Identificative copulatives in Southern Ndebele : evidence for diachronic postulations in Zulu 1

Identificative copulatives in Southern Ndebele: evidence for diachronic postulations in Zulu Southern Ndebele is the language with the smallest number of speakers of all eleven official languages of South Africa. It is thus not surprising that there is no comprehensive description of the copulatives of this language. This article offers an exposition of the grammatical structure of the identificative copulatives of Southern Ndebele. The basic grammatical structure of the copulatives of this language is very similar to that of Zulu. However, the unfailing occurrence of a copulative verb stem “-si” in all negative stative copulatives is an outstanding characteristic of Southern Ndebele. The occurrence of this copulative verb stem in Southern Ndebele offers the strongest language external evidence (in support of language internal evidence) of the underlying negative stative copulative of Zulu as being “-si” as well. The identificative copulative particle that occurs as “ngi” with the pronouns of all persons and noun classes as complement in Southern Ndebele, also serves as language external evidence for the postulation of the identificative copulative particle of Zulu as “ngi”. Zulu grammarians postulate this identificative copulative particle of Zulu 1 I am grateful to Ms. Nomsebenzi Skosana for verifying the Southern Ndebele data and to Prof. Robert Botne of Bloomington University for valuable comments on an earlier draft of this article. Identificative copulatives in Southern Ndebele: ... diachronic postulations in Zulu as “nga”, “ngu” or even “ng”. It is argued in this article that the identificative copulative particle of Zulu is indeed “ngi”. Opsomming Identifiserende kopulatiewe in Suid-Ndebele: ondersteuning vir diachroniese postulerings in Zoeloe Suid-Ndebele is die taal met die kleinste getal sprekers van die elf amptelike tale van Suid-Afrika. Dit is daarom nie verrassend dat daar geen omvattende beskrywing van die kopulatiewe van hierdie taal bestaan nie. Hierdie artikel bied ’n uiteensetting van die grammatiese struktuur van die identifiserende kopulatiewe van Suid-Ndebele. Die basiese grammatiese struktuur van die kopulatiewe van hierdie taal stem baie ooreen met dié van Zoeloe. Die reëlmatige voorkoms van die kopulatiewe werkwoordstam “-si” in alle negatiewe statiewe kopulatiewe is egter ’n uitstaande kenmerk van Suid-Ndebele. Die voorkoms van hierdie kopulatiewe werkwoordstam in Suid-Ndebele is die sterkste taaleksterne bewys (in aansluiting by die taalinterne bewyse) vir die postulering van ’n onderliggende negatiewe statiewe kopulatiewe werkwoordstam “-si” vir Zoeloe. Die identifiserende kopulatiewe partikel wat in Suid-Ndebele as “ngi” realiseer, saam met die voornaamwoorde van eerste en tweede persoon en die naamwoordklasse as komplement, dien ook as taaleksterne bewys vir die postulering van die identifiserende kopulatiewe partikel van Zoeloe as “ngi”. Zoeloe-grammatici postuleer hierdie identifiserende kopulatiewe partikel van Zoeloe as “nga”, “ngu” en selfs “ng”. In hierdie artikel word aangevoer dat die identifiserende kopulatiewe partikel van Zoeloe inderdaad “ngi” is.

Southern Ndebele is the language with the smallest number of speakers of all eleven official languages of South Africa.It is thus not surprising that there is no comprehensive description of the copulatives of this language.This article offers an exposition of the grammatical structure of the identificative copulatives of Southern Ndebele.The basic grammatical structure of the copulatives of this language is very similar to that of Zulu.However, the unfailing occurrence of a copulative verb stem "-si" in all negative stative copulatives is an outstanding characteristic of Southern Ndebele.The occurrence of this copulative verb stem in Southern Ndebele offers the strongest language external evidence (in support of language internal evidence) of the underlying negative stative copulative of Zulu as being "-si" as well.The identificative copulative particle that occurs as "ngi" with the pronouns of all persons and noun classes as complement in Southern Ndebele, also serves as language external evidence for the postulation of the identificative copulative particle of Zulu as "ngi".Zulu grammarians postulate this identificative copulative particle of Zulu 1 I am grateful to Ms. Nomsebenzi Skosana for verifying the Southern Ndebele data and to Prof. Robert Botne of Bloomington University for valuable comments on an earlier draft of this article.

