clement called
epistolographic; then, nally, Greek appears at the bottom. The
word demotic, the term most frequently employed, conveys well
the middle scripts extensive employment in all aspects of daily life:
letters, legal texts, decrees, and, eventually, literary documents too.
Once demotic had been introduced in the early Twenty-sixth Dynasty
c. 650 ncr, the hieratic script, which had for centuries previously
been the primary means of communication among literate Egyptians,
was thereafter rapidly relegated to what we might broadly describe
as the realm of religious and mortuary arcania: the Book of the
Dead, the so-called magico-medico papyri, hymns and prayers, etc.
In that sense, Clements term hieratic, i.e., priestly, was quite
accurate at the time he was writing. Hieratic had long since become
an esoteric form of writing used almost exclusively by priests and
scribes for copying out religious documents on papyri. In keeping
with the theme of this conference, I shall focus primarily on hier-
atic long before it became such a fossilized script, as well as on hier-
atics half-brother, so to speak, the cursive hieroglyphs.
By contrast, the term hieroglyphs had always been essentially
accuratethe translation of the Greek term, sacred writing, is a
good description of a script which may have had the widest range
of use, but was distinctly preferred when it came to religious texts,
i.e., for tombs, temples, stelae, and other inscriptional material, in the
literal sense. The term used in the Memphis Decree for the hiero-
glyphs, mdw nr the words of the god,
5
reveals much about the
Egyptians attitude towards this script and had been used to describe
the most pictorial form of the language for centuries before the
Ptolemies. It is striking that even at this late stage, there was still a
form of Egyptian that was written with rather concrete signs. Although
it would be inaccurate to call the hieroglyphs picture writing, never-
theless their connection with pictures was never lost. The pictures
ka Ellhnikow grmmasin in sacred and native and Greek characters; see Quirke
and Andrews, The Rosetta Stone, 22, with the accompanying original-size reproduc-
tion of the stela.
4
The Egyptian term used here for the hieroglyphs, mdw-nr, might be better lit-
erally rendered into Greek as theoglyphos, than as hieroglyphos. For a recon-
struction of the Rosetta Stone, see Parkinson, Cracking Codes, 26 (Fig. 8).
5
Wb. 2.180, 15181, 6; this was occasionally replaced by the expression mdt-nr
after Dyn. 18, see Wb. 2.182, 5.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 3
4 oorrx oorrr+, n.
underlying the hieroglyphs, furthermore, had a distinctly decorative
nature which meant that art and language were basically insepara-
ble in the Egyptian mind, whence the strong association between
the hieroglyphic script and monumentality which shall be another
theme in my presentation. Because of that association, the hiero-
glyphs never underwent an extensive abstraction of their underlying
forms as happened with cuneiform. Formal abstraction was conned
to hieratic and demotic. Although we have now covered the basic
relationship between the Egyptian script forms and sketched their
development, a most important factor is missing from this brief his-
torical overview.
Any attempt to explain under what circumstances one or another
of the various Egyptian script forms were used must take account
of register. Register is a term used to describe the variety of language
employed according to such social factors as class and context. For
example, the way in which people speak and write in academic dis-
course, in religious contexts, or in legal documents are all consider-
ably dierent from each other and dierent from how those same
individuals might speak in their daily lives. Each situation represents
a dierent register. Furthermore, in Egypt dierent registers of speech
had preferences for dierent scripts and, occasionally, even dierent
dialects as well. At the risk of stating the obvious, I would like to
emphasize that in the case of ancient Egypt, register is closely related
to the context in which a text is physically located as well as to the
type of surface on which it is found, that is, a tomb vs. a temple;
papyrus vs. stone stela.
6
Furthermore, the choice of script and sur-
face often combine in such as way as to provide us with valuable
insights into the Egyptians attitude towards a given text. Certainly,
such considerations played a similar role elsewhere in the ancient
Near East. An example which immediately comes to mind would
be the Prologue and the Epilogue of the Hammurapi Stela which
employ the so-called Hymnic-Epic dialect of Akkadian that is dis-
tinct from the dialect used for the laws of the Codex proper. Never-
theless, I believe that register had a much more inuential role in Egypt,
largely because of the wider varieties of script forms available.
6
For a good introduction to the inuence of context on the form of script and
other stylistic issues, see H.G. Fischer, Archaeological Aspects of Epigraphy and
Palaeography, Ancient Egyptian Epigraphy and Palaeography (H.G. Fischer and R.A.
Caminos; New York: Metropolitan Museum Press, 1979
2
) 2950.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 4
.xcirx+ rovr+i.x scnir+s 5
The Creation of Papyrus Documents
The complexity of the interrelationship between text and register is
perhaps best seen in the factors which inuenced the production of
literary, religious, and documentary papyri of the New Kingdom
when the art of writing reached its full ower. Papyrus and writing
on papyrus seem to have been Egyptian inventions, going back to
the earliest days of Egyptian historythat is, history in its very nar-
row sense of written texts. A sheet of uninscribed papyrus was
found in a First Dynasty tomb where it had presumably been put
at the deceaseds disposal for use in the afterlife. Sporadic nds
throughout the Old Kingdom attest to the fact that papyrus quickly
became the primary means of writing letters and keeping ocial
administrative documents during the Old Kingdom.
7
Thus, right at
the beginnings of Egyptian writing there developed a division between
the monumental and funerary registers, represented by hieroglyphs,
and the documentary and literary registers, represented by hieratic.
Despite the abundance of the papyrus plant in Egypt and the per-
vasive use of papyri, the Egyptians themselves have, ironically, left
no description of how it was made, perhaps because that may have
been a trade secret.
8
Several competing theories have been put forth
as to how strips of brous material were peeled away from the stalk
of the plant and treated. The details of papyrus technology as well
as the pros and cons of the various proposals are fascinating, but
shall not concern us here. What is more to the point is the man-
ner in which these strips were handled in forming sheets. First, strips
of papyrus ber were laid down on top of each other at right angles,
the join most likely relying only on the natural mucilage in the
papyrus itself. Once the sheets had been formed, dried, then bleached
naturally in the sunlight, scrolls were made by pasting sheets together
so that bers ran horizontally on the top inner surface of the roll.
9
7
During the Old Kingdom there were some rare examples of clay tablets used
for writing, see G. Soukiassian, A Governors Palace at 'Ayn Asil, Dakhla Oasis,
Egyptian Archaeology 11 (1997) 17. These rare objects were found at Balat in the
Dakhla Oasis where papyrus may have been rather hard to come by. The texts
were incised in cursive hieroglyphs by means of a wooden stylus, reminiscent of
the manner in which Mesopotamian texts were executed.
8
For a description and history of the manufacture of papyrus, see Parkinson and
Quirke, Papyrus, 23, esp. 1316.
9
Three suggestions for cutting the stem of the papyrus plant and the method of
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 5
6 oorrx oorrr+, n.
Let us now examine how register made its inuence felt. Normally,
when the Egyptians wished to write a religious or literary text, sheets
were cut out from the roll and inscribed so that the text ran hori-
zontally along with the bers. Of course, this practice varied over
the course of time. From the Old Kingdom to the late Middle King-
dom, for example, most texts were written in vertical columns run-
ning down the sheet. At all periods the script employed on papyri
was overwhelmingly hieratic. The major exception, the so-called cur-
sive hieroglyphs, I shall discuss along with my treatment of the
Book of the Dead shortly. However, for reasons we do not under-
standthere does not seem to be any practical reasona scribe
wishing to write a letter or ocial document turned his roll 90, cut
o a sheet, and then wrote his text normally in horizontal rows, but
with the bers now running vertically to the text.
10
Thus, with a
change of register from literary to ocial, the papyrus underwent a
dierent orientation of ber to writing direction at the same time.
In addition, there was an accompanying switch in dialects. Although
the dierence between the grammar and vocabulary of everyday let-
ters and that found in religious and literary texts was not particu-
larly great at rst, by the late Eighteenth Dynasty, the language of
daily business and the language of religious-literary writing had
diverged rather sharply. This distinction, moreover, was maintained
in inscribed materialstexts on stelae, tombs, and temple walls essen-
tially follow the diction of the literary-religious papyri. That register
utilized a dialect both archaic and synthetic which is conventionally
called late Middle Egyptian, a dialect whose articiality was increased
by its frequent association with hieroglyphic text, whether cursive or
fully-formed.
11
Since late Middle Egyptian was more or less divorced
from the speech of everyday life, it probably took on a quality that
the churchy and antiquated English of the King James Bible has
forming the sheets from the resulting sheets are illustrated in Parkinson and Quirke,
Papyrus, 14 Fig. 4.
10
Convenient illustrations of the relationship between the direction of writing and
the ber arrangement on papyri can be found in F. Junge, Einfhrung in die Grammatik
des Neugyptischen (Wiesbaden: Harrassowitz Verlag, 1999
2
) 2324.
11
A summary of the characteristics of the various stages of the late Middle
Egyptian and where it ts within the development of the Late Egyptian dialect in
particular appears along with a helpful table in Junge, Neugyptischen, 21. More
detailed expositions of this subject have been presented by J.-M. Kruchten, From
Middle Egyptian to Late Egyptian, LingAeg 6 (1999) 197 and K. Jansen-Winkeln,
Diglossie und Zweisprachigkeit im alten gypten, WZKM 85 (1995) 85115.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 6
.xcirx+ rovr+i.x scnir+s 7
to the ears of modern Americans. The dialect that came to be em-
ployed in ocial documents and letters is known as Late Egyptian
and was presumably fairly close to popular speech. Eventually the
gap between the language of papyri and that of stelae and monu-
ments closed considerably, with literary manuscripts, at least, adopt-
ing many features of Late Egyptian. Nevertheless, even then, literary
Late Egyptian and documentary Late Egyptian retained enough gram-
matical, orthographic, and phraseological distinctions so that, though
they became closer, they still represented clearly distinct registers.
At this juncture, one can add yet another aspect of register to our
portrait of ancient Egyptian writing, especially when dealing with
papyri. When texts are still living entities undergoing constant change,
we can speak of them as belonging to a productive stage. One
might say that the Con Texts, for example, with their many local
variants and constantly changing repertoire, always remained in the
productive stage of textual development. With the passage of time,
however, there was a pronounced tendency for some texts such as
literary classics or religious materialthe Book of the Dead, for exam-
pleto be produced primarily in the form of increasingly canonical
copies, at which point we can speak of them existing in a reproduc-
tive mode only.
12
Needless to say, the same terminology could be
applied fruitfully to texts elsewhere in the ancient Near East.
For a practical example of how such aspects of register can have
a signicant import when considering the nature of even familiar
texts, I would like to discuss a work well-known to most who have
studied the ancient Near EastThe Tale of Wenamun. This text
narrates the misfortunes of an Egyptian ocial conducting business
along the Syro-Palestian coast at the end of the Twentieth Dynasty.
The standard hieroglyphic text edition used by me and all others
teaching Late Egyptian appears in Gardiners Late Egyptian Stories.
The papyrus itself is derived from a single copy, which, on both
paleographic and archaeological grounds, is most likely to be dated
12
We owe the distinction between the productive and reproductive, which
is fundamental for our conception of the mechanisms of the transmission and can-
onization of texts, to J. Assmann, Egyptian Solar Religion in the New Kingdom: Re, Amun
and the Crisis of Polytheism (trans. A. Alcock; New York and London: Kegan Paul
International, 1995) 111. The Book of the Dead of the Saite Period, when both
the text and the order of the chapters became fairly canonical, would represent an
ideal example of the reproductive stage of a genre.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 7
8 oorrx oorrr+, n.
in the Twenty-rst Dynasty. Wenamun, however, is markedly dierent
in both its grammatical style and its choice of vocabulary from all
the rest of the works in the volume, so much so, in fact that it was
employed as a major source for the ern y-Groll, Late Egyptian Grammar,
a textbook focussed on the documentary register of Late Egyptian.
13
There is, furthermore, another aspect of this papyrus which indi-
cates even more strongly that the Tale of Wenamun is neither
ctitious nor truly literary in nature. I am referring here to the man-
ner in which the papyrus has been inscribed. As ern y pointed out,
the papyrus sheets of Wenamun, unlike those of every single other
work in Gardiners Late Egyptian Stories, have their bers running at
right angles to the text, that is, like a typical ocial document.
14
The Wenamun papyrus is, in fact, one of a group of three papyri
probably found together at el-Hibe, all evidentially the product of
a single scribal school, judging by the strong similarity in the
hieratic handwriting of these manuscripts. The other two papyri
demonstrateas the so-called Moscow Literary Letter shows
15
their
anity to the literary-religious register not only by their content and
grammar, but also by their non-documentary orientation of text and
bers. Taken together, this evidence suggests that the Tale of
Wenamun, highly colorful though the narrative may seem at points,
was in every respect conceived written out as an actual ocial report,
and, therefore, it should be credited with greater accuracy concern-
ing contemporary conditions than is sometimes done.
16
13
J. ern y and S.I. Groll, A Late Egyptian Grammar (Studia Pohl. Series Maior 4;
Rome: Pontical Biblical Institute Press, 1994
3
).
14
See the illustration of the papyri in M.A. Korostovstev, Puteshestvie Un-Amuna v
Bibl; Egipetskii ieraticheskii papirus no. 120 Gosudarst. Muzeia . . . im A.S. Pushkina (Pamiatniki
Literatury Narodov Vostoka, Teksty. Bolshaia Seriia 4; Moscow: Akademija nauk
SSSR, 1960) Pl. 1 where one can clearly distinguish between the ber direction of
the end-strengthening strip and the body of the papyrus itself.
15
R. Caminos, A Tale of Woe from a Hieratic Papyrus in the A.S. Pushkin Museum of
Fine Arts in Moscow (Oxford: Grith Institute. Ashmolean Museum Press, 1977) Pl.
3 shows clearly that the direction of the text and those of the bers coincide.
16
Whether this work represents a later copy of an actual ocial document or
whether it is a literary work, pure and simple, has been the source of considerable
controversy ever since the work was discovered. Three recent studies present many
of the arguments pro and con, see J. Baines, On Wenamun as a Literary Text,
Literatur und Politik im pharaonischen und ptolemischen gypten: Vortrge der Tagung zum
Gedenken an Georges Posener, 5.10. September 1996 in Leipzig (eds J. Assmann and
E. Blumenthal; Bibliothque dtude 127; Cairo: Institut Franais dArchologie Orien-
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 8
.xcirx+ rovr+i.x scnir+s 9
For another example of how a close examination of register can
add to our appreciation of how a text may have been perceived by
its ancient audience, I would like to discuss briey a remarkable doc-
ument that has entertained and fascinated both myself and our dis-
tinguished honoree, Professor Levine, for several years in a most
productive joint project. I am referring here to the Akkadian and
Egyptian versions of the Peace Treaty between Ramesses II and
Hattuili III.
17
Since the Egyptian version was inscribed in hiero-
glyphs in at least two temples and since the text invokes many deities
as witnesses, one might expect that it would adhere strictly to the
religious-literary register. A detailed examination of the grammar and
vocabulary, however, surprisingly revealed that the Treaty actually
closely conforms to the conventions of the Late Egyptian ocial and
documentary register. It fact, it conforms to the diction and gram-
mar of documentary Late Egyptian to a much greater extent than
most contemporary papyri from the Nineteenth Dynasty.
18
By avoid-
ing the archaic diction of Middle Egyptian religious texts and instead
opting for the more colloquial documentary register, those inscrib-
ing the text would have deliberately made the Treaty far more acces-
sible to the average educated Egyptian of the age. Such measures
would surely help publish and broadcast, so to speak, the news
that the two great rival powers of the age were now in a state of
amity. The choice of the documentary register also enhanced the
Treatys connection to certain aspects of those phenomena known
as popular religion and personal piety, but a closer examination
of those points would take us too far aeld today.
tale, 1999) 20933 and C.J. Eyre, Irony in the Story of Wenamun: The Politics
of Religion in the 21st Dynasty, Literatur und Politik im pharaonischen und ptolemischen
gypten, 23552; P. Vernus, Langue littrare et diglossie, Ancient Egyptian Literature:
History and Forms (ed. A. Loprieno; Probleme der gyptologie 10; Leiden: Brill, 1996)
56061.
17
O. Goelet and B. Levine, Making Peace in Heaven and on Earth: Religious
and Legal Aspects of the Treaty between Ramesses II and Hattuili III, Boundaries
of the Ancient Near Eastern World: A Tribute to Cyrus H. Gordon (eds M. Lubetsky et al.;
JSOTSup 273; Sheeld: Sheeld Academic Press, 1998) 25299.
18
For a discussion of the Late Egyptian dialectal features of the Treaty, see
Goelet and Levine, Making Peace, 25862.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 9
10 oorrx oorrr+, n.
Religious Texts and their Scripts
I shall now return to a vast genre of Egyptian papyri, the myriad
exemplars of the Book of the Dead, and examine aspects of regis-
ter as well as the intricate relationship between papyri and monu-
ments in ancient Egypt. I begin by expanding on a remark that I
made while sketching the place of hieratic within the Egyptian writ-
ing system. It was, I noted, related to yet another form of script
employed on papyrithe cursive hieroglyphs.
19
As far as their forms
are concerned, cursives occupy a position between hieratic and hiero-
glyphs and were developed primarily for use on papyri.
20
The rst
denitive examples may have occurred not on papyri, but rather on
the walls of a few Old Kingdom tombs, such as portions of the
inscriptions and decoration from the Sixth Dynasty tomb of Kai-
emankh at Giza.
21
Although most of the walls of the chamber where
this scene appears were decorated with representations of everyday
life and ornate colored hieroglyphs so characteristic of tombs of
important ocials during the Old Kingdom, for some reason the
decoration was not completed in the portion of the tombs under-
ground chamber where the cursives appear. If we consider the inten-
tion for using cursive hieroglyphs along with the strange stick-gure
people, one quickly understands that the cursive writing and gures
were meant as sketchy substitutes for normal tomb dcor. Similarly,
one sometimes encounters unnished spots in tombs where cursive
hieroglyphs have been used as the next-to-last stage before the actual
19
The relationship between various grades of script forms ranging from the
high prestige, elaborately painted or sculpted forms used primarily in tomb inscrip-
tions, outline forms for inscriptions on carved inscriptions on stelae, abbreviated
and cursive forms for papyri and occasionally stone, and nally hieratic for papyri
primarily for documentary purposes, is well-illustrated in Fischer, Archaeological
Aspects of Epigraphy and Palaeography, 41 Fig 4. A brief summary of under what
circumstances the cursive hieroglyphs were used can be found in the same work,
pp. 4042 and in a recent study.
20
One of the rst examples of cursive hieroglyphs, however, was on the walls of
the hypogeum in a Sixth Dynasty tomb at Giza; see H. Junker, Giza IV (DAWW;
Vienna and Leipzig: Hlder-Pichler-Tempsky, 1940). For discussion of the cursive
hieroglyphs and how they were used, see Parkinson and Quirke, Papyrus, 2428;
Fischer, Archaeological Aspects of Epigraphy and Palaeography, 3945; M.S. Ali,
Die Kursivhieroglyphen: Eine palographische Betrachtung, GM 180 (2001) 921.
21
Junker, Giza IV pls 9, 10. The scenes in the rest of the chamber are not only
complete, but in places they are colorfully executed, compare, for example, Color
Plate VIII.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 10
.xcirx+ rovr+i.x scnir+s 11
carving of the nal inscription as was done in the unnished tomb
of King Horemheb in the Valley of the Kings, a graphic indication
of their position in the hierarchy of Egyptian scripts.
22
A somewhat
dierent attitude underlies the occasional appearance of cursive hiero-
glyphs on Old Kingdom documentary papyri, but in this instance
moving, so to speak, in the opposite direction by supplying forms
that were larger, fuller, and more formal as the labels for more
abstract forms of the hieratic alongside, a usage which occurs as
early as the Abusir papyri in Dynasties VVI.
23
In this case, it is
clear by the relationship of cursive to hieratic script that this is a
conscious imitation of the format of a hieroglyphic royal decree on
a stone stela. Comparing these two examples we can sense other
aspects of the relationship between the cursive hieroglyphs and hier-
atic scriptin their essence the cursive hieroglyphs retained the dec-
orative and monumental associations of hieroglyphs, whereas hieratic
was intended for documentation. Generally speaking, those forms of
writing that were more detailed, more costly, and harder to execute
were at the same time the more prestigious, adding a dimension of
expense to the question of register. Cost played a similar role in a
few instances where inked cursive hieroglyphs were used because a
section of a stone surface was considerably harder than elsewhere
and therefore more expensive to carve as well. Finally, I would like
to note that at this stage in Egypts history, there was a strong ten-
dency for texts to be written in vertical columns, a practice that was
slowly abandoned in favor of the linear, horizontal format preferred
from the end of the Middle Kingdom onwards.
These factors all play a role in the development of the religious
texts which preceded and inuenced the Book of the Deadthe
Con Texts. As their very name implies, these texts appeared either
incised or inked on the inside surfaces of Middle Kingdom cons,
separated from the more ornamental inscriptions such as the oer-
ing formulae, the frise dobjets, or oering tables.
24
The script form
22
E. Hornung, Das Grab des Haremhab im Tal der Knige (Bern: Francke Verlag,
1971) Pl. 37 a, b.
23
P. Posener-Kriger and J.-L. de Cenival, The Abusir Papyri (Hieratic Papyri in
the British Museum, Fifth Series; London: British Museum Press, 1968) in many
of the larger accounts papyri, e.g., pls 13.
24
A great deal has been written on cons and con decoration in the period
between the First Intermediate Period to the New Kingdom. For a survey of this
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 11
12 oorrx oorrr+, n.
overwhelmingly chosen for the Con Texts themselves was cursive
hieroglyphs, partly because of the wooden surface on which they
were written and partly because during the Middle Kingdom that
script had become increasingly associated with religious or, more
generally speaking, arcane material, such as the medico-magic texts
from the Ramesseum,
25
a collection of short hymns to the god Sobek
from the same archive,
26
and a royal ritual papyrus. A large pro-
portion of the papyri employing cursive hieroglyphs exhibit two other
features which had been adopted by those inscribing the Con Texts:
they were written in the increasingly antiquated columnar fashion
and their texts were written in a retrograde direction. The last remark
requires a bit of explanation. Although Egyptian hieroglyphs may be
written from left-to-right or right-to-left depending, among other fac-
tors, on the exigencies of decoration, the ow of the text normally
proceeded against the direction in which the characters faced.
27
The
right-to-left direction so greatly predominated that we can consider
this the canonical direction of Egyptian writing. In a retrograde
inscription, however, the text ran in the same direction in which the
hieroglyphs faced, eectively backwards from the Egyptian point-of-
view. Both hieratic and demotic characters, incidentally, not only
always face towards the right, but they also always read from right
to left when written in horizontal rows.
topic with many helpful illustrations showing the placement the Con Texts in
relation to other components of con decoration, see G. Lapp, Die Entwicklung
der Srge von der 6. bis zur 13. Dynastie, The World of the Con Texts: Proceedings
on the Occasion of the 100th Birthdays of Adriaan de Buck, Leiden, December 1719, 1992
(ed. H. Willems; Egyptologische Uitgaven 9; Leiden: Nederlands Instituut voor het
Nabije Oosten, 1996) 7399.
25
A.H. Gardiner, The Ramesseum Payri: Plates (Oxford: Grith Institute at the
University Press, 1955) pls 1517; 2226.
26
A.H. Gardiner, Hymns to Sobek in a Ramesseum Papyrus, Rd 11 (1957)
4356.
27
The signicance and origins of the practice of using retrograde writing on
papyri is problematic. Retrograde writing was certainly not conned to the Book
of the Dead, nor was it a necessity in that corpus, since many early copies of the
work did not employ it, see A. Niwinski, The Problem of Retrograde Writing and
the Direction of Reading of the Book of the Dead in the New Kingdom, Studies
on the Illustrated Theban Funerary Papyri of the 11th and 10th Centuries BC (OBO 86;
Freiburg: Universittsverlag; Gttingen: Vandenhoeck & Ruprecht, 1989) 1317 and
M.A. Chegodaev, Some Remarks Regarding the So-called Retrograde Direction
of Writing in the Ancient Egyptian Book of the Dead, DE 35 (1996) 1924.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 12
.xcirx+ rovr+i.x scnir+s 13
In any case, one might ask why one would deliberately strive for
the arcane and archaic? To answer that question, one must bear in
mind that often in the ancient world and especially in Egypt, the
distant, near-mythical past could confer much validity on the activ-
ities of the here-and-now. Unlike our modern sensibility wherein that
which is most modern and cutting-edge is assumed to be, perforce,
intrinsically superior, the Egyptians revered the days of yore and
past forms. A constantly recurring theme in Egyptian texts involves
the king or a high ocial researching through ancient records in
order to reestablish the ideas and practices of the glorious past. A
wonderful example of this mentality is supplied by the Twenty-fth
Dynasty monarch, Shabako, who told of nding an ancient papyrus
in a temple and inscribing it on a stela even though it had been
eaten through by worms.
28
The text on this stela, known as the
Memphite Theology, was long thought to derive from a genuine Old
Kingdom source until F. Junges careful examination demonstrated
enough peculiarities to reveal its real nature as a pious forgery com-
missioned by Shabako.
29
Thus, the rather dubious tradition of cre-
ating faked antiques can be added to the long list of ancient Egypts
contributions to human heritage. Returning now to the Con Texts,
we can see that the script and its retrograde direction predominantly
employed in these spells was intended to convey an aura of great
antiquity and authenticity even though they most likely had been
composed fairly recently. In addition, the use of cursive hieroglyphs
made these texts more prestigious, an especially important consid-
eration for most of the con owners who could not aord the great
expense of a rock-cut tomb or a stone mastaba any more than my
wife and I would be able to aord the original antique on which
our dining room table is based.
The Book of the Dead and Its Production
In my sketch of the Book of the Dead which now follows, there are,
to be sure, numerous exceptions, but my remarks hold true for
28
Parkinson and Quirke, Papyrus, 7475, with Fig. 52 which shows the portion
of the inscription mentioning the worms.
29
F. Junge, Zur Frhdatierung des sog. Denkmal memphitischer Theologie,
MDAIK 29 (1973) 195204.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 13
14 oorrx oorrr+, n.
around eighty per cent of the corpus before the Third Intermediate
Period. When discussing Book of the Dead manuscripts, it is impor-
tant to bear in mind rst that these manuscripts were essentially bur-
ial equipment produced in funerary workshops, then placed in the
con. In this respect, the Egyptians anticipated a scribal tradition
later prevalent throughout the Near East, ancient and modern, by
developing a system of copying and producing illustrated religious
manuscripts for an elite clientele. In fact, the methods and some
details of the division of labor adopted by the scribal workshops
seems to have roughly paralleled the techniques used in the manu-
facture of cons and tombs.
30
The scribes who worked on these
papyri, much like their artisanal counterparts, developed specialties.
Some men were particularly good at the artwork which comprised
the vignettes; other scribes were essentially scriveners who copied
texts with varying, often lamentable, degrees of accuracy. Just as
tombs were decorated, dierent sections of a Book of the Dead
scroll might be illustrated by dierent teams who eventually met at
some point. Of course, the ability simply to paste parts of an entire
papyrus together was very much easier than teams working side-by-
side in the cramped quarters of a tomb. Once again we can see the
subtle interaction of papyrus and monumental text in the Egyptian
mind.
The lovely papyrus of Ani (BM 10470) is a particularly ne exam-
ple of such workshop manuscripts, especially since it was originally
made as a template papyrus and shows distinct traces of a team
eort.
31
At the last moment, the number of sheets that this mans
30
For example, H. Milde, The Vignettes in the Book of the Dead of Neferrenpet (Egypt-
ologische Uitgaven 7; Leiden: Nederlands Instituut voor het Nabije Oosten, 1991)
23940, has noted certain stylistic and thematic similarities between the Book of the
Dead of a certain Neferrenpet and the contemporary tomb of Queen Nefertari,
presumably because the scribes who composed the document were familiar with the
tomb.
31
The composite nature of this papyrus has been overlooked in a recent study of
the Ani Papyrus by T.G.H. James, Vignettes in the Papyrus of Ani, Colour and Paint-
ing in Ancient Egypt (ed. W.V. Davies; London: British Museum Press, 2001) 141
44. For a brief discussion of the physical composition of this papyrus, see O. Goelet,
A Commentary on the Corpus of Literature and Traditions which Constitutes The
Book of Going Forth by Day, The Egyptian Book of the Dead: The Book of Going Forth by
Day (ed. E. von Dassow; San Francisco: Chronicle Books, 1994) 142 and many
other remarks in the descriptions of the plates on pp. 15470. There were perhaps
as many as ve dierent scribes at work on this papyrus and at least two separate
vignette artists. The non-uniformity of the papyrus colored borders provides another
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 14
.xcirx+ rovr+i.x scnir+s 15
family could aord were pasted together, and his name and titles
inserted in the blank spaces. Even a cursory examination of photo-
graphic reproductions reveals the distinctively cruder and clumsier
hand of one of those delegated to ll in the names. There are sev-
eral instances where the space for the name was inadvertently left
blank. In another instance the scribe even misspelled Anis name in
a critical episode of his afterlife.
32
I have identied as many as ve
separate hands at work on Anis scroll.
This would be a good juncture to make a few brief remarks on
identifying both hieroglyphic and hieratic hands. Especially when
one is dealing with workshop- and school-connected materials,
identication of individual scribal hands should be done with the
greatest of caution.
33
Under these circumstances, scribes closely trained
other apprentice scribes to closely imitate their own book hand.
The situation is perhaps best understood by considering the unnat-
ural and studied hands exemplied by the modern uncial or italic
calligraphy used today for wedding announcements, formal procla-
mations, and the like.
34
These modern calligraphic hands, signicantly,
closely imitate Mediaeval or Renaissance models.
35
The result is hand-
writing that can be uncannily similar from document to document,
largely because the texts are written in an articial script far removed
from ones daily life and work. Drawing from this analogy, one wish-
ing to identify individual hands on ancient documents is therefore
advised to seek out the most common signs and most frequent words,
those which the scribe was apt to write unconsciously and with the
greatest uidity.
tell-tale sign in this respect. I hope to publish a more extensive description of the
manufacture of the Ani Papyrus at some point soon in the future.
32
See Goelet, Book of the Dead, pl. 4 over the gure of the standing Ani.
33
A note of caution should be injected here concerning attempts to identify var-
ious scribal hands in light of the fact that scribes tend to develop a style of hand-
writing based on the person who taught them, see J.J. Janssen, On Style in Egyptian
Handwriting, JEA 73 (1987) 16167. A careful, model approach towards the use
of identiable hieratic hands can be found in H. Van den Berg and K. Donker
van Heel, A Scribes Cache from the Valley of Queens? The Palaeography of
Documents from Deir el-Medina: Some Remarks, Deir el-Medina in the Third Millennium
AD: A Tribute to Jac.J. Janssen (eds R.J. Demare and E. Egberts; Leiden: Nederlands
Instituut voor het Nabije Oosten, 2000) 949.
34
J. Smith, Creative Calligraphy (New York: Lorenz Books, 1998) 59 and many
other places in this work.
35
Smith, Creative Calligraphy, 43.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 15
16 oorrx oorrr+, n.
The fact that many ancient manuscripts were to a certain degree
mass-produced does not mean that they could not also be of the
highest aesthetic quality. Textual accuracy, unfortunately, was seem-
ingly of much lesser importance. By combining the methods of tex-
tual criticism, art history, and paleography, it is possible to create
whole family trees of related manuscripts, perhaps representing
workshops.
36
This is not to say, however, that many Book of the
Dead papyri were not made to order, but these too would have
been produced by specialiststhis was the reproductive mode par
excellance.
Whether they were template manuscripts or specically commis-
sioned works, almost every Book of the Dead was an exercise in
articiality and produced by men who specialized in such works.
These scribes were basically copyists who were often working with
imperfect master editions. The articiality began with the very scrolls
themselves. By the middle of the Eighteenth Dynasty, a typical Book
of the Dead scroll, unlike a business, or even a literary papyrus,
unrolled towards the right, with the columns of text moving in the
same directionfrom left to right. It was, thus, a completely back-
wards, or even an upside-down, roll of papyrus. Metaphorically, how-
ever, the Book of the Dead and its writing was imitating the backwards
world of the underworld, the Duat, through which the sun was mov-
ing from West to East towards rebirth at sunrise, counter to its direc-
tion in the visible world of the living. The Book of the Dead tended
predominantly to be written with cursive hieroglyphs in columnar
format, had retrograde text for the most part, and used the archaic,
classical Middle Egyptian. This assemblage of peculiarities should not
surprise us since the Book of the Dead spells were in many ways a
derivation of the Con Texts and imitated the manner in which
the latter had been inscribed. In almost every aspect, Book of the
Dead papyri represented the mirror image of the standard scribal
environment. Normally, a scribe would write predominantly in the
business-ocial register, i.e., essentially contemporary Late Egyptian,
would employ hieratic written horizontally, his script would ow
from right to left, and his scroll would unroll leftwards.
36
See, for example, the studies of U. Rssler-Khler on Chapter 17, Kapitel 17
des gyptischen Totenbuches: Untersuchungen zur Textgeschichte und Funktion eines Textes der
altgyptischen Totenliteratur (GOF 10; Wiesbaden: Harrassowitz Verlag, 1979) and idem,
Sargtextspruch 335 und seine Tradierung, GM 163 (1998) 7193.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 16
.xcirx+ rovr+i.x scnir+s 17
Even if one comes close to making an argument from silence by
saying so, it may well be that a Book of the Dead was often pur-
chased as a substitute for a tomb and elaborate burial equipment
for those who could not aord one of those elaborate cli-tombs we
so much associate with New Kingdom Thebes.
37
Virtually none of
the lengthy and elaborate manuscripts preserved from the New
Kingdom can be traced back to those famous tombs. The layout
and content of an average manuscript, furthermore, recalled several
aspects of an elaborate burial in a tomb. Many papyri began with
a depiction of an elaborate funeral procession and the nal cere-
monies at the tomb entrance.
38
Additionally, near the beginning of
these papyri are hymns to the rising and setting sun, and to Osiris,
closely paralleling those which appear near the entrance of a typi-
cal New Kingdom tomb, and the Opening of the Mouth Ceremony,
so important to the eectuation of the tomb and its contents. In
addition, papyri normally contain chapters not only concerned with
numerous funerary amulets, but the entire set-up of a well-appointed
burial chamber as well. Finally, at the end of the scroll there was
often a depiction of the very purpose of the Book of the Dead, whose
title in Egyptian was the far cheerier The Chapters of Emerging
by Day, the deceaseds spirit coming forth from the tomb into the
daylight. All this for individuals who probably were not tomb own-
ers. I might add that papyri sometimes incorporate elements of con-
temporary tomb design such as the two-colored bands which likewise
appear at the top and bottom of the walls. A Book of the Dead
could supply, then, many features of a tomb and an elaborate bur-
ial and was certainly vastly cheaper.
The connection between tomb and papyrus can be seen in the
most elaborate funerary monuments of allthe royal tombs of the
New Kingdom. There, too, the texts, stated explicitly to be books
written by the gods themselves, frequently employed both columnar
texts and cursive hieroglyphs, and with the retrograde direction
predominating, all certainly in a deliberate imitation of the papyrus
37
For two views of this problem, see H. Beinlich, Das Totenbuch bei Tutan-
chamun, GM 102 (1988) 718 and H. Guksch, Totenbuchpapyrus vs. Grabeigaben,
GM 104 (1988) 8990.
38
A complete study of this chapter of the Book of the Dead has been made by
B. Lscher, Totenbuch Spruch 1 nach Quellen des Neuen Reiches (KT; Wiesbaden: Harras-
sowitz Verlag, 1986).
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 17
18 oorrx oorrr+, n.
originals.
39
Also, like the Middle Kingdom ritual papyrus mentioned
earlier, these scenes were accompanied by stick gures. This mode
of illustration may all seem quite peculiar, given that the pharaoh
could unquestionably aord the most elaborate and expensive forms
of decoration, until one considers that, to the Egyptian mind, divinely
authored books would most likely follow such archaic practices. Also,
note the predominantly yellow tone of the background of these royal
afterlife scenes, quite unlike the standard white of the walls of a con-
temporary private tomb in the Theban necropolis. These ancient
books on the walls have been painted to imitate deliberately the yel-
low of an old papyrus, just the same tone as Anis Book of the Dead
has naturally turned over the millennia since its manufacture.
40
Anis
papyrus is to some degree a book imitating a tomb, but in the royal
tombs we have the reversethe tomb as ancient book. Once more,
we encounter the Egyptian love of the false antique.
Some Closing Notes on Scribal Training
The studied and anachronistic nature of both royal tomb decora-
tion and the Book of the Dead lead me to some closing observa-
tions on scribal training and the signicance of an important genre
of inscribed objects in the New Kingdom. The royal necropolis work-
ers at Thebes lived in the special village of Deir el-Medina, a site
which has yielded a great number of hieratic school ostraca, writ-
ing tablets, and didactic papyri that oer insights into the way young
scribes learnt their trade. Although this material dates primarily to
the Nineteenth and Twentieth Dynasties, a high proportion of the
hieratic ostraca in particulara cheap and disposable medium
contain passages from the great literary texts of the Middle Kingdom,
composed several hundred years before. Considering that they were
supposedly written primarily for instructional purposes, the script
hands on the vast preponderance of these ostraca show little evi-
dence of the awkward characters which one might expect under
39
E. Hornung, Tal der Knige: Die Ruhesttte der Pharaonen (Zurich and Munich:
Artemis Verlag, 1983
2
) pls 95108, among many others illustrated in this work. The
yellow color is distinct.
40
Quirke and Parkinson, Papyrus, Color Plate V; also the Book of the Dead
papyrus of Amenemhat, illustrated on the same page, Color Plate VI.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 18
.xcirx+ rovr+i.x scnir+s 19
these circumstances.
41
In fact, some ostraca with Middle Kingdom
texts were rather elegantly and carefully executed in contemporary
Ramesside hieratic, exhibiting no traces the expected awkwardness
of a beginners hand.
42
On the other hand, the majority of literary
ostraca are fraught with errors in grammar and orthography, indi-
cating that the classical Middle Egyptian of these texts was as anti-
quated and unfamiliar as Chaucers English might be for a classroom
of American high school students. On one remarkable school papyrus,
in fact, the student was apparently given an exercise in which he
was to write some standard lines rst in colloquial Late Egyptian,
then to follow each of these with their equivalents in Middle Egyp-
tian.
43
A striking conrmation of the inuence of register on scribal
handwriting emerges from the corpus of Deir el-Medina ostraca as
a whole, for it is possible for experts to separate literary and docu-
mentary ostraca from one another with surprising accuracy on the
41
The writing skill displayed on many so-called school ostraca has been noted
by several scholars who have worked with these objects, see A.G. McDowell,
Teachers and Students at Deir el-Medina, Deir el-Medina in the Third Millennium
AD, 21733. A somewhat dierent approach to these texts has been taken by J.J.
Janssen, Literacy and Letters at Deir el-Medna, Village Voices: Proceedings of the
Symposium Texts from Deir el-Medna and their Interpretation: Leiden, May 31June 1,
1991 (eds R.J. Demare and A. Egberts; CNWS Publications 13; Leiden: Centre
of Non-Western Studies, Leiden University, 1992) 8687, who proposes that one
use of such material may have been to produce cheap editions of the text for
use in the village, thus reminiscent of our modern paperbacks. He also notes
(Literacy and Letters, 87 n. 87) that there is no reference to a school among
these so-called schoolboy-exercises. Recently, A. Gasse has once more raised the
question of the existence of a school in the village, see Le K2, un cas dcole?
Deir el-Medina in the Third Millennium AD, 10920.
42
For example, BM EA 5629, illustrated in Parkinson, Cracking Codes, Color Plate
28; or the more famous Ashmolean Ostracon containing a Ramesside copy of the
Story of Sinuhe, see J.W.B. Barns, The Ashmolean Ostracon of Sinuhe (Oxford: Oxford
University Press, 1952) Frontispiece and the more detailed photographs that accom-
pany the transcription. Barns, however, disagreed with ern y on whether a student
or a competent teacher-scribe executed the copy; see p. 35.
43
R.A. Caminos, A Fragmentary Hieratic School-book in the British Museum,
JEA 54 (1968) 11422. A similar papyrus exercise written in a form of early demotic
has recently been identied, see J.F. Quack, A New Bilingual Fragment from the
British Museum (Papyrus BM EA 69574), JEA 85 (1999) 15464, with pls XXI
and XXII. In this latter case, strangely enough, both the late Middle Egyptian and
the demotic text were written in a fossilized hieratic handwriting which was described
as uncial by the commentator. Textual material such as this in which there seems
to be a sharp disparity between written and spoken language, or the language of
formal texts and that of daily usage raises the fascinating problem of diglossia. A
survey of this wide-ranging problem is presented by P. Vernus, Langue littraire
et diglossie, Ancient Egyptian Literature, 55564.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 19
20 oorrx oorrr+, n.
basis of the appearance of their script alone, employing such objec-
tive criteria as the writing surface used, the size of handwriting, and
line spacing.
44
Of course, in a society which idealized the past as much as Egypt,
a grounding in the classics of their culture might well have been
the mark of an educated person, just as it is today.
45
But this sugges-
tion would leave unexplained why such a small percentage of Rames-
side literary and instructional ostraca are based on Late Egyptian,
the language of their everyday life. Oddly enough, there is an abun-
dance of lengthy Late Egyptian school texts, the so-called Late Egyptian
Miscellanies, but these are preserved predominantly on the more expen-
sive medium of papyri.
46
Although knowledge of the ancient Egyptian
equivalent of the classics and past forms of the language may well
have lent an individual the aura of an educated man, I believe that
the more likely and more practical reason for the emphasis on the
ostensibly impractical Middle Egyptian form of the language lies in
its use in the mortuary register, i.e., the Book of the Dead and tomb
inscriptions, both royal and private. I think that this usage is conrmed
by the type of texts generally found on a much rarer genre of ostraca
and writing boards, namely those written in hieroglyphs or cursive
hieroglyphs. The text on these ostraca almost always involve the type
of late Middle Egyptian-based material that appear in tombs, and,
furthermore, such exemplars will occasionally be accompanied by
sketched gural representations of the sort normally found in funer-
ary contexts. The texts on these hieroglyphic ostraca, moreover, are
frequently written in a vertical format with the column lines care-
fully indicated.
47
Not incidentally, such ostraca would also aord the
44
A. Gasse, Les ostraca hiratiques littraires de Deir el-Medina: Nouvelles ori-
entations de la publication, Village Voices, 5170.
45
The question of what might have constituted the mark of an educated man
and a classical education has long fascinated Egyptologists, for example, see
Janssen, Literacy and Letters at Deir el-Medina, 8591.
46
A.H. Gardiner, Late-Egyptian Miscellanies (Bibliotheca Aegyptiaca 7; Brussels: di-
tion de la Fondation gyptologique Reine lisabeth, 1937) and R.A. Caminos, Late
Egyptian Miscellanies (Brown Egyptological Studies 1; Oxford: Oxford University Press,
1954).
47
I am speaking here particularly of the ostraca exemplars of the didactic text
known as the Kemyt. This was a composition composed during the Middle Kingdom
in the current dialect of Egyptian. It was revived again during the Ramesside Period,
when it enjoyed a particular popularity among the village of royal necropolis work-
ers at Deir el-Medina despite the fact that it employed not only an obsolete dialect
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 20
.xcirx+ rovr+i.x scnir+s 21
student practice writing on the exact same type of limestone surface
he would be apt to encounter on a tomb wall. The division of both
script and content among ostraca suggests a process which would
accord with Clement of Alexandrias description of scribal training
rst, familiarization with the text genre using the script of everyday
life, then training in the hieroglyphs themselves. I would like to close
with this thought: even if fakery might come naturally to some indi-
viduals, fortunately, most people must to be trained to produce it.
and columnar format, but it also was written in an equally antiquated script whose
forms were close to that of cursive hieroglyphs. These characteristics made the Kemyt
an ideal text for training men whose occupation was production of texts and other
material for the Theban mortuary industry; see H. Brunner, Altgyptische Erziehung
(Wiesbaden: Otto Harrassowitz, 1957) 8388.
SHIFFMAN_f2_1-21 12/2/02 5:19 PM Page 21
22
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 22
This page intentionally left blank
THE SAMARIA PAPYRI AND THE
BABYLONIO
-
ARAMEAN SYMBIOSIS
Douglas M. Gropp
The Samaria papyri are a group of fragmentary remains of legal
documents once belonging to wealthy patricians of Samaria. They
come from the cave Abu Shinjeh in the Wdi ed-Dliyeh, about 14
km north of Jericho on the western rim of the Jordan rift. Most of
these were discovered in the early spring of 1962 by the Ta'mireh
Bedouin. Subsequent archaeological explorations in January 1963
and February 1964 contributed modestly to the initial nd and put
the papyri in a more denite context.
1
Despite the nd spot, the papyri are all legal documents originally
drafted in Samaria in the fourth century ncr. The place in which
the documents were executed is given either in the rst or last line
of the document. In the eleven documents in which the place is pre-
served, it is named as the city or province of Samaria. The docu-
ments were also dated by the reign of the current Persian king.
Where the name of the king is preserved, it is usually Artaxerxes (at
least ve times). One document is dated to sometime between the
thirtieth and thirty-ninth year, and therefore must come from the reign
of Artaxerxes II (Mnemon), between 375 and 365 ncr. The date
of WDSP 1 is fully preserved as 19 March 335 ncr, the second
year of Darius III (Codomannus).
2
Most of the papyri were prob-
ably written during the reign of Artaxerxes III (Ochus; 358337 ncr).
The late pre-Alexandrine coins found in the cave strongly corrobo-
rate the internal indications of dating,
3
the latest being of Tyrian issue
of 334 ncr. The script of the papyri is somewhat more advanced
23
1
P.W. and N.L. Lapp (eds), Discoveries in the Wdi ed-Dliyeh (AASOR 46; Cambridge,
MA: American Schools of Oriental Research, 1974) 116, 3032.
2
F.M. Cross, Samaria Papyrus 1: An Aramaic Slave Conveyance of 335 B.C.E.
found in the Wdi ed-Dliyeh, Nahman Avigad Volume (eds Y. Yadin and B. Mazar;
ErIsr 18; Jerusalem: Israel Exploration Society in cooperation with the Institute of
Archaeology, Hebrew University, 1985) 7*17*.
3
F.M. Cross, Coins, Discoveries in the Wdi ed-Dliyeh, 5759.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 23
24 rotor.s v. onorr
typologically than the script of the Aramaic corpora from the late
fth century and so ts well within this horizon.
The papyri are quite fragmentary. Eighteen of the fragments are
long enough, that is complete enough in their vertical dimension, to
be called papyri. The largest papyrus, WDSP 1, is no more than
48% extant. A few of these fragments are no more than a thin strip
of papyrus; a couple of others are in tatters. Nine or ten further
pieces are sizable enough to allow some assessment of their legal
import. Nine other museum plates contain nearly 150 additional
fragments of various shapes and sizes. All of the plates are housed
in the Rockefeller Museum in Jerusalem.
The language of the Samaria papyri is Ocial Aramaic, the
ideal standard language in which scribes of the Persian period (prob-
ably from Darius I to Darius III [522330 ncr]) would draft docu-
ments of an ocial nature.
4
The language of the Samaria papyri is
virtually identical to the language of the fth-century Elephantine
legal papyri and the Arsames correspondence. In fact, in spite of
being chronologically later, the language of the Samaria papyri is
even more consistently conservative in its conformity to the norm of
Ocial Aramaic than the language of the other two corpora. In
spite of its later provenience, it also reects little or no Persian
inuence in contrast to the Elephantine papyri, but especially to the
Arsames correspondence where Persian inuence is more extensive.
The Samaria papyri do show a greater proportion of specically late
Neo-Babylonian loans. But this is clearly related to the origins of its
legal formularies.
5
Of the fragments sizable enough to allow some assessment of their
legal genre, at least half are slave sales. Some of these represent the
sale of a single slave (WDSP 1, 3, 4, 11 recto, 18, 19, 26?), others
of multiple slaves (WDSP 2, 5, 6, 7, 8, 9, 20). But a variety of other
legal genres are represented. There is a clear instance of a house
sale (WDSP 15) and a conveyance of chambers in a public build-
ing (WDSP 14), in addition to several deeds of sale whose objects
cannot be ascertained (WDSP 21, 22, 24, 25). Two or more docu-
ments look like a pledge of a slave in exchange for a loan (WDSP
4
See D.M. Gropp, Imperial Aramaic, OEANE 3.14446.
5
D.M. Gropp, The Language of the Samaria Papyri: A Preliminary Study,
Maarav 56 (1990) 16987.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 24
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 25
10, 12, 13 recto?, 27?). It is impossible to be condent about the
terms of these fragmentary contracts apart from some hermeneuti-
cal key provided by closer legal parallels. One document conveys a
vineyard (possibly as a pledge rather than as a sale) (WDSP 16).
Several documents may resolve some contingency, but in most cases
the papyri are too fragmentary for condent interpretation. There
is a receipt for the repayment of a loan involving a pledge (WDSP
17: a double document), the release of a pledged slave (WDSP 13
verso), the settlement (?) of a dispute over a slave (WDSP 11 verso),
and possibly a judicial settlement by an oath (WDSP 23).
If it were not for some auspicious circumstances that provide extra-
ordinary possibilities for reconstructing the text of the papyri, their
signicance would have been greatly reduced. The possibilities of
reconstruction correlate directly with the legal genres represented.
The best represented type of deed is the slave sale. Despite their
fragmentary condition, it has proved possible to propose full recon-
structions for nine or ten of these. The slave sale deeds seem to
share a common formulary to such an extent that they provide us
with partially overlapping bits and pieces of the same formulary.
There is remarkably little variation in the verbal realization of each
formula, and even less variation in the sequence of formulas within
the sale formulary. The date, the names of the principals, the names
of the slaves, the sale price, and the amount of the penalty for con-
travention of the sale are basically the only elements that vary from
deed to deed. Each papyrus contributes a little to our knowledge of
that formulary. By constant comparison and rearranging of all the
bits of writing serially, F.M. Cross and I have been able to recon-
struct the entire formulary. Proposed reconstructions have been tested
against estimated line lengths for each papyrus. In addition to the
constraints of space, we have been aided by legal parallels from
Mesopotamia to Egypt. Circular reasoning, while not wholly escaped,
can be reduced to a minimum. Now that this formulary for the
deeds of slave sale is established, it can be applied as a kind of tem-
plate for interpreting other deeds of conveyance. Such a procedure
has been successful in reconstructing a deed of house sale (WDSP
15), but only modestly helpful in interpreting the texts of other
genres.
The Samaria papyri are signicant for the history of Aramaic lan-
guage and paleography. A modicum of information about the polit-
ical organization and historical situation of fourth-century Samaria
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 25
26 rotor.s v. onorr
can be eked out of these fragments. But the Samaria papyri will ulti-
mately prove most interesting for the light they cast on the history
of law. They provide an especially promising occasion for the study of
the contact between Aramaic and cuneiform traditions. Comparison
of legal formularies provides one of the most controllable instances
for the study of cultures in contact. I have tried to explore these
parallels in the hope that they will provide an initial heuristic for
the legal historians and cuneiformists who will doubtlessly nd the
Samaria papyri well worth their attention.
Although I do not pretend to be a jurist, I have found the task
of philological interpretation to be inseparable from a comparative
legal evaluation of the deeds. It was, in fact, only through a com-
parison with cuneiform law, particularly through a comparison with
Middle Assyrian deeds of sale, that Cross was able to divine the
basic structure of the sale formulary of the Samaria papyri. The legal
formulary of the slave sales is obviously dependent proximately or
ultimately on cuneiform antecedents. I have come to the conclusion
that the late Neo-Babylonian sale formulary for movables (from the
time of Darius I on) formed the basis of the Aramaic formulary.
6
Aside from a few important parallels, it is remarkable the extent
to which the formularies of the Samaria papyri dier from the for-
mularies of the Elephantine legal papyri. On the other hand, the
Samaria papyri share a larger number of features with the later
Murabba'at and Naal ever deeds. The Samaria papyri thus pro-
vide some counterbalance to the understandably heavy reliance on
the Elephantine legal papyri for reconstructing the early develop-
ment of Jewish law.
My concern in this paper is less with the structure of the formu-
lary for selling slaves in the Samaria papyri as such, or with the
legal import of individual legal formulas within that structure. I am
primarily concerned here with the genetic question: What is the ori-
gin of the formulary for selling slaves in the Samaria papyri and
what may that tell us about the cultures in contact in the Persian
period?
