Nizaa language

Nizaa (Nizaa pronunciation: [nɪ˦zʌː˧˨]^[1]), also known as Galim, Nyamnyam, and Suga, is an endangered Mambiloid language spoken in the Adamawa Region of northern Cameroon. Most of the language's speakers live in and around the village of Galim in the department of Faro-et-Déo.

Nizaa has a complex sound system with 60 consonant phonemes, eleven tones, and a contrast between oral and nasal vowels. In terms of grammar, it is the only Bantoid language that allows multiple verbal suffixes on one verb. It also is neither a head-initial nor head-final language (the head or main element of a clause does not prefer to come before or after its modifiers) and uses postpositions instead of prepositions (the adposition follows the noun it modifies).

Nizaa was first extensively documented in the 1980s by Norwegian linguists Rolf Theil Endresen [no] and Bjørghild Kjelsvik. The language is endangered, but the exact number of active speakers is unknown since the last census of speakers took place in 1985.

Nizaa is also referred to as Suga (also spelled Ssuga),^[2] Galim, Nyamnyam (also spelled Nyemnyem, Njemnjem, and Jemjem),^[2][3] "Sewe" and "Mengaka". Nizaa is the word the Nizaa call themselves, while Suga comes from Pero súgò 'stranger' or 'not Pero'.^[3] 'Nyamnyam' is a pejorative term likely derived from the Fula word nyaamnyaamjo 'cannibal', which may also mean 'sorcerer',^[4] though there is no evidence the Nizaa have ever practised cannabalism or sorcery.^[5][6] The name Galim comes from the main town of the Nizaa of the same name.^[3][7] Nizaa is referred to as "Mengaka" in the 1988 version of Ethnologue,^[8] but Endresen in 1991 did not recognise the name or know where it came from.^[1] "Sewe" may also be another alternate name for the language.^[2]

Nizaa is primarily spoken in and around the village of Galim, located in Faro-et-Déo, which itself is in the Adamawa Region of northern Cameroon; the village has roughly 2,000 inhabitants.^[9] The most recent census of speakers was carried out in 1985 and reported 10,000 people actively speaking the language. However, the Atlas Linguistique du Cameroun (ALCAM) estimated only 2,000 speakers in 1983, so the actual number of speakers is unknown.^[10][3] The language is endangered.^[11]

The Nizaa people are primarily farmers who grow maize, though other crops such as yams, millet, peanuts, and cassava are also grown, and their main food is a paste made from cooked flour, sorghum, millet, or cassava.^[12] Hunting and fishing are also regular food sources among the Nizaa, with beekeeping and ironsmithing being used as sources of income by some communities.^[13] Most Nizaa practise Islam, though some do practise Christianity or a traditional religion that incorporates ancestor worship.^[8][4] The Nizaa are divided into various different clans, each with their own sacred animal which they do not eat or harm; traditionally, it is thought that these animals helped a clan overcome a stressful situation in the mythological past.^[14]

Most Nizaa are not literate, and the few who are often only can read and write Fula in the Ajami Arabic script.^[8][15] The romanisation of Nizaa has not been widely adopted by the Nizaa people because of their low literacy rate.^[8][15] Several other languages are spoken in the region, and most Nizaa speakers are bilingual in Fula, specifically the Adamawa dialect, since it is essentially the lingua franca of northern Cameroon. Many also know Hausa, another regional language, or French, because it is an official language of Cameroon.^[9][16]

Several details of Nizaa were known as early as 1932, but the language was first documented extensively from 1979 to 1984 by Norwegian linguist Rolf Theil Endresen [no] at the University of Oslo.^[5][17][18] However, his research was not published until 1991.^[5] Endresen devised the romanisation system of Nizaa and published the first analysis of the language,^[18][14] and he also supervised later research on Nizaa by his student, Bjørghild Kjelsvik. Kjelsvik began her work in the Nizaa community via the Evangelical Lutheran Church of Cameroon, which documented local languages in addition to engaging in evangelical conversion work.^[18]

