Startseite / Archive / 2011 / I am Subordinate to Gbudwe, but Your Sovereign: Using a Subject Pronoun in Object Position in order to Claim Power


1. Introduction


In the period from 1954 until 1974, Evans-Pritchard published a great number of stories in the Zande language which document the history, cultural institutions and traditions as well as part of the oral literature of the Zande. The stories about Zande history, told by elders among the Azande of Sudan who memorized the days of King Gbudwe, were recorded in the Equatoria Province of Sudan either by Evans-Pritchard himself or by his several assistants. While these stories have long since been evaluated from an anthropological point of view, their linguistic evaluation, is still in a fairly early stage, just as stated by Evans-Pritchard already in 1962 (1962a:289), although they constitute an enormously rich inventory of data for almost all sub-disciplines of linguistics. [1]


On the basis of the excellent translations [2] given by Evans-Pritchard and by his assistants as well as the grammatical descriptions of Zande by Lagae (1921), Gore (1926), Tucker (1959) and Santandrea (1965) and several articles by Boyd (e.g. 1995, 1998), a linguistic evaluation of large sections of the texts is possible even in a situation of armchair fieldwork. Nevertheless, some constructions remain difficult to understand, and the exact meanings of certain lexemes and expressions and the function of certain elements are not easy to determine, above all morphosyntactic functions of certain particles which the given grammatical descriptions do not discuss in detail or not at all. In quite a number of cases a satisfactory analysis of certain passages is possible only with the help of native speakers. This is in particular true for constructions of rare occurrence as the one discussed in the present text, where a personal pronoun which normally marks the subject is used in object function, which constitutes an apparent for-function mismatch. It is found in the story “Installation of Gangura and Mutilation of a Noble (Kuagbiaru)” henceforth referred to as the Gangura-story (Evans-Pritchard 1963a:283-294).


Since no second example of such a construction could be found, it is not possible to decide whether the form is correct or ungrammatical. The highly divergent reactions of the eight speakers of Zande who were given the text to read allow only assumptions about the criteria according to which the construction is grammatical.

In order to determine the function of the surprising pronoun use, the personal pronouns of Zande are investigated in chapter 2. At the end of the chapter follows a discussion of the verb fu ‘give’ and its complements, since they are used in the Gangura-story in a very specific way. Chapter 3 analyses the apparently wrong use of a subject pronoun in object position, the pragmatic function of the expression, and the reaction by native speakers. Conclusions are drawn in chapter 3.

2. The personal pronouns of Zande


Zande has two series of pronouns [3] (see Table 1) which are distinguished in the pronouns of the 1st and 2nd person of the singular and the plural and on the pronouns of Animate (but non-Human) gender and of Inanimate gender. Only pronouns of the 3rd person of Human gender (which distinguish Masculine and Feminine in the singular, while in the plural there is a gender neutral pronoun yo with the variant i) do not make a distinction between the two series.

Table 1: Personal pronouns


Series 1

Series 2
















3 m



yo, i



3 f



3 AN











The anaphoric pronoun ni, neutral to series and number, refers to Inanimate antecedents in oblique case and optionally to Human antecedents in object position.


The general functions of the two series of personal pronouns are described in Pasch (2007, 2010). The first function of the pronouns in series 1 is to mark the subject, and in this function they more often than not precede the verb, but in case of inversion (Lagae 1921:122) they follow the verb. They also occur after the prepositions sa ‘towards’, na ‘with’ (ex.5) and ga 'of' marking alienable possession. They also indicate interpersonal inalienable relationships such as between family members, blood-brothers [4] etc. In all positions where the pronouns of series 1 occur, the respective persons have a high degree of control over the situations described.


The Pronouns of series 2 they are used in inalienable possessive constructions marking part-whole relationships, they are used in object position following the verb, and they are governed by prepositions other than the three just mentioned, e.g. fu 'for' (ex. 2, 3, 4, 5). In all of these constructions, the respective persons have only a low degree of control over the situations described. In order to clearly distinguish the two categories graphically the pronouns of series 1 are written in this paper [5] as separate words, while those of series 2 are written as suffixes of the governing verbs or prepositions.


