You are on page 1of 17

Language Sciences, Vdume 13, Number 1, pp. 21-37, 1991 0388-ooo1191 $3.00+.

00
Printed in Great Britain Pergsmlon Press plc

A Comparative Study of Swahili ni and Arabic


kibza as Copulative Elements

Hassan A. Max-shad Sal& M. Sufeiman

ilmm Al-Qura University Yarmouk University

K&xi and ni are here treated as linking verbs conjoining subjects to their predicative com-
plements. Their distributional patterns in simple statements, questions and negative sentences are
discussed, and in the process, their general properties are compared. Two observations are
established. Firstly, the two verbal elements qualify as true copulas; this inference is particularly
significant with regard to /c&a as there is a popular opinion contrary to the finding. Secondly,
verbless construction in the two languages are the result of a deletion rule. In the case of Arabic,
the deletion rule is preceded by another rule - topicalization.

INTRODUCTION

The main objective of this paper is to present a comparative descripeion of the


Swahili rai me-presents and the Arabic kiina (be-past) as major copulative elements in
the two languages. The approach that wifI serve as the basis for developing our
arguments and mounter-arguments will be data based. Our concern here is descriptive
rather than theoretical. But where a theoretical analysis of a phenomenon has been
attempted, Chomsky’s earlier transformational model of grammar is assumed.
Readers might wonder why we have chosen Arabic and Swahili. Is it because the
two languages share peculiar copular characteristics? The answer to this question is
no. Our interest was triggered by three factors. Firstly, the authors - one a native
speaker of Swahili and the other of Arabic - were keen about investigating (what they
consider) an important but not fulfy explored aspect of their languages. [This paper
constitutes a brief exploratory study of a major copuh in each language; a full report
of the entire copular system in the two languages will be presented in the future.1
Secondiy, we believe that a comparative study of any aspect of Arabic (an exceflent
22 Language Sciences, Volume 13, Number 1 (1991)

representative of the Semitic family of languages) and Swahili (a very good represen-
tative of the Bantu family) - irrespective of whether the same properties are found
in some other languages or not - is bound to enrich the presently available collection
of illustrative data across languages.
Thirdly, Arabic and Swahili are not just any two languages picked at random. These
languages are known to have been in contact for centuries; thanks to Islam and intra-
oceanic trade. And one of the by-products of this long term contact is the large number
of Arabic loanwords in Swahili.

COPULATIVE VERBS

Preliminaries

We will begin by explaining the term copula. According to David Crystal (1985:76),
a copulative element is a linking verb that has “little independent meaning,” and that
its “main function is to relate other elements of clause, especially subject and com-
plement.” Let us apply this definition to the following sentences.

(la) Ali likes green tea.


(lb) Ali is a doctor.

If we compare the two sentences, we can see that the predicator in (lb), unlike that
in (la), is semantically empty; its main purpose is to latch the subject and its com-
plement together. Its secondary function is to carry an appropriate tense marker as the
clause does not have any tense carrier. The copular element used in our illustration
is the verb “be.” English has other linking verbs (e.g. feel, look, appear, become,
etc.) which, although they behave similarly, are not as completely devoid of meaning
as the verb “be.” It can therefore be stated that “be,” when preceding non-verbal
predicative groups, will constitute an excellent example of a copula. In (lb), the non-
verbal predicative group used is a noun phrase; others which can follow “be” in
similar types of constructions include adjectives (he is sick), prepositional phrases (he
is in the library), adverbials (he is outside).
To sum up, a copula is a verbal element that will link a subject and its complement
in a clause; and since complements do not inflect for tenses, a copula also serves as
a tense carrier (or indicator).

Swahili Examples

Swahili has a set of true copulas (as defined above), and what can be called quasi-
copulas. True copulas include ni (be-present) and its substitutes +a (be), and ndi- (be-
present), whose uses are illustrated in (2).
cm Ali ni daktari.
Ali is a doctor.
(Zb) (i) Ali alikuwa tajiri.
sp-pt-inf-be rich
Ah was rich>
(ii) Ali atakuwa mrefu.
sp-ft-&f-be talf
Ah will be tab.
Hawa ndiwo wabmu.
these ndi-sp-o teachers
These are (emphatic) the teachers.
[sp = subject agreement prefix; pt = past tense prefix; ft = future tense
prefix; inf = infinitive marker.]