Furthermore, because another language exists in Adamawa Region also called "Nyamnyam", linguists have often confused the two languages, and the exact classification of Nizaa is still in doubt.^[19] However, in 1983, ALCAM classified Nizaa and Kwanja as Mambiloid languages.^[20] Finally, Roger Blench in 1988 classified the Mambiloid languages, along with another language family in Nigeria's Adamawa State, the Dakoid languages, as members of the Northern Bantoid languages, which are a subdivision of the Bantoid language family.^[20]

However, another linguist, Bruce Connell disagreed with this classification. In 1997, he suggested that the Mambiloid languages form a valid genetic group (or the languages are each others' closest relatives), but also noted that Nizaa and Ndoro are highly divergent and can only be tenatively considered members of the Mambiloid languages. He also suggested that lexical and grammatical similarities within the Mambiloid languages are partly due to contact, not just shared ancestry, and that evidence of a phylogenetic relationship between the Mambiloid languages and the Dakoid languages is weak. Connell also suggested that the Mambiloid languages may either belong within the South Bantoid languages or may branch off independently from the Bantoid languages, separate from both the Dakoid languages and the South Bantoid languages.^[21][22]

When the orthography, or the conventional spelling system of Nizaa, differs from the phonetic representation, it is shown in angle brackets.

The Nizaa language has a contrast between oral and nasal vowels, with five short oral vowels, ten long oral vowels, and seven long nasal vowels.^[23][1] In Endresen's original romanisation, nasalised vowels were indicated by adding an ogonek, but Kjelsvik's revised romanisation indicates such vowels by adding ⟨ŋ⟩ after the vowel.^[1]

Older speakers of Nizaa also distinguish an eighth nasalised vowel /ɤ̃ː/ from /ʌ̃ː/, but this distinction has been lost in younger speakers.^[24][25]

Each of the five short oral vowels except /a/ can be pronounced in two different ways, depending on the context.^[23][1] These different pronunciations are called allophones, but they represent the same underlying vowel phoneme.^[26] Each allophone is part of one to two of three sets: the back set [ɯ, ʊ, ɤ, o], used when a vowel comes before /ɾ, g, ɰ̃/, except if the vowel follows /tʃ, dʒ, ᶮdʒ, ɲ, j, ʃ/ (e.g., sèr 'broom' is pronounced [sɤɾ˨] while cir 'poverty' is pronounced [tʃɯɾ˧]); the front set [ɪ, ʏ, e, ø], used when a vowel follows /tʃ, dʒ, ᶮdʒ, ɲ, j, ʃ/ and is before /n/ (e.g., còn is pronounced [tʃøn˨]); and the "normal" set [ɪ, ʊ, e, o], used everywhere else (e.g., ci is pronounced [tʃɪ˧]).^[27]

Nizaa has complex consonantal inventory with 60 consonants, including five marginal phonemes, which occur infrequently in the language.^[29] In Nizaa, there are six main types of consonants: labial consonants, or consonants made with one or both lips; alveolar consonants, articulated with the tongue touching the alveolar ridge; post-alveolar consonants, articulated with the tongue behind the alveolar ridge; velar consonants, articulated with the tongue against the velum (the back part of the roof of the mouth); glottal consonants, articulated with the glottis (or the opening between the vocal folds); and labial–velar consonants; consonants that are simultaneously articulated at the lips and velum.^[28]

Nizaa also has prenasalised consonants, which refer to sequences of nasal and non-nasal consonants that act like a single consonant, and implosive consonants, which refer to consonants articulated by both moving the glottis downward and expelling air from the lungs.^[28] Both types of consonants are common in the languages of Sub-Saharan Africa.^[30][31]

