Documente Academic
Documente Profesional
Documente Cultură
JOURNAL OF
Semitic
Studies
VOLUME LIV. NO. 1
SPRING 2009
INFORMATION FOR AUTHORS
Semitic
Studies
V OLUME LIV. NO. 1
SPRING 2009
Contents
articles
JAN KEETMAN, Wechselwirkung von Vokalen und Gutturalen im
Semitischen unter dem Einfluss anderer Sprachen: die Beispiele des
Akkadischen und Hebräischen 1
NA’AMA PAT-EL, The Development of the Semitic Definite Article:
A Syntactic Approach 19
GORDON J. HAMILTON, A Proposal to Read the Legend of a Seal-Amulet
from Deir Rifa, Egypt as an Early West Semitic Alphabetic Inscription 51
VINCENT DECAEN, Theme and Variation in Psalm 111: Metrical Phrase
and Foot in Generative Perspective 81
ROBERT HOLMSTEDT, Word Order and Information Structure in Ruth
and Jonah: A Generative-Typological Analysis 111
FRANÇOIS BRON, Notes sur les inscriptions néo-puniques de Henchir
Medeina (Althiburos) 141
TZVI NOVICK, The Modality of Òarik in Tannaitic Hebrew 149
SHLOMY RAISKIN, Talmudic Aramaic Fauna Names: Murzema and Shaqi†na 161
NADEZHDA VIDRO, A Newly Reconstructed Karaite Work on Hebrew
Grammar 169
GERRIT BOS and Y. TZVI LANGERMANN, The Introduction of Sergius
of Resh¨aina to Galen’s Commentary on Hippocrates’ On nutriment 179
ALESSANDRA AVANZINI, Origin and Classification of the Ancient South
Arabian Languages 205
AARON D. RUBIN, The Functions of the Preposition k- in Mehri 221
HILLA PELED-SHAPIRA, From Conventional to Personal, or: What
Happened to Metaphor under the Influence of Ideology The case of Gha’ib
Tu¨ma Farman 227
reviews
Georges BOHAS and Mihai DAT, Une théorie de l’organisation du lexique des langues
sémitiques: matrices et étymons, Philippe CASSUTO and Pierre LARCHER (eds),
La formation des mots dans les langues sémitiques (Edward LIPINSKI) 251
Aaron D. RUBIN, Studies in Semitic Grammaticalization (David STEC) 253
Jean-Jacques GLASSNER, Translated and Edited by Zainab BAHRANI and Marc
VAN DE MIEROOP, The Invention of Cuneiform: Writing in Sumer (Jon TAYLOR) 254
J.A. HALLORAN, Sumerian Lexicon: A Dictionary Guide to the Ancient Sumerian
Language (J.N. POSTGATE) 255
Stefanie U. GULDE, Der Tod als Herrscher in Ugarit und Israel (W.G.E. WATSON) 258
Hallvard HAGELIA, The Tel Dan Inscription: A Critical Investigation of Recent
Research on Its Palaeography and Philology (Bob BECKING) 259
Ernst JENNI, Studien zur Sprachwelt des Alten Testaments II (Martin F.J. BAASTEN) 261
Robert D. MILLER II S.F.O., Chieftains of the Highland Clans: a History of Israel
in the 12th and 11th Centuries B.C., Ann E. KILLEBREW, Biblical Peoples and
Ethnicity: an Archaeological Study of Egyptians, Canaanites, Philistines, and Early
Israel 1300-1000 B.C.E. (Eveline J. VAN DER STEEN) 265
Bernard S. JACKSON, Wisdom Laws: A Study of the Mishpatim of Exodus
21:1 22:16 (Jeffrey STACKERT) 270
Hilary LIPKA, Sexual Transgression in the Hebrew Bible (Jenni WILLIAMS) 272
Kevin A. WILSON, The Campaign of Pharaoh Shoshenq I into Palestine
(K.A. KITCHEN) 274
Thomas B. DOZEMAN and Konrad SCHMID (eds), A Farewell to the Yahwist?
The Composition of the Pentateuch in Recent European Interpretation
(William JOHNSTONE) 276
Alice HUNT, Missing Priests: The Zadokites in Tradition and History (Deborah
ROOKE) 278
Lena-Sofia TIEMEYER, Priestly Rites and Prophetic Rage: Post-Exilic Prophetic
Critique of the Priesthood (Deborah ROOKE) 281
Nicholas P. LUNN, Word-Order Variation in Biblical Hebrew Poetry:
Differentiating Pragmatics and Poetics (Robert D. HOLMSTEDT) 283
D. GOODBLATT, Elements of Ancient Jewish Nationalism (Doron MENDELS) 285
Beverly P. MORTENSEN, The Priesthood in Targum Pseudo-Jonathan: Renewing
the Profession (Simon ADNAMS LASAIR) 287
Alexander SAMELY, Forms of Rabbinic Literature and Thought: An Introduction
(Günter STEMBERGER) 289
Emanuela TREVISAN SEMI, Jacques Faitlovitch and the Jews of Ethiopia (Steven
KAPLAN) 292
Alessandro BAUSI and Alessandro GORI, Tradizioni orientali del «Martirio di
Areta». La prima recensione Araba e la versione Etiopica. Edizione critica e
traduzione (Michael A. KNIBB) 294
John C. LAMOREAUX (translator), Theodore Abu Qurrah (Hugh GODDARD) 296
Garth FOWDEN, QuÒayr ¨Amra: Art and the Umayyad Elite in Late Antique Syria
(Andrew MARSHAM) 298
Paul STARKEY, Modern Arabic Literature (Ami ELAD-BOUSKILA) 302
short notes
Steven E. FASSBERG and A. HURVITZ (eds), Biblical Hebrew in its Northwest
Semitic Setting: Typological and Historical Perspectives (J.F. HEALEY) 307
Hélène LOZACHMEUR, La Collection Clermont-Ganneau: ostraca, épigraphes sur
jarre, étiquettes de bois (J.F. HEALEY) 308
Karel JONGELING, and ROBERT M. KERR, Late Punic Epigraphy: an Introduction
to the Study of Neo-Punic and Latino-Punic Inscriptions (J.F. HEALEY) 309
Jean-Claude HAELEWYCK, Grammaire comparée des langues sémitiques: élements
de phonétique, de morphologies et de syntaxe (J.F. HEALEY) 310
Suzanne SCHOLZ, Introducing the Women’s Hebrew Bible (Deborah ROOKE) 311
JAN KEETMAN
ISTANBUL
Abstract
In this article it is argued that the sound shifts (*ai >) e > i and
pharyngeal + a > e > e (both long and short) from Imperial Akkadian
to Old Babylonian prove the existence of two vowel qualities of e in
Imperial Akkadian. This may well be explained if we regard the sec
ond sound shift as the shift of the pressing of the pharynx from a
consonant to a vowel, probably under Sumerian influence, resulting
in a vowel near a, like German ä, which may have existed in
Sumerian. While in Akkadian some consonants where lost but left an
imprint on the vowels, in Hebrew auxiliary vowels evolved which sus
tained the pronunciation of nearly the same consonants. While
Akkadian made a compromise, Hebrew resisted when its daily use
faded or was even reconstructed in some communities in antiquity.
a) Akkadisch
1
Hasselbach 2005.
2
Neben Stärken hat diese Arbeit leider auch zwei grundsätzliche Schwächen.
Die erste ist die Vernachlässigung der Personennamen. Zwar hat Hasselbach Recht
damit, dass das Material der Personennamen vom übrigen sprachlichen Material zu
trennen ist, doch sollte auch es möglichst vollständig gesammelt und für sich er
schlossen werden. Manche Probleme und auch Missverständnisse kommen daher,
dass diese Arbeit nicht unternommen wurde. Die andere Schwäche ist die zum Teil
ungenügende Rezeption ihrer Vorgänger. Man lese sich dazu z. B. die zum Teil
1
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
2
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
3
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
muss auch berucksichtigt werden, dass zwischen Sarkalisarri von Akkad und der
zweiten Halfte der Regierungszeit Sulgis von Ur eine Spanne von 100 oder 150 Jah
ren liegt, fur die das Akkadische nur außerst sparlich dokumentiert ist.
11
Dazu ausführlich und mit Erörterung der akustischen Wirkung, Keetman 2004.
12
Vgl. den Exkurs am Ende dieses Artikels.
4
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
13
Diese könnte, statt mit dem hinteren Teil der Zunge auch mit den Pharynx
muskeln erzeugt werden. Doch wir gehen von nicht absichtlich herbeigeführten
Effekten bei der Bildung der Vokale aus und betrachten deshalb nur die Bewegung
der Zunge.
14
Die zentralisierende Wirkung der Pharyngale auf Vokale belegt auch das
Patach Furtivum im Hebräischen (s. u.).
15
Zur Vokalharmonie innerhalb sumerischer Wortwurzeln vgl. Keetman 2005:
11 12.
5
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
16
Vor diesem kurzen ä ist natürlich ein Stimmabsatz zu denken, der hier der
Übersichtlichkeit wegen ausgelassen wurde.
17
Tropper 1999 schlägt eine andere Erklärung, nämlich Metathese des
Pharyngals vor. Eine Metathese gerade dieser Konsonanten kurz vor ihrem Ausfall,
die im wesentlichen auf das Babylonische beschränkt bliebe, ist jedoch nicht sehr
wahrscheinlich. Vgl. auch Keetman 2004: 11 mit Anm. 14.
18
GAG3 §9b.
19
Hasselbach 2005: 120.
20
Steinkeller 2004: 12 14.
6
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
21
VS 25, 41 v 10; AWL 43 iii 8; 64 ii 4; Nik I 22 vii 2 passim.
22
Ukg. 4 vii 26 = 5 vii 9. Vgl. Selz 1998: 324 mit Anm. 195 und Literatur dort.
23
ELTS 15 i 24; ii 23; iii 21 passim.
24
Gelb, Steinkeller und Whiting 1991: 237 8 stellen fest, dass ENGAR hier für
ENGAR.US steht und dass ihnen und dem „(field) scribe“ bei Hausverkäufen der
„master house surveyor“ und der „street herold“ gegenüberstehen. Daraus schließen
sie, der ENGAR(.US) müsse „a high administrative official in charge with
agricultural activities“ sein. Andererseits ist die Bedeutung US = ús „angrenzen, be
nachbart sein“ altsumerisch gut bezeugt (Behrens und Steible 1983: 361 2). Neh
men wir nun eine Parallele zwischen dem „master house surveyor“ (um mi a lú é és
∞ar) und dem „(field)scribe“ (dub sar( gána)) an, so steht der „street herald“ (ni∞ir
sila) dem engar ( ús) gegenüber. Während die ersten beiden Berufe mit Vermessung,
bzw. Niederschrift zu tun haben, dient der ni∞ir sila wohl der Bekanntmachung der
neuen Verhältnisse im Viertel. Auf dem Land gibt es weniger Nachbarn und es
reicht vielleicht, wenn man sie einfach bittet, doch beim Kauf anwesend zu sein.
25
ELTS 15 xii 20; xiii 4.
26
Vgl. Di Vito 1993: 94, mit Hinweis u. a. auf sargonisch EN ì lí, be lí
DINGIR und Ur III EN.DINGIR.MU. Als ältere Form zum präsargonisch in
Lagas belegten ì lí be6 lí ist beli ili (bzw. ba‘li ili) wegen der Veränderung der Wort
stellung zu erwarten. Siehe Di Vito 1993: 289 90. Die einzige reale Möglichkeit
für einen Irrtum wäre, dass uns ein „banana name“ einen Streich spielt. Zwar gibt
es in ELTS 15 einen ähnlichen „banana name“ Ki lí lí, doch findet sich bé/be6 li li
auch in Fara, wo diese Namensbildung ziemlich unublich ist.
27
Cf. Sommerfeld 2003: 572.
28
NTSS 569 Rs.(?) iii 6’.
7
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
29
Der im gleichen Text Rs.(?) i 3’ genannten Namen ha li li ist wegen ha lí
lum, WF 22 ix, vermutlich als hal ili „Mein Gott ist ein Onkel (mütterlicherseits)“
zu deuten. Für ähnliche Namen vgl. AHw 314b oben.
30
RTC 12 iii 2.
31
Ad 3, 17 19: BE lí É wa a ti [l]i /is me\ „Mein Herr möge mein Wort hö
ren!“ Das letzte Wort ist stark beschädigt, aber nach der Kopie hinreichend sicher.
32
Kienast undVolk 1995: 43; cf. Michalowski 1993: 28 30.
33
Sommerfeld 2003: 585; ähnlich Westenholz 1999: 33. Allerdings spielt bei
beiden Autoren auch noch der Ubergang aj > e statt i wie im Altbabylonischen eine
Rolle bei der Trennung des Reichsakkadischen vom lokalen Akkadisch Babyloniens.
34
Eine ähnliche Analyse erwägt auch Hasselbach 2005: 232, zieht aber nicht
den Schluss, dass mithin die angenommene Vorreiterrolle der Diyala Region, nicht
zu beweisen ist.
8
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
9
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
Die E-Färbung ist daher besser von dem Ort her zu erklären, an
dem die beiden Pharyngale ursprünglich artikuliert wurden, als
durch eine hypothetische und zugleich weite Verschiebung eines die-
ser Laute. Ein Gleitlaut mag dabei als Folge der Vokalfärbung an
manchen Stellen hinterher entstanden sein.
Außerdem kann man, ausgehend von Kouwenbergs Beobachtung,
dass ¨ain mit dem Stimmabsatz zusammenfiel, argumentieren dass
deshalb Verben, die ursprünglich h oder Ì enthielten, einem anderen
Paradigma angeglichen wurden. Das wäre eine Verteilung auf ein be-
reits vorgegebenes System entsprechend dem Kriterium: Unterbre-
chung des Luftstromes oder keine Unterbrechung.
Die Palatalisierung von Ì steht auch im Widerspruch zum ortho-
graphischen Befund der reichsakkadischen Texte. Wegen des bereits
beginnenden Übergangs a > E (ä) wäre hier j für Ì insbesondere in
Gegenwart von a zu erwarten.
Die anschließende Tabelle gibt einen Überblick über den Zeichen-
gebrauch:42
’V A E Ì Ù, Ú
¨V A, Á E Ì Ú
hV A, Á, É - - -
ÌV É E E -
jV Ì+A E I, E? U (Ú)
Das Fragezeichen nach E für [ji] stammt von Hasselbach. Ú für [ju]
wurde vom Autor eingeklammert, da er es für wahrscheinlich hält,
dass die wenigen Texte, die das Präfix ju- als ú- schreiben, den begin-
nenden Lautwandel [ju-] > [u-] dokumentieren, der schon in der Ur
III-Zeit ganz vollzogen ist.43 Da dieser Wandel offenbar auch die Le-
sung des Zeichens I von ji zu i verändert, ist es möglich, dass der be-
ginnende Abfall von j zuerst beim Präfix [ju-] zu beobachten ist.44
42
Cf. Hasselbach 2005: 34. E für *hi wurde als unsicher ausgelassen. Vgl. ibid.
79. Nach ibid. 81 wurde E für *Ìe ergänzt (bei Hasselbach wohl ausgelassen, weil Ì
hier reichsakkadisch wahrscheinlich bereits ausgefallen ist).
43
In Ur III Texten gibt es keine für [ji], [ju] reservierte Zeichen mehr. Dazu
Hilgert 2002: 120 1, cf. GAG3 §22c*.
44
Die spätere Festlegung von I auf [i] beruht natürlich auch auf der Beobach
tung, dass I nicht nur beim Präfix, sondern auch an anderen Stellen für [i] gebraucht
wird und älteres ì hier weitgehend verdrängt. Doch das mag mit Verzögerung gesche
hen sein oder es ist bei den wenigen Texten, bei denen der Abfall von anlautendem j
eintritt, nur zufällig nicht bezeugt, was gut möglich ist, da i am Silbenanfang weit
10
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
Unsicher ist U für *Ìu weil der einzige Beleg li-ip-te-u-ma (ptÌ) sehr
wahrscheinlich als liptajuma „sie sollen öffnen“ mit einem nach dem
Abfall von Ì und Färbung des Vokals entstandenen Gleitlaut zu in-
terpretieren ist.45
Die Tabelle zeigt, dass das Reichsakkadische j relativ klar von allen
übrigen Zeilen trennt. Die einzige Ausnahme ist das zugleich am
wenigsten spezifische Zeichen E. Man könnte einwenden, dass
Kouwenberg die Palatalisierung nur nach Vokal zwingend verlangt hat-
te. Doch dann brauchen wir einen anderen sprachlichen Mechanismus,
um Ìa > E (ä) zu erklären und wenn ein solcher existiert, gibt es keinen
Grund mehr, die Palatalisierung nach einem Vokal anzunehmen.
Das etwas ältere Syllabar der Ebla-Texte ist weniger systematisch
aufgebaut als das reichsakkadische Syllabar. Davon heben sich die le-
xikalischen Texte mit ihrer mehr einheitlichen Schreibweise ab. Au-
ßerdem ist die Interpretation in den lexikalischen Listen zuverlässiger
als bei einsprachigen Texten oder gar in den Eigennamen. Daher
scheint es geraten das Syllabar der lexikalischen Texte zum Vergleich
heranzuziehen.46
’V A, Ì - Ì Ù
‘V A - Ì Ù
hV É, A - I U9
ÌV É, Ì - I, Ì U9
jV A - I U9, U4
seltener ist als das Verbalpräfix ji bzw. i . Dazu kommen orthographische Besonder
heiten, wie die noch lange übliche Schreibung von ili „mein Gott“ als NI.NI = ì lí.
45
Cf. Hasselbach 2005: 81 und Kouwenberg 2006: 151 mit der gleichen Inter
pretation.
46
Die Tabelle folgt der Zusammenstellung bei Rubio 2006: 115 16. Auf einige
unsichere Lesungen wurde verzichtet. Der Lautwert ji für I ist auch lexikalisch belegt
und zwar in dem Wort jidum, das in Ebla offenbar nicht wie sonst im Akkadischen
für „Arm“, „Seite“ steht, sondern (noch) wie westsemitisch jad, „Hand“ bedeutet.
Vgl. die Belege bei Pettinato 1982 unter VE 515, VE 531, VE 557, VE 012 und
kontrastierend ì sa tù (isatum) „Feuer“, VE 783. Für die interessante Gleichung VE
802: den líl = i li lu (besser: ji li lu) vgl. den möglichen Gebrauch von EN zur
Schreibung von [je/in], cf. Sommerfeld 1999: 20 Anm. 29. Auf die strittige Etymo
logie dieses Namens kann hier nicht eingegangen werden. Siehe zuletzt Sommerfeld
2006: 74. Cf. auch VE 799a/b dEN.ZI/ZU sú i nu (= sú ji nu) für den semitischen
Namen des Mondgottes (ganz genau: [tsujin] oder [tsujain] > [tsujin]?).
11
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
c) Hebräisch
12
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
13
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
Schließt der Pharyngal direkt an einen Vokal an, so sind ein Luft-
strom von der Lunge und schwingende Stimmbänder vorhanden,
weswegen tatsächlich kurz a hörbar wird. Steht der Pharyngal hinge-
gen am Silbenanfang, so ist die Bildungsphase nicht hörbar. Das Lö-
sen der hinteren Enge mag leichter fallen als die Bildung der Enge,
weswegen vor Vokal kein PataÌ furtivum zu hören ist.
Sieht man in PataÌ und dann auch in PataÌ furtivum kein a, son-
dern eher ein ä, so lässt sich dies durch Abschwächung der für die
Bildung einer hinteren Enge nicht gebrauchten, aber für a wichtigen
vorderen Öffnung erklären.50Wir wären dann in einer Situation ähn-
lich dem Akkadischen, nur dass die im Akkadischen anzunehmende
weitere Vorverlagerung zu einem in der Mitte zwischen a und i ste-
henden e durch Segol blockiert war. Segol selbst wird von ∑ere am
weiteren Ausweichen nach vorne gehindert.
Bei der Segolierung, bei der es keinen Grund gibt, den Vokal an-
ders als passend zu wählen, kann man im Prinzip den gleichen Effekt
wie beim PataÌ furtivum beobachten, wobei auch hier die Wirkung
vor dem Pharyngal größer ist.
Einwenden könnte man, dass הeigentlich kein Pharyngal ist. Doch
es ist möglich, dass הleicht pharyngalisiert gesprochen wurde, insbe-
sondere in einem sprachlichen Umfeld, in dem dieser Konsonant
nicht üblich war, was die Tendenz zu einer Art Betonung gefördert
haben mag. Vielleicht erklären sich so auch einige der gelegentlichen
Unregelmäßigkeiten beim Landwandel a > e im Akkadischen.51
Weil bei den Îatef-Vokalen nicht sogleich ein Vokal folgt, der
praktisch ein Loslassen der Zungenstellung bewirkt, tritt hier die Be-
günstigung eines Vokals nahe a bei den Pharyngalen und h wohl
ebenfalls ein. Wohl weil es mit Schwa praktisch kaum noch wahr-
nehmbar wäre, wurde auch אeinbezogen.
Um das Auftreten von QameÒ Ìa†uf unter den Îa†ef-Vokalen in un-
serem Modell zu erklären, muss man annehmen, dass es aus Redukti-
on von u (oder o) entstanden ist. Von u nach a steigen beide Formanten
kontinuierlich an. Eine hintere Enge hebt beide Formanten und ver-
schiebt einen hinteren Vokal also in Richtung a. Kommt hinzu, dass
durch die Flüchtigkeit des Lautes und den angenommenen Gebrauch
der Zunge zur Bildung der hinteren Enge, die Mundstellung von u
wohl auch insgesamt nicht mehr ganz eingenommen werden kann. Ein
kurzes å könnte sehr wohl das Ergebnis dieser Veränderung sein.
50
Ich lasse die Frage offen, denn die Nichtverwendung von QameÒ lässt sich
auch damit erklären, dass QameÒ in unbetonten, geschlossenen Silben å und nicht a
vertrat.
51
Für solche Unregelmäßigkeiten vgl. Kogan 2001.
14
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
52
Poebel 1931: 3.
53
Dazu ausführlich Keetman 2005.
54
Vorläufig Keetman 2005. Zwei weitere Artikel des Autors hierzu sind in Vor
bereitung, einer in WdO.
55
Smith 2007.
15
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
BIBLIOGRAPHIE
56
Trotz der von sumerisch é gal „Palast“ abgeleiteten westsemitischen Wörter
mit h kannte das Sumerische sehr wahrscheinlich kein h oder einen ähnlichen Laut.
Cf. Keetman 2006.
57
Sommerfeld 1999: 18 19; 2003: 572.
58
So Poebel 1931, Keetman 2005 und etwas anders formuliert auch Smith 2007.
59
Jedenfalls beim Plural m. im casus obliquus. Beispiele bei Sommerfeld 1999: 19.
16
DIE BEISPIELE DES AKKADISCHEN UND HEBRÄISCHEN
Keetman, J. 2004. ‚Der Verlust der „Kehllaute“ im Akkadischen und der Lautwan
del a > e’, Altorientalische Forschungen 31, 5 14
2005. ‚Die altsumerische Vokalharmonie und die Vokale des Sumerischen’.
Journal of Cuneiform Studies 57, 1 16
2006. ‚Gab es ein h im Sumerischen?’ Babel und Bibel 3, 9 30
Kienast, B und K. Volk. 1995. Die sumerischen und akkadischen Briefe. (FAOS 19.
Stuttgart)
Kogan, L. 2001. ‘*g in Akkadian'. Ugarit Forschungen 33, 263 7. (Ergänzungen in
UF 34)
Kogan, L. und K. Markina 2006, Rezension zu Hasselbach 2005. Babel und Bibel
3, 555 88.
Kouwenberg, N.J.C. 2006. ‘The Proto Semitic Gutturals in Old Assyrian’, in G.
Deutscher und N.J.C. Kouwenberg (Hrsg.) The Akkadian Language in its
Semitic Context, Studies in the Akkadian of the Third and Second Millennium BC
(Leiden). 150 76
Michalowski, P. 1993. Letters from Early Mesopotamia, translated by Piotr
Michalowski, edited by Erica Reiner. (Atlanta, Georgia)
Pettinato, G. 1982. Teste lessicali bilingui della bibliotheca L. 2769. (MEE 4, Napoli)
Poebel, A. 1931. The Sumerian Prefix Forms e and i in the Time of the Earlier
Princes of Lagas. (AS 2, Chicago)
Rubio, G. 2006. ‘Eblaite, Akkadian and Eastsemitic’, in G. Deutscher und N.J.C.
Kouwenberg (Hrsg.) The Akkadian Language in its Semitic Context, Studies in
the Akkadian of the Third and Second Millennium BC (Leiden), 110 39
Smith, E. 2007. ‘[ ATR] Harmony and the Vowel Inventory of Sumerian’. Journal
of Cuneiform Studies 59, 19 39
Soden, W. von 1995. Grundriss der Akkadischen Grammatik. (3. Auflage. AnOr 33.
Rom.) (= GAG3)
Sommerfeld, W. 1999. Die Texte der Akkade Zeit. I. Das Diyala Gebiet: Tutub.
(Münster)
2003. ‘Bemerkungen zur Dialektgliedrung, Altakkadisch, Assyrisch und
Babylonisch’, in G. Selz (Hrsg.) Festschrift für B. Kienast. (AOAT 274, Mün
ster). 569 86
2006. ‘Die ältesten semitischen Sprachzeugnisse eine kritische Bestandsauf
nahme’, in G. Deutscher und N.J.C. Kouwenberg (Hrsg.) The Akkadian
Language in its Semitic Context, Studies in the Akkadian of the Third and Second
Millennium BC (Leiden). 30 75
Steinkeller, P. 2004. ‘On the Writing of belum in Sargonic and Earlier Sources’.
Nouvelles Assyriologiques Brèves et Utilitaires 2004/1, 12 14
Tropper, J. 1999. ‘Zur Etymologie von Akkadisch sukênu/suhehhunu’. Welt des
Orients 30, 91 4
Vito, R. di. 1993. Studies in Third Millennium Sumerian and Akkadian Personal
Names. The Designation and Conception of the Personal God. (Studia Pohl SM
16. Rom)
17
NA}AMA PAT-EL
HARVARD UNIVERSITY
Abstract
1. Introduction
* Several people have commented on earlier versions of this paper. I would like
especially to thank John Huehnergard for his continued support and interest and for
his many thought provoking comments. I would also wish to thank Wolfhart
Heinrichs, Geoffrey Khan and Aaron Rubin for their helpful comments. Any errors
are mine alone. Abbreviations used in this paper are: Adj adjective; AdN adnominal;
Akk. Akkadian; Amh. Amharic; Arb. Arabic; Arm. Aramaic; BH Biblical
Hebrew; C Christian; CA Classical Arabic; Can. Canaanite; CS Central
Semitic; Def. definite; Dem demonstrative; ES East Semitic; G G¢¨¢z; IE
Indo European; Indef. indefinite; J Jewish; LateA Late Aramaic; M Muslim;
MidA Middle Aramaic; MH Mishnaic Hebrew; N noun; Nor. Norwegian;
OA Old Aramaic; OfA Official Aramaic; OSA Old South Arabian; Ph.
Phoenician; Poss. = possessive suffix; PS = proto Semitic; P WS proto West Semitic;
TO Targum Onqelos; Ug. Ugaritic.
1
In this paper, the term ‘definite’ refers to the grammatical marker, and not to its
pragmatic or semantic features, unless specifically noted, as the emphasis of the cur
rent paper is syntax. The term ‘demonstrative’ is a deictic expression; ‘attributive/
19
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
20
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
Ug.: spr hnd ‘this book (m.acc.)’ (2.19, 9); ’alpm sswm hnd ‘these 2000
horses (m.pl)' (2.33, 32); ml’akty hnd ‘this mission (f.) of mine’
(2.33, 35)
Arm.: dmwt’ z’t ‘this statue’ (Fekh. 15)6
Arb.: hada l-malik / l-malik ha∂a ‘this king’
zayd ha∂a ‘this Zeid’
¨ibadi ha’ula’i ‘these servants of mine’
In Arab. Dem-N is permissible only when the noun is marked with
the definite article, which is a late feature, at least later than proto-CS. In
any case, the Dem-N word order is conditioned by the syntax of the
noun.7 G. is under heavy influence of non-Semitic languages, about
which we know very little.8 Thus, since the preferred word order in
both ES and WS is N-Dem, it should be considered both a PS and a P-
WS feature, while the article is only common CS (Rubin 2005: 65).
This is a major weakness in the assumption that the article developed
from an attributive demonstrative. Note also, that according to Rubin’s
reconstruction, the Canaanite and Aramaic articles both go back to the
same demonstrative (*han), yet the article is post-positive in Aramaic
and pre-positive in Canaanite.9 While the syntax of the article is con-
sistent in all the CS languages, which no doubt points to a similarity in
the development of their article, the differences in its position cannot be
ignored.10
6
This word order changed by LateA and later to become Dem. N. See Pat El
(forthcoming).
7
The situation in OSA is hard to evaluate. There are not many examples of the
attributive demonstrative and the information regarding their behaviour in different
syntactic environments is therefore partial.
8
The order of N modifier in G. is variable, that is adjectives, genitive and relative
clauses may be positioned before, or after their head noun; however, in modern Ethio
Semitic, most languages exhibit modifier N word order, although some variation is
possible (Little 1974: 79, fig. 7). This is a strong indication that there has been sub
stantial shift of nominal modifiers from post position to pre position. Some elements
are more likely to gain a fixed position than others. Already in G. quantifying adjec
tives, such as b¢zuÌ ‘much’, and ordinal numbers are regularly pre posed: sab¢¨awit
hagar ‘the seventh city’. Thus, the pre position of the demonstrative in G. is a result of
a process of change, rather than an indication of the original situation.
9
A similar point was made by Lambdin (1971: 316). Firmage’s rejection (2002:
37, n. 14) of Lambdin in this regard should not be accepted. Lambdin is correct in
noting that demonstratives appear after the noun in Semitic. Firmage uses Arabic as
evidence that Lambdin is wrong, but Arabic is a clear innovation, and a late one at
that, and even there we find the demonstrative regularly after the noun.
10
It is well known by now that the order of morphemes does not necessarily
reflect their word order prior to grammaticalization (Harris and Campbell 1995:
199 200); however, in this case, it seems that the development of the article is much
later than the stabilization of the demonstrative word order.
21
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
11
Once more, Lambdin is right to note that ha is not an independent demonstra
tive in Hebrew and Phoenician (1971: 315). Here too, we must reject Firmage’s dis
missal of Lambdin who asserts that we simply do not have enough early material in
these two languages to make such a claim (2002: 37, n. 14). We do indeed have
comparative evidence which clearly corroborates Lambdin’s claim.
12
Lieberman (1986: 591) argues that this form is constructed out of several ele
ments: ’an , the first element in independent nominative pronouns (*’an a, ’an taˆ,
’an ti etc.) and ni, an Afroasiatic pronoun. For Hasselbach (2007: 25) *han is a prefix,
while Rubin (2005: 76) claims that the origin of this form is a nisbe *hanni, which was
reduced when cliticized. All these scholars agree that the Akkadian form is not an
inflected *han, but either a derivation of it or a complex form with a prefix *han.
13
Starting from MidA (Cook 1992: 10). See more below.
14
I would like to thank Rebecca Hasselbach for making her paper available to me
pre publication.
15
Even if we assume that the Targumic Aramaic ’ar¨a ha ∂a is an imitation of the
Hebrew ha ’areÒ haz zo(’)t, it still behoves us to note that while the Hebrew definite
article haC is represented by the Aramaic definite article a, the Aramaic translator
clearly identified haC before a demonstrative in Hebrew with ha in Aramaic.
16
Other forms (hndt, hnk, hnkt) appear in unclear contexts.
17
In MH the head noun is not necessarily marked with an article: s¢losa d¢ßarim
hallalu ‘these three things’ (Av. 5:19). See discussion in section 3.4.
22
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
ze seper tol¢∂o† ’a∂ am (Gen. 5:1) ‘this is the book of the history of
man’
Akk. (*han >) annûm is primarily used attributively, though not exclu-
sively.
Given that *ha(-n) is prefixed to attributive elements, it is strange
that in Aram. and OSA, *ha(-n) is a suffix.18 Hasselbach also notes that
these forms’ inflection, if it exists at all, is secondary (Hasselbach 2007:
21). If indeed these forms were not inflected, it is hard to regard them
as pronominal.
While the phonological arguments for reconstructing a form *ha(-n)
as the underlying form of the article are solid, there are some indica-
tions that *ha(-n) could not have been a demonstrative. Demonstratives
by definition mark degrees of proximity (Diessel 2005: 170–1). Yet
the element *ha(-n) not only lacks such a distinction19, it may also be
cliticized to various elements indiscriminately:20 the proximal demon-
strative in CA (ha∂a) and Akk. (anni-); the distal demonstrative in BH
(hallaze); to both in TO (ha-hu, ha-∂a), Syriac (hanna, haw < ha-w),
and Mand. (hahu, hax).21
A slightly less perplexing problem is the lack of inflection. The defi-
nite article in CS does not exhibit gender-number distinctions.22 This
is quite puzzling since even languages which have lost gender distinc-
tions in the nominal paradigm (e.g. Dutch23) still maintain gender dis-
18
Even Hasselbach who supports Rubin’s analysis notes that ‘already in early Se
mitic there existed the possibility to mark the demonstrative pronouns as adnominal
by affixing the element *ha(n)’ (2007: 21). With no evidence to the contrary, ‘affix
ing’ should be understood as prefixing.
19
Hasselbach (2007: 21) argues that n is probably a near deixis element in Se
mitic, but ha itself does not seem to have any such attribute.
20
This lack of proximity distinction is found with the relative determinative pro
noun as well; the determinative is, of course, not a demonstrative.
21
In addition, from the very beginning, the article and the demonstrative co
occur: Arb. ha∂a al malik ‘this king’, BH ha ’ama haz zo(’)t ‘this maidservant’ (Gen.
21:10). In a process of grammaticalization of Dem>article, there is usually a period
where the presence of the article excludes the demonstrative; this did not happen in
CS. In most Indo European languages it is still the case, and the demonstrative and
definite article are in complimentary distribution: *this the king.
22
Zaborski (2000) argues that the definite article was originally inflected for gen
der and number (m. ha n, f. ha t, pl. ha l), but subsequent assimilation levelled the
differences. However, there is no direct evidence that this is so and the entire premise
lies heavily on the assertion that the article must have its origin in a demonstrative.
23
The adnominal categories in the Germanic languages were more gender dis
tinctive than the nominals. This is especially evident in the article and demonstratives
(Dekeyser 1980: 98). While in English the noun, the definite article and even the
indefinite article (slightly later) lost all gender number case distinctions in Late Old
English, the demonstrative maintained a basic number marked paradigm. In Dutch
23
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
tinctions on the article; in fact in many languages the rise of the article
is assumed to be related to the collapse of the nominal system which
leads to the article carrying the relevant nominal inflection.24 In
Semitic, with its rich nominal morphology, there is no reason that the
attributive demonstrative should lose its inflection as it becomes a defi-
nite article.25
Thus, it would be highly problematic to reconstruct the CS definite
article from a demonstrative. The only way that a non-inflecting
adnominal prefix *ha(-n) could have been the origin of the CS article
is if it was attached to the attribute first, and only later to the noun by
analogy, as it is quite unlikely that an ad-nominal prefix will be at-
tached to a substantive. This point is crucial to the understanding of
the syntax of the article and its later developments in Semitic.
While many of the syntactical features of the article are shared, there
are some native, and apparently old, patterns in each language which
do not quite fit what we think we know about the article’s syntax. The
importance of relics to historical reconstruction has been highlighted
repeatedly by historical linguists dealing with morphology,26 but is
where the distinctions were lost much later than English, the demonstrative and defi
nite article are the only forms which maintain a distinction between genders (com
mon de and neuter het).
24
Lyons (1999: 67) mentions only one inflected language (Hungarian) where the
article is frozen, while the demonstrative and noun show inflection. However, Hun
garian has fewer nominal features than Semitic (e.g. it has no gender). Note also that
the innovated article in Amharic distinguishes gender: m.s.: l¢g u ‘the boy’ / bära w
‘the ox’; f.s: l¢g itwa ‘the girl’ / lam wa or lam itu or lam itwa ‘cow’; c.pl.: n¢gus occ u
‘kings’ (Leslau 1995: 155 6). The tendency of articles in world languages to carry
more grammatical load than the noun is so strong that it was the basis of a theory that
the innovation of articles is a way to ensure that the nominal phrase will carry the
grammatical markers, even when the morphological expression is lost (Lyons 1999:
324). There are of course known exceptions, as English, where the noun distinguishes
number but the article does not.
25
Compare, for example, the definite article in Central Neo Aramaic, which origi
nated from pronouns, where distinctions between the genders were not levelled: ™ur.
b u gab ano (Jastrow 1992, 6.10:8) On this side (m.); b i xaÒra†e (Jastrow 1992,
12.13:11) On this peg (f.). Typological studies have shown that attributive
demonstratives are the last grammatical entities in the nominal system to lose inflec
tion. Moravcsik’s proposed universal is: if the adnominal adjective agrees with the
noun, so does the adnominal demonstrative (1997: 317). Nevertheless, it is possible
that in the process of grammaticalization, a demonstrative may lose its inflection. The
difficulty in Semitic is that even outside the ‘article’ position hn does not show inflec
tion.
26
Hetzron (1976: 92ff.) labelled it ‘the principle of archaic heterogeneity’, ac
cording to which the relatively most heterogeneous system must be the most archaic.
Hetzron used this principle mostly in comparative historical linguistics to determine
the relative archaism of one system over the other. This, however, may very well apply
24
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
25
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
26
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
§178; Rubin 2005: 82), while true, does not by itself account for the
fact that only the second element is marked with the definite article;
pronouns and nouns do not have the same syntax nor the same level of
definiteness (Khan 1984: 470).
3. Relics
3.1 Arabic
3.1.1 Improper Annexation (al-’i∂afa gayr al-Ìaqiqiyya)
Improper annexation is a special construct which contains an adjective
as the regens and a noun as the rectum: Ìasanu l-waghi ‘beautiful (lit.
pretty of appearance)’ (cf. BH y¢pa† to’ar). In this pattern, the rectum is
normally marked with the definite article, yet the construct is not syn-
tactically definite, i.e. an adjective modifying it would not be marked
with an article and attributive demonstratives are excluded. In a regu-
lar type of annexation, whenever the rectum is marked with an article,
the annexation is syntactically definite. However, in the improper an-
nexation, if the construct is to be definite, the article must precede the
regens, i.e. the adjective, as well as the rectum:32
ga¨du s-sa¨ari ‘curly of hair, a person with curly hair’
al-ga¨du s-sa¨ari ‘the curly of hair, the person with curly hair’
It is hard to understand why a definite rectum does not mark the
entire construct as definite as it does with any normal construct pat-
tern.
3.1.2 The Relative Pronoun alla∂i
The Arabic relative pronoun (alla∂i, allati) is a combination of the
definite article al-, a particle la and the PS relative-determinative pro-
noun *∂u (Akk. sa, BH zu, Aramaic zi). There is no doubt that the
32
This is not the case in Hebrew, where the improper annexation imitates the
syntax of the regular annexation: s¢ßa¨ hap paro† y¢po† ham mar’e ‘the seven healthy
cows’ (Gen. 41:4). Note, however, that most examples in the Bible are predicative
and there are in effect very few examples of the improper annexation with a definite
article.
27
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
3.2 Hebrew
3.2.1 The Definite Article as a Subordinating Particle
The definite article may appear on participles whose head noun is not
definite:
s¢pipon ¨ale ’oraÌ han-noseÈ ¨iqq¢ße sus (Gen. 49:17) ‘a snake on a road
biting a horse’s heels’.
b¢-ya∂ mal’aÈim hab-ba’im Y¢rusalayim (Jer. 27:3) ‘by the hand of mes-
sengers coming to Jerusalem’.
The lack of article on the head noun causes great difficulty to
Hebraists. Davidson states that in this pattern the head noun is defi-
nite, and even when it is not formally marked, ‘the preceding word is
really definite’ (1902: 133, §99).
33
This basic pronoun is still in use in some Modern dialects, like the Zbala dialect
in Morocco. See Brustad (2000: 109 10).
34
Hopkins (1984: 240, §288 n. 1) notes that the cases where an asyndetic relative
clause follows an intrisincally or formally definite antecedent are much rarer than
alla∂i clauses following an indefinite noun.
28
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
35
See Gesenius (1910: 447i); Goshen Gottstein (1946: 29); Peretz (1967: 75,
108 9); Waltke and O’Connor (1990: 339d). There are some examples of finite
verbs; however, all the examples with finite verbs occur in late books, except Josh.
10:24: q¢Òine ’an¢se ham milÌama he hal¢Èu(’) ’itto (Josh. 10: 24) ‘officers, men of
war, who travelled with him’. This pattern with finite verbs disappeared all together
in later stages of Hebrew, while the participle with the definite article is used from the
earliest sources until Israeli Hebrew. In addition, note that almost all the finite verbs
found preceded by the definite article begin with h, and thus may be explained as
cases of dittography. Therefore, it is possible that the pattern with finite verbs is a late
analogy to the pattern with participles and developed after the participle became a
part of the verbal paradigm, thus ’aser+predicative participle: ’aser+ finite verb::
haC+participle:? > haC+finite verb.
36
That is, there can never be an overt subject in this ‘relative clause’ and there can
never be constructions of the type **mal’aÈimi hab ba’imj ’elehemi ‘messengers to whom
people come’. The Arabic na¨t sababi pattern allows the attribute to refer to a different
referent than the head noun. However, the na¨t is clearly a nominal pattern, as its
negation is the nominal negator gayr rather than the verbal negators ma/la (Polotsky
1978: 169, 171 2). Note also that the predicate of the alleged relative clause is as
29
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
attached only to the participle and never to any other element. All
these are substantial differences between the relative clause and the
participle clause, the latter being highly restricted compared to ’aser
relatives.37 Besides the lack of agreement in definition, the N-partici-
ple combination is syntactically identical to the N-Adj. combination
(for which see section 3.4).
In addition, there is no reason to treat an article on participles as
having a function distinct from the article on substantives. Therefore
hqdsm in h’lnm hqdsm cannot be a sentence, if h’lnm is not a sentence.
Moreover, if Gzella’s analysis is correct then the PS relative-determina-
tive *∂V could have similarly transferred to the substantive from its
attributive position, but this did not happen.38
Relative clauses, like adjectives, are exponents of the attributive rela-
tion, but this does not imply that their internal syntax is identical to
that of the adjective, nor that the article is identical to the relative
pronoun, as I hope the discussion above illustrates.
signed case by the head noun, not by its function within the clause, that is, it bears
regular nominal agreement: …l adwiyati l gayri muÒarraÌn bi bay¨i ha qanunan… ‘drugs,
the sale of which is prohibited by law’ (Qanun as sur†a 36, apud Polotsky 1978: 171,
e.g. 40).
37
Siloni (1995: 448, n. 3) acknowledges the possibility that participles can func
tion as nouns or adjectives, but maintains that they are verbal in this pattern, because
they take direct objects, while nouns cannot take direct objects. This approach is a bit
circular, as Siloni takes this to be a given, while this is what needs to be proven and in
any case, examples where participles take direct objects when they are non verbal
exist: ki m¢ragg¢lim ’e† ha ’areÒ (Gen. 42:30) ‘as spying over the land’; l¢ no†¢rim ’e†
piryo (Song 8:12) ‘people guarding its fruit’. Siloni’s sole reason to analyse the article
as a relativizer is her inability to explain the lack of article on the head noun, despite
the fact, that even according to her parameters, there is only one minor similarity
between relative clauses and participle clauses (see her table in p. 452). It is also re
grettable that Siloni’s data and analysis are so misleading. Many of the examples she
provides as ungrammatical may be grammatical with other lexemes; she uses sen
tences with the presentative hinne which show a different syntax than regular nomi
nal sentences in the language. She ignores some important facts concerning partici
ples in Hebrew, e.g. participles can occur without definite article in Modern Hebrew,
only when they have been fully substantivized, thus qonim ‘shoppers’ (her example) is
acceptable, but not ’oÈlim ‘diners’, mitnaÌalim ‘settlers’ but not yosßim ‘residers’.
Finally, she refers to Arabic na¨t sababi construction to prove that in some languages,
the participle may predicate a non subject element. Here too the participle is no
different than the attributive adjective and merits no special consideration. Moreover,
Siloni ignores the clear indication of case in Arabic, which undeniably points to a
nominal agreement with the head noun.
38
In some Neo Arabic dialects, the definite article is used as a relativizer, probably
due to contamination with the relative pronoun illi; this is, however, a secondary and
late innovation and should not be taken to represent the original function of the
Arabic article.
30
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
Note also that like the Arabic improper annexation, here too, the
pattern is not definite unless the head noun is also definite:
¨al ham-mizbeaÌ hab-banuy (Judg. 6:28) ‘on the already-built altar’.
ha-¨alma hay-yoÒe† lis’oß (Gen. 24:43) ‘the maiden going out to draw
(water)’.
This pattern occurs in Mishnaic Hebrew with the same syntax:
indefinite noun is followed by a definite participle (e.g. Er. 9:4). If
indeed the function of the particle *han was to mark definiteness, a
definite article on the attributive participle should have been enough
to render the entire construction definite.
39
Note that this syntactic re interpretation goes against the meaning of the clause:
yawmu s sabi¨i is ‘the seventh day’, not ‘the day of the seventh N’, daru l aÌirati is ‘the
other world’, not ‘the world of the other N’ and so on. In addition, note that in the
noun and adjective in this pattern are always in agreement, which is not the case for
the construct: (Lebanese Arb) sint il ma∂ye ‘last year’.
31
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
32
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
43
Blau (ibid.) notes that since the article on the adjective is the one marking the
opposition between predicative and attributive, the article on the noun was felt to be
redundant. Note also the positioning of the article on the noun alone is much rarer
(Blau, ibid §240).
44
See also Rhodokanakis (1911: II §89a) for Omani Arabic. In Maltese, the pat
tern is common mainly in toponyms and adverbials: wied il kbir ‘the great valley’;
blata l bajda ‘the white rock’. See Borg (1989) for details and discussion.
45
Grotzfeld (2000: 10) notes that speakers of Lebanese Arabic tend to ‘correct’
this pattern in writing according to what they think is normative, thus adding l
before the noun, while in speech the pattern with no l before the noun is productive.
Similarly, my colleague SN explained ktab ¢z zgir ‘small book’ in his Southern Leba
nese dialect as ‘incorrect fast speech’. See also Borg (1989: 78, n. 19) who quotes a
local scholar explaining a similar pattern in Maltese as ‘orthographical abbreviation’.
46
See also Grotzfeld: ‘adjektivische Attribute, die unterscheidenden Wert haben,
sind gewöhnlich wie untergeordnete Substantive konstruiert: selber mit Artikel bei
einem Beziehungswort ohne Artikel’ (1965: 93).
47
Note also Borg (2000: 30) who suggests that BH patterns such as Òom ha ¨asiri
33
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
• Iraqi Arabic
The pattern is quite common and productive in the Christian and
Jewish dialects of Baghdad, but can be found also in the Muslim dia-
lect, though there it occurs mostly in fixed combinations (Blanc 1964:
126; Mansour 1991: 44).48 The tendency to reanalyse it as a construct
is even more pronounced in this dialect, as can be seen in examples
with feminine nouns, where the construct differs morphologically from
the non-construct. All Iraqi dialects use the pattern for place names,
like bab es-sarji ‘the Eastern Gate (a name of a city quarter)’, and some
recurrent expressions, such as id el-yemna ‘the right hand’49 as well as
for temporal expressions, such as sahr ej-jay ‘next month’.
C: ¨id le-kbig ‘the great feast (Easter)’
qondart el-lexxi ‘the other shoe’
J: Òlat l¢-jdidi ‘the new synagogue’
dfater el-¨ettaq ‘the old note-books’
M: ¨id ec-cebir ‘the great feast (Feast of the Sacrifice)’
¨enab l-aswad ‘the black grapes’
The regular relative construction is le-blad s-sefnaha ‘the country we
saw’. However, in all dialects, though particularly in the Jewish dialect,
there are examples where the head noun is not marked (Blanc 1964:
127). Here too, the pattern is occasionally reinterpreted as a construct.
In the Christian and Muslim dialects the pattern is again common
with temporal expressions, while in the Jewish dialect it is very pro-
ductive:
C/M: sant ej-jina ‘the year we came’
waqt el-kent bej-jays ‘the time you were in the army’
J: sen es-suwwet-u ‘the thing you did’
ßaÌur ella∂i qa-yezzawwaj ‘the young man who is getting mar-
ried’
ma-ynam be-mkan el-yebgadlu ‘he doesn’t sleep where [lit. in a
place that] he’ll be cold’
Since this pattern occurs in CA and other modern dialects of Ara-
bic, Blanc (1964: 128) cautiously suggests that the Jewish and Chris-
(Zech. 8:19) may be normal constructs with two nouns and an attributive adjective,
i.e. not the tenth fast, but rather the fast of the tenth (month).
48
Geoffrey Khan noted (in a personal communication) that in the Eastern Arabic
dialects one cannot exclude a possible influence of a Persian substratum, where the
izafe pattern is similar to the pattern discussed here.
49
Note Syriac ’i∂eh da ymina ‘the right hand’ where the N Adj combination is
reinterpreted as a genitive construction.
34
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
tian dialect preserve traces of an older usage, while the Muslim dialect
shows widespread levelling.
3.4.2 Hebrew
The BH article may precede the attribute of a formally unmarked noun
(Lambdin 1971: 321; Waltke and O’Connor 1990: 260). Note that
the noun is not in the construct state, thus, unlike Arabic, this pattern
in Hebrew is synchronically understood as a noun-adjective construc-
tion:50
seßa¨ paro† ha†-†oßo† (Gen. 41:26) ‘the seven healthy cows’
qane ha†-†oß (Jer. 6:20) ‘the good cane’
ÌaÒer hag-g¢dola (1 Kg. 7:12) ‘the big courtyard’
ruaÌ ha-ra¨a (1 Sam. 16:23) ‘the evil spirit’
As in Arabic, this pattern is quite common with ordinals (Brockel-
mann 1961: II 209, §132c–d):51
miy-yom ha-rison ¨ad yom has-s¢ßi¨i (Exod. 12:15) ‘from the first to the
seventh day’.
’el mabo has-slisi (Jer. 38:14) ‘the third entrance’.
¨a∂ m¢qom sa¨ar ha-rison (Zech. 14:10) ‘as far as the place of the first
gate’.
Lambert (1895) notes that the word yom is always unmarked with
an article when followed by an ordinal number, while the number is
always marked. There are only two exceptions, both in late books (Neh.
8:18 and Dan. 10:12). In the vast majority of the cases the word yom is
preceded by a preposition, where the definiteness is only revealed in
the vocalization. Lambert (1895: 280) further claims that the masoretes
added the article wherever they could, without changing the conso-
nantal text, although originally no article was there. The original pat-
tern is, according to Lambert, a construct.
This situation is even more widespread in MH, where it seems that
when the article is used, the combination N-definite Adj. is more com-
mon than definite N-definite Adj (Sarfatti 1989:156).52
50
The consonantal text makes some examples ambiguous, e.g. b¢ dereÈ ha† †oßa
w¢ ha ysara (1 Sam. 12:23) ‘the good and upright way’.
51
Note that examples, where the noun is marked with article or possessive suffix
and the ordinal is not, are not definite: ha ros ’eÌa∂ (1Sam. 13:18). Similarly: ’aÌiÈem
’eÌa∂ (Gen. 42:19) ‘one of your brothers’, despite the suffix on the noun.
52
There is also a pattern where only the noun is marked with the article, however
in Sarfatti’s corpus there were only 7 such occurrences, compared to 63 of the pattern
under discussion. Sarfatti (1989: 157) suggests that the pattern where only the noun
is marked was a substandard form. Ben Hayyim (1992: 434) analyses the adjective in
35
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
36
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
3.3.5 Summary
All the languages surveyed above, except perhaps Phoenician, exhibit
this pattern as part of their system, not as an accidental form, despite
the fact that it is in contradiction with the regular agreement rules in
these languages.
