Navigation – Plan du site

AccueilNuméros8(1)The role of the Babanki noun phra...

The role of the Babanki noun phrase-final enclitic class marker

Pius W. Akumbu et Roland Kießling

Résumés

Le babanki, une langue bantoue des Grassfields, dans la région Nord-Ouest du Cameroun, présente un marqueur de classe enclitique au sein du syntagme nominal (SN), apparenté à celui que l’on trouve dans d’autres langues Ring comme l’aghem et le lamnso’. Comme c’est généralement le cas dans les systèmes de classes nominales des langues Ring, en babanki, tous les modifieurs dans le SN s’accordent avec la tête nominale via des marques d’accord préfixées ou procliticisées. De manière surprenante, dans certaines conditions, il existe un enclitique final supplémentaire dans le SN, qui exprime également la classe nominale du nom tête. La question qui se pose est de savoir pourquoi un tel enclitique supplémentaire exprimant la classe nominale est nécessaire. L’objectif de cet article est de fournir une description de cet enclitique en babanki, à la fois en ce qui concerne les paramètres syntaxiques et sémantico-pragmatiques – ce qui n’a pas été abordé auparavant dans les descriptions disponibles des structures grammaticales babanki. Concernant les conditions syntaxiques, deux paramètres cruciaux déterminent la présence ou l’absence du marqueur de classe enclitique dans le SN : (1) la composition interne du SN en ce qui concerne (a) le type de modifieur impliqué et (b) la présence de la tête, et (2) les conditions externes au SN, en particulier la position syntaxique du SN. Quant aux facteurs pragmatiques, une trace minimale de l’influence du statut de focalisation du SN sur l’occurrence de l’enclitique peut être détectée dans l’absence systématique de l’enclitique sur les modifieurs qui sont intrinsèquement focalisés.

Haut de page

Texte intégral

1. Introduction

1This paper explores the syntactic and pragmatic conditions which determine the presence or absence of the noun phrase (NP) level enclitic class marker in Babanki, a Central Ring Grassfields Bantu language of North-West Cameroon. The Babanki enclitic is clearly related to a formally and functionally similar item found in other Ring languages such as Aghem (Hyman 1979a and 2010), Isu (Kießling 2010), and Lamnso’ (McGarrity & Botne 2001).

2Babanki has twelve noun classes which form nine genders based on distinct agreement patterns. The noun classes are indexed by prefixes or proclitics across a full range of syntactic targets such as modifiers within the NP, e.g. pronominal possessives (poss), nominal possessives (ass), demonstratives (dem), and attributive adjectives (adj), as shown in Table 1. Noun class membership is, to a large extent, overtly marked by nominal affixes in the noun itself. The large majority of these affixes are prefixes with the single exception of class 10, which is marked by a suffix.

Table 1 — Babanki noun class system and some agreement markers

  • 1 Standard Bantu numbering has been used since the 12 noun classes that exist in Babanki correspond (...)
  • 2 A merger of classes 3 and 5 in Babanki can be seen in their identical schwa prefixes as well as va (...)
  • 3 The numbering of noun classes is not continuous, as 4, 11, 12, 14‑18 are missing because no Babank (...)
  • 4 It is puzzling that ‑m is retained after n in class 6a. All consultants we checked with did this c (...)
Class1 Affix Ex. poss
ɔ́m ‘my’
dem
ὲn ‘this’
adj
‘new’
am
‘of the…’
Gloss
1 Ø tsɔ̀ŋ ə̀ɣ‑ɔ̄m y‑ὲn ə́‑fí ə̀= thief
2 və̀‑ və̀‑tsɔ́ŋ ə̀vw‑ɔ́m=ə́ v‑ὲn=ə́ və̀‑fí=və́ və́= thieves
32 ə̀‑ ə̀‑tó ə̀ɣ‑ɔ́m=ə́ y‑ὲn=ə́ ə̀‑fí=ɣə́ ə́= hut
53 ə̀‑ ə̀‑wúm ə̀ɣ‑ɔ́m=ə́ y‑ὲn=ə́ ə̀‑fí=ɣə́ ə́= egg
6 à‑ à‑wúm àɣ‑ɔ́m=ə́ y‑ὲn=ə́ à‑fí=ɣə́ ə́= eggs
6a mə̀‑ mə̀‑ɲín ə̀mw‑ɔ́m=ə̀ m‑ὲn(=mə́)4 mə́‑m‑fí=mə̀ mə̀= birds
7 kə̀‑ kə̀‑mbò ə̀k‑ɔ́m(=kə́) k‑ὲn(=kə́) kə̀‑fí=kə́ kə́= bag
8 ə̀‑ ə̀‑mbò ə̀vw‑ɔ́m=ə́ v‑ὲn=ə́ ə̀‑fí=və́ ə́= bags
9 Ø byí ə̀ɣ‑ɔ̄m y‑ὲn ə́‑fí ə̀= goat
10 ‑sə́ byí‑ꜜsə́ ə̀ʃ‑ɔ́m(=sə́) s‑ὲn(=sə́) sə̀‑fí=sə́ sə́= goats
13 tə̀‑ tə̀‑tó ə̀ty‑ɔ́m(=tə́) t‑ὲn(=tə́) tə̀‑fí=tə́ tə́= huts
19 fə̀‑ fə̀‑ɲín ə̀fw‑ɔ́m(=fə́) f‑ὲn(=fə́) fə̀‑fí=fə́ fə́= bird

3Surprisingly, under certain conditions, there is an additional NP-final enclitic, which also indexes the noun class of the head noun, as indicated for the possessive pronouns, the demonstrative, and the attributive adjective in Table 1. One of the crucial conditions to trigger the presence of the enclitic is the type of modifier that follows the head noun. Thus, attributive adjectives (1a) require the enclitic, whereas the interrogative determiner (1b) requires the enclitic to be absent.

(1) Presence vs. absence of the enclitic
a. wàyn zὲn fə̄‑ɲín fə̄‑fí=fə́
1.child p1 buy 19‑bird 19‑new=19.enc
‘The child bought a new bird.’
b. wàyn zὲn fə̄‑ɲín fə̀‑kɔ̀ɔ̀ (*=fə́)
1.child p1 buy 19‑bird 19‑which
‘Which bird did the child buy?

4But the situation is more complex. Beside the modifier type, two other factors determine the presence or absence of the enclitic: the syntactic position of the NP, i.e. whether it is the subject, direct object, or prepositional complement, and the presence or absence of a lexical head in the NP. Moreover, the distribution of the enclitic across these morphosyntactic contexts is not simply a matter of yes or no. Rather, three scenarios have to be taken into account: (a) obligatory presence (e.g. with attributive adjectives and participles), (b) optional presence (e.g. with demonstratives, genuine pronominal possessives, and the quantifier tsέn ‘some, certain’), and (c) obligatory absence (e.g. with derived pronominal possessives, the quantifier tsὲm ‘whole, all’, numerals, and interrogative modifiers). We show in this paper that the enclitic is syntactically motivated, depending on the nominal modifier and on the syntactic role of the head noun. We proceed in §2 by outlining relevant facts about nominal modifiers and then examine the syntactic and pragmatic conditions of the enclitic in detail in §3. This is followed by a conclusion in §4.

2. Nominal modifiers

  • 5 Most of the data for this study has been provided by the first author and checked with five other (...)

5Previous work on the Babanki noun class system has centered around the morphosyntax and morphotonology of elements within the NP (Hyman 1979b and 1980; Mutaka & Chie 2006; Akumbu 2011, 2016 and 2019; Chie 2014) and the concordial prefixes and proclitics which index noun classes across the full range of syntactic targets, i.e. modifiers within the NP: demonstratives (§2.1), pronominal possessives (§2.2), nominal possessives (§2.3), adjectives and participles (§2.4), interrogatives (§2.5), numerals (§2.6), and other quantifiers (§2.7). These modifiers generally follow their head noun. Demonstratives and possessives may precede under special focus conditions (§3.6). In addition to the concordial prefixes and proclitics, a set of agreement markers in post-modifier position has also been identified and categorized as suffixes without providing any functional characterisation (Hyman 1980; Akumbu & Chibaka 2012). We reanalyze these “suffixes” as enclitics at the NP level, due to their morphosyntactic behaviour discussed in §3, and explore their function(s) in detail.5

2.1 Demonstratives

6Three demonstrative categories exist in Babanki, i.e. proximal (near speaker), distal (away from speaker), and anaphoric (previous reference). The underlying forms for these demonstrative roots proposed in Hyman (1980: 246) are ὲn, ì, and ì…á, respectively. Agreement c– prefixes are found in all classes and a set of additional =(c)v class markers, i.e. the enclitics discussed in §3, follow the demonstrative. The full forms of the demonstratives are given in Table 2 with the meanings ‘this’, ‘that’, and ‘the one referred to’. The =cv enclitics, i.e. those of classes 6a, 7, 10, 13, and 19 are optional, as seen with the demonstratives here and genuine possessives in §2.2.

Table 2 — Demonstratives

  • 6 When followed by a H tone, the L tone of the demonstrative root is raised to a M tone, as elaborat (...)
Class ‘this’ ‘that’ ‘the one referred to’
1 y‑ὲn y‑ì y‑ì‑á
2 v‑ὲn=ə́6 v‑ì=í v‑ì‑á
3 y‑ὲn=ə́ y‑ì=í y‑ì‑á
5 y‑ὲn=ə́ y‑ì=í y‑ì‑á
6 y‑ὲn=ə́ y‑ì=í y‑ì‑á
6a m‑ὲn(=mə́) m‑ì(=mə́) m‑ì(=mə́)‑á
7 k‑ὲn(=kə́) ky‑ì(=kə́) ky‑ì(=kə́)‑á
8 v‑ὲn=ə́ v‑ì=í v‑ì‑á
9 y‑ὲn y‑ì y‑ì‑á
10 s‑ὲn(=sə́) ʃ‑ì(=sə́) ʃ‑ì(=sə́)‑á
13 t‑ὲn(=tə́) ty‑ì(=tə́) ty‑ì(=tə́)‑á
19 f‑ὲn(=fə́) fy‑ì(=fə́) fy‑ì(=fə́)‑á

2.2 Pronominal possessives

  • 7 This suggests that the derived possessives originate from full-fledged nouns, retaining some nomin (...)

7Babanki pronominal possessives fall in two distinct groups which could be defined by morphosyntactic criteria. The first set of “genuine” possessives includes ɔ́m 1sg, yə̀ ~ wù 2sg, yέs 1pl.excl, and ə́ŋ 2pl (Table 3a), while the second set of “derived” possessives is made up of wέn 3sg, və́ɣə́ŋ 1pl.incl, and və̀wέnə́ 3pl (Table 3b). While genuine possessives have (v)c– prefixes and require possessive concords, the derived ones have (c)v‑ prefixes and require associative concords (discussed in §2.3).7 The additional =(c)v class marking enclitics seen on demonstratives also occur with possessives. The concord consonants of classes 2, 6a, 7, 8, 13, and 19, i.e. v, m, k, t, and f, are either labialized or palatalized caused by the vowel that follows the consonants. Class 13 also shows palatalization before a rounded back vowel revealing a trace of the Proto-Bantu palatal vowel in the prefix, i.e. *ti‑ (de Wolf 1971). 2sg exhibits variation in classes 1, 3, 5, 6, and 9 based on the alternation of y ~ w.

