Verbal Extensions in Bantu: (The Case of Swahili and Nyamwezi)
Verbal Extensions in Bantu: (The Case of Swahili and Nyamwezi)
Verbal Extensions in Bantu: (The Case of Swahili and Nyamwezi)
net/publication/228707562
CITATIONS READS
20 2,226
1 author:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Jikri - the sacred songs of the Sidi Afro.descendants of Gujarat, India. View project
All content following this page was uploaded by Abdulaziz Yusuf Lodhi on 28 May 2014.
Abdulaziz Y. Lodhi
Department of Asian and African Languages
Uppsala University
abdulaziz.lodhi@afro.uu.se
* This study is based on a seminar paper presented at the Dept. of Asian and African
Languages in December 1985, after field work conducted in June/July 1983 in a Nyamwezi
speaking immigrant settlement near Kizimbani, Zanzibar. In the preparation of the paper I
have freely used the phraseology of the authors consulted. The text would be over-loaded
with redundant notes and quotations if I were to be particular about this practice. Hence
references are given only where I have found it meaningful to do so.
E. Jonsson (1949) has been taken as the main source for the Nyamwezi material as it is
the most widely used handbook by Scandinavians and easily available too. For Swahili,
Ashton (1944) has been referred to for similar reasons. Additional data on Nyamwezi has
been supplied by Mr Kidumla Kamagi Kidumla of Stockholm, a native speaker of
Kinyang’wezi; and the Swahili material has been verified personally by the present author as
a native speaker of Kiunguja/Kiswahili. I would also like to extend my gratefulness to
Professor Thilo Schadeberg (Leiden) for his very useful comments and my colleagues Mr
Julian M. Kimaro (Tanzania) and Mr Mwandawiro Mghanga (Kenya) for much data on
Chagga and Taita respectively. Some more data and comments were generously offered by
members of the Comparlingafric forum, especially Karen Van Otterloo and Maddalena
Toscano.
Abdulaziz Lodhi · Verbal extensions in Bantu 5
2.1 Passive
This form indicates that the subject is acted upon by an agent. It is not treated in
Bantu as a “voice” opposed merely to the “active”. Apart from the Passive of the
Simple, there are Passive forms of the Applied/Applicative/Prepositional,
Contactive, Conversive/Inversive and Causative extensions. In Zulu even the
Associative or Reciprocal extension has a Passive extension, e.g. -bonana (see
one another) > -bonanwa (be seen mutually), and the Passive itself has a
Reciprocal extension, e.g. -bona (see) > -bonwa (be seen) > -bonwana (be seen
mutually); but very few verbs have these methods of formation and in such
cases there are no apparent differences in meaning. The passive reciprocals and
the reciprocal passives are used only with Class 17 (locative) markers, e.g. in
Zulu (Doke 1963: 138): Kuyazondwa lapha (There is mutual hatred there),
Kwabonanwa emfuleni (There was a seeing of one another at the river). This
possibility occurs also in Kifuliiru (J50), Nyamwezi and the neighbouring
languages with certain restriction. Generally there is no semantic difference in
the two variant forms of the passive.
The Passive is indicated by the post-radical/pre-final element (suffix) -wa,
-ewa, -iwa or -ibwa, and in a few cases by -o, e.g. in Chagga. Some Bantu
languages are known to have lost the Passive form, instead, as in Ngala and
Ngombe, the Stative extension with the suffix -ama is used.
2.2 Neuter
2.3 Applicative
This extension indicates that the action is applied on behalf of, towards or with
regard to some object. Its post-radical element is -ea, -ia, -ela, -ila, -ena or -ina
according to the established rules of vowel harmony and nasal stem. Frequently,
this extension acquires a specialised meaning and is then lexicalised. In the
conventional grammars, this form is called the Applicative, Applied or
Prepositional. In Chagga -iya is affixed even to verbs ending in -ua, e.g. -ondua
(take away) > -ondiya (take away from), -vungua (open) > vunguya (open for
someone, open with something).
