CC BY-NC-ND
20
Kirill Babaev
Russian State University for the Humanities
Reconstructing Benue-Congo person marking II
This paper is the second and last part of a comparative analysis of person marking systems in Benue-Congo (BC) languages, started in (Babaev 2008, available online for reference). The first part of the paper containing sections 1-2 gave an overview of the linguistic studies on the issue to date and presented a tentative reconstruction of person marking in the Proto-Bantoid language. In the second part of the paper, this work is continued by collecting data from all the other branches of BC and making the first step towards a reconstruction of the Proto-BC system of person marking.
Keywords: Niger-Congo, Benue-Congo, personal pronouns, comparative research, reconstruction, person marking.
The comparative outlook of person marking systems in the language families lying to the west of the Bantoid-speaking area is a challenge. These language stocks (the East BC families of Cross River, Plateau, Kainji and Jukunoid, and the West BC including Edoid, Nupoid, Defoid, Idomoid, Igboid and a few genetically isolated languages of Nigeria) are still far from being sufficiently studied or even described, and the amount of linguistic data for many of them remains quite scarce. In comparison with the Bantu family which has enjoyed much attention from comparative linguists within the last decades, there are very few papers researching the other subfamilies of BC from a comparative standpoint. This is especially true for studies in morphology, including person marking.
The aim here is therefore to make the very first step towards the comparative analysis and reconstruction of person markers in BC. This preliminary step will include compiling the data on as many languages as possible, systematising it into a single picture and making tentative conclusions about the general principles of person marking in the proto-language. By now, this seems the most ambitious objective we can achieve, so that the future, more complex work in this domain could have a basic foundation to move forward.
An important note here is that we had to omit the 3rd person markers in the present paper. First, in BC, they are mostly a part of the noun class system and therefore do not usually function as parts of the pronominal paradigms. Moreover, adding the 3rd person would have made the paper too huge. However, we will attract the 3rd person data where necessary in our analysis.
3. Cross River
The Cross River (CR) family, consisting of some 68 languages1, covers an area in southeastern Nigeria (Korop and Usakade are also spoken in the border districts of Cameroon), resembling a long arch stretching from SW to NE, from the Niger delta region to the upper
1 Figures hereinafter according to Lewis 2009. Journal of Language Relationship • Вопросы языкового родства • 4 (2010) • Pp. 1-45 • © Babaev K., 2010
Cross river valley. Its southernmost tongues border the Ijoid languages, while central and northern dialects are squeezed between the Igboid area to the east and the Bantoid area to the west.
Williamson (1988: 104) gave her indication on the homeland and early migrations of the CR speakers from the BC home area, which she located at the confluence of the Niger and the Benue. According to her, the break-up of Bendi and Delta Cross subbranches of CR took place some 5,000 years ago, when the speakers of the latter branch moved to the south from the hills of Ogoja. On their way, CR speaking people settled along the banks of the river and around, and the migratory movement only stopped approximately 1,500 years ago in Central Delta, where the CR newcomers pressed the Ijo people.
Since Greenberg (1963) classified CR as a branch of BC, the internal structuring of the family has not been changed much, and the most recent family tree scheme (Williamson & Blench 2000: 33) looks generally as follows:
Diagram 1
Still, the family cannot be called fully homogeneous, and the discussions are still under way whether the Bendi branch belongs to CR or is a closer relative with Bantoid (Connell 1988). This is the main issue about the CR classification, since the languages of the second major branch, Delta Cross, seem to share a number of common items in both lexicon and grammar, though, again, very few lexical innovations and almost no phonetic innovations have been described even for Proto-Delta Cross.
The reconstructed consonant system of Proto-CR is presented following Dimmendaal (1978) and contains two basic sets of voiced / voiceless obstruents plus a special series of "for-tis" obstruents, sometimes seen as geminates (Faraclas 1989: 386):
PP P b m
tt t d n
kk k
kp
(w)?
W?
Comparative works on CR include (Faraclas 1989) and (Connell 1988). In subbranch reconstructions, Dimmendaal (1978) and Sterk (1979) should be named for Upper Cross, Wolff (1969) and Alex (1989) for Central Delta, Williamson (1985) and Ikoro (1989) for Ogoni, Connell (1991, 1995) for Lower Cross. However, these works mainly deal with lexical and phonetic reconstruction, only briefly touching the morphology. The reconstruction of some elements of person marking are found in Bond & Anderson (2006) for Ogoni and Connell (ms.) for Lower Cross. Person markers for Proto-CR have not been subject to specific comparative research so far.
In most CR languages, there are four distinct series of markers denoting subject (predicative), object (predicative), possessive (nominal) and emphatic, also called "stressed" (independent pronouns). The subject markers are either prefixes or clitics — in the latter case, usually identical with the independent pronouns. Moreover, verb prefixes are often emphasised
by independent pronouns in a normal verb phrase. Object markers can be both independent and suffixed, and possessive markers are predominantly suffixed either to the noun or the class determiner in concord with the noun.
1st person singular (Table 3.1)
Some forms in 3.1 and subsequent charts have variable tone, depending on the tense or aspect of the predicate. For lack of space, we will unfortunately have to omit tonal distinctions in most cases if the segmental form does not change. Classificatory subgroup names are given in brackets following the name of the language.
The subject prefixed marker can be reconstructed as *N-, a homorganic syllabic nasal. It might be suggested that it appeared due to the reduction of an earlier *mV-: we see the same process in the system of noun class markers, where the Proto-BC pluralizer *ma- > *N- all over CR. When followed by a vowel, the initial consonant of the prefix is always m-. The vowel in unmarked forms is most often -i-. The *ma- form (a long vowel is possible) is a contraction with a perfective marker. The subject prefix is lengthened in the negative in Delta Cross languages (cf. Iwara 2002: 106-107).
The independent pronoun in the majority of languages is derived from *ami, which confirms the reconstruction by Connell (ms.) for Lower Cross. A similar form is used as a possessive suffix, often lacking the final vowel: *-am. In Kohumono a-n-vm 'I, me', a class determiner -n-is infixed between the association particle a- and the person marker. In Delta Cross, we see *mi-nV which is also found in some Northwest Bantu independent and object series (Basaa A men, Viya B meni, etc.) as well as in other Bantoid languages (Tikar mun, Yemba mey) (Babaev 2008: 153-154).
The object marker for Proto-CR could be *mi or *-m(i). Bond & Anderson (2006) reconstruct Proto-Ogoni *ml, with a variable front vowel. Therefore, the tentative reconstructed forms are:
*N-, *mi subject
*ma- subject (perfective)
*m(i) object
*ami independent, possessive
*minV independent (Proto-Delta Cross)
2nd person singular (Table 3.2)
The subject marker was certainly a vocalic one, though we cannot be sure whether *o or *a was the original vowel. The negative marker -a- lengthens the prefix in Ibibio and Legbo, the same as in the 1sg. The remaining three sets — object, possessive and independent — are mostly based on the proto-language root *wo / *wu / *we.
The object marker can be reconstructed as *-o / *wo / *we, the independent (and most probably possessive) as *owo which is *ofo in Proto-Lower Cross. In Kana, one of the two possessive pronouns is cognate to the emphatic pronoun, the other one (denoting body parts only) is the normal object pronoun (Ikoro 1996), and it seems that Tee shows the same situation. That means the absence of a specific possessive set in Proto-Ogoni.
There are signs of the Proto-CR suffix *-n(V) attached to non-subject markers (Bekwarra iwo-n 'your', Kukele we-ni 'you', Ibibio fie-n 'you'). It is also in use in other person / number forms, and may serve as an indirect case marker.
Table 3.1
subject object possessive independent / emphatic
Bekwarra (Bendi)2 N-, maa- (perf.) m (after CV verb), mi iyi ami
Boki (Bendi)3 me me
Kohumono (Delta, U)4 N- anvm -um / -vm anvm
Kukele (Delta, mê me -amê meni
U)5
Lokaa (Delta, U)6 n-, nn- (neg.) N, min -mi ami
Mbembe (Delta, U)7 N- -m -aam kaam
Ibibio (Delta, L)8 N- min, -eq- mmi / nmi ami
Efik (Delta, L)9 N- mi (o)mi ami
Obolo (Delta, L)10 mi-, mù muy mà emi
Legbo (Delta, L)11 N-, NN- (neg.) -m
Proto-Lower Cross12 *ami
Ogbia (Delta, C)13 mi / mi maami -ami / -ami ami
Abua (Delta, C)14 mi i-mi a-mi mi, mina
Eleme (Delta, N-, rt- (neg.) -mi -àmi, na ami
Ogoni)15 ma- (perf.)
m- / m-,
Kana (Delta, Ogoni)16 mm (neg.), maa (progr.), mâà- (perf.) me, mi nà, n-da (emph.), me (body parts) mm, n-da (emph.)
Tee (Delta, Ogoni)17 m me nà, ndâa daa
2 Stanford 1976 [1973]
3 Westermann & Bryan 1952: 120-121
4 Cook 1976 [1972]
5 Preston 1854: 12
6 Iwara 1990
7 Barnwell 1980, Oyama & Barnwell 1985
8 Cook 1976 [1968], Essien 1987
9 Westermann & Bryan 1952: 135-136, Cook 1976 [1968]
10 Faraclas 1984, Rowland Oke 2003
11 Hyman et al. 2002, Hyman & Udoh 2002
12 Connell [ms.]
13 Isukul [ms.]
14 Gardner 1976 [1966/67]
15 Bond 2003, 2006
16 Westermann & Bryan 1952: 137, Ikoro 1996: 114-132, Bond 2006, Williamson et al. 2007
17 Wilson Kpakpan Nwf-Bari 2002
Table 3.2
subject object possessive independent / emphatic
Bekwarra o- wo iwön awö
Boki wo'o У0
Kohumono 3- / a- àgwo -o / -6 agwo
Kukele wë wë, (a)gwë yo weni
Lokaa a-, aa- (neg.) 6, wo -wù awu
Mbembe a- -0 -0 ko
Ibibio è- (pos.), ù- (neg.) fîèn, -й- nfö afd
Efik à- / ö- / è- f (ö)fö afo
Obolo o-, ù- ùy ùo öwü
Legbo a-, aà- (neg.) -6
Proto-Lower Cross *ofo
Ogbia (a)nwé maanwé -îyom
Abua na pir-na a-na na
Eleme o-, à?ô, rö- (neg.) -rü -y°
Kana ö- / o- a o, a-lo (emph.), a (body parts) oo / oo18, ä-lö
Tee o o, öö o
Unclear forms include, among others, Abua -na and the Eleme object suffix -rü. The Eleme negative prefix ro- has the same origins as its 1sg. counterpart, but we have no evidence of its relation to the object marker.
*o- / *a- subject
*wo / *wu / *we object
*dwd independent, possessive
*wo-n(V) / *we-n(V) object / possessive / emphatic
1st person plural (Table 3.3)
In Delta Cross, prefixed predicative person markers are gradually replaced by independent pronouns. Still, they are sometimes found in V-shape as both subject prefixes and object suffixes.
befe in Kukele is almost identical to Proto-Bantoid *(a)bece 'we' < *(a-)ba-ice (Babaev 2008: 158). Another retention is ado 'we, us' in Kohumono which should go back to *ato with an intervocal voicing, probably cognate to the Proto-Bantu subject prefix *tv- 'we'. tu / ti are common 1pl. pronouns in Kwa and Gur languages as well (cf. Бабаев 2010), but whether Bekwarra iten 'our' is a related form is obscure.
Connell (ms.) reconstructs *ajit 'you (emph.)' for Proto-Lower Cross. We would also suggest *ajin here, which may be related to Abua a-yira 'our (incl.)'. Namely, the -ar / -rV -like forms which are found in the Central Delta and Ogoni languages are probably cognate with
18 33 occurs in nominal affirmative and focalised negative constructions; 33 is used in non-focus negative phrases, both verbal and nominal Ikoro 1996: 118-119.
Table 3.3
subject object possessive independent/emphatic
Bekwarra abere e- ti íten abere
Boki beve be
Kohumono e- / e- ado ado
Kukele beje ba befe -afi
Lokaa m5n -ml, -mon am5n
Mbembe mo- / mo- mina -amina mina
Ibibio i- npin, -i- np.in npin
Efik i- npin np.in npin
Obolo e-, mi- (past), eme emey eme ézi / eji
Legbo me-, me (neg.) -mán
Proto-L.Cross *ajit
Ogbia iyar / tyar maiyar -íyér
Abua yoor (excl.), yira (incl.) i-yoor (excl.), i-yira (incl.) a-yoor (excl.), a-yira (incl.) yoor (excl.), yira (incl.)
Eleme re- / ne- -í ébai
Kana i- / i- i i, á-li (emph.) ii, a-li (emph.)
Tee i i, ii plíi iri
them as well. The intervocal consonant j / y is seen as -z- in Obolo and can be related to -f- in Kukele, an Upper Cross tongue. Phonologically, it would be reasonable to reconstruct *aci(n) for Proto-CR, which would be close to the Bantoid possessive forms: cf. Proto-Manenguba *aci or Kenyang -ese. The final Lower Cross consonant acts in a number of Lower Cross dialects which we did not include in the 3.3: Ebughu ijin, Enwang and Uda ajin (Connell, ms.).
