Medumba language

Medumba is part of the Eastern Group of the Bamileke Cluster, which also include Fe'fe', Ghomálá', Kwa', and Nda'nda'. The Bamileke cluster — along with Ngemba, Nkambe and Nun — is part of the Eastern Grassfields subgroup which, together with the Ring languages and the Southwest Grassfields languages, constitute the Grassfield Bantu grouping.

Medumba figures prominently in linguistic research on the Bamileke cluster, partly because of the high quality of the work done by Jan Voorhoeve in the 1960s and 1970s, including work on (in chronological order):

• morpheme structure constraints (Voorhoeve 1965[10])
• personal pronouns (Voorhoeve 1967[11])
• noun classes (Voorhoeve 1967[12] and 1969[13])
• tone of nouns (Voorhoeve 1971[14])
• traditional Bamileke narratives (Voorhoeve 1976[15])
• general linguistic description of the Bangangte dialect (Voorhoeve 1977[16])

This work was pursued by L. Hyman in the 1980s, on the closely related language Fe'fe' (add refs). It was re-invigorated in the early 2010s by research groups at Boston University (lead by C. O'Connor) and at the University of British Columbia (lead by R.-M. Déchaine). Some of these publications include:

• [add refs]

Also notable are the scholarly contributions of Medumba speaker-linguists, including (in chronological order):

• Keupdjio 2011, MA thesis (University of Yaounde) on content questions[17]
• Kouankem 2011, PhD dissertation (University of Yaounde) on nominal syntax[18]
• Kouankem 2013 on concord[19]

Recent work on Medumba is part of a more general push towards documenting the languages of Africa, in the face of rising levels of language endangerment.[20] Cameroon — along with Nigeria, Sudan, and Ethiopia — is reported to have one of the highest language mortality rates in Africa.[21][22]

Efforts to develop a Medumba orthography date back to the beginning of the 20th century,[23] and are associated with the following milestones:

• 1926: Protestant missionaries develop an orthography for use in primary schools to facilitate the penetration of the Christian Bible in the region.
• 1957: The French administration prohibits the use of local languages in the territories of the Union Française.
• 1973, 20 December: CEPOM (Comité de Langue pour l'Etude et la Production des Œuvres Bamiléké-Medumba) is created; its mission is to promote research on Medumba and develop literacy materials.

Currently, educational materials, literature and dictionaries for the language are produced by CEPOM, based in Bangangte. The combined output of missionary and CEPOM work has produced more that 80 publications on Medumba language and culture; these are published in French, English, and Medumba. Over time, publications in Medumba have used six different orthographies:[24]

• 1926: published in a collection of 12 songs
• 1945: published in a book on religious history (Nu Yawe Nsi) and a syllabary
• 1967: publication of the New Testament (Kàn nswe) and Tshô Pangante
• 1973: publication of Cam Medumba and zi'te nkite.
• 1979: publication of Mbwog NkUd MedUmbA and Tons, Textes…
• 1985, 2 February: adoption of the current orthography at the 4th CEPOM Council

The current alphabet is given in Table 1.

In addition to simplex consonants, Medumba has numerous complex consonants, and these are represented as digraphs or trigraphs; see Table 2. Nasals, stops, and fricatives can be labialized; in the orthography this is represented as a CW digraph. Stops and fricatives can be pre-nasalized; in the orthography this is represented as an NC digraph or an NCW trigraph.

Orthographic conventions for tone-marking are as follows:

• High-tone is not marked
• Low-tone is marked as a grave accent on the vowel, e.g. à
• Rising Low-High is marked as a hatcheck on a single vowel, e.g. ɛˇ
• Falling High-Low is marked as circumflex on a single vowel, e.g. ʉ̂

Recent developments in digital literacy have had an impact on Medumba. For example, a seven-language electronic calendar — in French, English, and the five national languages (Medumba, Ghomala, Yemba, Meta, Kom) — sends information on time and date to an LCD screen via a VGA controller. The calendar has been developed as a way of mitigating the impact of language competition, and with the specific goal of raising the profile and prestige of the national languages in a context of language endangerment.[25]

Radio FM 100 Medumba was established in 2000 by Medumba Kum Ntsi.[26]

Medumba has a 10 simplex vowels, and 5 complex vowels (diphthongs).

Table 3 gives the inventory of simple vowels, which include 5 closed (high) vowels /i, ʉ, u, ɪ, ʊ/ and 5 non-closed (non-high) /e, ə, o, a, ɑ/. See (1-10) for examples.