As the focus ofthe paper is on ni, it will be w~~~w~~~~at this point to discuss briefly
the other copufas ~~~~strat~ above. We will start with EC%-.
In the standard variety of Swahili, ndi- is a bound morpheme appearing onfy in the
strtEture __i_ -SF-0’ where -cr is what iraditjona~ists call o-of-reference (e.g. Ashton
1984:99-1@3).2 As can be deduced from (2c), ndi- is an emphatic cop& referring
only to the present tense. It is emphatic because it impiies “this is what I (or you, we,
or they) say to be the case.” Thus the literal translation of (2~) would be something
like this: “These are what we (or some other people) say to be the teachers.” It is
precisely because of this underlying interpretation that the Swahili approximate
equivalent of the Engfish 4’e.ris n&o, that is, “it is what I say to be the case,” But
it is more complex than _vesas, depending on its immediate context, M&V can also
mean no! This possibi~j~ is due to the imptied meaning which it carries. For instance,
if nd&~ is uttered as a response to a negative question then it would mean tao; (3)
provides an ap~r~riate context for this second inte~re~t~on of ndz@.

(3) Q: Hutaki kwenda sokani?


Don’t you want to go to the market?
R: Ndiyo (sitaki) .
What you have assumed is the case (I don’t want to).
No (I don’t want to).

Unlike rzi and nbi-, -wa requires inflectional pregxes to mark subject a~~~rne~t as
well as tenses; it does not, however, accept present tense mo~h~mes* And because of
the presence of these prefixes, it is treated like any other rnonosy~~~bi~verb in that the
24 Language Sciences,Volume 13, Number 1 (1991)

infinitive -ku- is prefixed to the stem in most constructions in the order indicated in
(2b).
Then there are quasi-copulas. They are quasi because unlike true copulas they do
express - to a certain extent - a semantic predicate. The list of such items include -ko
(and its sisters: -PO, -mo) and -na. (4) provides illustrative sentences using -/co and -na.

(4a) Watoto wako nyumbani.


children sp-ko house-locative suffix
The children are at home.

(4b) Nyumba zina madirisha.


houses sp-na windows
Houses have windows.

As can be seen above, the common feature shared by quasi-copulas is that they have
to have a subject agreement prefix. It should be noted that -na here is not the pre-
position na which is never prefixed. The preposition is a free morpheme except when
it appears in its contracted form (as in ruzmiderived from na mirni),or when it is
written with suffixes (as in nacho derived from n&i-o).
In this paper, as it has already been pointed out, we will focus on the copula ni.

Arabic Examples

In this section we will assume without providing supportive arguments that Arabic
does have copulas; evidence to support this assumption will be given in Section 4.
The verbal element kiina (be-past) has been classified into three types: niiqisuh
(incomplete), ziiidah(redundant), and ti&.mah (complete)s - for instance, by Hassan
(1963:417), Abdul-Hamid (1964288). We consider kiim tr?mmah as non-copulative;
the incomplete and redundant k&a, together with their present/future forms, in our
opinion, qualify as true copulas, and the examples we have chosen in (5) bear this out.

(5a) niqisah: (i) kPna al-waladu mustayiddan


be-past def. boy ready
The boy was ready.

(ii) yakiinu al-waladu mustaTiddan


be-present def.boy ready
The boy is ready.

(iii) sayaktinu al-waladu mustafiddan


be-future def. boy ready
The boy will be ready.
SwaMii ni and Arabic k&a as Copdative E%ments 25

(5b) zlidah: al-qitaru k&a qadiman


def. train be-past come-attributive
The train was coming.
[This example is given by Hassan (1963).1

The role of the verbal elements in the above examples is twofold: they provide a link
connecting each subject to its predicative eiement and they indicate the tense. Hence,
according to the definition given at the beginning, ~~~/ya~~~~/suyak~n~ are true
copuias.
To understand the role of l&u in (Sb), we need to consider k&inaz&M within the
larger framework of copular elements in Arabic, and the Arabic assertive sentences.
Unlike some other languages, Arabic does not have a ruie stating that every categorial
sentence must have a verb. But according to Shehadi (1969), this is not to say that
it is not possible (and in certain cases it is in fact required) that there be a word which
upon analysis would be seen as performing a copular role. In (5b), such a word is kiEna
(z~%dah); its presence in the sentence does strengthen the connection between the
subject and the predicative complement.
Like Swahili and many other Ianguages, Arabic has in addition a set of quasi-
copulas. Included in this set are yuwj& ~~w~/~~~~,inlta, a&&r, etc. We are clas-
sifying these elements as quasi-copulas because each has its own independent meaning,
and - as illustrated in (6) - their copulative function is only logical.