Table 3a — Genuine possessives

  • 8 The lowering of the H tone of 1sg, 1pl.excl, and 2pl in classes 1 and 9 reveals that these two cla (...)
Class 1sg ɔ́m 2sg yə̀ ~ wù 1pl.excl yέs 2pl ə́ŋ
1 ə̀ɣ‑ɔ̄m8 ə̀‑yə̀/ə̀‑wù ə̀‑yɛ̄s ə̀ɣ‑ə̄ŋ
2 ə̀vw‑ɔ́m=ə́ ə̀v‑yə̄=ə́ ə̀v‑yέs=ə́ ə̀vy‑ə́ŋ=ə́
3 ə̀ɣ‑ɔ́m=ə́ ə̀‑yə̄=ə́/ə̀‑wū=ú ə̀‑yέs=ə́ ə̀ɣ‑ə́ŋ=ə́
5 ə̀ɣ‑ɔ́m=ə́ ə̀‑yə̄=ə́/ə̀‑wū=ú ə̀‑yέs=ə́ ə̀ɣ‑ə́ŋ=ə́
6 àɣ‑ɔ́m=ə́ à‑yə̄=ə́/à‑wū=ú à‑yέs=ə́ àɣ‑ə́ŋ=ə́
6a ə̀mw‑ɔ́m=ə̀ ə̀m‑yə̀(=mə́) ə̀m‑yέs=ə̀ ə̀my‑ə́ŋ=ə̀
7 ə̀k‑ɔ́m(=kə́) ə̀k‑yə̀(=kə́) ə̀k‑yέs(=kə́) ə̀ky‑ə́ŋ(=kə́)
8 ə̀vw‑ɔ́m=ə́ ə̀v‑yə̄=ə́ ə̀v‑yέs=ə́ ə̀vy‑ə́ŋ=ə́
9 ə̀ɣ‑ɔ̄m ə̀‑yə̀/ə̀‑wù ə̀‑yɛ̄s ə̀ɣ‑ə̄ŋ
10 ə̀ʃ‑ɔ́m(=sə́) ə̀ʃ‑yə̀(=sə́) ə̀‑ʃέs(=sə́) ə̀ʃ‑ə́ŋ(=sə́)
13 ə̀ty‑ɔ́m(=tə́) ə̀t‑yə̀(=tə́) ə̀t‑yέs(=tə́) ə̀ty‑ə́ŋ(=tə́)
19 ə̀fw‑ɔ́m(=fə́) ə̀f‑yə̀(=fə́) ə̀f‑yέs(=fə́) ə̀fy‑ə́ŋ(=kə́)

Table 3b — Derived possessives

Class 3sg ꜜwέn 1pl.incl və́ɣə́ŋ 3pl və̀wéꜜnə́
1 ə̀‑wέn ə̀‑və́ɣə́ŋ ə̀‑və̀wéꜜnə́
2 və́‑ꜜwέn və́‑və́ɣə́ŋ və́‑və́wéꜜnə́
3 ə́‑ꜜwέn ə́‑və́ɣə́ŋ ə́‑və́wéꜜnə́
5 ə́‑ꜜwέn ə́‑və́ɣə́ŋ ə́‑və́wénə́
6 ə́‑ꜜwέn ə́‑və́ɣə́ŋ ə́‑və́wéꜜnə́
6a mə̀‑wέn mə̀‑və́ɣə́ŋ mə̀‑və̀wéꜜnə́
7 kə́‑ꜜwέn=kə́ kə́‑və́ɣə́ŋ kə́‑və́wéꜜnə́
8 ə́‑ꜜwέn ə́‑və́ɣə́ŋ ə́‑və́wéꜜnə́
9 ə̀‑wέn ə̀‑və́ɣə́ŋ ə̀‑və̀wénə́
10 sə́‑ꜜwέn=sə́ sə́‑və́ɣə́ŋ sə́‑və́wéꜜnə́
13 tə́‑ꜜwέn=tə́ tə́‑və́ɣə́ŋ tə́‑və́wéꜜnə́
19 fə́‑ꜜwέn=fə́ fə́‑və́ɣə́ŋ fə́‑və́wéꜜnə́

8The behavior of the derived possessives is not uniform with respect to the enclitic, i.e. 3sg shows enclitics in 7, 10, 13, 19, but not in the rest, and deviates in this from the genuine possessives. 1pl.incl and 3pl do not show any enclitics throughout the entire paradigm.

2.3 Nominal possessives

9A possessive relationship between a possessed head noun and a following possessor noun is marked by associative morphemes (am), which index the noun class of the possessed noun, as listed in Table 4. Since morphophonological processes link the associative marker (am) with N2 rather than N1 (Hyman 1979a; Akumbu 2011), the associative markers are considered to be proclitics to N2. Associative markers reflect a tonal contrast of L tone for classes 1, 6a, and 9 vs. H tone for all other classes (Hyman 1980: 239), inherited from Proto-Grassfields (Watters 2003: 241).

Table 4 — Associative markers

Class am Class am
1 ə̀= 7 kə́=
2 və́= 8 ə́=
3 ə́= 9 ə̀=
5 ə́= 10 sə́=
6 ə́= 13 tə́=
6a mə̀= 19 fə́=

10The associative marker occurs between the two nouns and agrees with the modified head noun, which is in n1 position, e.g. kə̀mbò kə́=və̄tsɔ́ŋ ‘bag of thieves’.

2.4 Adjectives and participles

11Genuine adjectives and deverbal adjectives, i.e. participles, have (c)v‑ concordial prefixes as well as a set of additional =cv class markers, i.e. the enclitics, e.g. fə̀ɲín fə̄‑fí=fə́ ‘new bird’, fə̀ɲín fə̄‑fɨ́fə́=fə́ ‘white bird’. The full paradigm of genuine and derived adjectives is given in Table 5.

Table 5 — Genuine and deverbal adjectives

Class ‘new’ fɨ́f ‘be(come) white’
1 ə́‑fí ə́‑fɨ́fə́
2 və̀‑fí=və́ və̀‑fɨ́fə́=və́
3 ə̀‑fí=ɣə́ ə̀‑fɨ́fə́=ɣə́
5 ə̀‑fí=ɣə́ ə̀‑fɨ́fə́=ɣə́
6 à‑fí=ɣə́ à‑fɨ́fə́=ɣə́
6a mə́‑m‑fí=mə̀ mə́‑m‑fɨ́fə́=mə̀
7 kə̀‑fí=kə́ kə̀‑fɨ́fə́=kə́
8 ə̀‑fí=və́ ə̀‑fɨ́fə́=və́
9 ə́‑fí ə́‑fɨ́fə́
10 sə̀‑fí=sə́ sə̀‑fɨ́fə́=sə́
13 tə̀‑fí=tə́ tə̀‑fɨ́fə́=tə́
19 fə̀‑fí=fə́ fə̀‑fɨ́fə́=fə́
  • 9 Class 6a has an additional homorganic nasal following the ordinary agreement prefix. This addition (...)

12Classes 1 and 9 lack additional class markers following the adjectives and participles, and there is no indication of the presence of a floating concordial L tone, as with the genuine possessives (see fn. 8). Unlike the other classes where the prefix tone is L, classes 1, 6a, and 9 prefixes have H tone, an exact mirror image of the tonal patterns found in associative concords. This phenomenon remains a puzzle that needs further exploration as we are unable to explain it so far. A similar pattern is found with the quantifier tsέn ‘some, certain’ (§2.7).9

2.5 Interrogative modifiers

13The interrogative modifiers kɔ̀ɔ̀ ‘which’ and ʃə̀ʔ ‘how many’ are indexed for noun class by means of prefixes only, as shown in Table 6.

Table 6 — Interrogative modifiers

Class kɔ̀ɔ̀ ‘which’ Class kɔ̀ɔ̀ ‘which’ Class ʃə̀ʔ ‘how many’
1 kɔ̀ɔ̀/kɔ̄ɔ̄ 7 kə̀‑kɔ̀ɔ̀ 2 və̀‑ʃə̀ʔ
2 və̀‑kɔ̀ɔ̀ 8 ə̀‑kɔ̀ɔ̀ 6 à‑ʃə̀ʔ
3 ə̀‑kɔ̀ɔ̀ 9 kɔ̀ɔ̀/kɔ̄ɔ̄ 6a mə̀‑n‑ʃə̀ʔ
5 ə̀‑kɔ̀ɔ̀ 10 sə̀‑kɔ̀ɔ̀ 8 ə̀‑ʃə̀ʔ
6 à‑kɔ̀ɔ̀ 13 tə̀‑kɔ̀ɔ̀ 10 sə̀‑ʃə̀ʔ
6a mə̀‑ŋ‑kɔ̀ɔ̀ 19 fə̀‑kɔ̀ɔ̀ 13 tə̀‑ʃə̀ʔ
  • 10 The raising of the L tone of kɔ̀ɔ to M in classes 1 and 9 is due to the presence of a floating H b (...)

14With the interrogative modifier kɔ̀ɔ̀ ‘which’, classes 1 and 9 also have a form with M tone, which occurs after L tone nouns (Hyman 1980: 241), e.g. class 9: byí kɔ̀ɔ̀ ‘which goat?’ vs. ɲàm kɔ̄ɔ̄ ‘which animal?’.10 A similar pattern is seen with the numeral ‘one’ in the following section.

2.6 Numerals

15Only the numbers 1 through 5 exhibit noun class concord by a set of (c)v‑ prefixes identical to those on interrogative modifiers (except for class 6a), as shown in Table 7. The numerals 1, 2 and 4 come with a floating H tone, which prevents the L tone from downgliding.

Table 7 — Numerals

Class mùʔ  ‘one’ bò  ‘two’ táʔ ‘three’ kàʔ  ‘four’ tàn ‘five’
1 mùʔ/mūʔ
2 və̀‑bò və̀‑táʔ və̀‑kàʔ və̀‑tàn
3 ə̀‑mùʔ
5 ə̀‑mùʔ
6 à‑bò à‑táʔ à‑kàʔ à‑tàn
6a ə̀‑m‑bò ə̀‑n‑táʔ ə̀‑ŋ‑kàʔ ə̀‑n‑tàn
7 kə̀‑mùʔ
8 ə̀‑bò ə̀‑táʔ ə̀‑kàʔ ə̀‑tàn
9 mùʔ/mūʔ
10 sə̀‑bò sə̀‑táʔ sə̀‑kàʔ sə̀‑tàn
13 tə̀‑bò tə̀‑táʔ tə̀‑kàʔ tə̀‑tàn
19 fə̀‑mùʔ

2.7 Quantifiers

  • 11 Classes 2, 3, 5, 6, 6a, and 8 have an additional H schwa, seemingly an epenthetic vowel between tw (...)

16The quantifier tsέn ‘some, certain’ requires (c)v‑ concordial prefixes and a set of additional =cv enclitics identical to those found on adjectives and participles, discussed in §2.4.11

Table 8 — Quantifier tsέn ‘some, certain’

Class tsέn ‘some, certain’ Class tsέn ‘some, certain’
1 ə́‑tsέn 7 kə̀‑tsέn=kə́
2 və̀‑tsέnə́=və́ 8 ə̀‑tsέnə́=və́
3 ə̀‑tsέnə́=ɣə́ 9 ə́‑tsέn
5 ə̀‑tsέnə́=ɣə́ 10 sə̀‑tsέn=sə́
6 à‑tsέnə́=ɣə́ 13 tə̀‑tsέn=tə́
6a mə́‑n‑tsέnə́=mə̀ 19 fə̀‑tsέn=fə́

17Another quantifier tsὲm ‘whole’ (that occurs with singular nouns) and ‘all’ (which occurs with plural nouns) has only L tone (c)v‑ concordial prefixes, as in Table 9. The concordial prefixes are the same as with tsέn ‘some, certain’, except that the tone in classes 1, 6a, and 9 is also L.

Table 9 — Quantifier tsὲm ‘whole/all’

Class tsὲm ‘whole, all’ Class tsὲm ‘whole, all’
1 ə̀‑tsὲm 7 kə̀‑tsὲm
2 və̀‑tsὲm 8 ə̀‑tsὲm
3 ə̀‑tsὲm 9 ə̀‑tsὲm
5 ə̀‑tsὲm 10 sə̀‑tsὲm
6 à‑tsὲm 13 tə̀‑tsὲm
6a mə̀‑n‑tsὲm 19 fə̀‑tsὲm

18Other quantifiers such as fə̀ndέʔ (19/6a) ‘few’ and kə̀dɨ̀ŋ (7/8) ‘much/many’ are actually quantity nouns, which are typically used as heads in associative constructions under dependency reversal for attributing quantities, e.g. kə̀dɨ̄ŋ kə́=víʔí ‘a multitude of people’. Their concordial behaviour thus follows the one of nominal possessives (§2.3).