In Upper and Lower Delta languages person markers are me / mo (subject), mV(n) (object, possessive), *(a)mi(na) (non-subject, independent). These forms have parallels in languages as far away as Gur and Mande, but are rarely found in BC. So, the tentative reconstructions for Proto-CR are these:
*e- / *i- subject
*aci(n) independent
*mV subject / object / independent
2nd person plural
In Eleme, the subject marker is suffixed to the verbal predicate. To distinguish the exact meaning of the phrase, the independent pronoun of the respective person is used together with this prefix:
dbau dose de-i nja (1)
2.PL must eat-2.PL food 'you must eat food' (Bond 2003)
In Ibibio, markers of the 2nd and 3rd persons have been equalised:
Table 3.4
singular plural
1st p- i-
2nd è- è-
3rd é- é-
Table 3.5
subject object possessive independent / emphatic
Bekwarra amin amin inèn amin
Boki beva be
Kohumono a- / 3- and ano
Kukele be ba bewe, begwe -ani begwe
Lokaa ba: -wù aba:
Mbembe ma- boqa -aboqa boya
Ibibio e- (pos.), i- (neg.) ndufd, -i- / -e- ndùfè / mbùfo ndufó / mbufo
Efik e- (pos.), i- (neg.) mbufo mbùfo mbufo
Obolo e-, mi- (past) ey ébe épi
Legbo ba-, bo (neg.) -bo
Ogbia ¡pin m3ipin -îpîn
Abua pina i-pina a-pina pina
Eleme ne-, -I, ro- (neg.) -i obau
Kana bl- i bùï, a-bàâlô (emph.), ï (body parts) bui, a-baalo (emph.)
Tee bo boo plôo boro
We mostly see a vocalic prefix as subject marker, sometimes (but not everywhere) identical to the one we saw in the 1pl. Here, it is reflected as e- or i-, but we also find a- in Kohu-mono. These vocalic markers are never found in the possessive series or as independent pronouns.
The non-subject markers with the -Vnl- shape include Kohumono and (object / emphatic), Bekwarra inen and Kukele -ani (possessive), Obolo epi (emphatic), and Ogbia / Abua -ipin / pina. Similar forms are found in Bendi, Upper Cross, Lower Cross and Central Delta languages, encompassing most of CR. Eleme (the Ogoni group) has a subject prefix ne- which may complete the picture. We can only suggest *anV / *inl(n) with a non-subject meaning based on the comparison with the other person / number forms of the paradigm.
The nasality in Boki be 'us' suggests a cognate with *be-ne we reconstructed for Proto-Bantoid (Babaev 2008: 159), with the same root and a prefixed plural class marker *ba- / *be-. The independent pronoun beva 'we' in Boki may suggest *be-nue as well. In Kana, as Ikoro (1996: 115) suggested, emphatic pronoun a-baald 'you pl.' and d-badld 'your pl.' seem to be derived from the 3pl. ba 'they'. The widespread typological feature of replacing the 2pl. pronoun by the 3pl. one may be suggested for Kukele, Tee and Legbo.
The Lower Cross *mbufo < *o-bu-fo form is a probable pluralisation of the 2sg. *ofo with the same prefix marking the independent pronoun. A similar model can be seen in Kukele bewe 'you pl.' < *be-i-we, Legbo bo vs. 2sg. -5, Tee boo vs. 2sg. oo, and Kana d-baald vs. 2sg. a-ld. Ko-
humono has merged the 2sg. and 2pl. pronouns. The process of eliminating the suppletion between the two 2nd person pronouns must have started no later than in Proto-Delta Cross, and the proto-form might have been *be-wo / *ba-wo / *be-we for object and independent forms. The Eleme negative marker ro- lost a distinction in number and marks both the 2nd person singular and plural.
The 2pl. forms for Proto-CR might be given as follows:
*e- / *i- subject
*anV(n) / *inI(n) independent
*ba-wo / *be-wo subject / non-subject (Proto-Delta Cross)
The summarising table demonstrates the resulting tentative forms for Proto-CR:
Table 3.6
prefixed (subject) object independent (non-subject)
lsg. *N-, *mi, *ma (perf.) *m(i) *àml, *minV
2sg. *o- / *a- *wo / *wu / *we, *won(V) / *wen(V) *dwd, *wonV / *wenV
lpl. *e- / *i-, *mV *mV *àci(n), *mV
2pl. *e- / *i-, *ba-wo / *be-wo *ànVn / *lnI(n), *ba-wo / *be-wo
The comparison between the selected Proto-CR forms and the Proto-Bantoid person markers (Babaev 2008: 160) is shown below.
Table 3.7
Proto-Bantoid Proto-CR
lsg. subject *pi- *N-
lsg. independent *(à)me *àmi
2sg. subject *v- *o-
2sg. independent *(à)we *dwd
lpl. non-subject *(bè-)c(u)e *àci(n)
2pl. non-subject *(bè-)n(u)e *ànVn / *inI(n)
The similarities between the two columns seem evident. The independent pronouns in Proto-Bantoid and Proto-CR are virtually identical. Still, the variability of some vowels does matter and deserves a more detailed analysis following this brief survey. In the singular, *ml / *aml with a final front vowel is common for both families. The CR form *mi-nV finds exact parallels in a number of Bantoid tongues including Bantu Western zones A-C, H and K (Babaev 2008: 162-166). The 2sg. prefix *o- is also a clear cognate, and the independent CR pronoun *ôwô directly corresponds to widespread variants owo / àwe / ewe / wo / we in Bantu and Bantoid.
In the plural, person markers are more diverse. Both the Bantoid languages and some CR languages use pronouns with the prefixed pluraliser *be- or, alternatively, with a prefixed vowel of either front or back row: *acV / *icV for 1pl. and *anV / *inV for 2pl. The original meanings of this vowel prefix are still traceable in Bantu tongues where forms in *-a- are possessives, while *-i- marks the independent pronoun (Babaev 2008: 139).
While the Bantu prefixed subject markers differ in origin from the non-subject (originally independent) pronouns, this difference can only be seen in the 2sg. in CR. As well as in Ban-
toid (beyond Bantu), CR person markers are mostly syntactically independent. Proto-Bantu subject markers *pi- 'I', *tv- 'we', and *mu- 'you pl.' (Babaev 2008: 148) only have a few cognates in CR, such as Kohumono ado 'we'.
4. Plateau, Kainji and Jukunoid (Central Nigerian)
Since Greenberg (1963) first identified the Plateau and Jukunoid languages stretching as a wide belt from W to E through the center of Nigeria, too many pages have been covered with discussions on both the external and the internal classification of this language stock. This discussion was fuelled by the clear fact that the languages which Greenberg defined with these two terms (and which we now know as Plateau, Kainji and Jukunoid) are indeed very close to each other, and definitely to the other tongues of East BC, namely CR and Bantoid. However, the lack of reliable language data and, more essentially, the lack of comparative research did not allow to make any distinct conclusions on the classification of the language groups of the Nigerian Plateau.
The Plateau and Kainji languages in North Nigeria border an area inhabited by Chadic-speaking people, which makes many minority languages endangered due to the cultural advance of Hausa. In the south, besides Chadic, Nupoid and Idomoid neighbours are located. The Jukunoid area borders Bantoid to the south.
At least six different classifications of these three language families were presented between 1963 and 2000 (Williamson 1971; Hoffmann 1976; Bennett & Sterk 1977; Gerhardt & Jockers 1981; Gerhardt 1989; Crozier & Blench 1992) resulting in the most widely referred internal grouping suggested by Williamson & Blench (2000). The new idea of the latter paper was to suggest the existence of a Central Nigerian (CN) proto-language, from which the languages of all the three groups originated. However, many issues of this subgrouping still remain unsettled. The idea of the Proto-Plateau language is challenged, since there are only a handful of lexical innovations that we can attribute to it. The Jukunoid languages indeed form a very tight group of around 20 languages (Lewis 2009), but whether they once formed a community with Tarokoid, a Plateau group, as suggested by Shimizu (1975), is a question mark. The internal structure of the Plateau family, the most diversified of all the three, must comprise at least ten groups of the same chronological level, which is not fully acceptable. One of the recent analyses of the issue by Roger Blench presented in 2005 bears a telling — and, for the moment, rhetoric — heading: "Is there a boundary between Plateau and Jukunoid"?
The most recent classification of CN, which is a compilation of (Blench 2004), looks as follows:
Diagram 2
Blench is undoubtedly the most fruitful modern collector of Plateau language data, having gathered material from a great number of languages of the area and identified at least a dozen new ones. He has presented plenty of field notes with lexical data, which also include some morphological analysis. His online databases include initial steps in reconstructing Proto-Kainji and Proto-Plateau, which would have otherwise remained without any attempts at reconstruction. Beyond this, there has been no attempt to build a reliable system of proto-language phonology or morphology for either of the two. The most recent comparative survey of the Plateau and Kainji groups is contained in (Gerhardt 1989); however, it only briefly describes the morphology, mainly the nominal class systems, to which Gerhardt (1983) dedicated a special paper earlier. Williamson & Blench (2000: 32) give a short summary of morphological features characterising CN: "full or reduced noun class systems, mostly with prefixes; widespread verbal extensions; object pronouns following transitive verbs", etc. The reference material for comparative lexical data is contained in "Benue-Congo Comparative Wordlists" (Williamson & Shimizu 1968; Williamson 1973).
For Jukunoid, a large study was conducted by Shimizu (1971-80), who has reconstructed the core of Proto-Jukunoid and has also dealt with personal pronouns. It clearly demonstrates that the Jukunoid morphology was more innovative than that of both Plateau and Kainji: this is seen in the noun class system, of which Jukunoid languages have preserved only -V class suffixes where Kainji and Platoid have CV-/V- prefixes corresponding to the Bantu class markers, which are considered more archaic (Gerhardt 1989: 372). However, the pronominal paradigms of some Jukunoid tongues were not yet recorded by that time, and the decades following Shimizu's fundamental work filled some important gaps.
De Wolf (1968) suggested that the Proto-Kainji noun class system also showed signs of assimilating prefixed consonants. As we will see below, the situation in person marking is nearly the same.
In the present section, we will perform a brief comparison of person markers of all the three CN families, thus trying to draw a route towards the Proto-CN reconstruction, and further comparing the results with the rest of previously reconstructed East Bantoid systems. We must admit, however, that this analysis is still far from being complete.
Subject markers in CN languages precede the predicate. An object marker is usually suffixed to the verb, though in some Central Plateau languages it is placed between the tense/aspect element and the verb root, just like in Bantu. The difference between subject and object markers is often limited to suprasegmental features.
Possessive pronouns are mostly suffixed to the noun, but can also act independently following the noun (often called "possessive adjectives"), in which case they are sometimes grammaticalised compounds of a substantive stem with a possessive suffix, as in Berom or Tarok (Longtau 2008). Possessive pronouns can be linked with the nouns by an associative particle: Wapan atan bu u 'your house', lit. 'house of you'. And, similar to Bantoid, there is a general trend to bind the particle with the possessive: Jukun Takum bu 'your' < bu-u 'of-you'.
A phenomenon of "recapitulative", "copy" or "repeating" pronouns is witnessed in a number of languages of the Plateau. This means that the subject pronoun placed before the predicate is also copied following it, being either suffixed or independent. This construction can have different meanings. Copy pronouns can follow intransitive verbs and mark negation, as in Migili (Lijili, S.Plateau) or Kente (Jukunoid). They are marking the plural in Izere (C.Plateau):
pin rus pin agabu (2)
2.PL beat 2.PL dog
'you are beating the dog'
and the unexpected ("unanticipated") action, including the negative action, in Wapan (Jukunoid):
Be ci zhenzhen ku (3)
'they are royalty' (neutral)
vs.
Be ci be zhenzhen ku (4)
'they are royalty' (though you did not expect them to be).
In Kuteb (Jukunoid) it marks the completive aspect and the negative, and Koops (2007) calls this construction a "possessive verb", since the pospositive pronoun has the same form in Kuteb as the possessive one. Postpredicative repeating pronouns are an areal feature found in a number of languages of Central Sudan, including Adamaua-Ubangian (Gbaya, Sangi) and Chadic. Whether it originates from NC or beyond, is not clear (Gerhardt 1989: 374).
The number of series of person markers varies in CN languages. In Jukun, at least seven sets of personal pronouns can be distinguished, varying in tone, vowel length and prefixed particles, as shown below (Storch, field notes):
Table 4.1
subject emphatic subject object possessive emphatic possessive recapitulative possessive recapitulative
lsg. m am m m / bum abüm m. m m
2sg. u au u bü abu bü. bü ü
3sg. kü aku kü ba aku ba. ba a
lpl. ï ai / abi ï bi abi bi. bi i
2pl. ni ani nï bini abini bïni. .bini ni
3pl. bi abi bï bibi abibi bibi. .bibi bi
Note that tense / aspect / modality markers can also vary in tone, so the perfective paradigm will be suprasegmentally different from that of the progressive, which further increases the number of series. In Tables 4.2-4.6, we combined all subsets of person markers of selected CN languages into four columns: subject, object, possessive and independent (emphatic), marking specific semantics where necessary.
1st person singular (Table 4.2)
The subject pronoun for Proto-Plateau and and Proto-Kainji may be schematically reconstructed as *ml, with a front vowel. There are a number of languages demonstrating ma- / ma or mu, but the proto-language front vowel may be supported by Tesu, Kwanka, Toro mye / mya: these probably added the aspect marker -a- to the earlier *mi-. In the north, forms derived from *mi-n- demonstrate an extension similar to what we saw in CR in Section 3. The Proto-Jukunoid form is *mi / *me (Shimizu 1971-80). Hone preserved the prefixed subject marker n-, comparable to CR and Bantoid N-.
The object pronoun for Proto-CN can be reconstructed as *me / *me. Relic forms in *-n / *nl are also noted, marking the indirect object in Eggon, and the direct object in the present tense in Izere. In Fyam, *n- is found as the independent pronoun. Independent pronouns seem quite uniform and allow to reconstruct *me / *ame identical to the forms in CR.