2.4 Causative
This extension indicates cause to do, or cause to be. Its post-radical element
varies considerably from region to region in Bantu Africa. In the Causative
forms, Carl Meinhof’s Ur-Bantu *-a occurs as -esa, -isa, -esha, -isha, -eza or
-iza and -sa, -sha or -za. There are also less common but complicated forms with
the original -ya suffix with labial and nasal stems, e.g. in Lamba -ima (get up) >
-imya (raise), Swahili -ona (see) > -onya (warn), -ogopa (fear) > -ogofya
(frighten). Other rare usages are Zulu -lima (plough) > -limisa (help to plough),
Swahili -la (eat) > -lisha (graze) and Taita -ona (see) > -onya (show). In Chagga
there is an irregular -ra causative suffix, i.e. -lala (sleep) > -lara (cause to
sleep), and in verbs loaned from Swahili the original -sha is retained, e.g.
-torosha (to help someone to escape/elope). The most frequently used causative
particles are -sa/-isa and -sha/-isha in eastern African Bantu languages.
2.5 Augmentative
2.6 Intensive
generally the same as the Causative one but it does not contract as the latter
does, e.g. Shona -naka (be good) > Causative -nakisa/-natsa, but Intensive
-nakisa only. Whereas in Zulu, the Causative post-radical -isa is differentiated
with the Intensive -isisa, e.g. -buza (ask) > -buzisisa (ask insistently). At times
the Applied or the Conversive forms are also used to express intensity.
2.7 Extensive
This form indicates that the action is extended in time or space, or repeated
extensively. Extensiveness is expressed with Intransitive, Transitive and also
Causative verbs, e.g. Lamba has in the Intrasitive extension the elements -aka,
-aika or -auka; in the Transitive -ala, -aila or -aula; and in the Causative -asya,
-aisya or -ausya.
This extension is sometimes called the Durative and is very similar to the
Intensive form. The Applied and the Conversive forms are also used to express
extension or duration.
2.8 Reciprocal
This form indicates that the action is reciprocated, done ‘to one another’. Ashton
calls this the Associative (Reciprocal). The usual post-radical element is -na or
-ana and the extension often takes a conjunctive construction with -na or -no,
e.g. Zulu Ngihlangana nomuntu (I met with a person); compare with the Swahili
equivalent: Nilionana na mtu.
In Lamba there are two Reciprocal forms, one denoting reciprocity between
two bodies and the other denoting reciprocity among/between many/several
bodies, e.g. -lekana (divide into two parts) and -lekansyanya (divide into many
parts/units).
2.9 Associative
This form indicates two or more subjects associated in action. Its post-radical
elements are -akana, -ekana and -ikana. Many Bantu languages employ the
Reciprocal form to denote association. Sotho uses the form -ahana. Ashton calls
this form the Stative or Neuter and distinguishes it from that which uses only the
-na element, i.e. the Associative (Reciprocal).
2.10 Reversive
This form indicates an entire reversal of the action. Known also as the
Conversive, this is still a regularly formed extension where there is a marked
8 Africa & Asia, No 2, 2002
shift in the rule of vowel harmony, i.e. the root-vowel -e is grouped with the
vowels -a, -i and -u, and not with -o; thus it harmonises with the vowel -u of the
Reversive post-radical instead of -o. The various suffixes are -ua, -ula, -ulula,
-oa, -ola, -olola and the uncommon -uza.
The Intransitive extension (see also § 2.2) -ka/-eka/-ika, etc., becomes
-uka/-uluka/-oka/-oloka, etc., e.g. Bemba -funga (fasten) > -fungula/-fungulula
(unfasten), and Swahili -funga (fasten/shut) > -fungua (unfasten/open).
Occasionally, these same particles are used to form the Repetitive extension
which denotes a repeating or doing over again. This appears to be a subordinate
idea to the Reversive, e.g. Lamba -wyala (sow) > -wyalulua (resow).
2.11 Perfective
2.12. Reduplication
2.13 Stative
2.14 Contactive
This form expresses contact. Johnson (1939) uses the term Tenacious for this
form. Its characteristic suffix is -ata/-ta. It is not always identified as a separate
verb form in many Bantu languages, e.g. Lamba/Swahili -fumbata (grasp), Zulu
-namatha (stick, pierce).
2.15 Denominative
This form is made from nouns or adjectives, and though rare, a few verbs of this
type usually occur in each Bantu language. The post-radical element here is -pa
or -mpa; sometimes the longer forms -para or -pala are used, e.g.
This term is used to describe verbs formed from ideophones1 and they are
sometimes called Deideophonic Verbs.