3. The noun phrase enclitic

19In many cases, modified NPs require a final enclitic element that agrees in noun class with the head noun. The full set of enclitic noun class indexes is given in Table 10 in contrast to both the associative marker (am) set, which procliticizes to a nominal modifier, and the subject marker (sm), which procliticizes to the verb.

Table 10 — NP-final enclitics in contrast to associative and subject proclitics

  • 12 Bracketing indicates the deletion of the initial consonants of enclitics in classes 2, 3, 5, 6, 6a (...)
  • 13 As suggested by one of the reviewers, Proto-Bantu agreement prefixes of classes 3 and 6 have been (...)
Class enc12 am sm Class enc am sm
1 = ̀ ə̀= Ø= 7 =kə́ kə́= kə́=
2 =(v)ə́ və́= ə́= 8 =(v)ə́ ə́= ə́=
3 =(ɣ)ə́13 ə́= ə́= 9 = ̀ ə̀= Ø=
5 =(ɣ)ə́ ə́= ə́= 10 =sə́ sə́= sə́=
6 =(ɣ)ə́ ə́= ə́= 13 =tə́ tə́= tə́=
6a =(m)ə̀ mə̀= Ø= 19 =fə́ fə́= fə́=
  • 14 Outside the Babanki NP, agreement is seen on subject markers between the subject noun and the verb (...)

20While there is considerable formal overlap of the three concordial series, the following observations can be made: The =cv allormorph of the enclitic is identical in form with the associative markers of cv= structure, i.e. və́, mə̀, kə́, sə́, tə́, and fə́ of classes 2, 6a, 7, 10, 13, and 19, respectively. In all other classes, the enclitic is distinct from both the associative marker and the subject proclitic. Associative markers which consist of a single low-toned schwa vowel only, i.e. in classes 1 and 9, correspond to a floating L tone enclitic. In the remaining classes, i.e. 3~5, 6, and 8, the enclitic is distinct by the additional presence of an initial consonant, which is absent in the associative and subject proclitics, i.e. ɣ in classes 3~5 and 6 and v in class 8.14 In a historical perspective, the formal contrast for classes 3~5 and 6 vs. 8 has been retained only in the enclitic and is neutralized in the associative and subject markers. The full paradigm of the enclitic is illustrated by NPs modified by the adjective ‘new’ in (2).

  • 15 The floating enclitic L tone of classes 1 and 9 docks leftwards to form a HL contour tone, which s (...)
  • 16 When found between two H tones, the L tone on the agreement prefix is realized M following the gen (...)
  • 17 The plural form byí‑ꜜsə́ ‘goats’ is realized as byí without its plural class 10 suffix. Unlike the (...)
(2) The enclitic by noun class
a. wàyn ə́‑fí=̀15 b. və̀‑tsɔ́ŋ və̄‑fí=və́16
1.child 1‑new=1.enc 2‑thief 2‑new=2.enc
‘new child’ ‘new thieves’
c. ə̀‑lyə̀ŋ ə̀‑fí=ɣə́ d. ə̀‑wúm ə̄‑fí=ɣə́
3‑bamboo 3‑new=3.enc 5‑egg 5‑new=5.enc
‘new bamboo’ ‘new egg’
e. à‑wúm ā‑fí=ɣə́ f. mə̀‑ɲín mə̄‑m‑fí=mə̀
6‑egg 6‑new=6.enc 6a‑bird 6a‑n‑new=6a.enc
‘new eggs’ ‘new birds’
g. kə̀‑kɨ́ kə̄‑fí=kə́ h. ə̀‑kɨ́ ə̄‑fí=və́
7‑chair 7‑new=7.enc 8‑chair 8‑new=8.enc
‘new chair’ ‘new chairs’
i. byí ə́‑fí=̀ j. byí17 sə̄‑fí=sə́
9.goat 9‑new=9.enc goat.10 10‑new=10.enc
‘new goat’ ‘new goats’
k. tə̀‑lyə̀ŋ tə̀‑fí=tə́ l. fə̀‑ɲín fə̄‑fí=fə́
13.bamboo 13‑new=13.enc 19‑bird 19‑new=19.enc
‘new bamboos’ ‘new bird’

21When an NP occurs in subject position as in (3a‑e), a situation arises in which it becomes difficult to decide whether the agreement marker following the subject NP is the NP-final enclitic or rather the subject marking proclitic to the verb. Both never co-occur. Their distribution can be captured by the following generalization: The occurrence of the subject marker is the default, as shown in (3a‑b). Depending on the internal composition of the NP, however, the enclitic is required as final bracket of the NP, blocking the subject marker. This happens, for example, in case the subject head noun is modified by an attributive adjective as in (3c‑e).

(3)
a. və̀‑tʃí (*=və́) ə́= nə̀ntʃwì
2‑in.law (*=2.enc) 2.sm= p1 come afternoon
‘In-laws came in the afternoon.’
b. və̀‑tʃí və̀‑tsὲm (*=və́) ə́= nə̀ntʃwì
2‑in.law 2‑all (*=2.enc) 2.sm= p1 come afternoon
‘All in-laws came in the afternoon.’
c. və̀‑tʃí və̄‑fí=və́ (*ə́=) nə̀ntʃwì
2‑in.law 2‑new=2.enc (*2.sm=) p1 come afternoon
‘New in-laws came in the afternoon.’
d. ə̀‑lyə̀ŋ ə̀‑fí=ɣə́ (*ə́=) fə̀ŋ nə̀ntʃwì
3‑bamboo 3‑new=3.enc (*3.sm=) p1 fall afternoon
‘A new bamboo fell in the afternoon.’
e. mə̀‑ɲín mə́‑m‑fí=mə̀ (*Ø=) tàyn nə̀ntʃwì
6a‑bird 6a‑n‑new=6a.enc (*6a.sm=) p1 fly afternoon
‘New birds flew in the afternoon.’

22As can be seen in (3), the head noun of the subject NP, və̀-tʃí ‘in-laws’, triggers concordial agreement for class 2 in all cases (3a‑c), just as ə̀-lyə̀ŋ ‘bamboo’ does for class 3 in (3d), and mə̀-ɲín ‘birds’ for class 6a in (3e). With the bare noun as subject in (3a) and with the quantifier tsὲm ‘all’ as final modifier in (3b), no NP-final enclitic =və́ is required and instead the subject marker ə́= serves as agreement target for the head noun. As soon as there is an attributive adjective such as ‘new’ as final modifier of the subject in (3c‑e), however, agreement is realized in the NP-final enclitic =və́ for class 2 in (3c), =ɣə́ for class 3 in (3d), and =mə̀ for class 6a in (3e), all of which block the subject marker. This clearly shows that only a single syntactic slot is available for one of the two, i.e. the NP-final enclitic to the subject or the subject proclitic to the verb, and combining both of them, as in (4), is ungrammatical.

(4)
a. *və̀‑tʃí ə́= =və́ nə̀ntʃwì
2‑in.law 2.sm= =2.enc p1 come afternoon
b. *və̀‑tʃí =və́ ə́= nə̀ntʃwì
2‑in.law =2.enc 2.sm= p1 come afternoon
c. *və̀‑tʃí və̄‑fí ə́= =və́ nə̀ntʃwì
2‑in.law 2‑new 2.sm= =2.enc p1 come afternoon
d. *və̀‑tʃí və̄‑fí =və́ ə́= nə̀ntʃwì
2‑in.law 2‑new =2.enc 2.sm= p1 come afternoon
e. *və̀‑tʃí və̀‑tsὲm ə́= =və́ nə̀ntʃwì
2‑in.law 2‑all 2.sm= =2.enc p1 come afternoon
f. *və̀‑tʃí və̀‑tsὲm =və́ ə́= nə̀ntʃwì
2‑in.law 2‑all =2.enc 2.sm= p1 come afternoon

23In cases of ambiguity, i.e. where the contrast of enclitic and subject marker is neutralized, as in classes 1, 7, 9, 10, 13, and 19, the placement of a pause is the only way to distinguish the enclitic from the subject marker. If an intonational break or pause (ǀ) can occur before the element, it must be the subject marker as in (5a). If the break can be inserted after the element, it must be the enclitic, as in (5b).

(5)
a. və̀‑tʃí və̀‑tsὲm ǀ ə́= nə̀ntʃwì
2‑in.law 2‑all 2.sm= p1 come afternoon
‘All in-laws came in the afternoon.’
b. və̀‑tʃí və̄‑fí=və́ ǀ nə̀ntʃwì
2‑in.law 2‑new=2.enc p1 come afternoon
‘New in-laws came in the afternoon.’

24In general, the distribution of the enclitic is determined by two parameters, i.e. NP-internal conditions pertaining to the type of modifier, as specified in the columns of Table 11a, and NP-external conditions pertaining to the wider syntactic environment, as specified in the lines of Table 11a. These conditions determine whether the enclitic is either obligatorily present (indicated by +), optionally present (indicated by (+)) or obligatorily absent (indicated by –). Patterns of identical parametrical values are conflated in the condensed Table 11b.

Table 11a — Overview of syntactic contexts of the enclitic according to NP-internal and NP-external conditions

  • 18 Minus in this slot indicates the obligatory absence of the enclitic for all nouns except those of (...)
n18 adj/part dem poss quan n+n num int rel
gen der tsέn tsὲm
Subject
decl +
neg +
wh‑q +
Direct object
decl + (+) (+) (+)
neg + (+) (+) (+)
wh‑q + (+) (+) (+)
imp + (+) (+) (+)
Prepositional complement
decl + + (+) (+) (+)
neg + + (+) (+) (+)
wh‑q + + (+) (+) (+)
imp + + (+) (+) (+)

Table 11b — Condensed overview of syntactic contexts of the enclitic according to NP-internal and NP-external conditions

n adj/part dem, gen poss, quan tsέn der poss, quan tsὲm, n, num, int, rel
Subject +
Direct object + (+)
Prepositional complement + + (+)

25The patterns that emerge from Table 11 will be discussed and exemplified in the following sections, i.e. the obligatory presence of the enclitic with attributive adjectives, participles, and bare nouns (§3.1), its optional presence with demonstratives, genuine pronominal possessives, and the quantifier tsέn ‘some, certain’ (§3.2), and its obligatory absence with derived pronominal possessives, the quantifier tsὲm ‘whole, all’, numerals, interrogative modifiers, and relative clauses (§3.3). As for the NP-external syntactic conditions, the occurrence of the enclitic also depends on the syntactic role of the NP as subject, direct object or prepositional complement and its interaction with the clause type, i.e. whether it occurs in affirmative declarative (decl) or negative declarative clauses (neg), in wh-questions (wh‑q) or in imperatives (imp). In §3.4, we discuss the occurrence of the enclitic in complex NPs with combined modifiers, before proceeding to examine the relationship between the enclitic and the presence or absence of a nominal head in §3.5. Finally, an attempt is made in §3.6 to determine whether the distribution of the enclitic is influenced by focus conditions.

3.1 Obligatory presence of the enclitic

26The enclitic obligatorily follows an NP under two general conditions: (a) the NP terminates in an adjectival or participial modifier and (b) the non-modified noun occurs as prepositional complement. These conditions will be exemplified in the following.

  • 19 The obligatory presence of the enclitic after adjectival modifiers is also reported in Aghem (Hyma (...)

27First, NPs concluded by an adjectival or participial modifier require the presence of the enclitic under all syntactic conditions, i.e. in the affirmative declarative (6a, 7a, 8a), under negation (6b, 7b, 8b), in wh-questions (6c, 7c, 8c), in imperatives (7d, 8d) irrespective of whether they are in subject (6), direct object (7), or prepositional complement position (8).19

28As discussed above, subject position presents a tricky case since the enclitic can easily be confused with the subject marker due to a high degree of syncretism, as shown in Table 10. As soon as the subject NP terminates in an adjectival or participial modifier, the enclitic blocks the surfacing of the subject marker. This is clearly visible with the nouns və̀tʃí ‘in-laws’ (6a), ə̀wúm ‘egg’ (6b), and àwúm ‘eggs’ (6c), which require the concordial enclitic forms =və́ of class 2 and =ɣə́ of classes 5 and 6, respectively, and not their subject marking counterparts, which would converge in the form ə́= for all three classes.

