|
The two types of causatives in Kinyarwanda: From iconicity
to symbolicity
by Alexandre Kimenyi
This paper examines the two types of causatives found in Kinyarwanda, namely
the suffixes -y- and -iish-. Both causatives are also found in other Bantu languages
as well. So far no detailed and adequate study has been conducted to determine
their respective semantic differences and syntactic constraints. Some studies
suggest that these two morphemes are either doublets or that they are phonetically
related. Although they are phonologically independent of each other, and sometimes
seem to be used in complementary distribution there are cases also where they
are used together within the same verb stem as a discontinuous morpheme. Their
respective use depends on different factors which are either morphosyntactic,
morphophonological and semantic. The causative suffix -y- behaves like any other
derivational morpheme because its use is restricted and unpredictable. But -iish-
looks like other grammatical morphemes because it can be used with any verb
for causativization purposes. The present study is different from my previous
account of causatives in Kinyarwanda (Kimenyi 1980 and Kimenyi 1984). These
studies dealt mainly with the grammatical properties of causatives. Their specific
uses and constraints were not discussed. Multiple causativation was not mentioned
either. The behavior of grammatical relations of the base verb subject (causee)
and its objects was not examined, either. The causee (the former subject) behaves
differently either becoming a double or triple direct object, an adjunct (oblique
object) or an unspecified direct object (deletion). The present investigation
makes contribution to current issues in both syntactic theory and cognitive
studies in its discussion of the form of the underlying syntactic representation
and its relation to cognitive structures. The underlying structure is shown
to be very simple. It consists of the verb, grammatical relations and cases.
There are only three grammatical relations namely the subject, direct objects
which are inherent or structural and adjuncts. Cases of structural DOs are marked
on the verb. Because of the finite nature of structures that grammar can generate,
sentence processing has to use other mechanisms as well for proper semantic
decoding.
1. Constraints on the use of -y- causative morpheme
The causative suffix -y- has exactly the same meanings as its counterpart the
suffix -iish-.
The only difference is their respective degree of productivity. -y- has a lot
of constraints which is characteristic of lexical or derivational morphology
whereas -iish- can be added to any verb thus behaving like other grammatical
extensions namely the applicative and the comitative/reciprocal suffixes.
1.1. Phonetic factors responsible for the non-use of -y-:
The causative morpheme is not found with (i) monosyllabic verbs such as -rí-
'eat'; -nyó- 'drink'; -se- 'grind'; -gi- 'go'; etc., (ii) verb stems
which end with palatal consonants namely (ny, sh, c, j, shy, y, ) as in -béeshy-
'lie'; -reeshy - 'have the same size'; -meny- 'know'; -séeny- 'destroy';
-íic- 'kill'; -vúunj- 'exchange money'; -iinj- 'to substract information';
-maash- 'to practice shooting'; -báash- 'be able'; -gay- 'underestimate/despise'
and labiovelarized consonants (w, nyw, mw, njw,...) as in -aanjw- 'to odd chores';
-neebw- 'get lazy'; etc. This morpheme is not found with verb stems which end
with alveolar fricatives (s,z) and the nasal (n), either.
