ChapterPDF Available

The Omotic Language Family



Content may be subject to copyright.
The Omotic Language
Azeb Amha
26.1 Introduction
‘Omotic’ is one of the six language families within the Afroasiatic phylum.
It comprises about twenty-eight distinct languages, with some uncertainty
on the language–dialect distinction still remaining (cf. Unseth 1990;
Bender 2000, 2003; Hudson 2005). The languages are the following (alter-
native names in brackets): Anfillo, Aari, Basketo, Bench or Benchnon
(Gimira), Boro (Shinasha), Ch’aara, Dime, Dizi (Maji), Gamo, Ganza, Gofa,
Hamar, Haro (Gidicho), Hozo-Sezo, Kafa (Kefa), Koorete (Kore, Koyra,
Amaro), Kullo (Dawro), Ma¨le-Ganta, Maale, Malo, Mocha (Shekacho), Nayi
(Nao), Oyda, Sheko, Wolaitta (Welamo), Yemsa (Janjero, Yem), Zayse and
The family tree in Figure 26.1 is based on Fleming (1976).
With the exception of Ganza, which recent research has shown to be
spoken in the border areas between Sudan and Ethiopia, other Omotic
languages are found in Ethiopia. Most of these languages are spoken in
a contiguous area in the Southern Ethiopian Nations, Nationalities and
Peoples’ Regional State (SNNPRS). Outside of this state there are a few
enclave languages, namely Yemsa in Oromiya, Shinasha in Amhara
Regional state, Hozo-Sezo and Northern Mao in the Benishangul-Gumuz
Regional State.
The present contribution is based on my earlier overview paper on the
Omotic language family (Azeb Amha 2012). Some topics in the earlier
paper do not appear here for reasons of space. These include relative
clauses, content question words, ideophones as well as the classification
history of the language family and debate on its position within
Afroasiatic. The remaining topics are updated taking into account recent
publications. Some unreferenced data on Maale, Zargulla and Wolaitta are
based on my field notes and/or native-speaker intuition in the case of
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
In section 26.2, some salient features from the phonology of Omotic
languages are presented. A survey of the lexical and morphological cate-
gories of Omotic languages is made in section 26.3. Some aspects of the
structure of phrases and simple and complex clauses are discussed in
sections 26.4–26.9.
26.2 Phonology
26.2.1 Segments
The consonant segments in Table 26.1 are common to a large number of
Omotic languages.
The segments in brackets are well attested in the language family, but
their phonemic status or distribution differs across languages. Most of the
Kafa lgs.
Maji lgs.
Koyra (Koorete)
Janjero (Yem)
Gimira (Bench)
S.Mao (Anfillo)
Figure 26.1 Classification of Omotic languages
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
segments in Table 26.1 are included in Bender’s reconstruction of Proto-
Omotic consonant phonemes (2003: 310). The use of a full series of glotta-
lized consonants , ɗ, s’, ts’, tʃ’, k’ is common. Free variation among pand ɸ
at word-initial position is another common feature. However, in individual
languages or sub-branches, there are deviations from the common pattern.
These include:
0750500250 1000 km
0 100 200 300 400 500 miles
Konso Burji
Dizi Dime Oyda
Aari Maale
Yems a
Addis Ababa
See enlargement
Language groups
are shown with different
font styles
Extinct language
Map 26.1 Omotic languages. Reproduced with kind permission of Professor Dr Zygmunt
The Omotic Language Family 817
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
1 Absence or weak phonological status of some sibilants (s, s
voiceless labials (p, ɸ)and liquids r, l in the Gonga branch. In the
Gonga languages that have some of these consonants, it is claimed
that they are recent introductions through contact with Amharic (cf.
Theil 2007 on Kafa).
2. The presence of retroflex set ș,ʐ,ʨ, and ʨin Benchnon, Dizi, Nayi
and Sheko and their absence in other Omotic languages.
3. The lack of contrastive gemination in Benchnon and Dizi whereas
this is very common in other Omotic languages.
4. The segments ŋ,x,X,ɣand ʁin Dime, which are reported in
Mulugeta Seyoum (2008), that are absent in other Omotic languages.