(6) Noun class enclitics with subjects modified by adjectives
a. və̀‑tʃí və̄‑fí=və́
2‑in.law 2‑new=2.enc p1 come
‘New in-laws came.’
b. ə̀‑wúm ə̄‑kú=ɣə́ kó fə́ŋ
5‑egg 5‑raw=5.enc neg p1 fall
‘The raw egg didn’t fall.’
c. à‑wúm ā‑kú=ɣə́ fə̀ŋ
6‑egg 6‑raw=6.enc p1 fall when
‘When did raw eggs fall?’
(7) Noun class enclitics with direct objects modified by adjectives
a. kə̀‑kɔ̄s kə́= ə̄‑sáŋ ə̄‑fɨ́fə́=ɣə́ à
7‑slave 7.sm= p1 give 5‑corn 5‑white=5.enc to 2sg
‘The slave gave white corn to you.’
b. kə̀‑kɔ̄s kə́= kó ə̄‑sáŋ ə̄‑fɨ́fə́=ɣə́ à
7‑slave 7.sm= neg p1 give 5‑corn 5‑white=5.enc to 2sg
‘The slave didn’t give white corn to you.’
c. kə̀‑kɔ̄s kə́ yὲn ə̄‑káŋ ə̄‑fí=və́
7‑slave 7.sm p1 see 8‑dish 8‑new=8.enc where
‘Where did the slave see new dishes?’
d. fə̀‑ɲín fə̄‑fέnə́=fə́ à wàyn
give.imp 19‑bird 19‑black=19.enc to 1.child
‘Give the black bird to the child!’
(8) Noun class enclitics with prepositional complements modified by adjectives
a. kə̀‑kɔ̄s kə́= wʉ̀ ə̄‑sáŋ à və̀‑tʃí və̄‑fí=və́
7‑slave 7.sm= p1 keep 5‑corn for 2‑in.law 2‑new=2.enc
‘The slave kept corn for new in-laws.’
b. kə̀‑kɔ̄s kə́= kó wʉ̀ ə̄‑sáŋ à mə̀‑ɲín mə̄‑m‑fέnə́=mə̀
7‑slave 7.sm= neg p1 keep 5‑corn for 6a‑bird 6a‑n‑black=6a.enc
‘The slave didn’t keep corn for black birds.’
c. kə̀‑kɔ̄s kə́= wʉ̀ ə̄‑sáŋ à mə̀‑ɲín mə̄‑m‑fέnə́=mə̄
7‑slave 7.sm= p1 keep 5‑corn for 6a‑bird 6a‑n‑black=6a.enc when
‘When did the slave keep corn for black birds?’
d. wāyn nə̀ fə̀‑ɲín fə̄‑fí=fə́
give.imp 1.child with 19‑bird 19‑new=19.enc
‘Give the child a new bird!’

29With non-modified, i.e. bare nouns, the enclitic is obligatorily present only when it is the prepositional complement in affirmative declaratives (9a), negative declaratives (9b), wh-questions (9c), and imperatives (9d).

(9) Noun class enclitics with non-modified nouns as prepositional complements
a. kə̀‑kɔ̄s kə́= nə̀ kə̀‑mbò=kə́
7‑slave 7.sm= p1 come with 7‑bag=7.enc
‘The slave came with a bag.’
b. kə̀‑kɔ̄s kə́= kó à tə̀‑fɔ̄yn=tə́
7‑slave 7.sm= neg p1 give to 13‑chief=13.enc
‘The slave didn’t give to chiefs.’
c. kə̀‑kɔ̄s kə́= nə̀ fə̀‑ɲín=fə́
7‑slave 7.sm= p1 come with 19‑bird=19.enc when
‘When did the slave come with a bird?’
d. ʃí wāyn fə̀‑ɲín=fə́
take.imp 1.child from 19‑bird=19.enc
‘Take the child from a bird!’

30Class 10 nouns present a special case. In contrast to all other nouns, they never come with a noun class prefix, but instead with a final class marker =sə́, under various NP-internal and NP-external conditions, as shown in (10), where the NP (whether modified or not) is subject (10a‑b), direct object (10c‑d), or prepositional complement (10e‑f).

(10) Noun class 10 enclitic
a. byí=ꜜsə́ nə̀ntʃwì
goat=10.enc p1 come afternoon
‘Goats came in the afternoon.’
b. byí sə́= wāyn=ꜜsə́ nə̀ntʃwì
goat 10.am= 1.child=10.enc p1 come afternoon
‘The child’s goats came in the afternoon.’
c. lyɨ̀ ndzàm=sə́ kə̀‑mbò
1sg p1 take axe=10.enc from 7‑bag
‘I took axes from a bag.’
d. lyɨ̀ ndzàm sə̄‑fí=sə́ kə̀‑mbò
1sg p1 take axe 10‑new=10.enc from 7‑bag
‘I took new axes from a bag.’
e. wàyn wʉ̀ ā‑gə́ŋ á ɲám=ꜜsə́
1.child p1 keep 6‑grass for animal=10.enc
‘The child kept grass for cattle.’
f. kə̀‑kɔ̄s kə́= wʉ̀ ā‑gə́ŋ á ɲám sə̄‑kwáʔá=ꜜsə́
7‑slave 7.sm= p1 keep 6‑grass for animal 10‑sick=10.enc
‘The slave kept grass for sick cattle.’

31Previous analyses (Hyman 1980; Akumbu & Chibaka 2012) have identified the adnominal marker of class 10 as a suffix. We argue that this “suffix” is in fact identical to the enclitic. As seen, for example in (10d), the item sə́ that follows ndzàm is not a noun “suffix”, i.e. it is not ndzàm-sə́ fí, but rather the concordial prefix for the adjective, i.e ndzàm sə̄-fí. Similarly, in (10b), the item sə́ after byí is the associative proclitic on N2, i.e. it is not byí-sə́ wāyn, but rather byí sə́=wāyn. Under a “suffix” analysis this would force us to assume that class 10 nouns regularly lose their adnominal noun class “suffix” under modification (through replacement by the dependent agreement marker, as pointed out in fn. 16, i.e. ndzàm-sə́ sə̄-fí ndzàm sə̄-fí). Under the enclitic analysis, this would not be needed at all since the enclitic is actually retained, i.e. ndzàm sə̄-fí=sə́, and only separated from the noun by intervening modifiers. The only extravagance that sets class 10 apart from all other classes is the fact that the enclitic is obligatory present in unmodified class 10 nouns, i.e. in a context where it is obligatorily absent in all other classes.

32This is illustrated with the nouns byí=sə́ ‘goats (class 10)’ and tə̀-ɣàm ‘mats (class 13)’ when they occur in subject position, as direct objects, or as prepositional complements.

(11) Presence vs. absence of enclitic on non-modified subjects of class 10 and class 13
a. byí=ꜜsə́ fə̀ŋ á nsé
goat=10.enc p1 fall on 9.ground
‘Goats fell on the ground.’
b. tə̀‑ɣām (*=tə́) tə́ fə̀ŋ á nsé
13‑mat 13.sm p1 fall on 9.ground
‘Mats fell on the ground.’
(12) Presence vs. absence of enclitic on non-modified direct objects of class 10 and class 13
a. wàyn zὲn byí=ꜜsə́ tsɔ̀ŋ
1.child p1 buy goat=10.enc from 1.thief
‘The child bought goats from the thief.’
b. wàyn zὲn tə̄‑ɣàm (*=tə́) tsɔ̀ŋ
1.child p1 buy 13‑mat from 1.thief
‘The child bought mats from the thief.’
(13) Presence vs. absence of enclitic on non-modified prepositional complements of class 10 and class 13
a. wàyn wʉ̀ ā‑gə́ŋ á ɲám=ꜜsə́
1.child p1 keep 6‑grass for animal=10.ENC
‘The child kept grass for cattle.’
b. wàyn kíʔí ə̄‑kó tə́‑ɣàm (*=tə́)
1.child p1 have 5‑money from 13‑mat
‘The child made money from (selling) mats.’
  • 20 The reason for dropping the erstwhile prefix of class 10 as well as that of the singular class 9 i (...)

33The presence of the enclitic on non-modified class 10 nouns in contexts where the enclitic is obligatorily absent in other classes suggests that the enclitic originated as a device to explicitly mark the plurality of class 10 nouns.20 Hyman (1980: 229) notes that “the ‑sə́ noun suffix is actually introduced by rule just in case there is no overt mark distinguishing class 10 from its class 9 singular”. Bantoid languages with both prefixes and suffixes and the few Narrow Bantu languages with suffixes might have developed in similar ways like Babanki, but the issue requires further investigation.

3.2 Optional presence of the enclitic

34The enclitic is optionally present when the NP terminates in a demonstrative modifier, a genuine possessive (i.e. 1sg, 2sg, 1pl.excl, and 2pl) or the quantifier tsέn ‘some, certain’. This holds for all syntactic positions except for subject NPs, as demonstrated below for direct objects (14) and prepositional complements (15) in affirmative declaratives, negative declaratives, wh-questions, and imperatives. The optional presence of the enclitic refers to the fact that speakers freely use or leave out the enclitic without any functional difference, e.g. fə̀ɲín fὲn=fə́ ‘this bird’ alternates freely with fə̀ɲín fὲn. In the examples given here and elsewhere, the optional presence of the noun class enclitic is indicated by bracketing, e.g. fə̀ɲín fὲn(=fə́).

(14) Optional presence of noun class enclitics with modified direct objects
  • 21 Unlike in Aghem, where the enclitic and the demonstrative are mutually exclusive, both can co-occu (...)
a. kə̀‑kɔ̄s kə́= kə̄‑káŋ k‑ὲn=(kə́) à 21
7‑slave 7.sm= p1 give 7‑dish 7‑dem=(7.enc) to 2sg
‘The slave gave this dish to you.’
b. kə̀‑kɔ̄s kə́= kó fə̄‑ɲín fy‑ì=(fə́) à
7‑slave 7.sm= neg p1 give 19‑bird 19‑dem=(19.enc) to 2sg
‘The slave didn’t give that bird to you.’
c. kə̀‑kɔ̄s kə́= yὲn kə̄‑kɨ́ ə̄k‑yέs=(kə́)
7‑slave 7.sm= p1 see 7‑chair 7‑1pl.excl=(7.enc) where
‘Where did the slave see our chair?’
d. kə̀‑mbò kə̄‑tsέn=(kə́) á ꜜwέn
give.imp 7‑bag 7‑certain=(7.enc) to 3sg
‘Give a certain bag to him!’
(15) Optional presence of noun class enclitics with modified prepositional complements
a. kə̀‑kɔ̄s kə́= wʉ̀ ə̄‑sáŋ à tə̀‑fɔ́yn ty‑ī=(tə́)
7‑slave 7.sm= p1 keep 5‑corn for 13‑chief 13‑dem=(13.enc)
‘The slave kept corn for those chiefs.’
b. kə̀‑kɔ̄s kə́= kó wʉ̀ ə̄‑sáŋ à tə̀‑fɔ́yn t‑ɛ̄n=(tə́)
7‑slave 7.sm= neg p1 keep 5‑corn for 13‑chief 13‑dem=(13.enc)
‘The slave didn’t keep corn for these chiefs.’
c. wàyn ə̄‑sáŋ à kə̄‑kɔ̀s ə̀k‑ɔ́m=(kə́)
1.child p1 give 5‑corn to 7‑slave 7‑1sg=(7.enc) when
‘When did the child give corn to my slave?’
d. ɣáʔ mə̄‑ɲín mə̄‑n‑tsέnə́=(mə̀) á ꜜwέn
catch.imp 6a‑bird 6a‑n‑some=(6a.enc) for 3sg
‘Catch some birds for him!’

