Till startsida
Till innehåll Läs mer om hur kakor används på gu.se

A comparative study of Bantu noun classes

Participant: Jouni Maho

This is a comparative study of the noun class (grammatical gender) systems in the Bantu languages, the largest group of languages found in Africa. They are spoken by more than 150 million people in central, eastern and southern Africa. The total number of Bantu languages is difficult to estimate. There are at least 300 languages, perhaps as many as 600. (Compare this with the total number of languages in Africa, which is generally put somewhere between 1,500 and 2,000.) The data used in this study is drawn from more than 300 languages/dialects covering the entire Bantu-speaking area.

The main objective is to achieve two goals

  • to describe typological variation
  • to relate that variation to diachronic issues.

Three main methodological approaches are pursued: typology (both synchronic and diachronic), dialectology and linguistic reconstruction. The bulk of the thesis is organized into 6 chapters.

Following two introductory chapters, the first of which introduces some theoretical preliminaries and the second of which presents the Bantu languages, there are three main chapters devoted to a typological overview of what Bantu noun class systems are and how they work (chapter 3), the geographical patterning of a selection of features related to the noun class systems (chapter 4) and a partial reconstruction of the noun class system in an assumed Proto-Bantu language (chapter 5). Chapter 6 concludes the study. Supplementary sections with data and map methodology issues follow.

The first goal (typological variation) is the main concern of chapters 3 and 4. In chapter 3, the most important typological features of the Bantu noun class systems are discussed. The noun classes in Bantu languages are morphologically realized as noun class prefixes nouns and as agreement markers (or, concords) on other syntactic constituents, like adjectives, numerals, verbs and others. For instance, the Kiswahili noun kitabu 'book' is composed of a noun prefix ki- and a nominal root -tabu. Similarly, the adjective kidogo 'small' in the phrase kitabu kidogo 'small book' is composed of an agreement marker ki- and an adjectival root -dogo. Since the noun classes indicate also grammatical number, the plural form of a noun is normally classified in a different noun class than the singular form. Thus the plural of kitabu 'book' is vitabu, which bears a different noun prefix, namely vi-. Another noun belonging to different noun classes would take different prefixes, for instance, mti 'tree' and miti 'trees', which bear the prefixes m- and mi- respectively. Individual noun classes are customarily referred to with a numbering system originally devised in the 19th century. Thus we can compare the noun class systems in different Bantu languages with considerable ease. In Kiswahili, for instance, the above-exemplified noun classes are usually referred to as classes 7 (ki-), 8 (vi-), 3 (m-) and 4 (mi-). A noun class labelled 7 in one Bantu language is thus etymologically related to noun classes labelled 7 in other Bantu languages, irrespective of their phonological forms.

Among the more important features determining the use of noun class markers are thus 'noun class belonging' and 'grammatical number'. In a number of languages, however, a third feature is also important, namely 'animacy', that is, whether or not the "object" referred to by the noun is a human being and/or animal (in contrast to being a non-living thing, like a knife or a stone). An interesting fact about Bantu noun class systems is that these three features ('noun class belonging', 'grammatical number', 'animacy') are employed differently with noun prefixes and concords respectively. Thus in some languages we find noun class prefixes on nouns which have no counterparts in the agreement markers, which is to say that i certain contexts the noun prefix and the agreement marker do not refer to the same noun class, as in the Kiswahili phrase mafundi wabaya 'bad craftsmen'. Here the noun bears a class 6 prefix ma- while the adjective bears the marker wa- of class 2. (This is due to the fact that animate nouns take a special set of agreement markers.) Also other types of non-synchronized uses of noun prefixes and agreement markers occur. This leads to the presentation in chapter 3 of a typological matrix in which different types of noun class systems are placed in relation to each other, depending on how the features 'noun class belonging', 'grammatical number' and 'animacy' affect the use of noun prefixes and concords respectively. The matrix identifies a total of 30 logically possible noun class system types. Of these, only 9 can be attested.

