N-initial nouns in Landuma and their counterparts in Mande

Describes a group of kin terms in Landuma which have a non-standard phonological structure: they begin with the consonant cluster NC. It is shown that the properties of these nouns can be explained via their origin: all of them are borrowed from Mande.

48,3 K

. ,

, , , , .


N-initial nouns in Landuma and their counterparts in Mande

Nina Sumbatovaf, Valentin Vydrirt

+ Institute of Linguistics, Russian Academy of Sciences;

t LLACAN (CNRS), Paris / INALCO, Paris / St. Petersburg State University;

This paper describes a group of kin terms in Landuma (a Mel language spoken in northwestern Guinea) which have a non-standard phonological structure: they begin with the consonant cluster NC. It is shown that the anomalous properties of these nouns can be explained via their origin: all of them are borrowed from Mande.

In Western Mande languages, nouns for elder kin are also anomalous in that they are often unable to adjoin a definite or referential article. It has been suggested previously that this anomaly could be explained by the presence of an archaic nasal prefix, a grammatical marker of elder kin. At the same time, such a nasal prefix is not attested in any modern Mande language.

Two hypotheses can be advanced on the origin of the initial nasal element in the anomalous Landuma nouns. According to the first, this element goes back to a nasal prefix reconstructed for nouns referring to elder kin in Mande. If so, the Landuma data can be regarded as an argument for the relatively recent disappearance of this prefix in Mande (i.e., subsequent to the start of close contact between speakers of Proto-Landuma and speakers of Proto- Manding and Proto-Susu). Alternatively, the nasal element can be regarded as a reinterpreted Mande 1SG pronoun n which, in its possessive function, appears frequently with kin terms. It cannot be excluded that both sources may have been relevant.

Keywords: Landuma language; Mel languages; Western Mande; kinship terms; language contact.

.. , .. .

( ), - . : NC. , : .

, . , , . .

. , , . , ( , -). 1 , . , .

: ; ; ; .

General information on Landuma

Landuma is a language of the Mel family spoken by about 30 000 people (our estimate) in the north-west of the Republic of Guinea, in the vicinity of the city of Boke. The Landuma data analyzed in this paper was obtained in the course of fieldwork in Guinea in 2015-2018. A writing system for Landuma was elaborated by Kirk Rogers (Rogers 2005; 2008).

The phonological system of Landuma includes three front vowels i, e, e; four back vowels u, o, , ; and two central vowels d, a. The consonants are represented in Table 1. In this paper we use an IPA-based phonological transcription (with modifications typical for African linguistics), rather than the practical transcription introduced by Kirk Rogers.

Table 1. Landuma consonants







Voiceless plosives



k The sounds [k] and [g] are allophones of the same phoneme k. [k/g]

Voiced plosives















Oral sonorants


r, l


Landuma has a relatively simple agglutinative morphology, mainly in the verbal domain. The verbs have a rich derivation system (reflexive, reciprocal, instrumental, etc.). They inflect for TAM categories and agree with the subject in person/number, animacy/ noun class (see Section 2). Landuma is a right-branching SVO language.

Nominal morphology and agreement

The only regular grammatical affixes on the noun are nominal prefixes. These prefixes, which are present on most nouns, at first sight function as class markers. They show a number opposition between singular and plural (see Table 2). Some prefixes are also associated with other semantic features: for example, the prefixes ta-/ma- (SG/pl) express diminutivity, pa-/ na- (SG/pl) are augmentative, the pair wa-/a- is characteristic of human nouns, and deverbal nouns are always marked by the prefix ka-. Other prefixes have no obvious semantic value, besides signaling number.

Table 2. Nominal prefixes in Landuma






we-cames / a-cames `merchant'



dA-loko / se-loko `day'


dA-sek / yA-sek `tooth'



ke-babu / ce-babu `maize'


ke-ca / yA-ca `hand'



-bat / yA-bat `river'



tA-sar / mA-sar `small stone' (diminutive)



pA-sar / nA-sar `big stone' (augmentative)

At the same time, Landuma has a significant number of nouns that bear no prefix in the singular Every plural noun has a prefix.. Prefixes are absent in most recent borrowings, which are mainly from French: farip `flour' < French farine, plas `place' < French place, etc. Most proper names, both personal names and place names, also have no prefix: Barlande `Barlande' (village name), Fatu `Fatu' (woman's name). Finally, and most importantly, 12 to 20% of other nouns are prefixless without there being any obvious reason for this, cf. bumbi `hare', dis `body', gbundo `secret', jombo `hyena', kas `father', nenc `fire'. These nouns form their plurals by adding one of the plural prefixes: bumbi `hare' / PL ya-bumbi, dis `body' / pl sa-dis, gbundo `secret' / pl ya-gbundo or sa-gbundo, jombo `hyena' / pl ya-jombo, kas `father' / pl a-kas, nenc `fire' / pl ya-nenc.

The most intriguing property of Landuma nouns concerns how they control agreement. Agreement is found within the noun phrase (adjectives, demonstratives, the numeral `one' and some other words agree with the nominal head) and within the clause (the verb agrees with the subject). The choice of anaphoric pronouns generally follows the same rules as agreement proper.

Agreement in Landuma is basically agreement in animacy, but there is also a kind of agreement based on the phonology of the controlling noun. The basic (and somewhat simplified) agreement rule is as follows:

(1) a. semantic agreement: animate nouns trigger special animate prefixes (example 2abc). b. phonological (radical alliterative) agreement: the first consonant of the agreement prefix repeats the first consonant of the controller; if the controller begins with a vowel, the agreement prefix begins with q- Radical alliterative agreement is a typologically rare phenomenon found in the Kru language family (Sande 2019), the Arapesh languages of New Guinea (Dobrin 2012), and a few others., example (3abcd). nouns landuma mande phonological

(2a) jombo wdk-in

hyena AN-one

`one hyena'

(2b) knrmoko wdk-in teacher AN-one

`one teacher'

(2c) wd-be wdk-in

NP-king AN-one `one king'

(3a) tn-ldr t-in

NP-finger AGR-one `one finger'

(3b) lokuq l-in

week AGR-one

`one week'

(3c) gbundo gb-in secret AGR-one

`one secret'

(3d) n-sar q-in

NP-stone AGR-one `one stone'

The rules seen in (1) are strictly observed by all inanimate nouns and by human nouns with the prefix wd-. At the same time, there are some groups of nouns that combine semantic and phonological agreement and/or fluctuate between these two options in certain constructions.