Introduction
Southern Ndebele is the least spoken and least studied of the eleven official languages of South Africa.It is thus not surprising that the copulatives (probably the most perplexing grammatical structures in the Bantu languages) of Southern Ndebele have not been documented exhaustively.In some respects the identificative copulatives of Southern Ndebele are similar to those of Zulu.The Southern Ndebele copulatives have, nevertheless, some unique characteristics.

Aims
The article aims to: • offer an overview of the copulatives of Southern Ndebele; • use the language data from Southern Ndebele to support the postulation of the identificative copulative particle 2 of Zulu as ngi; and • illustrate how the language data of Southern Ndebele (as external language evidence) support the postulation of a stative negative copulative verb stem -si for all stative copulatives of Zulu.(The unfailing occurrence of this copulative verb stem in all subcategories of the stative copulative in Southern Ndebele offers the strongest proof, on the comparative level, for the justification of the postulation of an underlying negative stative copulative verb stem *-si for Zulu where this copulative verb stem has a very limited occurrence.)

Classification of copulatives in Southern Ndebele
Following Welmers (1973:328), the copulatives of Southern Ndebele are subcategorised into four subtypes, namely the identificative, associative, descriptive and locative copulatives.These copulative types can occur in either the inchoative or stative form.The inchoative and stative categories are in binary opposition.In this article the focus will be on the identificative copulatives.
The scheme below represents a summary of the patterns of the identificative copulatives of Southern Ndebele in the indicative mood.The following abbreviations and symbols are used in the tables below: • # = a scientific word boundary; • ø = a zero morpheme occurs in this slot; • subj agr = subject agreement morpheme; 2 Although most Bantu grammarians use the term copula to denote the element (which occurs as ngi (or yi) in Zulu) that distinguishes the identificative copulatives from the other types of copulative, the term identificative copulative particle is used instead.The term copula is avoided, because it refers to the verb to be.Consider for example Longman Concise English Dictionary (1987:307), Oxford Advanced Learner's Dictionary (1995:257), The Collins paperback English Dictionary (1986:190) and Pei and Gaynor (1954:48).The identificative copulative particle is not a verb.It can actually co-occur with the copulative verb stems -ba, -bi and -si in Southern Ndebele, as is evident from the examples in this article.These language forms, furthermore, conform to the separability test for word identification developed by Van Wyk (1958 and1968) and applied to Northern Sotho by Lombard et al. (1985:12-13) and are thus regarded as particle words.• ident cop = identificative copulative particle; • N = noun; • ku-= the impersonal subject agreement morpheme

The binary opposition of stative versus inchoative with reference to copulatives
The umbrella term implication is used by some linguists to refer to the binary opposition stative versus inchoative.

The inchoative copulatives
Southern Ndebele is typically Bantu in that it uses the copulative verb stem -ba to mark inchoativeness in copulatives, as is evident from example 1a below: 1a. Bona baba batjhumayeli.

The stative copulatives
The affirmative form of the stative copulatives in Southern Ndebele is characterised by the absence of a verb stem.The phonological evidence supporting the postulation of an affirmative stative copulative verb stem *-li for Zulu is absent in Southern Ndebele.
Hence it is posited that the affirmative form of the stative copulative in Southern Ndebele contains no (underlying) copulative verb stem.
In contrast to the affirmative stative copulative where no copulative verb stem occurs, the negative stative copulative verb stem -si occurs in all negative stative copulatives of Southern Ndebele.

The identificative copulatives of Southern Ndebele
The identificative copulatives are those copulatives where the referent is identified in some way.

The affirmative forms of the stative identificative copulative in the indicative mood with a noun as complement
While the personal copulatives contain the particular subject agreement morpheme, the impersonal forms contain no subject morpheme.The impersonal forms are discussed here.
Impersonal identificative copulatives with a noun from any noun class as a complement (except classes 1a, 8, 9 and 10) are formed by using one of two basic strategies.The choice between these strategies is phonologically conditioned.If the noun root is monosyllabic the pre-prefix is omitted, while the true prefix 3 is reduplicated.Consider the stative copulative forms listed below, derived from nouns from classes 3 and 5 respectively: In the case of class 1a nouns, the identificative copulative particle ngi is prefixed to the noun, thus: Class 1a noun: ugogo (grandmother)