Several explanations have been given to account for its existence:
compositum (Feghali 1928; Borg 2000), construct (Blau), relative (Ben
Hayyim 1992; Gzella 2006) and syntactic elative (Grotzfeld 2000).57
The assumption of compositum is not in accordance with Semitic syn-
tax: if the entire combination is understood as one word, why is the
article medial and not prefixed? The construct explanation is not in
accordance with the meaning of the pattern, and the regular gender-
number agreement between the noun and the adjective indicates that
the adjective modifies the noun and is not dependent on it. The claim
for a relative origin I have dealt with above. Grotzfeld’s suggestion is
interesting, but it does not resolve the historical origin of the pattern.
Analysis of the data presented above indicates that the article’s original
function was not to mark definiteness and that it belongs with the
attribute. The following is a summary of the arguments, based on the
examples and discussion above.
57
Borg’s explanation may also be closer to the one made by Grotzfeld. Borg (1989,
2000) claims that this pattern did not originate from the nominal construct and is
independent of it. He suggests that the pattern with what he calls ‘initial article’, [N
article Adj.], antedates the regular pattern [article N article Adj.], and was used for
specific combinations in order to mark the entire phrase as one lexeme.
37
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
58
Note that in some languages, a noun marked with an article is semantically
equivalent to a noun with possessive suffixes: Aram. ’aba, ’imma = ‘the father/mother,
my father/mother’ (similarly with brother).
38
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
59
The plural endings are a case in point: some form of broken plural is known to
exist in proto WS, but in NWS, excluding a group of relics, all nouns take regular
adjectival endings. Furthermore, the anding * at for f.pl. and * un for m.pl. are pro
ductive and predictable on nouns in NWS, despite the fact that nouns normally do
not have predictable plural forms. The Semitic feminine ending t/ at also likely origi
nated from the adjective. It is most likely that in PS substantives were not morpho
logically marked for gender, e.g. Arb. ’umm ‘mother’, BH ’eßen ‘stone’ and pairs such
as Arb. gamal ‘camel’ / naqa ‘she camel’, where natural gender is marked lexically.
60
Even Amharic, which uses the suffix pronoun to mark definiteness, does not
mark both noun and adjective with the article.
61
Arabic has fsg agreement with inanimate plural nouns of both genders, while
collective nouns with singular morphology usually take plural adjectives, though they
may occasionally be found with singular adjectives (Reckendorf 1921: 59). Some
sporadic examples of lack of agreement in case are also found. See Reckendorf (1921:
60 §41.8).
62
The article is not strictly speaking needed to nominalize adjectives. Yet, beyond
a small set of adjectives, it is hard to find an adjective without an article outside the
attributive position. Some adjectives may be fully substantivized, ex. BH Òaddiq ‘right
eous man’ and then their syntax is like that of substantives, i.e. no article is needed.
63
Note that even in Amharic the definite article nominalizes adjectives: t¢ll¢q u
‘the big one’ (Leslau 2000: 49 §44.5).
39
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
5. Origin
The most likely origin of the article is a deictic particle *ha and its
derivatives, *han and *hal, as is suggested by Hasselbach (2007).66 One
or both of these particles exist in all the Semitic languages and mostly
function as presentatives:67
Akk. allû, annû (anna?)
Amarna. allû, annû ‘presentative’ (Rainey 1988)
BH. hinne, hen (he’?)68
64
The pattern with PS *∂u should be reconstructed to Proto Semitic (Pat El and
Treiger 2008).
65
In Arabic, the article excludes nunation, but it does not determine the case of
the form; the case is determined by the syntax of the noun, regardless of definiteness
or the presence of the article.
66
Rubin (2005:75) emphasized that *han and *’ul are not etymologically re
lated. He also did not connect any of them to *ha. Other scholars, like Tropper (2001),
assumed a phonological process *han > *hal.
67
The development presentative > article is found elsewhere as well: In Creole
French, the French presentative la is used as a definite article, while the old article le/
la/les is not used as a definite article (Grant 1995). Sissala, a Niger Congo language,
spoken in Burkina Faso and Ghana, uses a derivation of the verb ‘to see’ as a definite
article (Blass 1990: 197). This form also functions as a presentative and is attached to
demonstratives to indicate spatial distance (Blass 1990: 198 9). Finally, in French
the demonstrative ce/cette/ces are compound forms which originate from an attach
ment of an uninflected Latin presentative ecce to demonstratives: ce < ecce hoc, cet/
cette < ecce iste, celui/celle < ecce ille (Harris 1980: 147).
68
Brown (1987) suggests that halo’ represents two separate particles: ha+lo’ (inter
rogative with the negative particle) and *halu’ etymologically related to Aram. hlw/’lw
and Akk. allû. He notes that translating halo’ as ‘is it not? nonne?’ cannot be applied
40
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
69
Arm. h’/ha/, ’alu (BA), hlw (EgA) (Muraoka and Porten 1998: 329)
Ug. hl /halli-/,70 hn
Arb. hal, ’inna
(G. h¢yya ‘there’ perhaps connected to BH he, na-yya ‘here (I am)’, nahu
‘presentative’)71
This function of the presentative may be attested in Ug. as well,
though Ug. did not develop an article. There are examples of a relative
pronoun preceded by the prefiix ha:
’anykn dt l’ikt mÒrm hndt b Òr mtt (KTU 1.19.10 13) ‘The group of ships
that you sent to Egypt, which was wrecked off Tyre’.
Pardee analyses hndt as an independent demonstrative referring to
the f. noun ’any (2003/2004: 136) against Tropper’s reading it as a
locative adverb (230). Pardee takes hnd to be the basic demonstrative
in Ug. However, it is quite possible that hnd and its derivatives are
themselves complex entities: hn-d, hn-dt, as in other Semitic languages,
and as can be seen in other Ug. deictic particles, like hnk etc.72 This is
also evident in the fact that hn is a presentative, and only when it takes
a demonstrative extension, like d, k, m, it is a demonstrative pronoun.
This pattern occurs with substantives as well:73
hn alpm sswm hnd (RS 16.402 31 2 and 37 8) ‘these 2000 horses’.
Both Tropper (2000: 233) and Pardee (2003/2004: 125, n. 455)
suggest that hn here functions as a locative adverb. This is possible, but
consistently to all the biblical examples (e.g. 1 Sam. 20:37; Ps. 54:2; Prov. 26:19).
Brown (1987: 219, n. 103) further suggests that the ’ was added to *halu’ because it
was confused with the homophonous halo’.
69
To this Brown (1987: 211 12) adds ’aru. The semantic similarities between
hlw/’lw and ’rw should not be confused with etymological connection; therefore,
I see no reason to include this form in the list. Another possible form is Syr. harka
‘here’, which Rubin (2007: 123) suggests is derived from hn ka, which underwent
*#n > #r (Testen 1985).
70
See also hlk, hlny (in some contexts interchangeable with hnny) in the opening
formula in epistolary usage (Brown 1987: 203 4; Tropper 2000: 738; Pardee 2003/
2004:365), hlm (Tropper 2000: 797, Huehnergard and Pat El 2007: n. 55). For
evidence of doubling see syllabic spelling in Huehnergard (1987: 33): al li ni ya.
71
Na/na is a very common presentative in G. and Ethio Semitic, and later
became the suppletive imperative ‘come’ in Amh. For more forms see Leslau (1987:
380 1).
72
Pardee himself suggests that the forms hnhmt (RS 15.128:8) and hnmt
(94.2965:20) are a combination of presentative hn and independent pronoun.
73
Note the similarity to the syntax of OA and Phoen., described in Lambdin,
where the definite article mostly appears with attributive demonstrative and relative
clauses.
41
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
may still show us an early stage of the development which led in other
languages to the grammaticalization of the article.74
The presentative hn may also precede a noun followed by an at-
tributive demonstrative: hn bns hw ‘that servant’ (RS 96.2039:8, 10,
15). In one example, the spelling hbt ‘(the) house’ (RS 29.093:16)
either points to an assimilation, or to the use of ha, instead of han.75
The connection between the prefix ha and the presentative is appar-
ent in other patterns as well. In Huehnergard and Pat-El (2007), we
have shown that Arab. hada and BH hinne ze have very similar syntax,
which we argued is cleft, where the presentative is marked as the predi-
cate:
ha-da bnuka wajadnahu bi-’a¨la makkata (Ibn Hisam Sirah 113) ‘Here is
you son whom we have found in the mountain of Mecca’.76
w-hinne-ze mal’aÈ nogea¨ bo (1Kg 19:5) ‘Then an angel touched him’.
This pattern is used to predicate the excitement of the speaker, simi-
lar to Fr. voici que, voila que, which also shows the pattern [Presenta-
tive]-[Subordinate clause]. In both languages, the presentative may be
syntactically separated from the relative pronoun, though in Arabic
this separation is less common than in Hebrew.77
6. A Proposed Scenario
74
In this I take Pardee’s analysis against Tropper’s (2000: 233, §42.73 4). While
Tropper argues that hn (which he vocalizes as hannV ) in the examples quoted above
functions as a demonstrative, Pardee (2003/2004: 135, 365 and elsewhere) insists
that there is no substantiation for a demonstrative function as ‘hn functions uniquely
as a presentative particle in Ugaritic, and the same is true of the expanded forms hnn
and hnny’ (p. 365). Pardee (ibid, n. 494) further suggests that French voici works best
for hn.
75
Pardee is certainly right that there are too few examples in Ug. to substantiate
Tropper’s claim for an article.
76
In Christian Arabic, ha∂a is regularly used as a presentative (Blau 1966: 463,
§363).
77
Note that in the article cited above, which dealt with the Semitic determinative
*∂V, we have claimed that the patterns are similar syntactically, though using
etymologically different elements. My claim here is that these elements are not
etymologically different.
78
This was already alluded to in Lambdin (1971: 324).
42
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
43
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
81
This feature of the determinative pronoun may be the reason why it was never
transfered to the noun like the article.
44
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
45
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
85
We may perhaps speculate that the return of the attributive prefix ha is related
to the process of article weakening. This process is evident in Eastern Aramaic, but
may have had its origins already in MidA. Thus, the prefix ha may have been added
to demonstratives to further strengthen their deictic features.
This process may be compared to a certain innovation in Creole French, where in
a large part of the nominal lexicon, the original French article was agglutinated to the
following noun and understood as a part of it (la vie > lavi ‘life’, les yeux > lizye ‘eyes’).
These languages then innovated an article based on the deictic particle la, which is
suffixed in all forms of French: Fr. le chien > Mauritian Creole lisye ‘a dog’, lisye la ‘the
dog’ (Grant 1995).
Posner (1997: 384) suggests that this process in Creole French can be seen as a
culmination of the process that led to the creation of the French definite article.
Possibly as a result of the collapse of the French case system, ille, a semantically bleached
demonstrative, was increasingly being used as a noun marker. Ille became the default
marker in the absence of other determiners. Posner posits that this is merely a redistri
bution of elements, rather than a structural change.
86
Note for example that more conservative dialects, like Q, do not use this prefix
(Cook 1992: 16). There is one occurrence of hdn in 4QAmramb, but since the text
after it is fragmentary it is hard to evaluate its exact meaning and analysis.
87
Partial list: JPA hdyn, hhyn, hhw’, hhy’; SamA hdh, hdyn, hhw’, hhy’; CPA hlyk,
hlyn, hdn, hd’, hlyn; Syr. hana, ha∂e, halen; Mand. hazin, haze, halin; JBA hada,
haden, ha’ilen.
46
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
syntax of the article in CS, I have argued that the article was first at-
tached to a non-predicative form, and was then reanalysed as an
adnominal marker. The existence of two patterns, one with a marker
and one lacking it, led to the understanding of the prefix as marking a
definite nominal as opposed to non-definite noun with ø-prefix. This
process happened with the Semitic relative-determinative, whose origi-
nal function had nothing to do with definiteness, but was reanalysed
as marking a relative clause with a definite head. These arguments are
substantiated by empirical data and explain the existing syntax and the
relic syntax.
I have further suggested that the article was derived from the pre-
sentative, which is a non-predicative particle. The presentative, or its
reflexes han/hal, exists in all the languages. In Ug., which did not de-
velop an article, the presentative hn may be attached to relative pro-
nouns and demonstratives, a stage that I have argued preceded the full
development of the article.
While Huehnergard’s claim that the article is not a common CS fea-
ture stands, we can see the beginning of the process described above in
Ug. although it did not lead to the creation of an article there. How-
ever, since the preliminary stages exist already in Ug., the use of ha or its
reflexes to mark attributes should be considered a common CS feature.
REFERENCES
47
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
48
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
49
THE DEVELOPMENT OF THE SEMITIC DEFINITE ARTICLE
50
GORDON J. HAMILTON
HURON UNIVERSITY COLLEGE
Abstract
51
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
52
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
53
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
1. The Artefacts
pp. 330 1 for a select bibliography on this relocated artifact (EA 70881 in the British
Museum). Based on the floruit of Lahun late in the Middle Kingdom, the use of this
type of weaving instrument in loom types known from that era (and not those docu
mented from the New Kingdom), and the minimal state of the development of its
incised Proto Canaanite alphabetic forms from Egyptian scripts, I would agree with
the most probable dating for this object proposed by Quirke (ibid. 92), c. 1850 1700
BCE (Origins, 296 9). The combination of those elements, including the more objec
tive radiocarbon dating results, precludes the possibility that consonantal alphabetic
writing began some four to five hundred years later, in the late fourteenth century
BCE, as recently argued by Sass, ‘Genesis of the Alphabet’.
6
See especially T.G.H. James, ‘Ancient Egyptian Seals’, in D. Collon (ed.), 7000
Years of Seals (London 1997), 33 4, 37. For multiple examples of these button shaped
seal amulets, see G. Brunton, Qau and Badari vol. 1 (London 1927), pls 32.37 96;
33.97 108. For a vestigial continuation of the older type of cylinder seal, see the fine
faience seal of Queen Sobekneferu of the late Twelfth Dynasty (photo: Bourriau,
‘Second Intermediate Period’, opposite p. 192).
7
See Darnell et al., Two Early Alphabetic Inscriptions, 83, and Origins, 328, on the
crude ankh sign inscribed beside Wadi el Îol Text 2.
8
Concerning the longevity of scroll and spiral motifs on scarabs, W.A. Ward (Studies
on Scarab Seals vol. 1 [Warminster 1978], 55) wrote: ‘All the basic patterns, ulti
mately used in varying degrees of complexity, which underlie the unlimited variations
of the Twelfth Dynasty and later, existed during the First Intermediate Period’. More
recently, see D. Ben Tor, ‘The Relations between Egypt and Palestine in the Middle
Kingdom as Reflected by Contemporary Canaanite Scarabs’, IEJ 47 (1997), 165,
n. 19. For examples of spiral motifs on scarabs from Egypt assigned to the Middle
Kingdom, see ibid. figs 1.4, 1.9, 1.16; of the double ankh motif, figs 1.5, 1.13 14.
Both motifs continue in the Palestinian series (ibid. figs. 3 6), the emergence of which
Ben Tor (ibid. 164) dates to the Thirteenth Dynasty.
54
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
With this seal Petrie (Rifeh, pl. 23.1) found a set of seventy-nine
small beads, probably of glazed ‘faience’.9 Since each of these ball beads
is also perforated, in all likelihood they formed a necklace surrounding
the perforated seal-amulet when worn by its owner (pl. 1). The entire
set of beads and button-seal was restrung with modern string at the
Petrie Museum of Egyptian Archaeology (pls 1, 2). The reconstruction
of this arrangement of beads and seal has an impact on how one un-
derstands the direction of writing originally intended for its legend.
Figure 1
Drawing of UC 51354
From Petrie, Rifeh, pl. 23.1
Scale: Actual Size
Figure 2
Drawing of UC 51354
Traced from a transparency of a digitized photograph (see pl. 2)
and collated with the original
9
Dr Quirke noted that while these beads have not been tested and are badly
soiled, glazed ‘faience’ would be their more probable material (email correspondence,
April 23, 2007). Concerning all of the beads found in these cemeteries, Petrie (Rifeh,
13) made only a summary statement: ‘But the greater part were purely of the XIIth
dynasty style of ball beads of blue or green glaze, carnelian, or amethyst. A few strings
of small garnet beads were found…’.
55
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
3. Palaeography
10
The positions of the ankh signs provide no clue concerning the direction in
which to read this text, since one of them would be upside down in the horizontal
arrangement supposed by Petrie (fig. 1 above) and both would be situated sideways in
the vertical arrangement that I am proposing (fig. 2). See R.H. Wilkinson (Reading
Egyptian Art: A Hieroglyphic Guide to Ancient Egyptian Painting and Sculpture [Lon
don 1992], 177) regarding the ankh sign as a symbol for ‘life’ and ‘a potent amulet
throughout all periods of Egyptian history’.
11
I am indebted for this interpretation to the arrangement of these objects in the
photograph taken by Ana Dori of the Petrie Museum (pl. 2). Following Egyptian
practice, early alphabetic letters could be arranged either as vertical columns or hori
zontal lines, although the former is much more commonly attested (see the figures in
B. Sass, The Genesis of the Alphabet and Its Development in the Second Millennium B.C.
[ÄAT 13, Wiesbaden 1988] and the drawings in Origins, 324 99).
12
The long standing issue of whether West Semites borrowed hieroglyphic and
hieratic forms of a limited number of Egyptian signs to use as forms of the letters in
their consonantal alphabet may now be answered affirmatively. Concerning the two
new inscriptions from Wadi el Îol, see: G.J. Hamilton, ‘W.F. Albright and Early
Alphabetic Epigraphy’, NEA 65 (2002), 39, n. 14; Darnell, ‘Die frühalphabetischen
Inschriften’, 165 71; Sass, ‘Genesis of the Alphabet’, 150; Darnell et al., Two Early
Alphabetic Inscriptions, 75 87; and Dobbs Allsopp, ‘Ancient Southwest Scripts’, 496
7. See Origins, 29 268 for a detailed reconstruction the nearly complete repertoire of
specific Egyptian forms that were adopted, most of which can be ascertained with
certitude in part on the basis of evidence from those recent discoveries or other relo
cated epigraphs, the Lahun Heddle Jack (n. 5 above), Sinai 375a and 375c (ibid. 1,
374 8 [new photographs: G.J. Hamilton et al., ‘Three Recently Relocated Early West
Semitic Alphabetic Texts: A Photographic Essay’, Maarav 14 (2007), pls 1 8]).
13
Various terms have been used to name these scripts and inscriptions, e.g. ‘Proto
Canaanite’; ‘Old Canaanite’; ‘Proto Sinaitic’; ‘proto alphabetic’; ‘early alphabetic’ (see
56
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
57
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
the brief overview in Origins, 4). I shall employ the term ‘Proto Canaanite’ in refer
ence to early alphabetic scripts and epigraphs written before c. 1400 BCE, no matter
whether they were found in Egypt, the western Sinai, or the southern Levant (with no
connotations intended regarding the ethnicity of those who wrote them) and ‘Old
Canaanite’ for those dated to between c. 1400 to 1050 BCE.
14
I shall employ the Hebrew acrophones, simply because they are the best known
to those interested in West Semitic epigraphy. See Origins (29 253) for a reconstruc
tion of the original letter names that integrates several epigraphic and traditional sources:
an incomplete list of clipped names from Ugarit; mostly clipped forms twice tran
scribed into syllabic cuneiform on a recently published tablet in The British Museum
(F.M. Cross and J. Huehnergard, ‘The Alphabet on a Late Babylonian Cuneiform
School Tablet’, Or 72 [2003], 223 8); recently published transcriptions of several of
the full Aramaic acrophones into cuneiform; and the letter names transmitted in Greek,
Ethiopic, several levels of Hebrew and Aramaic, and Arabic traditions. With few ex
ceptions, these names can now be shown to correlate closely with the earliest graphic
forms of the letters and their antecedents in the Egyptian sign list (summarized in
Origins, 283 9; contrast the minimal reconstruction recently proposed by W.W. Hallo,
‘Again the Abecedaries’, in C. Cohen, A. Hurvitz, and S.M. Paul [eds], Sefer Moshe:
The Moshe Weinfeld Jubilee Volume: Studies in the Bible and the Ancient Near East,
Qumran and Post Biblical Judaism [Winona Lake 2004], 285 302).
15
The sources of the drawings of these Egyptian attestations, the majority of which
were done by Egyptologists (whose accuracy was checked whenever possible against
the originals), will be cited on a letter by letter basis below. The asterisks preceding
two Egyptian forms indicate tracings of modern fonts by myself. For a fuller listing of
Egyptian hieroglyphic and hieratic forms as they pertain to these and the other early
alphabetic letters, see Origins, 29 268, especially fig. 2.77, ‘Summary Script Charts’
(254 68).
16
Apart from two lameds excerpted from Sinai 357, the drawings in the two col
umns on the right were done by the present writer, in most instances based on an exami
nation of the original inscriptions (the sources for these tracings, mostly from new pho
tographs of the glass negatives of the originals in the collection of the Harvard Semitic
Museum, are listed in the figures of Appendix 1, Origins, 324 98; any others will be
referenced below). The numbers in the second column from the right refer to Proto
Canaanite inscriptions from the western Sinai, often termed the ‘Proto Sinaitic’ texts.
58
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
59
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
352, a steliform panel from the same site.22 One of the Proto-Canaanite
lameds on Sinai 352 also shares with the proposed l from Deir Rifa an
upward thrust to its largely horizontal stem. That stance would repre-
sent a relatively small development away from the positioning of hier-
atic writings of this letter’s source in Egyptian scripts, V1.23 The only
difficulty in identifying this letter as a lamed is the exact point at which
the right end of the loop joins the upwardly thrusting line. One would
expect it to connect at more of a right angle to judge from those other
Proto- and Old Canaanite writings of l and not to continue at a slant;
but compare the angle on the slightly ‘open-headed’ hieratic form of
V1 excerpted as the highest writing in fig. 3 above.
Waw is an alternate reading for this letter. The letter waw descends
from T3, a mace usually with a pear-shaped head, exemplified in fig. 3
by a standard hieroglyphic form on a Twelfth Dynasty stele from
Abydos.24 Less often T3 has an elliptical head, as illustrated by a slightly
damaged semi-cursive hieroglyph on a papyrus dating to the Twelfth-
Thirteenth Dynasties from the Ramesseum.25 Although differing in
their stances, both of these hieroglyphic exemplars of mace-signs show
‘V’-shaped connections between their heads and stems that are similar
to that found on the first letter of the seal from Deir Rifa. In terms of
Proto-Canaanite inscriptions, Sinai 351 witnesses a horizontally situ-
ated waw with a pear-shaped head that is formed in a similar, but not
identical manner to this letter. The waw on Wadi el-Îol Text 2,26 al-
though manifesting a different shape for its head, has an upswing to its
22
Note the more developed stance of l found on a bowl with Old Canaanite script
found at Lachish (drawn from G.R. Driver, Semitic Writing3 [London 1976], pl. 43),
which is now assigned to the twelfth century BCE based on a lowered chronology for
Ramesses II and III (F.M. Cross, Leaves from an Epigrapher’s Notebook: Collected Pa
pers in Hebrew and West Semitic Palaeography and Epigraphy [= LEN; HSS 51, Winona
Lake 2003], 209, n. 11).
23
Changes in the stances of early alphabetic letters from their Egyptian graphic
prototypes are often greater than developments to their forms (see the summary in
Origins, 276 83).
24
See ibid. 86, n. 83 for a brief overview of the history of scholarship about this
letter and pp. 86 92 for the graphic overlap between forms of the Egyptian sign T3,
‘mace’ and early West Semitic waw. The first hieroglyphic form of T3 is reproduced
from E.A.T.W. Budge et al. (eds), Hieroglyphic Texts from Egyptian Stelae, etc., in the
British Museum (London, 1911 22), part 2, pl. 12 hereafter cited as HT, part
number, and plate number. I am indebted to the British Museum Company for per
mission to reproduce this and other signs from HT, 1 6, the accuracy of each of which
has been checked either on the original inscription or on a photograph of it in the
archives of the Department of Ancient Egypt and the Sudan, The British Museum.
25
This semi cursive hieroglyph of T3 was traced from A.H. Gardiner, Ramesseum
Papyri: Plates (Oxford 1955), pl. 22.
26
See Darnell et al., Two Early Alphabetic Inscriptions, pl. 7.
60
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
shaft that is much like that which is seen on the first letter of UC
51354. While a horizontal stance for the Egyptian sign T3, ‘mace’ is
attested,27 the upward thrust seen on the stem of this letter would rep-
resent a minor development away from the positioning of such an Egyp-
tian hieroglyph. Waw is thus a possible reading for this letter.
The difficulty in discerning whether the writer of this seal-amulet
intended a waw or lamed here should not obscure the fact that either
identification of this letter has good graphic parallels in known Proto-
Canaanite inscriptions. The formal change from either potential Egyp-
tian prototype, T3, ‘mace’ for waw, or, V1, ‘coil of rope’ for lamed,
would represent a relatively minor development in stance, the upswing
of its horizontal part. In fact, that small development helps to identify
this as a West Semitic letter (and not one of either of those two closely-
related forms of Egyptian signs).28
One can identify the second letter on this seal-amulet as an early
alphabetic qôp with certainty given the similarity of its form with other
Proto-Canaanite writings of that letter.29 It has a three-part form: an
27
The horizontal positioning of T3 is much more rarely attested than its usual
upright posture; see Origins (87, fig. 2.23) for a horizontal T3 from Wadi Îammamat.
28
One is probably dealing with ‘look alikes’ here (rare for these two letters in early
alphabetic scripts [ibid. 134 n. 156]). In many systems of writing some forms come
to appear so similar that they can only be differentiated by their linguistic contexts.
See Z. Zába (Rock Inscriptions of Lower Nubia [Czechoslovak Institute of Egyptology
in Prague and Cairo Publications 1, Prague 1974], sign list: D4 and D21) for exam
ples of four such ‘look alikes’ in Egyptian scripts. See Cross, LEN, 11, 29, for exam
ples of well known ‘look alikes’ in much later Jewish Aramaic scripts (i.e. d/r, w/y).
29
I am following an insight of Driver (Semitic Writing3, 167 9) regarding the
meaning of the acrophonic name *qop as ‘monkey’ back into Egyptian scripts to
isolate its probable, but not certain graphic prototype as E32, ‘sacred baboon’ (see
Origins, 209 21). E32 has both an upright posture (as exemplified in fig. 3 by an
idealized font traced from Wilkinson, Reading Egyptian Art, 73) and a horizontal one
where a monkey is portrayed on all four limbs (illustrated by a hieratic form dating to
the Twelfth Dynasty [Möller, I, Nachträge]). The form of qôp from Deir Rifa presum
ably descends from a horizontal form of E32 in which the monkey’s limbs have been
omitted; see Origins (217 8) for discussion of the omission of feet and legs on this
letter and on Egyptian signs during the late Middle Kingdom and Second Intermedi
ate Period. One awaits the discovery of actual hieroglyphs of E32, particularly semi
cursive ones, from Middle Kingdom inscriptions to test out this probable derivation.
The identification of these letter forms as qôps has recently been questioned, unneces
sarily in my view, by S. Wimmer and S. Wimmer Dweikat, ‘The Alphabetic Texts
from Wadi el Hôl: A First Try’, GM (2001), 108, 111, fig. 3 and in a more nuanced
fashion by F. Briquel Chatonnet, ‘Les inscriptions protosinaïtiques’, in D. Valbelle
and C. Bonnet (eds), Le Sinaï durant l’antiquité et le Moyen Âge: 4000 ans d’histoire
pour un désert: actes du colloque “Sinaï” qui s’est tenu l’UNESCO du 19 au 21 Septembre
1997 (Paris 1998), 57. On its occurrence in nqb in the complete Proto Canaanite
inscription of Sinai 346a, see especially: W.F. Albright, The Proto Sinaitic Texts and
61
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
angular head on the left; a slightly larger body on the right; and an
oblique line situated above its right end. The latter likely represents a
short tail, whose top happens to touch the preceding letter.30 The total
shape of this qôp is most comparable to a two-part qôp with a closed
neck and ovular body on Sinai 351. Insignificant contrasts can be seen
in the formations of the heads of those two letters, the Deir Rifa form
showing a blunter profile, that Sinaitic writing being shaped more like
a diamond. The short tail on the q of this seal may be compared with
less well-preserved occurrences of that feature on Proto-Canaanite qôps
from Sinai 346b and 349. A curled tail is well preserved on the re-
duced form of the certain q on the Grossman Seal, less well so on a new
reading of q on the abecedary line of the ¨Izbet ∑ar†ah Ostracon.31
Although the third letter on the seal-amulet from Deir Rifa is dam-
aged on its left, it may still be identified with confidence as a ‘viper’
form of nûn with one ‘horn’.32 This nûn shows four parts: a short head
set high on the right side; an oblique ‘horn’ of medium length on top
of that head; a longer slanting line for its neck; and part of the body
that is situated lower, almost on a horizontal plane. Each of those fea-
tures on this letter matches forms of ‘horned vipers’ found in Egyptian
Their Decipherment (HTS 22, Cambridge, MA 1966), 17; D. Pardee, ‘Proto Sinaitic’,
OEANE vol. 4 (1997), 354; and, without positing an emendation to that text, Ori
gins, 401. On its occurrence in the complete text of the Grossman/Goetze/St Louis
Seal, see: F.M. Cross, ‘The Evolution of the Proto Canaanite Alphabet’, BASOR 134
(1954), 21, n. 24 and ‘The Origin and Early Evolution of the Alphabet’, EI 7 (1967),
10* (= LEN, 312, n. 24; 319, n. 16); Albright, The Proto Sinaitic Texts, 11; and
Origins, 309, 398. For new photographs of this text and an impression of it, see
Hamilton, ‘W.F. Albright’, 38.
30
See fig. 4 below for other letters on this seal (Ì, z), Wadi el Îol Text 2 (t, y),
Sinai 352 (t, l), and the Lachish Dagger (∂, l) whose lines occasionally touch or over
lap. This phenomenon is relatively rare and should not be taken as evidence of a script
system that employs ligatured forms.
31
See the discussion of this new reading in Origins (218, n. 282), where it is pos
ited that q was written out of order after r on this beginner’s tablet (compare the
certain mistake in the ordering of z and Ì, and perhaps also of p and ¨, in this abecedary;
contrast A. Demsky, ‘Abecedaries’, in W.W. Hallo and K.L. Younger [eds], The Context
of Scripture: Canonical Compositions from the Biblical World vol. 1 [Leiden 1997], 363).
32
Both the Egyptian sign I9 and its Proto Canaanite descendants show variants
with two, one, and no ‘horns’ (Origins, 154 61). Nûn can now be shown to have two
sources in the Egyptian sign list, both I9, ‘horned viper’ and I10, ‘cobra in repose’
(with E. Puech, ‘Quelques remarques sur l’alphabet au deuxième millénaire’, Atti del
I congresso internazionale di studi fenici e punici vol. 2 [Rome 1983], 579, fig. 8; Darnell
et al., Two Early Alphabetic Inscriptions, 94, n. 44; see: Origins, 154 68; against, among
others, Sass, Genesis, 125 [contrast 106, n. 76] and W.V. Davies, ‘Egyptian Hieroglyphs’,
Reading the Past: Ancient Writing from Cuneiform to the Alphabet [London 1990],
131 2, Tables 2, 3).
62
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
inscriptions. Except for a slighter longer neck, its basic body shape
corresponds closely to a two-‘horned’ writing of hieroglyphic I9 in-
cised on an inscription dating to the First Intermediate Period from
Koptos (highest example in fig. 3).33 Its depiction with just one ‘horn’,
executed by an oblique line, is analogous to two occurrences of that
sign on an Egyptian stele assigned to the Middle Kingdom, whose
provenance is unknown.34 The most comparable, well-preserved Proto-
Canaanite nûn with one ‘horn’ occurs in Sinai 363, although the neck
and body parts have merged on that particular writing. The distinct
neck and body on the form from Deir Rifa more closely match those
features on a ‘hornless’ nûn from Sinai 347. One can thus find very
similar models of nûn in previously recognized Proto-Canaanite epi-
graphs and trace the clear parts of the nûn on UC 51354 to an Egyp-
tian hieroglyphic prototype, a variant of I9, with no need to postulate
any significant changes. This snake-pictograph was incised to face in
the opposite direction of the preceding letter, qôp (no matter how one
conceives of the stance of the n as having come about — see the next
paragraph). The use of opposing orientations for these two letters with
fronts and backs is another small indication that one is seeing here an
early West Semitic alphabetic, and not an Egyptian inscription.35
One then needs to account for the clearly incised, slightly curving
oblique line that is found high on the left side of this letter and the
outlined section extending beyond its body located lower on that side.
I would speculate that the seal-cutter initially incised the higher line as
the beginning of a nûn, its head and neck, intending to cut it to face
towards the left (as in the preceding qôp). A small chip of the steatite
then appears to have flaked off below that incision (see pl. 2, figs. 2, 3).
After that damage apparently occurred, the cutter then incised the com-
plete form of nûn, now oriented towards the right (presumably select-
ing that option rather than throwing away the already fashioned, deco-
rated, and partially-incised seal). The outlined figure on the lower left
could then be accounted for as the preliminary etching for the rest of
the tail of the complete nûn, which the seal-cutter did not finish incis-
ing (perhaps not wishing to risk extending the damaged section of the
steatite).36 While the preceding is a speculative reconstruction of what
33
Traced from H.M. Stewart, Reliefs and Paintings from the Petrie Collection, Part
Two: Archaic Period to Second Intermediate Period (Warminster 1979), pl. 11.1.
34
Reproduced from HT, 3, pl. 16.
35
See Origins (279 82) regarding letters with fronts and backs facing opposite
directions, on both vertically and horizontally arranged Proto Canaanite texts as
developments away from the usage of orientations in Egyptian scripts.
36
Compare the small fragment numbered Sinai 372a incised with a door sign
63
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
whose bottom lines were initially etched but apparently never completed (photo
graph: R.F. Butin, ‘The Proto Sinaitic Inscriptions’, HTR 25 [1932], pl. 26; Sass,
Genesis, fig. 120).
37
See Origins (Appendix 2, 401 6) concerning the various ways of arranging let
ters on early alphabetic texts, from group writings to placements as single files and the
likely origins of those arrangements in Egyptian practices.
38
Ibid. 97 102. With this derivation I have extended back into Egyptian writing
the understanding of the name(s) of Ì as ‘fence’ by F.M. Cross and T.O. Lambdin, ‘A
Ugaritic Abecedary and the Origins of the Proto Canaanite Alphabet’, BASOR 160
(1960) 26 (= Cross, LEN, 316). This required a separation of early writings of Ìêt and
dalet, which have often been confused but can now be differentiated on the basis of
their Egyptian graphic prototypes (Origins, 61 3).
39
The first and second letters on this seal also touch (fig. 2), as do at least two
letters on Wadi el Îol Text 2, Sinai 352, and the Lachish Dagger (fig. 4).
64
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
Linear forms of that sign typically show four verticals with long hori-
zontal lines cutting through their middles and running just below their
bases, as illustrated by two semi-cursive hieroglyphic exemplars and a
hieratic writing excerpted from texts dating to the Twelfth Dynasty.40
A conservatively derivative ‘four-bar’ form of Ìêt has recently been dis-
covered on Sinai 375a, a plaque from Serabi†.41 A similar form of that
letter with only three bars occurs on the dipinto with Old Canaanite
40
The hieroglyphic exemplars are reproduced from HT, 4, pl. 50, a statuette from
Deir el BaÌri. The hieratic form is excerpted from Möller, II, 368.
41
See Origins (98 100, figs. 2.26 7) for a discussion of this new reading, based
on a fresh collation of the original plaque that was recently relocated in the collection
of the Harvard Semitic Museum. This letter in fig. 3 was traced from a transparency
of an unpublished colour photograph of Sinai 375a made by Drs M. Lundberg and
B. Zuckerman, West Semitic Research: black and white photographs of that inscrip
tion are now published in Hamilton et al., ‘Three Recently Relocated Texts’, pls 6, 7.
65
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
66
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
46
I am following the identification of this variant form of T7, ‘axe’, as T7A in
Hieroglyphica Sign List; it is listed as T7* by Gardiner (Egyptian Grammar3 [London
1957], 511). The isolation of both hieratic forms of T7 and hieroglyphic forms of
T7A as sources for early alphabetic forms of zayin was achieved through the conver
gence of lines of research by several scholars (Origins, 92, n. 95). A clear derivative
of a hieratic form of that sign has also been identified on Sinai 345 (ibid. 93 4,
fig. 2.24). Only the hieroglyphic alphabetic correspondences are charted in
fig. 3.
47
Gardiner, Egyptian Grammar3, 511, T7*, n. 2.
48
Reproduced from Lacau, Sarcophages antérieurs au Nouvel Empire, pl. 39.
49
See Origins (94 5) for a discussion of this new reading, and Hamilton et al.,
‘Three Recently Relocated Texts’, pls 6, 7 for photographs of it on p. 19, section 5,
end of first full paragraph, after ‘horizontal line’, add: (read either from right to left
on the seal itself [pl. 2, fig. 2] or from left to right on an impression of it).
50
This follows the graphic interpretation of that letter, which was lightly incised,
by F.M. Cross, ‘Newly Found Inscriptions in Old Canaanite and Early Phoenician
Scripts’, BASOR 238 (1980), figs 10, 11 (= idem, LEN, 221, fig. 32.6; 223, fig. 32.7),
confirmed on the original by both E. Puech, ‘Origine de l’alphabet’, RB 93 (1986),
fig. 3 and the present writer, Origins, fig. 2.25; cf. the more limited perceptions of it
by M. Kochavi, ‘An Ostracon of the Period of the Judges from ¨Izbet ∑ar†ah’, Tel Aviv
4 (1977), fig. 3, and Sass, Genesis, fig. 175.
51
After the merger of *z and *∂ in transmitting tradents of the alphabet, this
ostracon contains only the zayin grapheme (Origins, 95 6). I would doubt that its
Ìêt zayin sequence should be taken as evidence of ‘pluriform primers’ (so B. Byrne,
‘The Refuge of Scribalism in Iron I Palestine’, BASOR 345 [2007], 18) even given the
same ordering on the recently discovered abecedary from Tel Zayit (R. Tappy et al.,
‘An Abecedary of the Mid Tenth Century B.C.E. from the Judean Shephelah’, BASOR
344 [2006], 5 45). Both of these beginners made multiple mistakes in ordering the
letters (cf. Byrne, ‘Scribalism’, 21, n. 77).
67
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
52
See the tracings by Puech, ‘Origine’, fig. 3 and the present writer, Origins,
fig. 2.25.
53
Gardiner, Egyptian Grammar3, 439, 511, T7*, n. 1; Fischer, ‘Archaeological
Aspects of Epigraphy and Palaeography’, 34, n. 31.
54
See Origins (289 90) for a discussion of this and two other palaeographic indi
cators of when Proto Canaanite writing started, c. 1900 BCE (similarly and independ
ently, Darnell et al., Two Early Alphabetic Inscriptions, 90; contrast Sass, ‘Genesis of
the Alphabet’, 157).
68
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
This short text also manifests both archaic and innovative arrange-
ments of its letters. It continues an archaic arrangement of the letters
in which Ìêt and zayin are written side by side in an otherwise single-
file column of characters. In contrast, the two letters with fronts and
backs, qôp and nûn, are arranged to face in opposite directions, an
arrangement of characters that is against Egyptian norms, but well at-
tested in other Proto-Canaanite inscriptions.
4. Dating
55
See the references in n. 53. The beginning of the Twelfth Dynasty has been
recently estimated at c. 1938 BCE (Baines and Malek, Cultural Atlas2, 9, 40), or, c.
1985 BCE (I. Shaw [ed.], The Oxford History of Ancient Egypt [Oxford 2000], 480, 459
[a select bibliography on the difficulties of Middle Kingdom chronology]).
69
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
that has rotated from the position of its hieratic forebear).56 Found in
Tomb 1502 at Lachish, that dagger can now be dated by co-ordinating
typological analyses of the ceramics found with it and its early alpha-
betic script to c. 1725 BCE (± 25 years).57 Given the large formal change
seen in the Ì of UC 51354, I would estimate that the legend on this
seal was inscribed at least somewhat later than the inscription on that
dagger, c. 1700 BCE.
Yet one also needs to acknowledge that unlike that dagger, this seal-
amulet’s discovery was so poorly recorded and published that one is
dealing with an artefact that is almost without an archaeological prov-
enance (so Quirke, personal communication). I would indicate that
unfortunate reality by leaving a very wide margin for error in estimat-
ing its time of writing, c. 1700 BCE (± 150 years). It was most likely cut
in the middle of Egypt’s Thirteenth Dynasty, a time when we know
that other West Semites began having seals engraved in Egypt and
scarab-shaped amulets start to appear in Palestine.58 Yet this seal could
have been inscribed as early as the two early alphabetic graffiti from
Wadi el-Îol or the West Semitic name on the heddle jack from Lahun,
c. 1850–1700 BCE.59 It seems very unlikely, however, that such a high-
quality seal of glazed steatite and probably ‘faience’ beads would have
been acquired by a West Semite and incised with Proto-Canaanite let-
ters after the expulsion of the Semitic Hyksos from Egypt, c. 1550 BCE
(or shortly thereafter).60
56
Origins, 303 4.
57
See especially Tubb’s analysis of the contents of that tomb (ibid. n. 37), includ
ing the elimination of a scarab thought to be of Neferhotep II, whose presence had
previously required a later dating for the whole assemblage. The continuing scepti
cism about the Proto Canaanite classification of the script on this dagger by Sass
(Genesis, 54; ‘Genesis of the alphabet’, 150) is no longer warranted; for good parallels
in Proto Canaanite and Egyptian scripts for each letter in the order in which it oc
curs, see Origins, 147 9; 132, n. 153; 223 5; 157 60; 248 51. For a new proposal
as to the meaning of this text in West Semitic, see ibid. 149, n. 181. The secure
assignment of this artefact to the Middle Bronze Age, which Sass acknowledges (ibid.),
makes his lowering of the origin of alphabetic writing to c. 1300 BCE not just unlikely
but impossible (see also nn. 4, 5 above).
58
For the seal of an ‘Asiatic’ in Egyptian from the Thirteenth Dynasty, see M.
Bietak, Avaris, The Capital of the Hyksos, Recent Finds at Tell el Dab¨a (London 1996),
41, fig. 35.12 and Section 6 below. Regarding the emergence of scarabs in Palestine at
that time, see Ben Tor, ‘Relations between Egypt and Palestine’, 164.
59
Also see the references in nn. 4 and 5.
60
The end of the Second Intermediate Period has recently been estimated at
c. 1550 BCE, specifically 1532 28 BCE for the conquest of Avaris (Bourriau, ‘Second
Intermediate Period’, 185), or, c. 1520 BCE (Baines and Malek, Cultural Atlas2, 9, 42).
70
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
71
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
72
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
context of this noun used as a proper name on a seal does not allow
one to sort out those semantic options with any certainty. Thus I will
transliterate the name as Qn.
F.M. Cross has recommended reading the lowest two letters, Ìz, as a
m. sg. G active participle traced to the root *Ìzy, ‘to see’, used as a
substantive to indicate the occupation of this seal’s owner, ‘(the) Seer’.72
Although one would expect such a third-weak form to be spelled with
a final y if the nominative case ending was /u/, i.e. *Ìaziyu, Cross would
explain its absence here as being due to a genitive /i/ following the
prepositional lamed, contracting /*iyi/ to /*î/, and left unmarked in
the orthography of this period, i.e *la/i-Ìaziyi > *la/i-Ìazî.73 Since words
traced to the root *Ìzy are well documented in later West Semitic lan-
guages and occupations following personal names are well attested on
seals belonging to West Semites, I shall follow Cross’s proposal for de-
ciphering this word as ‘(the) Seer’ on UC 51354.74
Participial forms traced to the root *Ìzy are well known from both
Biblical Hebrew and epigraphic sources in Aramaic and a Transjordanian
dialect.75 Of some 129 occurrences of words assigned to this root in
For evidence of metal working in the mines at that site, above the entrance to one of
which Sinai 380 was written (Sass, Genesis, figs. 100, 103, 104), see I. Beit Arieh,
‘Serâbî† el Khâdim: New Metallurgical and Chronological Aspects’, Levant 17 (1985),
89 116.
Halpern (‘Kenites’, 18 21) treated some of the biblical evidence as possibly re
flecting persons, places, and groups of the late second millennium BCE, but empha
sized that most biblical traditions about them stem from the Iron Age, c. 700 BCE or
later. Hess (‘Cain’, 806) noted that I. Eph¨al (The Ancient Arabs [Jerusalem 1982],
194, 211 12, 226 7) would date some of the South Semitic epigraphic attestations
to as early as the fifth century BCE.
72
Email correspondence, September 27, 2006. Prof. Cross corrected my attempt
to read the last two letters in the opposite direction, zÌ, as part of a short sentence
name, ‘(The clan) Cain moves (away)’ (‘to move’ or ‘to move away’ are meanings
assigned to *zwÌ/zÌÌ in later Aramaic, especially Syriac, and Arabic). That proposal
fails given the lack of attestations of that verb in the West Semitic onomasticon.
Cross’s superior interpretation is communicated with his kind permission.
73
Email correspondence, September 29, 2006.
74
The Arabic cognate Ìazay, ‘perceive with an inner vision’ (BDB, 302) assures
that one is dealing in this root with an etymological *Ì (and not *Ì) and *z (not *∂).
To our knowledge, *Ì is differentiated from *Ì and *z from *∂ in Proto Canaanite
inscriptions (for a discussion of their different graphemes and originally distinct letter
names, see Origins, 97 102, 57 60; 92 7, 145 54).
75
Although dated, see the still valuable review by A. Jepsen, ‘chazah’, TDOT (IV),
280 90. For parallels of titles on seals and sealings of the much later Iron Age, see B.
Sass in N. Avigad, Corpus of West Semitic Stamp Seals (Jerusalem 1997), 466 8, espe
cially four or five titles composed of m. sg. G active participles, hspr, ‘the Scribe’, hÒrp,
‘the Goldsmith’, (h)khn, ‘(the) Priest’, hrp’, ‘the Healer’, and probably lÌs ‘Charmer/
Whisperer of incantations’ (467 8, 509).
73
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
76
Jepsen, ‘chazah’, 282. For more recent literature particularly concerning attesta
tions in various levels of Aramaic, see HALOT (5), 1872.
77
Jepsen, ‘chazah’, 285 9. For a probable occurrence of the cognate noun Ìzn,
‘vision’ (less likely ‘inspector’ or ‘supervisor’) in a broken context in a Hebrew inscrip
tion from Lachish, see F.W. Dobbs Allsopp et al., Hebrew Inscriptions: Texts from the
Biblical Period of the Monarchy with Concordance (New Haven 2005), 337 8 and
J.P. Kang, ‘A Dictionary of Epigraphic Hebrew’, unpublished Ph.D. dissertation
(Princeton Theological Seminary 2006), 136. I would like to thank Dr Kang for both
of these references.
78
Recently see A. Millard, ‘The Inscription of Zakkur, King of Hamath’, in
W.W. Hallo and K.L. Younger (eds), The Context of Scripture: Monumental Inscrip
tions from the Biblical World vol. 2 (Leiden 2000), 155 and C. L. Seow, ‘West Semitic
Sources’, in M. Nissinen, Prophets and Prophecy in the Ancient near East (SBLWAW
12, Atlanta 2003), 203 7, n. a.
79
Millard, ‘Zakkur’, 155.
80
See the various dates reviewed by M. Dijkstra, ‘Is Balaam also among the
Prophets?’ JBL 114 (1995), 45 6, B. Levine, ‘The Deir ¨Alla Plaster Inscriptions’, in
W.W. Hallo and K.L. Younger (eds), The Context of Scripture: Monumental Inscrip
tions from the Biblical World vol. 2 (Leiden 2000), 141, and Seow, ‘West Semitic
Sources’, 207.
81
Similarly, Levine, ‘Deir ¨Alla’, 142; Dijkstra, ‘Balaam’, 47. Both a cognate verb,
wyÌz, ‘he saw’ and noun, mÌzh, ‘a vision’ can be reconstructed securely at the end of
line 1 (with J. Hackett, The Balaam Text from Deir ¨Alla [HSM 31, Chico 1984], 33,
against, among others, Dijkstra ‘Balaam’, 52, n. 20 and Seow, ‘West Semitic Sources’,
209 10).
74
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
6. Functions
75
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
86
James, ‘Ancient Egyptian Seals’, 38.
87
On the use of seal amulets without legends, the majority of which are shaped as
scarab beetles, James commented (ibid. 34): ‘Engraved with a distinctive device it
could be used as an individual badge of possession, a very suitable instrument for
marking and sealing purposes…It did not have to be “read” in its strictest sense, but
simply recognised’.
88
For a map showing the dozen sites in Egypt with Palestinian Middle Bronze
Age cultural finds, see Baines and Malek, Cultural Atlas2, 41. Also see n. 91 below.
89
The longest early alphabetic rock inscription to surface to date is Sinai 349,
originally with seven lines (Origins, 339 40). For an overview of semi cursive hiero
glyphic and hieratic documents on papyrus, in particular see R. Parkinson and S.
Quirke, Papyrus (London 1995).
90
Bietak, Avaris, 41. Ben Tor (‘Relations between Egypt and Palestine’, 187 8)
concluded that: ‘[T]he custom of using scarabs primarily as funerary amulets was
transmitted by the Asiatics who settled in the eastern Delta during the late Middle
Kingdom. The Tell el Dab¨a material culture clearly demonstrates that these settlers
adopted Egyptian customs, including the use of scarabs as funerary amulets, and
most probably initiated the import of the latter into Palestine’. One may now con
sider looking further south in Egypt for evidence of the adoption of that custom.
76
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
77
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
Texts’, in A. Leahy and J. Tait (eds), Studies on Ancient Egypt in Honour of H.S. Smith
(Egypt Exploration Society Occasional Publications 13, London 1999), 43 8. I would
like to thank Ms Bourriau for the latter reference.
95
Bourriau, ‘Second Intermediate Period’, 190.
96
Millard, ‘Zakkur’, 155, n. 10 (with literature).
97
For other examples of those tendencies, see Origins, 320.
78
A PROPOSAL TO READ THE LEGEND OF A SEAL-AMULET
98
Previously, the cylindrical Grossman Seal has been the earliest seal certainly
inscribed with alphabetic letters (see n. 62 above). I remain uncertain about the clas
sification of the writing on a seal amulet from Palestine proposed by K.A. Kitchen,
‘An Early West Semitic Epigraph on a Scarab from Tell Abu Zureiq?’, IEJ 39 (1989),
278 80. The problem in that proposal is palaeographic his proposed lamed does
not correspond to any early forms attested for that well known letter (cf. Origins,
128 9, figs. 2.35 8).
79
VINCENT DECAEN
UNIVERSITY OF TORONTO
Abstract
1
A special word of thanks to Elan Dresher, to whom this paper is dedicated,
and whose continued support and encouragement through the years have kept me
working away at generative grammar and Tiberian Hebrew phonology (or as I pre
fer, ‘generative Masoretics’). It should be obvious that Dresher’s seminal 1994 study
is my constant inspiration throughout.
I must also recognize John Hobbins and his invaluable online resources, espe
cially his detailed bibliographies and history of research (see http://ancienthebrew
poetry.typepad.com). He was quick with voluminous and insightful feedback on
earlier drafts, and the study is many times better as a result. I am especially grateful
for his pointing out those crucial footnotes in Dhorme’s commentary on Job, and
also for patiently explaining the details of Fokkelman’s syllable counting.
Thanks also to Jim Price who, in addition to offering feedback on an earlier
draft, generously supplied me with data for nesiga before silluq. Specific insights are
noted in the relevant footnotes.
Of course, these gentlemen can in no way be held responsible for the views
let alone the errors contained herein.
I should also thank an anonymous reviewer for helpful suggestions that have
been incorporated in this final version, especially for alerting me to Loretz’s count
for Psalm 111.
Finally, my work is made possible in part through a generous grant from the
nonprofit GRAMCORD Institute, dedicated to computer assisted analysis of bibli
cal languages ( http://www.gramcord.org).
81
THEME AND VARIATION IN PSALM 111
2
Khan furnishes a prototypical example, tô´¨aßo†êhém
´ (Ez. 6:9); see further
Khan (1987: 54ff ).
