Semitic Studies in Honour of Edward Ullendorff 9004148345, 9789004148345

Citation preview

SEMITIC STUDIES IN HONOUR OF EDWARD ULLENDORFF

STUDIES IN SEMITIC LANGUAGES AND LINGUISTICS EDITED BY

T. MURAOKA AND C.H.M. VERSTEEGH

VOLUME XLVII SEMITIC STUDIES IN HONOUR OF EDWARD ULLENDORFF

SEMITIC STUDIES IN HONOUR OF EDWARD ULLENDORFF EDITED BY

GEOFFREYKHAN

BRILL LEIDEN· BOSTON 2005

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data A C.LP. record for this book is available from the Library of Congress.

ISSN 0081-8461 ISBN 90 04 14834 5

© Copyright 2005 by Koninklijke Brill.Nv, Leiden, The Netherlands Koninklijke Brill NV incorporates the imprints Brill Academic Publishers, Martinus Nijhqif Publishers and VSP. All rights reserved. No part if this publication mqy be reproduced, translated, stored in a retrieval system, or transmitted in any form or by al!)' means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklyke Brill provided that the appropriate fies are paid directlY to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers, MA 01923, USA. Fees are suiject to change. PRINTED IN THE NETHERLANDS

CONTENTS INTRODUCTION ..................................................................................... 1 UPDATE TO THE BIBLIOGRAPHY OF THE WRITINGS OF EDWARD ULLENDORFF ................................................................. 4 GIDEON GOLDENBERG Semitic Triradicalism and the Biradical Question..................................... 7 JOHN HUEHNERGARD Reflexes of *Qatl Forms in G~\.:.,.

~ ~~I ~ ~;i

Js-

'YbJ 'J'J

IJIJ

I"I.\~!\~I~·I

. 0':!?

J!.o J?f J;.J :.li J!.o dl:!ll

~1J:-'\r10-' IJ'~J;.s.J";:"

~ ~ ~u.-

IJ:-'I ')J

~;! ,~I ~-;:~:... ~ •

[

~~.uJ

.

olAJ

.~, - - - ] '..:..>~..w. J

4ft ~ ..!.JWI ~~ lilJ ' ~J .r;J ~ J!.o J?i 4I~

J-pl ~\.:.,. w; ,a....tS:.l1 f"T ~ :;; ~4J ~~ ~4 ,

~.,Ai. ;l; j...AJIJ

•

'0''';'

e:J.-IJ

!.s;

..

J!.o

..

J.

(J~I I~~i;" J

W" «:J;» ...t..,..i ~I

J ...

lS'L...

(al-1jalil, Kitab al-'Ayn I, 49-50) ... I.A~ ~

*'-=

'J~ +- J~

li? }

~

~

'I, .. t Y' +- Y' J.

1)1.!.t'i

"'0"'\"" 1('\ . ).l +- ..... ).l :~_IJ '-'J')-I

t~).l

~IJ .~~

Ji J.,-u

,.-",,.

~I

:J~ JJ.:l1

IlA

4--J1J""iJ l::\IJ l:JI

~'Y llA :J~ W" ,~ti IlA :J~

'4---1

(JI5"'J

~

..,..,4

(~~f 0~)

~ ~I ..

J J? 0-'

~ 1)1 'Y !.Sf: !.S?'

,.!.m piJ

(Sibawayh 11,31-35)

.

~

~

:J?i 4I~ ~ (s> ~ yl!\I J 'Y! oA? (J~ 'Y ~I dl:!lIJ J! ~ ~ I)~

,.t,j

J~ J ~ L.;G ~

~

lilJ

'I,$'~ .sJ.IJ JG .lulll

~ oA? yl!\I J a...-...ll J J')-I ~i JJ\r1 ~.)-I (J15"' ~IJ .sJ.1 '"

•

,.....

~ ...t..,..1 (JI.) J

.I..ii 'C-= ~

~

, lii

,,~

,,~,

.

-

i

~ ~ :~ f"'Y 1 ~ ~..l...t 1J'..l>

(Ibn Durayd, Gamharah I, 53)

SEMITIC TRIRADICALISM

13

In the Hebrew linguistic tradition, even before the adoption of the Arab theory of 'weak letters', triradicality as a minimum requirement could not be ignored. M::mal:;tem ben Saruq, the 10th century lexicographer arranged the Biblical vocabulary in his lexicon Ma1;tbact under entries of one, two, three or more letters, each consisting of the unwavering common letters that are never omitted in any of the related forms. These entries are termed C'''CiI n",o' yasogo! ha-millfm 'bases of the words' (corresponding to AlrasI's 'ummahiit al-'alfiit wa-'ussuha), but they are not regarded as roots or arch-roots in any etymological or semantic or even formal sense; on the contrary, it is stated that words with less than three radical letters need to be supplemented to make a defined root, without which addition i1"~Pil i1"Oil ,oJ)n t6 10 ta'amog ha-milia ha-q$iira 'the short word would not stand'. The firm insistence in M~nal:;tem's introduction on the triliteral root as a minimum requirement came to ensure that the various sections (n'p"no) under a uniliteral or biliteral entry should be clearly recognized as separate and distinct lexical items. The entry '!l, for example, is divided in eight sections: (1) ,~, i11~ 'bull, cow'; (2) as in '.,~ 'fruit', the related verb (...Jpry) and i11~e; (3) as in c.; '!lW:;t l!1iJi!l i1n~ (ps. 74: 13) and in 'iTliiet;1\:1 'i~ (Isa. 24: 19) 'split, divide' etc.; (4) 'similar to 3' '~iJ 'break, violate' etc.; (5) ,~e 'lot'; (6) ~l~ (also i11~) 'wild ass'; (7) c:ne ; (8) "'~ 'pot' and ,~,~~. The main idea is that there is no etymological, semantic or grammatical relation between the 'sections' which only share the same biliteral kernel. The real lexical items are the 'sections', but within the entries of the Ma1;tbatt these are illustrated only by Biblical quotations, with no explanatory labels. Strong roots which have two radicals in common are listed separately. M~nal:;tem's main entries represent a technical device of arrangement and are lexically meaningless. They can be compared in principle to the main divisions in al-ijalIl's Kitiib al-] SS > ns sound change has yielded np.91iJ 'bringing up'. Some other fluctuations as well may occur in lexemes with weak and phonetically unstable radicals, and paradigms can be mixed simply due to phonetic similarity, but such common phenomena have nothing to do with reconstructing proto-roots like *-Vtb or *-vsq, and, as we all know, there are no biradical roots in Semitic, and no one can imagine forms derived from such roots. The traditional analysis of Semitic base-formation by interdigitation (or 'Durchdringung', or 'dovetailing') of root and scheme has been known in general linguistic literature since the 1980's as 'McCarthy's theory'.5 Two real innovations were added to the description, the one is the addition of a skeletal CV tier involving syllabic structure as regulating the interdigitation of root and scheme, and the other is the so-called 'Obligatory Contour Principle' (OCP). CV sequence and syllabic structure are clearly out of place here. I have written about this previously and find no need to repeat it. We may compare subjunctive and imperative forms like t-angar and nagar in Ethiopic, Hebrew variants like yiSma'u and yiSim'u and a lot of other evidence, passim. Within the same theoretical framework it was assumed that multiple occurrences of a consonant or a vowel in the stem must be represented by a single element on the melodic tiers (Le. root and scheme level), following the so-called 'Obligatory Contour Principle', according to which 'at the melodic level, adjacent identical elements are prohibited',6 so that the underlying scheme of katab is lal, and the underlying root of samam is Isml; the spreading is said to be dictated by the skeletal template, and all auto segmental spreading is said to be rightward7 (that is if written from left to right). An underlying biradical root is thus assumed in the synchronic description. A central argument used for applying the 'Obligatory Contour Principle' to Semitic derivational morphology is the claim that while there is in Semitic quite a number of roots whose second 5

For example, in Spencer (1991: 134) or Carstairs-McCarthy (1992: 82).

6

McCarthy (1986: 208)

7

McCarthy (1986: 209)

20

GIDEON GOLDENBERG

and third radicals are apparently identical (1-2-2), there are 'certainly in Arabic, and reasonably confidently in the other major Semitic languages' no roots (or a 'virtually complete absence' of roots) with identical initial radicals 1-1-2.8 This is said to prove that such roots with adjacent identical elements must be regarded as biradicals whose second radical is spread (from left to right) to fill the unassociated C-position in the prosodic template tier, i.e. to provide the third radical of the rootmorpheme. From among the adherents of this doctrine it was mainly Georges Bohas who showed that there were close lexical (or etymological) connexions between mediae geminatae and verbs with a weak radical after, between or before the R\R2 basis, and argued sensibly against the principle of obligatory rightward spreading and for an alternative theory of free association and reduplication, which should account for roots 1-2-2, l-w/y-2, 1-2-w/y, w-I-2, 1-2-1 or 1-2-1-2 (as well as for the Ethiopian 1-1-2 verbs), all regarded as being basically expanded biradicals, but making distinct roots. 9 Though little is left here of the 'obligatory contour', Bohas claims that free association still keeps in line with the obligatory requirement, 'tout en respectant les exigences du PCO',10 only because adjacent identical elements are still described as associated with a single radical (but see below). He later discovered that etymological connexions might apparently be identified not only between, e.g., Arabic ...J~++ and ...J~+y, ...Jgdd and ...Jgwd, ...Jrdd and ...Jwrd, or -vqlql and -vqlq, but also between ...Jgzz, ...Jgzr, ...Jgz', ...Jgzl, ...Jgzm, ...J9Z;', ...Jwgz, ...Jgrz and ...Jglz. These roots, as if derived by various types of J~I, are said to be related to a common biconsonantal primitive gz, defined as an etymon and not as a root. ll The greatest mistake of this assumption lies in the basic facts. In Ethiopian languages, which share with the other Semitic languages the same system of morphological derivation by root and pattern, roots 1-1-2, and not only 1-2-2 (or 1-2-1-2), are quite common, like the Amharic ...Jbbt,

...Jeer, ...Jddb, ...Jddr, ...Jddq, ...Jggb, ...Jggm, ...Jggr, ...JgWgWt, ...JgWgWt, ...Jggb, ...Jggr, ...Jkkm, ...Jkkr, ...Jkwkwl, ...Jqql, ...Jqqr, ...Jqqt, ...Jssn, ...Jssg, ...Jssr, ...Jttg, ...Jttr, ...Jttl, ...Jw, ...Jttt (...J$$t), ...Jzzn, or similarly in Tigrinya, see M. Denais ap. Bohas (1991: 135). Amharic ...Jqql 'cook', ...Jqll 'be light' and ...Jqlql 'mix' are independent 8

McCarthy (1981: 395-96; 1986: 209); Lowenstamm and Alaoui EIMhammedi (1995: 127)

9

See Bohas (1991; 1993), Bohas and Chekayri (1993).

10 11

Bohas and Chekayri (1993: 12) Bohas (1995)

SEMITIC TRIRADICALISM

21

roots, and so are the roots in most other similar series, like vssg 'hide' vsgg 'be fine' - vSgSg 'make room' or the like, vkkr 'be vigorous' 'twist' - vkrkr 'be acrid', vggr 'have no appetite' - vgrr 'leak out' - vgrgr 'be in a hurry', vtP" 'harden' - vtrr 'be hot' - vtrtr 'suspect', vt# 'abandon' - vtll 'be pure' /'shelter' - vtltl 'wander about', vddb 'be dull, hard' - vdbb 'give shade' - vdbdb 'beat', and many others. If a supposed 'germ' *ql can bring out not only vqll, but also vqql and vqlql as separate roots, then the OCP exercise is altogether baseless.

vkrr

7. There remains the question as to how to understand the relation between the roots in a set of semantically close triradicals which have two radicals in common and only the third is different. Here one might think of two possible explanations: (1) regarding the two common radicals as the Urwurzel or the real root, (2) seeing that various forms with semantic similarity may be phonetically close showing what one may wish to connect with various degrees of sound symbolism, iconicity or sheer analogy. Georg von der Gabelentz said a century ago that 'root' (if loosely defined) could be a very subjective concept. There have been scholars who have stated categorically, like Eduard Konig in his Hebrew grammar, that 'die Wurzel ist zweiconsonantig. Das dreiconsonantige Gebilde ist der einfachste Stamm (Grundstamm) der Urtheilsausserungen' (Konig Lehrgebiiude [1895] II 270). A clearer and preciser statement of an original Semitic biradicalism was that made by Solomon Theodore HaIevy Hurwitz, who made the most thorough and most elaborate study ever made on the subject. His conclusion was that 'the Semitic root was originally biliteral [- - - ]. The original biliteral developed its triliteral form by the addition of formative elements to express various functions of the stem' (Hurwitz 1913: 107). Such meaningful formative elements, or root-determinatives, cannot be established, and regarding the common sounds in partially similar lexemes as proto-roots may lead to results that imply pre-historic processes that cannot be observed in actual languages. Meaningful assonance, rhyme or alliteration are a widespread phenomenon, and association of sounds and meanings is not so rare, as in English slim, slender, slide, slush, slither, or glow, gleam, glitter, glaze, glade, etc.12 Semantic affinity between the words in which the sounds occur can well be recognized, but this does not imply any status of sounds such as sl- or gl- as representing by themselves etyma. Should we regard English tOiling and moiling as derived from *voil? G. von der Gabelentz, whose ideas we 12

Cf., for example, Haas (1976: § 1.2).

22

GIDEON GOLDENBERG

have already considered, drew attention to the great allomorphic variations in stem-forms and noted the common phenomenon observable everywhere of having similar word-forms associated semantically and phonetically. I should like to conclude with an apposite quotation from the second part of Goethe's Faust, in a dialogue ofMephistopheles and a griffin:! 3

7090

(ftn rotbrlg molkl ~ocl) batf mtcl)'s ntcl)t uerbrlesen, ~ls neuer G;aft anftanbig fie au gritSen ... G;U1ckau ben fd)onen ~raun, ben Wugen G;retfenl G;retf (fd)narrenb) .

7095

9Ucl)t G;retfenl G;retfenl-9Uemanb ~orl es gem, ~as man t~n G;reis nennt. ~ebem m30rte Wingt ~er Urfprung nacl), roo es ficl) ~er bebtngt:

G;rau, gramItcl), grlesgram, greuItcl), G;raber, grlmmtg, (ftl}mologtfd) gletcl)erroetfe fttmmig, merfttmmen uns. 9Repl)tftopl)eles..

Unb bocl), ntcl)t ab3ufd)roeifen, G;efaUt bas G;ret tm (f~rentttel G;retfen. (G;oet~e, ffau{t 11 ) What exactly a griffin's etymology might be I do not pretend to know, but even such a creature, I presume, would not mean to reconstruct gr as the Urwurzel of those words.

REFERENCES al-ijaIil, Kitiib Al-'Ayn -'AM 'Abd aI-Ral,mlan aI-ijalil b. ' Al;unad aI-Fariihidi [718-776~791], Kitiib Al-'Ayn, ed. Mahdi al-MabzUmI and 'lbriihim al-Samarr~PI, 8 vols, Baghdad, 1980-1985.

13

Attention to this passage was drawn by H. J. Polotsky, mentioned by Ullendorff (1978).

SEMITIC TRlRADICALISM

23

Baudouin de Courtenay, LA, 1897, [I1:BaH ArreKCaH.D;pOBH'I EO.D;y3H .D;e KYPTeH3--Jan Ignacy Niecislaw Baudouin de Courtenay], 'HeKoTopble H3 o6IIJ,Hx IIOJIO)l(eHHH, K KOTOPbIM .D;OBeJIH EO.D;y3Ha ere Ha6mo.D;eHH5! H HCCJIe.D;OBaHHH HBJIeHHH H3bIKa', if36paHl-lble mpyobl no o6UfeMY R3bIK03f1aflUlO, MocKBa, 1963, vol. I, pp. 348350. Benveniste, E., 1958, 'Les verbes delocutifs', ap. E. Benveniste, Problemes de Linguistique Generale, Evreux, 1966, pp. 277-85. Bloomfield, L., 1933, Language, New York. Bohas, G., 1991, 'Le PCO, la composition des racines et les conventions d'association', Bulletin d'Etudes Orientales de I'Institut Franyais de Damas 43, pp. 119-37. - - - - , 1993, 'Le PCO et la structure des racines' and 'Diverses conceptions de la morphologie arabe', Developpements Recents en Linguistique arabe et Semitique. Seminaire tenu au College de France, organise et presente par Georges Bohas [Publication editee par l'Institut Franyais d'Etudes Arabe de Damas N!! 142], Damas, pp. 9-59. - - - - , 1995, 'Au-deh\ de la racine', Proceedings of the Colloquium on Arabic Linguistics, Bucharest August 29-September 2 1994, Part One. Edited by N. Anghelescu and A A Avram, Bucharest, pp. 29-45. Bohas, G. and Chekayri, A, 1993, 'Les realisations des racines biliteres en arabe', Semitica: Serta philologica Constantino Tsereteli dicata, curaverunt Riccardo Contini, Fabrizio A. Pennacchietti, Mauro Tosco, Torino, pp. 1-13. Bottcher, F., 1866-1868, Ausfohrliches Lehrbuch der Hebriiischen Sprache. 2 vols, Leipzig. Brockelmann, c., 1908-13, Grundriss der Vergleichenden Grammatik der Semitischen Sprachen, 2 voIs (vol. i 1908, vol. ii 1913), Berlin. Carstairs-McCarthy, A, 1992, Current Morphology, London - New York. Cohen, D., 1978, 'A propos d'un dictionnaire des racines semitiques', Atti del Secondo Congresso Internazionale di Linguistica CamitoSemitica (Firenze, 16-19 aprile 1974), raccolti da Pelio Fronzaroli = Quaderni di Semitistica 5, Firenze, pp. 87-100. Cohen, M., 1936, Traite de Langue Amharique (Abyssinie), Paris.

24

GIDEON GOLDENBERG

Ewald, H., 1870, Ausfiihrliches Lehrbuch der Hebriiischen Sprache des Alten Bundes. 8. Ausg., Gottingen. Gabelentz, G. von der, 1901, Die Sprachwissenschajt, ihre Aufgaben, Methoden und Bisherigen Ergebnisse. Zweite, vermehrte und verbesserte Aufiage, hrsg. v. Albrecht Graf von der Schulenburg, Leipzig. Gesenius,W. and Kautzsch, E., 1910, Wilhelm Gesenius' Hebriiische Grammatik, 28th edition, ed. v. E. Kautzsch, Leipzig 1909 = Gesenius' Hebrew Grammar, edited and enlarged by E. Kautzsch, 2nd English edition by A. E. Cowley, Oxford, 1910. Goldenberg, G., 1994, 'Principles of Semitic word-structure', Semitic and Cushitic Studies, ed. G. Goldenberg and S. Raz, Wiesbaden, pp. 2964. - - - - " 1997, 'Conservative and Innovative Features in Semitic Languages', Afroasiatica Neapolitana [Istituto Universitario Orientale: Studi Africanistici, Serie Etiopica 6]: Papers from the 8th Italian Meeting of Afroasiatic (Hamito-Semitic) Linguistics, Napoli, ed. A. Bausi and M. Tosco, pp. 3-21. Haas, W., 1976, 'Writing: the basic options', in W. Haas (ed.), Writing Without Letters, Manchester and Totowa, N.J., pp. 131-208. Hurwitz, S.T.H., 1913, Root-Determinatives in Semitic Speech. A Contribution to Semitic Philology, New York. Ibn Durayd, Gamharah-'Abfi Bakr Mu1}ammad b. al-I:Iasan b. Durayd al-'Azdi [838-933], Gamharat al-Lugah, ed. Ramzj Munir Ba'labald,3 vols., Beirut, 1987. Konig, F.E., 1881-1897, Historisch-kritisches Lehrgebiiude der Hebriiischen Sprache. 3 vols (vol. i 1881, vol. ii 1895, vol. iii 1897), Leipzig. Lowenstamm, J. and Alaoui EIMhammedi, S., 1995, 'On the correctness of the biliteral analysis of mediae geminatae verbs', Langues Orientales Anciennes: Phi/ologie et Linguistique 5-6, pp. 127-31. McCarthy, J.J., 1981, 'A prosodic theory of nonconcatenative morphology', Linguistic Inquiry 12, pp. 373-418. - - - - , 1986, 'OCP effects: gemination and antigemination', Linguistic Inquiry 17, pp. 207-63. Olmo Lete, G. del, 2003, Questions de Linguist/que Semitique: Racine et Lexeme. Histoire de la Recherche (1940-2000) [Antiquites Semitiques V], Paris.

SEMITIC TRIRADICALISM

25

Prasse, K.-G., 1967-68, 'L'origine de l'imparfait 6thiopien 1.2 yafe~~am',

Comptes Rendus du Groupe Linguistique d'Etudes ChamitoSemitiques (GLECS) 12, pp. 52-53. Prunet, J.-F., 1996, 'Some core properties of Semitic morphology: Evidence from the far South', in J. Durand and B. Laks (eds.), Current Trends in Phonology, Salford, 11, pp. 617-52. SIbawayh-'Abii Bisr 'Amr b. 'U,t.man SIbawayhi [757?- c.796], Kitab Sibawayhi 2 vols, Biilaq, 1316-1318 [1898-1900]. Spencer, A., 1991, Morphological Theory. An Introduction to Word Structure in Generative Grammar, Cambridge MA. Ullendorff, E., 1978, 'Mephistophelian semantics: A postscript to BSOAS, xl, 3, 1977, 578', Bulletin of the School of Oriental and African Studies 41, p.582. Voigt, RM., 1988, Die infirmen Verbaltypen des Arabischen und das Biradikalismus-Problem, Wiesbaden.

REFLEXES OF *QATL FORMS IN GH'3Z John Huehnergard In the half century since its publication, Professor Ullendorffs groundbreaking study of the phonology of the Ethiopian Semitic languages has not been superseded, and it remains a rich source of information and insight (Ullendorff 1955). In the present note I hope to contribute in a small way to our understanding of the historical phonology of classical Ethiopic, with a review and analysis of the G 70tta dogge 'to the village'.2 However, adjectival modifiers and qualifier nominals with nay, may optionally follow the noun head, in which case the definite marker remains in initial position: la-bab labi 'the big gate', la-ba7at nay nayat 'the cave of the lion'. In the case that a qualifier with nay precedes its head, the definite marker attaches directly to the qualifying noun: nay la-mudiryat sayam 'the chiefs of the provinces'. In complex NPs without nay, the definite marker variously attaches to the head noun, to the qualifier, or to both the head and the qualifier: wolad la-dogge 'the boys of the village', la-gomat derho 'the advice of the chicken', or la-wolad la-dagge 'the boys of the village'. Additionally, the definite marker may be optionally attached to a noun with a pronominal possessive suffix: la-bo7os-a 'her husband', la7amfit-om 'their entrails', and similarly is optional when a deictic/demonstrative determiner is included in the NP: t0b la-lali laM 'on that night', but 7011an 7amlolat 'these days'. In Tigrinya, the definite marker is formally identical with the short form of the far demonstrative adjective, and as such marks gender and number in agreement with the noun: m.sg. 7atu/7oo, f.sg. 7ata, m.pI. 7otom, f.pl. 7otiin, with shortening to tu, etc., in rapid speech and after prepositions. As in Tigre, the definite marker in Tigrinya is attached to the front of the noun phrase, where it is normally only preceded by

nu

2

Tigre data is taken from Leslau (1945) and Raz (1983). The transcription of all the language data has been kept as in the sources except that the subscript dot indicating glottalic articulation has been replaced with a following , for typographical reasons (similarly k' for q), and h replaces h with a subscript dot. For the glottal stop 'l is used, and r for the voiced pharyngeal.