Ngugogo (< ngi # ugogo).
(It is a grandmother.) 3 In languages that employ a pre-prefix, such as the Nguni languages, it is essential to distinguish between the pre-prefix and the true prefix.In the case of class 7, for instance, the full prefix isi-comprises the pre-prefix i-followed by the true prefix -si-.
Literator 27( 2 4.2 The negative forms of the personal and impersonal stative copulatives in the indicative mood with a noun as complement The negative forms of the personal and impersonal stative identificative copulatives in the indicative mood with nouns from any noun class as complement (except classes 1a and 9) have the following structure: negative morpheme a-, the particular negative subject morpheme (the impersonal subject morpheme kuis used for the impersonal forms), the negative stative copulative verb stem -si, the identificative copulative form of the complementary noun.This implies that the pre-prefix is omitted and the true prefix is reduplicated with monosyllabic noun stems as in example 9a below, while in the case of polysyllabic noun stems, the pre-prefix is omitted, while the tone on the vowel of the true prefix changes to low as in example 9b below.
(You are not an old man.) If the complementary noun is a noun in class 1a, the copulative has the following structure: negative morpheme a-, the particular negative subject morpheme (the impersonal subject morpheme kuis used for the impersonal forms), the negative stative copulative verb stem -si, the identificative copulative particle ngi, the complementary noun in class 1a.Consider example 10 below: 10. Loyo msana akasi nguMabhena (< a-+ -ka-+ -si # ngi # uMabhena).(That boy is not Mabhena.) If the complementary noun is a noun in class 9 the copulative has the structure as explicated as follows: negative morpheme a-, the particular negative subject morpheme (the impersonal subject morpheme ku-is used for the impersonal forms), the negative stative copulative verb stem -si, the identificative copulative particle yi, the complementary noun in class 9. Consider example 11 below: 11. Isilwana lesi asisi yinja (< a-+ -si-+ -si # yi # inja).(This animal is not a dog.) The personal and impersonal forms of the stative identificative copulative in the indicative mood, with a noun from each noun class as complement, are listed in the tables below.The identificative copulative particle ngi occurs before all demonstrative pronouns as well.Consider example 12 below.
(It is that firewood that is used when a fire is made.)

The identificative copulative in the different moods
The only differences between the identificative copulative in the indicative mood and the moods other than the indicative are that the appropriate subject morphemes for that particular mood are used, while, in the negative, the particular subject morpheme is followed by the negative morpheme -ngainstead of being preceded by the negative morpheme a-.
Identificative copulative forms of the situative mood are supplied as examples of these copulatives in a mood other than the indicative mood.