See Table 4 for the inventory of the five diphthongs, which include /ia, iə, ʉa, ʉɑ, uɑ/. See (11-15) for examples.

Diphthongs involve a combination of a closed (high) vowel (V1) /i,ʉ,u/ with a non-closed (non-high) vowel (V2) /a,ə,ɑ/, as follows:

• front /i/ combines with front or centre /a/ and /ə/ to form /ia/ and
  • front /i/ does not combine with back /ɑ/, so */iɑ/ is not a possible diphthong
• central /ʉ/ combines with front or back /a/ and /ɑ/ to form /ʉa/ and /ʉɑ/
  • central /ʉ/ does not combine with central /ə/; so */ʉə/ is not a possible diphthong
• back /u/ combines with back /ɑ/ to form /uɑ/
  • back /u/ does not combine non-back vowels, so */ua/ and */uə/ are not possible diphthongs

The canonical morpheme in Medumba is a single syllable, either an open CV syllable or a closed CVC syllable (Voorhoeve 1965:319).[10] This morpheme structure constraint has consequences for the consonant inventory. Indeed, a notable property of Medumba is that the number of contrastive consonants differs according to whether one considers consonants in onset position (i.e., consonants that begin a CV or CVC syllable) or consonants in coda position (i.e. consonants that end a CVC syllable). Below, the consonant inventory is introduced, and the distributional differences between coda (C2) and onset (C1) consonants are described.

Medumba has 42 consonants, of which 18 are simplex consonants and 24 are complex consonants.

Table 5 gives the inventory of the 18 simplex consonants (Voorhoeve 1965); they are displayed according to place of articulation (labial, alveolar, palatal, velar, glottal), and manner of articulation (nasal stop, oral stop, fricative). There are 4 nasals [m, n], ɲ, ŋ], 4 voiced stops [b, d, j, g], 4 voiceless stops [t, c, k, ʔ]), and 2 voiced fricatives (v, z) and 3 voiceless fricatives (f, s, ʃ). Dashes (--) indicate gaps in the consonant inventory.

Note the absence of the following segments:

• there is a voiced bilabial stop /b/, but no counterpart voiceless bilabial stop */p/
  • but [p] occurs as an allophone of /b/
• there is a voiceless palatal fricative /ʃ/ (with limited distribution), but no counterpart voiced palatal fricative */ʒ/
  • but [ʒ] occurs as an allophone of /ɟ/
• there are velar nasal and oral stops, but no counterpart velar fricatives */ɣ/ or */x/
  • but [ɣ] occurs as a release in the /ᵑg/~[ᵑgˠ] alternation, and [x] occurs as a release in the /ᵑk/~[ᵑkˣ] alternation

Table 6 gives the inventory of the 24 complex consonants found in Medumba (Voorhoeve 1965:326, section 3.1.9); they are displayed according to the primary place of articulation (labial, alveolar, palatal and velar) and the nature of the secondary articulation (labialized, pre-nasalized, and pre-nasalized labialized). This includes 8 labialized consonants (Cᵂ), 10 pre-nasalized consonants (ᴺC), and 6 pre-nasalized labialized stops (ᴺCᵂ). Complex consonants only occur in onset position; see the next section for exemplification.

Note the following gaps in the inventory of complex consonants:

• no labialized labial nasal or voiceless stop: */mᵂ/, */pᵂ/, */ᵐpᵂ/
• no labialized alveolar (nasal or oral) stop: */nᵂ/, */dᵂ/, */ⁿdᵂ/, */tᵂ/, */ⁿtᵂ/
• no labialized voiceless labial, palatal or velar fricative: */fᵂ/, */ᵐfᵂ/, */ʃᵂ/, */ᶮʃᵂ/, */xᵂ/, */ᵑxᵂ/

Of the 40 consonants found in Medumba, only 7 can be coda consonants: the 3 nasal stops /m, n, ŋ/, the 3 counterpart voiced oral stops /b, d, ɡ/, and the glottal stop /ʔ/. See Table 7 for the list of coda consonants, ordered according to place of articulation (labial, alveolar, velar, glottal), and manner of articulation (nasal stop, oral stop). (16-18) present examples of nasals in coda posit. (19-22) present examples of oral stops in coda position, which display allophonic variation conditioned by the right-edge context.

Final nasals include bilabial /m/, alveolar /n/, and velar /ŋ/.

