AAL 5/2. — Appleyard Monographie Journals of the Near East General Editor: Giorgio Buccellati Afroastatic Gnguistics Editor: Robert Hetzron, Santa Barbara Associate Editor: Russell G. Schuh, Los Angeles Advisory Board: Atiel Bloch, Berkeley John B, Callender, Los Angeles ‘Talmy Givon, Los Angeles ‘Thomas G. Penchoen, Los Angeles Stanislav Segert, Los Angeles Volume 5 Issue 2 December 197 A Comparative Approach to the Amharic Lexicon by D. L. Appleyard S Undena Publications Malibu 1977AFROASIATIC LINGUISTICS AAL includes contributions in linguistics within the vast domain of Afroasiatic (Hamito-Semitic) languages. Articles of general, theoretical interest using Afroasiatic material, descriptive, historical and comparative studies are included. Editor: Robert Hetzron (1346 San Rafael, Santa Barbara, Ca. 93109, U.S.A.) Associate Editor: Russell G. Schuh (15337 Hart St., Van Nuys, Ca. 91406, U.S.A.) Advisory Board: A. Bloch, J. B. Callender, T. Givén, T. G. Penchoen, S. Segert. MONOGRAPHIC JOURNALS OF THE NEAR EAST MINE is a system of journals on the Near East, with each joumal devoted to a specialized study area, and each issue consisting normally of a single article. Current journals in the system are Afroasatic Linguistics, Assur, Computer Aided Research in Ancient Near Eastern Studies and Syro-Mesopotamian Studies. General Subscription. — For a prepayment of $15.00 the subscriber selects random issues from within the entire system as desired, up to a total of 200 pages. ‘The subscriber is also entitled to (1) periodical lists of abstracts from all journals in the system, and (2) reservation to any journal within the system, whereby issues of a given journal are sent on approval immediately upon publication (and may be returned within two weeks). Library Subscription. — A prepayment of $15.00 for each journal in the system secures all issues of a single volume as soon as they are published. This subscription schedule does not allow the selection of random issues. Library subscriptions are available to both institutions and individual scholars. Individual issues are numbered sequentially within each volume. Each issue has its own pagination. A volume is closed when a total of about 200 pages is reached. AA title page and a table of contents listing all issues within each volume are sent to all subscribers at the close of a volume. Periodicity in the order of appearance of issues is not predetermined. A volume, however, is generally completed within one year. Institutional and Professional discount of 20% on single subscriptions entered within six months of the beginning of any given volume (higher on larger orders). Payment must accompany orders from individuals. ‘A handling fee of 80/ will be charged to Libraries if order is not prepaid. Order from: UNDENA PUBLICATIONS, P.O. Box 97, Malibu, California 90265, U.S.A. ©1977 by Undena Publications. Al rights reserved. No part of this publication may be reproduced or transmitted in any form or by any ‘means, eleetronie or mechanical, including photo-copy, recording, or any information storage and retrieval system, without permission in writing from the publisher.Monognaphic Journals of the Near East Ajnoasiatic Linguistics §/2 (December 1977) A COMPARATIVE APPROACH TO THE AMHARIC LEXICON* by D.L, Appleyard School of Oriental and African Studies, University of London A glance at the morphology of Anharic does not reveal mich that is not plainly of Semitic origin, or that cannot be readily ex- plained in Semitic terms. The lexicon, however, has given some casual observers quite a different impression. "A close exanin- ation of the lexicon, as here of certain semmntic fields chosen in particular to give a fairly wide cover of vocabulary levels, "basic" and "non-basic," does, however, reveal a fairly high proportion of inherited Semitic material. The borrowed element of the vocabulary is shown to derive principally fron two gen- eral sources: i) other Semitic languages, especially Arabic; ii) Cushitic Languages, especially Agaw and East Cushitic (Calla, Sidamo, Saho-Afar, Somali, etc.). ‘CONTENTS ee INO 2 1, THE SEMANTIC FIELD "MIN" ee pee cde 6 Ll. Goneral tems 2. 22222222 2D PIT ID III Dl iil Dlo7 nite Kia cerms] gor ey ee re eee 8 1.3. Parts of the body) 1.21 1 Be u 2. THE SIMANTIC FIELD "THE DOMESTIC ENVIROWENT” 2. ee 2.1, Agricultural activities and implements... 2 21 Crees aris ess ees a Fee Gr agen h Geeta Seep ogeonesaee sons 24 2:4. Food and its preparation | 2222222222222 DI II DDIIII: 26 215. The house we eee Bee er ee ey sere oD 3. THE SEMANTIC FIELD "THE NATURAL ENVIROWENT” © 0 eee 32 3.1, Natural phenomena... eee ee 3.2. Flora... ee ae eee 39 B3l Fama 222 it 5 ee ee 4, THE SEMANTIC FIELD "SOCIAL ORGANIZATION" eee eee 4.1. Law end goverment... 2. cll ellie lllllnllile 48 M2, MOMMY ce eee ree eet 82 “the text of this paper forns part of a thesis titled "the Semitic Basis of the Anharic AALS, 432 D. L. Appleyard [AL 5/2 ‘ s 4.3. Warfare... 6. fee eee er ae irk (atti) Gaecacoouoodceoqoocedusceusuneans 56 GRAMMATICAL ITEMS: PRONOUNS, NUMERALS AND PARTICLES © 6... eee 5.1. Inseparable particles |... see ee eee sce eee 5.2. Separable particles . | |) | 556650 0000Gc000 B5600065 REFINES 6... ee ee Be ee eae Poe oa 0. INTRODUCTION The Ethiopian language area is particularly interesting within the Afroasiatic field, con- prising as it does, anongst others, representatives of three of the language families normilly included under the’tem Afroasiatié: Semitic, Cushitic, and Onotic. Moreover, what is espe- cially interesting to the historical linguist, and what will in part concern us in this paper, is the long period of contact and interaction between at least to of these language groups, (Ethiopian) Semitic and Cushitic. The length of this period of contact, based upon ‘the assumption that Semitic speech was introduced into Ethiopia from the other Side of the Red Sea at a broadly identifiable date, must ostensibly be around two millennia and is al- most certainly somewhat longer. It should, of course, be pointed out here that Ethiopian Senitic and Cushitic, or certain subgroups’ thereof, are not likely to have been the only Tanguage groups to have influenced one another within the Ethiopian language area. In Speaking about Ethiopian Semitic-Cushitic contacts I am being deliberately vague at this stage of the study; Cushitic is not, of course, a single unit, nor even a completely and satisfactorily defined one as yet. liowever, the results of this study of the lexicon reveal that in examining the history of the Anharic vocabulary we have to deal principally with two branches only of Cushitic: Central Cushitic, or Agaw, and to a lesser degree East Cushitic, both Highland (Burji-Sidano) and Lowland (Saho-Afar, Oromo, Sonali, etc.). Moreover, as the title of this paper indicates, we are concerned only with the Amharic vocabulaty, not with other Ethiopian Semitic languages, whose lexical histories mst neces- sarily be different from that of Anharic. Lexicon" accepted for the degree of Ph.D. by the University of London in Septenber 1975. ‘The languages referred to in the text are abbreviated as follows: Ak. Akkadian Caf. Gafat Sem. Semitic Anh. Amharic Gk.” Greek Sid. Sidamo Ar. Arabic Go’ Goggot si. Aram. Aranaic Gy. Gyeto Sod. Bil. Bilin Gz. Ge'ez Som. Ch.” Chaha,Caiha Har. Harari Soa. Cush. Cushitic Heb: Hebrew Syr: Ewa Kem. Kenant Te.” Tigre, Tegre End. andiigaii Khm. Khamir,Xamir Tha, Tigrinya, Tegrefifia Enn, nnamor M.— Muhor thuxor Ug. Ugaritic ESA’ Epigraphic South Arabian MS. Miskan WI. Walzine Eth. Ethiopian (Eth. Sen. S.Ag. Southern Agaw (Awmgi) 2.” Zway S. Eth., N. Eth.) AALS, 44197) A Comparative Approach to Amharic Lexicon 3 Im this study I shall deal with certain selected areas of the lexicon in accordance with the concept of semantic field. In each section a separate semantic field will be discussed first as a whole, from the point of view of the respective statistics of inherited Senitic and loan items and any extra-linguistic inferences that may be drawn therefrom, and secondly with regard to the individual items representing cach field, which will be examined in sone detail, providing an approach, at least, to an etymological treatment of parts of the Amharié lexicon. The notion of semantic field, Like mich of semantics, lacks the precision of definition of the levels of functional Linguistics. A senantic field nay be broadly defined as a con- Ceptual sphere. The individual lexical items which can be grouped under a given senantic field will, however, differ from language area to language area. To this extent a semantic field is not as readily or as clearly definable as the functional units of language. Never- theless, the concept is a useful and valid one in a study of this kind. Certain areas of the lexicon are bound to be more vulnerable to borrowing and innovation’ than others, as they reflect the cultural sensitivity of their referents. It is precisely for this’ reason that the division of the lexicon into semantic fields and the comparison of the retention Tate statistics of each field have such value here. A comparative analysis of phonology and morphology may reveal a genetic position of the language and its position vis-a-vis cognate Languages, but a comparative lexical study can provide not only finer details of linguistic connections and contacts, but also indications of extra-linguistic factors. It is, however, advisable to add a word of caution here. One should hesitate to draw sweeping historical or cultural conclusions from the history of an isolated iten, It would, for example, be Linguistically naive to conclude anything regarding the domestication and keeping of goats amongst the Ahara from the fact that the Anharic term for this animal, gayi, is of non- Senitic origin. It is rather the overall picture in each semantic field that’ should be exanined in this way. ‘The senantic fields for, this study were chosen in order, to provide a reasonably wide spectrum of culturally "specific" and “non-specific” vocabulary.? Five fairly wide fields were selected: (1) "hn," (2) "the domestic environment," (3) "the natural enviroment,” (4) "social organization,"® and (5) "grammatical itens,” a "field" closely involved with orph- ology, encompassing pronouns, numerals, and particles, This last "field" has been included to provide a lexical link with morphological analyses such as Hetzron's (1972). Each of these broad fields is then further subdivided into more specific sections; for example, the semantic field "nan" has three subdivisions, (1.1) general, (1.2) kinship terms, and (1.3) parts of the body. The first four broad seftantic fields range from what could be expected to be a conservative area of the lexicon ("nan") to the areas of social organization and donestic environment, more sensitive to borrowing and innovation. The field of natural en- Vironnent has been included because it is in part delimited and defined by the particular geographical and ecological context in which the language is spoken. The latter is a par- Licularly important consideration wien one approaches the question of Semitic origins within Ethiopia. The more “conservative” fields may be assumed to give an indication of the maximal retention of inherited Semitic lexical stock, whilst the more innovatory fields Will be especially valuable in providing clues to cultural influences and pressures exerted from outside the language comity. Thus, the inherited Semitic and loanword composition of these areas of the lexicon can be an important guide to extra-Linguistic developments. in the Anharic language area. For example, it cones as no surprise that much of the specific flora and fama vocabulary of Anharic is of non-Senitic, Cushitic origin (see below, under 2.2). On the other hand, the influence of Arabic is strongly felt in the fields of comerce and warfare (see below, inder 4). 2For a definition and discussion of these terms sce Hlymes (1960:4-5). °the inspiration for this kind of treatment of the lexicon is derived chiefly from Fronzaroli et sec. AAL 5, 454 D. L. Appleyard [AAL 5/2 An important theoretical concept throughout this discussion, which ought to be briefly ex- plained here, is the notion of "basic" vocabulary. ‘The concept of vocabulary as polarized into two groups, "cultural" and "non-cultural," or "basic" and "non-basic," is by no means anew idea. At’one end of the scale are itens whose referents are considered to be nearly universal and most resistant to innovation and replacement. At the other end are those whose referents are specific to the cultural environment and which are, therefore, suscep tible to innovation and replacement in accordance with developments in’ the society in which the language is spoken. The recognition of different levels of the lexicon subject to different rates of change has led to the composition of "basic" word-lists. Whatever the merit or otherwise of the glottochronological application of this notion of “basic” vocabu- lary is, the recognition of these two broadly identifiable poles of the lexicon cannot, I believe, be seriously objected to. ‘The individual lexical constituents of "basic" vocabu- lary necessarily differ from one imguage area to another. Rather than define individual lexical itens as "basic," it would seen more advisable to deal only with semantic fields at this level. ‘These are less specific than individual items and are nore applicable to the notion of iniversality, upon which the concept of "basic" vocabulary is partly built, and it is this contention that lies behind the choice of semantic fields used in this discussion. ‘Thus, one might reasonably expect to include kinship terms, nanes of parts of the body and certain natural phenomena, though not necessarily individual items from these fields, amongst "basic" vocabulary. Specific plant or animal nanes, many social terms, and 30 on are, on the other hand, "non-basic."” Throughout the discussions that follow I'shall use this concept of "basic" vocabulary in dealing with inherited Senitic and loanword proportions from one semantic field to another. I have already suggested above that the distribution of specific lexical items, whether of inherited Semitic origin or not, can provide valuable insight into areas outside the inmedi- ate scope of the description of the lexicon itself. Thus, we may reasonably look to lexical evidence to relate to the classification and interaction of the various Ethiopian Semitic languages anongst themselves. hinilst lexical isoglosses might not necessarily be indicative of genetic groupings, they may often provide tangible clues to former geographical contigui- ty, at least. Lexical criteria are, of course, not the first or the principal means that should be used in classification work, precisely because of the vulnerability of the lexicon ‘to change and outside influence. Nevertheless, a few important lexical isoglosses can be found which broadly seen to coincide with those established on morphological evidence, i.e. those drawn up by Hetzron (1972). Perhaps the most interesting of these concern the North South Ethiopian split. Most prominent amongst these are those instances where S. Ethiopian 95 a whole hs preserved a Senitic root absent, fron.N. Ethiopian: exx. Aue (nh, wizea) sweat, "e¢t_ (Rah, agin) "dust", *4bb (Anh. sbab) "snake", See (Anh. tal) "worn (Anh, née) ‘when’, *4-t (inh. {it} "face". “There are alSo instances where S. Ethiopian fs A hole Sharos a cusGerege fase pattaming or danelopomt, frat M. Eehlopian: ex banat as masculine ‘father-in-lay" with @ re-formation *hanaté as its feminine counterpar Eth. hamlu), hamat); *mos ‘husband’ and *masit 'wife' as against N. Eth. 692s and opouits Se*otali- "kidney" as against N. Eth. k*atit (Gz.Tna.), kolbat"at (Te.) tkale + @ ‘twenty’ as against N. Eth, *a8aa, Sosa, formed on the comon Toot ke* "two's S. Eth, Adabr ‘Seat! but N. Eth. dab ‘mountain's S."Eth. ‘bind® "right" from the Eth.Sem. root fn ‘be straight!» but N. Eth: yinan (Gz.tha.), man (Te.); S. Eth. *4sy ‘grind’ but'N. Eth. ggh. In addition to these’ Sonitic roots and patterns conion to S. Ethiopian only, there is a snail nunber of non- Semitic lon itens which occur throughout S; Ethiopian but not in N,,Ethfopian: | *52/Gbe + i (Anh. zégin) ‘nine’, *4/find/aka (Anh. Pértika) *noon',® g*azbit inee",® kur-a ‘crow, “The imherited Semitic item is preserved throughout S. Ethiopian in the restricted sense of ‘month’: Anh. wit, ‘this item does, in fact, occur in Tigrinya, but with the sense of ‘strength’, which is within the semantic range of many of the S. Ethiopian terms, including Anh. gufbét; the Tigrinya item is probably, therefore, an anharicism. MAL 5, 461977) A Comparative Approach to Amharic Lexicon 5 Sanndibait (Anh. andéibiét) ‘tongue’, *4/t0ag- 'calf". In each of these cases, with the probable exception of ‘crow’,® N’ Ethiopian preserves the inherited Semitic root. Of course, none of these in itself is necessarily proof of a genetic division between North and South Ethiopian, but they are interesting and, indeed, relevant in the light of the morphological isoglosses distinguishing the three northern languages from the rest of Ethiopian Senitic. The geographical contiguity, at the present or in former times, of the anguages concerned could explain several of thise examples, particularly the conn non- mnitic itens. Within S. Ethiopian there are several lexical isoglosses connecting Atharic with other nenbers of Transversal South Ethiopic (see Hetzron (1972:36)), i.e. Amharic, Argobba, Warari, and East Gurage. Often Gafat, and sometimes other Gutage languages, especially Soddo, “are included in these isoglosses. Gafat, etc., does not belong to the sane sub- group of S. Ethiopian as Anharic, but has for a’long period been contiguous with Auharic and under its influence. Examples of these locally restricted itens, including non- Senitic tems, are akost ‘aunt! (Anh.Arg.Har.Gaf,), for ‘thigh’ (Anh-Arg-Har.SI.WL.Z.Gaf. Sod.), 02 "beard! (Anh.Arg.Har.Gaf.W1.Ch.), ge ‘country place tom'” (Anh, Arg.tarS1,W1. Z.Gaf‘Sod.), zag ‘tree! (Anh.Arg.Har.Gaf.), 'waSia ‘dog’ (Anh. Arg.Gaf.Sod.}, ofgutit ‘viper’ (Ani.tiar.S1-W1.2.). Finally, in this connection, it should be mentioned that Anharic, occasionally together with other contiguous S. Ethiopian languages, often shares a lexical isogloss with Tigrinya, thereby cutting across alleged genetic boundaries. ‘This is particularly prevalent in the semantic field of social organization, a likely explanation for which is not hard to find. Speakers of Atharic and Tigrinya ate the direct inheritors of the Go'e2-Axmite cultural tradition and consequently have long shared in their cultural development. This is to sone extent refelcted in specifically "cultural" vocabulary. Aside from terms in herited directly from Ge'ez, Atharic and Tigrinya share the following social terms: atika : hatixa "head, superior’, daiiia, "judgo', geta + gtiyta ‘lord’, goadbet : gréndbet 'neigh- bour’, étiina "town', 34(sétt "revolt", dala ; doxa ‘poor’, wittaddin + wittohaddin ‘soldier’, dabtia in the sense’of "cantor, lay priest'.’ Some of these might simply be loans from Anharic into Tigrinya, as, for éxample, the form of dailia suggests. Others could equally well be conmon developments. Examples from other senantic fields where Anharic and Tigrinya share a common form, loan or developnent ofan inherited root, are: set : sdbéyti ‘woman’, ‘amiat + ono ‘nother', dinkoro : dink*éxo "deaf", gota : goraw "left (handed)', mites : rathas ongye’, onkubae « Sonkutatily ‘egg! besidé nokta, ete. (Tha. also has: “ark ax"oho) Soha: Zoxd "shout" dingsya + dang*étea "stone, rock’, gan ? g2°0 ~ gon ‘but'. Some o ‘these otcur throughout S. Ethiopian or in sone other $. Ethiopian languages besides Atharic, but are all found only in Tigrinya anong N. Ethiopian. ‘The receptiveness of the lexicon to change and outside influence, nore perhaps than any other level of linguistic analysis, means that this kind of lexical isogloss cannot be employed alone in language classification. ‘The importance of these isoglosses lies in ‘outlining geographical and/or cultural language areas, as denonstrated, for example, by the Tigrinya-Anharic or Atharic-Gurage isoglosses. Only in the case of those Semitic root isoglosses distinguishing $. Ethiopian fron N. Ethiopian does it appear that lexical evi- dence can be directly correlated with morphologically established groups. The few exanples relevant to the case of Transversal South Ethiopic are obscured by contacts with other S. Ethiopian languages like Gafat and Soddo. Lexical evidence for subgroups not involving Auharic has not been studied here. Gz. bYa*, Tha. kYax, Te. kawa* have close formal cognates in some of the Agaw languages. "cz. déibtina means 'tabernacle’ and is thus closer to the sense of the original Gk. déphthérai. MAL 5, 476 D. Le Appleyard IAAL 5/2 Each of the five semantic field sections will begin with a discussion on the overall field arising from the details of individual lexical itens that follow. This discussion will take the form of the statistics of inherited as against borrowed itens, any structural patterns that can be identified, any extra-Linguistic observations that can be made, and so on. This will be followed in each section by a more detailed etymological treatment of the individual lexical itens involved. Only in a closed set like kinship terms, numerals, or parts of the body can the list of items be anything like complete. Elsewhere the list of items is intended to be no more than representative of the semantic field. For this purpose, therefore, only the most “obvious” items were chosen. Those items with some particularly relevant or interesting contribution to the history of the language are discussed in full. Other more straightforvard itens need only be labelled as inherited Semitic, Cushitic, or whatever. 