4.7.1
The stative forms of the identificative copulative in the situative mood

Comparison between the identificative copulatives of Southern Ndebele and Zulu
In Southern Ndebele the basic strategy for the formation of the identificative copulatives with nouns as complement is to omit the pre-prefix and to reduplicate the /CV/ of the true prefix (if the noun root is monosyllabic) or to omit the pre-prefix and lower the tone on the true prefix (if the noun root is polysyllabic).This strategy is also prominent in Xhosa.In those instances where a noun from noun class 1a is used as complement, an identificative copulative particle ngi occurs before the noun, while in those instances where a noun from noun class 9 is used as complement, an identificative copulative particle yi occurs before it.
In Zulu, however, the basic strategy is to use the identificative copulative particle ngi (or yi which has developed from ngi) before all nouns.The identificative copulative particle ngi is used before nouns commencing with the vowels a, u and o, while yi is used before nouns commencing with the vowel i.In Southern Ndebele the identificative copulative particle ngi is used before all pronouns (except those of 2nd person singular and class 1 and 1a where ngi becomes ngu.Consider the forms nguwe and nguye in this regard).
However, in Zulu the identificative copulative particle yi is used before all pronouns (except those of 2nd person singular and classes 1 and 1a where, like Southern Ndebele, the identificative copulative particle ngi is used).In the latter instances the identificative copulative particle ngi becomes ngu in Zulu as well, thus resulting in the forms nguwe and nguye.
5.1 The postulation of the identificative copulative particle of Zulu as ng(i) Grammarians are not in agreement as far as the postulation of the form of the identificative copulative particle of Zulu is concerned.This form has been postulated as nga, ngu, ng and ngi.
Van Eeden (1956:397) contends that this copula is nga.However, according to the normal phonological rules of Zulu /a + u > o/.
Hence, one would expect the vowel coalescence in an example such as u-nga-uSabela to result in the form ungoSabela (you are Sabela).The resultant form is, however, unguSabela (you are Sabela).Hence, Van Eeden's postulation cannot be supported.
On the other hand, Poulos and Msimang (1998:356) postulate the identificative copula as ngu.Again, the general phonological rules for vowel coalescence in Zulu yield a problem since /u + a > wa/, yet the Zulu form of the phrase (they are our boys) is not *bangwabafana (< ba-ngu-abafana) bakithi but bangabafana bakithi.This postulation is therefore also rejected.Doke (1981:216), Ziervogel, Louw and Taljaard (1985:103) as well as Taljaard and Bosch (1988:91) identify the identificative copula as being ng.No Zulu word or morpheme ends on a consonant (except a few ideophones with a paranormal phonological structure).The postulation of the identificative copulative particle as ng is thus also rejected.Posthumus (1978:65 and1988:63) is the only contemporary scholar who postulates the identificative copulative particle of Zulu as ngi.This postulation is supported by language internal as well as language external evidence.
The language internal evidence in support of the postulation of the identificative copulative particle as ngi is the following: • The postulation of the identificative copulative particle as ngi is the only one that does not yield contradictions in terms of the normal phonological rules of Zulu for vowel coalescence.• The variant identificative copulative particle yi of Zulu (that occurs before nouns commencing with the vowel i and the majority of pronouns) seems to be a later development judging from forms such as a si ngimi (It is not me) supplied by Grout (1859:104) and Wanger (1917:644) and ngiti (it is us) noted by Wanger (1917:604).These forms occur as akusimina/akuyimi (it is not me) and yithi (it is us) respectively in present-day Zulu.
The language external evidence in support of the postulation of the identificative copulative particle as ngi is the following: • The identificative copulative particle in the Bantu languages is generally identical in form to the subject morpheme of the first person singular.In this regard one can thus expect the identificative copulative particle of Zulu to occur as ngi.
• Meinhof (1948:153) (1973); Wilkes (1974) and Posthumus (1978) and(1988) who have postulated an underlying stative copulative verb stem -*li (and *-si in the negative) for Zulu have either not motivated the postulation or have based their postulation on the presence of these copulative verb stems in languages such as the Sotho languages.
One scholar who rejects such a postulation outright is Cope (1963:83), who says: Again it is sometimes suggested that the contraction may be /nge/< *ngali, but again there is no evidence in Zulu.
For a detailed discussion of the justification for the postulation of an underlying stative copulative verb stem in Zulu, cf.Posthumus (1978:38-47 and1988:62).
Although Nkabinde supports the postulation of an underlying copulative verb stem *-li in the affirmative forms of the stative identificative copulatives, he rejects the existence of an underlying copulative verb stem in the negative forms of the copulative.He (Nkabinde, 1986:58) maintains: Unlike Posthumus (1980) we do not find justification for postulating a negative form of the lexical li in the deep structure.
It transpires from the examples cited by Gauton (2002:357) that she too does not recognise the existence of a negative copulative verb stem -si.In the analysis of her examples she postulates the underlying negative stative copulative verb stem as *-li.
The overwhelming evidence from a large number of Bantu languages where the negative stative copulative verb stem occurs as -si and the fact that this copulative verb stem actually occurs in examples such as 13 cited below, rule out the possibility of the negative stative copulative verb stem being *-li in Zulu.
The stative copulative verb stem generally has a very limited occurrence in those Bantu languages where it occurs in the surface structure.One of the strongest arguments used by linguists for the postulation of an underlying copulative verb stem *-li (and -si in the negative) for the stative copulatives of Zulu is based on comparative arguments.In this regard reference is generally made to the Sotho languages.However, in contrast to the Sotho languages, the negative copulative verb stem -si of Southern Ndebele occurs in all the subcategories of the stative copulatives (identificative, associative, descriptive and locative forms) and in all the moods.Southern Ndebele (being a language closely related to Zulu) thus offers the strongest language external evidence, on the comparative level, for the postulation of the negative stative copulative verb stem -si for Zulu.
Another consideration for the postulation of an underlying negative stative copulative verb stem -si in Zulu is the remnant form with -si occurring in a very limited number of cases in the identificative copulative as an alternative form.Consider the first three variants in example 13 below in this regard: 13. Angisinguye/Angisiye/Angisuye/Anginguye uThemba.
(Negative morph.a -1st p. sing.sm ngi -negative copulative verb stem -si (# identificative copulative particle (ngi)) # Cl. 1 shortened emphatic pronoun ye(na) (< u -e -na) (I am not him, Themba.) Further language internal support for such a postulation for Zulu is given by Posthumus (1988:62).He argues that the phonological change of the non-indicative negative morpheme from -ngato -ngeand the progressive aspectual morpheme from -sa-to -se-(in the stative copulatives) is due to the influence of the vowel [i] of the omitted copulative verb stems (*-li for the positive forms or *-si for the negative forms respectively).Consider the Zulu examples 14 and 15 below: copulative, only the progressive aspectual morpheme -sa-changes to -se-while the negative morpheme remains -nga-.This is due to the syntagma of the morphemes.The omitted vowel /i/ of the copulative verb stem -si will coalesce with the immediately preceding vowel /a/, of -sa-to become /e/, while the vowel /a/ of -nga-will not coalesce.Consider example 16 below.
16. Umfana ongaseyingane (< o -nga -sa -*si # y(i) # ingane) ... (The boy who is no longer a child ...) The occurrence of the non-indicative negative morpheme as -nga-in the stative copulative forms of Southern Ndebele, as opposed to the negative morpheme -ngeof Zulu, can be explained in view of the fact that the copulative verb stem -si is not omitted in this Southern Ndebele and hence no vowel coalescence /a + i > e/ takes place as is the case in Zulu.(Consider the explanation of the vowel coalescence in example 16 above.) The Southern Ndebele data offer the most systematic evidence (on the comparative level) of all the languages in the South-Eastern Bantu language zone for the postulation of an underlying stative copulative verb stem -si in the negative copulative forms of Zulu.

