African language studies

pdf.pdf

Collected Papers in Oriental and African Studies
AFRICAN LANGUAGE
STUDIES
IX
School of Oriental and African Studies University of London
[Sole Agents: Luzac & Company, Ltd., 46 Gt. Russell St., London, W.C. 1] 1968


1
1


COLLECTED PAPERS IN ORIENTAL AND AFRICAN STUDIES
AFRICAN LANGUAGE STUDIES IX
Editor:
Malcolm Guthrie
Assistant Editors:
E. C. Rowlands F. D. D. Winston




Collected Papers in Oriental and African Studies
AFRICAN LANGUAGE
STUDIES
IX
School of Oriental and African Studies University of London
[Sole Agents: Luzac & Company, Ltd., 46 Gt. Russell St., London, W.C. 1 ]
1968


© School of Oriental and African Studies University of London


CONTENTS
PAGE
Inflectional characteristics of the so-called ‘weak verbs’ in Somali, by B. W. Andrzejewski ... ........................1
The Hamziya deciphered, by Jan Knappert . 52
Nandi riddles from north-eastern Congo, by Lyndon Harries . 82
‘ The song of the rains ’: a Hausa poem by Na’ibi S. Wali, by D. W. Arnott...................................................................120
‘ The song of the rains ’ : metric values in performance, by A. V. King and Malam Rashid Ibrahim.................................................148
The indigenous scripts of West Africa and Surinam : their inspiration and design, by David Dalby . . . . . . . .156




INFLECTIONAL CHARACTERISTICS OF THE SO-CALLED ‘WEAK VERBS’ IN SOMALI By B. W. Andrzejewski
Essential Simplicity of Structure
As presented in the existing literature the structure of the Somali verb gives the impression of great complexity and irregularity of patterns. Yet the structure of the verb in this language is essentially simple and can be reduced to comparatively few concise formulations.
The so-called weak verbs,1 the most common verbal type, appear on the surface to be particularly complex, since their inflectional patterns are masked, as it were, by series of alternances. In these verbs, moreover, there are affixes which, though often identical in shape, have different grammatical functions and are subject to different distribution rules.
In this article an attempt is made to provide formulations which are much more economic than the existing ones and which at the same time throw into relief the structural framework ; this is quite straightforward in its characteristics once the alternating elements of the components are generalized and encoded. The new approach presented here appears to be fruitful also in some other Cushitic languages, such as Galla or ‘Afar (Danakil), where numerous alternances in the affixes obscure the inflectional patterns which these languages share with Somali.2
Terminology
As the number of works on the grammatical structure of Cushitic languages, including Somali, is fairly limited, no uniform grammatical terminology has been established, and anyone working in this field is forced to define his terms or to state whose terminology he is applying. For present purposes I have taken the terms of C. R. V. Bell as points of reference, only amplifying or departing from them when necessary.
The term ‘ weak verbs ’ applies to all verbs of Bell’s 1st, 2nd and 3rd conjugations, with all their subdivisions. These cover all the verbs of the language except for five irregular ‘ strong verbs ’ characterized by Semitic-like prefixes, and for Bell’s 4th conjugation, which consists of combinations of the strong verb ygthay
1 The term ‘ Starke und schwache verba ’ (strong and weak verbs) was introduced by Reinisch in his Die Somali Sprache. For details of all works referred to in this article see Bibliography, p. 50. For further works on Somali see John W. Johnson, A bibliography of Somali language materials.
2 For a detailed discussion of patterns of verbal conjugation in Cushitic languages and the whole Erythraic group, see the works by Tucker and Bryan listed in the Bibliography.


2
B. W. ANDRZEJEWSKI
‘ to be ’ with what have so far been regarded as adjectives.3 In this article all the adjectives which are capable of occurrence in such combinations are regarded as verbs, and jointly with the verbs of the 4th conjugation are referred to as hybrid verbs.
Tenses and moods of weak verbs, which I subsume under the term ‘ tense paradigms ’, are listed in Table I. The first column gives the name used in this article in full, and the second gives the abbreviated form. The third column shows the corresponding terms employed by C. R. V. Bell.
The most important departure from Bell’s terminology is the division of all verbal forms into ‘ main ’ and ‘ dependent ’. Apart from differences in inflectional characteristics and accentual patterns, this division is necessary on syntactic grounds : for an account of these, readers are referred to the sections ‘ Comparison between noun forms and nominal clusters ’ in The Declensions of Somali nouns, pp. 92-97, and ‘ Relative sentences containing a verb ’ in R. C. Abraham’s Somali-English dictionary, pp. 297-304, and also to the section ‘ Subsidiary agreement signs ’ in the introduction to Hikmad Soomaali, pp. 24-28.4
Table I
List of Tense Paradigms
Imperative Present general extensive lmper. Pres. gen. ext. Imperative Present general
£ Present continuous extensive Pres. cnt. ext. Present continuous
Present general restrictive Pres. gen. res. Present general
c3 Present continuous restrictive Pres. cnt. res. Present continuous
C3 CL Past general extensive Past gen. ext. Past general
o Past continuous extensive Past cnt. ext. Past continuous
Past general restrictive Past gen. res. Past general
"cZ Past continuous restrictive Past cnt. res. Past continuous
kJ CL Past independent Past indep.
.g Optative Optat. Subjunctive general
*c3 Rhetorical Rhet.
Potential Poten. Potential
Infinitive Inf. Infinitive
Negative imperative Neg. imper. Negative imperative
£ Negative present general Neg. pres. gen. Negative present general
c3 £ Negative present continuous variable Neg. pres. cnt. var.
Off 00 Negative present-past general Neg. pres.-past gen. Negative past general
s« Negative present continuous invariable Neg. pres. cnt. inv.
•S c3 Negative past continuous Neg. past cnt. Negative past continuous
1 Negative optative Neg. opt. Negative subjunctive
Negative conditional Neg. cond. Negative potential
3 Such ‘ adjectives ’ occupy the same positions in the sentence as the forms of the restrictive and convergent paradigms of weak verbs.
4 These syntactic reasons could be summarized as follows : unlike main verbal forms, dependent verbal forms are capable of bearing case signs (agreement signs) which are, figuratively speaking, ‘ displaced ’ from the nouns on which they are dependent.


3
CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
Present general divergent A Present continuous divergent A Pres. gen. dvg. A Pres. cnt. dvg. A Subjunctive general Subjunctive continuous
CZ5 E Present general divergent B Pres. gen. dvg. B Present general
Present continuous divergent B Pres. cnt. dvg. B Present continuous
73 Past general divergent A Past gen. dvg. A Past general
Ui OJ Past continuous divergent A Past cnt. dvg. A Past continuous
ex Past general divergent B Past gen. dvg. B Past general
_> Past continuous divergent B Past cnt. dvg. B Past continuous
So Present general convergent A Pres. gen. cvg. A Present general
O ex Present continuous convergent A Pres. cnt. cvg. A Present continuous
Present general convergent B Pres. gen. cvg. B Present general
c Past general convergent A Past gen. cvg. A Past general
ex Past continuous convergent A Past cnt. cvg. A Past continuous
czo Negative present-past general Neg. pres.-past gen. Negative past general and
£ .9? 43 aj J-l CtJ ex dependent A Negative present-past general dependent B dep. A Neg. pres.-past gen. dep. B negative subjunctive general
c3 OB G dependent A Negative present continuous dependent B A Neg. pres. cnt. dep. B continuous
G A Negative conditional dependent Neg. cond. dep. Negative potential
Division into extensive and restrictive tenses is based on both morphological and syntactic differences, which are explained in the ‘ Note on verbal forms ’ in The Declensions of Somali nouns, pp. 128-137,5 and in my article ‘Accentual patterns in verbal forms ... pp. 125-126.
The nature of divergent and convergent paradigms is described in the last-mentioned Note and in ‘ Notes on the substantive pronouns in Somali ’, pp. 88-93, where, however, instead of ‘ convergent ’ the term ‘ construct ’ is used. Readers should note that whatever has been said in these two publications about construct paradigms applies here to convergent paradigms.
The need for dividing dependent forms into A and B is explained in the publications previously referred to in connection with the division into main and dependent, and divergent and convergent paradigms.
6 To the definition of forms of restrictive paradigms given in that publication one should add that they also occur when they agree with the interrogative pronoun yaa ‘ who ? ’.


4
B. W. ANDRZEJEWSKI
For a description of the past indep., optat., rhet.,6 neg. pres. cnt. inv., neg. optat. and neg. cond. paradigms see ‘ Accentual patterns in verbal forms . . . ’, pp. 126-128. The rhet. paradigm is also described in note 8/6-7 in Hikmad Soomaali, pp. 76-77.
The neg. pres. var. is a continuous equivalent of the neg. pres, gen., and rarely occurs in Northern Somali.
The term neg. pres.-past gen. is used instead of negative past general since this tense, when preceded by the indicator baa (ayaa, yaa, iyaa) or miyaa combined with the indicator aan, can have the same time reference as either the present or the past tense, and only the context can remove ambiguity.
E.g. Lac&g b&anu hays&n.
fHe does not possess money.
[He did not possess money.
The same applies to the neg. pres.-past gen. dep. (which is always preceded by the negative indicator aan).
In all the tense paradigms the division into persons is based on the criterion of concord with substantive pronouns, which are discussed at some length in my article ‘ Notes on the substantive pronouns in Somali ’.
Bell’s division of verbs into conjugational types has been amplified here by arrangement into root extension classes. This is to a large extent a matter of difference in terminology, since Bell’s conjugational types are based on the characteristics of the stems, thus incorporating root extensions in his classification.
The grouping of verbs into root extension classes largely corresponds to the classification of‘ stems ’ given by Reinisch and Moreno. A different approach is adopted by Abraham, who uses root extensions and terminations, taken jointly, as a basis of his classification, and gives 3 sg. m. pres. cnt. ext. forms as headings of lexical entries.
In Table II each extension class is formally identified by code letters which represent all the alternants of each extension set out in full on pp. 27-29. Groups of code letters divided by an oblique stroke or strokes indicate multiple extensions, i.e. those which are regarded as combinations of two or more extensions. The code letters marked with + represent root extensions of hybrid verbs which occur as components of multiple extensions in weak verbs.7 A brief summary
6 The rhetorical paradigm was referred to in that article as ‘ negative rhetorical From the point of view of its structure this is not a negative tense since it is preceded by the indicator ma (interrogative) and not by the indicator ma (negative). These two indicators can be distinguished by their relative position in the sentence. The interrogative ma precedes prepositional particles, while the negative ma follows them. For further information see Kirk’s grammar (p. 117), Zholkovsky’s excellent article on the sequences of preverbal particles and my article on pronominal and prepositional particles.
7 A difficult notational problem is presented by the hybrid verb extension class Z+ ; this corresponds to Bell’s ‘ radical adjectives ’ (which he describes on p. 76 of his book as ‘ those which are not formed from any other words ’) together with their combinations with the verb y^hay ‘to be ’ within his 4th verbal conjugation. In this class the extension consists of the absence of affix (affix zero) and thus in the code letters in the list of root extensions given in this


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
5
of the notional associations of each extension class is given in the second column8 to facilitate its identification and in the third column Bell’s conjugational types are given for comparison.
Some root extension classes, such as Z, IN, AYN, AN, SAN, AYN/SAN, AN+/AAN and SAN+/AAN, contain a very large number of verbs, and it is to these classes that the majority of newly borrowed or newly coined words are assigned in the language. Other classes have fewer numbers, and in several classes only two or three examples have been found. Since, however, it seems desirable to place every verb in a particular class, the small extension classes are listed together with the large ones in order that any future lexicographical work may have a grid of reference as complete as possible in this respect, even though, clearly, extension classes with a small number of members have less importance in the system of the language than those with large ones.
Table II
List of Root Extension Classes
z No specific notional associations. 1
AM ‘ Agentless ’ passive9 of the corresponding verbs of class Z, combined with the notion of ‘ becoming i.e. achieving a change into a particular state, e.g. bderan ‘ to “ become ” cultivated ’ (said of land), cf. bder (Z) ‘ to cultivate ’.10
SAM ‘ Agentless ’ passive of the corresponding verbs of class IN, combined with the notion of ‘ becoming ’, e.g. bquhsan ‘ to “ become ” filled ’ cf. bquhi (IN) ‘ to fill ’. 1 ‘ Static ’ (§ 136)
SOOM Two examples only : doorsdon ‘to “ become ” changed in appearance or character ’, cf. d$ori (IN) ‘ to change into ’ (trans.). qarsdon ‘ to “ become ” hidden ’, cf. qdri (IN) ‘ to hide ’ (trans.).
OOB To become or to turn into what is denoted by the corresponding nominal, e.g. tqugdw ‘ to become a thief ’, cf. tqug (masc. noun) ‘ thief ’ ; dqmbasdw ‘ to turn into ashes ’, cf. dambas (masc. noun) ‘ ash ; ashes ’. IB
AAB Two examples only : ku m$gacdw ‘ to name ’, cf. mdgac (masc. noun) name, hallow ‘ to get lost ’, cf. halide (AYN) ‘ to lose ’.
article Z+ merely indicates a derivational relationship. On account of this there is no difference in the actual phonetic realizations of the extensions : AYN/SAM and Zf/AYN/SAM ; AYN and Z+/AYN ; AYN/SIIN and Z*/AYN/SIIN ; AYN/SAN and Z+/AYN/ SAN.
8 The notional associations of the extension classes of hybrid verbs are left out of the present account.
8 I.e. a passive which cannot be accompanied by a nominal or by a pronoun denoting an agent by whose activity the particular state is achieved. Note that in Somali there are no sentences which would structurally correspond to the English ‘ he was rescued by a fisherman ’.
10 For the grammatical status of the verbal forms used here as examples see section ‘ Exemplification ’, p. 10.


6
B. W. ANDRZEJEWSKI
TAM Reciprocal activity whose nature is suggested by the corre- sponding noun, e.g. ultan ‘ to fight one another with sticks cf. ul (fem. noun) ‘ stick ’ ; ddotan ‘ to dispute with one another cf. ddod (fem. noun) ‘ dispute ’. 1 ‘ Reciprocal ’ (§ 137)
R-AM Three examples only : Id dagaallan ‘ to fight against cf. dagdal (masc. noun) ‘ fighting, battle ’. Id gdrran ‘ to litigate with ’, cf. gdr (fem. noun) ‘ justice, lawsuit ’. wdrran ‘ to tell news ’, cf. wdr (masc. noun) ‘ news ’.
AYN/SAM ‘ Agentless ’ passive of the corresponding verbs of class AYN, combined with the notion of ‘ becoming ’, e.g. shaancjdysan ‘to “become” sieved’, cf. shaandde (AYN) ‘ to sieve ’ and shdan^o (fem. noun) ‘ sieve ’. Static (§ 136)
Z+/AYN/SAM ‘ Agentless ’ passive of the corresponding verbs of class Z+/AYN, combined with the notion of ‘ becoming ’, cf. samdysan ‘ to “ become ” made ’ cf. samde (Z+/AYN) ‘ to make ’ and sdn (Z+) ‘ to be good ’.
IN (a) Transitive of the corresponding intransitive verbs of classes Z and AN, e.g. c|aldali ‘ to melt ’ (trans.), cf. (jaldal (Z) ‘ to melt ’ (intrans.) ; ddadi ‘ to spill ’ (trans.), cf. daadd (AN) ‘ to spill ’ (intrans.). (b) Causative of the corresponding non-causative verbs of classes Z and AN, e.g. ddaji ‘ to cause to graze ’, cf. ddaq (Z) ‘ to graze ’ ; s$ehi ‘ to cause to go to sleep ’, cf. seefcd (AN) ‘ to go to sleep ’. (c) To perform an activity whose nature is suggested by the corresponding noun, e.g. qurhi ‘ to make beautiful, to adorn ’, cf. quruh (fem. noun) ‘ beauty ’. 2 ‘ Causative ’ (§§ 10 and 133)
AYN (a) To perform an activity whose nature is suggested by the meaning of the corresponding noun, e.g. caande ‘ to put milk in ’, cf. caano (masc. noun) ‘ milk ’ ; kabde ‘ to hit with a shoe ’, cf. kdb (fem. noun) ‘ shoe ’, herde ‘ to put into an enclosure ’, cf. hdro (fem. noun) ‘ enclosure ’. (b) To be in that position in time or space which is described by a corresponding attributive, e.g. sokde ‘ to be on this side (of something) ’, cf. sokd (attrib.) ‘ on this side ’ ; dambde ‘ to be behind ’, cf. dambd (attrib.) ‘ behind ’. 2 ‘ Applicative ’ (§§ 10 and 133)
Z+/AYN To cause someone or something to acquire the characteristics described by corresponding hybrid verbs, e.g. casde ‘ to cause to be red ’, cf. cds (Z+) ‘ to be red ’.
SUN To make someone (or, rarely, something) perform the activity or achieve a change into a state described by the corresponding verb of class Z or, rarely, AN, e.g. cunsli ‘ make someone eat ’, cf. cun (Z) ‘ to eat ’, dabbaalsli ‘ to make someone swim, to make something float ’, cf. dabbaald (AN) ‘ to swim, to float ’.


7
CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
SIN One example only : facjiisi ‘ to make a person or animal sit cf. (z) ‘to sit, to stay
AAB/IN One example only : hall&abi ‘ to cause to get lost cf. hall&w (AAB) ‘ to get lost
TAM/SIIN Combination of notional associations of classes TAM and SUN, e.g. ultansli ‘ to make people fight one another with sticks cf. ultan (TAM) ‘ to fight one another with sticks ’.
IN/SIIN1 One example only : g§linsli ‘ to explain, to lead into, to induce ’, cf. g£li (IN) ‘ to put in ’.
IN/SIIN2 One example only : heshiisli ‘ to make people agree with one another ’, cf. heshii (IN) ‘ to agree with one another ’.
AYN/SIIN Combination of notional associations of classes AYN and SUN, e.g. dambaysli ‘ to cause someone to be behind ’, cf. dambde (AYN) ‘ to be behind ’.
Z+/AYN/SIIN Combination of notional associations of classes Z+/AYN and SUN, e.g. jeTaysii ‘ to make someone like or love someone or something ’, cf. jec61 (Z+) ‘ to like, to love, to be fond of ’.
AN/SIIN1 One example only : dooransli ‘ to make someone choose ’, cf. doord (AN) ‘ to choose ’.
AN/SIIN2 One example only : qabadsii ‘ to make someone catch or take, to make someone go to a place ’, cf. qabd (AN) ‘ to catch, to take, to go to a place ’.
ON/SIIN One example only : socodsii ‘ to set in motion ’, cf. socd (ON) ‘ to be in motion, to walk ’.
Z+/OON+/ One example only :
AYN/SIIN cusboonaysli ‘ to renew, to renovate ’, cf. cusub (Z+) ‘ to be new ’.
AN (a) To perform for one’s own benefit an activity described by the corresponding verb of class Z, e.g. cjisd ‘ to build for oneself ’, cf. cps (Z) ‘ to build ’. (b) Intransitive of the corresponding transitive verbs of class Z and IN, e.g. waald ‘ to become mad ’, cf.^ w&al (Z) ‘ to madden ’ ; gqbd ‘ to burn ’ (intrans.), cf. gqb (Z) ‘ to burn ’ (trans.). (c) Non-causative of the corresponding causative verbs of class IN, e.g. tykd ‘ to pray ’, cf. tqji (IN) ‘ to cause to pray, to lead in prayers ’. |


8
B. W. ANDRZEJEWSKI
(d) To perform an activity whose nature is suggested by the meaning of the corresponding noun, e.g. canaan6 ‘ to reprove, to criticize ’, cf. canaan (fern, noun) 1 reproof, criticism (e) Reflexive of the corresponding verb of class Z, e.g. maycjd ‘ to wash oneself ’, cf. majr^ (Z) ‘ to wash
ON Two examples only : noqd ‘ to become ’ (with the particle ku and/or sdo ‘ to return ’), cf. soo-ndqod (masc. noun) ‘ return soc6 ‘ to be in motion, to walk ’, cf. s6cod (masc. noun) ‘ movement, walking 3
SAN (a) To perform for one’s own benefit an activity described by the corresponding verb in class IN, e.g. byufcsd ‘ to fill for oneself ’, cf. b$uhi (IN) ‘ to fill (b) To assume a posture or to achieve a state described by the corresponding verb of class Z, e.g. kqududsd ‘ to assume a crouching position ’, cf. kqudud (Z) ‘ to be in a crouching position ’; hubsd ‘ to become certain, to ascertain ’, cf. hub (Z) ‘ to be certain (c) To perform an activity whose nature is suggested by the meaning of the corresponding noun, e.g. h^beensd ‘ to spend a night away from home ’, cf. h^bden (masc. noun) ‘ night ’. (d) Non-causative of the corresponding causative verb of class IN, e.g. hanuunsd ‘ to experience pain ’, cf. hanuuji (IN) ‘ to cause pain ’. (e) Reflexive of the corresponding verb of class IN, e.g. $ngegs6 ‘ to dry oneself ’, cf. $ngdji (IN) ‘ to dry ’ (trans.).
OON (a) To have bodily sensations, emotions or thoughts whose nature is suggested by the meaning of the corresponding noun, (jahamdod ‘ to feel cold ’, cf. t(ah&n (fem. noun) ‘ cold ’. (d) To perform for one’s own benefit an activity described by the corresponding verb of class Z, e.g. §rydod ‘ to chase for oneself ’, cf. £ri (Z) ‘ to chase ’. 3B
SHOON Two examples only : hawshdod ‘ to toil, to work hard ’, cf. hdwl (fem. noun) ‘ toil, hard work ’. hishdod ‘ to be ashamed, to fear shame ’, cf. hil (masc. noun) ‘ sense of shame, sense of decorum ’.
TOON One example only : cjibtdod ‘ to experience trouble or difficulties ’, cf. c|ib (fem. noun) ‘ trouble, difficulty ’.
IN/SAN Combination of the notional associations of classes IN and SAN, e.g. §ngejisd ‘ to dry for oneself ’ (trans.), cf. §ngdji (IN) ‘ to dry ’ (trans.), £ngeg (Z) ‘ to dry ’ (intrans.).