6
See especially H. Petschow, Die neubabylonischen Kauormulare (Leipziger rechtswis-
senschaftliche Studien 118; Leipzig: Theodor Weicher, 1939) 4368. This mono-
graph has been invaluable to me in reconstructing the formulary of the Samaria
Papyri. The late Neo-Babylonian sale formulary for movables, itself, is evidently the
heir of Middle Assyrian and other peripheral cuneiform legal traditions.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 26
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 27
7
But there is a certain amount of inconsistency on this point. WDSP 1 main-
tains the subjective style throughout the Schlussklauseln. So do apparently WDSP 4,
5 (beginning already in 5:6), 6, 7, 8, 9. WDSP 2, on the other hand, is apparently
objectively styled as far as 2:7, where it shifts to subjective style continuing as far
as the phrase: [::: ::, :] : ::,: in 2:10. WDSP 3 maintains its objec-
tive style through the defension clause and until the quoted speech in 3:7. It, too,
then maintains this subjective orientation as far as the phrase: : : ::,:
[:::] ::, in 3:910.
8
Cross, Samaria Papyrus 1, 13*.
9
Cf. P. Koschaker, Neue keilschriftliche Rechtsurkunden aus der El-Amarna-Zeit (ASAW
The origin of the sale formulary of the Samaria papyri is three-
fold. (1) Aramaic scribes in Babylonia adopted the late Neo-Babylonian
formulary for the sale of movables (from the time of Darius I on)
as their basic model. (2) Aramaic scribes (still in Babylonia) creatively
modied this model by drawing on formulas from other types of late
Neo-Babylonian documents. (3) Aramaic scribes (probably in Palestine)
further modied the adopted formulary by partially assimilating it
to their own native legal traditions.
First I would like to mention a puzzling phenomenon that ought
to provide clues to the origin of the formulary, but which I do not
nd precisely paralleled anywhere. The formulary as a whole is objec-
tively stylized. But there is usually a shift from an objective formu-
lation in the operative section to subjective formulation in the Schluss-
klauseln.
7
While the obligations undertaken by the seller in the
Schlussklauseln are subjectively formulated, the framework of the
Schlussklauseln, consisting of the pivotal clause ::: : : :.:
at its beginning and often . . . ::: ::, : : ::,: at the
end, is objectively formulated. Cross notes that the mention of :
in the introduction and conclusion of the Schlussklauseln forms an
inclusio.
8
Both elements of the inclusio are in objective third person
discourse, while the nal clauses themselves tend to be formulated
in the rst person. There is no evidence to counter the presump-
tion that the deed always returns to an objective formulation before
the witnesses are listed in the text.
This situation in the Samaria papyriwith objective style in the
operative section shifting to subjective style in the Schlussklauseln
contrasts with the Elephantine legal documents, on the one hand,
which are subjectively styled throughout, and with the late Neo-
Babylonian deeds of sale, on the other, which are objectively styled
throughout. Neither the peripheral Sprech(linu-)urkunden
9
nor the
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 27
28 rotor.s v. onorr
somewhat later Zwiegesprchsurkunden
10
are really comparable. Still, it
makes legal sense for the seller or alienor to assume the obligations
of the nal clauses in the rst person.
(1) Aramaic scribes in Babylonia adopted the late Neo-Babylonian formulary
for the sale of movables ( from the time of Darius I on) as their basic model
This primary source of the formulary for selling slaves in the Samaria
papyri is evidenced in the overall structure of the document con-
sisting of a declaration of sale, a receipt-quittance clause, a defen-
sion clause, and list of witnesses, as well as in the objective formulation
of the deed as a whole. The late Neo-Babylonian formulary for the
sale of movables has also left a clear imprint in the word order, loan
words, and calques within the major clauses of the formulary. Let
us look at each of these three main constituents of the formulary in
turnthe declaration of sale, the receipt-quittance clause, and the
defension clauseto illustrate this proposition.
(a) The Declaration of Sale
The declaration of sale like the rest of the document is formulated
ex latere venditoris. Seller orientation versus buyer orientation is the
primary distinction between the sale of movables and the sale of
immovables in late Neo-Babylonian documents. Movables (including
slaves) are sold; immovables are bought.
11
The late Neo-Babylonian
39/5; Leipzig: S. Hirzel, 1928) 2123 [hereafter NRUA]; M. San Nicol, Beitrge
zur Rechtsgeschichte im Bereiche der keilschriftlichen Rechtsquellen (Oslo: H. Aschenhoug &
Co. [W. Nygaard], 1931) 15052; Y. Mus, Studies in the Aramaic Legal Papyri from
Elephantine (Studia et documenta ad jura Orientis antiqui pertinentia 8; Leiden: Brill,
1969) 22 n. 2, 17576.
10
Cf. H. Petschow, Die neubabylonische Zwiegesprchsurkunde und Genesis
23, JCS 19 (1965) 10320; idem, Mittelbabylonische Rechts- und Wirtschaftsurkunden der
Hilprechtsammlung Jena (ASAW 64/4; Berlin: Akademie, 1974) 3839. J.C. Greeneld
(Babylonian-Aramaic Relationship, Mesopotamien und seine Nachbarn: Akten des XXV
e
Rencontre Assyriologique Internationale Berlin July 37, 1978 [eds H.-J. Nissen and J.
Renger; Berlin: Dietrich Reimer, 1982] 2.47374) suggests that the increase in the
use of Zwiegesprchsurkunden may be due to West Semitic inuence. He fails, how-
ever, to distinguish between the various types of dialogue documents when he
lumps the Elephantine papyri, the Daliya papyri, the Bar Kosiba contracts and
the later Jewish rt together with the Neo-Babylonian Zwiegesprchsurkunden.
While the impetus for the scribes of Samaria to frame the Schlussklauseln may indeed
come from the West, the precise phenomenonobjectively formulated operative
section, but subjectively formulated Schlussklauselnremains unparalleled.
11
Cf. Petschow, Kauormulare, 43.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 28
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 29
12
The Ur III Sumerian and later Old Babylonian deeds of sale were formulated
from the point of view of the buyer. Cf. e.g., for slave sales, M. Schorr, Urkunden
des altbabylonischen Zivil- und Prozessrechts (VAB 5; Leipzig: J.C. Hinrichs, 1913) 12130.
Cf. the schema of Old Babylonian sales in general presented on p. 111. See, fur-
ther, the discussion of M. San Nicol, Die Schluklauseln der altbabylonischen Kauf- und
Tauschvertrge: Ein Beitrge zur Geschichte des Barkaufes ( Mnchener Beitrge zur
Papyrusforschung und antiken Rechtsgeschichte 4; Munich: C.H. Beck, 1922) 2628,
3637. The Middle Babylonian deeds of sale follow the Old Babylonian pattern.
Cf., e.g., the two slave sales transcribed and translated by H. Petschow, Mittelbabylonische
Rechts- und Wirtschaftsurkunden der Hilprechtsammlung Jena, 1116. The early Neo-
Babylonian slave sale, in the form of a Zwiegesprchsurkunde, transcribed and trans-
lated by M. San Nicol, Babylonische Rechtsurkunden des ausgehenden 8. und 7. Jahrhunderts
v. Chr. (ABAW 34; Munich: Bayerischen Akademie der Wissenschaften, 1951) 8889,
is also from the point of view of the buyer. On the other hand, the Middle Assyrian
deeds of sale are formulated from the point of view of the seller. Cf. E. Ebeling,
Urkunden des Archivs von Assur aus mittelassyrischer Zeit, Part III: Assyrische Kauf-,
Zessions- und Schenkungsurkunden (MAOG 7; Leipzig: Harrassowitz, 1933) 5589.
But apart from several early eighth-century deeds from Nimrud (e.g., CTN II 5,
17), Neo-Assyrian deeds of sale are consistently formulated from the buyers point
of view. Cf. J.N. Postgate, Fifty Neo-Assyrian Legal Documents (Warminster, England:
Aris & Phillips, 1976) 1314. The situation in the more peripheral cuneiform legal
traditions is more complicated, but seller-oriented deeds are attested at Susa, Ugarit,
and Nuzi. The seller orientation is rooted ultimately in Old Akkadian practice. Cf.
Mus, Studies, 22, 91 n. 4, 1068; and R. Yaron, Aramaic Deeds of Conveyance,
Bib 41 (1960) 38182.
13
Kauormulare, 25, 4344; and cf. p. 63.
14
Compare the late Neo-Babylonian declaration of sale ana m gamrti . . . iddin with
the Middle Assyrian ana m gamer . . . iddinma uappi. In the Middle Assyrian deeds
of sale, the receipt-quittance clause (e.g., annaka m amtunu mar apl zak) is nor-
mally preceded by the transfer clause (uppu laqi ) and the no-contest clause (turu u
dabbu lau). In the late Neo-Babylonian formulary, on the other hand, the receipt-
quittance clause (kasp m PN [the slave] PN [the seller] ina qt PN [the buyer]
mair eir) normally follows the declaration of sale directly. The late Neo-Babylonian
guarantee against eviction, pt s pqirnu a ina mui PN (the slave) PN (the
seller) nai the seller assumes guarantee/responsibility against a contestor/claimant
over the slave sold, must be historically related to the corresponding Middle Assyrian
clause pat puqurrn"e a PN (the slave) PN (the seller) nai the seller assumes
guarantee/responsibility for vindication of the (title of the) slave sold. The defension
clause, which often takes the place of the eviction guarantee from the time of Darius
I on, is much more closely related to the similar defension clause attested at Nuzi.
situation with its split formulary of sale for movables and immov-
ables stands isolated in cuneiform law. In the rest of the cuneiform
legal traditions movables and immovables share a common formu-
lary with a common orientation, whether from the point of view of
the seller or the buyer.
12
Petschow discusses the possibility that this
seller orientation in the sale of movables in the Neo-Babylonian for-
mulary may have been derived from the Middle Assyrian formu-
lary.
13
Besides the seller orientation, the Middle Assyrian sale formulary
shares with the late Neo-Babylonian sale of movables, a receipt-quit-
tance clause and a guarantee clause.
14
The Elephantine Legal papyri
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 29
30 rotor.s v. onorr
are also formulated ex latere alienatoris.
15
The Aramaic ::: :
16
in
the Samaria papyri corresponds to Akkadian ana kaspi iddin in late
Neo-Babylonian slave-sales. Although we have the verb positively
attested on the papyrus only twice in the declaration of sale (WDSP
20:4 and WDSP 26:2), we may infer that the Samaria papyri gen-
erally had simply :, and not :: : as in the Elephantine legal
papyri,
17
from the fact that in later references to the sale : alone
is used (WDSP 4:9; 8:7, 9),
18
from the close parallelism between the
declaration of sale in the Samaria papyri and the late Neo-Babylonian
deeds of sale, and from considerations of space. The meaning of the
clause in the Elephantine deeds of sale is we have sold and trans-
ferred.
19
But the notion of transference in the sale formulary of the
Samaria papyri is handled in a separate clause, and there with the
alienee rather than the alienor as actor: And PN (the buyer) took
possession of PN as slave in his (the sellers) presence.
The slave sold in the Samaria papyri is always placed rst in the
declaration of sale. The late Neo-Babylonian declaration of sale usu-
ally begins: PN (the seller) ina d libbu, followed by the slave sold.
But when the text lacks the voluntative phrase ina d/migir libbu
of his own free will, the object sold is regularly placed rst, and
the seller follows.
20
Middle Assyrian deeds of sale also put the object
of sale rst.
21
15
Yaron, Aramaic Deeds of Conveyance, Bib 41 (1960) 250, and see pp. 38182
for a comparative discussion. The few hints of the formulation of Palestinian deeds
of purchase (:,: c:) in the biblical period seem to point to a buyer orienta-
tion, e.g., Jer. 32:716, 25, 4344; Ruth 4:911. But cf. Genesis 23, 1 Kgs 21:6,
and 1 Chron. 21:2225 from a seller orientation. Other indications could point to
a double formulation (e.g., Gen. 47:1926; Exod. 21:2, 7; Lev. 25:1417, 2930,
50; Deut. 28:68 etc.).
16
Compare TAD B3.12:24 (Kraeling 12).
17
E.g., TAD B3.4 (Kraeling 3) 3, 10, 13; B3.12 (Kraeling 12) 3, 67, 9, 15, 17,
25, 29. Mus, Studies, 34 n. 1, rightly compares iddinma uappi from the Middle
Assyrian sale formulary.
18
:[: ]: [PN-:] could conceivably be reconstructed in WDSP 26:3, but I
would prefer a dierent interpretation.
19
So correctly E.G. Kraeling, The Brooklyn Museum Aramaic Papyri (New Haven,
CT: Published for The Brooklyn Museum by the Yale University Press, 1953) 159.
20
Petschow, Kauormulare, 45: In vereinzelten Urkunden fehlt die Wendung ina
hd libbiu, so besonders in chaldischer Zeit (Nbk 37; 97; 201; NRV I 63; TC 12,
27; persisch NRV I 70; 78; 96) und berwiegend in Urkunden, die andere bewegliche
Sachen als Sklaven zum Gegenstande haben (z.B. Tiere: NRV I 100; 101; 103;
Camb 1 u.a.; Schie: YOS VII 173). Der Text beginnt dann regelmig (anders
Nbk 97; NRV 78) mit dem Kaufobjekt, worauf Verkufer usw. wie oben folgen.
In addition to the examples of the slave sold placed rst cited by Petschow, cf.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 30
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 31
NRV I 64, 69, 83. NRV I 85 places the slave rst, even with the voluntative phrase
coming later in the declaration of sale.
21
In fact, most cuneiform traditions, including the Old Babylonian, Middle
Babylonian, and Neo-Assyrian, put the object of sale before the seller. It is only
the introduction of the voluntative element into the sale formulary in the Neo-
Babylonian period that reverses the order of object and seller. Cf. San Nicol,
Beitrge, 158.
22
Cross, Samaria Papyrus 1, 11*. Cf. CAD s.v. aru, 103; San Nicol, NRV
I 102 n. 2; Ungnad, NRV Glossar, 64, 15051; Petschow, Kauormulare, 4553; K.L.
Tallqvist, Die Sprache der Contracte Nab-N"ids (555538 v. Chr.) mit Bercksichtigung der
Contracte Nebukadrezars und Cyrus (Helsingfors: Druck von J.C. Frenckell & Sohn,
1890) 75.
23
In speaking of Babylonian and Assyrian I am only referring to conventional
transcriptionnot trying to represent actual phonetics.
24
Cf. S.A. Kaufman, The Akkadian Inuences on Aramaic (AS 19; Chicago and
London: The University of Chicago Press, 1974), 14042, 14344; GAG 31a. All
references cited in the CAD s.v. aru, p. 103, are written M ari. Perhaps, then,
it was pronounced iwari in late Neo-Babylonian. In the light of this possibility,
the evidence of the Babylonian loans in Aramaic showing post-vocalic m > w listed
by Kaufman, Akkadian Inuences, 144 n. 24, should be reconsidered.
25
Cf. Ungnad, NRV Glossar, 5556, 15051; San Nicol, NRV I 106 n. 3;
Petschow, Kauormulare, 4553; Tallqvist, Die Sprache, 61. Ungnad treated m in this
phrase as a plurale tantum. This recalls the plural form of m in Old Akkadian
deeds, and Middle Babylonian conveyances from Alalakh. Cf. Mus, Studies, 200.
Compare also ana m gamer in Middle Assyrian deeds of sale, e.g., Ebeling, Urkunden,
83, 84.
26
Cf. CAD s.v. aru A, 93b.
27
Kauormulare, 4653.
The declaration of sale normally ends with the compound phrase
:: : : the stipulated price, the full price. Cross rst deci-
phered the form : as a direct loan from Neo-Babylonian m
ari.
22
It shows the regular correspondence of Babylonian and
Aramaic (as opposed to the Assyrian = Aramaic s).
23
It seems to
show the late Neo-Babylonian development of intervocalic (here
apparently post-vocalic!) m > w.
24
:: : is calqued on the Neo-
Babylonian m gamrti.
25
I have translated : on the basis of the
evident meaning of the verbal form in Babylonian,
26
but Petschow
has demonstrated that a dierence in legal import between the two
phrases in Neo-Babylonian is untenable.
27
To my knowledge, m
ari and m gamrti existed only as alternatives in the antecedent
cuneiform legal formularies; they were never combined as in the
Samaria papyri. Perhaps, then, it was Aramaic scribes, who rst
combined the phrases in their own adapted sale formulary.
The distribution of these two phrases in cuneiform documents
may help us locate the time of origin for the derived Aramaic sale
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 31
32 rotor.s v. onorr
formulary. The geographical distribution of these phrases appears to
be insignicant.
28
In the reign of Nebuchadnezzar II, m ari is used
almost to the exclusion of m gamrti.
29
Only in the time of Nabonidus
does m gamrti become a common alternative. This gives us a ter-
minus a quo for the borrowing. On the other hand, sale documents
of the late Persian period are formulated almost exclusively with the
term m gamrti.
30
By the Seleucid period, m ari never appears.
31
In fact, Petschow does not record any instances of the phrase m
ari after the reign of Darius I. This gives us a terminus ad quem.
(b) The Receipt-Quittance Clause
The wording, word order, and position in the formulary of the
receipt-quittance clauses in the Samaria papyri correspond closely to
that of the late Neo-Babylonian sale of movables.
32
There are two
alternate realizations of this clause in our texts. In one c:: is the
object of the participle ::,:, with the seller as explicit subject (WDSP
2:3; 21:1). The form ::,: is evidently the Aramaic translation of
mair in the late Neo-Babylonian receipt-quittance formula.
33
In the other realization of the receipt-quittance clause, c:: seems
to be the subject of :: :, (The sale price) . . . is paid, received
28
Petschow, Kauormulare, 52.
29
Petschow, Kauormulare, 51, 52.
30
Petschow, Kauormulare, 52.
31
Petschow, Kauormulare, 53. (In accordance with his tabulation on p. 52, I think
he may mean after the close of the reign of Darius I [521486].)
32
Petschow, Kauormulare, 2425, 53, distinguishes between a receipt-quittance
(Kaufpreisquittung), from the sellers perspective, and a notice of payment (Zahlungvermerk),
objectively stated action of the buyer. The Old Babylonian, Middle Babylonian,
and Neo-Assyrian formularies know a notice of payment, but not a receipt-quit-
tance. In this light, the Middle Assyrian receipt-quittance clause stands out as the
true analogue of the late Neo-Babylonian formula.
33
So Cross, Samaria Papyrus 1, 12*. To express a past transaction we might
rather have expected the perfect tense qabbil. If we are not to see in this partici-
ple a narrative past tense usage as in the Aramaic of Daniel, we may assume the
accent is on the state in which the seller holds in receipt the sale price. Perhaps,
it would be better, in that case, to parse it as a passive participle mqabbal rather
than as the active participle mqabbil. This would nd some support in the use of
::, in a similar receipt clause in a document from the Naal ever published
by H.J. Polotsky, Three Greek Documents from the Family Archive of Babatha
(Hebrew), E.L. Sukenik Memorial Volume [18991953] (ed. N. Avigad; ErIsr 8; Jerusalem:
Israel Exploration Society, 1967) 50, pl. 11. In any case, we may compare the
receipt-quittance formula from the deed of land sale Xev/Se 50 + Mur 26, lines
1112 (originally published by J.T. Milik and re-edited by A. Yardeni in Aramaic,
Hebrew, and Greek Documentary Texts from Naal ever and Other Sites, with an Appendix
Containing Alleged Qumran Texts (The Seiyl Collection II) [DJD 27; Oxford: Clarendon
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 32
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 33
Press, 1997]): ::.: : ::,: : c::: and the sum I have received, (its) price
forever. Also here, the receipt-quittance formula immediately follows the declara-
tion of sale. Compare also Mur 25 frag. 1 line 5; 32, line 2; 33, line 2. Note the
use of qubbulu in Seleucid receipt clauses (CAD s.v., p. 292). Cf. W. von Soden,
Aramische Wrter in neuassyrischen und neu- und sptbabylonischen Texten: Ein
Vorbericht, Or 37 (1968) 264; 46 (1977) 193. Could the stative/passive formation
of qubbulu in Seleucid Babylonian reect Aramaic mqabbal in receipt clauses? The
D stem of qbl with this meaning is rare in Ocial Aramaic. It occurs twice in
TAD A4.2 (Cowley 37) 3 (a letter). Normally, lq is preferred.
34
See Petschow, Kauormulare, 16 n. 27, 5355; and Mus, Studies, 12526.
Compare also apil zaku in Middle Assyrian slave sales, e.g., KAJ 169:14; 171:23 in
Ebeling, Urkunden, 82, 83. KAJ 170:18 (p. 84) has mar apl zak. The corre-
sponding clause in the Neo-Assyrian slave sales is a notice of payment, kaspu gam-
mur taddin. Cf. ARU nos 472556 passim.
35
Cf. Cross, Samaria Papyrus 1, 12*; CAD s.v. eru, pp. 411b12a.
36
The root mkr is rare in Aramaic. It does, however, appear in Syriac. But
there it has the restricted meaning to acquire a wife (by paying the mohar, cf. the
Peshitta of 2 Sam. 3:14), to betroth, or to be betrothed. J. Levy, Chaldisches
Wrterbuch ber die Targumim und einen grossen Theil des rabbinischen Schriftums (Leipzig:
Baumgrtner; London: David Nutt, 1868) 2.36, cites one example of the Aramaic
verb mkr. It occurs in the Dt stem, and has the same meaning as the verb in
Syriac. The range of use of mkr in Aramaic does not t the context in the Samaria
papyri, where the subject of :: is c::.
37
E.g., eir nadin mair in V. Scheil, Carptim, RA 24 (1927) 39:19; eirtu marti
in BE X 73:6.
(WDSP 3:3; 7:4). Though I imagine this to be the less common vari-
ant of the receipt-quittance clause in the Samaria papyri, it may be
the older formula of the two. The phrase :: : corresponds to
the common late Neo-Babylonian receipt-quittance formula mair eir,
which replaced the earlier mair/nadin zaku.
34
The form : is a
direct Neo-Babylonian loan, albeit shaped into an Aramaic passive
participle.
35
Similarly, it is hard to dissociate :: from Akkadian mair.
36
Although the normal Neo-Babylonian sequence is mair eir, the
sequence found in the Samaria papyri is also attested.
37
The inexact correspondence of the Aramaic passive with the
Akkadian stative may account for the hesitation between these alter-
native realizations. A Babylonian, using a stative verb form, could
sustain a dual focus on the seller and the sale price. The statives eir
mair show number and gender concord with the seller. But when
the statives are rendered by the Aramaic passive, though a certain
formal resemblance is maintained, the seller is lost from perspective.
The regular occurrence of the receipt-quittance clause in our texts
also helps to establish the date of the crystalization of this Aramaic
sale formulary. In the early Neo-Babylonian period, this clause is
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 33
34 rotor.s v. onorr
not found in documents of sale.
38
It appears in isolated examples
during the Chaldean dynasty, but becomes an obligatory con-
stituent of the formulary for the sale of movables
39
beginning with
the reign of Darius I.
40
(c) The Defension Clause
The defension clause is a regular constituent of the sale formulary
of the Samaria papyri.
41
We do not nd the defension clause in the
late Neo-Babylonian deeds of sale of movables until the eighth year
of Darius I (514 ncr).
42
This provides another terminus a quo for the
crystalization of the Aramaic formulary. The late Neo-Babylonian
defension clause runs typically as follows: ina mu paqru ina mui
the Slave ittabu, the Seller the Slave umarraqamma ana the Buyer
inamdin When a claim arises concerning the slave, the seller will
clear the slave of all claims and give him back to the buyer. This
in turn is apparently derived from an earlier defension clause that
we nd attested at Nuzi: umma the Slave pqirna irtai the Sellers
uzakkma ana the Buyer inaddin.
43
Here as elsewhere the Aramaic
38
Petschow, Kauormulare, 53.
39
The Neo-Babylonian sale of immovables generally lacks the receipt-quittance
clause (cf. Petschow, Kauormulare, 53).
40
See the discussion and tabulation of Petschow, Kauormulare, 5354, and 55 n. 37.
41
Yaron, Introduction to the Law of the Aramaic Papyri (Oxford: Clarendon Press,
1961) 89, correlates the presence of a defension clause with the presence of a con-
sideration. H.Z. Szubin and B. Porten, Litigation Concerning Abandoned Property
at Elephantine (Kraeling 1), JNES 42 (1983) 284, try to narrow the criteria for
the appearance of the defension clause: Where title to the property is certain, a
defension clause is not mandated. . . . a defension clause is relevant and material
only where the alienor cannot produce clear title, such as in situations of aban-
doned property. Conversely, a defension clause is absent where clear title is pre-
sent. As far as I can see, this analysis can only apply to the Elephantine legal
papyri (TAD B3.2, B3.4 = Kraeling 1 and 3). It does not hold for the defension
clause at Nuzi or in the late Neo-Babylonian sale of movables in the Persian period.
It certainly does not hold for the Samaria papyri. Either Yarons analysis is cor-
rect or the defension clause may simply not have been a regular constituent of the
formulary for conveyances at Elephantine.
42
Petschow, Kauormulare, 57.
43
R.H. Pfeier and E.A. Speiser, One Hundred New Selected Nuzi Texts (AASOR
16; New Haven, CT: American Schools of Oriental Research, 1936) 37:1317,
p. 31; cf. also 19:1112; 21:1012; 34:1922; 42:2326; 52:1013; 55:2228;
58:1417; 65:2023; 66:2729; 96:1114. Cf. Koschaker, NRUA, 32 and text nos
22:1114; 24:1316; 25:1619; 26:1619 (all from Nuzi). For a similar clause in
Old Babylonian Susa texts with the verbal sequence ubbubu = nadnu cf. Yaron, The
Law of the Aramaic Papyri, 11718; and CAD s.v. ebbu, 7a.
SHIFFMAN_f3_22-49 12/2/02 5:17 PM Page 34
+nr s.v.ni. r.rvni .xr +nr n.nvroxio-.n.vr.x svvniosis 35
44
Cf. Mus, Studies, 2223.
45
First suggested by A.T. Clay, Legal and Commercial Transactions, Dated in the Assyrian,
Neo-Babylonian and Persian Periods, Chiey from Nippur (= BE VIII
1
; Philadelphia:
Department of Archaeology, University of Pennsylvania, 1908) 22 n., and then fol-
lowed by many others, e.g., P. Koschaker, Babylonisch-Assyrisches Brgschaftsrecht: Ein
Beitrag zur Schuld und Haftung (Leipzig/Berlin: B.G. Teubner, 1911) 19293; San
Nicol, Schlussklauseln, 171 n. 79; B. Landsberger, Bemerkungen zu San Nicol und
Ungnad, Neubabylonische Rechts- und Verwaltungsurkunden, Bd. I I.2, ZA 39
(1930) 28687; E.Y. Kutscher, On the Terminology of Documents in Talmudic
and Gaonic Literature (Hebrew), Tarbiz 19 (1947/8) 58; JAOS 74 (1954) 247; von
Soden, Aramische Wrter, Or 35 (1966) 18; 46 (1977) 190; Mus, Studies, 18788
n. Cf. also the discussion of Greeneld, Babylonian-Aramaic Relationship, 473.
46
Compare Xev/Se 50 + Mur 26, lines 15, 19 (defension clause); in Syriac,
P.Dura 28 (243 cr), lines 1215 (defension clause); Qumran Aramaic: ,: ::
[]::. Also, with them he burnishes the clouds (11QtgJob 29:1 = v+ Job 37:11);
Targumic Aramaic : :..: (:,:: :
XXVII
e
dynasties) (Bibliotheque de lEcole des hautes etudes. IV
e
section, Sciences
historiques et philologiques: fasc. 300; Paris: H. Champion, 1953) no. 18:1314.
sister, near or far, shall not be able to sue you (TAD B2.1:9, 2.2:13;
3.6:5; 5.1:6, 5.5:5; cf. 2.7:10; 3.2:9). Does the term mean relative
or alien, related or not related,
47
close or distant relative?
48
None of the above. It was R. Yaron who called attention to the
identical biblical polar pair,
49
which occurs ten times, in almost as
many associations, and in each case the meaning is one of locus,
a person or territory which is to hand and one which is far away
(1 Kgs 8:46, Ezek. 22:5, 2 Chron. 6:36 [lands]; Isa. 57:19 [persons];
Deut. 13:8 [nations]; Jer. 25:26 [kings]; Esth. 9:20 [ Jews]; Dan. 9:7
[Israel]; Jer. 48:24 [cities]). My favorite is Prov. 27:10A close
neighbor is better than a distant brother. Thus in our contracts the
expression should be translated, son or daughter, brother or sister,
whether near or far. Leaving the territory did not deprive the plain-
ti of his potential rights.
4. The fourth expression is the polar pair qyt/f[w tdj rps, a
new or old document, which occurs in some ve conveyances with
the meaning, no one shall be able to produce a new or old docu-
ment to contest this conveyance (TAD B2.3:16, 2.7:12; 3.10:22,
3.11:15, 3.12:29). Might new mean a future (document)? Again,
biblical parallels come to our assistance. The pair dj:y indicates
that new means recentin the blessing of abundance, you shall
have to clear out the old (grain) to make room for the new (which
is full-grown [Lev. 26:10]), and the delights of the maiden, all choice
fruits, both new (freshly picked) and old (long-stored) have I kept,
my beloved, for you (Song 7:14). In a similar vein, the biblical pair
rkw yy (Num. 6:3) is rendered by the Targum as qyt[w tdj rmj,
new and old wine. This interpretation is borne out by an abnor-
mal hieratic parallelyou have given us their old documents and
their new documents which have been written.
50
5. There are over twenty-ve contracts whose intact ending con-
tains the statement PN wrote this document (in Elephantine/Syene)
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 67
68 nrz.rrr ron+rx
pk PN.
51
From the rst, this term was translated at the dictation
of
52
and the clause came to be known as the dictation clause.
53
Yet the one Hebrew parallel to Aramaic pk occurs in a caption to
a list of soldiers who deserted Saul for David, in accordance with
the word of the Lonr (hwhy ypk) (1 Chron. 12:24). Our passages are
thus best translated at the instruction of. The average person was
not so versed in legal terminology that he could have dictated a
document to a professional scribe. Had that been the case, the term
would not have been pk but pm, from the mouth of. This is
the term that appears several times in the account of Jeremiahs
dictation of all his prophecies to the scribe Baruch son of Neriah
( Jer. 36:4, 27, 32; 45:1). When the royal ocials heard the contents
of the scroll, they interrogated Baruch to ascertain just how he
copied down the prophets words. From his mouth (wypm)? they
queried. To which the scribe replied, From his mouth (wypm). He would
call out to me all these words and I would write them down on a
scroll with ink (36:1718). This is not the way it happened at
Elephantine.
The Elephantine scribes were skilled practitioners who knew their
stylistics. I cite two examples. A grant written by Nathan son of
Ananiah shows that it is possible to produce a document with ortho-
graphic peculiarities and even spelling errors
54
and yet arrange the
transfer clauses in a chiastic arrangement, lacing through them the
Leitwort bhy, give seven times (TAD B2.7:27):
55
a. I gave you the house which Meshullam son of Zaccur son of Ater,
Aramean of Syene, gave me
b. and a document he wrote for me about it.
c. and I gave it to Miptahiah in exchange for her goods which
she gave me.
d. I consumed them but did not nd silver or goods
to repay you.
51
TAD B2.1:15, 2.2:17, 2.3:28, 2.4:16, 2.6:37, 2.7:17, 2.8:12, 2.9:16, 2.10:17,
2.11:15; 3.1:21, 3.2:10, 3.4:23, 3.5:22, 3.6:16, 3.8:43, 3.9:9, 3.10:23, 3.11:17, 3.12:33,
3.13:12; 4.6:18; 5.3:6, 5.5:11; [6.3:16], 6.4:9; D2.14 frag. d:2; pk is missing in
B3.3:14.
52
Sayce and Cowley, Aramaic Papyri Discovered at Assuan, A:15 (= TAD B2.1:15),
B:17 (= TAD B2.2:17), etc.
53
R. Yaron, Law of the Aramaic Papyri, 15.
54
Cowley, Aramaic Papyri, 37.
55
Porten, Elephantine Papyri in English, B29.
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 68
rrrrn.x+ixr .xr +nr ninrr 69
56
B. Porten, Elephantine Aramaic Contracts and the Priestly Literature, Minah
le-Naum: Biblical and Other Studies Presented to Nahum M. Sarna in Honour of his 70th
Birthday (eds M. Brettler and M. Fishbane; JSOTSup 154; Sheeld: Sheeld
Academic Press, 1993) 260.
c. I gave you this house in exchange for your goods worth 5
karsh.
b. and I gave you the old document which that Meshullam wrote
for me.
a. this houseI gave it to you and withdrew from it.
What we see at work here is the principle of graduated repetition,
where each time the key word recurs it is accompanied by an addi-
tional piece of necessary informationI gave a house which he gave
me (a) in exchange for goods which she gave, worth 5 karsh (c); I
gave an old document (b), gave the house and withdrew from it (a).
Rather than making a statement with a single verb and, where
required, multiple subjects, objects, or modiers, the writer repeated
the verb twice, thrice, or as often as necessary in order to highlight
each one of the nominal, adjectival, or adverbial elements.
56
This principle is well articulated in a bequest drawn up by that
other major Elephantine scribe, Haggai son of Shemaiah (TAD B3.10).
Refrain-like, he repeats the key word bhy, give nine times in seven-
teen lines in a variety of combinationsI gave to you (Transfer I),
whose value I gave (Pedigree), I gave it to you (Transfer II);
which I gave to Jehoishma my daughter in aection, which I,
Anani, gave to Jehoishma my daughter (Measurements); which I,
Anani, gave to Jehoishma my daughter (Boundaries); I, Anani son
of Azariah, gave it to you in aection (Transfer III; inclusio to
Measurements and Boundaries); I, Anani gave it to Jehoishma my
daughter at my death in aection, I gave it to her at my death
(Transfer IV). Each repetition incrementally enhances the bequest,
until a climatic letdown: (1) I gave to you what I bought; (2) I gave
to Jehoishma my daughter in aection; (3) I, Anani, gave to Jehoishma; (4)
I, Anani son of Azariah, gave it to you in aection; (5) I, Anani gave
it to Jehoishma my daughter at my death in aection. Three times he
says that the transfer is being made in aection, as bets a bequest
from a father to a daughter. Only at the end does he inform her
that this bequest is not made from this day and forever but at his
death. Such graduated repetition is well-attested in the Priestly
Code, particularly in Leviticus 115 where key expressions recur in
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 69
70 nrz.rrr ron+rx
varying verbal combinations (e.g., Lev. 1:9, 13, 17; 5:6, 10, 13).
57
I
cite one example of a threefold incremental repetition (Lev. 6:2, 5, 6):
and the re on the altar shall be kept burning on it.
The re on the altar shall be kept burning on it; it shall not go out.
A perpetual re shall be kept burning on the altar; it shall not go out.
I have elsewhere shown in detail that just as the priestly literature
illuminates this bequest drawn up by Haggai so Elephantine for-
mulae may be invoked to uncover the broader meaning of the grant
of priestly emoluments in Num. 18:819.
58
We turn now to the subject of religion. Though we have no reli-
gious literature as such, no Hebrew Scriptures, there is not a little
documentation for religious institutions and practices. We shall dis-
cuss ve itemsPassover, Sabbath, libation, impurity, and dreams.
1. The Passover Papyrus (TAD A4.1) is one of the better known
Aramaic texts, even though the word Passover does not occur in
it at all, as Kraeling was eager to point out. For him its restoration
was even unlikely. Most of the right half of the papyrus is miss-
ing and Kraeling thought that 10 cm was missing at the left edge
as well, so that realistic restoration was out of the question.
59
After
several failed attempts by P. Grelot to calculate the original width
of the papyrus, I was able, through a combination of papyrological
and epistolary considerations, and after having examined the papyrus
itself, to propose a minimalist solution.
60
It was thus possible to restore
the missing right half with great probability by relying heavily on
biblical verses, primarily in Exodus 1213 and Leviticus 23. We note
six restorations, ve biblically based:
1. wd[b[ ajsp] [ob]serve [the Passover] Num. 9:23
2. wdb[ ayryfp yz agj] [observe the Feast of Lev. 23:6
[wlka ryfp mwy t[b Unleavened Bread;
seven days eat mazzoth]
57
See M. Paran, Forms of the Priestly Style in the Pentateuch: Patterns, Linguistic Usages,
Syntactic Structures (Hebrew) ( Jerusalem: Magnes Press, 1989) 17578.
58
Porten, Elephantine Aramaic Contracts and the Priestly Literature, 26168.
59
Kraeling, Brooklyn Museum Aramaic Papyri, 9395.
60
B. Porten, The Aramaic Passover Papyrus: Physical Format and Textual
Reconstruction, Actes du XV
e
Congrs international de Papyrologie: BruxellesLouvain, 29
aout3 septembre 1977. Part 3: Problemes generaux. Papyrologie litteraire (eds J. Bingen and
G. Nachtergael; Papyrologia Bruxellensia 18; Brussels: Fondation egyptologique
Reine Elisabeth, 1979) 3945; idem, Aramaic Papyri and Parchments: A New
Look, BA 42 (1979) 8892; idem, Elephantine Papyri in English, B13.
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 70
rrrrn.x+ixr .xr +nr ninrr 71
61
The recent attempt at restoration by P. Frei (Persian Imperial Authorization:
A Summary, Persian and Torah: The Theory of Imperial Authorization of the Pentateuch
[ed. J.W. Watts; SBLSymS 17; Atlanta: Scholars Press, 2001] 16) need be rejected
out of hand. His restoration would translate, [I have given the order that you,
Arsames, should say to the Jew]ish [garrison.] Such a statement would never have
been followed by the next word preserved in our text, namely, the transition marker
t[k, now, but by the instruction itself.
62
Kraeling, Brooklyn Museum Aramaic Papyri, 96.
3. [wdb[t l]a hdyb[ work do n[ot do] Exod. 12:15;
Lev. 23:8
4. rymj yz [dnm lkw anything of leaven do
[wlkat l]a n[ot eat] Exod. 12:20
5. [kytbb yzjty law] [and let it not be seen in
your houses] Exod. 13:7
6. kynwtb wl[n[h] [br]ing into your chambers
The operative line of the letter (line 3) has no parallel, cannot be
condently restored, and so we do not know exactly why the letter
was written.
61
But Ezra 6:1922 presents us with a contemporary
account that reports three features identical with those in our letter:
performance of the Passover, purication, and observance of the
Feast of Unleavened Bread. Does the purication in our letter refer
to the Passover sacrice or to the Feast of Unleavened Bread? A
private ostracon letter that advises one Hoshaiah to personally look
after the kids, inquires when will you perform the Passover? (TAD
D7.6:89 [COS 3.87A]). Paleographically, this letter is dated some
fty years before the Passover Papyrus and this query has been
taken to mean that, following Deut. 16:1, the exact date of the
Passover had not yet been xed.
62
Yet, other explanations for this
slipped-in query are possible. Without a calendar, or guided only
by the Egyptian calendar, the writer may have sought to know the
day of the festival, or more probably, the hour at which the sacrice
would take place. Perhaps he wanted to get home in time. Alternately,
he may have wanted to know if a second Adar was to be interca-
lated and Passover delayed by a month.
The opposite of purity, of course, is impurity and one ostracon
admonishes the unknown recipient, Do not dispatch to me bread
without it being sealed. Lo, all the jars are impure. Behold, the bread
which [yo]u dispatch[ed] to me yesterday is im[pure] (TAD D7.44).
The association of unsealed jars with impurity is encountered in
Num. 19:1415. According to that law, if a man dies in a tent, any
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 71
72 nrz.rrr ron+rx
unsealed vessel there becomes impure. We may thus conjecture that
a recent death had occurred in the household of the writer in Ele-
phantine and so the bread he was receiving at Syene in unsealed
vessels was impure.
63
It is also the ostraca that inform us of the Sabbath.
64
Only one
of the half-dozen or so occurrences (TAD D7.10:5, 7.12:9, 7.28:4,
7.35:7, and perhaps 7.48:5) gives clear context and the result is not
a little puzzling. An unknown writer tells Isla, Now, behold, legumes
I shall dispatch tomorrow. Meet the boat tomorrow on Sabbath.
Lest, if they get lost, by the life of YHH, if not yo[ur] soul I shall
take. Do not rely on Meshullemeth or on Shemaiah (TAD D7.16:15
[COS 3.87G]). Here we have an oath taken hhylyj that if the legumes
get lost, the writer will kill Isla, so she better meet the boat on the
Sabbath. A couple decades later, Nehemiah would admonish his
Judean countrymen for treading winepresses and transporting pro-
duce on the Sabbath and prevented Tyrian merchants from bring-
ing sh and other wares into Jerusalem on that day (Neh. 13:1522).
In our case a Jew would be handing over legumes on the Sabbath
to an Egyptian boatman on the banks of Syene who would be turn-
ing them over to a Jewess on the shores of Elephantine. Would
Nehemiah have approved of this transaction?
The concave of an ostracon discovered by Sayce 100 years ago,
delivered by him to the Bodleian Library but since vanished, has
been interpreted to read, See to the slave-girl Uriyah gave me for
the weaving. Give her to Gemaryah son of Ahyo and let him deter-
mine her wages and send her back to Uriyah.
65
Following sugges-
tions by S.A. Kaufman that the enigmatic atnj should be derived
from nj and explained as a free-will gift and aksn and ark
should be assigned their patent meanings, libation and beer,
respectively (TAD C3.13:7; A6.9:3; C3.11:2, otherwise D7.20:3), I
have placed these lines at the beginning and not in the middle of
the letter (as bets the concave) and translated Now, regard the
63
For full discussion of the laws of impurity as relating to this ostracon, see
B. Porten and A. Yardeni, Ostracon Clermont-Ganneau 125(?): A Case of Ritual
Purity, JAOS 113 (1993) 45156.
64
For a comprehensive discussion of the Sabbath in the Elephantine ostraca, see
L. Doering, Schabbat: Sabbathalacha und -praxis im antiken Judentum und Urchristentum
(TSAJ 78; Tbingen: Mohr Siebeck, 1999) 2342.
65
J. Lindenberger, Ancient Aramaic and Hebrew Letters (SBLWAW; Atlanta: Scholars
Press, 1994) no. 18:1012.
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 72
rrrrn.x+ixr .xr +nr ninrr 73
66
Porten, Archives from Elephantine, 275.
67
The Structure and Orientation of the Jewish Temple at ElephantineA
Revised Plan of the Jewish District, JAOS 81 (1961) 3842.
68
Une Communaut Judo-Aramenne lphantine, en gypte, aux VI
e
et V
e
sicles av.
J.-C. (London: Published for the British Academy by H. Milford, Oxford University
Press, 1915) opp. p. 14.
oir+ which Uriah gave me for the libation. Give it to Gemariah son
of Ahio so that he may prepare it from the beer and bring it to
Uria (TAD D7.9:13 [COS 3.87I]). Beer as libation is what we nd
in Num. 28:7 as part of the twice-daily regular oering in the
Tabernacle. The word r[, prepare occurs frequently in cultic
contexts with regard to arranging wood and sacrice on the altar
(Gen. 22:9; Lev. 1:7; 1 Kgs 18:33), lamps in the lampstand (Exod.
27:21; Lev. 24:34), and bread of display on the table (Exod. 40:23;
Lev. 24:8). It is used here for the rst time with liquid. In another
ostracon, one named Uriah was associated with the house of the
House of YHH (TAD D7.18). Was he a priest?
An ostracon, whose interpretation had formerly been a matter of
dispute,
66
reads on the concave, Now, lo, a dream I saw and from
that time I am very hot. May Jamoliah see my welfare, and con-
tinues on the convex, Now, if you wish, do not sell them. Let the
children eat (them). Lo, there is no remainder of cucumbers (TAD
D7.17). From the agitated to the ordinary with a ip of the pot-
sherd! The two best-known dreams in the Bible are divine messages
sent to pagan monarchs and interpreted by inspired Israelites.
Nebuchadnezzar used the same expression as our writer, tyzj lj,
a dream I saw (Dan. 4:2). Both Pharaoh and Nebuchadnezzar
were agitated by their respective dreams, expressed in both cases
by the root [p (Gen. 41:8; Dan. 2:1, 3). Joseph introduced his inter-
pretation with the words, God will answer Pharaohs welfare, (Gen.
41:16), using the same word as in the ostracon, wl/l. Like God
for Pharaoh, Jamoliah was concerned for the welfare of our anony-
mous dreamer.
The highlight of the Elephantine Jewish community, of course,
was its Temple. The rst article I published was over forty years ago,
in 1961, where I rejected the plan Kraeling drew up for the Temple
and its environs and presented a revised plan.
67
The initial attempt
at a plan was made by A. van Hoonacker who delivered the Schweich
Lectures for 1914, in French nonetheless (Fig. 1).
68
His plan was
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 73
74 nrz.rrr ron+rx
then reproduced by Cowley in 1923 (Fig. 2).
69
Hoonacker cleverly
recognized that if all four house borders shifted 90 from one
document to another, then the true geographical direction must lie
at a 45 angle in between. Since the documents used the terms
above and below to indicate north and south, both schol-
ars assumed that a Nile orientation was being followed and that
above meant upstream, that is south, and below, down-
stream = north. When Kraeling published his documents in 1953
he argued that above meant north and below meant south
since the location of his tmy, district; isle of Khnum below the
house of Anani (TAD B3.4:8; cf. 3.5:8, 3.10:10, 3.11:6, 3.12:2021)
accorded well with the archeological presence of the temple of Khnum
69
Aramaic Papyri, 13.
Figure 1. Disposition of houses according to Hoonacker.
N.
W.
S.
E.
Egra.
Egra.
Zecharja
Hanul?
b.-Zech.
Qonja.
Peftonit
Espemet.
Maseja
Mibaja.
Dargman.
Hoea
b.-Urija.
Jezanja.
Mardouk?
bar-Pali.
Jair
b.-Penulja.
Meoullam.
Gadol
b.-Oea.
R
u
e
r
o
y
a
l
e
(
a
b
l
m
h
r
a
)
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 74
rrrrn.x+ixr .xr +nr ninrr 75
70
Brooklyn Museum Aramaic Papyri, 81.
south of the house ruins. Incorporating his new nds, he proposed
a revised plan (Fig. 3).
70
As I pondered his plan, I realized that it
was in need of further revision, one that would yield a new orien-
tation. We are dealing with the location of four houses, each one
having four neighbors. Three of them border on the Temple. If
arranged properly, they may have archeological signicance and lead
to the discovery of the Temple.
The rst house of Mahseiah which he gave to his daughter
Mibtahiah appears in three documents (TAD B2.13). In two of
them, the house of Mibtahiahs husband Jezaniah is located to the
south (TAD B2.12 [Fig. 4]), while the third orients the houses 90
Figure 2. Disposition of houses according to Cowley, following Hoonacker.
N
TEMPLE Gadol
b. Oshea'
13
S
t
r
e
e
t
K
i
n
g
s
S
t
r
e
e
t
Meshullam
b. Zaccur
13
Yeosh
b. Penuliah
13
X. b. Palo
13
Hosea
b. Uriah
25
8
Dargman
6
8
Jezaniah
b. Uriah
9 & 25
8
Mahseiah
b. Yedoniah
5
1
1
c
u
b
i
t
s
8
Zechariah
b. Nathan
& Hazul
b.Zech.
25
8
oniya
b. Zadok
5
p
o
r
t
i
c
o
S
t
r
e
e
t
1
3
c
u
b
i
t
s
1
h
a
n
d
Pef'onith
Espeme
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 75
76 nrz.rrr ron+rx
Figure 3. Disposition of houses according to Kraeling, equating above with
north and below with south.
TEMPLE
OF
YAHU Dargman
b. Harshin
Hosha'
b. Uriah
Jezaniah
b. Uriah
Maseiah
b. Yedoniah
Pef'onith
Espemet
Qoniya
b. Zadok
Zekariah
b. Nathan
Hal
b. Zekariah
R
o
a
d
S
t
r
e
e
t
T
m
o
f
K
h
n
u
m
M.N.
Ananiah
b. Azariah
Treasury
of the
King
Satibar
b. Bzw
Marduk
b. Palo
Meshullam
b. Zakkur
Gaddl
b. Osha'
Ye"osh
b. Penuliah
S
t
r
e
e
t
S
t
r
e
e
t
?
S
t
r
e
e
t
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 76
rrrrn.x+ixr .xr +nr ninrr 77
to the east (TAD B2.3 [Fig. 5]). The true orientation of the house
of Jezaniah, then, has to be southeast (Fig. 6) and not northeast, as
Kraeling placed him (Fig. 3). East of the house of Jezaniah lay the
Temple of YHW (TAD B2.10 [Fig. 7]), whose true position would
thus be southeast of those houses (Fig. 8). Like the house of Jezaniah,
a house of Meshullam also passed into the hands of Mibtahiah. That
house lay to the north of the Temple (TAD B2.7 [Fig. 9]). Kraeling
tilted it 45 NW and positioned it on the short end of the Temple
(Fig. 3). I observed that if it were placed on the long end (Fig. 10),
that end would come to roughly 60 cubits. This would accord well
with the dimensions of the fourth house, that of Anani, which lay
to the east and north of the Temple (TAD B3.45 [Fig. 11]). That
side would come to an estimated 20 cubits and a 60 20 building
would be in imitation of Solomons Temple (1 Kgs 6:2). Its NE
orientation would point toward Jerusalem (Fig. 12) and recall the
biblical passages in 1 Kgs 8:48 and Dan. 6:11 where Jews exiled
abroad turn in prayer toward Jerusalem. In my 1961 article, I stated,
Archeological conrmation of the Temple location and orientation
must await further excavation. Here was a nd waiting to be made,
so imagine my surprised delight when in 1998 I received and read
an article by C. von Pilgrim, Director of the Swiss Archaeological
Figure 4. House of Mahseiah/Mibtahiah with husband Jezaniah positioned
to the south (TAD B2.12).
Street aq
N
W
S
E
Below hfml
Above hl[ml
tmpsa
Espemett
hynwq
Konaiah
hysjm
Mahseiah
hyjfbm
Mibtahiah
mgrd
Dargamana
hyrkz
Zechariah
hynzy
Jezaniah
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 77
78 nrz.rrr ron+rx
Figure 5. House of Mahseiah/Mibtahiah with Jezaniah positioned to the east
(TAD B2.3).
Figure 6. True orientation of house of Mahseiah/Mibtahiah with Jezaniah
located southeast.
mgrd
Dargamana
Below hfml
N
E W
S
Above hl[ml
hynzy
Jezaniah
hyrkz
Zechariah
hynwq
Konaiah
tmpsa
Espemet
S
t
r
e
e
t
a
q
<11>
hysjm
Mahseiah
hyjfbm
Mibtahiah
<
1
3
>
tmpsa
Espemet
hynwq
Konaiah
hysjm
Mahseiah
hyjfbm
Mibtahiah
hynzy
Jezaniah
hyrkz
Zechariah
mgrd
Dargamana
S
t
r
e
e
t
a
q
s
H
i
g
h
w
a
y
a
k
l
m
j
r
a
why yz arwga
Temple
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 79
80 nrz.rrr ron+rx
Figure 8. True orientation of house of Jezaniah with Temple located southeast.
hynzy
Jezaniah
hysjm
Mahseiah
hyjfbm
Mibtahiah
lwxh
Hazzul
[wh
Hosea
why yz arwga
Temple
N
E W
S
NE
SE
K
i
n
g
s
H
i
g
h
w
a
y
a
k
l
m
j
r
a
Figure 9. House of Meshullam/Mibtahiah positioned to the north of the
Temple of YHW (TAD B2.7).
N
W
S
E
Below hfml
Above hl[ml
hhy yz arwga
Temple
arudj
wrj
Jaush
way
Gaddul
lwdg
S
t
r
e
e
t
a
q
Meshullam
lm
Mahseiah
hysjm
Mibtahiah
hyjfbm
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 80
rrrrn.x+ixr .xr +nr ninrr 81
Figure 10. True orientation of house of Meshullam/Mibtahiah northwest, and
along the long side of the Temple.
why yz arwga
Temple
arudj
wrj
Jaush
way
W
S
SE
E
NE
N
Gaddul
lwdg
S
t
r
e
e
t
a
q
Meshullam
lm
Mahseiah
hysjm
Mibtahiah
hyhfbm
Figure 11. House of Anani located northeast of the Temple and northwest of
(i.e. above) the tmy of Khnum (TAD B3.45).
N
NE
E
SE
S
W
W
a
y
/
T
o
w
n
o
f
K
h
n
u
m
w
n
j
y
z
y
m
t
K
i
n
g
s
S
t
r
e
e
t
a
k
l
m
q
w
aklm yz arxwa
Treasury
rbyt
Shatibara
ynn[
Anani
K
i
n
g
s
S
t
r
e
e
t
a
k
l
m
q
w
why rwga
Temple
SHIFFMAN_f4_50-84 12/17/02 9:40 AM Page 81
Mission at Elephantine and Aswan, with the non-revelatory title
Textzeugnis und archoligischer Befund: Zur Topographie Elephan-
tines in der 27. Dynastie.
71
There on p. 486 was the plan I had
drawn up in my 1961 article and on p. 491 a drawing which t the
archeological ndings into my plan (Fig. 13). And of course, there
was the Temple. But had he found it on the ground? He claims he
did and he told about it in a brief report dated 1999.
72
That year,
I visited von Pilgrim in Cairo and discussed his ndings at length.