Table 4.2
subject object possessive independent / emphatic
Plateau
Tesu (Alumic)19 mya
Ayu20 me
Berom (Beromic)21 ma-, me me, mé / mé hoy, -nay mé
Cara (C, NC)22 mini
Izere (C, SC)23 ml (pres.), ti (past) ni (pres.), Hin (fut.) -naan
Irigwe (C, SC)24 nje niy, N'- njé njé
Fyam (SE)25 naa, in -uy, -iy -nay, nay (emph) mé
Horom (SE)26 me
Hyam (Hyamic)27 mi
Kulu (N)28 amin, amii mi mii
Kanufi (Ninzic)29 inbo
Bu (Ninzic)30 ame
Mada (Ninzic)31 ygs yg3 yg3
Eggon (W, SW)32 me me (dir.), nè (indir.) me
Che (W, SW)33 ml
Kwanka (W, SW)34 mye
Ake (W, SW)35 àmù
Idun (W, NW)36 mi (fut., progr.), mum
Yeskwa (W, NW)37 ema / ma, ma (past), maa (fut.), maa (progr.), mma (cond.) ma -ma / ma gema
19 Blench & Kato 2007
20 Blench 2006e
21 Bouquiaux 1970
22 Blench 2006h
23 Blench & Kaze 2006
24 Blench & Gya 2008
25 Nettle 1998
26 Blench 2006g
27 Blench 2006b
28 Shimizu 1996
29 Blench 2006d
30 Blench 2006f
31 Blench 2006c
32 Sibomana 1985: 54-61
33 Blench et al. 2006
34 Blench 2007b
35 Blench 2007a
36 Blench 2008a
37 Blench 2008b
subject object possessive independent / emphatic
Tarok (Tarokoid)38 N, mi (fut.) mi mi, -jimi /-nyimi mmami
Sur (Tarokoid)39 mi
Toro40 mya
Lijili (S)41 n me lme
Kainji
Hun-Saare (W, Duka)42 em, me (neg., nominal predicate, fut.) me re, de me
C'Lela (W, Duka)43 ma, mi me, mi ri
Ror (W, Duka)44 am / me
Laru (W, Kainji Lake)45 ma
Basa (W, Basa)46 n (past), ma (pres.) ma
Pongu (W, Kamuku)47 gem
Tsuvadi (W, Kambari)48 omu
Cicipu (W, Kambari)49 mu, m- -n vo amu
Jukunoid
Jukun Takum (C, Jukun)50 m, m m m am
Jibu (C, Jukun)51 m, ml
Hone (C, Jukun)52 n-, in- (subj.), mll -l (dir.), -£ (dir.neg.), yam / yimll (indir.) -mii, -um amll
Wapan (C, Kororofa)53 m, n (before m, n, d) m, ml (indir.) -m aml
38 Sibomana 1981/82, Longtau & Blench, forthcoming, Longtau 2008
39 Blench 2006i
40 Blench 2006j
41 Stofberg 1978: 331
42 Cressmann & Skitch 1980 [1974]
43 Rikoto et al. 2002
44 Blench, manuscript a
45 Blench, manuscript b
46 Blench & al. 1991
47 Blench, manuscript a
48 Blench 2007c
49 McGill [online]
50 Welmers 1949, Koops 2007: 258
51 Priest [ms.]
52 Storch 2005
53 Evenhouse [ms.]
subject object possessive independent / emphatic
Etkywan (C, Kpan-Icen)54 âme
Bete55 mba
Yukuben (Yukuben-Kuteb) 56 m âmà
Kuteb (Yukuben-Kuteb)57 m m, me -'m, anám ame
Lufu58 m mma
Possessive suffixes in Plateau and Kainji often use the subject marker *ml. The Proto-Plateau form *-nay is derived from *-nam, found also in Jukunoid (Kuteb) -anám, comparable with Kohumono (CR) ànvm 'I, me' (Section 3 above). The Kuteb construction, according to Koops (2007: 107), is a merger of the nominalizer -a-, the possessive particle -na- and the person marker. A similar construction underlies Berom hoy. The Jukunoid languages have *-m as the possessive suffix.
Proto-CN markers can be suggested as such:
*ml subject
*me / *me object
*me / *ame independent / emphatic
For Proto-Jukunoid we may reconstruct *m (subject), *m(e) (object), *-m (possessive suffix) and *ame (emphatic pronoun).
2nd person singular (Table 4.3)
Subject markers of the 2sg. may be preliminarily suggested as *o and *wo in Proto-Plateau. The non-labialised form is seen in Berom, Fyam and Idun, sometimes with a "personal theme prefix" such as Berom h- (Bouquiaux 1970). The Jukunoid data and some Kainji tongues (Hun-Saare) support the hypothesis that the subject form in Proto-CN was *o / *u, while the object / possessive marker was *wo / *wu, and the independent emphatic pronoun is seen as *awo / *awu. This distribution finds support in all the three families, but too much diffusion still blurs the picture. -m- in Yukuben is a non-personal affix for all the singular paradigm: amà 'I', àmû 'you', ami 'he'.
Various nominal formants are used for creating the possessive markers: some of these are Berom m-, Tarok -ji- / -nyi-, Kuteb aná-, C'Lela ro-, etc.
Lijili (Plateau) and Cicipu (Kainji) emphatic pronouns iwo / ivo correspond to the independent pronouns of Bantoid (Babaev 2008: 144-145). In general, the paradigms of these two languages are strikingly similar to that of the Bantu 2sg. markers, which is noteworthy for their classification.
54 Koops 2007: 258
55 Koops 2007: 258
56 Rennison, p.c.
57 Koops 2007
58 Koops 2007: 258
Table 4.3
subject object possessive independent / emphatic
Plateau
Tesu non
Ayu puk
Berom ho-, o ho -mo, mo hó, ó
Cara wunmi
Izere wan (pres.), ta (past) ká (pres.), tàa (fut.) -fwan
Irigwe ywe ré ywê ywé
Fyam ti, wu -o -mé, náme (emph.) wéé
Horom awo awo
Hyam yu
Kulu aygu, un, yu y y
Kanufi wo
Bu awu
Mada wa W3 wa
Eggon ywo ywo (dir), dywo (indir) -yo, byo / bmo
Che yu i-min
Kwanka wo
Ake àyo
Idun ywu (fut., progress.), yo
Yeskwa emu, mu (past), muu (fut.), muu (progr.), mmu (cond.) mu -mu / mu gemu
Tarok 6u, u / upan (indirect) M dû, -jibû / -nyibu mmàbû
Sur 6u
Toro ano
Lijili o wo iwo
Kainji
Hun-Saare o, wo wo ru, du wo
C'Lela vo vo róvo
Ror bo
Pongu gà
Tsuvadi 'avó
Cicipu vù, v- àvù ivo
subject object possessive independent / emphatic
Jukunoid
Jukun Takum u u u au
Jibu wù
Hone 0-, vùù -ù (dir.), -6 (dir.neg.), yûù (indir.) -wuu avùù
Wapan ù ù -û aù
Etkywan abö
Bete owu
Yukuben mu àmu / omü
Kuteb u fu -'fu, anafu afu, abi (dial.)
Lufu u au
Proto-CN markers can be reconstructed as follows:
*o / *u subject
*wo / *wu object, possessive
*awo / *awu emphatic
1st person plural (Table 4.4)
It is possible to suggest two variants of proto-language person markers for the 1pl.: *ti / *ati. *a- was added to construct the emphatic pronoun. Proto-Jukunoid had *ti for subject and object, *-(a)ti for possessive, and *atï as a stressed pronoun. The dental obstruent has disappeared in much of Central Jukunoid regularly, cf. the Proto-BC plural class suffix *ti-, becoming *-i in Central Jukunoid (Shimizu 1971: 197). Based on the forms above, we can even see the direction of the reduction process: *t (Kuteb) > d (Bete) > j (Etkywan) > y (Hone) > 0 (Jukun).
The vocalism of Kainji and Plateau forms varies greatly, and the amount of data is not sufficient for any representation, so the tentative conclusions on Proto-CN markers can only be made according to the Jukunoid forms and the other person / number forms of the paradigm.
The *-Vt forms, with the back vowel, are suffixed to various nominalisers used to form the possessive markers, but in Plateau they are also widespread in the subject series. Tesu mboro, Berom wot, Cara futte and Kwanka work may all descend from Proto-Plateau *fiot(V), where is a reflex of the NC plural noun class marker *bV-. The possessive form in Proto-Plateau could be *mot, found in Beromic, Southeast and Southwest subgroups. Kanufi and Che have subject forms from *tot.
In Kainji, there is an evident lack of data that we could gather. However, the Cicipu subject marker tù correlates with Bantu tu-/-tu- 'we/us' (Babaev 2008: 139), and C'Lela cin 'we two' directly corresponds to Proto-CR *àcI(n) 'we'.
Hun-Saare and Ror are among the few Kainji languages with a distinction between exclusive and inclusive pronouns. This feature is a clear innovation.
*ti subject, object
*at(i) emphatic, possessive
Table 4.4
subject object possessive independent / emphatic
Plateau
Tesu mboro
Ayu bibtr
Berom wot < *hwot"59 wot, hot -mot, mot hot, wot
Cara futte
Izere yir...yir (pres.), ti...yir (past) ni...yir (pres), tiin...yir (fut) -feyir
Irigwe nji niy nji nji
Fyam tl, tik -té -mot, na-mun (emph.) moti
Horom tay
Hyam hera
Kulu andil yaa (e)nddl
Kanufi tot
Bu kita
Mada td td td
Eggon gi gi (dir.), dàgi (indir.) bi
Che tut i-mot
Kwanka wora
Ake ani
Idun men
Yeskwa embi, mbi (past), mbii (fut.), mbii (progr.), mbi (cond.) mbi -mbi / mbi gembi
Tarok i, i / i (fut.) yi yi, -jiyi / -nyiyi mmàyi
Sur y¥ y¥
Toro anoro
Lijili la la ila
Kainji
Hun-Saare a (incl.), te (excl.) na (incl.), te (excl.) na (incl.), te (excl.)
C'Lela cin (dual), mà, nà, cwan cwan a'co
Ror in (incl.), it (excl.)
Laru ti
Pongu géétù
Cicipu tù, ti- ttù otù
59 Bouquiaux 1970: 165
subject object possessive independent / emphatic
Jukunoid
JukunTakum i i Î at
Jibu yi
Hone i-, yii -yi (dm), yéyi (indir.) -yn ayii
Wapan i i -i ai
Etkywan ajë
Bete dëdë
Yukuben zï adi / ëdi / azi / ëzi
Kuteb tï tï -'tï, anatï atï
Lufu ï aï
2nd person plural (Table 4.6)
The Proto-Jukunoid forms can be suggested as *ni (predicative subject / object), *-ni (possessive), and *ânî (emphatic). The Jibu pronoun niy might reflect the original form with the final *-n lost in other Jukunoid languages. This pronoun is seen in Plateau, where it might be reconstructed as *p.i(n), corresponding to the Proto-CR *inI(n). Clear cognates within the Bantoid family include Vute (Mambiloid) nim and Manenguba (Bantu A15) *p.î- (subject) and *-àp.î (possessive) (Babaev 2008: 156).
The tentative Proto-Plateau form *bIn(I) is witnessed in Ayu bi-bin, Bu biyi, and probably some other languages as well. It is not present in either Kainji or Jukunoid. The possessives for Plateau languages are often derived from *-min, with the possessive particle *m-. Another form — *(a)nu — should be suggested on the basis of forms in Plateau (Ake, Toro, Kulu, Toro) and West Kainji (Hun-Saare, C'Lela and Ror).
The only two person markers of 1pl. which we can tentatively reconstruct for Proto-CN are *pi(n) and *(a)nu.
The combined chart of Proto-CN pronouns which may be proposed according to these thoughts is given in 4.5:
Table 4.5
subject object possessive independent / emphatic
1sg. *mI *me / *me Jukunoid *-m *me / *ame
2sg. *o / *u *wo / *wu *wo / *wu *awo / *awu
1pl. *ti *ti *-at(i) *at(i)
2pl. *pi(n), *(a)nu *pi(n), *(a)nu *pi(n), *(a)nu *pi(n), *(a)nu
The emphatic series is marked by the *a- prefix, and it would be logical to support the *ap.i(n) form for the 2pl., though it is only seen in Jukunoid.
Table 4.6
subject object possessive independent / emphatic
Plateau
Tesu mono
Ayu bibin
Berom yin < hyin hin, yin, én -min, min hin, yin, én
Cara yimi
Izere pin.pin (pres.), ta...pin (past) ka...pin (pres.), tàa...pin (past) -fepin
Irigwe pi niy pi pi
Fyam ti ...-n, wun -mun -mun, namot (emph.) muni
Horom min
Hyam pii
Kulu anuy ma (e)nuy anuy
Kanufi ayin
Bu biyi
Mada gya gya gya
Eggon gimi gymi / bmi / mo (dir.), dàmo (indir.) b"mi
Che yhepi bamin
Kwanka pina
Ake anù
Idun mbi
Yeskwa emi, mi (past), mii (fut.), mii (progr.), mi (cond.) mi -mi / mi gemi
Tarok o / opan (indirect) wo, opan (indirect) wo, -jiwo / -nyiwo mmàwô
Sur pipin pipin
Toro anono
Lijili yi yi iyi
Kainji
Hun-Saare no no no
C'Lela nwa
Ror no
Laru darni
Pongu gahi
Cicipu do, i- ddo idô
subject object possessive independent / emphatic
Jukunoid
JukunTakum ni ni ni ani
Jibu niy
Hone nan- / nan -nan (dir.), yanan (indir.) -nan anan
Wapan ni ni -ni ani
Etkywan ape
Bete lele
Yukuben pi api / epi
Kuteb ni ni -'ni, anani ani
Lufu ni ani
5. Edoid
The relatively small group of 33 Edoid languages is spoken in Southern Nigeria, in three areas separated from each other by either natural or ethnic borders. This division is the basis for the internal classification of Edoid: since Elugbe (1979, 1989) this family is usually divided into four main groups: North Western (NWE) and North Central (NCE) languages are spoken in the northern half of Bendel State; the South Western (SWE) languages border them from the south, and Delta Edoid (DE) is scattered in the language-dense district of the Niger Delta, to the east of the larger bulk of Edoid. The bordering ethnic groups speak West BC (Defoid, Ig-boid, Nupoid) and Ijoid languages.