Nouns denoting animals trigger animate prefixes on verbs and anaphoric pronouns (4), but, most usually, alliterative prefixes on the agreement targets within a noun phrase Within the noun phrase, some fluctuations between non-alliterative and alliterative agreement markers are attested. (5ab).

(4) Jombo n-yup le wd-sontle le do, dor dn-mop ko le,

hyena lPRS-turn.out asr 3SG.AN-run.fast asr there hunger AGR-catch him/her asr

wd-c-ko kd-ko dade qkon fdna, wd-sarn.

3SG.AN-C0NS-go NP-go village (s)he.suBj also 3SG.AN-carry.baggage

`(it turned out that) the hyena was running very fast, (after a while) it felt hungry, it was running and (finally) came to the village. it, too, was carrying its baggage.' [oral text]

(5a) jonbo ja-bi

hyena AGR-black `black hyena'

(5b) jonbo j-in or jonbo wak-in hyena AGR-one hyena AN-one

`one hyena'

The agreement behavior associated with prefixless animate nouns is less predictable: they allow phonological agreement in certain constructions, but ultimately their agreement varies across speakers.

Most human nouns without the prefix wo- (knrmoko `teacher', knrnndi `student', bobo `deaf- mute', imamu `imam'; diminutive and augmentative nouns referring to people like tnncay `little girl') allow alliterative agreement within the NP (6ab) but require semantic (animate) agreement markers on verbs (6c) and are antecedents of animate anaphoric pronouns.

(6a) knrmoko kd-tot or (6b) knrmoko wo-tot

teacher AGR-good teacher AN-good

`good teacher'

(6c) knrmoko wd-n-der le

teacher 3sg.AN-FACT-come ASR `The teacher came.'

A special group of animate nouns will be considered in the next section.

A special group of animate nouns (N-nouns)

Presentation of the N-nouns. In this section, we will turn to the main topic of this paper, namely a small group of animate nouns (henceforth "N-nouns") which are in several respects non-typical of Landuma. The full set, according to the dictionary (Rogers & Bryant 2012), includes nine nouns referring to close social relations and kinship terms: nna `mother', mbarirj `friend', mbenba `ancestor', ncoko `uncle, mother's brother', njatiki `host', ntana `father-in-law, mother-in-law', ntarn `elder sibling', ntener `aunt, father's sister', ntokma `namesake'; one noun referring to an animal: mbdrfi `wild boar'; and, finally, two proper nouns (Nfasori `Infasori', Nfnli `Enfali') and also Nkiln, the title of the prophet Muhammad (Nkiln Mnhnmndu `the prophet Muhammad').

All N-nouns begin with a consonant cluster of the type NC (where N is a nasal consonant). These are the only words in Landuma that begin with a consonant cluster.

The initial nasal consonant in all N-nouns agrees with the subsequent consonant in place of articulation. In the Landuma orthography elaborated by Kirk Rogers (2005; 2008), it is represented by the letter n in all cases.

Landuma has four nasal consonant phonemes: /m/, /n/, //, and /r/. The palatal consonant // is rare; it is only attested word-initially before a vowel, as in pnmnne `time' (certainly an Arabic borrowing). The consonants /m/ and /r/ are allowed in various positions; in particular, they are found before consonants (although infrequently, and never word-initially) and do not undergo assimilation: wakomsi `midwife'; cayne `carry on the head'. Finally, the consonant /n/ is allowed in various positions, but when it precedes an obstruent it undergoes place assimilation: /nb/ is realized as [mb], /nk/ as [pk], etc. For example, the factative marker n- is realized as [m] before labial consonants and as [p] before velars:

(7a) Wd-m-bnr le m-nlo m-i.

3SG.AN-FACT-add asr NP-rice agr-def `(S)he added some rice.'

(7b) Fatu wd-y-kudi mi le.

Fatu 3SG.AN-FACT-greet me ASR `Fatu greeted me.'

This means, in particular, that three different nasal sounds ([m], [n], [p]) can be met before dental and alveolar consonants but [n] is impossible before labial and velar consonants. In the latter case, the phonological oppositions /m/ vs. /n/ and /n/ vs. /^/ are neutralized Different phonological interpretations of the initial nasal consonants followed by another consonant can therefore be proposed, and we do not plan to discuss the issue in more detail here.. Henceforth, we will use a capital N to encode the initial nasal consonant in nouns with an initial consonant cluster (for example, Nbenba `ancestor', Ncoko `uncle', etc.).

Morphology of N-nouns

Postulating a common consonant in a group of semantically close nouns immediately raises the question: does this consonant represent a prefix or a part of the noun stem? Identifying it as a prefix does not seem to be a good decision in our case. For example, the plural forms of the N-nouns feature the plural prefix a-, which is quite usual for animate nouns: Ntana `father-in-law, mother-in-law' -- PL aNtana, Ntarn `elder sibling' -- pl aNtarn. One of our language consultants also allowed forms with the unique prefix ara-: araNtana, araNtarn. In any case, the plural forms of the N-nouns are invariably formed by adding a prefix to the singular form, which is quite normal for prefixless nouns. All prefixed nouns form their plurals by replacing the singular prefix, cf. the examples in Table 2.

Another important factor is agreement. As described in Section 2, prefixless animate nouns constitute a zone in which agreement patterns are fluctuating or mixed. This is not the case with the N-nouns (including the animal noun Nbdrfi `wild boar'), which invariably trigger animate agreement markers:

(8a) Ncoko qon uncle AN.DEF `the uncle'

(8b) Ncoko w-n uncle AN-that `that uncle'

(8c) Ncoko k-on

uncle AN-3SG.POSS `his/her uncle'

(8d) Ncoko k-a Mamadu uncle AN-POSS Mamadu `the uncle of Mamadu'

(9a) Nbarfi wak-in

wild.boar AN-one `one wild boar'

(9b) Nbarfi wa-paq

wild.boar AN-big `a/the big wild boar'

One of our language consultants allowed phonological agreement and used the consonants [m] and [n] in the agreeing units. This suggests that we ought to interpret the initial clusters of the N-nouns as representing different nasal phonemes. However, this data is not fully reliable and was not endorsed by other speakers.