3
Specifically, a ‘loose iambic’ rhythm. The loose iambic foot is analogous to
deuterium (one proton, one neutron) with its two isotopes, tritium (one proton,
two neutrons) and standard hydrogen (one proton, zero neutrons). For a general
review of the theory of iambic meter and its many variations, see, e.g., Steele
(1999). For comparable loose iambic verse in German and Russian, see further
Tarlinskaja (1993); see also Segert (1960: 289).
82
THEME AND VARIATION IN PSALM 111
83
THEME AND VARIATION IN PSALM 111
84
THEME AND VARIATION IN PSALM 111
Pursuing the strategy of divide and conquer, the scope of this pro-
grammatic paper is drastically limited to just the twenty-two lines of
Psalm 111 and those lines only (a traditional romanization with phrase
divisions added is provided in appendix 1). The paper explains how
the prosodic representation of the liturgical chant regulates the poetic
metre of this psalm. The analysis is proposed as the basis of an ex-
tended research programme in generative metrics and BH poetry.
Psalm 111 has been selected, first and obviously, because it is a
poem that has been annotated with the poetic accents, and so presum-
ably is representative of the metrical tradition. Second, it is an acrostic,
which removes the fiddle-factors in lineation as well as possible inter-
calation of extraneous lines (of potentially different metrical struc-
ture). Third, there are no emendations worth troubling about. Fourth,
it is short for a biblical acrostic, presenting a more tractable database.
Finally, and most importantly, Psalm 111 emerges as the most regu-
lar poem in the seminal syllable-counting study of Culley (1970).
Furthermore, because it is an acrostic, it also receives a detailed statis-
tical analysis in Vance’s massive doctoral study (2001) as an addi-
tional check.
Culley (1970) proposes a ‘strictly descriptive approach’ to BH po-
etry. He adopts a syllable count as the better measure of the length
and the contour of the colon. He provides a detailed analysis of sev-
eral key poems, the product of the application of his relatively crude
statistical approach.
Vance, however, is dismissive of the value of Culley’s paper (Vance
2001: Ch. 2, 182–4) — despite the broad overlap of the two ap-
proaches.
9
Segert (1960) provides an overview of the ‘alternating stress approach’, with
references for the history of this line of research (see esp. his note 1, p. 284).
85
THEME AND VARIATION IN PSALM 111
10
Interestingly, select lines from the so called P Source (Numbers 23 4) are
grouped here too. In a corpus linguistic study of biblical dialectology (DeCaen
2000 1), I suggest that the Pss congregate at the extreme end of the dialect spec
trum together with the P Source.
À propos the significance of the eight syllable count, Fokkelman (2000) writes,
‘And what if we should find more than 40 poems in the Psalter that score this per
fect 8.00 as the average number of syllables per colon? In that case, the proposition
that the occurrence of such a round number is a coincidence has become totally
untenable… I am toying with the idea that the 148 poems of the Psalter contain
1193 strophes, 2695 verses…, 5714 cola, 18,944 words and 45,733 syllables. If we
divide this last number by the number of cola we get 8.0036, a highly remarkable
number as it only deviates from the perfect number 8 (8.00) in its third decimal.
This is such a minimal deviation that we should consider whether the number 8
might not have been a guideline or normative figure for the poets in constructing
cola’ (Fokkelman 2000: 9 10).
86
THEME AND VARIATION IN PSALM 111
11
‘Word count’ here discounts white space counter indicated by the TH
hyphen or maqqef. The word count for the purely consonantal text increases to 72
by eliminating the five maqqefs, giving a mean of 3.3 words instead.
12
For a theoretical explanation of the nature and distribution of TH anaptyxis
in terms of syllable contact and sonority constraints, see DeCaen (2003).
87
THEME AND VARIATION IN PSALM 111
‘Meter is a contract between the poet and the reader. The poet de-
clares what he or she is going to do in the opening lines of the poem,
and this in turn, sets up the reader’s expectation’ (Vance 2001: 491).
Following Vance’s suggestion, the heuristic adopted for this research
programme is that the first line and first stanza must establish the un-
derlying metrical theme of the Psalm.
Further, the ‘fulfilling of the contract may involve permissible
variations to which the reader is sensitive and that give heightened
pleasure for the reader’ (491). Accordingly, the ‘permissible’ variation
will be sought in subsequent lines.
At the highest level, we observe two ‘bars’ or phonological
phrases in 111:1a (see the accentual parse provided in appendix 3).
Indeed, the poem generally approaches, in Kury¥owicz’s terms, a 2 +
2 verse.13 However, making due allowance for the lawful transforma-
tions of disjunctives into virtual disjunctives, made explicit in Price’s
phrase-structure analysis (1990, 1996),14 the acrostic becomes con-
sistently 2 + 2. (The ‘virtual’ disjunctives are marked below by an
exclamation mark (!), and all such cases are so indicated in appendix
3, section (B).)
This observation of 2 + 2 metrical regularity hardly scratches the
surface, however, and still might be compatible with the syntax-only
approach of conventional wisdom (see references in Vance 2001).
The linguistically significant generalization is the double dichotomy
of the accentual parse of 111:1a given in (1).15
13
Cf. Ps. 44:5 27 (Kury¥owicz 1975: §15, pp. 222f ), Prov. 8:22 31 (Cooper
1976: 199ff ); and Job generally (Vetter 1897). The apparent cases of 4 + 2 in verses
9 and 10 are reanalysed here as tripartite 2 + 2 + 2.
14
The Law of Transformation is similar to rhythm rule or nesiga, just one pro
sodic level higher: the clash of back to back disjunctives is resolved by demoting a
disjunctive to its ‘virtual’ disjunctive counterpart (an appropriate conjunctive).
Schematically, we could characterize all instances in Psalm 111 as C D D → C C D.
(Indeed, musically, the reality might even be more nesiga like: X Y Y → Y X Y. This
might explain why putative conjunctives, reserved exclusively for the output of such
a transformation, graphically resemble disjunctive counterparts in the prose system:
conspicuously, the conjunctive tarcha, which is identical to the prose D1f tiphcha.)
15
The standard notation employed here is borrowed from Dresher (1994), ulti
mately from Cohen (1969). Conjunctives (C) are distinguished from disjunctives
(D). The relative rank of the disjunctive is given {0 3}, where 0 marks the highest/
strongest grade among the disjunctives and 3 the lowest/weakest; D0 is verse final
silluq, for example, and Dn will be the general variable, where n ranges over {0 3}.
88
THEME AND VARIATION IN PSALM 111
(1) 1
2f 1
C D2f M16 D1
1f 0
C D1f C D0
The additional f for ‘final’ separates the minor disjunctives (Dnf ) from the major
disjunctives (simply Dn); in Price’s terminology, Dnf is the ‘subordinate’ disjunc
tive, while Dn is the ‘remote’ disjunctive. Finally, as a conjunctive place filler in the
representations, I am using M to mark maqqef.
16
The maqqef (M) is assumed throughout to do duty for the trope’s conjunc
tive: a ‘virtual munach’, as it were, in this case.
17
The rendering of the divine name is conventional, and the doubt regarding
the actual pronunciation of this name does not raise substantive questions in this
study, as will become clear in the sequel.
89
THEME AND VARIATION IN PSALM 111
FOOT (F) F F F F
SYLLABLE (s) s s s s s s s s
90
THEME AND VARIATION IN PSALM 111
3. Variation:
Foot-Syncopation and Foot-Substitution
(4) I
f f
F F Ø F
s s s s s s s s
be sô´∂ ye sa rîm
´ we ¨e ∂â¬
(5) I
f f
Ø F F F
s s s s s s s
91
THEME AND VARIATION IN PSALM 111
3.1. Foot-Syncopation
The upper end of the range is necessarily four feet per line, given the
dipodic nature of the phonological phrase. The variants mark the
lower end of the range as three feet per line.
Two questions arise once we admit the tripodic line under syncopa-
tion. First, how do we keep the lower range from dropping to two
feet per line, since permitting syncopation generally would obviously
overgenerate in Psalm 111? In other words, how do we prevent syn-
copation from applying to both phrases in the line? Second, more
technically, what is the prosodic nature of the tripodic input and
output of the ‘Law of Transformation’ (note 14) governing the TH
virtual disjunctives?
First, we obviously need a basic phrase-structure rule that intro-
duces asymmetry and syncopation. Accordingly we introduce
optionality (indicated by the round brackets) into a generalized
phrase-structure rule, obtaining thereby X → (W) S (read ‘some pro-
sodic unit X maps to an optional weak unit W and an obligatory
strong unit S on the lower prosodic level’). This rule permits only the
two mappings between levels shown in (6): the iambic structure in
(6a) and the so-called ‘degenerate’ structure in (6b).
(6) (a) X (b) X
W S S
To avoid overgenerating (i.e., obtaining the reduction to just two
feet per line through syncopation in both phrases), we require a gen-
eral constraint against ‘stress clash’. The constraint is formulated in
(7), abbreviated *CLASH (read ‘star clash’), to ensure that two feet are
insufficient to project the two phonological phrases demanded by
the poetic line. (Let us assume further that the repair strategy would
result in the structure in (8).)
(7) *CLASH X X
*S S
92
THEME AND VARIATION IN PSALM 111
(8) → X
W S
The clash constraint in (7)–(8) suggests in turn the proper treat-
ment of the TH Law of Transformation, whereby a disjunctive is de-
moted to its ‘virtual’ counterpart (note 13). The input in (9), with
back-to-back strong nodes, would obviously violate *CLASH; the out-
put must be similar to (8). I propose that the asymmetrical left-recur-
sive structure in (10a) captures the Janus-faced behaviour of the vir-
tual disjunctives, contrasting with the flat structure in (10b).
(9) I
fW fS
FW *FS FS
(10a) I (10b) I
f f
fW FS F W FW F S
FW FS
93
THEME AND VARIATION IN PSALM 111
UW IS
IW IS
the distribution and depth of TH phonological phrase nesting, derived from first
principles.
U
U I
I I
I f
f f
f F
20
Many thanks again to James Price, who kindly provided me with nesiga data
from his marked up database. His data relating to the environment before athnach
and silluq in the three poetic books are consistent with the complementary distribu
tion predicted by the conjecture here.
94
THEME AND VARIATION IN PSALM 111
FW FS
s s s <s>
(b) f
FW FS
s s s
(13) (a) f
fW FS
FS
s s s s <s>
21
A quick summary of Praetorius’s description of the environment is supplied
by Khan (1987: 62); see Revell (1987) for an exhaustive treatment.
95
THEME AND VARIATION IN PSALM 111
(b) f
fW FS
FS
s s s
F F
sS sW sS
hô¬∂ ’o ∂1h
96
THEME AND VARIATION IN PSALM 111
FS FW FS
FW sS sW sS sW sS
sW sS te hil la †ô´
ye sa rî¬m
22
It should be pointed out that the Tiberian cantillation distinguishes ‘short’
versus ‘long’ trisyllabic words (Dresher in press; 1994: §6.2, 34 6), whereas there is
no distinction here in Psalm 111 (they are all short anapests). Specifically, lîre’aw
(5a) and ne’emanîm (7b) are both ‘long words’ by rule, but this does not register in
the metre or rather, the Tiberian prosody results in supernumerary feet. This very
subtle mismatch is worth exploring as a diagnostic and perhaps key in future study.
23
While this ‘head dependent asymmetry’ in Psalm 111 is quite striking, in
cross linguistic perspective it is rather commonplace. A theoretical treatment within
current generative phonology can be found in Dresher and van der Hulst (1998);
they specifically treat TH under ‘visibility’ asymmetries (§2.1, 328ff ). (Notice fur
ther that this head dependent ‘complexity’ asymmetry here in Psalm 111 may in
part shed some light on the original differences in stress and trope between con
junctives and disjunctives.)
97
THEME AND VARIATION IN PSALM 111
98
THEME AND VARIATION IN PSALM 111
suggestions in the secondary literature (see further Scoralick 1997: 191 193).
Fokkelman finds Scoralick’s recent study ‘unsatisfactory’ (2003: 213, n. 1). His
analysis is admittedly inconsistent with that of the twin Psalm 112, as well as the
MT and ‘all subsequent translations and commentaries’, yet Scoralick is chided for
not spotting it in her ‘unsound’ analysis (p. 213). Indeed, his version is so obvious
that he can ‘leave it to the reader to find good poetical reasons’ for altering MT
(2002: 170)!
26
Assuming that the typographical error in his count is corrected.
27
This metrical grid can naturally and simply explain the relative length of
vowels in TH orthoepy. Khan gathers the details from the masoretico grammatical
literature with special reference to the accentuation (1987: esp. 59 66).
99
THEME AND VARIATION IN PSALM 111
fW fS
(FW) FS (FW) FS
sW sS F/sW sS sW sS F/sW sS
*
* *
* * * *
* * * * * * * *
100
THEME AND VARIATION IN PSALM 111
structure rule above, restated here: X → (W) S. The minimum of the range would
equal the number of phonological phrases in the layer above, while the maximum
of the range would be capped by the number of feet permitted in the layer below.
In Psalm 111 the range is predictably 2 4 prosodic words per line.
101
THEME AND VARIATION IN PSALM 111
typically distinguish at most four degrees of stress. There are also just
four basic elements, for example, in the Latin chant (initium, tenor,
mediant, finalis). There are in fact many sources of four without re-
sorting to numerology. Nevertheless, in Psalm 111 the maximal
expansion of the system to handle the final tricola only requires
branching to the depth of a ‘minor’ D3 (or D3f ): surely not a coin-
cidence, on the view presented here.
Finally, why ‘twin’ Tiberian cantillation systems? Perhaps because
the simpler poetic system was designed to handle the metrically sim-
ple and regular 2 + 2 verse that characterizes ‘truth’ (Job, Proverbs
and much of the Psalms), presumably with deep roots in the Second
Temple liturgy. Subsequently, it might be conjectured, the primitive
system30 had to be greatly expanded and articulated (including cru-
cially the promotion of athnach to D0 status) to handle the challenge
of reading extended prose passages in a similar fashion.
The next steps in this programme are the analysis of the remain-
ing acrostic psalms and the parallel track of investigating the so-
called 2 + 2 mashal metre (4 + 4 in terms of ‘beats’ or feet) of the
presumably homogeneous corpus in Job (Vetter 1897). Also on the
agenda is the relation of the two Tiberian systems to each other31 and
to the non-Tiberian accent systems. A particular question arising in
the present discussion is why the Law of Transformation (note 14)
should stand as an iron law for all degrees of disjunctive in the TH
poetic system, whereas the analogous transformation is confined to
geresh (D3f ) in the prose system.32
30
Revell points out (p.c.), in this regard, that the Palestinian and Babylonian
systems bear a strong family resemblance to the poetic, not the prose, Tiberian sys
tem.
31
Price points out that the prose in Job 32:1 6a is accentuated according to the
poetic system, and that ‘these verses exhibit minor deviations from the rules of ac
centuation due to their non poetic structure’ (1990: note 1, p. 161). These ‘minor
deviations’ might be a good place to begin.
32
A startling implication follows from this difference: two different liturgical
readings are possible, given identical Hebrew inputs. The difference would be con
spicuous in terms of spirantization (or lack thereof ).
102
THEME AND VARIATION IN PSALM 111
I
1a ’ô-∂1h yah-wé33 | be-Èol le-ß0ß 3/4/A
1b be-sô¬∂ ye-sa-rî´m | we-¨e-∂⬠3/3/B 6/7
2a ge-∂o-lî´m | ma¨-a-sê¬ y yah-wé 3/3/C
2b de-rû-sî´m | le-Èol Ìep-Òê-hém 2/3/C 5/6
II
5a †é-rep | na-†0n lî´-re’-0w 3/4/C
5b yiz-k3r le-¨ô-l0m | be-rî-†ô¬ 3/3/B 6/7
6a k3aÌ ma-¨a-s0w | hig-gî´∂ le-am-m¨ô¬ 4/4/A
6b la-†1† la-hém | na-Ìa-lᆠgô-yî´m 4/4/A 8/8
III
9a pe-∂û¬† | sa-láÌ le-¨am-mô¬ 3/3/C
9b Òiw-w⬠le-¨ô-l0m | be-rî-†ô¬ 3/3/B 6/6
9c qa-∂ô¬s we-nô-r0’| se-mô¬ 3/3/B 9/9
33
Here and throughout, one can read ’ãdonay instead.
103
THEME AND VARIATION IN PSALM 111
104
THEME AND VARIATION IN PSALM 111
n+1 n
C Dn+1 C Dn
n+1 n
C Dn+1 ! Dn
105
THEME AND VARIATION IN PSALM 111
n+1 n
Dn+1 C Dn
BIBLIOGRAPHY
Anstey, Matthew. 2006. ‘The Grammatical lexical Cline in Tiberian Hebrew’. Jour
nal of Semitic Studies 51:1, 59 84
Collins, Terry. 2005. Review of Colometry and Accentuation in Hebrew Prophetic
Poetry by Thomas Renz. Journal of Semitic Studies 50:2, 384 5
Cohen, Miles B. 1969. The System of Accentuation in the Hebrew Bible.
(Minneapolis)
Cooper, Alan Mitchell. 1976. ‘Biblical Poetics: A Linguistic Approach’. Unpub
lished Ph.D. thesis, Yale University
Culley, R.C. 1970. ‘Metrical Analysis of Classical Hebrew Poetry’, in J.W. Wevers
and D.B. Redford (eds), Essays on the Ancient Semitic World (Toronto Semitic
Texts and Studies, Toronto). 12 28
35
Technically, this line, as parsed by the TH accents, is the output of the vir
tual disjunctive transformation (that is, technically line type B). However, I am
reading here this munach as a virtual disjunctive dechi (hence, line type C) metri
causa. This reading removes the glaring irregularity of the trisyllabic semûÈîm on a
conjunctive, and removes an awkward B line from the strophe pattern. Regardless,
nothing substantive is at stake as far as the essential claims put forward in this pa
per; it does raise, however, the vexing question of just what sort of mismatches
might be permitted between the TH accents and a regular metre. (This question of
mismatches is not unrelated to the problem of finding an algorithm that can gener
ate the TH accentuation from the raw morphosyntactic input [and what sort of
mismatches result], and how this algorithm might differ between the two accent
systems[prose versus poetic].)
106
THEME AND VARIATION IN PSALM 111
107
THEME AND VARIATION IN PSALM 111
108
THEME AND VARIATION IN PSALM 111
Sanders, Paul. 2003. ‘Pausal Forms and the Delimitation of Cola in Biblical
Hebrew Poetry’, in Korpel and Oesch 264 78
Scoralick, Ruth. 1997. ‘Psalm 111 Bauplan und Gedankengang’. Biblica 78:2,
190 205
Segert, Stanislav. 1960. ‘Problems of Hebrew Prosody’, in Congress Volume, Oxford
1959 (Supplements to Vetus Testamentum 7. Leiden). 283 91
Steele, Timothy. 1999. All the Fun’s in How You Say a Thing: An Explanation of
Meter and Versification. (Athens, Ohio)
Tarlinskaja, Marina. 1993. Strict Stress Meter in English Poetry Compared with
German and Russian. (Calgary, Alberta)
Vance, Donald R. 2001. The Question of Meter in Biblical Hebrew Poetry. (Studies in
Bible and Early Christianity 46. Lewiston NY)
Vetter, Paul. 1897. ‘Die Metrik des Buches Job’. Biblische Studien 2:4, 377 464
Weil, Daniel Meir. 1995. The Masoretic Chant of the Bible. (Jerusalem)
109
ROBERT D. HOLMSTEDT
UNIVERSITY OF TORONTO
Abstract
Introduction
111
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
112
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
113
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
114
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
115
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
Clause Type
The criterion of clause type builds upon Greenberg’s initial approach;
however, typologists have since considerably refined the definition of
the appropriate basic clause. One example is in Anna Siewierska’s
study Word Order Rules, in which she defines the basic clause as ‘sty-
listically neutral, independent, indicative clauses with full noun
phrase (NP) participants, where the subject is definite, agentive and
human, the object is a definite semantic patient, and the verb repre-
sents an action, not a state or an event.’9 Admittedly, clauses of the
sort described by Siewierska may not occur in abundance in a typical
7
The overwhelming majority opinion is that VS is basic in BH. Advocates of
an SV approach are rare; see P. Joüon, Grammaire de l’Hebreu biblique (Rome
1923); V. DeCaen, ‘On the Placement and Interpretation of the Verb in Standard
Biblical Hebrew Prose’, unpublished Ph.D. thesis (University of Toronto 1995);
R.D. Holmstedt, ‘The Relative Clause in Biblical Hebrew: A Linguistic Analysis’,
unpublished Ph.D. thesis (University of Wisconsin 2002); idem, ‘Word Order in
the Book of Proverbs’, in R.L. Troxel, K.G. Friebel, and D.R. Magary (eds), Seeking
Out the Wisdom of the Ancients: Essays Offered to Honor Michael V. Fox on the Occa
sion of His Sixty Fifth Birthday (Winona Lake, IN 2005), 135 54; idem, ‘Issues in
the Linguistic Analysis of a Dead Language, with Particular Reference to Ancient
Hebrew’, The Journal of Hebrew Scriptures 6:11 (2006), 1 21.
8
M. Dryer, ‘Word Order’, in T. Shopen (ed.), Language Typology and Syntactic
Description, Vol. I: Clause Structure, 2nd edn (Cambridge 2007), 61 131.
9
A. Siewierska, Word Order Rules (London 1988), 8.
116
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
Frequency
Those who assign primacy, or at least significance, to statistics use
the frequency criterion. Whereas with the first criterion a basic clause
type may be in the statistical minority, the frequency approach de-
mands that the basic order designation be assigned to a statistically
dominant pattern.12 This criterion is one of the most common in
BH studies and Takamitsu Muraoka summarized it succinctly in his
study of emphatic structures in BH: ‘[W]e are not interested in dis-
cussing the theory that [VS] order is normal because action is the
most important piece of information to be conveyed by this sentence
type called verbal clause. In other words, by saying that V-S is the
normal word-order we do not mean that it is logically or intrinsically
so, but simply statistically.’13
The problems in using simple statistical dominance to determine
basic word order are weighty, however. For example, given a 2:1 ratio
of VS to SV order in a given corpus, are we justified in classifying
that language as VS? The problem is exacerbated when the statistics
10
Ibid., 8 14.
11
For a concise summary of the basic issues involved in the typological quest
for determining ‘basic word order’ in any given language, see F.J. Newmeyer, Lan
guage Form and Language Function (Cambridge, MA 1998), 330 7.
12
See, for example, J.A. Hawkins, Word Order Universals (New York 1983).
13
T. Muraoka, Emphatic Words and Structures in Biblical Hebrew (Jerusalem
1985), 30.
117
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
are even closer, as Matthew Dryer notes: ‘In the Auk dialect of
Tlingit, for example, a text count … for the order of subject and
verb revealed VS outnumbering SV by 177 to 156. In a case like
this, the difference in frequency is sufficiently small for it not to
seem reasonable to say that VS is more frequent than SV or that VS
is basic.’14 Additionally, since a given text type, or genre, may be asso-
ciated with specific clause types, such as the strong association be-
tween BH narrative and the wayyiqtol form, clause type frequency
must be qualified appropriately and may not necessarily represent
the basic word order in the grammar of that language (as opposed to
the narrower ‘grammar’ of a specific text or genre). For instance,
consider the raw numbers from Ruth and Jonah, provided in (13).
(13) Simple Count of Verbs and Clauses in Ruth and Jonah
Verb Count from the Book of Ruth Verb Count from the Book of Jonah
420 Verbs 201 Verbs
313 Finite Verbs 149 finite verbs
138 wayyiqtol 84 wayyiqtol
97 Perfective 38 Perfective
78 Imperfective 27 Imperfective
Clauses with Explicit Subjects Clauses with Explicit Subjects
47 VS clauses 44 VS clauses
19 VS wayyiqtol 35 VS wayyiqtol
26 XVS (10 modal, 3 X, 13 C) 8 XVS (5 modal, 3 C)
2 VS 1 VS
22 SV 17 SV
20 SV 13 SV
2 XSV 1 XSV
The relative numbers from both books are quite similar. Of the
total verbs, seventy-five percent are finite verbs, the type which are
typically identified as necessary for the basic clause type. Of the
finite verbs,15 forty-four percent in Ruth and fifty-six percent in
Jonah are the past narrative wayyiqtol. It is thus immediately clear
that one particular form is strongly associated with narrative. More
important are the clauses with overt subject constituents. At first
14
Dryer, ‘Word Order’, 74.
15
Since finite verbs can be associated with syntactic subjects, they meet the
minimal qualification for a word order study that is interested in the relative order
of the subject and verb. Non finite verbs (i.e. infinitives, imperatives, and partici
ples) exhibit additional syntactic complications and should be initially excluded in
basic word order analyses.
118
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
blush it appears that both books favour VS order by a 2:1 ratio over
SV. However, this ratio covers a great number of complicating factors
involving both the basic clause type and pragmatics criteria. At this
point, it will suffice to point out that the chart in (13) makes it clear
that when we consider only clauses without subordinating constitu-
ents (‘C’), fronted phrases (‘X’), or modal verbs, SV is favoured by
10:1, if not more, in both books.
Distribution
The third approach is based on the test of distribution. Given two or
more alternatives for a syntactic construction, the one that occurs in
the greater number of environments is the basic order. Note that this
is not the same as statistical dominance, because the issue at hand is
not ‘occurrence’ but ‘environment’. For instance, in English, manner
adverbs may both precede and follow the verb, as in (14)–(15):
(14) a. Ethan slowly walked into the room.
ADV V
b. Ethan walked into the room slowly.
V ADV
(15) a. ?*Ethan is slowly walking.
ADV V
b. Ethan is walking slowly.
V ADV
Although both options exist in English, based on the distribu-
tional patterns, it can be argued that V-ADV order is basic because
there are environments in which the order ADV-V is not used or is
less felicitous, e.g. (15a).16
Pragmatics
Finally, we come to the criterion of pragmatics. Attention to the
pragmatic features of clauses is particularly significant for so-called
‘free-order’ languages like Hebrew, that is, languages exhibiting a
great deal of word order variation and no immediately apparent basic
order. At the core of this approach is the recognition that the major-
ity of language data contains pragmatically ‘marked’, or ‘non-neu-
tral’, clauses. Even for languages that have a more rigid word order,
such as English, pragmatics can produce extreme but grammatically
16
Ibid., 69.
119
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
When the issue of basic word order in BH, specifically the order of
the verb and its specifier, the subject (i.e. VS or SV), is approached
from a generative perspective, constituent movement becomes a criti-
cal feature in the analysis. By way of a brief orientation, generative
analysis has determined that initial derivations (we could call these
‘clauses-in-the-making’) start with the subject preceding the verb.
Since within the generative approach many constituents in the clause
‘move’ from this starting position to higher positions in the clause
(that is, towards the front of the clause), it is possible for this deriva-
17
Ibid., 76.
18
Marianne Mithun (‘Is Basic Word Order Universal?’, in D.L. Payne (ed.),
Pragmatics of Word Order Flexibility [Amsterdam 1992], 15 61) questions whether
some languages can be assigned to a typologically word order category. In particu
lar, for languages with an apparently ‘free word order’, Mithun argues that we
should not be looking for a basic word order in terms of the position of subject,
verb and modifiers. Rather, she suggests that in these languages it might be the case
that the syntactic role of an item (subject, object, etc.) is less important than its dis
course role (e.g. topic hood, identifiability, ‘newsworthiness’). Thus, the order of
the constituents, subject noun phrase, verb, complements, etc., will change in a ‘ba
sic clause’, depending on the information status of the constituents. I do not think
that Mithun’s observations obviate a basic word order discussion for such lan
guages. I suggest that in clauses in which the constituents all share the same prag
matic marking, e.g. all the constituents are ‘new’, such as in presentative clauses, the
observable order could be identified as basic.
120
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
121
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
22
German, Dutch and standard Afrikaans are considered to be ‘well behaved
V2 languages’ (i.e., they never allow the co occurrence of a complementizer and V2
verb in complement clauses of matrix verbs); Danish, Swedish, Norwegian and Fri
sian are examples of languages categorized as ‘limited embedded V2 languages’ (i.e.
only with bridge verbs such as ‘know that…’); Yiddish and Icelandic are examples
of ‘general embedded V2 languages’ (i.e. they permit the co occurrence of
complementizers and V2 in all the complements of all matrix verbs). English is an
example of a ‘residual V2’ language, in which we find V2 phenomena in, e.g. ques
tions, as in example (19) in the main text above. See S. Vikner, Verb Movement and
Expletive Subjects in the Germanic Languages (New York 1995).
23
The derivational approach to clause construction has been shown over the
last four decades to fulfill the requirements of descriptive and explanatory adequacy
as well as theoretical economy. For instance, a grammar that includes a base word
order with a few operations motivating constituent movement, which in turn may
result in further movement, is inherently less burdensome than a taxonomic gram
mar, that is, one that simply lists the numerous, almost infinite permutations. And
if the derivation based grammar explains all and only the grammatical and felici
tous examples, then it is to be highly preferred.
122
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
(19) English
(a) She saw the new Hummer. (S-V-O)
(b) [Which new car] did she see? (O-V-S, where the tensed auxiliary = V)
(c) [What] did she see? (O-V-S, where the tensed auxiliary = V)
The indicative statement in (19a) represents the base word order
in English: SVO. But in both interrogative statements in (19b)–
(19c), we find not only the fronting of the WH-phrase (‘which new
car’ and ‘what’), but also the inversion to VS order (where the salient
verb is the tensed form of do). This variation is often labeled ‘residual
verb-second’ and, given the history of English, it is not surprising
that it has retained some Germanic syntactic features.
Constituent movement may also be semantically driven, as in
(20), where the examples from Kru, a Niger-Congo language, exhibit
a switch from SVO order in typical declarative clauses (20a) to SOV
in clauses with a negative function word (20b).24
(20) Kru (Niger-Congo)
(a) Nyeyu-na bla nyino-na (S-V-O)
man-the beat woman-the
the man beat the woman
(b) Nyeyu-na si nyino-na bla (S-O-V)
man-the NEG woman-the beat
the man did not beat the woman
Additionally, some African languages vary the word order depend-
ing on the tense and aspect of the verb used; the Sudanic languages
Lendu, Moru, Mangbetu, and the Gur languages Natioro and
Bagassi exhibit SVO order with the perfective verbs and SOV with
imperfective verbs.25 Similarly, the Sudanic language Anyuak/Anywa
appears to switch from SVO in the present tense to SOV in the past
and future.26
A helpful comparison for BH is that of the formal registers of
modern Israeli Hebrew: when a constituent precedes the subject and
verb, the normal SV order is inverted to VS, as in the clauses in ex-
ample (21).27
24
T. Givón, On Understanding Grammar (New York 1979), 124 5.
25
A. Siewierska, Word Order Rules (London 1988), 95.
26
C. Perner, Anyuak: A Luo Language of the Southern Sudan (New Haven, CT
1990); M. Reh, Anywa Language: Description and Internal Reconstructions (Köln
1996).
27
L. Glinert, The Grammar of Modern Hebrew (Cambridge 1989), 417; idem,
Modern Hebrew: An Essential Grammar, 3rd edn (New York 2005), 162 4.
123
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
28
U. Shlonsky and E. Doron, ‘Verb Second in Hebrew’, in D. Bates (ed.), The
Proceedings of the Tenth West Coast Conference on Formal Linguistics (Stanford 1992),
431 45; E. Doron, ‘Word Order in Hebrew’, in J. Lecarme, J. Lowenstamm, and
U. Shlonsky (eds), Research in Afroasiatic Grammar: Papers from the third conference
on Afroasiatic Languages (Amsterdam 1996), 41 56; U. Shlonsky, Clause Structure
and Word Order in Hebrew and Arabic (Oxford 1997).
29
Note that the use of SV order to mark narrative transitions, e.g. stages in the
plot development or ‘scene’ changes, is a literary device not a formal pragmatic op
eration.
30
Jon. 1:6 is the sole example in Ruth and Jonah of a clause with an explicit
modal word with a VS clause. For examples of triggered VS order with explicit
modal words outside of Ruth and Jonah (with the function word ’im ‘if ’), see Gen.
47:16, 18; Exod. 22:2; Lev. 13:56; Num. 14:8; 21:9; 30:6; 1 Sam. 21:5; 2 Sam.
124
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
5:6; Isa. 4:4 5; 6:11; 24:13; Ezek. 16:48; Job 31:9; 37:20; Prov. 23:15; Song
7:13; Eccl. 10:10.
31
See also 1:9; 2:4, 12 (2x), 19; 4:11, 12, 14; Jon. 1:11, 12; 3:8b, 9.
32
What the function word within the wayyiqtol was originally is unknown.
That it is a function word best explains the phenomenon of this form: the raised
verb (due to triggering) is then morphologically fused with the unknown function
word that the gemination represents as well as the introductory simple conjunction
‘and’. This accounts for the fixed VS order we always see with the wayyiqtol. For
surveys of both classical and modern proposals regarding the history and semantics
of the underspecified function word present in the wayyiqtol form see L. McFall,
The Enigma of the Hebrew Verbal System: Solutions from Ewald to the Present Day
(Sheffield 1982), 217 18; B.K. Waltke and M. O’Connor, An Introduction to Bibli
cal Hebrew Syntax (Winona Lake, IN 1990), 544 5; W.R. Garr, ‘Driver’s Treatise
and the Study of Hebrew: Then and Now’ (Preface to reprint of S.R. Driver’s A
Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions.
Grand Rapids, MI 1998), xviii lxxxvi.
125
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
Information Structure
33
J. Firbas, ‘On Defining the Theme in Functional Sentence Analysis’, Travaux
Linguistiques de Prague 1 (1966), 267 80, esp. p. 272; idem, ‘ On the Delimitation
of the Theme in Functional Sentence Perspective’, in R. Dirven and V. Fried (eds),
Functionalism in Linguistics (Amsterdam 1987), 137 56; idem, Functional Sentence
Perspective in Written and Spoken Communication (Cambridge 1992), 72 3; see also
M.A.K. Halliday, An Introduction to Functional Grammar (London 1985), 38.
34
See J. Lyons, Semantics (Cambridge 1977), 509; K. Lambrecht, Information
Structure and Sentence Form: Topic, Focus, and the Mental Representations of Dis
course Referents (Cambridge 1994), 206 18. For an explicitly generative formula
tion, see M.S. Rochemont, Focus in Generative Grammar (Amsterdam 1986), 9 10.
126
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
127
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
36
For discussions of Topic and Focus, particularly those set weakly or strongly
within a generative framework, see Rochemont, Focus; E. Vallduví and E. Engdahl,
‘The Linguistic Realization of Information Packaging’, Linguistics 34:3 (1996),
459 519; E. Vallduví and M. Vilkuna, ‘On Rheme and Kontrast’, in P. Culicover
and L. McNally (eds), The Limits of Syntax (San Diego 1998), 79 108;
Zubizarreta, Prosody. See also G. Rebuschi and L. Tuller (eds), The Grammar of Fo
cus (Amsterdam 1999); A. Meinunger, Syntactic Aspects of Topic and Comment (Am
sterdam 2000).
37
Neither Ruth nor Jonah contains ‘Scene setting’ Topic VS (triggered) exam
ples of the simple type like Gen. 8:14 (ûbaÌodes hassenî b¢sib¨â w¢¨esrîm yôm laÌodes
yabsâ ha’areÒ ‘And in the second month on the twenty seventh day of the month the
earth dried up’). The only qualifying example is perhaps Ruth 1:17 with the adverb
kô ‘thus, in this manner/way’, e.g. ko ya¨asê yhwh lî ‘In this way Yhwh will act to
wards me’.
128
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
129
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
fronting. But we find very few, and the simple reason for this is that
narratives are informationally complex. They always contain multi-
ple themes as they develop, and so the only place that can contain a
clause without at least a Topic constituent is at the beginning of the
narrative or scenes with new characters.39 A parade example is the
first verse of the book of Job, provided in (38).
(38) Job 1:1 (S-V-PP; All Rhematic, no Topic or Focus)
’is hayâ b¢’ereÒ ¨ûÒ
‘a man was in the land of Uz’
We have only one example like this in the book of Ruth, at the
outset of the genealogy in chapter four, given in (39), repeated from
(23).
(39) (=23) Ruth 4:18b (S-V-O; S=Theme, V-O=Rheme; No Topic or Focus)
pereÒ holid ’et ÌeÒron
‘Perez begat Hezron’
The SV clause presents us with one old entity, pereÒ, and two new
pieces of information, a new verb and object. Critically, we cannot
analyse the subject pereÒ as a Topic, because there are no other the-
matic entities from which to choose. The NP pereÒ is the only agentive
entity available from the preceding clause to serve as the subject of the
verb holid. Topic does not function redundantly in this way. Aside
from these examples, we, therefore, have clear cases of basic SV word
order in Ruth and Jonah, but, as I have explained, this is expected.
In accordance with the features of narrative, the remaining SV ex-
amples present us with either Topic or Focus information. Consider
example (40):
(40) Ruth 4:1 (S-V-ADV; S=Topic)
ûbo¨az ¨alâ hassa¨ar
‘and Boaz went up to the gate’
This SV example orients the reader to which character is acting at
a major transition in the book: Boaz. At the beginning of this new
scene, the use of Topic-fronting indicates which of the known char-
acters will carry the plot forward. Similarly, consider (41).
(41) Ruth 3:4 (S-V-PP-O; S=Topic)
w¢hû’ yaggid lak ’et ’aser ta¨asin
‘and he will tell you what you should do’
39
Besides the initiation of narratives or new scenes with new characters,
genealogies and proverbs are the only other consistent source of basic SV clauses.
130
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
This example makes it clear that personal pronouns used with fi-
nite verbs do not always present contrasts (or better, Focus). In fact,
most of the occurrences of personal pronouns are used to present
Topics (see also Ruth 1:22, 2:13).
Example (42) adds an important piece of information: some lexi-
cal and grammatical items do not function as syntactic operators,
that is, they do not trigger inversion; hinnê is one of these non-op-
erators (see also Ruth 3:2, 4:1, and example [48], below).
(42) Ruth 2:4 (S-V-PP; S=TOPIC)
w¢hinnê bo¨az ba’ mibbêt leÌem
‘and surprise! Boaz came from Bethlehem’
As in Ruth, the first two SV clauses in the book of Jonah present
us with Topic entities: in 1:4a it is Yhwh that is a fronted Topic, in
1:4c it is the ship upon which Jonah was sailing. Consider (43),
which provides 1:4c:
(43) Jon. 1:4c (S-V-O-PP; S=Topic)
w¢ha’oniyyâ Ìiss¢bâ l¢hissaber
‘and the ship was about to break up’
Here we have a case of a thematic entity, the ship (introduced al-
ready in v. 3), fronted in order to orient the reader to a new Topic.
The narrator had been talking about Yhwh, but has now shifted to a
new Topic, the ship. The only other simple Topic-fronting clause in
Jonah is in 3:3, given in (44).
(44) Jon. 3:3 (S-V-PREDNOM; S= Topic)
w¢ninwê haytâ ¨ir g¢dolâ le’lohim
‘and Nineveh was a great city to the gods’
Nineveh has been a thematic constituent since the second verse of
the book, and it is again employed immediately preceding this
clause, but as a oblique argument. In (44) it takes on a nominative
subject role and is fronted as a Topic entity, marking Nineveh as the
item out of all the possible thematic constituents to be modified by a
predication.
While (38)–(44) illustrate a subject-Topic structure, in (45) we see
an example of an object undergoing Topic-fronting.
(45) Ruth 4:3 (O-V-S; O=Topic)
Ìelqat hassadê ’aser l¢’aÌinû le’elimelek makrâ no¨omi
‘Naomi is selling40 the portion of the field that belongs to our kinsman,
Elimelek’
40
I take makrâ in this clause as a perfect verb used performatively, i.e. ‘Naomi
hereby puts up for sale’.
131
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
41
It is possible that the clause below, from Ruth 3:11, is another example of a
fronted Topic object, but is not entirely clear whether the fronted object functions
as the Topic or a Focus (see also Ruth 3:5).
Ruth 3:11 (O V PP; O=Topic)
kol ’aser to’mri ’e¨esê llak
‘all that you say I will do for you’ (Ruth 3:11)
132
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
133
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
subjects to be focused like this would require the pronoun ’att to ex-
ist in the second half. The pronoun does not exist in this clause, thus
such a reading is not available. Instead, Naomi, the speaker, is con-
trasting the courses of action that the two daughters-in-law have
taken: one returned, one stayed.
As a last example of single-item Focus-fronting, consider (49), in
which we see a Focus-fronted adverb. Notice, by the way, that the
Focus-fronting of the adverb triggers VS inversion.
(49) Ruth 1:21a (ADV-V-S; ADV=Focus)
w¢rêqam hesibani yhwh
‘(and I full went away) but empty Yhwh returned me’
The adverb in (49) is Focus-fronted to highlight the contrast be-
tween the manner in which Naomi left Israel, ‘full’, and in her opin-
ion the manner in which Yhwh has brought her back from Moab,
‘empty’. And, of course, this contrast establishes a dominant motif in
the book as a whole.
In example (50), the initial function word ki should trigger VS in-
version, yet we have SV order.
(50) Ruth 4:15 (C-S-V; S=Focus)
ki kallatek ’aser ’ahebatek y¢ladattû
‘because your daughter-in-law who loves you bore him’
It is clear that Topic and Focus-fronting are movement operations
that occur after the syntactic triggering process that produces VS in-
version. So, in this case, the Focus-fronted subject phrase kallatek
’aser ’ahebatek is moved to its position after VS inversion, a move
that results in a surface order of SV.
For example (51), the subject hammawet is fronted to contrast it
not with contextual alternatives, but with logical alternatives —
those established solely from the shared knowledge of the speaker-lis-
tener outside of a particular discourse.
(51) Ruth 1:17 (C-S-V-PP; S=Focus)
ki hammawet yaprid bêni ûbênek
‘indeed (only) death will separate me and you’
So hammawet is contrasted with, basically, anything else that typi-
cally might be a reason for a widowed daughter-in-law to leave her
mother-in-law, such as other family or new marriage. The addition
of the English restrictive adverb only better captures this particular
Focus structure for us than simply giving heavy stress to the word
hammawet. Crucially, the ability of the ancient Hebrew CP to con-
tain both a subordinating function word, such as ki in (51) and a
134
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
135
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
The clause in (55) presents us with the first part of Naomi’s com-
plaint, the second part of which we examined in (49). Not only is
the adjective m¢le’â (used adverbially here) placed before the verb, so
too is the personal pronoun ’ani. The pronoun orients the reader to
the desired Theme — in the preceding verse, Naomi had just stated
‘Shaddai has made me very bitter’, which means that there were at
least two possible thematic items from which to select. The fact that
Naomi changes from Shaddai to herself as the subject of the next
clause is the motivation for the Topic pronoun referring to herself.
The adverbial m¢le’â is then Focus-fronted and is used to create the
contrast between m¢le’â and rêqam that the next clause (presented in
[49]) completes.
While the author of Ruth did not frequently employ multiple-
fronting, the author of Jonah clearly found it to be a useful strategy.
There are three examples of double fronting in the shorter book of
Jonah (1:14; 2:9; 4:11), all of which are similar to what we saw in
Ruth 1:21. More complex is the example of what is likely a triple-
fronting in Jon. 2:10, provided in (56).
(56) Jon. 2:10a (S-PP-V-PP; S=Topic; PP=Focus; V=Focus)
wa’anî b¢qôl tôdã ’ezb¢Ìã la
‘but I, with a voice of thanks, shall sacrifice to you’
In (56) the unambiguously modal verb ’ezb¢Ìâ is preceded by
both the subject pronoun ’ani and prepositional phrase b¢qol todâ.
Why? First, the Topic-fronted subject pronoun orients the reader to
the fact that the next predication will concern the referent of ’ani
(which is, of course, Jonah), not the previously modified ‘adherents
of worthless idols’. Then the Focus-fronted PP b¢qol todâ presents a
contrast between what Jonah’s manner of action and the implied
manner of thanklessness or silence of those in the previous clause,
Jon. 2:9. Finally, the verb ’ezb¢Ìâ also carries Focus in order to con-
trast how those from v. 9 ‘abandon’ their faithfulness while Jonah not
only remains faithful but intends to ‘offer a sacrifice’ to God.44
verb obviously means that the verb cannot be in the second syntactic position in
the clause. Hence, the motivation for using the more general reference ‘triggered
inversion’ (after Shlonsky, Clause Structure).
44
It is possible that Ruth 1:14b is also a case of double Focus, with both Ruth
and the verb ‘clung’ marked for a contrast with Orpah and ‘kissed’. However, in
order for the verbal contrast to make sense logically, we must assume semantic
gapping and reconstruct it: Orpah kissed her mother in law and left, but Ruth did
not kiss (?) her mother in law and clung to her. Perhaps, but I am doubtful. Rather,
this clause is best taken as a case of subject Focus fronting, or perhaps Focus on
both the subject and verb.
136
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
From the data in Ruth and Jonah, it appears that for these authors
the domain of fronting, what is also called the ‘left periphery’, has
the structure given in (57), which nicely accords with a great deal of
current research on the architecture of the left periphery and
fronting phenomena.
(57)
3
0
CP Domain of Triggers
0
0
for raising V over
0
canonical S position
0
2
C TopP
0
0
0
0
0
Topic-fronted FocP
1
phrase(s)
Focus-fronted TP
phrase(s)
core SV clause
Conclusion
The majority of the clause types in Ruth and Jonah (or in any bibli-
cal book) do not fulfil the typological criteria used to determine ba-
sic word order, especially the basic clause type criterion. This means
that at the centre of the VS versus SV basic word order argument
stands a small set of clauses, but this is an important set if we want to
classify ancient Hebrew typologically and account for information
structure accurately. If we start with the VS position, a necessary po-
sition is that no SV clause lacks a Topic or Focus operator, but VS
45
For example, Beninca’ and Poletto, ‘Topic, Focus, and V2’, argue for Italian
that ‘the encoding of informational relations in the syntax of the left periphery fol
lows a very precise semantic path’, which they identify as: [Topic [Hanging Topic
[Scene Setting Constituents [Left Dislocation [List Interpretation]]]]][Focus
[Contr. Adverbs/Objects [Contr. Circum./Quant. Adverbs [Informational Fo
cus]]]].
137
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
46
Note that if there were any Focus marking for Naomi, which is tempting to
read, then we should have the active verb with Naomi as an explicit subject. In
stead, the use of the passive clause in Ruth 4:17 makes it clear that the Focus is on
the event, not the participants.
47
The sole VS example in Jonah is in the poem, in 2:6a:
Jon. 2:6a (V S; chiasm)
apapûnî mayim ¨ad nepes t¢hôm y¢sob¢benî
‘(the) waters have encompassed me, up to (my) throat/life, the deep has
surrounded me’
In this example, the syntactically and semantically non triggered verb is not Focus
or Topic fronted either. Rather, this is a perfect example of chiasm at its most el
egant. The first colon presents V S order and the second presents S V order, with
the PP ‘up to throat’ acting as a Janus member, facing both cola.
138
WORD ORDER AND INFORMATION STRUCTURE IN RUTH AND JONAH
left-periphery unless there exists a focus item that has triggered such
movement; in other words, VS order is rare unless something else
precedes the verb.
In this article I have sketched a working model for investigating
issues of word order variation in BH, a model that is built upon both
typological and generative linguistics. Significantly, I have concluded
that the data suggest a SV analysis for BH rather than the conven-
tional VS analysis. Taking this empirically-driven SV conclusion as a
starting point for an analysis of information structure, I described a
framework for understanding the interaction of four core pragmatic
concepts: Theme, Rheme, Topic, and Focus. The resulting model of
core BH syntax, and the left periphery in particular, allows for the
type of flexibility, including multiple-fronting structures, that BH
exhibits.
139
FRANÇOIS BRON
EPHE, PARIS
Abstract
141
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
cription d’Altiburos’, JA (1887), 457 71; idem, Inscription néo punique d'Altiburos
(Ligne 8 et 9) (Paris 1891).
8
Ch. Clermont Ganneau, ‘Le “mazrah” et les curiae, collegia ou ordines cartha
ginois dans le Tarif des sacrifices de Marseille et dans les inscriptions néo puniques
de Maktar et d’Altiburos’, RAO, t. III (Paris 1900), 22 40.
9
Ibid, 332 3 et n. 1.
10
G.A. Cooke, A Textbook of North West Semitic Inscriptions (Oxford 1903),
no 55.
11
H. Donner et W. Röllig, Kanaanäische und aramäische Inschriften (Wiesbaden
1962 4), no 159.
12
K. Jongeling et R.M. Kerr, Late Punic Epigraphy (Tübingen 2005), 39 40.
See also now K. Jongeling, Handbook of Neo Punic Inscriptions (Tübingen 2008),
155 7.
13
M. Ennaïfer, op. cit. à la n. 2, pl. V. Nous remercions très vivement madame
P. Kalensky, du Musée du Louvre, qui nous a procuré une nouvelle photographie de
cette inscription (Fig. 1).
142
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
4) fils de Lbw’, Z¨lgm, fils de S†w¨n, Y¨stbgw, fils de Mshb’ et leurs collè-
gues, (qui composent) le mzrÌ.
5) … sur les sanctuaires, au mois de Krr de l’année de Yll, le sacrifica-
teur, fils de .g†¨n, avec
6) les suffètes Mshb’, fils de Yzrm, ¨Azarba¨al, fils de Barik et S.sln, fils
de Z¨zbl; et Mbyw, le scribe, qui
7) est sur eux (?), fils de Y¨†mn; étant prêtre de Ba¨al Îammon Wrwsn,
fils de Aris; parce qu’il a entendu leur voix, il les a bénis.
8) (Eux) qui ont offert ici son holocauste et la minÌa dans le sanctuaire
9) et qui, par surcroît, ont offert spontanément là un sacrifice molek.’
143
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
17
J. Friedrich, W. Röllig, M.G. Amadasi Guzzo et W.R. Mayer, Phönizisch
Punische Grammatik (Rome 1999), par. 147a.
18
Cf. M.G. Amadasi Guzzo et V. Karageorghis, Fouilles de Kition III. Inscrip
tions phéniciennes (Nicosie 1977), 184 5 (F 2).
19
E. Lipinski, Dieux et déesses de l'univers phénicien et punique (Louvain 1995),
373.
20
Clermont Ganneau, op. cit. à la n. 8, p. 332 3, n. 1.
144
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
Les l. 8–9 ont été l’objet d’une étude particulière de Berger,21 puis
elles ont été réexaminées à plusieurs reprises par J. Février.22 Pour
Berger, le rapport avec le texte précédent est clair: ‘c’est la constata-
tion de l’accomplissement du vœu’. Cela est possible; cependant, à
une époque où il est à la mode de dénoncer des faux, ces deux lignes
ont de quoi susciter les soupçons: tout d’abord, leur graphie très ma-
ladroite surprend, à la suite d’un texte officiel. Ensuite, la présence
d’une forme de hiphil est un hapax en phénico-punique. Enfin, le
mot ¨lt, ‘holocauste’, n’est pas connu par ailleurs en phénico-puni-
que, pourtant riche en termes relatifs aux sacrifices; en revanche, la
paronomase he¨elah ¨ôlah, ‘offrir un holocauste’, est bien attestée en
hébreu biblique.
L’inscription bilingue (KAI 160), découverte en 1908, a connu un
sort funeste: exposée à l’air libre sur le site, elle a été effacée par les
intempéries.23 Par la suite, la pierre a été transportée au Musée du
Bardo et l’inscription republiée, à partir des photographies ancien-
nes, par Z. Ben Abdallah.24 L’inscription latine avait été signalée tout
d’abord par A. Merlin,25 puis le texte punique a été étudié par E.
Vassel.26 Chabot y est revenu brièvement dans le Bulletin archéologi-
que de 1944.27 Mais la notoriété de cette inscription est due à un ar-
ticle de Février, au titre frappant.28 Par la suite, malgré une note de
Levi Della Vida,29 son interprétation a été reprise dans la première
édition de KAI et par diverses publications archéologiques.30
Seules les deux premières lignes du texte latin se prêtent à une res-
titution:
QVOD BONVM FAV[stum, fe-
l]IXQVE SIT SOD[alibus
Le terme crucial est celui de sodalibus qui désigne les membres
d’une corporation, d’un sodalicium.