1, THE SEMANTIC FIELD “MAN” ‘The three subdivisions grouped under this heading are 1.1 general tems, 1,2 kinship terms, and 1.3 parts of the body. In the case of kinship terms, we are dealing with a relatively small’ set of items which is closely structured and in which all members are interrelated in a system, such that the loss or replacement of one item in that systen may affect the whole. For example, the terns aggot ‘uncle’ and akost ‘aunt’, both of non-Senitic origin, do not replace Single lexemes, but phrases in common Ethiopian Semitic, as Gz. of" *am ‘mother's brother', “shed “ub ‘father's sister', etc. It is, incidentally, interesting to note here that a sinilar restructuring also occirs amongst the nunerals, where Ath. <.'thousand" probably of Agaw origin, replaces the phrase ‘ten hundred’, as Gz, ‘aSdttu mo%. An interesting morphological feature of kinship tems in Amharic, as a system, is the suffix -at, common to several itens: abbat ‘father’, omat 'nother', ayat ‘grandparent’; in anat ~ anat and merat, however, the suffix -at is of’a different origin, being originally a feminine formative. This suffix -at is most likely related to the external plural formative -at, occurring in Ge'ez, Tigrinya, and Tigre, and in Amharic as -a88- in certain plural pro- nouns. Possible support for this’ view that “at in certain kinship terms is identical with the plural formative gccurs in other S. Ethiopian items like Sl.Wl. abot ‘father’, Ch. adot ‘nother’, Har. ‘indo’ ‘woman', S1.WL. oda’ 'wonan', all of which contain a suffix’ clearly connected with the plural formative #-at(i) ~ -ot(<), though used on singular nouns. These suffixes have lost their plural connotation in these items and the ordinary plural formative is added as on any other noun: Anh. abbatot, nnatof, ayatoé, etc. It can only be conjec- tured why an (originally) plural formative becane attached to’ these items, but, the most Likely explanation would seem tobe that the plural was employed as a kind of honorific, as is still the case in the "polite" forms in Anharic. Furthermore, the simple items “ab and 2mm acquired specifically theological connotations and, therefore, formally differentiated items might have been felt necessary in the purely kinship sense. OF the 17 items discussed under the heading "kinship," all but three are of Senitic origin. A possible explanation why two of these non-Senitic terms, aggot and akast, might have been taken over has already been suggested. In the instance of the third non-Semitic item, onnat ‘nother’, it is interesting to note that Anharic is not the only Ethiopian Semitic language that has’a non-Senitic term for 'nother'.® *For a detailed discussion on this and all other items in the preliminary sections to each semantic field sce the individual analyses that follow. AAL 5, 481977) A Comparative Approach to Amharic Lexicon 7 Amongst terms for parts of the body there is a somewhat higher percentage of non-Senitic itens--26.68 (16 itens out of a total of 60 studied). Anongst the Semitic itens are pre- served most of the comon Senitic terms (see Fronzaroli (1964: 18fF and 262ff)): ag ‘mouth’, aganga 'nose', afant ‘bone’, ayn 'eye', of "hand', dim 'blood', gét ‘face’, quadro throat’, hod “belty', kueatit 'kidney’, Lobb 'heart', mitas ‘tongue’, ads *head', Sabat 'groy hair’, ik "foetus, Sant ‘urine’, t14an ‘nail’, tat "finger", and’ so on. Of those items of non-Semitic origin onbant 'navel' and figut ‘hair’ are particularly wide- spread throughout Ethiopian Senitic, both North and South. Conmon to S. Ethiopian only are andabét 'tongue', Sanna 'shoe, sole’ of the foot’, guebét knec',® guig ‘cheek", riz ‘beard’. ‘The only item of denonstrably Cushitic origin, besides onbant and tigur, which has a cog- nate in Ge'ez as well as in modern Ethiopian Semitic is samba "lung". 1 do not think it would be very valid to try and draw any specific extra-linguistic conclu- sions from the relatively high percentage of non-Semitic itens in the field of parts of the body. If the concept and implications of "basic" vocabulary are accepted, then the occurrence of so may original loans in this particular field can be taken to’reflect the Jong and intinate symbiosis between Anharic (and Ethiopian Semitic) and Cushitic languages in the Ethiopian language area. The non-Semitic terms here are for the most part attributa- ble to specific Cushitic Languages or language groups, but a small nunber cannot be readily so defined. Such items, several of which occur throughout the lexicon, are clearly Cushitic in so far as, whilst lacking satisfactory Semitic cognates outside the Ethiopian area, they have acceptable cognates in more than one Cushitic Language group. Their shape, however, is not immediately relatable to one Cushitic group more than others, or is sufficiently divergent from any of the modern Cushitic forms as to prohibit identification with any particular one. 1.1, GENERAL TERMS anoge ‘old man" 1 Gz, Aunigawi ~ sandgay, Tha. sarigit,'° Har. rdga, Old. Anh. anige (Cohen M. (1939:44); Ludolf (1968:57) also has andje) Several Ethiopian Semitic, languages, including Amharic, also have @ corresponding verb: G2. Tha. ‘andgd, Anh.Arg. andg§d, S1.NI. xdje. ‘The S. Ethiopian verbal forms are probably de- noninatives, which would explain the palatalization g > § from the y of the noun suffix taken as final radical. The underlying form of the Toot is "tg, as appears in the N. Ethio- pian verb forms. Brockelmann (1928:49) connected this xg with’Sem. "rk ‘be long’, but this would seem doubtful, On the other hand, Cerulli (1936:28) suggests a derivation fron Cushitic, cf. Som. raag ‘renain, be late", Saho-Afar xa 'endure". It is not, of course, inconceivable that we are dealing with the confusion or the influence of the Sénitic and the (East) Cushitic root here in Ethiopian Semitic. onnast ‘female’ 2 Gz. anast, Tha, anvstiiyti, Te. 25st, Arg. onodia, Har. anosti, Gaf. ansoti, Ch.Gy. arast, $1.2. onnoet, Enn. ast, End. ast, etc. Common Semitic uf (Ar. ‘unta, Heb. *i8a, etc.). ‘Ta. qutbit is probably due to Amharic influence; the Tna. term for ‘knee’, borki, is inherited Semitic. "Feminine in form, but both masculine and feminine in meaning. AALS, a98 D. L. Appleyard [AAL 5/2 set ‘woman" : Tha, sdbéyti. A feminine derivative of sib? following item) man’ (see the Amharic and Tigrinya appear to be the only Ethiopian Semitic languages which employ a feminine form of sab° for ‘wonan'. Most of the other languages use a derivative of the Toot “iis-4 (see onnest, above), or 6°s-t (see mist, below). sao "nan" 1 Gz, s@b*, Tha, sdbeay (pl. sdb), Te. sib 'people', Arg. su, Har. usu>, Gaf. stimu, Ch. sab,’etc. ‘&m interesting speculation in whether Eth. si? might be connected with the South Arabian common noun s6° 'warrior' and the ethnic name sb? 'Saba>". wind 'male" + Gz. wiitd 'son', Tha, uiiddé, Te. uid, Arg. wind, Har, wiitdé. Argobba and Anharic share this unique instance of, the apparent development £d > nd. However, in Line with the not infrequent occurrence of an "intrusive" nasal throughout S. Ethiopian,” ~ it is probable that the innediate precursor of wind could have been something akin to Tna. widdé and Te. wid(d), ice. *wiidd, deriving from wild by total assimilation of the £ to the d. ‘The root is common Semitic wed "beget', which provides simple nominal derivations of ‘the meaning ‘son, boy, child", etc., throughout Semitic. 1.2. KINSHIP TERNS abbat 'father' Gz.Te., tab, Tna., ‘ab ~ ‘abbo, Arg. ait, Har. Ge, Gaf. ab™ Gh. aba, etc. ‘This is immediately identifiable with the common Semitic %, but it should be noted that a similar root occurs in various Cushitic languages: Bil. abba, Kem. aba, Galla abba, Som. abe, etc. aiSa 'betroth’ + Gz. hazdyé 'share, give a bride gift’, Tna. hagiyi, Te. hasga, Caf. ald ‘marry’. Senitic, cf. Ar. fasga ‘fall as a share’, Heb. lsd 'share, divide’ agibba "marry Gaf. atgibbi, Ch. agéipan. Literally "bring in’; the root Eth. gb? ‘enter’ occurs throughout Ethiopian Semitic and in Ar. §abata, ESA gb? agget ‘uncle’ Tha. *akko, Har. kaka. This is of Cushitic origin; for the Anharic item cf. Bil. *ég, Kem. ag, Khn. ig. For the Tigrinya and Harari items a slightly better formal cognate occurs in Galla akko’ 'grandnother’, Sid. atte ‘old’. Anharic makes no distinction between maternal and paternal relation ships, as are mile elsewhere in Ethiopian Semitic: Te. hat : ab(u), Har. kaka : £2&t, Gaf. vatabbd, S.Arg. abo- : abost ‘aunt’ 1 S.lrg. akista, Har. oxista, Caf. akkost. Reinisch's derivation (1887(1) :19) from sonething along the lines of Khn. ig zin ‘uncle's sister’ is not totally convincing. It seems probable that this item should in sone way be related to a similar root as that underlying the tem for 'imcle'; in this connection note also Gimira akes ‘grandparent’. Again Amharic makes no distinction between maternal and paternal relationships, but cf. Tha. hatanno : *ammo, Har. oxista : anna, Gaf. abhost : AAL 5, 501977) A Comparative Approach to Amharic Lexicon 9 , and so on. Note that in the three EthSem. languages that do have cognates to Anh. akost, the temn is restricted to the sense of matemal aunt. nat ~anai* + G2.Te. amet ‘nother-in-law", Tha. amat_‘nother/daughter- in-law", Har. fanat ‘nother-in-lae", fand&i ' father-in-law" Arg. hamaé ‘father-in-law’, Ch. anaiti ""father-in-law', MsSod.W. amaé "father-in-law", 2. anakBi, etc. ‘The form amat occurs throughout the rest of Gurage with the meaning ‘mother-in-law’, except in End. which has awit. The S. Ethiopian forms for ‘father-in-law’ suggest a reconstructed common form *hamati beside ‘mother-in-law’ hamat. This differentiation is secondary and culiar to S. Ethiopian, whilst N. Ethiopian shows the comon Semitic pattern, as in Cr. ham : hanat. "Of the S, Ethiopian languages Cafat stands alone in preserving the simple form of the root: an‘. ayat ‘grandparent A possible derivation of this item, which appears to have no formal cognates elsewhere in Eth.Sem., is from Sabéyy Gz. ‘great’ + the suffix -at discussed above. A similar semantic development can be observed in Ina; “addi Sabbay ‘grandmother’, lit. ‘great mother", and perhaps in Te. *abevb ‘grandfather’, in which -©ob might derive from the root “by. Note also Khn. xéy abba ‘grandfather’ and xdy oa ‘grandmother’, where xdy is the adjective ‘great’. | Amongst the other Ethiopian Semitic languages a Composite fon is used: Gaf. yab"é ab", tna, sabe haggo. A sex distinction is made in sone languages: Arg. baba’: Amahat, Har. bab : wnna, Caf. yab*é ab%é : am™étita. In connection with this last item note also Gondare Anh. méta "grandparent’ and an(m)it "great-grandparent'. annat ‘nother’ + Tha, *omo, Har. dndy "Lady, matron’, Praetorius (1879:59,43) derived this from Gz. arm and hence common Semitic n. Whilst the dovelopnent im >n is not unknom in Ethiopian Semitic, the typical environment for such a change immediately before a dental is absont here. Rather, 1 believe this should be derived from a non-Senitic source, e.g. Saho-Afar ind ‘nother’, Kh, soi. Senitic. °w is preserved elsewhere in Eth.Sem.: G2, 2am, Te. an, Arg. om, Gaf. mie, and in Anh. smmibet ‘nis tress", uindom(nj ‘brother’, and’ the vocatives smmd,~ snama~' gmaye. Other modern Eth.Som. Imguages also have probable non-Semitic itens for ‘nother’: Tha. ‘addé, Ch. adot, Enn.tnd. adod, S1.W1. ndit, to which such forms as Borana Calla héada, Benadir Som. lindo,’ Katfa indo} etc, may be’ compared. at "sister! Gz. ot, Tha, hawtd,'? Te hot, Arg. ohud, Har. ait, Old Ani. het ‘The dictionaries (Guidi, 1901; Baetenan, 1929; Gankin, 1969) show a range of overlapping meanings: anat ‘nother/brother/son/daughter-in“law'; anaé 'father/brother/sister/daughter- in-law". See Tubiana (1951-4:51): "la confusion peut s'expliquer par occurrence de deux formes, Ltune gudze fhanag (devenue normalenent anat en anharique) "belle-nére (ilim. 77), l'autre anh."anaZ, connue d6j2 de Dill, come ‘child or parent-in-law! *2A secondary development from masc. haw. MAL 5, SI10 D. L. Appleyard [AAL 5/2 The altematiye modem Anharic form shat is, I believe, more likely to be due to the in- fluence of Ge'ez than a directly inherited form, as the perservation of the laryngal into modem Anharic would be highly irregular. ‘The Toot is common Semitic + ‘brother’, %j-t ‘sister! bat "husband" : Gz.Tna.Te. bi%ad ‘master, husband", Gaf. bat. Conon Senitic be. ‘The other Eth.Sem. languages, use the sane item as ‘man’: Tha, sab say, Gz. 60%, mat, Te. byes, or “lord” as Har. aboé. Sone of the ScEth. languages have a special term, Che'mas Arg. més, a masculine counterpart of Anh. mist (sce belos). 2o§ ‘child’ : Gz, Lod in the phrase Ladd bet ‘slave born into a household (Gk. odkogents)', Har. L432, SL.WL.Arg. £19, ch. axe, M. yog, Enn.End. a, E.Ms. drag. ‘This is a primary nominal derivative of the common Semitic root wld 'beget'; for this par- Uicular pattern with this sense note especially kK. &Zdu "bastard'. Sex distinction is made in Atharic by compounding with wind ‘malo’ or set 'fenale’, but several of the other Eth.Sem. languages have separate lexemes: Har. Liji ; kahat, Ch. axt : giniid, Tha. widdé : grat, Gc. wild : wilatt. A feminine derivative of £o§ does exist in Amharic and is re- Corded already by Ludolf: togit ~ Lagot, but it appears to be not so common in modern ‘Anharic as the compound set 203. moat 'sister/daughter-in-law’ : Gz. mix‘at ‘sister-in-law, Tha.Te. mix‘at bride". The root is x*w to which Ar. utGuz "yoke of oxen", Heb. agG 'friend', 2°@ ‘join, befriend” may be compared. mist ~ mabe ‘wife’ 2 Har, méSté, Arg.Z, most, M.Ch.Sod. mab, N.Eth, cognates have initial 6-: Gz. 60st 'wonan', Among the S.Eth. languages, Harari, Argoba, Caha, and most (al1?) Gurage Languages have Corresponding masculine forms: | Har. "fellow', Arg. mis ‘husband’, Ch. mos, to which Gz. 69°44 ‘man’ and Te. 69°04 ‘husband’ ay be compared. Cohen (1939:421) ‘suggests interference with an item such’as Gz. mat "husband" to account for the 6 : m alternation between North and South Ethiopian. However, sporadic instances of a b : m alternation do occur elsewehre in Ethiopian Semitic. ‘The palatalization » > 3 in sone of the S.Eth. forms could have been conditioned by the follow ing & : mest > mol(9)t; those forms without palatalization may have arisen from a netathesized “misot > mist, most. The Eth. root bs (Gz. by2sd, "be harsh, bad') is common Senitic b°}. For the semantic developnent from "be bad, strong’ to 'man' compare Heb. geber ‘nan"’and gabar "be strong" wwindom(m) "brother" : Te. wid 2am beside fu "brother". This is 4 compound of wind 'nale, son’ and -am(m) ‘mother’, a bound morph in Amharic. The inherited Semitic item for ‘brother’ occurs elsewhere in Eth.Sem., but is lost in Anharic: Gz. *0h", Tha. haw, Te. hu, Har. oh, Arg. ah. This might be becatse of the inherent phonetic weakness’in the resultant Atharic form which would have been,*2. ‘The feminine derivative of this root does, however, survive in Amharic as ot ‘sister’. MAL 5, 52.1977) A Comparative Approach to Amharic Lexccon ct waxsa "brother/sister-in-Law' ‘The derivation of this item from Eth. wts (common Sem. wit) ‘inherit’ would seem to reflect the custom of levirate. 1.3. PARTS OF THE BODY ag ‘mouth" + Gz.Tna.Te. *af, Har.Arg.W1.2. ag, Ch. ag", etc. Common Semitic *p-, cf. Ar. gam, §Z, Heb. pe, Akk. pil. It is worth noting here that an Afroasiatic cognate occurs in various Cushitic languages , whose form is closer to the Ethiopian Semitic than that is to the rest of Semitic, cf.: Bil. ib, Saho-Afar af, Som, 4{, Galla dfaan{2), etc. agonga ‘nose’ + Gz. *ang, Hamasen Tha. “angi, Te. anog, Har. a, Ga. ag*é, Gh. agura, ete. The Anharic form may be explained as a metathesis of Sang + -2§(8Ja. Common Semitic np (Ar.,2ang, ESA. °ng, Heb. °3p, ARK, appu). anmdiné "hurt Gz. hammé 'be sick’, Tha. hand Te. amma, Common Senitic fm "be hot, feverish’. amot "bile' + Gena, hanog Gaf, anoti, Z. amut, Ch. anota, etc. Common Semitic, cf. Ar. funa 'venom', Heb. hema, Akk. dmtu, andiibéit "tongue" : Har, awit, M,Sod. ateiimit, Ch. andbiit, S1.W1.2. aniimiit, Go. atemiit, E. anndbéit, End. andid, etc. Of East Cushitic origin, cf. Galla axaaba, Som. “drnab, Sid. arnabo, Had. aftaabo, Saho anraab,etc. The form underiying the Eth.Sem. items may be teconstructed as *atndbit, which, with its -2 suffix, resembles the most closely such forms as Kanbatta anxdbétit, Alaba wi ‘This suffixed -£ in the Eth.Sem. forms could, however, be a development purely within Eth.Sem, angie ‘brain’ 1 Tha, hang*ét, Te. fangit, Var. hangutta, Arg. ang*it. Of East Cushitic origin, probably, cf. Saho-Afar hangat ~ hangat, Som. hargutta; the Bilin item hang”él is probably, however, taken fron Eth. Sem. angit ‘neck’ 1 Te. *angit, Har. angit, Arg. anaid, Old Anh. hangit, and allo fmgt in Tna. tihangitd 'to tie round the neck and shoulders. ‘This root *hng-t has been connected with the coon Senitic nq ~ Sng: Gz. haniihi: 'strangle', ‘Aah. anndikaé, Ar. ‘ung ‘neck’, Heb. ‘andq ‘necklace’. This is not impossible, allowing for the unpredictable shift of to 9. angat ‘innards’ 1 Gaf. anziiti, Ch.M.Ms. anzit, Sod. an¥ét, Old Anh. hanzit. ML 5, 532 D. L. Appleyard IAAL 5/2 ‘This would appear to be a derivative of the Eth, root hms (G2. fiom 'uterus', Anh. ams) , perhaps fron sorething like “hang + dyt with voicing of § > z and palatalization of z > 2/} in Anharic and Soddo. The root ln’ is Semitic, cf. Heb. ltgmeS. belly’, AKK. embu. an "faeces" Tha, har, Te. faro®, Ch.M.Sod. ard. Semitic, cf. Ar. jur®, Hed.” haat, atont "bone" + Gz. Sagm, Tha. Sagmé, Te. Sagam, Har. af, Arg. hagom, Gaf. agm*i, Ch. atam, etc. ‘The Atharic is the only forn to have a suffix -t, which probably originates from the plural, cf. Gz. *a*zumt, i.e. collective used as a singular, cf. tat ‘finger’, below. Common Semitic “gm. ayn ‘eye! + GzeTe, Sayn, Tha. Sayné, Har-S1.W. Zn, Arg.Ch. en, Gal. ini, End. én, ete. Connon Senitic "yn. agar "leg, foot" 1 Gz. *agt, Tha. 229d, Te. Sagar, Arg. ingén, Har. igin ~ Angin, Caf. agthd, Ch. agar, ete. This root, “yt, is Semitic, cf. ESA. *gt, Datina and Palestinian Ar. *ia, though it doubtful, I feel, whether it is imediately connected with Sem. ag2 of the same meaning, as has been suggested before, a ‘hand" Gz. 2ad, Jna. *id, Te. Pade, Arg.SL.WL. ang, Har. 42, Ch. ag, 2. ongé. The palatalization in the S. Ethiopian forms derives from the proto-fom *ade, which occurs still in Tigre and in the Ge'ez form used with pronoun suffixes. Ethiopian *d is the sane aS conmon Senitic yd; traces of an original 2d besides yd occur elsewhere in Semitic. aibart ‘navel! + G2. honbont, Tha. frombarti, Te. hombor, Arg. bone, Har. famburté, 2” amb sxbet. ‘This item is almost certainly of Cushitic origin, and specifically East Cushitic, cf.: Son. fund, Galla handuta, Saho hiéndub ~ handub? several Agaw languages have what may be ‘a cognate form, Kem. gdnbiita, Quara gunbera, which has been taken into Gafat as gunbond and Gondare Amharic’ as ganbota. It has been suggested that the Ethiopian Semitic forms should be connected with other Semitic terms such as Ar. nabaa ‘excrescence', or Heb. 4abbin ‘navel’, but this seems unlikely on formal grounds alone. ams "vagina! hhum$ . Conon Semitic, cf. Heb. figmeS ‘belly’, Aran. akk, embu, nba ~ nb ‘tear’ + Gz. *anbyt, Tha. nebsat, Te. *onba*, Arg. ambi» Har. obi> Gaf. mb*d } ch. “omba,, S1.2."anb', ete. AAL 5, $41977] A Comparative Approach to Amharic Lexicon 1B ‘The root is nb*, which is common Semitic, cf. Ar. naba‘a 'gush forth’, etc. saooe ‘blind’ 1 Gz.Tna.Te. Sawa, Arg. omar, Gaf. ameand, Sod, in-diwor, SLM. en-ut, etc. Comnon Semitic su. Jama ‘shoe, sole of the foot" : M.Sod.WL. Jama. OF S.Agaw origin, Eamé; other Agaw forms are Bil. angi, Kem. Samba, Khm. caba, etc. Eon ‘thigh’ 2 Arg.Gaf. Son "thigh", WL. Sow "back’, and probably also Ch’ ein ~ kin ‘behind There have been several attempts at the etymology of this iten fron Praetorius (1879:95) , ‘who sought to derive it from Gz. Sant ‘rib’, to Wajnberg (1937b:20), who proposed Cz. lash ‘ap! as its origin, and Cohen (1969, no. 182), and Cohen (1931:109}, who suggested as connection with Gz.’k*aana® ‘elbow'.” Ali threé are unlikely on phonetic grounds alone. Rather, if the connection made with’ch. kin is correct, one wonders whether "kan > Eon should'be derived from the root kn* ‘be Straight’ (see below, under fait); tho senahtic scatter of the various itens listed above would not, I think, be out of keeping with such an etymology. dim "blood" + so throughout Ethiopian Semitic (Gafat regularly has déin’a). Common Semitic dn. dinkoro 'deaf" : na. diink*éixo, Ch. tink*arra; probably to be connected vith Ge. danbau, Har.St. cuba, Gat. danurS, WL. donb, End. dan. Cushitic, cf. Gatla donko, donkono "stupid", Alaba dOnka, Som. dégon 'fool', and the common Agaw tems for 'donkey", Bil. dox*ana, Kem. dug*dna, S.Ag. dux"ari, etc. (Cohen 1951-4:15-16). Git "face" + Har, §2t, Arg. fd, ChE.Gy. yift, SLWL. aft, 2. M.MS.Go.Sod. gt, etc Common Semitic, cf, Soq. so 'forehead', AKK, piitu, probably also Heb. pe ‘corner, side (esp. of the head)’. gubbét 'Liver" 2 M.Sod.WL. gabbYot, Old Anh. gubbiid. Praetorius (1879:85) related this iten to Sem, kbd, which does of course, occur throughout Ethiopian Semitic including Anharic (kibd > hod "belly'). It is not inconceivable that gubbiéit should have derived from sone form of the root kbd by a variety of faulty restitu- tion and irregular developments, but it is, I feel, unlikely, though no alternative etymology would seem to offer itself. AALS, 58rT D. L. Appleyard IAAL 5/2 guebiit "knee" + Har, ga€éb, Arg.Gaf.Sod. guebit, Ch. g*onbit, Gy. g*amdt, End. gunad, etc. All the S.Eth. forms except Har. gatib can be derived fron *g*atbit; the item does not occur in N. Ethiopian except in'Ina. guebét, which has only the restricted sense of ‘strength’ and is almost certainly an Anharicisn, The item is of Cushitic origin, though it is not possible, I believe, to define one particular branch of Cushitic as the most likely sour cf. Bil,Kin,Ken. gob, S.Ag. gixib, Saho-Afar gulub, Son. §lib, Calla Sétba, Darasa gutbo, Kanbatta quéubéta, etc. “Attempts to connect the Eth.Sen. item to Semitic roots such as qb and 3b (Praetorius 1879:67 and 72) mst be discounted. ung. cheek" + Tha, gYonfi, Har. gunk, Arg. gumbof, Ch.E. gténBd, Sl. ging, etc. Najnberg_(1937b:19-23) ingeniously suggested this could be an old ‘dual’ of the sane root as seen in Gz. ga "face', i.e. “g"v[n)ge. This is not impossible; at least, it is diffi- cult not to connect this $. Ethiopian (and Tigrinya) item with the widespread Cushitic root from which G2, gés derives: cf. Bil.kem. gi3, Khm. gac, Saho gazaa, Som. géq ‘beard, chin', Sid. g@fo ‘chin’, etc. gina "left" + Tha. gokaw "left-handed', Arg. gata, Har. guta ‘left’, Gaf. gan, ch. gura, etc: Gushitic, cf. Afar guna, Som. giixey, Sid, guratio, Kem. gor. Leslau's (1956:203) suggestion ‘that gona should be related to Som,'gxb ‘wost' seems less likely than a Cushitic explanation. guriixo ‘throat’ 2 Gz, g*anse, Tha. gYonlino, g*ong*anit ‘goitre', Te. gore, Gak. quand, Ch. gY oniin. Goamen Semitic, a® (swallow and gngr “throat! ef» Ar. junta ‘gulp’, Heajana vb. ‘swallow’, tieb! gang®aot ‘throat’, AKK. gagguattu, etc. Goro ‘ear’ occurring only in Anharic. ‘The other Eth.Sem, languages all preserve the common Semitic item, cf. Gz. *vzn. Anh. Goro is almost certainly from Galla gutta, or something Like it. ginba "back" + Te. guabit. Praetorius (1879:84 and 94) derived this from the Semitic root doa. There are, however, several Cushitic itens, particularly Agaw, which provide a more satisfactory explanation: Bil, guaba, Khm. gétba, Quara gibra; also’ Som. géxab 'shoulderblade’. hod "belly" : Gz. kilbd 'liver', Tha, kébdé, Te. kiibed, Har. kid, Ch. xépt,, WL. hébd, ete. Conmon Semitic kbd. hu€atit "kidney" : Gz.Tna. RYolét, Te. kateet*vt, Har. kutiy, Arg. kutlay, Gaf. butlatit,’S1. bétayo, 2.” hatayo, etc; note also Old Anh. hotatit, AAL 5, 561977) A Comparative Approach to Amharic Lexicon 1s ALL the S. Ethiopian forms may be derived from a partially reduplicated stem *k" slati-. Connon Semitic key. and ‘forearm’ + Gz.Tha, kYoma®, Har, kuru, Arg. karna, Gaf. kandé, Ch.Gy. xondi, 81, kond, "WL. hard, 22 hand, ete. Amharic also has the item fain ‘elbow' which is apparently from the sane root, but does not show the usual development an >nd. Older Anharic forms of this latter item, urna and written kotna*, suggest perhaps the direct influence of Gz. k*o1na®, On the other hand, ond is the regular development of *koana®, The root is Common Senitic br*, cf. Ar. kuta® "foot, leg’, Heb. kena’, Akk. bitin. hingiix "Lip" : Gz.Tna. kang . hiimgain, Ch. bingin, etc. It is sonevhat difficult to decide whether this is an inherited Senitic iten to be connected with Moroccan Ar. kagina ‘groin’, or of Agaw origin: cf. Bil. hangar, Kom, + Quara hanpax, S.Ag. léimper; these Agaw items could, of course, be loans fron Ethiopian Senitic (Cohen'1969, no. 177). fuéa ‘testicles’ Old Anh, path, ‘This is of Semitic origin and represents a variant pattern of the same root @s in Anh. ant ege's cf. Soq. kehéeihen 'egg', Nehri kali, Sheri kahatit; note also Maghrebi Ar. (pl. qtavi)’ "testicle’. bértiin ‘clitoris! 