It appears that pronouns in Zande have always served to mark social distance just like in many other languages of the world (cf. Head 1968), but the criteria for the choice have changed considerably in the last 200 years. Before independence, the 2nd person plural pronoun was not used to indicate social distance as it is e.g. common in Romance languages. After independence, the pattern was borrowed from French as it was spoken in the country. As a consequence the criteria which determine the use of the polite and of the familiar forms differ from those in modern European French. In modern European French the V-form [6] is normally used among adults as a sign of respect and politeness when addressing people with whom there is no close personal relationship irrespective of their social status (horizontal distance) while in former centuries it was used as a sign of deference towards people of higher social status (vertical distance). The T-form is no longer used by the upper class to address commoners, but adults use it above all among family members and friends. The use of the T-form towards children and, by some people, as a sign of disdain towards other adults is a remnant of its function of marking vertical distance.


In Central African Republic as in many other francophone African countries, the French V-form is rather used to express deference in speaking to an individual of a higher social status [7] or to an elder, thus marking vertical distance. People without a high formal education are usually addressed by their employers with tu while they respond with vous. This usage of vous and tu reflects to a high degree the use of the respective pronouns in Sango, Zande and other local languages as could be observed. The man responsible for the laundry in a Catholic Mission in Bangui, e.g., would in no way accept to be addressed by the Sango pronoun ala (2nd person plural), but he considered this pronoun appropriate for addressing a high-ranking person, e.g. the bishop. [8]


In Zande, the pronoun of the 2nd person singular, mo, was formerly used to address a single interlocutor, irrespective of his/her social rank. This behavior is well documented in the conversations between master and servant in Lagae (1921:145-179), and it has been confirmed by speakers of the language in Bangassou (CAR). Since independence the pronoun oni (2nd person plural) has developed into a deferential address pronoun used to address a person of higher status. According to several consultants, this development has taken place due to the influence of Sango.

3. A subject pronoun in syntactic object position


It appears that in the days of the Zande kingdom (last third of the 18th century until the beginning of the 20th century) another grammatical device of marking social distance could be used. This device implies a deliberate violation of the above-mentioned morphosyntactic rules concerning the use of the two series of pronouns. It is, however, in agreement with the features +CONTROL and –CONTROL of the respective series. [9] An example of such rule violation is the fu-mi -construction in example (1) found in the story "Installation of Gangura and mutilation of a noble" (Evans-Pritchard 1963a:283).


... Gbudue














… and then Gbudwe sent me here [lit: gave me to this place].


The use of the pronoun mi of series 1 suffixed to the verb is striking: the pronoun -re of series 2, which among other things functions as object pronoun is to be expected in this position. Unfortunately, other usages of personal pronouns of Series 1 in object position are not (yet) known. Given that the texts published by Evans-Pritchard are well edited and practically devoid of mistakes, we are, however, inclined to regard this example as a clear indication that – at least with regard to the members of the aristocracy – pronouns of series 1 could formerly be used in object position as a very sophisticated way to express a claim of power which could not easily be expressed otherwise preserving all the subtle connotations. It is power of an intermediate level, more precisely that of a prince over a local population, a prince who himself is dependent on his superior, King Gbudwe.


The construction was discussed with seven consultants, speakers of Zande who had been given copies of the story and who had read the entire story before. The first consultant remarked spontaneously that this construction reflects the language of the ruling class. He added that it expresses that Gangura, the speaker, has been given power by Gbudwe and that he intends to remain in his position for good and not just for a limited period. Another consultant did not even notice the unusual pronoun use. When pointed to it he only stated that the construction is perfectly correct and confirmed the interpretation. A third one first regarded the construction as ungrammatical, but when given the explanation of the first consultant agreed to it as fully convincing.


The other consultants, however, quite decidedly declared the construction ungrammatical. No explanation of pragmatic conditions of the usage of the pronouns and the control features of the two series would make them change their opinion. According to them, the correct form of the pronoun of the first person after the verb fu cannot be but -re, belonging to series 2 as is the case in example (2), which is taken from the same story.










Rikita gave me a district to administer (lit. R. gave me to the village).