35It should be noted that the relevant modifer has to occur in final position of the entire NP since the situation might change as modifiers are combined in complex NPs (§3.4).

3.3 Obligatory absence of the enclitic

  • 22 It can be argued that the absence of the enclitic with the interrogative modifiers and numerals is (...)

36The enclitic is obligatorily absent when NPs terminate in derived possessive modifiers (i.e. 3sg, 1pl.incl, 3pl), nominal possessives, interrogative modifiers,22 the quantifier tsὲm ‘whole, all’, numerals, and relative clauses. This holds for all syntactic contexts, irrespective of whether the NP is in subject position (16), in direct object position (17), or occurs as prepositional complement (18) in affirmative declaratives, negative declaratives, wh-questions, and imperatives. The obligatory absence of the enclitics terminating the NP is shown by marking them with an asterisk and placing them in parentheses after the final modifier.

(16) Obligatory absence of noun class enclitics with modified subjects
a. kə̀‑kɔ̄s kə́=wέn (*=kə́) kə́= pfɨ́
7‑slave 7.am=3sg 7.sm= p1 die
‘His slave died.’
b. və̀‑tʃí və́=wùwì (*=və́) ə́= nə̀ntʃwì
2‑in.law 2.am=1.woman 2.sm= p1 come afternoon
‘The woman’s in-laws came in the afternoon.’
c. à fàŋ mə̄‑ɲín mə̀‑ŋ‑kɔ̀ɔ̀ (*=mə̀) á kə̀‑mbò á
it remain 6a‑bird 6a‑n‑which in 7‑bag q
‘Which birds are left in the bag?’
d. ə̀‑kɔ̀s ə̀‑tsὲm (*=ə́) ə́=
8‑slave 8‑all 8.sm= p1 come when
‘When did all the slaves come?’
e. kə̀‑kɔ̀s kə̀‑mùʔ (*=kə́) kə́= kó ʒɨ́
7‑slave 7‑one 7.sm= neg p1 eat
‘One slave didn’t eat.’
f. və̀‑tsɔ́ŋ (*=və́) á ʒís (*=və́) ə́= bʉ́nə́ ꜜlí
2‑thief rel 1sg p1 feed 2.sm= sleep p0
‘The thieves that I fed have slept.’
(17) Obligatory absence of noun class enclitics with modified direct objects
a. kə̀‑kɔ̄s kə́= kə̄‑kɨ́ kə́=wέn (*=kə́) à
7‑slave 7.sm= p1 give 7‑chair 7.am=3sg to 2sg
‘The slave gave his chair to you.’
b. kə̀‑kúm kə́= kó kə̄‑kɨ́ kə́=kə̄‑kɔ̀s (*=kə́) à
7‑juju 7.sm= neg p1 give 7‑chair 7.am=7‑slave to 2sg
‘The juju didn’t give the chair of a slave to you.’
c. kə̀‑kɔ̄s kə́= kó ɣàʔ və̄‑tsɔ́ŋ və̀‑ʃə̀ʔ (*=və́)
7‑slave 7.sm= neg p1 catch 2‑thief 2‑how.many
‘How many thieves did the slave not catch?’
d. wàyn mə̄‑ɲín mə̀‑n‑tsὲm (*=mə̀) à ndə̀
1.child p1 give 6a‑bird 6a‑n‑all to who
‘Who did the child give all birds to?’
e. fə̀‑ɲín fə̀‑mùʔ (*=fə́) à wàyn
give.imp 19‑bird 19‑one to 1.child
‘Give one bird to the child!’
f. bàʔlə́ kə̀‑kɔ̄s (*=kə́) á ʒís (*=kə́)
2sg sell 7‑slave rel 1sg p1 feed
‘You have sold the slave that I fed.’
(18) Obligatory absence of noun class enclitics with modified prepositional complements
a. wàyn wyè ə̄‑sáŋ á kə̄‑káŋ kə́=və́ɣéŋ (*=kə́)
1.child p1 put 5‑corn in 7‑dish 7.am=1pl.incl
‘The child put corn in our dish.’
b. kə̀‑kúm kə́= kó wʉ̀ byí á kə̄‑kɨ́ kə́=kə̄‑kɔ̀s (*=kə́)
7‑juju 7.sm= neg p1 keep 9.goat on 7‑chair 7.am=7‑slave
‘The juju didn’t keep a dog on the chair of a slave.’
c. kə̀‑kúm kə́= kó tὲm ɲàm à fə̄‑ɲín fə̀‑kɔ̀ɔ̀ (*=fə́)
7‑juju 7.sm= neg p1 shoot 9.animal for 19‑bird 19‑which
‘Which bird did the juju not shoot an animal for?’
d. wàyn wʉ̀ ə̄‑sáŋ á víʔ və̀‑tsὲm (*=və́)
1.child p1 keep 5‑corn for 2.people 2‑all where
‘Where did the child keep corn for everyone?’
e. ə̄‑sáŋ à fə̀‑ɲín fə̀‑mùʔ (*fə́)
give.imp 5‑corn to 19‑bird 19‑one
‘Give corn to one bird!’
f. ɲàm à kə̀‑kɔ̄s (*=kə́) á ʒís (*=və́)
2sg give 9.meat to 7‑slave rel 1sg p1 feed
‘You gave the meat to the slave that I fed.’

37Note that in contrast to the genuine possessive pronouns (14c, 15c) — which allow for the enclitic to be present — the derived possessive pronouns (17a, 18a) align with the nominal possessives (17b, 18b) in that they require the enclitic to be absent obligatorily. This actually points to the origin of the derived possessive pronouns in erstwhile nouns as argued in §2.2.

38The enclitic is also obligatorily absent when non-modified nouns occur as direct object in imperatives, as in (19).

(19) Obligatory absence of noun class enclitics with non-modified direct objects in imperatives
a. kə̄-káŋ (*=kə́)
give.imp 7-dish
‘Give the dish!’
b. ɣə̀ŋ ʒís fə̄‑ɲín (*=fə́)
2pl feed 19‑bird
‘You (pl) feed the bird!’
c. tɔ́ŋtə́ və̄‑tʃí (*=və́)
greet 2‑in.law
‘Greet the in-laws!’

3.4 Complex NPs with combined modifiers

  • 23 The free order of a possessive pronoun and an adjective in a complex NP is also reported in Aghem (...)

39When several modifiers combine in an NP the possessive pronoun and adjective occur closest to the modified noun in either order, followed by the participle, quantifier, demonstrative, numeral, interrogative, and the relative clause, as schematized in the formula in (20a) and illustrated in (20b‑d).23

(20) Order of modifiers in the NP
a. n {poss.pro adj / adj poss.pro} part quan dem num int rel
b. fə̀‑ɲín ə̄fw‑ɔ́m fə̄‑fí fə̄‑fɨ́fə́ fə̄‑tsέn f‑ὲn fə̀‑mùʔ
19‑bird 19‑1sg 19‑new 19‑white 19‑certain 19‑dem 19‑one
‘This one other new white bird of mine.’
c. mə̀‑ɲín mə̄‑m‑fɨ́fə́ mə̀‑wέn mə̄‑m‑fí m‑yì mə̀‑n‑ʃə̀ʔ
19‑bird 6a‑n‑white 6a‑3sg 6a‑n‑new 6a‑dem 6a‑n‑which
‘How many of those new white birds of his?’
d. mə̀‑ɲín mə̄‑m‑fɨ́fə́ mə̀‑wέn mə̄‑m‑fí m‑yì mə̀‑m‑bò á yέn
19‑bird 6a‑n‑white 6a‑3sg 6a‑n‑new 6a‑dem 6a‑n‑two rel 2sg p1 see
‘Those two new white birds of his that you saw.’

40When modifiers are combined, the presence or absence of the enclitic is determined by the final modifier in the NP. The enclitic is present if the NP-final modifier is one that requires the presence of the enclitic (21), optionally present if the NP-final modifier is one that permits the presence of the enclitic (22), and obligatorily absent if the NP-final modifier is one that requires the absence of the enclitic (23), whether the NP is in subject, direct object, or prepositional complement position.

(21) Obligatorily present when NP-final modifier requires the enclitic
  • 24 The class 5 enclitic =ɣə́ is formally different from the sm ə́. In addition, the intonational brea (...)
a. ə̀‑wúm ə̄ɣ‑ɔ́mə́ ə̄‑dʒʉ́ŋə́=ɣə́ |24 pfɨ́
5‑egg 5‑1sg 5‑good=5.enc p1 die
‘My nice egg got broken.’
b. wàyn zὲn fə̄‑ɲín ə̄fw‑ɔ́m fə̄‑fí=fə́
1.child p1 buy 19‑bird 19‑1sg 19‑new=19.enc
‘The child bought my new bird.’
c. wàyn ə̄‑sáŋ à fə̄‑ɲín ə̄fw‑ɔ́m fə̄‑dʒʉ́ŋə́ fə̄‑fí=fə́
1.child p1 give 5‑corn to 19‑bird 19‑1sg 19‑good 19‑new=19.enc
‘The child gave corn to my nice new bird.’
(22) Optionally present when NP-final modifier permits the enclitic
  • 25 For the enclitic, the brackets represent the status of optional presence, while for the sm, the br (...)
a. fə̀‑ɲín ə̄fw‑ɔ́m fə̄‑dʒʉ́ŋə́ fə̄‑fí f‑yì (=fə́) (fə́)25 pfɨ́
19‑bird 19‑1sg 19‑good 19‑new 19‑dem (=19.enc) (19.sm=) p1 die
‘That nice new bird of mine died.’
b. wàyn zὲn fə̄‑ɲín fə̄‑dʒʉ́ŋə́ fə̄‑fí fə̄‑tsέn (fə́)
1.child p1 buy 19‑bird 19‑good 19‑new 19‑certain 19.enc
‘The child bought a certain nice new bird.’
c. wàyn ə̄‑sáŋ à fə̄‑ɲín ə̄fw‑ɔ́m fə̄‑fí f‑ὲn=(fə́)
1.child p1 give 5‑corn to 19‑bird 19‑1sg 19‑new 19‑dem=(19.enc)
‘The child gave corn to this new bird of mine.’
(23) Obligatorily absent when NP-final modifier does not permit the enclitic
a. fə̀‑ɲín fə̄‑dʒʉ́ŋə́ fə̄‑fí f‑yì fə̀‑mùʔ (*=fə́) fə́ pfɨ́
19‑bird 19‑good 19‑new 19‑dem 19‑one 19.sm p1 die
‘That one nice new bird died.’
b. wàyn zὲn fə̄‑ɲín fə̄‑fέnə́ fə̀‑kɔ̀ɔ̀ (*=fə́)
1.child p1 buy 19‑bird 19‑black 19‑which
‘Which black bird did the child buy?’
c. wàyn ā‑sáŋ à mə̀‑ɲín ə̄m‑yə̀ m‑yì mə̀‑n‑tsὲm á
1.child p1 give 6‑corn to 6a‑bird 6a‑2sg 6a‑dem 6a‑n‑all rel
yέn (*=mə̀)
1sg p1 see
‘The child gave corn to all those your birds that I saw.’

3.5 The enclitic and headedness

41Another factor which determines the occurrence of the enclitic is the presence or absence of a lexical head in the NP. Its influence, however, is marginal since headless NPs differ only in one diagnostic environment from NPs with lexical heads with respect to the enclitic, i.e. modification by a derived possessive pronoun. This is shown in Table 12, which contrasts the distribution of the enclitic across all diagnostic modifier types of NPs in dependency from the presence of a lexical head.

Table 12 — Distribution of the enclitic across modifier types in dependency from the presence of a head

+ lexical head – lexical head
adj/part + +
dem (+) (+)
gen poss (+) (+)
der poss +
quan tsέn ‘some, certain’ (+) (+)
quan tsὲm ‘whole, all’
num

42In contrast to the obligatory absence of the enclitic in lexically headed NPs modified by derived possessive pronouns, as presented in §3.3 and exemplified in (17a), (18a), and (24a), the enclitic is obligatorily present in their headless counterparts, as shown in (24b).