1 Tha. kontiin, Te. binginit, Har. konton, etc. Semitic, cf. Datina Ar. qanfi@, Soq. kantar. ‘The same item also occurs in E. Cushitic: Som, kéntin, Galla béngin, etc” bain "right" 1 Arg, Bia, Gaf. kon, Har. béinct ‘warp’, WL. énmé, Ch.cy. ind | etc. ‘The Atharic form derives from a qatZé pattem of the root fyt* ‘be straight’ (Ina. ‘Anh. kéiuia), whicn is probably related to Ar. qt® (qanna‘a 'satisfy'), ESA. Zqn*. Praetorius (1879:44) connected this with Gz. g*évg, "4s ‘shin', which may be so, though the semantics are sonewhat aukward. In either case, the ultimate origin of this'item is alnost certainly Agaw, cf. Bil. qét ~ git, Kem. got, Kim. xuda; note also Galla huddu ‘anus’. 2obb "heart" + Gz. £abb, Tha. Lobbé, Te. Lab, etc. Comion Semitic 2b. mites ~ metas ‘tongue’ + Tha, mighas, Arg. miitas, Gaf. mofasd, Old. Har, mifhasan, ‘The root is common Ethiopian and general Somitic th (248) ‘lick'. The other S.Eth. Lan- guages use items related to Anh. anddbét, of Cushitic origin. AAL 5, 5716 D. L. Appleyard [AAL 5/2 mutak ‘saliva’ Gz. manak, Tha.Te. match, Har, maak, Arg. matai, Sod. tamagi,'Ch. ambaki, Enh. ambe i, End. mone*, etc. ‘The root is wrh (Gz. wixihd 'spit'), of common Semitic origin, cf. Heb. yiraq. ngs "soul" 2 Gz, nifs, Tna.Har. nigsi, etc. Common Semitic npi. nas ‘head’ 2 Gz, A0%, Tha. 198d, Te, adeus, Har. units. ‘The Amharic term ras derives from a qatag pattern, like Te. néas, whilst the element which provides the base of the Srd person pronouns preserves the pattern x9°s. Common Semitic 2°%. riz ‘beard’ : Arg. aniz, Gaf. siz, 2, aneda. ‘The Anharic, Argobba, and Gafat itens are probably of Sidano origin, cf. Tenbaro di whereas the’Zway item may be directly from Galla ateda. za sanba "lung! + Gz.tna. sinbut, Te. sdinba*, Ch. samb*a, Sod. sanb*o, etc. Of Gushitic, and probably specifically Agaw origin: cf. Bil. sdnbé, Kem. siimba, S.Ag. Sambi; note’ also Som. sdmbab, Galla somba, son ‘nerve’ Gz. Sow, Tha. sur ~ Son, Te.llar. sot, Arg. sored, Gaf. Soret, Ch. son, ete. Common Semitic, cf. Ar. surta ‘umbilical cord', Heb. 45% ‘nerve, mscle', Aram, Seryand [pulse’, etc, “This is probably a variant of the Semitic root 34% 'root'; the two meanings ‘nerve' and "root' have, for the most part, been collapsed in Ethiopian Semitic under the fone root, Saw. Scbiit "grey hair" 1 Gz, Sibiit, Tha, sibit ~ Sobiit, Te. Léb, Har. Libat, Arg. Sabie, ete. Conon Senitic yb. Anharic has fommed a denoninative yerb Sébbata in place of the original primary verb root form, as in Gz. Sebd, Ar. Saba, AKk. Sibu, etc. 3i "foetus" + Gz. doy, Tha. Satde ‘afterbirth', Te. satet, Common Semitic Sty, cf. Ar. sata, Hel, Situ, Akk. Léeztu, ete. Sanna ~ Séinna ‘urinate’ (n. Sent): Gz. Send, Sant, Tha. Sand, Su Common Semitic tun. ‘Also meaning ‘root’ in Amharic, as in some, but not all of the other ‘Languages: AALS, 381977) A Comparative Approach to Amharée Lexicon 7 4tigga ‘spit, vomit! + Gz.Te, ti§*a, Tha, tigre, Gat. tH§s%, Ch. tigan, etc. Several S. Ethiopian languages, including Anharic, have a descriptive compound derivative with the verb 'to say": lar. tug Gaya, Arg. ontog ata, Anh. ontogg ata, 2. Cagun bald, etc. Comon Semitic, cf. Ar. fa{fa, Aram. ¢pap.” Similar forms also o¢cur in Cushitic (Cohen 1969, no. 319}. tongai 'breath' + Tna, tongas, Har, tomgas. From the root ngs (Sem. np). tina ‘sleep! : Har, W@%2, Arg. enna, Chic) 2. ind, Wi. athe, E. noyyin, enn, nevi, End. nai, and probably also Tha. nihayii 'be tired of. Amharic treats the tas a radical except in the derived noun mainéta. The cognate forms clearly show that the underlying root form is nly (as Tigrinya) or *nyh (hence the palatal- ization to ft in sone $.Eth. languages). Praetorius (1890:43) connected this root nhy, including Gz. tinahay’ ‘confess sins" (2), with Ar. hn 'be weak, exhausted’ via a bi- radical nominal with the sten *hn-, restructured into a triradical verb root as nhy. ‘This seems a little contrived but is, perhaps, not impossible; at least, no satisfactory rival etymology can be found. The N: Ethiopian languages show a variety of roots for ‘sleep, Gz. mum, skb, Te. skb, Tha. skb, dks; of these mm and skb have numerous common Semitic’ cognates: ‘oyiim, Sod. anniin, S1. one, fabor Inail’ 1 Gas $06, Tha, gagni, Te. sagan, Har. fifin, Arg. kusgon, Gat. pfhd, Chidy. Pagan, oi Common Semitic zpr. 4tiigua "hair" 1 G2, gage, Ta. sag"né, Te. Sigan, War. Bigin, Arg. Sood Bae eee cates. danas, Ste It is an interesting point that Anharic shares the vocalization pattern fala with X. Ethiopian, whilst all the other S, Ethiopian forms can be derived from the pattem *s2géx. Eth. sqt'~ 49x is of Cushitic origin (E. Cushitic?): cf. Som, dégox, Saho daga, Afar dagiwru, Baiso ogorto; also Bil. Soqot, if this is a genuine, inherited Agaw item, and not a ldan from Ethiopian Semitic, This item has been connected with Sem. St, which may bbe correct at the level of Afroasiatic. The inmediate origin of the Eth.Sem. item, how- ever, is surely Cushitic rather than a purely Senitic variant of $*. fers "tooth" + Gz. zors 'molar', Semitic, cf. Ar. dits. AIL the other Eth.Sem. Languages preserve the common Semitic item, e.g. Gz. san. fat ‘Einger' Gz. *agbat, Tha, *agqbot, Te. Fobit, Har, afabiwia, Arg. fad, Gaf. Sati ‘hand, Sod. afabét, Ch. atebi(tl, etc. the Atharic, Argobba, and Cafat forms can all be derived from *s4b‘at, cf. Old Anh. $d2t, which Like the Tigre'item does not show the initial > of the renaining Eth.Sen. forms. Common Semitic 55°, AAL 5, 5918 D. L. Appleyard [AAL 5/2 fut ‘breast’ Gz.Te. fab, Tna, fub, lar. f0t, Arg. fut, Gaf. fuuui, Ch. tu, Gy. tao, ete. Only the Anharic, Harari, and Argobba items have a suffix -¢ and can be derived from Mgabt. Semitic, cf. Ar. féby "udder, teat’. wizza ‘sweat’ (vb) : Har, awiza?a, Ch.Enn.Gy. awzasa-, S1.W1.2. awiza-. Anh. wiz (n); Har. wi2e, Gaf. wied, Ch. woeat. Semitic wd*, cf. Heb. yeza*. Another Amharic item with the same meaning is fab (n), a£aba (vo), to which tna, Cahbat (n) may be conpared; cf. Sen. £hb ‘burn, be parched’. 2, THE SEMANTIC FIELD “THE DOMESTIC ENVIRONMENT” ‘The subdivisions under this heading are 2.1) agricultural activities and implements, 2.2) crops, 2.3) domestic animals, 2.4) food and its preparation, and 2.5) the house. In the field’ of agricultural terminology we are dealing with an area of the lexicon which is Likely to be susceptible to Linguistic borrowing and innovation in direct response to cultural borrowing, unlike the field of "nan" which has been exanined above. The basic agricultural nethods and processes, like ploughing, sowing, reaping, milling, and the nanes of the coumoner donestic animals may be regarded within the context of the given ecological area as being culturally so non-specific as to be classifiable in lexicographic terms as "basic" vocabulary. It nay, therefore, be reasonably assumed that the corresponding lexical items are not typically subject to ready borrowing or replacement. ‘this is in contrast to more specific itens, like the nanes of local corps, donestic animals, certain specialized tools, and 0 on., When we turn, then, to these areas of the Anharic lexicon, we find that whilst the “basic” vocabulary i mainly of inherited Semitic origin, there is a considerable number of non-Semitic terms anongst crop and animal nanes. In historical terms this can be inter- preted as evidence that the objects, but not necessarily the processes of agriculture in highland Ethiopia ove not a little to the indigenous, non-Semitic populations. Indeed, it appears (Canst 1969:11-12; also Simons 1970:124ff) that the Ethiopian plateau and surround- ing areas were @ centre of plant domestication and dispersal fron any early period, long before the earliest conjectured arrival of Semitic speaking peoples. Crops such a5 fe§ (Poa abyssinica), mug ,(Gudzotia abyssinica) were domesticated locally, whilst others like’ ‘wheat’ (soide), ‘barley’ (gébs), ‘finger millet’ (dagussa), and "flax’ (46a) seem to have reached Ethiopia early on.'* The names of these cereals aré of non-Senitic origin, most probably from Agaw. An exception to this is ‘barley’ (aibs), which may be Semitic.” The term for ‘wheat (Sonde) is probably of Cushitic origin, thoigh perhaps only as the contamination of a original Senitic form. Of course, this does not mean to say that the crop itself was un- Jnnown to the incoming Senites, Other crop nanes like atx 'pea', bakeda ‘bean’, mosson "lentil", béxbixe ‘red pepper’, and so on, are loans from outside Ethiopia, mostly either fron or through the medium of Arabic Amongst the nanes of donestic animals we find bine ‘ox', dibie ‘kid’, (dxis ‘horse’, gomiit "camel", gotgdt "young animal’, Zam 'cow', and kdbt ‘cattle’ of inherited Semitic origin, whilst tems Like alioyya ‘donkey', dammit’ ‘cat', doxo ‘chicken’, soyit ‘goat’, wabla ‘dog! ‘Conti Rossini (1928:106) attributes the introduction of “nolte piante utili specialmente per 1'alimentazione" to the South Arabians, but see Ganst (1969) and Simoons (1970). AAL 5, 60)1977) A Comparative Approach to Amharic Lexicon 19 and probably big 'sheep' and biipto ‘mile’ are of non-Semitic origin. Of these only the last, two, big and bilo, have cognates distributed throughout Ethiopian Semitic; doxo and donmit are shared by N. Ethiopian and Anharic; woSSa and foydit are common to Anharic and neighbouring S. Ethiopian languages; ahoyya, at the other end of the scale, occurs only in Amharic and has no cognates in Ethiopian Semitic. In the instance of these restricted loans we are dealing with localized items and, indeed, the 1ikely source languages are identifiable in each instance. When we examine the terms for these aninals in other Ethiopian Semitic Languages, we find a wide array of items from various sources, both Cushitic loans (Har. adwttu'cat' from Galla; Har. buf ‘dog' also fron Galla; Gaf. kutti ‘chicken’ from Sidamo, and $0 on), and inherited Semitic items not preserved in Amharic (Tha, hélbé 'dog'; Har. ‘fy "goat'}. As I said in the introduction to this paper, I do not deem it wise to try and Gray historical or cultural implications fron the history’ of individual jtems._ Nevertheless, T think it is a legitimate question to raise why, for example, are such "basic" tems as ‘dog’, 'goat", etc., of non-Semitic origin in the Amharic lexicon, whilst other Ethiopian Semitic languages preserve the comon Semitic iten? It would be naive to answer this in concrete terms; rather the significant point lies in the distribution of the non-Senitic languages from which these terms were taken. In the case of wob8a, for exanple, the apparent Source is Sidamo; foyi£, on the other hand, is seemingly of S. Agaw origin, Both of these languages were spread over wider areas in former times and have for sone tine been subject ‘to the encroachment of Anharic, These terms are concrete evidence of the substrate over which Anharic speech has spread and developed. the instance of the two terms big ‘sheep’ and békeo "mie", however, is somewhat different. Both terms are widespread throughout, Ethiopian Semitic, including Co'ez, and it can, therefore, be proven that their “introduction” into Ethiopian Semitic is of an early date. They aré not localized loans replacing Semitic terms preserved elsewhere, but are fundamental constituents of common Ethiopian Senitic vocabulary. In the example of the mile, békéo, a positive statement can be nade to the effect that this animal seems to have been arelatively late introduction to the Semitic speaking world, there being no common Semitic term tor the animal; indeed, Arabic borrowed the term bagé from the Ethiopian side of the Red Sea. Tho sheep certainly was well known to the Semites, there being several common Senitic terms for the aninal, none of which, hovever, occur in Ethiopian Senitic (5°, gn, ete.). In the field of food terms the following items are of clear Senitic origin: ayb ‘cheese’, stat ‘evening meal’, betz "honeyed water", b8£, "be cooked, ripe’, doket ‘flour’, b's "break bread’, max "honey', mosa 'midday meal", dago 'yoghurt', and fore "fresh, raw'.” Probably also of inherited Semitic origin are botA' "ferment", {th "boil', kobe 'butter’, misob "basket table", motad ‘griddle’, tbs "fry", and wit ‘stew’, though the precise derivation of these itens is not clear and the likely Semitic cognates’are of weak formal and/or semantic fit. Perhaps more than in the other areas of the lexicon discussed here, food terms are subject to borrowing and influence from outside as fashions fluctuate and trade introduces new food crops, Thus, the names of all the typical spices, otd, bérbiixe, zanjabat, table equipment Like jabina ‘coffee pot', fongat ‘coffee cup’, sdian ‘plate’, dost ‘cooking pot', Subka "fork", etc., and fruits’1ike mz "banana’ and foné '1ime', hot to mention more tecent intro- Guctions, are all loans mostly from or through the agency Of Arabic. | Probable Cushitic con-, {wibutions to, food terminology, in Amharic are bisso "roasted barley flour’, ona ‘fatty neat’, fio 'sait', btanga ‘dried meat’, Sono, "chickpea paste’, S.mbata ‘chickpeas’, "and possibly also ingixa 'bread', tigea 'beer', and 733 ‘honey wine’. OF the names of parts of the house etc., for which a satisfactory etymology can be established, the majority is, of inherited Semitic origin. These Senitic items are typically nares of the most basic and "primitive" architectural features, like ator 'fence', bitx ‘door, gate’, bet "house", dij ‘gateway’, mindér 'village', kodan ‘thatch’, mosiso ‘centre pole’, and probably also madaiga 'hearth'.’ The majority of the names of other parts of the house, Like g"ada, fara, watta, gutlsééa, and constructional features like gotina, gamb, hab, etc., remain AALS, 6120 D. L. Appleyard [AAL 5/2 unidentified as to origin. The few readily identifiable Cushitic terms here are gadgodda ‘wall of wattle and daub’, kot ‘loft, high shelf", gogo 'straw hut". More advanced and sophisticated terms, like’dixb ‘upper storey’, gok "storeyed building’, dindja ‘staircase’, wikan, etc., are of Arabic origin, In the field of the domestic enviomment, therefore, there is a relatively high proportion of non-Senitic material (25 itens out of’ total of 93 studied : 26.9%), all of which con- sists of nanes of specific animals, plants, or foodstuffs. To these may be added the three building tems noted above. Perhaps the most fundamental ‘observation that can be made here, and which, incidentally, is also valid in the field of the natural environment, is that’broadly speaking general’ terms are inherited Semitic, but the specific and typically Ethiopian tems are of non-Senitic origin, 2.1, AGRICULTURAL ACTIVITIES AND IMPLEMENTS afSidé "reap" + Gz.Tha. *apiidd, Gaf. appiidi, Sod. addicim. ‘The palatalization of the medial radical in Amharic may have arisen in the imperfect (yisagyud > yagod), or the gerundive (fapéde > afodo). Common Semitic “dd. aebiibit ik" 2 Gz.Tna. hage SLWL.2. , Te. hatha, Har. hatiba, Arg. hatediba, Ch. andbéin, etc. Common Semitic heb. anni weed" : Har. hakam (n), Arg. haxrdma, Sod. isndndim, Ch, andnin, etc. AIL the S.Eth. forms go back to a B-type conjugation of fm, which occurs in N.Eth. in the A-type conjugation (C2. haxdni, etc.) meaning ‘prohibit, be prohibited’. Common Semitic jvm. anniisé ‘plough’ (vb) Ge . fanisd, Te. harsa, Har. hanisa, Arg. hanadsa, S1.ML.Z. etc. Common Semitic rt. The instrumental noun from this root (Amh. mardS$a) provides the name of the plough, or occasionally more specifically the ploughshare (Tna.Te. mahriSa) in most Of the modern’ Ethiopian Senitic languages, Also frop the sane root and of fairly wide occurrence is the noun ‘agricultural land: Anh, ots, Tna, honda, Te. anos, Har. andi, Gaf. omnaba. ‘nil! : Gz, féigha "break into pieces’, Tha, fésgohe ‘grind’, Har. fifa, Arg. §a$fa, SI.WL. GaSe, 2. a2, Emn.End. fe*d, etc. ‘The S.Eth. forms all appear to derive fron a root *4gy, beside N.Eth. (sh. Semitic, cf. Ar. fadda’ "break open", gadaka ‘expose’, Heb. pasa ‘crush’, pas ‘split’, etc. bidda ‘draw water" : Ge.Te. tha, Tha, hédke, Har. hédaha, Arg. hiiddiha, Gat. baidda, Gh.” gétam, ‘etc. Senitic, cf. Ar. gadaha "bore', Heb. q@dal. bambéin "yoke" + Tna.Arg. finbix, Gaf. bimbéixd, and probably also Gz. finn197} A Comparative Approach to Anharie Lexicon a This is alnost certainly Semitic and related to Syr. gama ‘belt', though the item does occur (a5 a loan) in Cushitic, cf. Galla kanbani, Sid. kanband. bénniini 'glean' 2 Gz.Tna. pariind, Har. hiiama (n) ‘wheat, stalk used in basket weaving", Sl. kaame 'stubble", (Anh. bidum) . Semitic, cf. Ar. qaxama 'gnaw', quia "tree stump’. mij ‘upper grindstone’: Gz. mde, Har. mgd, Oh. mii, SL. miijGe, 2. mic. ‘The root is dhy (Gz. dahayé 'grind'), to which Heb. d@h@ ‘push, thrust’, Ar. dah@ ‘spread’ may be conpared. min} maSon 'winnowing basket' : Gz. méS*e, Tha, méste, The root is S*y (Gz. Sa€ayi), to which Ar. Ja°% may be compared. niigfa ‘sieve’ + Ga.Tna, niféyi, Te-Har. niga, 2. niga ‘The root in Anharic has shifted to the biradical pattem that generally presupposes an ori- ginal laryngal as third radical. In the instrumental noun (uingit), however, the original Toot form persists; cf. Gz. minge, Tna. mingit, Har. wingét, ete. Semitic, cf. Heb. néped (vb) , napa (n). sateibié ‘castrate’ : Gz. Sdtiba "pull out', Tha, sdeaba ‘castrate’, Te. sitba, Har. sééba, etc. Semitic, cf. Ar. sataba ‘pull out’. tikbiité "plant" i, Tna. . tikea, Arg. tekkiita, Caf. takeatd, Si.WL. Echt, ’Ns.Go. Sakkéein, etc. In most of the S.Eth. languages, though not in Anharic, this verb conjugates according to the B-type pattern. ‘Formal cognates occur in Ar. zakiéa 'trust', Aran, ttket; the semantic correlation between these and the Eth,Sem, tke ‘plant, fix’ is weak, but not impossible. tiimmiidit "yoke" (vb) Connon Semitic dnd. pind, Ina. samidi, Te. ganda, Har. tiinida. zinna "sow' : G2.Te. zén*a, Ta, zér*e, Har. zina*a, Ch. zlinam, ‘The’ noun 'seed' (Gz. zia*, Anh. zx, etc.) from the same root is fairly widespread throughout Ethiopian Semitic. Common Semitic das (Ar., Heb., Aram., Akk.) ~ dt? (ESA., Eth.Sem.). 2.2. CROPS atin "pea" : Tha,Te, ‘atin ‘chickpea’, Har. atin ‘pea’, Gaf. atind, 2. antind, Ch.Go. atind "bean", etc. AAL 5, 632 D. L. Appleyard IAAL 5/2 This is almost certainly a loan from Aden Ar. ‘atax. The tem also occurs in several Cushitic languages: Galla atata, Kanbatta atand, Kem. azdn, Kh, adin.’® ohst 'grain’ 2 Gz, tok, Tha. *oxti, Te. *obat, Har. oxé, Arg.W1.2. shot, Chi dxot "barley", ‘Sod. dkot, M. axl, etc. Common Semitic *ht ‘eat’. abit ‘unripe grain’ 1 Gz, Siivét, Tha, siinmuit ~ Séieit, Te. Har, asZta~ adita, Arg. obed, Caf. abat, ChiW1.2.cy. set, Sod. asit, eve. ‘The S.Eth. foms all derive from *giyét, with attraction of w to y under the influence of the following <. ‘The root is Suy "ripen’ (Gz. Sduwiyit), which is probably to be identi- fied with Ar. Soy Gaut ‘grill, roast", *alut ‘be ripe’for grinding (of grain)'). baketa "bean" : Te.WL.Z, bakeea, Har. baketa, Sl. bakeLea, Gy. bikéta, ete. A loan from Ar. baqitd*. The Senitic root bgt exists in Ethiopian Senitic (cf. Anh. baguété 'sprout'), but this nom is certainly a loa fron Arabic and not an inherited ‘om. bun ~ bunna ‘coffee’ So throughout Ethiopian Semitic and Cushitic; a loan fron Ar. bunn, perhaps. However, insofar as the coffee plant is itself of Gthiopian origin, one wonders whether the tem burn might not also be’ of Ethiopian origin. bixbixe "red pepper’ + Gz. binbiixe, pépine, Tna.Te.Arg.N1.2. banbixe, Har. binbiné, ete. ‘The alternative Ge'ez form, piipiixe, looks ike a direct loan from Gk. péperi, whereas banbéne must be from another source. Semitic languages other than Ethiopian Semitic have the Toot plp{~ sL4€. The form of the root brbr occurs outside Ethiopian Semitic only amongst Cushitic languages. Whatever the inmediate origin of the form binbéne is, the ultimate source of all these nanes is Indian, ef, Pali pippatt. Gat ‘catha edutis" So throughout Ethiopian Semitic. ‘This item is undoubtedly related to the Arabic name of the same plant, @@t. Since the plant itself appears to be of Ethiopian origin, it would Seem Teasonable to assume that the name is, too, and that Arabic has borrowed the term from Ethiopia. The item also occurs widely throughout Cushitic. dagussa ‘finger millet' :Tna.Te. dagussa. “see Conti Rossini (1912:174): "la variété qu'on a pu étudier dans le nord d'Ethiopie est le Cicer arietinum, et qui serble originaire du bassin oriental de la Méditerranée, d'od il aurait té importé dans le Yemen et en Abyssinie." AAL 5, 04197] Comparative Approach to Amharic Lexicon 23 Gushitic, cf. Bil. dagisa, Khm, dausa, Kem, dauSa, S.Ag. dagusé, Galla daguzd ~ dagui. donnsé "potato" + Tna.Te, dann’, Hgr. démi8éa, Ch. doniSa, S1.WL.Z. damasbia, Cy. dinéa, etc. ‘The nane originally referred to the edible root Coleus tuberosus, but is now applied to the potato, Solanum tuberosun. From Galla dannitéa ‘the Galla potato’, Coleus edulis. CE. also S.Ag. “dunézi ‘potato’ (see Mooney 1963:12). gits "barley" + Har, gis, Arg.Sod. gibs, 2. gébas. ‘This is generally (Praetorius 1879:48; Cerulli 1936:251; Leslau 1963:76) regarded as a metathesized form of the root gm (Gz. sigim, Tha. sogin); i.e. ‘sig.m> ‘gimlls > gibs. Gz. sdigdm, etc., is probably of Semitic origin, cf. Soq. Skinoh ‘grain of millet’, though certain Agaw languages have an apparent cognate form, perhaps an early loan fron N. Ethiopian: Bil. sind, Khm, sokamd, S.Ag. sinké. mossor ‘lentil’ 2 Tes masan~ minsan, ar. missin, Arg.Gaf. mossan, Sod. mabB und, S1.W1.2. mabon, A different root form occurs in Ge'ez and Tigrinya botson. Both root pattems occur in Cushitic: Son, méisix, Kanbatta mind, etc., and Saho bursén, Khm, béaséx, etc, The shape of the Ge'ez and Tigrinya itens is inmediately reminiscent of Ar. bulsut, ‘whilst the rest of Ethiopian Semitic forns with initial m- are formally closer to the ultimate origin of the root, Pali masina. This suggests that perhaps massa, etc., was taken directly from some Indian source. maSelfa ‘greater millet’: Tna,Sod. masta, ch. naXina. This is probably of Aga origin, at least compatible cognates occur in Bil. mata, Khm, ayes, Ken, mayla, Quara nila, $.Ag. mela. nag "Guizotia abyssinica’ Tna. néhug~ nog’, Te. nvhig~ nahag" . Cushitic and probably specifically Agaw, cf, Bil. Cohong"@, Quara Long’ a, Kim, nui, S.Ag. nugi, but note also Galla nugé, Saho nehig. ‘onde "wheat ' + Gz, Sprnay, Tha. sormay, Te, Sonnay, Har. ssaré, ‘wheat bread’, Ga. sond@, Ch.’ sand, S1.Z.'sone, Sod. Sarai, Gy. sonay, "MS. donne, etc. Tt has been suggested that this common Ethiopian Senitic iten is to be derived fron the Semitic root 4*x, i.e. from something like *So*aanay. This would not, however, be a regu- Yar process of derivation. Better formal (and semantic) fits occur throughout Cushiti Bil. Senay, Calla sénta, Som. saneon, Mar sca, and perhaps also Beja sexaan. This List certainly inctudes foans fron Ethiopian Semitic (Bilin, for example, is, the only Agaw language here; the others have a totally different item, c.g. Kem. Gang" a, Khm. zw, etc.), but others do have the appearance of "genuine" Cushitic items. Simbana chickpea! 