Most Bantu languages exhibit 15 or 16 noun classes, while some have more than 20. These normally group into 10 or more singular/plural pairings. In chapter 4, the geographical distribution of those pairings that can be attested in published sources are traced, and a number of geographically restricted distributions, so-called (typological) coherence areas, are identified. For instance, 7/4 pairings are much more common the north-western Bantu area than elsewhere, and pairings involving plural class 2 are much more common in the western Bantu area than elsewhere. Also other types of coherence areas are traced, like noun class system types, noun class mergers, the classification of diminutive and augmentative nouns, as well as others. A total of 70 putative coherence areas are identified.

The second goal (diachrony) is pursued mainly in chapters 3 and 5. Chapter 3 concludes with a tentative diachronic model describing how Bantu noun class systems (can) change over time. The model is based on the idea that 'animacy' is a feature increasingly working itself into - and out of - the Bantu noun class systems. The starting point for the model are those types of noun class systems that can be attested, which is then expanded by considering also hypothetical noun class system types. The accuracy of the diachronic model depends crucially on the existence of one particular noun class system type which ought to exist, but has not been attested (yet!). The typological matrix predicts that it marks 'noun class belonging', 'grammatical number' and 'animacy' on nouns but only 'grammatical number' and 'animacy' in its agreement system.

Chapter 5 is devoted solely to diachronic issues in that the geographical patternings traced in chapter 4 are tried for reconstructive purposes. Thus certain linguistic features are back-projected into an assumed Proto-Bantu noun clas system by analysing the present-day geographical distribution of specific features. This leads to a slight revision of the conventional reconstructions. In particular, a few features are added and a distinction is made between probable and less probable features.

The main results of the study include:

  1. a typological matrix for the analysis of Bantu noun class systems
  2. the identification of a number of coherence areas
  3. a tentative diachronic model describing changes in Bantu noun class systems
  4. a partial (re-)reconstruction of the Proto-Bantu noun class system

Project-related papers and reports

Maho, Jouni. 1994. Remarks on derivation and inflection in Bantu languages. Seminar held at the Department of Linguistics, Göteborg University. Pp 16.

Maho, Jouni. 1995. Calculating differences in phonological features of consonants in a sample of 25 Bantu languages. Department of Linguistics, Göteborg University. Pp 43.

Maho, Jouni. 1996. Remarks on derivation and inflection. Seminar held at the Department of Linguistics, Göteborg University. Pp 25.

Maho, Jouni. 1996. A small survey of derivation in nine languages: Ju/'hoan (= !Kung), Khoekhoe, Setswana, Babungo, Yoruba, Hindi, Russian, Swedish and Finnish. Department of Linguistics, Göteborg University. Pp 17.

Maho, Jouni. 1997. An overview of nominal morphology in Bantu languages. Seminar held at the Department of Oriental and African Languages, Göteborg University. Pp 24.

Maho, Jouni. 1998. Implications and speculations: a revision of the Proto-Bantu noun class system. Seminar held at the Department of Oriental and African Languages, Göteborg University. Pp 13 plus 23 maps.

Maho, Jouni. 1998. Typological coherence areas in subequatorial Africa. Seminar held at the Department of Oriental and African Languages, Göteborg University. Pp 7, plus 20 figures and maps.

Maho, Jouni. 1999. A comparative study of Bantu noun classes (= PhD dissertation). Orientalia et africana gothoburgensia, no 13. Göteborg: Acta Universitatis Gothoburgensis. ISSN 1404-3556; ISBN 91-7346-364-7. Pp xvi, 388.

  1. Theoretical preliminaries, 1
  2. The Bantu languages: an overview, 18
  3. Bantu noun class systems: a typological overview, 50
  4. Selected distributions and coherence areas, 146
  5. Historical implications and speculations, 243
  6. Summary, conclusions and comments, 271

Supplementary section A, 275
Supplementary section B, 329
Bibliography, 338
Indexes, 369

Distributors & Orders
Acta Universitatis Gothoburgensis
Box 222
SE-40530 Göteborg

Maho, Jouni Filip. 2003. Remarks on a few "polyplural" classes in Bantu. In: Africa & Asia: Göteborg working papers on Asian and African languages and literatures, no 3, pp 161-184.

Sidansvarig: Annika Andersson|Sidan uppdaterades: 2010-09-14

På Göteborgs universitet använder vi kakor (cookies) för att webbplatsen ska fungera på ett bra sätt för dig. Genom att surfa vidare godkänner du att vi använder kakor.  Vad är kakor?