The reason why it is not found after s and z is because they are mistakenly
considered as mutated consonants. When the perfective suffix -ye, the nominalizer
suffix- yi or the causative morpheme -y- is added to verb stems which end with
r, d or g, this consonant becomes z, and if it ends with t it becomes s (Kimenyi,
1979). As a matter of fact when the applicative suffix -ir- is added to verb
stems which end with z , the z is reanalysed as r+y and the applicative suffix
is inserted between the two as the following examples show:
1a gucúruuza 'to trade' > *gucúruuzira >gucúruuriza
kubaaza 'to sew' > *kubaazira >kubaariza
b. kuráara 'to spend the night' >kuráaza 'to spend the night
with' >kuráariza
kuvúga 'to talk/speak' >kuvúza 'to make talk/play instrument'>kuvúgiriza
kurwáara 'to be sick' >kurwáaza 'to assist somebody sick'>kurwáariza
These same verb stems have z changed into r before -iish-
2. aracúruuza 's/he trades' >aracúruuriisha 's/he trades with'
arabúguza 's/he plays igisoro the African stone game' >arabúguriisha
's/he plays igisoro with'
gusókoza 'to comb' > gusókoreesha 'to comb with'
kwíibaza 'to ask oneself' >kwíibariisha 'to ask oneself'
What is interesting, however, is that even if verb stems which end with the
alveolar nasal n don't allow -y- suffixation, the n of suffixes does as the
comitive suffix -an- and the stative -aan- show in (3):
3. kurwaana 'to fight' (intr) >kurwaanya 'to fight' (trans)
gutóongana 'to quarrel' (intr)>gutóonganya 'to quarrel' (trans)
gutúungaana 'to be well done'>gutúungaanya 'to cause to be
well done'
guhoombaana 'to be squeezed'>guhoombaanya 'to squeeze'
1.2. Morphosemantic factors:
There are many morphemes once they have been added to the verb stem prevent
the suffixation of -y-. Some of them are -ik-, -am-, -aang- and -iriz-. The
morpheme -ik- refers to the position in which something is put in as indicated
by the following examples: -manik- 'hang'; -raambik- 'lay down'; -hágarik-
'make stand'; -cúrik- 'put upside down'; even the ones which have been
lexicalized such as -ziik- 'put under water '; -viik- 'put something in the
water so that it doesn't drip'; -zirik- 'tie'; -hunik- 'store in a granary';
etc. It is thus appropriate to call it a positional morpheme. -am- is a stative
morpheme because it shows the state either physical or mental in which the thing
is in. -am- : -garam- 'to lie on the back'; -pfúkam- 'to kneel'; -ryáam-
'to lie down'; etc.
Interestingly also these verbs are transitivized by adding the positional morpheme
-ik-.
The suffix -aang- refers to iterative events or noisy activities: gucúuraanga
'to play the cithare'; gukómaanga 'to knock at the door'; kurúmaanga
'to eat without relish'; etc.
The non-occurrence of the causative morpheme -y- after the intensifier morpheme
-iriz- as in -vúgiriz- 'to whistle'; -hamiriz- 'to dance' is due to phonological
factors. It already contains the final alveolar fricative z and is thus reanalysed
as a reduplication of the applicative morpheme -ir- plus the causative morpheme
-y- which in combination with r mutate into z.
There are many other verbs which for semantic reasons never take the causative.
They are intrinsically causative thus being decomposable as to cause+V and having
only agentive subjects such as -kúbit- 'hit'; -híing- 'cultivate';
-mén- 'break'; -tém- 'cut'; -shiim- 'thank'; -kiing- 'close';
.... A full list of this category verbs is needed, however, to establish all
their shared semantic features because there are other verbs which have the
same semantic properties which get causativized with this particular morpheme.
2. Causative doubling
There are three types of causative doubling: -y- doubling-; -iish- doubling
and -iish + -y- combination. The -y- doubling and -iish- + -y- are phonetically
motivated whereas -iish- doubling is due to syntax.
2.1. -y- doubling.
The causative morpheme -y- is doubled when added to a verb stem which ends
with an alveolar consonant (s, z, t, d, n, r) with the perfective aspect -ye.
The applicative suffix -ir- is also inserted between the two causatives.
This rule is formulated in (1).
1 ...V+y+ye ---->...V+y+ir+y+ye
atogosheje 's/he just boiled' <a-togot-y-ir-y-ye < a-togot-y-ye
she-boil-caus-appl-caus-asp she-boil-caus-asp
arwaajije 's/he nursing' <a-rwaar-y-ir-y-ye < a-rwaar-y-ye
s/he-be sick-caus-appl-caus-asp s/he-be sick-caus-asp
bageenjeje 'they just rode' <ba-geend-y-ir-y-ye< ba-geend-y-ye
they-go-caus-asp
-y- is also doubled if the comitative/reciprocal morpheme -an- is added to verb
stem.
2. gutogota 'to boil' >gutogosa 'to boil' >gutogosanya (ku-togot-y-an-a)
'to boil with'
kurigita 'to disappear' >kurigisanya 'to cause to disappear with'> kurigisanya
'to cause to disappear with'
kwóoga 'to wash feet/swim' >kwóoza 'to wash' >kwóozanya
'to wash with'
kuramuka 'to survive the night'>kuramutsa 'to greet'> kuramutsanya 'to
greet with'
kurwáara 'to be sick' kurwáaza 'to nurse'kurwáazanya 'to
nurse each other'
gukáraba 'to wash hands'>gukárabya 'to help wash hands' >
gukárabyanya 'to help wash hands with'
2.2. -iish- + -y-
Both -iish- and -y- appear in the same verb whenever -iish- is followed by
an applicative.