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
9
AYN/SAN Combination of notional associations of class AYN (heading (a)) with that of class SAN (heading (a)), e.g. caanays6 ‘ to put milk in for oneself cf. caan6e ‘ to put milk in
z+/ayn/san Combination of notional associations of class Z+/AYN with that of class SAN (heading (a)), e.g. casays6 ‘ to cause to be red for oneself cf. cas6e (Z+/AYN) ‘ to cause to be red 3
SIIN/SAN Combination of notional associations of class SUN with that of class SAN (heading (a)), e.g. cunsiis6 ‘ to make someone eat in expectation of some later benefit (e.g. gratitude) cf. cunsli (SUN) ‘ to make someone eat
TAM/SIIN/ SAN Combination of the notional associations of classes TAM, SUN and of class SAN (heading (a)), e.g. ultansiisd ‘ to make people fight one another with sticks in expectation of deriving some benefit from this cf. ultansii (TAM/SIIN) ‘ to make people fight one another with sticks cf. ultan (TAM) ‘ to fight one another with sticks
Z+1 AN+ R-AN+ SAN+ OON+ TOON+ VAAN To acquire the characteristics or achieve the state described by the corresponding hybrid verbs, e.g. casdw (Z+/AAN) ‘ to become red ’, cf. cds (Z+)1 to be red ’. ladn&w (AN+/AAN) ‘ to get well ’, cf. laddn (AN+) ‘ to be well ’. d§ggandw (R-AN+/AAN) ‘ to take up residence, to settle or encamp ’, cf. d§ggan (R-AN+) ‘ to reside, to be settled or encamped ’. w^naagsan&w (SAN+/AAN) ‘ to become good ’, cf. wjpaags&n (SAN+) ‘ to be good ’. gydboondw (OON+/AAN) ‘ to become appropriate or aligned ’, cf. g^dbdon (OON+) ‘ to be appropriate or aligned ’. djgtoon&w (TOON+/AAN) ‘ to become alert ’, cf. djgtdon (TOON+) ‘ to be alert ’. 3A
Transcription and Territorial Limits
The transcription used is the same as in Hikmad Soomaali and The declensions of Somali nouns.
The descriptive statements given here apply in detail to Somali as spoken in the Northern Region of the Somali Republic (i.e. Hargeisa and Burao governorates). With some minor modifications, however, they could be extended to the dialects of the group described as Common Somali.11 The method of analysis could, in my view, be profitably applied to the remaining Somali dialects.
11 For a note on the grouping of dialects see Andrzejewski and Lewis, Somali poetry: An introduction, 37-38.
B


10
B. W. ANDRZEJEWSKI
Exemplification
Except for the illustrative text which forms the last section of this article, the examples of verbal forms are given within the framework of very short sentences, most of which contain components other than verbs, such as indicators and pronouns. It is important to note that verbal forms, with the exception of those of the imper. and past indep., cannot constitute by themselves a meaningful utterance in Somali.
It often happens that several English translations of the same Somali sentence are equally possible. In particular the absence of an object pronoun particle after a transitive or causative verb in Somali may correspond to English pronouns ‘ him ’, ‘ her ’, ‘ it ’ or ‘ them ’ and the absence of a preverbal subject pronoun in Somali before certain negative verbal pronouns may correspond to English ‘ I ’, ‘ you ’, ‘ he ’, ‘ she ’, ‘ we ’, or ‘ they ’, unless a larger context eliminates such ambiguity.
In the examples in this article only one translation is given, even if other translations are equally possible, since such ambiguities have no bearing on the main theme under discussion.
When whole verbs, in all their forms, are referred to, the 2 sg. imper. is employed as the representative form (lexical entry form) as in the vocabulary notes in Hikmad Soomaali. This applies to all weak verbs, but hybrid verbs are quoted in their pres. gen. res. form, e.g. wspaags&n ‘ to be good ’. All representative forms are followed by their root extension code letters in brackets and are translated into the infinitive in English.
It should be noted that in some weak verbs the 2 sg. imper. forms occur only in unusual contexts. This applies in particular to verbs which are associated with the notion of ‘ becoming ’. For example, a form like w§yndw ! ‘ become (grow) big ! ’ is almost exclusively used as a blessing addressed to children, when thanking them for something. Even less usual is the form bjydw ! ‘ become (turn into) water ! It was, however, commonly used by the Dervishes and was addressed to enemy bullets. This reflected their belief that a pious man could be granted such a miracle during a holy war. In poetry the 2 sg. and 2 pi. imper. forms are sometimes applied to inanimate objects.
Structure of the Weak Verb
All weak verbs have an identical structure and can be analysed into three morphemic components which always occur in the same order : root—extension— termination. While the root is present in every verbal form, some forms occur without an extension or termination. Their absence, however, has a function and


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
11
will be referred to as extension or termination zero, represented by the sign #. How these three components are identified will be seen from the following groups of examples :
Dis !
Wuu cfisay. W&y cfistay. Wuu (Jisayaa, M& 4isin-
Build it !
He built it.
She built it.
He is building it. He did not build it.
All the above verbal forms share the same root c|is but differ in their terminations -#, -ay, -tay, -ayaa and -in. They share these, however, with other verbs, e.g.
Wuu cunay. W&y cuntay. Wuu cunayaa.
cunin.
Eat it !
He ate it.
She ate it.
He is eating it. He did not eat it.
The forms of the verb Disd !
Wiiu cfistay. W&y cjisatay. Wuu cfisanayaa, Ma 4 is ann in.
Build it for yourself !
He built it for himself.
She built it for herself.
He is building it for himself. He did not build it for himself.
Assuming that the root is Such root extensions are shared by other verbs, e.g.
Beerd !
Wuu beertay. W&y beeratay. Wuu beer&nayaa. M& beerannin.
Cultivate it for yourself !
Cf. Bder ! Cultivate it !
He cultivated it for himself.
She cultivated it for herself.
He is cultivating it for himself.
He did not cultivate it for himself.
Verbal forms are arranged into extension classes according to the root extensions which they share, and on this account verbs with different extensions are regarded as separate verbs even when they share the same roots.
Verbal forms in all classes of the weak verbs are arranged into tense paradigms according to the terminations which they share.


12
B. W. ANDRZEJEWSKI
From the point of view of their notional associations the three components of the verb can be described as follows :
(a) The root gives the basic core of meaning and is often shared with word classes other than verbs.
(b) The extension amplifies the meaning of the root with such notions as performing an activity for one’s own benefit, causativeness, transitivity, intransitivity, etc.
(c) The terminations provide time reference, mood (commands, wishes, indicativeness, negation, etc.) and pronominal reference.
Terminations
Terminations are analysed further into three smaller components, a division which has great advantages in economy of statement. These components, which always occur in the same order, are : intermedium—link—ending.
There are three intermedia : I. ay(ey, aay)
II. ey(ay)
HI. #
The optional shape variants of the intermedia given in brackets indicate some vacillation in pronunciation from speaker to speaker. So far no one has succeeded in establishing rigid rules for their distribution, and my own researches suggest that these could only be stated in terms of the tendencies formulated below : Intermedium I:
Some speakers tend to use ay throughout, e.g.
Wuu Qis&yaa. He is building it.
Wuu djidayaa. He is refusing.
Miy&anu maaashiinayn ? Will he not make him hear it ?
Miy&anu j$og&yn ? Will he not stay ?
Others, however, tend to use ay with forms whose vowels belong to the ‘ back series ’,12 and ey with forms whose vowels belong to the ‘ front series ’, e.g.
Wuu ((isayaa.
Wuu djideyaa.
Miy&anu maciashiinayn ?
Miy&anu j$og6yn ?
The shape aay occurs mainly in poetry, and then only when a vowel follows immediately :
12 For the definition of these two series see Andrzejewski, ‘ The problem of vowel representation . .. ’. Forms with vowels of the front series are marked with a cedilla under their first vowel letter while those of the back series are left unmarked.


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
13
Wuu 4,Isaayaa. Wuu djidaayaa.
Intermedium II:
The majority of speakers use ey throughout, e.g.
Md 4js6yn. Md djiddyn. Md k^rindyn.
He was not building it. He was not refusing. He was not boiling it.
Sometimes ay occurs, e.g. : Md 41sdyn.
He was not building it.
Note that the vowels of all forms which have this intermedium belong to the ‘ front series ’.
In some dialects, notably that spoken in the Mijurtinia province of Somalia, the intermedia contain the consonant h (which is preceded in some cases by a short vowel), e.g.
Wuu 4,isahayaa. He is building it.
This fact confirms Moreno’s hypothesis that the intermedia were originally roots of the verb hdy ‘ to hold, to have, to keep ’ used as an auxiliary verb.
In grammatical descriptions and editions of texts there is a tendency to standardize the spelling of the optional variants of the intermedia in favour of either ay or ey. Thus Bell, for example, uses ey variants throughout for intermedia I and II, while ay is used in Hikmad Soomaali and in Abraham’s dictionary. It is interesting to note that in Moreno’s II somalo della Somalia and in texts written by Somalis in Somali Script13 no such standardization is attempted.
The positions of intermedia within tense paradigms are given in Table III, where it is understood that in the paradigms not listed the intermedium is #.
The significance of the arrangement in this table will become apparent later in the section ; this table and Tables V and VII ‘ Distribution of links within tense paradigms ’, and ‘ Distribution of endings within tense paradigms ’, are designed to be superimposed on one another in their proper order, i.e. intermedium— link—ending. Taken jointly, the three tables thus provide a complete list of terminations in weak verbs.
There are three links 14 and the first two of them consist of sets of alternants : I. t d Q sh s II. n ~ l(n) ~ r(n)
III. #
13 For an account of this script see Moreno, Il somalo della Somalia, 290-97.
14 The term ‘ link ’ was suggested to me by Kirk’s ‘ linking consonant Kirk, however, applies this term only to the initial consonant of definite articles and demonstratives when they combine with nouns.


14
B. W. ANDRZEJEWSKI
Table III
Distribution of Intermedia
1 sg. 2sg. 3 sg. m. 3 sg. f. 1 pi. 2 pi. 3 pi.
Main positive paradigms Pres. cnt. ext. Pres. cnt. res. Past cnt. ext. Past cnt. res. ay (ey, aay)
Main negative paradigms Neg. pres. cnt. var. Neg. pres. cnt. inv.
Neg. past cnt. ey (ay)
Dependent positive paradigms Pres. cnt. dvg. A Pres. cnt. dvg. B Past cnt. dvg. A Past cnt. dvg. B Pres. cnt. cvg. A Pres. cnt. cvg. B Past cnt. cvg. A Past cnt. cvg. B ay (ey, aay)
Dependent negative paradigms Neg. pres. cnt. dep. A Neg. pres. cnt. dep. B
Neg. past cnt. dep. A Neg. past cnt. dep. B ey (ay)
The distribution of the alternances is automatic and is determined by the nature and characteristics of the immediately preceding component, as shown in Table IV. Note that the root can immediately precede the link only in verbs with the extension Z and the intermedium #.


15
CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
Table IV
Distribution of Alternants in Links
1 Link I Immediately preceding component Alternant of the link
A root ending in : b, g, f, s, sh, n or r Any of the following extensions : AM, SAM, SOOM, TAM, R-AM, AYN/SAM, Z+/AYN/SAM, AN, ON, SAN, OON, SHOON, TOON, IN/SAN, AYN/SAN, Z+/AYN/ SAN, SIIN/SAN, TAM/SIIN/SAN, Z+/AAN, AN+/AAN, R-AN+/ A AN, SAN+/AAN, OON+/AAN, TOON+/AAN t
A root ending in : d, q, \ kh, h, c, h, w, y or in any vowel except for vowels described in the entry on the occurrence of the alternant sh (see below). Any of the following extensions : OOB, AAB d
A root ending in the consonant 4 4
A root ending in a vowel, but only in alternating roots in which one alternant ends in the consonant 1 and the other, while otherwise identical in shape, has not the final consonant 1, e.g. qal ~ qa (see section ‘ Roots ’). sh
The intermedia I and II. Any of the following extensions : IN, AYN, Z+/AYN, SUN, SIN, AAB/IN, TAM/SIIN, IN/SIIN1, IN/SIIN2, AYN/SIIN, Z+/AYN/SIIN, AN/SIIN1, AN/SIIN2, ON/ SUN, Z+/OON+/SIIN s
Link II Any component other than a root ending in the consonant 1 or r n
A root ending in the consonant 1 l(n)
A root ending in the consonant r r(n)
As the links occur very frequently in a large number of forms, it would be burdensome and uneconomic to enumerate their alternants each time they are discussed or listed. To obviate this difficulty the following notational device is introduced :
I. The symbol t* will refer to the whole series t~d~tj,~sh~s
II. The symbol n* will refer to the whole series n ~ l(n) ~ r(n)
The positions of links in tense paradigms is shown in Table V. It is to be understood that in those paradigms which are not listed the link is #.


16
B. W. ANDRZEJEWSKI
Table V
Distribution of Links in Tense Paradigms
1 sg. 2 sg. 3 sg. m. 3 sg. f. 1 pi. 2 pi. 3 pi.
1 Main positive paradigms Pres. gen. ext. Pres. cnt. ext. # t* # t* n& t* #
Pres. gen. res. Pres. cnt. res.
Past gen. ext. Past cnt. ext. t* t*
Past gen. res. Past cnt. res. # #
Past indep. Optat. Rhet. Poten. t* t*
Main negative paradigms Neg. pres. gen. Neg. pres. cnt. var. Neg. cond.
Dependent positive paradigms Pres. gen. dvg. A Pres. cnt. dvg. A Pres. gen. dvg. B Pres. cnt. dvg. B Past gen. dvg. A Past cnt. dvg. A Past gen. dvg. B Past cnt. dvg. B
Pres. gen. cvg. A Pres. cnt. cvg. A Pres. gen. cvg. B Pres. cnt. cvg. B Past gen. cvg. A Past cnt. cvg. A Past gen. cvg. B Past cnt. cvg. B # #
Dependent negative paradigms Neg. cond. dep. t* t*


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
17
There are sixteen endings : a, aa, aan, e, ee, een, ay(ey), o, id, i ~ #, in ~ #(in), ina, n(nin), ini ~ i(ini), ni(nini) and # (absence of an ending). Optional variants are given in brackets.
In the ending ay(ey) there is vacillation in the quality of the vowel and the same tendencies can be observed in the distribution of the two variants as in intermedium I. As in the case of intermedium I, some authors standardize in favour of ay and some in favour of ey, while in Moreno’s II somalo della Somalia and in the Somali Script no such standardization is attempted.
Table VI
Distribution of Alternants in Endings
Alternants selected
Immediately preceding component Ending i ~ # Ending in ~ # (in) Ending ini ~ i(ini)
Root with extension # (extension class Z). Any of the following extensions : AM, SAM, SOOM, OOB, AAB, TAM, R-AM, AYN/SAM, Z+/AYN/SAM. i in ini
Any extension other than those enumerated above. # # (in) i(ini)
In the endings i ~ in ~ #(in) and ini ~ i(ini), alternances take place and their distribution is determined by the nature and characteristics of the immediately preceding component. The details of their distribution are given in Table VI.


18
B. W. ANDRZEJEWSKI
The distribution of all the endings is shown in Table VII.
Table VII
Distribution of Endings Within Tense Paradigms
1 sg. 2 sg. 3 sg. m. 3 sg. f. 1 pi. 2 pi. 3 pi.
Main positive paradigms Imper. # a
Pres. gen. ext. Pres. cnt. ext. aa aan
Pres. gen. res. Pres. cnt. res. a
Past gen. ext. Past cnt. ext. ay (ey) een
Past gen. res. Past cnt. res. ay (ey)
Past indep. ay (ey) # ay (ey) e
Optat. 0 id (o) 0 aan (een) een
Rhet. 0 id 0 aan
Poten. ee een
Inf. i — #
Main negative paradigms Neg. imper. in ~ #(in) ina
Neg. pres. gen. Neg. pres. cnt. var. 0 id (o) 0 aan
Neg. pres.-past gen. in ~ #(in)
Neg. pres. cnt. inv. Neg. past cnt. n(nin)
Neg. optat. in ~ #(in)
Neg. cond. een


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
19
1 sg. 2 sg. 3 sg. m. 3 sg. f. 1 pi. 2 pi. 3 pi.
Dependent positive paradigms j Pres. gen. dvg. A Pres. cnt. dvg. A 0 id (o) 0 aan
Pres. gen. dvg. B Pres. cnt. dvg. B aa aan
Past gen. dvg. A Past cnt. dvg. A Past gen. dvg. B Past cnt. dvg. B ay (ey) een
Pres. gen. cvg. A Pres. cnt. cvg. A a
Pres. gen. cvg. B Pres. cnt. cvg. B aa
Past. gen. cvg. A Past cnt. cvg. A Past gen. cvg. B Past cnt. cvg. B ay (ey)
Inf. dep. i — #
Dependent negative paradigms Neg. pres.-past gen. dep. A in — # (in)
Neg. pres.-past gen. dep. B ini ~ i(ini)
Neg. pres. cnt. dep. A n(nin)
Neg. pres. cnt. dep. B ni(nini)
Neg. past cnt. dep. A n(nin)
Neg. past cnt. dep. B ni(nini)
Neg. cond. dep. een
When the tables showing the nature and distribution of intermedia, links and endings are superimposed on one another in the given order they produce a complete list of terminations of weak verbs, as in Table VIII.
It should be noted that when a particular termination is referred to later in this article it subsumes all its three components, i.e. intermedium, link and ending. If all the three components are # the termination will also be #.


20
B. W. ANDRZEJEWSKI
Table VIII List of Terminations
1 sg. 2sg. 3 sg. m. 3 sg. f. 1 pi. 2 pi. 3 pi.
Imper. # a
Pres. gen. ext. aa taa daa 4aa shaa saa aa taa daa c|aa shaa saa naa laa (naa) raa (naa) taan daan cjaan shaan saan aan
Pres. cnt. ext. ayaa aysaa ayaa aysaa aynaa aysaan ayaan
etc. etc. etc. etc. etc. etc. etc. etc.
The formulations given in this section are illustrated by short sentences
containing forms of the verbs :
l&ab „ to fold riih „ to push
sqg „ to wait kdc „ to rise
cun „ to eat hi4 „ to close
fur „ to open t61 „ to stitch (Root
shid „ to light
Examples of verbs in other extension classes will be given in the next section.
In the analysis of the examples the following abbreviations are used :
Rt. = root, Ex. = root extension, Int. = intermedium, Lk. = link, End. = ending.
Pres. cnt. ext. Rt. Ex. Int. Lk. End.
1 sg. Waan c(isayaa. I am'
2 sg. Waad (Jfeaysaa. You (sg.) are s
3 sg. m. Wuu ^lsayaa. He is # aa
3 sg. f. Way (Jfeaysaa. She is â– building it. 1 pi. Waynu cjfeaynaa. We are n
2 pi. Waydin ^lsaysaan. You (pi.) are s aan
3 pi. Way cjlsayaan. They are #


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI 21
Past gen. ext. Rt. Ex. Int. Lk. End.
1 sg. W&an ^isay. I
2 sg. Wdad ^istay. You (sg.) t
3 sg. m. Wiiu 1 pi. Wdynu Qisnay. We n
2 pi. Wiydin cjwteen. 3 pi. Wdy Past gen. res.
1 sg. Anig&a 15 cjteay. I
2 sg. Adig&a (Jfe&y. You (sg.) #
3 sg. m. Isdgda ^isdy. He
3 sg. f. Iyddda cjtet&y. She -built it. 1 pi. Inn&g&a ^isn&y. We n ay
2 pi. Idink&a <}is&y. 3 pi. Iydgda (Jte&y. You (pi.) #
Neg. pres.-past gen.
All) > M& disin. per.J H I did not build it. 4is # # # in
15 This is a contraction of aniga + b&a. Similarly, adig&a = adiga + b&a, is&g&a = is&ga + b&a, etc. The indicator b&a emphasizes here the substantive pronoun which precedes it. For a detailed account of this particle, see Hetzron’s article ‘ The particle b&a in Northern Somali ’.


22
B. W. ANDRZEJEWSKI
Neg. past cnt. Rt. Ex. Int. Lk. End.
All j Md 4fs6yn.16 1 per.J „ (4jseynin).J I was not building it. 4is # ey n nin
Past gen. ext.
r laabay. folded it. laab
sygay. waited. syg
cunay. ate it. cnn
furay. opened it. fur
shiday. He lighted it. shid
liqay. swallowed it. liq
riifcay. pushed it. riifc
kacay. rose. kac
fcic|ay. closed it. fci tolay. stitched it. tol
#
16 Note that the vowels of this form belong to the front series, while those of the other forms of the verb given above belong to the back series (see Andrzejewski, ‘ The problem of vowel representation It often happens that different forms of the same verb have
vowels belonging to different series, and numerous examples of this can be found in Hikmad Soomaali. The general rules could be summarized as follows :
I. When the 2 sg. imper. form has vowels of the front series, all the forms of the verb have vowels of that series.
II. When the 2 sg. imper. form has vowels of the back series all the forms of the verb have vowels of that series except for the following which have vowels of the front series :
(а) in all weak verbs :
(i) all forms of 3 sg. m. indep.,
(ii) all forms of past cnt. tense paradigms, positive and negative, main and dependent,
(iii) all forms which have the ending id ;
(б) in those weak verbs whose root or root extension ends in y or y : all forms in which the ending ay (ey) follows immediately such a root or root extension ;
(c) in those weak verbs in whose roots the last vowel is e, ee, o or oo : all forms in which the root is immediately followed by :
(i) an intermedium
(ii) any alternant of the extension IN
(iii) any ending which begins with or consists of the short vowel i.
Note moreover that :
All verbs of class OOB have vowels of the front series, even if the corresponding noun has vowels of the back series.
In the verbs c|&w (Z)4 to mate ’ (said of horses), g&l (Z) ‘ to enter ’, (is kd) rdw (Z) ‘ to guard oneself against, look out ’ and tag (Z) ‘ to go ’, the roots have the alternances c|aw ~

CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI 23
Rt. Ex. Int. Lk. End.
r laabtay. (folded it. laab
sygtay. waited. syg t
cuntay. ate it. cun
furtay. opened it. fur
shidday. lighted it. shid
3 sg. f. W&y liqday. She swallowed it. liq
riihday. pushed it. riih d
kacday. rose. kac
&i toshay. stitched it. to sh
r laabnay. folded it. laab
s^ignay. waited. syg # n ay
cunnay. ate it. cun
furray. opened it. fur r
(fumay). (n)
1 pi. W&ynu - shidnay. We - lighted it. shid
liqnay. swallowed it. liq
riibnay. pushed it. riih n
kacnay. rose. kac
hic|nay. closed it. hi^
tollay. stitched it. tol 1
(tolnay). (n)
Extensions
It is a characteristic feature of root extensions of weak verbs that each of them, except the extension Z, has a series of two or more alternants. The distribution of the alternants in each series is always determined in the first place by the type of termination which, figuratively speaking, acts as a ‘ selector ’, as can be seen from the following examples, where the root is kar and the extension has the alternants i, iy, in, inn which are selected by the different terminations :
Kar-i !
WUu kar-iy-aa. Wuu kar-in-ayaa. Ha kar-inn-in.
Boil (sg.) it!
He boils it.
He is boiling it.
Do not (sg.) boil it!
(Termination #)
( „ aa)
( „ ayaa)
(
in)


24
B. W. ANDRZEJEWSKI
In some extensions there are two or more alternants which occur with identical terminations ; their distribution is then determined by the nature of the root. Consider for example the alternants ad, d, t, t (4) and sh given below, which belong to the same extension and occur with the same termination aa, but are preceded by roots which differ in their phonological characteristics :
gunt-ad-aa.
<|aq-d-aa.
wM
fur-t-aa.
hi^-t-aa.
(bi4-4-aa).
qa-sh-aa.
He
ties it into a knot for himself, looks after it „ „
opens it „ „
closes it „ „
5?
slaughters it
5 J 5?
5? ??
The distribution of the alternants is predictable and can be described by reference to types of terminations and roots with which each alternant occurs. In view of this it is useful to arrange all terminations and all relevant types of roots into groups according to their role in the distribution of alternants in the extensions. Such groups, marked by serial numbers, can then be employed for constructing concise rules of distribution.
In Table IX all terminations are grouped in such a way.
Table IX
Terminations Grouped as Selectors of Extension Alternants
1. Terminations consisting of:
a, aa, aan, e, ee, een, ay (ey), 0.
2. Any of the following : #.
The first alternant, i.e. i, of the termination i ~
„ ,, in „ ,, in - - # (in).
„ „ ini „ ,, ini ~ i(ini).
3. The termination ina.
4. Any termination beginning with an intermedium.
5. Any of the following :
The second alternant, i.e. #, of the termination i ~ #.
„ „ # „ „ in ~ # (in).
„ „ i „ „ ini~i(ini).
This group excludes the optional variants of the second alternants.
6. Any of the following :
The optional variant, i.e. (in), of the second alternant of the termination in ~ # (in). The optional variant, i.e. (ini), of the second alternant of the termination ini ~ i(ini).
7. Any termination beginning with the link t*.


