Agreement in coordinated noun phrases in Runyankore-Rukiga
by Allen Asiimwe
Agreement morphology is one of the most pronounced and studied aspects of Bantu grammars. It is a cross-referencing mechanism between different parts of a sentence or a phrase. Agreement occurs depending on the arguments being cross-referenced; for a verbal predicate the subject and object in a sentence trigger agreement on the verb. In the nominal domain all noun modifiers must bear agreement markers, corresponding to the class of the noun they modify and this is exemplified in (1) below:
(1)
Ebijuma byangye bibiri bihango birungi bikagwa
“My two nice big fruits fell”
| èbìjùmà |
| e | bi | juma |
| IV | 8 | fruit |
| N |
| byàngyè |
| bi | angye |
| 8 | mine |
| PNposs |
| bíkàgwà |
| bi | ka | gw | a |
| 8 | PAST | fall | IND |
| V |
The subject ebijuma ‘fruits’ in (1) belongs to noun class 8; and the possessive, numeral, adjectives as well as the verb bear a corresponding agreement prefix.
In some cases, as for instance shown in the above example, the shape of the prefixes is identical, which is sometimes called ‘illiterative’ (Katamba 2003). However, agreement markers do not necessarily need to be identical with the class markers of the subject noun as for instance shown in (2):
(2)
Omuti muhango muraingwa gwagwa
“A big tall tree has fallen”
| muraingwa |
| mu | raingwa |
| 3 | tall |
| ADJ |
| gwágwà |
| gu | a | gw | a |
| 3 | PRESPFV | fall | IND |
| V |
Whereas omuti ‘tree’ belongs to class 3, the agreement prefix on the verb is for class 20, a class for augmentatives, although it is rendered as a class 3 marker, since all parts of speech in a sentence must be in agreement with the head noun.
Turning to coordination, when two NPs, belonging to the same noun class, are coordinated the corresponding plural concord prefix is used to express agreement, e.g.:
3
Omwana na nyina nibazaana
“A child and the mother are playing”
| nìbàzàánà |
| ni | ba | zaan | a |
| PRES | 2 | play | IND |
| V |
The verb in the example above 'nibazaana' triggers class 2 agreement (plural marker for class 1), since the coordinated nouns belong to the same noun class (i.e. class1). Schadeberg (1992) and Krifka (1995) refer to this as morphological agreement i.e. where two nouns of the same class trigger the corresponding plural concord on the verb. Yet, nouns belonging to class 9 and 10 present a different situation; they may occur with either cl8 or 10, as illustrated below:
4a)
Entaama n'embuzi nibirwana
“The sheep and the goat are fighting”
| èntààmà |
| E | n | taama |
| IV | 9 | sheep |
| N |
| nìbìrwánà |
| ni | bi | rwan | a |
| PRES | 8 | fight | IND |
| V |
(b)
#Entaama n'embuzi nizirwana
“The sheep and the goat are fighting”
| èntààmà |
| e | n | taama |
| IV | 9 | sheep |
| N |
| nìzìrwánà |
| ni | zi | rwan | a |
| PRES | 10 | fight | IND |
| V |
(c)
Entaama neerwana n'embuzi
“The sheep is fighting with the goat”
| èntààmà |
| e | n | taama |
| IV | 9 | sheep |
| N |
(4b) is not acceptable because though the ente ‘cow’ and embuzi ‘goat’ belong to class 9, they are different animals, which seem to make it impossible to use the plural agreement marker. (4a) is acceptable but most natural is an expression as shown in (2c), where now instead of a 'na' as coordonating conjuction, it is used as preposition.
The use of class8 is however natural as an agreement marker under the coordination of inanimate nouns. Look at (5):
5
Emeeza n'entebe bihendekire
“The table and the chair are broken”
| bìhèndèkírè |
| bi | hendek | ire |
| 8 | break | STAT |
| V |
The nouns coordinated in (5) are both inanimate and the agreement marker for class 8 (-bi-) is appropriate. An interesting question is: what will happen when the coordinated nouns do not fall under the same noun class? Let us look at the following schema in (6), where the choice of verbal prefix is left open:
(6)
What prefix would a native speaker be inclined to add to the verb?