This verbal extension is very common in Bantu and relatively little work
has been done on it. The most common formation is by the element -k a
(Intrasitive), -la/-ra/-na (Transitive), -sa/-sha/-za/-ja (Causative, with bisyllabic
ideophones), e.g. Lamba -awu (of crossing over, taking out) > -awuka (cross
over) > the causative -awusya (take across). The equivalent Swahili forms are
-vuka (go over) and -vusha (take across). In Zulu we get -bihli (of falling apart)
> -bihlika (knock apart) > -bihliza (scatter about), etc.
1 Doke (1935: 118) defines an ideophone as “[a] vivid representation of an idea in sound.
A word often onomatopoeic which describes a predicate, qualificative or adverb in respect to
manner, colour, sound, smell, action, state or intensity.”
10 Africa & Asia, No 2, 2002
2.17 Remarks
3.2.1. In Swahili, the characteristic post-radical elements are -(l)ea and -(l)ia,
and in Nyamwezi they are -ela and -ila. The rules of vowel harmony are well
illustrated here. For details, see Ashton (1944: 217) and Jonsson (1949: 92-3).
3.2.2. Both the Swahili and Nyamwezi Applicative forms are used to express: a)
to do to, for, or on behalf of someone, or to the benefit or detriment of someone
or something; b) motion towards; c) purpose; d) finality or completeness; e)
“why?” with the interrogative nini? or the enclitic -ni; and f) “of oneself, by
oneself” with the reflexive -ji-.
The areas of usage of the Applicative form in Swahili and Nyamwezi are
almost identical; the only exception appears to be in the use of the Reduplicated
or Double Applicative form with the meaning “for” or “for the sake of” which
does not occur in modern Swahili, e.g. Nyamwezi Yesu watuchilia biswe and
Swahili Yesu ametufilia (mbali) (Jesus has died (away) for our sake).
3.2.4. In a few cases in both Swahili and Nyamwezi, the Applicative form
conveys a Conversive meaning (see § 3.7.3), e.g. -hama (move away, emigrate)
> -hamia (move in/to, immigrate), -nuka (smell bad) > -nukia (smell good).
3.3.1. The characteristic post-radical stative element is -ka with the same
principles of vowel harmony applying here as in the Applicative form. The
secondary elements are -eka, -ika, -leka and -lika, e.g. Swahili -fanya (do) >
12 Africa & Asia, No 2, 2002
-fanyika (be done), -sema (say) > -semeka (be said), -twaa (take) > -twalika (be
taken), and -tembea (walk) > -tembeleka (be trodden).
3.3.2. In Nyamwezi the Neuter is built in 3 ways: a) Verbs with final -la have
the Neuter post-radical -ka, e.g. -kenagula (destroy) > -kenaguka (be destroyed,
break down); b) Verbs with the pre-final vowels -a, -i and -u take the final -ika,
e.g. -tina (cut down/off) > -tinika (be cut off), and -binza (break off), > -binzika
(be broken);2 and c) Verbs with the stem vowel -e or -o has the final -eka, e.g.
-nona (sweet) > -nonya (sweeten) > -noneka (be sweet/sweetenend).
3.3.3. In Swahili the final -la has disappeared, and identical forms are in -waka
(to burn, be burning), -choka (be tired), etc, where the simple forms -wala and
-chola do not exist any longer.
3.3.4. Both the Swahili and Nyamwezi Neuter forms are used in the same way,
i.e. to express state without reference to agency, and to express potentiality. In a
few cases, potentiality in Swahili is expressed also by the suffix -lika, e.g.
-pasuka (split/tear) > -pasulika; as in Skati lake limepasuka (Her skirt is
split/torn) and Kitambaa hiki hakifai, chapasuka rahisi (This cloth is no good, it
tears easily/It gets easily torn).
Furthermore, in about half a dozen cases, Swahili uses -ikana or -ekana
with the potential meaning, e.g. -pata (get) > -patikana (be available), -ona (see)
> -onekana (be visible). These additional forms do not seem to occur in
Nyamwezi, but are found in several other languages in eastern Africa, e.g. Taita
-patikana (be available) and -wonekana (be seen/sighted).
2 The Augmentative/Extended form of -binza with the final -gula follows rule a) e.g.
-binzagula (break into pieces) > -binzaguka (be broken into pieces).