3. V-iish-ir... > V-iish-ir-ir-y-
gukóra 'to work' >gukóreesha 'to cause to work'>gukóreeshereza
'to cause to work for'
/ku-kór-iish-ir-a/ >/ku-kór-iish-ir-ir-y-a/
kumesa 'to wash clothes'>kumesheesha 'to cause to wash clothes'> kumesheeshereza
'to cause to wash clothes for'
/ku-mes-iish-ir-a/>/ku-mes-iish-ir-ir-y-a/
gukina 'to play'>gukiniisha 'to cause to play'>gukiniishiriza 'to cause
to play for'
/ku-kin-iish-ir-a/>/ku-kin-iish-ir-ir-y-a/
The applicative suffix is also found with verb stems which end with the intensifier
morpheme -iir/-eer- after vowel shortening when this causative morpheme is added.
4. kumenya 'to know'>kumenyeera 'to get accustomed to'>kumenyereza 'to
accustom'
kureeba 'to watch'>kureebeera 'to watch attentively'>kureebereza 'to cause
to watch attentively'
3, Semantics of causative constructions:
The two Kinyarwanda causative suffixes have some of the meanings found in other
languages causative constructions such to cause S , to have S, to get S, to
make S but others as well in which there is no causer, causation and caused
event. There is thus a semantic fluidity which results in ambiguity. Only the
contect can help in deciding what the true meaning of the causative construction
is.
3. 1. Direct and Indirect causation
Direct causation implies a direct link between the subject and the causee such
as control or resultative event. Indirect causation, on the other hand, presupposes
an intermediary agent between the subject and the causee. Direct causation is
semantically similar to transitivity. In the case of transitivity, the subject
(agent) directly affects the object (patient). Similarly, the main subject of
causative construction directly affects the causee.
It is the morpheme -y- which shows most of the time indicates direct causation
and -iish- indirect causation as we see in the examples provided below:
1a. Umugabo a-ra-ambut-s-a umugoré urúuzi.
man he-pres-cross-caus-asp woman river
'The man is making the woman cross the river' (direct causation)
b. Umugabo a-ra-ambuk-iish-a umugoré urúuzi.
man he-pres-cross-caus-asp woman river
'The man is having the woman cross the river'.(indirect causation)
When -y- is used with verbs with the comitative morpheme -an- it also acquires
the direct causation meaning only.
2. kurwaana 'to fight' (intrans) >kurwaanya 'to fight' (trans)>kurwaaniisha
to fight with'/to cause to fight each other'
gutóongana 'to quarrel' (intrans)>gutóonganya 'to quarrel'(trans)
>gutóonganiisha 'to quarrel with means of'/to cause to quarrel each
other'
2. 2. Instrumental meaning
The other difference between the two causatives is that -iish- has among other
referents, the instrumental meaning but -y- doesn't have this one(1).
3a. Umugoré a-ra-ambuk-a urúuzi n'úubwáato.
'The woman is crossing the river with/in a boat'
b. Umugabo a-ra-ambut-s-a umugoré urúuzi n'úubwáato.
'The man is helping the woman cross the river with/in a boat'.
c. *Umugabo a-ra-ambut-s-a umugoré urúuzi ubwáato.
d. Umugabo a-ra-ambuk-iish-a umugoré urúuzi ubwáato.
'The man is helping the woman cross the river in a boat'.
In 3(d) the instrumental is the only direct object of the causative morpheme
whereas umugoré and urúuzi are direct objects of the base verb.
When both the instrumental and the former subject (the causee) are objects of
causativization then causative doubling takes place as we will see later as
the following example shows:
4. Umugabo a-ra-ambuk-iish-iish-a umugoré ubwáato urúuzi.
man he-pres-cross-caus-caus-asp woman boat river
This sentence means that the woman is made cross the river using a boat by
the man.