21
Op. cit. à la n. 7.
22
J. G. Février, ‘Molchomor’, RHR 143 (1953), 8 18; idem, ‘Essai de reconsti
tution du sacrifice molek’, JA 248 (1960), 167 87.
23
Cf. M. Ennaïfer, op. cit. à la n. 2, p. 20 et n. 51.
24
Z. Ben Abdallah, Catalogue des inscriptions latines païennes du Musée du Bardo
(Rome 1986), 148, no 380.
25
A. Merlin, BAC (1908), 234 et pl. 45.
26
E. Vassel, ‘Etudes puniques V. Sur la bilingue d’Althiburos’, Revue tunisienne
(1916), 278 85.
27
J. B. Chabot, BAC (1944), 289 90.
28
J. G. Février, ‘Une corporation de l’encens à Althiburos’, Semitica 4 (1951
2), 19 24.
29
G. Levi Della Vida, ‘Iscrizione punica da Lepcis’, ANLR, ser. VIII, 10 (1955),
550 61: cf. p. 556, n. 2.
145
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
146
NOTES SUR LES INSCRIPTIONS NEO-PUNIQUES
Les lignes suivantes ne sont pas plus faciles: le mot clé en est pro-
bablement s¨t. L’étymologie proposée par Février n’est guère convain-
cante, d’autant plus que le sens exact de la racine ™¨Y est obscur,
aussi bien en ougaritique qu’en minéen. Chabot avait suggéré que ce
pourrait être, ‘dans un dialecte indigène, l’équivalent de sodalis’.32
C’est aussi l’idée de Krahmalkov, qui compare un akkadien se’u,
‘ami’.33 Ne serait-ce pas plutôt le correspondant d’ar. sî¨at-, ‘parti,
réunion d’hommes’, sabéen s¨t, ‘communauté, partisans’? A la l. 2, byt
s¨t désignerait donc ‘la maison de la corporation’. A la même ligne,
précédé de l-, nqsmy ne peut être qu’un infinitif construit niphal,
suivi d’un pronom suffixe.
A la fin de la l. 3, il faut peut-être lire k-ns’ s¨t, ‘car la corporation a
offert’, ce qui correspondrait au in] templo po[nere du texte latin.
A la l. 4, Clermont-Ganneau lisait S¨†rnyn’ bn S¨†r, avec un samech,
ce qui paraît justifié, vu la longueur de la hampe.34 S¨†rnyn’ est déjà
attesté dans une inscription de Sardaigne (KAI 173/7), alors que
S¨†rnyn’ n’est pas connu par ailleurs; les deux orthographes se rencon-
trent pour S/S¨†r.
Enfin, à la l. 5, on aurait mention d’un ‘comptable de la corpora-
tion’, Ìsb ou m]Ìsb s¨t.
Fig. 1.
32
J. B. Chabot, op.cit. à la n. 27.
33
Krahmalkov, op. cit., 476.
34
Lettre du 28 juillet 1913, citée par E. Vassel, op. cit. à la n. 26, p. 279.
147
TZVI NOVICK
NOTRE DAME
Abstract
The Paper argues that the word Òarîk in tannaitic Hebrew does not,
despite claims to the contrary, indicate permissibility. Analysis of
apparent counter examples illustrates the relationship between tele
ological necessity and weak deontic obligation.
The etymological meaning of the root Ò-r-k lies in the notion of lack
or need.1 From here it is a short step to the core modal usage of Òarîk
* I thank Elitzur Bar Asher, Dr Eliezer Diamond and Professor Christine Hayes
for their comments on earlier drafts of this article. This essay was made possible in
part by a grant from the Memorial Foundation for Jewish Culture.
1
See Ben Yehuda’s dictionary (A Complete Dictionary of Ancient and Modern
Hebrew, [New York 1959]), 5631 6). For discussion of Ugaritic and Aramaic cog
nates see E.Y. Kutscher, A History of the Hebrew Language (Jerusalem 1982), §123;
Isaac Gluska, Hebrew and Aramaic in Contact During the Tannaitic Period: A Socio
linguistic Approach (Tel Aviv 1999), 57 n. 338. For a full account of the verbal
forms of the root in tannaitic Hebrew see Menahem Moreshet, A Lexicon of the
New Verbs in Tannaitic Hebrew (Ramat Gan 1980), 312 13.
149
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
2
Such conditional sentences have been called ‘anankastic’, from Greek ânágkj.
See George Henrik von Wright, Norm and Action (London 1963), 10; Kai von
Fintel and Sabine Iatridou, ‘What to Do If You Want to Go to Harlem: Notes on
Anankastic Conditionals and Related Matters’ (Cambridge, Mass. 2004), 1 15
(ms. available at http://web.mit.edu/fintel/www/harlem.pdf ).
3
Unless otherwise qualified, citations to the Mishnah use ms. Kaufman; to the
Tosefta, ms. Vienna; to Mek. R. Ish., ms. Oxford 151; to the Sifra, ms. Assemani;
to Sifre Num. and Sifre Deut., ms. Vatican 32. For another case of teleological Òarîk
see m. Kel. 5:7, where Òarîk occurs in response to the question: ‘how does one pu
rify an impure oven?’ See also MenaÌem Zevi Kaddari, ‘On Deontic Modality in
Mishnaic Hebrew’, in Moshe Bar Asher (ed.), Studies in Mishnaic Hebrew (Scripta
Hierosolymitana 37, Jerusalem 1998), 207 8, 213.
4
Outside of this case, lehakrîz ‘to announce’ is always preceded in tannaitic lit
erature by Ìayyab.
150
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
mon meaning of ’ên Òarîk, ‘he need not’, would be nonsensical here,
as it is in no case obligatory to save one’s food from a fire.5
The next instance, Sifre Num. 68 (Horovitz ed., p. 63), involves a
debate about the identity of the characters in Num. 9:6 who had in-
curred corpse impurity and therefore did not participate in the
Passover sacrifice. R. Ishmael opines that these were the bearers of
Joseph’s coffin, while, in R. Akiva’s view, they were Mishael and
ElÒafan, who had removed the bodies of Nadav and Avihu from the
Temple (cf. Lev. 10:4). R. Isaac counters: ’ên Òarîk ‘that is not per-
missible (to say)’, because any of these figures would have contracted
corpse impurity long enough before the Passover that ‘they could
have purified themselves’. Rather, these were individuals who, imme-
diately before Passover, had encountered and buried an abandoned
corpse. As R. Isaac attempts to refute the views of R. Ishmael and
R. Akiva, and not simply to offer his own alternative, it is clear that,
with ’ên Òarîk, he means to claim that their views are not simply not
necessary, but in fact not supportable.6
In Sifre Deut. 16 (Finkelstein ed., pp. 26–7), the last pericope
alleged to attest negative Òarîk ‘not permissible’, it is reported that
R. Ishmael, when acting as judge in a case between an Israelite and a
non-Israelite, would always decide in favour of the Israelite, whether
by applying Torah law or the law of the nations. R. Shimon b.
Gamaliel comments: ’ênû Òarîk ‘it should not be done so’, but rather
the judgment should be fair, under whatever law the litigants agree
upon. Given that R. Shimon champions a different practice, we may
rightly conclude that he means to characterize R. Ishmael’s approach
not simply as unnecessary, but as incorrect.7
I have found many other clear instances of negative Òarîk with the
approximate sense ‘not permissible’ in tannaitic literature. Two exam-
ples will suffice. In t. Hor. 1:7, an interpretation of a biblical verse is
5
On this passage see Saul Lieberman, Tosefta Ki Fshu†ah (3rd ed.; New York
2001), 2.211.
6
On this passage see Menahem I. Kahana, ‘The Halakhic Midrashim,’ in S.
Safrai (ed.), The Literature of the Sages, Second Part (Philadelphia 2006), 37 8 and
n. 157. In the Bavli parallel (b. Suk. 25b), the printed editions and ms. Munich 95
omit ’ên Òarîk, but mss Harley 5508 and Ebr 134 include it.
7
On this passage see Saul Lieberman, Tosefet Rishonim (Jerusalem 1937 9),
1.155; Kahana, ‘Halakhic Midrashim’, 37 n. 157. The medieval commentary of
Pseudo Rabad (Herbert W. Basser [ed.], Pseudo Rabad: Commentary to Sifre Deuter
onomy [Atlanta 1994], 19) glosses ’ênû Òarîk with ’ênû yakôl in his explication of the
passage: ‘R. Shimon disagrees and says that if they came before an Israelite and
said, judge us according to Israelite law, he is not permitted (’ênû yakôl) to judge
according to gentile law to make the Israelite victorious’.
151
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
152
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
then do they die?’ This case is quite certain, but its status as tannaitic is question
able.
I have collected five instances in post tannaitic literature where Lieberman
makes a claim for Òarîk ‘possible, permissible’. Three (y. ¨Eruv. 2:3 [20a]; b. Shab.
121a; b. Yoma 88a) involve negation and two (y. Pes. 1:1 [27a]; y. Gi†. 6:2 [48a])
are positive. See Saul Lieberman, Yerushalmi Ki Fshuto (Jerusalem 1934), 252, 369
71; idem, Tosefet Rishonim, 1.155; idem, Tosefta Ki Fshutah, 2.822; 3.855). The
three negative cases are plausible. On the two positive instances see n. 22 below.
10
The use of Òerôk in connection with divine power, as in the above examples, is
very prevalent in TJ: when a prophet rejects the divine status of foreign gods or
kings (e.g., Isa. 37:19; Ezek. 28:2; Hos. 8:6), TJ denies of them Òerôk. See also TJ
to Judg. 6:31; 1 Kgs 18:21; 18:24; 2 Kgs 19:18; Isa. 41:23; Jer. 2:11; 10:11;
16:20; Jon. 1:5; Hab. 2:18. In all of these cases, where Òerôk is not followed by an
infinitival complement, it is possible to interpret the word as ‘benefit’ rather than
‘power’, but in light of the examples cited in the body of the text, where an
infinitival complement follows, and the meaning ‘power’ seems clearly called for, it
seems reasonable to assign the meaning ‘power’ in these cases also.
11
Mek. R. Ish. Ba Ìôdes 6 s.v. lô’ yihyeh (ed.Horovitz, p. 223) = Midr. Tannaim
Deut. 5:7 (Hoffmann ed., p. 20); Mek. R. Ish. Ba Ìôdes 6 s.v. kî ’anôkî (Horovitz
ed., p. 226); Sifre Deut. 318 (Finkelstein ed., p. 364) = Midr. Tannaim Deut.
153
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
32:17 (Hoffmann ed., p. 195). In the second of the three cases, Òorek alternates
with koaÌ ‘strength, power’. Another possible example occurs in Sifre Deut. 43
(Finkelstein ed., p. 98), where a king instructs his son, who is about to visit a ban
quet: ‘do not eat more than your Òorek, do not drink more than your Òorek’. The son
ignores his father, and eats and drinks ‘more than his Òorek’, and ends up vomiting.
It seems more reasonable to assume that Òorek here indicates one’s capability rather
than one’s need, as one can ordinarily eat and drink far in excess of one’s need with
out vomiting. Presumably, the semantic development of Òôrek and its Aramaic cog
nate proceeded in two stages: ‘a want, a need’ > ‘that which supplies a need, some
thing that confers benefit’ > ‘power, capability’. The first stage is widely attested in
Hebrew and Aramaic, for example, in Sifre Deut. 324 (Finkelstein ed., p. 375) ‘fat
things in which there is Òôrek’, which Reuven Hammer (Sifre: A Tannaitic Commen
tary on the Book of Deuteronomy [New Haven 1986], 336) renders as ‘fat things
which are nourishing’; in TJ, e.g., at Jer. 13:7; and in Samaritan Aramaic, as per Z.
Ben Hayyim, The Literary and Oral Traditions of Hebrew and Aramaic Amongst the
Samaritans (Jerusalem 1957 79), 3.2.179, 197. I have not found evidence of the
second stage outside TJ and the noted rabbinic sources, and it appears to have gone
unnoticed by lexicographers. The entries under Òôrek in Ben Yehuda’s dictionary
(Dictionary, 5640 6) and under Òôrek and Òerôk in that of Marcus Jastrow (A Dic
tionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Litera
ture [New York 1992], 1271, 1300) do not include the meaning ‘power, capability’.
Nor is this meaning recognized in translations of Targum Jonathan. Thus Daniel J.
Harrington and Anthony J. Saldarini (Targum Jonathan of the Former Prophets: In
troduction, Translation and Notes [Aramaic Bible 10; Wilmington 1987], 273), for
example, render the above phrase from 2 Kgs 5:7 thus: ‘Is there need from before
the Lord for me to kill and to let live?’
12
Note that in the clearest case, Sifre Zuta Num. 5:31, it occurs in the mouth
of a woman conversing with her friends, i.e., in a colloquial context.
13
There is some confusion, in the transmission of Ben Sira and tannaitic litera
ture, between Òôrek and Òarîk. See Z. Ben Hayyim, מסורת השומרונים וזיקתה למסורת
הלשון של מגילות ים המלח וללשון חז"ל, Lesonenu 22 (1958), 229 n. 15 ;אHanoch Yalon,
Introduction to the Vocalization of the Mishna (Jerusalem 1964), 102 4. It is note
worthy, however, that, while mss Kaufman and Parma of the Mishnah usually fa
vour ’ên Òôrek lômar over ’ên Òarîk lômar (Yalon, Introduction, 103), in the three
Mishnah pericopes cited above, supra at note 9, as attesting negative Òarîk with the
meaning ‘not permissible’, namely, m. Ta¨an. 2:3; m. Yoma 4:1; and m. So†. 9:6,
mss Kaufman and Parma consistently record Òarîk.
154
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
155
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
156
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
19
Lieberman, Tosefta Ki Fshu†ah, 2.396.
20
Lieberman, Tosefta Ki Fshu†ah, 2.402.
157
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
the eruv, to cede control’ (t. ¨Eruv. 5:12).21 Thus, while the distinc-
tion between Jew and Gentile occurs in teleological universals — the
Jew can make the eruv work by ceding control, the Gentile cannot
— the distinction between observant Jew and apostate occurs in
weak deontics: the former should cede control, and the latter should
not.
The last pericope, t. Ma¨as. Sh. 5:16, tells a story involving R.
Eliezer, who owned a vineyard near Lod. Produce of a vineyard’s
fourth year must be consumed in Jerusalem. Grapes grown at some
distance from Jerusalem may be redeemed on money, which must
then be taken to Jerusalem and spent there on food. Rabbinic decree
at first permitted redemption only for vineyards beyond a day-trip’s
distance from Jerusalem, but later extended this license to all vine-
yards beyond the city’s walls (t. Ma¨as. Sh. 5:14–15). The Tosefta
tells that R. Eliezer ‘did not want’ to redeem his vineyard. His stu-
dents told him: ‘now that, by decree, all vineyards up to the wall of
Jerusalem are subject to redemption, you are Òarîk to redeem it’. R.
Eliezer then arose, and harvested and redeemed his crop. On
Lieberman’s account, R. Eliezer, a traditionalist, was reluctant to rely
on the license to redeem. But his students informed him that the li-
cense to redeem entailed a stricture — it prohibited the sale of one’s
fruit within Jerusalem — and therefore R. Eliezer could (Òarîk) re-
deem, and need have no qualms about relying on a lenient ruling.22
Lieberman’s reading requires extensive inference beyond the plain
text. It seems far easier to suppose that Òarîk here indicates propriety.
R. Eliezer’s students tell him that because the rabbis have now de-
creed redemption permissible even to the wall of Jerusalem, it is
meet that he do so.23
21
On t. ¨Eruv. 5:11 see Lieberman, Tosefta Ki fshuta, 2.397 8.
22
Lieberman, Tosefta Ki Fshu†ah, 1.785. For R. Eliezer’s tendency toward strin
gency in legal practice, see Yitzhak D. Gilat, R. Eliezer ben Hyrcanus: A Scholar
Outcast (Ramat Gan 1984), 60 7. Lieberman’s exegesis seems in part motivated by
the aim of aligning the Tosefta story with its parallel in the Babylonian Talmud (R.
ha Sh. 31b; BeiÒ. 5a). According to the Bavli, R. Eliezer wished to convey his vine
yard to the poor. His students informed him that his colleagues had already ‘per
mitted’ (hatîrûhû) his vineyard. For further explanation see Rashi, ad loc., adopted
by Lieberman. Lieberman sees an equivalence between ‘permitted’ in the Bavli ver
sion and the Tosefta’s ‘you are Òarîk to redeem it’.
23
The two cases of positive Òarîk ‘permissible’ in the Yerushalmi alleged by
Lieberman (see above n. 8) are likewise dubious. On y. Pes. 1:1 (27a), two alterna
tive interpretations of the question in this passage, ‘as for dark alleyways, need/may
(Òarîk) one inspect them ab initio by candlelight?’ are described ad loc. in the com
mentary Qorban Ha ¨edâ, and there seems no reason to favour one over the other.
The solution to the other positive Òarîk, y. Gi†. 6:2 (48a) (‘a minor girl who is Òerîkâ
158
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
In this last case, Òarîk shades almost into the notion of permissibil-
ity. David Daube’s analysis of Latin oportere offers an enlightening
parallel:
Oportere signifies ‘it is correct, proper’. Hence it may refer not only to
what is necessary, to duties, but also to what is permitted, to rights. I
may say ‘The correct thing is to wear evening dress’, i.e. it is necessary;
but I may also say, ‘It is perfectly correct to wear a lounge suit’, i.e. it is
permitted.24
Thus, for example, the jurist Ulpian determines that if a party wall is
defective, ‘it was assuredly proper (oportuit) to pull it down’. But
Daube acknowledges that ‘even when oportet refers to rights, it does
not express complete freedom of choice. “The proper” always retains
a trace of “the requisite”’.25 In t. Ma¨as. Sh. 5:16 too, R. Eliezer’s stu-
dents do not simply state that R. Eliezer may redeem his fruits, but
that he may properly do so. They are not offering legal advice, but at-
tempting to persuade him by characterizing redemption of the fruit
as the ‘correct thing’ to do.
Positive Òarîk ‘ought’ occurs in a number of other cases. For exam-
ple, R. Judah b. Baba is quoted (t. Ter. 5:10) as insisting that ‘courts
should (Òerîkîn) conclude that all become nullified in one hundred
and one’, i.e., that heave-offerings of all species lose their status when
they fall into a mixture one hundred times larger than it.26 Likewise,
t. B.Q. 9:29 dictates that one who is injured by another should
(Òarîk) seek God’s mercy on behalf of the injurer even if the injurer
has not sought it.
to be divorced’), is probably to be found in y. Yev. 13:2 (13c), where the same
pericope is introduced in connection with a structurally parallel line from the
Mishnah (‘a minor girl who is Òerîkâ to refuse’) that uses Òarîk in the ordinary sense
of ‘necessary’.
24
David Daube, Forms of Roman Legislation (Oxford 1956), 12.
25
Daube, Forms, 14.
26
Roger Wertheimer (The Significance of Sense: Meaning, Modality, and Moral
ity [Ithaca 1972], 120 note *) observes that ought is rare in contemporary legislative
English. The one exceptional case that he cites seems quite analogous to t. Ter.
5:10: ‘The grand jury ought to find an indictment when all the evidence before it,
taken together, is such as in its judgment would, if unexplained or uncontradicted,
warrant a conviction by trial jury’. Wertheimer explains the significance of ought
here: ‘Given the special role of the grand jury … the intent and rationale of this
section seem clear enough. While the grand jury is not bound (‘must’ or ‘shall’) to
find an indictment in such cases, it is not simply empowered or permitted to do so.
Nor should one say that it is left to the grand jury’s discretion. Rather the grand
jury is provisionally or presumptively bound: i.e., they will find an indictment in
such cases, but if, in their judgment, there is some good reason for not doing so,
they are permitted to refrain’.
159
THE MODALITY OF ∑ARÎK IN TANNAITIC HEBREW
Conclusion
27
See Kaddari, ‘Deontic Modality’, 211, on Òarîk: ‘Sometimes society is the
source of the obligation and its essence is good advice, as in … “Three things a
man must [Òarîk] say in his house on Sabbath eve, towards dusk” [m. Shab. 2:7]; …
“for a man must [Òarîk] satisfy people just as he has to satisfy the Almighty’ [m.
Sheq. 3:2]”.’ While many examples of Òarîk ‘ought’ do have an interpersonal ele
ment, it is questionable whether society should be described as ‘the source of obli
gation’; we may note that m. Sheq. 3:2 cites two scriptural prooftexts. In any case,
both of the pericopes cited by Kaddari begin with Òarîk ’adam ‘one should’. It is
probable that the occurrence of ’adam as subject plays at least some role in signal
ling the weak deontic force of these declarations. Statements beginning Ìayyab
’adam ‘one must’ likewise, though to a lesser extent, tend to convey obligations that
are in one or another sense weak; see, e.g., m. Ber. 9:5: ‘One must bless over the
bad as one blesses over the good’.
28
It seems that Òarîk expressing strong obligation also tends to occur with
slightly greater frequency in the Tosefta than in the Mishnah. Thus, for example, in
the Mishnah, le¨asser is preceded in three pericopes by Ìayyab (m. Dem. 6:11 12;
m. Gi†. 5:4) and never by Òarîk. In the Tosefta, Òarîk le¨asser occurs in nine pericopes
(t. Dem. 8:1 [paralleling m. Dem. 6:12]; t. Ma¨as. 2:5; 2:10 12; t. Ma¨as. Sh. 3:7
8; t. Shab. 15:14) and Ìayyab le¨asser in only two pericopes, one paralleling Ìayyab
le¨asser in the Mishnah (t. Dem. 1:4 [paralleling m. Dem. 6:11]; t. Ma¨as. 2:1).
Likewise, t. Pes. 5:7 corresponds to m. Pes. 8:2, but the former twice has ‘one need
(Òarîk) not perform the Second Passover’ while the latter twice has ‘one is exempt
(pa†ûr) from performing the Second Passover’. More generally, in the parallel dis
cussion of the Second Passover in t. Pes. 7 8 and m. Pes. 8 9, the Tosefta without
exception uses ’ên Òarîk while the Mishnah equally consistently uses pa†ûr or an
imperfect verb. This distribution may suggest that Òarîk ‘must’ was also perceived as
colloquial, though not to the same degree as Òarîk ‘ought’.
160
SHLOMY RAISKIN
BAR-ILAN UNIVERSITY
Abstract
This article investigates the origin and meaning of the Aramaic fauna
names murzema and shaqi†na, mentioned in the Babylonian
Talmud (Tractate Îullin, p. 63a). Beginning with Akkadian and
Sumerian, the article surveys various roots in Semitic languages, as
well as Persian. Various commentaries on the Talmud as well as seveal
more contemporary sources indicate that the two birds can both
be identified as the Greater Flamingo. The final part of the article
concludes that the murzema is indeed the Greater Flamingo, which
was called by hunters in Arabic al mirzam.
Introduction
161
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
169, Munich 95 and Vatican 123 it would seem that murzema is not designated to be
a mnemonic sign at all, but rather a separate issue altogether from shaqi†na. This is
deduced from the fact that the Hamburg and Munich mss have a mnemonic sign
appearing before the word murzema which is ‘ אדומין כשריןred are kosher’, and in
Vatican 123 there is no mention of a mnemonic sign ( )וסימנךat all. Regarding the
word murzema itself in the Vilna edition it appears as מורזמא, in the Vatican 123
ms. it appears as ( מזרזמאmezarzema, however, there are additional changes in this
ms.), in Vatican 122 it appears as ( מורמזאmurmeza), and in the Munich ms. it is
'( מורזימmurzim[a]).
2
CAD = The Assyrian Dictionary of the Oriental Institute of the University of
Chicago (John A. Brinkman et al. eds), Chicago Il., 1984, Vol. 15 (S), 168.
3
Niek Veldhuis, Religion, Literature, and Scholarship: The Sumerian composition
Nanse and the birds, with a catalogue of Sumerian bird names (Leiden 2004), 250.
4
Michael Sokoloff, A Dictionary of Jewish Babylonian Aramaic of the Talmudic and
Geonic Periods (Ramat Gan 2002), 1173, s.v. ‘’שקט.
162
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
163
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
11
The root r m z is Arabic as well, with a similar meaning. Regarding the various
meanings of r m z see at length in C.E. Bosworth et al. (eds), The Encyclopedia of
Islam: New Edition Vol. VIII (Leiden 1995), 426 30.
12
See Menahem Zvi Kaddari, Milon Ha’ivrit Hamikra’it (Ramat Gan 2006), 998;
Wilhelm Gesenius, A Hebrew and English Lexicon of the Old Testament: with an ap
pendix containing the biblical Aramaic (Francis Brown, S.R. Driver and Charles
A. Briggs, [eds], Oxford 1906), 931, s.v. ‘’רזם. See also: W.F. Albright, ‘Notes on
Egypto Semitic Etymology’, American Journal of Semitic Languages and Literatures,
34, 2 (1918), 84. An opposite meaning exists in R. Chaim Dov Shevel’s article ‘פירוש
(’רבינו מיוחס בן אליהו על ספר איוב)ב, Hadarom, 23 (Nisan 1966), 95, item 12 and foot
note 14: ואין לו עד במקרא,"ומה ירזמון מפני מה עיניך פקוחות ומביטות בעזות, namely the
meaning is not ‘winking’ or ‘hinting’ eyes, but rather ‘open eyes gazing fiercely’. Eliezer
Ben Yehuda (1858 1922), in his dictionary Milon Ha’ivirt Hayeshana VeHahadasha
(N.H. Tur Sinai [ed.]), Vol. 13 (14) (Jerusalem 1952), 6517, footnote 7, refers to the
Arabic, with the meaning ‘weak' or ‘weary', and the verse’s meaning is accordingly:
‘why have your eyes grown weary of sight’. However, see tenth century commentator
Jehuda ben Koreisch, Epistola de Studii Targum Utilitate (J.J.L. Barges and D.B.
Goldberg [eds], Paris 1857), 57: ומה ירזמון עיניך… יריד בה ירמזון והו רמז בלשון משנה ורמיזה
ורמיזות וכדאך הו באלערביה רמז וארמאז ותפסירה ואשארה וקד דכר מתל ומה ירמזון ומה ירזמון פי
…חרוף אלתקדים ואלתאכיר מתל כבשה וכשבה ושמלה ושלמה.
13
Ludwig Lewysohn, Die Zoologie des Talmuds: eine umfassende Darstellung der
rabbinischen Zoologie, unter steter Vergleichung der Forschungen alterer und neuerer
Schriftsteller (Frankfurt a. M. 1858), 170 1.
14
Jacob Levy, Wörterbuch uber die Talmudim und Midraschim, Vol. III ע מ
(Berlin Wien 1924), 247 8.
15
Jacob Levy, ibid., Vol. IV ת פ, p. 602, s.v. ְקטנ ָא
ְ שִ or ִקיטנ ָא
ְ ש ְ . Incidentally, David
Golomb suggested that the origin of the word shaqi†na is from the Hebrew maskit
164
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
()משכית, which seems highly unlikely (see D. Golomb, Targumna, part III: Leviticus,
Vol. II (Warsaw Jerusalem 1938), 433).
16
Marcus Jastrow, Dictionary of the Targumim, the Talmud Babli and Yerushalmi,
and the Midrashic Literature (New York Berlin 1926), 1622, s.v. ‘ִקיטנ ָא ְ ש ְ ’.
17
Marcus Jastrow, ibid., p. 749, s.v. ‘’מוְּרז ְָמה, referred from ‘( ’ַמְרז ְָמאibid., p. 840).
18
Lexikon Der Abbreviaturen (Gesamellt von G.H. Händler, ‘Anhang zum
Aramäisch neuhebräisches Wörterbuch' von Gustaf H. Dalman, Frankfurt a. M.
1897), 241, s.v. ‘מרזמא, ’מורזמא, simply left a question mark next to the word murzema,
and with regard to shaqi†na (ibid., p. 413, s.v. ‘ִקיטנ ָא
ָ ש ְ ’) simply notes ‘eine Vogelart’
‘a type of bird’, with no specific identification.
19
Krupnik’s lexicon just states that murzema is a type of bird, see: Baruch Krupnik,
Milon Shimushi LaTamud LaMidrash VeLatargum (A Practical Dictionary for the Tal
mud, Midrash and Targum), Vol. II (London 1927), 19, s.v. ‘’מוְּרז ְָמא.
20
Michael Sokoloff., op. cit. (footnote No. 4), p. 1174, s.v. ‘שקטנאה, ’שקיטנאה.
21
Michael Sokoloff, op cit., p. 649, s.v. ‘’מורזמא.
22
Sheqitan is the Modern Hebrew term for flamingo.
23
Israel Aharoni, Memoirs of a Hebrew Zoologist (Jerusalem 2001), 254 (Hebrew).
24
Moshe Henech Bernstein, Olam Hatushiya (World of Wisdom, London 1937),
66.
25
Yisrael Meir Levinger, ‘’עוף טהור נאכל במסורת, Sinai, Vol. 64, No. 1 6/386 91
(Tishre Adar 1969), 269.
165
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
166
TALMUDIC ARAMAIC FAUNA NAMES: MURZEMA AND SHAQI™NA
33
EI, op. cit.
167
NADEZHDA VIDRO
UNIVERSITY OF CAMBRIDGE
Abstract
169
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
מנה$תצארא ישמל אבואבה אלא אלשאד$תצרת אלכאפי פי אללגה אכ$קד כנת אכ
תצר$מנה אלכאפי וסאל סאיל אן אכ$סוא מא חצל פיה מן אלזיאדאת מא לם יתצ
מל פיה$ כור יעקד פיה עקוד פי אלתצאריף ויג$תצר אלמד$ מן אלמכ$ר אוגז$תצר אכ$מכ
ע אליהא תסתועב פי אליסיר מן אלזמאן$ מלא מנהא תכון אצולא ירג$ ג
I had prepared an abridgment of the Sufficient Book on the Hebrew
Language which comprised all its chapters, with a few exceptions, as
well as some additions not included in the Sufficient Book. Then some-
one asked me to write another abridgment even more concise than the
one mentioned above in which the rules pertaining to conjugations
would be established and all conjugations would be brought together
as a basis to which one could refer and which could be studied in a
short time (Or. 5565 E, fol. 13r.).
The fragment discovered by Hirschfeld in the British Library
Genizah Collection consists of two leaves (folios 13 and 14 of Or.
5565 E, henceforth MS O) and contains the following parts of the
treatise: (1) the title (fol. 13r.); (2) the beginning of the introduction
(fol. 13v.); (3) the end of the chapter on words belonging to one or
more parts of speech (fol. 14r.); (4) the beginning of the chapter on
the conditions for forming morphological patterns (fol. 14r.–14v.)
Until now it was believed that the above-mentioned fragment was
all that survived of Kitab al-¨uqud. However, when editing an anony-
mous Karaite grammatical treatise named in the colophon al-
MukhtaÒar,3 this work being carried out as part of my Ph.D. research
at the University of Cambridge under the supervision of Prof. G.
Khan, I was able to prove its identity with Kitab al-¨uqud on the
grounds of the following evidence.
The most important MSS containing the text of al-MukhtaÒar are
II Firk. Evr.-Arab. I 2581 (henceforth MS A, 7 folios), and II Firk.
Evr.-Arab. I 2591 (henceforth MS B, 110 folios) preserved in the
second Firkovitch collection in the National Library of Russia in St.
Petersburg.4 These MSS clearly belong together, MS B being the
continuation of MS A. Together they contain the text of al-
MukhtaÒar from some place in the middle of the treatise to its end
(with a few lacunae). A comparison of these MSS with MS O reveals
3
This treatise was first discovered by M.N. Zislin. Cf. M.N. Zislin, ‘Eastern
School of Jewish Grammarians in X XII cent. (Résumé)’ (in Russian), in Semitskie
yazyki 2:2 (1965), 246 9, esp. 246 7. See also Geoffrey Khan, ‘The Medieval
Karaite Tradition of Hebrew Grammar’, Boletín de la Asociacion Española de
Orientalistas 38 (2002), 51 76, esp. 65; Khan et al. al Kitab al kafi, vol. 1, xxx.
4
Two other MSS I am currently aware of are MSS II Firk. Evr. Arab. I 2724
(23 folios) and II Firk. Evr. Arab. I 2594 (2 folios). They contain text copied from
several different places in the treatise.
170
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
that the last six lines of MS O, fol. 13v. are nearly identical with the
first seven lines of MS A, fol. 1v. Moreover, the text of MS O, fol.
14r.–14v. is contained in MS B, fol. 6r.–7r. with minor variations in
wording and orthography. These two overlaps suggest that MS O,
fol. 13–14 and MSS A and B are fragments of one and the same
grammatical work. The only obstacle to this identification is the fact
that MS O, fol. 13–14 is entitled Kitab al-¨uqud fi taÒarif al-lugha al-
¨ibraniyya whereas the text in MSS A and B is referred to as al-
MukhtaÒar in the body of the treatise as well as in the colophon.
However, as was already mentioned, Kitab al-¨uqud was intended as
an abridgment of al-Kitab al-kafi so that the term al-MukhtaÒar
should probably be regarded as a description of the nature of the
treatise rather than its title.
All the above leads to the conclusion that Kitab al-¨uqud, which
was previously considered lost except for the fragment published by
Hirschfeld, has in fact survived almost entirely and can be recon-
structed as follows:
5
Hirschfeld, ‘An Unknown Grammatical Work’, 2.
6
For recent accounts on Abu al Faraj Harun’s contribution to the development
of Karaite grammatical tradition, cf. for example, G. Khan, ‘Abu al Faraj Harun
and the Early Karaite Grammatical Tradition’, JJS 48 (1997), 314 34; idem. ‘The
Medieval Karaite Tradition’, 62 9; Khan et al., al Kitab al kafi, vol. 1, xi xlvi.
7
Cf. for example, Khan et al., al Kitab al kafi, vol. 1, xii; N. Basal, ‘Excerpts
from the Abridgment (al MuÌtaÒar) of al Kitab al kafi by Abu al Faraj Harun in
Arabic Script’, Israel Oriental Studies 17 (1997), 197 225, esp. 201.
171
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
duction to Kitab al-¨uqud 8 also appears on this list under the title al-
MukhtaÒar.9 However, Abu al-Faraj Harun’s authorship of these
books must now be questioned. Indeed, it was already shown10 that
the treatise contained in MS B cannot have been written by Abu al-
Faraj Harun, for this grammarian is mentioned in its text as a con-
temporary of the author:
הרון בן אלפרג$ אבי אלפרג$א והו אלשיך$ודה פי זמאננא הד$ כר מן כתבה מוג$וקד ד
…חרסה אללה
Someone, whose books are present in our time, namely the master
Abu al-Faraj Harun ibn al-Faraj, may God protect him, mentioned…
(MS B, fol. 9r.)
In as much as it has now been established that this treatise is a
part of Kitab al-¨uqud both al-MukhtaÒar and Kitab al-¨uqud must be
considered anonymous until further evidence is discovered.
Along the same lines, the generally accepted opinion on the mean-
ing of the book’s title as well as its purpose should now be reconsid-
ered. Hirschfeld translated Kitab al-¨uqud fi taÒarif al-lugha al-
¨ibraniyya as Pearl-Strings on the Grammatical Inflections of the Hebrew
Language and thought that the intention of the author in writing this
book was to compose ‘a vade mecum in which the bare grammatical
facts were strung together’.11 However, the term ¨uqud and its singular
form ¨aqd are frequently used in the newly identified sections of the
book12 to describe rules concerning derivational relations between
different forms of a verb, referred to as ¨uqud la tanÌallu, literally
‘knots that cannot be untied’ (sg. ¨aqd la yanÌallu), e.g.:
לך אלאמר אבדא נקטתין$ר מלוך ד$הא גיר אצלי יכון אכֵ רה$ואעלם אן כל אמר אכ
א עקד לא$ פהד. ַרבֶה צבאך ו]צ[אה.ְדעה חכמה לנפשךֶ עין והמא כן$מא סוא מוצ
י פי$לך אלאמר אלד$ ועקד תֿאן הו אן אלנקטתין אלתי פי ד.ע$א אלמוצ$ינחל פי הד
ועקד תֿאלתֿ הו:בר אליחיד אבדא אלי קאמצה
ַ ע
ָ הא גיר אצלי תנקלב פי
ֵ רה חרף$אכ
ה עקוד לא$ר עבר אלמונתֿ אבדא בתו והד$רה הא גיר אצלי יכון אכ$אן כל אמר אכ
תנחל
8
See quotation above.
9
This al MukhtaÒar is to be distinguished from the text of the same title
discussed above.
10
Khan, ‘The Medieval Karaite Tradition’, 65; Khan et al., al Kitab al kafi, vol.
1, xxx; Zislin, ‘Eastern school’, 24; M.N. Zislin (ed.), Me’or ¨Ayin (Moscow 1990),
esp. 17.
11
Hirschfeld, ‘An Unknown Grammatical Work’, 2. Hirschfeld’s opinion was
based on the author’s remark in the introduction ( יעקד פיה עקוד פי אלתצאריךMS O,
fol. 13v.).
12
This fact provides additional evidence in support of our identification.
172
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
Take note that the last vowel of each imperative ending in a non-root
heh is always Òere except in two passages, i.e. ( כן ְדעֶה חכמה לנפשךProv.
24:14) and ( ַרבֶה צבאך וצאהJudg. 9:29). This is a rule which cannot be
dissolved regarding this matter. The second rule is that the Òere in the
end of such imperatives ending in a non-root heh always turns into a
qamaÒ in the masculine singular past verb form. The third rule is that
every feminine singular past verb form derived from an imperative
ending in a non-root heh always ends in a taw. These are rules which
can never be dissolved (MS B, fol. 18r.).
This passage shows that the Judaeo-Arabic עקדmust indeed be
read ¨aqd (lit. ‘knot’) and not ¨iqd (‘pearl-string’) and translated in
the technical sense of ‘link, connection, relation, rule’. To the best of
my knowledge, no other work on grammar uses the word ¨aqd -¨uqud
with the same meaning. A parallel Hebrew term qeser (pl. qisronim)
occurs in Me’or ¨Ayin, an anonymous Karaite grammatical treatise
composed in Hebrew in the second half of the eleventh century,
where it is used to describe dependences between grammatical facts
similar to those found in Kitab al-¨uqud.13 However, considering that
Kitab al-¨uqud is one of the main sources of Me’or ¨Ayin,14 qeser here
is in all probability a translation of ¨aqd.
If this meaning of the term ¨uqud is established, the title of our
treatise must be translated Book of Rules Regarding the Grammatical
Inflections of the Hebrew Language rather than Pearl-Strings on the
Grammatical Inflections of the Hebrew Language, even though it was
probably intentionally ambiguous for rhetorical effect. In this case it
seems likely that it was the purpose of the author not simply to col-
lect grammatical facts pertaining to the conjugational patterns of
Hebrew verbs but to link them to one another and in doing so to
create a set of practical rules to facilitate the acquisition of the system
of Hebrew verbal derivation.
An additional purpose of the author, announced in the introduc-
tion in the words ‘( ויגמל פיה גמלא מנהאall conjugations would be
brought together’ [MS O, fol. 13v.]), was to discuss all conjugational
patterns of Biblical Hebrew. He did this following the method of
symbols common in Karaite linguistics before the twelfth century.15
13
Zislin, Me’or ¨Ayin, 139 44: for Zislin's interpretation of qeser as ‘relationship
between grammatical forms’, cf. ibid., 104.
14
Ibid., 17.
15
On this method, cf. ibid., 17 19, 20; D. Becker, ‘The “Ways” of the Hebrew
Verb according to the Karaite Grammarians Abu al Faraj Harun and the Author of
Me’or ¨Ayin’ (in Hebrew), in M.A. Friedman (ed.), Te¨udah 7 (Studies in Judaica,
Tel Aviv 1991), 249 75; A. Maman, ‘Medieval Grammatical Thought: Karaites
versus Rabbanites’ (in Hebrew), Language Studies 7 (1996), 79 96, esp. 87 96.
173
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
The idea of this method is to divide Hebrew verbs into groups such
that all verbs in a group have one particular vowel in the first (or the
last) syllable of their imperatives and one particular vowel in the first
(or the last) syllable of their past forms, which is different from that
of the imperative. These classes of verbs are then assigned mnemonic
symbols (Ar. ¨alama), i.e. disyllabic Hebrew words such that the first
vowel of the symbol word corresponds to the vowel common to all
imperatives and the last vowel of the symbol word corresponds to the
vowel common to all past forms in the group. For instance, the sym-
bol ִשיָרהrepresents verbs in which the first vowel of the imperative is
Ìireq and the first vowel of the past form is qamaÒ, e.g. ָשם, נ ָכון, נ ַָמק,
etc. By their definition, the symbols can comprise verbs of different
types so further division into conjugational patterns (Ar. taÒrif ) is
performed in order to group together verbs with a similar structure.
Then each conjugational pattern is described by citing a number of
forms of a representative verb with this pattern. Apart from Kitab al-
¨uqud, this method is currently known from Abu al-Faraj Harun’s
books al-Kitab al-mushtamil and al-Kitab al-kafi as well as from
Me’or ¨Ayin.
A comparative study of the accounts found in these four works is
beyond the scope of this article. However, it is worth noting that
Kitab al-¨uqud contains a much more elaborate description of the
verbal system of Biblical Hebrew than that found in al-Kitab al-kafi,
of which it was intended to be an abridgment. Indeed, Kitab al-
¨uqud presents twelve symbols comprising fifty-eight conjugational
patterns built on the basis of the first vowels of the imperative and
the past, as opposed to seven symbols and twenty-one conjugations
found in al-Kitab al-kafi. It also mentions five symbols with thirty-
five conjugational patterns built on the basis of the last vowels of the
imperative and the past, as opposed to three symbols with fifteen
conjugations found in al-Kitab al-kafi. Moreover, Abu al-Faraj
Harun limits his description of a conjugational pattern to four forms
of a representative verb, namely masculine singular imperative, third
person masculine singular past, active participle masculine singular
and passive participle masculine singular. This number was signifi-
cantly augmented by the author of Kitab al-¨uqud, who lists eighteen
forms, i.e. masculine singular imperative, third person masculine sin-
gular past, active participle in both genders and numbers, passive
participle in both genders and numbers, feminine singular impera-
tive, third person feminine singular past, masculine plural impera-
tive, masculine plural past and four forms of the future in the order
of the prefix letters אינת. This set of uninflected forms is often sup-
174
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
16
The hierarchy of chapters and sections is not very strict in Kitab al ¨uqud so
that chapters sometimes include sections on deviant topics.
17
Titles of sections given in italic are mine, not the author’s.
175
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
18
For other works on grammar using this terminology, cf. N. Allony. ‘Qarqa¨ot
ha Diqduq by Joshua ben Abraham: Three Judaeo Persian Fragments from the
Cairo Genizah’ (in Hebrew), in Y.D. Gilat, Ch. Levine and Z.M. Rabinowitz (eds),
Studies in Rabbinic Literature, Bible and Jewish History (Jerusalem 1982), 291 311,
esp. 297 8.
19
The introduction of this concept was previously attributed to Me’or ¨Ayin;
Zislin, Me’or ¨Ayin, 14; Maman, ‘Medieval Grammatical Thought’, 81 4; idem,
‘The Hebrew Alphabet as a Grammatical Mnemotechnic Framework: Introduction
to al Kitab al mushtamil, Part III’ (in Hebrew), Language Studies 8 (2001), 95 139,
esp. 103.
176
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
20
This of course does not mean that all of them are completely novel and do
not rely on other books by Abu al Faraj Harun or other authors.
177
A NEWLY RECONSTRUCTED KARAITE WORK ON HEBREW GRAMMAR
Conclusions
178
Abstract
Introduction
1
Îunayn’s version of Galen’s commentary, not extant, is noted in Fuat Sezgin,
Geschichte des arabischen Schrifttums, vol. 3 (Frankfurt 1970), 137. The latest inven
tory of Syriac versions of Galenic writings is Rainer Degen, ‘Galen im Syrischen:
Eine Ubersicht uber syrische Überlieferung der Werke Galens’, in Vivian Nutton
(ed.), Galen: Problems and Prospects (London 1981), 131 66.
2
Ibn Abi UÒaibi¨a, {Uyun al anba} fi †abaqat al a†ibba (Beirut: Dar al thaqafa,
third printing, 1981), part 2, 173.
3
Concerning Yusuf al Khuri, cf. Îunayn ibn IsÌaq, Risala, no. 53 (Ed. G.
Bergsträsser, Îunain ibn IsÌaq. Über die syrischen und arabischen Galen Übersetz
ungen. (Abhandlungen für die Kunde des Morgenlandes hrsg. von der Deutschen
Morgenländischen Gesellschaft. XVII. Band. No. 2). Leipzig 1925). According to
Îunayn his translation of On simples is of a poor quality (khabitha radi’a). See as
well G. Strohmaier, ‘Der syrische und der arabische Galen’, Aufstieg und Niedergang
der Römischen Welt, hrsg. v. W. Haase u. H. Temporini, Teil II, Bd. 37, 2. Berlin,
New York 1994, 1987 2017, at p. 2001.
180
THE INTRODUCTION OF SERGIUS OF RESH{AINA
Sergius asserts that he has translated this book for a ‘brother’ who,
out of ‘love for all knowledge that the soul requires’, requested
Hippocrates’ book On Nutriment in particular. The ‘order’ to write
— given here, though, not by a superior but by an equal — is a very
common theme in introductions. Sergius goes on to say that he did
not want to undertake the task out of fear of incurring ‘the envy of
those who are not satisfied by anything other than amassing money.’
Apologies of this sort are common in Syriac literature.4 However, we
have not found any other case where the writer expresses his fear of
avaricious envy; does this mean that Sergius was well-paid for this
translation, and feared the envy of his rivals?
4
Eva Riad, Studies in the Syriac Preface (Uppsala 1988), 190 208.
5
L.G. Westerink, ‘The Alexandrian commentators and the introductions to
their commentaries’, in Richard Sorabji (ed.), Aristotle Transformed: The Ancient
Commentators and their Influence (London 1990), 325 48. See also the following
notes.
6
See now Sara Klein Braslavy, ‘The Alexandrian Prologue Paradigm in
Gersonides’ Writings’, Jewish Quarterly Review 95 (2006), 257 89, and the litera
ture cited there.
181
THE INTRODUCTION OF SERGIUS OF RESH{AINA
182
THE INTRODUCTION OF SERGIUS OF RESH{AINA
183
THE INTRODUCTION OF SERGIUS OF RESH{AINA
184
THE INTRODUCTION OF SERGIUS OF RESH{AINA
tion upon these rubrics. In list II our translations are somewhat free,
since a literal English translation would be dense and in some cases
simply impossible.
list I
1 importance (himma)
2 benefit (manfa¨a)
3 relationship (nisba)
4 genre (Òanf )
5 composition (ta’lif )
6 soundness (ÒiÌÌa)
7 pedigree (isnad)
8 method (tadbir)
list II
1 it should have some importance
2 it should also be of benefit
3 it should be related to something
4 (6) it should be sound
5 (4) it should belong to a known literary genre
6 (5) it should be composed of different parts
7 it should trace back to one of the aspects of wisdom
8 it should have a method that is manifest (literally ‘described’)
list III
(1) The importance of the Aphorisms lies in its comprehensiveness.
(2) Its utility comes to the fore in its suitability for both skilled and
unskilled readers. (3) Palladius states at the outset that it (the Apho-
risms) is related to Hippocrates’ ‘method’ (tadbir). He explains that
the Hippocratic corpus divides up into four categories: (a) books con-
cerned with the action of people, e.g., On Prognosis, which explains
how we may foretell medical developments on the basis of symptoms
now present; (b) books concerned with the action of substances, e.g.
On Barley Water; (c) books that concern that which affects us from
without, e.g. Airs, Waters, and Places; (d) a method for writing a
book, such as the Aphorisms, which is divided into short statements.
Clearly the criterion for establishing the fourth category differs from
that employed for the first three. (4) ‘Soundness’ may mean here au-
thenticity, one of the standard topoi, since Palladius remarks here that
the soundness is connected with Hippocrates’ reputation. He men-
tions four books — the same four noted under the previous rubric as
exemplars par excellence of the Hippocratic corpus — that are utterly
indispensable and have no substitute.
185
THE INTRODUCTION OF SERGIUS OF RESH{AINA
(5) The Aphorisms is the key to medicine, and it must be read first.
Hence this rubric addresses the common theme of the place of the
book in the order of study. (6) Composition: Hippocrates wrote the
book in the form of concise aphoristic statements, and Galen later
organized it into seven sections. (7) The Aphorisms trace back to the
branches of medicine, theory and practice. (8) The method is subtle
(la†if ). Hippocrates employs three methods: (a) prolix narrative
(Ìadith); (b) concise, aphoristic remarks; and (c) a combination of
the first two.15
Let us return to Sergius. After displaying the second list of eight
topics, Sergius describes briefly the range of material covered in On
Nutriment — not just nutrition, but also bone-setting and embryol-
ogy. Perhaps this is related to the first topic, the book’s utility; but it
is only after these remarks that Sergius informs us that he will now
list the book’s uses, and they are many:
i. Some patients avoid drugs completely.
ii. Foods penetrate and work from within in a fundamental way that
drugs do not.
iii. Foods are appropriate for sound bodies as well as for the sick.
iv. The physician who knows how to make proper use of foods needs
nothing else.
v. The indigent have no choice but to rely upon foods.
vi. Mistakes in nutrition are not as damaging as those involving drugs.
vii. Travellers, or someone who finds himself in a place where drugs are
unavailable, must rely upon foodstuffs.
Sergius concludes that On Nutriment is indispensable, much as
Palladius had said about the Aphorisms in (4) above, but for different
reasons.
The second item is the reason for the book’s title. This is a very
common theme in the prologues, but Sergius’ explanation here is not
clear to us. The third topic is the division of the Hippocratic corpus.
According to Sergius, some Hippocratic writings discuss natural
things (humours, chyme, foetus, etc.), while others treat of unnatural
things (fevers, fractures, etc.). On Nutriment deals with both. This
rubric is thus close in structure to Palladius’ ‘relation’ (3), but the
contents are entirely different. The fourth rubric is preserved very
15
Mutandi mutatis this remains the scholarly evaluation of the Hippocratic cor
pus. Jacques Jouanna, Hippocrates (Baltimore and London 1999), writes (p. 57):
‘the diversity of the corpus derives from the various audiences aimed at by the
treatises…Certain treatises are addressed to a broad readership composed of both
specialists and lay people; others are aimed at a specialist readership.’
186
THE INTRODUCTION OF SERGIUS OF RESH{AINA
16
Jouanna, Hippocrates, 401.
17
J. Schacht and M. Meyerhof, The Medico Philosophical Controversy Between
Ibn Butlan of Baghdad and Ibn Ridwan of Cairo: A Contribution to the History of
Greek Learning Among the Arabs (Cairo 1937), 23 4. A small section of Kitab al
Nafi¨ was translated into English by M.C. Lyons, ‘The Kitab al Nafi¨ of Ali ibn
Ri∂wan’, Islamic Quarterly 6 (1961), 65 71, who made use of a collation of the two
manuscripts prepared by the late A.J. Arberry.
187
THE INTRODUCTION OF SERGIUS OF RESH{AINA
18
It is not certain where the quotation from Plato ends; it is quite possible that
the last sentence in our citation is due to Sergius.
19
See Fritz Steckerl, ‘Plato, Hippocrates, and the Menon Papyrus’, Classical
Philology 40 (1945), 166 80.
20
Phillip De Lacy, ‘Galen’s Platonism’, American Journal of Philology 93 (1972),
27 39, esp. pp. 36 8.