Abdulaziz Lodhi · Verbal extensions in Bantu 13
t+ya > sa/sha -takata (be clean) > -takasa (be clean)
-pata (get) > -pasha (give)
-pita (pass) > -pisha (let pass, make way)
w+ya > vya -nawa (wash hands) > -navya (help to wash hands, etc.)
-lewa (be drunk) > -levya (intoxicate)
2 vowels + ya > za -jaa (be full) > -jaza (fill) -jaza (fill)
-legea (be loose) > -legeza (loosen)
-tulia (be quiet) > -tuliza (console)
3.4.2. Some verbs with -p-, -t-, -k- and -n- take the particle -isha or -iza:
The verbs -pata and -pita have also the variants -patisha and -pitisha; however,
alternative forms have developed special meanings and usage,3 e.g. -pasha moto
(heat, warm up something), -pasha habari (give/pass on news); -pisha njia
(make way, let someone pass); -pitisha sheria (pass a law/decree). Similarly,
from -ona (see) we get -onya (warn) and -onesha/-onyesha (show).
3.4.3. Loaned nouns and adjectives also take the final -isha in the Swahili
Causative, e.g.4
Arabic: bahati (luck, chance) > -bahatisha (guess, try one luck)
tayari (ready) > -tayarisha (make ready, prepare)
3 The question of the semantic values of the verbal extensions must be left to a separate
paper.
4 For details on the extension of Arabic loan verbs see Lodhi (2000a: 115-120; 2000b).
For the extension of Indic and Persian loan verbs, see Lodhi (2000a: 121).
14 Africa & Asia, No 2, 2002
Swahili has also at least one double causative from the Arabic loan adverb sawa
(properly) > -sawaza (put something properly) > -sawazisha (to balance, adjust,
equate).
3.4.4. The Causative -sa occurs in Old Swahili, e.g. in the Kiamu dialect, where
-pita (pass) > -pisa, and -soma (read) > -somesa. The Kiamu term mapisi
(history, the past), and the rarely used wapiseo (ancestors), found in several
northern Swahili dialects, are also derived from this verb.
3.4.5. In Nyamwezi, the assimilation of -ya is less marked and the Causative
form is more easily derived than in Swahili, e.g. verbs with the final syllable
-ba, -da, -ha, -ma, -na, -pa and -ta all take the suffix -ya in the place of the final
-a, e.g.
-huba (lose ones way) > -hubya (make one lose the way, mislead)
-ganda (get thin) > -gandya (make thin)
-luha (suffer) > -luhya (victimise, make suffer)
-goma (strike, rebel) > -gomya (cause/lead rebellion/strike)
-nona (sweet) > -nonya (sweeten)
-lipa (pay) > -lipya (cause to pay, demand)
-bita (pass) > -bitya (let pass, take round)
3.4.7. The post-radicals -ga and -la are realised as -ja, e.g.
The Applicative forms in Nyamwezi also take the Causative -ja, e.g.
3.4.9. A very interesting use of the Causal -isha post-radical is with the
Nyamwezi verbs ‘to be’ and ‘to have’ when expressing negation, e.g. -gaya (not
to exist) > -gayiwa (to miss, not to have), -gayisha (undo, cause not to exist).
3.4.10. The post-radical -nza is realised as -nja in the Nyamwezi Causative, e.g.
-shiminza (walk, wander) > -shiminja (cause/force to go/move).
-funga (tie, bind, shut) > -fungama (be tied/bound in a fixed position)
-shika (hold) > -shikama (be held tightly)
-ina (bend, archaic) > -inama (stoop, bend down, be bent)
-zala (drown, archaic) > -zama (sink, drown, be immersed)
-tala/-tira (look, archaic) > -tazama/-tizama (look at, gaze upon, examine)
3.5.2. More frequently, the Static -ma takes the Associative extension -na, the
two particles together expressing a state of inter-dependence or inter-action, e.g.
-fungama (be in a fixed position) > -fungamana (be interlaced, allied), and
-shikama (be in a state of holding) > -shikamana (hold together tightly).
