deixis in the Dagaare Nominal Group: Syntagmatic and Paradigmatic Perspectives

Abstract

Although the use of multiple deictic items such as definite articles, possessives and demonstratives in the nominal group of many languages has been noted in the literature, the motivation for this phenomenon is often unexplained. The present study examines this phenomenon in the Lobr dialect of Dagaare (Niger-Congo: Mabia/Gur) from both syntagmatic and paradigmatic perspectives. It first identifies three deictic functions in the structure of the Dagaare nominal group: Deictic₁ ^ Deictic₂ ^ Classifier ^ Thing ^ Epithet ^ Numerative ^ Deictic₃ ^ Qualifier. Second, the study shows that the system of deixis in the Dagaare nominal group involves a division of semiotic labor among deictic elements in the realization of (participant) identification, person systems and specificity. Specifically, the definite article is used as a participant tracking device, including the reactivation of participants with unique reference from the common memory of the speaker and addressee. Demonstratives are specificity marking devices and thus encode discrete semantic information of referents, while indefinite determiners are non-specificity markers. The study combines issues that have been pursued as separate topics in the literature to shed insights on deixis in the nominal group as well as related concepts such as definiteness and specificity.

Keywords:

• Dagaare
• definite article
• deixis
• demonstratives
• nominal group
• systemic functional linguistics

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Dagaare (or Dagara) is spoken in the intersection of three West African countries, namely southern Burkina Faso, north-eastern Cote d’Ivoire and north-western Ghana. It consists of six principal dialects, comprising Lobr, Northern Birifor, Southern Birifor, Central Dagaare, Wule, and Waali. There is also a contact variety called Daga-Dyula, a mixture of Dagara and Dyula (Niger-Congo: Mande), and it is spoken in parts of Diebougou in Burkina Faso (cf. Bemile 2000, 212). All data in the present study are from the Lobr dialect. Speakers of Lobr are found in Burkina Faso, around Dissin, Maria Tang and Nyebo (all in the Ioba Province), and in Ghana, around Lawra, Nandom and Hamele (all in the Upper West Region).

2 Abbreviations and symbols: AFFR – affirmative, ADP – adposition, ADV – adverbial particle, COM – comitative, CONJ – conjunction, CONT – continuative, COP – copula, DEF – definite, DEM – demonstrative, DIST – distal, DUP – reduplication; EMP – emphatic, FOC – focus, HM – human, IDENT – identifying pronoun, IDEO – ideophone, IND – indicative, INTJ – interjection, IPFV – imperfective, JUNC – juncture, MOD – modality, NAFFR – non-affirmative, NEG – negative, NFUT – non-future, NMLZ – nominalizer, OBJ – objective; PFV – perfective, PL – plural, PROX – proximal, PRT – particle, PST – past, Q – question particle, QUOT – quotative, REL – relativizer, REM – remote, RNG – range, SG – singular, VOC – vocative, 1 – first person, 2 – second person, 3 –third person, […] – suspension point.

3 For lack of better terms, I would like to make a distinction between “article” and “determiners” in the present study just to emphasize that the Dagaare “definite article” a and what I am calling “indefinite determiners” realize different sub-systems of deixis in the nominal group (cf. Section 5.2). In other words, the Dagaare definite article a and indefinite determiners do not belong to the same paradigm. Whereas the definite article a is essentially an anaphor and tracks referents (or participants) in discourse as presumed (see Dryer 2007, 2013b for a typological overview), the indefinite determiners are rather in semantic opposition with demonstrative determiners and realize non-specificity – the domain of reference of both the indefinite determiners and the demonstrative determiners is exophoric. See Section 5.2 for details.

4 kɔ̀ntɔ́-bìlì (translated into English as ‘beings of the wild’ by Jack Goody 1972, 19–22) are mysterious creatures that are a common motif in the mythology of West African cultures. They are believed to be diminutive human-like creatures with long beard and their feet turned backwards and live deep inside the forest or remote parts of the grassland. The chief being (kɔ̀ntɔ́-bὶε, see line 012) is believed to have extremely weird features. Among the Dagara, these beings are believed to be the source of invention, knowledge of divination and healing, spiritual power, and any other special knowledge or skill. They teach these to persons of good-will who encounter them in the forest for the benefit of the community. Nonetheless these beings are dreaded by both children and adults and nobody wishes to encounter them, but those who survive them are revered for their knowledge and skill.