2. 3. Causatives without a causer
Causatives in Kinyarwanda are used sometimes to relate not to causers of events
but indidivuals affected by the events instead. This type of causative doesn't
involve any voluntary participation at all in the resultative event. The subject
of the main verb is a beneficiary or a maleficiary of the event.
5. Kagabo y-a-pf-uush-ij-e ababyéeyi.
Kagabo he-pst-die-caus-appl+caus+asp-asp parents
'Kagabo's parents died'
6. Uno mugabo y-a-hi-ish-ij-e inzu.
this man he-pst-burn-caus-appl+caus+asp-asp house
'This man's house burned'.
7. Umugoré a-rwaa-j-ij-e umwáana.
woman she-be sick-caus-appl+caus+asp-asp child
'The woman has a sick child'
8. Kamaáli y-a-giz-e ibyáago a-hoomb-y-a amafaraanga a-na-rumb-iish-a
n'áabáana.
Kamali he-pst-have-asp misfortune he-go bankrupt-caus-asp money he-also- vagabond-caus-asp
also children
'Kamali became unlucky: he went bankrupt with his money and his children
also became vagabond'.
3. 4. Causation as assistance
Causation as "assistance" implies physical or moral assistance but
not control or manipulation of the causee.
9. Umwáana a-ra-kúbuu-z-a umukoóbwa inzu.
child s/he-pres-clean-caus-asp girl house
'The child is helping the girl to clean the house'.
'The child is making the girl clean the house'.
10.Umugoré y-eemer-e-ye umugabo ku-mu-huung-iish-a.
woman she-accept-appl-asp man to-him-escape-caus-asp
The woman promised the man to help him escape'.
11. Umukoóbwa a-ra-kám-iish-a umuhuúngu.
girl she-pres-milk-caus-asp boy'
'The girl is helping the boy to milk'.
'The girl is making the boy milk'.
3. 5. causation as seeking assistance
The other meaning of the causative morhemes is 'seeking assistance'
12. kuraguza 'to seek divination'<kuragura 'to divine'
kugabuza 'to beg food'<kugabura 'to feed'
gucuumbikiisha 'to ask for lodging'<gucuumbika 'to lodge'
In most of the cases, the would be causee behaves as an adjunct when the causative
morpheme has "the seeking assistance" meaning.
13.`kuraguza ku mupfumu 'to seek divination at the diviner'
kugabuza ku mutúuranyi 'to ask for food at the neighbor's
gucuumbikiisha ku munyámahaánga 'to ask for lodging at a foreigner's'.
3. 5. Causatives without a resultative caused event
Not only causative constructions in Kinyarwanda may lack causees, causers but
they also may not produce any caused resultative event, either. The proper translation
will in this case be 'to try'.
14. Abagabo ba-á-gii-ye gu-héemb-eesh-a , ariko amafaraanga
ba-ra-yá-b-iim-a.
men they-pst-go-asp to-pay-caus-asp but money they-pst-it-them-refuse- asp
'The men went to get paid but they refused them the money'.
15. Abagoré ba-a-vug-iish-ije umugabo ntí-y-a-ba-subiz-a.
women they-pst-talk-caus-asp man not-he-pst-respond-asp
'The women talked to the man and he didn't respond'.
Because of these different meanings, causative constructions in Kinyarwanda
are always ambiguous. In the following sentences, for instance, it is not possible
to tell which is the causee and which is the patient in the first example and
in the second example it is not possible to know whether imódoká
is instrument or patient.
16. Umugoré y-a-áng-iish-ije umugabo abáana.
woman she-pst-hate-caus-asp man children
'The woman made the man hate the children'.
'The woman made the children hate the man'.
17. Abapóolisí ba-ra-kurur-iish-a abanyurúru imódoká.
policemen they-pres-pull-caus-asp prisoners car
'The policemen are making the prisoners pull the car'.
'The policemen are pulling the prisoners with a car'.
4. The behavior of the causee in causative constructions.
Causativization affects the causee more that it does other grammatical relations.
The causee is realized as either as a structural DO, an adjunct or as unspecified.
4.1.Unspecified object deletion (UOD)
All Kinyarwanda verbs, besides ditransitive verbs, can appear with or without
direct objects. A small category of causativized verbs also may appear without
objects at all whether causees or patients.