188
THE INTRODUCTION OF SERGIUS OF RESH{AINA
tween the doctrines of the two thinkers, which was translated into
Arabic, though this translation is no longer extant.21
Nonetheless, we have not been able to find the apothegm attrib-
uted to Plato in any genuine Platonic writing, or in any of the many
doxographies and other repositories of Platonic wisdom. This is not
very surprising; there are many examples of sayings attributed to
Plato that display what seems to be authentic Platonic teachings, but
whose precise formulation is not found in any of Plato’s writings.22
One can find remarks that are suggestive of the saying found in our
text. In Phaedo 67 B, for example, Socrates argues that if we main-
tain our own purity, ‘we shall probably reach the company of others
like ourselves and gain direct knowledge of all that is pure and un-
contaminated — that is, presumably, of truth.’23 The gist of this
statement is found in Abu ’l-Faraj’s introduction to his own com-
mentary to the Categories: ‘As for Plato, it is found in statement, “Let
only the pure approach the pure”.’24 Al-Biruni cites from the Phaedo
passages that have no exact equivalent in the Greek text; his editor
and translator, Edward Sachau, speculates that he may have seen
some as yet unidentified recension.25 We may if we like indulge in a
similar speculation concerning Abu ’l-Faraj’s source.
Abu ’l-Faraj’s introduction is very relevant for another reason. He
too cites Hippocrates and Plato back to back. However, there is an
important difference in the point which each author ascribes to the
two sages. The passage from the Phaedo quoted by Abu ’l-Faraj con-
cerns the individual’s preparation for his own study, and the impuri-
ties that are warned against are those of the body. The individual
is warned to restrain his impulses and desires. The sayings of
21
See Sezgin, GAS, VI,105 6.
22
On this see the classic study of Franz Rosenthal, ‘On the Knowledge of Pla
to’s Philosophy in the Islamic World’, Islamic Culture 14 (1940), 387 422.
23
Edith Hamilton and Huntington Cairns, The Collected Dialogues of Plato,
Bollingen Series LXXI (Princeton 1961), 49; the translation is by Hugh
Tredennick.
24
Ferrari, Kategorienkommentar, 10 (Arabic).
25
See E. Sachau, Alberuni’s India, vol. 2 (London 1910), 277 8. Some of the
citations from Phaedo agree well with the Greek, but others do not; Sachau suggests
that they derive ‘from a work in which text and commentary were mixed together,
and the original form of the dialogue changed into a simple relation.’ This is of
course pure speculation; no such recension of the Phaedo is known to exist. In an
earlier passage from the same book, Sachau encountered three citations from Plato,
only one of which matched the Greek dialogues. He queried the great historian of
Greek philosophy E. Zeller, who suggested in response that the problematic pas
sages are likely to derive from a commentary to the Timaeus ‘by some Christian au
thor, as e.g. John Philoponus’.
189
THE INTRODUCTION OF SERGIUS OF RESH{AINA
190
THE INTRODUCTION OF SERGIUS OF RESH{AINA
191
THE INTRODUCTION OF SERGIUS OF RESH{AINA
Acknowledgments
Langermann’s research was supported by the Israel Science Founda-
tion (Grant no. 192/05). Publication of the text from the unique
manuscript is Courtesy of the Library of the Jewish Theological
Seminary.
The Text
The text has been copied out, line by line and column by column.
Lacunae are indicated by three dots (…). On a few occasions we sug-
gest a correction to the text within square brackets [ ] immediately
after the word or words that appear to be corrupt.
[1b]….
…..
…..
…..נט’ר פי
כתבה אלתי וצ’עהא פי גמיע אלאנואע מנהא
פי אלפצול ומנהא פי אלאמזגה ומנהא פי אל
דם ואלנטפה )?(ותרכיב אלאעצ’א וכון אלטפל
ופי אלמא ואלהוא ואלארץ ואלבלדאן ופי גבר
אלעט’אם ואלאצ’מדה .ואלקרוח ופי גמיע
192
THE INTRODUCTION OF SERGIUS OF RESH{AINA
…][2a
…
…..
…..
אלו’ עלי כל גז יתפצל הדא אלכתאב .אלז’
יבין פיה מן אי אנואע אלעלם הו
הדא אלכתאב אלח’ יכבר באי אנוא]ע[ אלתעלים
יסתעמל אברקאט הדא אלכתאב אי אם אבוקראט
יריד אן יבין פי הדא אלכתאב עלי טלב עלם
אלגדאה וגרץ’ אנואעה ותגיירהא וקואהא ופעלהא
ומנאפעהא ומצרהא ולא יגד בדא מן אלכרוג
ען חדה .ואלקול פי אנואע אלאמראץ’ אלתי
תך מן כתרה אלגדא או מן רדאתה או מן
כיפיתה או מן כמיתה באבינה .וקול איצ'י פי
גבר אלעטאם אלמכסורה ועדה אלאיאם אלתי
תברא פיהא ופי כון אלגנין ואנחבאלה .ופי
תחרכה פי אלרחם ותמאמה עלי אלכמאל
ואבאן אלכרוגה מע עדה איאם כל ואחד מן
הדה תם ירגע פיכמל עלם הדא אלמקאל והדה
הי אלמנאפע אלתי ינאלהא מן קרא הדא אלכתאב
193
THE INTRODUCTION OF SERGIUS OF RESH{AINA
194
THE INTRODUCTION OF SERGIUS OF RESH{AINA
…][3b
…
…
…
פלדלך הדא אלכתאב הו אלראבע מן כתב
אבקראט ובעדה יקרא .ואעלם אנה יסמי
באסם אבקראט סלתה נפר ]תלתה עשר[ וקד וצ’ע
כתבא בעצ’הא יכאד ישבה בעצ’א" פלדלך מא
שך מן פסר כתבהם וכאדוא לא ימיזון בעצ’הא
מן בעץ’ .פאמא נחן פנקול והו אלחק אן הדא
אלכתאב הו לאבקראט אלדי וצ’ע כתאב תקדים
195
THE INTRODUCTION OF SERGIUS OF RESH{AINA
…][4a
…
…
ממן וצ’ע אלכתב … יבין…
ג’ אגזא מן אלעלם לאן מן אלעלם מא ידל עלי
אלמנטק ופהם אלעקל .ומנה מא ידל עלי עמל
אלאיד … אלג’ עלי אסתעמאל אלנועין ועלי
הדא אלחאל תגד אכתר כתב אבוקראט פאן מנהא
מא הו פי טביעה אלאנסאן ומנהא מא הו עלי
גיר הדא אלכתאב אלכסר ואלגבר ואמא אלדי פי
אלכשך פידל עלי נועין לאן הדא אלכתאב
אנמא הו מערפה אלחמי ואלקיאס עליהא מן
אפעאלהא תם אלחכם עליהא במא יוגבה אלקיאס
תם בעד דלך דכר אלכשך ועמלה וסאיר אלאשיא
אלתי תעמל באלאידי פפי הדא אלכתאב נועאן
מן אלעלם ואלעמל לאנה אדא קאל פי אלגדא
ואנצראפה ליגדוא אלאעצ’א וינמיהא פי וקת אל
נמו ויכלף עליהא מא תחלל מנהא פי וקת דלך
אנמא יסתעמל פיה אלנט’ר ואלקיאס ואדא קיל
חלו ליס בחלו חלו באלקוה חלו באלפעל ומא
ישבה הדא אנמא יסתעמל איצ’י טריק אלקיאס
ואמא אלצ’מאדאת ואלאדהאן וגמיע מא יעצב
196
THE INTRODUCTION OF SERGIUS OF RESH{AINA
197
THE INTRODUCTION OF SERGIUS OF RESH{AINA
Translation
ing to these two ways, concerning natural things, and those that are
unnatural [literally: ‘lie outside of nature’]. Healthy people…
[3b]…
… Therefore, this book is the fourth of Hippocrates’ books, after
which he [the medical student?] reads [advanced texts?]. Know that
he [it?] was called by the name Hippocrates…He composed books,
some of which are almost like others. This is the reason why the
commentators had some doubts and could hardly distinguish the
one from the other.
As for us, we say — and it is the truth — that this book is by the
Hippocrates who wrote On Prognosis, the Aphorisms, On Bone-Set-
ting, On Joints, Epidemics, and Acute Diseases. We know this, because
of its conciseness and wariness of prolixity, along with the depth of
understanding and the rarity of anything superfluous, yet concerning
this [issue], there is argument that is more correct and more firm.
This is the testimony of Galen and his commentary to it, among all
of them.
This book is divided among three issues. The first is the nutrition
of the general populace, their food and its action, the organs that are
nourished by it, its benefit and harm. The second is the classification
of the types of food. Included in this is the clarification, which of
them are moist and which dry, knowledge of smells and tastes; the
classification of the organs that are harmed by them or benefit from
them; and the diseases that are produced by them [foodstuffs]. The
third is the formation of the embryo, how many days it takes for it
to be completed, how many days it moves within the womb, and in
how many days it will be birthed, as well as the explication of the
difference between male and female. He then went back to the gen-
eral topics, describing the types of foods, and explaining their ben-
efits and harms. He could…
[4a]
…
…from among those who wrote books …he explained
…three parts of science. For part of science is that which points to
logic and intellectual understanding. Part of it points to work with
the hands [surgery]. The third concerns making use of both kinds.
You will find most of Hippocrates’ books to be according to this de-
scription. One (?) of them is On the Nature of Man, and some are
different from that book, [such as] [On] Fractures and Bone-setting.
However, that which is comprised by On barley water indicates two
kinds. For that book comprises knowledge of fever and reasoning
202
THE INTRODUCTION OF SERGIUS OF RESH{AINA
about it on the basis of its actions, then the judgement about them
as to what reasoning mandates, and then, after that, mention of the
barley water, its preparation, and the other things that are do be
done with the hands.
Now this book has two kinds of science and practice. When he
talks about food and its utilization for nourishing the organs, causing
them to grow during the period of growth, and replenishing that
which wastes away of them, during the period that that [is applica-
ble, i.e. maturity], [all of ] this is to be done on the basis of investiga-
tion and reasoning. When ‘sweet’, ‘not sweet’, ‘sweet in potentia’,
‘sweet in actu’, and the like are mentioned, this too is to be done by
way of reasoning. However, the pastes, oils, and all the like that [are
used to] bandage, bind and cure, they are the work of the hands,
[though] falling under [the rule?] of reasoning, for it [reasoning] is
secondary to it [practice in these matters].
It is known that this book encompasses these parts. The physi-
cians at the time of Galen made a greater division of science [or:
knowledge]. However, those who were in the time of Hippocrates
made use of knowledge only according to three divisions — either
because they were sparing with words but deep in their concepts
[4b], so that one would rely upon one’s understanding [as in the
English expression, ‘a word to the wise…’], or else they were prolix
and lengthy in their words, intending the opposite of what was
thought in the first [case]; or they combined both of these.
Hippocrates […] composed books according to this method, such
as On [Fractures] and Joints and Epidemics; and he composed a book
on diseases that passed by him and he inspected. He composed an-
other book [perhaps: other books], similar to this with regard to
their great prolixity and length. Since their discourse is lengthy, not
much understanding is needed [since everything is set forth at
length]. However, his On Nature, On Temperaments, and Aphorisms
are much deeper and much more abstruse, because he made them
concise and succinct. But as for On Barley Water, and what was said
about ulcerations in the head, and ulcerations in the entire body, as
books two and six of Epidemics, they comprise a combination of
both methods; however, they contain a discourse that does not re-
quire great understanding.
You ought to know that this book [i.e. On Nutriment] is one of
the books that he wrote with conciseness. Scholars did not compose
their writings according to this tendency out of their stinginess and
meanness with regard to knowledge. Instead, they did all of this [de-
liberately], not relying upon a lucid discourse. For when a lazy per-
203
THE INTRODUCTION OF SERGIUS OF RESH{AINA
204
ALESSANDRA AVANZINI
UNIVERSITY OF PISA
Abstract
1
Despite south Arabian culture being rightfully within the history of the an
cient near east, the history of south Arabia at the beginning of the first millennium
has often been neglected by the many general treatises on this period. This lacuna is
why the need voiced by Liverani (Liverani 2005 b: 171) for some mention of the
history of ancient south Arabian kingdoms in the history of the near east is so im
portant.
2
The standpoints of the two scholars are assuredly not identical. That of N.
Nebes is much more subtle and problematic, rich in interesting linguistic observa
205
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
206
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
207
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
4
The history of south Semitic writing is still to be written. Some interesting
parallels between the writing on the oldest ASA sticks and Dedanite writing were
suggested by M.C.A. Macdonald during the 10th Rencontre sabéenne in Saint
Petersburg. Assuredly, ASA writing has no clear genealogical relationship with
north Semitic alphabetical writing but it does have a ‘link by inspiration or anal
ogy’, cf. Healey 1990: 61.
5
M.C.A. Macdonald (Macdonald 2000) has provided an accurate definition of
this and what the exact subdivisions of the term are from the epigraphic stand
point.
6
One of the main steps forward in south Arabian studies in recent years has
been to consider ASA as a number of separate languages forming a linguistic group
instead of a single language broken into a series of dialects.
7
For an in depth discussion on Nebes’s hypothesis see Mazzini 2005: 223 5.
208
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
8
ASA was also included in Central Semitic (see Voigt 1987).
9
Appleyard’s obvious embarrassment (Appleyard 1996, Appleyard 2002), in ac
cepting the incongruence between this isogloss that distinguishes ASA from south
ern Semitic and the other many isoglosses that they share, can also be seen in his
interesting articles on this issue.
10
For Marrassini, although the verb form was primary, it would be a ‘conserva
tion’ and not an ‘innovation’ and thus useless for classifying south Semitic lan
guages.
11
Personally, I find D. Cohen’s theory perfectly plausible according to which
yVqattVl is an independent development in MSA and Ethiopic. Cohen reached this
conclusion after studying the dynamics by which recent languages (for example
neo Aramaic) developed the verb system. See Knudsen 1998: 5 6: ‘The crucial
problem in the discussion has been the comparative assessment of the relationship
between the Akkadian present and the Ethiopic and Modern South Arabian
imperfects. A majority of scholars maintain that these forms are genetically related
and that they reflect a common Proto Semitic form, whereas a relatively large
minority of scholars dispute that view. […] In recent years the majority view has
prevailed in the literature and maybe Marrassini is right that ‘one has got rid too
hastly of the Rundgren hypothesis on the réemploi de l’intensif (Rundgren 1959)’
(Marrassini 1991: 1020)’.
209
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
12
Some of my earlier remarks (Avanzini 1991) are certainly to be corrected on
the basis of Nebes’s studies.
13
Using the same monumental script not only implies sharing the same writing
technique but also the same textual models, text syntax throughout the whole of
ASA culture.
14
The best know isogloss common to every ASA language is the three unvoiced
sibilants.
15
The pronoun system with the same ∂ base for the relative and demonstrative
pronouns.
The determination of the noun made by the suffix n is the isogloss that charac
terizes the ASA family within the languages of Arabia as was pointed out by
Beeston (Beeston 1982: 178 86) and repeated by M.C.A. Macdonald (Macdonald
2000) in his classification of the languages of Arabia. In my opinion, the laryngal h
in the HAD suffix hn is a mater lectionis despite what Tropper believes (Tropper
2001: 11 13). The m suffix in the substantive which is also defined in Stein’s
grammar as a non determination suffix, has a complex history. It was originally
used for adverb forms and for indicating subordination as well as for the syntactical
dependence of elements already marked as definite. See the interesting remarks in
Gzella 2006 on determination in Phoenician.
16
Syntax: the order of the words (verb subject object) is the same in all ASA
languages; the syntax of the prepositions is the same, but different compared to
Arabic, Ethiopic, MSA while, certainly by no coincidence, similar to Ugaritic. Cf.
Pennacchietti 1974 based on the V. Brøndal’s theory of prepositions.
17
The study of the ASA lexicon is also extremely important for the hypothesis
suggested here of a lengthy history of ancient south Arabian languages prior to its
earliest written attestations. For example, the juridical lexicon in ASA languages has
a great ‘technical’ wealth and specificity which only evolved within ASA and is not
found in other Semitic languages. In historical terms, too, a developed detailed ju
ridical lexicon has interesting applications and in my view presumes a centralised
bureaucratised state power in at least some periods of the long history of South
Arabia.
210
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
18
Even the matres lectionis were used differently in the various ASA epigraphic
corpora. SAB, especially in its most archaic phase, used a very limited number of
matres lectionis (Stein 2003: 41 7), while there were many more in non Sabaic ASA
languages and in late SAB (Avanzini 2006c: 257 8). This is very interesting for the
cultural history of ancient south Arabia. On the one hand, epigraphic texts are very
homogeneous both in their graphic layout and in their palaeographic development,
creating an epigraphic ‘style’ common to texts in every ASA language and on the
other there are graphic phenomena such as the different way of writing vowels
which points to different writing schools in the single regions.
One interesting aspect is the habit of writing vowels in non SAB languages
using two graphemes. For example, some plural suffixes in the substantives: QAT:
hw; HAD, MIN: hy.
19
In QAT and HAD two attached pronouns s1ww s1yw attached to a dual or
to an external plural are attested. It is not easy to explain the long script, but it as
suredly depends on the accent a probably pre suffix stress and consequent pre
suffix vowel length (Gragg 1997) as the following HAD example: Shabwa
chantier V, 1975 6 7 (Pirenne 1990: 76): s1qnyt ’Ì∂ ¨lh s1ww h wfy grb s1 ‘the dedi
cation that He took over him for the safety of his person’ shows with the long suffix
after the preposition ¨lh (the laryngal h surely indicates a vowel).
211
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
QAT verb system between the use of yf ¨l and b-yf ¨l: ‘b- is character-
istic of the indicative in QAT, while a jussive (and perhaps also a sub-
junctive) use is characterized by the absence of b-. Such a formula-
tion would bring the QAT imperfect usage very close to that of the
Syro-Palestinian dialect of Arabic’ (Beeston 1962: 24–5).20
A study of the QAT corpus has shown that no modal opposition21
can be surmised for the yf ¨l vs. b-yf ¨l opposition, but the b-yf ¨l
b-yf ¨lwn (pl.) forms are used to specify the present-future (Avanzini
2005b), in relative clauses in particular.
The interesting point in general comparison is that QAT has a
morphological opposition within the verbal prefix forms between
two masculine plural terminations –w / -wn, the first used for the
preterite: CSAI 1, 203 = RES 3858 4–5: w-ygb’ w-h[†]b Y∂mrmlk ’byt
w-’r∂ty w-’qny Qtbn ‘Y∂mrmlk gathered and returned the houses, the
lands, and the possessions of Qataban’. CSAI 1, 115 = ¨Arbach-
Sayun 1 l. 3: w-yhrgw w-s1lqÌ Î∂rmwt ‘and then they scattered death
and destruction on the Îa∂ramawt’, the second for the imperfective:
CSAI I 205 = R 4337: w-l-yÌr† Ìms1y wrqm l-mlk Qtbn w-¨hr S2mr
ys1mÂ’wn ‘must pay 50 pieces of gold to the king of Qataban and the
magistrate of S2mr whose task this is (whose task it was and will be
to receive payment of the fine)’.
QAT later morphologized a b- prefix for the form with the plural
in -wn to indicate the present-future.
In QAT we would have proof of two prefixed forms for the past: a
short form for the preterite: ‘he did’, and the other, long form for a
narrative imperfective action ‘he was doing’.
The QAT verb can therefore be reconstructed in accordance with
the following pattern:
f ¨l w-yf ¨l yf ¨l b-yf ¨l
w-yf ¨lw (pl.) yf ¨lwn (pl.) b-yf ¨lwn (pl.)
(suffix)- preterite (prefix)-preterite narrative-imperfective present-future
Not only QAT but all ASA languages have a prefix form for the
past. N. Nebes and P. Stein are assuredly correct in finding some of
Tropper’s conclusions hurried (Tropper 1994): there is no proof of a
20
Description was summarised in Beeston 1984: 64: ‘The indicative imperfect
normally shows a prefix b , e.g. R 3688 2 kbrm b ykbr. Without this prefix the im
perfect is usually a jussive, or conditional, or in some other way non indicative’.
21
This same modal opposition is probably recent and is not found in archaic
languages; cf. Petracek 1981.
213
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
214
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
The verb has three forms one with suffixes and two with prefixes.
The verb system which initially had a marked aspectual value soon
ran into difficulties. The short prefix form (preterite-aorist) became
integrated with the suffix form; the long form (present-narrative)
went through many phases of morphological specialization in which
a stronger temporal component than that of the early aspectual con-
notation was inserted thus becoming more specific.
This verb scheme is found upstream of all attested ASA languages;
indeed we could say that it is the proto-ASA verb system, a proto-
western not only a proto-north-western verb system25.
To drastically curtail the importance of the presumed loss of the
imperfect yVqattVl in ASA modifies the general frame of the history
of southern Semitic.
Having proved that south Arabian languages are testimony to a
history in the longue durée within south Arabia and do not come
from a recent colonization (late second-early first millennium) on
the part of populations from the north opens a broad historic and
linguistic debate. For instance, a repeated datum, which in my view
has been merely reiterated without being subjected to further discus-
24
The expression is frequent in late texts (see the examples mentioned by
Fransouzoff 2005: 251 2), but always has the divinity as the subject ‘that the god
listens to him, does his wish’. Frantsouzoff ’s explanation: ‘Il paraît que les auteurs
de notre inscription aient employé cette formule avec ys1m¨ sans en comprendre
profondement le sens réel’ does not seem very probable.
25
‘Typologically the reconstruction of the prefixed verb system of Proto ancient
South Arabian is very close to that of Northwest Semitic of the II Millennium’
(Mazzini 2007).
215
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
sion or critique, is that MSA does not derive from ASA.26 Separating
ancient south Arabian from modern south Arabian, again on the ba-
sis of the iprus-iparras/yenger-yenager isogloss and thus on the recent
arrival of populations who spoke ASA risks leading us astray histori-
cally and linguistically.
After lending support to the idea of non-affiliation between MSA
and ASA Simeone-Senelle had a meaningful lapsus and repeated that
the MSA languages were spoken in south Arabia ‘bien avant la
pénétration de l’Arabe, venu du nord’ (Simeone-Senelle 2002: 393).
Nobody had ever thought that MSA derived from classical Arabic
and it is therefore likely that Simeone-Senelle was getting ASA and
Arabic mixed up.
As we well know from studying the Romance languages which ob-
viously come from Latin, the former were born from a colloquial
language and not from Cicero’s Latin. If we compared a Latin in-
scription from the Republican period with a recording of spoken
Rumanian the difference would be no less great than if we were to
compare Sabaic and Mehri.
Epigraphic South Arabic is the most antique documentation of
southern Semitic. Why, then should we do without it, in principle at
least?
If my theory of an endogenous origin of ASA languages is correct,
it would have far-reaching consequences for the hitherto accepted
classification of Semitic languages: Hetzron’s map of the distribution
of Semitic languages would have to be redrawn. The oldest phase of
southern Semitic would be less southern because of the many
isoglosses between ASA and second millennium north-western
Semitic, and central Semitic would be less central and ‘innovative’,
comprising practically the whole western part of the Semitic world
from Syria to Yemen.
26
This hypothesis is asserted by some French scholars and I find the position
held by V. Porkomovkij (Porkomovkij 1997) more convincing. See also Huehner
gard 2005: 160 1: ‘Nebes demonstrated that none of the languages for which there
is sufficient evidence Sabaean, Minaean, Qatabanian exhibits the form
yVqattVl; the imperfective form of the verb is, instead, yaqtulu. […] It also means
that none of these languages can be the ancestor of either the Modern South
Arabian languages or the Ethiopian Semitic languages’.
Huehnergard’s article is very interesting for other reasons. He hypothesizes a
Central Semitic divided into two main groups, North Central Semitic and South
Central Semitic.
216
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
REFERENCES
Appleyard D.L. 1996. ‘Ethiopian Semitic and South Arabian towards a Re exami
nation of a Relationship’, in Sh. Izra’el and Sh. Raz (eds), Israel Oriental Studies
16. Studies in Modern Semitic Languages (Leiden). 203 28
2001. ‘New Finds in the 20th century: the South Semitic Languages’, in Sh.
Izra’el (ed.), Israel Oriental Studies 20. Semitic Linguistics: the State of the Art at
the Turn of the 21st Century (Tel Aviv). 401 30
Avanzini A. 1991. ‘Linguistic Data and Historical Reconstruction: between Semitic
and Epigraphic South Arabian’, in A.S. Kaye (ed.), Semitic Studies in Honor of
Wolf Leslau (Wiesbaden). 107 18
1995. Inventario delle iscrizioni sudarabiche 4. As Sawda’. (Paris Rome)
2004. Corpus of South Arabian Inscriptions I III. Qatabanic, Marginal
Qatabanic, Awsanite inscriptions. (Pisa)
2005a. ‘Some Thoughts on Ibex on Plinths in Early South Arabian Art’, Ara
bian Archaeology and Epigraphy 16, 144 53
2005b. ‘Some Brief Observations on the Qatabanic Verb System and on the
Relationship between the Ancient South Arabian and the Modern South Ara
bian’, in A.V. Sedov and I.M. Smiljanskaja (eds), Arabia Vitalis. Studies in Hon
our of V.V. Naumkin (Moscow). 318 23
2005c. ‘Some Remarks on the Classification of Ancient South Arabian Lan
guages’, in P. Fronzaroli (ed.), 10th Meeting of Hamito Semitic Linguistics,
Firenze 18 21 Aprile 2001 (Florence). 117 25
2006a. ‘Ancient South Arabian Anthroponomastics: Historical Remarks’. Pro
ceedings of the Seminar for Arabian Studies 36, 79 85
2006b. ‘To accompany a Recently Published Sabaic Text: Historical and
Grammatical Remarks’, in P.G. Borbone, A. Mengozzi and M. Tosco (eds),
Loquentes Linguis. Festschrift F. Pennacchietti (Wiesbaden). 35 44
2006c. ‘A Fresh Look at Sabaic'. Review of P. Stein. Untersuchungen zur
Phonologie und Morphologie des Sabäischen. (Rahden, Westf ) 2003. Journal of
American Oriental Society. 126: 2, 253 60
Beeston A.F.L. 1962. A Descriptive Grammar of Epigraphic South Arabian. (London)
1978. ‘A Minean Market Code’, Bulletin of the School of Oriental and African
Studies 41, 142 5
1982. ‘Languages in Pre Islamic Arabia’, Arabica 28, 178 86
1984. Sabaic Grammar. Journal of Semitic Studies Monograph 6. (Manches
ter)
Christian V. 1919 20. ‘Akkader und Südaraber als ältere Semitenschichte’,
Anthropos 15 16, 729 39
Cohen D. 1984. La phrase nominal et l’évolution du système verbal sémitique. (Paris)
Edens C. 2002. ‘Looking for Connections: Southwest Arabia in Late Prehistory’,
Man and Environment 27, 45 55
Frantsouzoff S. 1995. ‘The Inscriptions of the Temples of Δhat Îimyam at
Raybun’, Proceedings of the Seminar for Arabian Studies 25, 15 28
2001. ‘Le “tailleur de pierre” (grby n/hn) dans les inscriptions sudarabiques’,
Raydan 7, 125 43
2005. ‘Une nouvelle inscription Îa∂ramoutique datée par éponyme’, in
A.M. Sholan, S. Antonini and M. Arbach (eds), Sabaean Studies: Archaeological,
Epigraphical and Historical Studies in Honour of Y.M. ¨Abdallah, A. De Maigret
and Ch.J. Robin (Naples, ∑an¨a’). 247 58
217
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
Garbini G. 2004. ‘The Origins of South Arabians’, in A.V. Sedov (ed.), Scripta
Yemenica. Studies in Honor of M.B. Piotrovkij (Moscow). 203 9
2006. Introduzione all’epigrafia semitica. (Torino)
Gragg G. 1997. ‘Old South Arabian Phonology’, in A.S. Kaye (ed.), Phonology of
Asia and Africa 1 (Winona Lake, IN). 163 5
Gruntfest Y. 1999. ‘The Consecutive Imperfect in Semitic Epigraphy’, in Y.
Avishur and R. Deutsch (eds), Historical, Epigraphical and Biblical Studies in
Honor of Prof Michael Heltzer (Tel Aviv). 171 80
Gzella H. 2006. ‘Die Entstehung des Artikels im Semitischen: eine “Phönizische”
Perspektive’, Journal of Semitic Studies 51, 1 18
Healey J.F. 1990. The Early Alphabet. (London)
Huehnergard J. 2005. ‘Features of Central Semitic’, in A. Gianto (ed.), Biblical and
Oriental Essays in Memory of William L. Moran (Rome). 155 203
Joffe A. H. 2002. ‘The Rise of Secondary States in the Iron Age Levant’, Journal of
the Economic and Social History of the Orient 45, 425 67.
Keall E.J. 2004. ‘Possible Connections in Antiquity between the Red Sea Coast of
Yemen and the Horn of Africa’, in P. Lunde and A. Porter (eds), Trade and
Travel in the Red Sea Region. (Oxford). 43 55
Knudsen E.E. 1998. ‘Central Semitic *yaqtulum Reconsidered. A Rejoinder to J.
Tropper’, Journal of Semitic Studies 48, 1 9
Liverani M. 2005a. ‘Imperialism’, in S. Pollock and R. Bernbeck (eds), Archaeologies
of the Middle East. Critical Perspectives (Oxford). 223 43
2005b. Review of M. Van de Mieroop, A History of Ancient Near East (Ox
ford 2003), Orientalia 74, 171 3
Macdonald M.C.A. 2000. ‘Reflections on the Linguistic Map of Pre Islamic Ara
bia’, Arabian Archaeology and Epigraphy 11, 28 79
de Maigret A. 1981. ‘Two Prehistoric Cultures and a New Sabaean Site in the East
ern Highlands of North Yemen’, Raydan 4, 194 204
1990. The Bronze Age Culture of Hawlan al Tiyal and al Hada. (Rome)
1999. ‘The Arab Nomadic People and the Cultural Interface between the
“Fertile Crescent” and “Arabia Felix”’, Arabian Archaeology and Epigraphy 10,
220 4
Marrassini P. 1991. ‘Some Observations on South Semitic’, in A.S. Kaye (ed.),
Semitic Studies in Honor of Wolf Leslau (Wiesbaden). 1016 23
2003. ‘“Genetic Subgrouping” in Semitic’, in P. Marrassini (ed.) Semitic and
Assyriological Studies Presented to Pelio Fronzaroli (Wiesbaden). 369 80
Master D.M. 2001. ‘State Formation Theory and the Kingdom of Israel’, Journal of
Near Eastern Studies 60, 117 31
Mazzini G., 2005. ‘Ancient South Arabian Documentation and the Reconstruction
of Semitic’, in P. Fronzaroli (ed.), 10th Meeting of Hamito Semitic Linguistics,
Firenze 18 21 Aprile 2001 (Florence). 215 38
2006. ‘∂ base Pronominal System in Qatabanic’, in P.G. Borbone, A.
Mengozzi and M. Tosco (eds), Loquentes Linguis. Festschrift F. Pennacchietti
(Wiesbaden). 475 88
2007. ‘The System of Prefixed Verb in Proto ancient South Arabian’, in M.
Moriggi and M. Cassarino (eds), Proceedings of the 12th Meeting of Hamito Se
mitic Linguistic (Rubettino). 255 61
Nebes N. 1988. ‘The Infinitive in Sabaean and Qatabanian Inscriptions’, Proceed
ings of the Seminar for Arabian Studies 18, 63 78
218
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
219
CLASSIFICATION OF THE ANCIENT SOUTH ARABIAN LANGUAGES
Wilkinson T.J. 2005. ‘The Other Side of Sheba: Early Towns in the Highlands of
Yemen’ Bibliotheca Orientalis 62, 5 14
Wilkinson T.J., C. Edens and M. Gibson. 1997. ‘The Archaeology of the Yemen
High Plains: a Preliminary Chronology’, Arabian Archaeology and Epigraphy 8,
99 142
220
AARON D. RUBIN
PENNSYLVANIA STATE UNIVERSITY
Abstract
1
I am very grateful to Fabrizio Pennacchietti and Gary Rendsburg, who both
kindly provided me with very valuable comments to an earlier draft of this article.
2
All forms and examples come from the texts of T.M. Johnstone, as published
in Stroomer (1999), unless otherwise noted.
221
THE FUNCTIONS OF THE PREPOSITION K- IN MEHRI
Note that the 2ms form has the rare variant sik, formed presum-
ably on analogy with tik, the 2ms form of the pronominal direct ob-
ject marker t-.
As already noted, the basic meaning of this preposition, with or
without suffixes, is ‘with’, as illustrated in the following sample sen-
tences:
ho m¢g¢tayr k¢-Ìam¢´y ‘I will speak with my mother.’ (9:2)3
w¢kub k-'agawz ‘He went in with the old woman.’ (75:7)
t¢syer say ‘You will go with me.’ (55:2)
agayg s¢h sx¢w¢lul ‘The man (who was) with him stayed.’ (68:5)
s¢biw agayg sih¢m ‘They took the man prisoner with them.’ (69:2)
Etymologically, Mehri k- is connected with the widely attested
Semitic preposition k- ‘like, as’. On the correspondence of the Mehri
meaning to that found in the rest of Semitic, already Bittner (1914:
8) saw the parallel of Ethiopic m¢sla ‘with’ and Arabic mi†la ‘like’.4
A similar (but by no means identical) development of the preposi-
tion k- is also found in South Ethiopic. In Amharic, for example, the
preposition kä- has a wide variety of meanings, including ‘with’.5 At
this point, the idiomatic uses of this preposition shall be examined in
turn.
3
Numbers in parentheses following all examples refer to text and line numbers
in Stroomer (1999).
4
On Ethiopic m¢sla, see also Pennacchietti (1974: 189 n. 69). Pennacchietti
also briefly notes the importance of this parallel Ethiopic and South Arabian devel
opment (p. 198).
5
See further in Pennacchietti (1974: 195). On the uses of kä in Amharic, see
Leslau (1995: 605 7). Note here that in the meaning of ‘with’, the Amharic
kä usually combines with another pre or postposition.
6
This use was discussed very briefly by both Bittner (1913: 54 Anm. 1) and
Thomas (1937: 251).
222
THE FUNCTIONS OF THE PREPOSITION K- IN MEHRI
223
THE FUNCTIONS OF THE PREPOSITION K- IN MEHRI
11
This is the same root found in the word m¢wse ‘rain’.
224
THE FUNCTIONS OF THE PREPOSITION K- IN MEHRI
gathered from the context So, just as s¢h Èawt can mean ‘he has food’
or ‘he had food’, likewise say ÌarÈ can mean ‘I am hot’ or ‘I was
hot’.
3. Expressions of Time
4. Conclusions
This short article has outlined the various functions of the common
Mehri preposition k-. The particular possessive construction used for
12
This suggestion is found in Bittner (1914: 9) and in Johnstone (1987: 208),
under the heading kl¨n(y).
13
For examples of k occurring with several other temporal nouns, see Bittner
(1914: 8 9).
14
This sentence, and the one preceding, illustrate two different meanings of téˆ,
‘then’ and ‘until’. In fact, téˆ is also used as a temporal relative ‘when’. The phrase k¢
sob¢Ì is nearly always preceded by téˆ, in one of these three meanings.
15
Joüon (1996: §133g, §166m).
225
THE FUNCTIONS OF THE PREPOSITION K- IN MEHRI
REFERENCES
Bittner, Maximilian. 1913. ‘Studien zur Laut und Formenlehre der Mehri Sprache
in Südarabien. III. Zum Pronomen und zum Numerale’, Sitzungsberichte der
Kaiserlichen Akademie der Wissenschaften in Wien, Philosophisch Historische Klasse
172/5
1914. ‘Studien zur Laut und Formenlehre der Mehri Sprache in Südarabien.
IV. Zu den Partikeln’, Sitzungsberichte der Kaiserlichen Akademie der Wissen
schaften in Wien, Philosophisch Historische Klasse 174/4
Heine, Bernd, and Tanya Kuteva. 2002. World Lexicon of Grammaticalization.
(Cambridge)
Johnstone, T.M. 1987. Mehri Lexicon and English Mehri Word List. (London)
Joüon, Paul. 1996. A Grammar of Biblical Hebrew. 2 vols. Revised and translated by
T. Muraoka. (Rome)
Leslau, Wolf. 1995. Reference Grammar of Amharic. (Wiesbaden)
Maiden, Michael, and Cecilia Robustelli. 2000. A Reference Grammar of Modern
Italian. (Chicago)
Müller, David Heinrich. 1902. Die Mehri und Soqo†ri Sprache. I. Texte. Süd
arabische Expedition IV. (Vienna)
Rubin, Aaron D. 2005. Studies in Semitic Grammaticalization. (Winona Lake, IN)
Pennacchietti, Fabrizio. 1974. ‘Appunti per una storia comparata dei sistemi
preposizionali semitici’, Annali dell’Istituto Orientale di Napoli 34, 161 208
2005. ‘Sull’etimologia di arabo musa “rasoio”’, in A. Mengozzi (ed.), Studi
Afroasiatici: XI Incontro Italiano di Linguistica Camitosemitica (Milan), 231 37
Stroomer, Harry (ed.) 1999. Mehri Texts from Oman. Based on the Field Materials of
T.M. Johnstone. (Wiesbaden)
Thomas, Bertram. 1937. ‘Four Strange Tongues from Central South Arabia’, Pro
ceedings of the British Academy 23, 231 331
226
HILLA PELED-SHAPIRA
BAR-ILAN UNIVERSITY
Abstract
This article discusses metaphors derived from the animal world, and
used to describe relations between intellectuals and the authorities in
mid twentieth century Iraq, against the backdrop of the conservative
and conformist use of metaphor in Classical Arabic literature. The
topic in question is examined through the narrative works of the
twentieth century exiled Iraqi Communist writer Gha’ib Tu¨ma
Farman (1927 90), who used animal metaphors as an artistic device
through which he depicted himself as a Leftist intellectual persecuted
by the government. The examples from Farman’s own works are con
sidered in light of the use which other twentieth century Arab writers
make of animal metaphors, and the artistic needs which the latter
serve.
227
THE CASE OF GHA}IB TU{MA FARMAN
228
THE CASE OF GHA}IB TU{MA FARMAN
4
Menahem Milson, Medieval and Modern Intellectual Traditions in the Arab
World (Tel Aviv 1973), 36ff. (Hebrew).
5
Milson, Medieval and Modern Intellectual Traditions, 41.
229
THE CASE OF GHA}IB TU{MA FARMAN
6
W.P. Heinrichs, ‘majaz’, in Julie Scott Meisami and Paul Starkey (eds), Ency
clopedia of Arabic Literature, London, 1998, 495 6; B. Reinert, ‘Madjaz’, E.I2
(Leiden 1986), vol. V, pp. 1025 6. For more on the history of the term majaz see
the chapter ‘“U:« WHK' in MuÌammad ¨Abd al Mu††alib, al Balagha wa’l uslubiyya
(Cairo 1984), 57 70, esp. pp. 69 70, on the necessity of using majaz in order to
express spiritual needs.
7
On the widespread use of badi¨ in the Abbasid period and on the prominent
writers of that period who made frequent use of metaphor (…—UF
«), simile (tOA),
juxtaposition (‚U), paronomasia (”UM) and a variety of other rhetorical devices,
see: AÌmad Ibrahim Musa, al ∑abgh al badi¨i fi al lugha al ¨arabiyya (Cairo 1969),
50 116. On the use of these rhetorical devices by Òufi poets after the Abbasid pe
riod see MuÌammad ¨Abd al Mun¨im Khafaji, al Adab fi al turath al Òufi (Cairo
1980). ∑ufi poets would occasionally express themselves in indirect terms (‘ÊËdR
`
dB
« ÊËœ `OLK
« v« ÊËbLF
Ë ¨…—UF« vK …—Uô«'), in order to keep their thoughts and
philosophy secret (ibid., p. 250). This gave their poetry a symbolic dimension lack
ing in other poets, who preferred more concrete and direct language.
8
W.P. Heinrichs, ‘badi¨’, in Meisami and Starkey (eds), Encyclopedia of Arabic
Literature (London 1998), 122 3. See also Aghna†ius Kratshqufski, Kitab al badi¨
li ¨Abd Allah bin al Mu¨tazz (Damascus, no date).
230
THE CASE OF GHA}IB TU{MA FARMAN
231
THE CASE OF GHA}IB TU{MA FARMAN
are usually derived from farm animals, since most of his stories take
place in a rural setting.
The Lebanese woman writer Layla Ba¨labakki (b. 1934) describes
the heroine’s mental world and the status of women in the Arab
world in general using animal metaphors, as in the stories ‘Kuntu
muhra wa-Òirtu fa’ra’ (‘I was a Filly and became a Mouse’) and
‘al-Qi††a’ (‘The Cat’) (both in Safinat Ìanan ila al-qamar, Beirut
1964).14
When the Syrian woman writer Ghada al-Samman (b. 1942)
wants to describe the anxiety felt by the main character in the short
story ‘al-Fajr ¨inda al-nafidha’ (‘Dawn in the Window’), she makes
use of images borrowed from the avian world:15 the heroine fears
that her sensuous neighbour’s visits to her home every evening may
wreck her family. The neighbouring woman appears to the heroine
as a predator intent on preying on fearful forest birds flying around
hither and thither in frightened schools, which of course symbolize
the heroine herself.
The Egyptian woman writer and physician Nawal al-Sa¨dawi
(b. 1931) also uses many metaphors taken from the animal king-
dom. In her short story ‘Îubbi al-waÌid’ (‘My Only Love’)16 a black
dog shows more empathy towards the main character than does her
husband. In this way al-Sa¨dawi communicates to the reader her
criticism of Middle Eastern husbands’ indifference to and contempt
for their spouses, in contrast to the loyalty shown her by the dog. In
her Mudhakkirat †abiba (Memoirs of a Woman Doctor) al-Sa¨dawi
again uses metaphors taken from the animal world to describe the
heroine’s feelings in the alienated and unfriendly place where she
finds herself and her fears for her fate: the hospital at night seems to
her imagination to be like a wild hyena17 and her domineering hus-
band’s ears are described as large and flat as a rabbit’s, perhaps sym-
bolizing his cowardice and his concern for his ‘honour’ and status,
which hide under a cover of confident manhood.18 When a young
girl comes to her clinic in fear of her relatives, who threaten to mur-
der her for having brought ‘dishonour’ on her family, the doctor de-
14
On metaphorical images from the animal world in Ba¨labakki’s writings see
Olga Bramson, Revolutionary Attitudes of the Arab Women According to Layla
Baalbakki’s Narrative Works, M.A. Thesis (Bar Ilan University 1990), 94 9.
15
Ghada al Samman, ‘al Fajr ¨inda al nafidha’, in ¨Aynaka qadri (Beirut, 1985),
155.
16
Nawal al Sa¨dawi, ‘Îubbi al waÌid’, in LaÌÂat Òidq (Cairo 1962), 57 8, 65.
17
Al Sa¨dawi, Mudhakkirat †abiba (Beirut, 1988 [1st published in 1965]), 35.
18
Al Sa¨dawi, Mudhakkirat †abiba, 70.
232
THE CASE OF GHA}IB TU{MA FARMAN
cides to save her from the clutches of her relatives, whom she per-
ceives as ‘wild animals, snakes, rats and cockroaches’.19
Among the writings of Iraqi author ¨Adnan Ra’uf (1926–98)
metaphors taken from the animal kingdom are particularly in evi-
dence in his short story ‘Duda’ (‘Worm’, 1946),20 which is reminis-
cent of Franz Kafka’s The Metamorphosis. The story, a parable on
loneliness and the search for the meaning of life, revolves around the
worm, which symbolizes the pettiness and triviality of human exist-
ence and man’s feelings of insignificance in the face of the grand
forces which surround and move him and which he cannot affect in
any way. Ra’uf thus uses metaphors borrowed from the animal world
in order to raise and discuss existential philosophical questions.
Similar to the previously mentioned writers, the Iraqi Gha’ib
Tu¨ma Farman, who was born in Baghdad in 1927 and died in Mos-
cow in 1990, puts images from the animal world to a variety of uses
in unique ways. Whereas the authors mentioned above exploited ani-
mal metaphors mainly for characterizing their figures’ inner world
and outer looks, Farman expands their use to cover a great number
of topics, feelings and figures, thereby achieving four distinct aims.
First of all, the images he uses enable him to build up a complex,
detailed picture of how Leftist intellectuals in mid-twentieth-century
Iraq were persecuted and depicts their struggles against the authori-
ties. Secondly, animal metaphors make it possible for Farman to give
very special artistic expression to the educated individual’s feeling of
alienation from a society which refuses to accept him although he
grew up in it. The third function of images from the animal king-
dom in Farman’s works is to draw a picture of Baghdad as he remem-
bered it, a stifling city with narrow, winding alleys. Lastly, Farman
uses animal metaphors in order to describe unattainable peace and
happiness, as an allegory for his own life as a ceaseless wanderer who
never found a true place of rest and eventually died on foreign soil.
In the present article we shall examine these functions in Farman’s
works, in the context of the lives of Leftist Iraqi intellectuals in the
1950s, whom this writer represents so well.
Gha’ib Tu¨ma Farman began his writing career as a poet, then went
on to write short stories, which were published in two collections,
ÎaÒid al-raÌa (Grindstone Harvest, Baghdad 1954) and Mawlud akhar
(Another Newborn, Baghdad 1959), as well as eight novels, some of
which describe the Iraqi poor and their tribulations in different peri-
19
Al Sa¨dawi, Mudhakkirat †abiba, 86.
20
First published in the Iraqi journal al Waqt al ∂a’i¨ in 1946.
233
THE CASE OF GHA}IB TU{MA FARMAN
ods, while others deal with young Iraqi intellectuals, most of whom
had Socialist tendencies and were therefore persecuted, as was Farman
himself. He also wrote a historical essay on Iraq entitled al-Îukm al-
aswad fi al-¨Iraq (The Rule of Darkness in Iraq, Cairo 1957) and trans-
lated over thirty novels from Russian into Arabic. He wandered for
decades throughout the Arab world, in Europe and in Asia, partly
from choice, in order to study and work, and partly from necessity,
since his own government mistrusted him for his Socialist beliefs and
the Communist slogans he wove into his works, and did not let him
make a decent living in his own homeland.
While Iraq was still a monarchy Communists were already treated
badly. However, the situation worsened after it became a republic in
1958, and especially after the execution of ¨Abd al-Karim Qasim in
1963. The latter had been sympathetic towards the Communists,
who were subsequently treated harshly by the authorities, out of fear
that they would attempt to overthrow the regime and set up a Com-
munist-style ‘dictatorship of the proletariat’. The steps which the
authorities took against the Communists, and especially the intellec-
tuals, covered the gamut from economic sanctions and revoking their
citizenship in milder cases, to expulsion, imprisonment, torture and
even execution in the case of more intransigent opponents, some-
times involving a public trial ending in a hanging. Countless inno-
cent people fell victim to such actions. It even happened that the gov-
ernment sentenced a person to death, then reduced the sentence to
life imprisonment under public pressure, and later again sentenced
him to death on trumped-up charges, as in the case of the secretary of
the Communist Party in 1947, known as ‘fahd’ (‘panther’), according
to the memoirs of the Iraqi poet MuÌammad Mahdi al-Jawahiri.21
Gha’ib Tu¨ma Farman in his works describes these dark episodes in
the history of Iraq not only through the plot’s events, but also by
means of metaphors taken from the animal world, which the reader
can compare with the situation in Iraq during those years.
Leftist intellectuals’ sense of being persecuted – the individual versus the
authorities
As previously mentioned, the first aim of the imagery in question is
to make the reader aware of Leftist intellectuals’ sense of being perse-
cuted by the authorities to whose rule they were opposed. Their ob-
jections to the regime were expressed in ideologically driven writing
which the authorities perceived as seditious, and occasionally also in
positive acts of resistance, such as violent demonstrations, open criti-
21
MuÌammad Mahdi al Jawahiri, Dhikrayati vol. II (Damascus 1991), 53.
234
THE CASE OF GHA}IB TU{MA FARMAN
The manager: Where did you get this mistaken conception of history?
Ma¨ruf: Because I do not find in it anything of my own history, of my
life. Am I not human, sir?
The manager: I did begin to doubt whether you are human. […]
Silence animal, insect!
Ma¨ruf: And with these two epithets you threw me out of your history,
sir.
The manager: I’ll throw you out of this whole world, you rat. Get out
of here.
In this dialogue Ma¨ruf is equated with various animals, and de-
prived of his association with humankind and its history. When he
says to the manager that ‘you threw me out of your history’ he has
removed himself from the rest of mankind, to whom the manager
belongs. The manager treats the story’s hero with disdain, and refers
to him insultingly as an ‘animal’, ‘insect’ and ‘rat’. The use of the
names of animals which are considered pests is intended to mark in-
tellectuals as socially and politically harmful beings, who pose a dan-
ger to the very foundations of the regime.
The passage quoted above tells the reader that those in power os-
tracize and criminalize intellectuals; in other words, they refuse to
22
Gha’ib Tu¨ma Farman, Alam al sayyid Ma‘ruf (Beirut 1982), 113 14.
235
THE CASE OF GHA}IB TU{MA FARMAN
236
THE CASE OF GHA}IB TU{MA FARMAN
24
Robert B. Campbell (ed.), Contemporary Arab Writers: Biographies and
Autobiographiess (Beirut 1996), 394 5 (Arabic).
25
Farman, Alam al sayyid Ma¨ruf, 35.
237
THE CASE OF GHA}IB TU{MA FARMAN
ÆÆÆWOM ôË ¨UNO …Ëö ô ÆW"O)« WU
ô« fHM tO« Ê«dEM rU bO« UMO XU
26
Æ”«d
«
Mr. Hashim’s eyes looked at him with that same evil smile, completely
lacking in sweetness or pity … rapaciousness.
Farman’s choice of animal life, especially of predators in the forest,
for his imagery is not random. A predatory smile represents a preda-
tory person, in this case the boss lying in wait for the story’s hero.
This metaphor is not meant by the author to describe that person’s
actual smile, but rather to show the real intentions hiding behind the
smile; it is not a description of a specific kind of smile, but an at-
tempt to reveal the evil intentions it masks.
Farman’s novel al-Markab (The Ship, Beirut 1989) deals, among
other things, with dilemmas concerning the old versus the new, Is-
lamic tradition versus secularization, relations among young people,
and the status of women in the Arab world today. One of the novel’s
central characters is Suham, an educated young woman who must
cope with attempts by the management at her workplace to prevent
her promotion, using various means including the spread of false ru-
mours and slanders about her. Just as in Alam al-sayyid Ma¨ruf, here,
too, the directors in Suham’s workplace can be understood as alle-
gorically representing Iraq’s political leadership, which uses informers
to keep an eye on every stratum of society. Jabir is the ‘spy’ sent to
‘sniff ’ around Suham and to follow her around. When she sees him
she describes him as a predator lying in wait for its prey:
27
ÆÂœ s ÊU
FI ÁUMOË ¨W!OI« t
U
« r
Ád¬ w UH«Ë t
«d ¨Ã—b« bF« XM
I went up the staircase and then saw him standing at the top, smiling
his ugly smile, his eyes two spots of blood.
The two bloodspots (literally: ‘pools of blood’) in the eyes evoke
the image of a predator lying in wait for its prey, of a beast whose
eyes are bloodshot with full veins as it thirsts for blood. Iraq’s blood-
thirsty intelligence services have been widely described: the Iraqi
Communist poet Badr Shakir al-Sayyab (1926–64), for example,
wrote a famous poem entitled ‘al-Mukhbir’ (‘The Informer’), about
the Iraqi regime’s spying activities directed against its own people.
Some have gone so far as to call Iraq a ‘police and spy state’.28 Spies
26
Farman, Alam al sayyid Ma¨ruf, 40.
27
Farman, al Markab (Beirut 1989), 113.
28
Salim Fattal, In the Alleys of Bagdad (Jerusalem 2003), 310 11 (Hebrew).
238
THE CASE OF GHA}IB TU{MA FARMAN
reached into people’s very homes, so that family members were afraid
to express their thoughts even within their own four walls lest a fam-
ily member inform on his brother or father.
In addition to metaphors derived from the world of predators
Farman also mentions some wily creatures who entrap their prey,
such as the spider. When the hero of Alam al-sayyid Ma¨ruf spoke
with his superior,
29
Æt«— ‚u oKF uJM ZO ¡«—Ë s tOK qD
œUMF« ÊUDO È«—
He saw the devil of stubbornness observing him from behind a spider-
web hanging above his head.