Final stops includes bilabial /b/ (with allophone [p]), alveolar /d/ (with allophone [l,t]), velar /g/ (with allophone [ʁ, k]), and glottal /ʔ/.

Table 8 presents the inventory of onset consonants and their allophones. The only consonant excluded from onset position is the glottal stop /ʔ/. All other consonants occur in onset position, so there are 39 possible onset consonants. In onset position, nasals may be plain (C) or labialized (Cᵂ). All other consonant types (voiced plosives, voiceless plosives, fricative) occur as plain (C), labialized (Cᵂ), pre-nasalized (ᴺC), or pre-nasalized and labialized (ᴺCᵂ). In addition, onset consonants display allophonic variation that is conditioned by the following vowel. Examples are given below, as follows:

• onset labials; see (23-29)
• onset alveolars; see (30-38)
• onset palatals; see (39-49)
• onset velars; see (50-59)

Initial labials include:

• plain consonants (C: /m/, /b/ (with allophone [p]), and /f/ (with allophone [ᵐv])
• labialized consonants (C^w): /b^w/ (with allophone [bᵛ])
• pre-nasalized consonants (ᴺC): /ᵐb/ and /ᵐf/
• pre-nasalized labialized consonants (ᴺC^w): /ᵐb^w/

Initial alveolars include:

• plain consonants (C): /n/, /d/ (with allophone [l]), /t/ (with allophone [t^h]), and /s/
• labialized consonants (C^w): /s^w/
• pre-nasalized consonants (ᴺC): /ⁿd/, /ⁿt/ (with allophone [ⁿt^h]), and /ⁿs/ (with allophone [ⁿz])
• pre-nasalized labialized consonants (ᴺC^w): /ⁿs^w/

Initial palatals include:

• plain consonants (C): /ɲ/, /ɟ/ (with allophones [j], [ʒ], [z]), /c/ (with allophones [ts] and [cʃ]), and /ʃ/
• labialized consonants (C^w): /ɲ^w/, /ɟ^w/ (with allophones [ʒ^w],[z^w]), and /c^w/ (with allophones [ʃ] and [ts^w])
• pre-nasalized consonants (ᴺC): /ᶮɟ/ (with allophones [ᶮʒ], [ⁿz]), and /ᶮc/ (with allophones [ᶮcʃ] and [ⁿts])
• pre-nasalized labialized consonants (ᴺC^w): /ᶮc^w/ with allophone [ᶮʃ]

Initial velars include:

• plain consonants (C): /ŋ/, /g/ (with allophone [ʁ]), /k/ (with allophones [k^h] and [k^x])
• labialized consonants (C^w): /ŋ^w/ and /g^w/ (with allophones [w] and [v])
• pre-nasalized consonants (ᴺC): /ᵑg/ (with allophone [ᵑgˠ])and /ᵑk/ (with allophones [ᵑk^h] and [ᵑk^x])
• pre-nasalized labialized consonants (ᴺC^w): /ᵑg^w /and /ᵑk^w/

Consonant-final words — which are generally CVC because of the size constraint that favours CV or CVC words —are often augmented by a final vowel. This process of vowel insertion happens in one of two contexts: (i) before a pause; (ii) at the end of a sentence. The quality of the inserted vowel is conditioned by the final consonant: if the final C is a glottal stop, then the inserted vowel is schwa; elsewhere, the inserted vowel is a copy of the stem vowel. Examples illustrating vowel insertion are given in xx-yy.

(1) a. koo                          b.  cintEE
       ko-o                             cin-te-e
       love-FV                          xx-yy-FV
       'to want, to love'               'to urinated'
       (adapted from Voorhoeve 1965:332)

Consonants in onset position surface with different variants. This consonant allophone, a form of consonant mutation, is conditioned by the following vowel. There are seven conditioning contexts, as follows:

1. the non-closed vowels
2. the closed vowels
3. the high front vowel /i/
4. the high non-front vowels /ʉ/ and /u/
5. the high central (non-front, non-back) vowel /ʉ/
6. the high back vowel /u/
7. the vowels /o/ and /ə/

Table 9 lists these seven contexts, along with the set of segments and phonological class that they target, the effect that they have, and the allophonic alternation that they condition.

Medumba is described as a two-level tone system with low (L) and high (H) tones; examples are given in Table 16. Observe that the L/H contrast is found with all Lexical (open) class categories; this includes verbs, nouns and prepositions. Likewise, Functional (closed-class) categories show an L/H contrasts; this includes verbal F-categories (C, T, and Aspect) and nominal F-categories (Dem, Det, Pl). [Describe examples; also give minimal pairs]

Verb stems come in two shapes, CV and CVC, with each one contrasting Low and High tone. See (1-5) for examples of High/Low tone contrast with CV stems, and (6-6) for examples of High/Low tone contrast with CVC stems.