They seem to fall under the predicates which express daily activities such as
guhíingiisha 'to farm', kwúubakiisha 'to build', and verbs with
specific objects such as kuvúunjiisha 'to have (money) exchanged'; gusoreesha
'to tax', gutóoresha 'to make vote'etc.
1. Umugabo y-uúbak-iish-ij-e.
man he-build-caus-appl+caus+asp-asp
'The had (people) build'.
2. Abagoré ba-ra-híing-iish-a.
women they-pres-cultivate-caus-asp
'The women are having (people) cultivate'.
It is not possible at this point to tell which semantic properties all these
verbs share since both active and stative verbs are found in this category.
4. 2. Verbs with obligatory objects
A causativized verb has always to have a specified object whether causee or
patient if it is a ditransitive verb or has a structural object (benefactive
or comitative) or if it is an intransitive stative verb.
4. 3. Verbs with the patient object
The majority of causative constructions appear with the patient object and
the causee object (former subject) deleted.
3. Umugabo a-r-úubak-iish-a inzu.
The only complement of the causative suffix which is never deleted and therefore
has to be overtly expressed all the time is the instrumental.
4a. Umwáana a-ra-andik-iish-a ibáruwá ikáramú.
child s/he-pres-write-caus-asp letter pen
'The child is writing a letter with a pen.
b. Umwáana a-ra-andik-iish-a ibáruwá.
child s/he-pres-write-caus-asp letter
'The child is having a letter written'.
*'The child is writing a letter with'.
c. Umwáana a-ra-andik-iish-a ikáramú.
child s/he-pres-write-caus-asp pen
'The child is writing with a pen'
'The child is using a pen to write'.
Umugabo y-a-háa-ye abáana ibitabo.
Umugabo y-a-háa-ye abáana.
*Umugabo y-a-háa-ye ibitabo.
Umugoré y-a-h-éesh-ej-e abáana ibitabo umugabo.
Umugoré y-a-h-éesh-ej-e abáana ibitabo.
Umugoré y-a-h-éesh-ej-e abáana.
*Umugoré y-a-h-éesh-ej-e ibitabo.
*Umugoré y-a-h-éesh-ej-e umugabo.
Abáana ba-ra-híing-a umurimá w'úmugoré.
Umugabo a-ra-híing-iish-a abáana umuríma w'úmugoré.
Umugabo a-ra-híing-iish-ir-iz-a umugoré umurimá abáana.
Umugabo a-ra-híing-iish-ir-iz-a umugoré umurimá.
Umugabo a-ra-híing-iish-ir-iz-a umugoré.
*Umugabo a-ra-híing-iish-ir-iz-a umurimá.
*Umugabo a-ra-híing-iish-ir-iz-a abáana.
Abapóolisí ba-a-fat-an-ye abahuúngu urumogi.
Komíseeri y-a-fat-an-iish-ij-e abahuúngu urumogi abapóolisí.
Komíseeri y-a-fat-an-iish-ij-e abahuúngu urumogi.
Komíseeri y-a-fat-an-iish-ij-e abahuúngu.
*Komíseeri y-a-fat-an-iish-ij-e urumogi.
*Komíseeri y-a-fat-an-iish-ij-e abapóolisí.
With multiple causativization objects have to be expressed. Unlike one causativization
where may not be specified (both causee and causer).
Umugabo y-ar-íic-iish-ij-e.
'The man made (somebody) kill (somebody)'.
*Umugabo y-ar-íic-iish-iish-ij-e.
'The man made somebody kill with something'.
'The man caused somebody to have somebody killed'.
Y-ará-wu-mú-mw-iic-iish-iish-ij-e.
'He made him/her kill him/her with it'.
The 'have NP Ved' causatives have the causee not as direct objects but as oblique
objects with the preposition ku which is similar to causativization in Romance
languages in which the former direct object becomes an adjunct or oblique object.
Like in Chichewa, these adjunct agents are marked by the locative/goal suffix
ku.
1a. Umudózi a-ra-téer-a ishaáti ibiremo.
b. Umuhuúngu a-ra-téer-eesh-a ishaáti ibiremo ku mudózi.
2a. Baánk i-ra-vúunj-a amafaraanga.
b. Umucúruuzi a-ra-vúunj-iish-a amafaraanga kurí baánki.