It is true that all of these latter consultants speak Sango or French most of the time in their daily communications, as is common practice for basically all inhabitants of Bangassou. But they speak Zande quite regularly and they speak it fluently. Nevertheless, it must be noted that one of them grew up in a Sango and Yakoma-speaking environment and learned Zande only at the age of 11 years. He not only strictly rejected the fu-mi -construction as ungrammatical, but did so with regard to several other highly complex syntactic constructions in the stories collected by Evans-Pritchard. This might indicate a lesser degree of fluency in Zande due to imperfect acquisition of the language. Another consultant who was unhappy with the fu-mi -construction has been living in Bangui for many years, where Zande is not used as the daily means of communication. He comes to Bangassou only for a couple of weeks per year, and there he communicates – apart from his primary languages French and Sango – mostly in Nzakara, his second ethnic language. It is therefore likely that he either never learned the construction or that he does not remember them.


It appears that members of the aristocratic Avungara and Bandia clans are more likely to accept the fu-mi -construction as correct than descendants of commoners, even if their competence of Zande is not perfect. This assumption is supported by the fact that the first consultant who accepted the fu-mi -construction in example (1) as correct and as indicative of the speech of the ruling class had similar difficulties with the complex syntactic constructions as the one who learned Zande only as a youth. His membership in the Bandia-clan might explain why the fu-mi -construction is knows despite a lesser completence in other rare constructions. The fact that the two commoners who also approved of the construction have an outstanding knowledge of Zande grammar and lexicon is an indirect indication that the fu-mi -construction was once grammatical, but has become obsolete.


Before discussing the choice of the pronoun of series 1 in example (1), the specific usage of the verb fu 'give' in the text must be investigated, because this verb is used in examples (1) and (2) in a specific reading which is a precondition of the use of the pronoun mi. Fu is a ditransitive verb which in most contexts has an Inanimate object and optionally a Human recipient. Social hierarchies allow, however, for persons of a higher status to transfer persons of lower rank, a transfer which is described by the same verb fu. This applies most notably with regard to women being given to men as gratification or as forced tribute (ex. 3).




























… he did not take his subjects' wives, demanding that the prettiest of them be given to him (lit. it is [such that] she [who] is very beautiful, they give her to him).


Example (3) shows that fu has grammaticalized into the preposition fu 'for' which introduces a recipient or an addressee which is always of Human gender. The preposition is homophonous with the verb with the exception of the pronouns of the 1st and the 2nd person singular as recipients. Here the vowel of the preposition assimilates, giving fe-re 'for me' and fo-ro 'for you', an assimilation not found with the verb fu, cf. ex. (2).


Several examples in the Gangura-text confirm that the installation of a noble as governor of a domain by king Gbudwe is also expressed by the verb fu (ex. 4 and 5), and the respective noble is encoded as a direct object. He is, so to speak, Gbudwe’s present of which his new domain is the recipient. Note that in most of these contexts Evans-Pritchard translates fu as 'send'.


… i
















Gangura was taken and put to Ndukpo’s domain to rule it. (lit.: they took G. and gave him the rule of Ndukpo’s domain).






















Gbudwe has sent him (Gangura) to rule over you, to reside among you (G. gave him, he came to govern you …).

4. Gangura’s claim of power


It must be mentioned that all of these examples are related to Gangura who claims power over the inhabitants of the new domain where Gbudwe has installed him. In the long discussion between Gangura and his subjects, Ndinda, another noble, relates how he was given a domain by Ndukpo, Gangura’s predecessor. He uses a different construction, which describes the process as a normal transfer of alienable possession: the domain is encoded as direct object (ex. 6, 7). and the speaker as the recipient. There is no allusion that Ndinda claims power against the will of others.
























… he (Ndukpo) gave me (Ndinda) an administrative post and gave me two women (lit.: he gave me my village, then he gave my two women).




















[If …] he [Ndukpo] would certainly not have given me a position, (lit.: he gave not at all the village to me at all).


Examples (1) and (2), describe similar situations. They are taken from Gangura’s inaugural address to his new subjects, which is also a speech of self-defense because the people are very much opposed to him because of his bad reputation as an administrator. Gangura tells his new subjects that he was given to a certain domain on two occasions in order to rule it, the first time by Rikita (ex. 2) and the second time by king Gbudwe (ex. 1). The situation in example 2 dates back some years while the situation in example 1 is the one under discussion at the moment when Gangura is speaking. Gangura deliberately expresses the way Rikita granted him a district by means of the grammatical construction fu-re. By means of this wording he indicates quite clearly that this former installation was only a temporary solution, like something that may be given back later. He wants to give the impression that he was given to the respective district for a certain period, and later returned to Gbudwe, as if that had been intended from the beginning. While Rikita installed him only temporarily, Gbudwe now has installed him for good. The difference between the temporary and the irrevocable installation is expressed by the choice of the pronoun in object position. The pronoun of Series 2 having the feature –CONTROL, indicates that his rule is limited while the pronoun of Series 1 having the feature +CONTROL indicates irrevocable rule.