(6a) yuwjad maFana kutub


exist with us books
There are books with us. (We have books.)

(6b) al-ins&u huwa hayawanun n&q


def. man be animal with discourse
Man is an animal with discourse.

(6~) inna al-waIada Bakiyyun


emph. def. boy clever
The boy is clever.

(6d) asbaha al-taqsu jamiian


became def. weather nice
The weather became nice.

It should be noted that in constructions such as those in (6), yuwjad and ~MQ are
normally seen as redundant and cons~quendy are often deleted. Another point worth
noting is that inna and Iowa can co-occur in their contracted form to impart extra
26 Language Sciences, Volume 13, Number 1 (1991)

emphasis in an assertion: “a1-waladu innahu aakiyyun,” that is, The boy is indeed
clever.
Our discussion ahead will concentrate on Mna.

DISTRIBUTIONAI, PATTERNS

Here we will examine syntactic patterns of ni/-wa and k&a in simple statements,
questions, and negation. The purpose is to establish common grounds shared by these
elements in the two languages.

Simple Statements

As it can be concluded from (7) below and data given earlier, ni and its sub-
stitute -wa appear in structures approximately represented by the frame NP _
where Lot stands for locative complement (z PP), and that NP is ex-
tended to include noun substitutes.

(74 Kisahani ni kichafu.


The plate is dirty.

Vb) MLo huu ni wangu.


pillow this is mine
This pillow is mine.

(7c) Sisi ni kumi.


We are ten.

(7d) Hapa ni jikoni.


this (place) is kitchen-Lot
This is where the kitchen is. (i.e. This is the kitchen.)

The frame in which kiina appears, based on (5) and (8) given below, seems to be
somewhat similar to that of ni; it is _ NP [NP 1

@a) k&a zaydun taliban


be-past Zayd student
Zayd was a student.

(8b) kiina al-taqsu jamilan


def. weather nice
The weather was nice.
Swahili ni end Arabic Kim as Copulative Elements 27

(8~) k&a zaydun fi al-bayti


Zayd at def. house
Zayd was at home.

Simpte Questions

Questions in Swahili are mainly expressed by a rising intonation, and where needed,
a wh-word is inserted at an appropriate position in the string. Some illustrative
examples are shown in (9).

(Pa) Kile ni kitu gani? or Ni kitu gani kile?


that is thing what
What is that thing?

(Pb) Watoto wale ni wa nani? or Ni wa nani watoto wale?


children those are of who
Whose children are those?

(9c) (Je) Ali alikuwa wapi? or Alikuwa wapi Ah?


Q-word Ali was where
Where was Ali?

As one would expect, ni (and its substitute) maintains its basic frame. However, for
stylistic variation or in order to create emphasis, ni together with the constituents
following the copula may be fronted as indicated in the alternatives shown above.
In Arabic, too, questions are indicated by a rising intonation, and when wh-words
are needed they are inserted at the initial position of the string. It is also interesting
to note that the Swahili Je (9~) can be used with any kind of question, but the Arabic
interrogative words hat and - ?a are restricted to yes/no type of questions [{lOa)
alternative].

(loaf k&a zaydun fi al-bayti? or k&a zaydun fi al-bayti?


Was Zayd at home?

(lob) ayna k%na zaydun?


Where was Zayd?

(10~) man kana ti al-bayti?


Who was at home?

We would like to note at this point that the comparison made so far may not
distinguish these two languages from many others, but we hope that each set of data
28 Language Sciences, Volume 13, Number 1 (1991)

presented will mean - particularly to those not so familiar with Arabic and/or Swahili
- one more extent of similarity obtaining in the two languages.