In principle, given the possibility of a stem bearing associated with one of four tone melodies — namely L(L), L(H), H(L) and H(H) for nouns and L or H for verbs — one expects to find a four-way tone contrast for a given segmental base (either CV or CVC). No such examples are attested within a given word-class, but there is one instance elf a 4-way contrast across word-classes. In addition, there are a few three-way contrasts for a given noun base, and numerous many four-way tone contrasts with the same base, if one looks at tone melodies across word-classes.

• CV tɔ, which has L(H), H(L) and H(H) stems
• CVC m-vɛd, which has L(L)
• CVC lɛn, to, and mvdd.

Examples are given in (1-3).

(3)  Low-tone           fù                          (4) High-tone          tʃə́ŋ   
                       [fù]                                                [tʃə́ŋ]
                       'medicine'                                           'food'         

(From Kouankem 2013:60; Mucha 2017: 8)

(5)  Low-tone    (a)   mbàŋ                       (6) High-tone         mʙə́   
                       [mbàŋ]                                           [mʙə́]
                       'next to'                                        'in front of'    

                 (b)   nùm
                       [nùm]
                       'on'
  
                 (c)   ɲàm
                       [ɲàm]
                       'behind'

(from Hawkes et al. 2015:122)

(7)  Low-tone    (a)   ndà                        (8) High-tone         mbʉ   
                       [ndà]                                           [mbʉ́]
                       'C'                                              'C'    

(From)

(9)  Low-tone    (a)    s-ə̂n                     (10) High-tone         N yə́n   
                       [s-ə̂n]                                           [N  yə́n]
                       AGR-this                                          N    Dem.Dist
                                                                       'that N'    

(Kouankem 2013:60)

(11)  Low-tone    (a)   bà   N                    (12) High-tone        ba N 
                       [bà]                                             [bá]
                       'PL'                                            'PL'   

(Kouankem 2013:62)

In addition to level high and low tones, Medumba exhibits falling (HL) and rising (LH) contour tones. These contour tones are morphologically derived from floating H tones that occur as affixes preceding or following the stems they associate with. These floating tones make themselves known by docking to tone-bearing units (TBUs) associated with L-tone, thus forming a tone contour. [DESCRIBE EXAMPLES; add LH examples]

(13)  L-tone verb (put in sentence)
      ghʉ̀ 
      [ɣʉ̀] 
      do
      'do’


(14)  Derived HL-tone verb (put in sentence)
      nghʉ̀ 
      [N-ɣʉ̀] 
      N-do
      'do, consecutive’

(identify source)

Medumba shows downstep, where H is produced at a lower pitch than an immediately preceding H tone; downstep is represented as (ꜜ). Downstep is viewed as resulting from a floating Low tone that shifts the pitch level of a following High tone one step lower than the preceding High tone.[14] Downstep is syntactically conditioned in that it occurs at phrasal boundaries:

• Downstep occurs between Subject and following Verb,
• Downstep occurs between Verb and following Complement
• Downstep occurs between Noun and following Complement

[GIVE EXAMPLES]

Medumba has several affixes including:

• the suffix -te, which attaches to verbs to derive the counterpart iterative form ('to verb regularly or repetitively'). While many -te forms have a counterpart base to which the may or may not transparently related (1-10), many of them seem to be frozen forms, for which the underived no longer exists (11-22).
• homorganic nasal prefixal N-, which is found in two contexts:
  • it attaches to verbs to derive nouns
  • it attaches to verbs to derive non-initial verbs (e.g. after an auxiliary or the non-initial verb of serial verb construction)
• suffixal H-tone
• prefixal H-tone

Tone on Nouns: {LL, HH, HL, LH}: Voorhoeve introduced a non-segment tone in order to distinguish two different low tone noun groups and two different high tone noun groups. For instance, naʔ and mfən both bear a segment low tone, but their tonal realization is different in the context such as mə jən mfən ___. For instance, naʔ in mə jən mfən naʔ (I saw the child of the cow) bears a non-low tone, whereas mfen in mə jən mfən mfən (I saw the child of the chief) bears a low tone. He proposed a four-way distinction, L(L), L(H), H(L), and H(H) to account for the nominal tone groups (Voorhoeve 1971:44-53). Examples are given in Table 10.