5. The use of palatalized and/or labialized consonants in Benchnon,
Northern Mao as well as in the Dizoid languages: Dizi, Nayi and
Sheko, although scholars differ in their analysis of these as phone-
mic or phonetic. For example, Breeze (1990: 5) includes p
and labialized ones b
in the
phonemic inventory of Benchnon, but Rapold (2006: 101–3) con-
cludes that these are phonetic. Whatever their phonemic status,
these consonants are interesting since outside of the group of lan-
guages mentioned above, they are attested in Ethio-Semitic lan-
guages. Especially palatalization is known as a distinctive feature
of the Wollo dialect of Amharic.
A ten-vowel system comprising i, u, e, o and aplus corresponding long
vowels ii, uu, ee, oo and aa is common and it is reconstructed for Proto-
Omotic in Bender (2003). Benchnon is exceptional in not having long
vowels. It is claimed that this language compensated for loss of long vowels
with increased number of contrastive tones (cf. Wedekind 1985). Breathy
vowels are reported for Eastern Omotic languages Aari and Karo in
Hayward (1990) and Moges Yigezu (2007) respectively but are not attested
Table 26.1 Consonant inventory in Omotic languages
Bilabial Alveolar Palatal Velar Glottal
Stops Voiceless p t k ɂ
Voiced b d g
glottalized ɓɗ(ʧ)k
Fricatives Voiceless s (ʃ)h
Voiced (z) (ʒ)
Affricates Voiceless (ts) (ʧ)
Voiced (dz) (j)
Glottalized (ts)ʧ
Nasal m n
Flap r
Lateral l
Glides w y
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
outside of this branch. Contact with Surmic may have played a role here.
Finally, Dime, Hamar and Karo from East Omotic; Dizi, Nayi and Sheko
from the Maji-Dizoid branch; Anfillo and (in restricted words) Kafa from
the Kafa-Gonga branch are reported to have a phonemic low-central vowel
(schwa), which is absent in other Omotic languages. Example: ga
vs. gas
ˇot’ef, ‘grain type’ in Anfillo (cf. Debela Goshu and Girma Demeke
2005: 65). This widespread distribution suggests that, historically, schwa
probably had a more prominent presence in Omotic.
26.2.2 Tone
A large number of Omotic languages are tonal. Some languages, including
Ari, Gamo and Wolaitta have a tone-accent system. The latter have con-
trasting low and high tones but with a restricted and largely predictable
distribution of the high tone. Among the tonal languages the number of
distinctive tones varies, ranging from two to six phonemic tones. A two-
tone system is reported among others for Maale (Azeb Amha 2001), Kafa
(Taddese Ado 2001; Theil 2007) and Koorete (Theil 2013). Three-tone lan-
guages include Dizi (Beachy 2005), Nayi (Aklilu Yilma 1994, 2003),
Northern Mao (Ahland 2012) and Yemsa (Wedekind 1990). Sheko has
four tones (Hellenthal 2010). Benchnon has six tonemes: five level tones
plus a sixth rising tone from level two to three (Wedekind 1983; Breeze
1990; and Rapold 2006). A set of six minimal pairs in (1a) and a set of five
minimal pairs in (1b) (Rapold 2006: 119–20) illustrate the tonal contrast.
The numbers to the left of the examples represent the tone levels.
(1) a. 5 kar ‘clear’
4kari ‘enset or banana leaf’
3kar ‘to circle’
23 kar ‘a game with stones’
2kar ‘wasp’
1kar ‘loincloth’
b. 5 dab ‘many’
3dab ‘to follow’
23 dab ‘to work together (in communal work)’
2dab ‘a type of basket, made from leaves, hung over
a fireplace’
1dab ‘to sustain a sick person in walking’
Beside lexical contrast, tone designates grammatical information. For
example, gender distinction in third person singular pronouns is sig-
nalled by tone difference in some languages, e.g. in Shinasha:
‘she’ and in Yem: bar
‘he’ vs. bar
‘she’. In Dizi, possession can be marked
by tone, although an alternative possessive prefix k- exists. In Benchnon,
passive derivation of some verbs is marked by tone, e.g. k’a
´yts’ ‘Work!