Taking into consideration Gangura's difficult personal history, his bad reputation and the unfriendly reception by his new subjects, he is forced to present himself in a better light in order to gain their confidence. He does so by underlining that he claims power of an intermediate level, i.e. on a level between his new subjects and King Gbudwe as the power superordinate to him. One way to achieve this goal is to explicitly mention that Gbudwe himself installed him. By doing so he demonstrates recognition of Gbudwe’s sovereignty and puts pressure on the new subjects to accept him, Gangura, as their ruler. At the same time he alludes to the limits of his power, which is supposed to help to dispel the impression of him as a governor who ill-treats his subjects. A second, more refined way, is the use of the fu-mi -construction. Mi in object position indicates recognition of the superior position of Gbudwe. Gangura refers to himself as a governor only in the syntactic position of the direct object, i.e. as someone who is an undergoer of what the person in subject position does to him.


At the same time, it is important for him to manifest his authority without provoking the new subject. This intention is also expressed by the pronoun mi which – normally used in subject position – expresses control and agentivity. Its choice makes it quite clear that Gangura aims at being a person of power, a decision maker, someone who has control over the domain. His power shall characterize the future situation with regard to the local population. As their administrator, he does not want to be treated again as a mere messenger who can be sent to some place only temporarily and then be removed. Putting that whole argumentation in a one-syllable function word should make it difficult if not impossible for his interlocutors to react quickly to his speech.

5. Conclusion


Constructions like the one in example (1) are not used in modern Zande, but considered ungrammatical by the majority of speakers. Since several speakers recognized the construction in ex. (1) as a formerly acceptable way for princes to claim power over the population in their domains, we may assume that there was a time in the past when the form was correct. Then the features +CONTROL and –CONTROL must have had an impact on the grammaticality of pronominal forms and could overrule the morphosyntactic rules.


The assumption that the choice of the morphological form could be determined by pragmatic considerations has been documented Belanda Bor by Beatrix von Heyking (in prep.), albeit in a highly different construction. Belanda Bor is a “mixed”-Westerm Nilotic language with primarily Northern Lwoo lexicon which apparently has inherited quite a number of features from Zande and Bviri. In ex. 8 the use of the subject pronoun in object position indicates a high degree of control. This control is, however, not of the type which enables the speaker to do away with the headaches, but of the type which enables him/her to estimate the prospective situation of his/her head and forsee on the basis of certain circumstances whether s/he is going to have headaches or not. Headaches come or they do not, and the speaker does not have direct influence on the pains. When describing such situations when the pains are there (ex.9) or were there (ex.10) the speaker refers to him/herself by the object or undergoer pronoun, i.e. as a participant who does not have control over the actual situation.


t̪ɔɔ wic








‘I will have headaches tomorrow’





t̪ɔɔ wic





‘I have headaches’ (lit.: I am having headaches)





t̪ɔɔ wic





‘I had headaches’ (lit.: I was having headaches)


A more precise analysis of the Zande example (1) is only possible on the basis of further examples, which still have to be found.

Today, the feature CONTROL is fairly irrelevant with regard to the choice of pronouns from series 1 or 2, their morphological forms and the syntactic position are only determined by their grammatical function.



anaphoric pronoun


animate (but not HUMAN) gender




definite marker (noun-phrase initial)


demonstrative (noun-phrase final)






human gender


inanimate gender


negation marker


relative present


relative past




marker of sequential action or situation








marking of tense and aspect according to Boyd 1998

3m / 3f

3rd person singular masculine / feminine


3rd person plural

1 / 2s.1

1st/2nd person singular, series 1

1 / 2s.2

1st/2nd person singular, series 2

1 / 2p.1

1st/2nd person plural, series 1

1 / 2p.2

1st/2nd person plural, series 2


Helma Pasch wants to express her deeply felt gratitude to the German Research Foundation for sponsoring a project on the investigation of the Zande language, including a fieldwork trip to Central African Republic where she found the results presented in this paper. Special thanks go to Yvonne Treis, Anne Storch, and Martin Neef who read an earlier version of this paper for their insightful comments and questions. All weaknesses are, however, ours.