Negation

The negative of ni is si. Thus any sentence containing ni can be negated by replacing
ni with si (1 la-b). However, negating -wu is a bit complicated. It is done by affixing
the general negative prefix ha- to the verb structure. When this happens, some of the
tense markers are changed to reflect negation: past tense -li- becomes -ku-, and the
present perfect -me- changes its form to be -ja-. (The future tense marker da- does
not undergo any change in the standard dialect.) In imperative and subjunctive
structures, the -si- prefix is used as is illustrated in (11 e-f). For more details see
Ashton (1984:70-7, 279-83).

(1 la) Ali ni daktari. + Ali si daktari.


Ali is a doctor. Ali is not a doctor.

(llb) Ali alikuwa daktari. + Ali hakuwa daktari.


Ali was a doctor. Ali was not a doctor.

(1 Ic) Ali amekuwa mnene. + Ali hajawa mnene.


Ali has become fat. Ali has not become fat.

(lld) Ali atakuwa mrefu. * Ali hatakuwa mrefu.


Ali will be tall. Ali will not be tall.

(1 le) Usiwe mpumbavu.


sp-si-be stupid
Don’t be stupid.

(1 If) Wasiwe wavivu.


sp-si-be lazy
They shouldn’t be lazy.

The Arabic equivalent of the Swahili si is laysa (12a) which, like si, expresses the
state of being/existence negatively. K&a is negated by the particles mii (12b) or lam
(12c), and be-future requires the negative marker lun (12d). What is interesting to note
in the data is the somewhat similar corresponding behaviour between nil-wu and
yakihdkiina when the sentences in which the copulas appear are negated.

(12a) laysa al-waladu mustaFidan


The boy is not ready. [cf. @a)(ii).l
Swahili ni and Arabk k&w as Co~~ve lZkme~~ts 29

Wb) ma k&ra zaydun fi al-bayt


neg be-past Zayd at def.house
Zayd was not at home.

(12c) lam yakun zaydun fi al-bayt


neg be Zayd at def.house
Zayd was not at home.

(124 Ian yakiina zadun fi al-bayt


neg be Zayd at def. house
Zayd will not be at home.

What has been presented so far shows that ni and kiina share a number of linguistic
properties. These include the structures in which the two copulas appear in simple
statements, questions and - to a certain extent - negation. There are, however, two
basic differences. First, ni is be-present while i&m is be-past. Second, unlike ni, tina
(or rather its basic form) can be inflected to represent be-present (yaktinu) or be-future
(sayakiin); in Swahili, when a sentence has past or future relevance, -wa is used
instead.4

FURTHER EXAMINATION OF THE COPULAS

The case of ni

As seen in the Copulative Verbs Section, the frame in which ni often appears is
NP- NP
Adj
1 Lot >
However, ni-constructions without the subject NP such as those in (13) are quite
common.

(13a) Ni mtu.
is person
It is a person.

(13b) Ni nani?
is who?
Who is it?

Ashton (1984:92) translates Ni mtu as It is a mm, and claims that ni is only to be


used in such constructions when the (missing) subject refers to the third person.
Granted that her interpretation of the construction will be valid for a number of
30 Language Sclenws, Volume 13, Number 1 (1991)

situations, her claim is not acceptable as it excludes the use of the pattern when the
subject refers to the first or second person as exemplified in (14).

(14a) Speaker A: Wewe ni nani?


you are who?
Who are you?
Speaker B: Ni Ali.
I am Ali.
(14b) Speaker A: Mimi ni nani?
I am who?
Who am I?

Speaker B: Ni polisi.
You are a policeman.

What we would like to suggest is that Ni NP is an elliptical structure, and out of


context - such as the one given in (14) - it is ambiguous; for, the missing subject
can be referring to first person (14a) or second person (14b). The context shows
clearly that the utterance Ni Ali is derived from Mimi ni Ali, and Ni poiisi from Wewe
ni polisi.
Another interesting phenomenon observable in Swahili is the availability of the
following type of structure.

(15a) Ali mrefu.


Ali tall
(Ali is tall.)
(15b) Sasa saa mbfli.
now o’clock two
(It is now two o’clock.)
(1%~) Yule babako?
that your father?
(Is that your father?)