5 Nominal compounding often leads to the formation of new nouns. In Dagaare, nominal compounding shows morphological and phonological characteristics. Morphologically, only the noun in final position occurs in its full form, which consists of the root and number inflection; preceding nouns in the compound normally occur in only their root form and so do not show number inflection. Phonologically, the items in the compound normally constitute a single phonological word, indicated by vowel harmony – the vowels of West African languages can be classified into two groups based on their tongue root value: [advanced tongue root] (e.g., e, i, o) or [retracted tongue root] (e.g., ε, ɩ, ʋ). The vowels of a word will be derived from only one of these classes of vowels and not from both, e.g., in the nominal compound po-bile (‘young man’) from the nouns pɔl (‘young’) and bile (‘offspring’), the retracted tongue root vowel [ɔ] of the initial noun is assimilated to harmonise with the [advanced tongue root] values of the vowels in the second noun (bile).

6 The sense of intensification by the ideophones is indicated by the reduplication of syllables. Semantically, the use of ideophones with adjectival nouns tend to be a metaphorical extension. e.g., yòlyòlyòl in (11) mimics dripping (‘dripping with beauty’ – could metaphorically derive from dripping oil); ɓõ̀ɓõ̀ in (12) mimics oozing (metaphorically deriving from oozing blood); gòlòlò mimics dangling (as in a dangling stick/pole). Further research is needed on this phenomenon.

7 The common chicken feed in Dagara communities is termites and there is a method of trapping them using cow dung (see Kees van der Geest’s documentary Shit & Chicks: https://www.youtube.com/watch?v=D9VSkQstKX4).

8 The system of identification presented here is based on J.R. Martin’s work (e.g., Martin 1983, 1992, Ch.3). I consider it the most suitable characterization of the phenomenon described here for the Dagaare nominal group. However, while Martin (1992) considers identification as a discourse semantic system, I present it as lexicogrammatical system of the nominal group in Dagaare. Further research is needed on the discourse semantics of referent tracking in Dagaare comparable to Martin’s (1992) study on English.

9 Unlike the definite article a and the weak forms of personal pronouns (see Table 4) which always realize [presuming], the [presuming] reading of generic pronouns kã̀w (‘one’) and kã̀w za (‘everyone’) depends on context. They sometimes require the definite article to be interpreted as presumed (see Extract 2, lines 4 and 5).

10 In Dagaare and many Niger-Congo languages, the system of information is realized by lexicogrammatical resources rather than phonological resources. In the Lobr variety of Dagaare, unmarked (or default) focus is realized by the focus particle nɩ or its clitic forms =ɩ and n=. Other varieties have different focus particles (e.g., la in Central Dagaare). Contrastive (or marked) focus is realized by strong pronominal forms, exclusive particles, and cleft-constructions, depending on the context – see Mwinlaaru (2017, 222–254). See e.g., Amfo (2018) for a typology of the realization of focus in African languages.

11 Lyons (1986) makes a distinction between “determiner-genitive” and “adjectival-genitive”. “Determiner-genitives” are presuming while “adjectival-genitives” are not. Based on this typology, Dagaare possessive determiners are of the “adjectival-genitive” kind. We need to make a distinction between the nominal group system of identification and the system of cohesion in Dagaare. As phoric items, possessive determiners participate in the system of cohesion but not in the system of identification. Further research is needed on this phenomenon.

12 We need to distinguish between the indefinite determiners zà (‘any’) and kã̀w zà (‘any’) from the quantifiers zà (‘all’) and kã̀w zà (‘every’) as well as the generic pronouns kã̀w (‘one’) and kã̀w zà (everyone). These are possibly related diachronically in terms of grammaticalization, where the determiner and quantifier uses derive from the pronominal use (see Kuteva et al. 2019, 10, 299–305, on the development of numeral ONE > INDEFINITE). Note that the indefinite pronoun ONE (kaw) is different from the numeral ONE (been or bʋ-en). See Kuteva et al. (2019, 301) for development of numeral ONE to indefinite pronoun. We cannot attest this development in Dagaare with synchronic data.