Mande cognates of N-nouns in Landuma

It turns out that practically all N-nouns in Landuma have cognates in Mande languages. Let us consider them one by one.In the following list, when a Mande word is cited without gloss, its meaning is the same as the meaning of

the Landuma word.

1) Landuma Nna `mother'

Proto-Manding *'na, Mandinka (Creissels, Jatta & Jobarteh 1982; Anonym 1995) naa `mother' (address form), Xasonka (Tveit & Dansoko 1993) na `mother; mother's sister', Kita Maninka (Creissels 2009: 55) na, Maninka 'na (without the tonal article) `mother', Bamana n `mother, mummy' (address form)

Bozo-Soninke: Tieyaxo (Anonyme 1982) nan `mother'; (in combination with the name of a fruit) `fruit tree', Sorogama (Daget, Kanipo & Sanankoua 1953) na `mother'; (in combination with the name of a fruit) `fruit tree'

Bobo (le Bris & Prost 1981) n

Samogo: Duun (Hochstetler 1996) na, Dzuun (Solomiac, Traor & Traor 1998) na

Proto-Eastern (Schreiber 2008) *da / *n, Boko (Jones 2004) d

Southern Mande: Dan-Blo (Erman & Loh 2008) d, Dan Gwtaa, Kla-Dan d `grandmother, great-grandmother; elder paternal aunt; father's or mother's elder brother's wife; mother's elder co-wife; mother-in-law (for a man)', Tura (Idiatov ms.) d `grandmother; elder paternal aunt; father's or mother's elder brother's wife; elder sister-in-law; mother-in-law, grandmother-in-law', Guro (Kuznetsova & Kuznetsova) daa, Yowre (Hopkins 1982) daa, Mwan (Perekhvalskaya & Yegb 2018) d-le `mother-in-law (for a woman)', Wan (Nikitina) n `mother', Ben (Paperno) na `mother'

2) Landuma Nbariq `friend'

Southwestern Mande: Liberian Kpelle (Leidenfrost & McKay 2007) malq `sororal nephew'

Soso (Anonyme n.d.) barn (?) kin; intimate friend, Jalonke (Creissels 2010) bari-mx (?) `kin', bari-du (?) `child; compatriot'

Jeri (Trbs 1998) bli `uncle'

Mokole: Lele (Vydrin 2009a) brin, byin, Lele (Mara & Camara 1979) bayen, Koranko (Kastenholz 1987a) br

Manding: Mandinka (Creissels, Jatta & Jobarteh 1982) bariq, barimma, Nyokolo Maninka (Meyer 1983) bariq, Xasonka (Tveit & Dansoko 1993) bariq, Kita Maninka (Creissels 2009) barin (no article), Kagoro (Vydrine 2001) bri-no `nephew, niece (sister's child)', Maninka barin, bari, brin, brin `maternal uncle; maternal kin'

3) Landuma Nbenba `ancestor'

South-Western Mande: Mende (Innes 1969) mbmb `lineage, family'; Guinean Kpelle

(Leger 1975) bomo-kolo, beme-kolo `ancestor (from the same clan)'

Susu (Dian & Vydrine 2012) bnb, bnb `ancestor, forefather'; `grandfather; greatgrandfather'; Yalunka bnb-n

Mokole: Lele (Vydrin 2009b) bnba `grandfather', Koranko (Kastenholz 1987a) bnba `ancestor, forefather'

Manding *bnbaa `ancestor': Mandinka bnbaa, Xasonka, Guinean Maninka bnba, Bambara bnba

Bobo (le Bris & Prost 1981) bema `ancestor'

South Mande: Eastern Dan bba, Tura (Idiatov) bba

4) Landuma Ncoko `maternal uncle'

Susu soxo `maternal uncle', `sororal nephew', Jalonke (Creissels 2010) soqo, Yalunka soxo

5) Landuma Njatiki `host'

Mende (Innes 1969) yl (< Manding)

Susu (Willits n.d.) ytigi `protector', Jalonke (Creissels 2010) jtgi, Yalunka (Willits n.d.) yatigi

Mokole: Lele (Vydrin 2009a) yatii, Lele (Mara & Camara 1979) yy Manding: Kita Maninka (Creissels 2009) jtigi, Maninka jtii, Bamana jtigi Soninke-Bozo: Bozo-Tigemaho (Anonyme 1982) jadi, Bozo-Sorogama (Daget, Kanipo & Sanankoua 1953) jatigi

6) Landuma Ntana `father-in-law, mother-in-law'

Manding: Maninka tna `totem; nuisance', Bambara tne `taboo, interdiction'

7) Landuma Ntarn `elder sibling'

Soso (Dian & Vydrine 2012; Tour 1989) tr, -0, Jalonke (Creissels 2010) tr `elder sister' Mokole: Kakabe (Vydrina 2015) tata `elder sibling', Mogofin (Polinder, Janse & van Linden 2009) tat `elder sister'

Manding: Mandinka (Creissels, Jatta & Jobarteh 1982; Anonym 1995) tataa `elder sibling; husband', Xasonka (Tveit & Dansoko 1993) tata `elder sibling'

Soninke (Galtier & Dantioko 1979; Smeltzer & Smeltzer 2001) taata

8) Landuma Ntenerj `aunt, father's sister'

Southwestern Mande: Bandi (Grossmann, Rodewald & Covac 1991) tn `aunt'; Guinean Kpelle (Konoshenko 2019) tlr `paternal aunt; any elder woman of the paternal aunt's family' Susu (Dian & Vydrine 2012) tnn `paternal aunt'

Vai (Welmers & Kandakai 1974) tn `maternal aunt'

Mokole: Lele (Vydrin 2009a) tne `paternal aunt', Koranko (Kastenholz 1987a) tene `paternal aunt'