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
25
8. Any termination beginning with the link n*.
9. The termination # of the 2 sg. imper.
10. The termination # of the 3 sg. m. indep.
Relevant types of roots are grouped in Table X, and their serial numbers are given in square brackets to distinguish them from the serial numbers of groups of terminations.
In some of the groups reference is made to alternances in roots, a feature which is briefly discussed in the section on roots in this article.
The overlapping of definitions in some of the groups arises from the fact that the distribution of alternants in different extensions is determined by different characteristics of the root.
Table X
Types of Roots Grouped as Selectors of Extension Alternants
[1] Roots ending in :
(a) two different consonants, or (Z?) a doubled consonant, or
(c) the consonant k, or
(d) a vowel + the semivowel w, but not if the w is of the type described under [5] (Z>) below ; or
(e) a consonant + the semivowel y, and also :
the root qpy as it occurs in the forms of the verb qqy&n (AN+) ‘ to be wet ’.
[2] Roots ending in :
(a) a vowel + a single consonant other than k, or
(b) the semivowel w 4- a single consonant other than k, or
(c) jj »» y », jj j, »»
[3] Roots ending in :
(a) a vowel 4- the single consonant 1 or r, or
(fr) the semivowel w 4- the single consonant 1 or r, or
(c) jj jj y jj jj jj jj
[4] Roots ending in :
(a) a vowel 4- a single consonant, or
(b) sl vowel + the semivowel w, or
(c) jj jj jj y jj
(d) a vowel other than those described under [6] (d) or [8].
[5] Roots ending in :
(a) a vowel 4- the consonant q, 5, kh, h, c or h, or
(b) a vowel + the semivowel w, but only if the w belongs to an alternating root in which one alternant ends in w and the other in b, e.g. aw ~ ab, or
(c) a vowel 4- the semivowel y.
c


B. W. ANDRZEJEWSKI
26
[6] Roots ending in :
(a) a vowel + the single consonant b, g, f, s, sh, n or r, or
(b) the semivowel w + any of the consonants listed above, or
O') >> 99 y 99 99 99 99
(d) a vowel, but only in alternating roots in which one alternant ends in d and the other, while otherwise identical, has not the final consonant d, e.g. daad ~ daa (see section on roots).
[7] Roots ending in :
(a) a vowel + the single consonant or
(b) the semivowel w + the single consonant <(, or
(C) 99 99 ¥ 99 99 99 99
[8] Roots ending in a vowel which belongs to an alternating root in which one alternant ends in the consonant 1 and the other, while otherwise identical, has not the final consonant 1, e.g. qal ~ qa (see section on roots).
[9] Roots ending in :
(a) a vowel + a single consonant other than j, s, sh or 1, or (Z>) the semivowel w + a consonant specified above, or
(^) 99 99 y 99 99 99 99
(d) a vowel of the type described under [8].
[10] Roots ending in the consonant j.
[11] Roots ending in a vowel other than those described under [6] (d) or [8].
Defining the exact limits for the relevant groups of roots presents some difficulties. While it is comparatively easy to ascertain which roots occur with which extensions in those items of vocabulary which are frequently used, one can seldom be absolutely sure about the non-occurrence of certain roots with certain extensions in less common contexts. As each root usually occurs only with a small number of extensions (rarely more than seven), there are 6 gaps ’, i.e. potentialities of occurrence which are not normally fulfilled, yet people who are especially gifted in the use of language, whether in conversation, story-telling or poetry, often fill such gaps in ways unexpected even to their compatriots. For this reason the possibility of a future need for the augmenting or adjusting of the definitions of the groups of roots cannot be ruled out.
The grouping and numbering of terminations and relevant types of roots makes it possible to provide, in Table XI, details of the distribution of the alternants of each extension.
The figures placed above the alternants of extensions refer to the terminations with which these alternants occur, and figures in square brackets which precede them refer to the relevant types of roots. When no such figures are given it is to be assumed that there is no specific restriction on the type of root.
The capital letter R- used in some alternances represents the same consonant as the last consonant of the root; if the root, for example, ends in r the first


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI 27
Table XI
Distribution of Extension Alternants
z # with all terminations and roots
AM 1-4 [1] am [2] m 7-9 an
SAM 1-4 [2] sam [8] sham 7-9 [2] san [8] shan
SOOM 1-4 soom 7-9 soon
OOB 1^ oob 7-9 ow (aw)
AAB 1-4 aab 7-9 aw (ow)
TAM 1-4 tarn 7-9 tan
R-AM 1-4 R-am 7-9 R-an
AYN/SAM 17 and Z+/ AYN/SAM 1^1 aysm (eysm) 7-9 aysan (eysan)
IN 4, 5,8 in 7, 9, 10 i 1 iy [9] (sh) [10] (#) 3,6 inn
AYN and Z+/AYN 3-6 ayn (eyn) 7,8 ay (ey) 1 eey 9, 10 ee (ay)
SUN 4, 5,8 [4] siin [8] shim 7, 9, 10 [4] sii [8] shii 1 [4] siiy [8] shiiy 3,6 [4] siinn [8] shiinn
SIN 4,5,8 sin 7, 9, 10 si 1 siy 3,6 sinn
AAB/IN 4, 5,8 aabin 7, 9,10 aabi 1 aabiy (awsh) 3, 6 aabinn
17 See note 7, p. 4.


28
B. W. ANDRZEJEWSKI
TAM/SIIN 4, 5,8 tansiin 7, 9, 10 tansii 1 tansiiy 3,6 tansiinn
IN/SIIN1 4, 5,8 insiin 7, 9, 10 insii 1 insiiy 3,6 msiinn
IN/SIIN2 4, 5,8 isiin 7, 9, 10 isii 1 isiiy 3,6 isiinn
AYN/SIIN and Z+/AYN/SIIN 4, 5,8 aysiin (eysiin) 7, 9, 10 aysii (eysii) 1 aysiiy (eysiiy) 3,6 aysiinn (eysiinn)
AN/SIIN1 4, 5,8 ansiin 7, 9, 10 ansii 1 ansiiy 3,6 ansiinn
AN/SIIN2 4, 5,8 adsiin 7, 9, 10 adsii 1 adsiiy 3,6 adsiinn
ON/SIIN 4, 5,8 odsiin 7, 9, 10 odsii 1 odsiiy 3,6 odsiinn
z+/oon+/ayn/ SUN 4, 5, 8 oonaysiin (ooneysiin) 7, 9, 10 oonaysii (ooneysii) 1 oonaysiiy (ooneysiiy) 3,6 oonaysiinn (ooneysiinn)
AN 4, 5,8 an 9 0 1 [1] ad [5] d [6] t [7] t(4) [8] sh 7 a 10 ey 3,6 ann
ON 4, 5,8 on 9 0 1 d 7 0 10 oy 3,6 onn
SAN 4, 5, 8 9 [4] san [4] so [8] shan [8] sho 1 [2] sad [8] shad [11] St 7 [4] sa [8] sha 10 [4] sey [8] shey 3,6 [4] sann [8] shann
OON 4, 5,8 oon 1,9, ood 10 7 00 3,6 oonn
SHOON 4, 5,8 shoon 1,9,10 7 shood shoo 3,6 shoonn
TOON 4, 5,8 toon 1,9, tood 10 7 too 3,6 toonn
IN/SAN 4, 5,8 isan 9 iso 1 ist 7 isa 10 isey 3,6 isann


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
29
AYN/SAN and ZVAYN/SAN 4, 5,8 aysan (eysan) 9 1 7 ayso ayst aysa (eyso) (eyst) (eysa) 10 aysey (eysey) 3,6 aysann (eysann)
SIIN/SAN 4,5,8 9 1 7 [4] siisan [4] siiso [4] siist [4] siisa [8] shiisan [8] shiiso [8] shiist [8] shiisa 10 [4] siisey [8] shiisey 3,6 [4] siisann [8] shiisann
TAM/SIIN/SAN 4, 5,8 tansiisan 9 1 7 10 tansiiso tansiist tansiisa tansiisey 3,6 tfl.nsiisfl.nn
Z+/AAN 4, 5,8 aan 9 aw (ow) 1 aad 7, 10 aa 3,6 aann
AN+/AAN 4, 5,8 [1] anaan [2] naan [3] (R-aan) 9 [1] anaw [1] (anow) [2] naw [2] (now) [3] (R-aw) [3](R-ow) 1 [1] anaad [2] naad [3] (R-aad) 7, 10 [1] anaa [2] naa [3] (R-aa) 3,6 [1] anaann [2] naann [3] (R-aann)
R-AN+/AAN 4, 5,8 R-anaan 9 R-anaw 1 R-anaad 7, 10 R-anaa 3,6 R-anaann
SAN+/AAN 4, 5,8 [2] sanaan [8] shanaan 9 [2] sanaw [2](sanow) [8] shanaw [8](shanow) 1 [2] sanaad [8] shanaw 7, 10 [2] sanaa [8] shanaa 3,6 [2] sanaann [8] shanaann
OON+/AAN 4, 5,8 oonaan 9 oonaw (oonow) 1 oonaad 7, 10 oonaa 3,6 oonaann
TOONVAAN 4, 5,8 toonaan 9 toonaw (toonow) 1 toonaad 7, 10 toonaa 3,6 toonaann
consonant of the extension is also r : w&r-ran. ‘ to tell news cf. w&r (masc. noun) ‘ news This notational device thus represents doubling (gemination) of a consonant at a point of junction between the root and the extension.
The capital letters placed on the left are used as code letters for each extension, as in the section on terminology.
Optional variants of alternants in extensions are given in round brackets. The variants which contain the sound sequence ey tend to occur under the same phonological conditions as the variant ey of intermedium I.
The formulations given in this section are illustrated by short sentences containing forms of the verbs :


30
B. W. ANDRZEJEWSKI
(Z) to build
guntan (AM) to get tied into a knot or knots
hic|an (AM) to get closed or tied
d^mbasdw (OOB) to turn (intr.) into ashes
k&ri (IN) to boil (trans.) ; to cook
g£li (IN) to put in ; to cause to enter
[In this verb two root alternants occur : g?l ~ g j$oji (IN) to stop (trans.)
caande (AYN) to put milk into
yeelsii (SUN) to make someone agree to something
maqashii (SUN) to make someone hear something
[On analogy with m&qal (Z) ‘ to hear ’ we can assume
that the root maqa is an alternant within maqal ~maqa.]
horraysii (AYN/SIIN) to make someone go first; to put someone or some-
thing first
guntd (AN) to tie something into a knot or knots for oneself
kabbd (AN) to sip
daawo (AN) to watch ; to view
seebd (AN) to go to sleep
cabd (AN) to complain
[In this verb two root alternants occur : caw ~ cab.]
furd (AN) to open for oneself
shido (AN) to light for oneself
[In this verb two root alternants occur : shid ~ shi.]
bi told (AN) to stitch for oneself
[In this verb two root alternants occur : tol ~ to.]
karsd (SAN) to boil (trans.) or to cook for oneself
ku itaalsd (SAN) to exert one’s strength in something
gashd (SAN) to put on (clothes) ; ‘ to cause to enter upon oneself ’
[On analogy with gdl (Z) ‘ to enter ’ and g$li ‘ to
put in ; to cause to enter ’ we can assume that the root ga here is an alternant within gal
~ ga ~ g$l ~ g§.]
jqogsd (SAN) to stop (intrans.)
fac|iisd (SAN) to sit down ; to assume sitting position
casdw (ZVAAN) to become red
tjfen&w (ANf/AAN) to get built
ku filn&w ku (filldw) j(AN+/AAN) to become sufficient for
w§,naagsaniw(SANt/AAN)to become good


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
31
Pres. cnt. ext. Rt. Ex. Int. Lk. End.
^isayaa. He is building it. gunt&mayaa. It is getting tied into a knot, bl^mayaa. It is getting closed, d^mbasdobayaa. It is turning into ashes. karlnayaa. He is boiling it. g$linayaa. He is putting it in. jqojinayaa. He is stopping it. kar ipoj in
3 sg. m. Wuu - 'aan&ynayaa. He is putting milk into it. yeelsiinayaa. He is making him agree to it. maqashiinayaa. He is making him hear it. horraysiinayaa. He is making him go first. caan yeel maqa horr ayn siin shim aysiin ay aa
guntinayaa. He is tying it into a knot for himself. kabb&nayaa. He is sipping it. daaw&nayaa. He is watching it. seeb&nayaa. He is going to sleep. 'ab&nayaa. He is complaining. gunt kabb daaw seek cab an


32
B. W. ANDRZEJEWSKI
3 Sg. m. WilU
Rt. Ex. Int. Lk. End.
fur&nayaa. He is opening it for himself, shid&nayaa. He is lighting it for himself, hi^&nayaa. He is closing it for himself, tol&nayaa. He is stitching it for himself. fur shid tol an
kars&nayaa. He is boiling it for himself, ku itaals&nayaa. He is exerting his strength in it. kar itaal san
gash&nayaa. He is putting it on. ga shan ay aa
jpogs&nayaa. He is stopping (intrans.). fatjiis&nayaa. He is sitting down. jpog facjii san
cas&anayaa. It is becoming red. Qisn&anayaa. It is getting built. cas 4is aan naan
kii filn&anayaa. It is becoming sufficient for him. (ku fill&anayaa). The same meaning as above. fil naan (laan)
wspiaagsan&anayaa. It is becoming good. w$naag sanaan


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
33
Past gen. ext.
3 sg. m. Wuu
Rt. Ex.
^isay. 4is
guntamay. gunt am
hic|may. m
d^mbasoobay. dspnbas oob
k^riyay. kar 18 iy
(k^rshay). (sh)
g§liyay. g?l iy
(g$shay). se (sh)
jpojiyay. iQOj iy
(j?ojay). (#)
^aneeyay. caan eey
y§elsiiyay. yeel siiy
m^qashiiyay. maqa shiiy
hprraysiiyay. horr aysiiy
guntaday. gunt
kabbaday. kabb ad
daawaday. daaw
seehday. seek d
cawday. caw
furtay. fur t
shitay. shi
hicjtay. (&i<4ay). (4)
toshay. to sh
karsaday. kar sad
ku itaalsaday. itaal

#
Int.
Lk.

18 See note 16, p. 22.
End.
ay


34
B. W. ANDRZEJEWSKI
Rt. Ex. Int. Lk. End.
gashaday. ga shad
jpogsaday. jpog sad
fa ‘asaaday. cas aad
3 sg. m. Wiiu - ^isnaaday. ku filnaaday. fil naad
(ku fillaaday). laad
wqnaagsanaaday. wqnaag sanaad
Past gen. ext. Rt. Ex. Int. Lk. End.
(Jistay. guntantay. gunt an t
hi^antay.
dqmbasowday. dgtmbas ow d
karisay. kar
g§lisay. g§l i
jpojisay. JQOj
3 sg. f. W&y - caanaysay. yeelsiisay. caan yeel ay sii s
maqashiisay. maqa shii
horraysiisay. horr aysii
guntatay. gunt
kabbatay. kabb
daawatay. daaw
seehatay. seek
cabatay. furatay. fur
shidatay. shid
hi^atay.


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
35
3 sg. f. Wiy
Rt. Ex. Int. Lk. End.
tolatay. tol a
karsatay. kar sa
ku itaalsatay. itaal
gashatay. ga sha
jpogsatay. JQOg sa
- fa^iisatay. fa casaatay. cas aa
cjisnaatay. ku filnaatay. fil naa
(ku fillaatay). (laa)
w^naagsanaatay. w^naag Sanaa
Roots
The majority of roots in Somali are stable, i.e. they have the same shape in every environment. From the point of view of the study of inflectional characteristics of verbal forms they present no descriptive problems. However, a fair number of roots have alternances of various kinds, such as ‘ contraction ’, differences in final consonants or transposition of consonants, e.g. :
Orod !
Orda !
Wuu cawday. Wdy 'abatay. J$og !
Jpoji!
T&'ab !
Tab'6 !
Run (sg.) !
Run (pi.) !
He complained.
She complained.
Stay (sg.) !
Stop (sg.) it!
Work (sg.) hard !
Work hard for yourself !
With very few exceptions such alternances have only two alternants and their distribution is predictable in terms of simple phonological or morpho-phonological rules. For example, in the root alternance orod ~ ord, in the verb 6rod (Z) ‘ to run ’, the first alternant occurs before terminations of groups 7-9, while the second occurs before terminations of groups 1-4.


36
B. W. ANDRZEJEWSKI
Root alternances in verbs have received extensive treatment in Bell’s The Somali language, and in Abraham’s Somali-English dictionary ; in the latter work information on their occurrence, distribution and putative historical origin is provided in the grammatical appendix and incorporated in the entries on individual verbs.
My analysis of root alternances departs from that of these two authors at a point which is relevant to the main theme of this article, namely in the treatment of roots in which the first alternant ends in the consonant 1 or d and the second, while otherwise identical in shape, ends without the final consonant, e.g. qal ~ qa (1 ~ #) and qod ~ qo (d #).
The distribution of alternants within 1 ~ # can be summarized as follows :
(a) Every root which ends in 1, when immediately followed by the link # or n‘, ends in # (i.e. absence of 1) when it is immediately followed by the link t* (then realized as sh), e.g.
Wuu qal-ay.
fqal-lay.
^((qal-uay). Wdy qa-shay.
He slaughtered it. We slaughtered it. She slaughtered it.
(6) Every root which ends in a vowel or semivowel +1 when followed immediately by any of the alternants of extension IN which begin with or consist of a vowel, ends in the same vowel or semivowel + # (i.e. without 1) when immediately followed by the optional variant sh of the alternant iy, e.g.
G£l-i !
Wiy g§l-isay. Ha g§l-kmin !
Whu(®?1'iyay:
((g§-shay)
Put (sg.) it in !
She put it in.
Do not put (sg.) it in ! He put it in.
(c) Every root which ends in a vowel or semivowel + 1 when it is immediately followed by those alternants of extension AN which are selected by terminations of groups 3-10, ends in the same vowel or semivowel + # when it occurs with an extension alternant selected by terminations of group 1, e.g.
Qal-6 ! Slaughter it for yourself !
Wdy qal-atay. She slaughtered it for herself.
Whu qa-shay. He slaughtered it for himself.
(d) Two roots which end in vowels when they occur with extension SHOON, end in 1 when they occur with corresponding nouns :


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
37
Hawshdod ! Work hard !
Cf. h&wl (fem. noun) ‘ hard work ’
Hishdod ! Be ashamed !
Cf. hil (masc. noun) ‘ sense of shame or
respect; sense of decorum ’
Some roots which end in 1 when they occur with extensions Z, IN or AN under conditions specified above, end in # when they occur with those extensions whose first code letter is S, e.g.
fuul (Z) to climb
fuu-shii (SIIN) to make someone climb
m&qal (Z) to hear
maqa-shli (SIIN) to make someone hear
ku c$l-i (IN) to return (trans.) ; to repeat
ku c§sh6 (SAN) to return (trans.) or to repeat for one’s own benefit; to reclaim
g&l (Z) to enter
g£l-i (IN) to put in ; to cause to enter
ga-sh6 (SAN) to put on (clothes) (‘ to cause to enter upon oneself ’)
It is impossible to predict from any phonological features in what roots this alternance occurs. Compare, for example, with the above roots the root yeel in ydel (Z) ‘ to agree, to do ’ and yeelsli (SIIN) ‘ to make someone agree ’.
The occurrence and distribution of the alternance d <~ # can be described as follows :
Every root which ends in a vowel or a semivowel + d when it occurs with (a) the extension Z, or
(h) with any extension other than TAM, TAM/SIIN, TAM/SIIN/SAN and AN, or
(c) with those alternants of extension AN which are selected by terminations of groups 3-10,
ends in the same vowel or semivowel + # (i.e. absence of d) when it occurs with
(a) any of the extensions TAM, TAMSIIN and TAM/SIIN/SAN, or
(b) with those alternants of extension AN which are selected by terminations of group 1.
E.g.
Way 1& dood-een. They disputed it with him.
(Extension Z)
Wiy doo-tam-een. They disputed it with one another.
(Extension TAM)
Wdy qaad-a-tay. She took it for herself.
(Extension AN, termination in group 7)


38
B. W. ANDRZEJEWSKI
Wiiu qaa-t-ay.
Similarly :
W4y qod-a-tay. Wiiu qo-t-ay. W&y shid-a-tay. Wiiu shi-t-ay.
He took it for himself.
(Extension AN, termination in group 1)
She dug for herself.
He dug for himself.
She lighted it for herself.
He lighted it for himself.
In one case only, namely in the root i§ed ~ j§e the alternant # occurs with extension SAN :
j?e-s6 (SAN) to turn (intrans.) towards cf. i?ed (Z) to be facing towards
j$edi (IN) to turn (trans.) towards
There is a small number of verbs in which roots and extensions combine in such a way that it is not possible to draw boundaries between them. These ‘ combinative ’ verbs are listed below and each of them is assigned hypothetically to an extension class, as shown by its code letters.
bakhtli (IN) to extinguish, cf. bikhti (Z) to die (said of animals or of fire)
d§,a (AYN) to leave (trans.); to let alone ; to let go
d$ys6 (AYN/SAN) to leave (trans.) for oneself (i.e. for one’s own benefit or
future use)
dsjasd (AYN/SAN) an optional alternative of d$ys6
g§e (AYN) to take something or someone to a place
g§ysd (AYN/SAN) to take something or someone to a place for oneself; to
bring about (an event)
g$o (IN) to cut, cf. g6’ (Z) to ‘ become ’ cut
goosd (IN/SAN) to cut for oneself
h4y (IN) to hold ; to have control over
haysd (IN/SAN) to hold for oneself; to possess
heshli (IN) to agree with one another
heshiisli (IN/SIIN2) to make people agree with one another
h$y (IN) to take home
14a (AYN) to beat; to kill (used when referring to a group of persons or animals to whom the activity is applied)
q$o (IN) to wet, cf. q$y (Z) to become wet
ku r4y (IN) to become delighted with, to profit by
raysd (IN/SAN) to experience an improvement in one’s state of mind or body ; to experience a relief from pain or worry
sli (SUN) to give