The consonants in parentheses are marginal: the labial flap is only found in ideophones or words that evoke a certain sound, and the voiced velar fricative is only found between vowels to distinguish compounds from disyllabic words.^[23] The two glottal stops are also marginal and are not written in the modern orthography, although Endresen's old orthography used an apostrophe ⟨ʼ⟩ and an apostrophe followed by 'w' ⟨ʼw⟩ to represent /ʔ/ and /ʔʷ/ respectively.^[23][32] They are only found in interjections (e.g., íŋìŋ /ʔɯɰ̃˦ʔɯɰ̃˨/ 'no'), loanwords from Arabic (e.g., Nizaa alad /ʔa˧lad˧/ 'Sunday', from Arabic: الأحد, romanized: al-ʾaḥad /alʔaħad/, loaned via Fula 'alat /ʔalat/), and variants of the third-person plural pronouns (e.g., wáá [ʔʷaː˦] 'they', when combined with a to-be or auxiliary verb).^[33] /x/ is only distinguished from /h/ in one word root, and consequently, shares the same letter as /h/ in the orthography.^[23][34]

When before the rounded vowels /ʊ ~ ʏ, o ~ ø, uː, oː, ɔː, ũː, ɔ̃ː/, the consonants /p, b, ᵐb, f, v, ᶬv, k͡p, ɡ͡b, ͡ᵑᵐɡ͡b, h/ become labialized [pʷ, bʷ, ᵐbʷ, fʷ, vʷ, ᶬvʷ, k͡pʷ, ɡ͡bʷ, ͡ᵑᵐɡ͡bʷ, hʷ] (e.g., mbombóó 'candy' is pronounced [ᵐbʷo˧ᵐbʷoː˦]).^[35]

The distinction between /s/ and /ʃ/, and between /sʷ/ and /ʃʷ/ is uncertain. Minimal pairs, or pairs of words that differ in only one sound, do exist between these sounds (e.g., saw 'to cross' and shaw 'to tremble'), but many words both /s(ʷ)/ and /ʃ(ʷ)/ can be used without changing the meaning (e.g., sùú and shùú both mean 'god').^[36]

Nizaa has three allowed syllable types: CV (e.g., ge 'to go'), CVV (e.g., sìì 'house'), and CVC, (e.g., yîm 'medicine') where C represents a consonant, V a short vowel, and VV a long vowel (which may be nasalised or not). The syllable structure V (a single short vowel) exists only in the particle a, which has various meanings based on the tone used;^[37] these include the copula or "to-be verb" á, which takes a high tone.^[38] CV and V syllables are monomoraic, while CVV and CVC syllables are bimoraic and take twice the amount of time as monomoraic syllables to fully articulate.^[39] Monosyllabic nouns can only have the syllable structures CVV and CVC, while monosyllabic verbs can have all allowed syllable structures (CV, CVV, CVC).^[37]

Nizaa only permits certain consonants to act as codas, or consonants that end a syllable; these are /p, b, t, d, k, g, m, n, ɰ̃, w, w̃, j, ɾ/. The consonants /d/ and /j/ are fairly marginal as codas, with /d/ only being found in loanwords from Fula and Arabic (e.g., pád 'everything, all' from Fula pat), and in some indigenous words (e.g., dutàd 'Rüppel's vulture'). /j/ is only found in the word sèy, the imperative form (used for commands) of the verb se 'to see or perceive', and in interjections and ideophones (e.g., kây 'no') in all dialects of Nizaa. However, in the Galim dialect of Nizaa, /j/ as a coda is found in more words and is not marginal (e.g., swěȳ 'flour made of roasted corn and peanuts' in the Galim dialect is swěe in other dialects, and shey 'a taboo or forbidden word' in the Galim dialect is shee in other dialects).^[23] In 1991, Endresen also listed another marginal coda /j̃/,^[40] but Kjelsvik, in 2002, re-analysed the only word it was known to occur in as /jĩː/.^[41]

Nizaa is a tonal language.^[42] Tonal languages use shifts in the pitch of syllables to differentiate words from each other, independently from pragmatic considerations.^[43] Such languages are common across Sub-Saharan Africa.^[44]