With an emphatic pronoun as complement Stative Inchoative Personal copulative forms
Literator 27(2) Aug. 2006:109-134 In Southern Ndebele, like the other Bantu languages, the binary opposition of inchoative versus stative is basically marked by the copulative verb stems.However, as already stated, no copulative verb stem occurs in the affirmative stative form.The table below reveals the binary opposition of implication, namely inchoative versus stative and the opposition of actuality, namely affirmative versus negative, as expressed by the copulative verb stems.
(Negative morph.a+Cl. 8sm + negative copulative verb stem -si # reduplicated true prefix + Cl. 8 noun (without pre-prefix)) (They are not dishes.)3.1.3The copulative verb stems In Southern Ndebele (like Zulu) yi is prefixed to nouns in class 9 as identificative copulative particle to both mono-and polysyllabic noun roots, thus: Aug. 20066:109-134  ISSN 0258-2279In Southern Ndebele (like Xhosa) polysyllabic noun roots from noun classes 8 and 10 take the contracted prefix iin-instead of izin-.In Southern Ndebele noun roots of these two noun classes (with the contracted prefix iin-or iim-) still prefix the /CV/ of the true prefix to these contracted prefixes in the copulative form.Consider examples 8a and 8b with nouns from noun classes 8 and 10 respectively: 118 ISSN 0258-2279 Literator 27(2) Aug. 2006:109-134 Stative personal identificative copulatives in the indicative mood with a noun as complement: affirmative and negative formsThe inchoative forms of the identificative copulative are similar to the stative forms except that the inchoative verb stem -ba occurs after the subject morpheme in the affirmative and -bi in the negative forms.(Theinchoative verb stems behave in most respects like any regular verb stem.)Consider the examples below.The stative forms of the identificative copulative in the indicative mood with pronouns as complementIn Southern Ndebele the identificative copulatives with an emphatic pronoun as complement are characterised by the use of the identificative copulative particle ngi for all classes.The basic patterns for the affirmative and negative of the stative forms are indicated below.comprehensive list of stative copulative forms, with emphatic pronouns for the individual noun classes as complement, are listed in the table below.The only difference between the personal and impersonal forms is that in the impersonal forms the subject morpheme ku-is used. A actually states that the identificative copula of Zulu is ngi.He says:In vielen Sprachen wird ein besonderes Wort als pronominale Kopula für alle Klassen gebraucht, im Suahili: ni, ... Zulu: ngi, ... languages to have the same underlying identificative copulative particle).It is thus postulated that, diachronically, Southern Ndebele and Zulu had the same form for the identificative copulative particle, namely ngi.
• The Southern Ndebele data support the postulation of the identificative copulative particle as ngi.(Note that in Southern Ndebele the identificative copulative particle occurs as ngi before nouns of class 1a and before all pronouns.Since Southern Ndebele and Zulu are closely related languages, one can expect 130 ISSN 0258-2279 Literator 27(2) Aug. 2006:109-134 these