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI 39
siisd (SIIN/SAN) to give something in exchange for something else ; to pay for
w$ydli (IN) to ask (a favour, or a question)
weydiisd (IN/SAN) to ask (a favour or a question) for oneself
The most convenient way of describing these verbs is to list their roots together with their extensions and to provide information about the distribution of alternants in the same way as has been done for root extensions, and this is done in Table XII.
Table XII
4, 5, 8 bakhtiin heshiin w§ydiin 7, 9, 10 bakhtii heshii w§ydii 1 bakhtiiy heshiiy w^ydiiy 3,6 bakhtiinn heshiinn w§ydiinn
IN 3, 4-6 hayn 1, 7-10 hay
3-6 hpyn rayn 7-10 hpy ray 1 hpoy reey
3-6 gpyn qpyn 7, 8 g?y qpy 1 gpoy Q AYN 3-6 d§yn g$yn layn 7, 8 d§y g?y lay 1 d§,ay g?ey laay 9, 10 d$a g$e laa
SUN 4, 5,8 siin 7, 9, 10 sii 1 siiy 3,6 siinn
IN/SIIN2 4, 5, 8 heshiisiin 7, 9, 10 heshiisii 1 heshiisiiy 3,6 heshiisiinn
IN/SAN 4, 5,8 goosan haysan raysan w§ydiisan 9 gooso hayso rayso w§ydiiso 1 goost hayst rayst w$ydiist 7 goosa haysa raysa ; w$ydiisa 10 3,6 goosey goosann haysey haysann raysey raysann w§ydiisey w§ydiisann
AYN/SAN 4, 5,8 d$ysan (d^asan) g$ysan 9 d§yso (d$aso) g$yso 1 d§yst (d^ast) g§yst 7 d§ysa (d^asa) g$ysa 10 3,6 d$ysey d§ysann (d^asey) (d^asann) g§ysey g$ysann
SIIN/SAN 4,5,8 siisan 9 siiso 1 siist 7 siisa 10 siisey 3,6 siisann


40
B. W. ANDRZEJEWSKI
Accentual Patterns
Accentual patterns are an integral part of verbal forms and frequently distinguish forms which are otherwise identical in shape. Compare, for example, the forms keen&y (3 sg. m. past gen. cvg. A) with keenay (3 sg. m. past gen. cvg. B) in the following sentences :
Ninkii keenly miyiiu arkay ? Did he see the man who brought it ?
Ninkii keenay miyiiu arkay ? Did the man who brought it see him ?
Similar distinction can be seen between forms like arkden (3 pi. past gen. dvg. A) and arkeen (3 pi. past gen. dvg. B) :
Rkggli ay arkden miyey u yee^een ? Did they call the men whom they saw ?
R&ggii ay arkeen miyey u yeeQeen ? Did the men whom they saw call them ?
An account of accentual patterns in main tense paradigms is given in my article ‘ Accentual patterns in verbal forms ... ’.
A list of accentual patterns in dependent tense paradigms in all extension classes is given below, and the number references to patterns are the same as in that article. One pattern, however, which does not occur in main tense paradigms has been added : it consists of the Accentual Unit No. 1 on the fourth syllable from the end and the Accentual Unit No. 9 on all the remaining syllables. This pattern is marked as Accentual Pattern No. 10.
Pres. gen. dvg. A
(a) all persons sg. and 1 pi. 1
(b) 2 pi. and 3 pi. 6
Pres. cnt. dvg. A
(a) all persons sg. and 1 pi. 5
(h) 2 pi. and 3 pi. 8
Pres. gen. dvg. B 9
Pres. cnt. dvg. B 3
Past gen. dvg. A
(а) all persons sg. and 1 pi. 1
(б) 2 pi. and 3 pi. 6
Past cnt. dvg. A
(а) all persons sg. and 1 pi. 5
(б) 2 pi. and 3 pi. 8
Past gen. dvg. B 9
Past cnt. dvg. B 3
Pres. gen. cvg. A 1
Pres. cnt. cvg. A 5
Pres. gen. cvg. B 9
Pres. cnt. cvg. B 3


CHARACTERISTICS OF THE SO-CALLED 4 WEAK VERBS ’ IN SOMALI
41
Past gen. cvg. A 1
Past cnt. cvg. A 5
Past gen. cvg. B 9
Past cnt. cvg. B 3
Inf. dep.
(a) with the termination in i 2
(Z?) with the termination in # 1
Neg. pres.-past gen. dep. A 1
Neg. pres.-past gen. dep. B 9
Neg. pres. cnt. dep. A
(a) with the termination ayn (eyn) 4
(Z>) with the termination aynin (eynin) 5
Neg. pres. cnt. dep. B
(a) with the termination ayni (eyni) 3
(Z?) with the termination aynini (eynini) 10
Neg. past cnt. dep. A
(a) with the termination eyn (ayn) 4
(Z?) with the termination eynin (aynin) 5
Neg. past cnt. dep. B
(a) with the termination eyni (ayni) 3
(Z?) with the termination eynini (aynini) 10
Neg. cond. dep. 1
Illustrative Text
In addition to the examples already given the formulations concerning weak verbs are illustrated by verbal forms which occur in a traditional story narrated by Mr cUmar Hyseen, also known as cUmar OstrSeliya, who has a high reputation in Somali-speaking territories as a story teller, poetry reciter and broadcaster.
The text is a transcript of a tape recording of a story narrated spontaneously, without any preparation, to a small Somali audience. As often happens in spontaneous speech, on several occasions the narrator changed his mind as to what grammatical construction he was going to use, when he had already begun one, and thus verbally 4 cancelled ’ certain words and phrases. This is indicated by placing such4 cancelled ’ items in round brackets and leaving them untranslated.
The translation of the text is designed to give some indication to the reader of the role and position of the verbal forms under observation in each sentence of the original, and this results in a somewhat artificial rendering in English, which in no way reflects the artistry of the original. The words placed in square brackets do not correspond to any words in the original but are implied in the context.
In the notes, lexical information is provided only on those words which are not included in R. C. Abraham’s Somali-English dictionary or which require further elucidation. Moreover, grammatical constructions relevant to the study
D


42
B. W. ANDRZEJEWSKI
of weak verbs are commented on when necessary, and obscure passages of the text are explained.
Finally all the weak verbs which occur in the text are analysed in Table XIII at the end of this section.
Text
1. (Wdha) b^rigii hord (iyda wdha jjray) ruuh h^ween dh iyda w^ddad u tagtay oo wahay ti^i, “ Shdekhow, ninkdygii wuu igd baycjay oo ruuh h^ween dh do kald iyuu jeclaaday, (wdhaan, waa) aniga na wuu i nacay. 2. Wah aan dqonayaa inaad qarcjdas ii qorto ninkdyga igu s6o c§lisd.” 3. Wuhuu yjcji, “ Wda tahay, wdhaan (Ku, aan) Ku i^aahdo maad y£eli ? ” 4. “Hda,” bdy tieji. 5. Wuhuu yjc|i, u Wah a ad yeeshaa ljbdah w§jigjisa, halkaa w$jiga dushjisa timdha kagd y£al sdo gqo ydo ii kden ! ” iyuu ku yjc|i. 6. “ Kolkaad timdha ii keento ninkaagii qarc|aas uu Kugu jeclaado iydan Kuu qqri doonaa.” 7. Kdlkaasda gqri (-gu) ljbdah leh iyda reerkii sdo d§gay. 8. Kdlkaasdy ldh djbddda ku hi44aY- 9. H^bdenkii baa ljbdahii yjmi yoo ldhdii buu qaatay. 10. H^bdenkii dambda ljbaahii soo noqday. 11. Muskii reerka e djbddda iydy ldh kald ku hi44ay« 12. Lahdaa na h^bdenkaasuu qaatay. 13. Hgtbdenkii dambe iydy intay muskii mdel kd d§Lloolisay, hdggii gqriga e dqalka sdo hjgtdy ldh ku hi^^ay-
14. H§,bdenkda na wuu sdo galay do lahdii buu la bahay do lahdaa na wuu la t$gay.
15. Kdlkaasay hgbdenkii kale dqalkqedii 4ig4e^a^Hsh djbddda intay ldhdii ku hit}(Jay bay — hdggii aqalka dushjisa ldhdii jqogtaa, haejiggii na guddha wuu ku hiQdny^hay— h^bdenkaa na lahdaasuu goostay. 16. Kdlkaasdy labd saamdod ^hdqod bannaysay do ldh ba intay qashdy ba4 na djbddda ka d$ldeshay, ba4 na dqalka gudihjisii iydy soo g$lisay do kd d§ldeshay. 17. Kdlkaasuu ljbdahii bacjkii djbddda (Id) qaatoo goostay. 18. Kolkuu 4g44amiy^u hjlibkii cundy iyuu qdarkii hord Id soo galay. 19. Ayadu na sdantii iydy g§es is ku yar laabtay do manqas bdy haysatay do kolkii ljbaahii madahjisii ejaafay iydy mdnqaskii sidda dogortii w$jiga kd gqysay. 20. Ldhdii buu goostay bdcjk^edii ydo wuu la t$gay. 21. Kdlka uu ljbdahii t$gdy ayadii na wahay la sdo ha^4ay bulbushii w$jiga kagd tjil ey dy soo goosatay. 22. Aroortii dambe iydy ayddoo sjddaa u keentay shdekhii yoo wahay ti^i, “ Waa kdn, w£jigii Ijbdaha (sdantii ku tjil) dogortii ku tjil.” 23. Kdlkaasuu yjcji, “ Sidled ugd gpysay ? ” 24. Wahay u sheegtay farsamddii ay u qabatay eey koi ba mdel u sdo jqojisay. 25. Kdlkaasuu yic|i, “ Djbddda wdha meerd e dugaagga dh ljbdah bda ugu b$laayaysdn,” buu yjeji. 26. “ Dddku na ljbdahuu kd wada baqaa,” buu yjeji. 27. “ Ljbaahu na B$ni-Aadanka wuu kd baqaa,” buu yjcji- 28. “ Ljbaahu cdynkaa ahaa farsamada wahsiinta aad dqalkdagii ku sdo g$lisay ; (as-) kani wda ba B$ni-Aadan do wda nin. 29. Ninkdagii qarejaas kale ma laha. 30. Hddalka ood u m$caanaysd iyo gacdnta ood wah kd siiso iyda qarejdastii ydh. 31. Naa, ninka adigda qarejdastii hayd d dqalka sdo g£li ydo wah sii ! ” buu yj^i. 32. “ Sidaasuu Kugu imdn d.” 33. Kdlkaasdy qggolaatay sidii yoo ninkii bdy intay u yeeQcjdy hddal m$caan kula hadashay do ay koi ba wah siisay. 34. Kdlkaasda ninkii aqalkii sdo galay.
Translation
1. [Once] in times past a woman went to a man of religion and what she said


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
43
to him [was], ‘ O learned man, my husband has turned away from me and has come to love another woman and has come to hate me. 2. What I want [is] that you should write for me an amulet-scroll which will bring back my husband.’ 3. What he said [was], ‘ All right, will you do what I tell you ? ’ 4. ‘ Yes,’ she said. 5. What he said [was], ‘ What you should do [is this] : cut off the hair which is on the face of a lion and bring it to me,’ he said. 6. ‘ When you bring the hair to me, I shall write for you an amulet-scroll through which your husband will come to love you.’ 7. Then the family encamped on a site in which there were lions. 8. Then she tied a ewe outside. 9. In the night the lion came and took the ewe for himself. 10. The following night the lion came back. 11. She tied another ewe by the fence of the encampment, on the outer side. 12. On that night he also took that ewe for himself. 13. The following night, after she had made a hole in the fence in one place, she tied a ewe on the [inner] side of the encampment [in a place] which was beside the hut. 14. On that night also he went in, emerged with it and went away with it. 15. Then, on another night, after she had tied the ewe outside to the central beam of the hut—the ewe is outside the hut and the rope is tied inside the hut—on that night also he tore it away for himself. 16. Then she made a space between two skins [of the wall of the hut] and after she had slaughtered a ewe, she hung half outside and she put the other half inside the hut and hung it up. 17. Then the lion took the half outside for himself and tore it away for himself. 18. When he had tasted it and ate the meat, he came in [to the hut] with the front part [of his body]. 19. She folded the skin together a little— and she had a pair of scissors—and when the head of the lion passed her she cut off the woolly hair from his face like that! 20. The lion tore away for himself the half of the ewe and went away with it. 21. When the lion went away, what was left with her [was] the woolly hair which had been on his face and which she had cut off for herself. 22. Next morning, carrying it, she brought it to the learned man and said, ‘ Here it is, the woolly hair which was on the face of the lion.’ 23. Then he said, ‘ How did you cut it off him ? ’ 24. What she told him about [was] the ruse which she had used for it and how she halted him again and again. 25. Then he said, ‘ Of the creatures which roam abroad and which are beasts of prey the lion is the most ferocious. 26. People are afraid of the lion,’ he said. 27. ‘ And the lion is afraid of the sons of Adam,’ he said. 28. ‘ [Consider] the ruse of giving things by which you made the lion, who was like that, enter your hut; [but] this one, [your husband], is just a son of Adam and is a man. 29. Your husband needs no other amulet-scroll. 30. The words, while you make them sweet for him, and the hand, while you give him something from it, are the amulet-scroll. 31. O woman, as for the man, you have his amulet-scroll, make him come into the hut and give him something,’ he said. 32. ‘ In this way he will come to you.’ 33. Then she accepted it like this, and after she had called the man she spoke to him with sweet words and gave him something again and again. 34. Then the man came into the hut.


44
Notes
B. W. ANDRZEJEWSKI
1. haween
ruuh haween ah
iy&a
w&hay
ah oo
2. inaad . . . qorto
q arenas
3. w&a tahay
5. wdhaad yeeshaa
(masc. collective noun) women (polite form).
a person who is a woman ; lit. 4 a person who is (or belongs to) women ’.
an optional alternative of the particle bda ; iyda like b&a combines with preverbal subject pronouns. For an extensive and well reasoned account of this particle see Robert Hetzron’s article 6 The particle bda in northern Somali ’. wdha ay ; waha = wah (masc. noun) + ka (def. article) ; the noun wah has a wide range of meanings : 4 what. . . ; a thing ; things ; a person ; persons ; creature ; creatures ’. Note that verbal forms which are dependent on this noun as their headword, have special rules of distribution when they occur in the 4 inceptive part ’ of a 4 wah-sentence ’. (For an account of this type of sentence, composed of 4 inceptive ’ and 4 sequel ’ parts, neither of which contains a main verb, see The declensions of Somali nouns, Appendix VI, pp. 140-1.) When the dependent verb in an inceptive part is positive, B forms are used in divergent tense paradigms and A forms in convergent tense paradigms ; when the dependent form is negative A forms are used throughout. E.g. Wuhuu d$onayaa bjyo. ‘ What he wants [is] water.’ (divergent B) ; W&ha nald jpogd cAbdi. 4 The person who is staying with us [is] cAbdi.’ (convergent A) ; W&h&an La h§lin sgliid. 4 The thing which people have not found [is] oil.’ (negative dependent A). When they occur outside the inceptive parts of wah-sentences, the dependent verbal forms have the rules of distribution summarized in my 4 Notes on the substantive pronouns in Somali ’, p. 89, e.g. Wuhuu d$onay6 garan m&ynd. 4 We do not know what he wants.’ (divergent A).
if there is no pause between these two words the first of them is normally pronounced oh.
when a divergent verbal form is dependent on the particle in 4 that ’, it can be either A or B, the choice being optional, (fern, noun) an amulet-scroll, consisting of a piece of paper or parchment on which verses from the Koran are written.
4 all right ’, lit. 4 you are ’.
4 what you should do [is this]. . . ’. This construction can be regarded as the inceptive part of a wah-sentence. The


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI 45
libaah ... y£al remaining part of the sentence can be regarded as its sequel part (see note on w&hay sentence 1). lit. ‘ the hair which is on the face of a lion, there, on the surface of his face
7. d$gay in verbal forms whose vowels belong to the ‘ front series ’ (represented here with a cedilla under the first vowel letter) the narrator pronounces the vowel in the ending ay with a quality intermediate between [e] and [a] of the I.P.A.
10. damb&a dambd + b&a.
13. intay Inta ay ; for construction introduced by Inta see Note 1/10 in Hikmad Soomaali, p. 65.
fc&ggii . . . hjgtdy lit. ‘ on the side of the encampment which was beside the hut ’.
bdggii h&g (masc. noun) + kii (def. article); the noun big ‘ side ; vicinity ; direction ; place ’ usually stands in concord with the 3 sg. f. forms of the verb and not with the 3 sg. m. as might be expected.
15. 16. labd saamdod ‘ two skins ’ ; reference is made here to the skins which form the walls of a hut.
17. (Id) most probably used inadvertently instead of k& ‘ from ’.
qaatoo qaatay do.
18- 4^44amiydy biiu.
qaarkii q&ar (masc. noun) + kii (def. article) ; q&ar means ‘ part; portion ; small group ’ ; here reference is made to the front part of the lion’s body.
19. ayddu an optional alternative of iyddu ‘ she ’.
sdantii. . . laabtay ‘ she folded the skin together a little ’ ; this she apparently did to protect herself from the lion.
21. ey dy this sequence may be considered as equivalent to 6e ay or e ay ‘ and which she ’, which would be more usual in this context.
22. ayddoo sjddaa aydda oo sjddaa ; lit. ‘ while she is (/was) carrying it ’ ; for constructions of this kind see Note 8/7 A in Hikmad Soomaali, p. 77.
23. sidled sidge bdad.
24. eey e ay ‘ and which she ’.


46
25. b$laayaysan
26. ljb&ahuu
27. B§ni-Aadan
28. (as-)
29. ma lah&
30. ood
31. adig&a hayd 6
32. imdn 6
B. W. ANDRZEJEWSKI
(hybrid verb) to be ferocious ; cf. b^laayo (fern, noun) evil ; misfortune ; dangerous person or animal, ljb&aha buu.
(masc. noun, singular or collective) son of Adam ; children of Adam ; person ; people (from Arabic bani Adam, oblique case of banu Adam ‘ sons of Adam ’).
probably the first vowel and consonant of the word asdgu 6 he ’ ; it appears that the narrator changed his mind in the middle of the word and used kani ‘ this one ’ instead.
4 needs not ’ ; lit. 4 has not ’.
oo aad 4 and which you ’.
adiga + bda.
when there is no pause between these two words they are pronounced as hayd >6 or hayd e or hay6e ; in this sentence the pronunciation used is hayd 6.
in the pronunciation of the narrator the final consonant in imdn is doubled, i.e. imdnn ; such pronunciation sometimes occurs when the conjunction e or ee follows immediately.
Table XIII
Analysis of Verbal Forms in the Text
(Figures in brackets refer to sentences in the text.)
Verbal form Grammatical designation Rt. Ex. lnt. Lk. End.
tagt&y (1) 3 sg. f. past gen. res. of tig (Z) 'to go ’ (Root tag ~ t$g) tag & # t ay
bayQay (1) 3 sg. m. past gen. ext. of b&yQ (Z) 4 to turn aside ’ bayq, # # # ay
jeclaaday (1) 3 sg. m. past gen. ext. of jecl&w (Z+/AAN) 4 to develop love for ; to come to love ’ jecl aad # # ay
nacay (1) 3 sg. m. past gen. ext. of ndc (Z)4 to develop hatred for ; to come to hate ’ nac # # # ay
d qortd (2) 2 sg. pres. gen. dvg. A of qdr (Z)4 to write ’ qor # # t 0
c§lis& (2) 3 sg. f. pres. gen. cvg. A of sdo c$li (IN)4 to bring back ; to return (trans.) ’ C?1 i # s a
19 See note 16, p. 22.


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
47
Verbal form Grammatical designation Rt. Ex. Int. Lk. End.
y$eli (3) inf. of ydel (Z) ‘ to agree ; to obey ; to follow (instructions) ’ (Root yeel ~ yee) yeel # # & i
yeeshaa (5) 2 sg. pres. gen. dvg. B of ydel (Z), see above yee # # sh aa
gpo (5) 2 sg. imper. of gqo (IN) ‘ to cut ; to tear ’ goo # # #
k6en (5) 2 sg. imper. of kden (Z) ‘ to bring ’ # # # #
keentd (6) 2 sg. pres. gen. dvg. A of kden (Z), see above keen # # t 0
jeclaado (6) 3 sg. m. pres. gen. dvg. A of jecl&w (Z+/AAN), see under je'laaday (1) jecl aad # # 0
q$ri (6) inf. of qdr (Z), see under qortd (2) qor # # # i
doonaa (6) 1 sg. pres. gen. ext. of ddon (Z), see above under dponayaa (2); doonaa is used here as an auxiliary with the meaning ‘ shall ’ or ‘ will ’ doon # # # aa
d$gay (7) 3 sg. m. past gen. ext. of d§g (Z) ‘ to encamp ; to settle ; to alight ’ d?g # # # ay
bi44,ay (8) 3 sg. f. past gen. ext. of hi^ (Z) ‘ to tie ; to shut’ # # 4 ay
qaatay (9) 3 sg. m. past gen. ext. of qaadd (AN) ‘ to take for oneself ’ (Root qaad ~ qaa) qaa t # # ay
noqday (10) 3 sg. m. past gen. ext. of sdo noqd (ON) ‘ to come back ’ noq d # # ay
bit^ay (11) the same as hi((c|ay (8)
qaatay (12) the same as qaatay (9)
d^loolis&y (13) 3 sg. f. past gen. dvg. A of d§ldoli (IN) ‘ to make a hole or holes in (something) ’ dalool i # s ay
hjgt&y (13) 3 sg. f. past gen. cvg. A of hjg (Z) ‘ to be beside ; to be adjacent or near ’ bjg # # t ay
hi galay (14) 3 sg. m. past gen. ext. of gdl (Z) ‘ to come in ; to enter ’ gal # # # ay
bafcay (14) 3 sg. m. past gen. ext. of b&fc (Z) ‘ to come out; to emerge ’ bah # # # ay


48
B. W. ANDRZEJEWSKI
Verbal form Grammatical designation Rt. Ex. Int. Lk. End.
t?gay (14) 3 sg. m. past gen. ext. of tdg (Z), see under tagtay (1) (Root tag ~ t§g. Note that the t§g alternant of the root occurs when the root is immediately followed by an intermedium or the ending ay (ey), i, in, ina or ini. It also occurs in the 3 sg. m. past indep. The alternant tag is found under all other conditions.) t?g # # # ay
(15) 3 sg. f. past gen. dvg. A of hiQ (Z), see under &i44ay (8) # 4 ay
jpogtaa (15) 3 sg. f. pres. gen. ext. of j$og (Z) ‘ to stay ; to be (in a place) ’ ipog # t aa
goostay (15) 3 sg. m. past gen. ext. of goosd (IN/SAN) ‘ to cut for oneself ; to tear for oneself ’ goost # ay
bannaysay (16) 3 sg. f. past gen. ext. of bannde (AYN) ‘ to make space ; to clear ’ bann ay # s ay
qash&y (16) 3 sg. f, past gen. dvg. A of q&l (Z) ‘ to slaughter ’ (Root qal ~ qa) qa # # sh ay
d$ldeshay (16) 3 sg. f. past gen. ext. of d$ldel (Z) ‘ to hang ; (trans.) ’ (Root d$ldel ~ d§lde) d$lde # 77^ sh ay
g§lisay (16) 3 sg. f. past gen. ext. of g£li (IN) ‘ to put in ; to cause to enter ’ g?l i # s ay
d^ldeshay (16) the same as d$ldeshay (16) above
qaatay (17) the same as qaatay (9)
goostay (17) the same as goostay (15)
cun&y (18) 3 sg. m. past gen. dvg. A of cun (Z) ‘ to eat ’ cun # # # ay
galay (18) the same as galay (14)
laabtay (19) 3 sg. f. past gen. ext. of l&ab (Z) ‘ to fold ’ laab # # t ay
haysatay (19) 3 sg. f. past gen. ext. of haysd (IN/SAN) ‘ to hold for oneself ; to have ’ haysa # t ay
tjaaf&y (19) 3 sg. m. past gen. dvg. A of cj&af (Z) ‘ to pass (intrans.) ’ 4aaf # # # ay