DEFINITE MARKERS

53

prepositions: 7ata s'abbax'ti gWal 'the beautiful girl', 7atu nay 7astiimhari giiza 'the house of the teacher',3 7ata 7at-tamarfo gWal 'the girl who is getting married', 7ankab-tu nay siib7ay giiza 'from the husband's house'.4 Additionally, quantifiers may precede the Noun Phrase with the definite marker: kulliin 7atiin habtamat 7anasti 'all the rich women'. Also as in Tigre, the definite marker may be used with a noun + possessive pronoun suffix: 7atu s'abbux' k'iimis-iiy 'my beautiful dress'. Leslau (1941: 39) further suggests that 3rd person pronominal suffixes can be used as 'definite articles', though a different interpretation is possible, as with similar instances from some South Ethiopic languages: kalatti-7om tiimiiharti 'the two students', lit. '[the] two of them, [the] students', saftatat nab biiriixa-7om ya7atwu 'the bandits go into the bush', lit. ' ... into their bush'. In Amharic and Argobba, the definite marker has the form of a suffix, -u/-w 5 on masculine and plural nouns, and -wa (and in Amharic a variant -itu, or with double marking -itwa, in some dialects) on feminine nouns. The suffixes -u and -wa are formally identical with the 3rd person possessive suffixes, masculine and feminine, respectively. The definite marker is placed on the left-most nominal constituent of the NP, i.e. a qualifying adjective, noun or relative clause: Amharic tallak'-u bet 'the big house', yii-siirratiififia-w biik10 'the worker's mule', azzih yii-tiitiikkiilii-w dankwan 'the tent that is pitched here'. The definite marker may be followed by the object marking clitic, or a sentence or phrasal connective such as -mm or -ss: leba yii-siirratiififia-w-n biik10 siirriik'ii 'a thief stole the worker's mule'. Unlike Tigre and Tigrinya, a noun with a pronominal possessive suffix may not also have a definite marker; if the NP also contains a qualifier, however, the qualifier assumes definite marking in agreement with the noun + possessive pronoun, which is intrinsically definite: wiindamm-e 'my brother' but tallak'-u wiindamm-e 'my elder brother'. In Harari, the definite marker has the invariable form -zo,6 which is identical to the 3rd masculine possessive suffix. As in Amharic, the 3

Note, however, that the placement of the definite marker before the complementiser nay differs from the Tigre sequence.

4

Tigrinya data is taken from Leslau (1941) and Tesfay (2002).

5

-u after consonants, -wafter vowels, except in Argobba which additionally has

"yu/_wu after the vowel i. In Arnharic, the article added to relative clauses has the additional shapes -t and -iiw in specific formal contexts. The information from Argobba is insufficient to say whether something similar occurs there. 6

-zo is lengthened to -zo- before another clitic element.

54

DAVID L. APPLEYARD

Harari definite marker precedes any object marking or sentence connective clitic, but unlike Amharic it remains in NP final position and is not moved leftward to any qualifier: ge-Ie zi-giidiira miigala-zo 'the big market of Harar', gar-Ie k'ac'i-zo-be 'on the outside of the house', islam haliit-zo-be 'in the Islamic way of life'.7 It is often difficult to decide whether the suffix -zo is indeed a definite marker or a pronominal possessive. Thus in the following examples, where Leslau glosses -zo as a definite marker, it is equally possible to interpret it as a possessive: ligi wiildi t'iiba-zo 'the boy's voice', diliiga-aw gar-za-be 'in the boss's house', ligi-le miirfnac-zo 'the boy's friends', goyta amri-zo 'the Lord's commandments', etc. In Muher and Ezha, the definite marker has the invariable form -we, which is not obviously connected with any other determiner or pronoun in the language(s). It occupies the same position as the Amharic-Argobba marker, i.e. in complex NPs it is placed on the left-most qualifying adjective or noun preceding the head: Muher anassayii-we bet 'the small house', xoyt-we siib 'the two menlboth the men', b-igga-we bet 'in the boy's house', yabiisa-we mass 'the man who is coming', yii-fdkka-we mass 'the man who went', Ezha yii-mast-we miirkamma takii 'the woman's pretty child'.S There are also a few examples in the available Muher data where the possessive pronoun suffix of the 3rd masc. sing. -Wxta seems to have definite marking function, though, as in the Harari instances cited above, it is often not possible to determine whether the meaning is really simply +Definite, or whether a pronominal anaphoric reference is meant: gazyii-m-aWxta masiit banno 'and the time was night' (lit. 'and its time .. .') (see Leslau 1981: 10 for more examples). In a number of other languages, Zway and Silt'i from East Gurage, and Soddo, Goggot, and Masqlin, as well as traces in Gyeto and Ennemor (the last five of which belong to various branches of what Hetzron subsumes under the term Gunnlin-Gurage) the definite marker has the form -i/-y, which is similar to one form of the pronominal (object) suffix of the 3rd masc. sing., in some of the languages in question. Soddo further has an optional form -iti on feminine nouns, and Silt'i has -te. For the purposes of illustration, only data from Soddo and Silt'i are given here. 9 The facts, however, appear to be the same for the remaining 7

Harari data is taken from Leslau (1965).

8

Muher is data is taken from Hetzron (1977) and Leslau (1981). The Ezha example is taken from the former.

9

The Soddo data is taken from Hetzron (1977), and the Silt'e data from Gutt 1997.

DEFINITE MARKERS

55

languages. The definite marker occupies a similar position in the NP as in Amharic-ArgobbalO and Muher-Ezha. Soddo: bii-g()dd()r-i ank'iifo 'with the big spoon', yii-lelocc-i awreocc 'to the other animals', yii-biila-y m()ss 'the man who said', ba-t'ot'-iti ge 'in the macaque's house', yii-kitt-i gurzazii c'iiwata 'the two old people's conversation'. In the Silt'i data, there are additionally examples of the definite marker added to another determiner such as a demonstrative adjective: itta-te ya-saalo laam 'this cow of Saalo's'; and also to a noun with a possessive pronoun suffix: baingiir-kaa-y ragat'a-y 'he kicked him with his foot', 11 la-baad-naa-y sab 'to the people of our country'. Lastly, as in a number of other South Ethiopic languages, there are occasional instances of a 3rd person pronominal suffix used in what may be interpreted as definite marking function. This is particularly clear in the following Masqan examples where context does not provide for a 3rd person anaphora: yiibet()mmi m()ss, m()k'iir()nhii bet-iihW yii-ciififiiihii, biiram tiisaliinn '[and] the man of the house asked him [saying], 'who are you who came to the house?" lit. ' ... to his house', where the speaker is referring to his own house; or, tii-gobbabit-iih()nno-ye wiiram ... c'ok'k'iisiinno 'he went to the brothers and asked them .. .', lit. ' ... to their brothers .. .', where context dictates that a possessive referent should be singUlar (Hetzron 1977: 56).1 2 In Gafat, the definite marker is invariable -s[ii], occupying a similar position in the NP to what we have seen in Amharic and most other South Ethiopic languages. Leslau (1956: 44) remarks that it is, however, 'd'un emploi beaucoup plus frequent que celui de l'article des autres langues ethiopiennes', and that it is 'un element de mise en relief et de reference'. Thus: ()mmun-sii-n W()sSa diiriishunni 'I found the big dog', y-iimmwiiy-()ggii wiiy()-s giiggii 'my uncle's new house', yiiniibbarii-y;rs giigga 'the house that I had', ya-giizziyu-sa gii9gii 'the house that he bought'. 13 In the remaining South Ethiopic languages, definite marking, where used, is done by means of the appropriate 3rd person pronominal clitic: 10

11 12

13

Note, however, that unlike in Amharic, the definite marker may optionally be placed after a sentence connective such as -mm, as in bet-"mm-i '[and] the house', below, beside Soddo at'ayocc-i-m '[and] the sheep'. Note also here the object suffix -y 'him'. The latter example indicates that the pronominal suffix agrees in number and gender with the head noun and not with any possible referent in the discourse: ta-gobbabit-ah"nno-ye = to+brothers+their+to. The Oafat data is taken from Leslau (1956). His morpheme analysis of some of the data is occasionally erratic; thus, he analyses the word in the first example as OImmun-s-an, though he has said -5 has the variant shape -sa before another clitic.

DAVID L. APPLEYARD

S6

m.sg. -[alta, f.sg. -[alxYta, m.pI. -[arxno, f.pI. -[alxniima.l 4 These are in effect bound forms of the independent pronouns. For instance in Chliha the corresponding independent pronouns are: xuta, xYita, xano, xanama, respectively. In Chliha, the definite marker may either remain on the NP head, or may move on to a qualifier: ya-bet gabt-ata 'half of the house' (lit. 'of house, its half), but ya-bar-ata kobar 'the celebration of the festival'.15 In Chliha, the full, independent form of the pronoun may also be used as a definite marker: krostyan xono basra xWet sat yatakso 'the Christians light [it] at twelve o'clock' (lit. 'Christians, they .. .'), or yatakso Ximo krastyan ro 'the ones who light [it] are Christians' (lit. 'they [who] light, they ... ').

The situation in Ge'ez

Although the purpose of this short paper is to examine definite marking in modem Ethiopian Semitic, it is not totally out of place briefly to review the situation in Ge(ez, as in many ways the origins of most of the various modem forms may already be seen there. Ge(ez, of course, does not have a specific definite marker as such. It does, however, have various mechanisms whereby the category +Definite can be signalled. The most frequently described applies to the indicating of a definite direct object by means of the construction V + object pron. suffix la- + N: ra7ina-hu lii7agzi7ana 'we have seen Our Lord, tagiibbar-a lii-ya7ati tabot 'thou shalt make that ark'. However, Dillmann (1907: 427) also remarks that the same construction is also occasionally found where context dictates that the NP is indefinite, i.e. -Definite: ra7oy-a lii-7anatti 7amat 'he saw a maidservant' . The same construction is used with definite noun complements or prepositions of nominal origin, N + possessive pron. suffix la- + N: lafla la-giis's'-u la-7adam 'upon the face of Adam', k'iidamiha la-t'abab 'the beginning of wisdom'. Another device used by Ge(ez is to use a 3rd person or determinative pronoun: wanii57a 7am-wa7atu 7abn zawa7atu boner 'and he took [one] of the stones of that place'. Particularly interesting, however, for the present discussion are those instances where the category definite seems to be marked by a 3rd person pronominal suffix added to the NP, in the Bible translation corresponding to a definite article in the Greek: 7asmii fds's'iimku miiwafali-hu 'for I have completed 14

The bracketed vowel ii occurs after consonant-final nouns.

15

The Chliha data is taken from Leslau (1950).

DEFINITE MARKERS

57

the days',16 ka17e-homu 7axaw 'the two brothers', or even more obviously naliimku nalmii wiikiimiiza nalm-u '1 dreamed a dream and thus is the dream' (lit. ' ... his dream'). Dillmann (1907: 426), from where the preceding examples are taken, remarks that such usage of the 3rd person suffixes is 'by no means rare'.

The origin ofdefinite markers in modern Ethiopian Semitic

General studies that discuss the question of the development of definite marking in the world's languages describe the commonest origin to be a demonstrative (Heine and Kuteva 2002: 109-111; Hopper and Traugott 1993: 8; Knimsky 1968: 241), as is, of course, the case in English and other European languages. Amongst Semitic languages, a demonstrative origin is also the usual explanation for the definite marker, or article, in Arabic, Hebrew and Aramaic (Voigt 1998 passim ).17 Within Ethiopian Semitic, however, this is patently only the case in Tigrinya, where 7atu, etc., is identical with the short form of the far demonstrative, as well as in Ge'ez, where, insofar as it has a distinct definite marking system, the 3rd person pronoun wa]atu, etc., also functions as a marker of definiteness as well as a demonstrative. Definite markers elsewhere in Ethiopian Semitic are either formally identical to 3rd person pronoun suffixes (possessive suffixes in Amharic-Argobba, Harari, Chaha and other Outer South Ethiopic languages, and certain categories of object or complement suffixes in Silt'i, Zway, Soddo, Goggot, Masqan, with traces elsewhere), or appear to be distinct from other morphemes (as in Tigre, Muher, Ezha and Gafat). Of course, the formal distinction between deictics or demonstratives and 3rd person pronouns is blurred in Semitic as in many other languages. What is typologically comparatively rare is the use of otherwise 3rd person possessive clitics as definite markers. Vycichl (1957: 167) does cite instances from Ancient Egyptian and Old Turkish, though as we have seen with a number of the examples quoted above, the interpretation of a given clitic as either a marker of definiteness, or as an anaphoric pronominal reference is not always easy to decide even in context. Whilst invariable Harari -zo is formally identical to the possessive suffix of the 3rd masc., this is most likely simply a clitic form of the 16

In the Greek, lit. 'my days' at T)fJ.£pm fJ.o'U.

17

But one variety of Modern Aramaic (Turoyo) has developed an article from the independent 3rd person pronoun (Jastrow 1997: 357).

DAVID L. APPLEYARD

58

independent pronoun, which is azzo, which also functions as a far demonstrative. As such, it is exactly paralleled by what is found in Chaha and other Outer South Ethiopic languages. Here, it will be recalled, a bound form of the 3rd person pronouns is used in definite marking function, though unlike in Harari, as a definite clitic these vary for gender and number in agreement with the head noun. Interestingly, Hetzron (1977: 59) suggests that the linking vowel -a- (fiiraz-ata 'the horse') in Chaha, etc., may be a trace of the nominal subordinator *zCi- > ya-, and as such would be a further parallel with Harari, where the z is presumably of the same origin. Amharic-Argobba -u, -wa/-itu does not, of course, coincide with this pattern. Not only is there no trace of any intervening subordinator, as the suffix -u proceeds directly from the earlier suffix -[hju, as in Ge'ez, Tigrinya, Tigre, etc., but the suffix shows gender (but not number) concord with its head. The Amharic variant -itu for the feminine may be analysed as comprising a gender-number neutral marker -u added to a feminine derivational element -it, still seen on such as andit 'one' (fern.), ogolit 'so-and-so' (fern.), arogit 'old ,woman', muSorrit 'bride', /og[gjit 'girl', etc., which thus would not originally have formed part of the definite marking morph. The variant -wa, which is the sole form in Argobba, presents a different problem. Ostensibly it looks like a combination of -u and the earlier, 'original' 3rd feminine suffix -a (still so in the inflection of the gerundive), to which Ge'ez -[hja, etc., may of course be compared. Vycichl (1957: 168), somewhat fancifully, attempts to derive Amharic -wa from earlier *-uha, where *-u is the Proto-Semitic nominative case suffix.I 8 Such an explanation would totally ignore the usually accepted history of Ethiopian Semitic. Whatever the origin of -wa, it remains anomalous in terms of what we see elsewhere in Ethiopian Semitic inasmuch as it is a definite marker with specifically feminine concord where the plural has no special marker: biire-w 'the ox' and biirewocc-u 'the oxen', but lam-wa 'the cow'.l9 In the same way that Amharic definite -u proceeds directly from the 3rd masc. pronoun suffix, so definite -i of Silt'i, Soddo, Masqan and some other South Ethiopic languages, seems to derive from what now occurs only as an object or complement suffix of the 3rd rnasc., added to verbs. In Soddo and other languages, -i/-y occurs as a conditioned variant of -u/w as the 'heavy' object suffix, i.e. occurring after originally long vowels, 18 19

Thence also m. -u < *-u+ha. This explanation is taken up by Lipinski (1997: 271). Soddo -iti and Silt'i -te may, however, be compared more or less directly with Amharic -itu.

DEFINITE MARKERS

59

and arising from dissimilation after a rounded vowel (see Hetzron 1977: 65-67). Here Hetzron argues that whilst the -y form was originally the marked one, the distributional features of the two allomorphs must have been reinterpreted and the labial or -w form became the marked one, such that the palatal or -y form was adopted as the unmarked variant, hence its wider occurrence. If this is so, then it would be the natural choice for the pronoun suffix used as a definite marker. Subsequently, in possessive function, these languages later innovated and now use forms deriving from the independent pronouns, leaving the suffix -i/-y as a 'relic' in definite marking function. If this line of argument is correct, then all the South Ethiopic languages with the exception of Gafat, Muher and Ezha employ either originally 3rd person pronoun suffixes as definite markers, or, as a later development, independent pronouns now used in the former function. The background to this usage can obviously be seen already in Ge'ez with the 'by no means rare' employment of pronoun suffixes as definite markers. The origin of Muher-Ezha -we remains, however, more problematical, particularly as the two languages belong in other respects to different internal groupings of South Ethiopic, though they are geographically adjacent. Gafat -sea] is, as Leslau (1956: 44) suggests, in origin almost certainly a particle of insistence with cognates in amongst others Chaha and, of course, Amharic -ss and Ge'ez -ssa. Lastly, Tigre la- is usually explained as originally a demonstrative (Leslau 1945: 183) and is compared with ,,,lli, etc., 'this' and lahay /lohay, etc., 'that'. Testen (1998: 176-177), however, remarks that Tigre la- 'bears a close resemblance to Arabic [a]l-', and further speculates whether, like the latter,20 it is an apparent reflex of a common Semitic particle *1- which lies behind asseverative and precative particles in Semitic, as well as the definite article. One might wonder, though, whether Tigre la- derives in some way from a reinterpretation of the construction type: pronoun suffix 10.- + N seen in Ge'ez. After all, Tigre la- adheres to its noun in a way more reminiscent of a preposition than a demonstrative clitic. Contrast, for instance, the sequence in nay lamudiryat sayam, cited above, with Tigrinya nay ,astamhari giiza. The functioning preposition 'to, for' in Tigre, however, continuing earlier 10.is 'to, for', just as ,,,b- continues earlier bii-. It is evident, therefore, that the origins of all the Modem Ethiopian Semitic definite markers (except Muher-Ezha -we ?) can be traced to the oldest form of the branch as exemplified by Ge'ez, but where no specific

-s

,,,tu

,,,I

20

He also links this to Ge'ez la- in la-7~mma- 'if, whether'.

60

DAVID L. APPLEYARD

definite marker had yet developed, or perhaps to state the situation better, where no single definite marker had fully developed.

REFERENCES Dillmann, A., 1907, Ethiopic Grammar. Second Edition, edited by Carl Bezold, translated, with additions by James A. Crichton, London: Williams and Norgate. Gutt, E.A., 1997, 'Concise grammar of Silt' e', in E.A. Gutt and Hussein Mohammed, Silt'e-Amharic-English Dictionary, pp. 895-957, Addis Ababa: Addis Ababa University Press. Heine, B. and Kuteva, T., 2002, World Lexicon of Grammaticalization, Cambridge: Cambridge University Press. Hetzron, R, 1977, The Gunnan-Gurage Languages, Napoli: Istituto Orientale di Napoli. Hopper, P.J., and Traugott, E.C., 1993, Grammaticalization, Cambridge: Cambridge University Press. Jastrow, 0., 1997, 'The Neo-Aramaic languages', in Hetzron, R (ed.), The Semitic Languages, London-New York: Routledge. Knimsk)T, J., 1968, 'Some ways of expressing the category of determinedness', Travaux Linguistiques de Prague 3, pp. 241-53. Leslau, W., 1941, Documents tigrigna (ethiopien septentrional). Grammaire et textes, Paris: Librairie C. Kincksieck. - - - - , 1945, 'Grammatical sketches in Tigre (North Ethiopic), dialect of Mensa', Journal of the American Oriental Society 65, pp. 164203. - - - - , 1950, Ethiopic Documents: Gurage, New York: The Viking Fund. - - - - , 1956, Etude descriptive et comparative du Gafat (Ethiopien meridional), Paris: Librairie C. Kincksieck. - - - , 1965, Ethiopians Speak. Studies in Cultural Background 1 Harari, Berkeley and Los Angeles: University of California Press. - - - - , 1981, Ethiopians Speak. Studies in Cultural Background. IV. Muher, Wiesbaden: Franz Steiner Verlag. Lipiilski, E., 1997, Semitic Languages. Outline of a Comparative Grammar, Leuven: Peeters. Raz, S., 1983, Tigre Grammar and Texts, Malibu: Undena Publications.

DEFINITE MARKERS

61

Tesfay Tewolde Yohannes, 2002, A Modern Grammar of Tigrinya, Rome: Tipografia U. Detti. Testen, D.D., 1998, Parallels in Semitic Linguistics, Leiden-BostonKoln: Brill. Voigt, R.M., 1998, 'Der Artikel im Semitischen', Journal of Semitic Studies 43, pp. 221-58. Vycichl, W., 1957, 'Trois notes de linguistique ambarique', Annales d'Ethiopie 2, pp. 167-76.

IS NEO-ARAMAIC A SEMITIC LANGUAGE? Simon Hopkins

Introduction In addition to his studies of individual texts and specific features in particular Semitic languages E.U. has throughout his career, with a characteristically broad sweep, given much thought to the wider implications of his researches. He has written a good deal about several broader aspects of linguistic enquiry. He has always been interested not merely in the grammatical mechanisms and lexical resources of the Semitic languages themselves, but also in the speakers who spoke, and continue to speak them. Subjects such as mutual intelligibility between Semitic languages, bilingualism and trilingualism in the ancient Near East, degrees of kinship between related languages and the classification of the Semitic family, modes of thought and their linguistic expression, translatability etc. are recurring themes in his books and in the four volumes of his collected articles. A fine example is the essay 'Is Biblical Hebrew a language?' Ullendorff (1971), which later gave its title to the first volume ofE.U.'s collected papers, in which it is reprinted as the first item. 1

The Semitic Sprachtypus If there is such a thing as the basic character of Semitic (i 73) or a Semitic Sprachtypus (ii 228), we should be able to say in what this essential unity of the Semitic tongues (i 187) consists and to describe it in precise concrete, rather than vague impressionistic terms. Such a description, which remains a desideratum in the field (iv 107), should hold good for all Semitic languages, not just for a convenient selection of them. E.U.'s treatment of this question in a celebrated and thought-provoking article The four volumes of collected papers are: Ullendorff (1977), (1987), (1990) and (1995a). These are referred to in the following as i, ii, Hi and iv respectively.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

63

'What is a Semitic language? (A problem of linguistic identification), (Ullendorff (1958), reprinted in i 155-64),2 a study among the most widely read items of his oeuvre, was drawn forth, inter alia, by a somewhat loose statement published shortly beforehand to the effect that the bond binding all (sic) Semitic languages together is so tight that the family likeness can be recognized at first glance.3 In reaction to this claim E.U. pointed out (i 156 and similarly i 31; Ullendorff 1995b: 285), very justifiably, that while this may well be so for the classical Semitic languages,4 it hardly applies to the rest. 5 It is one thing to recognize without further ado the patent Semitic character of languages such as (Classical) Arabic, Ethiopic or Ugaritic; the situation would be very different with regard to Modern Hebrew, Amharic or Soqotri. And the relationship between Gurage and Ugaritic,6 or Mandaean and Akkadian can scarcely be called obvious. The description (as opposed to recognition) oflinguistic identity and the portrayal of family resemblance between languages in precise terms is a notoriously difficult task (i 72, 155/6; ii 225). Different scholars will reach different conclusions, depending upon the questions asked and the approaches adopted. E.U. has very properly pointed out the importance of keeping diachrony and synchrony well apart in this exercise (i 69, 156, 162; ii 226); those with a retrospective, diachronic outlook are, of course, likely to be more impressed by the inherited Semitic stock of a given modern Semitic language than those who approach the question in more synchronic terms, unburdened by genetic considerations or historical knowledge. A scholar of the former disposition will be particularly interested in etymology, shared isoglosses etc. and will deliver his verdict 2 3

In the reprint the footnotes to the last page of the article are missing. Spuler (1953: 3): 'dass sich alle semitischen Sprachen so nahestehen, dass ihre Verwandschaft auf den ersten Blick zu erkennen ist ... '. It should, in fairnesss to Spuler, be noticed that Amharic is explicitly excluded from his survey as no longer Semitic ('als nicht mehr den eigentlichen semitischen Typ repriisentierend ').

4

At least in their written forms; the spoken versions may have caused some difficulty. Cf. Fleischer (1861: 381, = idem 1885-1888: iii, 468): 'Dass wir morgenllindische Sprachen gewohnlich mehr mit dem Auge filr das Auge als mit dem Ohre filr die Zunge lemen, ist nicht zu lindern, hat aber manche Nachteile'.

5

Cf. the opening sentence of H.J. Polotsky's essay 'Semitics' (Polotsky 1964a: 99): 'The relationship between the older Semitic languages is such that no modem professional training is needed to perceive it' . In the parallel passage ii 226 Ugaritic has been replaced by Phoenician.