There are two facts that should be noted regarding the above sentences. Firstly,
though verbless, they are understood to refer to present tense situations. Secondly,
they can all be re-written to include ni as the overt link connecting the subject to its
predicative element. Thus, Ali ni mrefi is an exact equivalent of (15a). Therefore, we
would like to propose that sentences containing ni are basic and those in (15) are their
derived structures. That is, there is a deletion rule that optionally deletes the copula
SwahUi ni and Arabic k&a as ~o~lat~ve Elements 31

ni in such constructions. Is this a reasonable proposal? We think that it is, because


a similar deletion rule is operative in a number of other languages such as Mandarin
Chinese (Hashimoto 1969:83-5), Black American English (Labov 1969; Wolfram
and Fasold 1974:158-60) and, as we shall soon observe, Arabic.
The following data seem to agree with what we have just presented. The following
(16a) has the overt link provided by ni, and in the case of (16b), the optional deletion
rule appears to have applied leaving the structure verbless.

(16a) Mimi ni ndani.


I am inside (e.g. a house).

(16b) Mimi ndani.


I inside
(I am inside.)

However, if (16a) is negated by replacing ni with si (refer to the section on Negation),


the resulting string (17a) is not acceptable. Its negative rendition has to take the form
shown in (17b).

(17a) *Mimi si ndani.


I am not inside.

(17b) Mimi siko ndani.


I am not inside.

What this implies is that the ni in (16a) cannot be the familiar copula we are discussing,
and (17b) is suggestive of its underlying representation. The item siko is the negative
form of niko, where ni- is the first person singular subject agreement prefix, and the
base -ko is one of those quasi-copulas briefly mentioned in the section on Swahili
examples. The prefix si- is the negative form of the first person singular subject prefix,
and it should not be confused with the negative counterpart of the copuia ni rep-
resented in (17a). This means then that the ni in f 16a) is a reduced structure whose
full form is niko. Perhaps this is an appropriate place to state again that elements such
as niko (or any of its sisters) do not strictly speaking qualify as true copulas. The
reason for this is that they carry inherent meaning which somewhat restricts their use
to sentences containing locative complements.5

The Case of kiina/yakiinu

In this section we shall show that nominal sentences in Arabic are derived structures
whose basic forms do contain copulas, and that, like Swahili, Arabic has a copula deletion
32 Language Sciences, Volume 13, Number 1 (1991)

rule, but this rule is a little bit more complex compared to that available in Swahili.
We will begin by assuming that Arabic is a VSO language; argument supporting
this assumption can be found in Suleiman (1985, 1986). However, in Arabic nominal
sentences, the more widely observable word order is subject-predicate. If our assumption
that Arabic is a VSO language is to remain, then the subject-predicate order can be
explained by positing a subject fronting rule. This topicalization rule is quite common
in a number of languages (Li 1976; Anderson and Chung 1977). Under these assumptions
we would like to propose that the basic structure of (18a) is (lgb), and that (18~) is
derived from (18b) by the suggested topicalization rule. Finally, the surface structure
(18a) results from the application of the copula deletion rule on the intermediate
structure (18~).

(18a) zaydun qawiyyun


Zayd strong
(Zayd is strong.)

(18b) yakiinu zaydun qawiyyun


be Zayd strong
(Zayd is strong.)

(1%) zaydun yaktinu qawiyyun


Zayd is strong.

The proposal that (18a) has underlyingly the copula yakiinu ties in with the analysis
of the Swahili data presented in (15), as, although verbless, (18a) refers to the present
tense situation. In the case of Arabic nominal sentences, however, copula deletion
occurs after the topicalization rule has rendered yak&u (now between the subject and
its complement) redundant.
Another type of nominal sentence which can provide evidence pointing to the
existence of underlying copulas is that whose predicate is locative. The relationship
that connects the subject (u~-w~~~ to the predicate (fi uZ-~~~~~~) in (19a) is that of
existence or being. Such a relationship is only possible if the connector is the verb “to
be.” It is therefore logical that the surface structure that lacks an overt copula is
derived from a sentence with an underlying copular link (19b).

(19a) al-waladu fi al-madrasa


def. boy at def. school
The boy is at school.