1 Tna.Te. s@bbiéxe ‘type of pea - Lethyrus sativus’, Har. Sunbura ‘chickpea’, Ch. Sonbora, etc. AAL 5, 68En D. L. Appleyard [UAL 5/2 The S. Ethiopian itens are probably of Sidamo (or Galla) origin: Sid. Sunbunaa, Galla Sunbura; for the N.Eth, forms cf. Saho sabbanee. Sonkurt onion’ 1 Gazz Sag" and, tna. Sogtanti, Har, Zonkinta, Arg. Sonkunt, Gai. Kunkurt, O14 Anh. Song ant ‘The item also occurs widely throughout Cushitic (Bil, 42g" anti, Kom. surta, Galla Sunkuata, etc.). Gz. og" sad and all the other forms, both Semitic and Cushitic, are ultimately from Gk. shéndon ‘garlic’, 4titba ‘flax, Linseed’ 1 Gz, tobe ~ witbe ~ totabe ~tiitabe. A loan from Agaw, cf. Kem. tarba, Khm, taba, etc. fe§ "Poa abyssinica" : Tha.Te. fag, Har. 4242, Arg. fef, Ch. fag, End. 424, etc. Qushitic, probably Agaw, cf. Bil. taba, Kh, ib, Kem, 2b, S.Ag. 244% 'millet', but also in Galla’faafi and Saho"Afar déagi. It is likely that there has been a certain’ degree of lexical borrowing between various Ethiopian languages, Semitic and Cushitic, in the spread of this item. The forms of the Galla and Saho-Afar itens look more like original loans from Ethiopian Senitic than inherited Cushitic stock; the t/d initial would not be the regular correspondence to Agaw 2 (see Dolgopol 'skiy 1973:54ff), but would regularly represent ¢ in @ loan from Ethiopian Semitic. On the other hand, the representation of a non-ejective by an ejective in an Ethiopian Semitic borrowing from Agaw would not be improbable. ‘The dialect Arabic forms, Datina fahaj~ falas 'Myrica gate” and Hadrami tahag (Leslau 1963: 152) are perhaps merely coincindences. The plant is apparently indigenous to Ethiopia, and these two Arabic dialect forms imply a root with a medial laryngal radical, for which one of the Ethiopian Semitic terms show any evidence. It would not, therefore, seem un- reasonable to propose a Cushitic, and specifically Agaw origin for this item. 2.3. DOMESTIC ANIMALS ahayya ‘donkey' : Tha, *axoya besides *adgi is probably a loan from Aharic, Other than this there appear to be no cognates of Anh, aloyya in the rest of Ethiopian Senitic. This item is probably to be connected with Saho-Afar otdato, also meaning 'donkey': Fakot + ya > ahcyya. bag "sheep" + G2, béiggo®, Tha. biggi®, Te. baggu®, Arg. bagi, Gaf. big. A clearly related fom occurs in Agaw, cf. Bil. bigga, Kem. biiga, Khn. baga, which Conti Rossini (1912:175) regards as a loan from Ethiopian Semitic, The usual Semitic terns for ‘sheep’, $°, dn, etc, are absent from Ethiopian Semitic, though the root #ty (Gz. fité tgoat', Ar. ’tata’"Lanb!, etc.) does occur in certain S.Eth. languages with the meaning ‘sheep! (liar. fy, Sod.’ day, Ch. fe, etc.). ‘Those languages which employ the root bg° (N. Ethiopian, Anharic, Argobba, Gafat) are, however, within a likely radius of Agaw con- fact or substratum and’ though this root has been described as Senitic (cf. Ar. bagbaga ‘bleeting'), it would seem more likely to be an indigenous Agaw and Cushitic root (cf. Beja book ‘billy-goat'). AAL 5, 661977) A Comparative Approach to Anharic Lexicon 25 bieto 'mile" + Gz, bike, Tha. baxee, Gag. bogoed, Ch. bogae, etc. ‘This is almost certainly of Cushitic origin: cf. Bil. bixtA, Kem, bila, Khm, bagta, Saho-Afar baketa, Qabena bikulita, Som. bagat. Arabic bajt is a ioan from Ethiopian’ Semitic (Gz. bake) (Nldeke 1910:583 also’ Hommel 1879:112£f), -_biikal, Har. bade, Arg. bileo Sait, 81.1. "bobto, 7. bokolit, Sod. béxe "bull, ox" 1 Gz. butaavi~ boaay~ bora, Tha.Te. butray, Har. ba*ana ~ bana, Arg.W1. bara, Ch.Sod. bona, 2. baxd, etc. It is interesting that all the S. Ethiopian forms except Anh, bite can be derived from a Pattern ending in -a, whilst Anharic agrees with Tigrinya and Tigre in having the suffix “ay > -e. Connon Semitic ba (Ar. ba®ct ‘camel’, ESA. ben, Heb. beZ1 ‘cattle’, etc.). dibée "billy-goat' : Gz.Tha, dabela, Te, débita, lar. débay 'heifer’. Senitic, cf. Ar. dubt ~ dawbat "young donkey’, davbal ‘suckling pig'. CE. also Anh. dabbot “young domestic’ animal’. dammit "cat! : Tna,Te. damm, Of Agaw origin, cf. Kem. damaya, Quara danyit. dono ‘chicken’ + Gz. dorho, Tha. dirho, Te. derho, Arg. dono. Gushitic, cf. especially Saho-Afar doxho, Som. dono, Beja endirhoo; note also Agaw *ditw-a: Bil.Kem."dinua, Khm. Gia, S.Ag. dix. “An East Cushitic (cf. Saho-Afar) origin seems more Likely than an’Agaw one because of the presence of the thitd Tadical h. (ids horse’ : G2.Tha.Te, finds, Har.Arg.Ch. (ardz, Caf. (éndd, etc. Forms with final radical z~ d occur throughout Cushitic and it would seem likely that such Cushitic forms (originally loans from Semitic rather than inherited Cushitic items?) have influenced those Ethiopian Semitic forms in -z/-d. ‘The root is ultimately Semitic, pti. Soytit "goat" t Arg. géyel, Gaf. gait. This item is of Agaw, and specifically S.Agaw. origin: S.Ag. giyélz, cf. also Denbiya fiyata. , Whether these itens are inmedistely relatable to the rest of Agaw: Bil. gontina, fotora, Kem. fantiixa, or not is debatable. getgiit "young animal" 2 Tha, gatgi Har. 9% etc. Tip Js « smdopltcati gat ot tha tan rove an cocuring tn Gr, tay'f "ae, om Sonitic of (Eth.Sem. ‘A similar reduplication occurs in Syrian Beduin Arabic genai "large lanb’. "young mile or horse’, Arg.Gaf. g:tgée ‘Lanb', Fyotmg animal", MNS. gataie» Ch.8y ohangon, gone ‘camel" + Gz.Tna.Te. ganera, etc. , Nar, ginita~ gfimita, Arg. gameta, Ch, Common Semitic gmt. AAL 5, 6726 D. L, Appleyard [AL 5/2 tan ‘cow + Gz. tahm, Tha, Lahm, Har.Sl. Lam, Arg.WL.2. Lam, Sod. diam, Chicy. diam, Gat. iad, inn. axdm, Co. dnnam, End. anaw, etc. This has been connected (ULlendorf£ 1956:192; also Krotkoff 1969:72) with the Senitic root ‘lm : Ar. Lalm 'food", tieb. Lehiem 'bread’, Soq. Lehem ‘shark’. Despite the slight anomaly Of the |i + h correspondence this does not Seen inlikely. Dillimann (1865:col,25) cites an Arabic term Lihm "taurus annosus"; perhaps the Semitic root £lm (Ar. Lima ‘swallow greedily") has influenced the development of the Ethiopian Semitic root in sone Way, too. mmiinga "flock, herd" 2 Gat. minga. Praetorius (1879:169) derived this from the root wig "guide, conduct", but a Cushitic deri- vation seems to ne the more likely: cf. Afar mango ‘herd’ from the root mag- 'be full’. singe ‘castrated animal’ ©: Tha. sanga, Har. sanga. Probably from Galla sanga, if this is an indigenous root. Praetorius (1879:98) preferred a derivation from Gz. sg" ‘stab’. toa ‘cals’ + War, £éa~ toa, Arg. tala, Gaf. gé0"@, Sod. £29, ch.Gy. dak, Ms.Go. digg, etc. This item occurs only in $, Ethiopian; the N. Ethiopian languages use a variety of terms: Gz. *ug'€, Te. *ugal, faluy, Tha. motax, etc. ‘The S. Ethiopian root may be characterized as *fu/ég"a, for which no likely origin’ can be identified. wok$a ‘dog! + Arg, wolda, Sod.Gaf, waSSé. OF Sidano origin, cf. Sid. wod0, Kanbatta woliSG, Gudella wika. The other S. Ethiopian languages have various itens of non-Senitic origin: Har. buté, wie ep Ch. goyé, etc. Only in N. Ethiopian is the inherited Semitic term preserved: Gz. kab, etc. wigiifo "kid" : Tha. wiiozo. Cushitic, cf. Sid. wafa, wofitto, Galla wotiyo (2). One wonders whether this Cushitic | root is in some way cognate with’ that which seens to appear in Ge'ez as a loan: wiiysit ‘wild goat’, cf. Afar wadaat, wadaar. 2.4. FOOD AND ITE PREPARATION anbaiSa ‘wheat bread" + Tna,Te, hombaSa, Old Anh, hombadSa, Probably a derivative of the Eth.Sem. root bz ~ bs (Gz. fyebast) with the suffix *-ya and an "intrusive" nasal. ayb "cheese" + Gz.Tha.Te. ated "milk", Har. hy, Arg. hayu, SLWL. ayb, 2. ayab, Ch. eb, etc. Conmon Semitic 2b (see Anh. atsibi, above). AAL 5, 68,1977] A Comparative Approach to Amharic Lexicon 27 angina "bread" : Tha.Te. *ongeaa, Arg. gangix, Ga. gongiind, Sod. gangin. The Anh. verb gaggini "bake bread’ has the sane root form with initial g- as the nominal forms in other S. Ethiopian languages, which, according to Cohen (1939:388), may be the original form of the root. Praetorius (1879:169) attempted to derive andra from Ar. Sugayr ‘nillet'; this is not implausible. However, there exists an item in Kenant and Quara which may’ suggest an Agaw origin: gina ‘a type of bread". It would be easier to explain *ongena, ongdna as an Ethiopian Semitic extension of Agaw gixa, and the verb root and nominal forms in initial g- as a kind of reduplicated extension thereof, than the ankutat ‘egg" + Tha.*onkutatih, Arg. onkutat, Gy. onfura. ‘Another version of the same root occurs in other Eth.Sem. languages: Gz. *ankokho ~ onkokho, Tna. *onk*ax" oho, Te. *onkokho, Har. akul, Caf. ank*d, 2. onkau, Sod. anko, Wl. ankakot,'etc.' The root kik which lies behind Ath. onkutal, eté., also occurs in Anh! bua ‘testicles’ and has satisfactory Semitic cognates, cf. 'Soq. fehétihen, etc. Similar forms to the other Eth.Sen. root pattern (Gz. *ankobho, etc.), however, occur throughout Cushitic: Quara kajawind, Kem, x" andg"ina, Calla ankako, ankaku, Saho'unkokoho, Som, Sgah, Rendille kun, Werize ugahe, Konso ukb"a, etc. The two Toots of Ethiopian Semitic may not in fact bé directly related; it is not impossible that Gz. *ankobho, etc., is of Cushitic origin. axat ‘evening meal! + Te, hanab ‘meal’, Har. haat, Arg. horbad, Ch.Gy.Sod. dbat, S. uAbat, 2. tbat, etc., also Old Anh. hanat. ‘The root is xb, which also occurs in Anharic in the verb tarnibi ‘dine’. Cohen (1939: 25) suggested a connection with Sem, ‘xb ‘evening’, which seems very possible. bokka ‘ferment’ 1 Gz, boh*a, Tha. bax" se, S.Eth. bikka/bokka, etc. ‘The modern Eth.Sem, forms nay all be derived from *bb!*!2/e with pseudocorrection of ft to vis-a-vis the Ge'ez root. This root is almost certainly Senitic and may possibly be con- nected with Ar. nabaja’ "be sour’. batz ‘honeyed water" + Ge. mozn "beer", Tha. borzi "honeyed water’, Har. biazi, SLIWL. banz. ALL the modern Eth,Sem. forms have the root baz, beside Gz. and Semitic mar: cf. Ar. mézn ‘type of beer’, ESA, mzn. basso ‘roasted barley flour': Tha. bisso ~ bosso, HHar.W1. béisso, 2. béssu, Ch. biisui, Sod. bissuea, etc. Of Cushitic origin: cf. Saho bufa, Galla baso, Qabena basstita; apparently not connected with the Semitic root 6s¢ 'be cooked",’as suggested by Praetorius (1879:73). bissila "be cooked, ripe’ : G bésiti,, Te. basta, Har. bisa, Gaf. bissiitd, Ch. abisinin, etc. Conmon Semitic b8¢ (Heb. baat, Soq. behel, ESA, bs¢ (n) ‘sacrificial hearth"). AAL 5, 6928 D. L. Appleyard TAAL 5/2 ona "fatty neat! So throughout modern Ethiopian Semitic. From Galla Sona. ‘salt’ 2 Gz. gow, Tha.Arg. OF Agaw origin, cf. Bil.Kem. Soua, Kim. Zaua, ete. . Te. Foun, Caf. dabbo ‘wheat bread’ : Har. dabbo, Sod. dabbo, Ch. dap"a, Enn. dapa, etc. Of East Cushitic origin, cf. Galla dabo, Kanbatta dabbita. doket "flour" 2 Old Anh, date. CE, the Eth.Sem, root dp 'grind', Common Semitic dag. dost ‘cooking pot! 2 Tha, dasté, Har. disti. From Ar, dist. Gitta "boil" 2 Gz. ffitha, Tha. (lithe, Har. fétaha, Arg. séeeiha, Sl. fata, Win2. ata, Sod’Ms.Go. ¢ittalm), Ch. tagenam, etc. ‘The Eth.Sem. root lf) is a perfectly "respectable" Semitic root shape, though it lacks any satisfactory cognates in the rest of Semitic. Dillmann's (1865:col.1339) correlation with Ar. (wt (fata) ‘cui et saturiendi et bulliendi vis inest' is not satisfactory, only one out of the three radicals being shared. Despite the lack of a staisfactory cognate else- ‘where in Semitic, I'do not think it unreasonable to assune a Semitic origin for this root. abe "butter" + Gz. fob’, Tha, bab*d~ babed, Arg. kobi~ pawi, Gat. kab", Ch.Sod. ab, Enn. *2b, etc. The Toot fb? occurs throughout Ethiopian Semitic in the sense of ‘anoint, grease, butter’, etc., but lacks any truly satisfactory formal and semantic cognates in the rest of Semitic. Formally compatible roots occur in Heb. qaba* 'fix, insert’ and Ak. qabi ‘say’, but the semantic disparity between these and the Eth.Sem. toot militates against any reasonable correlation of the roots. Nevertheless, as in the case of the previous item, the root is in accordance with the standard Semitic patterns and there is no reason to refute the likelihood of a Semitic origin. fanga ‘dried meat" 2 Tha.trg.Sod. banta, Har. hanga, Te. kanta, Gat. und ‘raw meat', Z. bYonga ‘dried fish’. ‘A loan from E. Cushitic, cf. Afar ,"anfaa, Calla ganfa, Kambatta goontd. brduiisd ‘break bread’ «Tha. andisdi, Har. foniisa 'take a portion of s.th.'. Throughout Gurage the root 4s > ns has the meaning 'begin'. Semitic, cf. Ak, qaxdSu ‘split’, n. qirbu ‘slice of bread’, Heb. gered ‘plank rman "honey" + GzeTnasTe. misax ‘honeyconb, honey", Har. ma 'wax', Ch. max "beeswax', Go. man, etc, AAL 5, 70197} A Comparative Approach to Anharéc Lexicon 29 Most of the other Eth.Som. languages preserve the common Semitic term for ‘honey': Har. dis, Arg. dims, Gaf. dabsd; this root also occurs in epigraphic Ge'ez (see Drewes 1962:34 and’85) ‘as dbs, but is replaced in "classical" Ge'ez by mifax. The original meaning of ‘this item was 'honeyconb" from which the various meanings 'honey’, ‘wax", etc., have been derived in the modern languages. Semitic, cf. leb. yatax ~ yard’ "honeyconb'. mosa ‘midday neal" Gz. Tha, masah. Common Semitic mif, cf. Gz. missoha 'amoint’, Heb. maak, Ar. masaha. misob ‘basket table’ So, too, in Go.Tna.Te.Arg.W1.Z.Caf. Praetorius (1879:159) connected this with the Semitic root feb, in particular with’ Ar. matzb "locus in quem homines disgregati conveniunt'; note also Mod.Heb. m@sibbé ‘social gathering’. One hesitates to relate such a concrete tem as, misob with the abstract and general sense of mscbba and matt), though the idea is attractive and there can be little formal objection to it. motad ‘griddle’ + Har.Sl. mogid, Arg.Sod.W1.Z. mofad, Ch.Gy. madad, etc. The root of this S. Ethiopian item occurs in its simple form in the Anh. verb tadé ‘cook on a griddle’. This might tentatively be compared with Gz. shd~ g¢d (aghadd ‘soften, rub with oil"). soga "neat" 2 Gz. Saga, Tna.Te. saga. Of Agaw origin, cf. Bil. sixa, Kem. siya. ‘The other Eth.Sem. languages preserve the com- mon Semitic root bia: Har. isdn, Caf. basdrd, otc. Sono ‘mashed chickpeas’: Tna.S1.W1.2. Sono, Har. San, Gaf. Soni, Ch. Yomi, ete. Of East Cushitic origin, cf. Galla Suto, Sid. Haxo, Tenbaro Suri. fiibbisd "fry" : G2.Tna. fibisd, Har. fibisa, Ch. fibisim, etc. Probably Semitic, cf. Ar. fabbasa 'smear'. £893 ‘honey wine" + Har. #4352, Ch. $43 (a loan from Anh?), dg" "honeyed water’, End. degi, etc. ssible cognates for this item occur throughout the Omtic languages: Kaffa doofoo, ina’a déwsoo, Nota dé"g0, Anfille dad, meaning 'beer'. In Sidano and Qabena there occur forms similar’ to Anharic $434 (tag§(e) and fa53cta, respectively), but these look Like loans from Atharic; other Sidano languages use a different item, cf. Had. dégasa. The N. Ethiopian languages have Gz.Tna. mes, which has cognates in Aga: cf. Bil. md, Ken. méz, etc., and in Saho mees (a loan’). titta ‘beer’ : Gh, fatea, Perhaps also to be related to the Amharic item are Tha. fatka 'bira di Linseme', Te. satka. ‘These last two items can be readily derived from the Agaw root of the same meaning: Bil. MALS, 7130 D. L. Appleyard TAAL 5/2 gilaxd, Kem. sila ~ s2tag, Denbiya satay, Quara selaia, S.Ag. S052; the Khamir term 32Ua, ‘on the other hand, is more likely a’loan from Amharic than an inherited Agaw item. ‘The origin of Anh. fiita'is almost certainly to be sought anongst these Agaw forms and their developments in Ethiopian Semitic, Tna. fata and Te. sotka. fore ‘raw, fresh" + GzTe. foray, Tna.S1.W1.2, fare, Har, fini, Sod. tari. Common Semitic fry. wiitdit ‘milk : Possibly related are Gaf. af"atii, M. (tat, Go. dgtat, Sod. agat. ‘The source of these and the Amharic item cannot be identified. Most of the other Ethiopian Semitic languages have forms derived from the root filb, cf. Ath. ayb ‘cheese, curdled milk"; Tigrinya, however, has an item of Agaw origin, aba. wit 'stew" 2 Arg. witsh, Har. witti, Gaf. wisi, Ch. wiit, etc. This is derived from the Eth.Sen. root wh 'pour' (Gz. 2ausaha) of Semitic origin, cf. Heb. wasaq (2). ziiyt ‘oil' + Gz.Har, ziyt, Tha. ziti. Both Néldeke (1910:42) and Leslau (19574:235) regard this as a loan, the one from Aramaic, the other from Arabic. Could it not equally well be an inherited Semitic item? 2.8. THE HOUSE, agan "fence" 2 Gz. fagon, Tha, hagua, Har. fugun, Arg. hangino, Gat. agar, ete. Common Semitic hgr. bina "gate, door’ : Tha. biinné ‘passage’, Te. bin ‘outside’, Har. bari ‘gate’, Arg.Gaf. bin, etc. Semitic, cf. ESA. bx 'gate', Ak. baxau ‘explore’. bet "house" + Gz.Tha.Te.ch. bet, Arg. bed. Common Semitic byt. Several other $.Eth. languages use a different tem, which is probably also of Semitic origin: cf. Har. gat (Sem. gut). daj ‘gate, entrance’ : G2. dede; probably also Tha, d(33) to’. Gz. dede is usually explained as being related to the common Semitic root de-t (Heb. defet, ete.). \é by psewlocorrection of MLS, 721977] A Comparative Approach to Amharic Lexicon a godgodda 'wat1' ?* 2 Gz. giidgiid ~ gidiigidd, tna. gédgidda, Har. gidiigdd ~ digidag, 2. gadgadda. Probably of Cushitic (East Cushitic) origin, cf. Galla géxgida 'mud-wall, flat-roofed house’, Son. gidtar, gideangid, Kanbatta goagadda. gogo ‘grass hut! + Har, g6890, Arg. gonto, Gaf. gogo, etc. Of East Cushitic origin, cf, Galla, Son., Sid. goo. bediddéin "thatch" (vb) + Gz.Tha, kéidiind "cover', Te. kiidna, Har. xédéina 'thatch', Gh. xatindm, etc. ‘Throughout S; Ethiopian this item has been specialized to mean ‘cover a roof with grass’, i.e. "thatch'. Semitic, cf. Akk. kaddnu 'protect', Datina Ar. mukdana ‘oven cover’. hog "loft : Har. BBE, Gy.Sod.Go. kata, Ch. rata, etc. Of Sidamo origin, cf. Tenbaro bofa. modaija. ‘hearth’ 2 Arg, gndo§¥a, Har. afdéga, WL. médagQ@, Ch. magaSa, Co. mabaiga, eve. ‘This $.Eth. item is almost certainly an instrumental noun from the common Eth.Sem. root ndd ‘oun’, i.e. *mondo/adya > modagga, etc. This root ndd has no certain cognates in the rest of Semitic. However, the root shape is in accordance with Senitic pattems, as is the nominal derivative *mandad-ya, etc. mindin ‘village’ 2 Gh, miidiin 'place'. Praetorius (1879:28) first explained this asa nominal derivative of the root idx ‘spend the night", i.e. mahdir (Gz. “dwelling place'), which is almost certainly right. The first stage in the development to Anh. mindéx would be *madit, which could result in méindéa after the insertion of the "intrusive" nasal; for other examples where -aC- becomes -énC~ (and not -anC-) cf. winz ‘river’ from *uaz (root whiz), zandaao "last year’ from *za+dato, pethaps. On the other hand, one is intrigued to notice the large number of modern South Arabian and dialect Arabic forms with an almost identical shape to Anh. méndéx and not incompatible ranges of meaning: Soq. béndher ‘port’, Nehri mandéx, Hadrami Ar. bandar, nani Ar. bender, etc. (Leslau 1938:89). maséiso ‘centre pole Probably from a root 48, i.e. *ma2sds-o, to which Ar. *assasa ‘found’, Syr. °abiyata Ycolum', Heb. *aXeya, and Akk. asztu ‘tower’ may be compared. 