Though the Edoid language area is rather compact, and even the earliest scholars noticed their mutual resemblance (cf. Koelle 1854), common morphosyntactic features of Edoid were not recorded until Strub (1915-16) gave a brief grammar description of one of Yekhee dialects, and Westermann (1926) analysed nominal class prefixes of some of the family's languages. By that time, the Edoid languages were seen as members of the Kwa family, and this tradition lasted until the 1970's when a few groups of Nigerian languages were unified under the term 'West BC' and thus joined the (New) BC family (Bennett & Sterk 1977).
The phonetic system of Proto-Edoid must be characterised by an extremely rich consonant inventory, including the fortis/lenis distinction of all obstruents and nasals, and two implo-sives which are only preserved now in DE (Elugbe 1989: 297). A ten-vowel system is reconstructed, in which a, e, i, o, u are joined by their 'non-expanded pharynx', more open pairs (Elugbe 1989: 298). Vowel harmony is one of the distinctive features in Edoid morphemes, including person markers. +ATR and -ATR vowels vary within the same lexeme, influenced by vowel harmony and the consonant environment. These varying vowels will be indicated by the respective capital letters for Proto-Edoid (i.e. *ml for *mi / mi).
Materials discussing Edoid grammar reconstructions mostly belong to Elugbe (1979, 1980, 1983, 1984, 1989). Neither these nor any other materials available to us cover the reconstruction of person markers or pronouns.
Personal pronouns are mostly independent in Edoid, though the subject forms are sometimes used as prefixed markers. A subject marker serves as a clitic following a nominal group subject, e.g. in Ivbie:
Ogele o ma yese (5)
Ogele 3SG much know
'Ogele knows much' (Emuekpere-Masagbor 1997: 190).
Subject markers are obligatory after emphatic pronouns and in copulative constructions with a nominal predicate. In Degema, both the independent pronoun and the concord person prefix are used in a verb phrase to show the emphasis:
mée mi-mon oji (6)
I lSG-see.FACT him
'I saw him' (Kari 2007: 92).
Four series of pronouns are rarely observed: the object marker coincides either with the subject or the possessive one. In some NCE tongues, all four series of markers given in 5.1. are identical. Possessive markers are located post-nominally and are syntactically independent.
1st person singular
Table 5.1
subject object possessive independent / emphatic
Yekhee (NCE)60 y, i- / i- me, me me, mhe me, me
Edo (NCE)61 i ßs ßs (i)me
Ivbie (NCE)62 mhi / i mhe mhe mhemhe
Okpamheri mi, me me hme meme
(NWE)63
Epie (DE) méni, me me meme
Degema (DE)64 me- / me-, mi- / mi-(past, neg.) mé / mé mée / mée mée
Engenni (DE)65 mi / mi -me / -me me / me me / me
Urhobo (SWE)66 mi, mé ße, me me oma-me
Okpe (SWE)67 mi / mi
Uvbie (SWE)68 mi, meme (progr.)
There are three candidates for Proto-Edoid: *mi and *i as the subject pronoun, and *me as a non-subject marker. The emphasis is often expressed by reduplication: Okpamheri, Epie meme, Ivbie mhemhe.
60 Strub 1915-16: 458-459, Elimelech 1976
61 Amayo 1980 [1975]
62 Emuekpere-Masagbor 1997
63 Elugbe 1980
64 Kari 2002; 2003; 2004, Thomas & Williamson 1967
65 Thomas 1976 [1968]
66 Ukere 1986
67 Omamor 1988
68 Omamor 1988
The i- forms in NCE may be derived from lenition and further elimination of the initial *m- before a front vowel. However, in Ivbie, the lenited form mhi is interchanged with i, which may demonstrate a different origin of the latter. The independent pronoun (i)me is found throughout the family.
We may suggest the following Proto-Edoid forms:
*mI subject (independent)
*I- subject (prefixed)?
*mE object / possessive
*mE / *mEmE emphatic
2nd person singular
Table 5.2
subject object possessive independent / emphatic
Yekhee U ê, ugwe, ugwegue £, itsy£ ê, ugwe, ugwegue
Edo u wé, é, ó, ó, á, a, rué, rùé rué / rué (u)we
Ivbie u gie, £ gi£ gie
Okpamheri u W£ bw£ wewe
Epie wowo
Degema mv- / mu-v- / u- (neg.) e- / e- (neg. imp.) wó wóo woo, uwo (Atala)
Engenni 6u / 6v -wo / -wo wo / wo 6o / 6o
Urhobo we, wo w£ wén oma-wen
Okpe wù / wv
Uvbie ù, gue (progr.)
We might reconstruct *u and *wo / *we for subject and non-subject markers respectively. Emphatic pronouns which may be suggested are *w£w£ / *wowo or *uw£ / *uwo.
Yekhee and Ivbie demonstrate a development *w£ > *gw£ (> gi£). Edo's collection of object markers vary depending on the phonetic environment. Degema has mu, which consists of the original u attached to the affirmative prefix m-. In the reconstructions below, we show capital letters to avoid duplicating forms with alternating +ATR or -ATR vowels.
*U subject
*wO, *wE non-subject
*wEwE / *wOwO, *uwE / *uwO emphatic
1st person plural (Table 5.3)
Table 5.3
subject object possessive independent / emphatic
Yekhee ma ma ma ma
Edo (i)ma ima ima, ßa (i)ma
Ivbie eye eye eye
Okpamheri mani mani hmani
Degema me- / me-, néni
e- / e- (neg.) eni em
Engenni eni eni eni eni
Urhobo aßare aßären raßare oma-raßare
Okpe ami / ami
Uvbie orín
The only form which can be supposed for Proto-Edoid is *ma / *mE, observed in NCE and NWE. No external cognates for this form were found in West BC. However, we find some examples of subject *mV- in Central Nigerian: Idun (NW Plateau) men, C'Lela (W Kainji) ma; and in CR: Mbembe mo- / mo-, Legbo me / me (negative). Maybe at least some of them were generated under the influence of the 1sg. marker *me / *mi (which is clear for Okpe). Typologically, this seems reasonable. Building the plural of the pronouns by changing the vocalism of the singular one is a common feature in Indo-European and beyond (Indo-European 1sg. *me ~ 1pl. *mes / *mos, Turkic 1sg. *ben ~ 1pl. *bir, Mongolian 1sg. *bi ~ 1pl. *ba, Hungarian 2sg. *te ~ 2pl. *ti, etc.); this method of building the plural is one the most widespread in the languages of the world (Siewierska 2004). Moreover, a vowel change is the core method of building the plural in the Edoid nominal system, and it is no wonder that it might be used for person markers as well.
Outside the BC family, similar markers of 1pl. are found in Kwa, namely the Gbe group (*mi 'we') and in C and W Tano languages (*ame 'we incl.') (Бабаев 2010). Several Grusi (the Gur family) languages show ma as an object marker (Segerer 2002-07).
Other forms in Table 5.3 do not allow to suggest any immediate reconstructions. We should note the resemblance of Urhobo a^are 'we' to Bekwarra (Bendi, CR) abere and Tesu (Alumic, Plateau) mboro. However, in Edoid there are no traces of the NC plural class prefix *ba-, so this form seems to be a frozen relic of this plurality marker.
2nd person plural (Table 5.4)
Most languages show various stages of levelling in the four series of pronouns. Yekhee, Edo, Ivbie and Engenni have basically only one pronoun for all meanings, Degema has two.
We tentatively reconstruct the Proto-Edoid pronoun as *wa / *fia. *ba-forms in other BC languages must be cognate with this marker, which obviously originated from the plural noun class prefix.
Table 5.4
Yekhee ßa, ißa ßa, ißa ßa ßa, ißa
Edo (u)wa uwa uwa (u)wa
Ivbie vhe / vha vhe vhe vhevhe
Okpamheri we wa buba
Degema ma- / ma-, maap / maap maap / maay maay / maay
a- / a- (neg.)
Engenni 6a -6a 6a 6a
Urhobo wa, ßan wa, ßan rowan, roßan oma-rowan
Okpe ari / ari
Uvbie mama mama
Summing up Section 5, the following hypothetical reconstructions for Proto-Edoid are suggested:
Table 5.5
subject object / possessive independent / emphatic
lsg. *ml, *I-(?) *mE *mE, *mEmE
2sg. *U *wO / *wE *wOwO / *wEwE, *uwO / *uwE
lpl. *ma / *mE *ma / *mE *ma / *mE
2pl. *wa / *ßa *wa / *pa *wa / *pa
6. Defoid and Igboid
Yoruboid and Akokoid languages were unified under the Defoid brand (Crozier & Blench 1992), although no consensus has been reached on this among Africanists (Capo 1989), and Igboid languages are defined as their close relatives in (Williamson & Blench 2000). To date, there is no comparative grammar or complex phonetic research written on any of the families in West BC, except some hard-to-find dissertation manuscripts in Nigerian universities. The situation is complicated further by the lack of data on many languages, beyond 100/200-item wordlists which rarely include personal pronouns and never deal with paradigms.
Around two dozens of Defoid tongues (the exact number is still risky to claim) are spoken in Southeastern Nigeria, the adjacent border areas of Benin and some districts of Central Togo. Yoruba or Yoruboid lects dominate the family, Yoruba itself having the greatest number of speakers in Nigeria, Benin and as far as West Indies, Brazil and USA. Igala, an isolated tongue considered a close relative of Yoruboid, is in use on the Eastern bank of the Lower Niger, and is classified as a separate subgroup. The other branch of the proposed Defoid stock, Akokoid, is a cluster of as many as ten less known dialects spoken to the SW of the Niger/Benue confluence.
The internal subclassification of Yoruboid has been proposed by Akinkugbe (1976, 1978), analysed by Capo (1989) and is based on a number of innovations separating Igala from the main Yoruboid core. There is a draft Proto-Yoruboid Swadesh list put online by Guillaume
Segerer69 (later referred to as [online]), but the conclusions there are apparently made upon Yoruba and Igala only, and only subject markers are included.
An attempt to classify Akokoid is being made by Blench (in progress). Ohiri-Aniche (1991, 1999) showed that Akokoid is as close to Edoid as it is to Yoruboid. Of all the Akokoid languages, Arigidi has enjoyed the most attention (e.g., Abiodun 1984), but mainly in regard to phonetics, so that personal marking remained almost untouched. For Yoruba, personal pronouns were studied by Bamgbose (1966), Fresco (1968), Omamor (1976) and others. There is a brief comparison of four basic personal pronouns of western Nigerian languages in Ohiri-Aniche (1999: 90). For Akokoid tongues, we only give Arigidi data from his paper.
The Igboid language area lies in the left-bank half of the Lower Niger basin in Nigeria. The total number of languages (some of which should rather be called dialect clusters) does not exceed ten, but the dialects of one idiom may vary greatly. Igbo, the largest language of the family with over 25 million speakers, includes over 30 dialects recognised so far (Lewis 2009). However, they have some slight differences in morphology. Outside Igbo, on which there exists extensive linguistic literature (see Manfredi 1989 for a survey), the only language thought to represent a separate branch of Igboid is Ekpeye, a tiny language spoken between the Orashi and Sambreiro rivers. Ekpeye is surrounded by Delta Cross, Edoid and Ijoid languages and must have undergone some influence from them. Some of the few sources of language data are the two papers by Clark (1969, 1972) and an unpublished wordlist on Roger Blench's website (Blench 2006a).
The Proto-Igboid phonetic system, according to Manfredi (1989: 344), included palatals such as c, j, f, p, labiovelars kw, gw, nw, hw, kp, and gb, and a 9- or 10-item vocalic system which we find in all Igboid lects. The vowel harmony is widespread, and the vowel system with the +/-ATR opposition must go back to the proto-language. Person markers of both Igboid and De-foid have plenty of tense/aspect, polarity and modality variations, mostly for the subject, marked by both vowel alternations and difference in tone. Tones, widely used for building various syntactic meanings all over West BC and Kwa (Creissels 2000: 238), are important since a great lot of person markers in these languages have a V form, as a result of reduction of earlier CV markers. In a number of tongues, subject, object and possessive markers fully coincide on the segmental level, with only occasional tonal distinctions. Direct object markers usually resemble their subject counterparts (sometimes with a different tone as well), or coincide with the possessive pronouns.
1st person singular (Table 6.1)
We cannot go further than *ml (with an undefined front vowel) for the reconstruction of all the three right-hand sets: subject, object and possessive pronouns. Proto-Igboid must have had *mi, but the Igbo forms have undergone reduction everywhere, so we can only rely upon the stressed pronouns here. Segerer (online) suggested Proto-Yoruboid *mi without any supporting evidence. The independent pronoun is *aml for both Proto-Igboid and Proto-Defoid. Izi -bedva is a particle used to indicate emphasis in all person markers.