The following year, 2000, I visited him at the site of Elephantine,
and he showed me what he felt must be meager remains of the
Temple. At that spot was found a two-line ostracon reading hl bh
r[ hptp, a payment order to give him his ration, barley. O that
we might know the identity of payer and payee!
82 nrz.rrr ron+rx
71
H. Guksch and D. Polz (eds), Stationen: Beitrge zur Kulture Geschichte gyptens
Rainer Stadelmann Gewidmet (Mainz: P. von Zabern, 1998) 48597.
72
Der Tempel des Jahwe, MDAIK 55 (1999) 14245 + Abb. 24.
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 82
rrrrn.x+ixr .xr +nr ninrr 83
F
i
g
u
r
e
1
2
.
D
i
s
p
o
s
i
t
i
o
n
o
f
h
o
u
s
e
s
w
i
t
h
T
e
m
p
l
e
o
f
Y
H
W
o
r
i
e
n
t
e
d
t
o
w
a
r
d
J
e
r
u
s
a
l
e
m
.
T
e
m
p
l
e
w
h
y
y
z
a
r
w
g
a
P
e
f
t
u
a
u
n
e
i
t
h
t
y
n
w
[
f
p
p
K
o
n
a
i
a
h
h
y
n
w
q
M
a
h
s
e
i
a
h
h
y
s
j
m
M
i
b
t
a
h
i
a
h
h
y
j
f
b
m
Z
e
c
h
a
r
i
a
h
h
y
r
k
z
H
a
z
z
u
l
l
w
x
h
J
e
z
a
n
i
a
h
h
y
n
z
y
H
o
s
e
a
[
z
h
D
a
r
g
a
m
a
n
a
m
g
r
d
a
r
u
d
j
[
w
r
j
M
e
s
h
u
l
l
a
m
m
M
a
h
s
e
i
a
h
h
y
s
j
m
M
i
b
t
a
h
i
a
h
h
y
j
f
b
m
J
a
u
s
h
w
a
y
G
a
d
d
u
l
l
w
d
g
T
a
m
e
t
t
m
t
o
r
r
w
j
y
j
p
f
m
p
w
P
a
i
&
P
e
m
e
t
W
a
y
/
T
o
w
n
o
f
K
h
n
u
m
w
g
j
y
z
y
m
t
K
i
n
g
s
S
t
r
e
e
t
a
b
l
m
q
w
K
i
n
g
s
H
i
g
h
w
a
y
a
b
l
m
j
d
a
P
a
r
n
u
&
M
r
d
a
v
a
w
d
r
m
w
w
n
r
p
S
h
a
t
i
b
a
r
a
r
b
y
t
S
h
r
i
n
e
o
f
t
h
e
g
o
d
y
t
n
j
n
q
T
r
e
a
s
u
r
y
a
k
l
m
a
r
x
w
a
W
a
y
o
f
t
h
e
g
o
d
y
t
n
a
w
m
t
A
n
a
n
i
y
n
n
[
J
e
h
o
i
s
h
m
a
[
m
y
w
h
y
K
i
n
g
s
S
t
r
e
e
t
a
k
l
m
q
w
S
t
r
e
e
t
q
w
N
I
G
S
S
T
R
A
S
S
E
D
.
K
N
I
G
S
PAE/
PAME
OR
ANANIAH
PARNU
HEILIG-
TUM
MAGAZIN
SCHUTZMAUER
G
GADDUL
GA
OA
JAUSH
O
1,2
MAHSEIAH
O
4,5
ARWODJ
S
T
R
A
S
S
E
D
.
K
N
I
G
S
Q
Z
J
12
J
3
E
S
P
E
M
E
T
Z
a
b
KONAIAH
HAZZUL
JEZANIAH
MA
MIBTA-
HIAH
M
1
,2
,7
D
A
R
G
A
-
M
A
N
A
M
46
HOSEA
SHIFFMAN_f4_50-84 12/11/02 9:04 AM Page 84
THE CORPUS OF THE QUMRAN PAPYRI
Emanuel Tov
The number of papyrus fragments found in the various sites in the
Judean Desert does not reect the same number of papyri as has
been deposited there, but for the sake of convenience we assume
that the material has decayed at the same rate in all sites, so that
we should have a good impression of the relation between the dierent
corpora of papyri from each site as they were left in situ. A dierent
issue is the relation between the numbers of the preserved papyrus
and leather fragments, as it remains dicult to assess the number
of preserved compositions on papyrus as compared with the ones
preserved on leather found at each site, for example at Qumran.
After all, papyrus was often more susceptible to decay than leather,
and one notes that literary papyri are not well preserved at all: not
even a single Qumran papyrus has its beginnings and ends preserved.
In no instance has a complete column of a papyrus been preserved
together with its top and bottom margins, while partial information
is available for 4QpapTob
a
ar (4Q196) frags 2, 17, 18. Furthermore,
because of the fragmentary condition of the papyrus and leather
fragments, the total numbers of the compositions preserved will never
be known. Our own calculations follow the insights of the scholars
who published the texts. In the case of the Qumran papyri, some
scholars combined many or possibly too many fragments as one item,
while others designated almost each individual fragment as a sepa-
rate composition. Thus a single item like 1Q69 or 1Q70, each named
1QpapUnclassied fragments may represent many more texts than
these two numbers suggest, while the minute fragments written in
the cryptA script have been presented in DJD 36
1
as thirty-six indi-
vidual texts (4Q249, 249az, 250, 250aj).
The following gures relating to all the papyri found in the Ju-
dean Desert represent the numbers of the papyrus texts, including
85
1
S.J. Pfann, Cryptic Texts; P. Alexander and others, in consultation with J. Vander-
Kam and M. Brady, Miscellanea, Part 1: Qumran Cave 4.XXVI (DJD 36; Oxford:
Clarendon Press, 2000).
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 85
86 rv.xtrr +o\
opisthographs containing two separate texts, as they appear in the
inventories based on the publications of these texts. The two sides
of opisthographs are counted as a single item even if they contain
two separate compositions, while the number in parenthesis, which
counts both sides of the opisthographs as separate items, is not taken
into consideration in the statistics.
Table 1: Hebrew, Aramaic, Greek, Latin, and Nabatean Papyri from the Judean Desert
(Listed from North to South)
Site Total No. Non-documentary Papyri Documentary Papyri
of Papyri No. % of Total No. % of Total
Jericho 23+ 0 0 23+ 100
Qumran 131 (138) 121 (128) 92 10 8
Nar 3+ 0 0 3+ 100
Ghweir 1 0 0 1 100
Murabba'at 101+ 5 5 96+ 95
Sdeir 1 0 0 1 100
ever/Seiyal 166+ 0 0 166+ 100
Mishmar 3 0 0 3 100
e"elim 3 0 0 3 100
Masada 31 (34) 3 (5) 7 28 (29) 93
Table 1 shows that the situation in Qumran diers totally from the
other sites in the Judean Desert. While in almost all the other sites
in the Judean Desert documentary papyri form the majority among
the papyrus texts, in Qumran almost all papyri are non-documen-
tary (literary). Non-documentary papyri are found in only two other
sites, Murabba'at and Masada, and there too they form a small
minority.
Table 2 compares the papyrus texts with the leather texts found
in these sites:
Table 2: Comparison of Hebrew, Aramaic, Greek, Latin, and Nabatean Papyri and Leather
Texts from the Judean Desert (Listed from North to South)
Site Total No. of Papyri Leather Texts
Papyri and No. % of Total No. % of Total
Leather Texts
Jericho 23+ 23+ 100 0 0
Qumran 930 131 (138) 14 800 86
Nar 3+ 3+ 100 0 0
Ghweir 1 1 100 0 0
Murabba'at 151+ 101+ 67 50 33
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 86
+nr conrts or +nr tvn.x r.rvni 87
Sdeir 4 1 25 3 75
ever/Seiyal 179 166+ 93 13 7
Mishmar 3 3 100 0 0
e"elim 3 0 0 3 100
Masada 45 31 70 14 30
Table 2 shows that once again the situation in Qumran diers from
that in the other sites in the Judean Desert. In Qumran the papyri
form a minority of the texts found there (14.5%), while in all other
sites except for Sdeir they form a majority (see below for Masada).
This dierence points to a very basic dierence between the Qumran
corpus and that of the other sites. The Qumran corpus contains
almost exclusively literary texts bearing witness to intellectual activ-
ity, while the other sites bear witness to the daily life carried on in
these places, though with some intellectual activity recorded as well
(only a small number of non-documentary leather texts has been
found in these sites). The leather texts from Qumran do not reect
any daily activity, with the exception of 4QRebukes Reported by
the Overseer (4Q477) and the Greek 4QAccount gr (4Q350) writ-
ten on the back of the Hebrew 4QNarrative Work and Prayer
(4Q460) frag. 9. Masada reects a situation similar to that at Qumran,
when the papyri found there are separated into two groups, the
Hebrew papyri deposited in Masada prior to the siege (2 [3]) and
the Latin and Greek documentary papyri and other material left
there by the Roman army (28 [30]), including Mas 721 r + v con-
taining one or possibly two lines of Vergil on the recto and one or
possibly two lines of an Unidentied Poetical Text on the verso. If
the Masada fragments are separated in this fashion, we note that
only a few papyri were left by the Jewish inhabitants (2 [3]) as
opposed to the fourteen Hebrew leather texts (including one Aramaic
text?) left by them.
It is hard to know how many valid parallels to the almost exclu-
sively literary corpus of Qumran are known from antiquity. Most
collections of Greek papyri from Egypt contain more documentary
than non-documentary texts, but possibly the literary texts from
Oxyrhynchus came from a specic part of that city. In Antinoopolis
Table 2 (cont.)
Site Total No. of Papyri Leather Texts
Papyri and No. % of Total No. % of Total
Leather Texts
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 87
88 rv.xtrr +o\
some 60 percent of the papyri are literary. But the most valid par-
allels are probably in the libraries which have been lost, that is, the
collections stored in Alexandria, Pergamon, and Ephesus from the
Hellenistic period, Roman libraries from later periods, and Christian
libraries from the fourth century cr in Jerusalem, Alexandria, and
Caesarea.
2
From a slightly later period derives the Nag Hammadi
literary corpus. If the Jerusalem Temple contained a library, its con-
tents are unknown, but it would have contained at least the Scripture
scrolls on leather.
Beyond these statistics, even if the number of the papyri found at
Qumran is a mere 14.5 percent of the total number of texts found
there, their number is nevertheless impressive (131 texts, mainly non-
documentary). These texts were written on single sheets as well as
on papyrus scrolls. Complete scrolls have not been preserved in
Qumran, but the dimensions of some such scrolls may be recon-
structed from the preserved fragments.
3
The majority of the Qumran papyri were written in Hebrew (in
the square and Cryptic A scripts, not in the paleo-Hebrew script),
while some were written in Aramaic. Because of their fragmentary
status, it is sometimes hard to distinguish between these two lan-
guages in badly preserved texts. Twenty-seven Greek texts have been
preserved as well. On the whole, it is impossible to assess the exact
number of the papyrus texts discovered in the caves, since many
texts are very fragmentary, and it is often hard to distinguish between
the dierent handwritings on these fragments.
Table 3 records the papyrus fragments found in six Qumran caves:
Table 3: Papyrus Fragments Found in the Qumran Caves
Cave No. of Papyri
1 3
4 86 (of which 7 opisthographs)
6 21
7 19 (all in Greek)
9 1
11 1
2
See the discussion by H.Y. Gamble, Books and Readers in the Early Church: A
History of Early Christian Texts (New Haven and London: Yale University Press, 1995)
15470, 17696.
3
For descriptions of Aramaic, Greek, and Egyptian papyrus scrolls found else-
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 88
+nr conrts or +nr tvn.x r.rvni 89
A special case is 4Q51a (4QpapUnidentied frags), consisting of two
inscribed papyrus fragments which, together with other uninscribed
fragments, were applied to the back of several columns of the bib-
lical leather scroll 4QSam
a
.
4
Frag. a contains remains of six lines,
but neither this fragment nor frag. b can be deciphered.
It is hard to know whether the preserved papyrus fragments from
Qumran represent a proportionate and coherent picture of the papyri
left behind in the caves, but a few observations should be made:
Most of the texts from Cave 6 are Hebrew papyri (twenty-one
papyri out of a total of thirty-one items), including a few biblical
papyri. This collection of texts must have derived from a special
source, dierent from the main depository of texts in Cave 4.
Cave 7 contains only Greek papyrus fragments (nineteen items),
probably mainly biblical texts.
The number of biblical papyrus fragments in Hebrew is much
smaller than the proportion of Bible fragments among the Qumran
scrolls in general, viz., merely two, three, or four biblical papyrus
texts from Cave 6 and one, two, or three papyri from Cave 4 as
opposed to some 200 biblical texts written on leather within the cor-
pus of 930 Qumran texts. For the Greek fragments the proportion
where and of the production of papyrus, see E.M. Thompson, An Introduction to Greek
and Latin Palaeography (Oxford: Clarendon Press, 1912) 4451; J. ern y, Paper &
Books in Ancient Egypt: An Inaugural Lecture Delivered at University College, London, 29 May,
1947 (London: H.K. Lewis, 1952); F.G. Kenyon, Books and Readers in Ancient Greece
and Rome (Oxford: Clarendon Press, 1951) 4074; E.G. Kraeling, The Brooklyn Museum
Aramaic Papyri (New Haven, CT: Yale University Press, 1953) 127; B. Porten and
A. Yardeni, Textbook of Aramaic Documents from Ancient Egypt, vol. 3 ( Jerusalem:
Akademon, 1993) especially p. xiii; T.C. Skeat, Early Christian Book Production:
Papyri and Manuscripts, The Cambridge History of the Bible (ed. G.W.H. Lampe;
Cambridge: Cambridge University Press, 1969) 2.5479; E.F. Wente, The Scribes
of Ancient Egypt, Civilizations of the Ancient Near East (ed. J.M. Sasson; New York:
Scribner, 1995) 4.221121; P. Posener-Kriger, Old Kingdom Papyri: External
Features, Papyrus: Structure and Usage (ed. M.L. Bierbrier; Occasional Paper 60;
London: British Museum, 1986) 2541; R.A. Caminos, Some Comments on the
Reuse of Papyrus, Papyrus: Structure and Usage, 4361; S. Talmon, in S. Talmon
and Y. Yadin, Masada VI, The Yigael Yadin Excavations 19631965, Final Reports, Hebrew
Fragments from Masada ( Jerusalem: Israel Exploration Society, 1999) 2629; J. Ashton,
The Persistence, Diusion and Interchangeability of Scribal Habits in the Ancient Near East before
the Codex (University of Sydney Ph.D. Dissertation: Sydney, 1999) ch. 2; Gamble,
Books and Readers, 4454. An attempt at a detailed technical analysis of writing on
papyrus is provided by S.J. Pfann in the introduction to 4Q249250 written in the
Cryptic A script in S.J. Pfann, DJD 36.51522.
4
F.M. Cross, D. Parry, and E. Ulrich, Qumran Cave 4.XII: Samuel (DJD 17; Oxford:
Clarendon Press, in press) pl. XXIII.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 89
90 rv.xtrr +o\
of the biblical papyri is much greater, but no exact calculations can
be made because of the lack of clarity relating to the texts from
Cave 7 which probably contain the rxx, but which have been
identied also as Enoch or parts of the New Testament.
The great majority of the Qumran papyri contain literary texts.
There are also eight documentary texts from Caves 4 and 6 (see
Table 4), but possibly they did not derive from Qumran.
Table 4: Documentary Texts from Qumran (?)
4Q347 4QpapDeed F ar, part of Xev/Se 32, and hence probably not deriv-
ing from Qumran (see DJD 27.1067)
5
4Q352 4QpapAccount of Cereal B ar or heb
4Q352a 4QpapAccount A ar or heb
4Q353 4QpapAccount of Cereal or Liquid ar or heb
4Q358 4QpapAccount F? ar or heb
4Q359 4QpapDeed C? ar or heb; cf. Xev/Se 7
6Q26 6QpapAccount or Contract
6Q29 6QpapCursive Unclassied Fragment (containing gures)
Cf. also a Greek documentary text on leather, 4QAccount gr (4Q350)
written on the back of the Hebrew 4QNarrative Work and Prayer
(4Q460) frag. 9.
6
The case of a non-Qumranic origin for the two texts from Cave
6 cannot be made conclusively, but various arguments may be adduced
for an origin beyond Qumran of the Cave 4 fragments. A. Yardeni,
DJD 27.28384, surmises that these documents were purchased
from Bedouin who attributed them to cave 4. Yardeni refers to
joins between 4QpapDeed F ar (4Q347) and Xev/Se 32 and
between 4Q359 and Xev/Se 7; to late carbon-14 dates of the doc-
umentary leather texts 4Q342 (4QLetter? ar r + v) and 4Q344
(4QDebt Acknowledgement ar), viz., late rst and early second cen-
tury cr; and to the cursive script utilized in these documentary texts
as compared with the other Qumran documents.
Table 5 records the nonbiblical Hebrew and Aramaic papyri
that are parallelled by copies on leather of the same composition.
5
J. Strugnell (personal communication, February 2000) records his reservations
regarding this conclusion, since the assumed Naal ever fragment did not come
from the same type of controlled excavations as the Naal ever papyri found by
Yadin (5/6ev).
6
For an analysis, see E. Tov, The Nature of the Greek Texts from the Judean
Desert, NovT 43 (2001) 111.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 90
+nr conrts or +nr tvn.x r.rvni 91
Table 5: Nonbiblical Hebrew and Aramaic Papyri from Qumran
Parallelled by Copies on Leather
Papyri Additional Copies on Leather
4Qpap pIsa
c
(4Q163) 4 copies of the same work or cycle: 4Q161, 4Q162,
4Q164, 4Q165
4QpapTob
a
ar (4Q196) 4 copies: 4Q197200
4QpapJub
b?,h
(4Q217, 9 copies: 1Q1718, 4Q216, 4Q218222, 11Q12
4Q223224), 4QpapJub
i
?
(4Q482), 4QpapGen
o
or
papJub
j
? (4Q483)
4QpapS
a,c
(4Q255, 4Q257) 10 copies: 1Q28, 4Q256, 4Q258264, 5Q11
4QpapD
h
(4Q273) 9 copies: 4Q266272, 5Q12, 6Q15
4QpapCal. Doc. A? 6 or 5 copies: 4Q324g, 4Q324h, 4Q313c, 4Q326,
(4Q324b), 6QpapCal. 4Q337, 4Q394 (if frags 12 indeed represent a
Doc. (6Q17) separate composition)
4Qpap apocrJer B? (4Q384) 7 copies of the same composition or cycle: 4Q383,
4Q385a, 4Q387, 4Q387a, 4Q388a, 4Q389, 4Q390
4Qpap psEzek
e
(4Q391) 4 or 5 copies of the same work or cycle: 4Q385,
4Q385b, 4Q385c, 4Q386, 4Q388
4QpapMMT
e
(4Q398) 5 or 6 copies: 4Q394397, 4Q399 as well as
possibly 4Q313 (Cryptic A script)
4QpapH
f
(4Q432) 7 copies: 1QH
a
, 1Q35, 4Q427431
4QpapM
f
(4Q496) 6 copies: 1Q33, 4Q491495
4QpapDibHam
b,c
1 copy: 4Q504
(4Q505, 4Q506)
4QpapPrFtes
c
(4Q509) 2 copies: 4Q507508
6QpapGiants ar (6Q8) 9 copies of EnGiants: 1Q2324, 2Q26, 4Q203,
4Q206 frags 23, 4Q530533
4Qpap cryptA Serekh 1 copy: 1Q28a
ha-'Edah
a-i
(if indeed these
are separate copies of this
composition)
While the number and nature of the surviving Qumran texts are to
a great extent due to happenstance, it cannot be coincidental that
the aforementioned literary compositions are, as a rule, represented
by four-ve, sometimes seven-eight, copies on leather and one (and
in some cases, two or three) on papyrus. This numerical relationship
indicates that the major material used for the literary documents left
behind by the Qumran community was leather, supplemented by a
number of papyrus scrolls. Whether these two groups of texts derived
from dierent sources is unknown, but at least a case can be made
for a separate origin of the Greek biblical fragments from Cave 7
and the collection of papyri from Cave 6.
Most of the literary papyrus texts are in Hebrew, while eight are
in Aramaic and twenty-one in Greek.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 91
92 rv.xtrr +o\
The information concerning the papyrus and leather texts of the
same composition is supplemented by reverse information concern-
ing Qumran compositions which are well represented on leather, but
not at all on papyrus (see Table 6). Thus, not included among the
papyrus scrolls, possibly coincidentally, are the following nonbiblical
compositions, of which several copies have been found at Qumran,
on leather only. These compositions are singled out because of the
relatively large number of copies witnessed for these compositions.
Table 6: Well-represented Non-biblical Compositions from Qumran Not Extant on Papyrus
Text Cave No.; No. of Copies
New Jerusalem ar and heb Cave 4: 3; Caves 1, 2 and 5: 1 each
Mysteries Cave 4: 3; Cave 1: 1
Enoch ar Cave 4: 7
EnAstr ar Cave 4: 4
Levi ar Cave 4: 6
psDaniel ar Cave 4: 3
Instruction Cave 4: 7; Cave 1: 1
Tohorot Cave 4: 4
Berakhot Cave 4: 5
Narrative and Poetic Cave 4: 3
Composition (4Q371373)
Shirot 'Olat ha-Shabbat Cave 4: 8; Cave 11: 1; Masada: 1
Mishmarot Cave 4: 9
Barkhi Nafshi Cave 4: 5
Visions of Amram ar Cave 4: 7
prEsth ar Cave 4: 6
Temple Cave 4: 1 or 2; Cave 11: 3
4QReworked Pentateuch Cave 4: 5
Ordinances Cave 4: 3
The list of nonbiblical Hebrew and Aramaic papyri which are par-
allelled by several copies on leather of the same composition (see
Table 5) leads to some further thoughts about the nature of the
complete corpus of the Qumran papyri. The majority of the papyri
(cf. the compositions listed in Table 5) are sectarian or of interest
to the Qumran community ( Jubilees and Giants). These sectarian
texts include several literary genres of the communitys writing: Rules,
halakhot, liturgical works, poetical compositions, pesharim, and sapi-
ential works. Only a small number of papyri are non-sectarian (Ara-
maic texts, Hebrew and Greek biblical texts). For the Hebrew papyri
from Qumran these data suggest a close connection between the
writing on papyrus and the Qumran community. The Appendix
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 92
+nr conrts or +nr tvn.x r.rvni 93
records some data concerning the assumed background of the indi-
vidual papyri:
a. Sectarian compositions (twenty-one texts together with papyri
[below, g] written in the Cryptic A script which are esoteric and
hence sectarian) are indicated by a number in bold face in the Ap-
pendix. The sectarian nature of these compositions is accepted by
most scholars, including that of the liturgical texts 4QpapPrQuot
(4Q503), 4QpapDibHam
b,c
(4Q505506), and 4QpapPrFtes
c
(4Q509).
Our analysis of their sectarian character usually follows D. Dimant.
7
b. Texts which according to our own analysis
8
are written in the
Qumran system, not all with the same degree of certainty regard-
ing the evidence. This group of nineteen texts partially overlaps with
the texts of sectarian content (thirteen texts of group a); for six texts,
indicated in the list with hollow characters as 4Q500, the writing
in the Qumran system is their only presumed connection with the
Qumran community.
g. Texts written in the Cryptic A script. This group is rather size-
able (according to S.J. Pfann, thirty-six papyri from Cave 4, two of
which are opisthographs containing dierent texts), but this group
may represent a much smaller number of texts. The writing in the
Cryptic A (esoteric) script reects the Qumran community.
9
This script
is described by Pfann as a development from the Late Phoenician
7
D. Dimant, The Qumran Manuscripts: Contents and Signicance, A Time to
Prepare the Way in the Wilderness: Papers on the Qumran Scrolls by Fellows of the Institute
for Advanced Studies of the Hebrew University, Jerusalem, 19891990 (eds D. Dimant and
L.H. Schiman; STDJ 16; Leiden: Brill, 1995) 2358. The discussion is also advanced
much by the analysis by C.A. Newsom, Sectually Explicit Literature from Qumran,
The Hebrew Bible and Its Interpreters (eds W.H. Propp et al.; Winona Lake, IN:
Eisenbrauns, 1990) 16787.
8
See Further Evidence for the Existence of a Qumran Scribal School, The
Dead Sea Scrolls: Fifty Years After Their Discovery: Proceedings of the Jerusalem Congress, July
2025, 1997 (eds L.H. Schiman, E. Tov, and J.C. VanderKam; Jerusalem: Israel
Exploration Society and The Shrine of the Book, Israel Museum, 2000) 199216.
9
The main compositions are:
4Qpap cryptA Midrash Sefer Moshe (4Q249)
4QcryptA Words of the Maskil (4Q298)
4QcryptA Lunisolar Calendar (4Q317)
Several more fragmentary groups of inscribed remains are only tentatively identied:
4Qpap cryptA Serekh ha-'Edah
ai
(4Q249ai)
4Q249jz: sundry small papyrus fragments
4Qpap cryptA Text Concerning Cultic Service A, B (4Q250, 250a)
4Q250bj: sundry small papyrus fragments
4QcryptA Miqat Ma'ae Ha-Torah
g
? (4Q313)
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 93
94 rv.xtrr +o\
scripts, and it is used for several texts of a Qumran sectarian nature
as well as for other texts which must have had a special meaning
for the Qumran community.
10
According to Pfann, this script was
used especially by the Maskil, see especially 4QcryptA Words of the
Maskil to All Sons of Dawn (4Q298).
Several papyrus fragments in the list are irrelevant to the ques-
tion under discussion: unclassied and unidentied fragments (21),
texts in Greek (21) and Aramaic (8), and biblical texts (26). If these
groups are disregarded, the majority of the texts are indeed sectar-
ian (63), or of interest to the sect ( Jubilees [2] and Giants [1]). At
the same time, ten fragmentary texts are of undetermined nature
(see Table 7).
Table 7: Fragmentary Papyri of Undetermined Nature
Text No. Name Sectarian Qumran Scribal
School
4Q331 4QpapHistorical Text C no data no data
4Q391 4Qpap psEzek
e
no data no data
4Q465 4QpapText Mentioning Samson? no data no data
4Q478 4QpapFragment Mentioning Festivals no data no data
4Q484 4QpapTJud? (4QpapJub
k
?) no data
4Q485 4QpapProphetical/Sapiential Text no data
4Q486 4QpapSapiential Composition? no data
4Q487 4QpapSapiential Composition? no data
6Q10 6QpapProphecy no data
While many of the literary papyri are parallelled by leather copies
of the same composition (see Table 5), other papyri present the only
evidence of the composition contained in them (see Table 8).
4QcryptA Unidentied Texts P, Q (4Q313a, b)
4QcryptA Calendrical Document F (4Q313c)
4QcryptA Lunisolar Calendar (4Q317)
4QcryptA Calendrical Document B (4Q324c)
11QcryptA Unidentied Text (11Q23)
10
S.J. Pfann, 4Q298: The Maskls Address to All Sons of Dawn, JQR 85
(1994) 20335. See also Pfanns study, The Writings in Esoteric Script from
Qumran, The Dead Sea Scrolls: Fifty Years After Their Discovery, 17790.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 94
+nr conrts or +nr tvn.x r.rvni 95
Table 8: Compositions Known Only from Papyrus Fragments
4QpapAdmonitory Parable (4Q302)
4QpapHistorical Text C (4Q331)
4Qpap paraKings et al. (4Q382)
4QpapHodayot-likeText B (4Q433a)
4QpapText Mentioning Samson? (4Q465)
4QpapFragment Mentioning Festivals (4Q478)
4QpapTJud? (4QpapJub
k
?; 4Q484)
4QpapProphetical/Sapiential Text (4Q485)
4QpapSapiential Composition? (4Q486)
4QpapSapiential Composition? (4Q487)
4QpapApocryphon ar (4Q488)
4QpapApocalypse ar (4Q489)
4QpapWar Scroll-like Text A (4Q497)
4QpapSap/Hymn (4Q498)
4QpapHymns/Prayers (4Q499)
4QpapBened (4Q500)
4QpapRitMar (4Q502)
4QpapPrQuot (4Q503)
4QpapRitPur B (4Q512)
6Qpap apocrSam-Kgs (6Q9)
6QpapProphecy (6Q10)
6QpapBened (6Q16)
6QpapHymn (6Q18), possibly a copy of ShirShabb?
(thus J. Strugnell, personal communication, February 2000)
6QpapUnclassied frags ar (Words of Michael?) (6Q23)
The Qumran papyri consist of an almost negligible number of bib-
lical texts (two, three, four, ve, or six), mainly from Cave 6 (see
Table 9).
Table 9: Biblical Texts on Papyrus
4QpapIsa
p
(4Q76)
4QpapGen
o
or 4QpapJub
j
(4Q483)
6QpapDeut? (6Q3)
6QpapKgs (6Q4)
6QpapPs? (6Q5)
6QpapDan (6Q7)
While the evidence for the Cave 4 biblical papyri is very scanty and
does not necessarily indicate the existence of complete biblical scrolls
(note that 4QpapIsa
p
contains only a few words, and could there-
fore have represented a pesher like 4Qpap pIsa
c
), the group of Cave
6 biblical scrolls is slightly more signicant.
From the point of view of their content, it is dicult to cha-
racterise the corpus of the Qumran papyri which contains almost
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 95
96 rv.xtrr +o\
exclusively non-documentary texts (see Table 1). The non-documentary
papyrus texts represent several, if not most genres of texts repre-
sented in the Qumran corpus. The papyri listed in Table 5 reect
these genres, but in the main these texts are sectarian, as is further
underlined by the copies of non-sectarian texts written by the Qumran
scribal school listed in the Appendix and analysed beneath Table 6.
At the same time, the long list of texts frequently represented among
the leather manuscripts of Qumran listed in Table 6 shows that not
all genres of Qumran texts are represented among the papyri. Ex-
tremely rare within the corpus of Qumran papyri are eschatological
writings and Hebrew and Aramaic biblical papyri, of which only a
very small minority is found among the Qumran papyri (see Table 9).
We suggest that the collection of Qumran papyri is mainly sec-
tarian and liturgical, and usually nonbiblical. Most papyri may reect
personal copies owned by members of the Qumran community, while
some may have been imported from other sources.
As far as we can ascertain, the individual texts within the corpus
of the Qumran non-documentary papyri have nothing in common
when contrasted with the texts written on leather nor do they have
specic content features. Thus, for example, the content of 4Qpap-
MMT
e
(4Q398) does not display any features which set it aside from
the copies of MMT written on leather.
A further feature which may be recognised is that no papyri writ-
ten in the paleo-Hebrew script have been preserved at Qumran.
This is no coincidence, since this script is used mainly for Scripture
(Torah and Job), and as very few biblical papyri are found at Qumran,
paleo-Hebrew papyri are not expected.
The scribal practices reected in the Qumran papyri can be exam-
ined best in a few texts that have been preserved relatively well.
4Qpap pIsa
c
(4Q163)
4QpapTob
a
ar (4Q196)
4QpapS
c
(4Q257)
4QpapAdmonitory Parable (4Q302)
4Qpap paraKings et al. (4Q382)
4QpapMMT
e
(4Q398)
4QpapH
f
(4Q432), see E. Schuller, DJD 29, pls XIIIXIV and fold-
out pl. III
With some exceptions, the scribal conventions used in writing on
papyrus are similar to those used for writing on the leather texts
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 96
+nr conrts or +nr tvn.x r.rvni 97
from the Judean Desert, insofar as they relate to the spelling sys-
tems, the use of nal letters, word division, paragraphing, writing in
columns, use of margins, occasional writing on two sides, etc. Further-
more, paragraphos signs are evidenced in some papyri.
11
Scribal signs,
such as are known from the leather texts, are also found in a few
papyri, especially in 4Qpap pIsa
c
(4Q163) which is unique, as it dis-
plays several scribal signs in col. II that are not known from other
texts. Because of the fragmentary nature of the papyri, little is known
about the dimensions of the Qumran papyri or of their individual
columns or sheets.
The main dierences in scribal habits between texts written on
leather and those on papyrus result from the material used: the lack
of ruling on papyri, and therefore also the absence of guide dots on
them (horizontal bres must have provided some form of guidance),
the absence of cancellation dots and of crossing out letters or words
with a line (scribes of papyri preferred to use other systems of eras-
ing, mainly washing o letters or words, such as in 4QpapPrQuot
[4Q503] frag. 11, line 4).
The only case of a paleo-Hebrew divine name written on papyrus
is in 6QpapHymn (6Q18) frag. 6, line 5; 10, line 3 (la). Otherwise,
tetrapuncta are found in three papyri: 4Qpap paraKings et al. (4Q382)
frag. 9, line 5; 4QpapTob
a
ar (4Q196) 17 i 5; frag. 18, line 15;
4Qpap psEzek
e
(4Q391) frags 36, 55, 58, 65. But the distribution of
the writing of the divine names in papyri cannot be examined well,
as no instances where the tetragrammaton was written in square
characters have been preserved either, except for 4QpapAdmoni-
tory Parable (4Q302) 1 i 8 (hy).
The great majority of the papyrus texts found in the Judean Desert
outside Qumran are documentary, except for
Mas 1o (Mas pap paleoText of Sam. Origin [recto])
Mas 1o (Mas pap paleoUnidentied Text [verso])
Mas pap Literary Text? gr (Mas 739)
11
4Qpap ParaExod gr frag. 17 between lines 2 and 3. The shhook sign is evi-
denced in several papyri: 4QpapM
f
(4Q496) 10 iii 13; 4QpapSap/Hymn (4Q498)
frag. 15; 4QpapRitMar (4Q502) frags 19, line 5; 142; 318; 4QpapPrQuot (4Q503)
col. III:1, 6, 12, 18, 23; IV:6; VIII:2, 22; XI:1, 6; etc.; 4QpapPrFtes
c
(4Q509) frags
10 ii11, line 8 (according to Baillet); 49, 225, 265; 4QpapRitPur B (4Q512) frags 13;
15 ii (both col. IX); col. XII, 7; frags 4850, line 5.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 97
98 rv.xtrr +o\
Mas pap Virgil lat (Mas pap 721) [recto]
Mas papUnidentied Poetical Text lat [verso]
Mur 108112
12
The following papyrus texts are documentary:
All the texts found at Naal ever: from Cave 5/6 derive thirty-one
papyri in Greek, nine in Nabatean, sixteen in Aramaic, and seven in
Hebrew; one in Greek from Cave 8 as well as small fragments in
Hebrew. From Xev/Se derive fourteen and many unidentied frag-
ments in Greek, and from the same site also 3646 fragments in
Aramaic (mainly) and Hebrew (note also many unidentied fragments
from ev/Se?)
Naal e"elim: two in Greek and one in Aramaic
Wadi Murabba'at: sixty-plus in Greek, thirteen in Aramaic, and twenty-
three in Hebrew
Naal Mishmar: one in Greek and two in Hebrew
Jericho: seven-plus in Hebrew, six-plus in Aramaic, and ten-plus in
Greek
Sdeir: one in Aramaic
Wadi Gweir: one in Greek
Wadi Nar: three-plus in Greek and Hebrew/Aramaic
Masada: probably eleven texts in Greek and twenty in Latin
All the aforementioned numbers are approximate due to the frag-
mentary state of the material.
While it is hard to summarise all the details of this descriptive
study, the most important conclusions are:
While in almost all the other sites in the Judean Desert docu-
mentary papyri form the majority among the papyrus texts, in Qumran
almost all papyri are non-documentary (literary).
In Qumran the papyri form a minority of the texts found there
(14%), while in all other sites except for Sdeir they form a majority.
The number of biblical papyrus fragments in Hebrew is much
smaller than the proportion of Bible fragments among the Qumran
scrolls in general.
The case of a non-Qumranic origin for the two documentary
texts from Cave 6 cannot be made conclusively, but various argu-
12
One of these, Mur 108 (Mur papPhiliosophical Text gr) is a poetical text, pos-
sibly in iambic trimeters. This is possibly a tragedy, see J. Strugnell, The Antiquaries
Journal 43 (1963) 304; C. Austin, Comicorum graecorum fragmenta in papyris reperta (Berlin:
De Gruyter, 1973) no. 360; C.P. Thiede, The Dead Sea Scrolls and the Jewish Origins
of Christianity (Oxford: Lion, 2000) 7879 (Ezekiel the Tragedian).
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 98
+nr conrts or +nr tvn.x r.rvni 99
ments may be adduced for an origin beyond Qumran of the Cave
4 documentary texts.
The major material used for the literary documents left behind
by the Qumran community was leather, supplemented by a num-
ber of papyrus scrolls. Whether these two groups of texts derived
from dierent sources is unknown, but at least a case can be made
for a separate origin of the Greek biblical fragments from Cave 7
and the collection of papyri from Cave 6.
The majority of the papyri are sectarian or of interest to the
Qumran community ( Jubilees and Giants). These sectarian texts in-
clude several literary genres of the communitys writing: Rules, halakhot,
liturgical works, poetical compositions, and sapiential works. Only a
small number of papyri are non-sectarian (Aramaic texts, Hebrew
and Greek biblical texts). For the Hebrew papyri from Qumran these
data suggest a close connection between the writing on papyrus and
the Qumran community.
Not all genres of Qumran texts are represented among the
papyri. Extremely rare within the corpus of Qumran papyri are
eschatological writings and Hebrew and Aramaic biblical papyri, of
which only a very small minority is found among the Qumran papyri.
It is suggested that the collection of Qumran papyri is mainly
sectarian and liturgical, and usually nonbiblical. Most papyri may
reect personal copies owned by members of the Qumran commu-
nity, while some may have been imported from other sources.
As far as we can ascertain, the individual texts within the cor-
pus of the Qumran non-documentary papyri have nothing in com-
mon when contrasted with the texts written on leather nor do they
reect specic content features.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 99
100 rv.xtrr +o\
APPENDIX
PAPYRUS TEXTS FROM QUMRAN
The appendix lists all the papyrus texts found in the Judean Desert.
Due to the fragmentary state of the material, several inventory num-
bers, such as 1Q69, refer to an undetermined number of texts.
bold face (e.g., 4Q163) = sectarian text (including texts written
by the Qumran scribal school)
13
4Q223 = text probably written by the Qumran scribal school
14
a. Non-Documentary Papyrus Texts (Listed from North to South)
Text No. Name Reverse Side Sectarian Qumran
of Opistho- Work System
graph
1Q69 1QpapUnclassied frags no data no data
1Q70 1QpapUnclassied frags r + v no data no data
1Q70bis 1QpapUnclassied frags no data no data
4Q51a 4QpapUnclassied frags no data no data
4Q69 4QpapIsa
p
no data
4Q120 4QpapLXXLev
b
4Q127 4Qpap paraExod gr
4Q163 4Qpap pIsa
c
(cf. also 4Q515) y y
4Q196 4QpapTob
a
ar
4Q217 4QpapJub
b
? ? no data
4Q223224 4QpapJub
h
? y?
4Q249 4Qpap cryptA Midrash Sefer y no data
Moshe
4Q249a 4Qpap cryptA Serekh ha-'Edah
a
y no data
4Q249b 4Qpap cryptA Serekh ha-'Edah
b
y no data
4Q249c 4Qpap cryptA Serekh ha-'Edah
c
y no data
4Q249d 4Qpap cryptA Serekh ha-'Edah
d
y no data
4Q249e 4Qpap cryptA Serekh ha-'Edah
e
y no data
4Q249f 4Qpap cryptA Serekh ha-'Edah
f
y no data
4Q249g 4Qpap cryptA Serekh ha-'Edah
g
y no data
4Q249h 4Qpap cryptA Serekh ha-'Edah
h
y no data
4Q249i 4Qpap cryptA Serekh ha-'Edah
i
y no data
13
The determining of the sectarian character usually follows Dimant, Qumran
Manuscripts.
14
This list follows Tov, Further Evidence and idem, Scribal Practices and Approaches
Reected in the Texts found in the Judean Desert, in press.
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 100
+nr conrts or +nr tvn.x r.rvni 101
4Q249j 4Qpap cryptA Lev
h
? y no data
4Q249k 4Qpap cryptA Text Quoting y no data
Leviticus A
4Q249l 4Qpap cryptA Text Quoting y no data
Leviticus B
4Q249m 4Qpap cryptA Hodayot-like Text E y no data
4Q249n 4Qpap cryptA Liturgical Work E? y no data
4Q249o 4Qpap cryptA Liturgical Work F? y no data
4Q249p 4Qpap cryptA Prophecy? y no data
4Q249q 4Qpap cryptA Fragment y no data
Mentioning Planting
4Q249r 4Qpap cryptA Unidentied Text A y no data
4Q249s 4Qpap cryptA Unidentied Text B y no data
4Q249t 4Qpap cryptA Unidentied Text C y no data
4Q249u 4Qpap cryptA Unidentied Text D y no data
4Q249v 4Qpap cryptA Unidentied Text E y no data
4Q249w 4Qpap cryptA Unidentied Text F y no data
4Q249x 4Qpap cryptA Unidentied Text G y no data
4Q249y 4Qpap cryptA Unidentied Text H y no data
4Q249z 4Qpap cryptA Miscellaneous Texts A y no data
4Q250 4Qpap cryptA Text Concerning y no data
Cultic Service A
4Q250a 4Qpap cryptA Text Concerning r + v y no data
Cultic Service B?
4Q250b 4Qpap cryptA Text Related to Isa 11 r + v y no data
4Q250c 4Qpap cryptA Unidentied Unid. y no data
Text I = r Text J
4Q250d 4Qpap cryptA Unidentied Unid. y no data
Text J = v Text I
4Q250e 4Qpap cryptA Unidentied Unid. y no data
Text K = r Text L
4Q250f 4Qpap cryptA Unidentied Unid. y no data
Text L = v Text K
4Q250g 4Qpap cryptA Unidentied Text M r + v y no data
4Q250h 4Qpap cryptA Unidentied Text N r + v y no data
4Q250i 4Qpap cryptA Unidentied r + v y no data
Text O = r
4Q250j 4Qpap cryptA Miscellaneous Texts B r + v y no data
4Q255 4QpapS
a
= v 4Q433a y no data
4Q257 4QpapS
c
y y
4Q273 4QpapD
h
y y?
4Q302 4QpapAdmonitory Parable y no data
4Q324b 4QpapCalendrical Document A? y no data
4Q331 4QpapHistorical Text C no data no data
4Q382 4Qpap paraKings et al. no data y
4Q384 4Qpap apocrJer B? no data y
4Q391 4Qpap psEzek
e
no data no data
Table (cont.)
Text No. Name Reverse side Sectarian Qumran
of Opistho- Work System
graph
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 101
102 rv.xtrr +o\
4Q398 4QpapMMT
e
y y?
4Q432 4QpapH
f
y y?
4Q433a 4QpapHodayot-like Text B = r 4Q255 y no data
4Q465 4QpapText Mentioning Samson? no data no data
4Q468j 4QpapUnclassied frags no data no data
4Q478 4QpapFragment Mentioning Festivals no data no data
4Q482 4QpapJub
i
? ? no data
4Q483 4QpapGen or papJub
j
? no data
4Q484 4QpapTJud? (4QpapJub
k
?) ? no data
4Q485 4QpapProphetical/Sapiential Text no data no data
4Q486 4QpapSapiential Composition? no data no data
4Q487 4QpapSapiential Composition? no data no data
4Q488 4QpapApocryphon ar
4Q489 4QpapApocalypse? ar
4Q490 4QpapFragments ar
4Q496 4QpapM
f
= v 4Q509 y y
4Q497 4QpapWar Scroll-like Text A = v 4Q499 y no data
4Q498 4QpapSap/Hymn no data y?
4Q499 4QpapHymns/Prayers = r 4Q497 y y?
4Q500 4QpapBened y?
4Q502 4QpapRitMar y y
4Q503 4QpapPrQuot = r 4Q512 y y
4Q505 4QpapDibHam
b
= r 4Q506 y y?
4Q506 4QpapDibHam
c
= v 4Q505 y y
4Q509 4QpapPrFtes
c
= r 4Q496 y y
4Q512 4QpapRitPur B = v 4Q503 y y
4Q515 4QpapUnclassied frags no data no data
4Q516 4QpapUnclassied frags no data no data
4Q517 4QpapUnclassied frags no data no data
4Q518 4QpapUnclassied frags = r 4Q519 no data no data
4Q519 4QpapUnclassied frags = v 4Q518 no data no data
4Q520 4QpapUnclassied frags = v no data no data
4Q558 4QpapVision
b
ar
4Q559 4QpapBibChronology ar
6Q3 6QpapDeut? no data
6Q4 6QpapKgs no
6Q5 6QpapPs? no data
6Q7 6QpapDan no
6Q8 6QpapGiants ar
6Q9 6Qpap apocrSam-Kgs no data y?
6Q10 6QpapProphecy no data no
6Q16 6QpapBened y? no data
6Q17 6QpapCalendrical Document y? no data
6Q18 6QpapHymn y y
6Q22 6QpapUnclassied frags y? no data
6Q23 6QpapUnclassied frags ar (Words of Michael?)
6Q24 6QpapUnclassied frags no data no data
Table (cont.)
Text No. Name Reverse side Sectarian Qumran
of Opistho- Work System
graph
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 102
+nr conrts or +nr tvn.x r.rvni 103
6Q25 6QpapUnclassied frags no data no data
6Q27 6QpapCursive Unclassied frags no data no data
6Q28 6QpapCursive Unclassied frags no data no data
6Q30 6QpapCursive Unclassied frags no data no data
6Q31 6QpapUnclassied frags no data no data
6QX1 6QpapUnclassied frags no data no data
7Q1 7QpapLXXExod
7Q2 7QpapEpJer gr
7Q3 7QpapBiblical text? gr
7Q4 7QpapBiblical text? gr
7Q5 7QpapBiblical text? gr
7Q6 7QpapUnclassied frags gr
7Q7 7QpapUnclassied frags gr
7Q8 7QpapUnclassied frags gr (papEn gr?)
7Q9 7QpapUnclassied frags gr
7Q10 7QpapUnclassied frags gr
7Q11 7QpapUnclassied frags gr (papEn gr?)
7Q12 7QpapUnclassied frags gr (papEn gr?)
7Q13 7QpapUnclassied frags gr (papEn gr?)
7Q14 7QpapUnclassied frags gr (papEn gr?)
7Q15 7QpapUnclassied frags gr
7Q16 7QpapUnclassied frags gr
7Q17 7QpapUnclassied frags gr
7Q18 7QpapUnclassied frags gr
7Q19 7QpapImprint gr
9Q 9QpapUnclassied frag.
11Q28 11QpapUnidentied Text D no data no data
b. Documentary Papyrus Texts from Qumran (?)
Text No. Name
4Q347 4QpapDeed F ar (part of Xev/Se 32)
4Q352 4QpapAccount of Cereal B ar or heb
4Q352a 4QpapAccount A ar or heb
4Q353 4QpapAccount of Cereal or Liquid ar or heb
4Q358 4QpapAccount F? ar or heb
4Q359 4QpapDeed C? ar or heb
4Q360a 4QpapUnidentied Fragments B ar
4Q361 4QpapUnidentied Fragment gr
6Q26 6QpapAccount or Contract
6Q29 6QpapCursive Unclassied frag.
Table (cont.)
Text No. Name Reverse side Sectarian Qumran
of Opistho- Work System
graph
SHIFFMAN_f5_85-103 12/11/02 9:05 AM Page 103
104
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 104
This page intentionally left blank
THE ROMAN CENSUS IN THE PAPYRI FROM THE
JUDAEAN DESERT AND THE EGYPTIAN
kat okan pograf*
Hannah M. Cotton
I should like to start with a well-known passage from the New
Testament which has been the subject of endless discussions and
controversies in the attempt to salvage the tradition about Jesus year
of birth:
In those days a decree (dgma) was issued by Caesar Augustus for the
whole inhabited world to register (in a census) (pogrfesyai psan
tn okoumnhn). This census (pograf) took place for the rst time
(prth) when Quirinius was governor of Syria.
And everyone made his way to be registered, each to his own city
(Luke 2:13).
1
Ever since the discovery of papyri recording house-to-house censuses
at fourteen-year intervals in Egypt (the kat okan pograf), we can
be sure that a hard core of historical fact lies behind the passage
from Luke, even if we cannot reconcile the time of the census with
the traditional date of Jesus birth.
2
105
* This revised version of the lecture delivered in New York in March 2000 in
honour of Baruch Levine owes much to detailed criticism provided by Dominic
Rathbone, who kindly allowed me to see his PSI XI 1183: Record of a Roman
Census Declaration, in advance of publication in Essays and Texts in Honor of J.
David Thomas (eds T. Gagos and R.S. Bagnall; American Studies in Papyrology 42;
Oakville, CT: American Society of Papyrologists, 2001) 99115.
1
Egneto d n taw mraiw kenaiw, jlyen dgma par Kasarow Agostou,
pogrfesyai psan tn okoumnhn. ath pograf prth gneto gemoneontow tw
Suraw Kurhnou. ka poreonto pntew pogrfesyai, kastow ew tn auto plin.
2
See now B. Palme, Die gyptische kat okan pograf und Luke 2, 15,
Protokolle zur Bibel 2 (1993) 124; idem, Ein Nachtrag zum Artikel PzB 2 (1993)
124, Protokolle zur Bibel 3 (1994) 17; K. Rosen, Jesu Geburtsdatum, der Census
des Quirinius und eine jdische Steuererklrung aus dem Jahr 127 nC., JAC 38
(1995) 515; cf. H.-A. Rupprecht, Ein Mnchener Papyrus zum Provinzial-Zensus,
Bayern und die Antike: 150 Jahre Maximilians-Gymnasium in Mnchen (ed. W.-A. v.
Reitzenstein; Munich: C.H. Beck, 1999) 26271; M. Wolter, Erstmal unter Quirinus!
Zum Verstndnis von Lk 2,2, BN 102 (2000) 3541.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 105
106 n.xx.n v. co++ox
Although one can show that there was never a general census of
the entire Roman empire, i.e., one held simultaneously in all provinces,
nevertheless it is true that provincial censuses were an Augustan inno-
vation, as was the general imposition of the provincial poll tax, tri-
butum capitis, intimately connected with the census, albeit not the only
reason for it.
3
There is therefore good reason to believe that the
census which took place in year 6 cr under P. Sulpicius Quirinius
was indeed the rst provincial census to be conducted in the province
of Syriaand naturally the rst census was ordered by Augustus
and not by the governor.
4
Quirinius did no more than transmit the
imperial order in his edict; he may even have cited the emperors
ipsissima verba in his edict commanding the people to register.
5
It is
also entirely credible that the same edict, or perhaps rather a later
one, ordered the provincials to go back to their legal domicile to
be registered there, as did the often-quoted edict of C. Vibius Maxi-
mus, the prefect of Egypt in 104, where one notices that the word-
ing implies that the census had already started:
The house-to-house census having started, it is essential that all per-
sons who for any reason whatsoever are absent from their homes be
summoned to return to their own hearths, in order that they may per-
form the customary business of registration and apply themselves to
the cultivation which concerns them (P.Lond. III 904 = W.Chr. 202,
lines 1838).
6
We have independent epigraphic evidence for a census in Syria in
6 cr in the inscription of Quintus Aemilius Secundus, who as pre-
fect of an auxiliary cohort conducted a census in Apamea:
3
L. Neesen, Untersuchungen zu den direkten Staatsabgaben der rmischen Kaiserzeit (Bonn:
Habelt, 1980) 3945; cf. P.A. Brunt, The Revenues of Rome, Roman Imperial
Themes (Oxford: Clarendon Press, 1990) 32930; D. Rathbone, Egypt, Augustus,
and Roman Taxation, Cahiers du Centre G. Glotz 4 (1993) 8699.
4
Thus there is no reason to reject Lukes statement on the ground that a cen-
sus was ordered by the governor and not by the emperor.
5
Somewhat in this vein, see H.M. Cotton, H na parxea Araba: The New
Province of Arabia in the Papyri from the Judaean Desert, ZPE 116 (1997) 2068.
6
tw kat o[kan pografw ]nest[shw] nagkan [stin psin to]w kay [ntina]
dpote att[an podhmosin p tn] nomn prosa[gglle]syai pa[nel]yen ew t au[tn
]fstia, n[a] ka tn sunyh [o]konoman t[w po]grafw plhrsvsin ka t pros-
[hko]s atow gevrgai proskartersv[sin]. . . . There is often a reference to the
prefects edict in the declaration itself (kat t keleusynta or t prostetagmna vel
sim.): I register myself . . . as ordered by the prefect . . . etc.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 106
+nr nov.x crxsts ix +nr r.rvni 107
7
Q(uintus) Aemilius Q(uinti) f(ilius) Pal(atina) Secundus [in] castris divi Aug. s[ub] Publio)
Sulpi[c]io Quirinio le[gato] C[a]esaris Syriae honoribus decoratus, pr[a]efect. Cohort(is) Aug(ustae)
I, pr[a]efect. cohort. II classicae; idem iussu Quirini censum egi Apamenae civitatis millium homin.
civium CXVII. . . .