Do it!’ and k’a
`yts’ ‘Be done!’ In some Ometo varieties, e.g. Maale,
The Omotic Language Family 819
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
a nominative and accusative case distinction in masculine and plural
nouns is marked by high and low tones respectively. Finally, tone is
used in stem formation within a single syntactic category or to indicate
derivational-semantic relation between two word classes. An example of
the former is tone pattern in Sheko verbs, in which a verb form can have
three different tone patterns depending on the paradigm in which it is
used (cf. Hellenthal 2010: 298–9): (1) tone pattern when it is used in the
singular imperative and in the jussive, (2) tone pattern when the verb is
used in ‘factual paradigms’ (e.g. realis, implicative, etc.) and (3) tone
pattern when it is used in ‘non-factual paradigms’ (including the irrealis,
negative and plural imperative). See also Rapold (2006: 265–82) for
a similar role of tone in Benchnon. The second case is the use of tone in
several languages to distinguish noun- and semantically related verb
stems. Leaving aside word-final vowels on nouns (see section
on the analysis of such vowels), we mention just a couple of noun/verb
pairs in two languages: ma
´ɗ-/maɗ-‘work/to work’, ʔo
´l-/ʔol- ‘war/to fight’ in
Maale (cf. Azeb Amha 2001a: 75); kja
`l- ‘egg/ lay egg’, bu
`t’- ‘fear/be
afraid’ in Northern Mao (cf. Ahland 2012: 184–92).
26.2.3 Syllable Structure
Monosyllabic words are not that common in the tone-accent languages.
Exceptions include personal and demonstrative pronouns, e.g. ha ‘this’, ne
‘you’ in Wolaitta. In these languages, consonant clusters are allowed to
a maximum of two in word-medial position, which are then split so that
the first member of the cluster becomes the coda of the first syllable, while
the second member is syllabified as onset of the following syllable, e.g.
mar.ka ‘witness’ in Wolaitta.
In highly tonal languages, e.g. Benchnon, monosyllabic words are fre-
quent and they tend to have closed syllables. Moreover, branching onsets
and codas are allowed in tonal languages, e.g. Benchnon has CCVC, CVCC,
CVCCC and CyVC (which requires the vowel aas nucleus) syllables. Finally,
the tonal languages Benchnon, Dizi, Sheko and Nayi use syllabic nasals n
and/or m
˙, whereas in other Omotic languages the syllable nucleus is
exclusively represented by vowels.
26.2.4 Morpheme-Structure Condition: Sibilant Harmony
The expression “sibilant harmony” is coined by Hayward (1988) to refer to
a well-formedness condition in words which requires that sibilant conso-
nants that occur in a word must have the same point of articulation: either
all of them are +palatal or palatal. In Benchnon, retroflex consonants also
play a role. The phenomenon is widely attested in Omotic languages, as
illustrated in (2), and it also affects loanwords, e.g. t’ı¨loʃ‘bride price’ from
Amharic is adapted as c
ˇ’ilooʃain Zayse. For these reasons, Hayward (1988,
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
2003) suggests sibilant harmony be reconstructed for Proto-Omotic. It is
not reported for other Ethiopian Afroasiatic languages.