69 http://sumale.vjf.cnrs.fr/NC/docs.php
Table 6.1
subject object possessive independent / emphatic
Defoid
Igala70 mi, nn, nna (progr.) mi mi omi
Yoruba71 mo / mo / mà, n, y (neg., past) mi mi èmi, dmi (Oka)
Arigidi72 ame
Igboid
Igbo73 m-, m / mv m m àmî
Echie74 m(v) -m m(v)
Ogba75 m
Ikwere76 m m (before obj.), mé mé mé, me (contrastive)
Izi77 mu mu mu mbedva
Ekpeye78 me / me / ma / mo / mo- (tense), N- -m me me
2nd person singular (Table 6.2) Table 6.2
subject object possessive independent / emphatic
Defoid
Igala we, e, weé (progr.) we we uwe, ee
Yoruba o, o (before particle), wà (before a) o, e e, re iwo, ùwo (Oka)
Arigidi aro
Igboid
Igbo i- / 1-, gl gl i àgî
Echie I gl gl
Ikwere i i, i (before object) i ji, je (contrastive)
Ogba yv, iyv
Izi i / l ngu ngu gvbedva
Ekpeye i- /1- / e- / e- (tense) yo -l yd yô
70 Philpot 1935
71 Bamgbose 1966, Ohiri-Aniche 1999: 90
72 Ohiri-Aniche 1999: 90
73 OuxMaH 1975
74 Blench, manuscript c
75 Blench 2005
76 Segerer 2002-07
77 Meier 1976
78 Clark 1976
Yoruboid and Akokoid data are too scarce to make ultimate conclusions: emphatic *uwe / *uwo might be the only form reconstructible for Proto-Yoruboid. Segerer (online) reconstructs Proto-Yoruboid *o. Arigidi shows an interesting form which is comparable with Eleme (CR, Ogoni) object *-rü 'us' and Edo non-subject rué / rue 'our, us'. Let us also add here Ukaan (Section 8 below) ihye-rb 'you (sg.)' (Ohiri-Aniche 1999: 90).
Proto-Igboid markers can be reconstructed as *i- (subject) and probably *gi / *gu-.
1st person plural (Table 6.3)
Table 6.3
subject object possessive independent / emphatic
Defoid
Igala a, aá (progr.) wa wa awa
Yoruba a, à (before particle or á) wa wa àwa
Ede dialects wa, ya, a
Arigidi áo
Igboid
Igbo apl / apt api api apî
Echie pi pi pi
Ikwere a áyl áyl áyllé
Ogba yê
Izi api api api apibedva
Ekpeye a- (excl.), a-.-nl (incl.) ye ye yê
Yoruboid markers may be derived from *a for the subject marker, *wa for object and possessive, and *awa for the independent emphatic pronoun. Proto-Igboid must have had *(a)pi, though note *a in Ikwere and Ekpeye (plural exclusive). The *(a)p.i form has a number of notable cognates across BC: let us name Efik and Ibibio (Lower CR) npin 'we', Ake (South West Plateua) ani 'we', Degema and Engenni (Delta Edoid) eni 'we'.
2nd person plural (Table 6.4)
The Yoruba object and possessive pronoun yin 'you, your' reminds of *^in which we have suggested in CR (see Section 3 above). A slighly different marker is found in Proto-Igboid *(V)nu. This pronoun appears below in Idomoid (Section 7) and Oko (Section 8) as well.
Arigidi amo has much in common beyond Defoid. It must be cognate to the Proto-Bantu subject prefix *mv- (Babaev 2008: 147) and probably Ukaan (Ikakumo dialect) ïhyè-mo. This correspondence, if verified, would be enough to suggest a Proto-BC retention.
Table 6.4
subject object possessive independent / emphatic
Defoid
Igala ame, me, mee (progr.) me me ame
Yoruba e, e (before particle or a) yin yin eyin, e-pwï (Oka)
Arigidi amd
Igboid
Igbo unù unù unù unû
Echie nv nv nv
Ikwere anv anv anv anvlé
Ogba vnv
Izi unu (variable tone) unu unu unubedva
Ekpeye i- / i-.. -nl, e- / e-.. -nl yinl yd yi / yonl
We will briefly summarise our tentative reconstructions in the following charts:
Table 6.5
subject object / possessive independent / emphatic
lsg. *mI *amI
2sg. *i (Igboid) *gi / *gu (Igboid) *uwe / *uwo (Yoruboid)
lpl. *a (Yoruboid) *(a)pi (Igboid) *wa (Yoruboid) *awa (Yoruboid)
2pl. \in (Yoruboid) *(V)nu (Igboid)
It is obvious that, although we tend to fit person markers into a traditional "Bantu-style" four-column table, noting subject, object, possessive markers and independent pronouns, this does not seem realistic for West BC. Proto-Igboid must have had only two distinct series of person markers (subject and object), and Yoruboid could have up to three, including independent pronouns with a vocalic prefix.
7. Nupoid and Idomoid
The Nupoid family spoken at the Niger-Benue confluence and to the north of it includes around a dozen languages: Blench (1989a) names seventeen, while Ethnologue (Lewis 2009) gives only eleven. They are indeed tightly knit together, which was noticed long ago by Koelle (1854: 8-9), who was the first to describe some of them. The grouping of Nupoid seems to have never been in doubt, as well as its subclassification: the Ebira-Gade group must be quite distant from the Nupe-Gbagyi core, within which Nupe and Gwari (Gbagyi-Gbari) subgroups are usually identified.
Diagram 3
Nupoid comparative research was conducted by Sterk (1977, 1978), Bennett & Sterk (1977), Hyman & Magaji (1970), Madugu (1975, 1979, 1985), to name just a few. The overview is contained in Blench (1989a). It argues for a 10-vowel system for Proto-Nupoid, with a complete series of +ATR and -ATR slots, and a fortis/lenis counter-opposition of consonants. The proto-language consonantal system is given by Blench as follows:
p b
m
w
t d s z r l
n
y
c J
k g kp gb
h
Nupoid languages use at least five tones, with three level tones and two contour tones. Noun classes are mostly marked by innovated suffixes, while some languages only have suffixes marking the plural. The word order of the verb phrase is SVO, the subject and the object of the verb expressed by free-standing pronouns. The object marker may be infixed into the verb root, like in Nupe kpêyê 'to know' — mï-kpê-û-yê 'I know it'. Possessive markers are post-posited to nouns.
While the Nupoid language area is mostly surrounded by other West BC tongues, including Igboid and Yoruboid, the smaller Idomoid family of languages lies farther to the east, bordering some of the East BC lects: e.g., Tiv (Bantoid) and Bokyi (CR). Eight or nine Idomoid languages are usually identified, which are classified following Crozier & Blench (1992):
Diagram 4
The idea of close genetic relations between Nupoid and Idomoid was first expressed as early as the mid-nineteenth century, and further supported by Bennett & Sterk (1977) and Blench (1989a). However, Armstrong (1989: 323) considered the Idomoid branch coordinate with all other West BC families. Armstrong had been analysing Idomoid from a comparative
standpoint since 1955, and has described a number of dialects, along with some comparative research papers (1981, 1983). Apart from this, most works are synchronic overviews, some of which are used in the charts below.
Proto-Idomoid must have had a 10-vowel system which is still observed in Igede, but was reduced to seven items in most of the other languages of the family. The consonant system lacks a fortis/lenis opposition characteristic for Nupoid. As well as in the latter, tones play a vital role in morphology, including the pronominal forms which build negative, tense, and object meanings using various tones. Idomoid languages have up to four tones.
In Idomoid, person markers are usually bound to verbs. The only independent pronoun is used in the emphatic context. Possessive suffixes are often attached to the attributive particle which is a former noun, cf. Idoma ku-m 'my', ku-nu 'his' etc. This feature is also encountered in East BC languages, including Bantoid and CR languages.
1st person singular (Table 7.1)
Table 7.1
subject object possessive independent / emphatic
Nupoid
Ebira79 ma / me mi / mi ami / emi emi
Gade80 n
Nupe-Tako81 mi, mii (opt.), mi (neg.) mî mî mî
Gbagyi82 mi (tone) mi (tone) mi (tone) dmi
Gbari83 myi, mà (fut.), ma (past) myi myi
Idomoid
Idoma84 m -m -m ami
Eloyi85 mi- (tone) mi -mi imi, imiml
Igede86 m- -m / -m pam àm
For both Nupoid and Idomoid, only two forms may be reconstructed: *mi for the marker expressing the verb subject, object and possession, and *Vmi for the emphatic pronoun.
79 Scholz 1976 [1973]
80 Sterk 1977
81 Smith 1980 [1967]
82 Hyman & Magaji 1970
83 Blench & Doma, manuscript
84 Armstrong 1989, Segerer 2002-07
85 Mackay 1976 [1968], Blench 2007d
86 Bergman 1976 [1973]
2nd person singular (Table 7.2)
Table 7.2
subject object possessive independent / emphatic
Nupoid
Ebira u / V wu / wv awv/ewu ewv
Nupe-Tako wo, o / woo (opt.), wo (neg.) wô / ô wô / wê / ô wô
Gbagyi ho ho ho oho
Gbari he he he
Idomoid
Idoma à -à àwà
Eloyi yo- (tone) yö, ya -ya üy(w)ö
Igede à-, àhv -y / -y pay àhv
u / o / o markers and wu / wo in non-subject series are cognate to 2sg. markers in other BC languages. Pronouns with initial y- in Idomoid remind of the yo / ywo forms in Plateau languages, as well as some Igboid forms.
1st person plural (Table 7.3)
Table 7.3
subject object possessive independent / emphatic
Nupoid
Ebira ya / ye, i / i yi / yi ayi / eyi eyi
Nupe-Tako yi yii ^pt^ yi (neg.) yî yî(-zi) yî
Gbagyi yi yi yi oyi
Gbari yi yi yi
Idomoid
Idoma àlà -(à)là àlà
Eloyi ki- nzú -nzú kinzú
Igede àhl -hi pahi àhi
The Proto-Nupoid pronouns may be conjectured to have been *yi for subject / object / possession and *Vyi for the emphatic pronoun. The -zi suffix in Nupe is a nominal plurality marker.
The Idoma pronoun reminds of lo-r which we observed in Bantu zones A-C (Babaev 2008: 171-175), where they are derived from Proto-Bantu *tv-. Igede àhl may be a derivation from *aci or *ati, widely witnessed across BC.
2nd person plural (Table 7.4)
Table 7.4
subject object possessive independent / emphatic
Nupoid
Ebira wa / we nini wu / wu nini awv / ewu nini ewv nini
Gade pa
Nupe-Tako ye, e (before cons.), yee ^pt.^ yé (neg.) ê / yê ê / yê(-zï) yê
Gbagyi fye fye fye ôfye
Gbari se se se
Idomoid
Idoma aá áá, -no and
Eloyi lu-, lalu- (tone) yí -yí lú(n)yí
Igede ànv -nú panv ànv
For Idoma and Igede, the proto-form must be *-nv / *-no and *anv / *ano. Both *nu and *pi, as we have seen earlier, are frequent all across the BC area. But there is a clear distribution: *nu / *Vnu is seen in Igboid and Bantoid, while *p.i is in Yoruboid, Central Nigerian and CR (but note anu(y) in some Plateau dialects in Section 4). We need a much deeper insight to reconstruct any proto-forms here.
In conclusion, we must say that the pronominal systems of Nupoid and Idomoid do not share enough similarities to justify the hypothesis of a specific mutual genetic relationship. Beyond the 1sg., there is hardly a single pronoun that we can reconstruct for 'Proto-Nupoid-Idomoid'. The idea of the NOI unity which we mentioned earlier does not seem to find evident support in the systems of person marking.
8. Benue-Congo Isolates
Oko (Ogori) is an unclassified BC language spoken by some 10,000 people in the Kogi State of Central Nigeria. Together with Ukaan and Akpes, Oko has long been regarded as an isolate due to significant lexical differences with the surrounding BC tongues. However, Williamson & Blench (2000) included it into West BC, placing it into the same stock with Nupoid and Idomoid.
Table 8.1
subject object possessive independent / emphatic
lsg. ï- / è- (perf.) mè- / mà- (progr.) mame, -mu / -mo m- ame
2sg. u- / o- (perf.) wè- / wà- (progr.) wawo, -wu / -wo w- av, awo
lpl. tï- / tè- (perf.) tè- / tà- (progr.) tato, -tu / -to t- / tï- ato
2pl. nï-/ nè- (perf.) nè- / nà- (progr.) nano, -nu / -no n- ano
There are at least two works on personal pronouns in Oko. We quote their forms below in accordance with works by Elugbe & Ologori (1980) and Atoyebi (2008). The variants divided by a slash mark depend on vowel harmony. See Table 8.1.
There are a few remarkable things about the Oko person marking system. First, its subject markers differentiate between progressive (incompletive) and perfective (completive) aspects. The difference lies mainly in the vocalism of the forms, but in the 1sg. they are built upon suppletive roots. This seems to have been the impact of an old aspect marker which, as it usually happens in BC, used to follow the subject marker, but finally merged with it, changing the vocalism of the latter. However, there is some additional morphological sense here. Completive markers are immediately linked to the verb root, while progressive ones are prefixed to the tense / aspect markers:
i-su-ja (7)
1SG.S.COMP-marry-3SG.O
'I married him', but
ma-aka-gam obin usie (8)
1SG.S.PROGR-FUT-greet king tomorrow 'I will greet the king tomorrow' (Atoyebi 2008: 31-32).
The only predicative marker preceded by completive person markers is that of negation: here the completive marker is used to prevent homophony, because in Oko the negative prefix is -me-. It seems that the completive markers are more ancient, and the progressive ones may have originated from complex auxiliary constructions.