CHARACTERISTICS OF THE SO-CALLED ‘ WEAK VERBS ’ IN SOMALI
49
Verbal form Grammatical designation Rt. Ex. Int. Lk. End.
gpysay (19) 3 sg. f. past gen. ext. of g$o (IN), see under g$o (5) goy # s ay
goostay (20) the same as goostay (15)
t?gay (20) the same as t$gay (14)
t§giy (21) 3 sg. m. past gen. dvg. A of t&g (Z), see under tagtay (1) t?g # # ay
ha4 goosat&y (21) 3 sg. f. past gen. dvg. A of goosd (IN/SAN), see under goostay (15) goosa # t ay
sjddaa (22) 3 sg. f. pres. gen. cvg. B of sjd (Z) ‘ to carry ’ sjd # # d aa
keentay (22) 3 sg. f. past gen. ext. of kden (Z), see under kden (5) keen # # t ay
gpysay (23) 2 sg. past gen. ext. of g$o (IN), see under g$o (5) gqy # s ay
sheegtay (24) 3 sg. f. past gen. dvg. B of u shdeg (Z) ‘ to tell (someone something) ’ sheeg # # t ay
qabatdy (24) 3 sg. f. past gen. dvg. A of qab6 (AN) ‘ to catch ; to carry out ; to perform ; to use ’ qab a # t ay
jqojis&y (24) 3 sg. f. past gen. dvg. A of 3<$O]i (IN) ‘ to stop (trans.) ; to halt ’ jqoj i # s ay
meera (25) 3 sg. m. pres. gen. cvg. A of mder (Z) ‘ to roam ; to move around ’ meer # # # a
baqaa (26) 3 sg. m. pres. gen. ext. of b&q (Z) ‘ to be afraid of ’ baq # # # aa
baqaa (27) the same as baqaa (26) above
gglis&y (28) 3 sg. f. past gen. dvg. A of g£li (IN), see under gglisay (16) g$l i # s ay
mqcaanays6 (30) 2 sg. pres. gen. dvg. A of mg/aande (Z+/AYN) ‘ to sweeten ; to make sweet ’ mg/aan ay # s 0
siisd (30) 2 sg. pres. gen. dvg. A of sli (SUN) ‘ to give ’ sii # s 0
hayd (31) 2 sg. pres. gen. res. of h&y (IN) ‘ to hold ; to have ’ hay # # a


50
B. W. ANDRZEJEWSKI
Verbal form Grammatical designation Rt. Ex. Int. Lk. End.
g?li (31) 2 sg. imper. of g§li (IN), see under g$lisay (16) g?l i #
sli (31) 2 sg. imper. of sli (SUN), see under siisd (30) sii #
pggolaatay (33) 3 sg. f. past gen. ext. of pggol&w (Z+/AAN) ‘ to accept ; to agree to ’ Qggol aa # t ay
yee^ay (33) 3 sg. f. past gen. dvg. A of u yde^ (Z) ‘ to call ’ yee< # # a ay
hadashay (33) 3 sg. f. past gen. ext. of h&dal (Z) ‘ to speak ’ (Root hadal ~ hada ~ hadl) hada # # sh. ay
siisay (33) 3 sg. f. past gen. ext. of sli (SIIN), see under siisd (30) sii # s ay
galay (34) the same as galay (14)
Bibliography
Abraham, R. C., Somali-English dictionary and English-Somali dictionary, University of London Press, 1964 and 1967.
Andrzejewski, B. W., ‘ The problem of vowel representation in the Isaaq dialect of Somali,’ BSOAS, XVII, 3, 1955, 567-80.
Andrzejewski, B. W., ‘ Accentual patterns in verbal forms in the Isaaq dialect of Somali,’ BSOAS, XVIII, 1, 1956, 103-29.
Andrzejewski, B. W., ‘ Pronominal and prepositional particles in northern Somali,’ ALS, I, 1960, 96-108.
Andrzejewski, B. W., ‘ Notes on the substantive pronouns in Somali,’ ALS, II, 1961, 80-99.
Andrzejewski, B. W., The declensions of Somali nouns, S.O.A.S., London, 1964. (Sole Agents : Luzac and Company, Ltd., 46 Great Russell St., W.C.l.)
Andrzejewski, B. W., and Lewis, I. M., Somali poetry : An introduction, The Clarendon Press, Oxford, 1964.
Armstrong, Lilias E., The phonetic structure of Somali, Gregg Press, Famborough, 1964. (Reprinted from Mitteilungen des Seminars fur Orientalischen Sprachen zu Berlin, XXXVII, 3, 1934, 116-61.)
Bell, C. R. V., The Somali language, Longmans, London, 1953; Gregg Press, Farnborough, 1968.
Galaal, M. H. I., Hikmad Soomaali, edited with grammatical introduction and notes by B. W. Andrzejewski, O.U.P., 1956.
Hetzron, Robert, ‘ The particle b&a in northern Somali,’ Journal of African Languages, IV, 2, 1965, 118-30.


CHARACTERISTICS OF THE SO-CALLED 6 WEAK VERBS ’ IN SOMALI 51
Johnson, John W., A bibliography of Somali language materials, Peace Corps, Hargeysa (Somali Republic), 1967.
Kirk, J. W. C., A grammar of the Somali language, Cambridge University Press, 1905.
Moreno, M. M., Il somalo della Somalia : Grammatica e testi del Benadir, Darod e Dighil, Istituto Poligrafico dello Stato, Rome, 1955.
Reinisch, Leo, Die Somali Sprache, Vol. Ill — ‘ Grammatik ’, Alfred Holder, Vienna, 1903. Tucker, A. N., ‘ Fringe Cushitic,’ BSOAS, XXX, 3, 1967, 655-80.
Tucker, A. N., and Bryan, M. A., Linguistic analyses—The non-Bantu languages of North-Eastern Africa, International African Institute, O.U.P., 1966.
Tucker, A. N., and Bryan, M. A., The non-Bantu languages of North-Eastern Africa, International African Institute, O.U.P., 1956.
TKojikobckhh, A. K., «IIocjieztOBaTejibHOCTH npeflrjiaronbHbix nacTHix b B3biice coMajin», ftebiKu AcfpuKu, H3flaTeJicTBO «HayKa», MocKBa 1966, CTp. 143-66.
(Zholkovsky, A. K., ‘ Sequences of preverbal particles in the Somali language,’ Languages of Africa, publishers : ‘ Nauka ’, Moscow, 1966, 143-66.)


THE HAMZIYA DECIPHERED
By Jan Knappert
The name Hamziya has been known to Swahili students for quite some time. William Hichens quotes from stanzas 3 and 4 in his Inkishafi,1 although it is doubtful whether he was in possession of a full translation of the text. On p. 17 of the Inkishafi, Hichens mentions the ‘ archaic language ... in which Aidarus cast the vigorous hemistichs of his Hamziya ’.
What is the Hamziya ? The original Hamziya was a poem of 456 lines in the tawil metre of 22 syllables in the line. It was written in Arabic by the famous poet Sharafu ’d-Din Abu Abdallah bin Muhammad bin Sa'idi ’d-Dilasi al-Busiri as-Sanhaji. Al-Busiri, as he is known for short, was born in the village Abu Sir in Egypt; his name is a corruption of the nisbe adjectival form of the name of his native village. He was born there in 608/1213 and died in Alexandria, where he was mubashir, ‘ official ’, in the year 696/1296.2
Apart from the Hamziya, he wrote two other poems that brought him immortal fame ; both have been translated into Swahili.
(1) Kawakibu 'd-Durriyya fimadhi hayri 'l-bariyya ‘ Sparkling stars in praise of the best of men ’. This poem, which is known as al-Burda ‘ the Robe ’ 3 has been translated into Swahili verse under the title Burudai ; two MSS of it are extant, one in the library of the University College of Dar es Salaam, the other one is in my possession.
(2) Duhru 'l-maSadi Saia wazni banat Sifiad ‘ The treasure for the Future Life in the metre of Banat SuSad'. Banat SuSad, * Far away is Felicity,’ was a lyrical poem of the classical Arabic tradition ; al-Busiri wrote his own poem, which also begins with Banat SuSad, in the same metre as a eulogy in honour of the Prophet Mohammed. It has been translated into Swahili under the same title.4
The full title of the Hamziya in Arabic is Kasidatu 'l-Hamziya fi ’l-mada?ihi 'n-Nabawiya, ‘ Hamzated Eulogy in praise of the Prophet ’ ; an alternative title was Ummu 'l-kurd fi madhi hayri ’l-ward, ‘ The Mother of Books on the praises for the best of creation.’5 The name Hamziya refers to the rhyme of the poem which is in hamza (a?).
1 William Hichens, Inkishafi, London, 1939, pp. 156, 174.
2 Carl Brockelmann, Geschichte der Arabischen Literatur, Leiden, 1943, Band I, p. 264 (308) and Supplementband I, 467.
3 Rene Basset, La Borda, Paris, 1893.
1 I am preparing a publication of this poem.
6 Lyndon Harries, ‘ Cultural Verse Forms in Swahili,’ African Studies, XV, 4, 1956, 4, and again in Swahili Poetry, O.U.P., 1962, 5, translates ‘ Mother of Cities ’, following Hichens, Inkishafi, 15.


THE HAMZIYA DECIPHERED
53
The three poems of al-Busiri are extremely popular throughout the Islamic world. I found copies on sale in cheap editions in the bookstalls of Egypt and Morocco. Translations have been made into Turkic and Indonesian languages. Yet al-Busiri has not received the attention from European scholars which he deserves, possibly because his poems do not reflect the orthodox tenets of Islam. This is true, but the popularity of al-Busiri’s works suggests that they do reflect the beliefs of the Islamic peoples to the farthest corners of the Muslim world.
As far as I know there is no translation of the Hamziya into any European language. In order to decipher the Swahili text it was necessary to translate the Arabic first. To this end I used a printed text in the Maghrebi (Moroccan) variety of Arabic script, published in Fas (Fez). It is referred to as A.
Of the Swahili text there are three MSS in existence. One (referred to as O) is in the library of the School of Oriental and African Studies in London. It is the oldest copy known to exist and bears the number 53823. It has the Arabic text with an interlinear translation in Swahili verse, as well as an introduction in Swahili in the same metre (see below). The colophon, in Arabic, reads as follows : ‘ Praise God and His Prophet our Lord Mohammed, his family and companions ; peace. The completion of the blessed Hamziya occurred on a Thursday after the afternoon prayers, in the month Dhu ’1-Hijja of the year 1257 of the Hejira of the Prophet, peace and praises be upon him. By the hand of the poor and destitute servant of God, humble among the beggars, abstaining from sin and neglect, preparing and hoping for forgiveness from the generous One. Uthman bin al-Qadhi bin Haji bin al-marahumu (= son of the deceased) Uthman, al-Faqimi by descent, and Shafi‘1 by adherence, and Ghazi and Fazi by domicile and birth, (open space) My God ! Forgive me and my parents and all the Mohammedans, and may God cause Mohammed to shine and peace be upon him and his followers. Amen.’ The important fact in this colophon is the date, because it gives us the terminus ante quem of the poem ; we shall probably never know how many years elapsed between the composition of the poem and this copy of it. Hichens gives 1792/3 a.d. as the date of the MS, although 1257 a.h. equals 1841 in the Christian era. I think we can infer how Hichens arrived at this error.6 On p. 38 of his Inkishafi he gives as the date of the MS ‘. . . dhu-l-Hijjat 1207 a.h. (= a.d. 1792/3) ’. It would seem that Hichens read a figure 5 as a 0. But the MS gives clearly 1257, for Ar. o equals 5, not 0. It follows that the MS is 50 years younger than had hitherto been assumed.
The text which I refer to as B is a typed transliteration from a MS I have never seen. It was sold to the E.A. Swahili Committee (as it then was) in 1963, by Sharif Ahmad Abdu ’r-Rahman al-Badawy, religious leader, Koranic teacher, physician and astute businessman, who is highly revered in Mombasa as a scholar of great repute ; the Swahili refer to him as Sharifu Rehemani. The text is a
6 Hichens, Inkishafi, 17.


54
JAN KNAPPERT
failed attempt at transliteration ; 7 the typescript and the transcriptional convention used are strongly reminiscent of Hyder El-Kindy’s office, in which case it can be traced back to the library of Sir Mbarak Ali Hinawy.8 B is longer than O and Y, because after A 456 = O 457 = Y 463,9 which is B 466, B adds 17 stanzas of epilogue, the last five of which contain the colophon, as is usual in Swahili poetry :
479
Na tarikhi yake nisizepo kuzidawini
yali ishirini na tisiya mwezi wa mema (nema),
And its date, when I finished compiling it10 was twenty and nine (of the) month of bliss,
480
Wa kuzawa kwake Tumwa wetu jumaa-tatu mbwake Nairuzi yali mwaka mfuma-juma.
Of the birth of our Prophet, a Monday,
it is the year of which New Year’s Day is a Saturday.
481
Na hijira yake Musutafa yali alifu na nyaka sitini na miwili na yana nyuma.
It was a thousand (years) since the hejira of the Chosen One, and sixty and two years and a hundred in addition.11
482
Na mtarajimu ni faini na ye na dali re wau na sini zaziwae na Uthaima.
And the translator is cayn and y and d r, w and s ; he who was born by Uthayma.
7 It is probably this typed version from which Lyndon Harries quotes four stanzas in African Studies, XV, 4. In v. 1 (my BOY 1), Harries translates Muwawazi (this can be so read) as ‘ All-Wise although it is more probably related to mbawazi ‘ compassion In v. 9 (my BOY 9, A 1) he did not realize that mitume yonte ‘ all the (other) prophets ’ was the subject of wakwelaye ‘ how will they ascend as is abundantly clear from the Arabic original ; nor, that kukwelako is short for kukwela kwako ‘ your ascension He therefore missed the point that the whole stanza is intended to bring out the greatness of Mohammed in comparison to all the other prophets. No translation of this type of Swahili poetry should be attempted without a sound knowledge not only of Arabic but also of the Islamic tradition.
8 I have reasons to believe that this assumption is correct because I have seen copies of other texts which came from El-Kindy’s office but must once have belonged to Sir Mbarak Ali Hinawy. If this is correct, then it is probable that this version is identical with the one used by Hichens, for almost all the Hichens papers once belonged to Sir Mbarak. And most of Harries’ material once belonged to Hichens.
0 Y adds 7 stanzas of introduction, also found in B and O, but not counted by me in O. (Y is described below on p. 57.)
10 dawini (Ar. d-w-n II) ‘ to collect (writings), to register, compile ’.
11 lit. ‘ a hundred afterwards ’. Hichens’ computation shows that he assumed yana to mean a hundred. Sacleux gives gana ‘ cent ’. But see p. 55.


THE HAMZIYA DECIPHERED
55
483
Mola murahamu na wazele na aonao kosa kwa sahiha kimaliza umurahima.
Lord, have mercy on him and on his parents and those whom he sees, if he has really perpetrated a mistake, be lenient.
To translate Swahili poetry is not the same as to understand it. It is easy now to construct the name of the translator as Aidarus ; it is less easy to fix the date of the translation from his cryptic words. The 29th day of the month in which Mohammed was born : we know that Mohammed was born on the 12th of Rabi'u ’1-Awwal, a Monday. Now let us assume that 1162 is the correct interpretation of the year in which the poem was translated. 29th Rabi'u ’1-Awwal 1162 = 20th March, 1749, and this was indeed a Monday. The first day of that year, the Nairuzi or New Year’s Day, 1st of Muharram 1162, fell actually on a Saturday.12 These two facts seem to corroborate the assumption that the translation was made, or at least completed on 20th March, 1749.13 The translation of the Hamziya is thus twenty years later than the oldest MS of the Herekali,14 but its language is considerably more archaic, as has already been realized by Hichens.15 There are four possible explanations for this fact. Either the text as we have it today goes back on a much earlier tradition, and Aidarus did no more than put it on paper; or alternatively, he deliberately ‘ archaized ’ his language, using forms which in his time were already obsolete. Thirdly, it is possible that the date of the oldest Herekali MS is not 1728 but a later datealthough I studied it with the greatest care, one can never be sure of oneself. By far the most likely explanation is that the word yana in B 481 does not mean a hundred, as Hichens interpreted it, but that na yana nyuma is a loose remark meaning ‘ and yesterday is behind (us) ’ ; not an unusual note for a Muslim scholar. The date of the translation is then 29 Rabi'u ’1-Awwal 1062 = 9th March, 1652. Unfortunately, this day is a Wednesday. I do not think we shall ever solve this problem. We cannot even find a single fact from Aidarus’ life. All we know about him is that he lived in Pate and that his father was Athumani bin Ali. Although Hichens puts his date at 1749, he cannot have died much later than 1700, as one can see from his genealogy as given by Hichens.16
12 See G. S. P. Freeman Grenville, The Muslim and Christian Calendars, London, O.U.P., 1963.
” This would seem to confirm Hichens’ date for the translation (Inkishafi, 17), quoted without reference by Lyndon Harries, Swahili Poetry, p. 5.
11 See Jan Knappert, Het Epos van Heraklios, Leiden, 1958, p. 109.
15 loc. cit.
16 Inkishafi, p. 10. Hichens gives a genealogy of the Sayid family, probably based on research done for him by Sir Mbarak Ali Hinawy (on this point I agree with Harries, Swahili Poetry, p. 86). In this genealogy, Aidarus is shown as a grand-nephew of Husein bin Sh. Abubakar bin Salim who died a.d. 1635. Even allowing to a certain extent for the astonishing longevity which Swahili poets enjoyed according to Hichens, one cannot place Aidarus’ death


56
JAN KNAPPERT
Metre
The metre of the Swahili version is long Ukawafi : 17 15 syllables in the line, e.g., B463
Na salamu tena [ ishukiya | kaburi yako
And may yet more peace descend upon thy tomb.
The poem consists of 465 stanzas of two lines each, e.g.,
B462
Salamu aleka | ya kwa kulla | alumbilezo
Mola wa jama a | ukitaya | wakusallima.
Peace be upon you from everything He has created,
the Lord of mankind ; when you mention Him He will give you peace.
Throughout the whole poem, every stanza ends in -ma ; in Swahili verse, as in Arabic poetry, rhyme consists of the identity of the last syllable of a line or stanza.
Normally, there are two caesurae in these long lines ; one falls after the sixth syllable, and one after the tenth :
B464
Na sala ingayo | ja misiki | yatukuwayo kwangu yo shimali | au kusi | uliko mwema.
Prayer is like musk which is carried from Me
to thee by the Northwind or by the Southwind wherever thou art, good one.
Since a caesura can only occur at the end of a word, it follows that there are three fixed accented syllables in every line, viz., the fifth, the ninth and the fourteenth : 0 400
ukitumaVisa | kurejea | amali zibi
kakukufiriwa | ni Mulungu | ja vumbi kama.
If you desire ardently to return from your evil deeds they will be swept away like dust by God.
much later than around 1700. This tallies with a computation based on the average interval of time between the generations in two of the lines in which also the dates of descendants are mentioned, viz. between Sh. Abubakar (d. 1584) and Sayid Abdallah (d. 1820 at the age of 100) there are five generations, with an average duration of 39 years. This would again put Aidarus’ death around 1700, as he is a great-grandson of Sh. Abubakar. However, it is unlikely that Sayid Abdallah died as late as 1820 ; if we put his death at 1800, he still lived to be 80. His uncle Talib died as early as 1755. A computation on this basis would put Aidarus’ death in 1692, which tallies with those of his cousin Ahmad (1686) and their far cousin Abdallah (1693). The average duration of generations would then be 36 years, which is still considerably longer than normal. Modern Swahili poets die at a much earlier age : Shaaban Robert died at 51, Sh. Amri Abedi at 49 ; would they have a harder life than the poets of yore ?
17 See for the term Ukawafi Jan Knappert, ‘ Some Notes on Swahili Literature ’, ALS, VII, 1966. See also E. Dammann, Dichtungen in der Lamu-Mundart des Suahili, Hamburg, 1940, pp. viii and 328.


THE HAMZIYA DECIPHERED
57
The Language
The language of the poem is called kingozi by Hichens.18 There does not seem to be a Ngozi region anywhere along the Coast. More likely seems to me the derivation of kmgozi from ngozi ‘ skin, leather so that it would mean ‘ parchment language In any case, Kingozi is the traditional name for that dialect of Swahili in which the most obscure and impenetrable passages of its poetry are composed. The main features of this dialect are the following :
1. Vocabulary. The more one studies Swahili poetry, the more one realizes that Swahili is a very extensive language with an enormous lexicon. No one has yet attempted to compile a thesaurus of literary Swahili, for the simple reason that the preliminary studies for such a work have not yet been completed. What we need first in Swahili philology is complete and annotated editions of all the major literary works, with lexical concordances. Only after that task is accomplished shall we be in a position to assess the extent of the Swahili lexicon. As there is no opportunity to publish the whole text of the Hamziya (it has 465 stanzas 19), I can give here only a specimen of the first 60 stanzas as an illustration of its richness of expression. At the end of the text I have given a complete list of all the words and grammatical forms that are peculiar to this text. It will be realized that a harvest of just over a hundred linguistic peculiarities in only sixty stanzas signifies a considerable deviation from the average Swahili literary text. This list includes (a) Arabic words not found elsewhere in Swahili texts ; (b) some unusual grammatical forms such as archaic perfect formations from verbs ; (c) words of Bantu origin which do occur in modern Swahili but which are here for the first time recorded in a proto-Swahili shape, which is now only found in Kiungwana 20 and in other Bantu languages ; (d) words the meaning of which could be inferred from the parallel Arabic text.
2. Phonemes. Mu'allim Yahya Ali sent me a copy of the Hamziya in his own elegant handwriting, which is here referred to as Y. In it, the learned Mu'allim does not only distinguish the vowels e and o by special vowel signs, but also the alveolar from the dental consonants by drawing the letter t over the former, a device borrowed from Urdu convention and very appropriately adapted to his own language. Furthermore, he marks aspiration following the same convention quite consistently. In the Roman transliteration, I have indicated the aspirated k as kh, aspirated p and t as pc and tc. Dental consonants are underlined : t, tc, d. In unassimilated words the Arabic consonants have been indicated thus : emphatics : t d s z ; uvulars : k g h ; pharyngeals : h 18 Inkishafi, 17. Harries, Swahili Poetry, 14, quotes Sacleux’ Dictionnaire.
19 A has 456, Y 463, B 483, O 466. This includes the introductions and the epilogue of B.
20 See John and L. F. Whitehead, Manuel de Kingwana, le dialecte occidental de Swahili, Lualaba, Congo Beige, 1928.
E


58
JAN KNAPPERT
In the explanatory notes to the text, the Swahili words are always quoted in bold, the Arabic words in italics.
Abbreviations
mod. = modern St. = Standard Swahili
KASIDA YA HAMZIYA
BOY 1
Nanza kwa jin ale bisimillahi lenye adhama na ar-Rahamani muandazi na ar-Rahima.
In the name of God, I begin with His glorious name, the Merciful, the Provider, the Compassionate.
Muandazi translates ar-Razzaq 4 the Provider ’ (-andaa).
BOY 2
Swifa na matelo ya jamali na kuyoweya zisitahikile Moliwangu Mola Valama.
Praises and beautiful laudation
are due to my God, Lord of the Universe.
matelo : normally in Swahili praise poems himidi or hamdi4 praise ’ is found after swifa ; matelo has presumably the same meaning. Sacleux gives matewa 4 louanges ’, but this may be unconnected.
alama from Ar. Salam 4 world, cosmos ’.
-yowea, cf. yowe 4 crying, shouting ’.
BOY 3
Saia na salamu za dawama mada dahari zimutetelele Muhammadi mwenye makama.
Prayers and salutations perpetually for the duration of the world must illuminate Mohammed, the one of high rank.
mada see A 48.
dahari 4 the world, our time ’.
-tetelela 4 to sparkle for ’, cf. cheche 4 spark ’.
BOY 4
Na alize zote na sahaba na wandamizi na wandamizao wandamizi kwa sira njema.
And all his family, his friends and followers,
and the followers of the followers who led virtuous lives.
ali ‘ family ’, from Ar. ?al.