6

64

SIMON HOPKINS

accordingly; a scholar for whom the historical origin of the linguistic material is of less importance will naturally pay more attention to its present appearance and behaviour. The difference in their results arises from the fact that they are comparing different things, or, rather, looking at the same things from two different points of view: the one will concentrate upon the fact of Semiticity (kinship), the other upon its nature (typology). For the descriptive experiment to be valid, comparables must be compared in a comparable fashion. Kinship and typology co-exist and overlap, but they are two different things.7 In many modem Semitic languages the traditional components, often in vastly altered forms, frequently behave in ways unfamiliar to the orthodox Semitist brought up on the classical structures. It is precisely to this split personality that a language such as Amharic owes what has felicitously been called its 'charme piquant' Cii 199). This dual physiognomy, however, also prompts questions about the degree of Semiticity of these Neo-Semitic8 languages. In his discussion of the issue E. U. has drawn attention to an interesting theoretical point, viz. that Semiticity, being relative, can be lost and can be acquired. 9 Thus there may well be languages which are no longer Semitic just as there may be others that are not yet Semitic (i 162), and we may have to face losing some members of the group or also admitting new ones (iv 110). In other words, our subject may have to come to terms with concepts such as postor ex-Semitic,lO pre-Semitic or pseudo-Semitic languages. There are certain hints in the older literature implying relative, i.e. lost, acquired or impaired Semiticity-e.g. d' Abbadie (below n.30) speaks of 'langues sous-semitiques', Ewald of 'wirklich semitisch' character;l1 Renan uses the expression 'a demi semitique'12 and Hommel distinguishes between

7 8

9

10 11 12

Cf. e.g. Pedersen (1931:100). The tenn 'langues neo-semitiques' was casually used by Renan in 1855 (below n.42). As a technical term it was, I believe, introduced by H. Ewald in the 7th edition of his Ausfiihrliches lehrbuch der hebrtiischen sprache (Ewald 1863: 37/8, §6e): 'Wir konnen sie die Neusemitischen sprachen nennen'. The immediate source quoted is J. Fiick, but the idea originates with N.S. Trubetskoy (1939: 81-89) > Arens (1955: 498): 'Somit kann eine Sprache aufhoren, indogermanisch zu sein, und umgekehrt kann eine Sprache indogermanisch werden.' C£ above n.3 on Arnharic. Ewald (1827: 166 n.7): 'solange die semitische Sprache wirklich semitisch war'. Renan (1855 [I"t edition]: 68 = 1863 [3rd edition]: 75).

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

65

'ursemitisch' and 'vorsemitisch'13_but as far as I know the frontier of entry into or exit from Semitic identity (iv 109) has not received sustained discussion. Something along these lines seems to be adumbrated in Garbini's notion (ii 225) that the Hamitic languages may be characterized as those African languages which have somehow acquired certain Semitic properties. The question of Semiticity naturally implies classification; some languages are likely to be more Semitic than others, or to reveal their Semiticity in different ways. The task of linguistic classification is a delicate operation at the best of times. In Semitic studies this already difficult problem was rendered especially acute in the middle of the 19th century by the growing acquaintance with the modem Neo-Semitic languages on the one hand, and the decipherment of ancient languages such as Akkadian and South Arabian on the other. These languages did not fit neatly into the scheme, and in some cases even seemed not to belong to it at all. Some are clearly Semitic by genealogy, yet at the same time contain many features, even very basic features, which are not to be found in the classical Semitic languages which had been known hitherto, viz. Hebrew, Aramaic, Arabic and Ethiopic. This raises questions about the extent to which such languages participate in the essential Semitic character, the typological 14 linguistic Semitism of which we have been speaking. Is the Semitic genius (i 182) displayed by the Neo-Semitic languages such that it too may be recognized at first glance? In NeoSemitic much that is familiar in the classical languages has either disappeared altogether or been transformed beyond immediate recognition-this applies to all divisions of language: phonology, morphology, syntax and vocabulary. Once raised, the question of Semiticity and the classification of its representatives proved highly refractory and has to this day not received a satisfactory solution. There is not only much diversity of judgment among scholars, but even polarity of opinion. While there has long been general agreement about what languages to compare (kinship), there is much less agreement on how to go about it and upon what basis to classify them (typology). The subjective element here is often extreme, so that two different scholars may look at the same set of linguistic facts and reach not just 13

Hommel (1883: 51).

14

I may refer here to a remark of H.J. Polotsky in UUendorff (1992: no.106 [1976]): ' ... dass die "typologische" Sprachwissenschaft die eigentlich wahre ist.' For the syntactical typology of Semitic languages cf. Goldenberg (1998: 138-147).

66

SIMON HOPKINS

different but diametrically opposed conclusions. One example from the realm of the classical Semitic languages (without Neo-Semitic complications) will suffice. In Ewald's Abhandlung uber die geschichtliche folge der Semitischen sprachen (Ewald 1871) Aramaic appears as the first and oldest of the five historical stages ('stufen') identified by the author (e.g. p. 53), i.e. the closest to Proto-Semitic, while Arabic appears as the fifth and latest (p. 11). For Hommel, on the other hand, Die semitischen Viilker und Sprachen i (Hommel 1883) 16, Arabic by virtue of the 'Althertiimlichkeit und Urspriinglichkeit' of its forms takes pride of place and Aramaic as 'am meisten sich vom ursemitischen entfemend' is relegated to final position. When two such great scholars as Ewald and Hommel, both of whom were familiar with the material at first hand, can reach such fundamentally conflicting results, one cannot but wonder about the theoretical basis underlying the whole classificatory operation.1 5 To a not inconsiderable extent the situation is similar even today; we do not have a classification of the Semitic languages of proven validity, but rather a convenient broad general concensus which somehow feeds upon itself (cf. ii 232a). The history of research in this field is replete with oddities. It is after all (as has been very well said) comparatively easy to lump together any two languages and show the special affinities they bear (ii 221; iv 110). E.U. has accordingly steered sceptically clear of overall classification schemes of the Semitic family, or of parts of it (i 150, 152), and has even become bored by the topic (iii 86, 179).1 6 The question of linguistic identity, however, of Semiticity itself, has continued to hold particular fascination for him. What is it about a Semitic language that makes it Semitic? He has pondered this topic long and hard, and has addressed the matter in greater or lesser detail on a number of occasions. But E.U. has never, I think, come to regard any Neo-Semitic language as so far removed from its ancestral model as to warrant the label post- or ex-Semitic; these are for him theoretical categories only, not yet reached in practice. In all Semitic languages the one, crucial, area, in which the old Semitic character has remained fairly consistently intact is the morphology, especially that of the verb. The systeme fermi (Meillet) reflected in the remarkably constant morphological structure of the 15

ef. Polotsky (1964a: 108). The case of Ewald and Hommel is a striking example of what Polotsky (ibid., 111) has called 'the individualism of linguistic reasoning' .

16

An exception is the Hebrew article on the classification ofUgaritic reprinted in i 114-18. This is perhaps the 'sin' referred to in iv 110.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

67

Semitic verbal organization (i 83b) in all its ancient and modem manifestations has for E.U. in every case served to uphold Semitic identity, a point which he has stressed repeatedly in his writings, e.g. in Ullendorff (1995b: 287). Closely related to linguistic identity and recognizability is the question of mutual intelligibility. In the case of unrelated languages, such as the Philistine mother-tongue of Delilah and the native Semitic speech of Samson (i 38, 4112), mutual incomprehensibility must have been absolute; communication can only have taken place if one or both parties were bilingual. But with related languages and dialects, mutual intelligibility is relative. To what degree were speakers of different (Northwest) Semitic languages in the Biblical period able to understand one another? E.U. has concluded on the basis of2 Kg. 18:26 that Hebrew and Aramaic cannot have been mutually intelligible (i 35, 40),17 whereas Hebrew and Phoenician probably were (i 45). And as Moabite is but a variety of Hebrew, it stands to reason that Hebrews and Moabites could communicate without undue difficulty (i 35/6, 44/5).

Modern Hebrew

In addition to the coeval mutual intelligibility, or unintelligibility, between Biblical Hebrew and its closest Semitic neighbours, E.U. has wondered about the communicative relationship obtaining between different chronological layers of Hebrew. How would Deborah have managed with the Hebrew of Daniel some thousand years later (i 7) and how would King David have found conversation with Queen Esther (i 8)? Would the intervening changes have impeded comprehension? Such speculations led to a diverting piece entitled 'Could Isaiah understand the Ha'arets newspaper?' (Ullendorff 1994, reprinted in iv 131-45). The assumption expressly underlying this essay is that the resurrected Isaiah's exposure to the Israeli newspaper would be visual only, for were Ha'arets to be read out to him in modem pronunciation, it is very doubtful that he would recognize the language as Hebrew at all (iv 131 and cf. i 72). E.U. has more than once referred to Modem = Israeli Hebrew as a Neo-Semitic language in which some of the characteristics of the classical Semitic tongues are no longer always obvious. His most concentrated utterances on this topic are to be found in his reviews of two 17

ef. ii 44 n.3 on the assumed ability of speakers of Syriac to 'cope' with Hebrew as a related Semitic language.

68

SIMON HOPKINS

books by H.B. Rosen, the pioneer of the study of Modem Hebrew (iii 152a). Reflections on Rosen's 1J7!Zl n'i:J17:1 'Our Hebrew' (1955-56) can be found in 'Modem Hebrew as a subject of linguistic investigation' (Ullendorff 1957, reprinted in i 68-80), and detailed notes on the same author's Contemporary Hebrew (1977) appeared in Ullendorff (1979, reprinted in iii 150-155). These studies were written in the wake of a famous pessimistic 18 (and often misused) remark by G. Bergstrasser, who long before in 1928 had opined that Modem Hebrew in Palestine might turn out to be a Hebrew 'das in Wirklichkeit eine europaische Sprache in durchsichtiger hebraischer Verkleidung ist, mit gemeineuropaischen Ziigen und einzelsprachlichen Besonderheiten, aber nur ganz ausserlich hebraischem Charakter'.1 9 Other scholars have expressed themselves in similar vein, and it is difficult to say that subsequent developments have proved them entirely wrong.20 A like judgment, incidentally, applies in large part to modem journalistic Arabic. 21 When we look at the structure of modem Israeli Hebrew we cannot but agree that conspicuous changes have taken place, many of them indeed under the influence of what has aptly been called Standard Average European. 22 Developments such as the widespread loss of laryngals and emphatics (i 73), the toleration of initial consonant clusters (i 77, but cf. ii 278), the increasing tendency of the language towards an analytic structure (ii 206), the shift in word order from the old Semitic VSO to SVO and replacement of co-ordination by subordination (iv 133) have undeniably affected the Semitism of Modem Hebrew, giving it in certain respects an un-Hebraic, (Indo-)Europeanized aspect (i 72, 162; iv 109, 141; Ullendorff 1995b: 287), so much so that the style, syntax and 18

Somewhat too pessimistic according to Brockelmann (1932: 178).

19

Bergstrasser (1928: 47), quoted iv 151 b.

20

E.g. Plessner (1931: col. 805 n.4) spoke of the 'Entsemitisierung' of Modern Hebrew and ibid (col. 806) mentions 'Die Diskrepanz zwischen der gltnzlich europaisierten Umgangssprache und der noch relativ semitischen Schriftsprache'; Baumgartner (1940-41: 615, = idem 1959: 235) also uses the tenn 'Entsemitisierung' in this connection; Arberry (1943: 8): 'omitting Hebrew as being in its classical fonn a dead language and in its modern revival in reality a variety of western idiom' . H.L. Ginsberg was wont to refer to Modern Hebrew as 'Yidrew' (J. Blau).

21

Fleisch (1974: 34): '11 en sort une prose arevetement arabe, mais pensee dans la langue etrangere, tres loin de l'esprit de la 'Arabiyya.' Tenn coined by B.L. Whorf; see Hopkins (1999a: 88).

22

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

69

modes of expression employed by Ha'arets newspaper may differ little from those of The Times (i 80). In contrast to these changes in phonology and syntax, however, morphology has fared somewhat better (i 72), especially the morphology of the verb, which, still governed firmly by the triconsonantal root (i 73), has been maintained well-nigh intact. For E.U. it is the essentially unimpaired verbal system of Modem Hebrew which more than anything else serves to uphold the Semitic identification of the language as a whole (i 162). This conclusion is matched by the basically Semitic vocabulary of Modem Hebrew (i 73). Since the pioneering work of Rosen a number of scholars have for one reason or another paid attention to the ambivalent typological physiognomy, the 'un-Hebrewness' (Rosen, quoted iii 152a) of Modem Hebrew. Some of these studies, if only by virtue of their titles, deserve mention here, e.g. S. Wild, '1st Ivrit eine semitische Sprache?' ,23 which deals directly with points raised by E.U.; C. Rabin, 'Is Modem Hebrew still a Semitic language?';24 P. Wexler, The Schizoid Nature of Modern Hebrew: A Slavic Language in Search of a Semitic Past,25 and G. Goldenberg, ';'l'n n'/jwl'W':::l n'i:::ll1:1' .26

Ethiopian Neo-Semitic

The character of Ethiopian Neo-Semitic is, of course, very well represented in E.u.'s published work, not only in the four volumes of collected articles, but also in the books The Semitic Languages of Ethiopia (1955) [SEL), The Ethiopians (1960, here quoted according to the 3rd edition 1973), An Amharic Chrestomathy (1965) and A Tigrinya (Tdgrdnna) Chrestomathy (1985). Ge'ez, the ancestor of the modem Semitic languages of Ethiopia is by no means remote from the Semitic type (SEL 11), but its descendants, i.e. Modem Ethiopic (SEL x),27 show very great divergences indeed from 23 24 25 26 27

Wild (1977). Rabin (1992). Wexler (1990). Goldenberg (1996). The curiosity by U. Oman, 'Hebrew is not a Jewish language' (Oman 1985), deals with a rather different aspect of the identity of Modern Hebrew. 'Diese stark vom Hamitischen beeinflussten, so zu sagen "schwarzsemitischen" Sprachen' (Noldeke to Goldziher 24.11.1904 in Simon 1986: 272).

70

SIMON HOPKlNS

what may be called typically Semitic. Amharic in particular shows a marked lack of Semitic conservatism (ii 207; Amh. Chr. 8); Tigrinya is somewhat less progressive (Ti'ia Chr. 15). Compared with Ge'ez the phonetic character of Amharic has undergone profound changes; laryngals have been reduced (thus making many formerly triconsonantal roots biconsonantal) and palatalization has become a marked characteristic of the sound pattern (SEL 24; Amh. Chr. 9; ii 209/10). But it is in its syntax that Amharic, and to only a slightly lesser extent Tigrinya (SEL 22; Ethiopians 121; Ti'ia Chr. 15),28 has developed furthest from classical Semitic models (SEL x, 223; Ethiopians 120; Amh. Chr. 8; i 161; iii 123/4), very probably under Cushitic influence (SEL 10, 15). It may be said that Amharic syntax has, in fact, completely lost all Semitic conceptions (SEL 24). These developments in phonology and syntax have given to Amharic and other Modem Ethiopic tongues a sound and an appearance that seem strange and unfamiliar to the orthodox Semitist (Amh. Chr. 9). These languages were hence admitted into the Semitic fold only gradually and hesitantly. d'Abbadie, having observed in 1839 that certain modem Ethiopian languages, including Amharic, derive from the Semitic family yet are without many of the typical features,29 distinguished a few years later in a pioneering sketch of the linguistic situation in Ethiopia between 'langues presumees semitiques' (Tigrinya, Tigre) and 'langues soussemitiques' (Amharic, Gurage etc.).3o Munzinger, unable to decide what to do with a language Semitic in certain material ingredients but unSemitic in syntax and spirit, in 1865 left the classification of Amharic open.3 I Meanwhile, Gesenius, equipped with the necessary knowledge and method, had already in 1819 arrived at a clearly reasoned recognition of

28

29 30 31

In the opinion of Polotsky (1987: 238) the 'involved' character of Tigrinya syntax is here a little overplayed. E.U.'s unrepentant response can be found in Ullendorff (1988: 201). d'Abbadie (1839: 365): 'derive immediatement de la famille semitique, sans en avoir cependant tous les caracteres'. d'Abbadie (1843: 103/4). Renan (1858 [2nd edition]: 331 = 1863 [3rd edition]: 337) described these Ethiopian 'langues sous-semitiques' as a 'famille batarde' [the passage is absent from the 1st edition of 1855]. W. Munzinger (1865: iii) in the preface to his Vocabulaire de la langue Tigre appended to Dillmann's Lexicon Linguae Aethiopicae: 'en un mot, elle a la forme et le mot semitique sans en avoir la langue et surtout sans son esprit' .

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

71

the Amharic situation. His admirable assessment is well worth giving in detail (cf. Amh. Chr. 3 n. 1): 'Die Sprache ist nichts als ein ... entarteter semitischer DiaJekt .... Sowohl der lexicalische Sprachvorrath, als der grammatische Bau, und vorziiglich letzterer (was entscheidend ist) haben vollig semitischen Charakter, welches Adelung ... und Vater ... verneint haben, indem sie sie eine nichtsemitische, hOchstens gemischte Sprache, nennen. Sie enthlilt namlich nicht etwa eine fremdartige Grammatik ... sondern ... bewegen sich alle ihre Bildungen ausschlieBlich in dem Kreise der semitischen Grammatik, deren Eigenthiimlichkeiten iiberall nur zu deutlich durchscheinen .... man muss sie vielmehr eine semitische ... Sprache nennen. '32 In comparison to the classical Semitic languages, Gesenius here described Amharic as 'entartet'.3 3 Indeed, it was later to be Noldeke's opinion that no Semitic language is further removed from the old Semitic type than Amharic,34 a view categorically espoused by Bergstriisser: 'so ist das Amharische diejenige semitische Sprache, die am wenigsten vom altsemitischen Charakter besitzt'.3 5 For this reason Amharic was omitted from the study of Spuler, mentioned above n.3. The 'Entartung' of Amharic notwithstanding, Gesenius was in no doubt whatever as to the essential Semiticity of the language, an identity which certain of his predecessors had denied. The important thing to notice here is the clear-sighted way by which he reached his verdict. To Gesenius the significance above all else of the Semitic morphological core was absolutely clear; the systeme ferme of Amharic was his principal diagnostic instrument ('was entscheidend ist'). This is a crucial point, for it has since become a generally recognized principle of historical linguistics that morphology carries more classificatory weight and is of greater evidential value than other areas of the linguistic structure.3 6 As 32 33

34 35 36

Gesenius (1819: 355b). See further below n.44 Gesenius's wording is the probable source of'das ... entartete und gemischte Amharische' of Bottcher (1866: 6) and the 'entarteter semit. Dialekt' mentioned by 1. Goldziher in the draft of an unpublished youthful (c. 1868-9) Encyclopiidie der semi!. Philologie preserved in the library of the Hungarian Academy of Sciences, MS Goldziher-gyiijtemeny 105. 'Entartet' was applied to Amharic also by Philippi (1871: 105). Noldeke (1899: 73, 1904: 3). Bergstrasser (1928: 112). Thus e.g. explicitly Polotsky in a note to his essay 'Egyptian' (1964b: 360 n.lO).

SIMON HOPKINS

72

we saw in the case of Modem Hebrew, the morphology of Neo-Ethiopic too has proved much more resistant to change than the greatly transformed phonology and syntax. The method used by Gesenius and the conclusion he reached foreshadow those ofE.V., whose approach from the ensemble of Semitics (e.g. i 182; ii 197; iv 101)37 and concentration upon morphological criteria have led him to a very similar result. The Semitic Ethiopian languages display an essential homogeneity (SEL 224), a homogeneity within which none has yet (sic) lost its essentially Semitic character, its original identity (Ethiopians 112; ii 197; cf. i 182). And with regard to the question as to what extent Amharic itself can still (iii 86) be regarded as Semitic, E.U. has often stated his belief that the systeme ferme of Amharic, Le. the Semitism of its morphology, above all of its conjugation (ii 199),38 is of sufficient weight to uphold its Semitic identity (ii 208; Amh. Chr. 9). Indeed, the verbal system alone is sufficient to mark Amharic as a Semitic tongue (iii 86). The lexicon (Hi 86ff.) also has something to contribute to the issue. The surmise that of the stock of the Amharic lexicon at least 50% has a safe Semitic basis (ii 210; Amh. Chr. 10-the percentages in iii 89 are far lower) has turned out to be a considerable underestimation (Too Chr. 16). Again, then, with Amharic as with Modem Hebrew, the Semitic identity of the morphology is matched by the preponderantly Semitic basis of the lexicon.

Neo-Aramaic In the light of these reflections, based entirely on E.V.'s work, it occurs to one to bring Neo-Aramaic, especially North Eastern Neo-Aramaic (NENA), into the discussion of the question 'What is a Semitic language?' In fact, E.V. has himself in passing mentioned Modem Syriac in this connection (i 161, 187). When comparing the modem dialects of Aramaic with the classical types from which they sprang we may legitimately wonder to what extent Renan, having noted the alleged ease with which Neo-Aramaic-speaking Nestorians were able to learn Biblical 37

The value of this approach does not, of course, apply only to Neo-Ethiopic: i 31, 150. Indeed, Gesenius himself had already used it in his classic Versuch (Gesenius: 1810) to demonstrate the Arabic affinities of Maltese.

38

er. HJ. Polotsky,

'(la syntaxe de l'amharique reste assujettie au) semitisme de sa morphologie, surtout de sa conjugaison' in the passage quoted Amh. Chr. 8.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

73

Hebrew, was justified in saying 'tant il est vrai que, meme dans leurs rameaux les plus ecartes, les langues semitiques conservent toujours le sceau immuable de leur unite')9 Renan actually went further than this. He believed that the Semitic languages, solidly imprisoned within their metallic rigidity ('roideur metallique'), were incapable of any real development;40 after the 14th century (sic), with the decline ofSyriac and Ge'ez, 'les langues semitiques n'ont plus d'histoire'.41 Accordingly, he denied the existence of Neo-Semitic languages altogether with the simple ex cathedra statement: 'Il n'y a pas de langues neo-semitiques. '42 In approaching the NENA dialects, the orthodox Semitist will immediately encounter much for which his earlier studies had not prepared him. Indeed, in some respects the situation is here more acute even than in Amharic, for in NENA very deep inroads have been made into the systeme ferme of the verb as well. The very examples yaqtulu : yuqtalu, qatala : qutila which were used to illustrate the paradigmatic system of the Semitic verb (ii 228) do not in fact occur in NENA at all. In these languages the old Semitic prefix-conjugation (imperfect yaqtulu) and suffix-conjugation (perfect qatala) have completely disappeared and been replaced by forms of participial origin. Now this is fundamentally different from the situation in Neo-Semitic languages such as Modem Hebrew and Amharic, where profound changes occur in phonology, syntax and vocabulary, though hardly ever in the morphology of the verb (iv 109); in Modem Hebrew and Amharic the yaqtulu and qatala forms have their direct modem continuations. But not only are the very ingredients of the NENA verbal system highly metamorphosed, their comportment also is rather far removed from older Semitic norms. While Modem Hebrew has undergone much European influence and Amharic has been affected by the impact of Cushitic, the verbal systems of the various NENA languages have an undeniably Iranian appearance. 43 For this and other reasons these Neo-Semitic languages too have been described as 'entartet'; J. Friedrich devoted an article to this question entitled 'Das Neusyrische als Typus einer entarteten semitischen Sprache'44 and similar terms have been applied by other scholars, 39 40 41 42 43 44

Renan (1855 [1'1 edition]: 249 = 1863 [3,d edition]: 269). Ibid. 390 [= 3rd edition 416] Ibid. 389 [= 3,d edition 415] Ibid. 403 [= 3rd edition 429] Kapeliuk. (1996: 59-70), idem (2004: 182ff.) and below n.48. Friedrich (1962). This scholar had earlier (Friedrich 1952: 156b) referred to Egyptian as 'entartete' Semitic, and earlier still (Friedrich 1925: co1.808) to

74

SIMON HOPKINS

without, however, necessarily implying loss of Semiticity: 'gesunken',45 'degeneriert',46 'aufgelost'47 or 'entfremdet'.48 As in the case of Amharic, the 'Entartung' of NENA is so advanced that its Semitic origin has occasionally not been recognized at all, and certainly not at first glance. 49 And the diversity within the group itself is so pronounced that speakers of these dialects do not necessarily recognize at once other NENA dialects as even being related to their own. 50 In order that the discussion not remain on too abstract a level, it might be useful to look at a short NENA text. The following anecdote is in the (Jewish) Neo-Aramaic dialect of Kerend in Iranian Kurdistan. It is a translation, rendered extempore by the speaker from the Persian original in G. Rosen, Elementa Persica, neu bearbeitet von F. Rosen (Leipzig 1915) no. 35. 51

xa-gora kiixata kaluwii. xa-ntiSa begana qarwa-e 0 hiye-u yatiw w kiixatake xazyale. 0 gora ga qiixatake klUle: xa-na5a begana-e ahmaq ga la ili ytiway wa ay kiixata h{te qarela. m-ay sabab ana la-keli hie rt1z-e kaluna. 0 gora begana mIre: tama al ana ahmaq qalamdad kolet? tama kiixatox la-kalweta? ana kiixata at la-qaryali. mIre baqef. agar kiixati laqaryal6x, dax kaet, dax ba at ma'/um xiray ke ana asxa mfri, ke at ahmaqyet?