Tone on Verbs: {L, H}: The radical of the verb has only one tonal contrast, which is a Low and High contrast.[28][29] The tone of radical may be realized differently in different contexts. For instance, a low tone verb that has a nasal prefix has a different tone from its non-prenasalized counterpart. Examples are provided to illustrate this phenomenon. [Give examples of CV, CVC, CV-L and CVC-L)

The examples above show that the Low tone radical is realized as a high tone if the verb is prenasalized, whereas it is realized as a low-mid tone if the verb is not prenasalized. The High tone radical is realized as mid tone regardless of prenasalization. This effect is also found in Bamileke-FeʔFeʔ.

Voorhoeve identifies two characteristics of noun classes that surface in Medumba:[12]

• Systems of pairings between singular and plural prefixes
• Concording nominal and prenominal prefixes

Voorhoeve also assumes that pronominal prefixes exist within the noun class system, with these prefixes consisting of inherent tone morphemes such as the left-edge floating tone. The added prenominal tone creates a tonal difference between singular and plural noun class pairs, with generation of the plural form created by the singular.[12][30]

Noun classes can be detached by singular and plural pairs.[31]

Voorhoeve ascertains that the nasal prefix serves as a distinguishing factor between singular and plural noun pairs. As seen in Table 19, this nasal prefix does not surface in all constructions, especially with singular nouns that are already nasal word-initially.[12][31]

Other noun classes in Medumba, however, exhibit derivation from Proto-Bantu noun classes, which had a strict parallel between singular and plural classes.[31] In comparing agreement systems of Bamileke languages, Medumba behaves independently from the expected system. Compared to Proto-Bantu noun classes, there seems to be a noun class merger in Medumba.[14] This merger occurs through the compression of various Proto-Bantu noun classes into a generalized noun class in Medumba. This can be seen in Table 13.

Proto-Bantu noun classes typically assign particular words to certain noun classes, but this is not exhibited in Medumba. This would imply the dissolution of strict noun classes like the ones found in Proto-Bantu, as the set noun classes are merging into one class, albeit still maintaining a noun class-like form.[31] The flexibility of noun classes in Medumba could be correlated with inflectional morphemes acting as the noun class system, with these morphemes surfacing as left-edge floating tones.[30]

Loan words are normally inserted into unrestricted noun classes. /látrí/, the loan word for "light" from English, can either be viewed as a mass noun or not, depending on the speaker. In the case of Medumba, this allows for any noun class to take a loan word.[31]

In looking at interspeaker variation on the addition of loan words in Medumba noun classes, the instability of a formal noun class allows flexibility with speakers of various dialects This could be due to disagreement on how to lexicalize a new loan word between the various dialects. One such example is the loan word for "light", /látrí/. While one speaker chose to pluralize /látrí/ as /ndátrí/, the other speaker refused to pluralize /látrí/ as they believed it was a mass noun.[31]

Younger speakers of Medumba are beginning to use forms of words that do not account for noun class, such as the first person possessive form /jɔm/. The first person possessive form normally varies depending on the noun it is possessing, such as in /látrí -ɔm/ (my light) or /ndátrí -t͡ʃɔm/ (my lights). It is unknown as to how much younger speakers know about noun classes and agreement.[31]

Bamileke distinguishes four sets of personal pronouns: simplex, possessive, complex, and reciprocal.[11]

Simplex pronouns are differentiated according to syntactic position:

• subject forms are V, CV, or CVC, and surface with Low tone or High tone
• object forms are V, CV, or CVC, and surface with Low toner or High tone
• elsewhere forms occur as indirect object, object of P, or in non-argument position (e.g. when topicalized by kí or focus-marked by á)

Almost all plural pronouns often begin with b-, which is a marker of plurality; see Table 15. The only exception to this is the 2pl pronoun jin, which is the elsewhere form (i.e. it is used with indirect objects, objects of prepositions, and for topicalization and focalization).