THE HAMZIYA DECIPHERED 59
Wandamizi translates Ar. at-tabfiuna ‘ the followers ’, which is customary in this context.
sira, lit. ‘ journey, biography ’, is in Swahili mostly used for6 life of Mohammed, exemplary conduct ’.
BOY 5
Baada ya sayo nyi ndu zangu nataka dua Mola ayasiri nitakayo kuyawandama.
After these words, you my brothers, I want a prayer ;
May God make it easy for me to follow them.
-yasiri, from Ar. yassara, yuyassiru ‘ to make easy, facilitate ’.
-wandama, the w- is apparently part of the root.
BOY 6
Lianna niliko nina niya faka kutenda kitenzi kya siyi Hamziya zake kalima.
Because I am here with the intention wishing to make a versification of this Hamziya, (of) its words.
Lianna : BO li?anna 4 because ’.
kitenzi, cf. utenzi, the name of a type of poetry.
siyi = modern hii.
kalima, Ar. 4 word ’.
BOY 7
Azitungilewo Muhammadi mwane SaVidi ali Abusiri utoleo wenye nadhima.
He who composed them was Mohammed, son of Said, he was (from) Abusiri ; (it is) a poetic product.
-tungile, -ile perfect of -tunga 4 compose ’.
-wo rel. suff. cl. 1. mwane = mwana wake.
Abusiri, see the Introduction, pp. 52-3.
utoleo, lit. 4 gift ’.
nadhima 4 poetry ’, Ar. nazm.
BOY 8
Pcindi anzilepo kuhatibu mwenye makama kwa sitifahama ali mwenye rikaba thama.
When the one of high rank began to preach,
with a question he got into the stirrup there and then.
anzile, -ile perfect from -anza 4 begin ’.


60
JAN KNAPPERT
sitifahamu 4 question ’, from Ar. istifham. The first stanza begins with a question.
rikaba or rikabu 4 stirrup ’, Ar. rikab.
Here begins the text of the Arabic Hamziya, which is marked A.
A 1, BOY 9
Hali wakwelaye kukwelako mitume yonte uwingu usiwo kulotewa ni moja sama
How can any of the prophets climb like you did to heaven with which no heaven vies ?
Hali = Ar. hal, a particle for introducing a question.
In wakwelaye, -ye translates A kayfa ‘ how ? ’
-kwela translates A tarqa ‘ will they climb ’.
kukwelako translates ruqiyyaka ‘ your climbing ’. -kwela must therefore be the older form for modern Sw. -kwea 4 to climb ’.
yonte (class 4 because in older literary Swahili mitume retains its own class) has been 4 corrected ’ by Y to yote 4 all ’. The form yonte still shows the -n-, postulated in it by Meinhof {Bantu Phonology, Berlin, 1932, p. 117).
usiwo kulotewa translates A ma tawalatha sama?un ‘ with which no heaven will compete ’. Sacleux gives for -otea 4 guetter, epier qqn avec l’intention de le tuer... ou de surprendre un secret ’ ; -lotewa must be of the same root.
sama = uwingu, both translate A sama?.
This verse refers to the Mi‘raj, Mohammed’s ascension, which I have discussed in detail in Swahili XXXVI, 2, 1966, and in my Traditional Swahili Poetry, Ch. iv. None of the other prophets of history have been summoned by God to travel through the seven heavens and meet Him face to face without veil or screen.
A 2, BOY 10
Khawafani nawe rifaani ihajizile nuru za rufaa kati kwenu kulo adhima.
They are not your equals ; elevation held back
the lights of sublimity (from) in your midst, where there is greatness.
Khawafani translates A lam yusawuka ‘ they do not resemble you ’. Kha- is the older form for modern ha-, negative prefix.
rifaani translates fi Sul aka 4 in your elevated position nuru translates A sanan ‘ brilliance, splendor ’.
kati kwenu translates A dunahum ‘ outside them, leaving them aside ’. The translator apparently missed the point of the poet, that : qad hala sanan minka dunahum 4 a brilliant light passed from you leaving them aside ’ : hala can have the meaning of 4 to evade, fail to meet ’. The poet argues that the light of divine


THE HAMZIYA DECIPHERED 61
grace that shines forth from Mohammed’s appearance is not found in other prophets.
ihajizile, constructed with -ile perfect from -hajizi < Ar. hajz 1 to hinder, prevent, block, keep away from ’, cf. hajiz ‘ obstacle
kulo < kulio, kuliko.
A3, BOY 11
Walimithilile swifa zako kuliko wantu ja maa yalio kumathili ndani nujuma.
They have levelled your qualities with (those of other) people like water that (tries) to compare with the stars.
-mithilile is the verb -mithili ‘ to compare ’ with the -ile perfect ending, ja translates A kama ‘ like ’.
maa is given by Sacleux as the equivalent of mai ‘ water ’ : it translates A alma? ‘ the water ’. The probable meaning of the stanza is : Some people have attempted to ignore the exceptional qualities of Mohammed, and think of him as a normal person, but water can no more extinguish the stars than anyone can ignore the shining example of the Prophet. The Ar. original mattalu means ‘ they have made equal, have made similar ’.
A 4, BOY 12
Uwe ndiwe tala ya fadhila na mayonjeya nuru khazilawi ilia mwako nuruni njema.
Be thou a lamp of virtue and divine favour, all lights emanate from thy good light.
tala translates misbah, so it is the older form of modern taa ‘ lamp ’. mayonjeya : sic BY ; O has majujiya ; for the translation see A 19. khazilawi translates ma tasduru ‘ do not go out ’ ; kha- is the older form of
modern ha-, negative prefix, -lawa is the older form of literary -awa ‘ to come out, appear ’. This line is thus literally translated from the Ar. : ‘ There goes out no light unless it be from thy good light.’ This line is a reference to the belief generally held in Islam, that all light in the universe emanates from God via Mohammed’s mystic being. See M. Horten, Die Religiose Gedankenwelt des Volkes im heutigen Islam, I, Halle, 1917, pp. 12-13.
A 5, BOY 13
Dhatu ’l-?ulumi zilawazo kwa VAlimi ’1-Ghaibi ndako ni masimu ya Adama.
The Essence of all knowledge which emanates from the Knower of the Unseen is yours ; it is the names of Adam.


62
JAN KNAPPERT
Dhatu = dhati6 essence often used as an allusion to God.
Vulumi = ulumu ‘ sciences, knowledge (in the plural)
-lawa 6 to appear see the previous stanza.
?Alimu ’1-Ghaibi ‘ He Who knows the Invisible ’ is one of the names of God. masimu, pi. of isimu, from Ar. ?ism ‘ name ’. This passage refers to the legend
of Adam, who was instructed about all the names of all creatures in the universe by God Himself. This included all God’s own names, so that Adam partook in the knowledge of the Invisible, which is often used as being itself one of the names of God. For Adam being taught the names of the creation, see my Traditional Swahili Poetry, Ch. ii.
A 6, BOY 14
Tcumwa khulasile masitoni ya kuwoneka kuteluliwa-we umahati na abu wema.
O Prophet, you never ceased to exist in the concealment of being, for you there were chosen communities and good parents.
khulasile, khu- for modern hu- ‘ you not ’ ; -lasile, -ile perfect of modern -ata, -acha 6 to leave off ’ ; Sacleux gives a Ngazija form -latsa as the equivalent of -acha. Apparently, -wacha is a secondary form, not the older one.
Masitoni = mafichoni, translates fidamafiri.
kuwoneka 6 to be visible ’, is supposed to translate al-kaunu ‘ existence ’. Possibly the translator was groping for kuweko 6 Dasein ’. There is no doubt that the Arabic author meant the opposite of being visible. This is part of the Islamic ‘ Christology ’ : Mohammed never ceases to exist, although his existence is hidden from mortal eyes. For the post-existence of Mohammed see Horten, op. cit., p. 56 ; Tor Andrae, Die Person Muhammeds in Lehre und Glauben seiner Gemeinde, Uppsala, 1917, pp. 333, 354.
kuteluliwa-we translates tuhtaru ‘ thou art the chosen one ’. Hence we can assume connection with mod. kuteua ‘ to select ’, pass, kuteuliwa.
umahati, Ar. plural of ?ummah, adopted in Swahili in two forms, umma ‘ community, congregation ’, and umati ‘ crowd ’.
abu, lit. ‘ father ’.
A 7, BOY 15
Hapapisilepo tindikano la kulla tcumwa ilia kibashiri ambiyaa kwawe kauma.
There did not pass any period in which prophets appeared without these prophets prophesying about thee to their flocks.
Hapapisilepo translates A ma madat ‘ not has passed ’. It must therefore be the -ile perfect from the verb -pita ‘ to pass ’.
tindikano translates A fatratun ‘ a period, an interval of time ’.


THE HAMZIYA DECIPHERED
63
-bashiri ‘ to bring good tidings, bring a glad message ambiyaa, Ar. anbiya?, pi. of nabii ‘ prophet
This verse refers to the Islamic belief that references to the advent of Mohammed can be found in the holy books of all prophets, such as Jesus, Moses, Zarathustra and Buddha. These prophets, it is believed, have announced Mohammed as the last and truest prophet. See my Traditional Swahili Poetry, pp. 152, 163, 181.
A 8, BOY 16
Zatafahari azmani na kutukuka kwawe marutaba baadaye rutaba jema.
The times are proud and elated because of you (and go through) elevation after elevation.
azmani, from Ar. azman, pi. of zaman ‘ time, period ’ ; it translates A al-^usuru ‘ times, epochs ’, pi. of Sasr.
marutaba, from Ar. martaba ‘ rank, degree, station ’ = rutaba < Ar. rutba ; these words translate A A 9, BOY 17
Palidhihirile kuwoneka kwawe karimu utoseleo kwa karimu waishe kurama.
There appeared into existence from thee the noble one, and from the noble one noble fathers.
palidhihirile, see A 28.
kuwoneka translates lil-wujudi ‘ to existence ’ ; see A 6. utoseleo, ile- perfect from -toka 6 come out, appear ’. waishe, pi. of ishe ‘ father ’ ; it translates A dba?uhu ‘ his fathers kurama, Ar. kurama?, pi. of karim 6 noble, generous
A 10, B 19, OY 18
Nasabu wadhani yo sharafu kwa zipamboze nyota ya jauza i bangili ula nujuma.
You reckon your ancestry noble for its adornments, the constellation of Gemini is a necklace of sublime stars.
Nasabu ‘ line of ancestors, genealogy ’. wadhani translates tahsibu ‘ you reckon sharafu ‘ nobility
zipamboze, St. vipambovye. jauza = A jawza? ‘ Gemini ula = \ Sula ‘ august ’.


64
JAN KNAPPERT
bangili, so marked diacritically and vocalized by Y, it translates A qalladatha
‘ makes a necklace for her ’ (sc. out of the stars). One could also read ipangilwe 4 it has been arranged, strung ’. In either case the meaning is that Mohammed’s ancestry compares with that of other men like Gemini with other constellations, and at the same time the poet metaphorizes this constellation as a necklace of bright stars round the neck of his Prophet.
A 11, B20, OY 19
Pendaza kifungo kya siyada na ufahari uwe kifungoni u kijuwo muutasima.
Acclaim the unison of authority and fame, thou art in that unison the well-guarded book.
Pendaza, sic O and Y ; it translates habbada 6 acclaim ’. kifungo 4 knot ’, translates Siqdu ‘ knot ’. siyada, Ar. abstr. to say id i4 lord ’.
kijuwo can be either a diminutive of uo 4 sheath, book-cover or of chuo 4 book ’.
It could also be connected with -jua 4 to know ’, and would then have the meaning 4 instrument of knowledge ’, i.e. a book.
muutasima translates A aftasma? (of the same radicals) 4 precious, excellent ’. kijuo supposedly translates alyatimatu ‘ the unique one ’ (?) ; it is not clear
what the translator had in mind.
A 12, B21, OY 20
Napendaza tena uso wako ungawa juwa, ufunukiwao ni usiku mwelupe mwema.
I also acclaim thy face, it is like the sun,
when it is unveiled, the night becomes white and beautiful.
-pendaza, see All. Y has pendeza. ungawa, sic O ; perh. ungwa ; in Y absent.
ufunukiwao translates A asfarat ‘ begins to shine, discloses, unveils itself mwelupe = modern mweupe 4 white ’.
A 13, B22, OY21
Usiku wa kuzala ni sikuye ambao kwamba pawele sururu na fahari ya dini thama.
The night of (his) birth was (like) a day which there was joy and pride for the religion there.
kuzala translates maulidi, so it equals modern kuzaa. pawele translates kana ‘ was ’. sururu Ar. 4 joy ’.
fahari translates A izdiha ‘ pride, radiance, heyday thama Ar. 4 there, them ’.


THE HAMZIYA DECIPHERED
65
A 14, B 23, OY 22
Pakafuatana na bishara za mbilingano kwamba lizaziwa Mustafa pamele zema.
There followed the glad tidings of invisible voices,
that Mustafa ‘ the Purified One ’ had been born ; good things had
come true.
Mbilingano translates hawatifu 6 invisible voices
lizaziwa, modern alizaliwa ; it translates A milida ‘ was born
zema translates hand? ‘ happiness, bliss St. : vyema.
pamele translates haqqa ‘ to prove true, be confirmed ’ ; modern pamemalizika.
A 15, B24, OY 23
Ukalewalewa ukumbingwa wake Kisari, ni muujizao khapangali wako kunama.
The main gallery of the Emperor tumbled down,
it was a miraculous sign for you that your building did not (even)
rock.
-lewalewa ‘to stagger, like a drunkard’, translates A tadaSa ‘to cave in, collapse, subside ’.
ukumbingwa translates A liiwanu ‘ a sitting room opening on the courtyard through an arcade ’ ; this is reminiscent of ukumbi ‘ corridor, galerie ou veranda dans le sens de la longueur d’une habitation ’.
Kisari or kesari, normally kaisari ‘ emperor ’. AOY have kisray. This is a crasis between kaisari and hosrau, the name of the last powerful Sassanian emperor of Persia.
This stanza refers to the legend which narrates that when Mohammed was born, the thrones of the monarchs of the earth were overturned. (Tor Andrae, Die Person Muhammeds in Lehre und Glauben seiner Gemeinde, p. 62.) Horten {Die religiose Gedankenwelt des Volkes im heutigen Islam, p. 21) relates how the pillar galleries of the palace of Hosrau trembled and the pinnacles tumbled down, in the night when Mohammed was born. This latter version comes remarkably close to this verse.
A 16, B25, OY 24
Pakaswiri sapo kulla nyumba moya ya moto zijalile ghammu kwa kuzima na baa zima.
And it happened that every single house of fire
was filled with grief because the fires were extinguished in a cold trial.
zijalile, modern zimejaa.
ghammu ‘ grief ’, translates A kurba 6 grief, worry ’. kuzima translates humudi‘ extinction ’. baa = balaa ‘ trial, visitation ’.


66
JAN KNAPPERT
This episode refers to the tradition that the holy fires which the Persians worshipped, were suddenly extinguished in the night of Mohammed’s birth. See Horten, loc. cit.
A 17, B26, OY 25
Mato ya Furusi yafusile hal pawele kuzima kwa soyo moto wao wa kunguruma.
The eyes of the Persians became hollow : What was (the cause of) the extinction of this roaring fire ?
Furusi, Ar. Furs, one of the words for Persians.
yafusile translates gharat ‘ became hollow, sank deeply in ’.
A 18, B 27, OY 26
Kuzawa kukulu paliwele kufa na waba kwao zini zao ukafiri waliko Tama.
This great birth meant death and disease
for them ; adultery and infidelity are rife with them.
-kulu, modern -kuu.
paliwele translates kana 4 has been ’.
waba, Ar. waba? ‘ disease, epidemic ’ ; A also has wabalu ‘ curse, evil ’, which in literary Swahili may become wabaa and so merge with the other word, kufa is probably the translation for this.
kwao translates Salayhimu ‘ over them ’.
zini za ukafiri is supposed to translate A ft tai ft i 'l-kufri 4 the lucky star of heathendom ’ ; zini however can only mean 4 adultery ’, one of the gravest sins in Islam, and therefore presumably attributed especially to the Magi.
Varna (not in A) 4 general, common ’.
A 19, B 28, OY 27
Kuhaniwa kwayo kwa Amina na mayonjeya ambayo Hawa asharifuwa kwayo kadima.
Blessed was Amina with this by the same (divine) favour with which Eve was honoured in olden times.
Kuhaniwa translates harifia ‘ salutary, beneficial ’. O has kutangiwa. mayonjeya translates alfadlu ‘ the (divine) favour ’. See also A 4. asharifuwa translates shurrifat ‘ she was honoured ’.
A 20, B 29, OY 28
Nyani furahiyo mwana Hawa kwamba tukule mimba yake Tcumwa au ndake fadhili njema.
Who (describes) your joy, noble Eve, that she carried the Prophet in embryo, or that hers was good divine favour.


THE HAMZIYA DECIPHERED 67
Nyani furahiyo translates A man li Hawa ‘ Who to Eve . . . The meaning is not clear in either language, but see below.
tukule translates A hamalat ‘ she carried ’ ; it is -tukua + he.
au ndake etc. translates au annaha bihi nafasa? ‘ or that she lay in childbed
with him ’. The translator presumably confused nafasa? ‘ in childbed ’ with nufasa? (note the vocalization) ‘ precious things ’.
This line refers presumably to the legend which tells that Amina had a very light pregnancy (see Horten, op. cit., p. 22), so that we must rather translate : ‘ Who noticed with Eve ’, etc.
A 21, B 30, OY 29
Siku apeteo kwa kuzala binti Wahbi fahari ambazo khazipatwi ni wake wema.
(That was) the day at which she obtained for childbearing fame which is never obtained (even) by good women.
apeteo translates nalat ‘ she obtained
kwa kuzala translates biwadftihi ‘ for being delivered of him ’ (Mohammed). Wahb ‘ gift name of Amina’s father.
khazipatwi translates ma lam tanalhu ‘ which they (fern.) do not receive ’. wake translates annisa? ‘ the women ’.
A 22, B 31, OY 30
Amina kendeya kaumuye na afudhali kama tukuao kablaye ni Mariama.
Amina went to stay with her clan (pregnant) of a better (prophet) than Mary had carried before her.
kendeya translates atat6 she arrived ’.
afudhali translates btfafdala mimma ‘ with a better one than . . . ’.
Jesus, the son of Mary, is, of course, never referred to as ‘ the son of God ’ ;
in Islamic literature he is merely the last prophet before Mohammed.
A 23, B 32, OY31
Amuzazilepo amlaku kamurahamu
shafa kwa kwambakwe kafariji yetu mitcima.
When she had born him, He blessed him with possessions, his power to heal by his word comforts our hearts.
Amuzazilepo translates idh wadaSathu ‘ When she had been delivered of him ’. amlaku is left unchanged : Ar. amlak ‘ goods, possessions ’. shafa, Ar. sifa? ‘ healing ’.
kwa kwambakwe translates biqawlihi ‘ with his word ’.


JAN KNAPPERT
68
A 24, B 33, OY 32
Alimuzazile atukule kitwakye juu na kutukuani mo ishara ya kulla mema.
She bore him, and he raised his head high,
and in this raising there was a sign of all good things.
atukule kitwakye translates raffia ra?sahu ‘ raising his head
mema. The translator allowed himself considerable liberty here. A has wa fl
dalika 'r-rafii ila kulli sudadin ?lma? ‘ And in that raising there was a gesture to his whole dominion ’. This refers to the legend which says that Mohammed was shown to the four corners of the earth which his word would soon dominate (see Horten, op. cit., pp. 26-28). Another interpretation of this gesture is that Mohammed began to pray God as the first thing in his life (see Tor Andrae, op. cit., p. 63).
A 25, B 34, OY 33
Matoye kiyenga luwinguni na maloleo ya ambao lake yambo tukufu na ya Tadhama.
His eyes looking up to heaven and his glances to Him whose honour is elevated and of greatness.
kiyenga translates ramiqa ‘ staring, gazing ’.
luwinguni translates tarfuhu 's-sama?i ‘ in the direction of heaven’. This shows that it is identical with modern uwinguni4 to heaven, etc. ’. It is the first occurrence of the prefix of class 11 with the 1-phoneme preserved.
maloleo translates wa marmcf short for marma-nnazari ‘ range of vision ’. maloleo is therefore to be connected with *-lola, modern -oa 4 to look ’.
lake yambo is intended to translate sa?nuhu ‘ his glory ’. The word sa?n can also mean 4 thing, matter ’, and that is probably why yambo 4 thing, matter ’ was used for its translation.
A 26, B 35, Y 34
Pakakurubiya kulikoye nujumu mbazi yakata miyanga kwa angale majimbo thamma.
There came near to where she was a brilliant star causing its luminous rays to penetrate in all directions.
mbazi from *-bala, modern -waa 4 to shine, be light ’.
angale from *-angala, modern -angaa 4 to be light, bright ’.
jimbo 4 region ’, prob, from Ar. janba ‘ direction ’. A has arja? ‘ directions ’.
thamma lit. 4 there ’.