8
Cf. Ant. 17:355: t! d Arxhelou xra! potelo! pro!nemhye!h! t %rvn; cf.
Ant. 18:2: parn d ka Kurnio! e! tn Ioudaan pro!ykhn t! %ura! with H.M.
Cotton, Some Aspects of the Roman Administration of Judaea/Syria-Palaestina,
Lokale Autonomie und rmische Ordnungsmacht in den kaiserzeitlichen Provinzen vom 1. bis 3.
Jahrhundert (ed. W. Eck; Schriften des Historischen Kollegs. Kolloquien 42; Munich:
R. Oldenbourg Verlag, 1999) 7679.
9
F. Millar, State and Subject: The Impact of Monarchy, Caesar Augustus: Seven
Aspects (eds F. Millar and E. Segal; Oxford: Clarendon Press, 1984) 44 (= Rome, the
Greek World and the East I: The Roman Republic and the Augustan Revolution [eds H.M.
Cotton and G.M. Rogers; Chapel Hill, NC; London: University of North Carolina
Press, 2002] 299).
10
Kurnio! d . . . p %ura! parn, p Ka!aro! dikaiodth! to ynou! pe!talmno!
ka timht! tn o!in genh!meno!. Kvpni! te at !ugkatapmpetai tgmato! tn
ppvn, ge!meno! Ioudavn t p p!in jou!&. parn d ka Kurnio! e! tn Ioudaan
pro!ykhn t! %ura! genomnhn potimh!men! te atn t! o!a! ka podv!meno! t
Arxelou xrmata.
11
See Cotton, H na parxea Araba, 2068.
Quintus Aemilius Secundus, son of Quintus, of the tribe Palatina, I
received honours in the camps of the Divine Augustus under Publius
Sulpicius Quirinius, legate of Caesar in Syria, as prefect of the First
Cohort Augusta, as prefect of the Second Cohort Classica; I also con-
ducted by Quirinius command a census in the city of Apamea (count-
ing) 117,000 citizens . . . (ILS 2683).
7
It was precisely in the year 6 cr, when a census was conducted in
Syria by the governor, P. Sulpicius Quirinius, that Archelaus was
banished from Judaea, which became at that time part of the Roman
province of Syria under its own prefect.
8
Naturally the census spread
into the newly annexed territory.
9
Josephus explicitly combines the fact
of annexation with the administration of the census:
Quirinius . . . arrived in Syria, dispatched by Caesar to be governor
of the people and to be the assessor of their properties. Coponius, a
man of equestrian rank, was sent along with him to rule over the Jews
with full authority. Quirinius was also present in Judaea, which had
been made subject to Syria, in order to make an assessment of their
property and liquidate the estate of Archelaus (Ant. 18:12).
10
That a census took place in Judaea at the time of its provincializa-
tion was a mere coincidence and does not constitute proof for the
inauguration of Judaea as a new independent province. The notion
that a provincial census followed immediately upon the annexation
of a territory to the Roman empire is not well founded.
11
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 107
108 n.xx.n v. co++ox
Josephus gives little further information about the course and form
which the census took in Judaea, except to observe that the Jews
found it hard to tolerate the judicial (?) hearing, perhaps disputes,
which accompanied the registrations (tn p taw pografaw krasin,
Ant. 18:3), and that the more seditious claimed that the property
assessment (potmhsiw) was no better than downright slavery (Ant.
18:4).
12
No doubt the Jews found the tax itself oensive, but the pas-
sages from Antiquities suggest that they also resented the assessment
of their property and the examination of the verity of their state-
mentsperhaps also the need to take an oath.
13
How was the provincial census carried out? Was it universal
even if not simultaneousand uniform?
It seems reasonable to assume that the provincial census would
be patterned on the model of the Roman census in republican times
which, at least in theory, took place every ve years to determine
scal and military liability and, indirectly, political rights.
14
None of
these features was relevant under the Empire and the census of
Roman citizens as such lapsed altogether after the Flavian period,
if not already after Augustus.
15
However, already from the middle
of the last century of the republic the conduct of the census of
Roman citizens in Italy was decentralised. The local magistrates of
each city took the census and forwarded the census-returns to Rome.
The Tabula Heracleensis from the mid-rst century ncr
16
gives us
the ocial republican formula of the census:
12
tn d potmhsin odn llo ntikruw doulean pifrein lgontew.
13
On the oath in census declarations, see M. Hombert and Cl. Praux, Recherches
sur le recensement dans lgypte Romaine (P. Bruxelles Inv. E. 7616) (Papyrologica Lugduno-
Batava 5; Leiden: Brill, 1952) 123. On the taking of the oath by Jews in census
declarations, see H.M. Cotton, Fragments of a Declaration of Landed Property
from the Province of Arabia, ZPE 85 (1991) 26667. Josephus mentions a census
again in War 2:385 as means of obtaining statistics about the population: ! ne!tin
k t! kay k!thn kefaln e!for! tekmra!yai.
14
However, in the east the provincial censuses may have owed much to Hellenistic
practices.
15
But see now Rathbone, PSI XI 1183, 11113.
16
For the date, see M.H. Crawford, Roman Statutes (Bulletin of the Institute of
Classical Studies Supplement 64; London: Institute of Classical Studies, School of
Advanced Study, University of London, 1996) 1.36062; cf. also E. Lo Cascio, who
believes that the Tabula Heracleensis is of the time of Caesar and the decentralised
census is a novelty introduced by Caesar: Le professiones della Tabula Heracleensis
e le procedure del census in et cesariana, Athenaeum 78 (1990) 287317; idem,
Le procedure di recensus dalla tarda repubblica al tardoantico e il calcolo della
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 108
+nr nov.x crxsts ix +nr r.rvni 109
popolazione di Roma, La Rome imperiale: demographie et logistique (Rome: cole Franaise
de Rome, 1997) 376.
17
Trans. M.H. Crawford, Roman Statutes 1.377.
18
See Cl. Nicolet, Control of the Human Sphere: the Census, Space, Geography,
and Politics in the Early Roman Empire ( Jerome lectures; 19th ser.; Ann Arbor, MI:
University of Michigan Press, 1991) 12347.
19
See A. Aichinger, Zwei Arten des Provinzialcensus? berlegungen zu neupub-
lizierten israelischen Papyrusfunden, Chiron 22 (1992) 3845.
20
Brunt, The Revenues of Rome, 336.
And he (i.e., the magistrate) is to receive from them (the citizens) under
oath their nomina, their praenomina, their fathers or patrons, their tribes,
their cognomina, and how many years old each of them shall be and
an account of their property (rationem pecuniae), according to the form
of the census (ex formula census), which shall have been published at
Rome by whoever is then about to conduct the census of the people;
and he is to see that it all is entered in the public records of his
municipium (FIRA I
2
13, lines 14649).
17
Local censuses in Italy continued in imperial times, but the returns
were no longer forwarded to Rome. They were used to determine
scal and liturgical duties of their own citizens. It seems that the
Roman type of census extended to Roman colonies and Latin com-
munities in the provinces as well.
18
What about the rest of the empire, the non-citizens, the peregrine
subjects and communities? The documentation is patchy and incom-
plete. Apart from some references in the literary and legal sources
we are left with the sporadic evidence, unevenly distributed between
the dierent provinces, of career inscriptions of senators and eques-
trians involved in taking the census. It would seem that the provin-
cial governors in the imperial provinces administered the provincial
census in their provinces. However, whereas some of them mention
their having taken the census in their career inscriptions, others do
not.
19
To determine the frequency of the census in some provinces
and their absence in others on the basis of these career inscriptions
would be hazardous indeed. As P.A. Brunt put it: If we were to
trust arguments e silentio in relation to Roman taxation, we should
have to conclude that there were provinces in which Rome extracted
not a single penny from her subjects!
20
Recently E. Lo Cascio proposed that whereas the census in impe-
rial provinces was conducted by the governor and the individual was
directly responsible to Rome, in the public provinces, those governed
by proconsuls, on the other hand, the cities conducted the census,
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 109
110 n.xx.n v. co++ox
and it was they who were responsible to Rome for the payment of
taxes. To put it in a nutshell: communal assessment and liability in
the public provinces as against central assessment and individual lia-
bility in the imperial provinces.
21
However, evidence for the involvement of the cities in taking the
census is not restricted to the public provinces, nor is the concept
of communal responsibility. Furthermore, although we nd imperial
ocials, e.g., equestrians and some military personnel (like the pre-
fect Quintus Aemilius Secundus in Apamea), involved in taking the
census in the imperial provinces,
22
it is clear that the Roman provin-
cial government did not dispose of the necessary manpower to process
the declarations, make the lists and the evaluations as well as col-
lect the taxes. As far as the collection of taxes is concerned there is
no doubt that responsibility for collecting them devolved on the cities,
where these existed, or, as in Egypt, on other autonomous local bod-
ies and institutionsand later on it was done through the liturgical
system which operated under the supervision of state ocials.
23
We have to admit that we simply do not know enough about the
taking of the census and its processing in any province apart from
Egypt. The most recent treatment of the Egyptian census declara-
tions by R. Bagnall and B. Frier is based on just over 300 declara-
tions,
24
dating from 12 to 259 cr with fourteen-year intervals between
each census.
25
These declarations were submitted to ocials in the
villages and the metropoleis. The declaration includes personal details
(name, identifying features, age, status) of the declarant(s) and other
people living in the same household, including slaves. These house-
hold declarations are registrations of people but not of property.
Houses and other living quarters are mentioned solely for the pur-
21
E. Lo Cascio, Census provinciale, imposizione scale e amministrazioni citta-
dine nel Principato, Lokale Autonomie und rmische Ordnungsmacht, 197212.
22
For more examples of soldiers taking the census, see H. Zwicky, Zur Verwendung
des Militrs in der Verwaltung der rmischen Kaiserzeit (Winterthur: Buchdrckerei Winterthur,
1944) 7576, and Brunt, The Revenues of Rome, 33435.
23
See now M. Sharp, Shearing Sheep: Rome and the Collection of Taxes in
Egypt, 30 nc.r 200, Lokale Autonomie und rmische Ordnungsmacht, 21341.
24
The number has increased since to almost four hundred; see Rathbone, PSI
XI 1183, n. 37.
25
R. Bagnall and B. Frier, The Demography of Roman Egypt (Cambridge Studies in
Population, Economy, and Society in Past Time 23; Cambridge: Cambridge University
Press, 1994).
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 110
+nr nov.x crxsts ix +nr r.rvni 111
26
Rathbone, PSI XI 1183, 1067: Oikia here also carried the connotation of
civil domicile . . . for a man could only be liable in one place for dues on the per-
son, and a multiple property owner would only register himself as a resident of his
home property.
27
Rathbone, Roman Taxation, 88. Consequently Roman citizens, being exempt
from the poll tax, need not have led these declarations in Egypt, as now con-
vincingly argued by Rathbone in PSI XI 1183; contra Bagnall and Frier who
take the view that the function of the census in Egypt was not merely to facilitate
exaction of scal and liturgical dues which fell on a person, but rst and foremost
to control the population, and in the case of Egypt to maintain a rigidly xed
social structure, in which Romans, citizens of Greek cities, metropolitans, and other
Egyptians (not to mention freedmen and slaves) were kept clearly distinct and barred
by a complex of rules from many forms of interaction, Demography, 29.
28
See Rathbone, PSI XI 1183, 106 and n. 27.
29
A.M. Harmon, Egyptian Property Returns, YClS 4 (1934) 135230; H.J.
Wol, Das Recht der griechischen Papyri gyptens in der Zeit der Ptolemer und des Prinzipat.
Vol. 2: Organisation und Kontrolle des privaten Rechtsverkehrs (Handbuch der Altertums-
wissenschaft 10.5; Rechtsgeschichte des Altertums 5; Munich: Beck, 1978) 22255.
See recently J.L. Rowlandson, Landowners and Tenants in Roman Egypt (Oxford: Oxford
University Press, 1996) 12728; 14547; and the introductions to P.Hamb. IV 241
and 300 (B. Kramer and D. Hagedorn [eds], Griechische Papyri der Staats- und
Universittsbibliothek Hamburg: (P.Hamb. IV) [Archiv fr Papyrusforschung und ver-
wandte Gebiete Beiheft 4; Stuttgart: B.G. Teubner, 1998]).
30
Cf. H. Braunert, Cives romani und KAT OIKIAN APOGRAPHAI, Antidoron
pose of registering the people domiciled in them.
26
All this strength-
ens the impression of the existence of intimate connection between
these declarations and liability to the poll tax, with which they shared
the name of laographia,
27
even if women, men above 62 and children
under 14all of whom not liable to the poll taxwere also declared
in the census. The glaring absence in the Egyptian census returns
of the declaration of property, not to mention its evaluation, makes
it clear that taxation on property in Egypt was not done through
the census process. It must have been done in some other way.
28
The registration of land in Egypt in the biblioykh gktsevn was
intermittent and unsystematic; the language of the property returns
submitted, whether regular or general, hardly encourages a belief
in their eectiveness as a means for anything like assigning taxes
and liturgies.
29
Both omissions, i.e., of landed property and of its evaluation in
the Egyptian provincial census declarations, are striking in view of
the explicit provisions for the declaration of landed property and its
evaluation in the census form (the forma censualis) cited by Ulpian
in the third book On Censusesa form which must have been used
in the provincial census as well:
30
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 111
112 n.xx.n v. co++ox
The census form ( forma censualis) provides that lands should be regis-
tered as follows: the name of each farm, the civitas and pagus in which
it is situated and the names of its two closest neighbours. As regards
a eld, which shall have been sown within the last ten years, how
many iugera it measures; as regards a vineyard, how many iugera it
measures and how many vines it contains; as regards olives, how many
iugera there are and how many trees; as regards meadow land, which
shall have been cut in the last ten years, how many iugera; as regards
pasture land, how many iugera there seem to be, and the same with
woodland with trees suitable for felling. In all cases the person mak-
ing the return is himself to make the evaluation (omnia ipse qui defert
aestimet) (Dig. 50.15.4pr.).
No pografa of the Egyptian kind were found amongst the papyri
from the Judaean Desert. On the other hand, two land declarations,
P.Yadin 16 (completely preserved)
31
and P.Hever 62,
32
and the sub-
scription to a third one, P.Hever 61
33
all three submitted at the
census of 127 in the province of Arabiawere found in the Cave
of Letters in Naal ever. The two land declarations preserved in
P.Yadin 16 and in P.Hever 62 evince a striking correspondence to
the forma censualis of the Digest cited above. The declarants, Babatha
and Samouos respectively, after identifying themselves and their domi-
cile, declare their landed property giving the name of each plot, its
size, the amount of tax it pays and two of its abutters. The follow-
ing table brings out this striking correspondence:
34
Martino David oblatum miscellanea papyrologica (eds E. Boswinkel et al.; Papyrologica
Lugduno-Batava 17; Leiden: Brill, 1968) 11.
31
This is Babathas land declaration, see N. Lewis, The Documents from the Bar
Kokhba Period in the Cave of Letters: Greek Papyri, with Aramaic and Nabatean Signatures
and Subscriptions, edited by Y. Yadin and J.C. Greeneld ( JDS 2: Jerusalem: Israel
Exploration Society, 1989) 6570.
32
This is Samouos son of Shim'ons land declaration found with the documents
of his wife, Salome Komase daughter of Levi, see H.M. Cotton in H.M. Cotton
and A. Yardeni, Aramaic, Hebrew and Greek Documentary Texts from Naal ever and Other
Sites with an Appendix Containing Alleged Qumran Texts (The Seiyl Collection II) (DJD 27;
Oxford: Clarendon Press, 1997) 18194.
33
The declarant is probably Salome Komases brother, see Cotton in Cotton
and Yardeni, Aramaic, Hebrew and Greek Texts from Naal ever, 17480.
34
On the entire procedure see the excellent paper by B. Isaac: Tax Collection
in Roman Arabia, The Near East under Roman Rule, Selected Papers (Mnemosyne, bib-
liotheca classica Batava. Supplementum 177; Leiden: Brill, 1998) 32233 (originally
published in Mediterranean Historical Review 9 [1994] 25666).
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 112
+
n
r
n
o
v
.
x
c
r
x
s
t
s
i
x
+
n
r
r
.
r
v
n
i
1
1
3
Dig. 50.15.4pr.
Forma censuali cavetur ut agri sic in censum
referantur:
The census form provides that lands should
be registered as follows:
Nomen fundi cuiusque: et in qua civitate et in quo
pago sit: et quos duos vicinos proximos habeat. Et
arvum, quod in decem annos proximos satum erit,
quot iugerum sit; vinea quot vites habeat etc. . . .
omnia ipse qui defert aestimet
the name of each farm, the civitas and pagus
in which it is situated and the names of its
two closest neighbours. As regards a eld,
which shall have been sown within the last
ten years, how many iugera it measures; as
regards a vineyard, how many iugera it mea-
sures and how many vines it contains . . .
In all cases the person making the return
is himself to make the estimate.
P.Hever 62, frag. a
(%)am(m!o!uo! *%imvn[o]! Mavzhn! t! Zoa-
rhn! perimtrou Ptra!, !o!kn []n doi!
n at Mavz&, p!ogrfomai mautn tn
trikon!ta
I, Sammouos son of Simon, of Maoza in
the district of Zoar of the administrative
region of Petra, domiciled in my own pri-
vate property in the said Maoza, register
myself, thirty years old, [as owner of ?]
me)tox! t! prw Ivnayhn %imvno! mr#o!
mi! !tin !pr#o#u k&reiy! !tou #[n]##!
kbvn tri##n telon frou mlan n lept
te#!#!#arkont#a pnte, ge#t[on]e! Ma#n#a!
Mana ka yla!!a
half share of a eld, called Arenoaratha,
within the boundaries of the aforesaid
Maoza, in partnership with Ionathes son
of Shimeon, which half share is (the area)
of sowing one se"ah three qabs of barley,
paying as tax one black and forty-ve
lepta, abutters (being) Manaes son of Manaes
and the sea.
P.Yadin 16
Babya %mvno! Mavzhn t! Zoarhn!
perimtrou Ptra!, oko!a n doi! n
at Mavz&, pogrfom#a#i kkthmai
I, Babtha daughter of Simon, of Maoza
in the district of Zoar of the administrative
region of Petra, domiciled in my own pri-
vate property in the said Maoza, register
what I possess:
kpon foinikno! n roi! Mavzvn leg-
menon Algifiamma !prou kreiy! !tou
n! kbvn trin telonta foniko! !urou
ka megmato! !ta dekapnte pathto !ta
dka !tefaniko mlan n lept trikon
ta getone! d! ka yla!!a
a date orchard called Algiphimma, the area
of sowing one saton three kaboi of barley,
paying as tax, in dates, Syrian and mixed
fteen sata, splits ten sata, and for the
crown tax one black and thirty leptae,
abutters a road and the sea.
S
H
I
F
F
M
A
N
_
f
6
_
1
0
4
-
1
2
2
1
2
/
1
1
/
0
2
9
:
0
6
A
M
P
a
g
e
1
1
3
114 n.xx.n v. co++ox
Nothing like these two land declarations has ever been found in
Egypt prior to Diocletians reforms.
35
In 297 an imperial edict ordered
a general imperial census. In 298300 censitores make their appear-
ance in land declarations from Egypt which closely resemble those
from the rst half of the second century cr in Arabia. As Wilcken
observed long ago this is something completely new in Egypt.
36
Until
then, so far as we know, land declarations were not made in con-
nection with the census, nor addressed to the ocials in charge of
the census. In fact even after Diocletians reforms personal declara-
tions in Egypt were still separate.
37
The discrepancy in the wording of the two land declarations from
Arabia
38
gave rise to the hypothesis that the land declaration served
also as a registration of persons for the purpose of the poll tax. Like
the pografa from Egypt, the land declarations from Arabia begin
with the rst person pogrfomai. However, whereas Babatha merely
says pogrfom)a)i kkthmai: I register what I possess (P.Yadin
16:15), Samouos son of Simon introduces the registration of landed
property with his age at the time: p#ogrfomai mautn tn trikon#t#a:
I register myself thirty years old (P.Hever 62 frag. a, line 13). It
has been argued that women in Arabia, as in Egypt, were not sub-
ject to the tributum capitis, and thus Babatha, unlike Samouos, had
no need to register herself.
39
This does not convince: too many of
the features recurrent in the Egyptian pografa are missing here to
make Samouos a personal declaration similar to what we have in
Egypt. Furthermore, womens exemption from the poll tax in Egypt
did not dispense them from being registered properly together with
others who resided in the same household; the orphan, Yeshua' son
of Yeshua', Babathas rst husband, who was still a minor at the time
40
35
The old kat okan pograf system seems to have been discontinued after the
census of 257/8, cf. Bagnall and Frier, Demography, 911.
36
W.Chr. p. 226.
37
E.g., P.Cair.Isid. 8 (= ChLA 41, 1201) from 309 cr.
38
Pointed out by N. Lewis, A Jewish Landowner from the Province of Arabia,
SCI 89 (198588) 136.
39
Cf. Lo Cascio, Census provinciale, 201: there was no need to mention exemp-
tion or liability to pay the poll tax: the mere fact that someone was a thirty-year-
old male was sucient to make him automatically liable to it; and since the amount
was probably a xed annual cash levy, unlike the tributum soli, it did not have to
be mentioned in the declaration itself.
40
Cf. P.Yadin 27 of 19 August 132.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 114
+nr nov.x crxsts ix +nr r.rvni 115
41
H.M. Cotton, The Guardianship of Jesus son of Babatha: Roman and Local
Law in the Province of Arabia, JRS 83 (1993) 94113.
42
Ulpian, On Censuses, book 2: It is necessary to record ones age in the cen-
sus since their age dispenses some people from the tribute; for example in the
provinces of Syria men are liable to the poll tax (tributum capitis) from age 14 and
women from age 12, until they both reach 65. The relevant age is that which is
recorded at the time of taking the census, Dig. 50.15.3pr.
43
For the republican form see text above ad n. 17; personal details and prop-
erty are combined in Lucius Pompeius Nigers declaration, see Rathbone, PSI XI
1183, 11113.
44
For the Bostra papyri, see J. Gascou, Units administratives locales et fonc-
tionnaires romains. Les donnes des nouveaux papyrus du Moyen Euphrate et
dArabie, Lokale Autonomie und rmische Ordnungsmacht, 7173 and Cotton, Adminis-
tration, ibid., 9091; cf. H.M. Cotton, W. Cockle, and F. Millar, The Papyrology
of the Roman Near East: A Survey, JRS 85 (1995) nos 17172.
45
For the Petra papyri, see the various articles by A. Arjava, R. Daniel, J. Frsn,
T. Gagos, M. Kaimio, L. Koenen, M. Lehtinen, and M. Vesterinen in Atti del XXII
Congresso Internazionale di Papirologia (Firenze, 2329 agosto 1998) (Florence: Istituto
papirologico G. Vitelli, 2001); J. Frsn, A. Arjava, and M. Lehtinen (eds) with
contributions by others, The Petra Papyri I (American Center of Oriental Research
Publications 4; Amman: ACOR, 2002).
and seems to have resided with his mother,
41
should have been men-
tioned in her declaration, if it was intended also to serve as a dec-
laration of persons. Finally, we cannot be sure that women were
exempt from the poll tax in Arabia. They were liable to it in the
Syrian provinces in Ulpians time.
42
It must be stated categorically, then, that just as information con-
cerning property is lacking from ca. 400 declarations submitted in
the house-to-house censuses in Egypt between 18/19 and 257/8 cr,
so is all information concerning persons lacking from the two land
declarations taken at the census of 127 in Arabia. Since it is impos-
sible both that property was not declared and assessed in Egypt, and
that there was no registration of persons in Arabia, it would seem
that people and property were registered separately in the two
provincesin contrast to the old republican custom which combined
the two.
43
We should expect to nd separate land declarations from
Egypt and separate person declarations from Arabia. The fact that
no record of the latter has survived from Arabia is far less disturb-
ing than the total absence of property declarations from Egypt between
18/19 and 257/8 cr. With the exception of the two Bostra papyri,
44
there are no papyri from the province of Arabia between P.Yadin
27 of 19 August 132 and the Petra papyri of the sixth century cr.
45
The three papyri mentioned above (P.Yadin 16 and P.Hever 61 and
62) are the only written records connected with a census to have
survived from the province of Arabia. In the case of Arabia we can
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 115
116 n.xx.n v. co++ox
easily assume that people and property were declared separately dur-
ing the census, and that none of the declarations of persons has sur-
vived. On the other hand, the absence of details about property in
the pografa from Egypt combined with the fact that no property
census declarations have survived there strongly suggests, as stated
above, that the assessment of property for the purpose of taxation
and exacting liturgies in Egypt was totally divorced from the census
process; but we have no clear idea of the process which determined
the level of taxation and liturgies on properties.
46
Registration of
property existed in both provinces but its main purpose seems to
have been the establishment of ownership.
47
The striking similarity between the formula censualis of the provin-
cial census and the two land declarations from Arabia makes us
realise once again how swift the new province of Arabia was to
assimilate Roman forms twenty-one years after its annexation.
48
In
fact we witness in these two land declarations a fascinating interplay
between Romanization in the shape of the faithful adoption of the
formula censualis, and traditional localism expressed in the preserva-
tion of local standards and units of measurement. The Roman author-
ities must have imposed the former and tolerated the latter.
In one feature, however, the land declarations from Arabia diverge
sharply from the formula censualis. The jurists injunction that in all
cases the person making the return is himself to estimate its value
(omnia ipse qui defert aestimet) can hardly refer to an assessment by
the declarant himself or herself of the amount of tax to be paid. But
this is precisely what we nd in the land declarations from Arabia;
in addition to giving the measurements of the plots and their two
linear neighbours, the declarants from Arabia state how much each
plot pays as tax both in kind and in cash: e.g., telonta (i.e., kpon
foinikno!) foniko! !uro ka megmato! !ta dekapnte pathto !ta
dka !tefaniko mlan n lept trikonta (Babathas declaration in
P.Yadin 16:1921), and: telon [i.e. m#r)o)! mi!] frou mlan n
lept te)!)!)arkont)a pnte (Samouos declaration in P.Hever 62,
46
See text above ad nn. 2829.
47
For Egypt see above all A.M. Harmon, Egyptian Property-Returns, YClS 4
(1934) 135234. For Arabia one may cite P.Yadin 24 of 130 cr lines 46: pid
pegrcato Ioda! Eleaz)a#r)o[u Xyou!vno!] pogenmnou !ou nr p nmat! !ou
n )t )p[o]g&raf kpou! foinikno! n Mavz&. The pograp here has nothing to
do with the census, pace N. Lewis, In the World of P.Yadin, SCI 18 (1999) 126.
48
Cf. H.M. Cotton, Guardianship of Jesus, 94.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 116
+nr nov.x crxsts ix +nr r.rvni 117
49
See W. Weiser and H.M. Cotton, Gebt dem Kaiser, was des Kaisers ist: Die
Geldwhrungen der Griechen, Juden, Nabater und Rmer im syrisch-nabatischen
Raum unter besonderer Bercksichtigung des Kurses von Sela'/Melaina und Lepton
nach der Annexion des Knigreiches der Nabater durch Rom, ZPE 114 (1996)
23787.
50
Which is not aurum coronarium, which was not an annual tax and would not
be declared in a census. Note though that in P.Hever 62 cash payments are not
described as stephanikon, except in frags am, lines 1718.
51
Cf. Cotton, H na parxea Araba, 2068.
52
Admittedly this is not true of the lots described in P.Yadin 7, a deed of gift
in Jewish Aramaic of 120 cr, see Y. Yadin and J.C. Greeneld, A. Yardeni, and
B. Levine, The Documents from the Bar Kokhba Period in the Cave of Letters: Hebrew, Aramaic
and Nabataean-Aramaic Papyri ( JDS 3; Jerusalem: Israel Exploration Society, 2002)
73108.
53
)kpon foineiknvn kalome)non )G)annay A!adaia !n dato! to) ato kpou f
[]mern pt e! pt &mran tetrt mivr<>an man <> tel!ei kay to! e! lgon
kuriako f!kou {kay to!} foneiko)! pathto !ta dka ka !uro ka naarou !ta j,
! getvne! )natoln kpon kuriakn kalomenon Gann)ay Abb)aidaia du!mn klhron)moi
Areta! ntou d! bo&r&r) [kl]hrono&moi Iv!hpo! Baba.
54
See e.g., P.Cair.Isid. 2 and 3.
55
See PSI XI 1183 in Rathbone, PSI XI 1183, 1001 and ibid., 11113.
frag. a, lines 1617). The expression of the rate of taxation in terms
of the old Nabataean monetary unit, the melaina,
49
combined with
the fact that cash payments are described as stephanikon, crown money,
50
i.e., the old Nabataean royal tax, implies that the Romans adopted
without much ado the rates of taxation prevailing in the Nabataean
realm. This may lend force to the claim made elsewhere that this
was the rst census to be conducted in Arabia after its annexation
in 106.
51
The rate of taxation seems to have been part and parcel
of the description of a piece of land in Arabia, as implied by the
phrasing found in a deed of gift from 129 cr from the archive of
Salome Komase daughter of Levi:
52
a date orchard called the Garden of Asadaia with its [the] water
[allowance] (of that orchard), once a week on the fourth day, for one
half-hour which will pay every year to the account of the scus of our
Lord ten se"ah of splits, and six se"ah of the Syrian and the naaran
dates. The abutters on the east the orchard of our Lord [the emperor]
called the Garden of 'Abbaidaia, on the west the heirs of Aretas, on
the south a road and on the north the heirs of Yosef son of Baba
(P.Hever 64:2633).
53
Neither in the Diocletianic land declarations from Egypt
54
nor in the
census declaration of Roman citizens in Egypt under the Empire
55
do we nd the declarant giving the amount of tax on the piece of
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 117
118 n.xx.n v. co++ox
land declared in the census. The census in Arabia may have been
inuenced by patterns of describing property in the Nabataean king-
dom or by the status of the land and the nature of landholding
there.
56
Thus not only was the provincial census dierent from the
republican census and the census of Roman citizens under the Empire
in that persons and property were declared separately, as we have
seen before, but in addition it may be assumed that the nature of
the provincial census, determined as it was by conditions in each
province, may have varied from province to province. By 127 cr, when
Titus Aninius Sextius Florentinus conducted the census in Arabia,
over 150 years had elapsed since the rst provincial census in Gaul
in 27 ncr under Augustus. Nevertheless it is quite likely that there
was no norm, represented by the forma censualis of Ulpian: the lat-
ter, if not reecting some local variation, may have been an ide-
alised type.
57
What system operated in Judaea? Did the declarations of 6 cr
resemble the Egyptian pografa or the Arabian land declarations?
A new interpretation of a Greek papyrus from Naal e"elim, pub-
lished for the rst time by B. Lifshitz in 1961,
58
may throw some
light on the census process in Judaea.
59
The papyrus consists of six fragments surviving from a large doc-
ument. The extent of the loss cannot be established. Only one of
the two largest fragments, frag. a, is reproduced below.
The document contained at least four columns, consisting of two
sets of two columns, the left one of each pair being a list of per-
sons and the right one a list of their respective ages. This structure
is revealed in frag. a, which preserves three columns, although the
remains of the rst column consist of only the ends of two names
appearing in lines 7, 8 and 9. Cols i and ii form a pair followed by
col. iii which was paired with the following column, which has not
survived.
56
See H.M. Cotton, Land Tenure in the Documents from the Nabataean
Kingdom and the Roman Province of Arabia, ZPE 119 (1997) 25565, with no
positive conclusions.
57
The conclusions reached here owe a great deal to my discussions with Dominic
Rathbone.
58
B. Lifshitz, The Greek Documents from Naal e"elim and Naal Mishmar,
IEJ 11 (1961) 5362.
59
The papyrus is re-edited by the present author as 34e papCensus List from
Judaea or Arabia gr, Miscellaneous Texts from the Judaean Desert (DJD 38; Oxford:
Clarendon Press, 2000) 21725.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 118
+nr nov.x crxsts ix +nr r.rvni 119
60
See the introduction to Mur. 89107 in P. Benoit, J.T. Milik, and R. de Vaux,
Les Grottes de Murabba't (DJD 2; Oxford: Clarendon Press, 1961).
col. i col. ii col. iii col. i col. ii col. iii
1 [tn] ke [age] 25
2 [t]n ig .[. . .]..[ [ag]e 13 .[. . .]..[
3 [t]n iy Ih!ou! Lhou.[ [ag]e 19 Yeshua son of
Levi
4 [t]n ma [. . .]n%vr..[ [ag]e 41 [. . .]nor [ son]
5 [ ] Iv!hpo! [ [age] Yosepos
[another son]
6 [ ] Ih!ou! al[ [age] Yeshua
an[other son]
7 trace [ ] Iv!hpo! ..[ trace [age] Yosepos
[another son]
8 ]u tn ja Aneina! .[ son of age 61 Aneinas
[ o]s [another son]
9 ]no! tn jz Ell)hlo! son of age 67 Ellelos
)[#a)l[ [Shim]on (?) an[other son]
10 [ ] Gio! a[ [ ] Gaius
a[nother son]
11 [ ] %)e[.].o! %eim)a[ [ ] Se..os son of
Seima[
12 [tn] kb A[ [age] 22 A[
13 [tn] lw K[ [age] 36 K[
14 tn ]&b [age] [?]2
The ocial nature of the list is quite apparent, and is also implied
by the uent hand of the scribe. If we exclude the Bar Kokhba doc-
uments, this is one of the few ocial documents found in the Judaean
Desert, where most of the documents are private. The only other
group of ocial documents are the parchment fragments from Wadi
Murabba'at (Mur. 89107), which on the most plausible interpreta-
tion are lists of taxes in money and kind received by the adminis-
tration.
60
Like the present document and the land declarations from
Arabia they are written in Greek. The use of Greek as the ocial
language in a Roman province implies of course the active partici-
pation of local people in the routine of provincial administration.
The date and place of writing of the present document are unknown.
If, as is argued reasonably by the archaeologists, the documents found
in Cave 34 of Naal e"elim were hidden there by refugees of the
Bar Kokhba revolt, then the present document should be dated most
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 119
120 n.xx.n v. co++ox
probably to the rst half of the second century cr. There are at
least fourteen names in cols i and ii and a minimum of thirteen in
col. iii. Together with the twenty-two names appearing in frags bf,
the list must have contained the names and ages of at least forty-
nine persons. The preponderance of Jewish names inclines one to
think that the papyrus comes from Judaea rather than from the
province of Arabia.
There is no support for Lifshitzs speculation that this is a list of
soldiers, a fraternity of warriors, who constituted the army of Bar
Kokhba.
61
Not only is there no apparent reason to associate the list
with the Bar Kokhba revolt, but the idea of soldiers, as pointed out
already by Benoit,
62
seems to be excluded by the presence of peo-
ple aged thirteen years on the one hand and sixty-seven on the
other.
63
The true nature of the list is revealed in the format of col. iii
where a name and patronymic is followed by another name or names
indented by slightly over 1 cm. It seems to be a roster of house-
holds listing the name of the head of the household followed by
those of the other members (i.e., sons). Only males appear in what
is preserved of the document; it is therefore likely that the list was
restricted to the male members of the household. Similar lists, drafted
by local ocials, are known from Egypt. They were derived, or
rather abstracted, from the census declarations submitted every four-
teen years at the house-by-house registration in Egypt, the kat okan
pograf.
64
Some of these lists are clearly intended solely for the
purpose of taxation; summarising the material contained in the dec-
larations, they omit all details which are irrelevant to that purpose,
such as the names of persons who were exempted from the poll tax,
and thus contain only males between fourteen and sixty-two. In
Egypt, women were exempt from the poll tax that males between
the ages of fourteen and sixty-two had to pay.
61
Lifshitz, Greek Documents, 6061; his conclusion is based on an erroneous
interpretation of the term delf! here, and elsewhere; see B. Lifshitz, Papyrus
grecs du dsert de Juda, Aegyptus 42 (1962) 25254.
62
P. Benoit, Bulletin, RB 68 (1961) 467.
63
Cf. frag. a ii 9.
64
See Hombert and Praux, Recherches, 13547; Bagnall and Frier, Demography,
2627.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 120
+nr nov.x crxsts ix +nr r.rvni 121
65
Above text ad nn. 811.
66
Cited above in n. 42.
67
Like the penymerow in Egypt, on which see P.J. Sijpesteijn, Penthemeros-Certicates
in Graeco-Roman Egypt (Papyrologoa Lugduno-Batava 12; Leiden: Brill, 1964).
68
See Bagnall and Frier, Demography, 2728.
69
On capital villages see discussion and bibliography in Cotton, Administration,
8286.
70
For Egypt see F. Oertel, Die Liturgie: Studien zur ptolemaischen und kaiserlichen
Verwaltung Agyptens (Leipzig: B.G. Teubner, 1917; reprinted Aalen: Scientia-Verlag,
However, it is precisely the exclusion of women from the present
list that makes it unlikely to be a tax list. It is more likely than not
that women were liable for the poll tax in Judaea, and this is for
the following reason. The rst census in Judaea in 6 cr, as pointed
out before, was an extension of the census carried on in Syria at
the time into the annexed territory.
65
Consequently the same rules
of liability to the poll tax prevailing in the province of Syria at the
time must have been applied to the annexed territory. Now, unless
a change occurred between 6 cr and the rst quarter of the third
century cr, the exclusion of women from the poll tax cannot be rec-
onciled with what Ulpian tells us about the Syrian provinces in his
time, where males from the age of fourteen, females from the age
of twelve, and both till they reached the age of sixty-ve, were liable
for the poll tax.
66
I am therefore inclined to accept D. Rathbones
suggestion to me that this is a list of men liable to one or several
liturgies which were only imposed on men, probably therefore man-
ual liturgies.
67
Such lists had to be kept up-to-date,
68
and therefore, we can be
sure that, if deposited in the cave during the Bar Kokhba revolt,
they were made not long before the outbreak of the revolt. If I am
right in thinking that this is a list from Judaea from before the Bar
Kokhba revolt, we must conclude that lists based on census decla-
rations, like the pografa in Egypt, were prepared by local ocials
in Judaea using the Greek language. If the capital villages of the
toparchiae in the Jewish region in Judaea fullled functions similar
to those carried out by cities in other parts of the Roman Em-
pire, then we can assume that these lists were prepared by the civic
authorities in these villages.
69
But it is entirely possible that, as in
Egypt, the Romans employed in Judaea too the liturgical system
with local people working under state ocials.
70
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 121
122 n.xx.n v. co++ox
Unfortunately this is as far as one can go with the fragmentary
evidence. Some lessons can be learnt though.
Although we probably owe the presence of papyri in the caves of
the Judaean Desert to the two great revolts in 6673 and 132136,
these papyri should be studied dispassionately and in contextalways
bearing the Egyptian example in mindthat is without attempting
to read into them the unique history of the Jewish people. For the
romantic historian a fraternity of warriors ghting for Jewish free-
dom is probably a great deal more exciting than a list of people
liable to one or several liturgies. However, it should not come as a
great surprise that the refugees of the revolts hid in these caves those
documents attesting the routine of daily life in a Roman province,
demonstrating once again that Judaea was a normal province and
Jewish society part and parcel of the society of the Roman Near
East.
71
1965); J.D. Thomas, Compulsory Public Service in Roman Egypt, Das rmisch-
byzantinischen gypten: Akten des internationalen Symposions 26.30. September 1978 in Trier
(eds G. Grimm et al.; Aegyptiaca Treverensia 2; Mainz am Rhein: P. von Zabern,
1983) 3539; N. Lewis, The Compulsory Public Services of Roman Egypt (Payrologica
Florentina 28; Florence: Edizioni Gonnell, 1997
2
).
71
See H.M. Cotton, Introduction to the Greek Documentary Texts, Aramaic,
Hebrew and Greek Documentary Texts from Naal ever, 15357 with R. Katzo s review
of DJD 27 in SCI 19 (2000) 32327.
SHIFFMAN_f6_104-122 12/11/02 9:06 AM Page 122
THE LANGUAGE OF POWER: LATIN IN THE
INSCRIPTIONS OF IUDAEA/SYRIA PALAESTINA*
Werner Eck
The papyri from the Near East throw into relief a very prominent
characteristic of Roman domination: the absence of any deliberate
attempt on Romes part to impose Latin as the normal language of
communication with her subjects. This does not mean of course that
Rome used the local languages in her communications with the sub-
jects. On the contrary: the titulus on Jesus cross and the boastful
and arrogant inscription which the rst prefect of Egypt, Cornelius
Gallus, put up at Philae
1
both of which use in addition to Latin
and Greek also Hebrew and Egyptian respectivelyare obviously
the exception to the rule; normally, so it seems, the native languages
played no role at all for Rome.
2
In the Roman Near East this atti-
tude, i.e., the fact that Rome did not impose Latin nor used the
local languages, meant that Greek, the lingua franca of the Near East
since the Hellenistic period was the ocial language of communi-
cation between Rome and its subjects.
That this was so is demonstrated not only by the papyri from
Egypt, but also by those found in other parts of the Roman Near
East: in the Judaean Desert, near the Euphrates, and in Bostra.
Although many of these documents are addressed to representatives
of the Roman government: a beneciarius, a centurion, a praefectus
alae, the governor of Syria Coele or Arabia himselfthey are all
written in Greek.
3
The same is true of announcements made by
123
* I am grateful to Hannah Cotton, with whom I have shared the work on many
of the inscriptions discussed here, for the English translation. I have explicitly asked
her to maintain the lecture style of her original translation.
1
Matt. 27:37; Mark 15:26; Luke 23:38; John 19:19; CIL III 14147, 5 = ILS
8995.
2
Cf. A. Wacke, Gallisch, Punisch, Syrisch oder Griechisch statt Latein? ZRG
110 (1993) 1459.
3
D. Feissel and J. Gascou, Documents darchives romains indits du Moyen
Euphrate (III
e
s. aprs J.-C.), Journal des Savants (1995) 65119; N. Lewis, The
Documents from the Bar Kokhba Period in the Cave of Letters: Greek Papyri ( JDS 2; Jerusalem:
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 123
124 vrnxrn rck
Israel Exploration Society/Hebrew University of Jerusalem/Shrine of the Book,
1989); H.M. Cotton and A. Yardeni, Aramaic, Hebrew and Greek Documentary Texts from
Naal ever and Other Sites with an Appendix Containing Alleged Qumran Texts (The Seiyl
Collection II) (DJD 27; Oxford: Clarendon Press, 1997) 158279.
4
See the collection in R. Katzo, Sources of Law in Roman Egypt: The Role
of the Prefect, ANRW II.13 (Berlin: W. de Gruyter, 1980) 80744; G. Purpura,
Gli editti dei prefetti dEgitto, I sec. a. C.-I sec. d. C., Annali Sem. giur. Univ.
Palermo 42 (1992) 487671; further, SB XVIII 13849; BGU XVI 2558; IFAO III
34; P.Oxy. LI 3613 (I am grateful to Andrea Jrdens for her help).
5
P.Yadin 16:3638.
6
See H.M. Cotton, Subscriptions and Signatures in the Papyri from the Judaean
Desert: The XEIROXRH%TH%, JJP 25 (1996) 2940, where she discusses also the
Greek translation of the subscription to the declaration of Ignotus son of Levy in
P.Hever 61, published in Cotton and Yardeni, Aramaic, Hebrew and Greek Documentary
Texts.
7
See Feissel and Gascou, Documents darchives romains, no. 1.
8
SB XII 11043; ChLA XI 466; on which see W. Eck, Ein Prokuratorenpaar
von Syria Palaestina in P. Berol. 21652, ZPE 123 (1998) 24955.
9
PSI IX 1026 = CIL XVI p. 146, no. 13 = S. Daris, Documenti per la storia dell
esercito in Egitto (Milan: Societ editrice Vita e pensiero, 1964) 194200 = CPL 117
Romes re-presentatives to her subjectsthey are all in Greek. Not
one of the edicts published by the prefects of Egyptof which we
know more than sixty by nowuse Latin.
4
Altogether Latin is rarely used in communications with Romes
subjects. And often when they had been composed in Latin, by the
time they reach us they were translated into Greek. Thus the sub-
scription of the equestrian prefect Priscus at the bottom of P.Yadin
16, Babathas census declaration of 4 December 127, originally read
as follows: Priscus praefectus alae (equitum) accepi pridie nonas Decembres
Gallicano et Titiano consulibus.
5
In this form it was posted in the basil-
ica of Rabat Moab. But in Babathas copy the text is given in trans-
lation:
Pre!ko! ))parxo! ppvn dejmhn t pr mi! nvnn Dekembrvn pata!
Galli)k[an]#o [ka Titiano]&.
6
With the exception of brief subscriptions in Latin, e.g., by the prae-
fectus Mesopotamiae Iulius Priscus who substituted for the governor of
Syria Coele (dipvn tn patean) in 245
7
or by the imperial freed-
man procurator Aelius Amphigetes in the province of Syria Palaestina
in 152,
8
there are no Latin texts originating from the Roman provin-
cial administration in the Roman Near East. A libellus (petition)
addressed to the governor of Syria Palaestina, Velius Fidus, in Caesarea
and composed in Latin in 150,
9
or military papyri from Masada and
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 124
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 125
Dura-Europos written in Latin
10
are the exception which prove the
rule: the writers are soldiers of legio X Fretensis and the cohors XX
Palmyrenorum respectively whose language was Latin in any case.
All this should not mislead us into thinking that Rome was indierent
to her own language. These last-mentioned documents clearly dis-
play it. When it came to Rome as ruling power and her represen-
tatives, then Latin was called for.
11
Rome, that is, the emperor, the
imperial ocials in the provinces, and the Roman army made their
claims for being recognized as such in Latin. For the purpose of
showing the provincials what Rome was and wherein her power
residedfor that, Greek, the general language of communication in
the Near East, would not do, but Latin alone. The province of
Iudaea/Syria Palaestina demonstrates this attitude in its Latin inscrip-
tions in a typical manner. This is one side of the provincial world
for which the papyri show the other. Together they give us a better
picture of reality.
To begin with, I would like to give a preliminary survey, of what
we actually know now about Latin inscriptions from Iudaea/Syria
Palaestina. In addition to this survey which portrays the scene in
broad outlines, I shall concentrate on some signicant examples, in
which the use of Latin as a reection of Roman mentality and power
is at the same time natural and deliberate.
First, one may well ask how many Latin inscriptions do we know
from Iudaea/Syria Palaestina? I have tried, without complete suc-
cess, to nd out an approximate number of the inscriptions, leaving
aside stamps on tiles or on other ceramic products, as well as tituli
picti like the ones from Masada, published by H.M. Cotton and
J. Geiger; for these tituli were written on foreign imports and were
written outside Iudaea.
12
Only in the case of Caesarea can we be quite
= J. Rea, Two Legates and a Procurator of Syria Palaestina, ZPE 26 (1977)
21722.
10
C.B. Welles, R.O. Fink, and J. Frank Gilliam, The Excavations at Dura-Europos,
Final Report V, Part I: The Parchments and Papyri (New Haven, CT: Yale University
Press, 1959); H.M. Cotton and J. Geiger, Masada II: The Latin and Greek Documents
( Jerusalem: Israel Exploration Society/Hebrew University of Jerusalem, 1989) 3179.
11
See W. Eck, Latein als Sprache politischer Kommunikation in den Stdten
der stlichen Provinzen, Chiron 30 (2000) 64160; idem, Ein Spiegel der Macht:
Lateinische Inschriften rmischer Zeit in Iudaea/Syria Palaestina, ZDPV 117 (2001)
4770. The present discussion depends in many ways on this paper.
12
See Cotton and Geiger, Masada II, nos 795853.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 125
126 vrnxrn rck
sure about the number. There are sixty-nine Latin inscriptions in
the corpus of Greek and Latin inscriptions from Caesarea, edited by
C.H. Lehmann and K.G. Holum in a corpus in 2000
13
that includes
all the texts found until 1992; fteen of them are pitiful fragments.
In addition, there are about twenty milestones from Caesarea and
its municipal territory. Since 1992 many new Latin inscriptions have
come to light in Caesarea, especially during the excavations directed
by J. Patrich and J. Porath; some inscriptions reached us from other
sources.
14
All in all there should be at least about 120 Latin texts.
Of course many of them are very fragmentary. Together with those
published by Lehmann and Holum we possess now more or less 190
Latin inscriptions from Caesareanot an overwhelming number for
a big city like Caesarea. Nevertheless compared with the three inscrip-
tions recorded in CIL III which covers the material from Caesarea
up to 1902, it has to be admitted that immense progress has been
made since the beginning of the century. Caesarea can boast now
of the greatest concentration of Latin inscriptions in the whole of
Iudaea/Syria Palaestina.
We do not have comparable collections for other sites in the
province. Consequently I can only give here a very supercial and
incomplete picture. From the whole area of Israel I have been able
to collect, up to the present, some 530 Latin texts, including 150
milestones, on which at least some letters can be read. Excluding
the inscriptions from Caesarea and the milestones, we are left with
about 210 other Latin inscriptions from the rest of the country
not a very impressive number. But I am condent that many more
texts exist, which, although published, have not yet come to our
notice, as well as many others not yet published. One of the more
13
C.H. Lehmann and K.G. Holum, The Greek and Latin Inscriptions of Caesarea
Maritima ( Joint Expedition to Caesarea Maritima: Excavation reports 5; Boston:
American Schools of Oriental Research, 2000).
14
W. Eck, An Inscription from Naal aggit Honoring a Roman Ocial,
Final ReportExcavations in Naal aggit (forthcoming): this text too could have come
from Caesarea. We have access to fteen Latin fragments now in private collec-
tions, of which two have just been published: see H.M. Cotton and W. Eck, Governors
and Their Personnel on Latin Inscriptions from Caesarea Maritima (PIASH 7.7; Jerusalem:
Israel Academy of Sciences and Humanities, 2001) 21540 and iidem, A New
Inscription from Caesarea Maritima and the Local Elite of Caesarea Maritima,
What Athens Has to Do with Jerusalem: Essays on Classical, Jewish and Early Christian Art
and Archaeology in Honor of Gideon Foerster (ed. L.V. Rutgers; Interdisciplinary Studies
in Ancient Culture and Religion 1; Leuven: Peeters, 2002).
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 126
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 127
crucial tasks of the Corpus Inscriptionum Iudaeae/Palaestinae which is now
underway, is to nd and record this unpublished material.
15
These texts, besides the milestones and inscriptions from Caesarea,
were found at about thirty dierent sites, from Acco-Ptolemais in
the north to Yotvataor Costiain the south.
16
The second high-
est number after Caesarea comes from Jerusalem, about thirty Latin
inscriptions.
17
That Caesarea and Jerusalem take the lead in the num-
ber of Latin inscriptions found in them is to be expected: after all,
both cities were refounded as Roman colonies. But there is a marked
dierence between the two: rst, we have now by far more texts
from Caesarea than from Jerusalema fact partly accounted for by
the dierent conditions and possibilities for excavations in each city.
But if we look at the type of inscriptions from either site, the dierent
character of each of these two cities, as far as it is reected in its
Latin inscriptions, becomes eminently clear. In Caesarea we nd
some twenty texts honouring provincial governors, many of them on
columns used as statue-bases, as well as some fourteen similar inscrip-
tions for the nancial procurators; even one imperial freedman was
honoured there, probably in Flavian times.
18
There are tombstones
for soldiers of both legions stationed in the province and for mem-
bers of their families and dedications to divinities. And of course
there are inscriptions that were originally placed under statues for
emperors, as well as building inscriptionsalmost all from the aque-
duct erected by Hadrian, but also one monumental Latin text on
an architrave of a building, nanced by a mother of a city coun-
cillor of Caesarea, obviously because she was her sons heir; in his
testament he had bequeathed this building to his hometown and the
mother was carrying out the terms of his will.
19
15
H.M. Cotton, L. di Segni, W. Eck, and B. Isaac, Corpus Inscriptionum
Iudaeae/Palaestinae, ZPE 127 (1999) 3078; cf. SCI 18 (1999) 17576.
16
AE 1986, 699; cf. W. Eck, Alam Costia constituerunt. Zum Verstndnis einer
Militrinschrift aus dem sdlichen Negev, Klio 74 (1992) 395400.
17
Cf. P. Thomsen, Die lateinischen und griechischen Inschriften der Stadt Jeru-
salem und ihrer nchsten Umgebung, ZDPV 44 (1921) 13868, long out of date.
18
No need to mention all the texts here since many of them are so far unpub-
lished; those already published can be found in Lehmann and Holum, The Greek
and Latin Inscriptions; see also Cotton and Eck, Governors. The rest will be pub-
lished by W. Eck, New Inscriptions from Caesarea in Judaea/Syria Palaestina,
Final Report (ed. J. Patrich; forthcoming) and by H.M. Cotton and W. Eck in Final
Report (ed. J. Porath, forthcoming).