The tone simplex pronouns depends on the following verb or auxiliary (for subject pronouns) or the preceding verb (for object pronouns). Examples of tone variation of pronoun in subject position (provide fuller description and edit examples). Example (1) illustrates 1st me; 2sg u, 3sg o, and 3pl bu also behave as in (1). Example (2) illustrate 1pl bag; 2pl bin also behaves as in (2). [give orthography and IPA]

(1) a. mə  kem men                                  (2) a. bag jen men 
       1sg see child                                       1pl see child
       'I saw the child'                                   'We saw the child'

    b. mə  ke? jen men                                  b. bag ke? jen men
       1sg NEG see child                                   1pl NEG see child
       'I did not see the child'                           'We did not see the child'

    c. mə  do          jen men                          c. bag do          jen men
       1sg Past. Recent see child                           1pl Past.Recent see child       
       'I just saw the child'                              'We just saw the child'

    d. mə ke  njen men                                  d. bag ke  njen men
       1sg ?? see  child                                   1pl ??  see  child
       'It was me who saw the child'                       'It was us who saw the child'
       (adapted from Voorhoeve 1967:423f.)

Examples of tone variation of pronoun in object position ((provide fuller description, edit examples put both IPA and orthography). Example (3) illustrates 1sg am; example (4) illustrates 1pl jag(e). [give orthography and IPA

(3) a. a   dab  am                                  (4) a. a   dab  jage
       3sg beat 1sg                                        3sg beat 1pl
       'S/he beat me'                                      'S/he beat us'

    b. a na? ndab   am                                  b. bag ke? jen men
       1 ??  N-beat 1sg                                    1pl NEG see child
       'S/he beat me a bit'                                'S/he beat us a bit'

    c. a   ton    am                                    c. a   ton    yage
       3sg spy.on 1sg                                      3sg spy.on 1pl       
       'S/he spied on me'                                  'S/he spied on us'

    d. a   na? nton   am                                d. a   na? nton   yage 
       3sg ??  spy.on 1sg                                  3sg ??  spy.on 1pl
       'S/he spied on me a bit'                            'S/he spied on us a bit'
       (adapted from Voorhoeve 1967:425)

The tonal structure of possessive pronouns depends on the tonal structure of the preceding noun.

Provide examples (see Voorhoeve 1967:426); use template from subject and object pronouns above.

Complex plural pronouns specify the composition of the group of participants denoted by the pronoun. [Provide fuller description; give examples form Voorhoever 1967:428); regularize IPA transcription]

(1)   1pl   (a)  [báàg-jí]                 (b)  [bàág-ɑ̀-bu]                       (c)  [báàg-wʉ]            (d)   [bág-ɑ̀-bìn']
                  baag-ɟi                        bǎg-à-bo                                bâg-uɨ                     bág-à-bìn´
                  1pl-3sg                        1pl-LINK-3pl                            1pl-2sg                    1pl-LINK-2pl
                  'us: me+him'                   'us: them/us+her/him/them'              'us: me+you(sg)'           'us:me/us+you(pl)'
                                             CHECK  

(2)   2pl   (a)  [bíìn-jé]                 (b)  [bín-ɑ̀-bu]                         (c)   —                   (d)   —
                  bîn-jé                         bín-à-bo
                  2pl-3sg                        2pl-LINK-3pl
                  'you all:you(sg.)+him'         = (i) 'you all: you(sg)+them 
                 CHECK                           = (ii) 'you all: you(pl+her/him/them'
                                            CHECK

(3)   3pl   (a)  [búù-jí]                 (b)   [bú-ɑ̀-bu]                           (c)   —                   (d)   —
                  bô-jé                          bó-à-bo
                  3pl-3sg                        3pl-LINK-3pl
                  'them:her/him+3her/him'        = (i) 'them:her/him+them
                  CHECK                          = (ii) 'them: them+her/him/them' 
                                             CHECK

(adapted from Voorhoeve 1967:427)

[Provide description]

(1)  bǎgbàg
     bǎg-bàg
     1pl-1pl
     '1pl, exclusive: 'us (us/me and them) between ourselves' CHECK           

(2)  bənbən
     bən-bən
     1pl.in-1pl.in 
     '1pl, inclusive: us (me and you) between ourselves' CHECK
   
(3)  binbìn
     bin-bìn
     2pl-2pl
     '2pl: you all between yourselves'

(4)  bobo
     bó-bó
     3pl-3pl
     '3pl: 'them between themselves'

(say where there examples are from)

Medumba has a rich inventory of temporal and aspectual auxiliaries, and makes productive use of serial verb constructions. [Give examples]

Medumba has finely articulated temporal contrasts, with up to 9 distinct past tense auxiliaries, and 5 distinct future tense auxiliaries.[32][33]