(24) Occurrence of enclitic in NPs modified by derived possessive pronouns
a. Obligatory absence of enclitic in lexically headed NPs
kə̀‑kɔ̄s kə́= kə̄‑kɨ́ kə́=wέn (*kə́) à
7‑slave 7.sm= p1 give 7‑chair 7.am=3sg to 2sg
‘The slave gave his chair to you.’
b. Obligatory presence of enclitic in headless NPs
kə̀‑kɔ̄s kə́= kə́=wέn=kə́ à
7‑slave 7.sm= p1 give 7.am=3sg=7.enc to 2sg
‘The slave gave his own to you.’

43Derived possessives provide a clue to understanding the relationship between headedness and the enclitic. It seems that the absence of a lexical head gives room for the enclitic to appear. We speculate that historically there was a significant distinction between headed NPs and headless ones by the presence of the enclitic on headless NPs. It seems that an NP needs to be headed by a full noun, otherwise the enclitic has to be present. It is as if the enclitic is present to make a well-formed NP when the nominal head is absent. Since this relationship between headedness and the enclitic is manifest only on derived possessives, where the enclitic is obligatorily present with headless NPs and obligatorily absent with headed ones, it is unclear whether the heterogeneity is archaic or rather an innovation.

44Otherwise, the distribution of the enclitic with headless NPs matches precisely the one in NPs with lexical heads. The enclitic is obligatorily present with adjectives and participles whether there is a lexical head or not.

(25) Obligatory presence of enclitic with adjectives/participles in the presence or absence of a lexical head
a. kə̀‑kɔ̄s kə́= yὲn (fə̄‑ɲín) fə̄‑fí=fə́
7‑slave 7.sm= p1 see (19‑bird) 19‑new=19.enc
‘The slave saw a new bird/one.’
b. kə̀‑kɔ̄s kə́= zὲn (kə̄‑káŋ) kə̄‑ɣɔ́ʔə́=kə́
7‑slave 7.sm= p1 buy (7‑plate) 7‑big=7.enc
‘The slave bought a big plate/one.’
c. kə̀‑kɔ̄s kə́= zὲn (kə̄‑káŋ) kə́ ə̀=wέn
7‑slave 7.sm= p1 buy (7‑plate) 7.am 1.mother 1.am=3sg
kə̄‑ɣɔ́ʔə́=kə́
7‑big=7.enc
‘The slave bought his mother’s big plate/one.’

45The enclitic is optionally present on demonstratives, genuine possessives, and the quantifier tsέn ‘some, certain’, irrespective of whether there is a lexical head.

(26) Optional presence of enclitic with demonstratives and the quantifier tsέn ‘some, certain’ in the presence or absence of a lexical head
a. kə̀‑kɔ̄s kə́= yὲn (fə̄‑ɲín) f‑ὲn=(fə́)
7‑slave 7.sm= p1 see (19‑bird) 19‑dem=(19.enc)
‘The slave saw this bird/one.’
b. kə̀‑kɔ̄s kə́= yὲn (fə̄‑ɲín) ə̄fw‑ɔ́m=(fə́)
7‑slave 7.sm= p1 see (19‑bird) 19‑1sg=(19.enc)
‘The slave saw my bird/mine.’
c. kə̀‑kɔ̄s kə́= zὲn (kə̄‑káŋ) kə̄‑tsέn=(kə́)
7‑slave 7.sm= p1 buy (7‑plate) 7‑certain=(7.enc)
‘The slave bought a certain plate/one.’

46The obligatory absence of the enclitic is observed with numerals, the quantifier tsὲm ‘whole, all’, interrogative modifiers, and nominal possessives whether there is a lexical head or not.

(27) Obligatory absence of enclitic with numeral and quantifier tsὲm ‘whole, all’ in the presence or absence of a lexical head
a. kə̀‑kɔ̄s kə́= yὲn (fə̄‑ɲín) fə̀‑mùʔ (*=fə́)
7‑slave 7.sm= p1 see (19‑bird) 19‑one
‘The slave saw one (bird).’
b. kə̀‑kɔ̄s kə́= pfè (à-wúm) à‑tsὲm (*=ɣə́)
7‑slave 7.sm= p1 cook (6‑egg) 6‑all
‘The slave cooked all (eggs).’
c. kə̀‑kɔ̄s kə́= pfè (à‑wúm) à‑kɔ̀ɔ̀ (*=ɣə́)
7‑slave 7.sm= p1 cook (6‑egg) 6‑which
‘Which (eggs) did the slave cook?’
d. kə̀‑kɔ̄s kə́= pfè (kə̀‑tʉ́) kə́=fə́‑ɲìn (*=kə́)
7‑slave 7.sm= p1 cook (7‑head) 7.am=19‑bird
‘The slave cooked the bird’s (head).’

3.6 The enclitic and focus

47Focalisation does not seem to have any influence on the distribution of the NP-final enclitic. Evidence presented in §§3.1‑3.5 shows that the presence vs. absence of the enclitic is determined synchronically by NP-internal and NP-external syntactic factors, particularly the type of modifiers involved. When focused for contrastive purposes by means of a pseudo-cleft construction with the copula dìʔ, the enclitic is obligatorily present in NPs modified by adjectives and participles (28a‑b), optionally present on demonstratives (28c), the quantifier tsέn ‘some, certain’ (28d), and genuine possessives (28e), and obligatorily absent on derived possessives (28f), nominal possessives, the quantifier tsὲm ‘whole, all’, interrogative modifiers, numerals (28g), and relative clauses (28h).

(28) Pseudo-cleft constructions for contrastive focus
a. à dìʔ tə̄‑ɣɔ́ŋ tə̄‑fí=tə́ á ə̄‑ŋgə̀ŋ
ds cop 13‑spear 13‑new=13.enc in 5‑house
‘There are new spears in the house (correcting a presupposition that they are old).’
b. à dìʔ tə̄‑ɣɔ́ŋ tə̄‑fέnə́=tə́ á ə̄‑ŋgə̀ŋ
ds cop 13‑spear 13‑black=13.enc in 5‑house
‘There are black spears in the house (correcting a presupposition that they are red).’
c. à dìʔ tə̄‑ɣɔ́ŋ t‑ὲn(=tə́) á zέn
ds cop 13‑spear 13‑dem(=13.enc) rel 2sg p1 buy
‘It is these spears that you bought (correcting a presupposition that they are some others).’
d. à dìʔ tə̄‑ɣɔ́ŋ tə̄‑tsέn(=tə́) á wàyn kwàʔ
ds cop 13‑spear 13‑some (=13.enc) rel 1.child p1 break
‘The child broke some spears (correcting a presupposition that it did not break any).’
e. à dìʔ tə̄‑ɣɔ́ŋ ə̄ty‑ɔ́m(=tə́) á wàyn kwàʔ
ds cop 13‑spear 13‑poss (=13.enc) rel 1.child p1 break
‘The child broke my spears (correcting a presupposition that it broke yours).’
f. à dìʔ tə̄‑ɣɔ́ŋ tə́=wέn (*=tə́) á wàyn kwàʔ
ds cop 13‑spear 13.am=3sg rel 1.child p1 break
‘The child broke his spears (correcting a presupposition it broke mine).’
g. à dìʔ tə̄‑ɣɔ́ŋ tə̀‑bò (*=tə́) á wàyn kwàʔ
ds cop 13‑spear 13‑two rel 1.child p1 break
‘It is two spears that the child broke (correcting a presupposition that it broke ten).’
h. à dìʔ tə̄‑ɣɔ́ŋ á və̀‑tsɔ́ŋ ə́ tsɔ́ŋ (*=tə́)
ds cop 13‑spear rel 2‑thief 2.sm p1 steal
‘It is spears that the thieves stole (correcting a presupposition that they are spears that were bought).’

48Although the distribution indicates that the presence vs. absence of the enclitic in both focus and non-focus conditions matches, one difference is that demonstratives, genuine possessives and derived possessives can also be preposed under focus conditions. When demonstratives and genuine possessives are preposed in the presence of a lexical head, the enclitic is obligatorily absent (whereas it is optionally present in non-focus conditions), revealing a situation where focalization precludes the presence of the enclitic. Table 13 summarizes the distribution of the enclitic under focus conditions in dependency on the position of the modifier and the presence of a lexical head.

Table 13 — Distribution of the enclitic across modifier types under focus conditions

  • 26 “n.a.” (not applicable) indicates that the modifier does not occur in that position.
Postposed modifier Preposed modifier
+ lexical head – lexical head
adj/part + n.a.26 n.a.
dem (+) +
gen poss (+) +
quan tsέn ‘some, certain’ (+) n.a. n.a.
der poss +
nom poss n.a. n.a.
quan tsὲm ‘whole, all’ n.a. n.a.
int n.a. n.a.
num n.a. n.a.
rel n.a. n.a.

49The enclitic is obligatorily present when demonstratives (29a), genuine possessives (29b), and derived possessives (29c) occur without a lexical head.

(29) Obligatory presence of enclitic with preposed demonstratives, genuine possessives, and derived possessives in the absence of a lexical head
a. kə̀‑kúm kə́ kó kə́ bέn. à bɨ̀ʔ ndə́ ə̀k‑ὲn=kə́ á
7‑juju 7.sm neg can 7.sm dance ds play who 7‑dem=7.enc q
‘The juju cannot dance. Who has played this (kind of music)?’
b. à dìʔ ə̄k‑ɔ́m=kə́ lə̀. kə́ ɣὲŋ dzàŋ
ds cop 7‑poss=7.enc only such neg 2pl again f3 call 1sg
á ə̄‑ɲʉ̀ ə́ kə̄‑ŋkɔ́yn
for 8‑thing 8.am 7‑foolish
‘Such is my (fate). You (pl) should never call me for such trivial things.’
c. à sə́ n‑dìʔ kə́=wέn=kə́ və́ wyʉ́ ntéʔ
ds then ncop 7.am=3sg=7.enc 3pl put only little
‘When it was his (plate) they served him very little (food).’

50On the other hand, the enclitic is obligatorily absent when demonstratives (30a), genuine possessives (30b), and derived possessives (30c) are preposed for contrastive focus in the presence of a lexical head.

(30) Obligatory absence of enclitic with preposed demonstratives, genuine possessives, and derived possessives in the presence of a lexical head
a. kə̀‑kúm kə́ kó kə́ bέn. à bɨ́ʔ ndə́ ə̀k‑ὲn
7‑juju 7.sm neg can 7.sm dance. ds play who 7‑dem
kə̀‑bɔ́ʔ (*=kə́) á
7‑music q
‘The juju cannot dance. Who has played this kind of music?’
b. à dìʔ ə̄k‑ɔ́m kə̀‑ɲʉ̀ (*=kə́) á lə̀. kə́ ɣὲŋ dzàŋ
ds cop 7‑poss 7‑thing only such neg 2pl again f3 call
á ə̄‑ɲʉ̀ ə́ kə̄‑ŋkɔ́yn
1sg for 8‑thing 8.am 7‑foolish
‘Such is my fate. You (pl) should never call me for trivial things.’
c. à sə́ n‑dìʔ kə́=wέn kə̀‑káŋ (*=kə́) və́ wyʉ́ ntéʔ
ds then ncop 7.am=3sg 7‑plate 3pl put only little
‘When it was his plate they served him very little (food).’

51Since the absence of the enclitic with preposed demonstratives, genuine possessives, and derived possessives appears to be triggered by focus, we can assume that focused NPs reject the enclitic in the presence of lexical heads.

52Support for this analysis comes from the fact that some quantifiers (tsὲm ‘whole, all’, numerals) and the interrogative modifers are inherently focused and so consistently reject the enclitic, as seen in §§3.3‑3.5 and further exemplified in (31).