Manding: Kita Maninka (Creissels 2009) tnen `paternal aunt' (no article), Guinean Maninka tenen, Bambara tene `paternal aunt'; Segu Bambara, Beledugu Bambara, Kaarta Bambara `aunt' (either paternal or maternal)

Bozo-Tigemaxo (Anonyme 1982) tayen `paternal aunt'

South Mande: Tura (Idiatov) tee `paternal aunt'

9) Landuma Ntokma `namesake'

This is a borrowing from Manding, where the form *toGo-ma is morphologically transparent: *toGo `name' + -ma, a suffix of mutual kinship relation. The word also appears in many other Mande languages (where it can be also regarded as a Manding loan):

Southwestern Mande: Liberian Kpelle (Leidenfrost & McKay 2007) toom, Guinean Kpelle (Konoshenko 2019) towe, tYe

Susu (Willits n.d.) txm, (Dian & Vydrine 2012; Tour 1994) txmn, Jalonke (Creissels 2010) tqm

Southern Mande: Western Dan (Erman & Loh 2008) t, Eastern Dan tvvq, Kla-Dan (Makeeva ms.) turj

There is also a similar Soninke form toxora where the function of the final element -ra is not quite clear.

10) Landuma Nbdrfi `wild boar'

Susu (Willits n.d.) bti `pig' may have a common origin with the Southwestern Mande forms: Looma boi-g, bo-g, bu-g, Liberian Kpelle (Leidenfrost & McKay 2007) , Guinean Kpelle (Konoshenko 2019) (there are some less reliable forms in other Mande languages which may also be cognates). The final element -fi is unclear (however, it may be comparable to Manding fin `black', in which case the form would mean `black pig').

11) Landuma Nkiln `title of the prophet Muhammad'

Most probably, a borrowing from Manding, where *ki means `send' and *-la is an agentive suffix, giving *kila `messenger'. This word was borrowed into many modern Mande languages.

Manding: Mandinka (Creissels, Jatta & Jobarteh 1982) kiilaa `messenger; prophet', Guinean Maninka kila, kla, cla `messenger', Bambara kira `prophet' (with an irregular form of the suffix)

Southwestern Mande: Bandi (Grossmann, Rodewald & Covac 1991) kel `messenger', Looma kela, keela `messenger', Liberian Kpelle (Leidenfrost & McKay 2007) kl `messenger', Guinean Kpelle (Konoshenko 2019) kl, kl

Susu (Willits n.d.) xr `messenger', kiil `Prophet' (the latter form is probably borrowed from Manding).

Both proper nouns belonging to this group, Nfasori `Nfasori' and Nfnli `Nfali', may also be borrowed from Mande, and more precisely from Manding, where F Sori may be a honorific form of the male name Sori (f means `father'), and Fli may have been a heathen name meaning `donkey'.

1.1. N-nouns in the context of Landuma-Mande language contacts

The history of the Landuma (and their closest relatives the Kogoli The Kogoli are an ethnic group in the area of Kumbia in north-western Guinea and adjacent areas of Guinea-Bissau, speaking a language that is, reportedly, closely related to Landuma. Unfortunately, practically no data on their language is available; the rare publications on the Kogoli that do exist (Suret Canal 2000; Ferry &

Sande 2000) provide almost exclusively ethnohistorical data.) is characterized by very close contacts with Mande people. The earliest written sources mentioning them date back to the beginning of the 16th century, and they appear in these documents as subjects of the king of Manding (Suret Canal 2000: 334). One of the main trade routes connecting the medieval Mali Empire led from Siguiri to Boke (Iffono 2000), and there is even an oral tradition that places the origin of the Kogoli in Siguiri, a Maninka town (Suret Canal 2000: 336-337). Historical documents tell us less about contacts of the Landuma with the Susu and Jalonke. However, it can be taken for granted that these contacts continued for centuries, with the result that today most Landumas are bilingual in Susu. Moreover, the Mogofin people are immediate neighbours of the Landuma, therefore some contact between the two languages is to be expected.

The N-nouns are certainly not the only group of nouns borrowed from Mande. However, the initial N is not found in other borrowed nouns. In particular, it is absent in the borrowed kin term dimbore `cousin', from Susu dinbore (Raimbault 1923; Lacan 1942). This term appears in older sources on Susu (Raimbault 1923; Lacan 1942). In more recent ones we find another term for `cousin', dxo. Landuma also has three borrowed kin terms that begin with a nasal consonant, but not with the NC cluster: mama `grandfather, grandmother'; nnnde `father's second wife', and nimoko `spouse of one's elder sibling; younger sibling of one's wife'. These nouns do not show any phonological differences from other Landuma nouns (whether borrowed or not).

This shows that the peculiarities of N-nouns cannot be explained by the simple fact that these nouns are borrowed: other borrowed nouns (including certain kin terms) do not belong to this group. More than that, in the modern Mande languages which represent the most likely sources of the borrowings into Landuma -- that is Susu, Jalonke/Yalunka, Maninka and Mogofin -- the cognate words have no initial element N-. In what follows, we consider two alternative sources of the initial consonant cluster in these nouns.

N- as a prefix for elder kin terms in Western Mande languages

According to the first hypothesis, the presence of the nasal element at the beginning of kinship terms in Landuma can be explained by the fact that the Mande donor languages had, in earlier periods of their existence, a nasal prefix marking exactly this semantic group of words. A reconstruction of this prefix was advanced in (Vydrin 2006). Since this paper is available only in Russian, let us briefly survey the morphological peculiarities of West Mande languages that provide grounds for this reconstruction.

Elder kin in Manding languages

As mentioned in (Spears 1972), most terms for elder kin in Guinean Maninka are incompatible with the tonal article (which is normally represented by a floating low tone following the noun): 'n `mother', f `father', tn `paternal aunt', borin `maternal uncle', 'ma, mm `grandmother', bnba `grandfather, ancestor', k `husband'. Maninka texts written in Nko (where tones are accurately marked) confirm the absence of the articles with these words. Spears also points to the fact that 'n `mother' and 'm `grandmother' have a preceding floating low tone.

In some Kagoro dialects, at least certain terms (f `father', babaa `father', kdto or qodoke `elder brother', bidan `in-law') are also used without the article (Vydrine 2001: 104, 121, 128).