(19b) al-waladu yuwjad/yakt’inu fi al-madrasa


def. boy exist/be at def. school
The boy is at school. (cf. Shehadi 1969.)
Swahili ni and Arabic k&a as Copulative Elements 33

Our final evidence in favour of an underlying copulative verb will be provided by


the verbal type of sentences in Arabic. Let us consider the passive construction given
in (20a). The active counterpart of this sentence is (20b) which can alternatively be
represented as (20~).

(204 kusira al-zujaju


broken def. glass
The glass was broken.

GObI kasara al-waladu al-zujaja


broke def.boy def. glass
The boy broke the glass.

WC) kana al-waladu kasara al-zujaja


be-past def. boy broke def. glass
The boy had broken the glass.

K&a in (20~) functions as an emphatic tense marker making absolutely clear to the
hearer that the event took place “at some point in the past.” In this context, kiina is
deleted to allow for the lexical verb kusaru to take the initial position of the sentence
(20b). Thus, although Arabic sentences (both nominal and verbal) o&w appear with-
out surface copular verbs, the data presented strongly suggest that their underlying
structures have them. As discussed, they are required as tense/aspect markers, and
they provide a way of making explicit the relationship between a subject and its
predicate.
Before we conclude this section, we would like to refer to the assertion made in
the first paragraph of the section on kiirdyukiinu that the kiindyukiinu deletion
rule is more complicated compared with the ni deletion rule. What we are going to do
here is merely point out this complexity leaving its rigorous analysis for a future
presentation.
According to Abdul-Hamid (1964:293-4), /C&U, together with its subject, may be
deleted if it occurs after the conditional particle lin (if). The phenomenon is shown
in (21) - (21a) being a derived structure whose basic form is presented in (21b).

(214 qad qila ma qila 7in sidqan wa ?in kabiban


emph. said what said if true and if false
What was said was said, whether true or false.

@lb) qad qila ma qila 7in kana al-qawlu sidqan wa 7in kana al-qawlu kabiban
emph. said what said if be-past utterance true and if be-past utterance false
What was said was said. whether the utterance was true or false.
UC 13Cl-c
34 Language Sciences, Volume 13, Number 1 (1991)

But when tina is preceded by the particle law, then it is kiina together with its
predicative complement or at least the locative part of the complement. The example
given below (cited in Hassan 1963:421) illustrates the process vividly. The following
(22a) is considered to be the derived sentence after the deletion rule has acted upon
(22bf.

(22a) aFfi al-muhtaj wa law dirham


give def. poor and if dirham
Give the needy even if you have only one dirham.

(22b) aS’tial-muhtij wa law k&a dirhamun 3 &y&k


give def. poor and if be-past dirham in your pocket
Give the needy even if you only have one dirham in your pocket.

The deletion of phrases like Ji juybika in (22b) might appear that anything can be in
underlying structures. But this is not so because the elements which delete are those
that are present by implication - very much similar to the deletion of the subject in
imperative structures.

CONCLUSION

This paper has shown that ni (and its substitute -wa) and k&a (niiqisah or ziidah)
do not have any semantic role in a construction. Their function is to link a complement
to the subject, and to act as tense/aspect markers. This suggests that the two verbal
elements are on a par with the English copula “be.”
The paper has also argued that the copula-free structures observable in the two
languages are a surface manifestation. However, these structures do have underlying
copulas. The surface sentences result after a copula deletion rule has applied; in the
case of Arabic, topicalization occurs before copula deletion.

NOTES

1. Swahili verbs have complex structures made up of prefixes, stem and suffixes.
Our interest here is in the subject agreement prefix (sp) in its capacity as a
provider of agreement between verbs and their subjects in terms of class noun,
or person and number. However, the use of sp is not restricted to the verb
structure; it appears in a number of other enviro~ents as exemplified in the
following sample situations.
Swahili ni and Arabic &aa as copulative elements 35

Example Gloss Structure Derived from Referring to noun


class number

ndicho it is the one ndi-sp-o ndi-ki-o 7


wale those sp-le wa-le 2, 3rd person
yote all, whole sp-ote ya-ote 4 and 6, or 9
langu my sp-angu li-angu 5

The complete set of sp as well as other affixes and the order in which they
appear in the verb structure or elsewhere can be found in any elementary Swahili
grammar book.
2. Barrett-Keach (1987:263-98) has provided a new analysis of the so-called o-
of-reference. She says that it is an epenthetic 101that is suffixed to any final
pronominal clitic.
3. KBna naqisah and zaidah require both subject and predicative element; their role
in a string is to link the two constituents. (Kana zaidah may be omitted in a
sentence without changing meaning.) Kana t%mmah (complete), on the other
hand, acts as an intransitive verb, and consequently it does not require a predica-
tive complement. The following example will illustrate this fact.