(2) Aari: susa ‘relative’, zaazmi ‘cold, wet’; ʒotʃ’- ‘close’, ts’oots’i
Basketo: ʒungurʃa‘back bone’
Benchnon: sets’- ‘winnow’, zos ‘neighbour’; ʒatʃu‘maize flower’,
Dime: tʃ’aʒt’i ‘full’, sis’e ‘day’, k’aʃnaʃiʃ‘eight’ (from ¨stı¨n ‘one’ +
Dizi: tʃetʃ‘be safe’, tʃetʃı¨ʃtey ‘guard, make feel safe’ (causative
suffix is -s)
Mocha: ʃaatʃ’tʃ’- ‘bite’
Zayse: suuts ‘blood’, ts’unts’ ‘absorb’; tʃ’untʃ’ale ‘red ant’, ʒiitʃtʃe
(Hayward 1988: 267)
Hayward (1988: 267) notes that the harmony is also observed across mor-
pheme boundaries in extended stems or in inflected words. His examples
include s
ap-s- ‘make wet’, tʃ’ob-ʃ-‘make light’, șup-ș- ‘make soft’ in
Benchnon, in which the causative morpheme may be realized as -s,-ʃor
retroflex -șdepending on the sibilant consonant in the root. Likewise, in
Aari, the perfective aspect marker -sis changed to -ʃwhen the verb root has
a palatal sibilant. Compare: giʔer-s-it ‘I was hit’ vs. z
ˇ-it ‘I was thrown’
and s
ˇ-it ‘I was seen’.
Northern Mao has sibilant harmony as well as vowel place harmony in
terms of backness (cf. Ahland 2012: 67–70 and 76–9). The vowel harmony
involves only the stem of three or more syllable nouns (not the tv e): ı
‘pot’, t’e
`‘patch’, but: ts’u
`‘squirrel’, ʃu
`‘donkey’. Similar vowel
patterning is reported for ideophones in Maale (cf. Azeb Amha 2001a: 255)
and Wolaitta (Azeb Amha 2001b: 54). Zaugg-Coretti (2013: 51) reports
vowel assimilation in Yemsa.
26.3 Morphology
In the present section we examine some salient aspects of the morphology
of (pro)nouns, verbs, adjectives, demonstratives and numerals.
Omotic languages vary considerably with respect to synthesis and fusion
indexes (cf. Comrie 1989). The languages can be put in a different place in
a scale of isolating and fusional. Malo, Maale and Oyda, for example, would
be closer to isolating languages; Dizi, Hamar, Northern Mao, etc. closer to
polysynthetic-agglutinative ones, whereas Wolaitta, Gamo, Sheko, etc. can
be characterized as agglutinative with some fusional morphology.
The Omotic Language Family 821
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
Most Omotic languages use only suffixes. Dizi, Sheko and Northern Mao
use suffixes as well as some prefixes. Moreover, in Sheko, feminine gender
is marked by an infix.
26.3.1 Nominal Inflection
Throughout Omotic there is a strong interdependency among various
nominal inflectional categories. This is particularly obvious in the case of
definiteness, gender and case. Because of this we discuss these inflectional
categories together in subsections and Before dealing
with these topics, we introduce basic (uninflected) nouns in section Basic nouns
In many Omotic languages, basic uninflected nouns end in a short vowel,
labelled a ‘terminal vowel’ (hereafter tv) following Hayward (1987).
The number of vowels that are used as tvs varies from language to
language: Yemsa uses all five short vowels (and it also has some consonant-
final nouns): a
¯‘chicken/hen’, fa
¯‘sheep’, bo
¯‘slave’, ka
and ko
¯‘ball’. The number of words with tvseand iis much smaller,
and several of these are borrowed words from Amharic and Oromo
(cf. Zaugg-Coretti 2013: 73 and 333–65). Similarly, Dizi uses all five
short vowels as tvs: nibu ‘sorghum’, kasi ‘game’, kasa ‘sand’, gune ‘castrated
goat’, kilo ‘kilogram’ (cf. Beachy 2005). Maale uses four of its five vowels:
ˇi‘snake’, k’ase ‘elbow’, alo
´‘kidney’ and ye
´rga ‘axe’; the vowel uis not
used as a terminal vowel. Wolaitta uses only three of its five vowels for the
same purpose, e.g., zare
´‘lizard’, da
´bbo ‘relative’ and kana
´‘dog’, and there
are no nouns in Wolaitta that end in ior u. Northern Mao, on the other
hand, uses just one of its five short vowels: e, which might be realized with
an H, M or L tone depending on the tone class of the nominal (Ahland
2012: 41).
tvs are not entirely part of the nominal root because often they are
deleted when a morpheme is added to the noun (so-called ‘unstable tvs’).