19
Lehmann and Holum, The Greek and Latin Inscriptions, no. 44; cf. W. Eck, Zu
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 127
128 vrnxrn rck
In Jerusalem, on the other hand, there are only four inscriptions
that include the name of an emperor, three building inscriptions and
one which was once attached to an honorary statue of Antoninus
Pius.
20
One of these, displayed in the courtyard of the Church of
Flagellation, must have been intended for a city gate or for a small
arch, dedicated to Hadrian; unfortunately the date of this building
cannot be ascertained, because the essential part of Hadrians titu-
lature was not preserved.
21
Another fragment of an imperial inscrip-
tion was found during the excavations on the Temple Mount in
196768. M. Avi-Yonah, its rst publisher, oered no suggestion as
to what kind of monument or building the inscription originally be-
longed. A new study of the fragment shows that it belonged to an
arch, erected and paid for by the colony and dedicated to Septimius
Severus and his whole family: Caracalla, Geta, Iulia Domna, and
probably Fulvia Plautilla, his daughter-in-law.
22
It would be extremely
interesting to know its original location. The name of the governor
of the province, mentioned in this inscription, was not preserved. In
contrast with Caesarea, with the exception of a round statue-base
for a legate of the legio X Fretensis, M. Iunius Maximus, now stand-
ing in the old city near the Jaa gate, no text honouring a gover-
nor has so far been found in Jerusalem.
23
Likewise not one member
of the city council, the decuriones, of Aelia Capitolina is attested,
whereas in Caesarea members of this social group are mentioned in
at least eight inscriptions.
24
Most of the other texts from Jerusalem
are either dedications to pagan deities or funerary inscriptions for
lateinischen Inschriften aus Caesarea in Iudaea/Syria Palaestina, ZPE 113 (1996)
12943, esp. 13839; cf. Cotton and Eck, Local Elite of Caesarea Maritima.
Lehmann and Holum, The Greek and Latin Inscriptions, no. 27 too is a building inscrip-
tion. Cf. W. Ecks review of Lehmann and Holum, Topoi (2002) (forthcoming).
20
CIL III 116 = 6639; 6640; 12080; Thomsen, Inschriften der Stadt Jerusalem,
no. 3.
21
The interpretation of the remains of what seems to be an arch for Hadrian
is problematic: cf. inter alia, E.M. Smallwood, The Jews under Roman Rule: From
Pompey to Diocletian (SJLA 20; Leiden: Brill, 1974) 46162; C. Arnould, Les arcs
romaines de Jrusalem: Architecture, dcor et urbanisme (NTOA 35; Gttingen: Vandenhoeck
& Ruprecht, 1997) 24279.
22
H.M. Cotton and W. Eck, Ein Ehrenbogen fr Septimius Severus und seine
Familie in Jerusalem, Donum amicitiae: Studies in Ancient History: Published on Occasion
of the 75th Anniversary of Foundation of the Department of Ancient History of the Jagiellonian
University (ed. E. D[browa; Cracow: Jagiellonian University Press, 1997) 1120.
23
CIL III 6641 = 12080a.
24
See Cotton and Eck, Local Elite of Caesarea Maritima.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 128
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 129
soldiers and civilians. The social and political rank of the subjects
and dedicants of most of the Latin inscriptions from Jerusalem is
considerably lower than that of the average rank in Caesarea.
Nevertheless, all Latin inscriptions both from Caesarea and from
Jerusalem are without exception directly or indirectly associated with
people representing Rome and its imperial power. I will come back
to this topic later.
Considerably fewer Latin inscriptions come from all other sites:
about eight inscriptions come from Neapolis-Samaria, the same holds
true for Emmaus-Nicopolis; the harvest from Scythopolis is meagre
indeed, so far as I can tell: only four Latin texts. From other places,
we have at most only one or two Latin inscriptions. Thus, for exam-
ple, from Legio, where the legio VI Ferrata was stationed for at least
one century, we have two Latin inscriptions, an altar dedicated by
a primipilus to Serapis and a new one, also an altar, dedicated to
Silvanus sanctus.
25
This is hardly a reection of the original situation,
but rather due to the fact that until now no excavations have been
carried out there. The situation is reminiscent of that of Caesarea
before the intensive excavations conducted there in the last two
decades, when the great bulk of the texts we now possess was found.
Excavations in the legionary fort in Legio would in all likelihood
bring more Latin texts to light. The same is probably true of Ascalon
26
and other sites.
The projected corpus of all inscriptions from Hellenistic and Roman
Israel (above, n. 13) will span more than a thousand years of the
history of the country. Latin inscriptions, however, are restricted to
a much shorter period. The rst Latin text we know of may be once
more the titulus on the cross, mentioned at the beginning, or the
inscription of Pontius Pilatus, attesting the renovation of the Tiberi-
eum, now identied as a lighthouse in the harbour of Caesarea in
G. Alfldys compelling new reading of the inscription.
27
From about
the same time could originate the inscription on a sarcophagus for
25
AE 1948, 145; W. Eck and Y. Tepper, A Dedication to Silvanus near the
Camp of the legio VI Ferrata near Lajiun, SCI 20 (2001) 8588.
26
A new inscription from this site was published by W. Eck and B. Zissu, A
Nauclerus de oeco poreuticorum in a New Inscription from Ashkelon/Ascalon, SCI 20
(2001) 8996.
27
G. Alfldy, Pontius Pilatus und das Tiberieum von Caesarea Maritima, SCI
18 (1999) 85108; see now L. Feldmans popularizing version, Financing the
Colosseum, BAR 27/4 ( July/August 2001) 2031, 6061.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 129
130 vrnxrn rck
Iulia Grata, the wife of a Ti. Iulius Mellon, libertus Augusti and procu-
rator of the imperial property at Iamnia.
28
If we go by the chrono-
logical distribution of the material from Caesarea, it would seem that
the great bulk of Latin inscriptions is to be dated to the period
between Hadrian and the late third century.
The rst tetrarchy under Diocletian is also represented in the
inscriptions from Caesarea. But after Diocletian there was a clear
break. Inscriptions in Latin are no longer to be found in Caesarea.
So far as I can tell not a single Latin inscription from Caesarea is
to be dated to the time of Constantine or later. This is surprising
and in need of an explanation which cannot be oered here.
For sites other than Caesarea, I am aware of only four Latin texts
that can be dated after Diocletian with sucient precision.
29
The
rst is a milestone put up under Constantine between 324 and 326,
and found near Tel Tumis on the road from Legio to Scythopolis.
The inscription is remarkable for another reason as well: it was
painted on plaster, rather than incised.
30
The second inscription,
engraved on the tombstone of a Flavius Lucianus, comes from Umm
el-Jemal (Bath Shelomo). It reads militans inter scutarios d.n. Constant[i ].
31
This has to be Constantius II, which dates the inscription between
the years 337361. After this, we have a text from Gaza mention-
ing Iuvenalis, a bishop of Jerusalem in the sixth century.
32
Finally,
the grati from the house of St Peter in Capernaum may belong
either in the fourth or the fth century. We know, although the text
is not preserved, that the pilgrim from Piacenza, travelling in the
sixth century, left a prayer in the room, where according to tradi-
tion Jesus had transformed water into wine during the wedding at
Canaa.
33
28
AE 1948, 141.
29
An earlier date is more likely for the grato found in the Church of the Holy
Sepulchre and dated by its editor, M. Broshi, between 324 and 337; see M. Broshi,
Excavations in the Holy Sepulchre in the Chapel of St. Vartan and the Armenian
Martyrs, Ancient Churches Revealed (ed. Y. Tsafrir; Jerusalem: Israel Exploration Society,
1993) 11823, esp. 12122; but cf. now S. Gibson and J.E. Taylor, Beneath the Church
of the Holy Sepulchre Jerusalem: The Archaeology and Early History of Traditional Golgotha
(Palestine Exploration Fund monograph. Series maior 1; London: Palestine Exploration
Fund, 1994) 2545 for a discussion of all the pertinent problems.
30
B. Isaac and I. Roll, Roman Roads in Judaea, vol. 1: The Legio-Scythopolis Roads
(BAR International Series 141; Oxford: British Archaeological Reports, 1982) 55, 82.
31
AE 1959, 196 = 1960, 197.
32
CIL III 14155,1.
33
W. Eck, Grati an Pilgerorten im sptrmischen Reich, Akten des XII.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 130
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 131
Even this brief synopsis reveals that Latin inscriptions are com-
monly associated with the centres of the Roman military or admin-
istrative activities in the province. Nevertheless, I believe that examining
in some detail a few of the old and new epigraphical texts will make
the use of Latin in this context clearer.
I would like to begin with a fragmentary text found almost a hun-
dred years ago near Abu Gosh, situated a few kilometres west of
Jerusalem. The text was published by H. Vincent in 1907 and repro-
duced in 1996 in the important survey by M. Fischer, B. Isaac, and
I. Roll.
34
The Latin text is written on a tabula ansata; parts of four
lines are preserved, but not more than three to six letters were pre-
served in each line:
IMP CAE[---]
IMP [---]
SEX p L%V[---]
COH[---]
For some reason nobody had attempted to make sense of what was
left of the text. The fact that the inscription is written in a tabula
ansata makes it almost certain that we are dealing here with a build-
ing inscription. This assumption is strengthened by the ratio between
height and width: its width far exceeds its height. This fact taken
together with the presence of elements of the name of an emperor,
or rather of two emperors, in the rst two lines makes it clear that
we are dealing here with an imperial building inscription. Next, one
should expect the person who was responsible for the project and
the person or persons who executed it. In line four we can read
coh(ors). This was the military unit whose job it was to construct the
building. In line three we could expect to nd the commander of
that unit. In similar texts from other provinces, it is here that we
nd the name of the governor of the province under whose com-
mand all the military units stood. Comparing what remained of the
name in line three, Sex. Lu[---], with the list of known governors of
Iudaea, we hit upon Sex. Lucilius Bassus, familiar to us all from
Internationalen Kongresses fur Christliche Archologie: Bonn, 22.28. September, 1991 (eds E.
Dassmann and J. Engemann; JAC Ergnzungsband 20; Mnster: Aschendorsche
Verlagsbuchhandlung, 1995) 1.20622.
34
H. Vincent, RB 4 (1907) 41718; M. Fischer, B. Isaac, and I. Roll, Roman
Roads in Judaea, vol. 2: The Jaa-Jerusalem Roads (BAR International Series 628;
Oxford: British Archaeological Reports, 1996) 117 with plates 94, 95 on p. 429.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 131
132 vrnxrn rck
Josephus Jewish War as the second senatorial governor of Iudaea
under Vespasian, who governed the province from 71 to 73, and
who conquered Herodium in 72 AD. Restoring his full name and
title in line three makes this line reach the same length as the two
preceding lines where the names of Vespasian and Titus should be
restored in line one and two respectively:
Imp(erator) Caes(ar) [Vespasianus Augustus]
Imp(erator) T(itus) [Caesar Vesp(asiani) Aug(usti) l(ius)]
Sex(to) Lu[cilio Basso leg(ato) Aug(usti) pr(o) pr(aetore)]
coh(ors) [---].
The reconstructed text would have been about 1.5 m wide; not a
monumental inscription, albeit still an impressive one, which adver-
tised the emperors and what they were doing in the province. Just
what they are doing here is not instantly clear, but there are not
very many options: either a military installation, such as a small fort,
or a relay station (mutatio) for the ocial postal and messenger ser-
vice (cursus publicus) would have perfectly t Abu Goshs topograph-
ical position.
35
The inscription displays neatly the political structure
in Judaea at the time and those responsible for maintaining it: the
two emperors who had gained power over this province a few years
earlier, the governor who represented the emperors in the province,
and nally the cohors which together with other army units secured
peace and tranquillity in the province. The inscription displays per-
fectly and succinctly the Roman order. Little did it matter to those
responsible for erecting it that most of the provincials could not read
and understand the text. It was not intended for them; it was self-
regarding: it was intended for the Romans themselves and for this
reason the inscription was written in Latin.
The same is true of another Latin inscription found in the vicin-
ity of Jericho, now in the museum of the Armenian Patriarchy, from
the time of Marcus Aurelius and Lucius Verus (161169 AD):
36
[Imp(eratoribus) Marco] Anton(ino)
[et Lucio V ]ero Aug(ustis)
[? leg(io) X Fret(ensis) ] fecit
[sub Iulio Co]mmodo co(n)s(ulare).
35
W. Eck, Sextus Lucilius Bassus, der Eroberer von Herodium, in einer Bauinschrift
von Abu Gosh, SCI 18 (1999) 10920.
36
H.M. Cotton and W. Eck, Eine Bauinschrift unter Marc Aurel und Lucius
Verus aus Jericho, ZPE 127 (1999) 21115.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 132
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 133
Here too the army unit whose task it was to erect some structure
for the Roman administration or army (?) in the province is recorded
in a building inscription; this time it is a unit of the legio X Fretensis.
At the same time the fact that two Augusti share the power at the
time is also recorded, and the supreme commander, the governor
of the province, is mentioned in the last line. Again the use of Latin
is self-referential.
From Abu Gosh and Jericho we move on to Caesarea, the cen-
tre of Roman power in the province. As mentioned before, the major-
ity of all known Latin inscriptions from Iudaea/Syria Palaestina comes
from this Roman colony. Many of them concern the provincial
administration, the governors, the procurators of the scal depart-
ment, and the Roman army in the province. This was only to be
expected. Less expected, however, is the fact that members of
Caesareas aristocracy not only used Latin in dedications to Roman
ocials, but also in order to record their private acts in public. One
is struck by the strong Latin character presented by this colony after
its founding, specically in the second and third centuries, which can
only be accounted for if we allow for the settling there of a large
dose of native speakers at the time when the colony was founded.
This aspect of the life of the city of Caesarea has been dealt with
elsewhere.
37
Excavations in Caesarea in the last few decades have revealed sev-
eral ocial building complexes by the sea-shore. For what purpose
were they built and by whom? To a large extent the answers to
these questions rests on the interpretation of the epigraphic mate-
rial discovered during those excavations, some of which took place
as late as 1996 and 1997, but most of this material still remains
unpublished.
Let us rst consider the so-called governors palace on the promon-
tory. In 1990, during the excavations by the Institute of Archaeology
of the Hebrew University and the University of Pennsylvania, two
marble columns were found on the eastern side of the Promontory-
Palace. The columns lay within the buildings debris on a hypocaust-
oor. Three inscriptions, honouring various people, are visible on
each column.
38
The rst text on one of the two columns honours a
37
See Cotton and Eck, Local Elite of Caesarea Maritima.
38
B. Burrell, Two Inscribed Columns from Caesarea Maritima, ZPE 99 (1993)
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 133
134 vrnxrn rck
T. Flavius Maximus, a philosopher; the dedicator is a certain Varius
Seleucus, obviously a local person. This text is written in Greek and
can tell us nothing about the function of the palace. It is followed
by two more texts, this time in Latin, of an entirely dierent nature:
the rst a dedication to the emperor Probus by the senatorial gov-
ernor Passenianus, and the second is a dedication to Galerius by the
equestrian governor (designated here praeses) Audius Priscus.
39
That
the governors of the province put up a column with a statue of the
reigning emperor with an appropriate Latin inscription accompany-
ing it inside their palace can hardly come as a surprise.
The same is true of the three texts on the other column. In the
rst inscription, a member of the local ruling class, a Sex. Cornelius
Taurinus, posthumously honoured the senatorial governor D. Seius
Seneca: his son put up the dedication after his fathers death. The
inscriptions rst editor, B. Burrell, could not give a precise date for
the governorship of D. Seius Seneca. In the meantime, a military
diploma published by P. Wei, when read together with previously
known texts, xes his term of oce to 157/58.
40
More than one
hundred years later, the column was reused for a dedication to the
emperor Probus by the governor Acilius Cleobulus; and about fteen
or twenty years later, the column was once more recycled by
Audius Priscus to honour Constantius Caesar between 293 and 305.
Needless to say, all these inscriptions are written in Latin.
The governors palace would be a likely place for the erection of
two columns with the statues they must have borne. Not having
been found in situ, however, this likely assumption has remained
unproven until later discoveries could substantiate it. In the years
1996 and 1997 two mosaic inscriptions were found east of the pre-
sumed governors palace, but in a context clearly related to this com-
plex. A third text, also discovered in situ, is written on a cylindrical
stand which served as a statue base, found lying in front of a mensa
in a room built on the southern end of the circular wall of the cir-
28795 = Lehmann and Holum, The Greek and Latin Inscriptions, nos 1315 and
1618.
39
In fact there was a fourth text on the column, but it was erased before another
text was inscribed; nothing of this text is preserved; see Eck, review of Lehmann
and Holum.
40
P. Wei, Neue Militrdiplome, ZPE 117 (1997) 22768, esp. 25659.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 134
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 135
cus, but still part of the same complex.
41
One of the mosaic inscrip-
tions, discovered already in 1996, which due to its intriguing char-
acter, was somewhat too hastily made public and published for
the rst time in the daily papers in England and America, reads as
follows:
Spes bona| adiutorib(us) | oci | custodiar(um)
Good hope to the Assistants of the Oce in charge of Prisoners (or
the Prison).
The journalists text was quite accurate; less so their interpretation
of it: this can hardly be a reference to St Pauls prison. The mosaic
oor could hardly have been constructed before the second century
AD, and more likely in the next one.
The second mosaic inscription, discovered in the summer of 1997,
records a special group of Roman soldiers, called frumentarii, whose
intimate relationship with the provincial governor is well known from
other provincial contexts. This inscription was found not far away from
the rst one, in a room dedicated to the genius of the frumentarii:
42
Sanct[o] Genio fru[m]entarioru[m] omnia felicia
Whatever the precise function of the frumentarii was, a hotly disputed
subject,
43
it is clearly the case that the room where the mosaic was
found served for social gatherings; it was a schola, a club-house, of
the frumentarii within the praetorium complex.
The third inscription found, as was pointed out above, on a statue-
base in a room attached to the circular wall of the circus, is a ded-
ication by a Claudius Severus, warden of the centurions club-room
who has paid (for it) with his own money:
44
Cl(audius) Severus
cust(os) sc(olae) (centurionum)
s(ua) p(ecunia) f(ecit)
41
For a discussion of the texts of all these inscriptions see Cotton and Eck,
Governors, 21540.
42
Cotton and Eck, Governors, 23234.
43
It is more or less agreed that they were engaged in the requisitioning of grain
for troops or, more often after the rst century, in arresting people and delivering
messages.
44
Cotton and Eck, Governors, 23435.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 135
136 vrnxrn rck
This text suggests that another club-house for yet another group of
soldiers, a schola of centuriones, who stood under the direct command
of the provincial governors, was located in the presumed governors
palace.
This group of three texts is crucial for determining the function
of the promontory-palace, precisely because they were all found in
situ. All three groups of soldiers represented in these three inscrip-
tions with their oces or social clubs had to be at hand for the gov-
ernor, and consequently, their oces or scholae could not be located
far away from his headquarters. There can be no doubt entertained
anymore about the purpose of the promontory palace.
45
It is here
that the governors headquarters were located. From here he gov-
erned the rest of the province; perhaps this was also his place of
residence, but this the inscriptions cannot tell us.
All the inscriptions, without a single exception, that were found
there are in Latin, both dedicatory texts under statues intended for
emperors and governors, as well as those found in the scholae of the
soldiers attached to the governors headquarters, the ociales. Of
course the ociales were recruited from the army units stationed in
the province; and the language of the army was Latin. Nevertheless,
these rooms served also for social purposes, and probably the sol-
diers themselves paid for the mosaics; we may also recall that the
statue in the schola of the centurions was an act of euergetism on
the part of Claudius Severus. The use of Latin should not be taken
for granted, since most of the soldiers in the second and third cen-
tury were recruited from the local population, and were not unlikely
to have occasionally used their mother tongue, which was Greek or
Aramaic. Nevertheless, and this must be stressed, all the inscriptions
of this period here are in Latin. This is quite dierent to what took
place later on in the other praetorium located on the opposite side of
the Herodian circus, to which we now shall come.
Even before Patrichs recent excavations in eld CC, a few Latin
inscriptions were found whose evidence, when taken together, strongly
suggested that the nancial procurator had his seat in Caesarea.
46
45
See K.L. Gleason, The Promontory Palace at Caesarea Maritima: Preliminary
Evidence for Herods Praetorium, JRA 11 (1998) 2352, who knew about the rst
of the mosaic inscriptions mentioned above.
46
That the procurator did not have to reside at the same place as the governor
is demonstrated by the situation in Arabia, where the governor had his seat in
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 136
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 137
In 1961, during A. Negevs excavations just south of the crusader
city, an inscribed column intended for the statue of L. Valerius
Valerianus, procurator Syriae Palaestinae in the rst quarter of the third
century, was found, not in situ, but dragged to a lime kiln located
along the northern extension of the portico of the so called tax-
revenue oce,
47
which was later identied as part of the palace of the
late antique governor.
48
Another column bearing a dedication in
Latin to the procurator Furius Timesitheus (now published by Lehmann
and Holum) was found nearby,
49
and a third Latin inscription on a
pillar, also intended for a procurator and so far published only on
the Internet,
50
was likewise found in the vicinity of the other two.
Finally, not far away, once more inside the late antique palace a
statue-base, bearing a dedication to an imperial freedman, was found.
51
It should be pointed out that two of these dedications were erected
by subaltern personnel of the procurators. These texts, all of which
were written in Latin, when taken together, identied Caesarea as
the seat of the procurator of the province. Having been found in a
very restricted area and in close proximity to each other, these inscrip-
tions alone could have determined the precise location of the procu-
ratorial praetorium but no one had made the connection until recently.
The nal and decisive proof that what came to be the late antique
palace of the governor had previously been the seat of the nancial
procurator, was supplied by Patrichs excavations, during which more
than thirty, mostly fragmentary, inscriptions were found. The mes-
sage of these new texts, when taken together, is crystal clear.
52
With
one exception, they are all written in Latin. I shall start with the
exception:
53
Bostra, whereas the nancial procurator had his central oce in Gerasa. See
R. Haensch, Capita provinciarum: Statthaltersitze und Provinzialverwaltung in der rmischen Kaiser-
zeit (Klner Forschungen 7; Mainz: P. von Zabern, 1997) 55663.
47
Lehmann and Holum, The Greek and Latin Inscriptions, no. 5.
48
K. Holum, Caesarea Palaestinae: Inscriptions from the Imperial Revenue
Oce, The Roman and Byzantine Near East: Some Recent Archaeological Research (ed. J.H.
Humphrey; JRASup 14; Ann Arbor, MI: Journal of Roman Archaeology, 1995)
33345.
49
Lehmann and Holum, The Greek and Latin Inscriptions, no. 8.
50
Lehmann: http://www.usd.edu/clehmann/cmvpcol.html.
51
Lehmann and Holum, The Greek and Latin Inscriptions, no. 2.
52
See Eck, New Inscriptions from Caesarea.
53
Eck, New Inscriptions from Caesarea, no. 1.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 137
138 vrnxrn rck
[-] [---]
Antpatro[w]
ptropow
to Sebasto
[---] Antipatros, procurator of the Emperor
Although the beginning is missing, it is obvious that this is a dedi-
cation, probably to a deity. In this context, the use of the Greek
language, even when the dedicant is a Roman procurator, is not
altogether surprising. Such religious dedications from the private
sphere are often in evidence inside a building complex belonging to
the administration. For example, in Sarmizetegusa in the Roman
province of Dacia, a sanctuary that was part of the procuratorial
praetorium, displays no less than thirty-two dedications of dierent
procurators oered to a variety of deities.
54
Most of them present
the same simple formula as here: the name of the deity in the dative,
followed by the name of the procurator and his title, andmore to
the pointmany of them are in Greek.
All the other texts in the procuratorial praetorium in Caesarea, about
thirty, are in Latin. One of them is a dedication by the procurator
Valerius Valerianus, mentioned above.
55
The text is fragmentary,
56
but possibly the inscription was part of a dedication to a deity like
that of Antipater. Other texts are inscriptions put under statues or
portraits of procurators, like the one dedicated to a Calpurnius
Quintianus,
57
where the name of the dedicator is, unfortunately, lost:
[. . ?Calp]urnio Quin-
[?tian]o proc(uratori ) Aug (usti )
[prov]inc(iae) S[ y]r (iae) Pal (aestinae)
[---/---].
In another very fragmentary text only the remains of a cognomen
and the beginning of an equestrian cursus honorum are preserved, but
the entire context makes it probable that this equestrian ocer is
54
I. Piso, Inschriften von Prokuratoren aus Sarmizetegusa (I), ZPE 50 (1983)
23351; idem, Inschriften von Prokuratoren aus Sarmizetegusa (II), ZPE 120
(1998) 25371. Cf. Haensch, Capita, 73132.
55
Above, n. 47.
56
Eck, New Inscriptions from Caesarea, no. 3.
57
Eck, New Inscriptions from Caesarea, no. 2.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 138
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 139
being honoured as a procurator of Syria Palaestina, and that this
function stood at the end of the text, now lost:
58
[---]a Ti(beri) f(ilio) [---]
[---].ano eq(uo) pu[b( lico) ---]
[--- trib(uno)] mil(itum) leg(ionis) V [Mac(edonicae)]
[---].
A very similar text is preserved on a blue marble tablet. The hon-
orand, a man of equestrian rank, is likely to have been honoured
while serving as procurator of Syria Palaestina:
59
[---]nino
[equ(o) publ(ico) iud(ici) selecto ] ex V
[dec(uriis), ?trib(uno) mil(itum) leg(ionis) ---]ic(ae), praef(ecto)
[alae? . . . proc(uratori) ---u]rbis
[---, proc(uratori) ---]uriae/yriae/briae
[---]
[ procurat(ori) prov(inciae) Syriae Palaestinae].
Other small fragments belong to bases on which statues of emper-
ors were set up, at least in one case by one of the procurators.
60
This is yet another feature common to these administrative com-
plexes; it is in evidence in the governors praetorium, the promontory
palace discussed before, as well as in Sarmizetegusa.
Finally, I would like to mention two mosaics, one more or less
preserved, the other one very fragmentary.
61
Whereas the text of the
rst one, feliciter, is very simple, the second possibly demonstrates that
an optio, belonging to the sta (ocium) of a Roman ocial, proba-
bly the procurator, was responsible for laying down the mosaic. Both
mosaics belong to the pre-Byzantine period, and they are written in
Latin, like the two mosaics found in the vicinity of the governors
palace. This is quite signicant. The use of Latin in structures con-
nected to administrative oces or complexes was either natural or
considered necessary, even in mosaics, during the second and third
centuries. This contrasts sharply with later usage. Many mosaic
inscriptions from Late Antiquity were found in this building com-
plexall of them are written in Greek, regardless of the rank of the
58
Eck, New Inscriptions from Caesarea, no. 5.
59
Eck, New Inscriptions from Caesarea, no. 4.
60
Eck, New Inscriptions from Caesarea, nos 78.
61
Eck, New Inscriptions from Caesarea, nos 3839.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 139
140 vrnxrn rck
person responsible for its construction; one of them was the pro-
vincial governor himself, proconsul Palaestinae, a certain Andreas, the
others are nameless numerarii, ocials of the late antique governor.
62
What gave rise to this drastic change in using Greek instead of Latin
is not yet clear. This throws into deep relief, however, the consis-
tent use of Latin until the beginning of the fourth century.
We must envision the impact that a concentration in a small area
of many texts, all of them written in Latin, had on those who entered
the praetoriumeven the mosaics greeted them in Latin. The whole
area abounded in monuments speaking the language of the ruling
power, and the people mentioned in them were the representatives
of that power. Although all these people could speak Greek, the
lingua franca in the eastern provinces, they used Latin when repre-
senting Rome even in their epigraphical documents. Part of their
own identity and their own share in Roman power was thus expressed,
which otherwise would have been lost.
In daily life, in the ordinary exchange of letters between the Roman
emperors and the cities in the eastern provinces, Greek was used
freely. The best proof for this is the existence of the ab epistulis Graecis,
an imperial secretary, responsible for the Greek correspondence and
for pronouncements in this language.
63
From the time of Marcus
Aurelius onwards, this secretary became permanent, but even ear-
lier the emperors sometimes used experts for Greek in their chancery.
64
However, the imperial correspondence reveals that when it came
to grave matters both emperors and senate very often used Latin.
65
The same is true of inscriptions in a political and administrative
context.
66
Another example from Iudaea demonstrates the last point very
clearly. Near Tel Shalem, about 12 km to the south of Scythopolis,
62
Lehmann and Holum, The Greek and Latin Inscriptions, nos 8596.
63
Cf. F. Millar, The Emperor in the Roman World (31 BCAD 337) (Ithaca, NY:
Cornell University Press, 1992
2
) 83109.
64
A.R. Birley, Locus virtutibus patefactus?: Zum Befrderungssystem in der Hohen Kaiserzeit
(Opladen: Westdeutscher Verlag, 1992); W. Eck, P. Aelius Apollonides, ab epis-
tulis Graecis, und ein Brief des Cornelius Fronto, ZPE 91 (1992) 23642.
65
See the text of a senatus consultum transmitted in Latin in the time of Marcus
Aurelius to Milet: P. Herrmann, Eine Kaiserurkunde aus der Zeit Marc Aurels
aus Milet, IstMitt 25 (1975) 14966; idem, Fragment einer Senatsrede Marc Aurels
aus Milet (Nachtrag zu Ist. Mitt. 25, 1975, 149 .), IstMitt 38 (1988) 313.
66
Cf., for example, boundary inscriptions near Ephesus in Latin: Inschriften von
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 140
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 141
marble fragments of an inscription, recently published by G. Foerster
and myself, were found.
67
This is an extraordinary monument which
also demonstrates the importance of Latin in the political context.
Although more than twenty fragments of the inscription were pre-
served, they constitute not more than one-fth or one-fourth of the
original text. Nonetheless, due to the formulaic nature of such texts,
the oered reconstruction of the inscription is certain. The imperial
titulature, which constitutes almost the whole text, is more or less
xed, especially in ocial documents, and this inscription belonged
to an ocial monument. The rst proof for this is the use of Latin
in a region in which Greek was normally used in the second cen-
tury, at least in the public sphere. To judge from the epigraphic
material from Scythopolis so far published, almost all the inscrip-
tions found in this city are written in Greek; only ve Latin inscrip-
tions are known and not one of them was set up by the city itself.
68
This makes it very unlikely that Scythopolis had erected the monu-
ment to which the inscription from Tel Shalem once belonged. The
second proof for non-local origin is the monumentality of the inscrip-
tion. The reconstruction shows that the inscription was originally
between ten and eleven metres wide.
69
No other inscription in ancient
Iudaea can compete with its dimensions. But even more signicant
is the letter size. In the rst line they are 41 cm high, in the sec-
ond line 24 cm and in the third line 1819 cm. The size of the let-
ters is startling and exceptional, not only in this province
70
but in
general. Even in Rome itself most building inscriptions fail to match
this size. Only the inscriptions of the Pantheon, the Temple of Castor
and Pollux, and Tituss Arch have bigger letters.
Ephesus II 459 = G. Alfldy, Epigraphische Notizen aus Kleinasien I. Ein benecium
des Augustus in Ephesos, ZPE 87 (1991) 15762.
67
W. Eck and G. Foerster, Ein Triumphbogen fr Hadrian im Tal von Beth
Shean bei Tel Shalem, JRA 12 (1999) 294313. All details mentioned in the fol-
lowing text are documented in this article.
68
AE 1924, 131; 1939, 158; 1990, 1013; 1993, 1617; 1964, 198 = 1993, 1618;
a Roman governor was responsible for the last text.
69
See Eck and Foerster, Triumphbogen.
70
Until the discovery of this inscription the largest letters attested in a Latin
inscription from Iudaea belonged to an inscription from Caesarea; they are ca. 30
cm high, but only a pitiful fragment has remained of this text, see C.B. Brusa,
Iscrizioni, Scavi di Caesarea Maritima (ed. A. Frova; Rome: LErma di Bretschneider,
1966) no. 3; the text is missing in Lehmann and Holum, The Greek and Latin Inscriptions;
cf. Eck, Review of Lehmann-Holum.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 141
142 vrnxrn rck
To what type of monument did the inscription belong? There can
be no doubt that the inscription was part of an arch erected for
Hadrian. The reconstructed titulature of the emperor safely and pre-
cisely dates the arch: by the time it was erected, Hadrian had already
adopted his second acclamation as imperator. We are, therefore, after
the nal victory in the Bar Kokhba revolt.
The communis opinio so far has been that Hadrian accepted this
acclamation in the second half of 135. But there is no proof of that.
None of the inscriptions that allegedly attest this title in 135 can
bear the burden of proving this. On the other hand, there are some
ocial inscriptions from 136 without imperator iterum. Given these
facts, one can say no more than that there is no proof that Hadrian
took imp. II before 136.
Be this as it may, the arch near Tel Shalem was only erected
after Hadrian was imp. II. But what was the reason for erecting such
a monumental arch just then and there, and who was responsible
for its erection?
We should not forget that an arch with a triumphal statue on top
of it was the biggest public honorary monument one could get.
Unfortunately, the end of the third line of the inscription, where the
dedicator was mentioned, is not preserved. But a reconstruction in
correct scale demonstrates that only a few letters are missing after
the emperors titles, about four to six letters. As already pointed out,
Scythopolis had nothing to do with the arch; the city lay too far
away, and, most importantly, its ocial language was Greek. An
inscription from the northern entrance gate in Scythopolis, although
as yet undeciphered, is certainly written in Greek,
71
like the one on
an arch honouring Hadrian in Gerasa.
72
The use of Latin suggests
a Roman dedicator. Theoretically the governor could be responsi-
ble for this; but there is not enough space for the name of a gov-
ernor and his title. Another possibility could be one of the legions
stationed in the province. But we would probably need more space
at the end for the name of a legion; furthermore, we have only one
71
A complete and certain decipherment of the text from Scythopolis is proba-
bly impossible because only the dowel holes of the bronze letters are preserved and
because each letter, so it seems, was attached by one nail only to the stone, which
hampers the identifying of individual letters.
72
C.B. Welles, The Inscriptions, Gerasa: City of the Decapolis (ed. C.H. Kraeling;
New Haven, CT, American Schools of Oriental Research, 1938) 4012 no. 58.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 142
r.+ix ix +nr ixscnir+ioxs or itr.r.svni. r.r.rs+ix. 143
example of a legion honouring an emperor by the erection of an
arch, namely the III Cyrenaica during the Parthian war near Dura-
Europos. That rules out the restoration leg(io) X Fretensis or leg(io) VI
Ferrata at the end of line three.
I believe that there is only one possibility: SPQR = senatus Populusque
Romanusthe Senate and the Roman people. That the SPQR hon-
oured emperors by erecting an arch or another large monument in
the provinces after memorable achievements, especially after an impor-
tant victory, is attested several times. Augustus and Tiberius were
honoured in Pannonia with two arches after the victory over the
rebellious Pannonian tribes,
73
and Augustus alone received the tropaeum
Alpium, la Turbie, for his successful battles beyond the Alps in Raetia
with a long inscription in Latin.
74
For Claudius, the senate voted an
arch on the sea shore in Gesoriacum where he embarked for the
British island in 43 AD.
75
Germanicus, Tiberius adopted son as well
as destined successor, was commemorated by arches: in Rome, on
the bank of the Rhine near Mainz, and on Mt Amanus in Syria for
his res gestae in Germany and in the eastern provinces.
76
The com-
mon abbreviation SPQR ts admirably the remaining spaces in line
three. I believe we can be quite condent about the restoration of
SPQR as the dedicator of the arch for Hadrian near Tel Shalem.
What does the erection of this arch mean? Why would the Roman
Senate and people vote to erect a monument on such a scale at
such a long distance away from Rome, in one of the eastern provinces,
in a place removed from any civilian settlement, and in Latin, not
in Greek, the normal language of most cities in the province? The
chronology supplies the key. The arch was erected after the Bar
Kokhba revolt for the victory in this war, a war that had greatly
unsettled Rome, a war that scared the emperor and the Roman
army. A small nation, or rather only part of it, had challenged Rome
for the third time within a span of seventy years: rst, in the so-
called Great Revolt of 6673/4, then in the revolt in the Jewish
eastern diaspora and in Iudaea between 115 and 117, and now once
73
Cass. Dio 56, 17, 1.
74
CIL V 7817; Plin., n.h. 3:13637.
75
Cass. Dio 60, 22, 1.
76
Tac., ann. 2, 83; tab. Siarensis frag. a, lines 934 = M.H. Crawford (ed.), Roman
Statutes (Bulletin of the Institute of Classical Studies Supplement 64; London: Institute
of Classical Studies, School of Advanced Study, University of London, 1996) 1.515.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 143
144 vrnxrn rck
more in the Bar Kokhba revolt of 132136. And it is more than
likely that this last revolt was not restricted to this province alone,
but probably spilled across its borders to Arabia and perhaps also
to Syria. Enormous eorts and resources were invested in crushing
the rebellion.
77
All the greater was the relief after the nal victory
and the recovery of Roman self-esteem. The arch near Tel Shalem
with its big, carefully carved inscription is proof of that; it is an elo-
quent reection of Roman power, or better, the restoration of Roman
power in this rebellious province. It bears witness to newly gained
self-condence; it is a vindication of Romes tarnished greatness:
Greek would not do; for this, Latin alone could serve as an ade-
quate means of expression.
77
See W. Eck, The Bar Kokhba Revolt: The Roman Point of View, JRS 89
(1999) 7689.
SHIFFMAN_f7_123-144 12/11/02 9:06 AM Page 144
ORAL ESTABLISHMENT OF DOWRY IN JEWISH AND
ROMAN LAW: hrymab ynqnh yrbd AND DOTIS DICTIO
Ranon Katzo
The present study addresses the oral constitution of dowries in Jewish
law, looks at the analogous institution in Roman lawan excep-
tional institution, it should be said at the outset, in that a unilateral
obligation arises from mere speech without ritual or formality, and
as A. Watson perceptively observed, arises from exceptional cir-
cumstancesand then speculates on the possible causal relationship
between them. By its nature the subject lies at the very border of
the framework of this conference and collection of essays on Aramaic
papyri, for it has to do with what is not written on papyrus. It may,
however, provide some explanation of why certain type of docu-
ments are not found among the papyri, and hence what the signicance
may be of what is.
1
Though Roman law spread through the empire and swept before
it local legal systems and practices, with more or less accommoda-
tion to them, it appears to have had little impact on Jewish law. In
contrast to Greek law, from which important legal institutions were
adopted, one is hard put to nd a convincing instance of a partic-
ular Roman legal institution which became part of Jewish law.
2
An
145
1
I wish to express my thanks to the Institute for the Study of Jewish Law, the
Israel Science Foundation, and the National Endowment for the Humanities for
support of work on this study; and to the University of Michigan and its Department
of Classical Studies for the hospitality extended during my stay there while on sab-
batical leave. I also thank Professor Shalom Albeck for his comments on a manu-
script of this paper. The views expressed here are of course my own. My remarks
here are an expansion of, and a revision of the views expressed in, my supple-
mentary notes to the Hebrew translation of A. Gulak, Das Urkundenwesen im Talmud
im Lichte der griechisch-aegyptischen Papyri und des griechischen und roemischen Rechts ( Jerusalem:
Rubin Mass, 1935) (henceforth Urkundenwesen), which appeared as A. Gulak, Legal
Documents of the Talmud in Light of Greek Papyri and Greek and Roman Law (Hebrew)
(edited and supplemented by R. Katzo; Jerusalem: Magnes Press, 1994) (hence-
forth Legal Documents), and appeared in slightly dierent form in Critical Studies in
Ancient Law, Comparative Law & Legal History: Essays in Honour of Alan Watson (eds
J. Cairns and O. Robinson; Oxford/Portland, OR: Hart, 2001) 15771.
2
For the view that Jewish law did not adopt Roman institutions and some
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 145
146 n.xox k.+zorr
speculation as to why that might be the case, see R. Katzo, Sperbers Dictionary
of Greek and Latin Legal Terms in Rabbinic LiteratureA Review-Essay, JSJ 20
(1989) 2046.
3
Gulak, Urkundenwesen, 4452; Gulak, Legal Documents, 6370.
4
B. Cohen, Betrothal in Jewish and Roman Law, PAAJR 18 (1949) 107,
reprinted in B. Cohen, Jewish and Roman LawA Comparative Study (New York: Jewish
Theological Seminary of America, 1966) 1.319.
5
B. Cohen, Dowry in Jewish and Roman Law, Mlanges Isidore Lvy (Brussels:
Secrtariat des ditions de lInstitut, 1955) = AIPHOS 13 (1953) 69, reprinted in
Cohen, Jewish and Roman LawA Comparative Study, 360.
6
A detailed account of the institution in Jewish law, particularly in the post-
talmudic juristic literature, is provided in Talmudic Encyclopedia (eds M. Bar-Ilan
and S.Y. Zevin; Jerusalem: Talmudic Encyclopedia Press with the assistance of the
Rav Kook Institute, 1956) 7.13849 s.v. hrymab ynqnh yrbd (Hebrew).
7
To your son, or for your son. Cf. A.R.W. Harrison, The Law of Athens, vol. 1:
The Family and Property (Oxford: Clarendon Press, 1968) 49, on the ambiguity of the
dative in the analogous Greek expression.
8
And appears only in the version in b. Ket. 102ab. Whether the two clauses
are to be read disjunctively or conjunctively, that is whether each side may become
obligated in this manner even without the participation of the other side, or the
obligations take eect only if they are reciprocal, was subject to a post-talmudic
dispute. Menahem Me"iri, Qid. 9b (ed. Sofer, repr. Jerusalem, 1963, pp. 6869);
Talmudic Encyclopedia, s.v. hrymab ynqnh yrbd nn. 3537.
exceptional instance may be the recognition of mere speech as bind-
ing in the constitution of a dowry, known in Hebrew as yrbd
hrymab ynqnh, things acquired by speech. The two major scholars
of the earlier part of this century who addressed the issues of con-
tacts between the Roman and Jewish legal systems, Asher Gulak and
Boaz Cohen, both assigned a Roman origin to this Jewish institu-
tion. Gulak identied it as stipulatio, though he raised and immedi-
ately rejected the possibility that dotis dictio is involved.
3
Cohen rst
followed Gulak,
4
and later in a brief paragraph, identied the Jewish
institution as deriving from the Roman dotis dictio, but his repeated
use there of the verb stipulate in a non-Roman sense obscures the
point.
5
In the following I will present briey the basic texts and rules
on hrymab ynqnh yrbd,
6
suggest that dotis dictio rather than stipula-
tio be considered as the possible Roman model, and nally question
whether any Roman origin should be sought at all.
The talmudic text, as it appears in the Babylonian Talmud (BT),
is as follows:
twn hta hmkw ;kw k ?nbl twn hta hmk :br rma ldyg br rma
.hrymab ynqnh yrbdh h h .wnqwdqw wdm[ .kw k ?tbl
Rav Gidel said in the name of Rav: [The parents said,] How much
do you give your son?
7
Such and such. And
8
how much do you
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 146
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 147
give your daughter? Such and such. [If ] they proceeded to betroth,
they acquired. These are the very things which are acquired by speech
(hrymab ynqnh yrbd).
The text appears in three passages of the Babylonian Talmud and,
as will be seen, with some variation in one passage of the Palestinian
Talmud (PT). In two of the passages of the former, b. Mo'ed Q.
18b and b. Qid. 9b, the text is quoted without comment to illus-
trate the meaning of the term atqysp yrf, documents recording
nancial settlements in anticipation of marriage. The contexts are
discussions concerning, respectively, which sorts of documents may
be written on the intermediate days of festivals and which sorts of
documents may be written without the parties agreement to the fact
of writing (in contradistinction to their agreement on the transac-
tion). In the third passage, b. Ket. 102ab, the text is again quoted
to illustrate atqysp yrf, this time in the context of a discussion of
the views of Rabbi Yoanan and Resh Lakish, both Palestinian
amoraim,
9
on whether a document asserting an indebtedness for which
there is no factual background is binding. In this passage, however,
the text is also the object of two short discussions, to which we will
return.
In y. Ket. 5:1 (29c), a passage following closely after, though not
an integral part of, a discussion of the same issue as in b. Ketubot,
involving the same two amoraim, the text is as follows.
wnya bahw btkb hkzm l[bh ala ,qswp bah k qswp l[bh k :ynt
b lwdyg ?adyb[ yh .hrymab ynqn h yrbd dblbw ,yrbdb ala hkzm
wyk ;kw k ?tbl twn hta hmk ;kw k ?nbl twn hta hmk ,br
.twnbh yb tkz hdyq
It has been taught, just as the husband can make a marriage settle-
ment [for his wife] so the father can make a marriage settlement [for
his daughter], but whereas the husband makes a settlement in writ-
ing, the father makes a settlement by words (yrbdb),
10
and only such
9
Amoraim (singular: amora) are the rabbis of the post-mishnaic part of the Talmud,
dating, roughly speaking, from the third to the fth centuries of this era.
10
I translate as the text is quoted by Me"iri, Ket. 102b (ed. Sofer, Jerusalem,
1949, p. 472) and by Samuel ben Natronai, from Sefer efe, apud Hagahot Maim-
oniyot, H. Ishut 23.9, yrbdb bahw, which seems to be the required sense even of
the text as printed: yrbdb ala hkzm wnya bahw, literally the father settles only by
speech. ala . . . ya . . ., nothing but, is often used hyperbolically, e.g., b. Ber. 5a,
6a, 26b. See Pnei Moshe ad loc. On Sefer efe see N. Danzig, The First Dis-
covered Leaves of Sefer efe, JQR 82 (1991) 51136, who places the compilation
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 147
148 n.xox k.+zorr
things that are acquired by speech (hrymab ynqn h yrbd). How is
this done? Gidul said in the name of Rav: [The parents said,] How
much do you give your son? Such and such. How much do you
give your daughter? Such and such. Once he has betrothed her
she has acquired as against the other daughters.
The point of the nal words of the text in this version is that the
betrothed daughter acquires the rights to whatever the father said
he would give her even if that is more than her fair share of her
fathers wealth.
11
This is subsequently emended by the Talmud to
read in addition, and the son as against the other sons. At this
point the substantive content is identical with that of the statement
as quoted in the BT. This fuller form is then quoted also in the
name of Rabbi Yoanan, and followed by the limitation that it
applies only in the case of a rst marriage, presumably of the child
who is a recipient.
12
Whether the limitation is part of Rabbi Yoanans
words or by the Talmud is not clear.
13
In my translation I have supplemented the text with the more
idiomatic parents as the parties to the dialogue. To be more accu-
rate, however, I should have used fathers, for only fathers are
meant. The pronouns and the verbs in the dialogue are all mascu-
line. Although in Hebrew, masculine gender of words also serves as
the default gender when reference is to persons of both sexes, and
thus the statement could thereby easily be interpreted extensively to
include both, that is not the case here. The discussions of the text
in the Babylonian Talmud turn on the power of the father to give
his minor daughter in marriage and receive the money of qiddushin
for himself, from which mothers are excluded. Finally, there is
conrmation in t. Ket. 6:8 (ed. Lieberman): A man makes a mar-
of work at the beginning of the tenth century cr, perhaps in southern Italy. At 59
n. 24, Danzig reports a reading of a manuscript in the Casanatense Library in
Rome quoting the same passage in Sefer efe as yrbdb a bahw, the father set-
tles even by speech. On precision in quotation of the PT in Sefer efe as against
the authors freedom in reworking material, see Danzig, The First Discovered
Leaves, 8990.
11
Though a father could give any of his daughters as much or as little dowry
as he wanted, observing the minimum set by m. Ket. 6:5, the rabbis set reason-
able proportions to be applied as charges to the estate in case the father was
deceased (m. Ket. 6:56; t. Ket. 6:3; y. Ket. 6:6 [30d]; b. Ket. 68a69b).
12
So Maimonides, H. Ishut 23.14; Shulan Arukh, Even Ha'ezer 51.1; but cf.
Me"iri, Qid. 9b (ed. Sofer, p. 69).
13
Gulak, Urkundenwesen, 45 n. 41; Legal Documents, 63 n. 41.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 148
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 149
riage settlement ( qswp) for his daughter, a woman does not make a
marriage settlement for her daughter. At rst sight this is aston-
ishing, for surely a mother, or anyone else for that matter, can give
or promise any gift, and on whatever terms, she wishes to give her
daughter on the occasion of her marriage.
14
Clearly the term for
make a marriage settlement ( qswp) is used in a special sense.
Though a mother, then, may make a marriage settlement using the
procedures for gift and obligation, only a father becomes com-
mitted to a settlement made orally without the backing of a formal
transfer or a written document properly executed.
15
Signicantly,
when two lines earlier the Tosefta passage discusses marriage settle-
ments made for an orphan girl by her mother or brother the verb
used is not qswp, but btwk, write. The PT incorporates this state-
ment into its discussion of the oral settlement described by Rav, thus,
too, imputing to the term qswp the technical sense of making a mar-
riage settlement orally.
16
14
The transaction recorded in the papyrus P.Hever 64 (formerly known as
Xev/Se Gr. 1, or Xev/Se 64), a gift of real estate by a mother to her daugh-
ter, has been plausibly interpreted as having been occasioned by the marriage of
the latter, a later stage of which is documented in P.Yadin 37. For text and inter-
pretation see H.M. Cotton, The Archive of Salome Daughter of Levi, Another
Archive from the Cave of Letters, ZPE 105 (1995) 18489; and H.M. Cotton
and A. Yardeni, Aramaic, Hebrew and Greek Documentary Texts from Naal ever and Other
Sites with an Appendix Containing Alleged Qumran Texts (The Seiyl Collection II) (DJD 27;
Oxford: Clarendon Press, 1997) 20323. Since we do not know a priori whether
Jewish law is being observed in that document, it is not evidence on Jewish law,
but is evidence of what at least one Jewish woman wished to do. The model for
this interpretation of P.Hever 64 is P.Yadin 19, a gift of a house by a father to
his daughter, no doubt on the occasion of her marriage eleven days earlier recorded
in P.Yadin 18. For a discussion of P.Yadin 19 and in particular its testamentary
aspects in relation to the marriage, see R. Katzo, An Interpretation of P.Yadin
19: A Jewish Gift after Death, Proceedings of the 20th International Congress of Papyrologists,
Copenhagen, 2329 August, 1992 (ed. A. Blow-Jacobsen; Copenhagen: Museum
Tusculanum Press, 1994) 56265.
15
S. Lieberman, Tosefta Ki-Feshuta (New York: Jewish Theological Seminary of
America, 1967) 6.28081.
16
Post-talmudic authorities understood the passages in this way as well. The author
of the Tosafot, Ket. 109a s.v. bt, reports nding it explicit in the passage of the
PT under discussion that only a father can become so obligated, but neither a mother
nor a brother. On the identication of the author as Rabbi Shimshon of Sens (early
thirteenth century), see E.E. Urbach, The Tosaphists: Their History, Writings and Methods
(Hebrew) ( Jerusalem: Mossad Bialik, 1980
4
) 2.627 n. 35. Alfasi, Ketubbot ch. 12
(Vilna, p. 61a), followed by Rabbenu Asher, Ket. 12.3, Me"iri ad Ket. 102b (ed.
Sofer, p. 472), Ritva ad Ket. 102b (ed. Goldstein, p. 804) report the same. Maimonides,
H. Ishut 23.14, followed by the ur and Shulan Arukh, Even Ha'ezer 51.
An exceptional responsum by the rabbis of Bari, quoted and presumably endorsed
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 149
150 n.xox k.+zorr
Whether the spouses could in Ravs view also become obligated
by mere speech is less clear. Nothing is said explicitly on this point
in the BT. The comment of Rava in the discussion of Ravs state-
ment, however, that Ravs dictum is understandable in the case of
a father of a minor girl because he becomes obligated in return for
the value of the money of qiddushin, implies that a bride who had
achieved majority would also by the same reasoning be able to oblig-
ate herself to the dowry by mere speech since it is she who receives
the money of qiddushin. The conclusion of the discussion of the BT,
that the fathers even of adult spouses become obligated by mere
speech in return for the satisfaction of the contracting of the mar-
riage could easily be extended to the groom as well. This reasoning
is used explicitly of the groom in the PT; however, there it is a pro-
pos unilateral obligation by writing, not by speech. On the contrary
the PT attaches to Ravs statement the tannaitic one that a groom
makes a marriage settlement in writing. It is, however, not clear if
the statement refers to possibility or requirement, to what is neces-
sary or to what is sucient. Post-talmudic authorities split on this
question, and ultimately the view prevailed that both spouses could
obligate themselves by speech.
17
The result of the transaction described may be either the creation
of an obligation or transfer of title.
18
The two notions are not as
by Rabbi Eliezer ben Natan (Ravan) (early twelfth century Germany) Even Ha'ezer
38 (repr. Jerusalem, 1975, p. 30a), and more conveniently though at second hand
from Samuel ben Natronai, the son-in-law of Ravan, with slight variation in Hagahot
Maimoniyot, H. Ishut 23.9, decides on the contrary that anyone becomes obligated
to a dowry by mere speech, and that father is specied in Ravs statement only
because that would be the typical case. Rabbi Samuel comments that the rabbis
of Bari must not have known the passage in the PT. (My thanks to Prof. Zvi
Steinfeld for his help in directing me to the identication of the persons involved.)