3.5.3. Jonsson does not identify this verbal extension in his study of Nyamwezi.
Nevertheless, several entries in his verb-lists (p. 108-111) prove the fact that
both the -m a and -mana post-radicals do exist in Nyamwezi, however
uncommon they may be. The following verbs taken at random illustrate the
occurence of the Static forms in Nyamwezi:
3.6.1. The post-radical -ta in this extension expresses contact and/or indicates “a
culmination, the final conclusion and settlement of a process” (Marconnès 1931;
quoted by Ashton 1944: 237), e.g. Swahili -kama (squeeze) > -kamata (take
hold of, arrest), and -fumba (shut/close e.g. a door) > -fumbata (enclose with
hands/arms), -kumba (pull towards oneself) > -kumbata/-kumbatia (embrace).
In a few cases, the simple forms of the Contactive verbs have been lost, e.g.
-ambata (adhere to, stick to) < *-amba, but > -ambisa (to glue/paste), and
Abdulaziz Lodhi · Verbal extensions in Bantu 17
-guruta (smoothen with a press, mangle clothes) < archaic *-gura (stretch,
straighten, lengthen).
3.6.3. The enclitic -to sometimes used in Swahili is defined by Ashton (p. 158)
as an adverbial, e.g. in the proverb Ukitema kuni, temato! (If you cut firewood,
cut it well!, i.e. If you do something, do it well!). Thus Naona raha (I am
enjoying) > Naonato raha! (I am enjoying very much!), and Nafurahi (I am
happy) > Nafurahito! (I am extremely happy!).
The -to enclitic in Swahili is met with in poetry usually written in the
northern dialects. It is probably derived from the Contactive -ta and is rarely
used in modern Swahili. It is almost non-existent in Nyamwezi.
3.7.1. This form conveys the opposit meaning to that expressed in the root, and
is also spoken of as the Reversive. Here in Swahili, the general rules of vowel
harmony have been modified so that the vowel -e of the stem of the simple form
harmonises with the vowels -a, -i and -u, and not with -o, as is the case with the
other extensions. The Conversive element here is the vowel -o or -u and appears
as pre-final to -a, -ka, -sha and -za, i.e. the Conversive may appear in the Simple
indicative, the Stative/Neuter and the Causative environments, e.g.
3.7.2. Nyamwezi employes the particles -ola, -ula, and sometimes -ila:
3.7.3. The suffixes -ila and -la in both Swahili and Nyamwezi are rarely found
in the Conversive. In Swahili, however, it is unduly classified as a sub-category
of the Applicative form (see § 3.2.4).
3.8.2. The Inceptive extension does not seem to exist as a distinct form in
Nyamwezi, but certain verbs suggest its existence in an assimilated state. The
-pa suffix is rarely encountered, e.g.
The plosive -pa is reduced to -ha in most cases through loss of the bilabial initial
-p- in an intermediate aspirated form -ph- found in archaic speech.
3.9.1. This form is generally called the Reciprocal, but in addition to reciprocity,
its characteristic element -na expresses other aspects of association such as
concerted action, interaction and inter-dependence or disassociation. Generally,
the Associative of the Simple, Applicative or Causative forms expresses
Abdulaziz Lodhi · Verbal extensions in Bantu 19
3.9.2. Nyamwezi also uses the -na as its characteristic Associative element but
less frequently than in Swahili. Thus -lemba (cheat) > -lembana (cheat one
another), -lumba (thank) > -lumbana (thank one another), and -kolwa (hate) >
-kolwana (hate one another).
3.9.3. With some verbs, instead of the extension -na, a phrasal construction with
the reflexive marker -i- is employed in Nyamwezi, e.g. -togwa (love, like) >
-itogwa ng’wenekele (love/like onself) > -itogwa benekele ne benekele (love/like
one another), and kutula (to hit) > kwitula ng’wenekele (to hit oneself) > kwitula
jenekele na jenekele (to hit one another, to fight).
Interaction or Inter-dependence is clearly marked by the -i- (Swahili -ji-)
reflexive in the Associative form, e.g. Nyamwezi kuguna (to help) > kugunana
(to help one another), kwigunana (to help one another in a group), and
kwigunana benekele na benekele (to help one another - 2 persons only). These
forms are comparable to the Swahili kusaidia, kusaidiana, kujisaidiana, and
kujisaidiana wenyewe kwa wenyewe, respectively.
3.10.4. Sometimes both Nyamwezi and Swahili use Reduplication of the verb
stem to convey Augmentation or Continuousness, e.g. Nyamwezi kuyela (to
wander) > kuyelayela (to wander about/around), and Swahili kutanga (to
wander) > kutangatanga (to loiter).