Elder kin in Koranko and Susu

In these languages, nouns appear in most cases with a suffix -- or -. In Koranko this is a definite or specific article, and in Susu it has evolved into a nominal morpheme.

However, in both these languages, there is a group of nouns that cannot attach the suffix.

In Koranko (Kastenholz 1987b: 206), this group includes: bo `comrade' (age-mate?), diyenamoo `friend', kramoo `teacher', tooma `namesake', tn `paternal aunt', br `maternal uncle', bnba `grandfather, ancestor', f `father', n `mother', numo `younger brother-in-law'.

In Susu (Tour 1989), there are several groups of nouns that appear without the nominal morpheme -i: French and Arabic loans; some shifters; a few names for biological species; and, finally, kinship terms and some other words for social relations: ng `mother', bb `father', soxo `uncle', mm `grandmother', bnb `grandfather', tnun `grandfather', tnn `paternal aunt', nndn `mother's co-wife', tr `elder sibling', xuny `younger sibling', dx `cousin', nmx `younger brother-in-law', txmn `nickname', kl `friend; lover', ytigi `host'. I should be noted that ng `mother' is the only noun in Susu with an initial prenasalized consonant.

Elder kin in Southwestern Mande

All the languages of this group are characterized by the phenomenon of initial consonant alternation. As a rule, each content word (noun, verb, adjective) has two forms with different initial consonants. In most languages of the group, there are some consonants which stand outside the consonant alternation system. They will not be considered here. These forms appear in different syntactic contexts. The initial consonant alternation in Southwestern Mande has been widely discussed in the specialist literature; see, among many other publications, (Dwyer 1974; Dwyer 1986; Kastenholz 1997: 100-104, 125137; Vydrin 2006: 100-114). Historically, the trigger of this consonant alternation is a preceding nasal element. This element can be a syllable coda, as in the word ms(y) `chief' in (10b), or a syllabic nasal */- representing a grammatical morpheme: a 3SG pronoun, as in (11b), or a referential article (12b) going back to the same 3SG pronoun. In all Southwestern Mande languages, there are at least two articles: the referential article */-, going back to the 3SG pronoun, and the suffix -i, a definite article, which most probably goes back to a demonstrative pronoun/determinative. The latter is very likely to be etymologically identical with the article -i/-e in other Western Mande languages as discussed above; the former is specific to the Southwestern Mande languages. In what follows, the alternant appearing after the nasal element will be referred to as strong, and the other as weak.

Bandi (Rodewald 1989: 30)

(10a) ph lk

ref \ woman hand

`woman's hand', historically *ph tk.

(10b) ms tk

ref \ chief hand

`chief's hand', historically *msy tk.

Mende (Innes 1971: 146)

(lla) Ngi t l-.


`I have seen them', historically *Ngi t t-.

(llb) Ngi t-.


`I have seen it', historically *Ngi it-.

Mende (Innes 1971: 36)

(12a) P vnd-i m.

ref \ European cotton-DEF be

`That is the European cotton', historically *N-p fnd-i m.

(12b) Fnd - m.

REF \cotton-DEF be

`That is the cotton', historically *lN-fnd-i m.

At the same time, in Mende, Loko, and Bandi there is a group of nouns whose initial consonants, contrary to expectations, do not undergo alternation: in any context, they appear with a strong initial consonant. This group includes nouns for elder relatives and some semantically close words, e.g.:

Bandi (Covac 1978: 20)

(13) ni ky

1sg.poss father

`my father' (rather than *ni yy)

(14) ng nj

3sg.poss mother

`his mother' (rather than *ng y)

These nouns are given here (the lists may be incomplete).

Loko: ky `father', kep.a `maternal uncle', nje `mother', ndy `elder sibling'.

Mende: kk ~ k `father', kp `maternal uncle', nje `mother', ngoo `elder sibling', ndw ~ nd `brother', ndimo `friend', mba `age-mate'.

Bandi: ky `father', k `maternal uncle', nj `mother', nd `mother' (address form), nd ~ ndy `elder sibling', mm `grandmother',12 kwl `grandfather', tn `paternal aunt', ndmo `friend', mbl `age-mate'.

These words usually appear without the definite article -i (the available data is insufficient to show whether they are compatible with the article in principle).

As we can see, the forms of the terms for elder relatives look as if they constantly appear with the referential article *r), even in contexts where the referential article would not normally be expected. In (Vydrin 2006: 139) it was suggested that we have here an archaic noun prefix *N- (presumably homonymous with the 3sg pronoun and the referential article), which marks the semantic group of elder relatives.

The complementary distribution of this marker with the definite article -i is an evident parallel with the incompatibility of the elder kin terms with articles in other Western Mande languages mentioned in 4.1 and 4.2. That is why in (Vydrin 2006) the prefix *N- for elder kin terms is reconstructed for Proto-Western Mande.

Elder kin terms borrowing from Mande to Landuma

As shown in 3.3, the N-nouns in Landuma were undoubtedly borrowed from Mande languages. According to the first hypothesis, the source of the initial nasal consonant might be the reconstructed prefix *N- for elder kin terms.

For sociolinguistic reasons, the main candidates likely to have donated these borrowings are Susu and/or Jalonke (Susu is the dominant language of the littoral zone in Guinea) and Manding. Mogofin and Kakabe (two closely related languages of the Mokole group) also cannot be excluded, although they are less probable candidates, given their low social status: the role of their ancestor in the past was hardly more significant. The Southwestern Mande languages are too distant from Landuma to be considered as probable lexical donors.

The borrowed kinship terms in Landuma confirm this assumption. They can be subdivided into the following groups (see 3.3 for details):

-- a Susu loan: Ncoko `maternal uncle';

-- Manding loans: Nna `mother', Nbariy `friend', Njatiki `friend', Ntana `father/mother-inlaw', Ntokma `namesake', Nkiln `Prophet'; In Bandi m alternates with w, while in Mende m is a non-alternating consonant. For this reason we have no way of knowing whether the Mende word mm `grandmother' belongs to this group or not.