Medumba makes use of numerous clause-typing particles that occur at the beginning or end of the sentence: they are used to mark yes/no questions, content questions, relative clauses, as well as embedded clauses. In addition, there are two forms of negation, according to whether negation has scope over VP or CP: VP-scope negation is contrastive (e.g. He bought some books, but he did not sell pens); CP-scope negation denies the truth of a proposition p (e.g. NOT-p = it is not the case that p). [Give examples]

Associative noun constructions, which usually translated as 'Noun₁ of Noun₂', as in bǎm mɛ́n (‘belly of the child’), are analyzed with a floating tone interposed between the two nouns. The presence of this floating tone is reflected by the tone melodies borne by the nouns that precede or follow it.[14][34] This associative tone, first hypothesized by Jan Voorhoeve, may be high or low depending on the class and tone pattern of the preceding noun. The tone may lower the pitch-level of Noun₂ (a phenomenon called down step); or it may combine with the final vowel of Noun₁, resulting in a (rising LH or falling HL) contour tone.

For example, in (1), the first line represents the theorized underlying tonal melody of the two nouns, with the L tone associative marker in bold. The second line shows the two tones which are actually pronounced at the surface, with the pitch level shown in brackets (a very low pitch followed by a mid-range pitch). Likewise in (2), the first line shows the theoretical tonal melody of what at the surface are pronounced as a LH contour on Noun₁ and a down-stepped high tone on Noun₂. Thus, in brackets we see a slash indicating a rise from low to high pitch, and a mid-level line indicating a mid-range pitch.

(1)    L-(H)  (L) (L)-H-(L))                    (2)    L-(L) (H) (L)-H-(L)
       L            H          [ _ – ]                 LH   ↓  H          [ / – ]
      nzwi           mɛn                              bǎm           mɛn
      nzwí           mɛ́n                              bàám        ↓   mɛ́n
      woman    of    child                            belly   of      child
     'the woman of the child'                        'the belly of the child'

(adapted from Voorhoeve 1971:50)

Examples (3) to (6) illustrate the realization of the associative in contexts where Noun₁ is L-tone and Noun₂ is H-tone. Whether or not Noun₁ is theorized to be followed by a floating L or H tone has no effect on the surface form (i.e., the difference between (3) and (5), and (4) and (6)). What does have an effect is the tone of the associative marker—in (3) and (5) it is H, while in (4) and (6) it is L, resulting in an LH contour on Noun₁.

The effect that the associative tone has on either of the two nouns may be dependent on which floating tones accompany those nouns, which in turn is decided by their noun class. Examples (7) to (10) illustrate the realization of the associative in contexts where Noun₁ is H-tone and Noun₂ is H-tone. As with (3) to (6), in examples (7) and (9) we can see that the floating tone following the noun has no effect on the surface form—if the associative marker is H, then the surface form will always be two level H tones, or else an HL contour followed by an H tone, varying by speaker or dialect. In the case of (8) and (10), the floating tone which follows the noun does have a perceptible effect. In (8), the surface form is either an H tone followed by an L tone, or else an HL contour followed by a L tone (again, varying by speaker or dialect). However, in (10) the only available choice is a H tone followed by down-stepped H, demonstrating that the theorized underlying tones do have an effect on the pitch at which tones are actually pronounced.

Resolution of associative tone: L-tone Noun1 + H-tone Noun2
(tones in brackets () are floating)

(3)    L-(L)  (L) (L)-H-(L)                  (4)    L-(L)  (H)  (L)-H-(L)
       L              H          [ _ – ]            L       H    ↓  H           [ / – ]
     mfə̀n            mɛn                            bàm            mɛn
     mfə̀n            mɛ́n                            bàá          ↓ mɛ́n
     1.chief   of    child                          5.belly of     child
     'the chief of the child'                       'the belly of the child'     

(5)    L-(H)  (L) (L)-H-(L)                  (6)    L-(H)  (H) (L)-H-(L)   
       L              H          [ _ – ]            L       H   ↓  H            [ / – ]
      nà             mɛn                            kɔ̀              mɛn
      nà             mɛ́n                            kɔ̀O         ↓  mɛ́n
      3.cow of       child                          5.spear of     child
      'the cow of the child'                        'the spear of the child'
     
(adapted from Voorhoeve 1971:50)

Resolution of associative tone: H-tone Noun1 + H-tone Noun3
(tones in brackets () are floating)