'Especially a wall made of wattle and daub, MAL 5, 7332 D. L. Appleyard [AAL 5/2 3. THE SEMANTIC FIELD “THE NATURAL ENVIRONMENT” The subdivisions under this heading are 3.1) natural phenomena, 3.2) flora, and 3.3) fama. In this semantic field there are necessarily itens of particular local occurrence, as well as more universal itens. The naves of locally restricted items, especially plant and animal nanes, could reasonably be expected to be of non-Somitic origin, borroned fron the indigenous population wherever these items were unknown and new to the’ incoming Semitic speakers. Nevertheless, this is not necessarily always the case; an existing lexical item may have its semantic range extended or modified to encompass the new object. This is the ase with Eth.Sen; (Gz.) 20% ‘hyena’, the cognates of which in other Senitic languages mean ‘wolf! or ‘jackal’. Altematively, existing morphs may be used to create a new form to describe a new object, as Anh. kgdone 'giraffe' (Eth.Sem, root bfn "be thin"), Gz. -anwe axis > Anh. auaris 'rhinoceros'’ (lit. "rough-skinned beast"), and so on. However, the najority of names of specifically Ethiopian (or African) plants and animals in Amharic is of non-Seniite origin, whereas general terms like aute ‘wild aninal', wig ‘bird’, hong ‘wing’, féind ‘horn’, etc, are of inherited Semitic origin. Similarly, the nancs of many creatures that must already have been faniliar to Semitic speakers from the other side of ‘he Red Sea are, Semitic: fab ‘hyena! (but |wol{" or ‘Jackal in Asiatic Semitic), anbisa ‘ion’, nabst "Leopard", obab ‘snake’, nob 'bee', zamb ‘fly’, etc. This is essentially the sane pattern as was discussed above in the field’of the donestic environment, nanely that the general terms are Senitic, but the nanes of specific objects, plants and aninals, are often non-Senitic. Out of some 42 aninal names studied here, 19” (approximately 458) ‘are of non-Semitic origin. Of these nineteen, most have comates only in S.:thiopian language: anoxa "bird of prey", azo ‘crocodile’, zalion ‘elephant’, héthdnno ‘wild pig’, kuna ‘crow’, Zongdi0 ‘baboon’. “Those common to North and South Ethiopian, including Ge’ez, are asa ‘fish', Sogta "guinea fowl", sdgano ‘ostrich", and 5 "Spider"; perhaps also hole 'par- tridge!, Occurring throughout modern Ethiopian Senitic, but not recorded in Ge'ez, is gunare *hippopotamis'. Amongst the twenty-three items of Semitic origin, of special interest are Anh. ay ‘mouse’, abab 'snake', and tue ‘worm’. The last two, vbab and taf, have cognates throughout S. Ethiopian but not in N. Ethiopian, where a Variety of items of different origins occur. Anh. ayf has only one Eth.Semitic cognate in Arg. hent, whilst a different Semitic term occurs in the rest of S. Ethiopian (liar. {u°ut~ sux, etc.) and N. Ethiopian employs a non-Senitic tem (Gz. *ansewt, etc.). In these three instances, therefore, S. Ethiopian preserves a Senitic root not found in N. Ethiopian. Two tems for apparently different kinds of (wild) pig, asana and swa, are both formed from inherited Semitic roots (Sm "be hateful" and jw 'dig', respectively), whilst the common Semitic term (n}zt does not occur in modern Ethiopian Semitic, This would appear, pethaps, to reflect the substitution of the name of a taboo animal by a descriptive term, as in the fanous case of the name of the bear in sone Indo-European languages. The taboo status of the pig is, of course, a well known Semitic trait (Honmel 1879:319). It is possible, too, that the use of Cushitic terms for 'fish' throughout Ethiopian Semitic (Gz. ala, etc., Har. tulim, etc.) and the absence of any inherited Semitic term here might be due to a Similar cause, the taboo in this case being a Cushitic feature. Amongst the few specific plant names studied here,'? the proportion of non-Semitic terns to Semitic is greater than anongst animal nanes. ‘Much of the flora of the Ethiopian high- lands would perhaps be new to Semitic speakers and consequently, as the lexicon here sug- gests, the terms for these new plants were readily taken over and adapted from indigenous 171 have dealt here only with a small number of plant names, mostly trees. This is not the place to present a detailed etymological dictionary of flora, or, indeed, anything else. Only those itens that are widespread, prominent, or significant in cultural terms have been discussed. AAL 5, 74197] A Comparative Approach to Amharic Lexicon 33 Janguages: exx.:, gotar ‘acacia’, f4ek"EL ‘euphorbia', banka ‘banboo', Sota ‘fig tree’, wiyna “olive tree'. A glance at a list of plant nanes common to most other Semitic lan- guages (see Fronzaroli 1968:267-303) shows an alnost total absence of Ethiopian cognates. This is because many of the principal Asiatic Semitic languages are, in the most general terms, native to a fairly consistent ecological area, or group of ateas; the plants typical Of this region are, however, not native to the Ethiopia highlands: ex. ‘oak’ (Hleb. eatton), "terebinth' (Ar. bugm), ‘willow’ (Ar. (2a), 'tanerisk’ (iieb. 2p3eg), ‘date palm’ (Ar. tant), and so on. ‘The only typical, native tree’of the Ethiopian highlands with a probable Semitic nane is the ‘juniper’, tzd. On the other hand, as was the case with ani- mal names, genetal terms are nearly ali Semitic: exx. abiba ‘flower’, {ore ‘fruit’, sar ‘grass’, upét 'wood', bosie 'leaf', vf ‘bark’, etc. A notable exception in Anharic is zag ‘tree’, which is of Agaw origin. The Semitic term survives in nost Ethiopian Semitic languages both in the sense of ‘tree’ and ‘wood’, but is restricted in Anharic to the latter sense (unfit), a feature which is shared with Harari, Argobba, and Gafat. In the field of natural phenomena, i.e. geophysical, meteorological, and astronomical terms, the proportion of Semitic terms is slightly higher than in the other two subdivisions of the field of the natural environment; at a conservative estinate (that is, comting only those items of safe Semitic etymology), 34 out of a total of 57 (approxinately 608) are of Semitic origin, The non-Semitic iteys’ gre not contained within a particular semantic category, but range from items like ¢axdka 'moon' to woha ‘water’, from Faka "mud" to diiga ‘highlands’, and so on. Only one of these thirteen non-Senitic items is conmon to all of Ethiopian Semitic, nanely déimina ‘cloud’. Conon to S. Ethiopian only are Sétdka ‘moon’, kin "day", “and woha 'water', though in the case of all three the original Semitic terms do Survive in some S. Ethiopian languages: cf. wit ‘month’ in Anharic, Har. ma?altu ~ mattu ‘day’ (Gz. masate, we), and iar. mz water’, The remaining non-Semitic’ itens are typically of restricted, local occurrence, like faka’'mal’ found in Tigrinya, Argobba, Gafat, and Miskan, or 'dingwya~ dangay 'stone' occurring in the sane languages plus Muhor, or £28 ~ Gos" 'smoke' found in N. Ethiopian, Anharic and Harari. What conclusions, if any, can be dravn from the presence of non-Semitic items in the field of natural phenotena? Obviously the explanation, that served for plant and animal names is not appropriate here. Items like 'noon’, ‘water’, ‘day’, ‘stone’, etc., are the kind of vocabulary that can be justifiably classified as "basic", in accordance vith the principals outlined at the beginning of this paper. Insofar as "basic" vocabulary tends to be more conservative than other areas of the lexicon, the occurrence of these originally loan items here could be said to reflect the degree to which non-Semitic and Semitic speaking peoples, ie. their languages, have fused in the development of Anharic, The majority of these non” Semitic items is of Agaw origin: tis ~ §28 ‘smoke’, diga ‘highlands’, diinmdna ‘cloud’, aya, 'stone", baeéx "lowlands", woha ‘water’ (probably), whilst others like foka ‘mud’, g ‘moon', and kin 'day' are nore difficult to attribute to a particular Cushitic lan- guage or language group. 3.1. NATURAL PHENOMENA agin ‘soil, dust’ : Har.Ch. afin, Ga. afird, etc. Common Semitic ‘pt ~ ‘atin "world" 2 Gz.Tna. ‘agin, Conon Semitic £m. AALS, 75u D. L. Appleyard [AAL 5/2 anba '£lat-topped mountain’ : Tha, *anba~ *anba, Caf. anba, ambatagé ‘mountain’. Of Agaw origin: cf, Bil.Kem, anba, Khm. aba, S.Ag. ambe. amid 'ashes' : Gz.Tna.Te-Har. hamid, Arg. hamid, Ch. amid, End. auind, etc. ‘The root fmd occurs throughout Ethiopian Semitic and may be compared with two Arabic roots despite the disparity of the initial laryngal: Ar, hnd ‘cool off, die down (of fire)’ and tynd’ "be extinguished’. amiday 'hoar-frost' : aniida, na. hamiday Yusty atmospheric condition’, Te. finds “hoat-frost's Har, hanacly, Gat, imddags probably also Ch. andidax 'cold', etc. This is almost certainly an extension of the root find seen in the previous item. aniit 'year' 1 Gz.Tha. ‘amit, Har.Sod, amit, etc. Conmon Semitic *m. aXéwa 'sand" 2 Tha, faSéiwa; also Gz. hoga, Tha.Te. foga, Te. haghag. The form of the Aharic and the first Tigrinya item suggests that this might be taken from Agaw: Kem.Khn, aSiiia, Quara alaua. The other Eth. Sem. forns are more likely of inherited Semitic origin: cf. eb. hagas 'gravel', Ar. hasan, Akk. higgu, Soq. hdsahdshin. The ‘Agaw foms may in tum be either taken from Ethiopian Semitic, ox may be independent flexes of a common Afroasiatic item: cf. Bil. quba, Beja haak, Afar *iddé (2), Som. ‘id, Kaffa baado, etc. pat ‘fire’ 1 Gz.Te. 22sat, Har, isdt~ oat, Arg. osad, Ch.Go. osat, Gat. asati, Sod. dsaz, ete. Common Semitic °3{-2). biga ‘dry season’ + Praetorius (1879:140) derived this from *bé-hoga ‘im Somer’ (Gz. bihagay) . Such an etymology does not seem unlikely and is, indeed, supported to some extent by the form of the corresponding verb root in Anharic,’ bala, where the palatalized § indicates an original final radical y and the vowel a of the first syllable could be explained as due to a lost second radical laryngal, i.e. *bhgy, a denominative derivation of béhagay. The ‘comparison made by Leslau (1963:43) with Har. béyog, M.Go. beg ‘rainy season' (de) is surely incorrect. Rather, Anh. big ‘Little rains" should be compared with these. The Toot of bi-hagay, etc., would appear to be Semitic: cf., perhaps, Heb. fig ‘feast day’, Syr. haggd, and perhaps also Ar. /éaja ‘year’. bina "be Light! : Gz.Te. béitha, Tha. birhe, Gaf. bind; the other S. Ethiopian Imguages presuppose & root, form thbs beside bak: ef. Har. bandm, SL. bate, Sod. tannin, etc. Both pattems occur anongst the Semitic cognates: Heb. bahar, Ar. bahona and bantha. AAL 5, 761977} A Comparative Approach to Anharie Lexicon 35 binniidit "be cold" 1 Gz, binidd, Te. bitda, Har. binida, Arg. binriida, etc. Anominal derivative (Gz. brid, Anh. biniido, Ch. bind, etc.) occurs throughout Ethiopian Semitic with the meaning ‘hail’? Conmon Semitic brd. baniha ‘desert! : Tha. binéxa, A close formal cognate occurs in the Bilin toponym biraja~ bara "Barka' (Gz. biiaba) Perhaps the proper noun has given rise to the common noun here, if not vice-veksa, bavniihdé “flash' (lightning) : Gz. banda, Ta. A nominal derivative in the sense of ‘lightning’ occurs throughout Eth.Sem.: cf. Gz. mibade, Tha. bixké, Har. bordk, Z. bavrakat, etc. Common Semitic bag. bar ‘sea, lake" + Gz. bale, Tha. babaé, Te-Har. béhar, Ch.Sod. bar, etc. Semitic, cf. ESA. bit, Ar. bale. foka ‘mud’ + Tha, Foye, Arg.Ms. Eoba, Gaf. Fobi, W.cur. Fokana. This item is almost certainly of Cushitic origin, though the precise source cannot be readily identified: cf. Bil. dinag’a, Khm. xog"a, Quara dax‘a ‘clay’, Afar Aag"a, Son, d6oqo ‘turbid water", Galla doke ~ dofke, etc. A’different version of the sane root (Son. obo ‘md, Sid. obbaa) occurs as a ioan in Har. ${ba ‘mud for building houses’. | possible extension of the sane root occurs in Har. g@ka ‘stink’, and Tha. sdyydxd ‘spoil'. ¥éttiind "be dark" : ‘Tha. ai , gdlma, Har. $Elaima, Arg. fedelina, Gat. pibtin’, &. fhinina,’ etc. Throughout S, Ethiopian this verb follows the B-type pattem, as in Tigrinya, as against A-type in Ge'ez and Tigre. Common Semitic z£m. Sanaa "noon" 2 Har, Sania "moonlight", Arg. li ait, en Pot es a This item occurs only in S. Ethiopian, where two patterns predominate: “finda and *Yinaka. The only convincing likely cognate is’Beja terig ‘month’. “Reinisch quotes a Quara,iten zixkd, but this looks like a loan fron Amharic; the usual Agaw tern for ‘moon’ or 'nonth’ is *atb-a. It is, perhaps, interesting to note that a similar form occurs in certain Sudanic languages: Kunama’teera, Mursi t2agi, etc., both meaning 'noon', though how these fit in with the Cushitic and Ethiopian Seniti¢ itens, if, indeed, they are connected, is not clear (see Dolgopol 'skiy 1973:48-9). The Semitic term wth survives in Amharic and other S. Ethiopian languages in the sense of ‘month’. fos ‘smoke’ + Ga.Tha.Te, fis, Tha. also 728, Har. $08. Of Agaw origin, cf. Bil. tadé, Khm. fiya, Kem. taza. ja "highlands! : Tha. diga, Har. dig ‘elevated ground’, Gaf. diga, Ch. a : digago “high'. AAL 5, 7736 D. L. Appleyard IAAL 5/2 Cushitic, perhaps Agaw: cf. Bil. dag "summit, above’, Khm. dig, Kem. day; note also Afar daga ‘nowtainous region’, Galla daaga ‘escaryment’. dimiina ‘cloud" + Gz.Tna, dinmniina, Tha, also débina, Gat, darminé, Har. dana, Sl. dibina, Sod: dabina, Ch. dabiéita, etc. ‘This item is almost certainly of Cushitic origin, though again it is not possible to identify a particular source. Close formal cognates occur in Bil. dama, Khn. dénena, though these may well be taken from Ethiopian Semitic. Other Cushitic forms are: Galla dumesa, Hadiya duuba, duubanfoo, etc. dingoya~ dongay 'stone' : Tna. ding*itea "rock", Arg. dingay, Gaf. dinga, N. dang”dta, Ms." ding” td Agaw, cf. Bil. ding" sna, Kin. dog ta. ‘The N. Ethiopian languages and all the S. Ethiopian languages, with the exception of Amharic, Argobba, and Gafat, preserve the comon Senitic item *bn. dur ‘forest : Gz, dibs 'nountain', Tha. dibré, Te. dibor, Gaf. dabnit “forest', Ch.Gy.Sod: daban, End. dot, etc. Note that all the N.Eth. forms derive from the pattern dibs and mean ‘mountain’, whilst the S.Eth. forns derive’ from “debt and mean ‘forest’. Semitic dbx. Gissisd ‘flow! 2 Ga. fésdad, “pour out, gush forth’, Tha. géssd "flow", ch. be spilled’. From this root ($4 are derived several nominal forms such as Anh. {isa 'strean', Ch. {ass 'rainstomn'. Semitic, cf. Heb. pia, pled ‘spread out", Ar. gaba, ESA, (4? ‘aquae deductio". gum ‘mist’ : Ch. guna, Go. gum; note also Gz. géme, Tna. gimi, Te. gimat. Praetorius (1879:67) considered these forms to be pseulocorrections for *jym, cf. Ar. gaum, gayn. However, though this is not impossible, there are widespread Cushitic items formally Closer to Ethiopian Senitic gun-/gin-: cf. Beja glim, Sid. gonicéo, Kanbatta gooma; note also Owtic, Kaffa gumoo, Noa gun, Wolam> gunaa, etc. hokiib ‘star" 1 Gz,Te.Arg.Sod.2, kokiib, Tna. koxob, Gaf. kokobii, Ch. xt ab, Gy. x"dx"in, etc. Common’ Semitic kbkb'> kukb. kunt "rainy season" + Gz, kendnt, na. bondimté, Te. hiinim, Har. kiané, Arg. kadmt, Gai. brine, SLWL. kim, Z."eotom,"Ch.Gy. xatan "year, Ms. hotum, etc. Forms with the suffix -t occur in Ge'ez, Tigrinya, Anharic, Argobba, and Gafat; otherwise the patterns *kotm, *kim, and “kdxim occur, ‘The root kum is Semitic, Cf. Ar. kakuna ‘rain, be generous". ratte "Lowlands" 1 Tha. pratea, Te. pabiibhoe, Gaf. pata. AAL 5, 781977] A Comparative Approach to Anharéc Lexicon 37 Guidi (1901: col-254) related this to the verb root (dea, "roast", but ¢ more 2ikely ex Planation is that this is of Agar origin: cf. Bil. qtata ‘river, Valley’, Kim. aq'at, Ken. Peon in ‘day’ + Ang. bling, Gaf.Sod. pind, Ch. bind, MMS. kind, En. Riad, ete. Praetorius (1879:142) tried to derive this from the root fxs 'be straight", i.e. ‘Hochstehen der Sonne", “However, there is a small number of Cushitic forms that may be compared: Bil.Ken. taxa ‘sun’; Kim. k*dia, and perhaps Som. qoraah, Rendille hura, also meaning ‘sun'. “The N. Ethiopian and the other S. Ethiopian languages preserve an iten of Senitic origin: Gz. misatt ( wee). fet ‘night! 1 Gz, Le€ét, Tha, Uiyti, Te, Lali, Har. Layee~ LBLi, Caf. Couanon Semitic £ye(-t). meda 'plaii Praetorius (1879:26) derived this from *méheda ‘Ort wo man geht" (Anh. meda go"), citing an Argobba item hed with ‘the same meaning as Anh. meda. ‘This Argobba term does not, however, appear to be recorded elsewhere. On the other hand, there is a formal resemblance between Ath. meda and Har. m&dan, also meaning 'plain'. This latter item appears to be a loan from Ar. maydan "square, open place’. modat ‘earth! + Ga. mdr, Tha, modri. Semitic, cf. ESA. mdr, Akk. midnu "type of soil’, Aran. medxz 'clod'. mobi 'be warm, hot' + Gz. mobil, Tha. moyi, Te. moka, Har. mdf, Ch.Gy. m'dbéin, Sod. makin, Wi. moke, etc. This comon Ethiopian Semitic root mi has no satisfactory cognates elsewhere in Semitic; Dillmann (1865:col.202) tried to relate it to various Arabic roots, wrk "be red hot", kk and ‘kk 'be hot'(of weather). These Arabic itens are plainly formally remote from the Ethiopian Semitic root. It is not, of course, inconceivable that both mf and wk are ultimately developments of the sane primitive root or root nucleus, but a5 it is not, possible to relate them with any degree of certainty, the Ethiopian’ item mist. remain "unidentified" as to origin. However, it can be said yet again that, although the Ethiopian root lacks adequate cognates in the rest of Semitic, its root’ shape is in accord with Semitic patterns. mang ‘spring’ + Gz. mingg*, Tha. moxSé (a loan from Ath., as the form shows), Arg. mang, "Gal. monba The root is mkt (G2. nipsa 'burst"), to which Ar. naga‘a ‘pour out' may be compared. mi38a "become evening" + Gz. miisyi, Tha, mésiyé, Te. misa, Har. mia, Ch. miSim, etc. Many Languages also have a derived nominal form: Gz. maset, Anh. mobit, Arg, med, Gaf.ch. mobéti, etc., all meaning ‘evening'. Comon Senitic, cf. Ar. mast, AkK. milu night". AAL 8, 7938 D. L. Appleyard [AAL 5/2 mata ‘evening’ 2 Gag. mobet. ‘This is derived from the root but (Gz. etd ‘pass the night!) as, the form ofthe Gafat cognate shows, and not from *4w ‘return’, as suggested by Praétorius (1879:159; see also Cohen 1939:413). Thus *mobet-a > mata. Ge. niiddt mn Gy. niitidé, etc. Sem. languages employ a root elated to Anh. tébagtité of the same mean- ing (Tha. tixapatd, Arg. obbattita, Caf. tikagtia). Both roots, nda and byl, are per- fectly in accord with Semitic root ‘patterns, but lack cognates outside Ethiopian Semitic. ‘na. nididd, Te, nédda, Har. néidida, niigas ~ nofes ‘wind 2 G2, niigas, Tha. nifas~ nogas, Arg.Gaf. nogas, Ch. angas, ete. Comon Semitic np. nigga ‘dawn' (vb) + Gz. nigha, Tha. néighe. ‘The nominal derivative occurs widely throughout Eth.Sem.: (G2. niigh ‘dawn, morning! tig, niga ‘tomorrow’ noget ‘dawn’, Gaf-Arey nag, ttonorroe', Gh. Raghy Ge: moghats ‘damn’, ‘na. nygaho, and so on. Conon Semitic’ ngh niigga ‘freeze, congeal' Gz, aig‘a, Tha. niig?e, Te. nig?a, Har. aiiga?a, Sod. Gxdggan, Ch. nian, E: niélggam, ete. Senitic, cf. Heb. aagat "be benunbed, congealed’. siinay "sky" : Gz.Tna.Gaf. siimay, Te. sina, Arg.Har. séiné, Ch. sine. Common Semitic Smy. The Tigre item, however, looks more like a loan from Ar. samt? than an inherited Ethiopian Semitic form, tinara ‘mountain’ ‘There appear to be no formal cognates of this item in the rest of Ethiopian Semitic; a variety of forms occur elsewhere, the majority of which is probably of non-Semitic origi Gz. débr (cf. dur, above), Tna. ba (cf. anba, above), Har. sri, Gaf. séged, Ch. f° Arg. gubba, etc. ‘One wonders whether Ath. tiata might’be connected with the Conon Semitic root 222: cf. Ar, tate ‘hill", Aram, £2224 "high", Heb. tel, tH0UL, Akk. titeu. fibba dawn" (vb) : Gz. gdbha, Tha. gabhe, Gat. gaibba. ‘The nominal derivative occurs throughout Eth.Sen.: cf. Gz. gabalt, Anh. pont, Arg. foun, Go. tobi, 2. fabba, etc. Common Semitic sbf. fil ~ fot "dew! 1 Gz. #H€2, Gaf. agal, Te. tila 'be wet’. Semitic, cf. Heb. fat. AALS, 801977} A Comparative Approach to Amharic Lexicon 39 ‘fola ‘shade, shadow" 2 Ge, $olalot, Tna.Te. solat, Har. fya, Arg. fata, Caf. Feta, M. fote, Chicy. forar, SL. gat, etc. ‘The Eth.Sen. root is ££, of which the final radical £ is lost in Anharic (and Argobba) through palatalization: "£>y >. Common Semitic 720. fay ‘sun! : G2. zihay, Tha. géliay, Arg. Fahed, Gy. Sayit, Ch. Sot, etc. cherie. is ts caly 5, Bektopian Lengnge with vie vect Zam, without the muPflx -t. Sonera other S. Ethiopian languages use different roots: cf. Gaf. aymond, Sod. yimot (cf. G2. Panin); Har. Zt, S1.WL. aye, Z. antt, Go. aret (cf. Ge, *ex) (see Blazikowsky-Brauner 1913: 10). The root zy is Semitic, cf. AP. quha” forenoon, dalwa dden, dahiya "be struck by the rays of the sun’ waha 'water" : Arg. dha, Gaf. digi, Ch.E. oxa, End. oft> Go.Ma. aga, Sod. Aga, etc; Old Anh. also witha, If the Anharic form represents a metathesis of *ofi"a, dia, then it can be related directly to the other S.Eth. forms and thence to Agaw: cf. Bil. *aq*, Kh. aq’, Kem. ax", S.Ag. aju. Alternatively, the Anharic item (but not the other S, Ethiopian terms)’nay be taken’ from Sidamo: cf. fladiya wo°0, Kanbatta wu°a, Sid. wa, wake. ‘The N. Ethiopian languages and a few S. Ethiopian ones preserve the comon Semitic root! Gz.Tna.Te. may, Har. miy~ mi, SL.WL: miy, 2. may. wine ‘river’ 2 Gz.Tna.Te. wofiz 'strean’. ‘The root patterning underlying the Anharic item must be *wihaz or *uahoz; thence ‘waz > winz, The Eth.Sen, root whiz is almost certainly Semitic and represents a variant development of the same primitive root form as seen in Eth.Sem, uz® ‘sweat’. wt, ‘month Gz. wink "moon, month’, Tha. winhé, Te. winah, Har. wiithé ~ wij "month',’Go.Sod. wind, Sl. wind, Z. wir, etc. Conon Semitic wh. waiSa 'cave" + Gat, wakkd, Arg.Sod. waSSa, etc. Perhaps of Agaw origin, cf, S.Ag. wali, Quara, Kem. wala, zanniba "rain' (vb) + Gz. ztinnii, Tha, zindind ~ zinibi, Te. ziitma, Har. ziitéima, SL.WL. zabimi,” Ch. zindbéim, etc? ‘The nominal derivative also occurs widely throughout Eth. Sem. zonab, Te. zatam, Har. zon, Gat. zonaba, Ch. zonab, e Gz. zonam, Tha.Anh. zanam ~ Common Semitic dam, drm. 3.2. FLORA abiba ‘flower’ : Tha. Sanbaba, Te. ‘amboba, Gaf.Arg.Go.Sod. abiiba , 2. anbiiba . Comon Senitic: cf. Heb. *eb "bud", 2abZb ‘ripening ears of grain, harvest time’, Jewish Aran. *bba~ nba "fruit', Ak. inbu 'bul'. Forms very similar to the Eth.Sen. "items occur in several Cushitic languages: exx. Bil. anboba, Saho-Afar anbaab, Natéa Gaila ababo. These may be independent Cushitic reflexes of a shared’ Afroasiatic root, or they may be owed to Ethiopian Semitic influence. AAL 5, 81.40 D. L. Appleyard [AAL 5/2, anig ‘vine, clinbing plant' : Gz. fardg 'vino', Tna.Te. hardg ‘climbing plant’. One wonders whether this Eth.Sem, root vig is connected with Ar. Jj and Heb. lug "cone out", or, indeed, with Eth. ‘ag ‘climb". The sane root probably also occurs in Anh. anang"ade ‘green’ things, green’, Old Anh. haniigade, haxiing" ade, if Praetorius’ (1879:93) derivation from *hanig + g* inad’ (cf. Gz. g* ond, Ani. gand'log'} is correct. anit "wood" + Gz. fog "tree, wood", Tha, tongiiyeé ~ eonfiyté, Te. “okay, Har. ingi 'wood", Arg. anked, Gat. anfa. h a8a, 2. snpet, ete Common Senitic *¢. shoh~ Soh ‘thorn’ Gz. Sok, Tna. *s8ox, Te. Sokit, Har. usux, Arg. ofoh, Sod. sok, Ch.Gy. Sox, etc. Comnon Semitic swe. gare "fruit! : Gz.Tna.Te.Arg. gare, Har. guri, Gaf. ford, Common Semitic pry. gond "log, trunk! + Gz. gYand, Tha, g*andi, Te. gonday, Ch. gond. Dillgann, (1865:Col. 1189) related this item to the Eth. root gnd ‘cut, cut off", citing Ar. Gd "be solid, hard’ as a closer senantic cognate.” A regular got pattern nominal derivative of gnd’does, however, occur in Ge'ez: gamd "Pars abscissa, Segmentun'. OF course, g”ynd could ultimately be from the same root, but with medial 'm > n conditioned by the close juncture with d regularized and spread to other forms (g* onnud); *gund > g* ond (root established as g'md) beside gn: regular qztl derivative gad. gotan ‘acacia’ + Ch. genat, Ulbarag guranan. Perhaps from Som. galdo¢, or some similar form. oso ‘anthelmintic plant’ : Tha.Te. koso, Sod. habed. Perhaps from an Qnotic source, cf. Siria8a kosbo, Mota hd-ko, Kaffa kaks. praeptak~ b* oth"at ‘euphorbia' : Tha. p'defat, Te. foonbat, Go. kuth"at (2), Ulbarag otk" ae. OF Agaw origin, cf. Bil. q*a€ang*ata, Ken. ka bénkitha~ orbiha 'banboo' : Tha. farkah, lar. bidrkaha. OF Agaw origin, cf. Kem. karixaxa, S.Ag. genqadi. it "leaf" : «brag, Tha. b'ageé, Har. hutti, Arg. kotal, Caf. ie MZ. ote, Ch. basi, Semitic, cf. Ar. qagata 'mow', qasié ‘orge vert qu'on donne aux chevaux" AAL 5, 82,1977) A Comparative Approach to Amiaric Lexccon a Linkin "green, fertile' : Gz. Tna, Léintiim, Gat. Lament a. ‘This root £mén is probably to be connected with Eth.Sem. fm ‘prosper’, nji idem, and at the sane tine with Ime "grow green’. These are all extensions and developments of the same primitive root also found in Ar. Lanaha ‘flash, sparkle", hant "sauctus arboris'. Lat "bark" 2 Gz. Loh, Tha, Lohpd, Te. Loftas, S.Ang. Lihingo. Semitic, cf. Ar. Labhasa 'squeeze', Heb. alias. sar ‘grass! : Gz. ath, Tha, sacri, Te. sitar, Har. dar ~ stn, Arg. s0t, Z. S51, Gaf.Sod-W1. sar, Ch.Gy. sda, etc. Connon Semitic 32, sot 'root' (see above under sat 'nerve') Sota ‘fig tree’ : Har. sdbla, Arg. Sola, Sod.Go. sobla, WL. sobld, Ch.Gy. Sabra, M. Saboya, etc:; probably