-- Susu or Manding loans: Nbenba `ancestor', Ntara `elder sibling', Nteneq `paternal aunt' Kinship terms borrowed from Mande and retaining an initial n- are occasionally found in other Mel languages, cf. ntene `aunt' in Baga Tshi-temu (Lamp 2016); ndm `namesake' in Kisi (Childs 2000). These forms seem to be the only instances of the presence of the roots of our list in Mel outside Landuma (or at least we have not found anything else in the data available for the other languages of the family). These two forms certainly result from independent (and relatively recent) borrowing from Mande; there is no reason to postulate their presence in Proto-Mel (or even a proto-language at any lower taxonomic level)..

In both the Susu-Jalonke and Manding groups, the kinship terms have no prenasalization, although its presence in the proto-language can be reconstructed. In Southwestern Mande, traces of the nasal element are more tangible (although still elusive), but direct borrowing from these languages to Landuma is hardly probable.

If the archaic nasal prefix hypothesis is accepted, two main conclusions can be drawn concerning the history of the Mande languages.

First, the Landuma data confirms the reconstruction of a nasal prefix in Proto-Western Mande advanced in (Vydrin 2006). In (Vydrine 1994; Vydrin 2006), this Proto-Western Mande morpheme was interpreted as an archaic noun class marker. Alternatively, it might be regarded as a kind of grammaticalized "honorific marker". We are not going to delve here into discussion about its nature; our fundamental concern is simply the fact of the existence of this morpheme in Proto-Western Mande.

Second, the disappearance of the prenasalized consonants in kinship terms in Western Mande languages (in particular Manding and Susu-Jalonke) seems to be a much more recent phenomenon than one might suppose, most probably dating back less than 1000 years: it must have followed the period when the kinship terms were borrowed into Landuma.

N- as the 1st person pronoun in Mande

Another hypothesis is much more straightforward: the word-initial nasal consonant in the Landuma kinship terms can be interpreted as a reflex of a Mande 1SG pronoun. We are thankful to Denis Creissels for drawing our attention to the plausibility of this hypothesis.

A semi-vocalic nasal, most probably high-toned (i.e. *N), can be reconstructed at least for the Proto-Western Mande level (and very probably for the Proto-Mande level too). Kinship terms, being relational nouns by definition, are rarely used in Mande languages without indication of the anchor (i.e. the individual to whom they stand in a kinship relation). As indicated by Dahl & Koptjevskaja-Tamm (2001), for kin terms, "a common case is for the anchor to be identical to the speaker of the utterance". In Mande, the kin terms typically appear with a 1SG possessor We have tried to check this assumption in the Bambara Reference Corpus (Vydrin, Maslinsky & Mric 2011-2020) (accessed on April 10, 2020). F `father' appears in the disambiguated subcorpus 1000 times. It is preceded by the non-emphatic 1SG pronoun n 79 times, and by the emphatic 1SG pronoun n 68 times (147 occurrences in total). For b `mother', we find 1137 occurrences; it is preceded by n 87 times, and by n 41 times (128 in total). These numbers may seem not very convincing, but the relatively weak cooccurrence of these kin terms with 1SG pronouns can be explained by the predominance of narratives in the Bambara Corpus. In dialogs the figures would certainly be much higher. and, hence, could have been borrowed into Landuma in this form. In this relation, we would like to quote Denis Creissels' observation (p.c.):

... dans un des textes diola-fogny sur lesquels je travaille actuellement, le terme mandinka nndiri `co-pouse de la mre' revient au moins une dizaine de fois sans aucune rfrence un possesseur particulier, et toujours comme nnandir ou inandir. S'il y avait un possessif, ce serait forcment un suffixe. Or il s'agit d'un emprunt occasionnel au mandinka, pas de quelque chose qu'on pourrait faire remonter un contact ancien avec une hypothtique langue mand, puisque les dictionnaires diola n'enregistrent pas ce terme. Le n- initial ne peut donc s'expliquer que comme le figement du possessif de premire personne du mandinka.


We have to admit that, at the present state of our knowledge, we do not see decisive arguments in favour of either of the two hypotheses. Certainly, the cognitively sound 1SG pronoun hypothesis looks highly attractive and convincing. There are, however, some minor arguments which can be interpreted in favor of the archaic prefix hypothesis too.

First, all kin terms in the N-group refer to elder relations, which conforms with the proposed meaning of the archaic prefix. The absence of the nasal element in the Landuma word dimbore `cousin' borrowed from Susu can be regarded as such an argument.

Moreover, the 1SG pronoun hypothesis does not explain the initial nasal in the word Nkila `prophet'. Kira, kla, cla `prophet' in the Manding languages is not a relational noun: it is normally separated from the possessor by a possessive marker. In the Bambara Reference Corpus there are 1915 occurrences of kira `prophet', and among these there is not a single one immediately preceded by the 1SG pronoun n or 1PL pronoun an. The same result obtains for the Maninka Reference Corpus, which gives 8761 occurrences of kla `prophet', not a single one of which is immediately preceded by the 1SG (n) or 1PL (n, n) pronoun. In this particular case, the archaic prefix hypothesis fits better, if we assume that this prefix had some kind of honorific semantics.

Meanwhile, neither hypothesis provides any reasonable explanation for the prenasalization in the Landuma word Nbdrfi `wild boar', unless we embark on speculations concerning the role of wild boars in the spiritual life of Landuma and/or ancient Manding speakers.

It is quite probable that both sources of prenasalization were pertinent. For some of the prenasalized Landuma forms which are not true kin terms (Njatiki `host', Ntokma `namesake'), a pronominal origin for the nasal element seems more plausible. The same is true for the noun Nna `mother', which is mainly used in the appellative function (the standard referential term for `mother' is kam). At the same time, for some other nouns (e.g., Ntara `elder sibling', Ntenerj `aunt, father's sister' and the other kin terms), the prefixal hypothesis appears quite reasonable.


1, 2, 3 -- first, second, third person

INAN -- inanimate

AGR -- agreement prefix

IPRS -- impersonal

AN -- animate

NP -- nominal prefix

ASR -- assertive

POSS -- possessive

BAS -- basic pronominal series

PRF -- perfect

CONS -- consecutive

REF -- referential article

DEF -- definite

SG -- singular

FACT -- factative

SUBJ -- subject


Anonym. 1995. Mandinka-English Dictionary. Banjul: W.E.C. International.