3a. Kagabo y-a-guz-e imódoká
b. Gateté y-a-gur-iish-ij-e imódoká kurí Kagabo.
4a. Umuváandimwé a-ra-reer-a abáana.
b. Umugabo a-ra-reer-eesh-a abáana ku muváandimwé.
Objects of causative constructions behave like inherent objects. They acquire
all the properties that other objects NPs have such as the ability to undergo
passivization, relativization, pronoun-incorporation, topicalization, etc. As
far as word order is concerned, usually there is free word order in the verb
post position. But there is tendency for the object of the causative suffix
(the former subject of the embedded clause) to come next to the verb complex.
The exception is the instrumental which has tendency to come last.
As far as incorporated pronouns are concerned, causatives always occupy the
accusative position farther away from the verb stem that is before datives and
benefactives. It is again the causee or the object of the causative morpheme
which comes closer to the verb whereas that of the base verb occupies the leftmost
position.
The causative is different from other extensions, however, namely the applicative
and the comitative/reciprocal because it is the only one which allows the subject
to become an object.
This is because it has a different underlying structure. It is indeed a case
of clause-union phenomenon. When the union occurs, the former subject of the
embedded clause (causee) behaves exactly like that of other languages in which
it is found. A demotion process. It acquires a lower hierarchy as either direct
object, oblique object or deletion.
1.3.-iish- doubling
The causative morpheme sometimes appears more than once in the same verb stem.
It is common with lexicalized causative verbs when they are causativized. Monosyllabic
verbs also have it doubled when causativized but this is restricted to some
speakers only and finally -iish- appears twice if the verb is twice causativided.
1.3.1. Monosyllabic verbs
It is possible to have monosyllabic verbs with causative doubling when they
are causativized only ones. Ofcourse this creates ambiguity since the verb can
be causatived twice.
gusya 'to grind' > gusheesha 'to cause to grind' >gusheesheesha 'to cause
(to cause) to grind'
kunywá 'to drink' >kunywéesha 'to cause to drink'>kunywéesheesha
'to cause (to cause) to drink'
guhá 'to give' >guhéesha 'to cause to give' >guhéesheesha
'to cause (to cause) to give'
Double causatization of monosyllabic verbs is due to metrrcal reasons. Kinyawanda
foot is usually bipedal, consisting of two syllables. Monosyllabic verbs are
thus an exception in Kinyarwanda. The causativized stem is thus taken as stem.
1.3.2. Lexicalized causative verbs
Lexicalized causative verbs although they are metonymically related to their
basic verbs have lost this connection for some speakers. It is the reason why
they are easily causativized.
kuneesha' to defeat' <kunnya 'to defecate'
kuríisha 'to graze' >kuryá 'to eat'
kwíigiisha 'to teach' < kwíiga 'to learn'
kugeendeesha 'to ride' >kugeenda 'to walk'
Umwáana a-ra-geend-eesh-a igaáre.
child s/he-pres-walk-caus-asp bicycle
'The child is riding a bicycle'.
Umugabo a-ra-geend-eesh-eesh-a umwáana igaáre.
man he-pres-walk-caus-caus-asp child bicycle
'The man is making the child ride a bicycle'.
Gateté y-a-né-esh-ej-e Kagabo n'ámayerí.
Gatete he-pst-defecate-caus-appl+caus-asp Kagabo with tricks
'Gatete defeated Kagabo with tricks'
Gateté y-a-né-esh-eesh-ej-e Kagabo amayerí.
Gatete he-pst-defecate-caus-caus-appl+caus-asp Kagabo tricks
'Gatete defeated Kagabo using tricks'.
Inká zi-zaa-ri-ish-a mu kibaya.
cows they-fut-eat-caus-asp in meadow
'The cows will graze in the meadow'.
Umuhuúngu a-zaa-ri-ish-iish-a inká mu kibaya.
boy he-fut-eat-caus-caus-asp cows in meadow
'The boy will make cows graze in the meadow'.
4. Multiple causativization
Multiple causativization occurs only with the -iish- suffix. This is possible
because as we noticed earlier this morpheme can have many referents. When two
or more of these different roles appear within the same verb phrase the same
number of suffixes appear on the verb to mark them.