If the reference by Rabbi Samuel to the authors of the responsum as ytwbr, my
teachers, means that he personally studied with them, as Danzig, First Discovered
Leaves, 62 n. 31a, suggests, the authors must be of the early twelfth century. (Yo"el
in the rst line of that note is a slip for Natan.) Pinchas Halevi Horowitz (1730
1805), Sefer ha-Miqneh (Part II of Sefer Haa"ah), Kuntres Aaron 51.1 (Oenbach,
1801; repr. Jerusalem, 1975, p. 374) suggests that compiling the rules put forth on
constructive transfers of money of qiddushin in b. Qid. 6b7a may result in that any-
one could become obligated by speech alone in the monetary aspects of a dowry.
17
Talmudic Encyclopedia, s.v. hrymab ynqnh yrbd nn. 45 and 51. For the ulti-
mately prevailing view, Maimonides, H. Ishut 23.13; Shulan Arukh, Even Ha'ezer
51.1. For the contrary view, Me"iri, Qid. 9b (ed. Sofer, p. 68); but cf. Me"iri, Ket.
102b (ed. Sofer, pp. 47273).
18
B.M. Lewin, Otzar ha-Geonim VIII: Ketubbot, p. 358, 793, drawn from Ittur, s.v.
Psiqata (Warsaw, 1883; repr. Jerusalem, 1987, p. 132).
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 150
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 151
sharply distinguished in Jewish as in Roman law, and both can come
under the term ynqn, acquired. The language of the dialogue,
How much do you give your son? not What do you give, indi-
cates that what is primarily intended is a monetary obligation.
19
The
PT, in the lines immediately following the statement of Rabbi Yoanan,
associates the statement with a tannaitic statement concerning the
passing of title in real estate from the father to the groom.
20
It should
be noted too, that these marriage settlements are not conned to
dowry, that is transfers and obligations from the bride or her father
to the groom, for such transfers and obligations from the father of
the groom to his own son are also included.
Chronologically, the statements can probably be located in the
second and third centuries of this era. Rav and Rabbi Yoanan are
both early amoraim, Rav in the early third century, Rabbi Yoanan
somewhat younger. It should be noted that in the PT passage the
words hrymab ynqnh yrbd appear not as part of the statement of
Rav, but in the passage immediately preceding it. The word ynt, it
has been taught, normally introduces a statement of tannaim.
21
The
manner in which the phrase hrymab ynqnh yrbd is used, provided
that it is hrymab ynqnh yrbd, seems to indicate an allusion to an
known existing, hence tannaitic, institution. This conrms the simi-
lar conclusion we should draw from the way the phrase is used in
the BT passage, These are the very hrymab ynqnh yrbd, that it is
an allusion to a known institution. Since Rav is of the rst genera-
tion of amoraim, the allusion must be to a tannaitic institution. Even
if it be argued that the phrase is not part of the quotation of Ravs
words but the comment of the editor of the BT passage, the com-
ment must have been made at an early stage of the editorial process,
since the words appear in the identical position every time Ravs
statement is quoted in the BT; and considering the use of the phrase
in the PT, the balance of probability is that the supposed BT edi-
tor in question alluded to a tannaitic institution. Furthermore, now
19
Sefer ha-Miqneh Kuntres Aaron 51.1 (supra n. 16 at end: p. 374).
20
See, e.g., Geonim (early post-Talmudic authorities) cited by Yosef ibn Migash,
in Shiah Mequbbeet, Ket. 55a (repr. 1967, p. 7); and in Ritva, Ket. 102b (ed.
Goldstein, p. 804 at n. 96. See further references in that note). Maimonides, H.
Zekhiah Umatanah 6.17, followed by Shulan Arukh, Even Ha'ezer 51.1, restricts
this to what is in the promissors ownership at the time of the promise.
21
Tannaim (singular tanna") are the rabbis of the Mishnaic part of the Talmud,
dating, roughly speaking, from the rst two centuries of this era.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 151
152 n.xox k.+zorr
that we have seen that in at least one tannaitic source, the term
qswp in the context of betrothal and marriage can bear the special-
ized meaning of making marriage settlements orally, oral proceed-
ings may lie behind any number of the very frequent uses of the
term qswp in tannaitic literature.
22
Geographically, the statements can be placed in the Land of Israel,
hence within the bounds of the Roman Empire. Both Rav and Rabbi
Yoanan are from there, both had studied in the academy of Rabbi
Judah the Prince, though Rav subsequently founded an academy in
Babylonia.
It is dicult to say how common the reliance on ynqnh yrbd
hrymab was in the talmudic period.
23
Perhaps an indication of loss
of interest in it is that there is very little elaboration of the rules
concerning hrymab ynqnh yrbd in the Talmud, except insofar as they
are subsumed under rules for qswp in general. Even issues which
one would expect to be addressed, and are addressed in the context
of other similar rules, are not found in the recorded literature until
well into the post-talmudic period. For instance, there is no discus-
sion a propos the words wdqw wdm[, if they proceeded to betroth,
on whether the betrothal, qiddushin, must follow immediately for
hrymab ynqnh yrbd to be eective,
24
although there is such a dis-
cussion of the analogous situation in connection with the bequest of
a person on his deathbed (b. B.B. 114ab). Similarly, there is no
discussion of the status of these commitments during the period
between qiddushin, betrothal, and nissu"in, marriage, nor if they were
made at the nissu"in rather than at the qiddushin, despite the other-
wise ubiquitous concern with this interim period.
25
The enigmatic
interchange between Ravina and Rav Ashi on whether it is per-
22
Contra Gulak, Urkundenwesen, 45 n. 40, Legal Documents, 63 n. 40, without sup-
porting argument.
23
It was, however, the common practice in medieval Europe according to Moses
of Couci, Sefer Mivot Gadol (SeMaG) 'Aseh 48 s.v. haw ya (ed. Farber, 1991,
p. 113a) (France) and Meir ha-Kohen of Rothenburg, Hagahot Maimoniyot, H. Ishut
23.7 (Germany), and must have been so in southern Italy as well, to judge from
the responsum of the rabbis of Bari (supra n. 16).
24
Disputed by Samuel ben Meir (Rashbam) apud Tosafot, Ket. 109a s.v. bt,
and Qid. 9b s.v. h (betrothal must follow immediately); and Rabbenu Yitzhak
(Ri) apud Shiah Mequbbeet, Ket. 102b s.v. apwg (repr. 1967, p. 301); Mordechai,
Ketubbot 263; and Hagahot Maimoniyot, H. Ishut 23.7 (betrothal need not follow
immediately).
25
Y. Ket. 5:1 (29c), does discuss this, but only with respect to written documents.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 152
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 153
missible (if that is what is meant) to write up the oral transaction
on the grounds that writing entails eects on third parties which
may not have been intended in oral proceedings, and the relation
of that interchange to the existence of atqysp yrf (b. Ket. 102b:
btkl wntn al)
26
does not change this conclusion.
To summarize so far, statements from the Land of Israel of the
second and third centuries describe a dialogue consisting of ques-
tions and answers between fathers of spouses on the occasion of
betrothal, and probably between spouses as well, in which owner-
ship is transferred and obligation arises by mere speech, without the
necessity of writing, or of formal transfer (ynq).
The Talmuds characterization of these proceedings as things
acquired by speech caught the attention of Gulak for its similarity
to the Roman doctrine that verbis obligatio contrahitur ex interrogatione et
responsione, cum quid dari erive nobis stipulamur, an obligation by words
is contracted by means of a question and an answer when we stip-
ulate that anything shall be given to or done for us ( Justinian, Inst.
3.15.pr.; Gaius 3.92). Stipulate is here used in the technical Roman
sense of asking a question of the sort, Do you promise to give, or
do, so-and-so, to which a congruent armative answer must be
given. This sort of dialogue consisting of question and answer cre-
ates an obligation on the part of the promissor without any further
need for writing or formal transfer. Stipulatio, it should be noted is
the only contract included in the category of verbal contracts in
Justinians Institutes. Not surprisingly, then, Gulak xed on stipulatio
as the background for the rabbinic things acquired by words.
Several objections, however, immediately arise, some of which
Gulak himself addressed in greater or lesser degree. First, the ques-
tions in the Jewish dowry dialogue do not match the requirements
of the Roman stipulatio. Essential to stipulatio is that all the details of
the obligation must be spelled out in the question, at least by impli-
cation. The question in the Jewish dialogue, How much do you
give your son? fails this elementary requirement. Gulak attempted
to ward o this objection by pointing to increasing laxity in the rules
concerning stipulatio, but there is no evidence for this particular
laxity.
26
For the range of interpretations that have been given, see Talmudic Encyclopedia,
s.v. hrymab ynqnh yrbd section 5, at nn. 95116.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 153
154 n.xox k.+zorr
Second, it is generally agreed that in the Greek-speaking (or rather
Greek-writing) eastern provinces of the Roman Empire, documents
often contained clauses asserting a stipulatio, but that in fact no oral
proceedings actually took place. In Gulaks reconstruction, however,
rabbinic literature insists on oral proceedings but contains not a hint
of the recording of a stipulatio in a document. We are asked, then,
to believe that the development in Palestine was opposite to that of
the other eastern provinces: elsewhere Greeks did not ask formal
questions and merely wrote that they did; the rabbis required for-
mal questions but so far as appears in rabbinic literature did not
write so in their documents.
Third, on Gulaks thesis that the oral proceedings are stipulatio,
why do the rabbis recognize its binding nature only in connection
with marriage settlements? Any transaction could be stated in terms
of stipulatio, yet the rabbis insist that these purely oral proceedings
bind only in these particular hrymab ynqnh yrbd, things acquired
by speech, to wit, marriage settlements. Gulak averted this objec-
tion in two steps. First, he interpreted the sense of the phrase as
providing these exact words are used, rather than only such things
that are acquired by speech.
27
Gulaks conviction that the words
spoken are purely formal stemmed from his understanding that these
words were said at the wedding, which is the occasion for formali-
ties, not for substantive negotiation. He failed, however, to take cog-
nizance of the fact that in the talmudic period the qiddushin, betrothal,
at which this dialogue took place, was not part of the wedding cer-
emony, as it indeed has been since the Middle Ages, but occurred
much earlier. It was socially the equivalent of the rst meeting of
the parents of a newly engaged couple customary in some Jewish
circles today.
Having taken the limitation to obligation by speech to be to for-
mal words not to dowry proceedings, Gulak suggested that stipulatio
played a greater part in the practice of the Jews of Palestine than
is recorded in the Talmud.
28
Now, indeed it has become apparent
that at least some Jews in the region were familiar with the stipu-
lation clause by the early second century of this era. Seven Greek
documents from the Judaean Desert, written by Jews contain at or
27
Urkundenwesen, 4647; Legal Documents, 6465.
28
Urkundenwesen, 50; Legal Documents, 69.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 154
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 155
near their conclusion stipulation clauses, such as pstei phrvtyh
ka nyvmologyh tata otvw kalw genesyai, In good faith the for-
mal question was asked and it was acknowledged in reply that this
is thus rightly done.
29
Two of these, P.Yadin 18 and 37, are mar-
riage documents containing dowry provisions, and a third, P.Yadin
17, documents a deposit/loan given by a wife to her husband which
could conceivably be associated with their marriage settlement.
30
If
one were inclined to maintain Gulaks view on the relation of yrbd
hrymab ynqnh and stipulatio, one could nd the proceedings of the
rabbinic hrymab ynqnh yrbd behind these clauses. The other docu-
ments, however, P.Yadin 21 and 22, two complementary copies of
an agreement on crops, and P.Yadin 20 and P.Hever 63 (= Xev/Se
Gr. 5), both concession of rights, have nothing to do with marriage.
It should also be noted that all of these documents were written not
in the Land of Israel but in the province of Arabia. It is as yet
impossible to determine whether the use of the stipulatio clause as
early as these documents were written is one of the many strands
that unite the Jews of Provincia Arabia and Judaea, or is a pecu-
liarity of Arabia, newly organized as a Roman province in 106 cr.
31
Yet a further problem is linguistic. yrbd can mean words or
things. Gulak takes the word in the phrase hrymab ynqnh yrbd
to mean words, and the phrase as a whole to mean words which
cause acquisition by the speaking of them. yrbd thus parallels ver-
bis in the phrase verbis obligatio contrahitur or t ( Justinian, Inst. 3.15.pr.;
Gaius 3.92). Granted, there is a certain advantage in this reading in
that the sense of yrbd remains that which the word has in the
immediately preceding passage in the PT, and the father by words.
However, there is a two-fold diculty. On the Latin side, verba is
not the normal expression for stipulation. Rather, nouns as stipulatio
29
P.Yadin 17:3839; 18:2728, 6667; 20:1617, 40; 21:2627; 22:2930; 37:14;
P.Hever 63:13 (formerly known as Xev/Se Gr. 5, line 13, published by H. Cotton,
The Archive of Salome Komase Daughter of Levi: Another Archive from the
Cave of Letters, ZPE 105 [1995] 178). See her discussion of the restoration of
the phrase, at ibid., 18182 and in Cotton and Yardeni, DJD 27.2012.
30
For the possibility that the parties in P.Yadin 17 had only recently married,
see R. Katzo, Polygamy in P. Yadin? ZPE 109 (1995) 12832 n. 14.
31
Cf. H. Cotton, A Cancelled Marriage Contract from the Judaean Desert
(Xev/Se Gr. 2), JRS 84 (1994) 6486, at 65 and 8586; and in DJD 27.15556.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 155
156 n.xox k.+zorr
and promissio, and verbs as stipulare and promittere are. True, in intro-
ductory passages in the institutional works of Gaius and Justinian
the term verba is used, but for a category of transactions. The fact
that in Justinians Institutes stipulatio is the sole remaining member
of the category of verbal contracts can create the illusion of iden-
tity between the member and the category. In Gaius Institutes, how-
ever, more relevant to the Jewish texts of the second and third
centuries, the category of contracts verbis includes dotis dictio and the
oath of a freedman to his patron as well. On the Hebrew side, this
reading overlooks the passive form of the verb ynqnh, that are
acquired. Further, the appreciation that the words ynqnh yrbd
hrymab are a quotation, whether by the baraita or by the editor of
the PT, from an earlier text, as I have argued above, makes the
advantage of consistency of the sense of the word yrbd vanish. yrbd
here must mean things, objects, and obligations, and the phrase
as a whole things that are acquired by speech.
If, then, any Roman institution stands behind the rabbinic yrbd
hrymab ynqnh, it is more likely to be dotis dictio.
32
hryma in the phrase
things acquired by speech (hryma) would thus exactly parallel the
Latin dictio.
Of the very few texts mentioning dotis dictio explicitly which sur-
vived Justinians editors, the most important for our purpose are the
following:
Ulpian 6.12: Dos aut datur aut dicitur aut promittitur. Dotem dicere potest
mulier quae nuptura est et debitor mulieris, si iussu eius dicat. Item parens mulieris
virilis sexus . . . velut pater avus paternus. Dowry is either given, spoken,
or promised. The woman about to be married can speak a dowry,
as can her debtor if he speaks it at her order. So also the womans
male parent . . ., to wit her father or paternal grandfather.
Ep. Gaius 2.9.3: Sunt et aliae obligationes, quae nulla praecedenti interrogatione
contrahi possunt, id est, ut si mulier sive sponso uxor futura, sive iam marito,
dotem dicat. Quod tam de mobilibus rebus, quam de fundis eri potest. Et non
solum in hac obligatione ipsa mulier obligatur, sed et pater eius, et debitor ipsius
mulieris, si pecunia, quam illi debebat, sponso creditricis ipse debitor in dotem
dixerit. Hae tantum tres personae nulla interrogatione praecedente possunt dictione
32
Accounts of dotis dictio can be found in all textbooks of Roman law. The most
recent full-scale study is that of A. Ortega Carillo de Albornoz, Dotis Dictio (Spanish)
(Bologna: Real Colegio de Espaa, 1975), with extensive bibliography. Particularly
insightful is A. Watson, The Law of Persons in the Later Roman Republic (Oxford:
Clarendon Press, 1967) 5763.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 156
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 157
dotis legitime obligari. Aliae vero personae, si pro muliere dotem viro promiserint,
communi iure obligari debent, id est, ut et interrogata respondeant, et stipulata
promittant. There are [verbal] obligations other [than stipulation], and
these can be contracted without a preceding question, such as if a
woman speaks a dowry to her anc or husband. This can be done
with movables as well as with land. Not only the woman herself may
be obligated in this type of obligation, but also her father and the
debtor of the woman herself, if the debtor himself speaks the money
which he owes her to her anc as dowry. Only these three persons
can legally become obligated by the speaking of a dowry with no
preceding question. Indeed other persons who promise a man a dowry
for a woman ought to be obligated by the regular rules, that is that
they respond to questions and promise what was stipulated.
In both texts a sharp distinction is drawn between promising a
dowry according to the usual rules concerning unilateral obligations,
a procedure available to anyone who wishes to contribute to a dowry,
and speaking a dowry, that is constituting a dowry by speech alone,
without a formal dialogue, a procedure available only to the bride
herself, her father and grandfather, and her debtor. The eect of
the speaking was that the person who spoke the dowry was
obliged to give the groom as dowry what he said he would, be it
land or movables, such as slaves, clothes, or money.
33
Within Roman
law it is peculiar in that, except for the oath of a freedman, it is
the only verbal form which is unilateral and proceeds from the
donor.
34
How formal or informal this speech was to be is in dispute.
Most scholars have taken it to be a solemn declaration which required
xed words, phrased impersonally that something doti tibi erit (or erunt),
will be your dowry.
35
Others, represented today by Watson, hold
that the declaration was informal.
36
The similarity of dotis dictio to hrymab ynqnh yrbd is obvious. None
of the objections that were raised above to the identication of yrbd
hrymab ynqnh with stipulatio obtain here, except that on the view that
dotis dictio did have xed words, hrymab ynqnh yrbd fails that require-
ment, too. Particularly notable is the limitation of the procedure to
33
The literary texts refer to land, slaves, and money; the papyri to clothes, jew-
elry, household items, and real estate both residential and agricultural.
34
Watson, Law of Persons, 57.
35
Argued by Ortega, Dotis Dictio, 7983.
36
A. Watson, The Form and Nature of Acceptilatio in Classical Roman Law,
RIDA
3
8 (1961) 392 n. 5, reprinted in A. Watson, Studies in Roman Private Law
(London/Rio Grande, OH: Hambledon Press, 1991) 194 n. 5; So V. Georgesco,
Essai dune thorie gnral de leges privatae (Paris: Rousseau & C., 1932) 8687.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 157
158 n.xox k.+zorr
particular persons. The divergence is easily explainable. ynqnh yrbd
hrymab is available to the groom (probably) and to his father, but
dotis dictio is not, for the obvious reason that in Jewish practice it is
customary for the groom and his family to contribute to the dowry
but in classical Roman law the possibility is entirely excluded.
37
Com-
pare m. Ket. 6:3: If the bride settled (hqsp) to give the groom as
dowry a thousand dinars, he settles (qswp) correspondingly fteen
minas (= 1500 dinars), and Codex Justinianus 5.15.1: Delivery,
not the writing of a document, creates a dowry.
38
On the other
hand, dotis dictio is available to the debtor of the bride, and only of
the bride, but there is no corresponding rule in connection with
hrymab ynqnh yrbd. This Roman rule, in particular the exclusion
of the fathers debtor, caused such consternation among Romanists
that some were moved to deny its existence in classical law.
39
An
attractive explanation put forth relatively recently, as these things go,
by C. Russo Ruggeri, for this apparent anomaly will explain its
absence from the analogous Jewish rule as well. The rule, it is sug-
gested, is intended to provide a way for the bride, who in such an
instance must ex hypothese be sui iuris, to get around an uncoopera-
tive tutor. Her access to her invested assets becomes easier for the
purpose of facilitating her marriage, and thereby incidentally change
of tutor as well.
40
This special arrangement would not have been
needed for hrymab ynqnh yrbd since in Jewish law an adult woman
is not subject to a guardian at all.
37
P.Mich. VII 434 + P.Ryl. IV 612 = FIRA III 17 = CPL 2089 = ChLA IV
249 may contain an exceptional instance. At line 16 of P.Mich. 434 the husband
declares that he has brought in certain property in the vicinity of Philadelphia (et
ipse intulisse se dixit ad vic[um Philadelphiam). See F. De Visscher, Document sur la
donatio ante nuptias, CdE 37 (1944) 1017. The publication of P.Ryl. 612 pre-
cludes the restoration of the line as ad vic[em dotis, making the matter somewhat
more speculative than thought earlier. The document as a whole is remarkable for
containing many non-Roman elements.
38
See further CJ 5.3.1; D. Daube, Roman Law: Linguistic, Social and Philosophical
Aspects (Edinburgh: Edinburgh University Press, 1969) 10912; R. Katzo, Donatio
ante nuptias and Jewish Dowry Additions, Papyrology (ed. N. Lewis; YClS 28; Cambridge:
Cambridge University Press, 1985) 23144.
39
Ortega, Dotis Dictio, 10715, with bibliography.
40
C. Russo Ruggeri, Il debitor parentis e la dotis dictio, Labeo 25 (1979)
3946.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 158
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 159
Use of the verb dicere in the context of dowry in writings of Cicero
41
and the younger Pliny
42
shows that elite Roman society observed
these distinctions of the jurists. Two Latin papyri from Egypt of the
second century cr show that they were observed by lower levels of
society as well, and reached the eastern Mediterranean. In each it
is said that a wife dixit deditque, spoke and gave a dowry to her
husband.
43
Similarity of Jewish and Roman law on any matter and the avail-
ability of a Roman model for Jewish law would not be sucient
grounds on which to make a case for adoption by Jewish law if the
institution could be explained as a natural development within Jewish
law.
44
In this case, however, the institution of hrymab ynqnh yrbd
is exceptional, since in Jewish law on the whole, transactions require
for their eectiveness a transfer, at least constructive, of some object
of value or a document.
45
Indeed it is declared by both the Talmuds
41
E.g., Pro Flacco 86: Doti, inquit, Valeria pecuniam omnem suam dixerat. Valeria,
he says, had settled all her money upon him for a dowry, said of the bride her-
self.
42
Epistles 2.4.2: cumque . . . nubenti tibi in dotem centum milia contulerim, praeter eam
summam quam pater tuus . . . dixit. And since . . . I had contributed 100,000 sesterces
towards your dowry when you married in addition to what your father assigned
you . . . Note the change in the verbs used for giving the dowry: dixit, spoke, of
the father of the bride; but contulerim, conferred, gave of Pliny, neither father nor
creditor of the bride.
43
P.Mich. VII 434.4 + P.Ryl. IV 612.3 (dixit et debit [read: dedit]) = FIRA III 17
= CPL 2089 = ChLA IV 249 (The earlier published reading of P.Mich. VII 434.4
is changed by the discovery of the copy in P.Ryl. 612.3); P.Mich. VII 442.9 =
FIRA III 20 = CPL 210 = ChLA V 295, which for this purpose should be read
with A. Berger, Miscellanea Papyrologica. I: P.Mich. Inv. 4703 and Dictio Dotis
in Roman Law, JJP 1 (1946) 1328, reprinted in BIDR 5556 (1951) 98113.
Each of these papyri has aroused considerable discussion. Bibliographies accompany
the republication of each in CPL and ChLA. For a review of the various opinions
concerning the general nature of P.Mich. VII 442, see R.O. Fink, P.Mich. VII 422
(read 442) (inv. 4703): Betrothal, Marriage, or Divorce? Essays in Honor of C. Bradford
Welles (American Studies in Papyrology 1; New Haven, CT: American Society of
Papyrologists, 1966) 917.
44
For a discussion of general considerations on the possible inuence of Roman
on Jewish law, see B. Jackson, On the Problem of Roman Inuence on the Halakah
and Normative Self Denition in Judaism, Jewish and Christian Self Denition, vol. 2:
Aspects of Judaism in the Graeco-Roman Period (ed. E.P. Sanders with A.I. Baumgarten
and A. Mendelson; London: SCM Press, 1981) 157203.
45
Exceptional, but not unique. Other exceptions are tl dm[m, the transfer of
an obligation toward a creditor from a debtor to his own debtor, done orally in
the presence of all three (noted by Gulak, Urkundenwesen, 46 n. 43, Legal Documents,
64 n. 43); br[, guarantor; [rm byk yrbd, the bequest of a person on his deathbed;
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 159
160 n.xox k.+zorr
to be anomalous. In y. Ket. 5:1 (29c), Rabbi Hiya bar Yosef nds
it extraordinary that the father of the groom would become oblig-
ated since he, in contrast to the father of the bride, receives noth-
ing in return, and Rabbi Zeira nds it no less extraordinary that
the groom could acquire rights to the dowry in the same transac-
tion in which he acquires his wife.
46
In b. Ket. 102b, it is Rava
who raises the former of these considerations: Ravs statement makes
sense if the bride is a minor, for then the father receives a benet
[the money of qiddushin, in return for his obligation to the dowry],
but not if the bride is an adult [since then the father does not receive
the money of qiddushin]. Yet, by God, Rav did say that, for other-
wise how does the father of the groom become obligated [for cer-
tainly no money of qiddushin comes to him].
47
The legal source,
then, would be original rabbinic legislation, rather than juristic inter-
pretation.
48
The historical source, however, of the anomalous, and
so recognized within the system, institution of hrymab ynqnh yrbd,
it may be argued, is the Roman dotis dictio.
and hwbgl hryma, consecration. The latter two eect change of ownership but not
obligation. Pinchas Halevi Horowitz, Sefer ha-Miqneh (supra n. 16) suggests the appli-
cation of the limits of these latter two to hrymab ynqnh yrbd as a way of explain-
ing the problematic restriction set in Maimonides, H. Zekhiah Umatanah 6.17, cited
supra n. 20. Accordingly the phrase in the PT and only such things as are acquired
by speech, would mean that can be acquired in the other transactions-by-mere-
speech, that is, bequest and consecration.
46
The bracketed explanatory additions to paragraphs I and K in the translation
by J. Neusner, The Talmud of the Land of Israel: A Preliminary Translation and Explanation,
vol. 22: Ketubot (Chicago and London: University of Chicago Press, 1985) 167 obscure
the point, though the translation itself is accurate.
47
The subsequent explanation of the Talmud, by which the parties become oblig-
ated in consideration of their satisfaction at the transaction itself, even though they
may receive no material benet whatever, is said to be characteristic of Rav Ashi,
mid-fth century cr, but not earlier in the BT, though its roots may be traced to
earlier Eretz Israel, e.g., the pericope immediately preceding the one under dis-
cussion in the PT. See S. Friedman, dwmltb ynqw hanh, Dinei Israel 3 (1972) 11545.
It is dismissed entirely by Me"iri, Qid. 9b (ed. Sofer, p. 67), who stresses the extra-
ordinary character of hrymab ynqnh yrbd. Professor Albeck observes that even if
Friedmans late dating of the Talmuds explanation is not accepted, the explana-
tion still represents the eorts of the rabbis to put a juristic construction on a prac-
tice required by social realities.
48
For a strong statement of the position that hrymab ynqnh yrbd was rabbinic
original legislation in response to a perceived desire of the community that oral
commitments made at betrothal be binding, and not a juristic construction of the
will of the specic donor, see Yeezkel Landau, Noda' Biyehudah, oshen Mishpa
28, s.v. ybwha dw[ (Prague, 1776).
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 160
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 161
It is precisely Watsons perceptive observations on the origin of
dotis dictio, however, which gives pause to accepting dotis dictio as the
origin of hrymab ynqnh yrbd. The Roman dotis dictio, with its lack
of formality, developed precisely to give legal eect to what etiquette
required. Scenes of betrothal and dotis dictio in Roman comedy show
that the correct thing to do was for the father of the bride to declare
that he is betrothing his daughter to a man and at the same time
state, directly or obliquely, what the dowry would be. Though there
may be haggling later, it was not good form for the prospective
groom to start by questioning the father of his beloved if the dowry
was going to be so-and-so much. This etiquette, though as it hap-
pens not dotis dictio itself, is evident in plays of Plautus.
49
Both the
etiquette and the terminology of dotis dictio are evident in Terence,
most strikingly in Heaut. 93542:
50
vrxrrrvts: (the father of the groom): What dowry shall I say you
spoke (dixisse) for my son? Why are you silent?
cnnrvrs: (the father of the bride): Dowry?
vrxrrrvts: Thats what I said.
cnnrvrs: Ah.
vrxrrrvts: Chremes, dont worry about it if there isnt any. The
dowry doesnt matter to us.
cnnrvrs: I thought that two talents would be enough considering our
means.
Note the exact parallel to the statement of Rav: How much do you
give your daughter? Such-and-such. In the following lines Chremes
insists that the transaction be described so that it seems that all his
property will go to his daughters dowry at the expense of his sons
expectations. Note again the parallel to the end of Ravs statement,
she acquired as against the other daughters.
The fact that dotis dictio is most obvious in Terence, whose plays
are direct translations from Greek, should raise the question of whether
this institution was Greek. Romanists did raise this question with an
eye to whether this vitiated the value of these passages as evidence
for Roman law, or to whether this implied that the institution should
be seen as an importation from Greek law. Both questions have been
49
E.g., Aulularia 25561; Trinummus 115663.
50
Cf. Andria 95051: The dowry, Pamphilus, is ten talents. I agree; Alternative
ending 2021: So, I betroth my daughter Philumena to you and promise dowry
of six talents.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 161
162 n.xox k.+zorr
answered in the negative.
51
However, that does not diminish the use-
fulness of the passages as evidence of Greek practice. If earlier schol-
ars could dismiss the entire question because no such passage was
among the available fragments of Greek New Comedy, that is no
longer the case. Passages of Menander and other authors have come
to light to conrm that there is no dierence between Roman and
Greek etiquette in this matter. Three representative examples from
dierent sources may suce:
52
Menander, Dyscolus 84244: Well, I betroth my daughter now, young
man, to you to produce legitimate children. I add three talents dowry.
P.Oxy. XXXI 2533:46: I betroth her, my daughter, to you, Moschion,
to produce lawful children. The dowry you know yourself.
Chariton, Callirhoe 8.8.12: Before you all I and Callirhoe oer our
thanks to our friend, Polycharmus. He has shown us true devotion
and loyalty, and, if you consent, let us give him my sister as his bride,
and for a dowry he shall have a share of the spoils.
The condence with which scholars such as L. Beauchet
53
and
A. Berger
54
asserted that dotis dictio had nothing to do with Greek
practice owed from their conviction that dotis dictio required the
pronouncement of xed formulas, which would be alien to Greek
law. That does not mean, however, that there was not a completely
informal institution in Greek law which otherwise functioned as dotis
dictio.
If, then, it be accepted that Greek law recognized the binding
nature of an informal promise of dowry even without the benet of
writing,
55
similar to the Roman institution of dotis dictio, it is not
Roman law to which one should look for the background to the
Jewish rule on hrymab ynqnh yrbd. Rather, two other possibilities
should be considered.
51
See discussion in Ortega, Dotis Dictio, 3248.
52
For more examples see R. Katzo, Greek and Jewish Marriage Formulas,
Classical Studies in Honor of David Sohlberg (eds R. Katzo with Y. Petro and D. Schaps;
Ramat Gan: Bar Ilan University Press, 1996) 22334.
53
L. Beauchet, Histoire du droit priv de la rpublique Athnienne (Paris: Chevalier-
Maresq, 1897) 1.278.
54
A. Berger, Dotis Dictio im rmischen Recht (summary of Dotis dictio w
prawie rzymskiem), Bulletin International de lAcadmie des Sciences de Cracovie. Classe de
Philologie, Classe dHistoire et de Philosophie. 1909 (Cracow: Imprimerie de lUniversit,
1910) 83.
55
So Harrison, The Law of Athens, 50.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 162
on.r rs+.nrisnvrx+ or rovnv ix rvisn .xr nov.x r.v 163
One is that the rule, exceptional in Jewish law, that a promise or
transfer of dowry could be eected with neither writing nor formal
transfer, may have entered Jewish law from Greek legal practice
along with the practice of dowry itself. If so, it must have occurred
early in the Hellenistic period, for the shift from brideprice (rhwm),
the practice in biblical times, to dowry was complete by the time of
Shim'on ben Sheta, ca. 100 ncr.
56
This would explain the lack of any reference to oral proceedings
in the few Jewish marriage documents which survived in the caves
of the Judaean Desert. In contrast to Roman marriage documents,
which do refer to previous oral establishment of the dowry,
57
none
of the Greek marriage documents on papyrus reect the oral nature
of the proceedings that established a dowry, even though Hellenistic
comedy and ction can give us condence that oral proceedings did
indeed sometimes take place. Neither is it ever said explicitly that it
is the document per se which brings the dowry into existence. It
was simply not thought that it was signicant to record whether
there were oral proceedings before the writing of the legal document
or not.
The same may be said of the few Jewish marriage documents that
survive from antiquity. In none is there any reference to oral pro-
ceedings, nor for that matter to a constitution of the dowry by the
fact of writing. The Jewish marriage documents, then, do not give
evidence one way or another on whether there were oral proceed-
ings, of the sort one would expect from the baraita, the father set-
tles by words. Of course, if there were any transactions of this sort
which were deliberately not written, as one could interpret the baraita,
we would have no record of it.
56
E. Bickerman, Two Legal Interpretations of the Septuagint, RIDA 3 (1956)
94, reprinted in E. Bickerman, Studies in Jewish and Christian History (AGJU 9; Leiden:
Brill, 1976) 1.20124. For a contrary view seeing in Aramaic papyri from Elephantine
evidence of Semitic roots to dowry practice among Jews, see R. Yaron, Introduction
to the Law of the Aramaic Papyri (Oxford: Clarendon Press, 1961) 5051; B. Porten,
Archives from Elephantine: The Life of an Ancient Jewish Military Colony (Berkeley and Los
Angeles: University of California Press, 1968) 7476; M.A. Friedman, Jewish Marriage
in Palestine: A Cairo Geniza Study (Tel-Aviv and New York: Jewish Theological Seminary
of America, 1980) 1.292 n. 11.
57
To the documents cited in n. 43 may be added a Latin marriage document
which refers to establishment of dowry by promissio: PSI VI 730.4 = CPL 207 =
ChLA XXV 783. The sole instance known to me of a reference in a Greek papyrus
marriage document to an earlier promise of dowry is P.Tebt. I 104.1213 (92 ncr).
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 163
164 n.xox k.+zorr
Alternatively, and I think more likely, the rule of ynqnh yrbd
hrymab, exceptional as it may be, developed independently in Jewish
law, as it did in Greek law and in Roman law, and, we may sur-
mise, for the same reasons. At a betrothal, two families join in a
prospective marriage, and good manners require that behavior
between them be that which prevails within a single family, where
family goods are distributed by trust and aection and without for-
mality. Yet there must be provision for the real possibility that mem-
bers of the two families will not always get along. Hence the recognition
by the jurists of a binding quality to things said in good manners
on the occasion of betrothal. Whatever the fathers say at that most
delicate occasion as to what they intend to give for the support of
the new young couple, they will have to stand by. To paraphrase
Watson,
58
the form of hrymab ynqnh yrbd, odd though it may be
as a Jewish creation, is, like dotis dictio, natural in terms of the social
circumstances under which it arose.
58
Watson, Law of Persons, 57.
SHIFFMAN_f8_145-164 12/2/02 5:24 PM Page 164
WITNESSES AND SIGNATURES IN THE HEBREW
AND ARAMAIC DOCUMENTS FROM THE
BAR KOKHBA CAVES
Lawrence H. Schiman
I. Introduction
It is well known that the Hebrew and Aramaic legal documents from
the Judean Desert are for the most part double documents, also
known as tied deeds.
1
Tied deeds were witnessed on the verso by
several witnesses whereas simple deeds were witnessed by two wit-
nesses on the recto. The double deed, as its name implies, is writ-
ten twice with the upper version tied.
For double deeds, the rst signature was that of the person for
whom the text had been written. He started his signature on the
back of the document, Close to the tie, opposite the start of the
lines on the lower version on the recto.
2
In the deeds in Semitic
languages, the rst signature was opposite the right margin of the
recto, and the signatures were written downwards from the stitch-
ing if they were in Hebrew, Aramaic, or Nabatean. If they were in
Greek, they were written upwards toward the stitching.
3
While
our study concentrates on the Hebrew and Aramaic documents, in
order to compare them to Jewish legal texts, and omits Semitic
165
1
H.M. Cotton and A. Yardeni, Aramaic, Hebrew and Greek Documentary Texts from
Naal ever and other Sites: With an Appendix Containing Alleged Qumran Texts (The Seiyl
Collection II) (DJD 27; Oxford: Clarendon Press, 1997) 9. Hannah Cotton was kind
enough to share with me the draft of her paper, Scribes, Notaries, Subscribers,
Witnesses, Signatures, Dating and Other External Aspects of Legal Contracts in
Papyri and Rabbinic Sources, delivered at the conference Rabbinic Law in its
Roman and Near Eastern Context held at Trinity College, Dublin, 1112 March
2002. Her study was being written as the present article was being prepared for
publication and readers will do well to read them together.
2
Yardeni, DJD 27.11. Cf. C. Hezser, Jewish Literacy in Roman Palestine (TSAJ 81;
Tbingen: Mohr Siebeck, 2001) 3045.
3
Cf. Y. Yadin, Bar Kokhba: The Rediscovery of the Legendary Hero of the Second Jewish
Revolt against Rome (New York: Random House, 1971) 22931; J.T. Milik, Deux
documents indits du dsert de Juda, Bib 38 (1957) 25658.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 165
166 r.vnrxcr n. scnirrv.x
4
N. Lewis, The Documents from the Bar Kokhba Period in the Cave of Letters: Greek Papyri
( JDS 2; Jerusalem: Israel Exploration Society, Hebrew University, Shrine of the
Book, 1989) 7.
5
Lewis, Greek Papyri, 68.
6
Lewis, Greek Papyri, 9; Cotton, DJD 27.141. See R. Haensch, Zum Verstndnis
von P. Jericho 16gr, SCI 20 (2001) 16266.
7
Yardeni, DJD 27.9 and n. 1.
8
Cf. Haensch, Zum Verstndnis von P. Jericho 16gr, 163 and n. 20.
subscriptions in Greek documents, we should note that double doc-
uments were widespread in the Greek documents from the Judean
Desert as well. They go back to the usages of ancient Mesopotamia
(cf. Jer. 32:912 which refers to such a document). Such documents
represent only a small portion of the Greek texts that have been
published, fewer than 100.
4
When written on waxed tablets in Roman
usage, such a text was termed a diploma. Such double texts were
designed to protect the inner text from tampering while making avail-
able the outer text for consultation. By the second century ncr in
Egypt, the inner text was reduced to a docket and only the outer
was written fully. In Egypt this practice generally went out of use
by the early rst century cr after Egypt became a Roman province,
but it continued in the eastern provinces of the Empire. It is dur-
ing this period that we observe this phenomenon in Judea and Arabia
in the rst and second centuries in Hebrew, Aramaic, Nabatean,
and Greek documents.
5
This usage is also attested in papyri from
the Middle Euphrates. Generally, the outer text was produced rst.
6
These documents were generally written across the bers and rolled
from top to bottom or the reverse.
7
II. Evidence of the Greek Documents
This presentation will concentrate on the witnesses and signatures
on the Hebrew and Aramaic documents from the Judean Desert.
Before proceeding to the detailed study, it is worth sketching the
evidence regarding witnesses of the Greek materials from this col-
lection as discussed by H. Cotton.
Normal Ptolemaic double documents had six witnesses, whereas
Roman military diplomas required seven, and seven may appear in
other contexts as well.
8
The majority of Judean Desert texts, how-
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 166
nrnnrv .xr .n.v.ic roctvrx+s 167
ever, have ve witnesses as do many other deeds from the Greco-
Roman Near East. Some Greek double documents from the Babatha
archive have seven witnesses. No Greek document has six witnesses.
Eight Greek documents have ve witnesses.
9
P.Yadin 18, a Greek
marriage contract, has seven signatures on the verso, but the rst
two are the brides father and the groom. Thereafter follow one
Greek signature and four in Hebrew and Aramaic, making two prin-
cipals and ve witnesses most probably.
10
A variety of signatures appear in subscriptions. A Greek text
(Xev/Se 60) is subscribed hbtk hyr, Reisha underwrote this.
This is a Greek document so hbtk has to mean that he underwrote
the receipt certifying that the money had been paid, that is, had
it written. P.Yadin 20 and Mur 42 have a similar formula in which
the subject of the verb hbtk attests responsibility for the contents of
the document.
11
Xev/Se 61 contains entirely two subscriptions in
Greek. Xev/Se 64, a Greek deed of gift, has seven signatures of
witnesses. The rst is of the donor, and the second of the husband
who must have signed for her. The procedure of a persons signing
for someone else, then adding his own name, is common in Judean
Desert documents.
12
The signature on his own behalf is followed by
the words hpn l[, indicating that the signatory is one of the parties
to the deed and that he or she was present when it was written.
13
III. The Hebrew and Aramaic Documents
In what follows we examine each of the Hebrew and Aramaic legal
documents in terms of their witnesses and signatures, as well as
9
Not counting the Jericho Papyri.
10
Cotton, DJD 27.14243. The Nabatean texts in the Babatha archive had ve
witnesses. P.Yadin 1 rst gives the signature of the debtor and last that of the scribe
so that there are probably ve witnesses. P.Yadin 2 and 3 also have seven signa-
tures, the rst of the seller and the last of the scribe, leaving ve witnesses.
11
Cotton, DJD 27.17273.
12
Personal communication from Hannah Cotton. See also Cotton, DJD 27.144.
Cf. Y.Yadin, J.C. Greeneld, A. Yardeni, and B.A. Levine, The Documents from the
Bar Kokhba Period in the Cave of Letters: Hebrew, Aramaic and Nabatean-Aramaic Papyri
( JDS 3; Jerusalem: Israel Exploration Society; Institute of Archaeology, Hebrew
University; Shrine of the Book, Israel Museum, 2002) 1113; Hezser, Jewish Literacy,
28183.
13
Cf. Y. Yadin, Expedition DThe Cave of the Letters, The Expedition to the
Judean Desert, 1960 ( Jerusalem: Israel Exploration Society, 196162) 2.253.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 167
168 r.vnrxcr n. scnirrv.x
whether they are simple or tied deeds. Texts are in Aramaic unless
explicitly noted.
Mur 18Acknowledgment of Debt (5556 CE)
This is a double deed with the signatures on the verso. The debtor
signs his name followed by hpn l[, for himself, meaning that he
conrms the contents of the document, and then appears the sig-
nature of the one who signed for him, followed by hrmm, at his
command. Two witnesses then appear with the name followed by
dh and d[.
14
The signatures are on the verso. The rst witness is
the scribe of the document,
15
and his name is preceded by btk, indi-
cating that he wrote the document.
Mur 19Writ of Divorce (111 CE)
This is a double document with signatures on the verso. The rst
signature is that of the husband hpn l[, followed by three witnesses,
each of whose name is followed by dh. The rst of the three wit-
nesses is the scribe, and the second witness is most probably the
scribes brother.
16
Mur 20Marriage Contract (117 CE)
Only one signature is preserved on the verso of this double deed,
that of the groom. To be restored is hpn l[.
17
Mur 21Marriage Contract (1st half of 2nd century CE)
The verso of this double deed has seven signatures. The rst is that
of the groom who signs hpn[ l[]. Then comes the scribe ([a]rps).
The third is that of the bride followed by btk hpn m. The mas-
culine verb is dicult. Perhaps read [h]btk, or btk can refer to
the next signatory who may have signed for the bride. Then follow
14
Cf. Hezser, Jewish Literacy, 348.
15
J.T. Milik, in P. Benoit, J.T. Milik, and R. de Vaux, Les Grottes de Murabba't
(DJD 2; Oxford: Clarendon Press, 1961) 104.
16
Milik, DJD 2.109; cf. Cotton, DJD 27.143.
17
Milik, DJD 2.111.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 168
nrnnrv .xr .n.v.ic roctvrx+s 169
the names of four witnesses, the last two of which are designated by
d[ and dh, which must be restored for the rst two. For the rst
witness, his place of origin is given.
18
Mur 22Deed of Sale of Land (131 CE)
This is a Hebrew double deed with the signatures on the verso.
There seem to have been seven signatures originally. The rst three
are not preserved, but the fourth appears to be that of the scribe.
19
We can probably assume that the rst was the seller, and the sec-
ond may have been that of the buyer.
Mur 24Farming Contracts (133 CE)
This is a series of eleven Hebrew documents on one manuscript for
rental (hrykj) of farmland. On each contract there were two signa-
tures. The rst is the lessee (rkj) followed by hpn l[. Then comes
Simeon ben Kosiba, hrmam m, by his order. Probably his name
was signed, as authorized by his representative whose name appears
in the documents but who does not sign his own name, but that of
the ruler. It is probably the scribe who signed the name of the lessee
as well as that of Ben Kosiba (Bar Kokhba).
Mur 27Deed of Sale (ca. 68 CE?)
20
This is a simple deed with the signatures on the recto. Only one
name appears, that of the seller, followed by hpn l[, and the rest
is not preserved.
Mur 28Deed of Sale (ca. 6673 CE?)
21
There are two signatures on the recto of this double deed, and
both are followed by hpn l[. It is probable that the rst is the
18
But cf. Cotton, DJD 27.143 and n. 50 for Yardenis alternate reading.
19
Milik, DJD 2.119, 121.
20
A. Yardeni, Textbook of Aramaic, Hebrew and Nabataean Documentary Texts from the
Judaean Desert and Related Material, vol. B: Translation, Paleography, Concordance, part I:
The Documents ( Jerusalem: Hebrew University, Ben-Zion Dinur Center for Research
in Jewish History, 2000) [23].
21
Yardeni, Textbook, B.[23].
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 169
170 r.vnrxcr n. scnirrv.x
seller, and the second is the guarantor mentioned explicitly in the
contract.
Mur 29Deed of Sale (133 CE)
This double document written in Hebrew includes ve signatures on
the verso. The rst is in Greek and is of the seller who indicates
that he signed with his own hand (also in Greek). Then comes the
signature of a woman, apparently the wife of the owner, who states
that [hla] ybt[k]h hpn l[, for herself, [she borro]wed the le[t]ters,
meaning that she asked someone to sign on her behalf as she was
illiterate.
22
Then follow three witnesses. The document stated that
the sale was executed in front of signatories (ymtwj) who are listed
with four names appearing, but only two of them appear signed
below on the verso. The other two may have signed on the recto.
23
Mur 30Deed of Sale of Plot (134 CE)
On the verso of this double Hebrew document are signed the seller,
hpn l[, then his wife, hpn l[, then three witnesses. The rst wit-
ness is the scribe, one of the signatories mentioned in the docu-
ment, but the other two witnesses are not. Four signatories had
been mentioned above in the document. The wife had signed because
in the text she explicitly certies that she has no claims against
this sale.
Xev/Se 8aDeed of Sale (134 or 135 CE) (Kfar Baryu)
This text states explicitly that it is a simple deed, and that the sig-
natories are therefore on the inside (line 14). On the right side, ush
right, is the signature of the seller, a line above to the left is the
signature of a woman, the wife of the seller, who in the document
asserts that she has no claims against her husband. Her signature is
followed by btk hpn l[, perhaps showing that she could indeed
write her name. But A. Yardeni
24
takes the word btk, which is mas-
22
See H.M. Cotton, Subscriptions and Signatures in the Papyri from the Judaean
Desert: The Cherochrestes, JJP 25 (1996) 2940.
23
See Milik, DJD 2.14344.
24
DJD 27.37.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 170
nrnnrv .xr .n.v.ic roctvrx+s 171
culine, as referring to the next signatory, thus attributing literacy to
a male, perhaps the scribe, who wrote on behalf of the illiterate
woman. This assumes that the second signature in the left column
is the one who wrote the womans name hrmm, at her command.
25
Below the name of the witness who signed for her, there were three
witnesses, with the words d[, then twice dh after each name.
Xev/Se 9Deed of Sale (2nd half of 1st century CE)
26
On the verso of this double deed there are preserved remnants
of seven signatures. The rst signature (occupying two lines appar-
ently) is that of the seller who was also the scribe, followed by
[hbtk h]pn l[[]. Then the signature of the buyer must have orig-
inally appeared. Then there are remnants of ve witnesses signa-
tures, with the last two each preserving dh.
27
Xev/Se 10Receipt (1st century CE?)
28
This text seems to have a signature of the scribe only, with hbtk,
indicating that he wrote it.
Xev/Se 13Waiver of Claims (Receipt for Ketubbah) (134 or 135 CE)
This is a simple deed with signatures on the recto.
29
This document
is signed rst by the woman who asserts that she has no claims
against her former husband from whom she received a divorce writ
(fg). Her signature is followed by btk hla hpn l[, for herself,
she borrowed the writing.
30
The scribes name comes next. He had
signed for the woman and now signs hrmm, at her word. hpn l[
indicates that she was present when the document was executed and
that she agreed with its contents. btk hla indicates her use of
the service of the scribe to sign for her, according to her hrmm,
25
On this usage, see Cotton, DJD 27.295.
26
Yardeni, Textbook, B.[21].
27
Yardeni, DJD 27.38, 5051; cf. Cotton, DJD 27.143. See Yardenis list of doc-
uments with seven signatures, 51 and n. 62.
28
Yardeni, Textbook, B.[36].
29
For the interpretation of the text, see H.M. Cotton and E. Qimron, Xev/Se
ar 13 of 134 or 135 C.E.: A Wifes Renunciation of Claims, JJS 49 (1998) 10818.
30
Not present tense as in Yardeni, DJD 27.67.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 171
172 r.vnrxcr n. scnirrv.x
instruction. This process was used since she was illiterate. Then
follow two signatures of witnesses with the word d[.
31
Xev/Se 22Deed of Sale (late 1st or early 2nd century CE)
32
On the verso of this double deed, there are seven signatures. The
last four are marked dh, witness. We can assume that the rst
two signatures were of the seller and of the buyer or scribe. A wit-
ness must have been the third, making a total of ve witnesses.
33
Xev/Se 49Promissory Note (133 CE)
This is a Hebrew document with some admixture of Aramaic. It
seems to be a simple document. At the end the borrower states that
he will sign in his own handwriting below. Then there appears the
name of the borrower who makes out the note followed by h[pn] l[
hybtk. Then there are three witnesses, with d[, then dh, then d[.
Xev/Se 50 + Mur 26Deed of Sale (1st or early 2nd century CE)
34
This is a double document with signatures on the verso. The rst
signature, in Greek, is of the seller who testies, also in Greek, that
he signed in his own hand.
35
The second signature, in Aramaic, is
probably of the sellers wife whose signature was hrmm, at her instruc-
tion, that is, signed for her. Again, we would have to take btk as
going with the following signature, that of the man who signed on
her behalf. A clause indicating her agreement and that she had no
claims against her husband must have been in the document.
36
There
then follow the signatures (on four lines) of a witness (dh), the scribe,
31
Cf. Yardeni, DJD 27.70.
32
Yardeni, Textbook, B.[26].
33
Cotton, DJD 27.143. On the verso of the double deed, Xev/Se 24, there
are two names which cannot be classied. Remnants of four unclassied signatures
are on the verso of Xev/Se 25, a double deed. Apparently ve signatures sur-
vive on the recto of Xev/Se 26, a text dealing with deposits and barley, but there
may be remains of signatures on the verso (Yardeni, DJD 27.95).
34
Yardeni, Textbook, B.[24].
35
Yardeni, DJD 27.129.
36
Lines 1819, Yardeni, DJD 27.124.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 172
nrnnrv .xr .n.v.ic roctvrx+s 173
and a nal witness (dh), so that there are two witnesses besides the
scribe.
37
Thus, there were a total of six signatures.
P.Yadin 7Deed of Gift (120 CE)
This double document is signed on the verso by the donor who states,
hbtk hpn l[. He was therefore present and actually signed the doc-
ument.
38
There are a total of seven signatures including this one.
The second, third, and fourth are of witnesses (dh), the other three
signatures, although fragmentary, are probably also witnesses, total-
ing six witnesses. The last witness may be the scribe.
39
It may be
assumed that the wife who receives the gift (upon her husbands death)
did not have to sign as she incurred no obligation by this document.
P.Yadin 8Purchase Contract (122 CE)
This is a contract for the purchase of animals. There are four lines
of signatures. The rst line contains three names. The rst is that
of the buyer. Then there appears the signature of an unknown per-
son followed by hbtk. It cannot be the seller since his name appears
in the document (he is the buyers brother). The third signature is
that of the scribe. A superlinear title, perhaps h[rps], identies him
a scribe. Then there are two Jewish script signatures followed by
dh, and a Greek signature, either a witness or a Roman provin-
cial administrator.
40
P.Yadin 10Babathas Ketubbah (between 122 and 125 CE)
41
This double document preserves remains of seven signatures on the
verso. The rst is that of the groom who actually was the scribe of
37
Yardeni, DJD 27.126.