3.10.5. The complex suffix -aga commonly found in Eastern Bantu (but not in
Swahili, except for in the Kingwana dialect of Swahili) expresses the “habitual”
aspect or frequency, generally expressed in Swahili by the use of the hu-
tense/aspect marker, e.g. Nyamwezi kuenda (go) > kuendaga (often/usually go),
and Swahili kwenda (go) > huenda (often/usually go).
there/several times), Alimpigapiga tu (He hit her only gently), and Mtoto huyu
alialia kutwa kucha (This child whimpers day and night).
In Nyamwezi too the Reduplication of the verbform has both Augmentative
and Subtractive implication depending on the context: kulya (to eat) > kulyalya
(eat a little all the time), kugema (to try) > kugemagema (to try a little/not
enough), and kulima (to plough/farm) > kulimalima (loosen the surface, scratch
the soil).
With monosyllabic (short) verbs, in Swahili the whole infinite form is
repeated instead of only the verb root as is generally the rule, e.g. Swahili kula
(to eat) > kulakula (to eat continuously, a little now and then, at short intervals),
Nyamwezi kulya > kulyalya, and Taita kuja > kujaja.
3.12.1. Most Swahili and Nyamwezi verbs, in the Simple and derived forms
(except for the Neuter/Stative and the Associative) may be put into the Passive,
e.g. in Swahili:
gari (I went by car), and Nilikwenda na gari (I went by/with my own car).
Monosyllabic verbs generally take -iwa suffix, e.g. -fa (die) > -fiwa (be
bereaved), and -la (eat) > -liwa (be eaten), but -pa (give) > -pewa (be given),
-pawa in Old Swahili and its Northern dialects.
Occasionally in Swahili, a Passive idea is expressed by the Associative
-kana, e.g. -ona (see) > -onekana (be seen).
3.12.2. The Passive in Nyamwezi has a different set of phonetic rules with
almost no exception, e.g.
4. Conclusions
Verbal extensions are a phenomenon more complex than what it appears to be,
and under its seemingly regular system of vowel harmony and assimilation,
there are some complex modifications; and the post-radical elements in several
cases have more than one implication, sometimes contradicting, e.g. the
common Applicative element is used also to express the Conversive, or the
Augmentative and the Subtractive; and the Conversive is also expressed by the
post-radical elements of other forms.
Most of the verbal extension forms found in Bantu are encountered in both
Swahili and Nyamwezi, though not always to a similar extent in equivalent
forms. Swahili and Nyamwezi extensions are more similar than different.
Nyamwezi, having less foreign influence and more Bantu word-stock than
Swahili, naturally posseses a greater number of Simple forms, e.g. -ina (bend)
and -inga (go out).
Verbal extensions in Swahili are well documented, analysed and described;
whereas in the case of Nyamwezi, verbal derivations and extensions are not all
identified or distinguished satisfactorily. Moreover, for both Swahili and
Nyamwezi, more work needs to be done on the tendency of verbal extensions to
develop special meanings, and this aspect needs to be described with complete
and sensible contexts. The question of the productivity of derivation is again not
fully answered. Ahrenberg, Ashton, Polome, Loogman and Miehe have all dealt
with the the problem in some way as far as Swahili is concerned, but much work
is required for Nyamwezi. In this brief paper, the question of productivity of
verbal derivation and their limits have been dealt with only marginally.
Furthermore, the traditional lables Applicative, Conversive, etc., do not
describe well the meanings of the extension morphemes -el-, -il-, -ul-, -sh-, etc.,
and their semantic descriptions may not be correct. In a different kind of
24 Africa & Asia, No 2, 2002
linguistic study, one may be interested in universal semantic categories, and may
thus want to describe how such categories are expressed in a given language.
Such a study demands solving a host of preliminary problems. The present paper
investigates “verbal extensions” as a morphological and not a semantic category,
though I believe the question of verbal derivation in Bantu needs to be addressed
as syntactic and semantic categories.
Abdulaziz Lodhi · Verbal extensions in Bantu 25
REFERENCES
Rather extensive bibliographies are included in the works consulted for this
essay (except for Jonsson 1949). Many of these other works have also been
consulted for the final preparation of this essay.
Velten, Carl. 1901. Grammatik des Kinyamuesi, der Sprache der Wanyamuesi in
Deutsch-Ostafrika. Göttingen.