Note that omuhiigi ‘hunter’ belongs to class 1 while embwa ‘dog’ falls under class 9. Ashton (1994) observes that in case two nouns, belonging to two different noun classes, are conjoined (and especially if they are concrete nouns) the verb agrees with the last named noun. Following Ashton’s observation, in (6), the verb should agree with embwa ‘dog’, as shown in (7) below:
(7)
#Omuhiigi n'embwa yaagyenda
“The hunter and the dog have gone”
| òmùhììgì |
| o | mu | hiigi |
| IV | 1 | hunter |
| N |
| yàgyéndà |
| ya | a | gyend | a |
| 9AGR | PRES | go | PFV |
| V |
However, (5) leads to an unacceptable construction; if ya- is chosen to be the agreement marker for both omuhiigi and embwa, it would mean that omuhiigi also belongs to class 9 which is not the case. However, Ashton’s observation holds for constructions which involve nouns that are both inanimate, like in (8) below:
(8)
Emiyembe n'ebyokurya biri hanu
“Mangoes and food are here”
| èmìyèmbè |
| e | mi | yembe |
| IV | 4 | mango |
| N |
| èbyòkùryà |
| e | bi | okurya |
| IV | 8 | food |
| N |
The agreement marker, namely bi- of class 8, covers both nouns since also cl4 is inanimate, and accidentally agrees well with the last named noun ebyokurya ‘food’. Another suggestion, close to Ashton’s, comes from Schadeberg (1992). He believes that agreement will be with the corresponding plural class of the closest conjunct. In fact, most scholars have argued that either the plural prefix of class 8 will be used or the class of the last NP (Schadeberg 1992, Krifka 1995) to express agreement with a coordinated subject NP. However, this generalization only holds when the conjuncts do not involve humans, as shown below:
(9)
Efuka n'omuhoro biri nkahe?
“Where are the hoe and the panga”
| òmùhòrò |
| o | mu | horo |
| IV | 3 | panga |
| N |
(10)
Amahuri n’ebijuma ni birungi aha baana
“Eggs and fruits are good for children”
| èbìjùmà |
| e | bi | juma |
| IV | 8 | fruit |
| N |
(11)
#Omuhiigi n’embwa byagyenda
“The hunter and the dog have gone”
| òmùhììgì |
| o | mu | hiigi |
| IV | 1 | hunter |
| N |
| byàgyéndà |
| bi | a | gyend | a |
| 8 | PRES | go | IND |
| V |
Again (11) is unacceptable because one of the conjuncts is human and bi- is used when both conjuncts are inanimate. Katamba (2003) argues for Haya that in case of gender conflict in coordination of human with an animal, the best solution is to avoid coordination and opt for comitative construction, this works well in Runyankore-Rukiga as well, and thus construction (6) (which is equal to (11) above) could be presented as shown in (10):
(12)
However, the original meaning of the sentence may be distorted; (6) means that the hunter left and also the dog left: not necessarily meaning that they left together and for the same destination, while (12) clearly means that the hunter left together with the dog for the same destination; may be, they have gone hunting. Coming back to my original question, namely: Which prefix would a native speaker be inclined to add to the verb; the answer lies in the construction below:
(13)
Both conjuncts are coordinated but with a repetition of the verb. This would mean that both the hunter and the dog have left but not necessarily for the same destination which preserves the original meaning of (6).
References
Ashton, E. O. 1944. Swahili Grammar. Harlow: Longman
Clement, M. 1943. Outline Grammar of Bantu. Mimeographed: Johannesburg: University of the Witwatersrand (Revised 1982 as Communication No.2 from the Department of African Languages, Rhodes University at Grahamston)
Katamba, F. 2003. Bantu Nominal Morphology. In: The Bantu languages. Routlege. London and New York.
Krifka, M. 1995. Swahili. In Joachim Jacobs, Arnim von Stechow, Wolfgang
Sternefeld & Theo Vennemann, eds., Syntax: Ein internationales Handbuch
zeitgenoessischer Forschung, vol. 2, Berlin, New York: Walter de Gruyter, 1397- 1418.
Schadeberg, T. C. 1992. A Sketch of Swahili Morphology, 3rd ed., Köln: Köppe.