THE HAMZIYA DECIPHERED
69
A 27, B 36, Y 35
Zilungu za yumbe zikawona zilizo Rumu kawona na alo nyumbanimwe Maka mukima.
The palaces of the Emperor, which are in Byzance, became visible ; even someone who was resident in Mecca could see them from his
house.
Zilungu, modern viungu, translates A kusuru ‘ palaces, castles yumbe, modern jumbe, translates A kaysara ‘ emperor zikawona, St. zikaoneka, translates A tarctfat ‘ became visible ’.
Rumu : all the Swahili texts and A have this, but Y has in his Ar. text as-Sami
‘ of Syria
mukima from Ar. mukim 6 resident, domiciled ’ (Wehr). A has batha? 6 plain, level land, open country ’.
alo = mod. aliye.
The bright star which becomes visible at the moment of Mohammed’s birth is so brilliant that it throws its light as far as the palaces of the Byzantine emperor, foreshadowing the defeat of Christianity against Islam. See my Traditional Swahili Poetry, Ch. iii.
A 28, B 37, Y 36
Palidhihirile miujiza kamuswanimwe isositamana kulikoye mato mazima.
There appeared miracles during the time he was suckled which were not hidden for healthy eyes.
Palidhihirile, with the -ile suffix of the perfect.
kamuswanimwe, from -kamusa 6 cause to squeeze (milk) out of ’ (but see the next stanza), translates A firidcfiihi ‘ at the place of/during the time of his fosterrelationship ’. This refers to the time when Mohammed was suckled by Halima, a woman of the clan of the Banu Sacdin. See MajmuS Mawlidi Sharafi d-Anam, ed. Suleiman Marca, Aden-Singapur, n.d., p. 38.
-sitamana, St. -fichamana.
A 29, B 38, Y 37
Wamuyizazapo wamwisaji kwa wufuzana kamba khatu mafa kwa muswani suyo yatima.
The wet nurses refused (to suckle) him because of his orphanhood, saying : ‘ We have no breasts for the suckling of this orphan.’
Wamuyizazapo translates ida ?abat ‘ when they refused ’. It must therefore be connected with -iza ‘ to refuse ’.
Wamwisaji translates murdftatu ‘ wet nurses ’. It must therefore be connected


70
JAN KNAPPERT
with *-mwisa ‘ to suckle which is reminiscent of -nwesha ‘ to let drink caus. of St. nywa ‘ drink Meinhof writes {Bantu Phonology, Berlin, 1932, p. 55) : ‘ The root for “ to drink ” is probably *mwa.’ This could be the origin of -mwisa.
kwa wufuzana translates liyutmihi ‘ because of his being an orphan kamba translates kulna ‘ they (fem.) said khatu, modern hatuna.
mafa < mafua ‘ breasts
kwa muswani, see the previous stanza. An alternative etymology would be :
< mwiswa ‘ to be suckled ’. Cf. mu- (prefix 2nd person sing.) < mwi-. See Meinhof, op. cit., p. 123.
The MajmuS tells us (p. 38) that it was customary for the women of Mecca to have their infants suckled by the women of the Bedouin tribes. A has ma fi 'l-yutmi ?anna ghanctfu ‘ What profit is there for us in an orphan ? ’
A 30, B 39, OY 38
Kamuyiliyaye mwana muke min ali Saadi Wamuketelewo kwa wutofu wenye ghulama.
A woman from the clan of Saad came for him,
whom the (other) women with babies had refused because of his
poverty.
min ali Saadi, restored from the Ar. text, min ‘ from ’.
-mu- in kamuyiliya has written above it ai an-nabii ‘ this is the Prophet ’. -ketele, -ile- perfect from St. -kataa ‘ to refuse ’.
kwa wutofu translates A lifaqriha ‘ because of her poverty ’. ghulama ‘ boy, slave ’, but here ‘ a sucking infant ’.
The moral of the history is that it is bad to refuse to suckle babies without fathers (Mohammed’s mother Amina was still alive, but he is already called an orphan). The wet nurses expect money for their services, except Halima who, as we shall see, is richly rewarded.
A 31, B 40, OY 39
Akamwamusaye maziwaye naye kanosha lye na wanawe na mbuzi maziwa mema.
She suckled him and she herself received to drink, with her sons and goats, good milk.
-amusa translates ardaSathu ‘ she suckled him ’, cf. -amua in Meinhof, op. cit., p. 55. Y has akamumwisaye, see A 29.
kanosha translates fasaqatha ‘ and quenched her ’, prob. < -nwesha. iye = yeye.
na wanawe translates wa baniha ‘ and her sons ’.


THE HAMZIYA DECIPHERED
71
na mbuzi ‘ with the goats The translater misunderstood the Arabic : albana-hunna 's-sa?u ‘ the sheep with their milk (pi.) As-sa?u ‘ sheep ’ is the subject of fasaqatha (see above) ‘ and gave (fern.) her to drink Apart from the grammatical evidence, this is made explicit in O, which is glossed as-sa? qvzt saqat-, and Halima qnzv -ha.
A 32, B 41, OY 40
Mbuzi kbukutavu kyelewazo kutaka dume zikatwewa pashi muhitaji mwenye kurama.
Shrivelled goats in the morning needed the male,
and in the evening there was none in need (any longer) of the
Generous One.
khukutavu, cf. -kukutaa ‘ se ratatiner par la secheresse ’ (Sacleux). kyelewazo translates asbahat ‘ they (fern.) were in the morning ’. zikatwewa translates wa ?amsat ‘ they (fern.) were in the evening ’. pashi translates ma biha 6 there is not with them (fern.) ’. Mod. pasi.
Cf. MajmiiS p. 42 : Halima said : ‘ There had been no milk in my breast,
and my baby son had disturbed me all night long because of his severe hunger. When I took on Mohammed (peace be upon him), I had been weak, but I became at once stronger and the pain I had felt left me. I put my breast into his mouth and the milk was at once stirred up and began to flow over overwhelmingly. And I heard a heavenly voice (hatifu) saying : “ Blessed be thou, Halima, with the care for the Hashimite ”.
(p. 44 : Halima is still narrating :) ‘ And we travelled to our settlements and wherever we sat down under a tree in the desert, that tree became green and bore fruit in due course, owing to the bar aka of Mohammed (peace be upon him). We had some misshapen and emaciated (goats or ewes ?), so I took Mohammed’s hand and placed it on them and they became full of milk.’
This last sentence tallies almost verbatim with the Arabic text in this stanza. The Swahili translator has apparently misunderstood it.
A 33, B 42, OY 41
Pawele ulufu pa Halima kwisakwe shidda kamba paswirile kwazo kuwa kula kwa Tcumwa.
There were thousands with Halima, her hardship was finished ; she said : it happened with these, they were the food for the Prophet.
Pawele from -wa ‘ to be ’. ulufu, Ar. pi. of ?alfu ‘ 1000 ’.
-swiri, from Ar. s-y-r ‘ to become, begin, happen ’. kwazo ‘ with them ’ refers presumably to the goats.
Again, the Swahili translator did not understand the Arabic text : ‘ Halima’s


72
JAN ƘNAPPERT
life became fertile after her barrenness, when nourishment for the Prophet came forth from her.’
A 34, B 43, OY 42
Ai nema yake ya ajabu alongelezwa ajri na jaza kwa jinsi ya zake zema.
O, wonderful blessing was accumulated for her, reward and recompense of the kind of her own good gifts.
nema = neema.
alongelezwa from -longela, St. modern -ongea ‘ to increase ’.
jaza from -jazi.
zake zema = vyake vyema.
A 35, B 44, OY 43
Mola atiyapo ziwumbeze kumuhudumu alio saVidi nao piya su?udi njema.
When the Lord made His creatures subservient to him who is fortunate, they too will be fortunate.
kumuhudumu translates sahhara * He made subservient ’.
suudi in Swahili is a noun, ‘ happiness, good luck ’ ; the Ar. text has suSadalu,
which is the plural of the adj. saSid.
A 36, B 45, OY 44
Iwele upunde kitindokye uzazileo zisika fukara wakifuza jani kuuma.
It happened that the grain sprouted its shoots and the weak stalks raised a leaf to dry.
upunde (Y gives ubonde, but that is unlikely in view of what follows, and of the
Arabic original) translates A habbatun ‘ grain ’.
kitindo (Y gives kitendo ; apparently he did not understand this line) translates
A sanabila, pi. of sunbula ‘ ear of corn ’ ; Sacleux gives for -tinda ‘ percer ’ which can mean ‘ to sprout ’, of young plants.
uzazile translates anbatat ‘ grew said of plants and grass.
zisika translates A wa ’l-Sasfa ‘ and the stalks (of the grain) ’. Sacleux gives
kisiki ‘ souche, trognon, chicot, eteul ’, which must be related.
fukara translates ad-duSafa? * the weak (pi.) ’ ; literally it means ‘ the poor ’. wakifuza translates yastasrifu ‘ look up, raise (one’s glance) ’.
kuuma may have been placed here for the rhyme. The only word left in the
Arabic text is the key-word ladayhi‘ in his presence ’.
The significance of this verse is that in the presence of Mohammed plants


THE HAMZIYA DECIPHERED
73
began to grow richly and the grains in the earth began to sprout. He is not only a messenger of God’s word, but also master of life on earth. See my ‘ The figure of the Prophet Mohammed according to the popular literature of the Islamic peoples Swahili, XXXII, 1961, 29.
OY 45
Aliziwiliya mamwisoni hifadhi yake na kuteuliwa tangu mwando Mtumi mwema.
He withheld during the time of his being suckled, his keeping, and he was chosen from the beginning as the good Prophet.
Aliziwilia translates harrama ‘ he forbade, declared taboo ’. mamwisoni translates al-maradila ‘ the wet-nurses ’. hifadbi translates sawnun ‘ preservation ’. kuteuliwa translates ijtiba? ‘ selection ’.
The implications of this verse are not clear to me. Not in A.
A 37, BOY 46
Eamupelekea jaddi yake amazizepo
kwa mwana matungu kwa fisala yakimuluma.
She sent him to his grandfather, when she had finished (nursing him), for the boy (it meant) suffering, the weaning caused him pain.
amazizepo translates wa qad fasalathu ' when she had weaned him ’. kwa fisala translates minfiraqihi‘ since the separation ’.
yakimuluma from -luma, modern -uma ‘ to bite, hurt ’. It translates fiha al-buraha?u ‘ there was in her affliction ’. The Swahili translator overlooked the fact that -ha is the feminine object ending. The poet means that the wet-nurse feels the weaning of her foster-child more than he.
A 38, BOY 47
Wamuziugilepo malaika yake Muhmgu kadhanni kwambao ni majiimi kwakeye mwema.
The angels of God surrounded him,
she thought they were jinns near the good one.
Wamuzingilepo translates A id ahatat ‘ lo, they surrounded ’. -ile perfect from -zinga ‘ surround ’.
Mulungu translates Allahi ‘ God ’ ; it is the expected older form of modern Mungu, literary Muungu ; preserved in its original form in several other Bantu languages.
kwambao translates annahum i that they ’, so < kwamba wao.
majinni translates qurana?u ‘ companions ’. The translator apparently
misunderstood this word.
F


JAN KNAPPERT
74
A 39, BOY 48
Wakamurudia wawenepo mahaba yake na kwa pendo sizo moto ndani ukimukima.
They returned to her when they saw her love and by this love a fire inside him blazed.
Wakamurudia does not have its equivalent in the Arabic text, wawenepo translates wa ra?a ‘ and he (they) saw ’.
mahaba and pendo translate wajd ‘ passion, strong emotion, ecstasy of love ’.
ndani translates al-ahsa? ‘ the bowels ’.
moto translates lahibun ‘ fire ’.
ukimukima translates tasli bihi ‘ burnt in (for) him ’.
A 40, BOY 49
Farikene naye akiyiza kwamba liwele mukimu paliye ukimina wusho kukoma.
She parted from him with regret because he had been a guest with her and his guesthood never tired her.
Farikene translates faraqathu ‘ she parted from him ’ ; < -farikana. akiyiza 4 reluctant ’ translates kurhan ‘ with aversion ’. liwele for aliwele translates kana ‘ he has been ’. mukimu (Ar. 4 resident ’) translates tawiyan ‘ as a guest ’. ukimina translates tawa? ‘ status of being a guest ’.
wusho (written wuso ; a better transliteration would be wusyo) from wusio
4 which not... Cl. 14 ’.
kukoma translates yumallu ‘ to tire ’.
paliye 4 there was for her ’ translates ladayhi ‘ with her ’.
A 41, BOY 50
Kcatiwa kuliko moyo wake kyake kifuwa katolewa nyama kuyoshani nyelusi dama.
Where the heart was his breast was cut open
and by washing the black bile was removed from the flesh.
kcatiwa translates suqqa ‘ was split open ’. The verb -k'ata is in Mombasa always pronounced with the first C aspirated, as Y has it. Note that no prefixes are required for tense- or person-indication.
kuyoshani translates ^inda gaslihi ‘ by washing him ’. It has the locative suffix -ni which can be employed with infinitives to indicate a temporal or causal clause. See my 4 Some notes on the locative in Swahili ’, Swahili, XXXVII, 1, 1967, pp. 65 ff. -yosha, modern -osha 4 to wash ’.
dama translates mudgatun ‘ bit, morsel, small piece of meat, embryo ’. It is no


THE HAMZIYA DECIPHERED
75
doubt identical with damu, from Ar. dam 4 blood In the legend the reference is to 4 black clot of blood ’, see my 4 Ukawafi wa Miiraji Swahili, XXXVI, 6, 1966, p. 114 and Traditional Swahili Poetry, Leiden, 1967, Ch. iv.
nyelusi translates sawda?u ‘ black It must be the older form of modern nyeusi 4 black
This stanza refers to the well-known episode of the breast-cleaning of Mohammed.
A 42, BOY 51
Fumbile watuu mukonongwa wa Jiburili uweshezwa ambayo ahbari k'azikusema.
The right hand of Gabriel closed (his breast ?)
and the Tidings that were not pronounced were placed in him.
Fumbile translates hatamathu 6 she (the hand) sealed him ’. -ile perfect from -fumba 4 to close ’.
mukonongwa translates yumna ‘ right hand ’ ; -ngwa may be just a syllable to fill the line, see A 15.
watuu could not be identified.
uweshezwa translates udfta ‘ was placed ’ ; like the latter, it is the passive of the causative ; the basic form is -weka.
ahabari translates anba?u ‘ news, tidings, prophecy ’.
kcazikusema translates lam tudaS ‘ was not disseminated ’. The meaning of this stanza is that the Archangel Gabriel enclosed in Mohammed’s heart all the divine messages that were not revealed to him : he knew them by divine operation.
A 43, BOY 52
Kilihifadhile siri zake sikyo kifungo kwamba uvundifu wa kifungo wakuyuyuma.
This seal preserved His secrets
because breaking the seal would cause (the secrets) to leak out.
Kilihifadhile translates Sana 4 he (it) preserved ’. sikyo older form of hicho 4 this ’. kifungo translates hitamu ‘ seal ’.
wakuyuyuma sic O, Y. My translation is a conjecture. The Arabic text uses a negative construction ; the Swahili translation is rather free.
A 44, BOY 53
Lizowele taVa na ibada na tafarrughi tangu ali mwana nayo ndiyo shani ya wema.
He became familiar with obedience and service and devotion since he was a child and that is the honour of the good.


76
JAN KNAPPERT
Lizowele translates ?alifa ‘ he became familiar It is the -ile perfect of -zowea. There is no subject prefix.
ta tafarrughi translates al-halwata ‘ seclusion, retirement ’. wema translates an-nujaba?, pi. of najib 1 noble, distinguished ’.
A 45, BOY 54
P'indi wulongofu ungiyapo katika moyo kuliko ibada zawaziba zilungo kwima.
When the divine guidance entered his heart his limbs became energetic in worship.
P'indi translates wa?ida ‘ when
wulongofu translates hidaya ‘ guidance, esp. from God ’ ; it must be the older form of modern uongofu ‘ guidance ’.
ungiyapo for Standard uingiapo.
kuliko ibada translates fi Sibadati ‘ in worship ’.
zilungo translates al-a$da?u ‘ the limbs ’ ; it must be the older form of Standard viungo ‘ limbs ’.
zawaziba translates nasitat 1 (they) become lively, energetic ’. It cannot so far be identified with a modern Swahili word.
A 46, BOY 55
Mola tumiliza zamanini ya kuletwakwe zimondo kulinda pakundule pakakundama.
God sent at the time of his mission
brilliant stars to guard him ; there they unfolded and blinked.
Mola tumiliza translates ba Sata Allahu ‘ God sent ’. Note the absence of pronominal and temporal prefixes.
kuletwakwe ‘ his being sent ’ translates maWatihi ‘ his mission ’.
zimondo translates suhuba ‘ shooting stars, meteors I cannot trace any
Swahili etymology of it.
pakundule is the -ile perfect of -kundua ‘ to unfold, open the eyes ’. It is, however, possible to regard it as an -e- noun in class 16. It would then translate A alfadctfu i space ’. A has wa daqa Sanha al-fada?u ‘ space narrowed because of these (lights) ’. If this is correct, then pakakundama translates daqa ‘ became narrow ’ ; Sacleux gives for -kundama ‘ se plier, se replier ’. One thinks here rather of curtains, although this image can be used for the sky, see Koran 21, 104 ; 39, 67. The idea of the Arabic poet is that while Mohammed lived in seclusion in the desert, God sent him stars as lamps lest the dark sky would seem terrifyingly unfathomable to him.


THE HAMZIYA DECIPHERED
77
A 47, BOY 56
Kitokoza jinni makaloui ya kusikiza kama tokozwao mbwa-mwitu wayinga-nyama.
They chased away the demons from their listening position just as the wolves are chased, the meat-hunters.
Kitokoza translates tatrudu ‘ they drive away
makaloni translates San maqaSidi ‘ from their positions ’ ; it is therefore the older form of makaoni, from -kaa ‘ sit, dwell
mbwa-mwitu translates ad-di?abu ‘ the wolves
This stanza refers to a legend according to which some jinns are always eagerly listening to the word of God. The Koran (46, 29 ; 72, 1) relates how they were sent by Allah to listen to Mohammed’s preaching and returned to their own people to spread his message.
A 48, BOY 57
Kwa sababu siyo zikamaya za makuhani aya za wahayi zisimaya madai dawama.
For this reason are the miraculous signs of the priests wiped out the signs of divine inspiration never disappear.
The verb -maya which occurs twice in this stanza, translates the Arabic verb maha, inf. VII immiha? ‘ to be obliterated, to vanish ’. It is not in the Swahili dictionaries, but this double occurrence establishes its meaning.
dawama, lit. ‘ eternity, perpetuity ’, here used adverbially.
madai, from Ar. mada ‘ duration ’, (and as prep.) ‘ for the duration of ’.
It will be noticed that the last line has one syllable too many ; the solution
will be either in reading wahi or mada.
A 49, BOY 58
Hadija muwene na takwaa na kupa nyongo dunia na haya ili zake huluku njema.
Hadija saw him with awe, and it gave her a hatred (for the) world, and modesty became (one of) her good qualities.
muwene translates ra?athu ‘ she saw him ’ ; it is the perfect of -wona, modern -ona ; note the absence of pronominal prefix.
takwaa, — A taqwa ‘ awe, respect, devoutness, piety ’.
nyongo-dunia ‘ gall of the world ’ translates az-zuhdu ‘ renunciation, indifference to worldly things, abstemiousness, abstinence, ascetism ’.
huluku translates sajiya * nature, natural disposition ’.
In the Utenzi wa Mwana Hadija we read how Hadija first met the Prophet Mohammed, her future husband, and was at once struck by the divine light that


78
JAN KNAPPERT
shone from his forehead : kazagaa kama nyota ‘ he was as bright as a star She was so overcome by his beauty that she invited him to her house and offered him all her money as an investment for his commerce.
A 50, BOY 59
Kajiwa habari kwamba wingu na muti mule zilimukingizize zivulize za jio zema.
The news reached her that a cloud and a tall tree protected him (like) the pleasant shades of the evening.
muti mule translates as-sarha ‘ arbre tres grand ’ ; so mule is in class 3 and not in class 18.
zilimukingizize translates azallathu ‘ gave him shade ’.
zivuli za jio translates afya?u ‘ evening shadows ’. Jio = modern jioni.
This episode is also found in the Utenzi wa Mwana Hadija. Mohammed and his uncle were travelling through the desert, when the Almighty saw that His Prophet was suffering from the fierceness of the sun (kwa ukalifu wa yuwa). Allah then orders Gabriel to fly quickly to Paradise and to tell Ridhiwani, its guardian : Nakupe hima kiwingu ‘ Let him give you quickly the little cloud nalikyo kiwumba tangu which I have created in the Beginning, send it to My Prophet,
and protect him from the sun ’.
mupekeye Tume wangu ukamufunike yuwa.
Jiburili akatoka kwa upesi na haraka kitwani mwa Nabiyaka kiwingu kamuwekeya.
Gabriel went out
with speed and celerity and over the Prophet’s head placed the little cloud for him.
Alipovuwa manenga When he (Mohammed’s uncle) raised his eyes uwene yuu la yanga he saw high in the sky kitwani mwake Mmanga over the head of the Arab (Mohammed) kiwingu kikizagaa. a small bright cloud.
It is one of those touching scenes that are intended to show how much God loved his Prophet. In Islamic literatures, shade is a symbol of divine protection ; in southerly countries, shadow is regarded as a blessing : the souls of the faithful in Paradise enjoy a vast expanse of shadow (Koran 56, 30).
A 51, BOY 60
Akajiwa tena ni hadithi kwamba waadi wa Tcumwa kuletwa tekelezo lime kwejema.
She was also told that the promise of the Prophet of God being sent was about to be fulfilled.