Anonyme. 1982. Guide de Transcription et Lexique Bozo -- Tieyaxo sawananbaana yee a xarabuye. Bamako: DNAFLA. Anonyme. n.d. Lexique susu-franais. Rep. de Guine: Service Alphabtisation et Education des Adultes.

Bris, Pierre le, Andr Prost. 1981. Dictionnaire bobo-franais, prcd d'une introduction grammaticale et suivi d'un Lexique franais-bobo. Paris: SELAF.

Childs, George Tucker. 2000. A dictionary of the Kisi language, with an English-Kisi index. Kln: Rdiger Kppe Verlag.

Covac, Don. 1978. A preliminary phonology of Bandi. Ms. Monrovia.

Creissels, Denis. 2009. Le malink de Kita (Mande languages and linguistics 9). Kln: Rdiger Kppe Verlag.

Creissels, Denis. 2010. Liste lexicale du dialonk de Falya. Mandenkan 46: 49-71.

Creissels, Denis, Sidia Jatta, Kalifa Jobarteh. 1982. Lexique mandinka-franais. Mandenkan 3: 1-207.

Daget, Jacques, J. Kanipo, M. Sanankoua. 1953. La langue bozo (tudes soudaniennes 1). Bamako: IFAN.

Dahl, Osten, Masja Koptjevskaja-Tamm. 2001. Kinship in grammar. In: Irne Baron, Michael Herslund, Finn Serensen (eds.). Dimensions of possession: 201-225. Philadelphia, PA: John Benjamins Publishing Company.

Dian, Mamadi, Valentin Vydrine. 2012. Le systme des termes de parent soso: une tude ethnolinguistique. Revue Horizons Facult des Lettres et Sciences du langage 10: 61-75.

Dobrin, Lise. 2012. Concreteness in grammar. The noun class systems of the Arapesh languages (Stanford Studies in Mor-phology and the Lexicon). Stanford: CSLI Publications.

Dwyer, David J. 1974. The historical development of Southwestern Mande consonants. Studies in African Linguistics 5(1). 59-94.

Dwyer, David J. 1986. Evolutionary morphology of definite articles in Southwestern Mande. In: Gerrit Dimmen- daal (ed.). Current approaches to African linguistics Vol. 3: 149-170. Dordrecht: Foris Publications.

Erman, Anna, Japhet Kahou Loh. 2008. Dictionnaire Dan-Franais (dan de l'Ouest) avec un index franais-dan. St P- tersbourg: Nestor-Istoria.

Ferry, Marie-Paule, Lansana Sande. 2000. Le pass des langues: Tyapi autrefois, Kogoli aujourd'hui. In: Grald Gaillard (ed.). Migrations anciennes et peuplement actuel des Ctes guinennes: 343-352. Paris: Harmattan.

Galtier, Grard, Makan Dantioko. 1979. Lexique sonink-franais. Bamako: DNAFLA.

Grossmann, Rebecca, Michael K. Rodewald, Don Covac. 1991. Bandi Dictionary. Ms. Monrovia.

Hochstetler, Lee. 1996. Enqute linguistique sur le duungoma: une langue samogo parle au Burkina Faso et au Mali. Mandenkan 31: 1-57.

Hopkins, Bradley. 1982. Etude tonologique du yaour (Cahiers ivoiriens de recherches linguistiques 11). Abidjan: SIL.

Idiatov, Dmitry. n.d. Dictionnaire toura-franais. Ms.

Iffono, Aly Gilbert. 2000. Migrations mandingues dans la zone littorale guinenne du XVme au XIXme sicle. In: Grald Gaillard (ed.). Migrations anciennes et peuplement actuel des Ctes guinennes: 77-80. Paris: Harmattan.

Innes, Gordon. 1969. A Mende-English dictionary. Cambridge University Press.

Innes, Gordon. 1971. A practical introduction to Mende. London: University of London, School of Oriental and Afri-can Studies.

Jones, Ross. 2004. Boko dictionary with reversed English-Boko finderlist. Mnchen: Lincom Europa.

Kastenholz, Raimund. 1987a. Materialen zum Koranko. Afrikanistische Arbeitspapiere Sondernummer: 1-181.

Kastenholz, Raimund. 1987b. Das Koranko. Ein Beitrag zur Erforschung der Nord-Mande-Sprachen. Kln: Universitt zu Kln Ph.D. dissertation.

Kastenholz, Raimund. 1997. Sprachgeschichte im West-Mande. Methoden und Rekonstruktionen. Kln: Rdiger Kppe Verlag.

Konoshenko, Maria. 2019. Dictionnaire kpele de la Guine (guerz) -- franais avec un index franais -- kpele. Mandenkan 62: 3-164.

Kuznetsova, Natalia, Olga Kuznetsova. n.d. Dictionnaire gouro-franais. Ms.

Lacan, Ph. L. 1942. Grammaire et dictionnaire franais-soussou et soussou-franais. Bordeaux / Conakry / Kindia: Procure des PP. du Saint-Esprit -- Mission Catholique.

Lamp, Frederick John. 2016. Baga Tshi-tem dictionary (Guinea). Ms.

Leger, Jean. 1975. Dictionnaire guerz. Nzrkor.

Leidenfrost, Theodore E., John S. McKay. 2007. Kpelle-English dictionary with a grammar sketch and English-Kpelle finderlist. Moscow (USA): Palaverhut Press.

Makeeva, Nadezhda. n.d. Dictionnaire kla-dan -- franais. Ms. (Toolbox file).

Mara, Facely, Camara Facely. 1979. tude descriptive du Ll. Kankan: Institut polytechnique Julius Nyerere M-moire de master.

Meyer, Grard. 1983. Lexique lmentaire Malink-Franais, suivi d'lements de grammaire. Dialecte maninkaa du Niokolo (sous-prfecture de Bandafassi). Kdougou.

Nikitina, Tatiana. n.d. Dictionnaire wan-franais-anglais-russe. Ms.