38
It is doubtful that he wrote it as suggested by K. Beyer, Die aramischen Texte
vom Toten Meer: Ergnzungsband: samt den Inschriften aus Palstina, dem Testament Levis aus
der Kairoer Genisa, der Fastenrolle und den alten talmudischen Zitaten: aramistische Einleitung,
Text, bersetzung, Deutung, Grammatik/Wrterbuch, deutsch-aramische Wortliste, Register
(Gottingen: Vandenhoeck & Ruprecht, 1994) 173. btk here means that he signed
his own name.
39
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 75.
40
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 10811, 11617.
41
Yardeni, Textbook, B.[56].
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 173
174 r.vnrxcr n. scnirrv.x
the document, followed by hbtk h[pn l[], indicating that he actu-
ally wrote it and accepts its conditions. The next signature is that
of the bride, but as she was illiterate someone must have signed for
her. Her signature is followed by hpn l[. The signature after hers
is that of the person who signed for her, followed by [h]rmm, at
her command. There then follow the three witnesses whose names
were followed by dh, witness.
42
P.Yadin 42Lease Agreement (132 CE)
In this simple document two of Bar Kokhbas administrators grant
a lease. There are only two signatures here, those of the adminis-
trators. Actually, it seems that the second also signed for the rst.
The rst name is followed by hbtk h[pn] l[ and the second only
by hb[tk], but the document makes clear that both are equally grant-
ing the lease. Apparently, such administrative leases did not require
witnesses.
43
P.Yadin 43Receipt (132 CE)
This receipt (for the transaction in P.Yadin 42) has two signatures,
one being one of the two administrators, and the other not pre-
served. The preserved name is followed by hbtk. (The signature on
the verso, line 10, is from another document.)
44
P.Yadin 44Distribution of Leased Land (135 CE)
This is a Hebrew document of distribution of leased parcels of land
among two sets of partners. It is in the form of a simple deed. The
deed is followed by seven lines of signatures. The rst four lines are
for the signatures of the four partners dividing the leased property.
Then there follow signatures of three witnesses. The rst lessee signed
his own name, followed by wpn l[, indicating his assent to the con-
tents. The names of the second, third, and fourth lessees were signed
by others, but still the formula wpn l[ appears after each name.
42
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 121, superseding
Y. Yadin, J.C. Greeneld and A. Yardeni, Babathas Ketubba, IEJ 44 (1994)
7677 and Cotton, DJD 27.143.
43
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 14243, 149.
44
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 15051, 155.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 174
nrnnrv .xr .n.v.ic roctvrx+s 175
The three principals who did not sign their own names are followed,
on the same line, with signatures of those who signed for them, each
preceded by btk. They are followed, in turn, by wnwxr[m], of his
free will, wrmam and hrmm, at his command or instruction. The
writer of the fourth signature was the scribe of the document. Then
come the names of three witnesses, each followed by d[.
45
P.Yadin 45Redistribution of Leased Land (135 CE)
In this Hebrew simple deed, one of the members of the two sets of
lessees in P.Yadin 44 cedes his rights to his partner. This document
has four signatures. The rst, the partner who leases his share to
the others, signs, followed by [h]pn l[. Then follow three witnesses
whose names are followed by d[.
46
P.Yadin 46Lease of Redistributed Land (135 CE)
A simple Hebrew document in which both partners lease the same
property referred to in P.Yadin 44 and 45 to a third party. The
rst signature, on the same line as the end of the document, is the
lessee who signs hpn[ = hpn l[, indicating his assent to the doc-
ument. Three witnesses then follow and their names are followed by
d(y)[.
47
Jer 2Deed of Sale or Lease (1st or early 2nd century CE)
48
This is a simple document which probably represents the obligation
of the unpaid quarter of a lease agreement for a eld. There were
four signatures, each followed by hbtk.
49
The editors assume that
all four signatories are witnesses, but this is unlikely in a simple doc-
ument. Most probably the rst two are the parties, the lessor and
the lessee, and the last two are the witnesses. Since no names can
be read at all in the document, we cannot be certain.
45
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 42, 5253.
46
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 55.
47
Yadin et al., Hebrew, Aramaic and Nabatean-Aramaic Papyri, 65.
48
Yardeni, Textbook, B.[40].
49
E. Eshel and H. Eshel, Jericho papDeed of Sale ar, in J. Charlesworth
et al., Miscellaneous Texts from the Judaean Desert (DJD 38; Oxford: Clarendon Press,
2000) 31, 36.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 175
176 r.vnrxcr n. scnirrv.x
Jer 3Deed of Sale (1st or 2nd century CE)
50
This is a double document since the signatures appear on the back.
Remains of these names appear on the verso. A space between the
second and third signatures seems to indicate that a signature may
have originally stood there, giving a total of four signatures.
51
The
fragmentary state of the text makes it impossible to tell if all four
signatories are witnesses or if, most likely, the rst two are the prin-
cipals to the contract.
Jer 7Deed of Sale (84 CE)
This is a double document of purchase of a date crop, which was
in the possession of orphans. It seems that there were originally ve
signatures. The rst is that of the seller, as usual, in this case the
guardian of the orphans.
52
After his name appears [h]pn [l[], indi-
cating that he conrms the contents of the document. The second
signature, not preserved, was probably that of the buyer, followed
by three witnesses whose names were most probably followed by
d[ or dh.
Jer 9Deed (1st century CE)
53
This is a very poorly preserved double document. The remains of
the signatures cannot be read, except for one. There is evidence of
three signatures.
54
We would presume that there were originally at
least four signatures, the seller, the buyer, and two witnesses.
Sdeir 2Promissory Note? (135 CE)
This is apparently a promissory note, and, like other promissory
notes from this period, would have required two witnesses.
55
Lines
78 expressly refer to writing the document and its being signed
50
Yardeni, Textbook, B.[40].
51
Eshel and Eshel, DJD 38.36.
52
E. Eshel, H. Eshel, and H. Misgav, Jericho papSale of Date Crop ar, DJD
38.61.
53
A. Yardeni, Jericho papDeed A heb?, DJD 38.67.
54
Yardeni, DJD 38.69.
55
A. Yardeni, Sdeir papPromissory Note? ar, DJD 38.12526.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 176
nrnnrv .xr .n.v.ic roctvrx+s 177
inside, meaning that it is a simple document, not a double one.
56
There are four signatures. The rst is that of the one who took on
the nancial obligation.
57
His name is followed by hpn l[, indicat-
ing his acceptance of the obligation. Then there appears the signa-
ture of the scribe who had signed for him, followed by hrmm, at
his verbal order or command. Then there are two witnesses whose
names were probably followed by the word d[ or dh.
4Q344Debt Acknowledgment (1st century CE?)
58
This promissory note was a double document.
59
There are three sig-
natories. The rst is the person taking on the nancial obligation.
60
Then followed two witnesses, most probably followed by d[ or dh.
4Q345Deed (1st century CE?)
61
This is a double deed that has two signatures at the end. The rst
is one of the parties to the agreement, followed by [hbtk h]pn l[,
indicating that he accepted responsibility for the contents of the deed.
The second signature is probably the other party, followed by hrmm,
indicating that he could not sign for himself. In this case, it appears
that both signatures were written by the scribe of the document.
62
If we are correct, the signatures of the two witnesses would have to
appear below the two preserved signatures.
IV. Comparison with Rabbinic Law
From the examination of the evidence above, a number of features
emerge that need to be compared with the practices mandated or
56
Yardeni, DJD 38.129.
57
See line 3.
58
Yardeni, Textbook, B.[20]. H. Eshel, Hebrew in Economic Documents from
the Judean Desert (Hebrew), Le 63 (20002001) 4344 argues for the Qumran
provenance and pre-70 cr dating of the economic documents in the 4Q collec-
tion published in DJD 27.
59
Cf. Yardeni, DJD 27.289.
60
See line 2.
61
Yardeni, Textbook, B.[40].
62
Yardeni, DJD 27.295.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 177
178 r.vnrxcr n. scnirrv.x
described in tannaitic sources and Talmudic literature. We will see
that in the area of signatures and witnesses the documents before
us are greatly at variance with the Rabbinic texts, even though the
legal formulae of the contracts are often extremely similar.
1. Signature by Parties to Contract
Tannaitic sources testify to the currency of both simple and tied
deeds. These two types of contracts were described as diering in
that the simple deed had its witnesses signatures inside the text
and the tied deed, on the back. The sages disputed over what
happens if the practice were reversed and the signatures were incor-
rectly located. M. B.B. 10:1 presents the majority view as disqua-
lifying such deeds and a minority, under certain circumstances,
accepting a tied deed with the witnesses on the inside (cf. t. B.B.
11:1; t. Gi. 6(8):9.
A fundamental feature of the documents from the Judean Desert
is the presence of signatures of the two parties to the transaction,
as, for example, the groom and bride, or the seller and buyer. This
is in contradistinction to the usual form of contracts known from
the medieval halakhic tradition, itself derived from the Babylonian
Talmudic prescriptions and from the customs of the Jews of Babylonia.
In such documents, only the witnesses sign.
It is important to remember that the question of who signs a
legal document is intimately linked with its overall diction. If the
text describes in the third person (objective) a transaction or an act
of formal acquisition (yynq), then the text will sensibly require only
the signatures of the witnesses who attest to the validity of the trans-
action described in the document. On the other hand, as is the case
with the documents from the Judean Desert, if the speaker in the
document is the seller or groom (subjective), for example, and the
document is itself his assertion of undertaking specic obligations,
then his signature (and often that of the other party) will be required
to validate the document.
63
63
Cf. A. Gulak, Legal Documents in the Talmud in Light of Greek Papyri and Greek and
Roman Law (Hebrew) (ed. and suppl. by R. Katzo; Jerusalem: Magnes Press, The
Hebrew University, 1994) 4044.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 178
nrnnrv .xr .n.v.ic roctvrx+s 179
To put it another way: a contract phrased in the third person is
a description of a transaction eected by an action of some kind
(yynq), whereas in a rst-person document, such as those from the
Judean Desert, the execution of the document eects the transac-
tion, and, hence, it must be signed as subscribed to by the parties.
Further, the function of the witnesses may dier also. In the third-
person contracts, they serve to record a transaction already under-
taken in their presence. In the rst-person documents, the witnesses
validate the document, which itself eects the transfer or obligation
described therein.
64
Rabbinic contracts that follow Babylonian Jewish usage are third-
person texts. Hence they required no signatures besides the witnesses.
Our texts are rst-person and so do require the signatures of the
parties. But note that a rst-person text, like the largest part of the
traditional Jewish marriage contract, can be rendered third-person
with a short introduction saying, The groom said to the bride. . . .
Such adaptations allowed later halakhah to avoid all signatures except
those of the witnesses, although later on there did arise a custom in
some localities to have the groom sign the marriage contract. But
we must also note that an essentially third-person contract text can
be turned into a rst-person text by adding a clause at the end in
which one or both parties undertake an obligation. Such clauses are
prominent in the Judean Desert contracts, but these are rst-person
documents.
It is therefore natural that our texts, phrased as they are, require
signatures of the parties. In regard to marriage contracts, this issue
was dealt with in detail,
65
and a survey of the evidence will help us
to clarify matters.
It was generally agreed that the consent of both parties to the
marriage contract had to be stated. In Palestinian ketubbah texts
from the Middle Ages, this even meant that the formal act of acqui-
sition included both the groom and bride. The statement of m. B.B.
10:4 was taken to mean that betrothal and ketubbah documents may
64
Cf. Gulak, Legal Documents in the Talmud, 2632.
65
M.A. Friedman, Jewish Marriage in Palestine: A Cairo Geniza Study (Tel Aviv and
New York: Tel-Aviv University, The Chaim Rosenberg School of Jewish Studies,
Moreshet Project for the Study of Eastern Jewry and The Jewish Theological
Seminary of America, 1980) 1.46385. This and the following paragraph are a sum-
mary of Friedmans study.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 179
180 r.vnrxcr n. scnirrv.x
not be written except with the consent of both parties. Nevertheless,
some Babylonian Amoraim accepted documents that did not indi-
cate the brides consent, but this was not the case in Palestinian tra-
dition, where bride and groom had to give explicit instructions. T.
Gi. 2:8
66
required that the witnesses hear the groom commanding
the scribe to write and the witnesses to sign. Such clauses are known
from Babylonian Jewish formulary for deeds of conveyance (b. Ket.
55a and Gaonic sources).
67
Such Babylonian clauses were not employed
in the Palestinian ketubbot from the Cairo Genizah. What these
texts have is rather a declaration at the end by the groom indicat-
ing his obligation, similar to clauses found in the Judean Desert texts.
In the texts from the Judean Desert, hpn l[ indicates that he is
taking the obligation on himself in conrming the validity of the
document.
Such subscription formulae, found in our Judean Desert texts, do
not appear anywhere in Rabbinic literature in relation to marriage
contracts but do appear for other documents.
68
Apparently, as we
see in our documents, they were common in the usage of Jews in
Late Antiquity. The witnesses signed below this subscription of the
party who was eecting the transfer. In the Genizah, as in our texts,
such a subscription is found in some marriage contracts.
69
Accordingly, the practice in our documents, in which the groom
or seller signs before the witnesses, may be in accord with usage
described in Rabbinic literature for the Land of Israel. Specically,
in rst-person formulations, such as are found in our documents, the
grooms signature simply came at the end, but if the text was third-
person, a subscription with the signature of the groom followed the
signatures of the witnesses.
66
S. Lieberman, Tosefta: Nashim (New York: Jewish Theological Seminary of
America, 1973) 251; S. Lieberman (ed.), Tosefta Ki-Fshuah, Seder Nashim (New York:
Jewish Theological Seminary of America, 1973) 8.81012.
67
Friedman, Jewish Marriage in Palestine, 471 n. 82.
68
T. B.M. 1:13; y. Gi. 8:12 (49b); y. B.B. 10:1 (17c).
69
The practice of some medieval authorities to require the grooms signature
may be a remnant of this procedure. See Friedman, Jewish Marriage in Palestine,
1.483; S.E. Stern, Seder Ketubah ke-Hilkhatah (Bene Berak: Makhon le-Meqar Torani
Naalat Yeezqel, 1995/6) 4041; B. Adler, Sefer ha-Nisu"in ke-Hilkhatam (Hilkhot
ve-Halikhot be-Yahadut; Jerusalem: Hamesorah, 1985
2
) 1.334.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 180
nrnnrv .xr .n.v.ic roctvrx+s 181
The signatures of the buyer or the bride, however, do not seem
to have any Rabbinic parallel. It is clear that such signatures, in the
case of the bride, solve the problem of guaranteeing her assent to
the proceedings. But no such practice is documented in our sources.
Since in our Judean Desert texts the buyers also seem to sign (not
just the brides), it seems that we must look elsewhere for the expla-
nation of this phenomenon. It is certainly not in accord with Rabbinic
usage.
70
2. Signature of Wife
In some documents we nd that when a husband is selling land, his
wife has to make a declaration in the contract and sign to the eect
that she has no claims against her husband. Womens signatures also
appear in the case of a receipt for a divorce document where it is
indicated that the woman has received all she is owed, what is called
a rbw in Rabbinic usage.
71
Taken together, in all the documents
from the Judean Desert, womens signatures, or at least their names
signed for them, are quite common.
From the point of view of comparison to Rabbinic usage, there
simply is no parallel to the wifes subscription. In Talmudic law, as
realized in later documents from the Genizah, only witnesses sign
legal documents. We have seen that sellers may also sign. The prac-
tice followed in our documents is based on the requirement of sub-
scriptions in Hellenistic law, a policy not part of Jewish law and
obviated in Rabbinic practice by the third-person formulation and
the role of the witnesses.
3. Number of Witnesses
Above we surveyed the number of witnesses in the various texts.
They are summarized in the following table:
70
Cf. Yadin et al., Babathas Ketubba, 77 n. 7 in which they cannot explain
why she signs.
71
Cf. Gulak, Legal Documents in the Talmud, 17577.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 181
182 r.vnrxcr n. scnirrv.x
Text Simple or Number of Number of Signatures of:
Double Signatures Witnesses
Mur 18 Double 4 (on verso) 2 Debtor, Scribe, W, W.
Mur 19 Double 4 3 Husband, Scribe, W, W.
Mur 20 Double 1 ? Groom
Mur 21 Double 7 4(5) Groom, Scribe, Bride, W,
W, W, W.
Mur 22 Double 7 4(5) [Seller (?)], [Buyer (?)],
[W], Scribe, W, W, W.
Mur 24 Simple 2 Simeon ben Kosiba, Lessee
(11 contracts)
Mur 27 Simple 1 ? Seller
Mur 28 Double 2 ? Seller, Guarantor (?)
Mur 29 Double 5 3 Seller (Greek), Wife, W,
W, W.
Mur 30 Double 5 2(3) Seller, Wife, Scribe, W, W
Xev/ Simple 6 3(4) Wife, Substitute, Seller, W,
Se 8a W, W.
Xev/Se 9 Double 7 5 Seller (scribe), [Buyer], W,
W, W, W, W.
Xev/Se 10 Simple 1 1 Scribe
Xev/Se 13 Simple 4 2 Woman, Scribe (substitute),
W, W.
Xev/Se 22 Double 7 5 Seller (?), Buyer (?), W (?)
Xev/Se 49 Simple 4 3 Borrower, W, W, W.
Xev/Se 50 Double 6 3 Seller (Greek), Wife,
+ Mur 26 Substitute, Scribe, W, W.
P.Yadin 7 Double 7 6 Donor, W, W, W (?), W
(?), W (?), W (?).
P.Yadin 8 Simple 6 2(3) Buyer, ?, Scribe, W, W, W (?)
P.Yadin 10 Double 7 5 Groom, Bride, W, W, W,
W (?), W (?).
P.Yadin 42 Simple 2 0 Administrators
P.Yadin 43 Simple 2 0 Administrator,
Administrator (?)
P.Yadin 44 Simple 7 3 4 Partners, W, W, W.
P.Yadin 45 Simple 4 3 Partner, W, W, W.
P.Yadin 46 Simple 4 3 Lessee, W, W, W.
Jer 2 Simple 4 2(4) Lessor (?), Lessee (?), W, W.
Jer 3 Double 4 2(4) 2 Principals (?), W, W.
Jer 7 Double 5 3(4) Seller, Buyer (?), W, W, W.
Jer 9 Double 4(?) 2(?) Seller, Buyer (?), W, W.
Sdeir 2 Simple 4 2 Obligatee, Scribe, W, W.
4Q344 Double 3 2 Obligatee, W, W.
4Q345 Double 4(?) 2(?) Seller, Buyer (?), W, W.
Talmudic law requires at least two witnesses for a simple deed, and
three for a tied deed. Any less for either type was disqualied (m.
B.B. 10:2). A tied deed had to have a signature for each knot. One
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 182
nrnnrv .xr .n.v.ic roctvrx+s 183
that had a knot without a signature was termed by the Tannaim a
bald document (jrq fg) and was considered not to be valid (m.
Gi. 8:9; cf. t. Gi. 6(8):9). But in the case in which a tied deed was
being executed, it was permissible according to some for relatives,
or even disqualied witnesses, to sign after two valid witnesses (m.
Gi. 8:10).
Interesting for our purpose is the denition given for a bald doc-
ument (one with insucient witnesses) in t. Gi. 6(8):9. It is described
as having seven knots and only six witnesses, six and ve, ve and
four, four and three, three and two. Signatures even of relatives may
be added to ll the missing slots.
72
From this passage it seems that
only one of the signatories may be a relative who would not nor-
mally be permitted to testify.
These passages certainly testify to the practice of various numbers
of signatories as found in the texts from the Judean Desert. S.
Lieberman, however, took the view, because the number seven was
mentioned rst, that these contracts followed the Roman method
which required seven,
73
rather than a lower number, such as six
which was the Ptolemaic practice or ve, which is found in many
of our texts from the Judean Desert and elsewhere in the Roman
Near East.
74
It seems, however, that this Tosefta passage simply rec-
ognizes that seven was the maximum in practice, and that various
lower numbers occurred also, three being the minimum allowed. It
cannot be taken to favor the Roman practice.
Tannaitic sources attest to the fact that one (according to the
Tosefta) or more (according to the Mishnah) of the signatories may
not be actual legal witnesses. But whereas the deeds from the Judean
Desert use the parties and the scribe to ll knots, the tannaitic sources
expect family members, apparently present at the execution of the
document, to add their signatures. The Tannaim have absolutely no
expectation, as these tannaitic texts prove, of signatures by the par-
ties (seller, buyer, bride or groom) but only by others. Yet for the
Judean Desert texts, relatives occur in subscriptions to Greek texts,
but not as signatories to Aramaic and Hebrew contracts.
72
On this text, see Lieberman, Tosefta Ki-Fshuah, Seder Nashim, 8.899900 whose
restoration we follow.
73
Roman wills and testaments required seven witnesses (Cotton, Scribes, Nota-
ries . . ., 6).
74
Cotton, DJD 27.14142.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 183
184 r.vnrxcr n. scnirrv.x
Mur 29 and 30 specically mention in the document the signa-
tories, ymtwj, who do not in fact all appear at the bottom of the
document. This clause apparently was in Babathas ketubbah and
tended to indicate that those who signed were actually present at
the execution of the document. Presence of the witnesses is of course
required in Rabbinic law as well and often is indicated in the lan-
guage of post-Talmudic contracts.
75
4. Signatures of Illiterates
It is well known that the texts from the Judean Desert allow others
to write the names of illiterate witnesses or parties to the contract
at their direction. Various formulae in the Aramaic and Greek doc-
uments are used to indicate this method of substitute signature; nec-
essary in light of the lack of literacy among many of the Jews of
Late Antiquity, including even Babatha herself. Other witnesses might
be able to write in Hebrew/Aramaic, but not in Greek, or the
reverse.
76
Talmudic sources proposed two solutions to this very problem.
One is to create a stencil with the name of the witness that he could
then ll in with ink. The other was to write his name on the doc-
ument in a light or clear liquid and he would then trace his signa-
ture over it.
77
These techniques were accepted by the Rabbis only
in order to enable a woman to receive a divorce document. The
Geonim created an alternative procedure for allowing testimony by
deposition in court, and the court would then write a document to
record the testimony. In general, deposition of witnesses was a devel-
opment in medieval halakhic practice, and did not exist in Late
Antiquity.
78
These procedures are in marked contrast to those noted in the
Judean Desert documents. The use of terms like hrmam or btk hla
to indicate substitute signatures, known from Greek papyri and Dura
Europos texts as well as our texts, was totally unknown to the Rabbis
75
S. Assaf, The Book of Shetaroth (Formulary) of R. Hai Gaon (Hebrew) (Supplement
to Tarbiz I, 3. Jerusalem: Azriel Press, 1930) 20 (loan document), 22 (writ of quit-
tance), 24 (deed of sale of houses) and 26 (elds), and passim.
76
See Cotton, DJD 27.14445.
77
See Friedman, Jewish Marriage in Palestine, 1.485.
78
Jacob ben Asher, ur oshen Mishpa 28 and J(oseph) Caro, Bet Yosef,
ad loc.
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 184
nrnnrv .xr .n.v.ic roctvrx+s 185
79
Friedman, Jewish Marriage in Palestine, 1.48588.
80
For citations of this widespread text, see Friedman, Jewish Marriage in Palestine,
1.491 n. 165.
81
On the language of signatures in tannaitic law, see m. Gi. 9:6, 8 (. Albeck,
Shishah Sidre Mishnah, Seder Nashim [ Jerusalem: Mosad Bialik; Tel Aviv: Dvir, 1954])
4067; t. Gi. 7(9):11; Hezser, Jewish Literacy, 18081, 329.
or rejected by them. For them, only the individual himself could
sign his name. But a number of Palestinian Genizah marriage con-
tracts do follow the practice of substitute signatures, in one case fol-
lowed by wyp rmamb, by his instructions, in others by an abbreviation
for wyp l[ wtwd[ hbtkn, his testimony was written by his dictation,
and elsewhere with no indication. This practice is known from other
Genizah texts as well, but Rabbinic authorities in the Middle Ages
condemned this practice.
79
What is clear is that the practice found in our documents is essen-
tially a Greco-Roman practice that was being used by Jews even for
contracts written in Hebrew and Aramaic, which in many details
accorded with Rabbinic law. But in this aspect, they diverged from
the practices of the Rabbis who did not accept this solution, believ-
ing that all witnesses had to sign. Against the view of medieval legal
authorities and Talmudic law before it, this practice of substitute sig-
natures persisted in Jewish practice in some segments of the com-
munity that continued to follow ancient Palestinian Jewish usage.
5. Signature of the Scribe
In a number of our texts, the scribe signs along with the witnesses,
sometimes identied as the scribe, other times known to be the scribe
only by his handwriting. In some ancient Near Eastern texts the
scribe also signed. Some Genizah documents indicate that the scribe
copied the document and served as a witness. This same practice
underlies a midrash attributed to the third-century Palestinian Amora
R. Yoanan, indicating that it was a widespread practice in this
period.
80
In the notarial practice of Late Antiquity it must have been
very convenient to simply have the scribe serve as a witness, espe-
cially in light of the illiteracy problem described above, which for
the Rabbis posed a complex problem. Thus, in this case, the prac-
tice of our documents accorded with Rabbinic law and usage.
81
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 185
186 r.vnrxcr n. scnirrv.x
V. Conclusion
This investigation demonstrates that the comparison of the proce-
dures for signing and witnessing contracts from the Judean Desert
to those described in Rabbinic law is a complex task. Many of the
practices in use here do follow what Rabbinic law prescribes. Yet
others dier markedly.
82
How can this apparent inconsistency be explained? Mishnaic prac-
tice should not be taken as a set of prescriptions derived from Jewish
law. Rather, they are a reection of the customary legal practices,
derived from both ancient Near Eastern and Greco-Roman sources,
as they were practiced in Palestine by Rabbinic courts. Many of
these were eventually recorded in the Mishnah, especially if they
were usages that required adjudication in the courts. Other Jews no
doubt followed similar but dierent practices, also accommodating
the Torahs values, as they understood them, to the legal process
of the day. To be sure, those Jews living in the Dead Sea region,
whether in Judea or Arabia, lived at a crossroads of cultures, and
it is natural to expect them to have assimilated a substantial amount
of Greco-Roman legal usage. Further, tannaitic law was constantly
developing, and the collections of Mishnah and Tosefta were only
later brought to completion. In many cases, our contracts may reect
earlier stages in the history of Jewish law than the redacted texts of
tannaitic tradition.
Put otherwise, good Jews in our region executed the contracts
we have studied and witnessed them. Like all Jews of Late Antiquity,
they lived as Jews, an ancient Semitic people, in a Greco-Roman
context. No wonder that their documents show thiseven their sig-
natures.
82
Cf. Cotton, DJD 27.15357; Z. Safrai, Shemirat ha-Halakhah be-Mismekhe
Midbar Yehudah, iddushim be-eqer Mered Bar-Kokhba" (ed. H. Eshel, B. Zissu;
Ramat Gan: Bar-Ilan University, 2001) 11337; L.H. Schiman, Reections on
the Deeds of Sale from the Judean Desert in Light of Rabbinic Literature, Bar
Ilan University Conference Volume, Bar Ilan Univ. Press (forthcoming).
SHIFFMAN_f9_165-186 12/11/02 9:07 AM Page 186
THE ROMAN NEAR EAST: THE VIEW FROM BELOW
Frank E. Peters
The Roman Empire was an enormous political organism bounded
only by the impertinent Scots and Irish on the west, the ungovern-
able Berbers on the south, the indomitable Iranians on the east, and
whatever barbarians were currently making mischief on the far bank
of the Danube. It was large, it was impressive, and it was organized.
We know a great deal about its organization because we possess a
substantial body of the Roman republican and imperial law that gov-
erned it and an even more substantial body of discourse ranging
from analysis to chatter, from Livy and Tacitus to Cicero and Pliny,
on how the system operated, at least at its highest levels. As often
in antiquity, the higher we mount, the clearer the vision; it is only
when we turn downwards from the macrocosm beloved of Stoic
philosophers and Roman historians alike that we begin to encounter
problems of understanding. The view from below is random, scat-
tered, and occasional; there are sudden brightly-lit corners surrounded
by vast areas of darkness, and we do not know how far we are
justied in extrapolating that light into the obscurity that surrounds
it. Is what was found in this cave in Palestine or scratched on that
wall in Britain typical or eccentric, an everyday occurrence or a
hapax drmenon?
The papers collected here are about this latter microcosmic per-
spective, a stroll in the sudden aleatory light that shone forth from
remote areas around the Dead Sea. Like the earlier, equally fortu-
itous discoveries from another edge of the Dead Sea, these papyri
unexpectedly pull back the curtain not only on events and person-
alities about which we had some previous information, the Essenes
there, Bar Kokhba here, but also on the lives and dealings of indi-
viduals who have lain mute and unidentied for nearly two millen-
nia. We have burst into their lives, inspected their possessions, listened
to their problems.
I am not one of the voyeurs in the present enterprise, but I have
tapped phones in other places in the Roman Near East and on other
occasions in the history of that empire, and I shall use some of what
187
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 187
188 rn.xk r. rr+rns
1
H. Waddington, Inscriptions grecques et latines de Syrie (Paris: Didot, 1870).
2
H.C. Butler, Syria: Publications of the Princeton University Archaeological Expedition to
Syria in 19041905 and 1909, vol. 2: Architecture. Section A: Southern Syria (Leiden:
Brill, 1919).
3
R. Bruennow and A. von Domaszewski, Die Provincia Arabia (Strassburg: Tuebner,
19049).
I observed then and there to broaden somewhat the context of the
present discussion.
As I have already suggested, the larger picture of the Roman Near
East is generally constructed out of literary sources as varied as
Josephus in the rst Christian century to Ammianus Marcellinus in
the fourth and Procopius and John Malalas in the sixth century.
Diverse documents like the Notitia Dignitatum and the Peutinger
Table help ll in the picture, as do later ecclesiastical notitiae and
the acts of church synods and councils. But for the details, or at
least those that occur below the sight-lines of the imperial and church
historians, we must generally look elsewhere, to the remains of mate-
rial culture, architectural, epigraphical, and papyrological. These are
scattered all across the empire from Britain to the Danube and
Euphrates frontiers and the remote outposts that guarded the ap-
proaches from Inner Arabia. Our present concern is the Near East,
however, and I shall conne my remarks to that area.
European travelers have tramped across the Near and Middle East
for centuries and usually with a watchful eye upon the terrain. But
they were mostly pilgrims and they were looking for something quite
specic out of the past, sites connected with well-known personages
and events out of their own sacred history, whether biblical or New
Testament, and they often passed the Greek and Roman antiqui-
ties, most of them far more obvious to the eye then than they are
now, with scarcely a glance.
The search for these remnants of the secular past began only in
the nineteenth century, and a crucial date for our purposes is 1870,
which marked the publication of H. Waddingtons rst comprehen-
sive collection of the Greek and Latin inscriptions of Syria,
1
followed
in the early years of the next century by two comprehensive surveys
of the surviving monuments, site by site, with additional epigraphi-
cal material that had been turned up by the survey: rst, the Princeton
survey of Roman Syria under the direction of H.C. Butler,
2
and the
second by R. Brnnow and A. von Domaszewski of Roman Arabia.
3
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 188
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 189
Thus, by 1910 scholars had at their disposal a considerable canon
of site- and time-tied inscriptions and a repertoire of buildings, pub-
lic and private, in and around which this evidence occurred. What
was chiey remarkable about this evidence, particularly that from
Syria, is that so much of it was microcosmic, that is, it had to do
with small-scale buildings and village culture. It is little wonder that
something as monumental and well-wrought as the Acropolis still
stands, or the baths of Caracalla or Herods Temple platform, but
in present-day Syria there are still private houses from the Roman
era, a great many of them in fact, and some indeed are still inhab-
ited. There are two areas of Roman Syria, the Hawran, the south-
ern lava lands that stretch almost from the gates of Damascus
southward across the Jordanian border and, in the north, the region
surveyed by G. Tchalenko that lay east of Antioch and north of
Aleppo,
4
where chance, climate, and an almost indestructible build-
ing material has preserved unrivaled glimpses into Roman private
life and village culture in the early and the late Roman Empire.
The eects of this collective evidence were not long in appearing.
In 1912, W.K. Prentice, the editor of the Greek and Latin inscrip-
tions found during the Princeton Expedition to Syria, published an
article on ocials responsible for public works in Roman Syria.
5
Six
such are named in the preserved inscriptions attached to the public
buildings of Syria, either in their planning (prnoia) or their execu-
tion (spoud) stage. Here, for example, is what Prentice says with
admirable precision of those called pisto:
From these inscriptions it appears that the pistoi were public ocials,
sometimes ve or more in number. They existed in certain parts of
southern Syria, particularly in Nabatea and Batanea, in the fourth cen-
tury ..r., perhaps for a longer time. They had charge of certain pub-
lic works, civil and religious. They were high ocials for they had the
pronoia (the planning responsibility), with or without spoud (supervision
of the actual construction). At least in some instances they did not
provide the funds . . . I believe that the pistoi constituted in certain
Syrian towns in the fourth century of our era the highest executive
board of the local community . . .
6
4
G. Tchalenko, Villages antiques de la Syrie du Nord: Le Massif du Belus lpoque
romain (Paris: Paul Geuthner, 195558).
5
W.K. Prentice, Ocials Charged with the Conduct of Public Works in Roman
and Byzantine Syria, Transactions and Proceedings of the American Philological Association
43 (1912) 11323.
6
Prentice, Ocials, 118.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 189
190 rn.xk r. rr+rns
Thus a door is opened onto a level of municipal administration
located far below that of the central authority; the inscriptions re-
viewed by Prentice allow us to see, or rather to glimpse, albeit in static
form and only in outline, local ocials at work in a local commu-
nity. But Prentice was well aware of how narrow were the limits of
that portrait:
This investigation has been conned to a limited number of inscrip-
tions, namely the Greek inscriptions of Syria. Its conclusions therefore
should be regarded as somewhat tentative and preliminary. A com-
plete investigation must include not only the Syrian inscriptions in
other languages, especially in Latin[ Jalabert and Mouterde under-
took the publication of all the Greek and Latin inscriptions of Syria
beginning in 1929]but also all that can be discovered about such
matters in other provinces. Such an inquiry will lead to a more com-
plete understanding of the local organization and administration of
towns in the Roman and Byzantine empires . . .
7
Another, more complex line of approach on the local level of activ-
ity in the Roman Empire was undertaken by G.M. Harper in his
1928 article on Village Administration in the Roman Province of
Syria,
8
where he attempted to correlate the literary evidence with
the epigraphical. It was an essay in urbanization, an attempt to dis-
tinguish a polis and a kmh in terms of the functional evidence. As
for the local ocials, he was far less conclusive than Prentice, par-
ticularly when he tried to square the Syrian evidence with that then
available from Roman Asia: it was clear that on the level of local
administration, titles of ocials and their functions varied from region
to region. Harper had somewhat more success with corporate bod-
ies. His analysis showed numerous examples of t konon tw kvmw
operating as a collective municipal unit, taking the initiative in the
construction of public buildings and electing ocials. The expres-
sion by the village (p tw kvmw) occurs often in the inscriptions,
from which Harper concludes that the only way in which the vil-
lagers may reasonably be supposed to have formed common pro-
jects is by a common assembly.
9
There is, Harper continues, very
little evidence of cities controlling large territories in which villages
7
Prentice, Ocials, 123.
8
YClS 1 (1928) 10558.
9
Harper, Village Administration, 143.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 190
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 191
were situated. Each village seems to be a more or less independent
unit, with comparatively little dependence upon a unit higher than
itself, except of course the central administration of the Romans.
Of course. One of the chief ties of the Roman center with its
peripheries was money. This we knew, but what Harper put into
relief through his analysis of the inscriptions is the other, local side
of the money question. Villages had their own revenues out of which
they pursued their own self-interest. The money came in the rst
place from the summae honoriae, the often quite substantial sums paid
to the municipality by ocials upon entering oce. There were in
addition nes for violations of the law, not in this case for cruising
unwittingly into the village speed trap but, interestingly, almost always
for the violations of tombs and not, as we might expect, for viola-
tions of property boundaries or corruption. Gifts were another source
of local revenue, as was the rent or use fees for public facilities like
the village spring and the birkehs or reservoirs that are still a com-
mon a feature in the villages of southern Syria, or the rentals of
public land for cultivation or pasturage.
And where did the money go? A lot of it went to paying the taxes
assessed on the village by the city on which it depended and, of
course, to Rome. Local expenditures were limited, it would appear,
chiey to the construction of public buildings, secular like a council
house or a bath, or religious, a shrine or a temple. The construc-
tion of these still preserved modest but elegant buildings may not
have been as expensive as might appear. The koinn of the towns
of Syria, like those elsewhere in the Empire, often possessed public
slaves who worked as unskilledand unpaidlabor on these build-
ing projects. In much the same way, the Empire cut its construction
costs by using the far more skilled labor of the army, which was
more like an engineer corps than an infantry regiment, in the con-
struction of public projects in the provinces where they were stationed.
There is no sign that these villages expended any public monies
on education, public health, or charity, and on the evidence of the
inscriptions, there were no local police in these Syrian villages. There
were, however, gendarmes. Syria was a frontier province, and in
addition to the legions stationed throughout the eastern provinces,
the frontiers were guarded in many instances by local militias, bur-
garii or frouro, who manned those lonely towers whose remains still
stand silently facing the Syrian steppe. Providing and maintaining
this irregular soldiery appears to have been a local obligation, one
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 191
192 rn.xk r. rr+rns
other of the waysthe quartering and provisioning of regular troops
was still anotherin which the Empire laid o some of its military
costs upon the communities it was presumably protecting.
As remarked, the city in whose territory the village was situated
and the emperors delighted in rewarding and punishing cities by
adding or subtracting villages from their jurisdictiontook its share
of village revenues. This occurred in a number of dierent ways,
but the one that concerns us here is the city tax on village markets.
The buying and selling of basic commodities went on all the time,
of course, in the village suqs of Syria, but in an article published in
1970, R. MacMullen of Yale collected and analyzed the evidence,
some literary but most of it epigraphical, for those special markets
the Latins called nundinae and the Greeks panhgreiw, and we call
fairs.
10
No aspect of commerce in the Roman Empire has been more
neglected than that which involves the movement of goods within
any given fteen-mile radius, MacMullen began his article, and
yet this local trade must have accounted for a good three-quarters
of the value of exchange throughout the economy as a whole.
11
It
is, we might add, another example of an exceedingly widespread
activity, in this case an economic one, that was too local or too com-
mon or too banal to catch the notice of the ancient historians.
The nundinae took place in three dierent settings, at isolated road-
side shops or crossroads venues to serve a population too thin to
live even in villages, in the villages themselves, and, nally, in cities.
It is the village fairs that interest us here. They were special markets,
mostly for manufactured items, and scheduled on regularly spaced
market-daysnundina is elided from novem dies, the typical interval
for the Latin marketsso that the peddlersa North African inscrip-
tion has turned up two nundinariae, two women market hucksters
could make a continuous circuit of villages in a circumscribed area.
A church synod held in Spain in 300 attempted to discourage cler-
ics who, in search of business and fairs, scout around the province
looking for protable fairs. And at about the same time, on the
other end of the Mediterranean, the public orator Libanius describes
large villages . . . exchanging their goods through festivals (panhgreiw),
each playing host in turn . . .; enriched by them through giving of
10
R. MacMullen, Roman Market Days, Phoenix 24 (1970) 33341.
11
MacMullen, Roman Market Days, 333.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 192
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 193
its surplus or lling its needs, setting out some things for sale, buy-
ing others . . . . They need little from the city because of this exchange
among themselves.
12
The Greek panguriw summons up rather more than nundina a
special characteristic of these fairs. They were as much religious as
they were commercial occasions, and the time and place of their
occurrence often owed as much to sacred days or a sacred place as
it did to the need to exchange goods: buying, selling, and worship-
ping were extremely common bedfellows. One has only to think of
Mecca and environs where, in pre-Islamic days the sacred months
when the treuga Dei prevailed were marked not only by the rituals
of the hajj but by a series of fairs (mawsim/mawsim; unlike nundina,
which has embedded in it the notion of a xed time, mawsim seems
rather to suggest a xed place). There is no epigraphical testimony
to these pre-Islamic fairsnor of anything else in and around Mecca.
We have only the memory of al-Azraqi, a historian writing two cen-
turies after the mawsim had disappeared:
The hajj was in the month of Dhu al-Hijja. People went out with
their goods and they ended up in a place called Uqaz on the day of
the new moon of the (next month). They stayed there twenty nights
during which they set up in Uqaz their market of all colors and all
goods in small houses. The leaders and foremen of each tribe over-
see the buying and selling among the tribes where they congregate in
the middle of the market.
13
When Azraqi was writing in the rst half of the ninth century, the
historical memory of the Muslims could no longer identify the actual
site of Uqaz and the other fairs that had once accompanied the hajj.
Little wonder. It is clear even from Azraqis brief description that
the site was what the Romans called a vicus, roadside clusters of
shabby shops, to use MacMullens language.
14
In Italy and else-
where in the Roman Empire physical evidence has come to light
for these transient roadside fairs through chance nds of coins, pot-
tery and small objects at rural crossroads.
15
Meccan archeologists, a
club that at present has no members, are scarcely holding their
12
Oration 11.230.
13
Azraqi, Akhbr Makka 129, cited in F.E. Peters, Mecca: A Literary History of the
Muslim Holy Land (Princeton: Princeton University Press, 1994) 44.
14
MacMullan, Roman Market Days, 334.
15
MacMullan, Roman Market Days, 333.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 193
194 rn.xk r. rr+rns
breaths waiting for that to occur in the Hijaz; one can only hope
that the Bedouin who herd sheep in the caves around Mecca are
keeping their eyes open.
In all the works I have so far mentioned, the use of epigraphical
material was almost exclusively descriptive, to ll in the blanks of
how things worked in the Roman Empire, particularly here on the
local level. That was still MacMullens approach in his 1970 article
on Roman Market Days, but by 1974 his work had tacked o
into a slightly dierent direction. He begins his Roman Social Relations
with the programmatic remark My aim in this essay is to get at
the feelings that governed the behavior of broad social groups or
conditions,
16
from which he excludes, solely for his own purposes,
family, slave and client relations, race relations, which I take to mean
dealings among dierent ethnic groups within the empire, and nally,
though he does not mention it, there is no treatment of relations
between religious groups.
MacMullen treated urban, suburban, and rural relations within
the Roman Empire separately, and it is into this latter category that
I suppose the matter of the present conference chiey falls. Trans-
humance or seasonal migration was a fact of life in many provinces
of the Empire, not least in the Syrias, Palestines, and Provincia
Arabia, and we learn from MacMullens analysis of laws, inscrip-
tions, and the occasional piece of literary evidence that the old enmity
between the homesteaders and the cattle and sheep herders was
already in full swing in the Roman Empire long before Hollywood
discovered it. So too was rustling, and that activity that once occurred
with some frequency in New York City, but now no longer, being
mugged not only on the public ways but even in your own hotel
room.
17
If you stop for the night in a Palestinian motel, the Jerusalem
Talmud sagely warns, you would be well advised to make your will
beforehand.
18
But for all this interesting detail, there is still something remote
and impersonal about these social relations laid out by MacMullen
and by others who have gone down the same way. Neither MacMullen
nor even M. Rostovtze will ever sound like S.D. Goitein: the clos-
16
R. MacMullen, Roman Social Relations (New Haven: Yale University Press,
1974) vii.
17
MacMullen, Roman Social Relations, 24.
18
Y. Ber. 4:4 (8b); cf. MacMullen, Roman Social Relations, 4 and n. 13.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 194
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 195
est the rst two could come to their subjects was through inscrip-
tions; the latter had an archive. Inscriptions are about people; an
archive often gives us an opportunity to hear the individuals own
voice, or at least a more authentic report of it than we can man-
age with an inscription. Inscriptions are by their nature ocial, for-
mal, and expensive, and only occasionally, and under very special
circumstances, bi-lingual. Epigraphical names are often space-makers,
the matter of prosopography and not the stu of social history:
X is honored by the people of Y; N at his own expense erected this
building with the prnoia of A and the spoud of B. Spontaneity is
proscribed and the emotions registered are as prescribed as a Byzantine
cursus honorum. No one is bad-mouthed in an inscription; his name
is simply chiseled o, like the removal of a retirees name from a
building directory.
Tchalenko, Butler and Brnnow, and Domaszewski all turned up
no archives, nor did Jaussen and Savignac in northern Arabia.
19
Archeology is as often a matter of luck as it is of hard work. How
would H.D. Colt know in 1935, while excavating an already well
known siteit had already been surveyed by L. Wooley and T.E.
Lawrence before the First World Warcalled in Arabic Awja al-
Khar, that he would uncover in two ruined churches ve distinct
papyrus archives from this place the ancients called Nessana. One
of the archives was a library of literary texts,
20
but the other four
were deeds, documents, and letters, containing ninety-six major doc-
uments in all, covering matters economic, legal, military, social and
religious.
21
The Nitzana archives are more than familiar to this company, I
am certain, but I adduce them here simply to underline the quali-
tative dierence between the epigraphical evidence I have so far
been discussing, and the archival, or, to put it another way, to mea-
sure the distance between two emblematic projects like M. Rostovtze s
Social and Economic History of the Roman Empire,
22
for which the works
19
A. Jaussen and R. Savignac, Mission archologie en Arabie, vol. I: Textes; vol. II:
Cartes et Photographes (Paris: Paul Geuthner, 191420).
20
L. Casson and E. Hettich, Excavations at Nessana, vol. 2: Literary Papyri (Princeton:
Princeton University Press, 1952).
21
C. Kraemer, Excavations at Nessana, vol. 3: Non-Literary Papyri (Princeton: Princeton
University Press, 1954).
22
(Oxford: Clarendon Press, 1957
2
).
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 195
196 rn.xk r. rr+rns
of Prentice and Harper which I discussed earlier were building blocks,
or even MacMullens more modest and less ideological Roman Social
Relations, on the one hand and of Goiteins Mediterranean Society on
the other.
23
The comparison is a little unfair, of course, since Goitein
was drawing upon one of the richest of medieval archives in the
Middle East, but the type holds true, no matter what the size. The
papyrological moment, if I may use that expression, is both more
ephemeral and more occasional than the epigraphical. The inscribed
text is commissioned; the written one is more often merely paid for.
Both inditings are meant to record and preserve, of course, but the
epigraphical text is meant to be displayed, publicly and permanently,
hence the formality of the language and the marked absence of the
local and vernacular tongues. The inscriptions were intended to be
as permanent as the stones upon which they were inscribed; that a
text put on a wall or a monument in the second century is still
being read today is no more than the ideal fulllment of the inten-
tions of the people who put it there. The preservation of papyri,
and later of paper, on the other hand, is largely due, as C. Kraemer
put it, to the deep-seated reluctance to destroy written material of
any sort.
24
All these dierences speak to a dierent subject matter in inscrip-
tions and papyri: more grave in one, more commonplace in the other;
and because of the relative ease of writing, more concise in one,
more diuse, and so, in many ways more informative, in the other.
But they reveal as well a dierent clientele. The population displayed
on the walls of buildings, are a substantial lot: sovereigns, high
prelates, public ocials and the always popular class of donors and
public philanthropists, men and women who have left their mark on
society and their names on public buildings. In the papyri, however,
the high and mighty rub shoulders with the humbler classes, mostly
soldiers and village priests and the occasional traveler at Nitzana,
but brides and grooms elsewhere, or just ordinary folk like you and
me looking for a good dose of that old black magic for their ene-
mies or in-laws.
23
S.D. Goitein, A Mediterranean Society: The Jewish Communities of the Arab World as
Portrayed in the Documents of the Cairo Geniza (Berkeley: University of California Press,
196793).
24
Kraemer, Non-Literary Papyri, 3.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 196
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 197
I will restrict my remarks to two of the items about travelers in
the Nitzana collection, but before that, note should be taken of a
rather remarkable feature of the collection as a whole, and perhaps
those at Petra and elsewhere as well. Most of the Nitzana docu-
ments date from the sixth and seventh centuries, the eras on either
side of the Muslim era: of the ninety-six major pieces, fty-ve date
from before the Muslim conquest of the 630s and forty-one date
from after that event. They come from a place on the remote fringe
of the Hellenized world, indeed, of the Roman Empire, a small town
in the southern Negev, a frontier settlement where there were few
Greeks and the Aramaic-speaking Nabatean majority was already
being replaced, as they were elsewhere in the limotrophe regions of
the Near East, by Arabic speakers from northern Arabia. But for all
that, the documents written down there, whether the papyri in these
archives, the tombstones in its cemeteries, or the grati scribbled
on its walls, whether the subject or the writers were Christian or
Muslim, are almost all of them in Greek.
25
And it is somewhere in this
complex of elements that is buried the answer, I suspect, to the ques-
tion of why those Arabs, many of them Christians, did not require
an Arabic translation of the Gospels until the ninth century.
And so it is to two Greek papyri that I shall refer. The rst, P.Ness.
89, dates from the sixth or early seventh century, before the com-
ing of the Muslims in any event. It has to do with the business of
a partnership of traders (kointhw). There are three of them, and
their names speak volumes about that time and that place: Sergius,
Abraham and I, Zunayn (line 33). They are likely camel traders,
though they dealt in other goods as well, and on this occasion they
made a trip to Mount Sinai, t vgion ron (line 23), accompanied,
at a priceat a considerable price, it would seem, half the value of
a camelby one of the Sinai Bedouin, who serves as their sukofn-
thw, a rather peculiar usage of that word for guide as Kraemer
notes.
26
They were going to Sinai to pray (w exn), that is, as pil-
grims, but also, as it turns out, to hand over a donation from Nessana
and to do a little business with the monks there. Zunayn, and per-
haps Abraham as well, may have been ethnic Arabsor could
Abraham have been a Jew?though they are not called such, but
25
Kraemer, Non-Literary Papyri, 18.
26
Kraemer, Non-Literary Papyri, 259.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 197
198 rn.xk r. rr+rns
throughout the text the Bedouin Arabs are explicitly referred to as
Sarak enoi, whether serving as a guide or, later in the text, as tribal
raiders, the Banu Udayyid (line 35, o Eialodeeid) who made o
with one of the companys camels (line 35). These were ordinary,
everyday occurrences at that remote outpost; piety, danger, and the
cost of doing business, together with the names of the not very
remarkable people who were engaged in it, all unfold with remark-
able clarity in the forty-odd lines of preserved text.
The other papyrus, P.Ness. 73, is a letter and can be dated with
some certainty to 683 CE. The Muslims by now have been in con-
trol of the Negev for half a century and yet this brief letter is still
written in Greek, even though it is from Abu Rashid, the amir
(smboulow) of the province, to the people of Nessana, which is still
called that in the Greek text. In the name of God Almighty, it
begins, n nmati to yeo pantokrtorow, and then exarist t ye,
which is followed by the adverb peita, which Kramer translates as
Etc. etc. and which corresponds almost exactly to the amm ba'd,
the phrase that invariably follows formulaic invocations of Gods
name in formal writing in Arabic. When my wife Ubayya arrives
thereUbayya probably arrived in Nessana with this letter in her
handprovide her with a man contracted to show her the way to
Tur Sina. And pay his salary too.
What we have, then, here and in the very similar P.Ness. 72 from
a year later, which is addressed more explicitly to a certain George,
who is obviously a Christian and apparently the administrator of
Nessana, are other trips from Nessana, now a reduced settlement of
no more than a 1000 souls but still the last jumping-o point, to
Sinai. Now it is the wife of the governor and in no. 72 by a mawla
(maule) of one Urwa ibn Abu Sufyan who wish to go. Their objec-
tive seems to be pilgrimage, as it was in Colt no. 89, but we can no
longer be certain of the religious aliation. Both the wife of the
governor and the mawla or freeman could be Christians or even
recent Christian converts, as Kraemer surmises,
27
who are continu-
ing the long-standing Christian custom of pilgrimage to the monastery
at Sinai. Or they could be Muslims making a ziyra, a pious visit to
Tur Sina, a place already venerated in the Quran, as in 95:2, where
it is bracketed with Mecca itself.
27
Kraemer, Non-Literary Papyri, 205.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 198
+nr nov.x xr.n r.s+: +nr \irv rnov nrrov 199
I adduce these two documents not to unpack from them all these
riches they hold for the historian of the Middle EastP. Mayerson
has already done that in quite exemplary fashionbut merely to
illustrate the dierence that papyrological evidence makes to the his-
torian, in this case, to the historian of the Roman Near East. They
and the rest of the Nitzana archives tell us invaluable things like the
prices of ordinary commodities, and so the cost of living, in the late
sixth and early seventh centuries. They tell us too about shifts, per-
manent shifts as it turns out, in the population of the region, a shift
not yet reected at Nitzana in the language of record, which is one
of your chief concerns in this conference, but certainly in the ono-
masticon that such a collection puts before us and that the writings
of the high historical tradition, which is not very interested in places
like Nitzana or Engeddi, never provide.