also to be compared here are Gz.Tha.Te. sdgla. Origin unidentified. fod "juniper" : Gz. gohd~ gohd, Tha. gohdé, Gaf. $odd, SL.NL. tid, 2. tala, Ch. dit, Ms.Go. dad, etc. Almost certainly Senitic and perhaps to be compared with Ar. gu‘d ‘height’. wiyna ~ wera ‘olive tree’ : Te. wigne, Har.S1.W1.2, wiigit, Ch. unyna, Go. wera. Gushitic, cf. Bil. uiixand, Kem. uiiyna, Hadiya weraa. Som. wiigat. The development g > y is a particularly Agaw phenortenon and its presence in the Amharic item vis-a-vis Tigre, Harari, etc., suggests particular Agaw influence in the development of the Atharic item. zag ‘tree’ : Har, 28 "large tree", Arg. zag, Gaf. za"a. Of Agaw origin, cf. Khm. zig, Kem. zag, Quara Jaga. 3.3. FAUNA abiibt ~ abobt~ okt "leech" » Tha. Satipté, Te, Satie, Har. Opti, Sod. cy. andbot, Wh. abit, etc. Common Semitic *Zq-t. anona "bird of prey’ | Samora, Arg. amona, Ch.SL.NL. anita, 2. anata, M.Co. ant ina, ete. Unidentified. anbisa "Lion' 2 Gz. Sanbiisa~ *anbiisa, Tha. *anbésa, S.Arg. hanbassa, AAL 5, 85a2 D. L. Appleyard [AL 5/2 Senitic, cf. Ar. Sanbas ~ abbas, The other Eth.Sem. languages enploy a variety of terns: Te. hayat, tar. winag, Caf. zébbd, Ch. Zap, ete. ¥ anbiifa "locust" + Gena, *anbiifa, Te. ‘ambifa, Gaf. onbiit , Arg. anbiita. ‘This probably derived from the Semitic root nbf ‘come out, emanate’. asa 'fish" Gz. ‘aa, Tha.Te. ‘asa, Arg.Ch.Gy.Sod. asa, Gaf. asd, etc. Gushitic, perhaps Agaw: cf. Bil. ‘aza, Kem, asa, S.Ag. asi; but note also Saho Saasa and Beja aba. sana 'pig’ + Tna. hastina, Te. hasana. This item is to be related to the Eth. root (6m (G2. faSiini "be hateful, bad"), to which Ar. abina 'be angry’, haSana ‘say unpleasant things’ nay be compared. "the pig is, of course, a taboo aninai in traditional Ethiopian culture, as anongst other Senitic speaking peoples. qwre ‘wild animal’ + Gz. ?wwe, Tha. 2arqwit (formally a plural, cf. G2. *arawit), Te. *auce’'snake', Gaf.Arg. ame, Har. Tad. From the same root derives Anh; ata ‘male (animal), chief, main’. Common Semitic *au(-y): cf. Ar. *auit, Heb. °4az Lion’, Akk. cai ‘eagle’, étc. The same iten, 2aule, also occurs in the compound noun Gz. “awe aris’ (Cosmas anout arisi) 'rhinoceros', Anh. aunaris. ayt ‘mouse’ 2 Arg. hong, Old An. hays. Semitic, cf. Akk. afasu 'weasel'. The N.Ethiopian languages employ an item of Agaw origin: Gz. Pangema, etc. (Cf. Bil. *onfame, Khm. seowa, Kem. ondou, S.Ag. one), whilst the rest of S.Ethiopian has an item of a different Senitic origin: lar. gueur~ fix, etc. (cf. Ar. fa°n). azzo ‘crocodile! + MiMs.Go. azzo, Old Anh. hazzo, etc. Note also Har. is, which, though doubtless related, is probably from a different imediate source. Cushitic,’cf. Kem. azo, S.Ag. azzu; cf. also Som. yaidas. bab ‘snake" 1 Arg. fiowaw~ how, Har. fubab, Gaf. abab"é, WL. ambab, O1d Anh. habab. Ndldeke (1910:89) regardod this as a derivative of the root (iob ‘love’ as an example of Gegensént., However, the root fibb occurs in Tigre with the meaning ‘wind, curve’; cf. also Ar. fubab’ ‘serpent "? g6uitit "viper" + Har, é§Géii, SL. 066aitna, WL. unfaiie’, 2. mpétivé. One wonders whether this S. Ethiopian item is related to the N. Ethiopian and conmon Semitic root meaning 'viper': Gz. °a4®ot, Ar. °a4*@, Heb. *epte, etc. The underlying form common to the S. Ethiopian items is’*ha§gon-; perhaps an extension in -n of a variant of the root °¢° ? AAL 5, 841977] A Comparative Approach to Amharic Lexécon 43 onkurarit "frog" f Riduiindcat ~ pdananotat ~ deat ta bone SL.WL.2. onbunarit. ALL these forms are ultimately extensions and reduplications of the root *k*a® ~ bth. This is clearly partially onomatopoeic, but can nevertheless be compared with other Semitic tems: Ar. quad ~ gina ~ qaraa’ ‘frog’, Talmudic Heb. qiingix ‘the croaking of frogs’. ‘inndnaat, Ta. tanto Ponkono®, Har. ankuraralti, onbotakit "Lizard" + Ch. onanet~ onanondt, E. on¥oronyit; perhaps also Har. abliinya n abhiya, 2. astyya, End. udznni, Gy. wobsni, etc. ‘The pattern of the Auharic item on+C)oCgaCz+it (/E2) occurs in this and similar forms through- cut Ethiopian Senitic in anigal nanes; Gf. Anh. ‘onfuranit ‘frog’, Arg. ankonarct ‘spider’, WL. ungaiiet "viper", Ch. ondonondit "lizard". “The root of this item (*5(h)£ ?) cannot be readily identified as to origin. ongab "dove! + Gz. Aagb, Tha, ragbé, Te. xagb, Har. etgib. Semitic, cf. Ak. Aigabu~ rigapu ‘sp. of bird’. anya "pig! 2 Ge, handioya ~ harauya, Te. harauya, Har. hariyya, Sl. dniiya, Old Ah. Irotya? ‘The S, Ethiopian forms may be derived from *ha/ saya > fa/anayya. The root is ow (G2. handui ‘dig'). Connon Semitic, cf. Ar. hata, Heb. hin. fotat ‘havik Har. f20é, 2. fututte. Of Bast Cushitic origin: cf. Sid. fublitee, Qabena fitcitita, Kanbatta fiecteitiza, Galla fugwlle; the Kenant itén t.0éy looks nore Like an old loan fron Atharic than an in- herited Agaw item, daleoyye 'duck" Har, From Galla dakiya. kiya, MS. dakkiyéi, Sod. dakiyye, etc. (tit ~ goegae ‘mole’ 2 Arg. gotgal, Har. (744, Caf. goegied, SL. gaegae. Probably from the root 4062 (G2. (de4iité "gush forth’, Anh. (iLiggied 'dehusk, bite (of insects)'), to which Semitic ptt may be compared: cf. Ar. gatla ‘break’, Heb. palal ‘cut, rend". and’particularly ESA. 4l¢ (n) "ditch, excavation’. Interestingly, the iten {dLgi does occur in Ge'ez, but with the meaning ‘elephant’. This latter item is probably to be connected with Senitic pz-, etc. gunaxe "hippopotamus" 2 Tha. gunare, Te. gunare 'thinoceros", Har. gundvre *hippopotamis', Ch. gomana, Go. gomani. Cushitic, cf. Afar guntarti, Saho gunfiaee, Som. §éer; also in Agav, Khm. gunéxi, Ken. guna. go’ "buffalo" : Har, 96%, Arg. go3, Ch.Gy. gi, Sod. gii88, S1. gio, etc. Exact formal cognates for this item occur in Sid. gooke, Khm. guia, Kem. gol. These may, AALS, 854 D. L. Appleyard IAAL 5/2 however, be taken from Ethiopian Semitic, especially since another term occurs more widely throughout Cushitic: Beja agabaa, Bil. 'kabga, Hadiya kobiaa, etc. On the other hand, it is hard not to associate gol, etc., with Gz. gam of the sano meaning. This latter ite is a loan through Ar. Janis from Persian gaunt ~ gintl ~ gains. ub "hyena! Gz. zo*b, Tha. za*bi, S.Arg. Su. Connon Semitic db ‘wolf, jackal'. The other Ethiopian Senitic languages have a variet tems, probably of non-Sénitic origin: Te. kay, Har. wiitba, Ga. kansind, Ch. gata, 30 on. of Gagne ‘guinea fowl’ Ge, zigha, Tha. zagna, Har, zigna~ zibra, Ch. fogra, ete. O£ Agaw origin: cf. Bil. Jdgrina, Kem. Jindna, S.Ag. zagrana. Sorat ‘tail! Probably to be connected with the iten,fona 'fly-whisk", several of the cognates of which hhave the meaning ‘tail’ (Ina.Arg.Gaf. fora). Both itens are of Cushitic origin: for Sorat cl. Ken. gindy~ Soriy, Kh. gona; for Zona cf. alla fina, Qubena giraa, S.Ag. Siri. ‘The N. Ethiopian languages preserve the ‘common Semitic term dib:” Gz.Te, and Hanasen Tha, zandb. bong ‘wing! + Gz. kong, Tha. kongé, Ch. kiinga, otc. Conmon Semitic knp. biinhinno ‘wild pig" + Har. katkarro, Arg. kaxkaro, Gat. hitahiin. East Cushitic, cf. Galla, Som. karkano. bob ‘partridge’ + Tna. kokah, Arg.ch. bob, Gaf. kub™a3¥i, Sod. fubié, Si. foke, Wi. for, etc. ‘This item is also recorded in Ge'ez as gokah, as in Tigrinya, but it is not certain to what extent this is an original Ge'ez item of an introduction from sone vernacular. The item is of Cushitic origin: cf. Khn. qoqaya, Kem. kawoya, Qabena kook’; perhaps also Afar ukkaa*e "guinea fowl’. feemat "louse! 2 G2.Tha, et mag, Te. homie, Har. bundy, Arg. komat, M. ume, CheGy. bimar, etc. Comon Semitic gne~ gin, funoé$a, Wl. kinatfo, Ch.Gy. kona’, etc. ‘The root is Eth. b'ng.~ bng (Gz. bindisd, binnigd “leap, junp'). Semitic, cf. Ar. qanasa ‘spring, jump, gallop". “The medial radical n of the Ethiopian Semitic root as against m of the Arabic’may be explained as having arisen through assimilation with the following $ in sone such environment as *qumy > bang. fun.fSe ‘flea! 1 Gz. tong, Tna. bY ongi, Te. hag ~,biis, Har. fund, Are. AAL 5, 801977) A Comparative Approach to Amharic Lexécon 45 fend "horn" 2 G2. pitt, Tha, hiané, Te. bite (pl. 2akanndit), Var. bir, Arg. bind, Gat.” kindd, Ch.ly. kan, ete. Common Semitic qrn. ura ‘crow + Arg. kuta, Gaf. huni, Har. kunaa, Ch. bY 2rd, M. hurd, etc. Praetorius (1879:67) tried to derive this from the Semitic root gnb (Ar. guxab, etc.). There are, however, widespread Cushitic items which provide a better formal fit: Bil. braxtaa, Kem. x*@udy, Maba ura, Tenbaro kira; note also Kaffa huneBéo, nub "bee" + Gz. nohb, Tha. nohbé, Te. nohob, Arg-Ch.Sod. nob, Gaf. nab", End. naa, etc: Other Semitic languages have the root nilb-; perhaps the medial h of the Ethiopian Semitic root can be explained as due to contamination with another root: cf. nhb (Ar. nakaba ‘plunder, move rapidly"). niibon “leopard! + Gz. ndimt, Tha. nébré, ChM. nébar. Common Semitic nm. sigon ‘ostrich’ 1 Gz. séigino, Ma.Te. sdgiin. Of Agaw origin, cf. Bil. sagan, Kim. sag” ana, Quara sagand. ‘sina "animal leg" 1 Gz. asik"dna "heel, horse's hoof", Tha. séx"éna, Har. Sdxana ‘shin’, Ch. sanxara, etc. Cushitic, probably specifically Agaw, cf. Bil. six"ana, Khm. sox" ona, Quara sukand; note also Beja sit"ina, Afar dit" a®oonoo. 3andnit 'spider' 1 Gz. Saret, Tna.Te. saret, Arg. undgrarit, Har. aikirirahti, Sod. ‘Sarayit, ch.60. Set, Si. o8ée, wl. anbananit, etc. ‘he underlying root of all the Eth.Sem. forms may be reconstructed as *Sn(y}, with internal reduplication in Anharic and other S. Ethiopian, languages to *suuly}. Cerulli (1936:257), relates this to a conmon Cushitic root meaning 'to wear, clothe’, derivatives of which occur in several Cushitic languages in the sense of ‘spider’: Bil. sarino, Kem. sari, Galla sarariti, Saho saxoo-baahayta 'il portatore di veste'. ug "tapenom' 2 Har. sufi, Sod, sof, Ch. sonfa, WL. sefo, etc. Of East Cushitic origin, cf. Hadiya suufo, Darasa heotd, Galla heto. tok"an "bed bug’ : Tha. toxtan, Te. tokan, Har. tuxdin, Arg. tuhan, Ch. toxan, etc. Gushitic, cf. Bil. taxana, Afar té*an, Galla tukana. MAL 5, 8740 D. L. Appleyard. [AAL 5/2 ot "worn" + Har. tutu*, Arg. tulir, Sod. tati, 2. tuk, Ch.Gy. Sorit, Go. Buti, M. diya, etc? Conmon Senitic t(w)£*-t (Hed. GLes%, Ak. tiiltu, Soq. ta°ateh, etc.); the Eth.Sem. forms (S. Ethiopian only) all derive from a pattern *¢2lo*, fota ‘Vervet monkey’ From Agaw: cf. Khn, EdEwtl, Ken. Sofdioa, § "bird! 1 GasTna.tTe, of, Har. af, Arg. wos, Gat. yosta, Ch.Gy. 4, Sod. of, SL.WZ. a, etc. Coumon Semitic ‘up. wata 'sp. of mountain antelope’ : Gz. wiSta~ wiSaga. Common Semitic we "mountain goat’. ton ~ zohon ‘elephant’ : Arg.Sgd._zithon, M. ziixond, Ch.Gy. zix*énd, End.Go. zihondi, Enn. Zix"ind, etc. In Har. doxon, SL. dihano, and W1. déhéno an item from a related, but different source ‘occurs. ‘The foms in initial z- can be derived from something 1ike *2Ak"dn, which bears no glose resemblance to any recorded Cushitic form, though almost identical to’ the reconstructed “common Cushitic” form (Cerulli 1938:198; Dolgopol'skiy 1973:107). The,forms in initial d-, however, are very similar to several East Cushitic forns: Hadiya daneé%o, Son. dagon, Saho dakaano, etc.; from the sane Cushitic root are Bil.Kem. gana. It isnot possible, therefore, to ascribe the Amharic item to a particular Cushitic source. The N. Ethiopian Janguages’all use a form haxnaz, and Ge'ez has a further fom, nége, seeningly of Indian origin. zanb ‘fly! : Tna.Har. zombé, Gaf. zombé, Arg.Ch. zomb, otc. Conon Semitic dbb. Ge'ez does not preserve this Senitic root, but employs an item of Agaw origin: gaponya (cf. also Tna. gongoya, Te. SanSay) . zanjdixo. "baboon" + Har. zligiau, Gaf. Sonjord, Sl. zOnero, Ms.Sod. zangind, etc. Of Agaw origin, cf, Bil. Jéguaa, Kem. Jagzna, Quara Jagind, S.Ag. zagri; note also Hadiya dagicraa, Som. ‘danger, daayer. 4. THE SEMANTIC FIELD “SOCIAL ORGANIZATION” ‘The subdivisions under this heading are 4.1) law and government, 4.2) economy, 4.3) warfare, and 4.4) religion and superstition. The areas of the lexicon covering culturally sensitive categories such as these are, as might be expected, more fluid in tum-over, in acceptance of loan elenents and influences, than sone of the other semantic fields discussed hitherto in this paper. Moreover, pethaps more in this field than in any other, can extra- linguistic conclusions about the history, ethnic contacts, and pattems of cultural influence of the speakers of the language be dram. Thus, when’we examine vocabulary from the sphere of trade AL 5, 88197] A Comparative Apphoach to Amharic Lexicon 47 and economy in Atharic, we find a large muber of items taken from Arabic or fron elsewhere (lurkish, Persian, for’ example) through the mediun of Arabic. The traditional role of out- Siders, particularly the Arabs, in the comerce of Ethiopia is, of course, well known. ‘Anongst these comercial tems’we may include hésab ‘account’, 'batek ‘gratis, free", gumwk "customs', gorS~ bead ‘small coin’, kutay ‘rent’, mosbin "poor", mézan ‘scales", suk stall shop", wifket ‘ounce’, and so on (seé Lesiau (1970}). Similarly,’a number of military terns are of Arabic origin, including itens like gonade ‘scimitar’, sdeea4é ‘draw up in battle Line’, haxb ‘army’ and its purely Amharic extension axbaivia warrior’, aswell as the nanos 9f products of more recent military technology like miidag ‘cannon’, famdnja 'rifle', négt ‘rifle’, etc. On the other hand, most traditional military titles’ in Atharic appear to be indigenous, the only exception being baa, taken from Arabic, though ultimately of Turkish origin, as is #anin§a, cited above. ‘The contribution of the non-Senitic languages of Ethiopia in this field is mich smaller. In economy there is ance ‘salt bar currency’, daha ‘poor’; in military terminology, gaika ‘shield’, Loe 'soldier' (also meaning ‘servant', cf. abkix (of Arabic origin) with the same range of meaning), and more specific terms like dutta ‘club’, yibbo ‘short spear’, jhe "large shield’, “Many of the basic terms, however, remain inherited Semitic: dot ictory", adddita 'ki11", sig¢dnd 'camp', sdhawit ‘army’, sdyg ‘sword’, téLat ‘enemy’, fon ‘spear, arny, war" (foronnat), zinndgd ‘plunder’, and’so on. ‘The samo, is essentially ‘true of the field of economy and trade: cf. gdbaya ‘market', kiiggéied ‘pay’, 2étdbka ‘borrow’, ndggidi 'trade', Biti 'sell', and so on. Religious terminology includes a high proportion of long-established loanwords, which entered the Ethiopian language area through Ge'ez during the early centuries of the Christian era. A mmber of specifically Christian terns are of Greek origin: p ‘metropolitan, bishop", mindkuse (and variants) ‘monk", dibtixa "lay priest, cantor’, ganna ‘Christmas’, diyakonos ‘deacon’, etc. The majority Of borrowed religious’ terms in’Atharic is, however, of Hebrew or Aramaic origin (see especially Néldcke (1910:32-46); also Ullendorft (1968: 121-5)). This includes both specifically Christian and more general terns that may be characterized as "Judeo-Christian". Of course, it need hardly be said that a whole body of religious terminology has been taken over fron’ Arabic by Islamic commities. ‘This does not Concern us here, however, It is not always easy to be certain either about the direction of @ loan-movenent or even about the process of borrowing itself between closely related Senitic languages (Ullendorff 1968:120), This is readily illustrated in the field of reli- gious terminology. Items such as miZ*ak ‘angel’, mélékot ‘divinity’, hagirat 'sin"'® are all Telated to original common Ethiopian roots for which there is no reason to suspect borrowing from other Semitic languages, though the particular semantic value given to these derivatives in the religious figld may be attributed to the theological colouring current in other Semitic languages.!? ‘Thus, the specific value of mitak as ‘angel’ alongside the conmon Eth. root £& "send' is attributed to Heb. mat*2k; or haférat as ‘sin’ besides Eth. ‘not find, not have’, is owed to Aran. AGiz0%. However, a nunber of basic roots Like (is "be holy’, bbr ‘bury’, xgm ‘curse’, bxk (C-type) 'bless', mik ‘bless’, are most likely of inherited’ Semitic origin semantically as well as formally; there appears to be no need to imply outside influence in the semantic development. The component roots of the nanes of See NSldeke (1910) where a full list of items of Hebrew and Aramaic origin is given. Ullendorff (1968) gives a similar list. L3Perhaps not all the itens Listed by Néldeke (1910) need have received the influence of Hebrew or Aramaic; I cannot see why kds, for example, should be so influenced when its semantics are perfectly in accord with the general Semitic theme for this root. AALS, 8948 D. L. Appleyard (MAL 5/2 the Deity, mek (amtak) and gz? + bin (s92é"abaher), are similarly of inherited Semitic origin. ‘The common Semitic root *2 'god', which occurs in all other branches of Semitic, is not found as an inherited item in Ethiopian Semitic. ‘Two important terms from the field of mgic and the supernatural are of non-Semitic origin, namely buda and zat, The latter seems to be derived from the old nane of a pagan, (Agaw?) sky-god, typically “denoted” to the position of a (malevolent) spirit in the "new" religion. ‘The former, buda, has cognates throughout the Ethiopian language area, including in a nutber of Sudani¢ Languages, to which its origin has been ascribed (Cerulli 1931-2:346) ‘The List of items from the field of lax, government and social organization in general presents a rather different picture from that of economy or religion, in that the over- whelming majority of items is of inherited Semitic origin, and obvious loanwords are few in number. Out of 50 terms studied, 36 (724) are of safe, directly inherited Senitic origin, whilst only five appear to be of non-Semitic, Cushitic origin. Especially interesting anong these five items are the two terms ase ~ afe’ ‘emperor’ and the (originally) vocative Janhoy. ‘The appearance of both terms in Anharic can be roughly dated, the first to the 14th century, to the reign of ‘Andi Seyon, and the second possibly to the ioth century, to the reign of Sirs Dongol, at least’ according to native tradition. Both terms are of Agaw origin and appear originally to have been titles or appellatives of Agaw rulers. Amongst the itens of inherited Semitic origin are several for which the closest senantic parallel occurs particularly in South Arabian.*® The roots of many of these do occur else- where in Semitic, but the point of interest here lies in the close senantic correlation with S.Arabian. "Thus, agit ‘country’ (Gz. hagia 'town', ESA. hgn ‘tom, state"), gibbar ‘vassal’, gibite "Farner" (ESA. gbt-m ‘group of serfs',”*gbt’'servants', Gz. gibt ‘slave, servant'}, guét 'fief" (ESA. giwle idem.), ng4 ‘be king’ (ESA. ng} impose tribute’), Sum ‘chieé' (Gz. Soum, ESA. Sum "pracfectus'}, wisdin "boundary" (ESA. wen), hoag "law" (ESA. » Sit "Sell" (Go. Seti, ESA. 3(y)¢ ‘trade, carry on business"), and’ from the field of military terminology séravit ‘army’ (ESA. snct). 4.1, LAW AND GOVERNMENT agin ‘country’ + Gz. haga "town", Tha. hagéa 'country'. Semitic, cf. ESA. gt ‘tom, state’, Yemeni Ar. haar "ruined toon". atika ‘head, superior’: Tha. hatéiya, Old Anh, hadiika. There are several formally compatible roots in Eth-Sen., none of which, however is really suitable as a cognate to this item on senantic grounds:' hee "be destroyed", ill ep ‘count’, htk "be round’. “Rather, one wonders if this root ffl ‘chief’ might not be connected with Eth.Sem. Zhe "be more, elder’, despite the difference in sequence of radicals and the dif- ferent laryngal. ‘he root Lik is of safe Semitic origin and may be compared with Heb. 4ak’gai ‘senior ones".?! 2%Se0 Ullendor££ (1955:156) where some S. Arabian lexical parallels are listed; see also Hefner (1960-435). 2For the Semitic connections of Lhk see Ullendorff (1956:194ff.), where Heb. tah%ya ‘senior ones! is compared. AAL 5, 901977] A Comparative Approach to Amharic Lexicon 49 age ‘emperor! + Old Anh, hage~ hae, ‘This item appears to be of Agaw origin, cf. Kem. aSena (Kem. X frequently corresponds to Eth,Sem, $ in loans of both directions}. The introduction of the term has been dated to the reign’ of Andi Soyon (1314-43). It should, however, be noted that aSena does not appear to have any purely Agaw derivation and lacks cognates in other Agaw languages. On the other hand, it bears a close resemblance to the Ge'ez tem fiazani, which appears on later Axumite insctiptions in the sense of ‘administrator’ (Littmann 1915744 and 4, inser. 