They cannot be treated as suffixes either because largely no function is
associated with their presence synchronically, although in some lan-
guages, e.g. in Yemsa, they indicate (feminine) gender (cf. Zaugg-Coretti
2013: 73–5). In Northern Mao the tv has acquired a function of distinguish-
ing nominals (including nouns, pronouns, deverbal nominalizations) from
verbs (Ahland 2012: 41). In languages that have two or more tvs, the
distribution of the tvs is lexically determined and alternation between
forms is generally not possible. A few exceptions are reported for Koorete,
e.g. s
ˇeor s
ˇito refer to stone’ (Hayward 1987). In Wolaitta the cognate
of the same word is s
ˇaonly; and in Maale just s
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available Definiteness, Gender and Case
Generally, an indefinite form of a noun is not morphologically marked;
a definite form of a noun is marked either by a distinct morphological form
or in combination with other nominal grammatical categories such as
gender and case depending on the language. Semantic-functional nuances
of definiteness, its difference or overlap with specificity are not fully
understood in every language. Often the form interpreted as ‘definite’ by
linguists is distinctly marked, and it is established in discourse by a first
mention and/or by context.
Most Omotic languages make a two-way (masculine and feminine) gen-
der distinction in nouns, pronouns, on modifying categories such as
demonstratives and, in some languages, on verbs. However, Northern
Mao ‘does not exhibit any gender marking on any element’ (Ahland
2012: 41).
In pronoun and verb paradigms, gender distinction is made only in the
third person singular. Yemsa is exceptional in distinguishing feminine and
masculine singular as well as plural in pronouns and in some verb para-
digms (cf. Zaugg-Coretti 2013: 96, and Chapter 6 of this volume). Gender
assignment is based on biological sex in large animate nouns: nouns
referring to female beings are feminine and those that refer to male ones
are masculine. Largely, inanimate nouns or lower animate nouns are
categorically masculine or feminine depending on the language, with
the possibility for gender switch to express certain meanings, e.g. diminu-
tive, augmentative or affection (Azeb Amha 2006 and forthcoming, respec-
tively, discuss gender assignment in Maale and a text-based study on
gender-switch in Wolaitta).
Nominative, accusative and genitive cases feature as core case types in
many Omotic languages. These cases have varying forms depending on
definiteness and/or gender.
Cases such as the dative and ablative are often
built on either the accusative or the genitive form (see §
Typically, nominative and accusative case marking is differential since
these cases may not be affixed to nouns in certain grammatical contexts,
even though the nouns in question fulfil the syntactic roles which other-
wise lead to the assignment of the case morphemes in question. Most
Omotic languages have a marked-nominative system (cf. Ko¨ nig 2006).
However, Dime, Sheko and Yemsa have a (morphologically and function-
ally) unmarked nominative. Originally ‘differential case marking’ was
prominently discussed in relation to the object noun. However, in
Omotic languages both subject and object marking can be differential
(cf. Dimmendaal 2007 suggests differential object marking (DOM) as one
of the areal features North-East Africa).
In Gamo, indefinite object nouns, such as dangarsa ‘elephant’ in (3a),
occur without a morphological case marker. But when the noun is marked
for definite, it is also marked for case, as can be seen in (3b).
The Omotic Language Family 823
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
(3) a. s
ˇa-z-ii dangarsa woɗides
hunter-m:def-nom elephant:acc killed:3msg:pfv
‘The hunter killed an elephant.’
b. s
ˇa-z-ii dangarsa-z-aa woɗı
hunter-m:def-nom elephant-m:def-acc killed:3msg:pfv
‘The hunter killed the elephant.’
c. dangarsa-z-ii s
ˇa-z-aa woɗı
elephant-def-nom hunter-def-acc killed:3msg:pfv
‘The elephant killed the hunter.’
´1990: 364) Number and Definiteness
Most Omotic languages make a two-way, singular–plural distinction in
nouns and pronouns. Northern Mao is exceptional in distinguishing sin-
gular, dual and plural in both nouns and pronouns (cf. Ahland 2012).
A similar three-way number distinction is reported for Dizi but only in
the personal pronoun system (Beachy 2005).