In the Qur’an the spider symbolizes an unreliable crutch, on
which those who followed other gods instead of Allah leaned,30 a
creature which wages war against its potential prey by weaving its
web to entrap it. In the same way Farman’s hero feels as if he were
prey about to fall into a waiting spider-web. There is an obvious as-
sociation between the manager, symbolizing the Iraqi regime looking
for a chance to pounce on any man of conscience standing in its way,
and the spider, a crawling predator which entraps insects and de-
vours them after they are caught in its web.
The phrase ‘hanging above his head’ creates an association with
the hangman’s rope, with which not a few Leftist intellectuals made
their acquaintance during the second half of the twentieth century.
Furthermore, the image of the spider-web adds to the sense of hope-
less entanglement, a feeling which accompanied Farman throughout
his years of wandering, for he had the unpleasant experience of los-
ing his citizenship, and of course his passport, too, because of his
political opinions. To a large measure his Leftist stand drove him
into homelessness.
By his skillful manipulation of similes and metaphors borrowed
from the domain of predators and relations in the wild, Farman
manages to convey the sense of persecution felt by Leftist intellectu-
als in Iraq, who live under a regime which constantly threatens their
freedom of expression and of action, and often their very lives. In
fact, it can in fairness be said that the use of metaphors borrowed
from the animal world enables the writer to express his displeasure,
to say the least, at the regime’s despotic attitude towards Iraqi intel-
lectuals, in a way which circumvents the censorship which is often
imposed on written works; this is so because as long as the criticism
29
Farman, Alam al sayyid Ma¨ruf, 112.
30
Qur’an 29:41.
239
THE CASE OF GHA}IB TU{MA FARMAN
is in the form of metaphor and not explicit, the authorities have little
excuse to attack the author, although writers have been subject to
limitations even with no apparent justifiable cause in Iraq and in
other Arab countries. Metaphors thus serve in this case both as a
medium of expression and as a kind of armour, protecting the author
from being stifled by the authorities.
A sense of alienation and of not belonging — the experience of exile
Farman’s second aim in his use of imagery borrowed from the animal
world is to express the kind of feelings of alienation and of being a
stranger which can be experienced by a private individual even within
his own family and people. Exile, physical or metaphorical, has been
the fate of many Leftist intellectuals in Iraq throughout the second
half of the twentieth century. Many went into exile; some were forced
to leave, others went of their own volition, but often it was a combi-
nation of the two. The process of looking for a way out of the intellec-
tual siege under which the Iraqi regime placed its intellectuals in-
volved much pain, which began in the homeland, where intellectuals
felt they were in internal exile, since they were unable to find work
because of orders from above, they were persecuted for their political
opinions, and their freedom of expression was curtailed. The next step
was exile in another Arab country; there the intellectual would try to
earn a living, but usually found that the restrictions on his freedom of
expression were no less stifling than before. Some intellectuals there-
fore left the Arab world and went into exile in Western Europe. There,
despite being relatively free to express themselves, they pined for their
homeland and their native tongue and did not therefore become fully
integrated into the local society and its culture. Farman’s friend, the
Jordanian-born Saudi author ¨Abd al-RaÌman Munif (1933–2004),
added that as if these difficulties were not enough the host country, if
it did not extradite the exiles to their country of origin, placed restric-
tions on their activity that turned them into virtual prisoners or hos-
tages.31 At any rate, the usual path which the life of an intellectual
took was: internal exile → exile in another Arab country → exile in
the West.
The experience of exile, with its attendant feelings of alienation
and of being an outsider, is often expressed in Farman’s writings by
means of metaphors borrowed from the animal kingdom. In the
novel al-Makha∂ (Labour Pains, Baghdad 1974) Farman tells the
story of Da’ud Karim, an Iraqi student abroad. The reader is exposed
31
¨Abd al RaÌman Munif, al Katib wa’l manfa humum wa afaq al riwaya al
¨arabiyya (Beirut 1992), 89.
240
THE CASE OF GHA}IB TU{MA FARMAN
241
THE CASE OF GHA}IB TU{MA FARMAN
37ÆÆÆwFOD«
I seem like a turtle on its back near the shores of life, kicking my feet
up in the empty air, greeted by the silent face of the sky. I move in
desperation on the sand of the shore, wishing I would return to my
natural position …
Such are the feelings of a person who returns to his homeland and
does not find his place: he is detached from his surroundings and
34
Farman, Åilal ¨ala al nafidha, 182 3.
35
Ibid.
36
Farman, Åilal ¨ala al nafidha, 285.
37
Farman, Åilal ¨ala al nafidha, 182.
242
THE CASE OF GHA}IB TU{MA FARMAN
subsists helplessly on life’s margins, at their side and not within. The
animal-world imagery also reflects the author’s personal experiences
in a natural way: like the hero of Åilal ¨ala al-nafidha, Farman was
not welcomed by the authorities when he visited Iraq after having
been exiled from it, and was therefore unable to renew his contacts
there and become reintegrated into its society. Farman was always
considered an undesirable and, as he often admitted to his friends,
he was constantly afraid that his passport would not be renewed.38
The travails, the alienation and the loneliness of the intellectual in
Farman’s world can be seen in most of his works. His characters of-
ten feel trapped and in a mental dead-end. Despite the optimism
which he was so careful to preserve in his conversations with his
friends even after decades of exile, the depressive mood and eco-
nomic and physical difficulties inherent in a life of wandering like
his become evident through his imagery. His works are full of auto-
biographical allusions, which can be deciphered by means of the
metaphors he uses:
¨¡uC« XU ¨qOK« w ¨W d{ô« ÊuJ sU XO« Æl Ub t√ WU t
KI
«
v« VI s W
d! »—UIF« qI
MË ¨WLKE« U
«Ë“ w dO«dB« wMG ¨¡UA ô« sH
39ÆVI
His mother’s clock greeted him with nine strokes. The house is as quiet
as a grave at night, with weak light and rotten innards. Grasshoppers
sing in the dark corners and scorpions run freely from one hole to an-
other.
Here Farman creates a pun, since the word ¨aqarib means both
‘hands of a clock’ and ‘scorpions’. Throughout the story the hero is
overwhelmed by his feeling of being persecuted by superiors and
of living in conditions which make it impossible for him to progress,
just as the scurrying scorpions symbolize an immediate and grave
danger.
A person’s inner feelings are often projected unto his surroundings,
and here the hero’s distress is reflected also in what he sees:
40 ÆWd
H U«uO ”˃— q" uO« WO« ‚u s qD —Uô«Ë
The trees tower above the fences of the houses like the heads of beasts
of prey.
38
Mentioned by the writer ¨Abd al RaÌman Munif at a gathering held after
Farman’s death: Munif, al Katib wa’l manfa, 118.
39
Farman, Alam al sayyid Ma¨ruf, 94.
40
Farman, Åilal ¨ala al nafidha, 286.
243
THE CASE OF GHA}IB TU{MA FARMAN
The hero feels out of place in his family and within Iraqi society.
When he walks along the street he feels that even the Baghdadi land-
scape, the trees lining its streets, are threatening and fraught with
danger. The sense of pressure and hopelessness is expressed also
through pseudo-legendary images, such as that of the monster
tinnin, a metaphor for the fictional narrator’s claustrophobia, which
takes away the air from the hero’s lungs:
s t U oM
œUJ
Ë ¨5
zd« s ¡«uN« h
1Ë ¨fHM« q«œ gOF
5M XLB«Ë
41
Æq«b«
Silence is like a tinnin living inside the soul, drawing the air out of the
lungs and almost choking their owner from the inside.
This metaphor induces a sense of suffocation and of being
hemmed in. The tinnin, a snake-like or fish-like sea monster, which
is here represented as a creature whose thirst knows no bounds, is
considered a harbinger of misfortune. According to one of the beliefs
associated with it, if a sick person dreams about a tinnin it means
that his end is drawing near.42 The novel’s hero is in internal exile,
surrounded by a sense of being a stranger in his own country; his
silence feels to him like a tinnin within the soul, sucking away all his
strength.
In this way Farman uses the tinnin and the other creatures in or-
der to describe the feelings of estrangement and alienation which
surround the characters in his writings even inside their homes and,
through them, he characterizes his own feelings as well.
The image of the city in Farman’s works
Samia Mehrez has described Nagib MaÌfuz (1911–2006) and Gamal
al-Ghi†ani (b. 1945) as historians of Cairo, since both in their writ-
ings have described that city over an extended period, in a sense
documenting its history.43 Like these two, Farman can be considered
a documenter of Baghdad. One of the ways in which he conveys the
way he remembers the city is by the use of imagery borrowed from
the animal world, through which he depicts Baghdad’s winding al-
leys as a place which strangles and imprisons, and in which the indi-
41
Farman, Åilal ¨ala al nafidha, 284.
42
Kamal al Din MuÌammad bin Musa bin ¨Isa al Damiri, (notes and intro
duction by AÌmad Îasan Basj), Îayat al Ìayawan al kubra (Beirut, 1994), vol. I,
238 9.
43
Samia Mehrez, ‘Re Writing the City: The Case of Khitat al Ghitani’, in
Egyptian Writers between History and Fiction (Cairo, 1994), 58 77.
244
THE CASE OF GHA}IB TU{MA FARMAN
44
Shmuel Moreh, ‘Town and Country in Modern Arabic Poetry from Shawqi
to Al Sayyab’ in Studies in Modern Arabic Prose and Poetry, (Leiden 1988), 136 60.
45
Farman, Khamsat aÒwat (Beirut 1967), 8.
46
Farman, Khamsat aÒwat, 7.
47
Farman, Alam al sayyid Ma¨ruf, 8.
245
THE CASE OF GHA}IB TU{MA FARMAN
The sun would rise from behind a thousand walls in his city which
wound around itself like a cocoon, with its closefitting walls and nar-
row alleys.
The metaphor of the cocoon is somewhat misleading, since a co-
coon protects the pupa as it is transformed into a butterfly, whereas
the phrase ‘behind a thousand walls’ indicates fear, reservation and
confinement, in a way which is reminiscent of a number of lines by
the Iraqi poet Buland al-Îaydari (1926–96), also an exile:
œ«bG wœ—UD
wdU%
48Æ…¬d*« U
«Ë“ q w
Baghdad is pursuing me
It is besieging me
In every angle of the mirror.
In his very first novel, al-Nakhla wa’l-jiran (The Palm Tree and the
Neighbours, Beirut 1965), Farman tells the story of Salima the baker,
her stepson Îusayn and other wretched characters living on the mar-
gins of Baghdadi society during World War II. This work is consid-
ered the first mature novel ever written in Iraq, addressing the com-
plex inner world of its characters. As the main character, Îusayn,
walks through the streets of Baghdad at night the narrator provides
the reader with a depiction of poverty, filth, disorder and negligence:
q bMË ÆWULI« «uU …¡uK2Ë ¨Wu« …UO*U WKKË …—cË ¨WA u qOK« w œ«bG XU
49ÆWU{ WDË ¨VzU VK Wu
Baghdad at night was desolate, filthy, moist with dirty water and filled
with heaps of garbage. At every heap there was a stray dog and a lost
cat.
In this passage Farman makes use of animals to stress the neglect
found in Baghdad, the piles of garbage which attract hungry stray
beasts. Farman emphasizes the city’s unsightliness, in the tradition of
Socialist writers who view the city as a place where individuals are re-
pressed by the authorities, where the neglected masses are incapable
of taking their fate into their own hands and to make for themselves
48
Buland al Îaydari, ‘Fi †ariq al hijra min Baghdad’, al A¨mal al kamila (Ku
wait, 1992), 671.
49
Farman, al Nakhla wal jiran (Baghdad [?] 1988), p. 134 (first published in
Beirut, 1965).
246
THE CASE OF GHA}IB TU{MA FARMAN
The sunset possesses a charm for Mr. Ma¨ruf which no other charm in
the world can equal. Whenever he encountered it from his isolated
corner in the coffee-shop overlooking the river nothing else around
him could lure him: not the birds which hovered above him and dived
into the river, rising high into the clear skies to disappear from the
world, for he knew that the birds would eventually return to their
nests […]
Birds have nests to which they return and from which they fly
away whenever they want. But Farman is denied this right; every
time he comes back to Iraq for a visit he is unsure whether the au-
thorities will let him leave again. Iraqi intellectuals were usually not
permitted to make private journeys abroad, and were never certain
whether they could enter or leave again.51
Another metaphor for unrealized freedom and happiness taken
from the animal world is the butterfly, symbolizing beauty, lightness
50
Farman, Alam al sayyid Ma¨ruf, 7.
51
Sulayman al ¨Askari, ‘al Muthaqqaf al ¨iraqi min al ÌiÒar al siyasi ila al manfa
wa’l mawt’, al ¨Arabi, 2001, issue 511, 8 13.
247
THE CASE OF GHA}IB TU{MA FARMAN
and mobility. Ironically, Farman, who throughout his adult life wan-
dered from one country to another until he settled in Moscow, did
so without the freedom enjoyed by the unfettered butterfly:
b
d
ôË ¨ÂUM*«Ë WEIO« 5 Á¬— rK w tQË ¨Ác WLU« …uAM« WE( qOD
Ê« b
d
ÊU
nOD« tM VOG
ô v
¨dU rK ô« u U p– Ê« dJH
Ê« b
d
ôË ¨tOMO `
H
Ê«
Ë« Æb
dA« ‰UO)« «eHË ¨”UM« ÂU “ 5 ¨se« s WKH w ¨UHD t d Íc« VO(«
Ê« ·U
uË ¨wU“ Õb w UNF{u ¨WuK W«d t
b
w XFË qH q" ÊU
UNd
Æu9Ë W«dH« oM
Ê« WU ÕbI« oM vK tH lC
Ê« UC
« vA
Ë ¨XKH
52Æ»dNË ¨W!
H« v« Íb
N Ê« v« ÕbI« w »dDC
248
THE CASE OF GHA}IB TU{MA FARMAN
Gha’ib Tu¨ma Farman also exploits them for conveying to the reader
as authentic a sense as possible of his own inner world and his frus-
trations as an Arab intellectual, as well as a window into some topics
which were of paramount importance to him as a Communist Iraqi
writer. Unlike medieval Arab authors, Gha’ib Tu¨ma Farman’s im-
agery is an instrument in the service of his ideology. In other words,
Farman provides a case-study of how Arab metaphor has changed
under the influence of Western culture, especially Communism.
First of all, Farman’s animal-world imagery provides a concise,
precise and fascinating description of the relationship between intel-
lectuals and the regime in mid-twentieth-century Iraq. It emphasizes
the sense of persecution which Leftist Iraqi intellectuals suffered dur-
ing the period of intense anti-Communism, and also the regime’s
arrogant behaviour towards Leftist intellectuals, whom it demonized
and de-legitimized. Farman with his metaphors conveys to the reader
the emotional world of an intellectual who wishes to deliver his
country from its backwardness and his people from its degenerate
religious beliefs, by means of an ideology which would free it from
the shackles of the past. This man was persecuted inside his own
homeland and as a result was forced into exile, just as Farman him-
self was.
It is not by chance, therefore, that the political leadership is com-
pared to a beast of prey, since the Iraqi authorities treated Farman
and his colleagues in a ‘predatory’ way, just as they treated Iraq’s reli-
gious minorities. Nor is it by chance that Farman’s heroes are com-
pared to hunted prey. The hunt is a metaphor for the regime’s ‘cap-
ture’ of intellectuals, whom it does not permit to disseminate their
ideas, to make a living in Iraq, and to freely publish their works.
The species of animals chosen by Farman are also intentional. It is
not by chance that the animals he uses bite, lure their prey cleverly
into a web, or possess many long arms. All these are intended to
serve his purpose, namely to provide a concrete description of the
lives of Leftist Iraqi intellectuals from the nineteen-fifties on, who
were trapped in the talons of the regime. In addition, the animal
metaphors can be said to have provided the author with a modicum
of protection against accusations by the authorities. After all, if noth-
ing explicit is said, and all is expressed in the form of metaphor and
simile, the regime supposedly cannot charge the author with any-
thing, although in fact Arab regimes have never been in need of con-
crete proof in order to arrest or exile a writer.
Secondly, animal metaphors help Farman describe the alienation
and solitude felt by his characters, in Iraq and abroad. An intellectu-
249
REVIEWS
GEORGES BOHAS and MIHAI DAT, Une théorie de l’organisation du lexique des langues
sémitiques: matrices et étymons (Collection Langages). ENS Éditions, Lyon 2007.
Pp. 235. Price: /24.00 paperback. ISBN: 978 2 84788 076 2.
PHILIPPE CASSUTO and PIERRE LARCHER (eds), La formation des mots dans les langues
sémitiques (Langues et langage 15). Publications de l’Université de Provence,
Aix en Provence 2007. Pp. 202. Price: /25.00 paperback. ISBN: 978 2 85399
660 0.
Two French publications dealing with word formation in Semitic languages ap
peared in 2007, one by G. Bohas and M. Dat, the other one by a team of ten
authors, who have participated in a syposium held in 2003 at Aix en Provence. As
expected, the first work provides further developments and illustrations of Bohas’s
theory on the ‘matrix’ and the ‘etymon’ (cf. JSS 50, 2005, pp. 205 7), especially in
Arabic and Hebrew. The starting point is the existence of primitive biconsonantal
bases C1C2, called ‘etymons’. The third radical is an ‘increment’ serving to express
a particular connotation of the basic meaning and leading to the formation of
three consonantal radicals. This is realized by various means, in particular by gemi
nation, prefixion, suffixion, metathesis and reduplication. Chapter I presents ex
amples in Arabic (pp. 41 68), while Chapter II examines the question in Hebrew
(pp. 69 97). An important change in Bohas’s theory is the welcome recognition
that also w and y may be constituent phonemes of the ‘etymon’ (p. 49, n. 13). The
‘matrix’ consisting of reversible phonological elements is then analysed in Chapters
III (pp. 99 132), IV (pp. 133 46), and V (pp. 147 54) on the basis of several ex
amples in Arabic. Chapter VI deals thereafter with similar cases in Hebrew (pp.
155 96). The consequences of the theory are developed in Chapter VII (pp. 197
222). They lead to the conclusion that a ‘matrix’ obtaining a similar meaning in
various language families does not prove their common origin. In fact, the ‘matrix’
as linguistic sign preserves some perceptible features of the object or action signi
fied, unless the ‘mimophonic’ weight of the sign is lost in the course of linguistic
evolution, distorting the phonological ‘mimicry’ of the ‘matrix’. In other words,
the original linguistic sign is not arbitrary, contrary to Saussure’s opinion, but has
onomatopoeic origins. The theory is based on correct observations, unnoticed by
R. Weipert (ZDMG 150, 2000, pp. 656 60). However, it operates with abstrac
tions, since exclusively consonantal ‘matrices’, ‘etymons’, ‘increments’, and ‘radi
cals’ cannot shape a living language. The question is how vocalic phonemes at
least a and i/u can be integrated in this system. The use of epenthetic vowels,
common in Maghrebine Arabic and in Gurage dialects, does not answer the ques
tion, for vowels can be structural elements of the language, just like the semi
vowels w and y.
Related questions are discussed by some authors in the second work, divided in
four sections. In the first one, Ph. Cassuto deals with the classification of words ac
cording to mediaeval grammarians of Hebrew (pp. 15 43). V. Porkhomovsky then
presents the conception of root and word formation according to earlier Russian
Semitists (pp. 45 52), in particular A.M. Gazov Ginzberg, who in 1965 antici
pated Bohas’s theory on the onomatopoeic origins of the Proto Semitic roots, N.V.
Yushmanov (1896 1946), who assumed that Proto Semitic had archiphonemes or
251
REVIEWS
‘diffused’ phonemes, S. Mayzel (1900 52), who accepted the triconsonantal origin
of the Semitic root, and I.M. Diakonoff, whose Afrasian Languages (Moscow 1988)
reflects a mainly diachronic perspective. Asking in his study Whether Language was
Imitative in Its Origins? (Moscow 1965 [in Russian]), Gazov Ginzberg distinguishes
four types of word formations by imitation of natural sounds: the ‘internal imita
tion’, based on human sounds, like snoring, laughing, crying, etc., the ‘external imi
tation’ of animal voices and all kind of noises, the gestures of oral organs, like open
ing or shutting of the mouth, and the babbling or nursery words. The question is,
of course, whether onomatopoeia was the principal source in the origin of language
or has played only a limited part. Attested phonetic differences provide the starting
point of Yushmanov’s research. In his posthumous Studies in General Phonetics,
Semitics, and the Morphology of Classical Arabic (Moscow 1998, pp. 126 90 [in
Russian]) he assumes that five ‘archiphonemes’ existed originally in Proto Semitic: a
dental (*t), a velar (*k), a liquid (*r/l), a labial (*p), and a nasal (*n). This means, for
instance, that the distinction of voiced, unvoiced, and emphatic consonants repre
sents a differentiation of the basic sounds in the course of linguistic evolution. The
same ought to be said about the spirantized and palatalized occlusives, changes that
can happen merely in order to ease the articulation. Yushmanov’s hypothesis is thus
basically diachronic and it could be reformulated nowadays in the broader frame of
the Afro Asiatic language family. Therefore no fundamental difference seems to ex
ist between Yushmanov’s theory and Diakonoff ’s approach, contrary to what
Porkhomovsky would seem to suggest (p. 52).
The next contribution, by J. F. Prunet, provides a critical overview of linguistic
approaches to the problem of the Semitic root, adding a large bibliography
(pp. 53 80). Prunet considers the ‘consonantal root’ and the ‘vocalized root’ to be
a single, indivisible morpheme, but he seems to regard the choice of the vowel in
particular roots as an empirical factor, which is at least questionable. In the second
section, the accurate analysis of roots and morphemes by Chr. Touratier (pp. 83
95) shows that the grammatical system of Semitic consonantal roots results from
an improper projection of the consonantal script into the realm of spoken lan
guage. P. Larcher’s study of the interrelation between root and pattern leads to the
conclusion that these grammatical categories do not suffice to explain the forma
tion and meaning of particular words (pp. 97 112), while L. Edzard examines
blends and compounds in the word formation of modern Semitic languages,
i.e. Arabic, Hebrew and Amharic (pp. 113 47). In the third section, M. Tosco
discusses the crisis of Semitic morphology in Maltese (pp. 151 63), while A.
Zaborski formulates some risky hypotheses concerning verbal formations in
Semitic and Cushitic (pp. 165 72). The last section of the work contains a discus
sion of problems related to the gender and number of nouns by H. Gabrion
(pp. 175 84) and an examination of the construct state in Babylonian by R.
Mugnaioni (pp. 185 200).
The study by G. Bohas and M. Dat contains a comprehensive bibliography
(pp. 223 31) and an index (pp. 233 5), while each contribution to the proceedings
of the Aix en Provence symposium is followed by a bibliography. Both works have
a highly theoretical profile, but provide some useful material and new insights for
the understanding of spoken and written Semitic languages.
doi:10.1093/jss/fgn052 EDWARD LIPINSKI
UNIVERSITY OF LEUVEN
252
REVIEWS
253
REVIEWS
JEAN JACQUES GLASSNER. Translated and edited by ZAINAB BAHRANI and MARC
VAN DE MIEROOP, The Invention of Cuneiform: Writing in Sumer. The Johns
Hopkins University Press, Baltimore 2003. Pp. xvii + 266. Price: £31.00 hard
back. ISBN: 0 8018 7389 4.
Glassner’s book is dedicated to a fascinating topic of interest far beyond the com
munity of Near Eastern specialists. The origins and impact of writing have attracted
the attention of scholars in a great variety of fields in recent years. Writing in Sumer
has aroused particular interest, since this is traditionally held to be the earliest in
vention of writing. Here Glassner reassesses the Near Eastern material in the light
of research conducted in other fields.
254
REVIEWS
Before launching into the main thrust of the book, Glassner rightly uses the first
chapter to detail native traditions about the invention of their writing system. And
here he proposes new interpretations of some of those traditions. Glassner brings
new ideas elsewhere too, some of them provocative (such as the hypothesis on
p. 199 that writing in Mesopotamia might have been invented for the purposes of
divination).
Occasionally, greater support would have been desirable for certain claims, as for
example when the author states: (p. 224) ‘Especially with the invention of the
Greek alphabet, writing lost its autonomy and placed itself purely at the service
of the spoken language’; (p. 225) ‘Next to prayers, offerings, and sacrifices, men
address letters or written requests to the gods, which the gods are more likely to
respond to than to oral requests’; or (p. 204) ‘to them [scribes and scholars], all
phenomena became first of all a graphic sign’, which signs were ‘perceived as the
source from which power emanated' and to which ‘the entirety of reality was
ascribed’.
Chapter 10 sees the author expound on the power of writing during the third
millennium BCE. In a lengthy section dedicated to the significance of writing the
divine determinative in front of certain royal names, the author hypothesises that
(p. 200) ‘The use of the determinative tends to bring the human ruler near to the
divine sphere, while the title digir/ilum [“god”] draws him away from it, as no
divinity ever bore it.’ Thus while rulers are described as ‘god of his city’ (pp. 200
1), ‘A god is always the “king”, lugal/sarrum, of his city or state’. Furthermore
(p. 201): ‘As that term [“god”] was never used for gods themselves, it removed
Naram Sin from the divine sphere and inserted him into the vocabulary of human
institutions’. The reviewer would see the titulature more as a unifying than as a dis
tinguishing feature. Rulers are, of course, most commonly referred to as kings
rather than gods of their city; meanwhile gods may be particular to a ruler (‘his/my
god’), bear an attribute (‘angry goddess’, ‘goddess with patient mercy’, ‘goddess
whose station shines from pure heaven’) or be associated with a place. Note ensí ke4
digir uru na ke4 rá zu im ma bé ‘the ruler prayed to the gods of his city’ (or ‘the
ruler [Gudea, whose name is written without a divine determinative], the god of his
city, prayed’; Gudea cylinder B i 15); ki tus maÌ zu sè digir ki lagaski a im si gam e
dè es ‘the gods of the land of Lagash bow down before your august residence’
(Ishme Dagan B 26); or á ba uruki ba digir bé ne ki bi sè ba an gam e es ‘On ac
count of its might, the gods of those cities bow down towards it’ (Lugal e 40).
More generally note that the epithet digir kalam ma ‘god(s) of the Land’ is not un
common and may be applied either to a deity (such as Haia or Ninshubur) or to a
king (such as Shu Sin or Ibbi Sin).
The reviewer agrees with the translators’ concluding sentiments (p. xvii). For
while this book will draw criticism (see most recently and most stridently
Englund’s review in JAOS 125 [2005] pp. 113 16), it challenges readers to con
template their views on writing, and the basis on which those views are held.
doi:10.1093/jss/fgn054 JON TAYLOR
THE BRITISH MUSEUM
255
REVIEWS
ary of the University Museum of the University of Pennsylvania (PSD for short)
achieved B, then A in three volumes, and was then overtaken by the electronic age
and transmogrified itself into a website.
There are of course good reasons for this unsatisfactory state of affairs. Principal
among these are the difficulty of eliciting from a given cuneiform sign an idea of its
pronunciation which can be rendered in a romanized transcription. Such transcrip
tions will not of course be accurate phonetic realizations, but should have a demon
strable validity in terms of the system of transcription agreed by Assyriologists,
which has been constructed through the mediation of the Akkadian syllabary and
uses the Latin alphabet to represent the Sumerian phonemes, which our transcrip
tions attempt to identify and approximately render. Sometimes, especially in earlier
texts of the 3rd millennium, we simply have no idea of the ‘pronunciation’ of a
given sign in a given context, or even in any context, and then we are reduced to
referring the sign by its sequential number in a modern list (e.g. LAK 321, meaning
A. Deimel, Liste der archaischen Keilschriftzeichen, No. 321). When we add to this
the discrepancies between different modern transcriptions of the same word, either
as a result of changing opinions through time, or simply disagreement among
scholars, either about proper transcription procedures or about the substantive ‘pro
nunciation’, it is clear that the identity of a Sumerian ‘word’ can be rather elusive,
and that it is difficult to construct a dictionary along the lines of the romanized
alphabet. This was pointed out to me many years ago by Edmond Sollberger, who
himself produced a very innovative glossary in his TCS 1, when he said that the
great advantage of Deimel’s Sumerisches Lexikon was its arrangement by the
cuneiform signs, since they do not change with the whims of modern Sumerolo
gists, unlike romanized versions which are subject to frequent change in response to
genuine advances in our knowledge of the language, and fashions in the principles
of transcription.
Reasons there may therefore be for the lack of a dictionary, but it remains unsat
isfactory. Until the PSD went on line, apart from glossaries to groups of texts, such
as Römer's Sumerische Königshymnen der Isin Zeit, or Sollberger’s TCS 1, the stand
ard short cut for finding the meaning of a transcribed Sumerian word was to look
in one of the Akkadian sign lists (primarily R. Labat [and F. Malbran Labat],
Manuel d’Épigraphie akkadienne, or R. Borger, Assyrisch Babylonische Zeichenliste).
These show Sumerian words, but only when the cuneiform signs were used as logo
grams ‘Sumerograms’ to write Akkadian words. Sumerian words were also
fundamental to the rich lexical tradition of indigenous Mesopotamian scholarship,
and much of our knowledge of the meaning and usage of the Sumerian vocabulary
comes from lexical texts and bilinguals. These equivalences do not however feature
in the Akkadian sign lists, unless they were also used as logograms. In recent years
to track them down the publication of B. Hübner and A. Reizammer, Inim Kiengi
(1985) has been a useful aid, since it brought together readings and meanings of
Sumerian words from the recent scholarly literature, including the lexical and bilin
gual sections of the two major Akkadian dictionaries, and other glossaries. It did
not have pretensions to be a dictionary, but was a very handy index which enabled
the Sumerologist to track down the hard facts behind the current accepted realiza
tion of a word.
There was therefore no such thing as a modern Sumerian dictionary in book
form, and clearly the volume under review seeks to fill the gap. It derives from 20
years of work by its author, John Alan Halloran, during which he put together a
web site offering a ‘Sumerian lexicon’. Like Inim Kiengi, it started not so much
256
REVIEWS
257
REVIEWS
STEFANIE U. GULDE, Der Tod als Herrscher in Ugarit und Israel (Forschungen zum
Alten Testament 2 Reihe 22). Mohr Siebeck, Tübingen 2007. Pp. xiv + 283.
Price: /54.00. ISBN: 978 3 16 149214 3.
As indicated by the title, the subject of this dissertation is the portrayal of death as
a ruler in Israel, set against an ancient Near Eastern background, chiefly the
Ugaritic texts. There is also reference to the evidence, including glyptic material,
from Anatolia, Egypt, Mesopotamia and Phoenicia. The book has three principal
parts: a lengthy introduction, which runs to 62 pages, on the main topic, i.e. death
as ruler, an analysis of all the relevant texts and lastly, a short summary.
The first part includes an account of how the death of man is portrayed in the
OT: death is negative, makes men equal and belongs in the context of worshipping
a single god. Next comes a section on figurative (bildsprachlich) elements in general:
metaphor, personification and so on, and then there is an excursus on the concept
of myth in OT studies. The conclusion is that myth and metaphor are similar in
function, differing only in context. Myth may be defined as a narrative metaphor.
This is followed by a brief survey of studies on Death as a figure or character
(Figur) in the ancient Near East, including the OT. The main part of the book is an
account of how Death is represented, first in ancient Near Eastern texts, with par
ticular emphasis on the Ugaritic texts, and then in OT texts. There, Death is a glut
ton (Hab. 2:5), a robber (Jer. 9:20), a shepherd (Ps. 49:15) and a partner in a cov
enant (Isa. 28:15, 18). A second excursus is on texts depicting the underworld as
greedy, without reference to Death, showing that this concept is not alien to an
cient Hebrew thought. The texts discussed are Deut. 11:6, Isa. 5:14; 9:19; Num.
16; 26:10 11; Prov. 1:12; 30:15 16; Pss 73:9, 106:17 and 124:3. The short con
cluding section is followed by a bibliography and indexes.
The ancient Near Eastern texts represent Death either as a creeping and furtive
being, especially in Mesopotamia and Egypt, or as a powerful monster, as in
Ugaritic mythology, a depiction that is the result of a long process of development.
In Northwest Semitic texts, Death belongs to the present world, not to the afterlife.
An unexpected conclusion is that in respect of Death as a ruler, the direction of in
fluence was from Syria Palestine to Egypt and Mesopotamia and not the other way
round (p. 240).
In the course of the book quite a few passages are considered, chiefly texts in
Ugaritic and Hebrew. In addition, where relevant, parallels are drawn between the
two sets of texts. Surprisingly, there is little reference to KTU 1.161, even though it
deals specifically with the death of kings (but see p. 187, n. 298). Curiously, there is
no discussion of the sea monster Tnn’s epithet sly†, commonly translated ‘tyrant’
(p. 89), which would seem to belong to the main topic of the book. Two items not
mentioned are P. Xella, Archeologia dell’inferno (Verona 1987) and my ‘Love and
Death Once More (Song of Songs VIII 6)’, VT 47, 1997, 385 7. On the Akkadian
text from Ras Shamra (discussed on pp. 105 7) see now A. Chalmers, ‘RS 25.460
and Early Hebrew Poetry’, UF 36 (2004), 1 9. The Hebrew expression ¨asa Ìozê in
Isa. 28:15b (discussed pp. 215 16) remains difficult, but my proposed solution
(Bib 59 [1978], 133) is not considered. There are a few mistakes: p. 77: ‘Cunéi
forme’ for ‘Cunéiformes’; p. 90, n. 77: the references to Del Olmo Lete, Diccio
nario, 106 and to a Hebrew cognate are incorrect; p. 99: ‘sdatrt’ for ‘sda†rt’; p. 157,
n. 207 ‘canaanite’ for ‘Canaanite’; p. 274: ‘Urprünge’ for ‘Ursprünge’; p. 276: ‘La
Terra die Baal’ for ‘La Terra di Baal’.
Some new suggestions are buried in the footnotes, many of which are quite long
(e.g. p. 200, n. 328). For example, it is proposed that Death’s title mdd il, ‘beloved
258
REVIEWS
of Ilu’, which he shares with Yam, the sea god, is not to be considered as euphemis
tic or ironic but as conferring legitimacy of status in the cosmic order (p. 110, n.
135). Also interesting are the Hittite parallel to Ps. 49:1 13 and 17 21 (cited p.
210, n. 354) and the explanation of the many votive sea anchors found in Ugarit
(p. 81, n. 39). There is much material in this survey that will provoke readers to re
examine the evidence on Death as a ruler in the ancient Near East and perhaps to
resolve the many unanswered questions that remain.
doi:10.1093/jss/fgn056 WILFRED G.E. WATSON
NORTHUMBERLAND
259
REVIEWS
260
REVIEWS
• Even if one accepts the classical join, Hagelia’s reconstruction of line 1 is more
than problematical: [ ] מר ע]די[ וגזר.[[ …‘ ]אsa]id …? tr[eaty …] and he cut [ ]'.
In my view, there is too much room between [( ע]דיFrag. A) and ( וגזרFrag.
B1+2) to see them as parts of one sentence. Besides, the waw before גזרmakes it
impossible to read here the syntagma גזר עד, ‘to make an agreement’, as parallel
to Hebr. כרת ברית.
• Hagelia (pp. 60 2) follows the generally accepted construction of Tel Dan in
scription lines 4 5: [אנה. [א]יתי.] הדד. … ו[יהמלך, ‘Hadad made [me] king, yes
me!’, interpreting the clause as a legitimation formula for a king who did not re
ceive the throne by inheritance. There are, however, a few problems with this
view. The addition א]יתי, supplying the object ‘me’ is far from certain. Within the
parameters of the joined text the addition א]ביresulting in the translation
‘Hadad made [my] fa[ther] king’, would yield an intelligent reading. In my view
hmlk should be construed as a Haph. ‘perfect’ 3.m.s. of the verb mlk: ‘he made
king’. For syntactical reasons it is plausible to suggest that the subject of this
clause preceded the verb and that hdd is the object: ‘X made Hadad … king’.5
Since the word divider after hdd is not certain, Hadad most probably should be
construed as the theophoric element of a personal name belonging to the person
who was made king. Suggesting that I read ‘Hadad’ as the divine name would
indeed turn the idea upside down, but Hagelia is misrepresenting my view
(p. 62).
• In his discussion on phonemes, Hagelia (p. 125) states that in the Tell Dan in
scription the phoneme /ê/ is found. This is correct for the final open syllable in
בה[תלחמה, and אלפי, but incorrect in the form אמריand אלפי. This noun should
be construed as masc. plur. det. and read as, e.g., ’asirayya(’ ), ‘the prisoners’.
• Page 134: construing הם. ארצas a construct chain, ‘the land of them’, might ac
count for the presence of a word divider.
• I would prefer to label איתas the Aramaic nota objectiva, and not as a nota
accusativum (p. 159), since the Aramaic language does not have cases.
• The book is disfigured by some typographical errors of which I only mention a
few: ‘Mobite’ (p.3); ‘Ahutiv’ (p.3); ‘stratiography’ (p. 122); ‘Joün’ (p. 136). At
page 68 the word dividing dots have disappeared.
Despite my critical remarks, I would like to thank Hallvard Hagelia for sharing
with the scholarly world his insights on the various aspects of the Tell Dan inscrip
tion. Nobody interested in West Semitic epigraphy or ancient Near Eastern history
can leave this book unread.
doi:10.1093/jss/fgn057 BOB BECKING
UNIVERSITY OF UTRECHT
ERNST JENNI, Studien zur Sprachwelt des Alten Testaments II. Kohlhammer, Stuttgart
2005. Pp. 351. Price: /45.00 hardback. ISBN: 3 17 018746 5.
This is the second volume of collected essays by the well known Swiss Hebraist
Ernst Jenni edited by Jürg Luchsinger, Hans Peter Mathys and Markus Saur. Eight
years after the first one came out, the present collection of solid studies testifies to
J.’s unabated scholarly activity and, according to the editors’ preface, should again
be considered as a present from the author ‘to the friends of the Hebrew Bible and
its language’. The gift is highly appreciated.
261
REVIEWS
As explained in the author’s preface, the overall purpose of the studies was not to
corroborate or illustrate a specific theory, but rather inductively to take the Biblical
Hebrew text and its grammatical peculiarities as point of departure and from there
to arrive at more general conclusions on the basis of a methodical approach. True to
his usual style, J. refrains from making his point by means of some eclectic exam
ples, but endeavours to be comprehensive in his taxonomy and exhaustive in his
listing of the features in question, a fact that adds considerably to the value of his
writings.
In all, the collection comprises eleven essays from the period 1998 2004, one of
which is published here for the first time. Several articles are tightly or loosely con
nected with J.’s earlier work on the preposition lamed; they refine at some points
the conclusions arrived at in the author’s famous monograph. Others are concerned
with the use of particles, whereas quite a few delve into the realm of modality.
The first study, entitled ‘Vollverb und Hilfsverb mit Infinitiv Ergänzungen im
Hebräischen’ (pp. 11 35; originally published in ZAH 11 [1998], 50 67), dis
cusses cases in which a modal verb such as י ָֹכל, נ ַָתן, ָאָבה, ֵמֵאן, הוִֹסיףor ֵהֵחל is
used together with an infinitive either with or without the preposition l , as in
Deut. 10:10 אבה יהוה השחיתך- לאor Deut. 23:6 בלעם-אבה יהוה לשמע אל-ולא, a fea
ture that standard grammars indiscriminately juxtapose as alternatives. J. shows that
infinitives without l should be considered the direct object of the modal verb,
which itself functions as a full fledged main verb, whereas constructions containing
infinitives with l are to be regarded as a single complex predicate, in which the fi
nite verb serves as an auxiliary. Accordingly, the construction without l is used for
general utterances, while, that with the preposition l refers to a specific, concrete
action or state. The modal verbs used with l are divided into verbs of partial reali
zation (Teilverwirklichung), possibility and volitivity. A critical reader may be left
with the question of whether indeed diachronic considerations play no role here,
but that is not something that will easily come up within J.’s strictly synchronic ap
proach.
The second essay, ‘Epistemische Modalitäten im Proverbienbuch’ (pp. 36 47;
previously published in the Fs. Hans Peter Müller, Mythos im Alten Testament und
seiner Umwelt [1999], 107 11), is concerned with about 80, predominantly late,
cases of infinitive construct with l that should be considered to be the predicate of
a non verbal clause. The function of such infinitives is usually adding a deontic
modal nuance to the predication (i.e. necessity or certain expectation), as in Eccl.
3:15 ‘ ואשר להיות כבר היהwhat must happen, has already happened’. In a small
number of cases in Proverbs, however, we should assume epistemic modality. For
instance, Prov. 16:30 עצה עיניו לחשב תהפכותcould be translated as ‘whoever narrows
the eyes is likely to contrive froward things’. To the literature quoted in this article
should now be added A. Gianto, ‘Mood and modality in classical Hebrew’, Israel
Oriental Studies 18 (1998), 183 98 and id., ‘Some notes on evidentiality in Biblical
Hebrew’ in Biblical and Oriental Essays in Memory of William L. Moran (2005),
133 53.
The third article, entitled ‘Einleitung formeller und familiärer Rede im Alten
Testament durch ’mr ’l und ’mr l ’ (pp. 48 64; published in the Fs. Georg Sauer,
Vielseitigkeit des Alten Testaments [1999], 17 33) suggests that, in the main, the dis
tinction between אמר אלand אמר לhas to do with different levels of politeness;
even though some of the nuances seem to escape the modern reader, a tendency to
prefer אלover לwhen ‘talking up’ is apparent. For a quite different approach, see
M. Malessa, ‘Biblisch Hebräisch ְל/ ִדּבֶּר אֶלund ֵאת/ ִדֶּבּר עִםim Vergleich’ in the Fs.
262
REVIEWS
Muraoka, Hamlet on a Hill (2003), 333 40 and id., Untersuchungen zur verbalen
Valenz im biblischen Hebräisch (Assen 2006; revision of diss. Leiden, 2003), esp.
167 205; there is doubtless room for further discussion.
The fourth essay, ‘Subjektive und objektive Klassifikation im althebräischen
Nominalsatz’ (pp. 65 76; previously in Theologische Zeitschrift 55 [1999], 103 11),
aims to determine the functional difference between classifying non verbal clauses,
such as Gen. 44:16 ‘ הננו עבדים לאדניbehold, we are my lord’s servants’, and what J.
calls ‘re classifying’ non verbal clauses such as Gen. 50:18 ‘ הננו לך לעבדיםbehold,
we are as your servants’, in which the predicate is preceded by the preposition l .
J. concludes that in the latter case there is no complete identity between the subject
(‘we’) and the predicate class to which the subject is re classified (‘servants’). In
other words, the identity is not yet real, but present in the mind of the speaker; the
latter clause could therefore be paraphrased as ‘as of now, we may be considered to
be or become your servants’; the same interpretation holds true for many other
such cases.
The fifth article, ‘Aktionsarten und Stammformen im Althebräischen: das Pi¨el
in verbesserter Sicht’ (pp. 77 106; originally ZAH 13 [2000], 67 90), in fact con
stitutes an appendix to J.’s well known monograph Das hebräische Pi¨el (1968), in
which the author revises earlier conclusions and takes issue with other publications
on the subject, such as Kouwenberg’s Gemination in the Akkadian verb (1997) and
Joosten’s ‘The function of the Semitic D stem’, Orientalia 67 (1998), 202 30.
J. abandons the division into transitive and intransitive verbs as a pivotal feature
and reformulates the function of the Pi¨el on the basis of Vendler’s (‘Verbs and
times’ [1957]) four lexical or situation aspects (Aktionsarten): state, activity, accom
plishment and achievement. With verbs expressing complex actions, the Pi¨el leaves
the process (Hergang) of the action implicit (unausgedrückt): with stative verbs it
serves as a factitive, with accomplishment verbs a resultative, with verbs expressing
complex movements or sound utterances it denotes the successive. Achievement
verbs and transitive activity verbs have no Pi¨el. With verbs of complex activity, a
repetitive action in the Pi¨el is rendered as a punctual achievement.
The sixth contribution, ‘Textinterne Epexegese im Alten Testament’ (pp. 107
17; previously in Fs. Odil Hannes Steck, Schriftauslegung in der Schrift [2000], 23
32) discusses the use of lamed with infinitive construct to modify a preceding utter
ance, such as לאמר: ‘by saying, …’ or Gen. 2:3 ברא אלהים לעשות-מלאכתו אשר-מכל
‘from all the work which He had created by making (it)’ (i.e. the Latin gerund
faciendo), with special attention to the logical semantic relation between the main
verb and the infinitive. In roughly one fifth out of 300 cases of such infinitives with
lamed we have to do with ‘epexegesis’; the infinitive explains the main verb by
specifying it or adding precision.
In the seventh essay, ‘Eine hebräische Abtönungspartikel: ¨al ken’ (pp. 118 33;
originally in Fs. Klaus Seybold, Prophetie und Psalmen [2001], 201 15), the author
scrutinizes the semantic range of כן-על. Whereas the expression usually (144 out of
155 times) introduces a logical consequence and hence may be rendered as ‘there
fore’, the constructions כן-( כי על10×) and כן-( אשר על1×) rather convey an under
lying reason (German: da nun einmal; Latin: quia enim). J. deftly places the two
functions within a single theoretical framework (with an interesting detour to Swiss
German dialectology!).
Essay number eight, ‘Semantische Gesichtspunkte des Hebräischen und deutscher
Übersetzungen am Beispiel von Num 10, 29 31’ (pp. 134 50; previously pub
lished in W. Gross [ed.], Bibelübersetzung heute. Geschichtliche Entwicklungen und
263
REVIEWS
264
REVIEWS
tion of essays by this learned octogenarian constitutes a gift to the academic com
munity, it is a gift that should be gratefully acknowledged.
doi:10.1093/jss/fgn058 MARTIN F.J. BAASTEN
UNIVERSITEIT LEIDEN
265
REVIEWS
Chapter 6 investigates written sources, mainly the Merneptah stela; and chapter
7 defines the hypothesis, in the form of a ‘narrative’. The last chapter (apart from
the very short conclusions at the end) looks at the biblical narrative and compares it
with the hypothesis.
This book is a laudable effort to use archaeological and anthropological models
in the study of a highly charged and dark period, applying a strict methodology.
Unfortunately, the result is the opposite of what the author intended: the method
ology is too fraught with difficulties and uncertainty in the basic data to lead to a
convincing hypothesis.
Apart from that, the method as applied by the author, is also flawed. My main
problem with Miller’s application of the complex chiefdom model is that instead of
using it as a hypothesis, which is subsequently tested for its validity, it is used as the
basis for a hypothetical ‘narrative’ that defines the boundaries of the various sys
tems. In other words, the question is not whether Early Israel consisted of complex
chiefdoms but rather where the boundaries were.
The Gravity Model, used by Miller for his primary analysis, is a way of quantify
ing the idea that interaction between sites (villages, towns, camps) is negatively cor
related to the distance (measured as walking time) between them, and positively to
the product of their size. This model has been tested by archaeologists and anthro
pologists, mainly in the New World, and a summary of results can be found in Plog
(S. Plog, ‘Measurement of prehistoric interaction between communities’, in K.
Flannery [ed.], The Early Mesoamerican Village [New York 1976], 255 71) or
Hodder and Orton (I. Hodder and C. Orton, Spatial Analysis in Archaeology [Cam
bridge 1976]). These summaries show that the model can quantify interaction, and
predict artefact distributions. There are difficulties however. Miller sums them up
for the Highlands: the size of multiperiod sites is far from certain, surveys have of
ten missed smaller sites and there are serious dating issues. Even so, whilst the
Gravity Model is useful to explain distribution patterns of artefacts, Hodder and
Orton (Spatial analysis, 195 6) state explicitly that it cannot be used to analyse so
cial (or ethnic) patterning or territoriality, as artefact distribution patterns do not
necessarily follow ethnic boundaries.
Miller uses the Gravity Model to create maps of interaction clusters, often com
plex systems with three or more tiers. These systems, however, are purely theoretical
constructs, created by applying a mathematical formula to the assumed size and
relative location of every site and then drawing lines between those with the highest
interaction factor. This ‘interaction factor’ varies and can be high, or relatively low
(as long as it is the highest outcome for that specific site), something that is not
expressed in Miller’s maps, where every line drawn between sites has the same
weight.
His application of the model also implies that every site necessarily interacts with
at least one other, introducing a factor of artificiality. Miller then skips a few steps,
and defines the centres and subcentres of these systems as centres of administrative
control (p. 30). So he jumps from a geographical model to a social one, without
any explanation.
Thiessen polygons are then drawn, as corroborating evidence for the distribution
of the clusters. As Thiessen polygons are based on the physical distance between
sites, and therefore a simplified version of the Gravity Model, they necessarily con
firm the boundaries of the various systems, and their corroborative value is zero. A
supportive method quantifying bimodal distribution as an indication of a complex
chiefdom, namely the plotting of the site size against the number of sites, is ren
266
REVIEWS
dered unreliable by the small number of sites involved in each system (56 for the
largest system of Balatah, if I counted right, and much smaller for the others) mak
ing the use of statistics meaningless. In other words, the application of the Gravity
Model, its interpretation, and its ‘support’ from other models, are all highly flawed.
There is no evidence here for complex chiefdoms.
Miller’s extensive (and very useful) enumeration of import artefacts shows evi
dence for outside contacts, possibly for markets and various forms of distribution,
but there are too many anomalies in this record to use as an argument for complex
chiefdoms.
His decision to include the archaeological data from Benjamin, as they are better
documented than those from the area of study, is not a happy one, as it creates an
extra assumption: that the population in Benjamin came from the same back
ground as that in the study area and was organized in the same way.
An extensive description of the ecology of the region shows that every ‘bailiwick’
defined by Miller extended over a variety of ecological environments, suggesting a
mixed economy. It outlines the interaction between the various ecological regions
for the exchange of subsistence goods and is useful regardless of whether the ‘baili
wicks’ actually existed or not.
The other areas of evidence, mortuary data and architectural data, are too scanty
to be of use. Miller’s postulate that monumental architecture necessitates
conscriptive labour, and therefore the existence of chiefdoms, seems outdated, as it
becomes clearer that public monumental projects (such as town walls, or temples)
could well be accomplished by corporate groups, such as tribes (see G. Philip, ‘The
Early Bronze I III Ages’, in B. MacDonald, R. Adams and P. Bienkowski [eds], The
Archaeology of Jordan [Sheffield 2001], 177 8; or E.J. van der Steen, Tribes and
Territories in Transition [Leuven 2004], 111 12 for recent examples), whereas
conscriptive labour can well indicate a feudal system.
The remainder of the book explores the written sources, particularly the mention
of ‘Israel’ in the Merneptah Stela and the biblical narrative. I agree with Miller that
the main conclusion from the Stela is the fact that Israel must have been an impor
tant factor, to be mentioned at all. However, the ‘people’ determinative seems to me
to have more significance than Miller grants it, and unfortunately, one of the things
the Stela does not provide is the geographic location of ‘Israel’.
The biblical narrative is laid beside Miller's own ‘narrative’ (chapters 7 and 8).
Miller is extremely careful to make clear that he does not take the biblical narrative
at face value, but that there may still be truth in it. Nevertheless Miller has missed a
chance here: the classic studies of Lord and Parry (A. Lord and A. Parry, The singer
of tales [Cambridge, Mass 1960]), or, closer to home, the study of Bedouin oral tra
ditions (P. Heath, The Thirsty Sword. Sirat Antar and the Arabic popular epic [Salt
Lake City 1996], B. Connelly, Arab folk epic and identity [Berkeley 1986]), or A.
Shryock’s, Nationalism and the genealogical imagination. Oral history and textual au
thority in Tribal Jordan [Berkeley 1997], a study of the oral and written traditions of
the Adwan and the Abbadi in central Transjordan), would have been enlightening.
The complex chiefdom ‘narrative’ is then used to explain some of the biblical narra
tive. It makes for an interesting story but is hardly proof of the complex chiefdom
society that Miller suggests here.
One of the main problems with most of these studies, that use anthropological
models in the explanation of Early Israel or the archaeology and history of the re
gion in general, is that they use models that apply to a number of ‘primitive’ cul
tures, often from the New World, Africa and Oceania. This is where these models
267
REVIEWS
have been developed and tested, and there is nothing comparable for the southern
Levant.