45 46 47 48

Hittite as 'entartete' Indo-European. The tenu 'entartet' has a long history in Semitic studies-we have already seen it applied to Amharic, above n.33. 'Entartung' is an important concept in Olshausen (1861, e.g. vii, ix, 3,4,7 ult., 106). Bottcher (1866: 6) calls Maltese 'ein entarteter Sprossling' of colloquial Arabic and NOldeke (1875: xxiv) speaks (in a somewhat different sense) of the 'sprachliche Entartung' of late Mandaic. With reference to Neo-Aramaic the word was first used by ROdiger (1839: 84), this being the immediate source of its occurrence in Ritter's Erdkunde (Ritter 1840: 682). Goldziher in the draft Encyc/opadie der semit. Philoiogie mentioned above n.33. G. Jacob, quoted by Polotsky (1979: 204 n.3). Philippi (1871: 122). Macuch (1962: col. 121), where also the tenus 'iranisiert' and 'europaisiert' are used.

49 50 51

Details in Hopkins (1999b: 319/20). Hopkins (l999b: 322). The whole collection, and other material in the same dialect, has been prepared for publication by the present writer.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

75

A man was writing a letter. A stranger came close by him and sat down and saw the letter. The man wrote in the letter 'An idiot stranger is sitting at my hand and is reading this letter. For this reason I cannot write any secret(s).' The stranger said 'Why do you consider me an idiot? Why don't you write your letter? I haven't read your letter.' He said to him 'If you haven't read my letter, how do you know, how has it become known to you that I said so, (Le.) that you are an idiot?' I am not sure that the Semitic ancestry of this language would be evident at first glance to somebody approaching such a thing for the first time from a background of classical Semitic studies. Apart from the fact that much of the vocabulary here is not Aramaic, viz. Persian (rru:,52 qalamdiid), Kurdish (begana), loaned via these languages from Arabic (ahmaq, sabab, ma'lum) or of Iranian origin yet adapted to Aramaic morphology (k/qilxa:ta), the native Aramaic stock is far from transparent. While xa-gora 'a man' = ~1~~ (i)0*,53 nasa 'a person' = ~~i(~)* or yatiw 'he sat' = ::l't:l;* may be obvious enough, ili 'my hand' = ('i') ",~* and dax 'how' = T~-i* are less so. Hiye 'he came' = 'Ij~*, kaluna 'I write' = ~J~-::lIj~*, mire 'he said' = i'I'?-"~(~)*, kolet 'you do' = n~-i~~-:;'* and kaet 'you know' = n~-ll'>:;'* cannot, I think, be said to be immediately selfevident. Nor is it clear at once that the root of xiray 'it has become' is Aramaic 'in 'to (re)turn'. The preterite verbs with incorporated objects54 xqzyale 'he saw it' = i'I'?-~-'Tn*, la-qqryalox 'you haven't read it' = -"P ~"* 1"-~ are unknown in older Aramaic. As far as etymology and kinship are concerned we are dealing with objective facts; there is no room for discussion regarding the genetic affinities of gora, nasa etc. and nobody would dispute their Semitic origin. But when we enter the vaguer and more subjective area of typology we are certainly entitled to wonder how far such an estranged Semitic language as this can still be classified as Semitic. Renan's 'sceau immuable' of Semitic unity indeed seems somehow rather far off. Even in this very short text there is much that is syntactically unheard of in older forms of Aramaic, e.g. the governed independent pronouns ql ana 'me (acc.)' and kiixata at 'the letter of you(rs)' (alternating with the Semitically more familiar kiixatox with pronominal suffix); the older 52

53 54

Here Persian riiz J1.), but equally often pronounced in the dialect in Kurdish fashion as riiz. Gora has been mentioned by E.U. in illustration of the 'complete dissolution of labials' in Neo-Semitic: SEL 106; i 161, 187. Polotsky (1994).

76

SIMON HOPKlNS

Semitic languages do not behave in this way, but in Neo-Semitic languages such constructions are rather widespread,55 in the present case under the impact of Iranian syntax, especially the eziife. The entirely Aramaic components of Me qare1a 'he is reading it' = i'l7-l(1p' i'l7-n'l( lit. 'there is to him [= he has] he reads it' likewise behave here in Iranian fashion, imitating precisely the Persian 'periphrase progressive' based upon auxiliary 'to have'.56 Reflected here is Persian ~IH .l.)b, which does not, however, occur in the original story in Elementa Persica, showing just how natural a part of Kerendi Neo-Aramaic this new continuous tense is. These dialects of Iranian Kurdistan and vicinity have lost the old Aramaic root m$)' and express the notion 'to be able' by a periphrasis based upon 'to come': 'it comes to him' = 'he can';57 this is what lies behind ana la-keti 'I cannot' = '?-'lJl$-~ l(? l(~l(*. qalamdiid kolet 'you consider (as)' is composed of Persian qalamdad and Aramaic kolet < '1::l11, but the syntax is entirely according to the Persian compound verb (j.lfi .l1.l.A.l! 'to consider (sth.) as (sth.),. E. U. has declared that we can usually recognize a Semitic language when we meet one (i 156; ii 225/6). This will apply even to extreme ca'ses such as NENA, provided that we know what to look for and how to find it. An examination of (North Eastern) Neo-Aramaic in the light of what E.U. has repeatedly advocated, viz. an approach from the ensemble of Semitics (above n. 37) with attention directed firmly upon the relatively impervious systeme ferme of the basic morphological material, will surely lead us to the conclusion that Neo-Aramaic, all Neo-Aramaic, is as much a part of the Semitic family as Modern Hebrew, Amharic or any other of the Neo-Semitic languages. The technical term 'Semitic' has since its inception over two hundred years ago always implied genetic linguistic relationship rather than typological similarity; to say that a language is Semitic has traditionally been to affirm the Semitic genealogy of its main constituent parts, especially its basic morphological parts. If we retain the term 'Semitic' in this its traditional (and, I think, only meaningful, verifiable) sense, the question posed as the title of the present article will be found to be redundant, for the answer is already contained in the appellation 'Neo-Aramaic'. In determining the Semiticity of a given language a certain amount of attention should also be paid to the lexicon, for inherited basic vocabulary, as opposed to adventitious loanwords, must surely also play a 55 56 57

Brockelmann (1908-13, ii, §174.) Lazard (1957: §153). Khan (2004: 290).

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

77

part in identification. 58 At the risk of banality, it is worth reminding ourselves that a language in which 'day' and 'night' appear as ywm and lyl is likely to be Semitic, and a new inscription unearthed on, say, Syrian soil where 'x son of x' takes the form x br x is on the strength of the word br alone very probably Aramaic. Beyond this, each word in the grammatical mechanism, i.e. the systeme ferme, of a language has not only a morphological shape but also a lexical meaning, and it is hardly possible to attach weight to the one while neglecting the other. Thus in the NENA languages the presence on a huge scale of casual, unadapted loanwords such as Kurdish begiina 'stranger' and Persian < Arabic ahmaq 'idiot' in the text above prove nothing whatever about the identity of the language in which they happen to occur. Against this, however, the marriage of Semitic morphology and Semitic lexicon according to precisely statable morphophonemic rules in words expressing such basic grammatical and semantic relations as e.g. hfye 'he came' ('MN), kalUna 'I write' (:m~), mIre 'he said' (iON), kolet 'you do' ('::ll1), kaet 'you know' (11"), xazy le 'he saw it' ('Tn), la-qaryiilox 'you haven't read it' (Nip) proves a very great deal. In summing up this issue, we shall find ourselves, as usual, in agreement with Noldeke, who in one of his earliest masterpieces remarked that with NENA we are witness to a Semitic language which, although in an advanced degree of decomposition ('in einem schon weit fortgeschrittnen Zersetzungsprocess'), nonetheless never yields up its basic Semitic character (' giebt die Sprache doch nie den semitischen Grundcharacter auf).59 Nothing more really need be said. Conclusion

The evolved and disguised Semiticity of languages such as Israeli Hebrew, Amharic and Neo-Aramaic has had some practical implications for the progress of Semitic philology. While nobody would advocate the omission from any comparative grammar of the Semitic languages of, say, Classical Arabic or Biblical Hebrew, many of the Neo-Semitic languages have progressed so far and in such new directions that it is 58

Pace E.U. i. 7; iii 87. On the other hand, vocabulary is given its due i 73, 163, 187 (where the existence in Tigrinya of nfr 'fly', 'bird' is even said to go far towards underlining the essential unity of the Semitic tongues); ii 228, 232b; iv 108.

59

Ntildeke (1868: xxxiii, xxxiv).

78

SIMON HOPKINS

difficult to do them justice within the same framework as the classical Semitic languages. This is the reason that the two large volumes of Brockelmann's Grundriss (1908-1913) include rather generous coverage, especially syntactic, of Neo-Semitic material, whereas the smaller versions of the work (Semitische SprachwissenschaJt [Leipzig 1sI edition 1906, 2nd edition 1916]; Kurzgefasste vergleichende Grammatik der semitischen Sprachen [Berlin 1908]), apart from excluding syntax, are restricted to the classical languages only. This reduced format, which has been followed by later works in the field, e.g. the widely used An Introduction to the Comparative Grammar of the Semitic Languages by E.U. et al. (Wiesbaden 1964), certainly has its conveniences. But it also has its disadvantages, for the user of such works may justly protest that the half was not told him (cf. i 156 bottom). Since the publication of the great Grundriss in 1908-1913 research has advanced, a very large amount of further Neo-Semitic material has become available and quite new NeoSemitic languages have been discovered and described. 60 The comparative study of the Neo-Semitic Sprachtypus in its various manifestations (Arabic, Aramaic, Ethiopian, South Arabian and, to a lesser extent, Modem Hebrew) is among the most important tasks of Semitic linguistics.61 Such study will not only infuse a breath of welcome fresh air into comparative Semitics. 62 It is also bound to enrich our understanding of the classical Semitic languages and thus to benefit Semitic philology as a whole.

REFERENCES Arberry, A.J., 1943, British Orientalists, London. Arens, H., 1955, SprachwissenschaJt, Freiburg-Mlinchen.

60

Of particular importance in this connection is the series Semitica Viva (Wiesbaden 1987ff.) under the editorship of O. Jastrow.

61

A start, virtually unnoticed, had already been made by E. Konig (1901: 21ff.): 'Haupttendenzen der Entwickelung in den neueren semitischen Sprachen'. Polotsky taught a course at the Hebrew University entitled 'Verbale Kategorien in den neusemitischen Sprachen (Ullendorff 1992: no. 25 [1956]) and special note must be made here of the comparison between Neo-Aramaic and NeoEthiopian by O. Kapeliuk (1989).

62

Cf. Polotsky in Ullendorff (1992: no. 93 [1972]).

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

79

Baumgartner, W., 1940-41, 'Was wir heute von der hebraischen Sprache und ihrer Geschichte wissen', Anthropos 35/36 (pub!. 1942-4), pp. 593-616, reprinted in Baumgartner (1959: 208-239). - - - - " 1959, Zum Alten Testament und seiner Umwelt, Leiden. Bergstrasser, G., 1928, Einfohrung in die semitischen Sprachen, MUnchen. Bottcher, F., 1866, Ausfohrliches Lehrbuch der hebraischen Sprache, i, Leipzig. Brockelmann, C., 1908-13, Grundriss der Vergleichenden Grammatik der Semitischen Sprachen, 2 vols (vo!. i 1908, vo!. ii 1913), Berlin. - - - - " 1932, Review of C. Levias, n'?:J:J n'~iK p"p, (New York 1930), MonatsschriJt for Geschichte und Wissenschaft des Judentums 76, pp. 173-78. d'Abbadie, A, 1839, 'Lettre de M. d'Abbadie aM. Garcin de Tassy', Journal Asiatique III vii, pp. 364-66. - - - - " 1843, 'Sur les langue ethiopiennes' Journal Asiatique IV H, pp. 102-107. Dillmann, A, 1865, Lexicon Linguae Aethiopicae, Leipzig. Ewald, G.H.A., 1827, Kritische Grammatik der hebraischen Sprache, Leipzig. - - - - , 1863, Ausfohrliches lehrbuch der Hebriiischen sprache, 7th edition, Gottingen. - - - , 1871, Abhandlung uber die geschichtliche folge der Semitischen sprachen, Gottingen. Fleisch, H., 1974, Etudes d'arabe dialectal, Beyrouth. Fleischer, H.L., 1861, 'Vermischtes' (item no.2), ZeitschriJt der Deutschen Morgenlandischen Gesellschaft 15, pp. 381-83, reprinted in Fleishcer (1885-1888 iii: 468-70). - - - - , 1885-1888, Kleinere SchriJten, i-iii, Leipzig. Friedrich, J., 1925, Review of A Meillet and M. Cohen, Les langues du monde (Paris 1924), Orientalistische Literaturzeitung 28, cols. 806809. - - - - , 1952, 'Semitisch und Hamitisch', Bibliotheca Orientalis 9, pp. 154-57. - - - - , 1962, 'Das Neusyrische als Typus einer entarteten semitischen Sprache', istituto Universitario Orientale di Napo/i, Annali, Sezione Linguistica 4, pp. 4-106.

80

SIMON HOPKINS

Gesenius, W., 1810, Versuch uber die maltesische Sprache ... als Beytrag zur arabischen Dialektologie, Leipzig. - - - - , 1819, 'Die Amharische Sprache', in J.S. Ersch and J.G. Gruber (eds.), AUegemeine Encyclopadie iii, Leipzig, pp. 355-58. Goldenberg, G., 1996, ';'l'n l1'l:l1tl l11tl;:::l 11'1::131:1', in Evolution and Renewal. Trends in the Development of the Hebrew Language (= 11'1::131:1111tl;:1 :1T111tl1nl1:1::l1 :1T11n11!:l11:1::l), The Israel Academy of Sciences and Humanities, Jerusalem, pp. 148-190. - - - - , 1998, Studies in Semitic Linguistics, Jerusalem. Hommel, F., 1883, Die semitischen VDlleer und Sprachen i, Leipzig. Hopkins, S., 1999a, Review of C. Holes, Modern Arabic (London-New York 1995), Zeitschriftfor Arabische Linguistik36, pp. 85-96. - - - - , 1999b, 'The Neo-Aramaic dialects of Iran', Irano-Judaica 4 (ed. S. Shaked and A. Netzer, Jerusalem), pp. 311-27. Kapeliuk,O., 1989, 'Some common traits in the evolution ofNeo-Syriac and of Neo-Ethiopian', Jerusalem Studies in Arabic and Islam 12, pp. 294-320. - - - - , 1996, 'Is Modem Hebrew the only "Indo-Europeanized" Semitic language? And what about Neo-Aramaic?', Israel Oriental Studies 16, pp. 59-70. - - - - , 2004, 'Iranian and Turkic structural interference in Arabic and Aramaic dialects', Jerusalem Studies in Arabic and Islam 29, pp. 176-94. Khan, G., 2004, The Jewish Neo-Aramaic Dialect oj Sulemaniyya and Jfalabja, Leiden. Konig, E., 1901, Hebraisch und Semitisch, Berlin. Lazard, G., 1957, Grammaire du persan contemporain, Paris [= idem, A Grammar ojContemporary Persian, trans. S.A. Lyon (Costa-New York 1992)]. Macuch, R., 1962, 'Zu J. Friedrichs Arbeiten iiber das Neusyrische in Lateinschrift', Orientalistische Literaturzeitung 57, cols. 117-125. Munzinger, W. , 1865, Vocabulaire de la langue Tigre appended to Dillmann (1865). Noldeke, Th., 1868, Grammatik der neusyrischen Sprache, Leipzig. - - - - . , 1875, Mandiiische Grammatik, Halle. - - - - . , 1899, Die semitischen Sprachen, 2nd edition, Leipzig.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

81

1904, Beitriige zur semitischen Sprachwissenschajt, Strassburg. Olshausen, J., 1861, Lehrbuch der hebriiischen Sprache, Braunschweig. Oman, u., 1985, 'Hebrew is not a Jewish language', LesomJnu 48-49, pp. 199-205 [in Hebrew]. Pedersen, H., 1931, Linguistic Science in the Nineteenth Century, trans. J.W. Spargo, Cambridge, Mass. Philippi, F.W.M., 1871, Wesen und Ursprung des Status Constructus im Hebriiischen, Weimar. Plessner, M., 1931, 'Modemes Hebraisch', Orientalistische Literaturzeitung 34, cols. 803-808. Polotsky, H.J., 1964a, 'Semitics' in E.A. Speiser (ed.), The World History of the Jewish People ill, At the Dawn of Civilization, London, pp. 99-111,357/8.

- - - , 1964b, 'Egyptian' in E.A. Speiser (ed.), The World History of the Jewish People ill, At the Dawn of Civilization, London, pp. 121-34,359-63. - - - - , 1979, 'Verbs with two objects in Modern Syriac (Urmi)', Israel Oriental Studies 9, pp. 204-27. - - - , 1987, Review ofUllendorff(1985), Journal ofSemitic Studies 32, pp. 237043, reprinted in Ullendorff(1992: 139-45). - - - - , 1994, 'Incorporation in Modern Syriac', in G. Goldenberg and S. Raz (eds.), Semitic and Cushitic Studies, Wiesbaden, pp. 90-102. Rabin, C., 1992, 'Is Modern Hebrew still a Semitic language?', Hebrew Linguistics (= n"::l31 I1lJtZl?::l) 33-35 [Gad B. Sarfatti Volume], pp. 93-99 [in Hebrew]. Renan, E., 1855, Histoire generale et systeme compare des langues semitiques, 1st edition, Paris, [2nd edition 1858, 3rd edition 1863] Ritter, C.; 1840, Die Erdkunde von Asien, Band vil2. Drittes Buch = Erdkunde, Neunter Teil. Drittes Buch, 2nd edition, Berlin. Rodiger, E., 1839, 'Ueber die aramaische Vulgarsprache der heutigen syrischen Christen', Zeitschrijt for die Kunde des Morgenlandes 2, pp. 77-93,314-16. Rosen, H., 1955-56, 1J?tZl n"::lii:1, Tel-Aviv. - - - - . , 1977, Contemporary Hebrew, The Hague-Paris. Simon, R., 1986, /gnac Goldziher. His Life and Scholarship as Reflected in his Works and Correspondence, Budapest-Leiden.

82

SIMON HOPKINS

Spuler, B., 1953, 'Der semitische Sprachtypus', in Semitistik [Handbuch der Orientalistik III i, mit Beitragen von Carl Brockelmann et al.], Leiden-Koln, pp. 3-25. Trubetskoy, N.S., 1939, 'Gedanken iiber das Indogennanen Problem', Acta Linguistica [Copenhagen] 1, pp. 81-89. Ullendorff, E., 1955, The Semitic Languages o/Ethiopia, London. - - - - , 1957, 'Modem Hebrew as a subject of linguistic investigation', JSS 2, pp. 251-63, reprinted in Ullendorff (1977: 6880). - - - - , 1958, 'What is a Semitic language? (A problem of linguistic identification)" Orientalia 27, pp. 66--75, reprinted in Ullendorff (1977: 155-64). - - - - , 1960, The Ethiopians, 1sI edition, Oxford [3rd edition 1973]. - - - - , 1965, An Amharic Chrestomathy, 1st edition, London (O.u.P.) [2nd edition, London (S.O.A.S.) 1978]. - - - - , 1971, 'Is Biblical Hebrew a language?', Bulletin o/the School 0/ Oriental and African Studies 34, pp. 241-55, reprinted in Ullendorff (1977: 3-17). - - - - , 1977, Is Biblical Hebrew a language? Studies in Semitic Languages and Civilizations, Wiesbaden. ---,1979, Review of Rosen (1955-56), Bulletin o/the School 0/ Oriental and African Studies 42, pp. 548-53, reprinted in Ullendorff (1990: 150-55). - - - - , 1985, A Tigrinya (bgraiUia) Chrestomathy, Stuttgart. - - - - , 1987, Studia Aethiopica et Semitica, Stuttgart. - - - - , 1988, 'Along the margins of the Tigrinya Chrestomathy', in S. Uhlig and B. Tafla (eds.), Collectanea Aethiopica, Stuttgart, pp. 193-201. - - - - , 1990, From the Bible to Enrico Cerulli. A Miscellany of Ethiopian and Semitic Papers, Stuttgart. - - - , 1992, HJ. Polotsky (1905-1991). Ausgewahlte Brie/e. Introduced, edited, and annotated by Edward Ullendorff, Stuttgart. - - - - , 1994, 'Could Isaiah understand the Ha'arets newspaper?' in S.E. Balentine and J. Barton (eds.), Language, Theology and the Bible. Essays in Honour of J. Barr, Oxford, pp. 120-34, reprinted in Ullendorff (1995: 131-45). - - - - , 1995a, From Emperor Haile Selassie to HJ. Polotsky. An Ethiopian and Semitic Miscellany, Wiesbaden.

IS NEO-ARAMAIC A SEMITIC LANGUAGE?

83

- - - - " 1995b, Review of A. Saenz-Badillos, A History o/the Hebrew Language, trans. J. Ewolde (Cambridge 1993), Journal 0/ Jewish Studies 46, pp. 283-92. Wexler, P., 1990, The Schizoid Nature o/Modern Hebrew, Wiesbaden. Wild, S., 1977, '1st Ivrit eine semitische Sprache?', ZDMG Supplement III,l. XIX. Deutscher Orientalistentag ... , Vortriige, ed. W. Voigt, Wiesbaden, pp. 757-61.

SOME PARALLELS IN LINGUISTIC DEVELOPMENT BETWEEN BIBLICAL HEBREW AND NEO-ARAMAIC Geoffrey Khan One of the fascinations of studying the neo-Semitic languages is the occurrence in them of numerous developments that have parallels in the classical Semitic languages. Some of these parallels can be regarded as direct developments of features in the earlier languages, though most are independent typological developments. Even if only typological, they are, nevertheless, often of considerable heuristic value in the understanding of the history of the classical Semitic languages. There are several reasons for this. Many proposed historical developments for the classical Semitic languages that involve hypothetical pre-historical reconstructions have parallels in neo-Semitic languages that have a historically documented starting point. Furthermore, some of the parallel developments have evolved further in the neo-Semitic languages, which makes their true character and trajectory clear, whereas these may remain unclear in their less evolved state in the classical Semitic languages. Finally, although we are dealing largely with typological parallels rather than direct diachronic development, the phonological and morphological similarities across the languages that arise due to their genetic connection provide analogous contexts for these developments. In this paper I shall present a few selected parallels between the development of Biblical Hebrew and Neo-Aramaic, with special attention to North Eastern Neo-Aramaic (NENA).l

The BGDKPT consonants A student of Biblical Hebrew is faced with numerous exceptions to the general rule that the plosive variant of the BGDKPT consonants occurs Several other scholars have been intrigued by typologically parallel developments between various Semitic languages, e.g. Blau (1969 and 1985), who describes features of Arabic dialects that are paralleled in North West Semitic, and Kapeliuk (1989), who examines common features in the evolution of neo-Aramaic and neo-Ethiopian.