(31)
a. ɣə̀ dzàŋ dʒī bwín ə̀ ə̀
3sg p1 call 1sg from road qt 1sg return csc come csc
byèʔè kə̀‑mbò kə́ ndɔ̀ŋ kə̀‑mùʔ (*=kə́)
carry 7‑bag 7.am 1.potato 7‑one
‘He called me to come back from the road and carry one bag of potatoes.’
b. tɔ̀lɔ̀kyí gàʔ à ə́ wέn ɣə̀ tsέn‑tə́ ə̄‑fó
tortoise say to 1.wife 1.am 3sg qt 3sg gather‑rep 8‑thing
ə̀‑tsὲm (*=ə́) ə̀ wyʉ̀ á nsé
8‑all csc keep on 9.ground
‘Tortoise told his wife to gather everything and place it on the ground.’
c. mbì wàyn tʃò ə̀ bὲm á wέn ɣə̀ byéʔé kə̀‑mbò
first 1.child pass csc ask to 3sg qt 3sg carry 7‑bag
kə̀‑kɔ̀ɔ̀ (*=kə́) ə̀ dʒʉ̀ á ə̄‑wɔ́ŋ lɔ́
7‑which csc keep on 5‑market emph
‘The first child moved forward and asked him which bag she should carry to the market.’

53In a historical Ring perspective, it makes sense to interpret these few instances of focus-driven absence of the enclitic as residual. The enclitic may have started out in Babanki as a pragmatically driven strategy to mark non-focalised NPs, possibly reflecting the contemporary situation of the West Ring languages Aghem (Hyman 1979b and 2010) and Isu (Kießling 2010), where the defocalising function of the enclitic is more prominent. In a pan-Ring trend that involves the progressive syntacticization of the enclitic and its concomittant depragmaticization, Babanki seems to have advanced considerably further than West Ring, allowing purely syntactic conditions for the appearance of the enclitic to almost completely replace erstwhile pragmatic ones.

4. Conclusion

54In the preceding sections various syntactic and pragmatic conditions of the NP-final enclitic have been presented. Specifically, we have shown that the type of nominal modifier almost completely determines the presence or absence of the enclitic. As seen, it is obligatorily present on attributive adjectives, participles, and bare nouns, optionally present on demonstratives, genuine pronominal possessives, quantifier tsέn ‘some, certain’, and obligatorily absent on derived pronominal possessives, quantifier tsὲm ‘whole, all’, numerals, and interrogative modifiers.

552Noun class enclitic markers also crop up in other branches of Ring, where they serve similar functions. Thus, evidence from Lamnso’ (East Ring) (McGarrity & Botne 2001) suggests that the presence vs. absence of an NP-final enclitic is conditioned by NP-internal and NP-external factors, similar as in Babanki. For example, the enclitic attaches to adjectives/participles (32a) and the quantifier ‘some’ (32b).

(32)
a. ʃi‑nən ʃ‑e teri=ʃi
19‑bird 19‑rel small=enc
‘small bird’
b. mi‑ndzəv‑mi dzə laav mi‑moʔ=mi
6‑water‑sm be.in house 6‑some=enc
‘Some water is in the house.’

56Nouns in prepositional phrases exhibit the enclitic (33a‑d) unless they are in locative phrases (33e), probably because locative phrases are not headed by prepositions but retain nominal properties, so that the construction may rather have to be analysed as a nominal possessive.

(33)
a. Ntaŋ naa vi‑faveyi fo ʃi‑nən=ʃi
Ntang cook 8‑food for 19‑bird=19.enc
‘Ntang is cooking food for the bird.’
b. Ntaŋ fo vi‑faveyi i ʃi‑nən=ʃi
Ntang give 8‑food to 19‑bird=19.enc
‘Ntang is giving food to the bird.’
c. wu ki koʔ ki‑tʃi i ki‑raŋ=ki
3sg pst climb 7‑tree with 7‑ladder=7.enc
‘He climbed the tree with a ladder.’
d. wu ki wiy i mi‑luʔ=mi
3sg pst climb with 6‑wine=6.enc
‘He came up with wine.’
e. a‑ŋwaʔ‑a dzə sar ki‑daŋ
2‑book‑sm be under 7‑desk
‘The books are under the desk.’

57A comitative adjunct that modifies a direct object requires the enclitic, as in (34b‑c).

(34)
a. m yen ʃi‑nən
1sg see 19‑bird
‘I see a bird.’
b. m yen ʃi‑nən wuna ki‑tam=ki
1sg see 19‑bird com 7‑elephant=7.enc
‘I see a bird and an elephant.’
c. m ker ki‑soo ʃi‑nən wuna ki‑ɣin=ki
1sg have 7‑hoe 19‑bird com 7‑bowl=7.enc
‘I have a hoe, a bird, and a bowl.’

58While the occurrence of the enclitic in Babanki (Central Ring) and Lamnso’ (East Ring) appears to be largely dependent on purely syntactic NP-internal and NP-external factors, it is largely driven by pragmatic conditions in West Ring in that the enclitic that characterizes the so-called “B-forms” is required for nouns in out-of-focus positions in Aghem (Hyman 1979a and 2010) and Isu (Kießling 2010). As illustration, consider the Aghem example in (35) taken from Hyman (1979a: 56), in which the direct object kɨ́-bέ ‘fufu’ forms part of the focus of the clause, i.e. the sentence could answer either questions such as ‘what did you eat?’ or ‘what did you do?’.

(35)
mɔ̂ zɨ̀ kɨ́‑bέ
1sg p1 eat 7‑fufu today
‘I ate fufu today.’

59When focused, the noun kɨ́-bέ ‘fufu’ occurs in its “A-form”, i.e. the enclitic is obligatorily absent, while the noun retains its class 7 prefix kɨ́‑. If, however, the focus is placed on other elements of the clause, e.g. the verb (36a), an adverb (36b), or the subject (36c), the direct object kɨ́-bέ ‘fufu’ is required in its out-of-focus “B-form” bέ=kɔ́, i.e. it has to take the enclitic =kɔ́ of class 7 and drop its noun class prefix.

(36)
a. mɔ̂ zɨ̀ nô bέ=ꜜkɔ́
1sg p1 eat foc fufu=7.enc today
‘I ate fufu today (not sell it).’
b. mɔ̂ zɨ̀ bέ=ꜜkɔ́
1sg p1 eat today fufu=7.enc
‘I ate fufu today (not yesterday).’
c. à mɔ̂ zɨ̀ mùɔ bέ=ꜜkɔ́
ds p1 eat 1sg fufu=7.enc today
‘I ate fufu today (not someone else).’

60This is also seen in negation (37a) and in imperatives (37b), which in Aghem are inherently focused independent of information structure.

(37)
a. ŋ̀ kâ zɨ̀ bέ=ꜜkɔ́
1sg neg eat fufu=7.enc today
‘I didn’t eat fufu today.’
b. zɨ́ bέ=ꜜkɔ́
eat fufu=7.enc
‘Eat fufu!

61From a general Bantu perspective, the role of the Babanki NP-final enclitic appears to be reminiscent of that of the augment in Bantu languages (De Blois 1970; Grégoire & Janssens 1999; Van de Velde 2019). Even though the Bantu augments are (pre)prefixes while the Babanki enclitics are NP-final, their functions and distribution are similar in some ways. Although pervasive in Bantu, certain factors, e.g. propositional act function, and syntactic context determine “whether the use of the augment is prohibited, obligatory or syntactically optional” (Van de Velde 2019: 251). The obligatory absence of the Babanki enclitic under focus when demonstratives and pronominal possessives are preposed can be likened to the absence of the augment on an object noun when it is under focus, e.g. in Ganda JE15 (Hyman & Katamba 1993).

(38) Ganda JE15 (ibid.: 228).
a. y‑à‑gúl‑ìr‑à à‑bá‑ànà è‑bí‑tábó
sp1pst‑buy‑applfv aug‑2‑child aug‑8‑book
‘He bought the children books.’
b. y‑à‑gúl‑ìr‑à bá‑ànà è‑bí‑tábó
sp1pst‑buy‑applfv 2‑child aug‑8‑book
‘He bought the children books.’

62Van de Velde (2019: 251) further points out that “on the level of the noun phrase, head nouns often lose their augment in the presence of (certain types of) preposed modifiers. The augment can then either function as a phrasal affix and appear on the prenominal modifier (39) or be left out (40)”.

(39) Nande JD42 (ibid.)
a. ɔ̀‑mʊ̀‑tɩ́ ‘the tree’
b. ɔ̀‑ɣʊ̀‑ntɩ̀ mʊ̀‑tɩ́ ‘the other tree’
(40) Zulu S42 (ibid.)
a. i‑ndawo leyo ‘that place’
b. le ndawo ‘this place’

63As seen in Babanki, numeral and interrogative modifiers are incompatible with the enclitic, just as in some Northwestern Bantu languages such as Zamba C322, where the interrogative ‘which’ (41) and numerals (42) are incompatible with the augment (Grégoire & Janssens 1999: 415).

(41) Zamba C322 (ibid.)
a. ízíkèì ‘an egg’
b. zíkèì zídò ‘which egg?’
(42) Zamba C322 (ibid.)
a. bábâ:ná ‘children’
b. bâ:ná díbòà ‘nine children’
c. ínswě: ‘fishes’
d. ńswě: nsámbó ‘seven fishes’

64A detailed discussion of the conditions on the presence vs. absence of the augment in Bantu is given in Van de Velde (2019: 249‑254).

65Regarding the role of headedness in Babanki, it has been shown that the enclitic is obligatorily absent on derived possessives if the NP is headed but obligatorily present if the NP is headless, making it the only NP-internal syntactic condition where the behaviour of headed NPs deviates with respect to the presence or absence of the enclitic. Since an NP needs to be headed by a full noun (keeping aside the quantifier ‘whole, all’ and numerals which do not require the enclitic even if the lexical head is missing), the enclitic is present to make a well-formed NP when the nominal head is absent but is not needed if the head noun is present.

66Concerning focus, it appears that the pragmatic influence on the enclitic has almost given way entirely to syntactic factors with the only indicator seen in the obligatory absence of the enclitic under focus when demonstratives and pronominal possessives move to the position immediately before the noun, as opposed to when it is optionally present on demonstratives and genuine possessives in non-focus conditions. In contrast to West Ring languages, the Babanki enclitic is overwhelmingly driven by syntactic conditions, while pragmatic conditions such as focalization only marginally interfere when the enclitic is required to be absent. This is seen with demonstrative and possessive modifiers under focus conditions, specifically when preposed to their nominal heads.

Abbreviations

downstep
ǀ intonational break or pause
1sg first person singular
2sg second person singular
3sg third person singular
1pl first person plural
2pl second person plural
3pl third person plural
1…19 noun classes
adj adjective
am associative marker
appl applicative
aug augment
com comitative
cop copula
csc consecutive
decl declarative
dem demonstrative
der derived
ds dummy subject
emph emphatic
enc enclitic
excl exclusive
f3 distant future tense
foc focus
fv final vowel
gen genuine
imp imperative
int interrogative
mod modifier
N noun
n nasal
neg negative
num numeral
p0 present/perfect tense
p1 immediate past tense
part participle
poss possessive
pro pronoun
pst past
q question
quan quantifier
qt quotative
rel relativizer/relative clause
rep repetitive
sm subject marker
sp subject prefix
wh-q wh-question
Haut de page

Bibliographie

Akumbu, Pius W. 2011. Tone in Kejom (Babanki) associative constructions. Journal of West African Languages 38. 1‑20.

Akumbu, Pius W. 2016. Babanki coda consonant deletion and vowel raising: A case of allomorphy. In Emily Clem, Virginia Dawson, Alice Shen, Amalia Horan Skilton, Geoff Bacon, Andrew Cheng & Erik Hans Maier (eds.), Proceedings of the Forty-second annual meeting of the Berkeley Linguistics Society, February 5‑7, 2016, 3‑20. Berkeley, CA: Berkeley Linguistics Society (accessed 5 July, 2022).

Akumbu, Pius W. 2019. A featural analysis of mid and downstepped high tone in Babanki. In Emily Clem, Peter Jenks & Hannah Sande (eds.), Theory and description in African linguistics: Selected papers from the 47th Annual Conference on African Linguistics, 3‑20. Berlin: Language Science Press (accessed 5 July, 2022).