Without the help of Olaf Derenthal, working in Alindao, the whole project could not have taken place. He convinced her that carrying out fieldwork in the South-West of CAR is not a risky enterprise. He allowed her to make use of his social network in order to get transport and accommodation and to find the first speakers of Zande. Bishop Peter Marzinkowski, Abbé Alain, Abbé Nestor, and Abbé Philippe of the Catholic Mission in Alindao allowed her a stopover in their guest house on her way to Bangassou and on the way back to Bangui, before she could continue the journey. Soeur Thérèse Béatrice Moustey allowed her to live in the community of the Soeurs de St. Paul de Chartres in Bangassou. Here she was given a comfortable working place on the terrassa, where the consultants enjoyed working with us. Bishop Aguirre from the Catholic Mission of Bangassou who has been working in the Zande territory for 20 years gave us many important pieces of information on Zande culture. We want to express our deeply felt gratitude to all of them. We also owe thanks to the Ministère de l'Éducation National et de la Recherche Scientifique who gave a research permit in two days without any complication. We are particularly indebted to our consultants Abbé François, Pierre Chrysostome, Gervais Zanga, Jerôme Goto, Katawa, Scholastika, Charlemagne L., and Abbé Benjamin who worked happily with us and who provided many valuable ideas and pieces of information, and to Landi Germain, who registered stories and conversations in Zande. Tambuahe gbe fu-roni.


Boyd, Raymond 1995

'Zande.' In: Raymond Boyd (ed.) Les système verbal dans les langues oubangiennes, pp.165-197. Munich: Lincom Europa

Boyd, Raymond 1998

'Zande.' In: Roulon-Doko, Paulette (ed.) Les manières d'« être » et les mots pour le dire dans les langues d'Afrique Centrale, pp.37-62. Munich: Lincom Europa

Brown, R. & A. Gilman 1968

'The pronouns of power and solidarity.' In: J.A. Fishman (ed.) Readings in the Sociology of Language, pp.252–281. Berlin: Mouton de Gruyter

Comrie, Bernard 1985

'Reflections on subject and object control.' Journal of Semantics 4,1: 47-65

Gore, Canon E.C. 1926

A Grammar of Zande. London: Sheldon Press

Evans-Pritchard, E.E. 1954

'Zande Texts.' Man. A Record of Anthropological Science. 54: 164

Evans-Pritchard, E.E. 1955

'Zande Historical Texts (I).' Sudan Notes and Records 36,2: 123-145

Evans-Pritchard, E.E. 1956a

'Zande Historical Texts (II).' Sudan Notes and Records 37,1: 20-47

Evans-Pritchard, E.E. 1956b

'Cannibalism: A Zande Text.' Africa. Journal of the International African Institute 26,1: 73-74

Evans-Pritchard, E.E. 1957

'Zande Historical Texts III.' Sudan Notes and Records 38,1: 74-99

Evans-Pritchard, E.E. 1962a

'Some Zande Texts – Part 1.' Kush. Journal of the Sudan National Board for Antiquities and Museums [Khartoum] 10: 289-314

Evans-Pritchard, E.E. 1962b

'Three Zande Texts.' Man. A Record of Anthropological Science 62: 149-152

Evans-Pritchard, E.E. 1963a

'Some Zande Texts – Part 2.' Kush 11: 273-301

Evans-Pritchard, E.E. 1963b

'Some Zande texts – Part 3.' Kush 12: 251-281

Evans-Pritchard, E.E. 1963c

'Some Zande Folktales-1.' Sudan Notes and Records 44: 43-68

Evans-Pritchard, E.E. 1964

'Some Zande Folk-Tales: 2.' Sudan Notes and Records 45: 59-78

Evans-Pritchard, E.E. 1965a

'Some Zande texts – Part 4.' Kush 13: 213-242

Evans-Pritchard, E.E. 1965b

'Some Zande Folk-Tales: 3.' Sudan Notes and Records 46: 50-66

Evans-Pritchard, E.E. 1966

'Some Zande texts – Part 5.' Kush 14: 300-325

Evans-Pritchard, E.E. 1967

'Some Zande Texts about Family and Kin.' Sudan Notes and Records 48: 99-110

Evans-Pritchard, E.E. 1970

'Plaints: a Zande text.' African Studies. A quarterly journal devoted to the study of African anthropology, government and languages [Johannesburg] 29,3: 200-205