Nevertheless, this approach is fraught with problems: not only are the cultures
themselves unrelated to those under study but so are the climates, ecosystems and
geology. This renders applications of the models, such as Miller does, theoretical
and tenuous. One area of study that may remedy that shortcoming is that of
Levantine eighteenth early twentieth century societies described in travel accounts
and by Arab historians. They do not provide ready made models, but they describe
a society that has more in common with that of the Iron Age, or any archaeological
period, than those of, for example, Tahiti.
There are some minor irritations in the book. The bibliography is very incom
plete: many publications mentioned in the text are missing or misquoted in the
bibliography; there is the consistent misspelling of names, such as Deir Allah (Deir
¨Alla); or Theissen polygons (Thiessen polygons). It does contain useful informa
tion, especially in the overviews of material culture and ecology. However, as a
model, it remains unconvincing.
Ann Killebrew’s book is a study of the various peoples that populate the Hebrew
Bible, their history, culture and social organization, as can be construed from the
archaeology and the written sources. The book deals with Egyptians, Canaanites,
Philistines and Israelites.
It starts with an introduction into the thirteenth century breakdown of the
Egyptian empire and the consequent collapse of the Late Bronze Age society. The
frame of reference used is World Systems Theory, and the various cores and periph
eries (and semiperipheries) are outlined. Canaan is defined as a periphery, largely of
Egypt. The ‘crisis’ is viewed in terms of systems collapse, although Killebrew prefers
to see the end of the Late Bronze Age as a ‘transformation’ rather than a ‘crisis’. I
tend to agree with her. The ‘crisis’ was rather a restructuring of elements of the sub
systems into a new, less integrated formation which was the Early Iron age.
The next chapter looks at Egypt in Canaan in the thirteenth century and asks if
this was a case of imperialism or colonialism. An extensive survey of the evidence
includes written Egyptian sources both in Egypt and in Canaan, architecture, pot
tery and other artefacts and burial customs. Egyptian presence in Canaan fluctu
ated and was strongest during the reign of Ramesses III. The purpose of Egypt’s
presence in Canaan was to collect a maximum tribute and the military presence at
sites such as Beth Shean, Deir el Balah and others, was to maintain order and se
cure the loyalty and cooperation of the local rulers.
Chapter 3 looks at the Canaanites, a culturally not very well defined group.
Killebrew calls them a geopolitical entity, a ‘multiethnic’ group that shared a lan
guage, a common culture, and formed a ‘social boundary’. They are the ancestors of
various Iron Age sociopolitical groups, such as the Phoenicians, Ammonites,
Moabites and Israelites. Consequently, Killebrew focuses on the history and archae
ology of the Canaanites during the Middle and Late Bronze Ages, as she sees here
the roots of the later Iron Age polities. An extensive survey of the architecture, pal
aces, temples and domestic structures, as well as burial customs, pottery and other
artefacts underline the wide cultural variety. One of the weaknesses of this chapter
is that it focuses entirely on the region west of the Jordan, excluding a large part of
the Canaanite territories, namely Transjordan. Consequently, some conclusions
about Canaanite culture are incorrect, such as the remark that there were no
Canaanite walled towns. There were several towns in Transjordan with Late Bronze
town walls, such as Sahab and Abu Kharaz, and there was a Late Bronze market
268
REVIEWS
centre on the Amman plateau. The region east of the Jordan formed a political and
economic entity with the western region in the Late Bronze Age and should not be
excluded.
Chapter 4 deals with Israel, the ‘Mixed Multitude’. It has a large section on re
gional surveys, on the basis of which Killebrew concludes that the collapse of Egyp
tian control must have created a power vacuum resulting in the settling of nomadic
populations in the region. The Transjordanian Plateau is dealt with in this section
but, unfortunately, the Jordan Valley is not. This is an important omission, as the
Jordan Valley was not only the link between the two highlands but also a focus of
political and ethnic activity both in the Late Bronze and Early Iron Ages (van der
Steen, Tribes and Territories, 301 8).
One of the problems with interpreting the material culture of the Early Iron Age
in the Highlands, as pointed out by Killebrew, is that most interpretations and
analyses are governed by a need to ‘discover’ Early Israel in it. This chapter is there
fore as much a survey of interpretations as of material remains. It looks extensively
at village planning, the pillared building, silos and terraces; traditional ‘markers’ of
Early Israel. After a brief survey of existing hypotheses Killebrew adds her own, the
‘Mixed Multitude’ theory. This defines Israel as a heterogeneous collection of
largely Canaanite tribal and kin based groups, comprising nomads, farmers, Apiru
and Shasu, and gradually merged into an ethnic entity embracing an epic narrative,
recounted in the Bible, and a shared religion, that of Yahweh.
Chapter 5 deals with the Philistines. Killebrew looks at the phenomena of cul
tural diffusion versus colonization, and concludes that the Philistines were colo
nists, originating probably in Cyprus but with roots in the eastern Aegean and per
haps Anatolia. The material culture is analysed using a ‘pilot site’, Tel Miqne /
Ekron, which has been excavated and published extensively, and which has the
longest uninterrupted occupation history of any Philistine site. The synthesis looks
at the origins of the Philistines, and at the chronology here the ‘low chronology
debate’ is touched upon briefly but no conclusions are drawn although Killebrew
seems to favour the ‘middle chronology’. The book concludes with a summary of
the reviewed evidence and an overall reiteration of the conclusions of the various
chapters.
The ‘Mixed Multitude’ model has much to recommend it. It incorporates the
different strands of material culture of Early Israel as well as the widely varying tra
ditions that lay at the root of the biblical narrative and it has the flexibility to adapt
to new evidence, or new interpretations of evidence, without becoming meaning
less. Whether the Yahweh cult was the core ideology of Early Israel seems to me
doubtful, as the Yahweh cult may have been the religion of one specific group from
Edom (Shasu?), and there is evidence for other cults as well. This in fact illustrates
my point that the model is versatile.
Interestingly enough, it is Miller’s section on imported goods in the highland
sites (pp. 45 51) rather than Killebrew’s own section on the archaeology of the
highlands that sums up material evidence for her theory: localized Transjordanian
influences in Cisjordanian sites, the presence of Egyptian objects, the bull on the
Bull site, copper objects from the Wadi Arabah, and even the locally produced Phil
istine pottery from Tell en Nasbeh. These may all be evidence for trade, as Miller
assumes, or they may have been brought and produced by small groups from a
heterogeneous background, Killebrew’s ‘mixed multitude’.
The question that is not answered satisfactorily is ‘why?’. Why did this ‘mixed
multitude’ of social and ethnic groups, migrants, runaway slaves, peasants and
269
REVIEWS
herders, merge into new ethnic or socio ethnic entity? Was it, as suggested in the
Bible, the presence of a common enemy such as the Philistines that made them
stick together? Was it, indeed, a sweeping religious mass conversion comparable to
that of early Islam? Or was it simply a question of power and conquest, one group
subjecting all the others? And what about the other polities in the region, such as
the Ammonites and the Moabites? Did their development into unified polities fol
low the same paths?
These questions still need answers and, despite the vast amounts of literature on
the subject (as provided in Killebrew’s bibliography), we are far from having those
answers. The question of ethnicity in this context is an important one. Accepting
Killebrew’s definition of ethnicity as group identification, the question remains:
why do people identify with a certain group and how and why does that self iden
tification change over time? This has been an important area of study in recent an
thropology. The work of Anthony Smith (A.D. Smith, National Identity [London
1991], The Nation in History: historiographical debates about ethnicity and national
ism [Cambridge 2000]) has been particularly useful in this context and may pro
vide some of the answers.
Both books, Miller’s study of the complex chiefdoms and Killebrew’s mixed mul
titude, present a model for the Early Iron Age in the highlands. Both use the same
sources but their conclusions are radically different.
In my opinion, Killebrew’s model is the more convincing, largely because its
point of departure is the archaeological record (including the written sources),
whereas Miller’s point of departure is the model itself to which the data are sub
jected. Because Miller does not look at the history or the origins of his ‘complex
chiefdoms’, they remain an isolated phenomenon with no roots in reality.
Killebrew’s mixed multitudes have names and histories in the Late Bronze Age as
well as a history of their own; therefore, they are more real, and as such, more con
vincing.
doi:10.1093/jss/fgn059 EVELINE J. VAN DER STEEN
UNIVERSITY OF LIVERPOOL
270
REVIEWS
271
REVIEWS
laws to form the Covenant Code. Jackson also suggests that the two sections of the
Mishpatim likely existed independently before being combined with the rest of the
Covenant Code. In his view, each of these stages of compilation like all biblical
law is influenced by wisdom circles.
It is ultimately difficult to evaluate the main thesis of Wisdom Laws, for the re
construction of earlier redactional strata through indirect evidence and especially
the recovery of originally oral sayings is a provisional exercise. To his credit,
Jackson is careful to note that the institutional and literary histories that he pro
poses are speculative at points and open to charges of circular reasoning (e.g., 387;
433). Nonetheless, the complexity of the compositional history that he proposes for
the Covenant Code and his precise identification of Deuteronomic and Priestly re
visions seem overreaching.
Also problematic is the lack of a strong thesis regarding the relationship between
the ancient Near Eastern legal collections and the Mishpatim. Jackson employs the
matically related ancient Near Eastern laws in his exegesis of the Covenant Code;
however, it is not clear how he understands the nature of their similarities. In his
case for emending Exod. 21:29’s ‘ ישמרנוhe has not guarded it’ according to the
LXX’s reading ‘he has not destroyed it’ (implied )ישמדנו, Jackson argues against fol
lowing the ancient Near Eastern parallels (e.g., CH §251 reads alapÒu la usanniq
ma ‘but he does not control his ox’), stating that such ancient Near Eastern parallels
cannot ‘control one’s judgment of the semantics and institutional context of the
biblical text’ (271 n. 78). He also ‘doubts’ that the original wisdom sayings of the
Mishpatim were formulated casuistically (31). This means that the current form of
the Mishpatim that corresponds with other ancient Near Eastern legal paragraphs is
independent of the content that is likewise parallel among these legal corpora. Such
a view seems unlikely. Nevertheless, these objections hardly undermine this work’s
significant contribution to the study of the Covenant Code, and Jackson is to be
commended for his sophisticated and nuanced analysis.
doi:10.1093/jss/fgn060 JEFFREY STACKERT
UNIVERSITY OF CHICAGO
HILARY LIPKA, Sexual Transgression in the Hebrew Bible. Sheffield Phoenix Press,
Sheffield 2006. Pp. xii + 285. Price: £60.00 hardback. ISBN: 978 1 905048
34 2.
Lipka’s excellent monograph examines the nature and consequence of sexual trans
gression in three boundary areas: religious, communal and personal. Lipka argues
that, contrary to common assumption, personal boundary violation was recognized
and acknowledged in the Hebrew Bible. In the two transgressions she examines,
rape and adultery, she argues that different texts portray differently the nature of
the offence, the impact on the victim, whose rights are violated, who should punish
the transgressor and what the punishment should be.
Although there are places in the discussion where questions of redaction are
raised and she also uses other ANE sources to illuminate where they can, the main
focus of her approach is essentially text based, with a series of close readings and
comparisons of words via intertexts. Although her approach intends to infer some
thing about Israelite sexual attitudes ‘in certain times and in certain places’, she fo
cuses on the community as portrayed within the Hebrew Bible, and its Sitz im
Leben. This approach ensures the text does not float free of its moorings. It looks
for the implied author rather than the empirical, although Lipka herself does not
use this terminology. She underscores the importance of using all genres of writings
272
REVIEWS
273
REVIEWS
274
REVIEWS
ing his known major works at Karnak including, thus, the great triumphal relief,
as Wilson admits (p. 66). Second, the Karnak stela of Shoshenq I is definitely a mili
tary record, gives a reason for armed action (‘vile things’ somebody [else] had
done), followed by [the king going out], ‘his chariots following him’ (it would not
be anybody else!) And so ‘His Majesty made great slaughter’ of the foe, and men
tion is made of the Bitter Lakes. This is the E. border with a foreign foe, not some
minor local squabble (W.’s Dakhla ‘parallel’ is not apposite). Third, the biographical
allusion of Hori (rather than Heru) to following the king on campaign in Retenu
may reflect his presence in the Levant with Shoshenq I but (with Wilson, p. 70)
not necessarily; we can agree that it is indecisive.
Fourth, the Megiddo fragment of a former royal stela of Shoshenq I. Here, again,
Wilson wriggles away from the obvious significance of its presence in Megiddo.
Such stelae, whether erected in towns or on rock faces always indicate Egyptian
overlordship, and often direct control, including the direct result of campaigning by
pharaohs at the time. So, the Shoshenq I stela can be no different: it represents,
however briefly, that king’s dominance over Megiddo and over its local ruler that
is, Jeroboam, king of Israel, Shoshenq’s protégé and intended vassal.
In Chapter 5, Wilson deals with the two biblical references. He accepts the essen
tial features in these, but fails to understand that mention of a Shoshenq reduction
of Israel to vassaldom is not necessary to the purposes of the Kings writer(s), still
less in Chronicles. That the campaign was against Jerusalem alone is also contra
dicted by the presence of the Negeb and south Judean sites included in the List.
Contrary to Wilson (very careless here), the Chronicler does not claim either that
Shoshenq’s conquest ‘devastated… the entire [Judean] countryside’ or that it ‘en
compassed all of Judah’ (p. 85).
Chapter 6 brings us to the book’s final conclusions. These begin with the errone
ous statement that the List can offer nothing on the campaign, merely because the
Relief is merely a formal work. He is right to say that the Megiddo stela does not
imply a Shoshenq destruction of Megiddo; quite the contrary, it was there to re
mind a live population where their ultimate loyalty should go. It does indicate
Egyptian overlordship.
Finally, as an Appendix, Wilson usefully provides a full conspectus of the surviv
ing place names in the List, and of possible readings and interpretations. P. 103
n. 8: contra Wilson, the El Hiba relief was definitely a triumph scene, but is 98%
lost. For List No. 12, in agreement with W, M[aq(q)eda?] is possible epigraphically
because the first bird is not 3 but m. Gezer is ruled out, and Rainey (as cited by W.)
cannot overrule the Chicago edition on this. However, M[aqqeda] is rather off
track; M[aresha] might be geographically superior. With Nos. 15 18, 22 7, 32 5,
38 9, we would be largely in happy agreement. Nos. 53 6 form a clear group
(p3 ndjskt, error in hieratic for p3 n( )skt, ‘one of Succoth’?). For Nos. 57 9,
Zemaraim is beyond reproach; restoring [Ti]rzah seems inevitable, hence ‘the
Tower’ should be Shechem, known from Judg. 9:46; No. 65 ‘the Vale’, that of
Esdraelon. Again, most of this is agreed widely, leaving lost entries aside. Thereafter,
in rows VI X, it is broadly agreed that we have names in the Negeb and around,
but not all solvable in detail. Not surprisingly, W. objects to my reading ‘height of
David’ in Nos. 105 6, where he rides rough shod over both the reading and the
comparative data. The reading of hadibiyat, ‘heights’ in No. 105 is perfectly fairly
restorable on the total known traces. In No. 106, Dwt is also certain. I had never
thought of reading it as ‘David’ (Dwd) until noticing ‘the psalms of Dawit’ in a
context of Psalm 65 in an Ethiopic victory inscription of Abreha in Old South Ara
275
REVIEWS
bia. Here, Dawit IS ‘David’, and cannot be any other. If one Afro Asiatic language
can do this, so can others. And they do. In Egyptian, voiced terminal consonants in
Semitic loanwords can become voiceless e.g., Hebrew hereb, ‘sword’, appears as
H r p in Egyptian. See J.E. Hoch, Semitic Words in Egyptian Texts of the New King
dom and Third Intermediate Period, (Princeton 1997), 233 5, No. 324. So, there is
no practical problem whatsoever in Dwd becoming Dwt here. On Nos. 110 12, all
three should be taken as one (against Na’aman), as ‘Arad of Beth Yeroham’; there is
no ancient rule against triplets! The interpretations of Nos. 121, 124 7, 130, 133,
139, 150 find wide acceptance as cited by Wilson (omitting, again, lost and dubi
ous entries). Finally, in row X, 2a 3a have wide acceptance; then, ¨Ain goren
(Wilson) for 4a is certainly feasible, if still unattested.
Space limits preclude further comments on this topic. Wilson’s work is a useful
discussion, but is not sufficiently au fait with the full available knowledge about
topographical lists to provide a fully balanced view of that of Shoshenq I.
doi:10.1093/jss/fgn062 K.A. KITCHEN
UNIVERSITY OF LIVERPOOL
THOMAS B. DOZEMAN and KONRAD SCHMID (eds), A Farewell to the Yahwist? The
Composition of the Pentateuch in Recent European Interpretation (Society of Bibli
cal Literature Symposium Series 34). Society of Biblical Literature, Atlanta 2006.
Pp. viii + 197. Price: $24.95 paperback. ISBN: 1 58983 163 2.
This collection is the fruit of a special session of the Pentateuch Seminar of the So
ciety of Biblical Literature held in November 2004. Six of the papers included were
read at that meeting (those by Thomas Römer, Konrad Schmid, Jan Christian
Gertz, Thomas B. Dozeman, Christoph Levin and David M. Carr); three (by
Albert de Pury, Erhard Blum and John Van Seters) have been added. The session
was organized by Dozeman, Römer and Schmid in the wake of the publication of
the similarly entitled controversial series of essays edited by Gertz, Schmid and
Markus Witte, Abschied vom Jahwisten (BZAW 315; Berlin 2002), itself prompted
by monographs by its three editors. The present work provides not a mere transla
tion of some of the German contributions to that earlier collection but new essays
(in English) by several of its contributors and new responses to them that are in
tended to carry the discussion forward. Compact but widely ramifying in content,
this collection thus provides in relatively brief compass valuable surveys of ‘the state
of the question’ in key areas of current Pentateuchal criticism.
The hope of the editors is that this collection will ‘facilitate communication be
tween European and North American scholars’ (perhaps one should say, ‘some
mainland European scholars’; North America includes both USA and Canada).
The marked difference between these Continental and North American approaches
is already signalled in the title of the book: the question mark in the title of the
American published volume is missing in the 2002 German published collection
(though one was present in its originally projected form; its eventual omission pre
sumably marked increasing conviction about the thesis).
The view promoted by Römer, Schmid, de Pury, Gertz and Blum, in a variety of
ways and on the basis of a variety of biblical sources, is that it was the Priestly au
thor, or at least the author of PG, the ‘Grundschrift’ that probably ends at Exodus
40, not ‘J’ in any of its guises, who was the first to combine in literary form, in the
exilic period or shortly after, the up till then independent primeval history and tra
ditions of the ancestors contained in Genesis with the alternative and competing
story of Israel’s origins in the exodus under Moses in Exodus. The creation of ‘the
276
REVIEWS
master narrative of salvation history’ is the work of P; the non P material that pre
supposes this master narrative is thus necessarily post P redactional.
Some examples of the Continentals’ argumentation may be given. There are sur
prisingly few overt literary links between Genesis and Exodus (and beyond) and
these turn out to be late (e.g., Genesis 15) (Schmid). Hosea 12 demonstrates pre
exilic rivalry between the Jacob and Moses traditions (de Pury). The respect enjoyed
by Joseph in Egypt and the enslavement of the Israelites there are mutually exclu
sive; the attempt to link Gen. 50:24 with Exod. 1:6 produces tragicomic results
(Gertz). Undoubtedly, the motif of the burial of Joseph’s bones interlinks Gen.
50:25, Exod. 13:19, Josh. 24:32, with its further relation to the purchase of
ground in Gen. 33:19, but that does not necessarily imply early literary connection
(Blum, in a change of mind from his earlier published position). The continuation
of Exod. 2:23aa in Exod. 4:19 shows that the burning bush episode, Exod. 3:1
4:18, that in any case presupposes Israel’s cry in Exod. 2:23 5, is a post-P insertion
(Schmid); the first and third signs in Exod. 4:1 9 are modelled on P’s plagues
(Blum).
Such views, and many besides, are taken up point by point and met with vigor
ous counterarguments by, especially, the Americans. Levin too diverges from his fel
low Continentals by defending the existence of a coherently edited collection of
non P material to which he attaches the label ‘J’. Essentially, for the Americans at
least, non P is pre P. Dozeman presents a strong argument that Moses’ ‘wilderness’
vocation in Exod. 3:1 4:18 is prior to his ‘Egypt’ vocation in P in Exod. 6:2 7:13
and provides thematic links for a pre P narrative running forward into Numbers.
Defending his view of J as author and historian, Van Seters launches a telling as
sault on the notion of ‘redactor’. In its current form, he traces that notion to
Rendtorff ’s application in redactio historical terms of Noth’s traditio historical
‘block’ model of originally independent Pentateuchal traditions.
This reviewer, an off shore European, will not seek to redress the balance; indeed
would suggest that the points made by the Americans should be taken considerably
further. A striking feature is the generally eirenic tone in which the discussion is
conducted and much credit is due to the organizers of the Seminar for the patient
maintenance of debate where views are so divergent and the stakes are so high. It
may be that there is a fundamental difference in mindset between the two groups.
The commitment of the Continentals is to analytical study: the ambition is to give
as complete an analysis of the literature as possible as a necessary preliminary to his
torical reconstruction. Thus even slight divergence and subtle distinction in the text
provide valuable evidence for the complexity of a process of literary growth that
must not be overlooked but be given due weight. No premature synthesis can do
this complexity justice. As Carr remarks with regard to Schmid and Gertz: ‘the
starting point is that the texts are separate until “proven combined”’ (pp. 162f.).
These Continental contributions are to be respected as challenges to easy synthesis.
But, it seems to this reviewer, the results of analysis often go against the grain of the
biblical narrative. The separation of Exodus 1 and 2, for instance, by K. Schmid
(and others; cited p. 83) leaves the motivation for Moses’ exposure on the Nile un
clear. He surmises shame on his mother’s part at an illegitimate birth (see his earlier
work Erzväter und Exodus, 1999, p. 154). Thus a slur on the Aaronic matriarch is
proposed to promote a theory which is highly likely never to have corresponded to
the way in which she was regarded at any point in the evolution of the text.
An unsatisfactory feature of the Continental consensus evident in this collection
is the term ‘non P’. The material in the Pentateuch that originates in independence
277
REVIEWS
ALICE HUNT, Missing Priests: The Zadokites in Tradition and History (Library of
Hebrew Bible/Old Testament Studies 452). T. & T. Clark, New York/London
2006. Pp. xiv + 218. Price: £65.00 hardback. ISBN: 978 0 567 02852 5.
This book’s central proposition is that the supposed leading role of the Zadokite
priesthood throughout the first millennium BCE in Israel is a scholarly fiction based
278
REVIEWS
on unsubstantiated assumptions that date from pre critical scholarship and were re
inforced in the late nineteenth century by Wellhausen’s synthesis of the history of
the priesthood which adopted the same assumptions. Hunt argues that even though
the ‘sons of Zadok’ were, and still are, assumed to have dominated the Jerusalem
Temple from the time of David down to the Hasmonaean period, there is simply
no evidence to this effect, either in the Old Testament/Apocrypha or in other re
lated literature of the period such as the Letter of Aristeas, Jubilees and Josephus.
The term ‘sons of Zadok’ appears four times in Ezekiel 40 8, and apart from that is
found only in the Dead Sea Scrolls, where it appears in the Community Rule (CD/
1QS), the Rule of the Congregation (1QSa) and the Blessings (1QSb) and in
4QFlorilegium in a midrash on Ezekiel 44. Hence, it cannot be assumed that a
priestly group known as ‘sons of Zadok’ even existed at a period significantly prior
to the date of the Dead Sea Scrolls, let alone that such a group dominated the cleri
cal hierarchy of the First and Second Temples.
The strategy adopted in order to demonstrate this proposition is as follows. The
first chapter is a brief review of modern scholarship together with a statement of
the issues, and declares, ‘The present study makes an assessment of the Zadokites
by locating them historiographically in biblical scholarship, by analysing
historiographic methods used in this scholarship, and then by offering a portrait
of the Zadokites through historiographic, literary, and social scientific analysis’
(p. 12). Chapter 2 then gives a much broader review of scholarship on the
Zadokites and chapter 3 presents a discussion of historiography, ending with a cou
ple of pages on social scientific methodologies, before the textual discussion begins.
Chapters 4 and 5 cover the biblical evidence for Zadokites, chapter 6 covers the
extra biblical evidence (1 and 2 Esdras, Josephus and the Dead Sea Scrolls), chapter
7 offers ten pages on the Zadokites at Qumran, and then finally chapter 8 discusses
Gerhard Lenski’s macro sociological theory and how it might be fitted to the
Hasmonaean kingdom, before concluding on the final page that the ‘sons of Zadok’
were a sectarian group that developed in the Hasmonaean period.
As a piece of scholarship, this is less than impressive. As already stated, the main
thrust of the work is that the supposed centuries long dominance of the Zadokites
as an identifiable group is purely an assumption with no supporting textual evi
dence; and it is true that there is very little direct textual support for such a con
struction. Inasmuch as it challenges scholars to think again about received para
digms instead of simply accepting them unquestioningly, the book serves a useful
function. However, there is little else to commend it. In fact, the overwhelming
impression is that this one basic insight has been padded out to book length pro
portions by the addition of large amounts of extraneous material, which in the final
reckoning add little or nothing to the main case being argued. Chapter 3 on
historiography, for example, reviews the work of Lemche, Thompson, Dever and
others, and attempts to show how their approaches and viewpoints are relevant to a
study of the Zadokites; but the supposed links between these scholars’ ideas and the
Zadokite question are truisms. Consider the following insights gleaned from
Lemche’s work: ‘[B]iblical references to the Zadokites… reflect the socio political
milieu of the time in which the references were written. … The evidence of the
presence and role of the Zadokites depends heavily upon the dating of the
pericopes’ (p. 59). Perhaps I am missing something here, but these seem to me to
be basic tenets of sound biblical scholarship, rather than startling new methodo
logical insights that need pages of introduction and explanation. (But then again,
perhaps that is an unwarranted assumption on my part.) The textual treatments in
279
REVIEWS
the following four chapters are equally disappointing. Calling them ‘analyses’ would
be a misnomer, as more often than not they consist of a quotation of the relevant
passage in Hunt’s own translation a self indulgence that is not justified by any
apparent scholarly or exegetical rationale in the subsequent discussions plus a re
statement of the content of the passage with no additional critique or comment.
Sometimes there is a review of scholarly opinions about a given quotation, but the
amount of actual analysis and original thought about the texts is disappointingly
small.
The most disappointing chapter from this point of view is chapter 5, on the ref
erences in Ezekiel to the sons of Zadok. Given that this is a crucial chapter in the
book, in that it deals with the only occurrences of the phrase ‘sons of Zadok’ that
appear in the Old Testament, I would have expected a rigorous analysis and lively
interaction with existing scholarly work on the topic. Instead, in a twenty page
chapter, what we get is a review of scholarship on Ezekiel 40 8 (6 pages), emphasis
on the crucial nature of the Ezekiel material plus author’s translation of Ezek.
40:45 6, 43:18 19, 44:6 16 and 48:9 11 (2 pages), further reviews of the schol
arship on Ezekiel produced by Blenkinsopp and Cook (four and a half pages), a
suggestion to read Ezekiel 44 in the light of Ezekiel 23 plus another author’s trans
lation of Ezek. 44:6 16 (one and a half pages), author’s translation of Ezek. 23:1
49 plus extensive bibliographical footnote on the whore metaphor in Ezekiel 16
and 23 (3 pages), followed by paraphrasing of the content of the translated chapter
23 (half a page), and finally the suggestion (accompanied by more author’s transla
tions of Ezek. 23:38 9 and 44:6 9) that Ezekiel 44 is a ‘midrash’ on Ezekiel 23 and
that it should be linked with concerns about foreigners serving in the Temple
(1 page). The remaining page or so of the chapter consists of summaries of the case
so far, in which Hunt comments of the Ezekiel material, ‘Whether we place it in
the late sixth century BCE or the Hasmonean period, we must conclude that at the
time of writing there was some discussion of the duties and rights of priests and
that the author of the material considered priests called ‘sons of Zadok’ as central to
service of YHWH’ (p. 142) hardly an earth shattering conclusion. Where are
the historiographical and social scientific methodologies trumpeted so fervently
earlier on in the book? How does all that discussion in chapter 3 feed into the tex
tual treatments in chapters 4 to 7? As far as I can see, not at all; Hunt is doing
nothing more than many historiographically unregenerate but intellectually aware
biblical critics have been doing for decades. True, she has altered her starting (or
concluding?) assumption on the basis of her observations about the textual evi
dence, but the way in which she treats the texts is certainly not revolutionary, not
least because in places she hardly treats them at all. Lenski’s social scientific model
does eventually make an appearance in chapter 8, which is headed ‘Conclusions’
does she not know that conclusions are not the place to introduce new mate
rial? but even here the model is applied to the Hasmonaeans without any men
tion of the Zadokites, and its purported relevance to the Zadokites is only finally
made semi explicit on the final page of the book in a summary of the whole work.
Hunt is also unable to contemplate any other explanation but her own thesis for
the absence of the designation ‘sons of Zadok’ from the textual record that has sur
vived, and strenuously denies the possibility of any group within the priesthood
being recognized even informally as ‘sons of Zadok’, to the point that she is forced
to misrepresent my own published comments on the subject by taking some of
them out of context and giving a false description of the context of others of them.
While this misrepresentation is disappointing, a more serious flaw in her scholar
280
REVIEWS
ship is that she makes no attempt to address the issue of the ideology of Samuel
Kings; such an examination might suggest an alternative reason for the dearth of
references to priestly houses in general and sons of Zadok in particular in the pre
exilic record, which is the most significant absence of evidence for the so called ‘tra
ditional view’. But Hunt is content simply to treat this Deuteronomistic silence as
evidence of absence, and move on to other texts without a backward glance
hardly the act of a responsible historiographer, one would have thought.
My initial assumption about this book was that it originated as a doctoral thesis.
However, this was not stated explicitly in the front matter, and upon encountering
the elementary level flaws discussed above, together with the book’s excessive reli
ance on often quite lengthy quotations from secondary works and a raft of typo
graphical and grammatical errors and obscurities, I found myself questioning this
assumption. But further investigation proved my initial assumption correct (vae
tibi, Vanderbilt!), thereby confirming the principle that absence of evidence should
not necessarily be taken as evidence of absence. If only the same principle could
have been applied to Hunt’s investigation of the sons of Zadok.
doi:10.1093/jss/fgn064 DEBORAH ROOKE
KING’S COLLEGE LONDON
LENA SOFIA TIEMEYER, Priestly Rites and Prophetic Rage: Post Exilic Prophetic Cri
tique of the Priesthood (Forschungen zum Alten Testament 2. Reihe 19). Mohr
Siebeck, Tübingen 2006. Pp. xvii + 318. Price: /59.00. ISBN: 3 16 149059 2.
The relationship between priests and prophets is a perennial source of speculation
for scholars, and although thankfully we have now moved away from the denigra
tion of ritual that once pervaded Old Testament studies, so that prophetic criticism
of cultic activities is no longer automatically assumed to be opposition to the cult
per se, there is still the issue of how the two areas of religious specialism related to
each other, and precisely how any prophetic critique of cultic activities should be
interpreted. These are the questions addressed in the present work. A revision of
Tiemeyer’s 2002 Oxford doctoral thesis, the book offers a close textual reading, in
the historical critical tradition, of four post exilic prophetic writings which
Tiemeyer argues contain criticism of the priesthood: Haggai, Zechariah 1 8,
Malachi and Trito Isaiah. The reading offered of these four central texts is sup
ported by comparison with additional biblical material, both pre and post exilic,
in which the same criticisms of the priesthood appear as in the central four. Follow
ing the mandatory review of scholarship, Tiemeyer begins by defining and defend
ing her view of precisely which parts of the four central texts should be understood
as a critique of the priesthood, before moving on to examine the nature of this cri
tique. First, she interprets Trito Isaiah and Malachi to show that the priests them
selves are conscious of no wrong doing, and then she sets out what she sees as the
content of the critique against the priests: they lack knowledge and have failed to
fulfil their duty to teach the people (Isa. 56:9 12, Mal. 2:1 9); they are encourag
ing social injustice (Isa. 58:3 5; Zech. 5:1 4; Mal. 3:5; cf. Nehemiah 5); they are
carrying out unorthodox sacrificial rites (Isa. 57:6 8; Isa.65:3 4; Isa. 66:3); they
have intermarried with foreign women (Ezra Nehemiah; Mal. 2:10 16); they are
supporting foreign political alliances (Isa. 57:9 10); they are neglecting the
Yahwistic cultic regulations (Mal. 1:6 14; Isa. 61:8); and they are themselves ritu
ally impure and so unable to function to purify the people (Hag. 2:10 14;
Zechariah; Malachi). However, despite this raft of criticisms, argues Tiemeyer, most
of the prophets examined do not urge the destruction of the priesthood, but have a
281
REVIEWS
range of views on what should happen to it. Zechariah and Malachi envisage a
purified and cleansed (Levitical) priesthood that will function as God intended it
to; and although Isa. 66:1 6 predicts the destruction of the priesthood, other
voices within Trito Isaiah advocate changes in its composition: a generalized priest
hood of all Judaeans (Isa. 61:6), a priesthood that can include proselytes as well as
native Jews (Isa. 56:1 8), and a priesthood that includes Gentiles (Isa. 66:18 24).
As a revision of a doctoral thesis, the book bears the hallmarks of the thesis genre
in the rather self conscious reviewing of scholarly opinions at every step and the
minute setting out of each stage in the argument. There is a good deal of close tex
tual discussion, which does not always make for easy reading, and an almost text
book like breaking down of the subject matter into bite sized paragraphs demar
cated by headings and sub (sub sub )headings, with surveys of arguments before
they commence and summaries of them once they have been presented. While the
evident desire to present the material in a clearly organized fashion is commend
able, and the format does allow the reader to home in on precisely the paragraph
that may be of interest for whatever reason, it does not really encourage one into
the sweep of the argument, and has a rather laboured feel to it. In particular, the
introductory review of scholarship on Isaiah 56 66 I found somewhat abstruse,
owing in part to the fact that the way in which the headings are used gives an erro
neous impression of the direction of the argument; and the highly detailed discus
sion in chapter 2 about which parts of Trito Isaiah address the priests is broken
down using as many as five levels of headings in places, an unnecessary extra fussi
ness which serves to atomize, rather than to clarify, the flow of thought. Easier
reading are the parts of the book where Tiemeyer is drawing conclusions and dis
cussing the implications of the detailed argumentation here there is a good sense
of what is driving the thesis and how the exegetical work feeds into the overall pic
ture. Indeed, where the work is able to keep sight of the provisional nature of the
conclusions that can be drawn as conclusions about the text rather than necessarily
about the reality behind the text, it is careful and solidly based, and the result is
generally convincing.
My main reservation about this project is, however, its use of prophetic critique
to establish ‘facts’ about the priesthood. It is not that Tiemeyer is unaware of the
difficulties involved in her chosen approach; indeed, she addresses them overtly in
the Introduction, where she says, ‘[W]hile the result of our enquiry will tell us a lot
about how the priests were perceived by the prophetic writers, we shall learn signifi
cantly less about the actual behaviour of the priests in this era’ (p. 2). She even re
peats this sentiment at the end of chapter 4, where, having argued that despite the
prophetic critique the priests believed themselves to be in the right, she moves on
to explore precisely what the prophetic accusations consisted of, with the proviso
that we must keep in mind that accusations are not the same as unbiased accounts
of historical events (p. 112). Yet in the conclusions to the following chapters that
deal with particular aspects of the prophetic critique, there is often a tacit assump
tion that the priests actually are in the wrong and the prophets are right, an as
sumption that comes through most clearly in the concluding remarks for the vol
ume as a whole. In the course of the book, Tiemeyer has compared the post exilic
critique of the priests with similar critiques found in the pre exilic prophets, for
example, in Amos and Ezekiel, and in the concluding remarks she comments,
‘Rather than attributing the post exilic priests with orthodoxy, reformed through
their suffering following the destruction of Jerusalem and/or through their exilic
experience, I have endeavoured to show that the emerging new clergy were not so
282
REVIEWS
very different from their pre exilic predecessors’ (p. 287). There seems to be here
the implied conception of a kind of pristine agreed orthodoxy from which pre
exilic clergy deviated, to which the exile was intended to recall them, and according
to which they are now legitimately being judged and condemned by the prophets.
However, maybe what is evidenced by the continuation of the prophetic critique
from pre exilic to post exilic times is that orthodox Yahwism/Judaism as tradition
ally understood took much longer to prevail than is generally assumed, and that
during the sixth and fifth centuries BCE it was still being negotiated between oppos
ing groups with different conceptions of what was legitimate. Certainly, to that ex
tent Tiemeyer is correct to say that the post exilic priests seem not to be so very
different from their pre exilic predecessors, but in assessing the implications of that
position righteousness should perhaps not quite so readily be assumed to lie with
the prophets. On the other hand, the ease with which such an assumption is so of
ten made in all sorts of scholarly contexts, not just here certainly demon
strates the potency of the prophetic rhetoric, which was arguably one of the factors
in creating the concept of ‘orthodoxy’ as we know it (or assume it).
A final niggle about the production of the volume: as well as a bibliography, it
includes three indexes, Source Index, Subject Index and Authors Index. However,
none of the three indexes covers material in the footnotes, which given that foot
notes are a major location of author and text citations seems odd to say the least,
and limits the usefulness of the indexes. In this computerized age of book produc
tion it should surely be possible to create indexes that cover the whole text of the
book including the footnotes.
doi:10.1093/jss/fgn065 DEBORAH ROOKE
KING’S COLLEGE LONDON
283
REVIEWS
given poetic line and other poetic influences, such as chiasm. To do so, he builds
especially on an information structure model proposed by Knud Lambrecht (Infor
mation Structure and Sentence Form: Topic, Focus, and the Mental Representation of
Discourse Referents, Cambridge 1994), which he summarizes and applies to BH
prose in chapter 3 before applying it to biblical poetry in the following chapters.
I find most of Lunn’s explanations in chapters 4 9 to be plausible and many to
be likely. For example, on p. 123 he explains the structure of the three lines in Ps.
138:6 ki ram yhwh // wesapal yir’eh // wegaboah mimmerÌaq yeyeda¨ ‘Because Yhwh
is exalted // yet he sees the lowly // but the haughty he knows at a distance’ (my
translation). The contrast between the first two lines is apparent (though God is
exalted he still notices the humble), but it is the third line that illustrates the essen
tial correctness of Lunn’s distinction between syntactic constraints and techniques
of poetic defamiliarization. He convincingly argues that the O PP V order of the
third line is not due to defamiliarization but rather has a syntactic pragmatic expla
nation; specifically, the initial object NP gaboah is a contrastive topic, the fronted
PP mimmerÌaq carries focus, and the final V yeyeda¨ is presuppositional, since it is
broadly related to the preceding verb yir’eh. The result of such pragmatically moti
vated reordering of the syntax is that, ironically, while God from his high position
takes notice of the ‘low’, he maintains his distance from the ‘high’, or as Lunn ex
plains, ‘God does know these people, just as he knows all human beings, but these
haughty ones he knows from afar’ (p. 123; italics are Lunn’s).
That said, I question Lunn’s wholesale acceptance of Lambrecht’s model of infor
mation structure; Lambrecht’s theory, while admirable in its scope and refinement,
has serious problems, especially relating to the definition of focus. For instance,
Lunn (following the work of his thesis advisor, Heimerdinger) categorizes the S V
clause hannaÌas hissî’anî ‘the serpent deceived me’ (Gen. 3:13, my translation) as an
‘event reporting’ clause with ‘sentence focus’ and asserts that this ‘is not a comment
about [Eve’s] own action (What did you do?), but a clause in which both subject and
predicate share equal focus (What happened?)’ (p. 80). I have two issues with this
analysis. First, if BH is a V S language (as Lunn asserts on p. 4), why are event
reporting clauses not V S? Lunn never addresses this oddity that basic word or
der does not fit into the very category of ‘focus structure’ (sentence focus) that lacks
any ‘pragmatic presuppositions’. Into which of Lambrecht’s three focus structures,
then, does a basic V S clause fit? Second, it is highly counterintuitive that all mean
ingful propositions fit into one of Lambrecht’s three focus structures, or to put it
another way, that there are no ‘sans focus’ meaningful propositions (note that
Lambrecht explicitly distinguishes focus from ‘new’ information). It is a question
able model that does not allow for the existence of a basic word order, pragmatically
neutral clause (which are admittedly rare but nonetheless possible). Lunn notes that
parallel S V clauses, such as pî Òaddîq yehgeh Ìokma // ûlesônô tedabber mispa† ‘the
mouth of the righteous utters wisdom // and his tongue speaks justice’ in Ps. 37:30
(my translation), are common in Proverbs, likely due to the lack of significant prag
matic context (p. 132; also p. 78). Why these clauses should be marked for focus is
unclear (contrast Lunn’s analysis of such clauses with the analysis I provide in my
‘Word Order in the Book of Proverbs’ [Pp. 135 54 in L. Troxel, D.R. Magary and
K.G. Friebel (eds), Seeking out the Wisdom of the Ancients: Essays Offered to Honor
Michael V. Fox on the Occasion of His Sixty Fifth Birthday. [Winona Lake, Ind.
2005]).
Lunn’s work is indeed valuable, if for no other reason than its defence of the es
sential syntactic sameness of BH prose and poetry, which I take to be its primary
284
REVIEWS
contribution. The book should be read and recognized for taking the critical next
step in our study of ancient Hebrew grammar applying advances in BH prose
syntax to BH poetry and in the process to distinguish pragmatic influences on word
order from poetic influences. Unfortunately, I cannot agree with Heimerdinger’s ef
fusive recommendation of the book on the front cover, back cover, and in the fore
word: the devotion to Lambrecht’s information structure model, which Lunn care
fully and skillfully applies to the BH poetic data, is also the primary weakness of his
analysis. I therefore suggest that it is time for those interested in pragmatics, infor
mation structure, and BH word order to set Lambrecht’s book aside and look for
other, more insightful models within general linguistics. It may very well be the
other lasting contribution of Lunn’s study that we in BH linguistics relegate
Lambrecht’s book to the status of supporting material rather than the central theo
retical role it plays in the BH word order and information structure monographs by
Heimerdinger, Shimasaki and Lunn.
doi:10.1093/jss/fgn066 ROBERT D. HOLMSTEDT
UNIVERSITY OF TORONTO
285
REVIEWS
using Hebrew (comparable to a present bilingual Indian whose use of English for
communication does not make him English in terms of nationalistic feelings).
Interestingly, G. pays no attention at all to the Greek and later Latin speaking
Jewish Diaspora (which was not bilingual). Does he assume that all of these Greek
speaking Jews were not ‘nationalistic’? If he does, he is right; this latter view would
support the thesis of A. Edrei and myself (published recently in the Journal for the
Study of Pseudepigrapha 16.2 [2007], 91 137 and 17.3 [2008], 163 87) claiming
that the Greek speaking Jewish Diaspora lost touch with the Eastern Hebrew
speaking Judaism and its rabbis as a result of the difference in language (e.g. that
the Western Jews did not know Hebrew or Aramaic, and the latter did not care to
translate their corpus of Halakhic literature into Greek or Latin). Thus, the
disconnectedness with the eastern (nationalistic) Jews was even stronger due to the
‘lack’ of nationalistic ideas in the Western Jewish Diaspora. Later in the history of
Europe the use of vernacular turned out to be a crucial element in the construction
of nationalistic ideas. This process can be observed in the actions of Jan Hus (and
later Martin Luther). These and other reformers translated Holy Scripture as well as
significant theological books into the vernacular. Such translations were an impor
tant factor in arousing national feelings. Bohemian Czechs during the fourteenth
and fifteenth centuries crystallized their national feelings against the background of
the German segment there. Much of this creation of Czech nationalism was based
on texts in the vernacular circulating already in the fourteenth century, a process
culminating in the foundation of the Bethlehem chapel in Prague in 1394 and the
vernacular preaching for which it was founded. Jan Hus became rector of the Beth
lehem chapel in Prague in addition to being a professor at the university. He was
conscious of the national and linguistic differences between the Czech and German
populations. Thus all his life he associated this nationalistic aspect with his new re
ligious reform (M. Lambert, Medieval Heresy3. Oxford 2002, chapters 15 16). It
would have been useful for G. to present some analogies to this use of language,
taking us beyond the Jews in the Second Temple period.
3. G.’s enormous efforts to distinguish between the three different ‘titles’ of na
tionalism, Israel, Judah and Zion are very interesting indeed (chapters 5 7). How
ever, if his description of the three types (and therefore, a distinction between dif
ferent stages of nationalism in the history of the Second Temple period) is true,
then we hardly hear what the differences between these ‘types’ were in the context
of the nationalistic idea, and what expression they had on the ground (namely
among the Jewish people). If they are not after all different in their content (ideo
logically and practically), and are just synonyms of the same thing, then the histori
cal evaluation of nationalism in ancient Israel has gained nothing at all, and the
chapters remain but a discussion, however important, of the nomenclature.
4. The discussion concerning the antiquity of much of the biblical corpus (in
chapter 5) is in itself important and should be addressed even more frequently by
Hellenistic Judaists. In developing this argument it would have been helpful to in
clude an in depth discussion of the Pseudepigrapha and the remnants of Jewish
Hellenistic literature. In this literature, which quite often rewrites the Bible, we can
actually see very clearly how the Jews lived with their Bible in daily life. An elabora
tion on some of this literature would have supported G.’s thesis about the Bible as a
canonic text, which helped build the Jewish nation (Eupolemus uses only ‘Judea’
for describing the Jews). For G.’s case it would not matter whether Jubilees was
written in the early Maccabean period (VanderKam) or later in the century
(Mendels), but it shows how ‘Israel’ used its heritage to endow the new created na
286
REVIEWS
tion in the Hasmonean period with legitimacy. G. would have profited from a dis
cussion of this particular work (By the way, both ‘Zion’ and ‘Israel’ are used in the
Book of Jubilees. For instance, 1:28; 2:31).
In conclusion, G’s book has strengthened in many of its points what has been
said about Jewish nationalism in recent years; the issue of its ‘fall’ (chapter 8)
should not bother one too much since, as G. rightly argues, there was a continuity
of the concept after the destruction of the Temple by what I have called ‘passive’
nationalism. Here one should add a general comment. It seems to me that at
present scholars in the humanities should avail themselves to a much greater extent
of the research carried out by social scientists. For instance, it would have been use
ful in this particular book to deal extensively (and not just in passing, as on p. 115)
with the problem of memory in its public and social context, in line with recent
approaches of social scientists (see for a very recent overview: D. Mendels [ed.], On
Memory. An Interdisciplinary Approach, Oxford and Vienna 2007), or to tackle in
depth some of the (wrong) ideas of Benedict Anderson (in his Imagined Communi
ties. London and New York 1991; see for example the sharp criticism of A. Hast
ings, The Construction of Nationhood. Cambridge 1997, p. 194).
doi:10.1093/jss/fgn067 DORON MENDELS
THE HEBREW UNIVERSITY OF JERUSALEM
287
REVIEWS
first volume, most of my comments will be directed toward that portion of this
book.
The book manifests a number of methodological problems. Many have to do
with Mortensen’s concrete engagements with her sample PJu passages. Yet one of
the more significant problems becomes evident on the first page of this book. As
was implied by the introductory paragraph above, one of the main distinctions in
this book is between Targum Pseudo Jonathan as a whole and the Pseudo Jonathan
unique material or PJu. As Mortensen states in her first sentence, ‘Targum Pseudo
Jonathan projects a single overriding purpose: it speaks to priests’ (p. 1). A few sen
tences later she writes, ‘This [priestly] agenda only stands out when the unique
Pseudo Jonathan material is separated from the rest’ (p. 1). By distinguishing and
separating the material she will later call PJu from Pseudo Jonathan as a whole,
Mortensen hopes to produce ‘a clear view of the obviously focused agenda of the
targumic writer’ (p. 1). Although this move may initially seem straightforward,
upon inspection it can be seen to complicate and confuse Mortensen’s analysis of
both Pseudo Jonathan and PJu. On the one hand she attempts to argue that
Pseudo Jonathan as a whole has a priest centered agenda. Yet on the other hand she
states that this agenda only becomes evident when the PJu material is examined in
isolation from the other Pseudo Jonathan material. In making this move,
Mortensen undermines the force of her argument before she even begins making it.
In effect she admits that when taken as an integral whole, Pseudo Jonathan oc
cludes the priest centered agenda that is manifest in PJu.
Mortensen does not address this particular methodological problem, and gener
ally seems to give methodology very little discussion. She does explain the process
by which she isolated the PJu material, and also gives full chapter treatment to
‘Pseudo Jonathan’s Unique Character’ (Chapter 1). However, she does not explain
how the priest centred agenda of PJu expands to include the entirety of Pseudo
Jonathan. In fact in her concrete analyses of specific passages she often conflates PJu
with Pseudo Jonathan as a whole as if the two have a single identity. In the con
cluding chapter of volume one, ‘Implications of Unique Pseudo Jonathan’s Priest
Centered Agenda’, she partially explains this conflation when she discusses the
composition of Pseudo Jonathan. She states that in Pseudo Jonathan ‘scholar
Cohens undertake to resurrect the profession of the priesthood in a fresh way. They
translate the Pentateuch. They underline all relevant passages, repeating them where
possible, and, where they find fertile ground, they insert current ideas and their
own insights. As with all targums, they make no indication of interpolations or
sources, allowing the entire targum to serve as the “translated word of God”. They
appoint a final editor who fashions the document into the smooth text we have to
day hence the single ‘targumist’ referred to above’ (p. 448). Although this hypoth
esis of a single targumist has the advantage of explaining how Pseudo Jonathan
might have had a priestly origin, it by no means indicates how a priest centred
agenda was made manifest throughout Pseudo Jonathan as a whole. This is espe
cially the case when considering the fact that much of Pseudo Jonathan, including
much PJu material, has parallels in rabbinic literature. Mortensen mentions some
of these parallels in several footnotes in her chapter on Esoteric Knowledge (cf. e.g.
p. 401, n. 8), but nowhere does she discuss these parallels at length. The most ex
tended discussion of PJu’s parallels appears in a footnote on page 435 where she
states in three sentences, ‘While many expansions unique to Pseudo Jonathan ap
pear in other rabbinic texts as well, the genius of the compiler lies in the selection
and arrangement of this material. Since the parallels occur in a wide variety of
288
REVIEWS
documents a simple transfer of ideas does not explain PJu. The independent agenda
of the targumist makes PJu a text deserving of attention as a separate phenomenon’
(p. 435, n. 1). In other words Pseudo Jonathan’s targumist has redacted these ex
pansions to fit with his priest centered agenda.
Mortensen’s statements concerning the redaction of the PJu material provide
some explanation for what could be perceived as a significant problem with her
study. Yet they do not account for the most problematic aspect of her book. In
short, much of Mortensen’s evidence does not seem to confirm the hypothesis of a
priestly origin and priestly audience for Pseudo Jonathan, or even PJu. In fact many
of the passages Mortensen examines only appear to be priestly because Mortensen
says they are so. In many of Mortensen’s examples there are no textual markers to
indicate that the passages in question are directed toward priests or someone of a
priestly family. Rather a priestly origin for such passages seems to be assumed sim
ply because they are found in PJu. Some passages studied by Mortensen do indeed
appear to have a priest centred message (cf. e.g. many of the passages in her ‘Tem
ple Job’ chapter). However these passages are the minority of cases examined by this
book. Mortensen could have provided a partial corrective for this problem by con
ducting a survey of priestly beliefs and practices in post temple Judaism, thus dem
onstrating the breadth of priestly interest and engagement. Yet given that informa
tion concerning such beliefs and practices is scarce, if not non existent, Mortensen
is placed in the difficult position of arguing that PJu and Pseudo Jonathan are
priestly documents with no evidence to corroborate her claims other than that
which is found in these targums. As a result, it is difficult to determine how
Mortensen can demonstrate her hypothesis with any great level of certainty.