BIBLICAL HEBREW AND NEO-ARAMAIC

85

after consonants and the fricative variant after vowels. Some of these exceptions can be understood in the framework of the general rule, as, for example, the fact that a BGDKPT consonant is plosive at the beginning of word when preceded by a word ending in vowel that has a disjunctive accent (e.g. '!??~i' '~;1 Gen. 12:11) whereas it is fricative when preceding word has a conjunctive (e.g. '!??~~ 'l:!;1 Neh. 4:1). Here the plosive pronunciation could be explained by the occurrence of a slight pause and separation between the consonant and the preceding vowel brought about by the disjunctive accent. What, however, is the student to make of forms such as '?7~ (Gen. 17:16), n~~7~ (1 Chron. 12:24), n;~tP (Exod. 16:13), '?'!ltP (Lam. 2:19), rziit;l1~ (Jer. 18:7) and b~~~7 (Num. 15:39), where a fricative occurs after a silent sewa. Attempts have been made to interpret the sewa in some of these case as vocalic,2 but it is clear from masoretic sources that in the Tiberian reading tradition, which the vocalisation signs reflect, the sewa was silent) Also the sewa that occurs after a long vowel in cases such :tl~l (1 Sam. 9: 10), b~1'~ (Deut. 1:32), ~~t;l~ (Ezra 4:6) and 6!JlO (Psa. 78:9), was silent according to the Tiberian masoretic sources.4 Conversely, a plosive is occasionally found after a vowel, as in the 2fs. perfect offmal guttural verbs such as t;1!JP-7 'you (fs.) took'. This situation has arisen from fact that the 'rule' of the BGDKPT consonants is no longer fully operational in the Tiberian pronuncation tradition that is reflected by the vocalisation and the plosive or fricative realisation is in the process of becoming fixed irrespective of the phonetic environment. Where a vowel has been elided before the consonant after the rule ceased to operate the consonant generally remains fricative: *mala~e > mal~ ('~7~), sfo~f > sif!sf ('::;l'!ltP), etc. There is, however, a certain degree of inconsistency in this, since a plosive realisation is occasionally found in the same morphological context, e.g. '''l~~ (Isa. 5:10), '~9l.'t (Jer. 10:17), :::li;l;>~ (Deut. 31 :24), and even in the same word, e.g. '$tpi (Cant. 8:6) vs. '~tpi (psa. 76:4), ,,~~~ (Num. 14:3) vs. ,,~~~ (1 Sam. 17:9). This inconsistency and opacity relating to the BGDKPT rule is a reflection of a historical change that was in the process of taking place when the Tiberian reading tradition was fixed. Such opacity is generally not 2

Ewald (1870: §89), Konig (1881 i, 69) and the Hebrew grammar of Gesenius up until the 27th edition (§ I 0), for example, regarded the syllable as loosely closed and the sewa as neither silent or vocalic but rather intermediate between the two (sewa medium).

3

The crucial evidence is that vowel before the sewa is not lengthened, which is the fate of all vowels in open syllables in the Tiberian reading tradition (Khan 1987: 33ff.).

4

Khan (1987: 54-55)

86

GEOFFREY KHAN

tolerated in living languages and developments occur that resolve it. One solution would be for the fricative and plosive forms to become totally independent of their phonetic context and acquire phonemic status. There are, indeed, incipient signs of the acquisition of phonemic status, since a few possible minimal pairs can be identified in Tiberian Hebrew where a plosive contrasts phonemic ally with a fricative, e.g. MljP-7 'to take' vs. t;1ljP-7 'you (2fs.) took'; ';l7~ 'thousands of' vs. ,~'?~ 'two thousand of'; ':;liP 'approach (fs.)' vs. '~iP 'my inside'.5 In the last two pairs one of the items is not actually attested, but its existence can be inferred (,~,?~ from c:g,?~, ':;liP from ~::JiP [Exod. 16:9]).6 The BGDGPT consonants developed in a similar way in Aramaic. The fricative realisation was initially conditioned by a preceding vowel, but this rule gradually ceased to operate. In some of the literary forms of Aramaic, as in Tiberian Hebrew, there is evidence of an incipient phonemicisation of the fricative and plosive variants, as in Syriac, for example, where one finds such pairs as garQa 'scabies' garba 'scabious'.1 In the neo-Aramaic dialects this process has reached its conclusion. The two erstwhile phonetic variants of BGDKPT consonants have become independent phonemes which contrast in numerous minimal pairs. 8 The following examples are from NENA dialects, where the new phonemes contrast after vowels or consonants and also at the beginning of a word, e.g. Qaraqosh: sata 'year'-sa.ta 'fever'; katwi 'they sit'-ka.twi 'they write'; giida 'wall'-giida 'churn'; Christian Barwar marta 'saying'mar.ta 'mistress'; tela 'fox'-.tela 'she came'. When a verbal root contains a BGDKPT consonant, fricatives and plosives do not alternate throughout its inflections as in earlier phases of North West Semitic, but rather either the fricative or the plosive comes to be treated as a radical consonantal phoneme and it occurs in all inflections and derivations of the root, irrespective of the current or historical phonetic environment, e.g. k.tw 'to write' « *ktb): ka.twa 'she writes', mak!i5wu 'to register', k.tawa 'book', ka.tawa 'writer' « *kattaQa). This process has taken place historically by selecting the variant that occurs in one of its inflections or derivations and generalising it by analogy to all inflections and derivations of the root. In the root k.tw the two second radicals are fricatives, which would have been generalised from a context such as the active participle *ka!i~. In some verbs a plosive has been generalised in these positions, e.g. Qaraqosh sty 5

Cf. Jouon and Muraoka (1991: 31-32).

6

Muraoka (1995: 21) has noted, however, that the theoretically possible dual construct form '!il715 is unlikely to have been used in reality.

7

Noldeke (1898: 16).

8

For the general phenomenon see Tserete1i (1980).

BIBLICAL HEBREW AND NEO-ARAMAIC

87

'to drink', stq 'to be silent', str 'to cover', Sxt 'to become dirty', rkw 'to ride'. In the case of the first three, the initial sequence of sibilant + stop (st- or st-) is likely to have existed not only in the prefix conjugation of earlier Aramaic (e.g. Syriac neste 'he will drink'), but also elsewhere due to the ease with which such a sequence is pronounced as a cluster (cf. the Syriac suffix conjugation leSti 'he drank', where the prosthetic vowel reflects the existence of the cluster. This stop was then generalised to include positions where there is a preceding vowel, e.g. siita 'He drinks' « *sa!e). In the case of Sxt 'to be dirty' and rkw 'to ride' a noun form appears to be the source of the analogy, namely saxta 'dirt', in which the stop ItI occurs after a vowelless consonant, and rakiiwa « *rakkii'Qii) 'rider', in which the stop /k/ is the reflex of a historically geminated consonant. As a result of this analogical generalisation in verbal roots and lexemes, a stop at the beginning of the word remains a stop when a particle that ends in a vowel is prefixed to the word, e.g. ka!iiwa 'the writer', lu-ka!iiwa 'and the writer', tiima 'there', lu-tiima 'and there'.9 We see, then, that in the NENA dialects the inconsistency and opacity with regard to the plosive and fricative BGDKPT consonants in earlier North West Semitic that were caused by the onset of a historical change have been completely resolved. 10 (Tuffuralconsonan~

The letters ~e! and fayin in the Tiberian reading tradition of Biblical Hebrew were pronounced as an unvoiced pharyngal and a voiced 9

10

Cf. Khan (2002: 30ff.), Kapeliuk (1997). Similar developments are found in reading traditions of Hebrew that evolved further than the Tiberian biblical tradition. The most conspicuous case is that of the Samaritan reading tradition, in which the BGDKPT consonants are all pronounced as plosives in all contexts, except the letter peh, which is regularly pronounced as a fricative. There is evidence, however, that in the Middle Ages a distinction between plosive and fricative variants existed as in the Tiberian tradition (Ben ijayyim 1954: 97-101). An example of a lesser degree of development is the tendency in some reading traditions of the Mishna to pronounce BGDKPT letters as stops at the beginning of a word after a prepositions with vocalic sewa (Katz 1981: 27) and also in spoken Israeli Hebrew, at least in its non-standard form, e.g. be-kita 'in a class'. In Israeli Hebrew, indeed, the fricative and plosive opposition is phonemic, e.g. saval 'he suffered' vs. sabal 'porter', and in its non-standard form a fricative or a plosive is occasionally generalised throughout a verbal root, e.g. /i-Shor 'to break', nisoar 'broken', sabarti '1 broke'; le-xabes 'to wash', xibasti '1 washed' (Berman 1997:

316).

.

GEOFFREY KHAN

88

pharyngal respectively.ll As is well know, comparison with Semitic languages that have preserved a larger consonantal inventory, such as Arabic and Ugaritic, shows that in some cases these pharyngals have developed from original velar fricatives, e.g. Hebrew,icq 'ass' = Arabic ~ima:r, but Hebrew n~ 'brother' = Arabic 'lag; Hebrew 'eye' = Arabic fayn, but Hebrew ::l"W 'evening' = Arabic garb. These correspondences show that at some stage in the history of Hebrew the original velar fricatives *g and *g merged with the pharyngal fricatives ~ and f. This merger took place in some other languages in the Syria-Palestine area already in the 2nd millennium B.C.E., as is shown, for example, by the Phoenician alphabet, which lacks separate symbols for the velar fricatives and, it seems, also by the shorter Ugaritic 'Mirror ABC', which was used in the Palestine area. 12 Egyptian transcriptions of Palestinian proper names from the beginning of the 1st millennium B.C.E. also reflect such a merger. 13 There is no consensus, however, as to when the merger took place in Hebrew. The argument turns on the extent to which the Phoenician alphabet was used polyphonously for the representation of Hebrew. The later masoretic pronunciation of the letter was /s/ or /s/ is cited as evidence for such polyphony at an earlier period. Did the letters ~e! and fayin disguise an original distinction between velar and pharyngal fricatives in pronunciation or had they merged, as in Phoenician? Important evidence for this issue is constituted by the Greek transcriptions in the Septuagint of proper names containing these letters, which often appear to reflect a distinction between etymological velar fricatives and etymological pharyngal fricatives, e.g. Axa/; = lJ;!~ (cf. Arabic 'lagaga), but IcraaK = PJ;!~; (cf. Arabic ~~ika); ra1;a = i1W (cf. Arabic Gazza) but BaaA. = ,,p~ (cf. Arabic ball). Some scholars have taken the view that the transcriptions cannot be relied upon as evidence of a genuine distinction between velar and pharyngal phonemes at this period, since the Greek khi and gamma occur in some cases where there is an etymological pharyngal. 14 Some have maintained that the alternation between khi, gamma and zero resulted from the attempt to use the inadequately equipped Greek alphabet to represent the Hebrew pharyngals, whose pronunciation lay somewhere between these. 1S

r,p

11

ef. Khan (1996: 7, 9).

12

Gordon (1967: 16).

13

This appears to be the case in the 'Shoshenq list' (c. 950); cf. Harris (1939: 63).

14

E.g. Bergstrasser (1918: 36-38), Harris (1939: 63).

15

E.g Lisowsky (1940: 149). This appears to be the view also of Murtonen (1986: 5-6; 1990: 28). A similar argument is presented by Ruzicka (1908), who contended that the voiced velar fricative never existed in earlier proto-North West Semitic.

BIBLICAL HEBREW AND NEO-ARAMAIC

89

According to Harris (1939: 63), the use of khi and gamma in the transcriptions reflect an allophonic range of the phonemes ~ and f that included both velar and pharyngal realisations. Flashar (1908) argued that the distinction between gamma and zero represented a genuine distinction in pronunciation, but this was introduced artificially by the translators and was not inherited from an earlier tradition. The study of Wevers (1970) on the transcriptions of the ~e! in the Septuagint showed that the inconsistency noted by previous scholars was often due to scribal corruptions in the manuscripts and could be eliminated by textual criticism. 16 He demonstrated that in the transcription of ~e! genuine etymological correspondences occur with a considerable degree of regularity in good manuscripts. After this study it is now widely accepted that the transcriptions reflect the existence of k and ~ as separate phonemes at the time of the Septuagint. 17 The situation, however, is still not clear cut. Wevers did not examine the internal chronology of the variant transcriptions within the Septuagint nor could he eliminate all the inconsistencies. Moreover he was concerned only with ~e!. The use of gamma to represent etymological 9 is less consistent. The frequency of its use, furthermore, diminishes in the Septuagint translation of the later biblical books. 18 A lesser degree of consistency in the transcription of etymological k by khi is also found in the later books. This has been interpreted by Blau (1982) as reflecting a historical change in progress, whereby the velars, which existed at the time of the early books, had become merged by the time of the latest books. Murtonen (1986: 5-6; 1990: 28) considers this to reflect the gradual weakening of the articulation of the fayin and ~e!, but notes that 'the more frequently used names' are more conservative of their original pronunciation. An examination of parallels in the development of Neo-Aramaic are of significance with regard to two main issues in this debate. (1) Does the reflection of velar fricatives in the Greek transcriptions necessarily demonstrate their existence in Hebrew in general at the same or earlier periods? (2) Can the inconsistencies in the transcriptions have a genuine linguistic basis? 16

His remarks applied especially to the work of Riizicka, whose inadequate text critical methods had already been noted by Bergstriisser (1918: 37).

17

Cf. the remarks of Steiner (1977 :120). Both Blau (1966: 141) and Steiner (ibid) point out that the counterexample cited in Moscati et al. (1964: 40) is based on a false etymology. There is, indeed, now evidence from an Aramaic text written in Demotic script that Aramaic spoken in Egypt at approximately the same period also made a distinction between velar and pharyngal fricatives; see Nims and Steiner (1983: 263) and Vleeming and Wesselius (1985: 25).

18

Flashar (1908: 313), Wutz (1925: 109, 121).

GEOFFREY KHAN

90

After the development of fricative variants of the BGDKPT consonants, it can be assumed that both the pharyngal fricatives ~ and f and also the velar fricatives k and 9 existed in Aramaic. Although the latter pair were originally allophonic alternants of plosives, we have seen that they gained the status of independent phonemes in the neo-Aramaic dialects. In the NENA group of dialects these velar fricative consonants underwent a further development in that they merged with the pharyngal fricatives ~ and f, thus recycling the earlier historical phonology of North West Semitic. In other dialect groups of neo-Aramaic, however, the two pairs of consonants were kept distinct. 19 The attested historical development of Aramaic shows us, therefore, that this merger of consonants was not consistent across all dialects. Within the NENA group some differences are found in the nature of the merger of the two groups of consonants. In most dialects the result of the merger of k and ~ is the velar k (conventionally transcribed as x), e.g. Qaraqosh:

baxa xmiira

'he weeps' 'ass'

< *bake < *~miirii

In some dialects in the north western periphery of the NENA area, however, the result of the merger is the pharyngal ~, as in earlier North West Semitic, e.g. Hertevin20 bii~e ~miira

'he weeps' 'ass'

< *biike < *~miirii

In all NENA dialects the voiced velar fricative 9 merged with the voiced pharyngal f. This voiced pharyngal, however, in most cases became subsequently weakened to a laryngal 'lor to zero. It tends to be retained as the laryngal 'l in the more archaic dialects such as Qaraqosh:

srii'la rii'las piila'l tar'la siima'l 19

20

'lamp' 'he wakes' 'he divides' 'door' 'he hears'

< *sriifii < *riifiS

< *piilif

< *Sriigii < *riigis < *piilig

< *tarfii < *siimaf

Velar fricatives and pharyngal fricatives are retained in Turoyo (Jastrow 1993: 8) and Western Neo-Aramaic (Arnold 1990: 11-14). In Neo-Mandaic they have not merged, though the pharyngals have been weakened (Macuch 1965: 79ff.). Jastrow (1988: 6).

BIBLICAL HEBREW AND NEO-ARAMAIC

91

In some dialects where it is weakened, there is a vestige of the pharyngal in the form of the pharyngalised pronunciation of the remaining word, e.g. Christian Urmi: 21 'he wakes' 'he divides'

< *rafiS < *paUf

< *ragiS < *pa/ig

In others in many contexts it is weakened completely, without a trace, e.g. Jewish Arbel:

res

pale

'he wakes' 'he divides'

< *rafiS < *ragiS < *palif < *paUg

The mergers of velar and pharyngal fricatives do not occur with complete consistency in the dialects. In many dialects where the pharyngal 1; generally merges with the velar x, the earlier pharyngal pronunciation is retained in the environment of emphatic consonants or q, e.g. Qaraqosh I;iizfqa 'strong' ral;oqa 'distant' In the Qaraqosh dialect, furthermore, the pharyngal I; has been preserved outside of this phonetic context in a number of words associated with religious tradition, e.g. rol;ad qudSa 'the Holy Spirit', ms~a 'the Messiah', slfl;e 'apostles', magabl;a 'altar' .22 The consonant 9 retains its velar articulation in a few words and does not merge with the pharyngal: 23

pagra siigid

'body' 'bow down'

< *pagra < *siigtg

The 9 in such words is often devoiced to x in some dialects, e.g. Christian Barwar: paxra, siixag. In some dialects a certain degree of free variation between pharygnal and velar realisations of historical *1; is attested in some words, e.g. Jewish SUlemaniyya: 24

gaxak - gal;ak xola-I;ola

'he laughs' 'rope'

21

Cf. Tsereteli (1990).

22

Cf. Khan (2002: 40-41).

23

Tsereteli (1990: 36-38).

24

Khan (2004: 33).

< *gl;k < *I;agla

92

GEOFFREY KHAN

In the case of historical *g, some cases of variation have been lexicalised in the form of a doublets. In one item of the pair the velar g has been retained but in the other it has merged with the pharygnal and the two items have subsequently become distinguished semantically, e.g. Christian Barwar: sagas 'he is disturbed'

*nafr

>

*nafar

>

*nafar

Possible evidence for this is found in the Greek transcriptions of the Hexapla, where segolates with $ere or bolem in Tiberian Hebrew are transcribed by epsilon and omicron, e.g. Xecr'A 1(OP

32

Fox (2002: 167).

'9? (Psa. 49:14) nil' (Psa. 49: 1)

96

GEOFFREY KHAN

The Hexapla evidence is somewhat problematic, since it does not reflect anaptyctic vowels in the transcription. If we follow Blau (1979: 52-53), however, we could attribute this to an orthographic phenomenon rather than to the lack of anaptyctics in actual pronunciation. As is well known, when an a vowel in Hebrew occurs in a pretonic open syllable, it is lengthened and is represented in the Tiberian vocalisation by qame$ on account of the fact that it has undergone the quality shift of *a >

5:

*dabdru >

*daMr

>

d5I]Dr

When a pretonic syllable was originally closed but has become open by the weakening of the gemination of a following guttural consonant, it often contains a pata1;t vowel sign in Tiberian Hebrew, e.g. tzi1hiJ « *ha1;t1;todes), ~~i1iJ « *hahhii), c'r:rl:.' « *7a1;t1;tim).3 3 This is traditionally referred to as 'virtual doubling' of the guttural (DageS forte implicitum, occultum or delitescens, Gesenius-Kautzsch 1910: § 22b). What is not so well known is that pata1;t vowels in such pretonic syllables were pronounced long in the Tiberian reading tradition.3 4 This lengthening must have taken place after the quality shift *a> 5 had ceased to operate, otherwise the vowel would be represented by qame$. It can be hypothesised that this lengthening was preceded by a period in which the vowel remained short in the open syllable by a historical drag effect:

*hahhU

>

*hahU

>

*hahU

A parallel to such a development can be identified in the NENA dialects. In this dialect group there is widespread weakening of consonant gemination. This has resulted in the opening up of syllables that were originally closed. In some dialects when this takes place a vowel that was short in the closed syllable regularly remains short when it is opened, e.g. Alqosh:35

*sapplra *zammara

33 34 35

>

>

saplTa zamara ..I

'beautiful' 'singer'

For the historical gemination in a word such as C'J:ll.C see Blau (1980: 33). See Khan (1987: 33ff.). Data regarding the Alqosh dialect are taken from CoghiIl (2003).

BIBLICAL HEBREW AND NEO-ARAMAIC

97

Although in the Alqosh dialect pretonic vowels exhibit a general tendency to be shortened in fast speech, a historically long a vowel in a pretonic syllable regularly retains its length in some forms, e.g. *nat~ra

>

*pa.le~wa

>

_

.t:

natora palexwa

'guard' 'he used to work'

In some other dialects the a vowel in an originally closed pretonic syllable is optionally pronounced long, e.g. Christian Barwar:

*sapptra *zammara *nat~ra

> > >

saptra - saptra zamara - zamara nat~ra

'beautiful' 'singer' 'guard'

In dialects such as Christian Barwar, therefore, the historical drag has weakened and originally short vowels in pretonic syllables can now be lengthened. Why is there a pata~ in the final stressed syllable of the 3ms. perfect ,rbi? whereas there is a qame$ in the final stressed syllable of a noun such as i~1? Why, furthermore, are the 3fs. and 3pl. of the perfect verb in Hebrew vocalised iT7tpi? and 1'tpi? whereas forms with pronominal suffixes are vocalised 1iT7t?i?, i'17t?i?, C7t?i?, etc.? The forms with pronominal suffixes conform to the normal rules of pretonic lengthening and propretonic reduction whereas the 3fs. and 3pl. forms without suffixes do not. As we have seen, the qame$ in the stressed syllable of a noun such as i~1 is the result the rule that an originally short a vowel was lengthened in a stressed open syllable (*dabciru > *dabclru > *da~ar). It must be assumed that this lengthening took place when the noun still had a final case vowel, which was subsequently lost, and that the stress fell on the penultimate syllable. The forms of the 3rd person perfect verb are generally reconstructed as follows: 3ms. 3fs. 3pl.

*qatala *qatalat *qatalu

The final short -a of the 3ms. and the final -t of the 3fs. were subsequently elided. Why did these not undergo the same development as nouns, with the vowel in the penultimate open syllable being lengthened by stress (*qatala > *qatclla> *qatcll; *qatalat> *qatcllat> *qatclla; *qata1u > *qatcllu )?

98

GEOFFREY KHAN

Some scholars have explained this by the theory that the original stress in verbs was not penutlimate, as in nouns, but rather antepenultimate: 36

*qata1a > *qatalat > *qata1u >

*qata1a > *qatalat > *qata1ii >

*qata1 *qatalii *qatalii

There are some problems with this reconstruction. A form with a pronominal suffix, such as '1ii7~i?, for example, has to be reconstructed as having penultimate stress in order to explain its vocalism. It is not clear why this should be if there was general antepenultimate stress in verbs:

*qatalahu

>

*qatalclhii

Others have argued that there was general penultimate stress but when the lengthening took place in the stressed syllable of nouns, the final vowel had already been elided in the 3ms. verb, so, in the verb, the stress fell on a closed syllable and the vowel was not lengthened (*dabaru but *qata1).37 It is not clear, however, why there should have been a difference in the development of final short vowels in these two grammatical categories. Brockelmann considered the lengthened stressed vowel in nouns to be originally a pausal lengthening that was extended into context forms. Such as an extension of pausal lengthening did not occur in verbs, due to their lesser tendency to occur in pause and, therefore, the stressed vowel of verbs remained short in context. 38 This, however, still does not fully explain the 3fs. and 3pl. forms, unless one follows the view of Blau (1978: 99-100) that the expected pretonic lengthening did not take place (*qiita1ii> *qiita1a, *qatalii > *qatah1) due to the influence of Aramaic. I shall propose here another explanation, which was suggested to me by a development in the verbal system ofneo-Aramaic. The pausal forms of the 3fs. and 3pl. have the stress on the penultimate: ii7~i?, '1"~i? As is usually the case with pausal forms, it appears that these preserve a more archaic stress position than the context 36

Bergstrasser (1918: i, 116-117), Bauer-Leander (1922: 179), Harris (1939: 60). Poebel (1939) argued that at some stage in the history of Hebrew stress was placed on the antepenultimate suyllable of all words, including nouns as well as verbs, but this cannot explain the difference in vocalisation of the two word categories.

37

This was the explanation of Sarauw (1939: 68) and Blau (1976: 31).

38

Brockelmann (1908-13, i, 106) and (1940). Birkeland (1940: 21) also inclines towards this view.

ef.

BIBLICAL HEBREW AND NEO-ARAMAIC

99

forms. Indeed, if the stress falls on the penultimate, the qame~ in the first syllable conforms to the normal rule of pretonic lengthening. It can be assumed, therefore, that the stress on verbs was originally penultimate, as in nouns, and not antepenultimate. In the 1SI and 2n person inflections of the perfect the pata1;t can be simply explained by the fact that the a vowel is in a closed syllable in these forms and so would not have been lengthened if stressed, just as an a vowel in the closed syllable of a noun was not lengthened:

*qat&ta > *qata1t > *qata1tf >

*fammu >

*qatalta *qata1t *qataltf

*fammu >

*fam

The long a vowels shifted quality to 5, then after this quality ceased to operate all stress vowels were lengthened, including those in a closed syllable, which resulted in the forms that are found in the Tiberian reading tradtion:

*qata1ta > *fam

*da~ar

> >

*q5talt5 >

fam (c~)

d5~5r

('*"1)

If the stress in the 3rd person perfect forms fell originally on the penultmate syllable, then it appears that the normal rule of vowel lengthening in a stressed open syllable was blocked:

*qata1a > *qatalat > *qata1u >

*q5tal *q5ta15 *q5talii

Subsequently the stress shifted to the final syllable of the 3fs. and 3pJ. forms but the vocalism remained fixed: This stress shift was the result of a generalisation of final stress in the language: 39

*q5ta15 *q5ta15

> >

*q5ta15 *q5talu

These forms are close to what is reflected by the Tiberian vocalisation il7tpl? and ~"tpl?, since in the Tiberian pronunciation tradition a vocalic sewa

39

For this phenomenon in the history of Hebrew see B1au (1978: 99-100).