Second, there are bound and independent object forms, and the latter are built by means of reduplication. Finally, the possessive markers are prefixed: a feature rarely met across BC.
The roots used for marking persons are definitely cognate with those found in East BC. See the comparison in 8.1. showing structural, segmental and tonal similarities between Oko and NW Bantu independent pronouns:
Table 8.1
Oko Viya (Bantu B301)
lsg. ami mi
2sg. av / awo we
lpl. ato itu
2pl. ano inu
Note that Oko has no Bantu neighbours — it is surrounded by West BC languages, so that the chance of language contact with Bantu as the reason for such a striking similarity is quite low. Since the Bantu system is considered archaic compared to other BC families, Oko pronouns seem to enjoy a common retention of the original pronouns with Bantu.
Akpes is another isolate tongue of C Nigeria, spoken to the southwest of the Niger-Benue confluence squeezed between larger Edoid and Yoruboid areas. It consists of about ten dialects, most of which are mutually intelligible. Disputes on the external classification of Akpes have lasted ever since Hansford & al. (1976) first described it as 'Kwa unclassified'. The view of Akpes as a BC language was expressed in (Williamson 1989) and reproduced in (Williamson & Blench 2000), supported by Elugbe (2001). Some classify Akpes as an Edoid language (Agoyi 1997).
The independent subject pronouns of Akpes are cited by Ohiri-Aniche (1999: 90).
Table 8.3
lsg. örp
2sg. ösi
lpl. abes
2pl. abet
The 1sg. form is probably derived from *dmi. The two plural forms correspond perfectly to East BC: we saw (a)bese < *(a)be-c(u)e in both Bantoid (Babaev 2008: 148) and CR languages (Section 3), there are remnants of this form in the languages of the Plateau (Section 4) and beyond. The same goes for *(a)be-n(u)e as the 2pl. form. The 2sg. form is a mystery, never found among BC languages. It can be a loanword, e.g. from some Gur languages of the area: cf. Ntcham -si, Akasele ise or Konkomba si, all denoting 2sg. and all spoken in Togo.
However, the three cognate forms out of four strongly suggest a common retention shared by Akpes and East BC languages.
Another unclassified BC language of C Nigeria is Ukaan, known by only a few papers published or placed online so far (Jungraithmayr 1973: 47-48), (Ohiri-Aniche 1999: 90), (Blench 2005b). Four dialects are identified, and each possesses its own grammatical peculiarities making them rather dissimilar to each other. Relying on the abovementioned works, we hereby present the forms from three of them: Ishe, Ikakumo and Ikan (only for the 1sg., taken from [Adekanye & Salffner 2007]).
Table 8.4
subject object possessive independent / emphatic
1sg. do (Ishe) ihye-ji (Ikakumo) ja (Ikan) wity (Ishe) -j / -ji (Ikan) joo (Ikan)
2sg. ho (Ishe) ihye-ro (Ikakumo)
lpl. ba (Ishe) ihye-bo (Ikakumo) ubo (Ishe)
2pl. ma (Ishe) ihye-mo (Ikakumo)
The Ikakumo dialect, whose subject pronouns are actually nouns with possessive suffixes, shares three forms of its pronominal markers with Arigidi (Akokoid) (see Section 6), and the Ishe dialect also corresponds with them. Based on this, we can confirm that the pronominal system of Ukaan is that of a West BC language quite close to Akokoid.
9. Conclusive notes
We have compiled data on person markers in some 300 languages of the BC family which might lead us to some preliminary conclusions on the Proto-BC system of person marking. Even
1st person singular
Table 9.1
subject object possessive independent / emphatic
Bantoid *pi- *(a)me *(a)me *(a)me
Bantu *pi- *(a)me *(T)me
CR *N-, *mi *m(i) ami *(a)mi, *minV
CN *me *ame
Edoid *mi, *i- *me *me *meme
Yoruboid *ml *amI
Igboid *ml *amI
Nupoid *mi *mi *mi *Vmi
Idomoid *mi *mi *mi *Vmi
Oko i-, me- mame, -mu m- ame
Akpes dpi
Ukaan do, ja, -ji -ji joo
2nd person singular Table 9.2
subject object possessive independent / emphatic
Bantoid *o- *(a)ue *(a)ue *(a)ue
Bantu *v- *-kv- *(a)we *(i)we
CR *o-, *a- *wo / *wu / *we, *wo-n(V) / *we-n(V) *owo, *wo-n(V) / *we-n(V) *dwd
CN *o, *u *wo, *wu *wo, *wu *awo, *awu
Edoid *U *wO, *wE *wO, *wE *wEwE / *wOwO, *uwE / *uwO
Yoruboid *uwe, *uwo
Igboid *i *gu *g° *gu *go *gu, *g°
Nupoid Nupe wo Nupe wo Nupe wo Nupe wo
Idomoid Idoma a Idoma -o Idoma awo
Oko Ü-, we- wawo, -wo w- av, awo
Akpes dsi
Ukaan ho, -ro
tentatively, this is no easy task. BC is an old family (over 6,000 years of age, according to [Williamson 1988: 104]), and the diversity of pronominal systems in all of its 900+ languages is much greater than that of Indo-European or Semitic. The data is scarce or incomplete, there is no historical track of the language development, and the lack of regular phonetic correspondences and reconstructed phonetic systems makes any comparative analysis highly vulnerable. Finally, various phonological processes of merging person markers with predicative markers
1st person plural
*e- / *i- subject
*aci(n) independent
*mV subject / object / independent
Table 9.3
subject object possessive independent / emphatic
Bantoid *tv- *(be-)ce *(bé-)ce *(be-)ce
Bantu *tv- *(a)cue *icue, *(be-)cue
CR *e- / *i-, *mV *mV *aci(n), *mV
CN *ti *ti *-at(i) *at(i)
Edoid *ma / *mE *ma / *mE *ma / *mE *ma / *mE
Yoruboid *a *wa *wa *awa
Igboid *(a)pi
Nupoid *yi *yi *yi *Vyi
Idomoid Idoma aid Idoma (a)ld Idoma aid
Oko ti- / te- tato , -tu t- atd
Akpes abes
Ukaan bd / ba ubd
2nd person plural Table 9.4
subject object possessive independent / emphatic
Bantoid *mv- *(bé-)ne *(bé-)ne *(be-)ne
Bantu *mv- *(a)nue *(i)nue
CR *e-, *i- *anl, *inl(n) *anl, *inl(n) *anI, *inI(n)
CN *pi(n), *(a)nu *pi(n), *(a)nu *pi(n), *(a)nu *pi(n), *(a)nu Jukunoid *dni
Edoid *wa / *pa *wa / *pa *wa / *pa *wa / *pa
Yoruboid *pin
Igboid *(V)nu
Idomoid *nv / *no *nv / *no *anv / *and
Oko ni- / ne- nano, -nu / -no n- and
Akpes abep
Ukaan md / ma
Ukaan bd / ba ubd
of tense, aspect, modality and polarity have made the situation in many languages obscure. We are confident that much more work is needed to create a reliable reconstruction of the proto-language system of person marking for BC. The following are just some conclusive
charts following the discussion in this paper. Bantu and Bantoid forms are given from (Babaev 2008: 148).
We may conclude that the object and possessive markers are mostly marginal compared to subject markers and emphatic pronouns. The object is marked in all languages of the stock, but in the majority there is no specific object series, and the subject markers or emphatic pronouns are used instead. For a number of languages (Bantoid, CR) we reconstruct a cumulative "non-subject" series that marks all of the remaining three meanings.
Similarly, the possessive meaning is built by means of the emphatic pronoun, usually suffixed or postposited to the possessed noun. Where proto-languages seem to have distinguished between emphatic and possessive pronouns (in Bantu, CN), the distinction was only a matter of a vowel prefix (or particle).
In Bantu, the independent pronouns constitute a separate person marking series which does not seem to be etymologically cognate with subject markers: *ame and *pi- in the 1sg. (Ba-baev 2008). Bantu is considered to be the most archaic of all BC branches, since it has preserved a number of important morphological retentions lost in the western part of the BC area. One of the examples is certainly the noun class system: Bantu preserved old class prefixes in their original CV-form, while most of its relatives in Nigeria either retained only V- prefixes or dropped every trace of the original system, sometimes (as in Jukunoid) elaborating a new one.
We face a similar situation in person marking. In Proto-BC, there seem to have been prefixed person markers of the verb subject and the independent pronouns, which were based on the same root as the non-subject markers (i. e. object and possessive). Originally this root was probably a nominal one: note that non-subject pronouns in Bantoid and CR languages can use the plural noun class prefix *bV-, which the subject markers do not show.
Later the prefixes were weakened in the majority of BC branches: they were probably reduced to V-shape already in Proto-West BC. Just as it happened with noun classes, the CV-marker was weakened to V- or -V, and then finally was lost. The process was encouraged by the presence of a clear "substitute", i. e. an independent pronoun with emphatic semantics that could instantly replace the prefixed subject marker. The BC languages of present-day Nigeria mostly use subject pronouns independently, and they are generally identical with or at least related to the emphatic pronouns.
Some traces of the former subject prefixed series are still found. In some languages, they are found in V-shape prefixes (e.g. for the 1sg. i- or n- which is a syllabic nasal bearing its own tone).
It is typologically common that when a language merges two series of pronouns in one, the resulting paradigm is mixed. Turkic languages, for instance, have adopted person markers of the preterite from the ancient possessive paradigm only in the 2sg. and 1pl. forms, using the original subject marker series for all the other slots. Unifications of several pronominal paradigms into one are commonly known in the world's languages, and non-subject forms adopted to mark the subject are also not unknown: let us recall Indo-European *me- 'me', which replaced older *eg'Hom for the nominative case in Irish, Hindi or Farsi, or Indo-European *nos 'us' which was generalised for the subject meaning in Latin and all Romance languages.
Syntactic shifts of this and other kinds are quite habitual for BC languages. In a number of languages, the pronominal paradigm is further unified by merging plural pronouns together as it happened in CR (cf. the Old English plural personal affix -ap). The plural particle *bV-which in Proto-BC marked the 3pl. spread to the 2pl. and further to 1pl., squeezing out the original person markers; this was noted as early as Delafosse (1904: 29). Another common means of building the plural innovations is to add *bV- to the corresponding singular pronoun, as in Tee (CR) boo 'you pl.' vs. oo 'you sg.' Proto-Edoid chose to analogically level the paradigm based on the singular pronoun roots: *mi / *me 'I' ~ *ma 'we' in the 1st person and *we / *wo
'you sg.' ~ *wa 'you pl.' in the second, and Delta Cross languages use the same affixes for marking singular and plural of the 2nd and 3rd persons.
We may suggest that there were only two series of person markers in Proto-BC: the subject prefixed series and the non-subject independent series:
Table 9.5
subject non-subject
lsg. *N- *mI
2sg. *o- / *u- *wV
1pl. *tu- / *ti- *(bV)ce
2pl. *mV- ? *(bV)nV
Some possible reflexes of these tentative reconstructed person markers are given below. Bantu and Bantoid forms are taken from (Babaev 2008), all other forms are mentioned in the charts of the present article.
*N-
Bantu: Oroko N-, Ndumu n, Enya n-, Shi n-, Digo n-, Tongwe N-, Kongo N-, Haya N-, Pimbwe n-, Yao n-, Tswana n-;
Other Bantoid: Samba-Daka m, Tiv m, Noni N-, Mbe n-, Kenyangp- / n-, Yemba N, Aghem N, Mundani nCR: Bekwarra N-, Kohumono N-, Mbembe N-, Ibibio N-, Eleme N-CN: Tarok n / m,, Lijili n, Basa n (past), Hone n-, Jukun m Defoid: Igala np, Yoruba n, y (neg., past) Igboid: Igbo m-, Ikwere m, Ekpeye N-Nupoid: Gade n Idomoid: Idoma m, Igede m-
*ml
Bantu: Basaa me, Tsogo me, Kele eme, Bila ime, Shambala imi, Kongo -ame, Luvale ami, Herero ami. Zulu mina
Other Bantoid: Samba-Daka mee, Mambila me, Esimbi me-, Kenyang me, Bamileke me, Jarawa mi / mi, Bwazza mi
CR: Boki me, Kukele me, Bekwarra ami, Lokaa -mi, Obolo mi-, Efik mi, Ogbia mi / mi, Eleme -mi
CN: Berom me, Horom me, Eggon me, Tarok mi, Lijili me, Hun-Saare me, Etkywan ame, Kuteb me
Edoid: Yekhee me / me, Edo (i)me, Epie me, Engenni mi / mi, Urhobo mi / me
Defoid: Igala mi, Yoruba mi, Arigidi ame
Igboid: Igbo ami, Ikwere me, Ekpeye me
Nupoid: Ebira mi / mi, Gbagyi mi
Idomoid: Eloyi mi, Idomo ami
Oko ame
Akpes dpi
*o-/*u-
Bantu: Oroko d-, Bafia Ü-, Mpongwee d-, Nkengo u-, Kumu v-, Digo v-, Sukuma v-, Shambala u-, Vili Ü-, Shi
Ü-, Kwangari o-, Songe o-, Pimbwe v-, Matuumbi v-, Herero Ü-, Venda Ü-, Tswana d-Other Bantoid: Tikar Ü, Tiw ü, Mbe d-, Kenyang 5-, Limbum a, Yemba d, Lamnso' a CR: Bekwarra o-, Kohumono a-, Mbembe a-, Efik d-, Eleme o-Central Nigerian: Berom ho-, Tarok u, Lijili o, Hun-Saare o, Wapan Ü, Kuteb u Edoid: Edo u, Uvbie Ü Defoid: Yoruba o Nupoid: Ebira v Idomoid: Igede a-, Idoma a Oko Ü- (perf.)