SHIFFMAN_f10_187-199 12/2/02 5:27 PM Page 199
200
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 200
This page intentionally left blank
THE DECIPHERMENT AND EDITION OF THE PETRA
PAPYRI: PRELIMINARY OBSERVATIONS
Ludwig Koenen
in collaboration with R.Ch. Caldwell,
R.W. Daniel, and T. Gagos
The carbonized Petra papyri, found at the end of 1993 in a side
room of a Byzantine church during an excavation of the American
Center of Oriental Research (ACOR), were opened and conserved
by a Finnish team in 1994/1995. This conservation team was directed
by Jaakko Frsn of the Finnish Academy and the University of
Helsinki. Its work led to a still continuing second phase in which
the papyri are deciphered, reconstructed, and prepared for publica-
tion by two teams, one from Helsinki, directed by Frsn, the other
from Michigan, led by the present writer. The rst volume has just
been published by the Finnish team (P.Petra I).
1
About thirty-ve or
201
1
J. Frsn, A. Arjava, and M. Lehtinen (eds) with contributions by others, The
Petra Papyri I (American Center of Oriental Research Publications 4; Amman: ACOR,
2002). For the excavation and the churchs archaeological and historical context,
see now Z.T. Fiema, Chr. Kanellopoulos, T. Waliszewski, and R. Schick (with con-
tributions by many other scholars), The Petra Church (ed. P.M. Bikai; American Center
of Oriental Research Publications 3; Amman: ACOR, 2001); specically on the
papyrus nd, see ibid., pp. 13850; J. Frsn, The Petra Papyri: Information and
Signicance, A City Forgotten and Rediscovered (eds J. Frsn and Z.T. Fiema; Publica-
tions of the Amos Anderson Museum: Helsinki University Press, 2002) 1824; cf. also
J. Taylor, Petra and the Lost Kingdom of the Nabataeans (London/New York: I.B. Tauris
Publishers, 2001) 20711; for a discussion of the papyrological information in its
historical context, see R.Ch. Caldwell, Between State and Steppe: New Evidence for Society
in Sixth-Century Southern Transjordan (University of Michigan Ph.D. Dissertation, Ann
Arbor, MI, 2001). The nal vs of the rst volume by the Michigan team, an edi-
tion of the long roll Inv. 10 is presently under revision.
The American team is supported by the National Endowment for the Humanities
(NEH) and the University of Michigan, the Finnish team by the Finish Academy
and the University of Helsinki. The work of both teams is greatly facilitated by
The American Center of Oriental Research (ACOR) and its Directors, Pierre and
Patricia Bikai, as well as by the Department of Antiquities of Jordan and its for-
mer Director General Ghazi Bisheh as well as its present Director General Fawwaz
al-Khraysheh. Our thanks go also to Omar al-Ghul (Yarmouk University) and Hani
Ali Falahat (Antiquities Inspectorate, Petra). O. al-Ghul advised us on matters of
Semitic philology; H.A. Falahat provided a wealth of information on toponyms used
today in the Petra and Wadi Musa region. Zakaria al-Nuaimmat at ACOR referred
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 201
202 rtrvio korxrx
us for information on Ail to Gluecks article cited in n. 8. Of greatest help were
the prosopographical les and the databank of all deciphered texts in searchable
format, which were created, regularly up-dated and managed by M. Lehtinen of
the Finnish team. I thank R.Ch. Caldwell also for correcting my English.
The best-preserved papyri were adopted by donors who named these rolls either
after themselves or somebody they wished to honor. Except in the case of short
and repeated references, I add the adoption name to the inventory or publication
number. Since the Finnish team has assigned papyri to individual scholars for deci-
pherment and publication, their name is also added in parentheses; the Michigan
team works as a group.
2
R.Ch. Caldwell had joined the team for shorter periods already in previous
years.
3
In Inv. 85+89G, also called P.Petra James Garland and Carol Andreae
(M. Lehtinen), dated to 578582, perhaps to 581/2, and Inv. 72+79, also called
P.Petra Gladys J. and Frank J. Vocci (M. Vierros), written between 582591,
Theodoros is still alive, but he seems to have died some time before 592, as
A. Arjava has argued by letter (cf. n. 6).
4
The papyri were probably brought into the storage room after the death of
Theodoros heirs. This theory may explain why the nd contains rolls that had no
legal function. The roll with copies of Obodianos donatio post mortem was never cut
apart for distribution of the copies to the interested parties (Inv. 6a, also called
forty of the originally ca. 140 rolls yield enough text and informa-
tion for publication. The two teams work in collaboration. For this
reason, my report will make use of information from papyri assigned
to both teams. It is based on, and updated from, an earlier paper
written by T. Gagos, R.W. Daniel, and the present writer for a pub-
lication to be edited by G. Markoe and to accompany an exhibi-
tion of the Cincinnati Art Museum on the Nabataeans. In addition
I am much obliged to R.Ch. Caldwell, who joined the Michigan
Petra Papyri Project in 2001.
2
Deciphering and interpreting continue.
A new reading, a new location of a fragment, a new parallel may
still change not only the reconstruction of the text of the papyrus
at hand, but also aect the decipherment of, and conclusions drawn
from, other papyri of the nd. Hence, any attempt to evaluate the
new texts is tentative as long as the primary texts are subject to
change.
1. Theodoros son of Obodianos and his Wealthy Family
The Petra papyri are property-related, private papers of Theodoros
son of Obodianos (514591),
3
a property owner and deacon, later
archdeacon of St. Mary, the see church of Petra, presumably the
church where the papyrus rolls were found.
4
He originated from,
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 202
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 203
Fig. 1: Petra and hinterland. Map adapted from Z.Th. Fiema, Military Architecture
and the Defense System of Roman-Byzantine Southern Jordan, Studies in the History
and Archaeology of Jordan V (Amman: Department of Antiquities of Jordan, 1985) 262.
Names attested in the Petra papyri are in italics.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 203
204 rtrvio korxrx
and later returned to, Kastron Zadakatha (modern adaqa, 20 km
SSE of Petra; see Fig. 1),
5
where he owned property inherited from
his father. The papyri document some of his private transactions
mostly in Petras hinterland.
6
Extant dates span from 537 to 593/4,
just a little beyond Theodoros death. Many of the documents involve
other members of his large family (see the family tree, Fig. 2).
Theodoros family was relatively well-to-do. Men from this family as
well as other upper class people in the Petra papyri added the Roman
gentilic name Flavius to their own name as in, for example, Flavios
P.Petra Daniel C. and Nancy E. Gamber [ J. Frsn]); and the end of Inv. 10
(below) seems to have been lost when the papyrus was rolled together for the last
time. Without the signatures, which in the Petra papyri seem always to have been
at the end of the documents, this roll had no legal value. Cf. L. Koenen, Preliminary
Observations on Legal Matters in P.Petra <Inv.> 10 (P.Petra Khaled and Suha
Shoman), Atti del XXII Congresso internazionale di papirologia: Firenze, 2329 agosto 1998
(Florence: Istituto Papirologico G. Vitelli, 2001) 728; Fiema in The Petra Church
[above, n. 1], 14849. Both publications compare the Nessana papyri found in
churches. Terry G. Wilfong directed my attention to the monastery of Phoibamon
in Deir el-Bari (Egypt) where also private papers had apparently been stored
(W. Godlewski, Le monastre de St. Phoibammon [Deir el Baari 5; Warsaw: PWN,
Editions scientique de Pologne 1986] 57).
5
Here as well as in the following pages, compare the tentative family tree (Fig.
2) and the map of Petra environment (Fig. 1).
6
Inv. 71.1 and 2 are receipts for tax arrears paid between 591 and 593 by the
heirs of the estate of Hierios son of Patrophilos on behalf of its share of the estate
of Panolbios son of Theodoros, Hierios nephew and cousin once removed. This
situation implies that part of the estate of Panolbios had fallen to the son of his
uncle Hierios, when Panolbios at his death (according to M. Lehtinen, between 565
and 582) had no living children and his father Theodoros was dead.
Valens
Alphios Dusarios Palladia Bassos Thaaious Obodianos
Euthenios daughter Patrophilos son Hieria (?) Obodianos Gesios Leontios
Bassos Epiphanios Sabinos
Hierios Stephanous Theodoros Obodianos
Panolbios Georgios
Fig. 2: The extended family of Theodoros son of Obodianos, partially based on
M. Lehtinen, Preliminary Remarks on the Prosopography of the Petra Papyri,
Atti del XXII Congresso (see n. 4) 78794 (a tree on 793).
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 204
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 205
Theodoros son of Obodianos. In Petra and elsewhere, Flavius served
as a status designation. It marked its bearer as a member of the
elite by relating him to Constantine the Great and the second Flavian
dynasty of the fourth century. The aluence of Theodoros family
may be illustrated by three examples.
(1) The rst testimony comes from P.Petra Inv. 10 (also called
P.Petra Khaled and Suha Schoman; cf. n. 1), a division of property.
This large roll concerns Theodoros cousins, the brothers Bassos,
Epiphanios, and Sabinos, who at some time between 527 and 537
divided their joint property among themselves: vineyards, seed land,
apartments, agricultural buildings, and four slaves.
7
Most of this prop-
erty was probably inherited from their father. It was situated in two
places, the village Serila (possibly on the Shera, the escarpment east
of the town of Wadi Musa), and in an area called Ogbana (arguably
in the now largely expanded area of that town).
8
Damage to the
papyrus does not allow us to calculate the size of the property with
precision, but it can be said that Epiphanios received a total of, at
least, 13.3 iugera of vineyards and 33 iugera of sown land and Sabinus,
7
Koenen, Preliminary Observations, Atti del XXII Congresso [n. 4], 72931; see
there also for the procedures of the division. The brothers rst divided the entire
property into three equal shares without yet knowing who would receive which
share. Then, by casting the lot, the pre-divided shares were assigned to their new
owners. Thus fairness was in the hands of God. At the end of the document, the
brothers swear an oath by the Holy Trinity and the Imperial Salvation that they
will abide by the terms of the division. The results were denitive and unques-
tionable.
8
Also Patrophilos, the brothers paternal uncle, owned land in Serila that he had
inherited from his father Bassos, probably the grandfather of the brothers of Inv.
10 (Inv. 63+65, also called P.Petra Wesley and Virginia Egan). On the basis of
place names, R.W. Daniel (P.Petra Inv. 10 and its Arabic, Atti del XXII Congresso
[n. 4], 33141, esp. 33438) proposed our working hypothesis that Serila and
Ogbana as well as el-J" were in the area that now is covered by the rapidly grow-
ing town of Wadi Musa (accepted by me in the oral presentation of this paper,
but see now sect. 3). Following Caldwells suggestion and additional comments by
Omar al-Ghul (see n. 1), I am now inclined to locate Serila in the Shera, the rolling
country on top of the escarpment to the east of Wadi Musa. The name may be
explained as er(")-l, the Shera belongs to l (see Caldwell, Between State and
Steppe, 17172 line 41); cf. the god Dushara, the one who belongs to (du) Shara"
(G.W. Bowersock, Roman Arabia [London and Cambridge, MA: Harvard University
Press, 1983] 8 n. 16). A village al-Ail, marked by its Nabataean nds and Roman
ruins (including a tower, and the spring 'Ain al-Ail, and a series of threshing oors)
is situated on the slopes of a long-stretched hill overlooking the Wadi al-Ail of the
Shera (see map; N. Glueck, Explorations in Eastern Palestine, II [AASOR 15; Philadelphia:
American Schools of Oriental Research, 193435] 7374). Its name may still pre-
serve the name of the god l.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 205
206 rtrvio korxrx
at least 16 iugera of vineyards and 26.5 iugera of sown land.
9
The
share of Bassus is practically lost, but he must have received prop-
erty roughly comparable to his two brothers, say ca. 15 iugera of
vineyards and close to 30 iugera of sown land. In total, the brothers
divided at least 134 iugera (close to eighty-four acres), and probably
more likely a minimum of about 140 iugera or eighty-seven acres.
10
For modern ears, this may not sound like much. In the absence of
economic data from Petra, it is impossible to say what these num-
bers really mean in terms of wealth, particularly since Inv. 10 does
not indicate the quality of the land. But vineyards represented almost
one third of the land that the brothers divided and may be assumed
to be top quality land.
11
Despite the dierence of Egypts economy
and yield per arura, one might get a rough impression by compar-
ing ready available data from fourth-century Egypt, where ca. 11
percent of landowners living in Hermopolis and Antinoopolis owned
more than 100 aruras or 109 iugera. As R. Bagnall observed, every
Hermopolitan male holding more than 100 aruras would have to
belong to the council for that body to number as much as 100
members.
12
Each of the brothers probably owned land already before the divi-
sion of Inv. 10. A substantial part of their wealth as well as that of
Theodoros extended family was clearly based on private landown-
ership. In this regard Petra was not much dierent from other
Mediterranean cities. New vineyards had been planted on 6 iugera
9
At least means that, in a few ambiguous readings, the lowest possible num-
ber of iugera went into the calculation. Where the size is not indicated or the num-
ber of iugera is broken o (in twelve instances), one iugerum is assumed, unless the
extant plural ogera indicates a minimum of two iugera. The calculation is only
possible for Epiphanios and Sabinos. The gures that above are given for Bassos
are the average of the numbers for his brothers.
10
For calculation I assume that the standard Roman iugerum of about 2523 square
meters was used, though there is some ambiguous evidence for the use of a so-
called small iugerum of about 1261 square meters in the Eastern Mediterranean;
see E. Schilbach, Byzantinische Metrologie (Byzantinisches Handbuch 4; Handbuch der
Altertumswissenschaft 12.4; Munich: C.H. Beck, 1970) 7780.
11
According to the Syro-Roman Law Book (eds K. Bruns and E. Sachau; FIRA II
pp. 795996 121), 5 iugera of vineyards carry the same tax load as 20 iugera arable
land. See Caldwell, Between State and Steppe, 79.
12
R.S. Bagnall, Egypt in Late Antiquity (Princeton: Princeton University Press, 1993)
6970; idem, Landholding in Late Roman Egypt, JRS 82 (1992) 12849.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 206
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 207
in Serila. It took three or four years for newly planted vines to
mature and yield fruits.
13
These were investments in the future.
In addition, the brothers divided between themselves a total of
twenty or twenty-one apartments, mostly subunits on the main-level
(okoi) of extended structures, partly on their top-level (pera),
14
twelve in Petra and nine in the village Serila, also a stable, a cis-
tern, and some part of a farmhouse as well as three threshing oors
with granaries, a piece of land, two dung depositories (very valuable
assets in ancient agriculture), and one large and one smaller dry-
garden (i.e., orchards without articial irrigation, Arabic: ganna) in
Serila and one large and at least four smaller dry-gardens in Petra.
It sounds as if the private gardens in Petra made it a garden city.
15
The apartments came with numerous attached structures varying
from case to case: bedchambers, entrances, courtyard entrances, court-
yards, stairs and balconies,
16
vestibules, door leaves, and a small
tower. Most of the structures were part of larger farmhouses or, in
Petra, part of one large villa. Dividing the structures owned by the
brothers created smaller units, each either with its private entry or
with the right to build one. Whether, however, all the apartments
were to be used by the brothers and their families or were rented
out or sold, we do not know. While it is also impossible to estimate
the size and value of the divided structures, we can assume that,
around the thirties of the sixth century, these structures as well as
the land under division indicated some wealth for the family of
Theodoros cousins.
(2) In 537, at the age of twenty-four, Theodoros son of Obodianos
married Stephanous daughter of Patrophilos. Two years later, in
539/40, Patrophilos and Theodoros reached an agreement about a
13
H.P. Olmo, Grapes, Evolution of Crop Plants (ed. N.W. Simmonds; Essex:
Longman Scientic & Technical, 1976) 294; C.E. Walsh, The Fruit of the Vine:
Viticulture in Ancient Israel (HSM 60; Winona Lake, IN: Eisenbrauns, 2000) 99.1.
14
For the architectural terminology, see sect. 3.
15
Gardens seem to have been part of early Nabataean Petra. See the recently
identied ornamental public garden with a pool located in Petras civic center south
of the Colonnaded Street: L.-A. Bedal, A Paradeisos in Petra: New Light on the
Lower Market, ADAJ 43 (1999) 22739; eadem, Paradise Found: Petras Urban
Oasis, Expedition 42/2 (2000) 2336; and In Search of Petras Buried Garden,
ACOR Newsletter 13/1 (Summer 2001) 13.
16
The balconies were part of the access way. They were reached by outside
stairs from the courtyard.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 207
208 rtrvio korxrx
dowry of three Roman pounds of gold and parapherna of 3' pounds
of gold (Inv. 63+65, see n. 8). The total of 6' pounds amounts to
2.128 kg or about 5 pounds 8' oz troy. The value given for dowries
can be ctitious in the sense that not the full amount may have
been paid by the father of the bride. But if Stephanous dowry should
have to be returned, Theodoros would have to pay the full 6'
pounds of gold. As A. Arjava remarked, this was certainly a con-
siderable sum of money in the province of Third Palestine, although
it was naturally not on the same level as the dowries of the impe-
rial aristocracies: their largest dowries were estimated at 100 pounds
of gold.
17
(3) Also early, in 538, Theodoros son of Obodianos, temporar-
ily moved from Petra to Gaza where he participated in the settle-
ment of an inheritance (P.Petra I 2, also called P.Petra P.E. MacAllister
[ J. Frsn]). The properties were spread over a large area, as is indi-
cated by the place-names that we recognize: Eleutheropolis (45 km
NE of Gaza), possibly Bero[saba] (Beersheba; 45 km SW of Gaza).
No doubt, the interests and activities of this Petraean family extended
far to the west of the Wadi Araba.
Despite the prosperity of this single family, we must be cautious
in drawing conclusions regarding the general health of Petras econ-
omy. It should, however, be obvious that Petra in the sixth century
was still a place where relatively wealthy people would live.
2. Petra, a Proud City of the Roman and Byzantine Empire in the
Sixth Century
The city maintained a healthy self-image during the sixth century.
She continued to style herself an important part of the Byzantine
Empire. Her full titulature in the dating formula used in some ocial
documents is:
18
17
A. Arjava, Family Finances in Byzantine Near East: P. Petra inv. 68, Atti
del XXII Congresso [n. 4], 6570 with the quotation on p. 68; also T. Gagos,
Negotiating Money and Space in Sixth Century Petra, Atti del XXII Congresso
[n. 4], 495509. P.Petra Inv. 68 is now P.Petra I 1.
18
Petras titulature is part of the opening dating formula. As one should expect,
most of the beginnings of the rolls are lost and the remaining ones are heavily
damaged. The full version is extant or can be reconstructed in ve papyri (P.Petra
I 1, Inv. 6a, 48 [see n. 39; beginning lost], 67, and 85+89G), covering a period
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 208
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 209
n Agou!tokolvn& Antvnian,
pi!m ka e&a
.
ge (?) mhtr kolvnin
Adrian Ptr& mhtroplei
t! Trth! Palai!tnh! %aloutara!.
In Imperial Colony Antoniana, Distinguished and Holy (?), Mother of
Colonies, Hadriana Petra, Metropolis of (the province) Tertia Palaestina
Salutaris.
The main components of this titulature are attested elsewhere, but
its entirety appears in the Petra papyri for the rst time. The com-
bined titles amount to a conscious catchword history of the Roman-
ization of the city. After Rome had annexed the Nabataean kingdom,
Petra received the title Metropolis of (the province) Arabia by
114. Some years later, Hadrian bestowed the title Hadriana upon
Petra, probably in connection with his visit to the city in 129/30.
Some ninety years later, Elagabalus, who was ocially called M. Aure-
lius Antoninus (218222), honored the city with the status of Roman
Colony: hence Agou!tokolvna Antvnian (sic), Imperial Colony
Antoniana. Rival Bostra, as well as other important cities of the
East, also had the epithets Hadriana and Imperial Colony. It
was probably also Elabagalus who bestowed the title mthr kolvnin,
Mother of Colonies, upon Petra. This title that again was used by
several cities in the east is also attested in an earlier inscription from
the area of Petra. The epithet recasts the Greek concept of mhtr-
poli!, mother-city, which denotes the main, or one of the main,
cities of a province. But while the rst part, mother, was retained,
the second part transformed the Greek concept of polis into the
Roman concept of colony. The result was a hybrid term
19
within
the distinctively Roman terminology of colonization. But, in the tit-
ulature, Mother of Colonies does not replace the constituent terms,
Mother-city and Colony, the latter present in kolvna, but rather
supplements and tops them, as all these titles stand side-by-side.
from 537 to 573. There existed also a shorter version, beginning with n mhtroplei
kt (ten extant examples from 538 to 593/4). See already L. Koenen, The Carbonized
Archive from Petra, JRA 9 (1996) 187.
19
So F. Millar, The Roman Coloniae of the Near East: A Study of Cultural
Relations, Roman Eastern Policy and Other Studies in Roman History: Proceedings of a
Colloqium at Tvrminne, 23 Oct. 1987 (eds H. Solin and M. Kajava; Commentationes
Humanarum Litterarum 91; Helsinki: Finnish Society of Sciences and Letters, 1990)
758; for mhtrokolvnea as a hybrid Greek-Latin term, see p. 41; idem, The
Roman Near East, 31 BCAD 337 (Cambridge, MA and London: Harvard University
Press, 1993) 651 and 308.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 209
210 rtrvio korxrx
Together, they are historical markers for the polarity of the Greek
and Roman cultures as they established themselves in the eastern
part of the empire.
Even the title of the province, Tertia Palaestina Salutaris, Third
Palestine the Benecial, sounds grandiose. Hitherto we knew the
province either as Palaestina Tertia or as Palaestina Salutaris. The
rhetorically combined accumulation of the three components is new
and probably a local phenomenon in a city that never could get a
title long enough to match its self-felt importance. New, too, are the
epithets distinguished and holy (?). The latter, e&a
.
g!, is an un-
certain reading
20
and to my knowledge not attested in the title of
any other city. It would have been understood as illustrious (eag!,
conspicuous) before the fourth century, but during the Christian
epoch, euagh! expressed predominantly religious purity and holiness
(eag!). In the Petra papyri and elsewhere, it marks churches, chapels,
monasteries, and other buildings used by the church. Hence, this
epithet may have become an emblem of Christian Petra.
By listing the titles bestowed by emperors, Petra appeared equal
to the most important cities of the Middle East. Even though, in
concrete terms, these past imperial favors did not signify a special
relationship between the emperors in Constantinople and Petra, they
were still important for the Petraeans. They still gave the city its ele-
vated status within the region.
The documents have not yielded substantial information on city
government. While they mention seven curiales (politeumenoi) of the
cities of Petra and Augustopolis (modern Udruh) and thus attest to
the continued existence of the bouleutic class,
21
this fact does not
20
Inv. 6a e#age (thus Frsn), 67:3 e&a
.
ge; P.Petra I 1:4 has only a lacuna for
both letters; and in Inv. 48 and 85+89G the word is not extant.
21
For the related questions as to how long bouleutai remained functioning in the
East, see H. Geremek, Les politeumenoi gyptiens, sont-ils identiques aux bouleuta?
Anagennesis 1 (1981) 23147; A. Laniado, bouleuta et politeumenoi, Chronique
dgypte 72 (1997) 13044; K.A. Worp, Arjante! and politeumenoi in Papyri
from Graeco-Roman Egypt, ZPE 115 (1997) 20120; Bagnall, Egypt in Late
Antiquity. For Petra, Fiema (The Petra Church [above, n. 1], 42627; cf. eundem,
P.Petra I, p. 3) argues that its city council still existed at the beginning of the sixth
cent. (and longer). He refers to a dedication to a councilor of the Petraeans
(bouleut! tn Petravn), deceased at the age of thirty-six, on an inscription, found
in el-J" (now the center of Wadi Musa; M. Sartre, Inscriptions de la Jordanie IV,
Inscr. Grecques et Latines de la Syrie [Paris: Librairie Orientaliste Paul Geuthner,
1993] 68). In his text, Sartre accepts the rst editors (A. Musil) reading t[ou]!
uia
. . .
Eboulo!, where the year 411 of the provincial era dates the inscription to
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 210
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 211
imply a city council (boul) making its own decisions. By the fth
century, city councils had lost their importance throughout the east-
ern provinces. The last fully functioning city council mentioned in
Egypt dates to 370 (P.Oxy. XVII 2110), and an epigram honoring
Herculios, the praetorian praefect of the oriental diocese of Illyricum,
implies that in 408/412 in Megara, and more generally in Greece
as well as in the rest of Herculeos diocese, councils were still active.
22
City councils may have continued for some time but certainly lost
their governing power. Even so, however, members of the property-
owning families, once the backbone of the city councils, retained
administrative functions in their provinces. They often held oces
and, most importantly, continued their involvement in collecting the
taxes that had been assigned to their communities by the central
government.
23
In this function, they guaranteed the collection with
their person and their private property.
This general system is now found in Petra and Augustopolis.
24
At
least four of the seven curiales mentioned (politeumenoi) were mem-
bers of tax committees that also kept the tax schedules (podktai,
xru!upodktai), while other politeumenoi carry this title when they
are attending to their private business. They all belong to property-
owning families whose ancestors once must have sat in Petras city
council (boul). Their names begin usually with the status designa-
tion Flavius. In two cases, father and son are each called politeu-
meno!, suggesting either that the titles were hereditary or that
516/17 (R.E. Brnnow and A. v. Domaszewski, Die Provincia Arabia [Straburg: K.J.
Trubner, 1909] 3.350). At that date I should expect that the deceased person would
be called makaritato!. In his comments, Sartre indeed argues for an earlier date
and, without having seen the inscription, suggests that, possibly, the number was
misread for ria, namely 216/7. Correcting the reading is tempting. The rst edi-
tor misread some other letters, and the three letters unread after the number cause
further doubts. But without inspection of the original, neither Sartres nor any other
correction can be veried. In any case, the new information from the papyri nei-
ther supports a sixth century council nor rules it out.
22
IG VII 93:4: He (Herculius) has lled cities and councils (!tea ka boul!)
with wisdom of mortal men; for the text of this inscription, see W. Peek, ZPE 31
(1978) 252; L. Robert, Hellenica IV (1948) 6061; SEG XXVIII 439 and XLVI 518;
cf. A. Laniado, bouleuta et politeumenoi, 141 n. 59.
23
The continued importance of the curiales in the eyes of the central government
is illustrated by a constitution of Anastasios: CJ 7.39.5 (around 500) and again Nov.
8 (of 535).
24
The occurrences of politeumenoi in the Petra papyri are discussed by Caldwell,
Between State and Steppe, 6872.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 211
212 rtrvio korxrx
the sons followed the example of their fathers in serving as tax
collectors.
At this point, two related papyri, P.Petra I 3, also called P.Petra
Zbigniew T. Fiema and Deborah Kooring (ed. M. Vierros), and 4
(ed. T. Rankinen) call for brief attention. They belong to a docu-
mentary type (called p!talma), in which one or both parties of a
transaction request the collegium of tax collectors to transfer the tax
responsibilities from the previous owner of landed property to the
new owner. In the mentioned two documents, Flavios Patrikios
addresses the head (or the acting member) of the tax collegium with
the words: to the most venerable (ade!imtato!) Euthenios son of
Dusarios, curialis and tax collector of the same (Petra): poli[teu]om[n]&
ka podkt t! at!. Because of the connecting and, it may seem
that politeumeno! signies another oce in addition to tax collec-
tor, but the language is ambiguous. politeumeno! may rather indi-
cate the high ocial standing of Euthenios, which then, by way of
explanation, is specied as tax collector. Taken in this way, poli-
teumeno! (curialis) neither indicates an oce in its own right nor is
the term identical with the oce of tax collector.
25
Not all tax col-
lectors in the Petra papyri were politeumenoi.
In sum, members of the families that once had curial rank con-
tinued to serve in ocial functions, especially in collecting taxes.
This function was important for the state, but also for the commu-
nity and the property owning families. The continuity, however, does
not imply that the boul, the city council, still existed. Its essential
administrative responsibilities had passed to its constituent private
families and, presumably, to the church and the military.
The preceding remarks implied already that the centralized land-
tax system of Byzantine Empire was implemented in Petra. The sys-
tem, essentially going back to Diocletian, was based on the iugum, a
theoretical and uniform measure for the calculation of taxes that, at
once, took quality, size, and probably also geographical dierences
into account. The agricultural land was categorized as vineyards,
sown land, land of lower qualities, and unfertile land. Depending on
the land quality, a dierent number of iugera or other local measures
25
The same usage occurs in Inv. 71.3, a tax receipt, where the signature of the
tax collector is followed by the signature of Flavios Victor son of Georgios(?),
pol(iteumeno!) ka lo[. The precise oce indicated by lo[ is not clear.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 212
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 213
counts as one iugum. For example, in the Syro-Roman Lawbook (above,
n. 11) one iugum was dened as the tax obligation either of ve iugera
of vineyards or of twenty iugera of sown land or of forty iugera of
mountainous and unfertile land. The numbers vary in attestations
from dierent areas. Within this system, Constantinoples annual tax
request needed merely to be divided by the number of iuga of any
specic province in order to calculate the at rate per iugum. On
this basis, the governor communicated the assessment for which indi-
vidual communities were responsible according to the total (in Petra,
the m!) of land. Similarly, the communities assessed the landown-
ers based on the total of the land (also m!) registered with them.
The system based on iuga, the iugatio, was implemented in Petra,
as R.Ch. Caldwell recognized. But, in the Petra papyri, we hear
very rarely about it. When the owners of land reported changes of
ownership to the committees of tax collectors (podktai, xru!upodktai;
see above), they used Roman iugera or Semitic measures like kor, se"ah
and kab. Whether these committees recorded every persons total of
land in these common measures or in iuga or in both, as one may
suspect, we do not know. One may compare the Egyptian practice
where the higher administration used calculation in iuga, while on
the local level the tax load was calculated in traditional rourai.
26
In the emerging picture, the burden of the tax system fell heav-
ily on the landowners, but they also controlled its implementation.
The details conform to Byzantine law and practice. This is also the
case in the following observation. In Petras hinterland most of the
agricultural land in private ownership was under the tax schedule
of the city. Such land was called free as, for example in the phrase
ogera leuyerik, free iugera. But there was another rare cate-
gory called patrimonal land, i.e., belonging to the patrimonium sci,
as in, e.g., ogera patrimvnli[a,
27
namely land that paid to the
26
Caldwell, Between State and Steppe, 7794. The evidence is mainly provided by
four receipts that are part of P.Petra I 710 (ed. M. Vesterinen; cf. her article
Theft and Taxes, A Series of Short Documents, Atti del XXII Congresso [n. 4],
128185); 10:10 mentions the ougatvn; Caldwell, Between State and Steppe, 8892.
It is not clear why in these casescontrary to the usual practice of the Petra
papyrithe iuga are mentioned, possibly in an exceptional obedience to Nov. 17.8
of 535 (see A. Ajarva and M. Vesterinen, P.Petra I, p. 104). For Egypt, see R.
Bagnall, P.Oxy. XVI 1905, SB V 7756 and Fourth-Century Taxation, ZPE 37
(1980) 18595; eundem, Egypt in Late Antiquity, 156.
27
In Inv. 8 pl. 38, an exchange of property, patrimvnali[ (twice) follows after
the text has just mentioned every free tax (p!an !untleian leuy[erikn) with
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 213
214 rtrvio korxrx
private treasury of the Imperial House. Hence, free, refers to land
that was free from the scal administration of the Imperial House.
This easily misleading terminology was previously known only from
Byzantine Caesarea in Cappodocia and Cyrrhus, a city in Syria, east
of Antioch.
28
The distinction between the two taxing authorities was
apparently important for the tax rate, but in Petra the taxation was
handled by the same committees.
There are also indications of a continued presence of the Byzantine
military in the camps of Zadakatha (now adaqa, 20 km S of Petra)
and, probably, Ammatha (now ammm, 32 km SE of Petra, 29
km east of adaqa).
29
The previous view held that in the sixth cen-
tury the castella of these places were no longer used by the military.
One of the parties in Inv. 67 (also called P.Petra Selz Foundation
II), written in 544 in Augustopolis (Udruh), is Dusarios, formerly of
Petra, a grand-uncle of Theodoros and a former prefect of Kastron
Ammatha: p praifktvn K!trou Ammayvn. Since remains of the
castellum, ascribed to the fourth-fth centuries, still exist, it is per-
suasive to see Dusarios as a former military prefect. The castellum
must then still have been in use in the sixth century.
The attestation for adaqa is unambiguous. The Petra papyri refer
to some private and civic activities of two categories of military
ocers of Kastron Zadakatha: prvre! ( priores), senior soldiers, and
mbaymo, soldiers of entering rank, hitherto known only from Nes-
sana.
30
One prior of Zadaqatha, Flavius Barakhos, occurs in Inv.
44a, a contract of 593/4 concerning the dowry that Kyra the daugh-
ter of Georgios brought to her marriage with Flavius Tomas. Also
one of the parties, the husband Flavios Thomas son of Nikephoros,
regard to another piece of land; cf. Inv. 64+65 and patrimvnou ogera in P.Petra
I 4:12; 5:4, 5, 13; T. Rankinen and M. Vierros, pp. 7577. The patrimonial land
is, or once was, Imperial domain land. For private land paying taxes to the Imperial
Fiscus already much earlier, see P.Hever 64:2630 and the editors (H.M. Cottons)
discussion.
28
See J. Gascou, Les grands domaines, la cit et ltat en gypte byzantine,
Travaux et Mmoires 9 (1985) n. 229 referring to Theodoret, epist. 42 (distinguishing
between oga leuyerik and tamiak) and Basil, ep. 104 (leuyra pograf
ba!ilik! oko!).
29
For the following cf. Caldwell, Between State and Steppe, 99104.
30
For mbaym! see P.Ness. 24 with Kraemers note on line 3 ( pace LSJ Rev.
Suppl. s.v.); cf. baym! in the meaning rank (G.W.H. Lampe, A Patristic Greek
Lexicon [Oxford: Clarendon Press, 1961] 28182); for prior as a military rank in the
Byzantine army see, e.g., P.Ness. 26:7 (570 cr) and 35:3 (sixth cent.); the priores
were organized in clubs (P. van Minnen, Prisoners of War and Hostages in Graeco-
Roman Egypt, JJP 30 [2000] 15960).
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 214
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 215
31
As M. Lehtinen thinks, because he was an mbaym!, recruit, in Inv. 43.
Since the title is a great honor, I doubt whether this unit or group had any recruits.
Hence, I follow T. Gagos in promoting Flavios Thomas to prior. domestici was also
the name of an elite corps of honor guards in Constantinople (organized in scholae).
See O. Seek, RE 4.64850 s.v. Comites 25; H.-J. Diesmer, RE Suppl. 11.111323
s.v. Protectores (domestici); Ch. Gizewski, DNP 3.739 s.v. Domesticus.
32
Fl(ouio!) #y[vm! Nikhfrou prvr tn | [kayv!ivmnvn dom]e!t(kvn) [to
K!trou Zadakayvn. But for his title prior, here given to him, see n. 31.
33
Inv. 24 refers to a local gure as Fl(ouio!) #Alian[! ] #m#egalopre(p!tato!)
kme! to y#e[o]u #k[on!i!tvrou].
34
Around 400, the military commander of Palestine (the dux Palaestinae) had native
cavalry stationed at Zadaqathon under his command; these elite units had been
extracted from regular legions (Not. dign. 1.34.24 p. 73 Seeck).
35
A !tratith! is mentioned in two other papyri (Inv. 25a and 40 r+v).
36
P.Petra I 710 (cf. n. 26) contains receipts for the payment of back taxes:
everything assessed plus extra taxes in crop, vine, meat, and in outstanding taxes,
pnta t paithy(nta) ka pikla!y(nta) n te genma(!i) ka on(o)kr(oi!) ka
l(oi)p(azomnoi!) (?) ede!in (for the tentative resolution of the last abbreviation,
cf. P.Flor. III 377:13 mhden! trou edouw loipazomnou). The rst two categories
refer to the annona militaris, i.e., the in-kind portion of the military pay. In Egypt, we
nd the categories grain, vine, meat, barley, and sta (n te !t ka onokroi!
ka kriyaxr, P.Flor. III 377:14, where other taxes follow. On the annona see now
F. Mitthof, Annona Militaris: Die Heeresversorgung im sptantiken gypten: Ein Beitrag zur
Verwaltungs- und Heeresgeschichte des Rmischen Reiches im 3. bis 6. Jahrh. n. Chr. (Papyrologica
is perhaps a prior. The passage was ingeniously reconstructed by
T. Gagos; Fl(avios) Thomas [ prior (or recruit)
31
of [the most devote]
domestici [of Kastron Zadakatha].
32
Two years earlier, though, he
was still a recruit (mbaym!; Inv. 43). If the reconstruction is essen-
tially correct, as I believe it is, then Thomas belonged to the Imperial
Honor Guards of Kastron Zadakatha. The title of this elite unit
must have been honorary,
33
as so many high titles at the time, yet
it was connected with active military forces at Kastron Zadakatha
that probably were attached to the dux (military commander) of
Palestine.
34
It is particularly signicant, that both ocers, Barakhos
as well as Thomas, carried the status symbol Flavios and, therefore,
belonged to the elite of the local population. Thomas was married
to Kyra, possibly a granddaughter of Theodoros.
35
This case close
to the end of the sixth century points to the social ties between the
local military and well-to-do people.
The continued presence of the military is doubtless connected with
the fact that, according to the new papyri, the annona militaris, the
tax that paid the in-kind component of the military salary, contin-
ued to be collected, just as in other provinces.
36
While in Palestina
these taxes in money and in kind were used for the maintenance of
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 215
216 rtrvio korxrx
both the regular Byzantine army and its Arabic allies ( foederati ),
37
it
seems that in the geographical area of Petra in the sixth century
they must mainly have proted the regular army stationed in this
area. In the Petra papyri, troops of the allied Ghassanids do not
appear. This needs to be taken up in a larger context and on another
occasion.
38
3. Multicultural Petra: Nabataeans, Arabs, Greeks, and Romans
The Petra papyri are written in Greek, as was stated above. Latin
occurs only once in a formulaic phrase.
39
The Greek naturally con-
tained a number of Latin terms transliterated into Greek. People
used, and the communal administration accepted, not only the Roman
land measure of the iugerum, but also Semitic land measures such as,
in their Grecized form, koriaai, !atiaai and kabiaai (above, sect.
2), measures that are also found in the Babatha archive and at
Nessana. In names of the elite, the Roman gentilic name and sta-
tus symbol Flavius was followed sometimes by historically Nabataean
names, but more frequently by names that primarily sounded Greek,
Roman or specically Christian. Already this mixture of names used
Florentina 32; Florence: Gonnelli, 2001); on on()kre(on) and kriyxuron as single
accounting categories for two dierent real items (wine/meat and barley/sta) see
Mitthof, Annona Militaris, 1.20812 and 2.560 (Katalog) no. 194. For the continued
levy of the annona under Arab rule, see M. el Abbadi, Annona militaris and riz
of Nessana, Atti del XVII Congresso Internazionale di Papirologia [Naples: Centro
Internazionale per la Studio dei Papiri Ercolanesi, 1984] 3.105762).
37
In alliances between the Byzantine government and the leaders of confederate
Arab tribes the emperor was obliged to specify the amount of annona to be given
to the Arabic leader and his troops. See P. Mayerson, Monks, Martyrs, Soldiers and
Saracens: Papers on the Near East in Late Antiquity ( Jerusalem: Israel Exploration Society
in Association with New York University, 1994) 344; cf. R. Scharf, Foederati: Von
der vlkerrechtlichen Kategorie zur byzantinischen Truppengattung (Tyche Suppl. 4; Vienna:
Verlag Holzhausens Nfg., 2001) 4548.
38
In the meantime, see Caldwell, Between State and Steppe, 11141.
39
Inv. 6a, also called P.Petra Daniel C. and Nancy E. Gamber ( J. Frsn).
Section 3 is by and large a selective summary of Daniels paper on P. Petra Inv.
10 and its Arabic, but presentscontrary to the oral presentation of this paper
reservations about the identication of some places in the town of Wadi Musa with
places mentioned in Inv. 10. Daniel also drafted the part on this section for the
joint Cincinnati paper, that in turn is the basis for the present paper (see n. 1). He
collaborated in these matters with Omar al-Ghul and Hani Ali Falahat. Also see
Caldwell, Between State and Steppe, app. 2, where he focuses on the Nabataean attes-
tations of the toponyms of Inv. 10. The commentary on roll 10 will oer further
discussions of the toponyms.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 216
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 217
40
K. Dijkstra, Life and Loyalty: A Study in the Socio-religious Culture of Syria and
Mesopotamia in the Graeco-Roman Period Based on Epigraphical Evidence (Religions in the
Graeco-Roman World 128; Leiden/New York/Kln: Brill, 1995) 31921. This and
the following examples are taken from Inv. 10 and will be discussed in the com-
mentary; see also Caldwell, Between State and Steppe, 16982.
41
This was brought to my attention by Glen W. Bowersock. For Tainos, the name
of an Arab, see Libanius, or. 24.6; Bowersock, Julian the Apostate (Cambridge, MA:
by the elite points to the culture of a city that once was the capi-
tal of the Nabataean kingdom, but that had come under Greek cul-
tural inuence and Roman power, and nally had converted to
Christianity. But a closer look points to a particularly strong Arabic
character of the language that the people of Petra and her hinter-
land spoke in sixth-century Petra, when at the same time they wrote
their administrative and legal papers in Greek. Nabataean already
contained strong elements of later Arabic, particularly the use of the
denite article al and a number of personal names that can be ana-
lyzed as Arabic. By the time of the Petra papyri, however, Nabataean
had long ceased to be used in written texts. The personal names
and toponyms of the Petra papyri must reect the language that
people spoke. They show characteristic features of Arabic such as
the use of the denite article al (as mentioned above), as well as the
formation of the diminutives and plural forms (the broken plural).
But decoding Greek transcriptions of Arabic is dicult, as we already
know from some terms, names, and toponyms in the Nessana papyri,
written around the same time as the Petra papyri. The sound of the
Greek language had changed over the centuries, aecting consonants
as well as vowels and thereby making it dicult to determine the
Arabic equivalent of Greek letters. Often several possibilities must be
tested, and ambiguity therefore remains.
The full study of these matters is outside the competence of the
Michigan team and must be left to Semitists. Here, I can oer only
a few examples. As was briey mentioned, some names used in the
large family of Theodoros are Nabataean and Arabic. Obodianos,
the name of Theodoros father and grandfather, is obviously derived
from Obodas, the Greco-Latin form of 'bdt. This name was borne
by two dierent Nabataean kings of the rst century ncr as well as
by a city in the Negev and, apparently, by a Nabataean god.
40
Dusarios, the name of a great uncle of Theodoros, was based on
the name Dushara (Dou!rh!), a Nabataean and Arabic deity. And
the name of Thaaious, Theodoross grandmother, apparently comes
from the name of an Arabic tribe.
41
Other names are Greco-Roman,
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 217
218 rtrvio korxrx
but may have been chosen because of their meaning in the other
language, for example, Alfeios (Lat. Alus rather than Greek Alfei!,
Nab. lfw, successor), Bassos (Lat. Bassus, Sem./Arab. bss, cat),
and Valens (Latin; probably w'l, he who takes refuge or w"l, stag,
plus Latin ending -ens). The Semitic derivations have long been sug-
gested by Semitists, because there is no other reason for the popu-
larity of these Latin names in the ancient Middle East.
42
Other Greek names in Theodoros family with a specically Greek
air were clearly chosen because of their Christian signicance. This
applies to his own name, Gift-of-God, a very common name, as
well as to the name of his paternal uncle Leontios, derived from
Greek lvn and Lat. leo, lion. Christian martyrs of these names
were widely worshipped and at Ammatha (ammm), the already
mentioned place SE of Petra, was a chapel of Theodoros (Inv. 64+65,
also called P.Petra Kenneth W. Russell and Wendy Steward, writ-
ten in 559). Nevertheless, as R.W. Daniel argues, the family may
well have been aware that the meaning of the two names were also
expressed by Arabic names: Nab./Arab. "Ausallh, gift of Allah
and "Asad, lion, respectively.
43
As was said earlier, Theodoros and
Leontios father had a name in the Nabataean tradition, Obodianos,
and another Leontios (Inv. 83; see n. 62) was so named by his father
Abdallas, Nab./Arab. 'Abdallh, Servant-of-Allah.
44
The names in
the family of Theodoros son of Obodianos show pride in being part
of several worlds: the local traditional Semitic, the Greco-Roman,
and the Christian.
Topographical names and the names of apartments are an addi-
tional source for personal names, even though they transmit names
used in a previous generation. Known from Nessana and common
Harvard University Press, 1978) 117. He refers to the Aramaic/Syriac ayyy, appar-
ently based on the name of the same tribe that came to be used for all nomadic
Arabs better known by the other name Saracens.
42
See F. al-Khraysheh, Die Personennamen in den nabatischen Inschriften des Corpus
Inscriptionum Semiticarum (Inaugural Dissertation, Philipps Universitt, Marburg/Lahn,
Fachbereich Aussereuropische Sprachen und Kulturen, 1986) 84 (for lfw); M. Sar-
tre, Bostra: Des origines lIslam (Bibliotheque archeologique et historique 117; Paris:
Librairie Orientaliste Paul Geuthner, 1985) 17273, 188, 22527; for further liter-
ature, see Daniel, P.Petra Inv. 10 and Its Arabic, 33940.
43
al-Khraysheh, Die Personennamen, 2829 s.n. "wyw and "ww; Sartre, Bostra, 183
s.n. A!ado!; 18687, also 199 and 202 s.nn. Dvro! and Zabdo!; Daniel, P. Petra
Inv. 10 and Its Arabic, 34041.
44
al-Khraysheh, Die Personennamen, 127 s.n. 'bd"lhy.
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 218
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 219
45
R.W. Daniel, Toponomastic Mal in P. Nessana 22 and P. Petra Inv. 10 . . .,
ZPE 122 (1998) 19596; cf. al-Khraysheh, Die Personennamen, 14445 s.n. 'mrw.
Another toponym formed with mal in Inv. 10 is Mal Orsiat.
46
usaina is Nabataean (see A. Negev, Personal Names in the Nabataean Realm
[Qedem 32; Jerusalem: Institute of Archaeology, Hebrew University of Jerusalem,
1991] no. 497), Laila is pre-Islamic. She is a gure in the prophet Muhammads
ancestry. See Daniel, P. Petra Inv. 10 and Its Arabic, 333; Caldwell, Between State
and Steppe, 175 line 73, and 177 line 96. In late Greek, Laila can be spelled Lela.
Another such name is May al-Louza.
47
The Greek name is in the nominative; in Roman and Byzantine times, the
nominative -io! appears often as -i!, especially in names (F.Th. Gignac, A Grammar
of the Greek Papyri of the Roman and Byzantine Time [Testi e documenti per lo studio
dellantichita 55; Milan: Istituto Editoriale Cisalpino-La Goliardica, 1981] 2.2527).
For the upper-story apartment, see at the end of this section.
48
There are four elds that have their own names, while usually the toponyms
cover areas of more than one eld. Out of a total of twelve resulting sown elds
in Ogbana with extant measures, ten are 1 iugerum, only one is 1' iugerum (the
integer may be larger) and one is 3 iugera. By comparison, of eighteen sown elds
with extant measures in Serila, only seven measure 1 iugerum, the rest are larger
(ve have 2 iugera, one has 2 or more iugera, three measure 3 iugera, and two have
5 iugera. In Ogbana, the abutting neighbors are only given for three plots, one of
in Petra are names beginning with mal, property, such as Mal
Amar al-%aroua, the property of Amar al-Sarwa, i.e., of 'Amer
or 'mir al-Sarwa. Amer/Amir is probably a common Semitic name.
45
Also the names of women appear in toponyms: May O!aina and May
Lela are the tract of land of usaina and of Laila, respec-
tively.
46
By contrast, Elliay Afyoni! is the upper-story apartment
of Aphthonios. The Greek name Aphthonis, which in Inv. 10 also
occurs in its fuller form Aphthonios, is here connected with the sta-
tus constructus of Arabic 'illyya in order to form the Semitic name
of the upper-story apartment.
47
In general, toponyms are Semitic, mostly Arabic, as one should
expect. A total of ve coincide with, or are similar to, toponyms
that are still used in Wadi Musa, as Daniel found with the assis-
tance of Hani Ali Falahat of the Antiquities Inspectorate, Petra (see
nn. 1 and 39). The modern names and their descriptive pattern illus-
trate the ancient toponyms. Here are the three closest pairs:
(1) Xaffay al-Aouaouer (Inv. 10:159f.)Kiat al-awwir,
(2) al-Oraem (Inv. 10:80, 157; cf. Inv. 8 el-Xoreum)uraim, and
(3) al-Ba!!a (Inv. 10:137)al-Baa.
According to the papyrus, the rst two places are in Ogbana, where
the three brothers divide mostly small and isolated plots of sown
land.
48
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 219
220 rtrvio korxrx
On (1): Kiat al-awwir is situated 250 m NW from the cen-
ter of the town (el-J")
49
on a steep, broadly terraced slope with
whitish, limestone-containing soil; near-by is a still running spring.
The name of this altogether fertile spot means the hand (lit. the
palm or the pan) of white earth. The place is now used for
orchards and elds. The name seems recondite, as Daniel called
it, but nevertheless descriptive of the place.
On (2): uraima name also attested for a village near Medina
is about 1000 m south of Kiat al-awwir, again in an elevated
location, now partially in agricultural use. The word, a diminutive
plural of arm, seems to refer to projecting patches of vegetation.
50
In Inv. 10, a piece of this land includes dryland and the other is
on three sides surrounded by roads (see n. 48). While this name is
more common than Kiat al-awwir, its meaning, present loca-
tion, and the indications of the papyrus produce a consistent
picture.
These two ancient and present-day toponyms and their present-
day location may indicate that ancient Ogbana
51
was in this area,
but two out of fourteen extant or partially Semitic names for land
in Ogbana can hardly carry the burden of proof. Nevertheless, these
two pieces of evidence for correspondence between the ancient and
current names are intriguing and, perhaps, indicative of the geo-
graphical identity of these places.
The third pair of correspondence between the ancient and the
current place is placed in Serila, a village, according to Inv. 10, rich
on vineyards and sown lands, houses, agricultural buildings, and dry-
gardens (orchards; see above, sect. 1), where most of the three broth-
ers agricultural property was located.
which has roads on three sides; another is bordered by dryland [?] on two sides,
and only one has real neighbors on all four sides including one of the brothers and
their grandfather). Cf. Caldwell, Between State and Steppe, 3738 and 168.
49
El-J" has not yet been found in the Petra papyri. For the identications of
ancient and current locations see Daniel, P. Petra Inv. 10 and Its Arabic, 33437
and Caldwell, Between State and Steppe, 178 on line 137, 175 on line 80, 177 on lines
97 and 98, 181 on lines 15960.
50
In Jordan, there also exist al-uraiyim and Wd al-uraiyim (formed as
diminutive of the diminutive). For the meaning see N. Groom, Arabic Topography and
Placenames: A Transliterated Arabic-English Dictionary with an Arabic Glossary of Topographical
Words and Placenames (Beirut and London: Longman, 1983) 142. One of his sug-
gestions is steppe corridor (adopted by Daniel).
51
This name is derived from the name of a person, family or tribe; cf. P.Euphr.
2:2 (D. Feissel and J. Gascou, Documents darchives romains indits du Moyen
SHIFFMAN_f11_200-226 12/11/02 9:08 AM Page 220
+nr rrcirnrnvrx+ .xr rri+iox or +nr rr+n. r.rvni 221
Euphrate [III
e
s. aprs J.-C.], Journal des Savants [1995] 8494); Caldwell, Between
State and Steppe, 3841; Daniel, P. Petra Inv. 10 and Its Arabic, 33637, n. 16.
52
No measures are given.
53
For the details see Daniel, P. Petra Inv. 10 and Its Arabic, 336 with n. 13.
On (3): al-Baa is almost at the bottom of the valley some 400
m SSE of Kiat al-awwir and perhaps 300 m south of the cen-
ter of the town of Wadi Musa. It is a lush, very fertile area with
orchards and gardens, abundantly watered by its own spring. As
Daniel reports, 40 m to the west, Nabataean or Roman piping was
found. Its name may be understood as the moist place. The name
can be applied to almost any place blessed with a spring, and springs
are very common in the area. In the passage of Inv. 10 that inter-
ests us here, one of the brothers (Sabinos) received an apparently
larger piece of sown land
52
in the area of Math Osaina and al-Bassa,
below the area of Ar[ ]. All this has no obvious resemblance to
what now is al-Baa.
Two other places that, according to the papyrus, were situated in
Serila, have been also suspected of belonging to the area of present-
day Wadi Musa, al-%ira (Inv. 10:18, 97, 98; al-ra) and al-Ber Am
al-Xaffa [ (Inv. 10:69;
Guwairat Umm al aa). The rst, al-ra,
situated on the top of a hill just about 250 m from al-Baa o the
center of Wadi Musa (el-J"), is named after a very common feature,
an enclosure for animals. The second,