GEOFFREY KHAN

100

was pronounced as a short a vowel.40 There is an accumulation of evidence, however, that in the Tiberian tradition the sewa occuring after a long vowel was generally silent,41 so we must assume that in a final stage of development the short a in the unstressed syllable was elided:

*q5tal5 *q5tahl

> >

q5tl5 q5tld

The crucial question in this reconstruction is why the usual rule of vowel lengthening in stressed open syllables was blocked in the forms *q5tru, *q5tal5 and *q5talii. Scholars who follow the theory of Grimme (1896: 46) that the lengthening of the vowel in a noun form such as *dabcfru was compensatory for the elision of the final vowel have proposed that the lack of lengthening of the stressed vowel in the 3fs. and 3pl. forms *qat6lat and *qat6lii was due to the fact that the final syllable was not elided. Since there is some evidence for this phenomenon elsewhere in Hebrew,42 we can accept it as a satisfactory explanation here. But can the lack of lengthening of the stressed vowel in the 3ms. form qatala be explained without resorting to the theory that the final vowel was elided earlier than the nominal case vowels? Some of the NENA dialects exhibit a development in verb paradigms that may be identified as a typological parallel and so cast light on this problem. The base of the present verb in NENA is the ertswhile active participle qiitil. This is inflected by a series of suffixes that express the pronominal subject. In the 3fs and 3pI. forms these are the fs. and pI. nominal inflections of earlier Aramaic. The 1SI and 2nd person suffixes are cliticised pronouns. Most of these suffixes begin with a vowel and their affixation causes the vowel after the second radical of the base to be elided. As a result the long a vowel after the first radical is in a short syllable and, by the usual principles of vowel quantity, it is shortened, e.g. Qaraqosh

qatal + a qatal + i qatal + at qatal + at qatal + itu qatal + an 40 41

42

> > > > > >

qatla qatli qatlat qatlat qatlitu qatlan

'she kills' 'they kill' 'you (ms.) kill' 'you (fs.) kill' 'you (pI.) kill' 'I (m.) kill'

Cf. the detailed treatment of this question in Morag (1963: 160-66) and the references in Khan (1987: 27 n.12). Khan (1987: 54-55). Cf. Blau (1976:31).

BIBLICAL HEBREW AND NEO-ARAMAIC

qiit"l + an qiit"l + ax

> >

qatlan qat/ax

101

'I (f.) kill' 'we kill'

In the Qaraqosh dialect, as in many other NENA dialects, the 3ms form is the uninflected base qiit"l with a long a vowel in the first syllable. What is of interest in this connection is that in some dialects the a vowel also in the 3ms. form is short. Such is the case, for example, in several of the Jewish dialects in the south eastern NENA area, e.g. Jewish Arbel:

qat"l qatla qatli, etc.

3ms. 3fs. 3pI.

The most plausible explanation is that the short a vowel in the 3ms. form has arisen by analogy with the short vowel in the corresponding syllable in the rest of the paradigm.43 A similar phenomenon of paradigm analogy is found in the paradigm of weak verbs with an original final pharyngal *1' in some of the Christian dialects. In the Christian Barwar dialect, for example, the 3ms present form of the verb 'to hear' is same, with a short a. In the strong verb, on the other hand, this vowel is long in the 3ms (qiit"l). The full paradigm of 'to hear' is as follows: 3rd pers. ms. fs. pI. 2nd pers. ms. fs. pI.

same sama same sam"t sam"t samitu

1st pers. ms. fs. pI.

sam"n sam"n Sam"x

In all forms except the 3ms. the first syllable was original closed due to the presence of a pharyngal consonant as third radical and this conditioned the shortness of the a vowel:

sama sami

43

<
e, rispettivamente, con una , tutte lettere che corrispondono appunto a fonemi esplosivi non aspirati. Considerate tutte queste particolarim della fonetica araba, si puo supporre che in arabo Veneticum-Venetici suonasse [banadiq]. Assunta 4 5

6

Cf. VAI, vol. I, p. 95. In islandese vengono registrate tre forme per 'Venezia': Fenedi, Fenidi e Feney. La terza risponde curiosamente al significato di 'isola acquitrinosa', dato che fen (anche in inglese) significa 'acquitrino, palude, pantano'; cf. De Vries (1977: 117). Cf. Nallino (1963).

VENEZIA E LIVORNO

155

questa veste fonetica tipicamente araba, il toponimo e stato analizzato in arabo come un sostantivo plurale poiche 10 schema sillabico lXaXiiXiXI che soggiace a [banadiq] in arabo e tipico di questo numero. All'orecchio di un parlante la lingua araba *baniidiq e dunque un nome pluraIe. Di qui I' esigenza di ricavarne, per retroformazione, il corrispettivo singolare. E a questo punto che e scattata una fortuita convergenza con bunduq 'noccioli; nocciole'. Questa parola al plurale fa banadiq e coincide pertanto con *banadiq 'Veneticum'. Per ottenere il singolare del secondo *banadiq e venuto quindi spontaneo di ricorrere alIo schema sillabico di bunduq ed e nata la forma al-bunduqiyyah 'la (citta) venetica', con il suffisso aggettivale femminile -iyyah. Non solo, ma da al-Bunduqiyyah e stato a sua volta derivato il gentilizio bunduqanr 'veneziano, abitante di Venezia' (f. bunduqaniyyah, pI. m. bunduqaniyyiin, pI. f. bunduq-aniyyat).7 La lingua era ormai pronta ad applicare a bunduqanr 10 schema sillabico lXaXiiXiXah! che serve alla formazione del plurale di un gentilizio senza distinzione di generi. E nata cosi la forma baniidiqah con il significato 'abitanti di Venezia'. Allo stesso modo a{ariqah 'africani' e il pIurale di ifriqr 'africano', mawii$ilah 'mossulioti' 10 e di maw~ilr 'abitante di Mossul' e baghadidah 'bagdadeni' e il plurale di baghdadr 'abitante di Bagdad'. Il fatto che in arabo il nome di Venezia continui ad essere accompagnato dell'articolo determinativo (al-Bunduqiyyah 'la [citta] venetica') puo avere comunque facilitato, nella coscienza linguistica dei parlanti, il suo collegamento con il concetto militaresco di pallottole e reIativi fucili: 'quella che ha a che fare con le pallottole'. Ancora un'informazione e in ultimo un avvertimento. Da al-Bunduqiyyah e stato derivato l'aggettivo bunduqr che significa tanto 'zecchino d'oro di Venezia' quanto 'oggetto laccato in oro'.8 E opportuno distinguere il collettivo bunduq 'nocciole' e tutti i suoi derivati botanici, bellici, venatori 0 veneziani che siano, dal sostantivo singolare bunduq (pI. banadrq). Quest'ultimo infatti significa 'bastardo'.9 Livomo e un'altra citta marinara italiana che stupisce per la fantasiosa pluralita dei suoi nomi, che non e tuttavia paragonabile a quella di Venezia. La famosa citta portuale, fondata dal granduca di Toscana Ferdinando I de' Medici (1587-1609), trae il suo nome da quello di un precedente villaggio di pescatori che fin dall'Xl sec. (documenti del 1017) era noto come Livorna.1 0 Oltre al villaggio ci doveva essere una 7

8 9 10

Cf. VAI, vol. I, p. 95. Cf. VAI, ibidem. Cf. VAI, vol. I, p. 96. Cf. Pieri (1969: 25).

156

F ABRIZIO A. PENNACCHIETII

fortificazione, eretta tra il IX e il X secolo per difendere dai saraceni la costa di Porto Pisano, che era l'emporio di Pisa.l1 Si e ritenuto che l'insediamento di Livoma risalisse all' epoca romana, per cui ne e stato ricostruito it nome originario in *Libuma, che alluderebbe a un tipo di imbarcazione da trasporto dell'antichita.l 2 It toponimo femminite Livoma e abbastanza bene riprodotto da Lioma, antico toponimo spagnolo per Livomo.13 Al contrario, la versione francese Livoume ha neutralizzato la distinzione morfologica tra il toponimo italiano piu recente al maschile (Livomo) e quello piu antico al femminile (Livoma). Caratteristica del toponimo spagnolo Lioma e anche quella di aver convogliato il significato di "luogo dove regnano confusione e agitazione", in riferimento al gran traffico di merci che si svolgeva nel porto franco di Livomo tra it XVII e il XIX secolo.l 4 Quest'ultimo significato, pero ormai svincolato dal riferimento alIa citta di Livomo, 10 si ritrova anche nel termine portogese lioma 'confusao, barafunda, enredo', ormai desueto.l 5 Sorprendente e l'aspetto che assume il nome di Livomo in inglese: Leghom [le'go:n / 'legho:n].l6 It toponimo sembra composto dalla giustapposizione di due termini inglesi, leg 'gamba' e horn 'como', che, a HvelIo semantico, non appaiono compatibili. Oltre che come toponimo, in inglese Leghom serve a designare (1) il particolare intreccio di paglia con cui dal1740 si confezionano i famosi 'cappelli di paglia di Firenze', e (2) una razza pregiata di polli (razza livomese 0 razza Livomo) importata negli Stati Uniti d' America da Livomo nel 1869 e da li diffusa in tutto it mondo. L' Oxford English Dictionary interpreta Leghom come un adattamento, mediante etimologia popolare, della forma italiana Legomo che sarebbe stata in uso tra it XVI e it XVII sec. 17 Siccome, pero, la forma Legomo non risulta documentata per scritto da nessuna parte, sara necessario ricercare l'origine di Leghom altrove. Un primo indizio sull'origine di Leghom 10 troviamo in Turchia dove, nelle mappe redatte tra la fine del XVII sec. e l'inizio del XVIII

12

Cf. Bausani (1990:38, nota 117). Cf. Pieri, ibidem.

13

Cf. EUIEA, tomo XXX, pp. 951-53; EUIEA, Ap6ndice, p. 823.

14

Cf. EUIEA, tomo XXX, p. 951. Cf. GDLP, p.263.

11

15 16

17

Cf. OED, vo!. VIII, p. 808c. Cf. ibidem.

VENEZIA E LIVORNO

157

sec., Livomo e indicata come 'Golfo di Belvedere'); cf. Bausani (1990, tavv. XI e XIII). Cf. Sebag (1991 :80). Cf. Sebag (1991:87). Cf. Zavadovskij (1979: 60). Cf. Sebag (1991: 80,316). Cf. VAI, vol. rn, p. 1162.

158

F ABRIZIO A. PENNACCHIETTI

riscatto di viaggiatori catturati dai corsari, la citta di Livomo viene comunque ricordata nelIa forma ~V :rtaoU oaprd, OlL el,~ 10V al,oova 10 EAEO~ aUlou. He dated it to the second century C.E., but did not indicate his grounds for doing so. No photograph was available of the inscription until a few years ago, when Steven Fine and Leonard Rutgers re-examined the style of the menorah which they dated to the period between the fourth and sixth centuries C.E..4 They stressed the importance of the combination of the menorah and the first inscription as the only evidence for a Jewish community in ancient Nicaea and as an important source for the history of Judaism generally in Asia Minor. On the basis of the photographs they provided an improved reading for the Jewish Greek inscription, substituting apl0v 'bread' for uya96v, and commenting that this equivalence for cry,? is both closer to the Hebrew and also found in the readings of the later Jewish revisers of the Greek Bible Aquila and Symmachus for this very verse, for which the Hebrew reads: cry~ 1P~ :i"!C?1J c1i.l'7 .~ 'W~-"~7· Almost simultaneously Annie Pralong published an article focusing on the issue of the Christian reuse of the marble block, accompanied by several photos. 5 Her dating of the Psalm inscription and menorah to the fourth century appears to rest largely on the iconography of the motif. In 2001 Constantine Zuckerman presented a further improvement to the reading, using photos provided by Pralong. 6 He pointed out that there was no room for the definite article before the participle at the beginning and that one should read 6L6ou~ aplOv 1ft :rtaoU oaprd, OlL Et~ atoova EAEO~ aulOu, 'giving bread to all flesh, because his mercy is for ever.' This is the reading accepted by the most recent editor, Ameling, in Inscriptiones Judaicae Orientis VoU!. 7 The verse cited in the inscription, Ps 135(136):25, is often used in benedictions. In fact it occurs in the first and probably the earliest blessing of the Birkat Ha-Mazon, the blessing after meals. Tradition attributes this section to Moses when the manna fell from heaven. (According to Mishna Sotah 7.1, the Birkat Ha-Mazon could be recited in 4 5 6 7

Fine and Rutgers (l996: 1-23,1998: 123-25). Pralong (1998: 603--609 and plates I, Ill, IV and V). Zuckerman(2001: 105-111). Ameling (2004: 321-24, no.I53).

214

ALISON SALVESEN

any language, not just Hebrew.} This connection would seem to rule out Sahin's suggestion that the block was originally a funerary monument and to support Fine and Rutgers's argument that the ashlar was either part of a synagogue, and the verse engraved on it came from a Greek language liturgy, or it came from a soup kitchen. Zuckerman favours the synagogue hypothesis. He speculates that the block was once part of a Torah shrine, and that it formed the front column of an aedicula. However, both he and Fine and Rutgers emphasise that the ashlar provides evidence for a synagogue in Nicaea in the Later Empire. As for the wording of the inscription, with Fine and Rutgers Zuckerman points to the discrepancy between the LXX version of Ps 135:25 and that of the inscription, and the resemblance to the recorded versions of Aquila and Symmachus for this verse. Furthermore he notes that his revised reading, without the definite article before OLOOUr;, corresponds to the reading attributed to Symmachus in certain medieval manuscripts of John Chrysostom's Homilies on the Psalms, edited by Mercati, OLOOUr; ap'tov "giving bread".8 Aquila's reading is given as or; OLOWOLV, 'who gives', but a slightly longer version of each reading also appears in two further manuscripts as the work of 'other' translators: aUor;' or; O(.OWOLV ap'tOv 'who gives bread" and aUor;' OLOOUr; ap'tov 'giving bread'.9 Zuckerman states, 'there is hardly a doubt that the inscription from Iznik provides for the first time the complete translation by Symmachus of Psalms 135:25.'10 Zuckerman notes that there has been some dispute from antiquity to the present day over the religious identity of Symmachus, some authorities believing him to be an Ebionite,ll others a Jew.l 2 He draws attention to the place in Palladius' Lausiac History 64 where the author appears to be quoting directly from a note by Origen referring to Symmachus as the translator of the Jews. Zuckerman believes that the use 8 9

10 11

12

Mercati (1952: 87-88). See Mercati (1952: 88), and Field (1875: 290b and n.8), where Field notes that de Montfaucon attributed the first reading to Aquila and the second to Symmachus. Zuckerman (2001: 107). Including Eusebius of Caesarea (Hist. Eeel. VI.17, Dem. Evang. VII.l.33), Jerome (De Vir. m., 54), and in the mid-twentieth century Hans-Joachlm Schoeps. Epiphanius (de Mens. et Pond. 16), Palladius (Hist. Laus. 64), and more recently Dominique Barthelemy (1974: 451-{)5), Salvesen (1991: 289-90), van der Kooij (1988: 1-20).

A JEWISH GREEK INSCRIPTION FROM NICAEA

215

of a verse from Psalms in Symmachus' version in a synagogue context in Asia Minor provides 'the first empirical proof of the use of Symmachus' translation of the Bible in a Jewish, arguably synagogal, context' , and that Symmachus was indeed a Jewish translator of the Jews.13 Attributions to the later revisions of the LXX are, however, often contradictory or confused.l 4 This is not surprising as many were copied rather mechanically over the centuries, in the margins of manuscripts, by scribes who had no idea about their provenance or significance. In this particular case the attributions appear in medieval manuscripts copied long after the lifetime not just of Aquila and Symmachus but also of Chrysostom. It is often possible to suggest an attribution for an anonymous reading, or to correct a named attribution, where the word or phrase is typical of the style of one of the revisers in other places, especially if the treatment of the Hebrew lexeme or idiomatic expression is characteristic of a particular translator elsewhere. Unfortunately, in this particular case, the Hebrew of Ps 136:25 is so simple to translate that it invites no idiosyncratic lexical choices that would enable us to identify it with a known revision. For instance, it is not at all remarkable that CO? is rendered by ap'to£ in more than one revision of LXX. However, it is possible to examine the syntactical translation technique to see whether it is in harmony with the approach of either Aquila or Symmachus in other passages. Though the preserved readings of the 'Three', Aquila, Symmachus and Theodotion are fragmentary, and often only the more interesting changes from LXX have been recorded, we have here four features that can be studied to see if they are paralleled elsewhere by the Jewish revisers: a)

the translation of the active participle without the article or relative pronoun.

b)

the rendering of "~~-'=?? by article

c)

the rendering ofc?i'!)? by El~ at&va without the article

d)

the rendering of i"ltjm by ~A.EO~ autou without the article

tfi

lljj

by

bLbou~

Jtaon oapd with the

13

Zuckerman (2001: 111).

14

For instance, Frederick Field (1875 passim) often discusses attributions in his extensive Latin footnotes, and suggests corrections to them on the basis of the style of the rendering.

ALISON SALVESEN

216

All four items can be compared with usage in other places. Kyosti Hyvarinen has published a study of Aquila's style. Busto Saiz and Gonz8lez Luis have worked on Symmachus' translation technique in respectively the Psalms and the Major Prophets. a) Aquila tends to translate active participles by a Greek finite verb in the present, especially when the Hebrew participle stands as a predicate. However, he uses a Greek participle when the Hebrew participle has an attributive sense,15 In Ps 102 (MT 103), for which we have substantial remains from a palimpsest in the Cairo Geniza, there is a series of active participles describing God's nature and activity. The first few (vv.3-5) are preceded by the Hebrew article, which Aquila faithfully reproduces in his Greek. But in v.6 nip1l:' ilf!;ll is rendered nmwv ('>LI f9" J'1l"C 28 J thl1 (sic) .eft."

29 J~'ttJ..C 30 J "-9"'1 1l000(IIo.(.eo 31 J f.,Li'tA 09".eoC

.e-~'i" A thi'tC

0-ft.32

1I,..llfto

h,e.",.. "",)(1t\. fDtt:C I 38. *Mt1'" ht\. "ilC;33

39. OOfto:f" IJAilOC+ IJAO~OCJ

*,eft.U il,..OCJ20

*O'"lb ,eft.U foo-f:C 172 106. *l1il"l' .e~tJ73 Octt: *f+AfI ~(J)r hl)ihCI r'}r J\,f~hC I 146. J\fD"2\"+r J\,fC;'7CI 147. ,ffD"AUA "'JI~-flCI 148. It.eoo''-flU J\.erc 13 9. 140. 141. 142. 143. 144. 145.

85 J .,..dJC::

GETATCHEW HAILE

264

304. f.c;.,.f-fll\-il A'fl- 006C~ 305. ooAta:fdJ< 'I'ckC~ 306. 307. 308. 309. 310. 311.

"~"':fdJ< 4s'l'C1

"X,.f:fdJ< oo9"hC; 'I'C,..:fdJ< 1IC1IC+ h,.f1l'l-9" i\ooilOC1 Offio"'i OhC"}:: ,.f(t~i\c mf- OCJi: "'1')0"}

332. ~f(f. 79v)

HC:: 313.9""} tJ'Hlli "CJi: ,.f"}1:i\:" "'1 J.>..C O"'l-flli~ 314. f~ HC oo""'oc 1.11. O"'~lidJ< i\-fI~li+

315. hoc:"9" lIi\li 01i\..41 i\-fI~li::

316. ,,"}.c;. erc 00~1I~ "'1i\:",.f

317. 318. 319. 320. 321. 3 22. 3 23.

,.f 1.erldJ< ",..:,. 1.041~ A Hm:"c+ f/bi\dJ< ilt;:C:: i\~'I''' fi\dJ< il oo 1IhCI 1.mC1:lI~A It."}h-lh-CI Hi\'M9" 1.'illC: ",,}"'I.c;.u flM9" oohllOc+ f"'~il "'1'illOC + ft;:6~ ooilOC +

324.01l4."} 411. oo7i"X'fPC+ 3 25. itC{l 1.tf1lC + 326. hil;JCP. 00'i;J1C+

.,.m

327. fll.;J "'111~1C+ 328. "}", .,.t..l1: ,,.. .,..,.~~

"(tC+ 3 29. mi\lI (t1lC" "ilh. (tllC + 330. "'"'1J.>..U .,.t..1t00 .,.ooC:: 331. 9" P'~:" "ilh 9"6~-fI (t~1 Mlh J.>..(I.-fI

'loo:"

fH'iOOdJ< 1ICJ9"+ 333. fO41:" lI;t-101:A+ 348. lIml ",..:,."} OO'P"'IJ.>.. ",..:,."} "''if"'lt,

349. h,.fi\1Jo Oh"}t;: "'J.>.. OlC+ 350. hoo "}ilC H'ifC:: 351. 1.U9" f'l'Oll:""} 'i1C 352. "'9"C" f~,.fil"'9"UC+ 353. :J,.;t-dJ. h t...,.:t: 92 Sic, for """'9"IJC

AN ARCHAIC AMHARIC POEM

f'l:"r t\rAC::

"':".,.cr Ht\«}t\r 11.f).

1\.-l-Cil().'i,

cfI°Oi: t\"'l.fl1t\oil Him h~n aJ(f. 1t h"}.('I aJnilc;"'i,V- ~:: H""~hO

"").~ HOI'lD?.f-l-::

80r)~~i: ~f aJnf.v"oo~thL P'm.C::

r

fg

265

aJ"'()~"}tv-

Translation (f. 74v) In the name o/the Father and the Son and Holy Spirit,

One God.

1.

This tract is a condemner 0/

glory/wealth. 1 2.

I know that which passes on

earth. 3.

I shall proclaim the thing( s)

(f. 7Sr) by which people are 4.

justified. You who do not listen to this, I adjure you in the

11. It will remain incomplete. 12. Now, desire not glory. 13. Regarding the world, its height is short. 14. It likes to fornicate and to be lascivious, 15. To roam around (?) and 16.

nameo/God. 5.

Regarding the world, its

17.

height2 is short. 6.

7.

It does not amount to a span. It is less than a fingernail. Wheat with pea,

8. 9. Te! with lentil, 10. However one counts

(together, their number) will not add up to ten,3

2

The primary meaning of hllC (bbr) is 'glory'. Since glory and wealth are related, it is also used to refer to 'wealth'. Furthermore, bbr and bbur are interchangeably used to refer to the person who has k~br: 'glorious', 'magnificent', 'honourable' or 'wealthy'. See the Introduction. 1. e. 'life' or 'duration'.

18. 19.

20.

21. 22. 23. 24.

3 4

defraud. 4 Its leg(s)!feet look like that of a soldier, When it comes back to be subservient, When it shakes the wealthy; When it makes the poor wealthy; When it makes lascivious the one who tries to be righteous, Bringing (him) out of the desert and wilderness. Now, desire not wealth: Be it sternum or leg (meat), Be it flesh of the hip or stomach lining,

10 is a symbolic number for 'completeness' . Perhaps ma~iirbar.

GETATCHEW HAILE

266 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

38. 39. 40. 41. 42. 43. 44.