(7)    H-(L)  (L) (L)-H-(L)                  (8)    H-(L)  (H) (L)-H-(L)    [TYPO?, check] 
       H              H          [ ¯ ¯ ] or         H           L              [ ¯ _ ] or 
       H       L      H          [ \ ¯ ]            H  L        L              [\ _ ]
      mɛn            mɛn                           ju          mɛn
      mɛ́(ɛ)n         mɛ́n                           ʒú(ù)       mɛ́n
      1.child of     1.child                       3.thing of  1.child
      'the child of the child'                     'the thing of the child'    

(9)    H-(H)  (L) (L)-H-(L)                  (10)    H-(H)  (H) (L)-H-(L)   
       H              H          [ ¯ ¯ ]             H           ↓  H            [ ¯ – ]
       H       L      H          [ \ ¯ ]            cʉ                 mɛn
      nzwi           mɛn                            tɨ́         ↓  mɛ́n
      nʒwí(ì)        mɛ́n                            3.tree of     1.child
      nzwí(ì)        mɛ́n                            'the tree of the child'
      1.woman  of    1.child                                                 
      'the woman of the child'                        
     
(adapted from Voorhoeve 1971:50)

Medumba demonstratives (the equivalent of English "this" or "that") include three elements:[35]

• a noun (N);
• a demonstrative (Dem), which shows noun class with the noun;
• an obligatory "adverbial locative reinforcer" (Loc)

The normal (unmarked) order or demondtratives is postnominal, yieding [Noun-Dem-Loc]. Divergence from this canonical order is possible, so that [Dem-N-Loc] is also attested, and yields a contrastive focus reading, where the noun is contrasted with other deictic alternatives (this man, as opposed to that man).[35]

Within the DP, there are 8 possible word order variations, taking into account the order of nouns, adjectives, demonstratives, and numerals. These are divided by Kouankem (2011) into post-nominal and pre-nominal variations.[36] The possible word orders are (where N = noun; A = adjective; Dem = demonstrative; and Num = number):

Post-nominal word order variation
(N = noun; A = adjective; Dem = demonstrative; Num = numeral)

(1)  N        A     Num  Dem
     bun      sɛŋkɛ kua   cən li
     give IPA
     children black four  DEM.PRX
     'these four black children'          

(2)  N        Num   A      Dem
     bun      kua   sɛŋkɛd cən li
     give IPA
     children four  black  DEM.PRX
     'these four black children' 
   
(3)  N        A      Dem     Num
     bun      sɛŋkɛd cən li  kua
     give IPA
     children black  DEM.PRX four
     'these four black children'

(adapted from Kouankem 2011:234-236; identify pp. & example nos.)

Pre-nominal word order variation
(N = noun; A = adjective; Dem = demonstrative; Num = numeral)

(1)  A         N        Num  Dem
     mbwo      bun      kua  cən li
     give IPA
     beautiful children four DEM.PRX
     'these four beautiful children'          

(2)  A          N       Dem     Num
     mbwo      bun      cən li  kua
     give IPA
     beautiful children DEM.PRX four
     'these four beautiful children' 
   
(3)  Dem     A         N        Num
     cən     mbwo      bun      kua  li
     DEM.PRX beautiful children four ??
     'these four beautiful children'

(4)  Dem     N        A      Num
     cən     bun      sɛŋkɛd kua li
     DEM.PRX children black four ??
     'these four black children'

(5)  Dem     N        Num  A
     cən     bun      kua sɛŋkɛd li
     DEM.PRX children four black ??
     'these four black children'

(adapted from Kouankem 2011:234-236; identify pp. & example nos.)

There are three kinds of adjective classes in Medumba, which differ in their order relative to the noun they modify. [GIVE EXAMPLES] [18]

There are a limited number of pure adjectives.[18] These are further divided into two classes. Class 1 pure adjectives always appear before their noun. Class 2 pure adjectives typically appear after the noun, but can appear before the noun in a contrastive context. [GIVE EXAMPLES]

[GIVE EXAMPLES]

[GIVE EXAMPLES]

Medumba has four locative prepositions: mbàŋ (next to), mʙəә́ (in front of), ɲàm (behind), and nùm (on).[37]

While there are several strategies for forming prepositional phrases involving overt prepositions or “locative specifications”, there are no overt locative prefixes which correspond to the Bantu locative classes (16), (17), and (18) (pa-, ku-, and mu- respectively).[28] However, there is a ''covert'' prefix in the form of a floating H tone whose presence may be detected in certain contexts by its effects on the pitch of surrounding tones, which Voorhoeve hypothesizes that it may be the remnant of those locative classes. The interpretation of a noun marked with this tone is variable and context dependent, generally corresponding to prepositions like on, at, or in. [Give examples]