In most Omotic languages singular is unmarked, whereas plural is
morphologically marked. In Aari, however, the singulative is marked on
the noun when it is definite (4c) while plural is unmarked. Definite nouns
that are not singulative can be interpreted as plural, and they may trigger
plural agreement in the verb (4b), but the noun itself is not marked for
plurality. The following examples are from Hayward (1990b: 444), glossing
added by the present author:
(4) a. tiile
´[unmarked indefinite]
water_pot be_present
‘There is a water-pot.’
b. tiile-na
´-k [definite, non-singulative =pl]
water_pot-def be_present-pl
‘There are water pots.’
c. tiile-s-ı
´yse [definite, singulative]
water_pot-sglt-def broke
‘The water-pot broke.’ (Hayward 1990b: 444)
Most languages have just one plural marker for all nouns (e.g. only -t-in
Wolaitta, see below). Others, including Koorete, Haro, Maale, Zayse and
Zargulla, have two or more plural markers, one used exclusively with
kinship terms. In Zayse, for example, the general plural is marked by the
morpheme -ı
´rand kinship terms are marked by -aats’ (Hayward
1990a: 247).
(5) ʔa
ˇa‘bed’ bars ‘younger kinsman’
ˇ-ir ‘beds’ bars
´ats’ ‘younger kinsmen’
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
In Wolaitta, all plural nouns are interpreted as definite, e.g.: keetta
house’, ʔaawa
´‘(a) father’ vs. keetta-t-a
´‘the houses (acc), ʔaawa-t-a
fathers’. Indefinite plural is expressed using quantifiers such as da
´ro ‘a
lot’ tʃ’o
´ra ‘numerous’ plus the singular form of the noun, e.g. tʃ’o
´ra keetta
‘(several) houses’. Secondary Cases or Case Stacking
In many languages locative, ablative, dative and instrumental cases obli-
gatorily follow accusative or genitive case. The following examples are
from Woliatta:
(6) a. naʔa
´-a-ssi mat’aa
´fa ʔimm-aa
child-m:acc-dat book:acc give-1sg:past:aff:decl
‘I gave a book to the boy.’
b. naʔ-e
´e-ppe mat’aa
´fa ʔekk-aa
child-f:gen-abl book:acc take=1sg:past:aff:decl
‘I took a book from the girl.’
c. naa-t-uu
´-ra mat’aa
´fa s
child-pl-pl:gen-comit book:acc buy-1sg:past:aff:decl
‘I bought a book together with/in company with the children.’
In Dizi the dative may be attached to the base form independently (7a),
or it may follow the accusative case -s- as in (7b):
(7) a. jaab-is s
ˇojt utn-k’aŋk ta-dɛ-ki
person-dat all love-inst give-ipfv-pfv:rel
‘the one who has been giving to all people with love’
b. tɛmari-a astɛmari-a-s-is t’agŋwɛrk’at wɛs-o
student-prox teacher-prox-acc-dat two paper send-3msg
‘The student sent two papers to the teacher.’ (Beachy 2005: 70)
Beachy (2005: 70–1) illustrates the dative and inessive case combination, as
in (8), and notes that ‘[t]his stack is narrower semantically than dative is
alone, since it always has the sense of being a location which the subject
goes to and enters’.
(8) gab-g-is giam tiɛ-nno
market-insv-dat yesterday go-1pl
‘We went to the market yesterday.’
There is no apparent semantic motivation for the combination of genitive
or accusative case with dative, ablative, etc. in the other languages
The Omotic Language Family 825
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
26.3.2 Verbal Inflection
In most Omotic languages the basic uninflected verbal lexeme is a bound
CVC(C) form to which affixes marking tense-aspect and other verbal cate-
gories are attached. But in Benchnon and Dime, the singular imperative
verb is not marked for subject/addressee; it is the simplest independent
verb form and can be considered as the basic verbal lexeme. As in a number
of other grammatical characteristics, Benchnon, Dizi, Northern Mao,
Sheko and Yemsa are united in distinguishing realis and irrealis verbs,
whereas such distinction is not reported for other Omotic languages.