Bantu: Oroko owa, Tsogo ewe, Babole awé, Lega ugwe, Rombo -afo, Nilamba ve, Shambala iwe, Ha we(we),
Kwangari ove, Nyakyusa-Ngonde ugwe, Matuumbi weé-nga, Herero ove, Venda iwe, Copi awe Other Bantoid: Mambila wo, Tikar wu, Esimbi wo-, Noni wo, Mbe -we, Kenyang wo, Limbum we, Yemba wú,
Mundani wee, Mbula wé CR: Bekwarra wo, Kohumono agwo, Kukele we, Lokaa awú, Obolo dwú
CN: Izere wan, Fyam wéé, Horom awo, Bu awu, Tarok wa, Lijili iwo, Hun-Saare wo, Jibu wu, Bete owu
Edoid: Edo wé, Okpamheri we, Degema wo, Urhobo we, Okpe wú / wu
Defoid: Igala we, Yoruba iwo
Nupoid: Ebira wu / wu, Nupe-Tako wo
Idomoid: Idoma awo
Oko -wu / -wo
*tu- / *ti-
Bantu: Bafia ti-, Viya tú-, Babole tó-, Nande tu-, Kuria tu-/to-, Sukuma tu-, Swahili tu-, Vili tú-, Rwanda tu-,
Dciriku tu-, Lunda tu-, Pimbwe tu-, Sena ti-, Matuumbi tu-, Umbundu tú-Other Bantoid: Meta ti, Bamum -úta (incl.) CR: Kohumono ado
CN: Berom wo-t, Fyam tí, Kanufi tot, Mada ta, Che tút, Hun-Saare te (excl.), Laru ti, Pongu géétú, Cicipu tú,
Kuteb ti Oko ti- (perf.)
*(bV)ce
Bantu: Oroko ise, Manenguba -ací, Basaa pes, Pinji asé, Mboshi bísí, Lega bíswé, Taita ísi, Nyamwezi iswe, Gogo
ase, Haya icwe, Luyana aci, Nyakyusa-Ngonde uswe, Ndonga tse Other Bantoid: Tiv sé, Noni bese, Kenyang besé, Aghem se (incl.), Duguri sú CR: Kukele bese, Obolo ezi CN: C'Lela cín (dual) Akpes abes
Bantu: Bulu mi-, Mituku mu-, Kamba mu-, Langi mu-, Zalamo mu-, Haya mu-, Luyana mu-, Kwezo mú-, Lungu
mu-, Tumbuka mu-, Matuumbi mu-, Herero mú-, Zulu mu-Other Bantoid: Dong me-re, Meta mbú CN: Tesu mono, Horom min, Yeskwa mí (past), Edoid: Degema ma- / ma-. Uvbie mama Defoid: Igala me, Arigidi amo Ukaan ma / mo
*(bV)nV
Bantu: Nen nú, Viya -anó, Kele enú, Lengola -anú, Bila pénú, Rombo pwé, Sukuma piywe, Gogo abe, Vili benó,
Haya ipwe, Luvale enu, Luba nu- (subject), Herero ene, Copi anu Other Bantoid: Mambila ben, Mesaka ben, Noni ben, Mbe eno, Bamum -ún, Bamileke biné, Lamnso' ven, Bwazza wún
CR: Bekwarra ínen, Kohumono ano, Mbembe boya, Obolo epi, Abua pina
CN: Ayu binin, Izere pín, Irigwe pí, Kulu anuy, Ake anú, Toro anono, Hun-Saare no, C'Lela nwa, Wapan ani,
Kuteb ani Edoid: Urhobo pan
Igboid: Igbo unú, Ikwere anu, Ekpeye yoni Idomoid: Idoma ano, Igede anu Oko ano Akpes abep
Abbreviations
C central fut. future perf. perfective
comp. completive imp. imperative pl. plural
cond. conditional incl. inclusive pos. positive
cons. consonant indir. indirect speech pres. present
dial. dialectal L Lower progr. progressive
dir. direct speech N north(ern) S south(ern)
E east(ern) neg. negative sg. singular
emph. emphatic opt. optative U Upper
excl. exclusive p.c. personal communication W west(ern)
Literature
Abiodun, Michael. 1984. Phonology of Arigidi. B.A. Essay. Ilorin: University of Ilorin.
Adekanye, Fred & Sophie Salffner (eds.). 2007. Ikann Proverbs, Riddles and a Story. Ikakumo-Aworo.
Agoyi, Taiwo Odeyemi. 1997. The category of number and the genetic classification of Ekirdmi. Seminar paper, University of Ilorin.
Akinkugbe, Femi. 1976. An Internal Classification of the Yoruboid Group (Yoruba, Isekiri, Igala). In: JWAL, 11.1-2, pp. 1-19.
Akinkugbe, Femi. 1978. A Comparative Phonology of Yoruba Dialects, Isekiri, and Igala. Ph.D. Thesis, University of Ibadan.
Alex, Inoma. 1989. A Reconstruction of the Segmental Phonology of Proto-Central Delta. M.A. Thesis. Port Harcourt: University of Port Harcourt.
Amayo, Airen. 1980 [1975]. Edo. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 2. West African Linguistic Society.
Armstrong, Robert. 1981. The Idomoid language sub-family of the Eastern Kwa borderland. In: Jungraith-mayr, H. (ed.), Berliner Afrikanische Vorträge. Berlin: Reimer. Pp. 7-33.
Armstrong, Robert. 1983. The Idomoid languages of the Benue and CR valleys. In: JWAL, 13.1, pp. 91-149.
Armstrong, Robert. 1989. Idomoid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 323-336. Lan-ham — New York — London: University Press of America.
Atoyeb!, Joseph Dele. 2008. Morphophonemic Determination of Person Forms in Oko. In: Newcastle Working Papers in Linguistics, University of Newcastle. Vol. 14: Pp. 23-38.
Babaev, Kirill. 2008. Reconstructing Niger-Congo Personal Pronouns I: Proto-Bantoid. In: Journal of West African Linguistics. Vol. XXXV: 1-2. Pp. 131-183. [http://www.nostratic.ru/benue-congo1.pdf]
Bamgbose, Ayo. 1966. A Grammar of Yoruba. In: West African Languages Monographs, 5. Cambridge University Press.
Barnwell, Katharine. 1980. Mbembe. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 2. West African Linguistic Society.
Bennett, P. & J. Sterk. 1977. South-Central Niger-Congo: a Reclassification. In: Studies in African Linguistics 8, 1977: pp. 241-273.
Bergman, Richard. 1976 [1973]. Igede. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Blench, Roger. 1989. Nupoid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 305-322. Lanham — New York — London: University Press of America.
Blench, Roger. 2005a. A Dictionary of Ogba, an Igboid Language of Southern Nigeria. Manuscript [http://www.roger-blench.info].
Blench, Roger. 2005b [1984]. The Ukaan Language: Bantu in South Western Nigeria? Revised paper presented at the Hamburg Conference 'Trends in the Historical Study of African Languages', Hamburg, 3-7 Sep 1984.
Blench, Roger. 2006a. A Dictionary of Ekpeye, an Igboid Language of Southern Nigeria. Manuscript [http://www.roger-blench.info].
Blench, Roger. 2006b. A Wordlist of the Hyam Language of Central Nigeria and Its Affinities. Manuscript [http://www. rogerblench.info].
Blench, Roger. 2006c. Mada Pronouns. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2006d. The Anib (Kanufi) Language of Central Nigeria and Its Affinities. Manuscript [http://www. rogerblench.info].
Blench, Roger. 2006e. The Ayu Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2006f. The Bu Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2006g. The Cara Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2006h. The Horom Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2006i. The Sur (Tapshin) Language of Central Nigeria and Its Affinities. Manuscript [http://www.roger-blench.info].
Blench, Roger. 2006j. The Toro Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2007a. The Ake Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2007b. The Kway Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2007c. A Sketch of Phonology and an Extended Wordlist of TsuVadi, a Kambari Language of Western Nigeria. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2007d. The Eloyi Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2008a. The Idü Language of Central Nigeria. Phonology, Wordlist and Suggestions for Orthography Changes. Manuscript [http://www.rogerblench.info].
Blench, Roger. 2008b. The Nyankpa (Yeskwa) Language of Central Nigeria. Manuscript [http://www.roger-blench.info].
Blench, Roger. (Manuscript a). The West Kainji Language of Northwestern Nigeria: Comparative Wordlist. [http://www.rogerblench.info].
Blench, Roger. (Manuscript b). The Laru People and Their Language. [http://www.rogerblench.info].
Blench, Roger. (Manuscript c). A Dictionary of Echie, an Igboid Language of Southern Nigeria. [http://www.roger-blench.info].
Blench, Roger & Barau Kato. 2007. The Tasu Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Blench, Roger & Bitrus Bulus Kaze. 2006. A Dictionary of the Izere Language of Fobur. Manuscript [http://www.rogerblench.info].
Blench, Roger & Musa Doma. A Dictionary of the Gbari Language. Manuscript [http://www.rogerblench.info].
Blench, Roger & Gya Daniel. 2008. Rigwe Pronouns. Manuscript [http://www.rogerblench.info].
Blench, Roger & al. 1991. Dictionary of the Language of the Bas(s)a People. Manuscript [http://www.roger-blench.info].
Blench, Roger, Ruth Adiwu & Gideon Asukutu. 2006. The Ce (Rukuba) Language of Central Nigeria and Its Affinities. Manuscript [http://www.rogerblench.info].
Bond, Oliver. 2003. Personal Inflection in Eleme. Manuscipt.
Bond, Oliver. 2006. Selected Proceedings of the 35th Annual Conference on African Linguistics, ed. John Mugane et al., 234-244. Somerville: Cascadilla Proceedings Project.
Bond, Oliver & Gregory Anderson. 2006. Divergent Structure in Ogonoid Languages. In: Cover, Rebecca T. & Kim, Yuni, (eds.), Proceedings of 31st Annual Meeting of the Berkeley Linguistic Society: Special Session on Languages of West Africa. Berkeley: Berkeley Linguistics Society.
Bouquiaux, Luc. 1970. La langue birom. Paris: Les belles lettres.
Capo, Hounkpati. 1989. Defoid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 275-290. Lanham — New York — London: University Press of America.
Cassetta, Pete & Lydia Cassetta. 1994. The Ugare Verb System. Manuscript.
Clark, David. 1969. A grammatical study of Ekpeye. Ph.D. Department of Linguistics and Phonetics, SOAS.
Clark, David. 1972. A four-term person system and its ramifications. Studies in African Linguistics, 3: 97-105.
Clark, David. Ekpeye. 1976. Izi. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Connell, Bruce. 1988. Lower Cross Reconstructions. Manuscript.
Connell, Bruce. 1991. Phonetic aspects of the Lower Cross Languages and their implications for sound change. Ph.D. Linguistics. Edinburgh: University of Edinburgh.
Connell, Bruce. 1995. The Lower Cross languages: a prolegomena to the classification of the CR languages. Journal of West African Languages, XXIV 1: pp. 3-46.
Connell, Bruce. Lower Cross Wordlists. Manuscript.
Cook, T.L. 1976 [1968]. Efik. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Cook, T.L. 1971. Report of the Benue-Congo Working Group of the West African Linguistic Society. Actes du 8e Congrès de la Société linguistique de l'Afrique Occidentale. Annales de l'Universite d'Abidjan, Serie H, 2: 559-574, Abidjan: Université d'Abidjan.
Cook, T.L. 1976 [1972]. Kohumono. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Creissels, Denis. 2000. Typology. In: Heine, Bernd and Nurse, Derek (eds). African Languages — An Introduction. Cambridge: Cambridge University press, pp. 231-258.
Cressmann, Esther & Donna Skitch. 1980 [1974]. Duka. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 2. West African Linguistic Society.
Crozier, D. & Roger Blench. 1992. Index of Nigerian Languages (edition 2). SIL, Dallas.
Delafosse, Maurice. 1904. Vocabulaires comparatifs de plus de 60 langues ou dialectes parlés à la Cote d'Ivoire et dans la régions limitrophes. Paris: E.Leroux.
De Wolf, Paul. 1968. Problems in Reconstructing the Noun Class System of the Languages of Plateau 1-a and 1-b. In: Benue-Congo Working Group of the West African Linguistic Society. Leiden Conference Draft.
Dimmendaal, Gerrit. 1978. The Consonant of Proto-Upper Cross and Their Implications for the Classification of Upper Cross languages. Leiden: Department of African Linguistics.
Elimelech, Baruch. 1976. A Tonal Grammar of Etsako. WPP, no. 35. Los Angeles: University of California.
Elugbe, Ben Ohi. 1979. Some Tentatitve Historical Inferences from Comparative Edoid Studies. In: Kiabàrà, vol. 2.1, pp. 82-101.
Elugbe, Ben Ohi. 1980. Reconstructing the Lenis Feature in Proto-Edoid. In: JALL, vol. 2, pp. 39-67.
Elugbe, Ben Ohi. 1983. Noun Class Prefixes in Proto-Edoid. In: PALL, vol. 1, pp. 59-75.
Elugbe, Ben Ohi. 1984. Morphology of the Gerund in Degema and Its Reconstruction in Proto-Edoid. In: SAL, vol. 15, pp. 77-90.
Elugbe, Ben Ohi. 1989. Edoid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 291-304. Lanham — New York — London: University Press of America.
Elugbe, Ben Ohi. 2001. The Classification of Akpes and Aika. Delivered at CALL 31, Leiden.
Elugbe, Ben Ohi & Ologori of Ogori. 1980. Ogori. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 2. West African Linguistic Society.