Akumbu, Pius W. & Evelyn Fogwe Chibaka. 2012. A pedagogic grammar of Babanki. Cologne: Rüdiger Köppe.

Chie, Esther P. 2014. The phrasal phonology of Gáʔá Kə̀dʒɔ̀m (Babanki): An outgrowth of the other components of the grammar of the language. Yaounde: University of Yaounde I (PhD dissertation).

De Blois, F.K. 1970. The augment in the Bantu languages. Africana Linguistica 4. 85‑165.

Grégoire, Claire & Baudoin Janssens. 1999. L’augment en bantou du nord-ouest. In Jean-Marie Hombert & Larry M. Hyman (eds.), Bantu historical linguistics: Theoretical and empirical perspectives, 413‑429. Stanford, CA: CSLI.

Hyman, Larry M. (ed.). 1979a. Aghem grammatical structure. With special reference to noun classes, tense-aspect and focus marking (Southern California Occasional Papers in Linguistics 7). Los Angeles, CA: Department of Linguistics, University of Southern California (accessed 5 July, 2022).

Hyman, Larry M. 1979b. Tonology of the Babanki noun. Studies in African Linguistics 10(2). 159‑178.

Hyman, Larry M. 1980. Babanki and the Ring group. In Larry M. Hyman & Jan Voorhoeve (eds.), Noun classes in Grassfields Bantu, 225‑258. Paris: SELAF.

Hyman, Larry M. 2010. Focus marking in Aghem: Syntax or semantics? In Ines Fiedler & Anne Schwarz (eds.), The expression of information structure: A documentation of its diversity across Africa (Typological Studies in Language 91), 95‑116. Amsterdam: John Benjamins (accessed 5 July, 2022).

Hyman, Larry M. 2018. Third person pronouns in Grassfields Bantu. In John R. Watters (ed.), Eastern Benue-Congo: Nouns, pronouns and verbs, 199‑221. Berlin: Language Science Press (accessed 5 July, 2022).

Hyman, Larry M. & Francis X. Katamba. 1993. The Augment in Luganda: Syntax or Pragmatics? In Sam A. Mchombo (ed.), Theoretical Aspects of Bantu Grammar, 209‑256. Stanford, CA: CSLI.

Kießling, Roland. 2010. Focalisation and defocalisation in Isu. In Ines Fiedler & Anne Schwarz (eds.), The expression of information structure: A documentation of its diversity across Africa (Typological Studies in Language 91), 145‑164. Amsterdam: John Benjamins (accessed 5 July, 2022).

McGarrity, Laura W. & Robert Botne. 2001. Between agreement and case marking in Lamnso. In Robert Botne & Rose Vandrasek (eds.), IUWPL 3: Explorations in African linguistics: From Lamnso’ to Sesotho, 53‑70. Bloomington, IN: IULC Publications (accessed 5 July, 2022).

Meeussen, A.E. 1967. Bantu grammatical reconstructions. Africana Linguistica 3. 79‑121.

Mutaka, Ngessimo M. & Esther Phubon Chie. 2006. Vowel raising in Babanki. Journal of West African Languages 33(1). 71‑88.

Van de Velde, Mark. 2019. Nominal morphology and syntax. In Mark Van de Velde, Koen Bostoen, Derek Nurse & Gérard Philippson (eds.), The Bantu languages, 237‑269. 2nd ed. London: Routledge.

Watters, John R. 2003. Grassfields Bantu. In Derek Nurse & Gérard Philippson (eds.), The Bantu languages, 225‑256. 1st ed. London: Routledge.

Wolf, Paul de. 1971. The noun-class system of Proto-Benue-Congo. The Hague: Mouton (accessed 5 July, 2022).

Haut de page

Notes

1 Standard Bantu numbering has been used since the 12 noun classes that exist in Babanki correspond to those reconstructed for Proto-Bantu (Meeussen 1967).

2 A merger of classes 3 and 5 in Babanki can be seen in their identical schwa prefixes as well as various agreement markers. While it is not completely transparent, the pairing with plural classes provides a basis to distinguish between the two singular classes, i.e. a 3~5 sg which pairs with a 6 pl is categorized as class 5, whereas a 3~5 sg which pairs with a 13 pl is categorized as class 3.

3 The numbering of noun classes is not continuous, as 4, 11, 12, 14‑18 are missing because no Babanki nouns are assigned to these classes.

4 It is puzzling that ‑m is retained after n in class 6a. All consultants we checked with did this consistently.

5 Most of the data for this study has been provided by the first author and checked with five other native speakers of Babanki Tungo (Kejom Ketinguh) during WhatsApp conversation sessions in March 2021. The four WhatsApp sessions dedicated for this purpose lasted 3 hours. Many thanks to Vivian Ba-ah, Regina Phubong, Cornelius Wuchu, Stanley Amuh, and Benjamin Nkwenti for sharing their knowledge of Babanki with us. Some examples have been taken from the existing literature as well as from texts collected by the first author. Babanki Tungo differs from Big Babanki (Kejom Keku) in only minor ways. There are a few lexical differences (examples to the left are from Big Babanki while those to the right are from Babanki Tungo): fə̀mváŋ vs. túbù ‘jigger’; kə̀bwín vs. dʒì ‘road’; wúŋ vs. fə̀làm ‘fishing net’. The following sound differences have also been identified: kə̀bònə̀ vs. kə̀bòlə̀ ‘calabash’; kə̀bwə́’ vs. kə̀byə́’ ‘half’; wàn vs. wàyn ‘child’; kə̀nʃɨ̀f vs. kə̀nfyɨ̀f ‘blindness’. Babanki has 26 consonant phonemes /b, t, d, k, g, ʔ, m, n, ɲ, ŋ, f, v, s, z, ʃ, ʒ, ɣ, pf, bv, ts, dz, tʃ, dʒ, l, w, y/, 8 vowel phonemes /i, ɨ, ʉ, u, e, o, ə, a/, and two tonemes /L, H/. Babanki is an SVO language, where the direct object, whether nominal or pronominal, follows the verb. There is a subject-verb agreement marker (sm) which follows the nominal subject and differs from subject pronouns. The sm is most visible in classes which have a Cə shape since the schwa that marks other classes may completely fuse with a neighboring vowel.

6 When followed by a H tone, the L tone of the demonstrative root is raised to a M tone, as elaborated in Hyman (1979b) and Akumbu (2019). This general process is also displayed by possessives (§2.2) and the numeral ‘three’ (§2.6) among others.

7 This suggests that the derived possessives originate from full-fledged nouns, retaining some nominal properties on their way to becoming grammaticalized as pronouns, as demonstrated by Hyman (1979a and 2018) for Aghem. While it is evident that 3sg and 3pl originate from ə̀wέn ‘body’, the source of 1pl.incl is not equally obvious.

8 The lowering of the H tone of 1sg, 1pl.excl, and 2pl in classes 1 and 9 reveals that these two classes come with an additional suffix, i.e. a floating L tone class marker in genuine possessives. This might be the same for the proximal and distal demonstratives, where the L suffix of classes 1 and 9 merges with the demonstrative roots.

9 Class 6a has an additional homorganic nasal following the ordinary agreement prefix. This additional nasal is also seen on numerals (§2.6) and quantifiers (§2.7). Its origin and function remain unclear, requiring further investigation beyond the scope of this study.

10 The raising of the L tone of kɔ̀ɔ to M in classes 1 and 9 is due to the presence of a floating H between the noun and the interrogative, which merges with the interrogative L to form a HL and subsequently simplifies to M. This points to a reflex of some H tone concordial element which triggers L raising in a preceding L environment, but which otherwise merges with the preceding root H tone. The details of this tonal behaviour as well as how it is linked to the deviant H tones in the concordial prefixes of classes 1, 6a, and 9 with adjectives need further investigation.

11 Classes 2, 3, 5, 6, 6a, and 8 have an additional H schwa, seemingly an epenthetic vowel between two voiced consonants, and which prevents the deletion of the enclitic consonant of the sort seen in demonstratives and genuine possessives.

12 Bracketing indicates the deletion of the initial consonants of enclitics in classes 2, 3, 5, 6, 6a, and 8, as triggered by immediately preceding voiced consonants (Hyman 1980: 229).

13 As suggested by one of the reviewers, Proto-Bantu agreement prefixes of classes 3 and 6 have been reconstructed with an initial *g consonant, of which ɣ is a reflex in Babanki. The fact that ɣ is also observed in the enclitic of class 5 shows that, as mentioned in fn. 2, classes 3 and 5 have actually merged in Babanki due to the phonological merger of their noun class prefixes, and only the enclitic of class 3 based on Proto-Bantu *gʊ́‑ was maintained. If they had not merged, one would have expected here an enclitic based on the Proto-Bantu concord of class 5, i.e. *dɩ́‑.

14 Outside the Babanki NP, agreement is seen on subject markers between the subject noun and the verb, as in other Ring languages (McGarrity & Botne 2001: 60). The subject marker always follows the preverbal subject noun in a main clause and precedes the following verb or tense/aspect marker (Hyman 1980: 237).

15 The floating enclitic L tone of classes 1 and 9 docks leftwards to form a HL contour tone, which subsequently simplifies to M.

16 When found between two H tones, the L tone on the agreement prefix is realized M following the general Prefix L-Raising rule in Babanki (Akumbu 2019: 6f.).

17 The plural form byí‑ꜜsə́ ‘goats’ is realized as byí without its plural class 10 suffix. Unlike the other noun classes which employ prefixes, class 10 nouns appear in isolation with a plural suffix. When followed by a -initial modifier a situation arises where the ‑plural suffix and the sə‑ concord prefix are contiguous. Since doubling of is not permitted a general rule deletes “the suffix of byí‑ꜜsə́ and other class 10 nouns whenever followed by a sə‑ concord” (Hyman 1980: 228). This type of noun suffix deletion to prevent doubling is also found in Lamnso’ (McGarrity & Botne 2001: 57). We argue in §3.1 that this Babanki class 10 “suffix” is actually the enclitic.

18 Minus in this slot indicates the obligatory absence of the enclitic for all nouns except those of class 10, which behave differently in that they retain the enclitic in their unmodified form under all conditions, as discussed in §3.1.

19 The obligatory presence of the enclitic after adjectival modifiers is also reported in Aghem (Hyman 1979a: 32, 67).

20 The reason for dropping the erstwhile prefix of class 10 as well as that of the singular class 9 is not obvious. Classes 9/10 are reconstructed with *ì‑/í‑ prefixes, respectively, in Proto-Grassfields (Hyman 1980: 253). Both classes lost their prefixes, and the singular/plural distinction was re-established by class 10 taking an extra overt marker, i.e. the enclitic.

21 Unlike in Aghem, where the enclitic and the demonstrative are mutually exclusive, both can co-occur, suggesting that the Babanki enclitic does not originate from the demonstrative.

22 It can be argued that the absence of the enclitic with the interrogative modifiers and numerals is due to the fact that they are inherently focalized, as discussed in §3.6.

23 The free order of a possessive pronoun and an adjective in a complex NP is also reported in Aghem (Hyman 2010: 105).

24 The class 5 enclitic =ɣə́ is formally different from the sm ə́. In addition, the intonational break after =ɣə́ confirms that it is enclitic to the NP and not to the verb.

25 For the enclitic, the brackets represent the status of optional presence, while for the sm, the brackets indicate that it must occur in the absence of the enclitic. Since the two are mutually exclusive, the presence of the enclitic blocks the sm.

26 “n.a.” (not applicable) indicates that the modifier does not occur in that position.

Haut de page

Pour citer cet article

Référence électronique

Pius W. Akumbu et Roland Kießling, « The role of the Babanki noun phrase-final enclitic class marker »Linguistique et langues africaines [En ligne], 8(1) | 2022, mis en ligne le 30 juin 2022, consulté le 17 mai 2026. URL : http://journals.openedition.org/lla/2202 ; DOI : https://doi.org/10.4000/lla.2202

Haut de page

Droits d’auteur

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC BY-NC-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search