In the present section we briefly discuss tense-aspect and negation mark-
ing. Subject agreement is discussed in sections 26.4.4–26.4.6. Tenseaspect distinction
Leaving aside terminological differences among researchers, some lan-
guages are reported to have aspect, e.g. Benchnon, Dime, Maale and
Sheko. Others are analysed as tense-marking languages, e.g. Koorete,
Wolaitta and Zargulla.
The three-way distinction in example (9a–c) may suggest that Maale has
a tense system with past, present and future tenses. However, the system
can be subsumed under two aspects: perfective and imperfective, because
the affirmative and negative future-intentional forms (9c and 10c) must
take the same morphemes (-a
´and –a
´nd) as the non-future, present and
present-progressive forms in (9b) and (10b).
(9) a. mukk-e
´-ne ‘I, you, he/she etc. came’ [Perfective]
b. mukk-a
´-ne ‘I, you, he/she etc. comes, is coming’ [Imperfective]
c. mukk-a
´-ne ‘I, you, he/she etc. will come’ [Imperfective]
(10) a. mukk-ı
´-se ‘I, you, he/she etc. did not come’ [Perfective]
b. mukk-uwa
´-se ‘I, you, he/she etc. does not come’ [Imperfective]
c. mukk-ı
‘I, you he/she etc. will not come’ [Imperfective]
Zargulla in contrast is best analysed as a tense-marking language
because of distinct affixes used for present/habitual, progressive, future
and past tenses, as illustrated in (11):
(11) ye
´wo ‘to come’ Infinitive
´nne ‘she came’ Past
´ne ‘she comes’ Present/habitual
´ne ‘she will come’ Future
´ne ‘she is coming’ Progressive
All Zargulla tense markers in example (11) end in -ne. It is possible to
analyse the morphemes further as, for example, tense-aspect markers -ı
and -e
´plus mood/modality or polarity marker -(n)ne. However, synchronic
morphosyntactic or semantic motivation for such analysis is not strong.
The morphemes in question are used in affirmative declarative and (polar)
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
interrogative clauses. In the negative, tense is reduced to past and non-past
distinction only (cf. Azeb Amha 2009b). Polarity
In many Omotic languages, negative polarity is morphologically marked
and positive polarity is unmarked. In Northern Mao, however, both are
formally marked. Compare the bolded forms in the future affirmative and
negative sentences in (12). The former involves the affirmative prefix ha-
and a lexical verb (12a), whereas in the negative (12b) the affirmative prefix
is dropped and the negative marker -a
´is suffixed to the infinitive stem.
The remaining verbal affixes are identical in the two sentences.
(12) a. ha-tjam-ga
‘I will count.’
‘I will not count.’
(Ahland 2012: 385)
Negation in declarative/indicative clauses is expressed by a suffix attached
to the verb. Some languages including Benchnon, Hamar, Shinasha and
Dime use a single negation marker in various tenses or aspects. Others,
e.g. Aari, have two morphemes whose distribution is determined by aspect:
perfective negative -k- and imperfective negative -y-, as in its-k-ite ‘I did not
eat’ vs. itsa-y-ite ‘I am not eating’ (cf. Bender 1991: 97). Gamo and Wolaitta
from the North Ometo branch, the East Ometo language Koorete, and
Yemsa, have special negative verb paradigms in declarative, interrogative
and imperative sentences. See sections 26.5.2 and 26.5.4 for examples.
Another widely attested characteristic is that a special negative mor-
pheme is used for marking a dependent negative verb, often translated as
‘without verb-ing’, among others, in Koorete, Wolaitta, Shinasha, Maale
and Haro. Moreover, in many languages negative non-verbal clauses are
marked by affixes distinct from those used in verbal clauses (see §26.5.5 on
non-verbal clauses).
26.3.3 Pronouns
The most commonly attested basic pronoun system among Omotic lan-
guages is one comprising eight pronoun forms, which indicate three
person distinctions (first, second and third), singular–plural distinction
for each of the three persons, masculine–feminine distinction only for
the third person singular, and a widely attested special third person ‘logo-
phoric’ or ‘reflexive’ pronoun. However, Dizi and Northern Mao distin-
guish dual pronouns as well; Yemsa distinguishes gender in plural
pronouns and Northern Mao lacks the logophoric/reflexive pronoun.