Evans-Pritchard, E.E. 1973

'Some Zande Texts on Vengeance for Death.' Africa 43: 236-243

Evans-Pritchard, E.E. 1974

'Four Zande Texts.' BSOAS (Bulletin of the School of Oriental and African Studies) 37,1: 41-51

Head, Brian F.

'Respect degrees in pronominal reference.' In: Greenberg, Joseph H. (ed.) Universals of Human Language Vol. 3. Word Structure , pp.151-211. Stanford: Stanford University Press

Heine, Bernd and Kyung-An Song 2010

'On the genesis of personal pronouns: some conceptual sources.' Language and Cognition 2,1: 117-147

Heyking, Beatrix von (in preparation)

A Grammar of Belanda Bor. PhD dissertation (Universität zu Köln)

Lagae, Constant Robert 1921

La langue des Azande, Vol.1. Grammaire, exercices, légendes. Bibliothèque Congo 6. Gand: Veritas

Pasch, Helma 2007

'Grammar of Location and Motion in Zande.' In: Mietzner, Angelika and Yvonne Treis (eds.) Encoding Motion. Case Studies from Africa. Annual Papers in African Linguistics 5: 165-181

Pasch, Helma forthcoming

'From Anaphor to Copula in Zande.' In: Proceedings of the 6th World Congress of African Linguistics, ed. by Matthias Brenzinger and Anne-Maria Fehn. Cologne: Rüdiger Köppe

Samarin, W. J. 2002

'Plurality and deference in urban Sango.' In V. Carstens and F. Parkinson (eds.), Advances in African linguistics. Trends in African Linguistics 4, pp.299–311. Trenton, NJ and Asmara, Eritrea: Africa World Press

Samarin, William n.d.

'Central African Republic.' ( , 02.01.2011) In: Countries and their Cultures ( )

Santandrea, Stefano 1965

Languages of the Banda and Zande Groups. A Contribution to a Comparative Study. Naples: Istituto Universitario Orientale

Tucker, A.N. 1959

Le groupe linguistique Zande. Tervuren: MRCB

[1] Apart from this, these stories constitute a highly appreciated cultural heritage which is in the process of getting lost among the Azande. Our consultants were very interested in getting photo-copies of the stories to take them home.

[2] Evans-Pritchard deliberately chose a highly literary style for the translations as opposed to Lagae (1921) whose translations of Zande texts reflect the Zande constructions to such a degree that the result is not always correct French.

[3] A third series, possessive pronouns, is emerging. These are composed of the possessive marker ga plus a pronoun of the 1st series. The possessive pronouns of the 1st and 2nd person singular have developed merged forms: gi or gi-mi ‘my’ and ga, next to ga-mo ‘your’ [gi-mi > gi ‘my’ and ga-mo > ga ‘your’].

[4] Exceptions are the lexicalized forms buba 'my father' and nina 'my mother'.

[5] The orthographic conventions

[6] The terms V-form and T-form, coined by Brown and Gilman (1968), are derived from the plural and singular pronouns of the 2nd person in some Romance languages.

[7] Samarin (n.d.) speaks of 'role'.

[8] This observation is not in full agreement with Samarin (2002, n.d.) who describes the development of ala as a deference expressing pronoun similar to the pronoun vous in modern French, which Heine & Kyung-An Song (2010:133) accordingly characterizes as V-form. Samarin states that any speaker of Sango can address anybody else with the V-form ala instead of using the traditional T-form mo (2nd person singular). Exceptions are children younger than 13 who do not address younger siblings with the V-form (Samarin 2002:307).

[9] The feature CONTROL as discussed in this paper is restricted to the choice of the form of a pronoun. It is unrelated to the question of subject-control and object-control verbs (cf. Comrie 1985).




Es liegen noch keine Kommentare vor.

Möchten Sie Stellung zu diesem Artikel nehmen oder haben Sie Ergänzungen?

Kommentar einreichen.