5

6

Be it lamb or leg of a fattened animal, (Or) a cut of a capon, A cooked fenugreek, Boiled in salt, pepper, butter and ginger. Now, desire not wealth: Be it a boiled5 provision for ajourney6 Be it matured mead, Be it inebriating wine, Be it milk and honey, Be it banana and sugar-cane, They will be wasted; They will be spilt on earth. All this would not be obtained in vain, without toil, Without having others work hard, Without an assault with a club, without looting, Without fighting with the spear. Make no (f. 75v) effort for this, For the sake of insignificant thing(s), To have the dance (of praise) of attendants, To be surrounded like a fence (by attendants) carrying arms,

45. To be followed like a national militia.7 46. To boast being on horseback, 47. To heap scalps, 48. To assault with ajavelin,8 49. To promote and demote, 50. To bind and knock down, 51. To cut (to pieces) without justice. 52. Now (desire not wealth), to throw out one in the field, 53. To plot by night,9 54. To tell lies, 55. To give a false witness, 56. To plot evil against others, 57. To hurry up to death,1O 58. To appropriate what one does not own,11 59. To be arrogant (?) toward a priest.l 2 60. Regarding a woman, whether she is rich, 61. Old or a girl, 62. Pubertal or juvenile, 63. Red or black, 64. Tall or short, 7 8

Or 'fusiliers' . According to sitwasaw manuscripts, siiiapn is riimb 'javelin', 'spear'. 9 Or 'To get up early in the morning'. 10 I. e. 'to kill'; cf. Ps. 108:17,

109:16.

Or 'purified'. 'Wegzehrung', see Littmann

(1914: 483).

11 Or 'saw'. 12 Or 'be bound by the priest,' see

Littmann (1914: 490).

AN ARCHAIC AMHARIC POEM 65. And her teeth could look like a chest, 66. And her eyes could look like the moon, 67. The height of her neck could be neither a (full) cubit nor a span. 68. (But) Satan will show you her appearance (brightening) it like a sunray. 69. She may have her hair braided, and make it bedazzling, 70. She may stand up and sing, 71. She may talk seductively, 72. She may be omamented and dance carrying a drum, 73. She may be wispy and showy, 74. She may kiss the lips, 75. She may approach and cool, 76. Her (face) may drip oil, 77. Her fragrance (f. 76a) may make you delirious, 78. She will drop you like excrement; 79. She will expel you from the house of God. 80. Regarding all this, 81. Let it not attract you. 82. It is Satan's counsel. 83. It is the Devil's love. 84. Regarding the end,13

267

85. It is (worse than) a bitter

koso, 14 86. It is (worse than) (amqaql5 87. 88. 89. 90.

or aloe. Having all this matter Heard, stay guarded. Fornicate not; and be not lascivious like me. Wealth is bad; it is poison

(?): 16 91. Be it sword or dagger (?), 92. Be it silk belt or headdress, 93. Be it a tent or a cloth 17 canopy, 94. Be it fluffed cotton or wispy cotton 18 95. Now, desire not a heap of shirts, 96. To be enriched with widesleeved shirts. 97. Regarding the coming glory, it is brighter than the land, 98. One that is glaring for the eye. 99. Now, desire not gold and silver, 100. Which would be kept in a box. 14 A bitter plant used as medicine against tapeworm. 15 A variant of pmq, 'poison extracted from euphorhia' . 16 The word is siidiir, perhaps for

sediir.

13 J adds 'when you rise from the grave'.

17 Or weaver-made. 18 diifar; J has yiitiiniiddiifd •carded cotton'.

268

GETATCHEW HAILE

101. It is a pile (?) that has no example. 102. Now, desire not braces of left and right, 103. A ring for the finger, 104. Be it a throne or a tautened (seat), 105. It is but a stretched wood. 106. Be it carpet or bed cover, it is flax plant. 107. Be it guh 19 or biirnos,20 it is sheep's fleece. 108. Be it qaft.an 21 or gamga,22 it is (f. 76v) flax plant. 109. A flag or silk niigarit,23 it is made like a drum. 110. A mule they call samburs,24 111. Which does not shake (but moves) with dignity, 112. Even if it flows like water, (or) if it is swift like a horse, 113. They call its name mantamar,25 114. Decorated in gold, woolen garment and in silver. 115. Now, desire not wealth, 116. Dressing a horse in leopard (skin),

19 Red woolen cloth. 20 Black felt cloak. 21 Silk shirt. 22 Muslin. 23 Kettle drum. 24 Resembling the chickpea, from sambra 'chickpea'. 25 'What a beauty!'

117. Or /:lambiiz26 or siirf27 or a

siimma28 spread 118. Ambiilay29 or asgar30 119. Borboti31 or cuirass.

120. Silk niigarit32 or silk 121. Javelin or coronet 122. Rifle or bullet,33 123. Which flies 34 very high. 124. All goes to the grass; it will not be for glory, 125. It will pass on (this) earth; 126. Worldly treasure is money tied and knotted. 127. It will not help one to live by being a guarantor; it will not redeem you 35 128. Before God. 129. Let me give you a word of advice: 130. Spend not the night sleeping.

26 27 28 29

30 31 32 33 34

35

Military ceremonial dress. Cover. Carpet. Garment. White horse. A kind of horse. Iron helmet, variant ofborborti. The word is labar (?). J has 'ariir 'bullet' . The word is yiimmisiirr; but it could be yammisiibr. The word in E is ayabbiiggiirakiimm, perhaps a corrupted one. In J it is tiiiiito, obviously a corruption of tiibiiZto, from G;l';lZ beziiwii 'to redeem'.

AN ARCHAIC AMHARIC POEM 131. Before the time-cycle is completed upon you, 132. Before the leather cord (for your shrouding) is cut for you, 133. Before the merciless comes upon you, 134. (Before) what they call a statue, a pile of stone (rests on you), 135. Sewn with iron, 136. A timber tied tightly with leather cord, 137. One without limit (?)36 138. Before the merciless comes upon you. 139. He who flies away like a bird, 140. Who climbs a tree, 141. Like an eagle of the air, 142. Cannot enter in the wilderness, 143. Or in a cave or ditch (to escape). 144. (f. 77r) He, too, who sinks in the sea 145. Cannot estrange himself. 146. He will bring one out but will not let one talk. 147. He will rather keep you all day investigating. 148. Before the merciless comes upon you, 149. Which is insolent (even) to a king.

36 Or 'The other with tar' (?).

269

150. It does not care for a ruler or a teacher, 151. It does not care for the poor or the rich, 152. It does not care for the tall or short (person), 153. It does not care for the weak or blind, broken or split. 154. The angel of death, the creature, does not care. 155. The appearance ofhis face is black. 156. He binds with chains. 157. Whatever you bought and looted will remain where it is. 158. A stranger will consume it. 159. Now, let me give you a word of advice: 160. Before they shroud you in grass, 161. Before they send you off by means of the tomb, 162. Before they load earth on you, 163. Oh warrior, ofa white horse, (before you go into) a sea from which you cannot come out, 164. Clean the debt on you; 165. Search your soul and body; 166. With a good teacher, 167. Who speaks from the Scriptures, 168. Beginning from the root, 169. That what God has commanded,

GETATCHEW HAILE

270

170. That by which people are justified, 171. That urges one to flee (to the desert), 172. Before your mouth is shut, 173. Before your eyes are blinded, 174. Before your mind is lost. 175. Commemoration37 would not justify you. 176. Give lodging (?) to those who come. 177. Have for the hungry (food) baked; 178. Cool the thirsty; 179. Clothe the naked, give lodging to the tired. 180. Do not tell lies. 181. Help the prisoner to be released. 182. Investigate if there is a sick person. 183. Love your friend. 184. (f. 77v) Honour your father and mother. 185. Fear God, 186. Counsel; the name of this is harvest. 187. Let not the harvest elude you, 188. TalJSas and Ij()dar. 38 189. The city of the Kingdom of Heaven 37

190. Is brighter than the sun. 191. And if you are caught (by the angel of death), do not be horrified. 192. The blessed God justifies one by (his) generosity 193. What is important to him is 194. That you abhor a campaign leader39 195. His superfluous goodness is without measure. 196. He will keep you in glory, 197. Together with the Apostles, 198. Reminding one of the Holy Spirit, 199. (At) the Kingdom of Salvation, the mountain. 200. He will make you king on the throne, 201. With your countenance being comely, 202. So that your countenance may live looking like the sun and the moon. 203. The beloved one of Our LadyMary 204. Will always satiate you. 205. Evil will find him, 206. Who violated the commandments of the Gospel, 207. Who passed (his time) in disgraceful acts, 208. Who destroyed a fence 40

tiizkar.

38 These are the two months of the harvest season, roughly December and November.

39 Perhaps 'a warmonger'.

AN ARCHAIC AMHARIC POEM 209. Of a knowledgeable man. 210. Let me give a word of advice: 211. (Regarding) the impending unrest, 212. Going to the top of the mountain 213. Should have been earlier. 214. Still, there will come from Dagen the worst rolling object, 215. Worst one from Zator, in Baranta 216. Worst one from Mugar, in 5awa, 217. Worst one from Zanmalat, in the (whole) country. 218. When the good (Lord) comes, 219. Worker o/miracle and wonder, 220. Having fulfilled everything, 221. In order that he may separate the wheat from the chaff. 41 222. (Treating all equally), as equal as (the two) spades (of the plough),42 223. (f. 78r)(To judge) those on the right for glory and on the left for disgrace.

40 This may refer to farmers who push back the border of a neighbor's farmland. 41 sinnar 'oats' (?).

42 dagar.

271

224. A witness would not benefit you,43 225. For litigation at (the judgment for) eternal (life), 226. (Where you wish to stay) let me stay only this day, 227. Let me counsel with my children and wife. 228. The word of God is tested. 229. If one has entered a house of (monastic) community, 230. If one has fasted, 231. If one has one baked (bread) when one is hungry, 232. The priest will speak out as a witness, 233. His mind and heart will gush (words). 234. (The angel of death) will drive you away in haste (as) with a stick. 235. Just as a lot is ill placed at the joint,44 236. (The fool) does not talk with wise people; 237. He does not clear the field during the dry season. 238. He does not plough during the rainy season, 239. (He is) always engaged in rivalry/competition, 240. Spoiling the faith, 241. Dancing with the wicked ones, 43 Literally: I shall not look for a witness for you. 44 Not clear.

272

GETATCHEWHAILE

242. Living in (the sin of) fornication. 243. He does not have mercy when asked to have mercy. 244. He does not remember his God. 245. He does not fear God. 246. His mind is lascivious. 247. He tries (?) not to die. 248. He depends on others. 249. He does not have for others (food) baked. 250. He does not cool the thirsty. 251. He does not give lodging for the one who is tired. 252. He does not help obtain a prisoner's release. 253. He does not investigate if there is a sick person. 254. He does not honour his friend. 255. He abolishes the Sabbath and holy days. 256. He returns to do violence. 45 257. He is a disciple of Satan, 258. He is the first born of the Devil, 259. The disgraced will not be released. 260. He has abandoned his creator, and has been subservient to a creature. 261. Now, let me learn from a teacher. 262. You should not think that (every thing) is all right. 45 ef.lines 16 and 17.

263. You will (f. 78v) depart for good. 264. You will not come back next month. 265 . You will go through the wilderness, 266. Being confused, 267. In the hot sun, in tearful sorrow, withyow46 bodily parts boiling, 268. As a purifier would fire with pebbles of Adaqo,47 269. When the soul and her body are purified, when they melt48 like wax. 270. The yoke of Hades 271. Never breaks however one tries, 272. Sitting on you like (on) a chair. 273. As with a pickax, 274. They would split acacia (wood), 275. You will bum like a bush, 276. Like an olive tree and adiir. 49 277. Regarding the manner of letting one fall down, 278. It is more than (what) a rifle and a dagger50 (can do); 46 Literally: 'his'. 47 As butter is purified by heat. 48 Literally: 'he melts'. 49 This is presumably a kind of tree, like the acacia. 50 qWi;irr is apparently a variant of

qarra.

AN ARCHAIC AMHARIC POEM 279. It is more than (what) cannon and shell51 (can do). 280. The teeth of Hades are not comely. 281. They will stretch you like a string 282. They will suspend you like a baldachin. 283. Woe that there is no good deed in you! 284. There will be no hiding place, when a heap of fire will sit on you, 285. Heavier than heaven and earth, 286. (Even) than mountains and a city. 287. It will come down on you, ascending (on you), 288. Heaping itself on your head. 289. There will be no one to rescue you; 290. And you will not die and be buried. 291. Nor will you flyaway like a bird. 292. You will just live on in anguish and agony, 293. With your countenance

273

295. Piercing you like an

asbur. 54 296. Regarding the thing that will devour you, 297. (It is like) the desert snake that kills; 298. It is more than a snake (f. 79r) and a wood tick; 299. It is worse than a plague; 300. It is more than cow's hair (?). 301. It makes one feel dripping (urine) 302. As a cow of (the month of) T~rr that

is caught (in)

294. Making a stink at you like a marten 53

drought, 303. Is eaten by (?) a cyst. 55 304. The children of the Devil are bitter. 305. Their countenance is black 306. Their heads are tied (?). 307. Their engagement is in plotting. 308. Their teeth are apart from each other. 309. They never feel sorry when assaulting, 310. With fist or elbow. 311. They make one say, woe, I wish in my mother's womb 312. I failed like the seed that fails lacking rain.

Probably related to Arabic na'ara 'to pierce'. 52 One possessed with a prophesizing evil spirit. 53 fifar is apparently a variant of faro.

54 'Porcupine' (?), not attested in the dictionary. 55 Perhaps 'Their hair is curly'.

looking like a zar52

51

274

GETATCHEW HAILE

313. What would have happened to you, my mother, if you had not spent that night! 314. At the time that my seed was conceived, I wish you had slept alone, 315. Coiling yourself, separately on your own. 316. There came upon me to live an endless fire; it will devour me eternally, 317. That which is without measure. 318. The sinner does not have a memorable name. 319. (Torture) will come down on me rolling, 320. To live forever. 321. Now, the worldly mutual aggrandizement, 322. Heaping of shirts, 323. Killing a fattened steer, 324. Showing off on a throne, 325. Having jest with ladies, 326. Talking behind a curtain, 327. Mistreating the poor, 328. (Commanding,) come, be judged; come, be bound with him, bind. 329. As hard as breaking a wiiliiba (?). 330. Now, the time cycle is finished. 331. From east to west, from north to south,

332. (With) the rain (f. 79v) that fell eight hundred thousand years, 333. The grain and grass that grew, 334. Added (to it) pebbles and woods, 335. People and birds of the sky which fly with wings, 336. And all the stars that are in the sky, 337. Ifone counted from (one) year to another, 338. The cry that one sinner will cry 339. Will be more in measure. 340. Forever with the rod of Hades 341. They will flog him seventeen (times every) night. 342. One two liibiir 56 343. Are a ladder (of) doubt, 344. Like a seal57 in high seas, 345. As a wise person would cross swimming and diving, 346. Without his mind fearing (?). 347. You, too, if you deduce from these words which I have spoken to you, 348. Crossing the sea of fire, and the waves of fire, 56 The word could be laba: 'One or two labar', rather than labiirr 'for the door'. 57 qWiitlam (?).

AN ARCHAIC AMHARIC POEM 349. I shall see (arriving) like one who flies with wings, 350. Like an eagle ofthe air. 351. And these words of wisdom 352. (For) him who learns and teaches, 353. Who teaches believing and fearing that the Lord will come. 354. Let me tell you about the fools, too:

275

355. Clothed in darkness, fire will emit eternally from their bodies. 356. The earth will swallow them opening its mouth. It will not remain open forever; it will never have judgment of mercy.

Forget me not (but recite on my behalf) the Our Father in Heaven. The height of the Devil is 10,750 (f. 80r) (cubit); his chest, 700; his mouth, 400; his eye, 80 cubits; and his teeth, 60. (This is) what is found in the Mii~hafii Mas(ir (,Book of Mystery').

REFERENCES EMML = Ethiopia Manuscript Microfilm Library, Hill Monastic Manuscript Library. Getatchew Haile, 1983, 'Old Amharic features in a manuscript from Wollo (EMML 7000)" in S. Segert and AJ.E. Bodrogligeti (eds.), Ethiopian Studies Dedicated to WolfLeslau, Wiesbaden, pp. 157-58. Getatchew Haile and Misrak Amare, 1991, Beauty of the Creation (Sanii Fatriit), Manchester. Guidi, I., 1889, 'Le canzoni geez-amarina in onore di Re Abisini', Rendiconti della R. Accademia dei Lincei, Classe di Scienze Morali, Storiche e Filologiche, vol. V, Sem. 1, Rome, pp. 54-66. Littmann, E., 1914, Die Altamharischen Kaiserlieder, Strassburg. - - - - " 1943, 'Altamharisches Glossar. Der Wortschhatz in den "Canzoni Geez-Amarina"', Rivista degli Studi Orientali 20, pp. 473505. Strelcyn, S., 1968-1973, Medecine et plantes d'Ethiopie, Warsaw and Naples.

INNOVATION AND MISONEISM DURING THE REIGN OF EMPEROR YOHANNES IV OF ETHIOPIA (1872-1889) Richard Pankhurst Introduction

The process of Ethiopian modernisation, which Emperor Tewodros had envisaged and had been interrupted by his suicide on 13 April 1868, continued during the ensuing reign of Emperor Yohannes. The object of this paper is to assess such innovation against the back-drop of the country's misoneism, i.e. opposition to change. Some of the developments of this period, such as the arrival of new imports, including medicines, were due to improved transportation and increased contacts with the outside world; others, among them opposition to slavery and the slave trade, resulted from events on the international stage; others again, in the field of land tenure, class relations, the growing diffusion of fire-arms, and depletion of wild life, originated in the internal dynamics of the age. Most such changes occurred independently of Emperor Yohannes, and were market-motivated. Other developments, notably the allocation of land to the Church, the forced conversions of Muslims, the banning of missionaries, the prohibition of tobacco, slavery, and the slave trade, the introduction of a national flag, the appointment of the country's first consul abroad, the establishment of a church in Jerusalem, and increased use of foreign-educated officials, all reflected the Emperor's personal predilections, and were mainly effected by imperial command.

Conflict, Tranquillity and Renewed Conflict

The period immediately after Tewodros's death was a time of difficulty. Dajazmach Kassa, the future Emperor Yohannes, defeated his rival Wagshum Gobaze at the battle of Assam, on 11 July 1871, and adopted the throne name of Yohannes. This victory opened a short but auspicious

INNOVATION AND MISONEISM

277

period of peace, conducive for development. Civil war, long the bane of the country, came to an end. This was a time, as the Earl of Mayo noted in the 1870s, when 'tranquillity prevailed', and people could 'cultivate their land in peace' .1 This period of peace was, however, soon interrupted by destructive warfare. Fighting with the Egyptians, in the 1870s, led in the north ofthe country to considerable destruction of buildings, and depopulation, as well as to the disruption of trade with the Red Sea coast. The British traveller Emelius De Cosson reported that the old settlement of Asmara was 'almost deserted, and the plains around it depopulated, for, being so near the frontier [with the Egyptians at the coast], they [the people of the area] had been exposed to the ravages of the Egyptian soldiers, who had made many pillaging expeditions and caused most of the inhabitants to fly into the interior for safety'.2 Emperor Yohannes's victories over the Egyptians, at the battles of Gundat in 1875 and Gura in 1875, preserved Ethiopia's independence, but did not restore stability. Raiding was continued by a rebel chief, Ras Walda Mika'el, who, according to A.B. Wylde, 'commenced to raid and devastate the seat of his old Government, and turned the Hamasien plateau, formerly known as the plain of the thousand villages, into a howling wilderness of ruined houses, with a few half-starved peasantry' .3 The subsequent war with the Dervishes in the 1880s, which resulted in the sacking of Gondar, led to major ravages in the north-west, and to the disruption of trade with Sudan and Egypt. Powell-Cotton noted over a decade later that evidence of destruction was still visible: 'The nearer we approached Gondar, the more numerous became the ruins of the villages, farmsteads, and churches. Up to the Dervish invasion Gondar was a considerable place. Many rich Nagadis [Le. traders] lived there, but nearly all these lost, not only their wealth, but their lives. Now there is only one large merchant residing in the place'.4 Such destruction naturally militated against development and the introduction of innovations, outside the purely military field.

Mayo (1876: 219).

n, 150).

2

De Cosson (1877:

3

Wylde (1901: 28). See also Pankhurst (1963: 151-5).

4

Powell-Cotton (1902: 301).

278

RICHARD P ANKHURST

Advances in Transport and Communications

Ethiopia's contacts with the outside world were facilitated by the cutting in 1869 of the Suez Canal, which turned the Red Sea into an extension of the Mediterranean. Communications were further improved by the Egyptians, who developed the port of Massawa in the 1870s, by building a causeway to link the island with the mainland. In order to penetrate militarily, they also constructed a road from Addi Qwala to Gundat. 5 The Emperor, on the other hand, was slow to improve the road from the Ethiopian highlands to the coast. According to his British assistant John Kirkham he 'preferred to keep his money hoarded up'.6 The German traveller Gerhard Rohlfs declared that Y ohannes was more interested in building churches than roads.7 Some road-work to the coast was, however, carried out by Swedish missionaries at Monkulu, inland from Massawa, who cleared a track through jungle country.8 Wylde nevertheless declared that the road up to Ginda'e was 'a nasty climb'. The Abyssinians, he explained, had not improved it, for fear of foreign invasion, and he comments 'I suppose they are right'.9 Transportation was later facilitated by the opening, by the Emperor's commander Ras Alula, of a good road between the highlands and the coast. Francis Harrison Smith was surprised in 1890 to find that the track had been 'cleared of obstructions', and was 'in many places so broad and level' that one could safely drive a coach along it.10 The Italians, after seizing Massawa in 1885, also actively improved transport facilities in the growing area they occupied. Gerald Portal commented a few years later that 'the horrors and dangers' of the old mountain-path from Asmara to the coast had been replaced by a 'wide and well-constructed road'. This he noted was 'fit for the passage of cavalry and artillery', both of which were soon advancing into the Ethiopian interior'! 1 Improvements in the Massawa road led to the growth of the nearby settlement of Monkulu, where merchants tarried when the heat at the 5 6 7 8 9

10 11

Mayo (1876: 42); Wylde (1888: n, 45); Pankhurst (1968: 288). Mayo (1876: 219). Rohlfs (1883: 238). Mayo (1876: 42). Wylde (1888: 1,204). Smith (1890: 239). Portal (1892: 250).

INNOVATION AND MISONEISM

279

coast was unbearable. The place, outside Emperor Yohannes's control, was the site, as we shall see, of a Swedish mission station. 12

Expanding Trade and New Imports

The principal focus of commercial innovation in northern Ethiopia at this time was Adwa, which for a century had been Tegray's main commercial centre. Its market handled an influx of new imports. De Cosson, in the mid-1870s, remarked that 'traders from the coast' brought 'common glass bottles for drinking tej [honey wine]; and even round lookingglasses'.13 A decade later Wylde noted that Adwa's imports included 'cottons of all sorts from England and the Continent, cotton prints of many sorts, silks, satins, Birmingham sundries in the way of cutlery, beads, needles and Bohemian glass-ware, kerosene lamps' and 'Africa trade in general' .14 Harrison Smith confirmed the import of' little French mirrors', as well as 'unbleached cotton goods' from Bombay and superior Egyptian shirting, bearing a trademark of Pyramids and camels. Manchester goods were 'chiefly represented by the manufactures of Messrs. Tabbush'.15 Increasing European imports were also reported at other markets, notably at Kudo Felassi, in Saraye. By the late 1860s its traders were using Egyptian coins, and by the early 1870s dealt in many foreign imports, including needles and nick-nacks.I 6 New imports were taken to lesser markets far and wide by itinerant merchants. Their stockin-trade, consisted, according to Wylde, of pins and needles, cotton cloth and a 'variety of Birmingham goods'. 17 The profit on such articles was 'enormous' .18 Some of the most colourful imports in this period were silk umbrellas, which were carried above the Emperor, great church leaders, and nobles. 'The King's umbrella and that of the Chief Priest', Wylde exclaimed, 'were the most gorgeous and wonderful fabrications, made in

12 13 14 15 16 17 18

Pankhurst (1984: 138-9). De Cosson (1876: I, 112). Wylde (1888: I, 284). Smith (1890: 155-6.) See also Simon (1885: 150); Rohlfs (1883: 161-2). Girard (1873: 63-4); Raffray (1880: 62). Wylde (1888: I, 292). Wylde (1888: 1,284).