Nizaa has three phonemic tone levels: high, mid, and low, as well as a number of two- and three-tone tone contours (combinations of tones within a syllable), which are indicated in the orthography using a variety of diacritical marks.^{[42][note 2]} Verb roots can only use the high or mid tones, unlike nouns, which may use any of the three levels.^[46] Tones regularly participate in grammatical processes.^[42]

Endresen groups the tones into "primary tones" and "secondary tones". The primary tones, high (H), mid (M), low (L), and rising (LH), are found on all types of syllables (CV, CVV, and CVC). The "secondary tones", which include falling (HM, HL, and ML), rising (MH) and peaking (LHM) tones, are found on word-final syllables and can be understood as one of the primary tones plus an additional tone, though MH and ML tones can be found on non-final syllables in the augumentative and diminutive forms of words.^[48] Kjelsvik, in 2002, recognised two more peaking tones, MHL and LHL.^[47]

Nizaa has floating high tones, which are high tones that are not attached to a specific syllable and are written with a small circle (°), e.g., M° or L°. Before a pause, floating tones are not realized, so oppositions such as M versus M° or L versus L° are neutralized (e.g., ŋun 'boy/girl', with a tone of M, and ɓʉʉ, with a tone of M° are articulated with the same tone). However, before another word, floating tones surface as a high tone on the preceding syllable (e.g., ŋun mum 'one boy/girl' retains its mid tone, while in ɓʉʉ́ mum 'one head', ɓʉʉ́ is a rising tone MH). As a result, floating tones expand the inventory of possible tonal realizations by combining with other tones, although some combinations of tones collapse and merge with other tone combinations.^[49]

A tone at the end of a syllable can shift to the following syllable. When the shifted tone is mid and the following syllable also carries mid tone, there is no change (e.g., when words bɔ̌ɔŋ 'egg' and mum 'one' are placed one after the other, the mid tone in LHM tone sequence of bɔ̌ɔŋ merges with the mid tone in mum, creating the sequence bɔ̀ɔ́ŋ mum). However, when the shifted tone is high, the shift is audible (e.g., when gòŋ 'elephant', which has a L° tone, is before the word ɓaara 'two', the floating high tone moves, creating the sequence gòŋ ɓáara 'two elephants'). Tone shift also interacts with suffixation: when a final L is added to mark the definite form of a noun (e.g., 'the house' versus 'house'), some peaking tones (e.g., MHM) simplify to LHM.^[50]

Nizaa also exhibits downstep (ꜜH), which is a high tone realized lower than the preceding high tone but higher than a mid tone. Downstep arises when a sequence like MH is reinterpreted as ꜜH after tone shift (e.g., in the word ɓóbcún 'large dog', the augmentative of ɓo᷇w 'dog', the mid tone in the falling HM tone shifts to the following suffix cún and this MH sequence becomes a downstep ꜜH).^[51]

In some types of suffixes, including some irregular noun plurals, the root vowel lowers to /a/ in Nizaa.^[1] Coda consonants in word-final syllables change in the imperfective aspect: syllables ending in /b/ instead end with /w/; syllables ending in /w̃/ end with /m/; syllables ending in /n/ nasalise the preceding vowel; syllables ending in /g/ end in a high tone and raise the vowel (i.e. /a/ to /ʌ́/); and syllables ending in nasal vowels raise and lengthen the preceding vowel. However, when the coda augmentative suffix is appended to a word with a weakened final consonant, the final consonant fortifies back to its "strong" form.^[52][48]

Nizaa generally uses subject–verb–object (SVO) word order and is neither strongly head-initial nor strongly head-final^[53] (in the English phrase eat an apple, the verb and head of the verb phrase eat precedes its complements, making English head-initial).^[54] In noun phrases, some nouns are head-initial while others are head-final.^{[55][note 3]} Relative clauses carry no overt grammatical marking and are head-initial, because they follow the noun they modify.^[58] The language primarily uses postpositions which follow their object, which is another head-initial aspect of the language.^[59] However, there is evidence of at least one preposition (the comitative, which denotes accompaniment).^[59] The possessor always precedes the possessee and most adjectives, demonstratives, and numerals also precede the noun they modify, head-final aspects of the language. Kjelsvik lists the following examples of Nizaa noun constructions in her 2002 thesis:^[60]