THE HAMZIYA DECIPHERED
79
tekelezo translates al-wafa?u ‘ the fulfilment
lime kwejema translates hana ’’ approached (in time) -ejema is in modern Sw. -egema.
List of Words and Grammatical Forms not Recorded Outside the Hamziya, vv. 1-60
alama world BOY 2
ali family BOY 4, 8
amusa to suckle A 31
an gale brightness A 26
bashiri to prophesy A7
dahari world, era BOY 3
ejema to approach A 51
elupe white A 12
elusi black A 41
farikana to part (recip.) A 40
fisala separation A 37
fusile to become hollow A 17
fuza to raise (eyes) A 36
hajizi to obstruct A 2, 3
haniwa to be blessed A 19
huluku nature A 49
-ile (perfect tense suffix) BOY 2, 8 ; A 2, 6, 7. 13, 14, 16, 17, 18, 28, 33, 36, 38, 40, 43, 44
ishe father A9
iye she A 31
ja like (adj.) A3
jalile filled A 16
jauza Gemini (constellation) A 10
kadima old times, the past A 19
kalima word BOY 6
ketele refused A 30
kha (negative prefix) A 2, 4, 15, 21, 29, 42
khu (do., 2nd pers. sing.) A6
kijuwo book A 11
kilungo limb A 45
kilungu palace A 27
kima to burn A 39
kimondo meteor A 46


80
JAN KNAPPERT
kisika stalk A 36
kitenzi poem BOY 6
kit in do shoot(noun) A 36
kukutavu shrivelled (adj.) A 32
kuyuyuma to leak A 43
kwela to climb A 1
lasile ceased A6
lawa to appear, come out A 4, 5
lewalewa to stagger A 15
lianna because BOY 6
longeleza to accumulate A 34
lotewa to be spied A 1
luma to bite, hurt A 37
luwingu heaven A 25
maa water A3
ma da, madai for the duration of (prep.) BOY 3, A 48
mafa = mafwa breasts A 29
makalo position A 47
malole looks, glances A 25
matelo praises BOY 2
maya to vanish A 57
mayonjea divine favour, i.e. wisdom A 4, 19
mazize finished A 37
mbazi < wazi bright A 26
mbilingano heavenly voice A 14
Mulungu God A 38
mumwisaji wet nurse A 29
muswa suckling period A 28, 29, 31
muutasima precious, well-guarded A 11
mwisa to suckle A 29
mwiso 5/6 = muswa OY 45
nadhima poetry BOY 7
-ngwa (nominal suffix (?)) A 15, 51
nosha to quench A 31
nyongo-dunia renunciation A 49
pakundule space A 46
pendaza to acclaim A 11, 12
rifaa elevation A2
rikaba stirrup BOY 8
rutaba exaltation A 8
sharifuwa to be honoured A 19
sira biography ; good example BOY 4


THE HAMZIYA DECIPHERED
81
sitahiki to be due ; to deserve BOY 2
siyada authority, rule A 11
tafarrughi seclusion A44
takwaa devotion A 49
tala lamp A4
tekelezo fulfilment A 51
teluliwa to be chosen A6
tetela to sparkle BOY 3
tindikano period A7
tokoza to drive out A 47
tukule carried, raised A 20, 24
ukimina guestship A 40
ukumbingwa palace hall A 15
ulumu knowledge, science A5
upunde grain A 36
utoleo gift BOY 7
uvundifu breach A 51
wahayi divine inspiration A 48
wandama to follow BOY 4, 5
wantu people A3
waziba to become energetic A 45
wele was A 40
wesheza to place A 51
woneka to exist A 6, 8
wufuzana orphanhood A 29
wulongofu guidance A44
wutofu poverty A 30
yasiri to make easy, facilitate BOY 5
yiza, yizaza to refuse A 29, 40
yonte all (cl. 4) A 1
yosha to wash A 41
yowea to cry BOY 2
zazile gave birth A 22, 36
zaziwa to be born A 14


NANDI RIDDLES FROM NORTH-EASTERN CONGO
By Lyndon Harries
Introduction
Riddles are an important genre of the folk-literature of the African peoples. While the primary purpose of riddles appears to be an exercise in the association of ideas based upon close observation of natural life, they also extend to ideas about the supernatural and about what could not possibly exist. From what can be observed, associated ideas arise of what cannot be observed. The element of didacticism is not a primary motive in riddles, but they do reflect the canons of normal behaviour within the tribal context.
Riddle-contests are not exclusively an activity of child-life. The present collection of some 200 riddles was made in a single afternoon, and my informant was a man. There is no reason to believe that he was especially gifted as a teller of riddles. Other men of his tribe could have provided an equally rich store of riddles, because they are the possession of a people rather than of a single individual. Memory would seem to be more important than creative thinking in this respect. Any creative activity associated with the telling of riddles is a communal one.
The Bantu languages spoken on the Congo side of the Ruwenzori Mountains— the Mountains of the Moon—in the districts of Lubero and Beni have been provisionally classified as belonging to ‘ the Nande group ’. Bryan states, ‘ The name Nande is in common use as a general term covering a number of dialects, including those known as Yira.’ 1 The present collection was taken down in the town of Kitsombiro in the Yira dialect, and the name of the language was given as ekiNandi, with a close final vowel. It is a 7-vowel language with contrast of vowel quality so narrow that in continuous contexts it is very difficult to record accurately. This is a characteristic of these north-eastern Congo languages, especially of Nyanga, spoken in the Masisi area, and which is included by Bryan in the Nande group, but which for various reasons should have separate classification.
Guthrie has mentioned Ndandi (close vowel in suffix) as a Northern Nande language,2 with prefix oru-, but the material on the languages of this group is so sparse that no comparative analysis is yet possible. The most that we know is the names of the dialects. Fr. Guibert Baudet has written an outline grammar and vocabulary of 6 Nande ’, but he does not state which dialect is the basis for his observations.3 He mentions that the final vowel in 6 Nande ’ is a close vowel, but
1 M. A. Bryan, The Bantu Languages of Africa, O.U.P., 1948, 95-6.
2 M. Guthrie, The Classification of Bantu Languages, O.U.P., 1948, 40-1.
3 Guibert Baudet, Elements de grammaire kinande, Brussels, [1947?], pp. 185.


NANDI RIDDLES FROM NORTH-EASTERN CONGO
83
he does not even note that the language is a 7-vowel one. Comparing the grammatical structure evidenced in the following riddles with Fr. Baudet’s notes, one finds some marked differences, e.g. tense-signs do not always coincide, and the differences are clearly dialectal. There is an unusual variety of tense-signs in this language. Fr. Baudet gives no less than 24 tense-signs in primary affirmative tenses, and a further 10 or so, recorded in these riddles, may be added to this number.
Besides recording the riddles, I was able to do some work with informants on the structure of the language, for without this the grammatical interpretation of the riddles would not have been possible. No examples were recorded of lexical tone in nominals and verbals. An unusual feature in the general Bantu field is that in Nandi nomino-verbals are in Cl. 5, e.g. eritsuka ‘ to begin ’, with a negative form eritenditsuka ‘ not to begin Grammatical tone does occur, but is comparatively rare because of the abundance of tense-signs. Tonal alternance was observed on verbal prefixes.
The symbols adopted here for indicating vowel qualities go from open to close reading left to right, viz. e/i/i and o/u/u. Note the following :
omuviri body omuviri work enguvi skin
enguvi orphan
-tola pour, shed
-tula break, tear
-tula hold, press
Particular features of the language will be noted as they occur in the riddles. Although Fr. Baudet notes two separate phonemes, viz. g and gh, this is really a single phoneme, a voiced velar fricative pronounced very back and very softly. Note also that the symbol v stands for a bilabial which often is written as b, and that 1 and r are allophones.4
The Formulae
The preliminaries to a riddle-contest may be as follows :
X. Tutsuke esyondekerano.
Let’s begin some riddles.
Y. Inga.
O.K.
X. Vutsura.
A bit of food.
Y. Nyenyi.
Vegetable.
4 This collection of riddles was made on study-leave in 1961 in the Congo on behalf of the SOAS, University of London, to whom I am grateful.


84
LYNDON HARRIES
When a second meaning or solution for the same riddle is wanted, the poser of the riddle says :
Vunga-lwe One like that.
Or : Lune-yo nilunanga
There is another one.
Both of these sentences consist of particles, viz. vu-, prefix, -nga- ‘ like ’, -lwe, particle of reference Cl. 11. Note that the word for ‘ riddle ’, viz. endekerano, may be in Cl. 9/10 or with prefix of Cl. 11. Particle -ne with prefix expresses ‘ being ’ or ‘ having ’; Baudet does not include this in his paradigm of the verb ‘ to be ’, but it occurs in his adverbial examples, e.g. tune mo vuguga ‘ we have nothing in the house ’, lit. ‘ we are inside nothing ’. Note adverbial use of Cl. 14 prefix. Particle ni- is the copula ; runanga may be used separately meaning ‘ another riddle ’, lit. ‘ one-with-like ’.
When a participant is tired and wants to leave off he will say :
Sing’aho.
Leave off there.
Response : Omukali utakivuta.
A woman who does not give birth.
Elision is of very frequent occurrence in Nandi, as in the elision of the verbal suffix in : Sing’aho. It can be established that it is the suffix which is elided, because the same thing occurs when the verbal suffix is modified to -e and then elided, e.g. tutsuk’omuviri ‘ let us begin work ’. In the relative utakivuta the infix -ta- is the negative infix, and -ki- is the tense-sign.
The Riddles
In setting out the riddles we have placed them under the following headings according to the meanings of the answers :
(1) Natural Phenomena
(2) Unnatural Phenomena
(3) Domestic Life
(4) The Animal World
(5) The Vegetable World
(6) The Human Body
(7) Utensils, etc.
These categories are meant only as a general guide, and indicate the general nature of the subject-matter of the riddles.


NANDI RIDDLES FROM NORTH-EASTERN CONGO
85
(1) Natural Phenomena
1. R. Kaenda vungongo.
It goes quietly.
A. Akalusi k’omu’mugende.
A little stream at its source.
Nominals in full isolated form have double prefixes. In k’omu’mugende the prefixes are triple, viz. ka-, the connective of Cl. 13 with vowel elided (the element -a- is not used in Nandi connectives, i.e. not ka-a-), followed by the locative prefix of Cl. 18, and the short form of the nominal prefix of Cl. 3.
2. R. Katido.
It sounds like a waterfall.
A. Eritundi ry’omwiriba.
Drop (of rain) on a pool.
The ideophone do 4 sound of water falling ’, is introduced by the defective verbal -ti.
3. R. Katugira mungiri.
It falls down 4 squash A. Eritundi ry’omuvohya.
Drop (of rain) on earthed trap.
Mungiri is ideophonic.
4. R. Avalya okw’isirya oku vasiasa itwalwira ah’ikw’aha.
Those on the other side will come, we will quarrel here beneath.
A. Emigeri.
Streams of rain-water.
Of streams meeting with consequent troubling of the waters. Cl. 5 nominals have prefix eri-, occurring also as i-, as in isirya. The infix -si- in vasiasa is future tense-sign, not mentioned in Baudet, and distinct from the negative sign -si- which is prefixal except in negative subjunctive. Radical -asa. In itwalwira, future tense sign i- -a- implies obligation. Locative Cl. 16 with ikwa 4 underneath ’.
5. R. Mutwaya y’ikw’eyi mutwahwira’yo.
We went there below and finished up there.
A. Emigeri.
Streams of rain-water.
Of rain falling. In mutwaya the tense-sign is mu- -a- and the verbal stem is -ya ; this tense is repeated in the last word. The prefix and extra suffix of y’ikw’eyi and the extra suffix -yo in the next word are Cl. 9 affixes used of location.


86
LYNDON HARRIES
6. R. Ngwite ngalye syage iviri siringirirene.
I have my two winnowing-baskets of the same size.
A. Oluvula n’ekitaka.
Heaven and earth.
A common way of beginning a riddle is to use the equivalent of 61 have my . . . ’. The stem -wite 4 have ’ is used with zero preradical tense-sign for present time, and to express past time the verb -vya ‘ be ’ is used as an auxiliary, e.g. twavya tuwite 4 we had ’. Syage, possessive stem 1st pers. with prefix of Cl. 10. Siringirirene, the stem is -ringirirana 4 be alike ’, and si- is Cl. 10 verbal prefix.
7. R. Ngwite mbene yage siririra omu’kiharo kiguma.
I have my goat, he does not graze in one valley.
A. Eriuva.
The sun.
The sun shines on many valleys. Siririra, the subjectival concord -i- is assimilated in the vowel of the negative prefix -si-, tense-sign is -ri-, and -rira is applied form of stem -rya 4 eat ’.
8. R. Ngwite muti wage ali ku’nyumu na mbisi.
I have my tree, it is there where it is dry and fresh.
A. Embula n’omwisi.
Rain and sunshine.
The contrast between 6 dry and fresh ’ is that between sunshine and rain. Note that in ali the Cl. 3 subjectival prefix a- does not agree with the short-series nominals nyumu and mbisi.
9. R. Sokulu ayire yaholira ey’isirya ngetsi.
Grandpa has died across the lake.
A. Omulumo w’embula.
Rain-thunder.
Ayire, verbal -ya 4 go ’ used as auxiliary ; yaholira, applied form of -hola 4 die ’, -y- is tense-sign, cf. vayire ivaholira 4 they died at. . . ’. Ey’isirya, locative extra prefix Cl. 9 with nominal of Cl. 5 isirya (full form erisirya) 4 opposite bank ’.
10. R. Kyavundavunda e’mutiri.
It moves away gently on the plain.
A. Ekihohe.
Fog.
Locative extra-prefix of Cl. 9, e in e’mutiri.


NANDI RIDDLES FROM NORTH-EASTERN CONGO
87
11. R. Ngwite kiteyo kyage ngahanda omuvanga w’e’vukeni inategera ky’eno.
I have my trap, I fasten the cord on foreign soil and it goes off here.
A. Ekima.
A rainbow.
The arc of a corded trap is compared to the arc of a rainbow. Ngahanda, contraction of nyikahanda in which -ka- is tense-sign ; w’e’vukeni, triple prefix w-, Cl. 3 in agreement with omuvanga, e-, locative Cl. 9, vu-, Cl. 14, cf. vukeni or vugeni, 4 exile ’ ; inategera, tense-sign -na- is one of at least five present tense-signs in Nandi, viz. -a-, -na-, -ka-, -ama-, -nemu- and -kina-. Note absence of agreement in this verbal with nominal kiteyo.
12. R. Okw’ibwe ryaNdongwe kuli muliro.
At the stone of Ndongwe there is fire.
A. Engununu.
A star.
The reference to ‘ the stone of Ndongwe ’ was not investigated, but belongs to a folk-tale. Note connective -a- in rya-, an unusual feature in this language, and occurring only before personal names.
13. R. Ekihugo kyahiire kalimba yo’yanasigere.
The countryside is burnt, a long path is what is left.
A. Olusi.
A river.
An alternative answer is : Enzira 6 a path ’. Kyahiire, perfect tense of -hia 4 be burnt ’ ; kalimba is Cl. 9, not Cl. 13, cf. keli 4 a serpent ’ ; yo’yanasigere, relative construction with perfect of -sigala, and tense-sign -ana- which usually implies an action still continuing. Subject relatives in Nandi have prefix of the same shape as the 1st demonstrative, and the example here is emphatic relative.
14. R. Kongo lumwe.
A full river drying up.
A. Olusi silwehuma n’amabwe.
A river does not carry away the stones.
Kongo 4 a river in full spate ’, an interesting connotation of a well-known name ; lumwe, a long-series nominal, cf. -uma 6 dry up ’ ; silwehuma n’amabwe, lit. 4 it does not move with the stones ’. Cf. the next riddle.
15. R. Ngulangwa meno akuka.
My dirty teeth are being pulled out.
A. Olusi numulwangasula siluliheka amabwe.
Even though the river is full it cannot carry away the stones.


88
LYNDON HARRIES
Dirty teeth may still be whole and are difficult to remove, and similarly large muddy stones in the river bed are not easily dislodged. Ngulangwa, passive of -kulanga 4 pull out with contraction of nyi- -k-, viz. ng-. Akuka, cf. -kuka 4 become dirty the prefix is Cl. 5 and the construction here corresponds to that which, with nominal stems, has extra dependent prefix. Numulwangasula, here numu- could be treated as a separate word, but it is better to regard this as a single verbal form with tense-sign numu- -anga-. Siluliheka, negative present, -li- is tense-sign.
16. R. Ngwite kuusi kage kakagenda vuhongohongo.
I have my stream, it goes along quietly.
A. Ol’omu’misola.
The one (stream) in the yam-leaves.
17. R. Hane kiteyato.
There is something which does not sleep.
A. Olusi.
A river.
Hane, Cl. 16 prefix with stem -ne 4 be ’ ; kiteyato, the stem is -yato 4 sleep, lie down ’, with negative infix -te-.
18. R. Obutakondeka.
What cannot be cut.
A. Omwirimya.
The darkness.
19. R. Twahumba kuno kosi.
We hit every place hereabouts.
A. Omwirimya.
Darkness. (It covers everything.)
20. R. Muno muli kalume kage avali kuno vamakahuruvana sivandikatuka.
Here inside is my man, those who are here are pulling hard, they will not manage it.
A. Omululu.
The cold.
An alternative answer is : Darkness. It is not possible to hold on to the darkness or the cold. The verbals here are -huruvana 4 pull roughly ’, -tuka 4 be able, and the tense-signs are -amaka- and -ndika-. The element -ka- occurs as a separate tense-sign, and Baudet gives -ama- (present) and -kandi- (future), but not the forms noted here.


NANDI RIDDLES FROM NORTH-EASTERN CONGO
89
21. R. Hari omundu ngasiva ngamutivita’ko sindimuhika’ko.
There is a person, I spend the day trotting after him but I do not reach him there.
A. Ekitutu.
Shadow.
Nga- in the verbals is contraction of nyika- ; extra suffix -ko (not -ako) is locative, cf. extra suffix -ko as substitute object, an alternative for the object infix, e.g. akatwala’ko 4 he sends it ’ Cl. 13 ; -ndi- is contraction of -nyili, and -mu- is object infix in both verbals. Another answer : Path.
22. R. Ngwite mundu wage ngamuvwira indi tutwe ovwira inyagana.
I have my man, I tell him let’s be friends, he refuses.
A. Ekitutu.
Shadow.
Stem -vwira 4 tell ’ ; indi, conjunctive, cf. indi ‘ I am ’ ; tutwe ovwira, lit. 4 let us cut a friendship ’ ; inyagana, cf. itwahana 4 we must invariably refuse ’, -gana 4 refuse ’, also a future tense implying obligation.
23. R. Engoko-lume sokulu asiga kera ikinavika.
The cock grandpa slaughtered long ago still crows.
A. Omusalala.
A waterfall.
Of the continuing sound of a waterfall. Asiga, for mwasiga, cf. -siga 4 slaughter ’, also -terya ; kera, adverbial ; ikinavika, tense-sign -kina-.
24. R. Ngwite lusi lwage lukagenda vw’iluka.
I have my river, it goes upwards.
A. Amagetsi g’omuseke.
Water through a straw.
Vw’iluka, adverbial, cf. -Iuka 4 ascend ’ ; the Cl. 14 prefix here is an extra prefix before short form of the nomino-verbal.
25. R. Ye sokulu na sokulu evy’agenda kalya nage oyo’nginarya.
He is grandpa and what grandpa eats I also eat.
A. Amagetsi.
Water.
In evy’agenda kalya, object relative evi- (Cl. 8 referring to foods) preceding the subject prefix of the auxiliary -genda ; oyo, emphatic use of demonstrative ; nginalya, 1st pers. nyi- with tense-sign -kina-, a contraction.
G


LYNDON HARRIES
90
26. R. Kavutuka.
Something unexpected, a surprise.
A. Omuliro g’omu’ngunza.
Fire in a pocket/bag.
27. R. Omwetu mulya akaponire tata.
There in our home it scorned my father. A. Esyongulukiro.
Ashes.
Ashes flying about have no respect for persons.
(2) Unnatural Phenomena
28. R. Ihembe ry’omulangala.
A horn of the mulangala tree.
A. Omusya watemo lusi.
A valley without a river.
The smooth mulangala tree has no horns, neither is there a valley which at some time did not have a river. Watemo, a non-verbal form with negative infix -te- and locative -mo.
29. R. Nasondire navulire.
I looked for it and missed it.
A. Omulihi w’eribwe.
The root of a stone.
30. R. Ndimba hano mbulirana.
I go round here, I disappear.
A. Olugove lw’ekirimu.
The trail of a spirit.
Note -timba 4 go round ’ ; -vulirana 4 disappear ’.
31. R. Ndulanga virya.
I break up foods.
A. Wanganarya obongo bw’ekirimu ?
Would you have eaten the brain of a spirit ?
Ndulanga, stem -tula 4 break ’, with intensive stem-extension ; obongo, probably a borrowing from Swahili ubongo.


NANDI RIDDLES FROM NORTH-EASTERN CONGO
91
32. R. Syawite hari.
He has no place.
A. Omungu.
God.
God is omnipresent, with no special place. Other names for God are : Nyamuhanga, Omuhima, Omusavuli, Agalisya.
33. R. Kanyondo mutogotya.
Kanyondo is lost.
A. Oyuholire syali syalola wako.
He who is dead is not seen again.
Oyuholire, subject relative with prefix of same shape as 1st demonstrative ; syali syalola, negative sign si- is repeated in this compound tense, and the verbal -Iola is passive in this language, the active form ‘ see ’ being -vana.
34. R. Erivala ry’eka.
Something impossible belonging to the village.
A. Ekyana ky’embuli sikyekuningirawa n’ekitiri.
A lamb is not covered with a basket.
Erivala, Cl. 5 nominal, ‘ something which is impossible to get or find, a thing that does not exist ’ ; eka, Cl. 9 ‘ town ’ ; -kuningira ‘ cover ’, and note passive in -awa.
35. R. Pupurubulinda.
Something that does not exist.
A. Enda siwite matema.
A louse has no cheeks.
A louse with cheeks is what does not exist. My informant would not accept the riddle here as two words, but even so it is very probable that pupuru is a separate ideophone for ‘ absolutely impossible ’, and that bulinda is an adverbial meaning ‘ improbably ’, though it has not been possible to identify the latter as a Cl. 14 nominal.
36. R. Erivala ry’e’vuhimba.
Something impossible at a homestead.
A. Esyonyumu.
Ants.
The kind of ant indicated here is found only in a banana plantation ; vuhimha ‘ dwelling-place ’, cf. Baudet, vuhimba sesene ‘ a new dwelling place ’. Vuhimba is also the name of a district, see next riddle.


LYNDON HARRIES
92
37. R. Erivala ryaMusindi.
Something impossible in Musindi.
A. Igutavana oko kingeli.
You will not see a kingeli snake there.
Kingeli4 harmless, grass snake ’. Musindi is a district.
38. R. Omuli mune katemo.
In the bush there is something impossible.
A. Olutu lw’eyisuki.
The nest of a house-fly.
Omuli, cf. iri4 bush country ’, also ekisuki; katemo, stemless word with prefix, negative tense-sign and locative, lit. 4 it-not-in ’. Olutu 4 nest ’, pi. esyondu.
(3) Domestic Life
39. R. Tulande visangi vyaNyalanga.
Let us sew Nyalanga’s old clothes.
A. Wanganalanda lu’musoni ?
Would you sew my mother-in-law’s cloth (unspecified) ?
Nyalanga was explained as a name for a mother-in-law. The riddle reflects the respect due to in-laws. Note connective -a- in vya-, not a normal feature in Nandi.
40. R. Twa kanupe.
Spit.
A. Ekinyambo ky’omusoni.
My mother-in-law’s bladder.
Twa, simple imperative, 4 cut ’, cf. also -twa matandi 4 spit ’. Explained as follows : Do not spit if you should see your mother-in-law’s urine.
41. R. Ekyaswa yamasira kayitogaya kw’eminyoku.
There outside it appears, he drops it at its meaty parts.
A. Omuli w’eviseke.
Eater of sugar-canes.
Ekyaswa yamasira, a common way of beginning a riddle ; ekyaswa 4 there outside ’, Cl. 9, cf. ali ekyaswa 4 he is outside/beyond/on the edge of’ ; amasira, cf. -sira 4 appear ’ ; eminyoku, Cl. 4, 4 meaty parts/muscles ’ ; -togaya, causative of -toga 4 fall down ’. A person eating sugar-cane drops the outer covering to get at the sugary pith.