Despite the flaws with this work, Mortensen has provided scholars with an inter
esting hypothesis that deserves further exploration in another context, using a differ
ent methodology. Not only does it articulate a new theory concerning the origins of
Targum Pseudo Jonathan, but it also raises the methodological question of how to
examine source material in conjunction with a completed document. For these rea
sons this volume will be of interest to targum scholars who desire to enter into the
discussion concerning the problem of Targum Pseudo Jonathan’s origins, as well as
the discussion of what methodology is the most appropriate to examine this problem.
doi:10.1093/jss/fgn068 SIMON ADNAMS LASAIR
EDMONTON
289
REVIEWS
Samely first describes the smallest self contained literary forms and the diverse
ways how statements are linked to each other by mere juxtaposition, series of say
ings varying a common theme, or clusters with an unclear thematic continuity.
Since criteria of coherence are infinitely flexible, the mutual relationship between
statements mostly remains unclear; the statements are open to connections of all
sorts. Syntheses of rabbinic thought on whatever topic thus have no direct basis in
the primary evidence. Although there are features of unity across rabbinic literature
(parallels, names of rabbis, biblical quotations), to claim its unity is a hermeneutic
and not necessary assumption, especially in the extreme form of an encompassing
Oral Torah.
In order to understand a text and its single terms, we have to decide whether the
text presents an open or a closed set of statements, if it is exhaustive or offers only
examples. Can we take document boundaries as evidence for a unified rabbinic
message, as Jacob Neusner proposes, or must we assume open clusters so that in the
extreme everything in rabbinic literature is connected with everything else? This is a
basic hermeneutic problem when reading rabbinic texts. Many rabbinic statements
have multiple thematic links and may change their primary focus with the chang
ing environment; it is up to the reader how far (s)he can go in making connections.
As Samely rightly emphasizes, the ambiguity of rabbinic textuality is frequently un
derestimated because the interpretation of the Bavli or Rashi etc. is taken for
granted (p. 62).
The following chapters on the use of the Bible in rabbinic literature mainly sum
marize Samely’s Rabbinic Interpretation, expanded to cover all of rabbinic literature.
In a very dense analysis of the surface of the Mishnaic text, Samely then discusses
the constant literary device of quoting rabbis. These speech reports and small narra
tives always refer to a reality outside the text, offer some kind of story including
disagreements among the rabbis. The amoraim assume that the anonymous voice
or stam represents the unanimous or majority opinion since the presenter sides with
it and only rarely explicitly takes sides. The dispute format where a single rabbi
counters the stam, signals that it ‘must have been public before the text was created’
(p. 106). The stam seems to have a constant identity throughout the text; thus ‘that
voice claims implicitly an authority greater than that of any single rabbi. As a result,
the stam suggests itself as the voice of rabbinic knowledge’ (p. 107). When the
anonymous voice of later documents binds together parts of the same and of differ
ent documents, their united stams give the impression of a coherent halakhah. Rab
binic texts have been arranged and composed into documents, but their composi
tion ‘is de emphasized on every page by the very literary structures which constitute
it’; the reader thus ‘becomes seduced into thinking that the documents are really
not constituted as fixed texts at all’ (p. 111).
Quoting rabbis depicts them as speaking and turns them all into contemporaries
of the reporting voice and of each other. This becomes more complex in later texts
which require an oral transmission before a saying reaches the reporting voice. The
literary form thus emphasizes the oral mediation of rabbinic knowledge. Oral reci
tation provides its own hermeneutic context, a certain place and time. It also de
mands dividing large texts for oral use on communal occasions, the size depending
on the category of text, longer for narrative texts, smaller for halakhic discussions.
The oral setting also influences the ordering of mishnaic sentences, complementing
pre existing clusters or series and commenting on them. A cluster like m. Ber 2 3
looks like a first draft, fitting a conversation with several speakers: ‘No one’s person
constant outlook, experiences, or preferences would act as a filter for what is re
290
REVIEWS
corded’ (p. 132). Why has this been maintained in the redaction of the tractates?
One might point to the inertia of existing structures in absence of an urgent incen
tive to change them. But this certainly is only a possible hypothesis, as Samely con
cedes.
Much of the Mishnah and other halakhic texts are characterized by a large scale
use of the hypothetical legal case. The reader assumes the individual halakhic cases
to be compatible with each other; it is up to him to decide how they are linked to
each other. Verbal articulations limit the meaning of existing practices and replace
the former openness inherent in the practice; every new case developed on the basis
of an earlier one also limits the meaning of the former case. But the texts ‘do not
give descriptions of Jewish life, but prescriptions’ (p. 152); Jewish practice at any
given time or place cannot be deduced from them.
A chapter is devoted to the Talmud as conversation and repository. Principles of
text arrangements tend to become weaker towards the end of a chapter: ‘This sug
gests that the Gemara to some extent merely stores, rather than organizes, the infor
mation it contains’ (p. 158). It is problematical to assume the unity of the docu
ment ‘Babylonian Talmud’; its unity is less in the text than in the rabbinic
tradition. Of particular interest is the description how quotations other than bibli
cal or Mishnaic passages are usually presented as if they came from a pool of tradi
tion without textual frame, thus deliberately distinguishing between the Mishnah
and other material used in the Gemara. Using baraitot as commentary on the
Mishnah, the Talmud suggests a unity of tradition. ‘The creators of the Gemara
may have (in my view, must have) obliterated whatever textual constitution their
own sources had in order to reconstitute them as a Mishnah commentary made up
from quotations… a literary conceit of the highest order’ (pp. 170f). In the
Gemara the Mishnah assumes a much greater complexity than in the original. Its
gaps are exploited and filled in to such an extent that the reader, once used to the
Gemara, sees the Mishnah no longer on its own, but as a pointer to a much fuller
text, as a summary of its commentary; we thus witness a ‘reversal of roles between
text and commentary’ (p. 173).
A last chapter deals with ‘Hermeneutic Models of Story and History’. Here
Samely describes the relationship between Midrash and biblical history and the way
later events are understood as repetitions of biblical paradigms. The present is seen
through biblical eyes. Biblical events are constructed as if they were first hand per
sonal history; coherent accounts of the contemporary world thus become unneces
sary.
Every chapter is followed by a list of select further reading. Briefly annotated
sample texts of rabbinic literature illustrate what the main text describes and
analyses. A glossary, bibliography and indices conclude this concise and very dense
description of literary forms of rabbinic literature and their deeper meaning.
The book is highly innovative in its description of how rabbinic texts function as
texts and constitute meaningful wholes. It describes them in a way which for the
most part is accessible without knowing the texts or their specific problems, thus
transferable to other sets of texts as well. As such it works very well for the outsider
and the beginner. But it is even more useful for somebody who already has experi
enced the frustration when trying to find connections in rabbinic texts and to un
derstand how a text is structured. To such a reader Samely’s description not only
explains the causes of one’s frustration but also helps to find meaning in such texts
with so many open ends. The relationship between text and tradition which alone
is the real context of the single works, is very well described, the much discussed
291
REVIEWS
question of the redaction of rabbinic texts receives much illumination. This impor
tant book will be of great help for beginners in rabbinic studies. But it is even more
recommended to everybody who already has some experience with rabbinic texts or
who has even studied them for a long time. People dealing with similar texts from
other cultures will also profit from its insights.
doi:10.1093/jss/fgn069 GÜNTER STEMBERGER
UNIVERSITY OF VIENNA
EMANUELA TREVISAN SEMI, Jacques Faitlovitch and the Jews of Ethiopia. Vallentine
Michael, Edgware 2007. Pp. xvii + 204. Price: £19.99. ISBN: 978 0 85303
655 5.
Despite the existence of a vast and ever growing literature on the Beta Israel
(Falasha) of Ethiopia, there are still significant gaps in the scholarly literature about
this group. None of these has been as significant as the lack of a proper scholarly
biography of Jacques (Ya’acov) Faitlovitch. Faitlovitch, more than any other figure
in the twentieth century, studied their traditions, shaped their image and promoted
their cause. In many ways, the aliyah of over 75,000 ‘Ethiopian Jews’ and their
descendants and relatives are the fulfillment of Faitlovitch’s dreams.
It is hard to think of any scholar better qualified than Emanuela Trevisan Semi of
the University of Venice to write about Faitlovitch. She has already written on Ital
ian fascist anthropology about the ‘Falasha’, edited a book of Taamrat Emmanuel’s
letters, and co edited several volumes of conference papers from the Society for the
Study of Ethiopian Jewry. She has a broad knowledge of the sources, as well as a
command of numerous relevant languages including Italian, Hebrew, French and
German.
The volume she has prepared follows Faitlovitch from his early years as a student
of the great Orientalist and fellow Falasha phile, Joseph Halévy, through his nu
merous journeys to Ethiopia. It chronicles his many years of public activity on be
half of ‘his people’. It brings together the relevant published sources as well as archi
val sources and newspaper and journals. Moreover, it departs sharply from the
romanticized vision of Faitlovitch as the ‘Father of the Falasha’ and offers a nuanced
portrait of this scholar, activist, educator, reformer, missionary and fundraiser. To
her credit, Trevisan Semi does not attempt to resolve the contradictions in
Faitlovitch’s character, but allows him to remain as both a romantic and an oppor
tunist, a champion of the ‘Falasha’ and one of their harshest critics.
To be sure even a scholar as well prepared as Trevisan Semi was presented with a
number of challenges in preparing this volume. Given Faitlovitch’s many roles and
the breadth of activities, a simple chronological narrative would not do justice to
the many larger issues which arose during half a century of work. Most of the four
teen chapters in the book explore a particular period, but go further by examining a
particular theme in Faitlovitch’s life.
Thus, throughout the book, but particularly in Chapter 6, Trevisan Semi stresses
the anti missionary focus of much of Faitlovitch’s activity. Indeed, the missionaries
were probably the first to treat the ‘Falasha’ as Jews in the broader sense of the term.
However, her reference to the ‘Church Mission among the Jews’ is confusing. The
relevant mission society was historically known as the ‘London Society for the Pro
motion of Christianity amongst the Jews’, it later became the Church Ministry
among the Jewish People. It was not the Church Missionary Society (pp. 162, 181),
which was founded a decade earlier, although representatives of the CMS were
present in Ethiopia in the early nineteenth century.
292
REVIEWS
293
REVIEWS
ver, numerous inconsistencies in the volume, which an editor should have caught.
Mesguid (p. 17), but mesgid (p. 60), Aescoly, (p. 52), but Aeskoly, (pp. 3, 4)
Halévy (passim) but on the back cover of the book Halévi. In the glossary tevet is
not capitalized as are other Hebrew months and Ger/gerim are ‘convert/s’ not ‘con
vert’s’. H.J. Polozky appears, (p. 3) but as readers of this journal know, the name
almost always appears as Polotsky. Martin Wurmbrand is mentioned (pp. 3, 6, 49),
but he is usually referred to as either Max or Mordechai in his publications. There
is, moreover, no consistency in the presentation of personal names. In some cases
full names are provided, in others family names with initials; in still others just last
names. Thus in the first paragraph of the first page we have Nahum Slouschz and
Joseph Halévy, but M. Nurock and Rabbi (no first name) Berman.
If the overall tone of this review appears somewhat critical, it is not because I
doubt the tremendous value of Trevisan Semi’s work; quite the opposite. Tudor
Parfitt is certainly correct when he writes in the forward (xiv) that this work ‘is
likely to remain the definitive work [on Faitlovitch] for some time to come.’ Every
serious scholar of the Beta Israel will read and cite this work. Accordingly, it must
be held to a very high standard.
doi:10.1093/jss/fgn070 STEVEN KAPLAN
HEBREW UNIVERSITY OF JERUSALEM
294
REVIEWS
dell’area sud semitca’, which is led by Paolo Marrassini. In a short preface Professor
Marrassini discusses the significance of the volume, highlighting the value of pub
lishing the Ethiopic version with the Arabic and the importance of the fact that the
editions of the Arabic and Ethiopic very properly, as he argues follow stand
ard procedures for the production of critical editions of texts and do not, as is often
the case in the editing of oriental texts, rely on the concept of a ‘base manuscript’.
The introduction is the work of Alessandro Bausi and provides, after a short ac
count of the events, a masterly survey of all the sources relating to the events at
Nagran and of the research that has been undertaken on them. There then follow
editions and translations of the two texts.
The first Arabic recension of the Martyrdom, although known and discussed, had
not previously been edited. The edition and translation by Alessandro Gori (pp. 19
89) is based on four manuscripts: BL Or. 5019 (eleventh cent.); Sin. Ar. 469 (thir
teenth cent.), which is damaged and contains only a fifth of the text; Birmingham,
Selly Oak Colleges, Mingana Chr. Ar. Add. 236 (thirteenth cent.), a fragment of
two leaves that forms the immediate continuation of the text in Sin. Ar. 469; and
Sin. Ar. 535 (thirteenth cent.), a copy of BL Or. 5019. The apparatus is divided
into two registers (in the upper register a list of the witnesses, in the lower the vari
ant readings), but there are only a few brief notes attached to the translation.
An edition and Portuguese translation of the Ethiopic version of the Martydom
of Areta was produced by Pereira in 1899, but the edition and translation by
Alessandro Bausi published here (pp. 91 306) places study of this text on a totally
different level. The ‘History of the men of Nagran and martyrdom of Saint Îirut
and his companions’ forms one of the writings included in the traditional Ethiopic
collection Gadla sama¨tat (‘Contendings of the Martyrs’). The Ethiopic text of the
Martyrdom is clearly based on an Arabic Vorlage, more precisely on an Arabic ver
sion that was related to, but certainly not identical with, the first Arabic recension.
Bausi’s edition is based on eleven manuscripts, which range in date from the four
teenth to the eighteenth century, and Bausi argues that the text that we have of the
Martyrdom is not too far removed from the original Ethiopic composition. The
apparatus to the text is divided into four registers: first, a list of the witnesses;
second, extracts from other Ethiopic writings that refer to Areta and the events
at Nagran, particularly extracts from the reading in the Synaxarium for 26 Îedar,
the day on which the martyrdom of Areta and the men of Nagran was com
memorated; third, the variant readings; and finally, an apparatus of the punctua
tion in the MSS. This last item is not normally included in editions of Ethiopic
texts, but can be important for the assessment of the MSS and the interpretations
they offer.
Bausi has attached to the translation some fairly extensive notes on the philologi
cal, historical and literary problems raised by the text. He makes full reference to
the other versions of the Martyrdom and to the extensive secondary literature, and
for the Greek version of the Martyrdom (BHG 166) he was able to make use of the
valuable new edition of the text prepared by Marina Detoraki in her Sorbonne doc
toral dissertation of 2000.
The number and variety of the sources that bear on the cluster of events con
nected with the martyrdoms at Nagran raise a whole host of problems of interpreta
tion that have attracted the attention of scholars for the last century and a half. The
present work quite deliberately does not attempt to offer new solutions to all these
problems, but, much more modestly, only to make available critical editions of two
witnesses of the oriental tradition of the Martyrdom. In fulfilling this aim the two
295
REVIEWS
authors have set high standards and provided a very valuable basis for further study
of these texts and of the issues raised by the events at Nagran.
doi:10.1093/jss/fgn071 MICHAEL A. KNIBB
KING’S COLLEGE LONDON
296
REVIEWS
mally on the basis either that the work is not actually by Theodore (e.g. the Summa
Theologiae Arabica), or that a good English translation has recently appeared else
where (e.g. Sidney Griffith’s translation of his work defending icons), or that the
work is so technical in its language that it is likely to be of interest only to real spe
cialists (e.g. the Letter to David the Monophysite [in Arabic]) and his discussion (in
Greek) of philosophical names.
What remains, therefore, is nineteen works, five (in Part I) originally in Arabic,
grouped under the title ‘Discerning the True Religion’, directed particularly against
Judaism but also including reference to Magians (Zoroastrians), Samaritans,
Manicheans, Marcionites, Muslims and others; six (in Part II), some originally in
Arabic and some in Greek, on ‘Discerning the True Church’, mainly directed
against the Armenians but also including discussion of the views of the Nestorians
and the Jacobites, the main alternative Christian groupings in Haran; five (in Part
III), all originally in Arabic and intended mainly for an internal (i.e. Christian)
readership while also beginning to reflect the beginnings of inter religious discus
sion of some of the themes, on ‘Topics in Controversial Theology’, that is, episte
mology, natural theology, the Trinity, and two texts on free will; and finally two,
both in Greek, linked under the title of ‘The Byzantine Legacy’, which include
guidance for Christians about how to relate to other religious communities, the
‘Refutation of the Saracens’ (as reported by John the Deacon) and the Greek Frag
ments (which, despite its title, is, the editor suggests, an integral composition prob
ably compiled within a century of Theodore’s death). The translator’s introduction
consists of a helpful six page outline of what is known (which is not a lot) of
Theodore’s biography, an eight page summary of the main arguments found in the
works, and ten pages giving fuller details of the manuscripts, editions, and other
translations of the works included in the volume. There is a full bibliography, index
of scriptural (including twenty Quranic) citations and index. The cover is beauti
fully illustrated with a picture of the four evangelists from a Syriac lectionary from
Mosul produced in 1499 CE, a timely reminder of the contemporary relevance of
the book in the light of the death, following kidnap, of Paulos Faraj Rahbo, the
Chaldean (Catholic) Archbishop of that city, in March 2008.
Specialists in Christian Muslim relations will look particularly to Chapter 17,
the short text where Theodore responds to questions from a Muslim enquirer about
the Christian view of free will (pp. 207 8), a theme which is also discussed on
pp. 196 8, Chapter 18, the Refutation of the Saracens, in which Theodore dis
cusses both Christian beliefs (about God, Christ and Muhammad) and practices
(such as the Eucharist and monogamy), and also pp. 238 41, where Theodore an
swers the questions in quick succession of the Arabs, of an unbeliever, and of an
Agarene (i.e. Muslim). There is much else of interest, however, not least as evidence
of the thoroughly religiously plural context of Haran around the end of the eighth
century, and it is also very interesting to see which of Theodore’s works were writ
ten in each language, and therefore primarily for whom, Greek for a primarily
Christian audience both inside and beyond the world of Islam, and then Arabic for
a more mixed, even a multi religious, audience within the world of Islam.
I have one technical quibble, namely that the arrangement of the footnotes is not
quite as helpful as it might be, since although they are easily accessible at the foot of
each page, they are run together, rather than each note starting on a new line of its
own, and this means that an individual note is not quite as easy to locate as it
might be. But this is a small quibble when put alongside the translator’s achieve
ment in making available for the first time in book form in English the collected
297
REVIEWS
works of an author who, in the translator’s words, was ‘arguably one of the most
creative and imaginative Christian theologians of the early Middle Ages’ (p. xii).
Put another way, John of Damascus, who is commonly described in a Western con
text as ‘the last of the Church Fathers’, did not represent the end of Chalcedonian
Christianity, or indeed Christianity as a whole, in the Middle East. Rather it lived
on, even if under radically different circumstances, especially after the transition
from Umayyad rule (with Damascus as capital) to ¨Abbasid rule (with Baghdad as
capital) in 750 CE, and the works of Theodore Abu Qurrah are powerful evidence
of this, both as one of the first Christians to write in Arabic and as one of the first
to undertake a sustained theological defence of Christianity against the rival claims
of Islam. It is therefore excellent to have an English language collection of most of
his works published in order to help make his name and works much more widely
known.
doi:10.1093/jss/fgn072 HUGH GODDARD
UNIVERSITY OF NOTTINGHAM
GARTH FOWDEN, QuÒayr ¨Amra: Art and the Umayyad Elite in Late Antique Syria
(The transformation of the classical heritage 36). University of California Press,
Berkeley and Los Angeles, 2004. Pp. xxix + 390. Price: $55.00 hardback. ISBN:
0 520 23665 3.
QuÒayr ¨Amra (‘the little castle of ¨Amra’) is located in the basalt desert known as
the Balqa’, in modern Jordan, about 60 kilometres southeast of ¨Amman. It is a
small, well preserved example of the many Romano Syrian and early Islamic quÒur
(conventionally, ‘desert castles’), most of which are found on the arid inner edge of
the Syrian half of the Fertile Crescent, on the border with the Syrian and Arabian
desert. QuÒayr ¨Amra’s fame stems from the murals that adorn the interior of the
bathhouse and audience hall (ruins of other structures, including a mosque and
caravanserai stand nearby). The figural representations of profane subjects associ
ated with Near Eastern kingship are startling exceptions to an ‘Islamic’ artistic tra
dition that is usually taken to be predominantly abstract and aniconic. Scholarly
discussions of the building have tended either to be art historical, and concerned
with the murals and their patron, or archaeological, treating the building’s function
and its wider geographical and architectural context.
Fowden’s analysis is art historical. He seeks to situate this single, particularly full,
example of Umayyad visual culture in its late antique, Syrian context, and to use
the frescoes to understand more about late Umayyad court culture and ideology.
Questions about the quÒur as a category of buildings and of their function and con
text are by no means ignored, but they are subordinated to questions of patronage
and artistic interpretation. The result is a timely and thought provoking contribu
tion to late antique studies and Islamic art history. Fowden assembles and synthe
sizes a very large amount of evidence pertinent to interpretation of QuÒayr ¨Amra’s
frescoes. In so doing, he brings out the historical distinctiveness of the decades be
fore 750, when a newly victorious Îijazi Arabian nobility ruled a world empire
from the former Roman diocese of Oriens.
Fowden’s primary concern is with six prominent groups of paintings within the
audience hall: nudes and entertainers; the hunt; the enthroned prince; the reclining
woman with children; the six kings; the bathing woman. Studies of each of these
in turn form the core of the book: chapters two, three, four, six, seven and eight,
respectively, are devoted to each image or set of images. The remainder of the
book’s ten chapters set the publication in the context of previous investigations
298
REVIEWS
(chapter one), examine other contexts for interpreting the building and its murals
(Greek culture and the archaeology of other quÒur, in chapter nine) and present
some conclusions about ‘Umayyad self representation’ (chapter ten). The book piv
ots around the attribution of the building to its patron, and the implications of this
proposed provenance (chapter five).
Much of the interpretation stands or falls on this question of patronage. Fowden is
not only sure that the building is indeed late Umayyad, he is near certain that it was
constructed in c. 735 by the crown prince al Walid b. Yazid, later caliph al Walid II
(r. 743 4). As he acknowledges, in the absence of a full foundation inscription such
an attribution will never be absolutely certain, but he builds a fairly strong case. The
(by now widely recognised) termini post and ante quem are 711 12 and 750 (the fall
of Roderic, king of Spain, who appears in one of the frescoes paying homage to the
hall’s prince, and the Abbasid revolution, respectively). It is argued that biographical
evidence makes al Walid b. Yazid the strongest candidate for patron by virtue of his
association with the Balqa’, including his exile there during the caliphate of his uncle
Hisham, and his reputation for cultural sophistication and hedonism (the other two
main candidates are al Walid’s father, Yazid b. ¨Abd al Malik, and his uncle and
namesake al Walid b. ¨Abd al Malik). Fowden proposes the beginning of al Walid b.
Yazid’s exile from Hisham’s court in c. 735 8, when he is known to have resided in
the Balqa’, as the most likely moment for QuÒayr ¨Amra’s construction.
Indeed, he argues, after Imbert, that it is possible (though very far from certain)
that a very damaged inscription within QuÒayr ¨Amra refers to its patron as wali
¨ahd al muslimin wa’l muslima, ‘heir apparent of the Muslim men and women’. In
other places, he also notes al Walid’s reputation as a provider of services to the pil
grims, and that he is said to have led the Ìajj in 735. It may be worth adding that
leadership of the campaigns against Rome and of the Ìajj were two very prestigious
public appointments, frequently given to the crown princes or to candidates for the
succession, and that the only qaÒr that can definitely be linked to an heir apparent
(by a clear inscription) is QaÒr Burqu’, also on a route between Syria and Arabia.
This was developed by al Walid b. ¨Abd al Malik in 700 1, the first year in five that
he had not campaigned against the Romans and the year in which his junior
brother (and thus rival), Sulayman, led the Ìajj; QuÒayr ¨Amra may fit into a pat
tern of Marwanid succession politics, where the outcomes of heirs’ (and potential
heirs’) contests depended in part upon a physical presence among the Syrian tribes
and on the Ìajj route.
Although Fowden claims that the written evidence is brought to test the inde
pendently formulated interpretation of the visual material, he occasionally appears
in fact to have been led by the later Arabic literary sources. He acknowledges that
his interpretation of the image of the reclining woman with two children as Umm
al Îakam, the concubine mother of al Walid II’s nominated heir (chapter six), and
the large bathing nude woman as Shah i Afrid, the mother of his rival, Yazid b. al
Walid (chapter eight), are ‘speculative compared [to the interpretation of other im
ages in the hall]’ (p. 250). In each case, anecdotes about Hisham’s court (r. 724 43)
from al Mada’ini (d. c. 830 50) and others of his generation, preserved only in late
ninth and tenth century compilations, are the basis of the reading. Furthermore,
as Fowden notes in his discussions of these particular images, for his interpretations
to stand, these murals would have to have been painted some years after the pro
posed foundation date of c. 735.
In chapter four, interpretation genuinely seems to be led by the art history and
then corroborated by the texts and as a result it is much more persuasive. Through
299
REVIEWS
a wide ranging analysis of the painting of the seated monarch, which would have
appeared above the prince’s actual throne, Fowden demonstrates the importance of
the Roman and Sasanian visual heritage to Umayyad expressions of their authority.
As he reminds us (p. 123), ‘the prince in his alcove is, first and foremost, a reflec
tion of the late Roman iconography of imperium’ (or, in the East, perhaps basileia,
or mulk). In this connection, he notes parallels with Roman consular diptychs, but
also those between QuÒayr ¨Amra’s prince and the ruler on the post Sasanian
Qazwin plate, who is also depicted in a fusion of Roman and Iranian styles, which
owes more to Sasanian Iran than the image at QuÒayr ¨Amra. He also notes evidence
from pre Islamic poetry. (Here one might add al Nabigha’s pre Islamic panegyric
for the Lakhmid king, al Nu¨man.)
However, for Fowden, the most important precedents for such images are East
Roman images of Adam and Christ as world rulers. Adam was much less problem
atic than Christ in Quranic terms, and he also neatly solved the impossibility of
depicting God in Islam: ‘an Umayyad prince was neatly circumventing the obstacle
Islam had placed in the way of depictions of divine sanction or investiture, by al
luding to an icon that was widely known in late antique Syria and depicted the
common father of all mankind, whom God had created in his own image’ (p. 140).
The argument works very well, free of any textual support, but Fowden adduces
quite a compelling piece of textual evidence in the letter about the caliphal succes
sion that al Walid II is said to have had publicly proclaimed in Iraq and Khurasan
in 743, in which the Quranic verse that portrays Adam as the first khalifa fi’l ar∂,
‘the caliph on the earth’ has a central place as the model of the caliph (i.e. ‘succes
sor’, or ‘deputy’).
Circumstantial evidence from the extant Syrian and Arabian literary tradition
suggests that Fowden may well be right to emphasize the importance of Adam
(alongside David, Solomon and other biblical and Quranic patriarchs) to Umayyad
self representation as ‘the ultimate symbol of legitimacy’ (p. 138). In much of the
early Arabic Islamic historical tradition (eighth tenth century), as in late antique
Syriac texts, such as the Book of the Cave of Treasures (probably sixth century or
later), Adam’s role as the holder of the first covenant with God, and hence as the
first ruler of the world is emphasised, as it is in East Roman visual evidence ad
duced by Fowden. For the Arabians, as for many other late antique peoples (nota
bly, the Armenians, at the other end of the Roman Iranian frontier), the idea of
God’s covenant as uniting all true monotheists was central to political theory, as
was the notion that divinely ordained rulership could be traced from the Creation,
via great historical figures, down to their sovereignty in the present. (The parallel
importance of Gayomard/Jamshid in the Iranian world, and then in Islam, should
also be explored further.) It is to exactly this effect that the letter of al Walid II in
vokes the Quranic Adam: the reference is to the ‘primordial covenant’ (¨ahd or
mithaq Allah) between God and Humanity which underpins (and is in some sense
coterminous with) the covenant between the Muslims and the Umayyad caliph,
who is now khalifat Allah fi’l ar∂ ‘the representative of God on earth’. This ‘cov
enant’ (¨ahd, mithaq, Ìabl, et al.) was absolutely central to all Umayyad self repre
sentation in poetry and prose.
For Fowden, al Walid is very much the driving personality behind the imagery at
the palace complex: the composition of the images ‘suggests an original, imagina
tive mind at work … likely … the patron’s rather than the artists’ (p. 254); ‘al
Walid … chose Adam, the lord of creation, as the model for his portrait’ (p. 323).
QuÒayr ¨Amra is evidence of an ‘explicit and self confident vision’ of Arabian
300
REVIEWS
monotheist monarchy (p. 324), which uses Adam (and, as he notes, also David) as
visual symbols of Umayyad authority. However, elsewhere he concedes that it is also
a rare chance survival, and most likely not that unusual (pp. xxv, 275f.). The prob
lem here is that once again the later, highly literary portrait of the eloquent and im
aginative figure of al Walid that we find in the ninth , and especially the tenth cen
tury tradition appears to determine the interpretation of the images. One wonders
what precedents for Umayyad use of the prophets and patriarchs we would find if
more of the Umayyad Dar al Imara south of the Temple Mount in Jerusalem still
stood, or if anything of Mu¨awiya’s al Kha∂ra’ remained in Damascus.
Certainly, the use of the patriarchs as models was not new at all. Just as Adam
was God’s first earthly delegate and covenantee, David’s particular value, for the
Umayyads as for the Roman emperors, was in his status as God’s first dynast; Solo
mon’s was as David’s successor (also, like Adam, a lord over the elements). The
Umayyad poetry makes David and Solomon’s (and others’, including Jacob’s) sig
nificance to the political theory of succession plain, and both the evidence of the
pre Islamic poetry and now Robin’s work (2005) on the South Arabian inscriptions
prove that this understanding of Hebrew/Old Testament kingship was not new to
the peoples of the Arabian peninsula in the seventh century.
How this Umayyad self representation was received is a second problem that ab
sence of evidence makes almost unanswerable. For Fowden, al Walid II’s innovative
use of Adam in image and text is ironic in that it was formulated on the eve of the
fall of the Umayyads (p. 324). He argues that because ‘Quranic exegetes active un
der the Umayyads, when called upon to explain the … word khalifa as applied to
Adam, never with one possible exception suggested any connection between
this term and the head of the Islamic state [khalifa, ‘caliph’]’, the ‘Muslim Estab
lishment’ (his capital ‘e’) never accepted Umayyad legitimation of their rule (p.
140). However, in the absence of sources, the existence in the Umayyad period of a
‘Muslim Establishment’ must remain an assumption. It is an argument from silence
that al Qadi (1988), who compiled the Islamic exegetical evidence on which
Fowden depends, was loathe to make: the Umayyad interpretation lost out, and
with it many of the arguments and interpretations of any ‘establishment’ that they
had patronised; indeed, the Qur’an itself suggests that the Umayyads’ reading of
the tradition in this instance was not particularly weak or wilful.
Although he is not very explicit about it, it would appear that Fowden’s ‘Muslim
Establishment’ existed in the Iraqi garrisons of Kufa and Basra and in the old capi
tal of Medina (we have to wait until page 315 for this to become apparent). This
may be after all, Medina and Kufa were centres of early traditionist religious law
but there is perhaps a hint of teleology here. ‘Muslim Establishment’ implies a
rather homogenous and static community with shared, normative values and ideals
perhaps the ninth and tenth century ¨ulama’, who developed and preserved the
common tradition and values of what would become Sunni Islam. It might be ar
gued that this scholarly ‘Establishment’ was only embryonic in this period com
peting with other epistemic communities and not yet ‘established’. The Umayyad
caliphate fell not so much as a result of an ideological failure to persuade this con
stituency (‘the Muslim Establishment’s refusal to legitimize the excessively “kingly”
Umayyads’ in the face of which al Walid was ‘defiant’ [p. 138]), but because of
structural weaknesses in the organisation of its tribal armies, material grievances in
the populations of its frontier provinces and a fatal fragility in its very patrimonial
(and Arabian) institutions of succession. The representatives of the victorious
strains of Islamic tradition then wrote the only extant histories, in which the
301
REVIEWS
302
REVIEWS
303
REVIEWS
(The Sand’s Labyrinth, 1994), and this is the case with the Moroccan writer
MuÌammad Zifzaf (1945 2001), in his novel al Mar’a wa’l Warda (The Woman
and the Rose, 1972), among other Arabic literary works. Therefore, this issue is the
encounter between the culture of the ‘East’ with that of the ‘West’, or in other
words between the colonizer and the colonized or the oppressor and the oppressed.
The corpus of Arabic literary works and particularly novels, which may be called
‘between East and West’, can be divided into two main categories: the first includes
those novels that meet the definition of the Eastern hero leaving for the ‘West’ who
in most cases returns to his own country. The second category includes novels in
which the subject of ‘East’ and ‘West’ is employed only as a partial axis of the plot.
Another important issue of modern Arabic literatures is writing outside the Arab
world, or in other words writing in exile. Starkey describes the question of exile, or
more precisely the phenomenon of political, cultural, economic and problems of
freedom of speech, as leading some prominent poets and writers from the begin
ning of the writing of modern Arabic literatures to leave their territory in the Arab
world for other territories, and particularly to North and South America. These
writers created a very unique literature, either in Arabic or in the local language, i.e.
English, Spanish and Portuguese, that is known as Adab al Mahjar, and Jubran
Khalil Jubran (1883 1931) is the most famous figure among other prominent writ
ers such as Mikha’il Nu¨ayma (1889 1988) and Ilya Abu Ma∂i (1889 1957) who
were its exponents. While most of the Mahjar writers were Lebanese and Syrians,
the Arab writers who were forced to emigrate outside the Arab world during the
nineteenth and the beginning of the twentieth centuries were Egyptians and Iraqis.
It is worthy of note that since the second half of the twentieth century a new wave
of Arab intellectuals has emigrated outside the Arab world and may be termed post
Mahjar writers. They came both from the Arab Mashriq and Maghrib and mostly
preferred to live in Europe. Some of them continue to write in Arabic (al ™ayyib
∑aliÌ, Îanan al Shaykh, Zakariyya Tamir, Salim Barakat, Adunis), while others
preferred to write in the language of their new territory, such as French (al ™ahir b.
Jallun, Asia Jabbar) or German (Rafiq al Shami).
Paul Starkey uses the term ‘Western Reader’ (pp. 12, 14, 17 and so on), and the
question that arises is what about other readers, such as those from Asia and Africa?
It is quite natural that each of ‘us’ writes for ‘his/her’ reader, but we should bear in
mind that readers from other territories are part and parcel of ‘our’ addressees, par
ticularly if like the author we state that ‘[…] I hope that it may also prove of inter
est to readers interested in comparative literature, as well as to those whose prime
interests may not be literary, but who are in one way or another concerned with
different aspects of the Middle East a region of obvious and growing importance
in the world today’ (p.vii).
In addition, one should be aware that most of the bibliography of this book is in
English with only a few sources in Arabic, and it must be said that this is somewhat
strange since the research under discussion deals with modern Arabic literatures,
and especially because this book will also be very useful in the Arab world and out
side it.
I should mention some mistakes and errors that have occurred in this book, first
and foremost repetition of the same information. This is the case on the same page
(p. 110), where it is mentioned that the Egyptian journal al Fajr was founded in
1925. Other cases include the repetition of mentioning the lifetime of writers such
as MaÌmud Sami al Barudi (1839 1904), on page 43 and again on page 44, and so
on. One more remark regards the need to update information connected with some
304
REVIEWS
writers, such as in the case of the Syrian writer ¨Umar Abu Risha who was born in
1910 (p. 72), but it should also be noted that he died in 1990. This is also the case
with the Egyptian writer ¨Abd al Îamid Juda al SaÌÌar who was born in 1913
(p. 100), but it is not mentioned that he died in 1974. The same applies to the
Tunisian writer MaÌmud al Mas¨adi who was born in 1911 (pp. 121, 194), but it is
not mentioned that he died in 1981.
Other errors occur in the publication dates of several literary works. This is the
case with the Egyptian writer MaÌmud ™ahir Lashin’s short story ‘Îadith al
Qarya’, that was published on 19 November, 1928, not in 1929 (p. 111). This is
also the case in the short novel by YaÌya Îaqqi Qindil Umm Hashim that was pub
lished in 1944 and not in 1946 (p. 111). Other errors occur such as place, in the
case of the Palestinian poet MaÌmud Darwish who was born in al Birweh (or al
Berweh) but not Barwa (p. 92), and that of the Palestinian writer Emil Îabibi who
was awarded the Israel Prize for literature in 1992, not in 1972 (p. 151).
Paul Starkey is very right when he points out regarding Maghribian literature
‘[…] that of nah∂a in North Africa has been seriously under studied, at least by
English speaking researchers and deserves more serious consideration’ (p. 37).
I have three comments to add on this issue: first, that it is characteristic not only of
the nah∂a but also after it, and as a matter of fact until the present day. Second, it is
not sufficiently studied not only by English speaking scholars but also by research
ers in Arabic from the Mashriq, and this is the case of composing anthologies of
modern Arabic literatures that are edited either in Arabic by Mashriqian authors or
in English and other languages. And third, although Starkey devotes a relatively
large part of his book to the Maghribian literatures (pp. 36 8, 57, 121 2, 133 5,
159 9, 193 7), his study of the Mashriq literatures still proves his own statement.
Nonetheless, it should be said that the tremendous contribution of modern Moroc
can literature to the North African literatures is also noteworthy since it does not
occupy a marginal place compared with the Tunisian and Algerian literatures, as is
incorrectly indicated in this study.
In conclusion, Paul Starkey's book is very useful and well organized. Although,
as he states, ‘the particular readership that I have had in mind has been that of the
university undergraduate coming to the subject for the first time, I hope that it may
also prove of interest to readers interested in comparative literature […]’ (p. vii).
He truly achieves this goal.
doi:10.1093/jss/fgn074 AMI ELAD-BOUSKILA
BEIT BERL COLLEGE
305
SHORT NOTES
STEVEN E. FASSBERG and A. HURVITZ, (eds), Biblical Hebrew in its Northwest
Semitic Setting: Typological and Historical Perspectives (Publication of the Institute
for Advanced Studies 1, The Hebrew University Jerusalem). Hebrew University
Magnes Press/Eisenbrauns, Jerusalem/Winona Lake, IN. 2006. Pp. 324. Price:
$45.00 hardback. ISBN: 1 57506 116 3.
The papers contained in this volume arose from a meeting of an international re
search group on Biblical Hebrew in its Northwest Semitic Setting convened at the
Institute for Advanced Studies at the Hebrew University of Jerusalem during the
2001 2 academic year. The Preface provides the setting: a new look at the research
which led to William L. Moran’s seminal paper on ‘The Hebrew Language in Its
Northwest Semitic Background’, published in the Albright Festschrift edited by
G. Ernest Wright (in 1961, though the article was finalized in 1958). So much new
evidence has been discovered and published since then that a new evaluation was
deemed necessary. The editors mention the Deir ¨Alla text, the Tell Fekheriye bilin
gual, the Yavneh Yam, Arad and Kuntillet ¨Ajrud inscriptions and (of later date) a
vast quantity of ‘Judaean Desert’ material (Dead Sea Scrolls, Bar Kosiba letters, etc.).
The essays included cover a wide variety of topics. They tend to be specific, per
haps preparing some of the ground for a new summary to rival Moran’s, but readers
will not really find a new synthesis here. There are twenty papers. Though they are
printed in alphabetical order of the name of the author, they may be grouped as
follows:
A. Lemaire’s contribution on ‘Hebrew and Aramaic in the First Milennium
B.C.E. in the Light of Epigraphic Evidence’ corresponds directly to the agenda of
updating Moran. Lemaire is a master of the epigraphic environment and in this sur
vey shows how epigraphy can help in the understanding of the development of He
brew. Another contribution by him, produced jointly with A. Yardeni, is focussed on
the publication for the first time of new Hebrew ostraca from c. 600 BCE.
Most of the other papers are directly related to the agenda of reviewing the lin
guistic setting of Biblical Hebrew, but deal with very specific aspects: M. Bar Asher
on the qal passive of geminate verbs; S.E. Fassberg on command sequences; W.R.
Garr on paragogic nun; J. Huehnergard on the etymology of se (connection with
ˆaser); A. Hurvitz on prostration formulae; J. Joosten on the disappearance of the
iterative weqatal (as a diachronic marker in Classical and Late Biblical Hebrew);
M.Z. Kaddari on homonymy and polysemy in the new Modern Hebrew Lexicon of
Biblical Hebrew; G. Khan on aspects of the copula; M. Mishor on the language
and text of Exodus 18; A. Moshavi on object/adverbial fronting (Genesis as cor
pus); A. Niccacci on the verbal system evidenced in poetry; F.H. Polak on epic for
mulae; E. Qimron on pausal pataÌ; G.A. Rendsburg on Israelian Hebrew in the
Song of Songs.
Some of the contributions are of wider Semitic interest but have little to do with
the theme. That of J. Blau is concerned with nominal inflection in Arabic. Two con
cern inner Ugaritic issues, the forms and functions of the finite verb in Ugaritic nar
rative verse (E.L. Greenstein) and onomastics in Ugaritic poetry (M. O’Connor).
J.A. Emerton’s contribution is concerned with tracing the origins of Israelite histori
cal writing.
307
SHORT NOTES
The volume reflects the vitality of research in this area and the editors are to be
congratulated on a fine production which will ultimately point the way for a new
Moran.
doi:10.1093/jss/fgn075 JOHN F. HEALEY
UNIVERSITY OF MANCHESTER
308
SHORT NOTES
She has done a remarkable job, displaying her archaeological and epigraphic ex
pertise in full measure. Inevitably the editorial task involved much archival research
and this research is well represented in the volumes, with detailed accounts of the
history of the study of the ostraca, the archives of the excavations and other files.
Indeed the book contains an important collection of evidence on the history of
French scholarship during the twentieth century.
All of the above is covered in the Introduction, along with summary information
on the genres of the ostraca, a bibliography of items already published and a reca
pitulation of the work done by the different contributors named above. There is
also a chapter devoted to the study of the material (ceramic) aspects of the ostraca,
a modern feature which would certainly not have appeared in the 1920s!
We then move to what must here be a very brief description of the edition of the
texts. Chapter II is devoted to palaeography. There follows in Chapter III the full
publication of the 280 items (mostly ostraca, some jar inscriptions) of the main
collection. Then there are 36 belonging to subsidiary collections of unknown origin
(items X1 X33 and Y1 Y3) and 10 joined fragmentary texts are re listed as a sup
plement. There are also five ‘étiquettes’ written on wood.
For each item, there is a full physical description and transliterations showing the
readings of the different scholars who have worked on the material. Better readings
have been achieved by Lozachmeur through the use of digital technology (Adobe
Photoshop). There are letters, lists of names, delivery notes, writing exercises, etc.
Translations are provided, though these are often insufficiently complete to estab
lish the precise nature of the message or transaction. A large proportion concern
foodstuffs, textiles, utensils, tools, building materials, etc. A few allude to the Sab
bath and ritual impurities.
Chapters V and VI contain an ‘approche prosopographique’ and brief grammati
cal comments. After a short conclusion, indexes, including an index of Aramaic
words, complete the first volume.
The second volume is mainly devoted to palaeographic tables and photographic
plates and drawings (and some digital enhancements) of the texts. These are, how
ever, preceded by some coloured plates illustrating ceramic groups, drawings of ce
ramics and a collection of plates of archival documents related to the earlier work
which had been done by Clermont Ganneau, Cowley and Dupont Sommer.
This is a magnificent publication. Both the Académie des Inscriptions et Belles
lettres and the wider scholarly community of Aramaicists will be ever in debt to
Hélène Lozachmeur, who has brought to splendid completion the work of many
great scholars, while making her own significant (and modern) contribution to
their labours. The publication is welcomed in a preface by M. Jean Leclant as a
model of publication: that is what it is.
doi:10.1093/jss/fgn076 JOHN F. HEALEY
UNIVERSITY OF MANCHESTER
309
SHORT NOTES
(2003), 193 201 and Orientalia 71 (2002), 173 81. The best grammar remains
that by J. Friedrich and W. Röllig (revised by M.G. Amadasi Guzzo) (Phönizisch
Punische Grammatik3, 1999). Until now the revival was, however, largely limited to
interest in Phoenician rather than the later Punic: John Gibson’s well known Text
book of Syrian Semitic Inscriptions series included a Phoenician volume (1982), but
Punic (which had formed part of Cooke’s Text book of North Semitic Inscriptions,
1903) was not covered. This relative neglect is now rectified: the introduction here
under review was followed in 2008 by the publication of Jongeling’s comprehensive
Handbook of Neo Punic Inscriptions (Mohr Siebeck). The latter gives the impression
of being the foundation on which the introductory volume of 2005 was based and
Jongeling’s involvement in Neo Punic research goes back at least as far as his 1984
Groningen thesis on Names in Neo Punic Inscriptions. (In addition, of course, to his
part in the J. Hoftijzer and Jongeling Dictionary of the North West Semitic Inscrip
tions of 1995.)
It is to be noted that Late Punic Epigraphy (like the Handbook of Neo Punic In
scriptions) is concerned with the subset of Punic epigraphy which is called ‘Late’
and written in a distinctive script of cursive origin (called ‘Neo Punic’). There are
also many Late Punic inscriptions in Latin and Greek script and a selection is in
cluded. The time limitation is that the inscriptions all postdate 146 BCE, the date of
the destruction of Carthage at the end of the Third Punic War. The inscriptions
come mostly from North Africa, though it was no doubt a special pleasure for
Jongeling, who lectures on Welsh as well as Semitics, to include the sole Neo Punic
inscription from north of the Alps, one from Holt in north east Wales.
Both the introduction under review here and the bigger Handbook are important
contributions to the study of this neglected branch of the Semitic languages.
doi:10.1093/jss/fgn077 JOHN F. HEALEY
UNIVERSITY OF MANCHESTER
310
SHORT NOTES
311
SUBSCRIPTIONS
A subscription to Journal of Semitic Studies comprises 2 issues. Prices include postage by surface mail, or for
subscribers in the USA and Canada by airfreight, or in India, Japan, Australia and New Zealand, by Air
Speeded Post. Airmail rates are available on request.
Annual Subscription Rate (Volume LIII, 2 issues, 2008)
Institutional
Print edition and site-wide online access: £156//234/US$296
Print edition only: £185//222/US$281
Site-wide online access only: £148//222/US$281
Personal
Print edition and individual online access: £51//77/US$102
Please note: £ Sterling rates apply in UK, / rates apply in Europe, US$ elsewhere
There may be other subscription rates available, for a complete listing please visit www.jss.oupjournals.org/
subinfo.
Full prepayment, in the correct currency, is required for all orders. Orders are regarded as firm and payments
are not refundable. Subscriptions are accepted and entered on a complete volume basis. Claims cannot be
considered more than FOUR months after publication or date of order, whichever is later. All subscriptions in
Canada are subject to GST. Subscriptions in the EU may be subject to European VAT. If registered, please
supply details to avoid unnecessary charges. For subscriptions that include online versions, a proportion of the
subscription price may be subject to UK VAT. Personal rate subscriptions are only available if payment is made
by personal cheque or credit card and delivery is to a private address.
The current year and two previous years’ issues are available from Oxford University Press. Previous volumes
can be obtained from the Periodicals Service Company, 11 Main Street, Germantown, NY 12526, USA.
Email: psc@periodicals.com. Tel: +1 (518) 537 4700. Fax: +1 (518) 537 5899.
For further information, please contact: Journals Customer Service Department, Oxford University Press,
Great Clarendon Street, Oxford OX2 6DP, UK. Email: jnls.cust.serv@oxfordjournals.org. Tel (and answer-
phone outside normal working hours): +44 (0)1865 353907. Fax: + 44 (0)1865 353485. In the US, please
contact: Journals Customer Service Department, Oxford University Press, 2001 Evans Road, Cary, NC
27513, USA. Email: jnlorders@oxfordjournals.org. Tel (and answerphone outside normal working hours):
800 852 7323 (toll-free in USA/Canada). Fax: 919 677 1714. In Japan, please contact: Journals Customer
Services, Oxford University Press, 1-1-17-5F, Mukogaoka, Bunkyo-ku, Tokyo, 113-0023, Japan. Email: oku-
daoup@po.iijnet.or.jp. Tel: (03) 3813 1461. Fax: (03) 3818 1522.
Methods of payment. (i) Cheque (payable to Oxford University Press, to Oxford University Press, Cashiers
Office, Great Clarendon Street, Oxford OX2 6DP, UK) in GB£ Sterling (drawn on a UK bank), US$ Dollars
(drawn on a US bank), or EU/ Euros. (ii) Bank transfer to Barclays Bank Plc, Oxford Group Office, Oxford
(bank sort code 20-65-18) (UK), overseas only Swift code BARC GB 22 (GB£ Sterling to account no.
70299332, IBAN GB89BARC20651870299332; US$ Dollars to account no. 66014600, IBAN GB27BARC
20651866014600; EU/ Euros to account no. 78923655, IBAN GB16BARC20651878923655). (iii) Credit
card (Mastercard, Visa, Switch or American Express).
Journal of Semitic Studies is published two times a year in April and October by Oxford University Press,
Oxford, UK. Annual subscription price is £156//234/US$296. Journal of Semitic Studies is distributed by
Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. Periodicals postage paid at Rahway, NJ and
at additional entry points.
US Postmaster: send address changes to Journal of Semitic Studies (ISSN 0022-4480), c/o Mercury Interna-
tional, 365 Blair Road, Avenel, NJ 07001, USA.
PERMISSIONS
For information on how to request permissions to reproduce articles/information from this journal, please visit
www.oxfordjournals.org/permissions.
ADVERTISING
Inquiries about advertising should be sent to Helen Pearson, Oxford Journals Advertising, PO Box 347,
Abingdon OX14 1GJ, UK. Email: helen@oxfordads.com. Tel: +44 (0)1235 201904; Fax: +44 (0)8704
296864.
DISCLAIMER
Statements of fact and opinion in the articles in Journal of Semitic Studies are those of the respective authors
and contributors and not of Journal of Semitic Studies or Oxford University Press. Neither Oxford University
Press nor Journal of Semitic Studies make any representation, express or implied, in respect of the accuracy of
the material in this journal and cannot accept any legal responsibility or liability for any errors or omissions
that may be made. The reader should make his/her own evaluation as to the appropriateness or otherwise of
any experimental technique described.
© The University of Manchester 2008
All rights reserved; no part of this publication may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without prior
written permission of the Publishers, or a licence permitting restricted copying issued in the UK by the
Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1P 9HE, or in the USA by the
Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923.
Typeset by Drukkerij Peeters, Herent, Belgium
Printed by Bell & Bain Ltd, Glasgow, UK
JAN KEETMAN
Wechselwirkung von Vokalen und Gutturalen im Semitischen
unter dem Einfluss anderer Sprachen: die Beispiele des Akkadi-
schen und Hebräischen 1
NA’AMA PAT EL
The Development of the Semitic Definite Article: A Syntactic
Approach 19
GORDON J. HAMILTON
A Proposal to Read the Legend of a Seal-Amulet from Deir Rifa,
Egypt as an Early West Semitic Alphabetic Inscription 51
VINCENT DECAEN
Theme and Variation in Psalm 111: Metrical Phrase and Foot in
Generative Perspective 81
ROBERT D. HOLMSTEDT
Word Order and Information Structure in Ruth and Jonah:
A Generative-Typological Analysis 111
FRANÇOIS BRON
Notes sur les inscriptions néo-puniques de Henchir Medeina
(Althiburos) 141
TZVI NOVICK
The Modality of Òarik in Tannaitic Hebrew 149
SHLOMY RAISKIN
Talmudic Aramaic Fauna Names: Murzema and Shaqi†na 161
NADEZHDA VIDRO
A Newly Reconstructed Karaite Work on Hebrew Grammar 169
GERRIT BOS and Y. TZVI LANGERMANN
The Introduction of Sergius of Resh¨aina to Galen’s Commentary
on Hippocrates’ On nutriment 179
ALESSANDRA AVANZINI
Origin and Classification of the Ancient South Arabian Languages 205
AARON D. RUBIN
The Functions of the Preposition k- in Mehri 221
HILLA PELED SHAPIRA
From Conventional to Personal, or: What Happened to Metaphor
under the Influence of Ideology – The case of Gha’ib Tu¨ma
Farman 227
REVIEWS 251
SHORT NOTICES 307