Kana al-amr
be-past def.matter
The matter was (i.e. The matter came to pass).
[Source: Hassan (1963:394).1

4. What is interesting is that wa- and ni can co-occur to produce an added emphasis
in tenses other than present. In such constructions the syntactic role of -wu is
reduced to that of a tense carrier, and, either ni or wu- can be negated (c-d) but
not both (e).

(4 Ali alikuwa ni mwalimu.


Ali sp-past-infinitive-wa be-present teacher
Ali was a teacher.

(b) Ali atakuwa ni mwalimu.


sp-future-infinitive-wa be-present
Ali will be a teacher.
Ah hakuwa ni mwahmu.
neg-past-wa be-present
Ali was not a teacher.

(4 Ali alikuwa si mwalimu.


sp-past-infinitive-wa neg be present
Ali was nor a teacher.

(e) *Ali hakuwa si mwalimu


neg-past-wa neg be-present
Ali was not a teacher.

5. There are, in addition to locative complements, a few other predicatives which


express condition or state that are compatible with -ko and its sisters. The list of
such predicatives include tayari (ready), wazi (state of being open), taabuni
(suffering/state of being in great pain). (We would like to point out that the final
-ni in the word tuabuni is not the locative -ni as is suggested by F. Johnson’s A
Standard Swahili En&h Dictionary.) The following sentences illustrate the use
of -ku when the predicatives are those cited.

(a) Mimi niko tayari.


I am ready.

to) Mlango uko wazi,


The door is open.

(c) Hali yake iko taabani.


condition her is in great pain
She is in great pain.

Abdu~-H~id, M. M.
1964 SIaarA Zbn %qit, Beirut, Lebanon: Dar Al-%hrrn AI-Hadithah.
Anderson, S. and Sandra Chung
1977 ‘“On Grammatical Order in Verb-initial Languages,” in Syntax: and
Sonantics, Vol. 8, pp. l-25, P. Cole and J. Sadock (eds.), New York:
Academic Press.
Ashton, E. 0.
1984 Swahili Grammar, London: Burnt Mill.
Swahili ni and Arabic k&a as Copulative Elements 37

Barrett-Keach, C. N.
1987 “Phonological Allomorphy in Swahili: on the Form of Inanimate Pro-
nominal Clitics,” Studies in Ajkicun Linguistics 18. 263-98.
Crystal, David.
1985 A Dictionary of Linguistics and Phonetics, New York: Basil Blackwell.
Hashimoto, Anne Yue
1969 “The Verb ‘to be’ in Modem Chinese,” in Z7re verb ‘be ‘and Its Synonyms,
Part 4, pp. 72-110, J. W. M. Verhaar (ed.), Dordrecht: Reidel.
Hassan, Abbas
1963 AZ-Nuhw Al-W@, Vol. 1, Cairo: Dar Al-May&if.
Labov , William
1969 “Contraction, Deletion, and Inherent Variability of the English Copula,”
Language 45. 715-62.
Li, Charles N. (ed.)
1976 Subject and Topic, New York: Academic Press.
Shehadi, Fadlou
1969 “Arabic and To Be,” in The Verb ‘Be ’ and Its Synonyms, Part 4, pp.
112-25.
Suleiman, S. M.
1985 Jordanian Arabic between Diglossia and Bilingualism: Linguistic Analysis,
Amsterdam: John Benjamins.
1986 “A Relationally-based Study of Subject-Object Properties in Standard
Arabic, ’ ’ Arab Journal of Language Studies 4. 133-50.
Wolfram, W. and R. W. Fasold
1974 ZJre Study of Social Dialects in American English, Englewood Cliffs, NJ:
Prentice-Hall.

You might also like