280

RICHARD PANKHURST

Paris, purple silk covered with handwork flowers and gold sticks',19 Theodore Bent notes that such umbrellas were 'used by great men solely as a mark of distinction'. Red umbrellas could thus be worn only by 'great princes', and white ones by 'ordinary officials', while the common people had locally-made ones constructed out of grass. 20 Another spectacular import at Adwa towards the end of the century was the metal roof of Sellase church, the first such roof in the country. Bent observed that the city had been 'spoilt irretrievably', and adds: 'This horrible roof catches the sun', 'gleams provokingly', and was a 'perpetual eye sore')1

Urban Developments The reign of Yohannes witnessed limited urban development. 22 The Emperor, on assuming the throne, travelled to Diibrii Tabor, and established a new palace at nearby Samma, where he also constructed the modem-style church of Heruy Giyorgis, erected by an Italian craftsman Giacomo Naretti. 23 Yohannes also stayed at Miiqale, a new settlement in Tegray, which he virtually founded. There he built the church of Miidhane Alam, and a fortress-like palace, likewise designed by Naretti. 24 At the end of the reign, Wylde reported that 'nearly every house' had 'a large garden with an irrigation channel' .25 Significant building was also carried out at both Adwa and Aksum. At the former Yohannes erected the great church of Sellase, and rebuilt the 18th. century palace. 26 Construction work was also carried out at Aksum, where at least one building was put up by Naretti. 27 Some important construction work likewise took place in the country of the Bahr Nagash, 19 20 21 22 23

Wylde (1888: n, 6-7). Bent (1896: 29-30). See also Rohlfs (1883: 217-18); Wylde (1888: I, 281-2). Bent (1896: 99, also 126). For a picture of the building when still thatched see Vigoni (1881: opposite p.100). Pankhurst (1995: 356). Rohlfs (1883: 196-202); Bianchi (1884: 44-5). See also Pankhurst (1995: 757).

24 25 26 27

Bairu Tafla (1977: 121-3). See also Pankhurst (1995: 120-4). Wylde (1901: 306). Bairu Tafla (1977: 93); Bent (1896: 126); Pankhurst (1995: 90-3). Bent (189: 108-9); Pankhurst (1995: 108).

INNOVATION AND MISONEISM

281

where Ras Alula established a palace at Asmara. 28 Portal describes it as 45 feet in diameter, with a 'lofty domed-shaped roof.29 Yohannes had one other notable urban achievement, namely the founding ofDase, soon to be capital of Wallo. The settlement, established in 1882, commemorated that year's 'Great Comet'. 30 Increasing Foreign Contacts

Innovation was also fueled by expanding foreign contacts. Increasing numbers of Ethiopians were then travelling abroad or to the coast, or consorting with foreign missionaries or other foreigners. The impact of foreign travel was apparent by the mid-1880s, when Wylde noted that the 'few' Ethiopians who had been to Massawa wore European-made trousers, 'either white or of some coloured washing material', with a coat buttoned to the neck. Such clothing was, however, 'hardly ever' worn without the traditional shamma, or toga) 1 Though banned by the Emperor, foreign missionaries played a modernising role on the country's periphery, where they provided a little European-style skilled training. The principal missionary centres were the Swedish Protestant school at Monkulu, and the French Lazarist institution at Karan. Wylde described the former as 'very well conducted'. It provided male pupils with 'a very useful education', training them as blacksmiths, carpenters, and masons, while females were taught cooking, sewing, and household work. 32 The value of such training was, however, limited by the static character of the economy, and the restricted demand for new skills. This was recognised by Frank James, who observed: 'the Fathers and Sisters confessed to us the difficulty of finding situtions for their proteges. Outcast from their own people and unable to find employment, they are thrown on their own resources, which proves more fatal to the women than the men')3

28 29 30 31 32 33

Wylde (1888: I, 222); Smith (1890: 107); Pankhurst (1995: 113-15). Portal (1892:

74-6).

Chaine (1913: 188-9); Pankhurst (1995: 235-{j). Wylde (1888: I, 283). Wylde (1888: I, 7). James (1883: London).

282

RICHARD PANKHURST

Ethiopians Educated Abroad

The idea of sending youngsters abroad for study was then still in its infancy, and did not in any case much appeal to Yohannes, who probably felt that it might endanger their Christian faith. Like Tewodros, he nevertheless made use of a handful of Ethiopians who had been abroad. Yohannes had three main foreign-educated Ethiopian assistants: Mahdara Qal Tawalda Madhen from Tegray and two men of partial Armenian descent, Mercha Warqe, and Yohannes Mashasha. Mahdara Qal, who had studied in Paris and Malta, played a notable role at court. He translated the Emperor's correspondence with foreign Powers and served as interpreter on several official occasions, including the visit of Colonel Charles Gordon,34 Mercha Warqe, who had been educated in Bombay, advised Dajazmac Kassa at the time of the British expedition to Maqdala. He was later accorded the title of Liqiimiiqwas, and despatched in 1870 on a mission to London. He subsequently negotiated with the Egyptians in 1875-6, and was involved in concluding the Tripartite Treaty of 1884, after which he undertook a mission to England accompanied by his nephew Yohannes Mashasha. Mercha also served as the Emperor's chief treasurer, in which post he was succeeded by the latter,35 Though utilising these three foreign-educated compatriots, Yohannes was apparently not enamoured by mission-educated youngsters. Martin Flad reports that when two returnees, Agagawi and Agagi, appeared before the monarch, the latter commented unfavourably on the fact that they had adopted the European practice of wearing shoes, saying: 'If you appear again before me, come barefoot: We Ethiopians do not wear shoes' ,36

Foreign Entrepreneurs and Craftsmen

By the reign of Yohannes several foreign entrepreneurs and craftsmen were active in Tegray. The existence of a saw-mill at Sabaguma was reported by De Cosson in the 1870s. Run by a Frenchman, it was situated in a valley 'surrounded on all sides by thickly wooded mountains',37 34 35 36 37

Pankhurst (199: 253-71). See also De Coursac (1926: 159, 169, 185,258). Wylde (1901: 42, 44,169,296-8,321); Pankhurst (1990: 259-64). Flad (1922: 280). De Cosson (1877: I, 36).

INNOVATION AND MISONEISM

283

Harrison Smith likewise met a German who had reportedly set up 'a number of power-mills' at Adwa.38 Several foreigners were likewise prominent as gunsmiths. The best known was Jean Baraglion, a Frenchman, who had lived in Ethiopia for over a decade, and, according to Wylde, enjoyed a monopoly at Adwa.39 Baraglion had, however, at least two rivals: A Hungarian called Andre, who also made artificial limbs; 40 and a Greek, who had lived in Shawa for several years. 41 Another Greek, Christopholos, had wielded an elephant gun for Ras Gobaze and had established himself at Gondar. People whispered that he was 'a magician', who practiced the 'black art' .42 An even more remarkable foreigner at Adwa was the German botanist Dr. Wilhelm Schimper, who had come to Ethiopia in 1836 and subsequently introduced water-cress. 43 His courtyard, De Cosson says, was 'full of inventions', including candles made of beeswax, and 'good champagne' made by bottling taj, with brandy and sugar. Schimper also assisted Y ohannes by writing to the German chancellor Otto von Bismarck.44 Yet another foreign innovator was an unidentified Frenchman, who introduced the cultivation of potatoes. Mayo described them as 'very small', but 'excellent eating'.45

Three Traditional Fields ofInnovation

Other innovations were the personal achievements of Emperor Yohannes. Many, like those of previous rulers, were in three classical Ethiopian fields of innovation: Palace-building, medicine and armament. 46

38 39 40 41 42 43 44 45 46

Smith (1890: 244-5). Wylde (1888: I, 293-5). Girard (1873: 240-1). Smith (1890: p. 195).

De Cosson (1877, n, 128). Wylde (1888, I, 245). Girard (1873: 17, 100, 121-9); De Cosson (1877: 1,119-20); De Coursac (1926: 91-3). See also Bairn Tafla (1981: 325, index). Mayo (1876: 121). Pankhurst (1964: 287-320).

284

RICHARD PANKHURST

The Palace and Imperial Dress y ohannes arranged for the construction of an entirely new palace at Maqiile. Designed by Naretti, it was an immense structure, at least partially inspired by the 17th century castles of Gondiir. 47 Harrison Smith recalls: 'The basement, which is entered by a heavy double door of Abyssinian wood, forms a banqueting hall. Leading out of this hall is a large audience or council chamber. From the basement one ascends by a double flight of broad, well-built stairs. The roof forms a fine promenade, from which an extensive panoramic view of the surrounding country is obtained, while the turret at each corner, and the castellated parapet have all been designed with a view to defence' .48 Another notable building constructed at around this time was at Saqota, where the local ruler, Wagshum Tiifari, had a Swiss carpenter, M. Dubois, erect a royal church. 49 Yohannes for his part was not averse to wearing new types of clothing. Though hitherto shoeless and dressed only in a shamma, according to Harrison Smith he subsequently wore a 'suit of blue and white striped tick' with thick woolen stockings, heavy boots, and a tweed coat. 50

Medical Innovation: The Emperor's Court Physician The second field of traditional innovation in which Yohannes played a prominent role was medicine. He had at his court in 1885-6 a Greek physician, Dr. Nicholas Parisis, lent by King George of the Hellenes. Yohannes was thus the first Ethiopian ruler to have his personal physician. Parisis introduced European-style vaccination, based on French serum, during the smallpox epidemic of 1886. Many prominent figures were vaccinated, including Emperor Yohannes, the Abun, or head of the church, King Menilek of Shiiwa, King Tiiklii Haymanot of Gojjam, Ras Alula, and 'many generals, officials, soldiers, and children'. 51 Yohannes was so convinced of the superiority of vaccination over traditional Ethiopian variolation that he forbade his subjects from practicing the 47

Vigoni (1881: 18-19).

48 49

Smith (1890: 228-9). Raffray (1880: 184).

50

Smith (1890: 96-7).

51

Parisis (1888: 130).

INNOVATION AND MISONEISM

285

latter. The force of this decree is confinned by Harrison Smith, who declares that the old practice was declared a 'heinous' offence. 52 Other medical deVelopments took place in this period throughout northern Ethiopia. In Adwa the aforementioned Hungarian armourer, Andre, made artificial limbs for persons who had undergone amputation in accordance with the traditional Ethiopian legal code. He charged ten Maria Theresa dollars per limb, but was also given presents of grain, meat, honey, etc. Girard53 reported that when Yohannes first saw Andre's artificial limbs he 'could not believe his eyes'.54 European missionaries, though excluded from the Ethiopian empire, had a medical impact on peripheral areas. The Roman Catholic Filles de la Charite treated around a hundred patients a day at Massawa and Karan, while the Swedish missionaries at Monkulu were also active. 55 New medicines were imported. Mercury sulphate was increasingly used in the treatment of syphilis, and by the early 1880s was widely used in northern Ethiopia. 56 The Emperor, Military Training; and the Acquisition of Fire-arms Yohannes, like his predecessor, was fully aware of the importance of modern fire-arms, which had been displayed for his benefit by members of the British expedition to Maqdala, and had proved decisive in their assault on Tewodros's citadel. Yohannes, like Tewodros, appreciated the value of modern European military skills, and was interested in European-style training. At the close of the expedition, he requested the British commander, General Robert Napier, to lend him a few soldiers to teach his men how to use the weapons he received from the British. Napier, however, refused, saying that the soldiers could not be left behind without the Queen of England's special orders. 57 Undeterred by 52 53

Parisis (1888:130); Smith (1890:113). Girard had many ideas for Ethiopia's modernisation. These included the publication of a monthly periodical to bring the country into contact with the outside world, the establishment of clinics in Adwa and Gondiir, an Ethiopian school in Paris, a trained royal bodyguard to render the monarch independent of feudal chiefs, a French consulate at Massawa and the cultivation of silk-worms (Girard, 1873: 17-19).

54 55 56 57

Girard (1873: 240-1, 262). Piolet (n.d.: 38). Parisis (1888:135-6). Hozier (1869: 260).

286

RICHARD PANKHURST

this rebuff, Yohannes sent emissaries to England in 1870 with a letter, reminiscent of Tewodros, requesting someone to teach 'arts or wisdom'. General Edward Stanton, the British consul in Egypt, reported, that the envoys were 'very anxious to induce English Engineers and Artizans to go to Abyssinia,' and added that they were 'desirous of obtaining the services of people clever at working mines, as Abyssinia produces gold, silver, tin, lead, and coal', but were unable to exploit them. 58 The British Government, unwilling to be further involved in Ethiopia, left the letter unanswered for over a year, and then ignored the monarch's request for craftsmen. Yohannes nevertheless succeeded privately in obtaining the services of a member of the expedition, Sergeant John Kirkham, who became the Emperor's military adviser, with the title of General. Recompensed with an estate at Ginda'e, he was entrusted with training young military trainees, but this was apparently not a success. William McEntyre Dye reports: 'King John authorized him to drill about one thousand soldiers. These soon showed great improvement... [but] were impatient under rigid discipline; and this the King did not permit [Kirkham] to enforce. After a few months the men complained of the restraint, and the system was abandoned.'59 Kirkham was also despatched on an important diplomatic mission to Europe. 60 Yohannes for time employed also a Frenchman, called Rene, who served in the administration ofKariin. 61 This period also coincided with a great expansion in the armament of the Emperor's forces. The British, on their departure in 1868, gave Y ohannes six mortars and six howitzers, both then scarcely known in Ethiopia, as well as 725 muskets with fixable bayonets. 62 Some of the British rifles were allocated, according to Girard, to an elite force of Taltal soldiers. 63 The victories over the Egyptians at Gundat in 1875 and Gura in 1876 subsequently led to the capture of 20,000 Remington breech-loading rifles, hitherto a monopoly of foreign soldiers. 64 58 59

Great Britain, P.R.O., F.O., 1/28, Stanton to F.O., 13 January 1871. Dye (1880: 472).

60

Simon (1885: 277-9); De Cosson (1877: I, 302, 107-180); Wylde (1901: 216-17). Girard (1873; 31-2, 35-8, 271). Holland and Hozier (1870: n, 94--6). Girard (1873: 104-5).

61 62 63 64

n, 63-4); De Coursac (1926:

Dye (1880: 141); De Cosson (1877: n, 165-6); Rohlfs (1883: 62-3); Wylde (1901: 24-8); Rubinson (1976: 323,335).

INNOVATION AND MISONEISM

287

The transformation resulting from this immense acquisition of firearms was noted by Rohlfs. Recalling that in Tewodros's time there were only a very few primitive rifles, he states that Ethiopian soldiers then relied mainly on spears, swords and shields. Such weapons could still be seen towards the end ofYohannes's reign, but were going out offashion. Shields omamented with gold and silver filigree were seldom any longer seen. Almost all Ethiopian soldiers by then carried breech-loading rifles, or models using percussion caps. 65 The value of such weapons was, however, limited, first by the Egyptian, and later by the Italian blockade on ammunition imports at the coast. 66

Hunting and Its Effect on Wild Life

The influx of fire-arms resulted in intensified hunting, particularly in the north of the country. This led to a steady decline in wild life. Evidence of this in the Karan area was provided by James. He observed that 'a few years ago elephants and rhinoceros were plentiful in places that the former now visit only at rare and uncertain intervals and where the presence of the latter is a thing entirely of the past'. 67 Most hunting was carried out by the people of the country, but the travels of the Earl of Mayo, author of Sport in Abyssinia, or the Mareb and Tackazee, of 1876, was a portent of foreign hunters to come. Support for the elimination of wild life was voiced by the aforementioned Mercha Warqe, who claimed that 'the people of Abyssinia were making strenuous efforts to clear the country of lions and other beasts'.68

Land Grants to the Church and Democratisation ofLand Tenure

The structure of land tenure and relations between lords and peasants, during the reign of Yohannes, underwent significant changes. These came in the wake of a long period of civil war, which had witnessed the rise of a class of soldiers-cum-bandits, who lived by looting the countryside. Instability may also have been engendered by Tewodros's 65 66 67 68

Rohlfs (1883: 136, 146-7,204).

Wylde (1901: 263). James (1890: 113-114). The Globe, 27 August 1884.

288

RICHARD PANKHURST

attempts to curtail church land. The period after the accession of Yohannes was, however, also potentially unstable, on account of three successive major invasions, by the Egyptians, Dervishes and Italians. The Emperor, a strong supporter of the Church, reversed his predecessor's anti-clerical policy, by reinforcing and apparently extending church control over the land. He confirmed the priests of Aksum in their former possessions, and allocated them further estates, notably in Tamben. He also issued a decree re-establishing the monastery of Dabra Bizlin in its old fiefdoms, and granting lands to several other religious establishments. 69 Y ohannes also presided over a democratisation of land holdings. On the eve of the fighting with the Dervishes in 1888, he and Ras Alula issued two far-reaching decrees. One, designed to raise money for the war, stated that anyone paying tribute on land was considered to hold it as resti, i.e. inheritable private property. This decision considerably extended the ranks of restenyatat, or rest-holders. The other edict granted 'squatters' rights', akin ~o those ofresti, to anyone who had held land for forty years or more. Women, in some areas, likewise received the right to inherit land from their fathers}O A more benevolent attitude towards the peasantry is likewise suggested by Wylde. He declares that by the end of Yohannes's reign 'the peasant and cultivator were ... better off and less molested by the soldiery, as only enough men were kept permanently under arms to enable the king to enforce his rule, and it was only in war time when expeditions had to be undertaken that the able-bodied peasantry were called out'}1

Status of the Nobility and Reduced Oppression ofthe Peasantry The hitherto exalted status of the nobility was challenged, apparently towards the end of the reign, in two separate incidents involving the collapse of restrictions on the peasantry. The first occurred in the Maqet area of Bagemder and is recalled by the Ethiopian scholar Alaqa Umma Haylu. Getahun Gesse, a prosperous peasant, wishing to hold a feast for the Virgin Mary, challenged the traditional law prohibiting the peasantry from slaughtering cattle or making tal. The local peasants were all invited. One of the guests, 69 70 71

Perini (1905: 85); Villari (1938: 1432-40); Huntingford (1965: 79). Nadel (1946: 7, 11, 17-18). Wylde (1901: 44).

INNOVATION AND MISONEISM

289

however, reported the matter to Yohannes, who summoned Getahun. fier the charge was levied, he replied: 'I farm extensively, so I have much grain. Since my forests are good for bees I have plenty of honey. My land is so large that I have many cattle. I do not sell all my crops as I have enough money. If I try to keep what 1 produce the grain will be consumed by the nakiiz [a grain-eating insect], the honey and butter will spoil, and the cattle will become too numerous for my land. I therefore decided to give a banquet for the Church, the priests, and the congregation. I have not killed frida [cattle] for myself, or prepared faj for my own use'. The Tegray nobles at court were strongly opposed to Getahun's plea. 'In the time of our fathers and grandfathers', they said, 'persons found with taj in their houses were deprived of their land; persons discovered sleeping on leather beds had their riches taken away from them. People followed their masters from qolla [lowland] to daga [highland] in order to win the privilege of the qamis [shirt], and the right to kill frida and make taj. Many people have become rich by farming and cattle-rearing, and if they were granted these privileges no one would attend court any more. The law clearly specifies that the defendant is guilty; and should be punished'. Yohannes, however, ruled against the nobles, declaring: 'The defendant was not found stealing money or behaving like a robber in the jungle. If he is not going to lack oxen with which to plough, it does not matter ifhe kills his ownfrida. Providing he does not become destitute, there is no objection to his drinking his own honey. Moreover, what he did was done in honour of St. Mary. He is, therefore, not liable to punishment, and henceforward people may live as they wish'. When the populace heard this epoch-making ruling they reportedly returned home rejoicing. The ballabats, or squires, and other rich people prepared large quantities of taj, killed as much cattle as they wished, and there was much rejoicing in the land' .72 The second. no less important, incident, involved Ras AlIa, the ruler of Hamasen, who issued a decree confirming the old law preventing the peasantry from preparing taj for their own use or for weddings or funerals. He justified this edict on the ground that taj was 'the drink of the King' and that it was 'not appropriate for everyone to drink it'. 72

Miingestu Uimma (1959: 50-56).

290

RICHARD PANKHURST

Shortly after this Hadgembesa, a man of good but not noble family, decided to make tiij for his marriage feast. His friends appealed to him not to break the law, but he refused. He was found gUilty by the Ras, and banished to the Habab county, after which he defected to the Italians who had by then established themselves in the territory soon to be named Eritrea.73 Opposition to Slavery and the Slave Trade

Yohannes was significantly influenced by the 19th. century British agitation against slavery and the slave trade, and was virtually the first Ethiopian ruler to take an official stand against them. Opposition to both institutions was apparently first brought to the Emperor's attention in 1873 by the British traveller De Cosson. Yohannes, according to the latter, 'listened attentively' to his interpreter's translation of the arguments against slavery, after which he declared that he had 'thought gravely over these matters and that it was true that slavery was distasteful to him as a Christian sovereign'. 'No European Power', he said, had, however, 'ever requested him to abolish the slave trade, and that while he was only Prince of Tegre it would have been out of his power to do so, as the measure would have been opposed by many of his chiefs, whom he could not have controlled'. 'Power', he explained, was, however, 'now in his hands', and he trusted that God would favour his efforts 'to deliver his country from the attacks of the Mussulmans', as well as to gain the friendship of Britain. De Cosson proceeded to argue that abolition would give great pleasure to the British, and continued: '1 had his promise that he would take immediate measures to enforce the punishment of death against all traders, Mahometan or Christian, who should in future attempt to buy, sell, or kidnap slaves in his country, or attempt to pass them through it. I begged his Majesty to complete the work by declaring free the slaves then existing in Abyssinia, as those who were contented with their lot would be no worse, while those who pined for liberty would be able to enjoy it like their fellow creatures. This also he said he would do, and, at my request, promised to pledge himself to do these things in letters written under his great seal' .74 73 74

Kolmodin (1915: 192-3, 195). De Cosson (1877, IT, 35-9). See also De Coursac (1926: 215-19).

INNOVATION AND MISONEISM

291

De Cosson was so confident in the Emperor's support that he reported to the British Foreign Secretary, Earl Granville, on 30 July, that Yohannes had agreed, 'not only to execute sentence of death on slave traders', but 'also to declare all slaves now existing free' .75 The Emperor duly instructed Kirkham to write to Granville under the imperial seal. The letter declared that the Emperor, as a Christian sovereign, found slavery 'distasteful', and promised 'to put an end to all traffic in slaves and to declare all slaves free men'. Yohannes also announced his intention of proclaiming that 'if any of his subjects shall buy or sell a slave, they will be punished by death without mercy'. Though the efficacy of these decrees, like those subsequently issued by Menilek, was limited, Yohannes had thus committed himself to the antislavery cause. Rohlfs confirmed that the Emperor had declared the trade illegal, but feared that the monarch closed his eyes to the capture of slaves who were not Christians. On the positive side, however, the great slave market at Galabat on the Sudan frontier, was, according to the British traveller William Winstanley, officially closed.76 Yohannes shortly afterwards committed himself by international treaty to the abolition of both slavery and the slave trade. In the Tripartite Treaty of 3 June 1884 he undertook 'to prohibit and to prevent, to the best of his ability, the buying and selling of slaves', as well as 'the import or export of slaves to or from his dominions'. He agreed also 'to protect, to the utmost of his power, all liberated slaves, and to punish severely any attempts to molest them or to reduce them again to slavery'. Wylde comments that the Emperor 'faithfully carried out this treaty' and that there was 'no known case of slaves passing through his dominions from that time until his death' in 1889. 77 The underground slave trade nevertheless continued. The tiny port of Embarami near Massawa, according to Wylde, was 'noted for the shipment of slaves', and 'many a kidnapped Abyssinian child' had there 'seen the last of Africa'. Many Abyssinian women were likewise lured to Galabat by promises that they could 'make plenty of money by immoral practices. Once taken there, they were regularly sold to the slave dealers, and ... never saw their native country again'. Women were similarly enticed by promises of work to go to Massawa, where they were exported

75 76 77

Great Britain, P.R.O., F.O. 1I27B, De Cosson to Foreign Office, 30 July 1873. Rohlfs (1883: 267-9); Winstanley (1888: I, 50, Il, 77-8). Wylde (1901: 474-5).

292

RICHARD PANKHURST

as slaves to Arabia.78 The last years of the Emperor's reign also witnessed considerable slave-raiding by the Dervishes. 79

Imposition ofReligious Orthodoxy Several of Emperor Yohannes' s most important innovations were in the religious field. A loyal child of the Church, he believed that the political unification of the Ethiopian empire should be accompanied, as Dajazmach Zewde Gabre Sellassie declares, 'by measures designed to establish religious uniformity' .80 Following his victories over the Egyptians at Gundat and Gura in 1875-6, Yohannes insisted that the principal Muslim leaders of Wallo, MuQ.ammad