Paperno, Denis. n.d. Dictionnaire beng-franais. Ms.

Perekhvalskaya, Elena, Mose Yegb. 2018. Dictionnaire mwan-franais. Mandenkan 60: 3-122.

Polinder, Wim, Jacqueline Janse, Kees Jan van Linden. 2009. Lexique Mogofin-Franais. Garama: Mission vang-lique Rforme Nerlandaise.

Raimbault, R.P. 1923. Dictionnaire franais-soso et soso-franais. 2nd edition. Rome: Sodalit de St. Pierre Claver.

Rodewald, Michael K. 1989. A grammar of Bandi and Mende tone. Arlington: University of Texas M.A.Thesis.

Rogers, Kirk. 2005. Introduction to Landuma Language and Culture Lessons. Ms.

Rogers, Kirk. 2008. Phonemic Analysis of the Landuma Language. Ms.

Rogers, Kirk, Daniel Bryant. 2012. Dictionnaire landouma -- franais. Bok: Mission Evanglique de Bok.

Sande, Hannah L. 2019. Phonologically determined nominal concord as post-syntactic: Evidence from Gubie. Journal of Linguistics 55(4): 831-878.

Schreiber, Henning. 2008. Eine historische Phonologie der Niger-Volta-Sprachen (Mande Languages and Linguistics / Lan-gues et Linguistique Mand 7). Kln: Rdiger Kppe Verlag.

Smeltzer, Brad, Susan Smeltzer. 2001. Lexique Sonink-Franais-Anglais. Soninke-French-English Dictionary. Index Franais-Sonink, English-Soninke. Bamako.

Solomiac, Paul, Ali Traor, Fab Traor. 1998. Dzngoo dzuon tsnfypn ko sbun'n kur'la. Lexique orthographique dzngoo-franais. Ouagadougou: SIL.

Spears, Richard. 1972. The form of Mandingo kinship terms. Anthropological Linguistics 14(7): 281-285.

Suret Canal, Jean. 2000. Note sur les Kogoli. In: Grald Gaillard (ed.). Migrations anciennes et peuplement actuel des Ctes guinennes: 333-341. Paris: Harmattan.

Tour, Aboubacar. 1989. lments de phonologie et morphologie de la langue ss (tudes mand 3). Grenoble: Universit Stendhal.

Tour, Aboubacar. 1994. Elments de description de la langue soso. Grenoble: Universit Stendhal-Grenoble III Ph.D. dissertation.

Trbs, Holger. 1998. Funktionale Sprachbeschreibung des Jeli (West-Mande) (Mande Languages and Linguistics / Langues et Linguistique Mand 3). Kln: Rdiger Kppe Verlag.

Tveit, Harald, Gagny Dansoko. 1993. Petit dictionnaire khassonk-franais. Oussoubidiagna: Mission Protestante Norvgienne.

Vydrin, Valentin. 2006. K rekonstrukcii fonologicheskogo tipa i imennoj morfologii pra-mande [Toward recon-struction of the phonological type and nominal morphology of Proto-Mande]. In: Nikolaj Kazanskiy (ed.). Mandeica Petropolitana (Acta Linguistica Petropolitana. Trudy Instituta lingvisticheskikh issledovanij RAN [Acta Linguistica Petropolitana. Transaction of the Instiute for linguistic studies] II (2)): 3-246. St. Petersburg: Nauka. Online at: https://alp.iling.spb.ru/static/alp_II_2.pdf.

Vydrin, Valentin. 2009a. Esquisse de la langue ll (groupe mokol). Mandenkan 45: 29-104.

Vydrin, Valentin. 2009b. Lele: predstavlenije malogo jazyka semji mande [Lele: presentation of a minor Mande language]. In: Valentin Vydrin (ed.). African Collection - 2009: 235-268. St. Petersburg: Muzej antropologii i etnografii RAN (Kunstkamera).

Vydrin, Valentin, Kirill Maslinsky, Jean-Jacques Mric. 2011-2020. Corpus Bambara de Rfrence. Online at: http://cormand.huma-num.fr/index.html.

Vydrina, Alexandra. 2015. Dictionnaire de la langue kakab suivi d'un index franais-kakab. Mandenkan: Bulletin d'tudes linguistiques mand. N 53. Ms., online at: https://llacan.cnrs.fr/PDF/Mandenkan53/53vydrina.pdf.

Vydrine, Valentin. 1994. Traces of Nominal Classification in the Mande Languages: the Soninke Evidence. St. Petersburg Journal of African Studies 3: 63-93.

Vydrine, Valentin. 2001. Esquisse contrastive du kagoro (Manding) (Mande languages and linguistics 4). Kln: Rdiger Kppe.

Welmers, W. E., C. K. Kandakai. 1974. A Vai-English dictionary (preliminary draft). Ms. Monrovia.

Willits, Brad. n.d. Dictionnaire susu-franais-anglais. Ms. (Toolbox file).



  • Definition. Categories of Nouns. Forms of Nouns. Assaying for Noun. Collective Nouns, Company Names, Family Names, Sports Teams. Plural noun forms. Plural compound nouns. Special cases. Plurals and apostrophes. Singular subjecst, plural predicates.

    [34,6 K], 21.01.2008

  • The discovery of nouns. Introduction. Classification of nouns in English. Nouns and pronouns. Semantic vs. grammatical number. Number in specific languages. Obligatoriness of number marking. Number agreement. Types of number.

    [31,2 K], 21.01.2008

  • Features of English Nouns. The Category of Case. The Category of Number of English Nouns. Structural Semantic Characteristics of English, morphological, syntactical Characteristics of Nouns. The Use of Articles with Nouns in Some Set Expsessions.

    [96,9 K], 10.07.2009

  • The problems as definition of nouns, main features of English nouns, their grammatical categories. Semantical characteristics of nouns and the category of number of english nouns. The lexicon-grammatical meaning of a class or of a subclass of words.

    [27,6 K], 07.07.2009

  • The case of the combination of a preposition with a noun in the initial form and description of cases in the English language: nominative, genitive, dative and accusative. Morphological and semantic features of nouns in English and Russian languages.

    [80,1 K], 05.05.2011

, , ..