Definiteness on nouns is marked by adding a lowering tone contour. The marking for plurals depends on the animacy of the noun: when the noun is animate, i.e. refers to a human or an animal, the suffix -wu is added, and when the noun is inanimate, another suffix -ya is added. The word ɓaara 'two' may also be used to represent plurality.^[61] No case-marking exists in Nizaa, with the possible exception of the locative, which marks location, although this may be a clitic instead because it may behave more like a word-like element that cannot stand on its own in speech and must attach to a neighboring word for pronunciation.^[1][62] Kjelsvik lists the following examples of noun morphology (forms of nouns):^[62]

Pronouns have three forms: their isolated versions, the versions when combined with the copula á and their versions in context (i.e. in a sentence with a finite noun). Some pronouns in Nizaa have different honorific forms to show respect, different logophoric forms to refer to someone speaking or being spoken about, or different vocative forms to address someone directly. Pronouns have singular and plural forms.^[46] The third-person plural pronoun in context has three forms, the older form ɓu, a newer form u (pronounced with an initial glottal stop, [ʔʊ˧]), and a long form búsúúŋwu.^[46][33] The third person plural pronoun form with copula two forms as well: the older form ɓwáá, and a newer form with a glottal stop wáá [ʔʷaː˦].^[33] Kjelsvik lists the following examples of pronouns:^[46][33]

Nizaa verb roots are monosyllabic, and always have a mid or high tone, but extra elements can be added to change the original meaning of the verb or for grammatical purposes.^[46] These suffixes include a "habitual/imperfective" -cì, signifying a repeated action; a "perfective/stative" -wu᷄, signifying a completed action or a current state of being; and two additional perfect tenses.^{[note 4]} Progressive (indicating an ongoing action) -ri and imperative (indicating a demand, marked with a low tone) suffixes also exist, as well as suffixes negating the original verb. There is also a "detransitivizer" suffix, -re᷇, that demotes the direct object of a verb to an oblique argument or removes it altogether.^[63] Many suffixes, like the imperative, have negative counterparts. For example, the perfective/stative's negative counterpart is -ŋwa.^[64]

In addition to these verb tense suffixes, Kjelsvik also describes several verb number indicators and a variety of suffixes indicating location.^[65] For example, four directional suffixes, which serve to identify the path of motion verbs, have been described by Kjelsvik. These are the illative -a, which indicates "motion into an enclosure"; the allative -ri, indicating "motion towards a location"; the distinctive -wa, indicating "motion away from a location"; and the sublative suffix -sa, indicating "motion towards a lower location".^[66]

Nizaa is the only Bantoid language that permits strings of verbal suffixes.^[67] A stacking of up to three suffixes to a single verb is grammatical in Nizaa,^[65] and up to four verbs may occur in one sentence in Nizaa.^[68] Kjelsvik lists the following examples of Nizaa verb constructions:^[69]

The Nizaa language has an Iroquois kinship system; that is, it uses the same word maaŋ for both mothers and maternal aunts, and the same word táá for both fathers and paternal uncles. However, it has different words for maternal uncles and paternal aunts, distinguishing them from the other relatives. The Nizaa language does not distinguish maternal and paternal grandparents either. Separate terms exist for older sisters and brothers, but there is no distinction of sex for younger siblings. The terms for 'older sister' díí and 'older brother' daà are also used as generic terms for polite address, while addressing someone as a 'younger sibling' na᷇m is seen as disrespectful regardless of gender. The words for cousins are the same as the ones for siblings and also depend on age. A generic term for in-laws, jwììŋ, also exists.^[70]

Kjelsvik gives a sample sentence in Nizaa:^[71]

• Languages of Cameroon
• Adamawa languages
• Mambila language
• Vute language