12 and 13). awa "proclamation" An agent nom (jiitaté) pattern from the Eth.Sem. root ‘ud 'go round’. In Go'e2 Sawadé has the meaning ‘bandit, one who roams the country’, but the root has a wide range of meanings to which ‘proclamation’ can easily be related. “This etymology would seen preferable to a derivation from the $. Ethiopian root **ud (liar. ida, Arg. aweda, Ch. odin, Gaf. autidéi) meaning ‘speak, tell unguda "guest, stranger’: Gz.Te. *ongada, Har. nugda, Gaf. ongodé, Sod. niigda, etc. From the Eth.Sem. root ngd in the primary sense of 'travel'. bad ~ bada ‘stranger’ : Gz. batd, Tha. basdi. Semitic, cf. Ar. ba*td ‘strange, distant’, ESA, bed 'renote’. dati established order’, daiiia ‘judge’ (n): Ge. diy daiifia, Har. daiina. ‘The Tigrinya noun daia, and perhaps also the Harari noun, are probably anharicisms. The Atharic root would appear to derive from *dny with transposed second and third radicals, Alternatively, if the verb root daiifié is a secondary formation from the noun dajlfia and not a primary formation, then daiifia may be explained as an extension in -ya of a noun *dané (? dé@yly)ani). Common Semitic dyn. id (vb), déiyyané (n), Tna. diiyyiind, ginadda "judge" 1 Gz.Tna, gird’, Har. findida, Ch. ginidim, etc. Senitic prd 'separate, distinguish’. fiitta release, divorce’ 1 Gz, fdtha, Tha. fathe, Har. fataha, Ch. gatam, etc. Connon Semitic pth 'open’. gabbind ‘pay tax, tribute’ (gibéxe 'famer', gibbar ‘vassal’, gobaa 'tax'): Gz. *agibbind ‘impose tribute’, gain ‘slave’, gibar ‘worker’, etc., Tha. gabbind ‘pay tax’, liar. gibaxa ‘pay an amount Of mney or cloth to. the bride at the conclusion of the en- gagement", Sod. gibbérdm "tame", WL. gebénd, etc. ‘The root gbx also means ‘work, do, make" in N. Ethiopian (G2. gibni), which would appear to be the starting point of the Various semantic’ developments in the rest of Ethiopian Senitic. Conon Semitic gbx "force, be strong’; with reference to the Ethiopian developments note especially ESA. °gbt 'servants', ghtm 'a group of serfs". AALS, 91so D. L. Appleyard TAAL 5/2 gutt 'fief" : Gz. g*obt, Tha. g” s€ti, Te. gutt. Semitic, cf. ESA. glu}, and perhaps also Ar. $i ‘the wall round a well or grave", geta ‘master, lord’ + Tha, g"duta, Har. gdyta, Gaf. g*iti, Arg. geta, ch. vata, Old Anh. g*eta, etc. Various attempts at the etymology of this item have been made; Cerulli (1936:252) considers it to be of Agaw origin, deriving it from a form analogous to Kh. qauata ‘leader’. Leslau (1963:76) suggests a connection with Galla gogta, which is, however, a purely Galla form and derivative of gosa ‘tribe’. Rather, the etymology proposed by Cohen (1939:88) is the most likely, nanely a derivative of *g"dy (Anh. -ge 'place'), which appears in Old Atharic as an independent item gay ~ gay ~ gey, and elsewhere in S. Ethiopian with meanings ranging from ‘cotry’ to ‘town’, ‘village’, and as a locative postfix, It is almost certainly Semitic and related to ESA. ’gw ‘collegium’, Hleb. gdy (Cohen 1933:54). gizza ‘rule, buy’ : Tha. giz*e "possess, buy', Te. géz°a, Har. giza*a ‘govern’, Arg. gazza, etc. ‘The root occurs in Ge'ez in the noninal form °agzé? ‘lord, master’. Semitic, cf. Ar. Gaza*a ‘distribute, share’. Jan(hoy) ‘emperor’ Composed of ay and the vocative particle hoy. The element Jan is of Agow origin: Bil.Ken. Gana ‘elephant’, used as a royal epithet. The introduction of this term into Amharic is traditionally dated to the reign of Sirsi Dongel (1563-97) .2? ogg "aw" : Gz. ogg, Tha. huggé, Te. frag ‘Limit’. The presence of the initial k- in the Auharic term corresponding to {t- elsewhere indicates that this is a Ge'ez take-over and not a genuinely inherited item. The root is Semitic, cf. especially ESA. ‘hg 'law'; this appears to be a peculiarly South Semitic development of figas which occurs in the rest of Semitic. hozb 'people' : Ge. A take-over from Ge'ez. Semitic, cf. 2b, Tha. fozbi. Ar. hézb ‘crowd, group of people’, hazaba ‘collect’. asi "compensate" ‘a, kéhasi ,. Te. kéhasa, Har. k@hasa, Arg. hithasa, Ch. hasan, etc. Praetorius (1879:7) suggested connecting khs (and Fth. ksh ‘grow lean’) with Heb. kaha¥ also meaning 'grow lean’. kissd ‘accuse’ : Tha, kissd, Te. kissa, Har. biisiisa, etc. Semitic, cf. Syr. bases 'blane, reprove', and perhaps also Heb. ksas ‘compute’, Akk. asdsu 'cut up’, Ar. kassa ‘pulverize’. *yiewoch (1911:280): "die Abessinier glauben, dass das Wort San(hdZ) erst seit dem 16. Jahthundert im Gebrauch sei, wihrend man frther’dafitr danzd gesagt habe. AAL 5, 921977) A Comparative Approach to Amharic Lexicon SL katiina "town" : Tha. kédéima, Har. kétim- in kétimbéxi ‘entrance to a ‘compound In Ge'ez kiitina means ‘edge, end, summit’, but is used later in the sense of ‘royal camp", as in earlier Anharic, from’which the meaning 'town' has developed. Probably Senitic, cf. Ar. katana 'hold,’hide, conceal’ and Gz. kétémi 'seal, close’. eiigta ‘punish’ : Gz. bisa ‘humiliate’, Tha, figte "punish', Te. pigsa, Har. bajata, etc. Semitic, cf. Ar. qaga‘a ‘grind, squash", Heb. qaga® ‘scrape’. {eba thief" + Ang. Leba, Gaf. 2ébi Perhaps of Galla origin, cf. abobu ‘steal’. B. neba, End. nisi nati 'swear' + Ge mijatd, To. miata, Gaf. madd, etc. The root mi€ occurs in other Semitic languages but with a range of meanings difficult to relate to Eth.Sem. 'swear': Ar, mahata ‘be barren", Heb. mahat ‘renounce’, Syr. mehet ‘be weak", ESA. mt ‘misfortune’. ‘The root shape is plainly of Semitic type, but {5 not readily relatable on semantic grounds to the other Semitic root(s) mht. gist "become king’ + Gz. néigSi, and hence modern Eth.Sem. ngs. Semitic, cf. Ar. na§aka ‘compel’, Heb. ndgas ‘urge, drive’, ESA. ng¥ ‘impose tribute’. niga ~ niga "free" If, as seems likely, niga is to be derived from Eth.Sem. ngf "be pure’, then the Tigrinya and Tigre forms mist be loans from Anharic; the protoform would then be ‘néigak. The root ngh is Semitic; cf. Ar. nagaha 'bepure', Heb. ndgah ‘sparkle’, etc. ‘Another term with the sane meaning in Anharic is ata, occurring in Old Anharic as hana, and in Ge'ez as hata ‘amy, troops, officers", "free, noble, warrior’. ‘The root is “it, ‘to which Heb. figt 'nobles', Ar. harsa ‘be’ of noble stock", Aran. trax ‘liberate’, etc., may be compared. . adtta ‘win a lawsuit! ndtsa "be straight, prosper’, T natsa ‘succeed’, Har. Adta*a, Arg. Adtta, W1.2. niiti, etc. Semitic, cf. Akk. rezh 'be fortified’. séovea (i) ‘work', (ii)'prescribe' : (i) G2zTe. siha, ‘na, sinhes (ii) Gz. Sénsa, Tha. sinse, Te. sina, Har. sinara. The root A* may be compared with Ar. Janata, ESA Une ‘order’. For the root sah, on the other hand, I can find no satisfactory cognate. MAL 5, 9352 D. L. Appleyard IAAL 5/2 Saggata ‘revolt! + Tha. Baggaitit, Arg. Befgiita, Sod, Bésgitin, etc. In some S.Eth, languages the root appears as 34¢ and it is this form of the root that may be compared with Gz. st (°asfdta ‘deceive, seduce’). Probably Semitic, cf. Ar. safuta "be Liberal'.#? ‘Sieeiinie "award" 1 Tha, Saeedind, Har. Setina, Arg. SoeLina, Caf. Léveiimi, etc. A'S, Ethiopian B-type pattern from the root sm 'peace'. Common Semitic Sem. Sun ‘chies" + Gz. Soyun (p.p. of Sem "put, appoint") ‘The item occurs throughout Ethiopian Semitic. The root Sym is, of course, common Semitic, but for an especially close senantic parallel to the noun Sojum, Sun compare ESA. Sum "praefectus, dominus'. S.mag tle ‘elder* + Tha.Te. Sumagute. Of Agaw origin, cf. Bil. singin (pl. Sénégal), Khm. sdingot, Kem, singéx. Anh. Somgot(Qle appears to be derived from the plural stem sénagiif~ plus an individualizing suffix -e. wittadisé "inherit' : GzeTna, Te. wiinsa, etc. From this root is derived the noun 19st ‘hereditary landrights'. Common Semitic wrt. was "guarantor" 2 G2.Tna. alos, Te. waftsa, Har. wis, Ch. was, ete. Semitic, cf. ESA. whS in Qatabanian suiS "gift, donation’. wistn "border’ + Gz."na.Te. witsin, etc. Semitic, cf. BSA, won. Another Amharic item of the same meaning, dinbéix, was related by Practorius (1879:100) to the root dba ‘mountain’, zega "subject 2 Gz. zega ‘poor", Har. 28ga. ‘This is probably of Agaw origin: cf. Bil. Jéxa 'poor', Ken. Ji "be ruined"; note also Beja Yehaana ‘beggar’. Another form of the same Cushitic root appears as 2 ioan in Anh. (a ‘poor’. 4.2. ECONOWY anoge ‘salt-bar currency' : Tha. *anole, Har. andte, Arg.S1.W.2. anole, etc. Of East Cushitic origin, cf. Galla anole, Sid. andte, Afar anole. See Rundgren (195: (937a:39) 21) for a detailed discussion on this root; see also Wajnberg AAL 5, 941977] A Comparative Approach to Amharic Lexicon 53 bor 'silver, dollar’ i: Tha, bow, Gaf. bornii, etc; a different root pattern occurs in Gz. baru, Semitic, cf. Ar. barta ‘be just’, lleb. barat "be pure", and especially Akk. bard ‘shine’. batya slave! This item is traditionally connected with the ethnic name Barya, an Eastern Sudanic speaking people of N.W.Ethiopia, already mentioned in the 4th century inscriptions of *Ezana. These people do not, however, use the nane of thomselves. It is not inconceivable that the ethnic nane (of whatever origin) cane to be used as a comin noun 'slave", since these and related Peoples have long been the traditional source of slaves for highland Ethiopia. Praetorius (2879:171) suggested a derivation from Sem. ba ‘cattle’ (cf. Anh. bine 'ox'), but I do not think that this is very likely; in the inscription of *Ezana the nane occurs in the same form as in Anharic today, barya. Zant "load Tha. géSand, Te. 2agcana, Har. farana~ fina, WL. tani, Sod. Sand, ChE. Zardm, etc. Senitic z¢n, "travel, migrate’ (Ar. zaSana, Heb. saan, Akk. gdm "load"), but note also Heb. #i°an "load". ‘Ta. doxa, Gaf. doha. Of Cushitic origin; cf. the Agaw root “Ze (Bil. Jéxa ‘poor’, etc.) taken into Anh. (as zega 'subject'. Forns with initial d- do occur in Agaw (Kem. dixa, Khm. doxa 'poor') but these look moré like loans from Anharic; but note also Galla dega ‘poor’. daha ‘poor’ gibiya ~ géboya ‘warket" Gaf. giboya, Go. gebé, Ch. gibéiya, etc. The root is almost certainly gb* ‘enter’; thus *gibo* + ya would originally and Literally have meant ‘meeting, reunion, assembly’. ginzib ‘money, property’ Gz. gdnzdb ‘treasure’, Tha. ginzib ‘money’. ‘The inmediate origin of this item is an Aramic term such as g@niizb (also occurring as gize®bar) ‘treasury’. ‘The ultinate source of all these is O1d Persian ganza~ ‘treasure’ and ganzabara ‘treasury’. g’az ‘caravan’, tig"azd ‘travel’: Gz. gota ~ giitazi ‘change camp, migrage', Tna. gitazd, Har, g@za, Sod, gazi, etc.; for the nou gaz cf. G2. gratcz, Sod. g*az, Har. giz. Semitic, cf. Ar, §dza “travel', Heb. gf "pass, change’ (giz), rather than gz* as cited by Dilltann’ (1865:col. 1187), meaning "cut habt ‘wealth’ ‘This is a take-over from Gz. habt (Vahb give"); a purely Anharic development of this occurs as kibe ‘cattle! with false restitution of for h, and subsequent differentiation of meaning. bagged "pay, divide’ : Gz. kigdtd "divide, assign’, Tha. bégéea 'pay', Te. kigla, Har. kagata, etc. MAL 5, 954 D. L. Appleyard (AAL 5/2 Semitic, cf. Ar. kafata ‘support, maintain', Heb, kapat ‘double, mltiply’. br atpind ‘count! + Ina. fagind, Arg. batting, Gat. féttiind, Sod. bottiinin, Ch.Gy. hafinin, etc. — This root fx ‘count! is probably identical with fn. "tie a knot' (Gz. fagdnd ~ biagini, Aah Waging, etc.). If this assumption is correct, it gives an interesting glimpse into carly methods of recording mmbers. Senitic, cf. perhaps Syr. qefat ‘attach’. os dibha "lend", té-Rébka 'borrow' : Gz. Lékkoha "loan", ‘na, *adibl Semitic, cf. Ar. Lagiha ‘conceive’, Heb. fagah ‘take'. che, Har. attpiha, ete. naggitdi "trade" 2 Tna. niigida, Tha, néigda, Har. négdé Sa, SL. niigidi., Ms. neggiidin, etc. The root in Ge'ez maintains a wider range of meanings than the modern Ethiopian Semitic forms: Gz. nigidé ‘travel, migrate, trade’, which reoccur in several derivatives in Atharic: mingid "road", ongida "stranger'. Semitic, cf. Aram, ngd '£low', Siti soi." 2 Gz. Sofi, tna. Sayiitii~ bagi, Senitic, cf. ESA. Slylf 'sell', and perhaps also Ar. Sauafa ‘make a long journey"; for the semantic development compare the previous item, 4.3. WARFARE aikin ‘soldier, servant' : Tna.Te, ‘askin ~ able From Ar. ‘askar. » Mar. akin. det ‘victory' Tha, dot, Har. dif ~ dot. Both the Tigrinya and the Harari items are probably loans from Atharic. The root is dhe (Gz. tidotd "retreat, flee) and *doke > dat would originally have meant something like (flight, rout’, the apparent reversal of meaning coming from a phrase such as dot adinnigé ‘make a’rout' > "be victorious’. Semitic, cf. Ar. dajata ‘flee’, Syr. dehee ‘be afraid’. giaddaed "Ki" 1 Har, gidéta, Arg. géiddita, Sod. giddiitin. ‘This is almost cortainly not connected, at least directly, with Eth.Sem. pet "kill" as Suggested by Practorius (1879;72). It'is rather derived fron gd¢: Cf. Ge- tigaddd, "fight", gidata ‘carcass’, Te. gadéta ‘tight'. Semitic, cf. Ar. §adata ‘quarrel’, ajadata ‘iden. '. The root kee does not occur in Anharic, but is ‘preserved elsewhere in Eth.Sem: Gz.Tna. bataied, Ch. °atindm, inn. -Atind , etc. gorade ‘scimitar’ 2 Tna, g'onade, Te. gorade, Har. guidde, Ms. g*énade, etc. Probably from Aden Arabic gutad. AL 5, 961977] A Comparative Approach to Amharic Lexicon 55 gakSa ‘shield’ : ThaArg. gaSSa, Gaf. gaki, etc. OF East Cushitic origin: cf. Galla gafana, galéa, Som. gaaldan; note also Kem. gala, S.Ag. gaki. "take prisoners’ Gz. mahndhi, Tha. mandxi, Te. mardha, Har. mardxa, Ch. maniixiim, etc. Throughout modern Eth,Sem. this item is construed as a C-type triliteral (/miR), a remodelling of the A-type quadriliteral (mhtk) of Ge'ez. Sa§gind ‘set up camp" + GeTna. sigind, Te. adigna, Har. sigana, etc. ‘This is probably the sane root as Eth.Sem. 4{t ‘measure’ and common Semitic spt: Ar. Sagara "travel', Heb. sdpar 'comt', ESA. sft (n) ‘measure’. Sinawit ‘aray' + Gz.Tna. sirawit. Semitic, cf. ESA. snot. siyg ‘sword’ 2 Gz. sdyfy Tha. stiygé, Har. 4%4, Semitic, cf. Ar. says. Babki “fee' 1 Gz. sihiyi ‘take refuge’, Te. ska, ar. sika ‘flee’, Arg. sihka, Sod. Sa8Kin, Ch. Lakin, S12 sade, WL. Babe, ote. The initial 3- may be explained as due to removed assimilation to the medial 3; the latter would seem to derive from k through palatalization to & and thence 3, An alternative etymology was proposed by Praetorius (1879:132), who connected Anh, 34884 with Gz. sis (Cansostiod 'walk"), Heb. Sc882* "Iead on’, The’ former derivation, from sky, is, 1 think, preferable. Josie 'dagger' 1 Gr. séiwtit, Tha. Sotie~ Iutol, Te. sotit, Har. Jodie, etc. Perhaps from an East Cushitic source: cf. Saho Sota, Som. Sootat. #ilta ‘hate’, fitat ‘enemy' : Gz.Te. saf*a, Tha. $al%e, Har. fata*a, Caf. tiga, Arg. fitta, th.cy: dinam, etc: ‘The common Semitic, cf. AKK. gel "be hostile’. The comon Semitic term for enemy, "dan, occurs in N.Ethiopian: Gz. zit, Tha. sin. ton ‘spear’ This might be connected with the Eth.Sem. root sux (Gz. gord 'carry', gor ‘burden, load"). Alternatively, it might be connected with an identical root occurring in Gaf. sawoind "strong, rigid’. The former has a Semitic cognate in Sogotri gon ‘wear’. uiigga ‘stab’, ttivagga ‘fight’: Gz. uiigta, tiuaago%a ‘wound o.a.', Tha. wiige, Ch. sileam ‘crush’, etc. Semitic, cf. Ar. unjara "hit", Soq. *ége. AAL 5, 9756 D. L. Appleyard TAAL 5/2 witttaddin 'soldier' 1: Tha. witttshaddin . wiittahaddin, Guidi (1901:col. 579) suggested derivation from a phrase wiltto addéx ‘che vive andando qua e 18"; this is ingenious, but sounds a little Like popular etymologizing. zinndgé 'plunder' 2 Tna. zimndifii, Te. zénga, Har. zindiga, etc. Semitic, cf. Ar. zanaga ‘cone upon s.th.', Aram. nezképiita "impetus, attack’, ESA. zat "incursio bellica’. 4.4, RELIGION AND SUPERSTITION antak "God" + Gz, Yamtak, Tha. *antax. Formally a plural of *mitk 'king'. The same development of the common Semitic root mek to express ‘divine ruler’ as well as "secular ruler’ occurs outside Ethiopian Semitic. agziabvher "Cod" From Gz. *9gzé*aboer lit. "lord of the earth’. buda ‘one who has the power of casting the evil eye’ ‘This term occurs throughout modern Ethiopian Semitic and in various Cushitic and E. Sudanic languages. It is to the latter that Cerulli (1931-2:346 note) ascribed the origin of this item, deriving it from a noun originally meaning ‘smith’: Shilluk b3do, Bari bod0, Bongo bodo. Smiths and metalworkers were traditionally associated with the power of the evil eye throughout much of Africa and elsewhere, barniieit "bless" + Gz. bani, Tha. bandxit, Te. banika. Common Semitic bak. dibtina “lay priest, cantor’: Tha, dibtiaa In Ge'ez dibtiina has the meaning 'tent, tabemacle' and diibtinawi ‘one who Lives in a tent’, of which Diliman (1865:col. 1106), Says ' vulyo etiam vocatur Caonccus «++ homo Léteratus." The item is ultimately of Greek origin: déphthérad ‘skins made into a tent" (pl. of déphthéra ‘skin, leather"). bébbiind "bury" : G2.Tha, hibénd, Te. bibaa, Har. pibéira, etc.; a nominal derivative occirs in'most’Eth.Sen, languages with the mean- ing 'g ia, Anh. mikabox, Har. kabni, Connon Semitic gor. iiddiisi ‘consecrate" 2 GacTna, piiddéisd. Néldeke (1910:35) regarded Eth.Sen. fds as influenced by Jewish-Coristian usage. Conon Semitic qd. AL 5, 981977) A Comparative Approach to Amharic Lexicon 87 bes ‘priest! 1 Gr. fiisis, tna. Ra8¥é, Te, fis, Har. Es. ‘The modern Eth.Sem, forms may be derived fron a pattern “hiss. A loan from Syriac qakeka ‘elder’. minnia "bless" + Tha, minixdi, Ang. merriba, Gat. mova, E. ninniihi, etc. A denominative formation from marak, ‘saliva’. Spitting as a mark of benediction occurs Widely in Ethiopia, cf. Galla duga 'benedizione di augurio 0 di riconoscenz manifestata 4 nezzo della saliva soffiata leggermente verso la facia 0 1'oggetto che si vuol benedire' (Da Thiene 1939:323). liga ‘curse’ : Geetha. Semitic, cf. Ar. rafana 'stone', Heb. adgam, AkK. nagdmu 'protest, cry’. toni fast" + Ge.Tna. gomd, Te. soma, Arg. toma, Gaf. simi, Ch.E. fomdim, ete. The sane root, gum, also occurs in Arabic; both it and the Ethiopian Semitic root are regarded as a’loan’from Heb. gam (Néldeke 1910:36) . zan ‘evil spirit" This item occurs throughout Ethiopian Semitic. It is almost certainly derived from the name of a pagan Cushitic sky god, probably Agaw: cf. Bil. Jax ‘sky, sky-god’, Khn. 20x ‘evil spirit’ (perhaps rather from Anharic?); the safe Cushitic root occurs in Rendille 4gne 'sky', and perhaps in several Onotic languages, too. The term zax also occurs in Arabic as @ loan from Ethiopian Semitic. 5, GRAMMATICAL ITEMS: PRONOUNS, NUMERALS, AND PARTICLES This final section covers what is, strictly speaking, not a semantic but a gramatical field, in so far as the referents of the items discussed are not definable objects, actions, or qualities, but aro (in the case of pronouns) part of the relation of the speaker to his audience, or’ (in the case of particles) the internal structurals of the language itself. These two categories, pronouns and particles, and numerals which in many ways lend thenselves to grouping with the other two, are typically included under the "basic'' vocabularly pole of the lexicon. For example, the 1952 Swadesh 200-item "basic" word-list (Swadesh 1952: 452-63) includes the cardinal numerals from 'one' to 'five', the pronouns of the Ist, 2nd, and 3rd persons, singular and plural, as well as other pronominals Like ‘here', 'there', ‘when', 'how', and the particles ‘and’, 'at', ‘if’, ‘in’, ‘not’. The principal behind the inclusion of these itens is in part that such elenénts ténd to be among the most conser- vative areas of the lexicon because of their intimate connection with morphology. whether all the itens selected by Swadesh can be correctly classified as lexical universals or not does not immediately affect this discussion. Some of the items, especially 'not', if’, ‘and', cannot, I feel, be thought of as lexical universals and should not be included in a "pasie’” word-iist without due consideration of the typology of the language(s) under exan- ination. However, in most languages the lover numerals and the primary pronominal forms are, on the whole, derived from the inherited stock of the language. There are naturally exceptions; within the field of Afroasiatic alone one need only note the use of Arabic mumerals above 'two', or ‘ten’, in many of the Berber dialects. However, the data from AAL 5, 99)58 D. L. Appleyard [AL 5/2 Amharic follow the expected trend in that all the morphemic elements (though not the actual ‘items thenselves) of the personal and other pronouns are inherited Semitic, as are all numerals except ‘nine’ and ‘thousand’. ‘The personal pronouns of the 2nd person, anti, ani, antu, the interrogatives, man, mon, and the denonstratives, yah ~ -z2th, and ya ~'-zz0ya, present no problems in derivation from common Semitic forms. the 15t person pronouns, one and ovia, are clearly derived fron inherited Semitic components, but not directly. "the final -e of one 'I' appears td be due to the influence of the corresponding possessive suffix, -¢; a Similar confusion of independent and dependent pronoun forms occurs throughout S. Ethiopian (Hetzron 1972:32-3). The initial o-, on the other hand, instead of an expected a- (cf. Gz. *and, Har- an, Arg. ay, etc.), is probably taken over by analogy from the plural di/ia, @ feature which, ‘again, occurs in’several other S. Ethiopian languages: Ch. aya, etc. The Ist person plural pro- Roun, vfiia, together with’ its cognates throughout S. Ethiopian and in Tigre in N. Ethiopian, lackS the initial n- found in Gz. nofni, Tha. nolna, and the generally reconstructed Proto-Senitic form. The loss of this initial n-, whether it occurred at the common Ethio- Plan Semitic stage or independently im S. Ethiopian and Tigre, can be attributed to a kind of haplology: nolina > *chna~ fron. (cf. Egyptian Arabic ij, "Yoneni Arabic hina, etc. , for parallel developments). Hetzron (1972:33-4) derives the’S. Ethiopian forms ftom nokia > "nina > “iatia > iia, etc. and regards S. Ethiopian forms like Arg. anna as having been depalatalized later. while there is some evidence for the change -ahC- > ~ic- in Amharic and more so in certain Gurage languages,"* there would appear to be no other instance of palatalization caused by a preceeding ¢lenent, as opposed to a following one, in Amharic. Might not all the S. Ethiopian forms, then, be derived from a common *hna ~ yma (as Te. frona) and the palatalization n > i be secondary, Arg. anna, etc., preserving a mre original, unpalatalized form? However, be this as it miy, the pronom is ultimately of inherited Semitic origin, which is the main point here. ‘The 3rd person pronoun base 2t5-~ 254- derives from the noun ‘head', Gz. 19° and almost certainly originated in the use of a noun denoting a part of the body together with the appropriate pronominal suffix as_an emphatic pronoun, a construction which is still current in Anh. one Aase "I myself! and Tha. “and 10*sdy. Thus, otau ‘he’ derives from 40°s + u "his head" and similarly Tha, nossu ‘he’ from né{s + u his soul', and perhaps also along similar lines Arg. kassu 'he' from kits + u ‘his belly’. The original independent pronouns Of the 3rd person in Ethiopian Semitic (Gz. woPatu, yo*oti, etc.) survive in Gafat (wot, yot), way (ut, it), and a variant form inh occurs in Tigre (hotu, Iota) and apparently in nany’of the Gurage languages. The old plural pronouns are mostly teplaced in S. Ethiopian including Anharic by a compound of omni- (Ge'ez plural denonstrative/relative ste) and ‘the singular pronoun (Hetzron 1972:29). The numerals in Anharic are of inherited Semitic origin, except for 'nine', whose origin is enignatic, and ‘thousand’, which is of Agaw origin. Sone of the Semitic mmerals, whilst presenting no etymological problems, do exhibit interesting phonetic developments, especially and 'one", hutiit(t) 'two', 40st "three", and axat(t) four’. The numeral and ‘one’ (Old Anh. hand} derives from hadi (Ina. hadi ~ fade, Te. fen. hatte, Arg. hand), an ellipsis of *ahad-, vhich occurs in G2. *ahadu and Har. akad. ‘The other S. Ethiopian forms could derive 2"The elenent -dot ~ -dan in the interrogative mondst ‘what?’ is, however, of Agaw origin. See Tubiana (1963:15-17). 2*the reconstructed protoforms of the demonstratives are *ziki and *zé*a, to which Gz. zoku "that" and zé°a~, possessive pronoun base, may be formally compared 26rxx: Anh. tim besides tam "beard’ : Gz. ohm; Anh. mize "best man" from *mohz + dys see Hetzron’(1972:33). AAL 5, 1001977) A Comparative Ayproach to Amiaric Lexccon 59 either from the shortened or the longer form: Gy, @t, Ch. at, Sl. ad, etc. The numerals ‘two", ‘three’, and 'four' exhibit the following "weak" sound’ changes!*” hued() from set otPetit (ke >'h); s08t from sétaistd (€ > F', ive. rounding ;possibly the only instance of this sound change in Anharic); ataz(t] from tanba®zi (b > g; one of two instances of this change in postconsonantal. position). ‘The Ethiopian Semitic languages all use the root ké? for the mmeral 'two': Gz. kut?eltu), Tha. bvtytti, Te. bul-oz, Anh. hutiit(t), Arg. ket, Har. ko*ot~ kit, Ch. tet, etc. The numeral ‘twenty’ is also’ formed fron this root in’S. Ethiopian (Anh. haya, liar. kiya, Arg. kiya, Ch. xaya, etc.), whilst the original inherited Semitic term survives only in N. Ethiopian (Gz. ana, tha.Te. asa). “The form haya, etc., is analogous to the other tens, being formed from the root of the unit numeral plus the suffix -a: *kI*)s£e + a. The root fy, from which all other Semitic languages derive the numeral 'two',?