Emuekpere Masagbor, G.A. 1997. Preverbal Subject Markers in Ivie. Ph.D. Thesis. University of Sherbrooke.
Essien O.E. 1987. The Aspectual System of Ibibio. In: Odden D. (ed.), Current Approaches to African Linguistics. Vol. 4. Pp. 151-165.
Evenhouse, W. Jukun Wukari: a Brief Description in English. Manuscript.
Faraclas, Nicholas. 1984. A Grammar of Obolo. Bloomington: Indiana University Linguistics Club.
Faraclas, Nicholas. 1989. CR. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 377-400. Lanham — New York — London: University Press of America.
Fresco, E. 1968. Two Dialects of Igala and Yoruba: Some Comparisons. In: Research Notes, 4, pp. 32-46. Ibadan: University of Ibadan.
Gardner, Ian. 1976 [1966/67]. Abua. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Gerhardt, Ludwig & Heinz Jockers. 1981. Leksikostatistischen Klassifikazionen von Plateausprachen. In: Berliner Afrikanische Vorträge (Akten des 21 DOT), pp. 25-54.
Gerhardt, Ludwig. 1983. Beträge zur Kenntnis der Sprachen des Nigerianischen Plateaus. Glückstadt: J. J. Augustin.
Gerhardt, Ludwig. 1989. Kainji and Platoid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 359376. Lanham — New York — London: University Press of America.
Greenberg, Joseph. 1963. The Languages of Africa. The Hague: Mouton; IJAL 29.1, Bloomington: Indiana University Press.
Hansford, Keir, John Bendor-Samuel & Ron Stanford (eds.). 1976. An Index of Nigerian Languages. Ghana: SIL.
Hoffmann, Carl. 1976. The Languages of Nigeria by Language Family. In: K. Hansford et al. (eds.), An Index of Nigerian Languages. Ghana SIL. Pp. 169-190.
Hyman, Larry & Daniel Magaji. 1970. Essentials of Gwari Grammar. Occasional Publication, 27. IAS: University of Ibadan.
Hyman, Larry & Imelda Udoh. 2002. Length harmony in Leggbo: A counter-universal? In: Workshop on the Phonology of African Languages (WOPAL), Vienna.
Hyman, Larry, Narrog Heiko, Mary Paster & Imelda Udoh. 2002. Legbo Verb Inflection: A Semantic and Phonological Particle Analysis. In: Proceedings of 28th Annual Meeting of the Berkeley Linguistics Society.
Ikoro, Suana M. 1989. Segmental phonology and lexicon of Proto-Keggoid. University of Port Harcourt: M.A. thesis.
Ikoro, Suana. 1996. The Kana Language. Leiden: Research School CNWS.
Isukul, Caroline. A Grammar of Agholo (Ogbia). Manuscript [http://www.rogerblench.info].
Iwara, Alexander. 1990. A phonemic analysis of long vocoids and phonetic diphtongs in Lokaa. African notes: bulletin of the Institute of African Studies. Ibadan: University of Ibadan, pp. 74-79.
Iwara, Alexander. 2002. The Tonal Phonology and Phonetics in the Future Negative in Lokaa. In: Mugan, J. (ed.), The Linguistic Typology and Representation of African Languages, vol. 5. Africa World Press. Pp. 103-116.
Jungraithmayr, H. 1973. Notes on the Ishe dialect of Ukaan (Akoko, Western State, Nigeria). Africana Marburgen-sia, 6,1:39-57.
Kari, Ethelbert. 2002. Distinguishing between Clitics and Words in Degema, Nigeria. In: African Study Monographs, 23(4), pp. 177-192.
Kari, Ethelbert. 2003. Clitic Doubling: Insights from Degema. In: Journal of Asian and African Studies, no. 66, pp. 235-248.
Kari, Ethelbert. 2004. A Reference Grammar of Degema. Köln: Rüdiger Köppe.
Kari, Ethelbert. 2007. Vowel Harmony in Degema, Nigeria. In: African Study Monographs, 28(2), pp. 87-97.
Koelle, Sigismund. 1854. Polyglotta Africana. London: Church Missionary House.
Koops, Robert. 2007. A Grammar of Kuteb, a Jukunoid Language of East-Central Nigeria. Köln: Rüdiger Köppe. In Press.
Kropp Dakubu, Mary Esther. West African Language Data Sheets. Vols 1-2. West African Linguistic Society. 19761980.
Lewis, M. Paul (ed.). 2009. Ethnologue: Languages of the World, 16th edition. Dallas: SIL International [http://www. ethnologue.com].
Longtau, Selbut. 2008. The Tarok Language: Its Basic Principles and Grammar. Dart-Jos: KWEF.
Longtau, Selbut & Roger Blench. Tarok Dictionary and English-Tarok Finderlist. Forthcoming.
Mackay, V. 1976 [1968]. Eloyi. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Madugu, Isaac. 1975. A grammar of Kwa-type verb serialisation; its nature and significance for current generative theory. Ph.D. thesis. Los Angeles: UCLA.
Madugu, Isaac. 1979. The Nupe verb and diachrony. In: Ibadan Journal of Humanistic Studies, 1, pp. 74-97.
Madugu, Isaac. 1985. Complex verbs in Nupe and Yoruba. SAL, 16:3, pp. 295-321.
Manfredi, Victor. 1989. Igboid. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 337-358. Lanham — New York — London: University Press of America.
McGill, Stuart. A Cicipu / English / Hausa Lexicon. [http://www.cicipu.org /dictionary.html]
Meeussen, Achille. 1965. Reconstructions grammaticales du bantou. Tervuren: Musée Royal de l'Afrique Centrale (MRAC).
Meier, Paul. 1976. Izi. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Nettle, Daniel. 1998. Materials from the South-Eastern Plateau Languages of Nigeria. In: Afrika und Übersee, Band 81. Pp. 253-279.
Ohiri-Aniche, Chinyere. 1991. A reconstruction of Proto-Igboid-Yoruboid-Edoid. Ph.D. Thesis. Port Harcourt: University of Port Harcourt.
Ohiri-Aniche, Chinyere. 1999. Language Diversification in the Akoko Area of Western Nigeria. In: Roger Blench, Matthew Spriggs (eds.). Archaeology and Language IV: Language Change and Cultural Transformation, pp. 79-94. Routledge.
Omamor, Augusta. 1976. The Place of Isekiri in Relation to Yoruba. In: Research Notes, vol. 7, nos. 1-2. Ibadan: University of Ibadan.
Omamor, Augusta. 1988. Okpe and Uvwie: a Case of Vowel Harmony Galore. In: Journal of West African Languages, 18.1, pp. 47-64.
Oyama, Ebinda, & Katharine Barnwell. 1985/1995. Mbembe-English Dictionary. Manuscript.
Philpot, W. T. A. 1935. Notes on the Igala Language. In: Bulletin of the School of Oriental and African Studies, 7, pp. 897-912.
Preston, B. 1854. A Grammar of the Bakêle Language, with Vocabularies. New York: J. P. Prall.
Priest, Perry. Jibu Dictionary, ed. by Roger Blench. Manuscript. [http://www.rogerblench.info]
Rikoto, B., P. Senchi, Steve Dettweiler & Sonia Dettweiler. 2002. K'Batksa: C'Lela—C'Turanci—C'Gana Dictionary. Leina Language Development / Translation Association.
Rowland Oke, Mary. 2003. Description systematique de la langue obolo-andoni. Paris: L'Harmattan.
Scholz, Hans-Jürgen. 1976 [1973]. Igbirra. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Vol. 1. West African Linguistic Society.
Segerer, Guillaume. 2002-2007. Les marques personnelles dans les langues africaines, [base de données en ligne]. http://sumaie.vjf.cnrs.fr/pronoms.
Segerer, Guillaume. 2008. Les langues Niger-Congo. In: Желтов А. Ю. [Zheltov, A. Y.] 2008. Языки нигер-конго: структурно-динамическая типология. СПб: Издательство Санкт-Петербургского университета. [The Niger-Congo Languages: Structural and Dynamic Typology. Saint-Petersburg: Saint-Petersburg University.]
Shimizu, Kiyoshi. 1971. Comparative Jukunoid: an Introduction. Ph.D. thesis. Ibadan: University of Ibadan.
Shimizu, Kiyoshi. 1971-1980. Comparative Jukunoid, vois 1-3. Vienna: Afro-Pub.
Shimizu, Kiyoshi. 1975. A Lexicostatisticai Study of Plateau Languages and Jukun. Anthropological Linguistics, 17, pp. 413-418.
Shimizu, Kiyoshi. 1996. A Kulu vocabulary and fragments of Kulu grammatical structures. AFLANG [http://www. aa.tufs.ac.jp/~P_afiang/].
Sibomana, Leo. 1981/82. Tarok III: Das Verbaisystem und der Satz. In: Afrika und Übersee, Band 64, 2, pp. 237-247.
Sibomana, Leo. 1985. A Phoiogicai and Grammaticai Outiine of Eggon. In: Afrika und Übersee, Band 68, pp. 43-68.
Siewierska, Anna. 2004. Person. Cambridge: Cambridge University Press.
Smith, Neii. 1980 [1967]. Nupe. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Voi. 2. West African Linguistic Society.
Stanford, B. 1976 [1973]. Bekwarra. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Voi. 1. West African Linguistic Society.
Sterk, Jan Pieter. 1977. Elements of Gade Grammar. Doctorai dissertation, University of Wisconsin, Madison.
Sterk, Jan Pieter. 1979. Fortis / Lenis Contrast in Upper Cross Consonants: a Survey. In: Kiabàrà: 2.2: 57-68.
Stofberg, Yvonne R. 1978. Migiii grammar. In: Language Data, African Series, 12. Daiias: Summer Institute of Linguistics.
Storch, Anne. 2005. Jukun Pronominai Systems from an Areai Perspective. Paper presented at the Jukun Conference, Vienna, 18-20 Nov 2005.
Strub, E. 1915-1916. Essai d'une grammaire de ia iangue kukuruku. In: Anthropos, voi. 10, pp. 450-465; voi. 11, pp. 888-907.
Thomas, Eiaine. 1976 [1968]. Engenni. In: Kropp Dakubu, M.E. (ed.) West African Language Data Sheets. Voi. 1. West African Linguistic Society.
Thomas, Eiaine & Kay Williamson. Wordiists of Deita Edo: Epie, Engenni, Degema. In: Occasional Papers 8, p. 105. Accra: Institute of African Studies, University of Ibadan.
Ukere, Anthony. 2005 [1986]. Urhobo-English Dictionary. Benin City: Eiupeju Press.
Welmers, Wiiiiam. 1949. Speaking Jukun (Takum Dialect). Lupwe.
Westermann, Diedrich. 1926. Das Edo in Nigerien. In: MSOS, voi. 29.3, pp. 32-58.
Westermann, Diedrich, & Margaret Bryan. 1952. The Languages of West Africa. Handbook of African Languages 2. London: Oxford University Press.
Williamson, Kay. 1971. The Benue-Congo Languages and Ijo. CTL: 7, pp. 245-306.
Williamson, Kay. 1973. Benue-Congo Comparative Wordlist, vol. 2. Ibadan: West African Linguistic Society.
Williamson, Kay. 1985. How to become a Kwa language. In: Linguistics and Philosophy. Essays in Honor of Ruben S. Wells. Eds. A. Makkai and A. Melby. Current Issues in Linguistic Theory, 42. Benjamins, Amsterdam.
Williamson, Kay. 1988. Linguistic evidence for the prehistory of the Niger Delta. The early history of the Niger Delta, edited by E.J. Alagoa, F.N. Anozie and N. Nzewunwa. Hamburg: Helmut Buske Verlag.
Williamson, Kay. 1989. Niger-Congo Overview. In: The Niger-Congo languages, ed. by John Bendor-Samuel, pp. 345. Lanham — New York — London: University Press of America.
Williamson, Kay & Roger Blench. 2000. Niger-Congo. In: Heine, Bernd and Nurse, Derek (eds). African Languages — An Introduction. Cambridge: Cambridge University press, pp. 11-42.
Williamson, Kay & Kiyoshi Shimizu. 1968. Benue-Congo Comparative Wordlist, vol. 1. Ibadan: West African Linguistic Society.
Williamson, Nwinee, Kay Williamson & Roger Blench. 2007. A Dictionary of Kana, an Ogoni Language of the Niger Delta, Nigeria. Manuscript.
Wilson Kpakpan Nwi-Bari. 2002. Tee-English Dictionary. Port Harcourt: Onyoma Research Publications.
Wolff, Hans. 1969. A Comparative Vocabulary of Abuan Dialects. Evanston: Northwestern University Press.
Бабаев К. В. 2010. О реконструкции личных местоимений в праязыке ква // Вопросы языкознания, № 4. [Babaev, Kirill. On the Reconstruction of Personal Pronouns in Proto-Kwa. In: Voprosy yazykoznaniya, no. 4, 2010. Moscow.]
Фихман Б. С. 1975. Язык игбо. М., «Наука». [Fikhman, B. S. The Igbo Language. Moscow: Nauka.]
Статья представляет собой вторую, завершающую часть сравнительного исследования систем личных показателей в языках бенуэ-конго, начатого в [Бабаев 2008] (работа доступна в он-лайн режиме). Первая часть содержала общий обзор работ по данной тематике, существующих на сегодняшний день, а также опыт реконструкции системы личных показателей в прабантоидном языке. Вторая часть тематически развивает предыдущую; в ней собраны данные по всем остальным ветвям семьи бенуэ-конго и сделан первый шаг на пути к реконструкции прабенуэ-конголезской системы личных показателей.