Umwáana a-ra-andik-a ibáruwá n'íikáramú.
child s/he-pres-write-asp letter with pen
'The child is writing a letter with a pen'.
Umwáana a-ra-andik-iish-a ibáruwá ikáramú.
Umugabo a-ra-andik-iish-iish-a umwáana ibáruwá ikáramú.
Consider the following English sentence.
We assume that it should be easily translated into Kinyarwanda because all languages
are capable of expressing any concept. Although paraphrases are possible, they
would make the sentence much longer and the translation would not be accurate.
The teacher made the parents have the children read books with glasses.
Umwáalimú y-a-sóm-eesh-eesh-eesh-ej-e ababyéeyi
abáana ibitabo indórerwamó.
Abáana ba-á-som-ye ibitabo n'íindórerwamó.
Abáana ba-á-som-eesh-ej-e ibitabo indórerwamó.
Ababyéeyi ba-á-som-eesh-eesh-ej-e abáana ibitabo indórerwamó.
Umwáalimú y-a-sóm-eesh-eesh-eesh-ej-e ababyéeyi
abáana ibitabo indórerwamó.
This type of complex causative construction is rare in every day speech due
to both complex sentence generation and processing. It is easy in writing because
of added cognitive devices namely vision and extended time to process. In speaking
the problem is caused by the restricted number of grammatical relations and
cases. Since there is no strict word order in the object position and since
the same case marker can refer to different semantic roles, speakers have to
use simple sentences.
The behavior of causative constructions clearly shows the rigidity of the syntactic
structure. Kinyarwanda has only three grammatical relations, namely subjects,
inherent direct objects (accusative and datives) and structural direct objects
which are marked on the verb for case and adjuncts. Because of the finite nature
of syntax in both forms and structures, like other structures, the same causative
construction can refer to different semantic interpretations. Extralinguistic
factors are needed for proper decoding. This suggests therefore that indeed
syntactic structures, semantic structures and conceptual structures are independent
of each other. This also supports, in a way, Chomsky's concept of the autonomy
of syntax and Saussure's idea of the arbitrariness of the linguistic sign. These
observations also suggest that to understand cognitive structures, more emphasis
should be put on semantics and pragmatics than syntax. The so-called argument
structure or theta-role doesn't seem to play any significant role at all in
the syntax of Kinyarwanda because as examples have shown grammatical relations
and cases in Kinyarwanda happen to be formal features which many times are used
indiscriminately to refer to unrelated semantic roles. The causative morpheme
can be used to show an agent, an instrument, a patient, a beneficiary, etc.
Clearly, to get the intended meaning of the causative construction, like in
other cases of different types of ambiguities, extralinguistic factors are needed.
Note:
(1). There are a handful of verbs which allow the instrumental meaning but
all of them end with the liquid r:
kugaruza (ku-garur-y-a)abaantu umuheto. 'to bring back people with a bow'
guháruza (ku-harur-y-a) umuhaánda isúka. 'to dig the road
with a hoe'
(1). Some causative verbs have null objects:
arahíingiisha/arúubakiisha/
(2). Either object or both may be expressed:
arúubakiisha (inzu) (abakózi)
(3). Stative or intransitive verbs have only one object.
gucíikiisha; gukeneesha; gutéteesha; gusariisha; gushoonjeesha
'to cause to be hungry';
gukóonjeesha 'to cause to get cold';
(4)
Some transitive verbs sound better without the agent being expressed
gufuungiisha; kwíirukaniisha; kudódeesha; gucúriisha; kuvúunjiisha;
-after the reversive suffix -uur- , -y- also receives an instrumental meaning:
gufuunga/gufuunguura uruugi/gufuunguuza uruugi icyúuma;
kwaandika/kwaandukuura/kwaandukuuza ibibázo ikáramú;
gufásha 'to help': gufáta 'to hold'
kurusha 'to surpass': kuruta 'to be older'
gusoma 'to read', 'to kiss' doesn't get all -y- suffixation when it means 'to
sip' or 'to puff', however, it does gusomya 'to offer a sip or a puff'
References
Alsina, Alex. 1992. On the argument structure of causatives. Linguistic Inquiry
23, 4: 517-555.
Kimenyi, Alexandre. 1979. A semiotic account of causative constructions. Linguistics.
| 