* surviyes in Ethiopian Semitic only in the day nane Gz. sdnuy, Anh. sdiio 'onday' and’ the verb, ‘accom. pany" (Ing. sdéinndyé, Anh. Said, Arg. Seviia, Har. ab@ia, etc.), to which Ar. tana ‘double’ and Heb. Sand 'repeat" may be compared. Various attempts have been made at the etymology of the mmeral ‘nine’, Anh. zégéii, which has cognates throughout S. Ethiopian, whilst N. Ethiopian preserves the inherited Senitic root (Gz. tus%atu~ téssatu, Tna. tlatti, Te. 40). Praetorius (1879:203) tried to derive it from a compound of the denonstrative zd plus a form of the root hss 'be small" plus an adjectival ending -di: *zdlropsdn ‘the snaller', presumably referring to a kind of subtraction method, like Lat. undeviginté ‘nineteen’. {eslau (1949:278-9), on the other hand, tries to relate S. Ethiopian ziti to N. Ethiopian 2s° + -ai by a rather tortuous com- bination of metathesis and irregular sound changes. Neither is, to say the least, satis- factory, nor can any cognate be found for this S. Ethiopian root outside the Semitic lan- guages of Ethiopia. The mmeral 3é(h) 'thousand" is of Agaw origin: cf. Bil. Six, Kem. Bé, S.Ag. Yay, and is not from the Senitic root sy as suggested by Praetorius’ (1879:203). the borrowing of this Agaw item here nay be neatly explained as "filling a gap" in the inherited mmerical system, ‘The Semitic root “tp, which is used for ‘thousand’ in the other Semitic languages, has the meaning 'ten thousand! in Ethiopian Semitic (Gz. *24, etc.), whilst ‘thousand’ is expressed by the phrase "ten hundred’ (Gz. caSdxtu mod). Vie mst now turn to the slightly more complicated sphere of particles, which may be either separable or inseparable, and which includes a wide range of items such as time and place adverbs, syntactic markers (conjunctions, etc.) and prepositions and postpositions. » The najority of these is of inherited Somitic origin. Sone of the more fundagental ones are common to all or most Semitic languages: bé- 'in, by’, €é- 'to, for’, miée 'when?", ata "withut" £28 "below", etc. Most, however, are peculiar to Ethiopian Semitic: | 49-’ 'when', vskii 'as far as’, sont ‘how many?', care ‘today’, ondd ‘as, like', onja "I don't know", bseéa ‘only’, ete. ‘The nimber of items that may’be attribiited to’a Cushitic source in’this field is quite small: alun ‘now", gun ‘but', on ‘but rather’, na ‘come!", gar(a) ‘with’, ‘and not all of these are incontestable. 271.e, sound changes that do not occur wherever the appropriate phonetic environment is present, but are restricted to certain lexical items only. The weak sound changes of Amharic’ are: (i) fricativization of k>h, b>w~ FP ~ 9, m> uw; (ii) the intrusive or inorganic nasal, ...VC... > -VnC-. *"Occasionally in ESA. kt*y and ké*ty are used as cardinal numerals besides more usual, tay, tlnlty. AAL 5, 10160 D. L. Appleyard [MAL 5/2. 5.1. INSEPARABLE PARTICLES ba-/bu- ‘in, by/if', Li-/Lo~ "to, for/in order to’, ki- ~ hd- "from, by, with'?® and si: ‘of" are ali straightforward common Senitic.°? The conjunctive suffix “m(m) and its wide- spread cognates in Eth.Sem. (Gaf. -rma, Arg.liar. -m, and the particle of insistance in Gaf. -m, Har. -m{o), Tna. “mmo, Gz.Te. ma) are related to the Semitic element -m which occurs as an enclitic in various languages: Ak. -ma, ESA. -m, -mv. ‘The prefixed particle s0- ‘when, whilst", which also occurs in Argobba, i probably,to be identified with the first elenent in the’free standing preposition oské ‘as far as’ and is, perhaps, ultimately derived from another single pronotinal element in Senitic, ¢-. The enclitic particles -a(8) ‘but’ and -ma "and" nay be of Cushitic origin; at least, Somali 4¢ and na have almost exactly parallel functions. Within Ethiopian Semitic Gz. -da and Tna. =n may be compared. Similarly, the interrogativizing suffix -no (Gz. -nu) has a parallel in Qushitic, cf. Agaw -né (Kemint). These Cushitic enclitics may, of course, be loans from Ethiopian Semitic and not véce vetsa. ith most of these particles that are not straightforward conmon Semitic like bi-, “m(m), etc., one is on rather unsure ground in trying to establish etymologies for individual items, especially where only one consonant, OF one consonant plus a voxel, is involved. A more positive statement can, however, be nade over the point that enclitic particles marking such syntactic functions as coordina- tion, emphasis, interrogation, etc., are a feature more of Cushitic languages. Whilst the actual forms of such particles in Atharic cannot be readily identified as to origin, the principal behind such particles is surely to be sought anongst the Cushitic languages Tather than be attributed to an inherited Semitic development. 5.2. SEPARABLE PARTICLES hun "now" : Arg. aha, Hor, axxa*, Gaf. ahuit, Sod. ahu, Ch. at, M. dxuitna, Si. akku, etc: Praetorius (1879:262) derived this from the root tum 'be': "haku, bakun ‘im Zustand’. It would bea little difficult, however, to relate all the S. Ethiopian forms to such a derivation; a case could just possibly be made out for a development *bi-kun > ahun in Anharic, but not so for the others. Cerulli (1936:232, under aha), on the other hand, pre~ fers a derivation fron a Sidano denonstrative elenent ak, hak. The absence of the final n/t in the Harari and other forns leads one to suspect that it is an added element and not part of the root and, therefore, if this assumption is correct, a derivation from hum be- cones even more unlikely. ati ~ yatit ‘without' + Ang.Gaf. ati, Sod. yatit, 2. bat, ete. This is most Likely derived from the elenent bité occurring in Gz. onbiiti, Te. *mbit, also meaning ‘without’, rather than fron the negative verb prefix’a€-. Compare Ar.Heb. bat. In the Zway item bat the b- would be the preposition and not part of the root; similarly the y- in Anh. yatd is the preposition wi-. 2°according to Praetorius (1879:267) the preposition 9- is also ultimately from hi bi > hi- > d- > o-. For the Semitic cognates of kd- cf. in particular ESA. (Minaean) = "to, towards, for’, (Sabaean) k- ‘at the time of" 3°ya- derives from the denonstrative/relative zi. MAL 8, 1021977) A Comparative Approach to Amharic Lexicon 61 ama ‘last year’ + Arg.Hlar. ama, Gaf. yaymon, Ch. ema, Sod. yamana, 2. ami, ete. Composed of Eth.Sem. ‘am "year" and an elenent -na glso occurring in other time adverbs in S. Ethiopian: Anh. tonantonna "yesterday", Har. 2@éuna, s@stina ‘the day before yesterday’. ohko emphatic particle: Tha, *akko~ -[k)ko, Gaf. -ko; also Gz. -ke, Te. *ake. Probably to be related with Heb. *ak ‘surely, indeed", but note also the ESA. -k severative particle, As with several other particles discussed here, the nucleus is a common Semitic elenent é with various functions. ‘The extensions on this’ thene, however, renain obscure; Heb. ah, ESA. -k, and Eth.Sem. (2a/]ko/e may all be derived from the same original elenent, but as always’ in the case of one-radical etymologies, no definite statement can really be ond 'as, like', ondo- "in order to': Arg.Gaf. andi Praetorius (1879:86) derived this from Gz. *anti ‘as The etymology is attractive on senan- ‘tic grounds, but slightly irregular on phonetic grounds, as the development nt > nd cannot be established elsewhere in Anharic. Gz. *ondii has been connected with lleb. °¢¢ ‘with! (Praetorius 1873:645; also Barth 1893:17), and also, less successfully I feel, with Ar. Sinda ‘at, near’. Heb. pt and Ak, été, which would appear to be cognate, tay have de- veloped fron **ént-, or conversely Gz. *ondi may be a dissimilation of *é¢t-. anga'T don't know" 2 Gz. *ondaté, Tha. onde. ‘anh. :nja must derive fron a form **mndéta to account for the palatalization d > J. Gz. enda* is usually explained as a compound of negative *an + dasé ‘ny knowledge’ (Sen. Jad*). ong contrastive particle Praetorius (1879:85) connected this with the last item, particularly Tha. ?ondvté, which, it is true, could be expected to correspond to an Anhatic form angé. However, the exact semantic equivalent of onjé in Tigrinya is not *mdo®é, but *ondo. Anh. ongé’is fundamentally an enphasizing, affirmative particle placed at the end’of a clause, which may optionally be followed by @ Contrasting clause. Hence, the function of gnjé in sentences of the type tagabiiz mngi ‘do help yourself." and totkab néw angi tanno’ aydotédm ‘it's big, not small" is esentially the sane and agrees semantically very well with a particle occurring in other S. Ethiopian languages, the form of which is entirely compatible with Anh. ongé: Har. ange, Ch. dé, Oy. dgiu» Enn. atiya, End. abtiyi, etc., all meaning "indeed" and apparently of Cushitic origin: cf. Kambatta, Tembaro dtku, The apparent formal convergence of wnja and ongé may, therefore, be accidental and the two may not be directly related. oshii, vsko- ‘up to, til’ : Gz, *ushi, Te. *asak, Arg. vsti, Gal. vsko-~ abba , etc. This is almost certainly composed of the sane elenent as the conjunction s9- and the prepo- sition ki-. Praetorius (1879:296) suggested a connection between °0s- and Ar. haytu. ‘An inmediate connection between the two, however, seems unlikely on phonetic grounds alone. If hayéu is to be analysed as composed of a separable elenent hiay + £ + w (see Fleisch (2968:146)) then perhaps the element ¢ and [th.Sem. *9s- are ultimately related. AL S, 10362 D. L. Appleyard [AAL 5/2 ayy (4) distributive particle, (41) ‘wailst' These two formally identical but semantically separate particles are most likely to be dis- tinguished fron one another etymologically, too. The distributive ayya is probaly merely ‘a reduplication of yi-, itself from zd, the demonstrative/relative pronoun, The temporal conjunction yyéi, on the other hand, i8 probably to be connected with Gz. *anzd, to which Ar. *éd@ may be compared. bo8Sa ‘only" Gz. bakty ‘only’, bohut ‘alone’, Tna, bafité, Te. boot, Arg. botéa. Asharic and Argobba ba8éa derive fron *balit + ya. Praetorius (1879:140) explained this as a contraction of Gz. bi%ahatté "in one’, This seems highly unlikely both on phonetic and structural grounds. Rather Eth.Sem, bit should be compared with Ar. bat 'pure' and ESA, bit of the sane meaning. gen "but! : Tha, gon~ go*an, Arg.Gaf.Go.Sod. gan. Praetorius (1879:149) derived this from the root ugn (wiigin 'side"). However, a similar item occurs in Agaw, Kem. gin, vhich is, moreover, often combined with the item dit (dita thing") as ddagdn.” This appears as a loan in Amharic as_daxu gon, or, with daw trans lated into Anharic tems, as négéx gon (Tubiana 1970:343-7). gar(a) ‘with : Gaf. gaxa. ‘This postposition is probably a loan from Galla gana. tay "upon! 2 Gz, fated, Tha. L0t2é, Te. Laat, Har. Léay~ My, Caf. Lassi, Sod. tai, Ch.Go. ‘The common ith.Sem. root £*t is composed of the Semitic prepositional elenents £ 'to, for' and © 'on'. ‘The palatalization in the Anharic, Harari, and Gafat forms arises from 4ake-, as occurring in Go'e2 before pronoun suffixes. mae "when?" : Har. maBi, Arg. mae, Gaf.ch. méBa, etc. ALL derive from *nite, which Ludol£ (1098:13) actually records for Old Anahric. Common Semitic mty. The N. Ethiopian languages use a term of different origin: Gz. ma*ze, Tha. mi?as, Te. mi?aze. né-w is" Functionally this is a verb, but its etymology and features of its inflection betray its origin as a declarative particle. The Stem né-, to which various pronominal endings are added (originally object markers), is common to’S. Ethiopian in the function of copula (ietzron 1972:80). ‘The sane pattern, nd + pronoun suffixes, occurs in Ge'ez as a declara~ tive or deictic: “nal ‘behold (hin)!"" (rarer ndyo) , néyonu’ "behold then!", etc. Ethiopian femitic ni- is rotated to the common Senitic dectaritive particle, cf. Ar." *ani-y Heb. me. na ‘come’ 1 Gz.Te. nda, Tna, nea, Har. na*. AAL 5, 1081977) A Comparative Approach to Anhanic Lexicon 63 $n irregular imperative of the verb 'to come' occurs in most Cushitic Languages, as well as in other Afroasiatic Ianguages. The Eth.Sem. form né® , the inflection of which follows the usual inperative pattern (Gz. Sg.m. nda, Sg.f. nod, PLim. nou, PL.f. ndta, Anh. Sg.m, na, Sg.f. ndy, PL. mu), is most probably to be connécted with the comparable irregular imperative in’most of the Agaw languages: Bil. Zax", Lax"a, Ken. Lag, Laga, etc. It is not possible to say outright which has borrowed from which here, if, indeed, borrowing has taken place. However, in the absence of any formal cognates of Eth,Sem! naé- in the rest of Semitic, an Agaw oFigin would not seem unlikely. sont ‘how much?" Arg. sont is the only direct formal cognate, though other Ethiopian Semitic languages exhibit forms which are almost certainly ultimately related to sant: G2. *os4ntu, perhaps derived from interrogative *o{(o) + Sont-u; also Har. mésti, SL.WL.2. most, Gat. anmasti, MS. anmast, pethaps composed of an interrogative ma- and sain). The Clement. sant itself would Seem to be a prinary nominal derivative of the root win 'linit'. There is no nocd to assume, as Praetorius (1879:129) does, that Anh. sunt is a development of an earlier 23s fonts Sunt could merely be the noun without any prefixed interrogative. tai "betow'*! + Gz. tajiti, Tha. tabed, Te. tihat, Har. tahay, Caf. taX8i, ChiGy. tite, Sod. tati, Wi. tat,’etc. Common Semitic ht. The palatalization in Anh. ta and elsewhere derives from tajite-, as occurring in Ge'éz before pronoun suffixes. tonant(onna) ‘yesterday’ Gz. tomazom , Tha, tanaki, Arg. tomay, Gaf. totam( anna), Ghicy. tanami, Go. tandfiia, etc. An altemative form Zotant(onna) also occurs in Amharic. Amharic tonant, etc. , is the only Eth.Sem, form with a suffix -t; both nt and metathesized 2ln occur’ in Ethiopian Semitic. The suffix -(on}na also occurs on ama "last year’. Common Semitic tml: cf. Hob. 2@ngt~ *etmgl, Aran, 2°miley ~ *Ltmdloy, kk. tintli~ itinali. wid "towards" 1 Gz. wil?odd ~ wo%adé 'next to, by the side of", Arg. uiidi ‘towards'. ‘This is composed of the conjunction wi and an adverbial accusative of *ed 'hand'. Similar constructions employing the noun ‘hand’ occur in other Semitic languages; cf. especially ‘Ak. ida 'by the side of", Soq. *éd "towards, into’, Sheri id. wwiyom(m) ~ wiiyas(s} ‘or’ Tha, wliyim~ wily, Arg. wem, Gaf.Ch. wiiy, Sod. wiiyad, S1.2. we, etc. This ig derived from something akin to Gz. i used in alternate questions. The form wimérmd > wiiyam(n) has been analysed as wiy + conjunctive enclitic -n(m) , hence the substitution of -8(8) in some forms, and the dropping of -m(m) in others. “Also hat from *ka’, exhibiting the k/t interchange found elsewhere in Atharic, AL 5, 10564 D. L. Appleyard [AAL 5/2 yet 'where?" + Gz. *ayte, Tha. ayté, Te. “aya, Arg. yed, M. ott, Ch. ete. An. yet and Arg. yed mist derive from **ayt) as the Tigrinya form, The root is the conmon Semitic interrogative 2y, to which various suffixes are added to form the interrogative adverb: cf. Ar. °ayna, Heb. °2~ °Bka, Aram. “ayht, AKK. ayutinu, zane 'today' This appears to have no cognates elsewhere in Ethiopian Senitic, or in Semitic in general. Practorius (1879:57, 169) ingeniously suggested a derivation from a compound of the demonstrative zd + *obne(t] (Gz. *vbret "alternation, turn, period of office’) and com- pared Ar. at-mutabariyant 'day and night' as a derivative of the same Semitic root, bry, with a specialized temporal sense. A form like *z@?ubxe could, indeed, develop into Amh. zaxe. The N. Ethiopian languages preserve the conmon Semitic term for’'day' here: Gz.Te. Tna. Lomi ~ Lom. Mogt of the S, Ethiopian languages use items cognate with Anh. ahut except for liar. hogi, S1.W1. age, Z. auf, which may be cognate with Tna, fto2i ~ REFERENCES Bateman, J. 1929. Déctionnaire anariana- frangais. Diré-Daoua: Inprinerie Saint Lazare des RR. PP. Capucins. Barth, J. 1893. EtmoLogische Studien zum senitischen énsbesondenre zum hobadischen Lexicon. Leipzig: J.C. Hinrichs'sche Buchhandlung. Brockelnann, C. 1928. Lexicon syriacum, 2nd edition. Halis Saxom Niemeyer. Cerulli, E. 1931-2. "Canti burleschi di studenti delle scuole abissine." Révista degté ‘studi ondentaké 13:342-50. + 1936, La Lingua e ta storia dé Hanan. Studi etiopici I. Roma: Istituto per rorionte. 1938. La Lingua ¢ ta storia ded Sidano. Stuti etiopici 11, Roma: Istituto per 1'Oriente. Cohen, M. 1931. Etudes d'éthiopien méridionag. Paris: Libraire orientaliste Paul Geuthner. 1933, "Gy?, ge, etc. 'vallée, pays’. aigtudes’ chanito-sémitiques 1:34. 1939. Nouvettes Etudes d'éthiopien méridionat. Bibliothéque de 1'Ecole des Hautes Etudes, fasc. 275. Paris: Libraire ancienne Honoré Champion. + 1951-4, "Une dénomination conmme de 1'ane et de 1a surdité en chamito- sénitique.” Comptes rendus du Groupe Linguistique d'études chamito- Bnitiques 6215-16. - 1969, Essai comparatif sur Le vocabutaine et La phonétique du chamito- ‘Sénitique. Paris: Libraire Honoré Chanpion. (Reprint of 1947 edition.| ‘Sumptibus Max "Comptes rendus du Groupe Linguistique AAL 5, 1001977} A Comparative Approach to Amharic Lexicon 65 Conti Rossini, C. 1912. La Langue des Kemant on Abyssinie, Kalserliche Akademie der ‘Wissenschaften. Schriften der Sprachkommission, Bd. IV. Wien: in Konmission bei’ Alfred Hélder. + 1928. La storia d'Ethiopia, Africa Italiana, Collezione di monografie a cura del Ministero delle colonie, Vol. TIT: Bergano: Istituto Italiano d'arti grafiche. Dilimann, A. 1865. Lexicon Linguae acthiopicae. Lipsiae: T.0, Weigel. Dolgopol 'skiy, A.B. 1973, Sraunitel’no-éstonicheskaya fonetika kushitskikh yazykov. Moskva: "Nauka". Drewes, A.J. 1962. Inscriptions de L'Ethéopie antique. Leiden: E.J. Brill. Fleisch, H. 1968. L'arabe classique: esquisse d'une structure Linguistique. Recherches publiges sous 1a direction de 1'institut de lettres orientales de Beyrouth, Série 2: langue et littérature arabes, Tone 5. Beyrouth: Dar el-Machreq Editeurs. Fronzaroli, P. 19644, "Studi sul lessico comme semitico, 1, Introduzione.” Rendiconté dott ‘Accademia Nazionate dei Linced, Chasse di Scienze monati, stoniche ¢ Gilotogiche, VIII. 19:155-72. + 1964. "Studi sul Lessico comme semitico, II. L'uomo e 1'et." RAL, VIII. 19:245-80., 196Sa. "Studi sul lessico comme semitico, III. I fenomani naturali." RANL, VIII, 20:135-50. + 1965b. "Studi sul lessico comme semitio, IV. La religione." RANL, VIII. 20:299-69.. 1968. "Studi sul lessico comme semitico, V. La natura selvatica." RANL, VIII. 25:267-303. . 1969. "Studi sul lessico comme semitico, VI. La natura domestica.” RANL, VIII. 24:285-320. + 1971, "Studi sul lessico comme semitico, VI. L'alimentazione." RANL, VIII. 20:603-42, Gamst, F. 1969. The Qemant: A Pagan-Hebraic Peasantry of Ethiopia, Case Studies in Cultural Anthropology. New York: Holt, Rinehart and Winston. Gankin, E.B. 1969. Ankharsho-russkiy s£ovan". Moskva: Sovetskaya Entsiklopediya. Guidi, I. 1901. Vocabotario anarico-itatiano. Rona: Casa editrice italiana. [Reprinted in 1953 by Istituto per 1'Oriente, Rona}. Hetzron, R. 1972, Ethiopian Semitic: Studies in Classigication. Journal of Semitic Studées, Monograph no 2. Manchester: Manchester University Press. Winer, M. 1960. “Uber sprachliche und kulturelle Bezichungen 2\ischen Sidarabien nd Kthiopien in Altertun." Atté del convegno internazionage di. studi etiopici, Accademia nazdonate ded Linced, anno 357, p. 435-45. Howmel, F. 1879. Die Namen dex Séugethiere bei den siidsenctischen Vétkern. Leipzig: J.C. Hinrichs'sche Buchhandlung. Hymes, D. 1960.. “Lexicostatistics so far." Cusnent Anthtopology 1:5-44. Krotkoff, 6. 1969. "Lahn "Fleisch und leben "Brot'." Wiener Zectachrige gi dée Kunde des Mongentandes 62:76-82. AAL 5, 10766 D. L. Appleyard [AAL 5/2 Leslau, W. 1938. Lexéque Soqopré (sudarabique moderne). Collection de 1a Société de linguistique de Paris, no. 41. Paris: Libraire C. Klincksieck. 1949. "Notes de gramaire et d'Stymologie éthiopienne." ond $:273-79. + 1956. Etude descriptive et comparative du Gafat. Collection de 1a Société de Linguistique de Paris, no $7, Paris: Libraire C. Klincksieck. 19S7a. “Arabic loawords in Amharic." Bugletin of the School of Oriental and African Studies 19:221-44. 1987. "Some mitilated roots in Ethiopic." Lingua 6:268-86. + 1963. An Etymologicat Dictionary of Harari. University of California Near Lastem Studies, Vol. 1. Los Ageles: University of California Press. Littmann, E, 1913. Dée deutsche Aum-Expedition, Vol. 4. Berlin: Koenigliche Museen, Verlag von Georg Reiner. Ludolf, H. 1698. Lexécon anharico-Zatinum cum indice Latino copioso inquirendis voeabutis amharicis in hoc opere contentis, Francofurti ad Moenum: Prostat apud Johannem David Zunnerun, Mittwoch, E. 1911. "Dschanhoi - die anharische Bezeichnng fiir 'Majestit'." Zectachrige Git Assyriotogéc 25:281-86, Mooney, M.F. 1963. A Glossary 0f Ethiopian PLant Names. Dublin: Dublin University Press. Néldeke, T. 1910. Neue Beétniige zur semitischen Sprachwissenschagt. Strassburg: Karl J. Tribner. Plazikowski-Brauner, H. 1913, Ein dthiopisch-amharisches Glossar (Sauisew}. Berlin. Praetorius, F. 1873. "Zur thiopisch-arabischen Gramatik." Zeitsolnigt den deutschen mongentiindischen Gesettschagt 27:639-44. 1879. Dée amharische Sprache. Halle: Verlag der Buchhandlung des Waisenhauses. + 1890, "Beitriige aur athiopischen Gramatik und Etymologie.” Beétndge zur Assyriotogie und vergleichenden semtischen Sprachwissenschagt 1:21-97, 369-278, Reinisch, L. 1887. Die Biin-Sprache, Bd.2, Wéaterbuch. Wien: Alfred Hélder. Rundgren, F. 1953. "The root 34t in the modern Ethiopic languages (Tigré, Tigrifia and Anharic) and Old Exyptian {fty, Coptic It." Onientatéa Suceana 2:19-21. Simoons, FJ. 1970. "Economic prehistory of Ethiopia." In Papers in Ajnican Prehistory, ed. J.D. Fage and R.A. Oliver. Cambridge: Canbridge University Press. Swadesh, M. 1952. "Lexicostatistic dating of prehistoric ethnic contacts." Proceedings of the Anonican Philosophical Society 96:152-65. da Thiene, G. 1939. Décionario defla Lingua gatea, Hlarar: Vicariato Apostolico. Tubiana, J. 1951-4. "Les noms de parenté en anharique." Comptes rendu du Groupe Lin- guistique d'études chanito-sémitiques 6:51. _ + 1965, "A propos, de Ianharique andan."” Comptes nendus du Groupe Lingucstique d'études chanito-sénitiques 7. 1970. "A propos de 1'anharique niigén gan." In NéLanges Marcel Cohen, ed. D. Cohen. Paris: Mouton. Ullendorff, E, 1955. "The Semitic Languages of Ethiopia and their contribution to general Semitic studies.” Afréca 25:154-60. AL 5, 1081977] A Companative Approach to Amharic Lexicon 67 + 1956. "the contribution of South Semitic to llebrew Lexicography." Vetus Testanentun 6:190-98, 1968. Ethiopia and the Bible. The Schweich lectures of the British ‘Academy. London: Oxford University Press, Wajnberg, I. 1937a. "Abessinische Etymologien." Roczndk otdentatistyczny 15:24-41. + 1937. "Dualreste und Dualspuren im Neusbessinischen." Rocenik ordentag- dstyceny 13:19-23. AL 5, 109