The Omotic Language Family 827
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
Generally, the pronoun sets mentioned above are manifested in at least
three case paradigms: possessive, nominative/subject and accusative/
object, as illustrated in (13) from Maale (Azeb Amha 2001a: 80):
(13) Possessive Subject Object
1sg ta ta
2sg ne ne
3msg ʔı
3fsg ʔı
1pl nu nu
2pl ʔı
´ntsi ʔı
3pl ʔiya
3log pe- – pe
The list in (13) shows that unlike other third person pronouns, the logo-
phoric pronoun, which is used in combination with an obligatory third
person pronoun or noun antecedent, does not distinguish gender and
number. Moreover, this pronoun does not have a subject form in many
Omotic languages. In Benchnon, however, the pronoun has a subject form
¯(14a), distinct from object and possessive form ba
(14) a. wȕs-a
refl:nom work-
said that she
worked.’ (Rapold 2006: 383)
b. peben-a
Pheben-nom:f refl:non_sbj cut_with_knife:sc-fs-newsit
‘Pheben cut herself.’ (Rapold 2006: 372)
c. dʒo
John-nom refl car trade-pfv- donkey trade-m
‘Having sold his
car, John
bought a donkey.’ (Rapold 2006: 374)
Inclusive–exclusive distinction in first person pronouns is made in
Benchnon (Rapold 2006), Zayse (Hayward 1990a), Koorete (Binyam
S. Mendisu 2008) and in Zargulla. The pronouns are given in Table 26.2.
Table 26.2 First person plural inclusive and exclusive pronouns in four
1pl exclusive (1+3) 1 pl inclusive (1+2)
Language Subject Object Subject Object
Benchnon nȍn
nú n nȉn
Koorete nuni nuu / numba nini nii / nimba
Zargulla núní núná níní/níi níná/nii
Zayse nú[j] núna níi nína
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
Benchnon has seven different paradigms of personal pronouns; six of
these function as subject pronouns, whereas one of the paradigms repre-
sents object pronouns (cf. Rapold 2006: 342; Breeze 1986, 1990).
The distribution of the subject pronouns depends on the discourse prag-
matic function.
26.3.4 Adjectives
Northern Mao is exceptional in not having adjectives as a lexical category
(cf. Ahland 2012: 232–6). In many Omotic languages the adjective is recog-
nized as a word class on morphosyntactic and semantic grounds, e.g. the
expression of property concepts (cf. Dixon 1982 and Dixon 2004).
However, the formal distinction between nouns and adjectives is not
straightforward, as these two categories share some grammatical features
such as taking obligatory TVs (terminal vowels) and the possibility to
undergo inchoative derivation. However, in modifying position, adjectives
must occur in their basic lexical form whereas nominal modifiers such as
possessor nouns take a genitive case marker or a linker (cf. §
In some languages, including Dime, Sheko and Yemsa, modifying adjec-
tives must agree with the head noun wherea modifying nouns do not. Also,
nouns are marked for plural by suffixes, whereas adjectives are redupli-
cated to express plurality.
In many languages adjectives have corresponding verbal forms. In some
cases the two have identical stems, e.g. in Wolaitta tʃ’ı
´m- ‘old’ in tʃ’ı
´ma ʔasa
‘old person’ and tʃ’im-iı
´si ‘he became old’. In others they are related but
distinct ʔaa
´ho ‘wide’ and ʔaa
´isi ‘it became wide’.
26.3.5 Demonstratives
A two-way distinction between proximal and distal (bound) demonstrative
forms is common in Omotic languages. In Wolaitta simple proximal and
distal demonstratives fall into three categories, with partial formal simi-
larity among some as can be seen in Table 26.3.
Table 26.3 Wolaitta basic demonstratives
Proximal (Deictic/
Deictic or anaphric Anaphoric
ha he híní ʔíní
ayáa híni
yaán- (intr.)
yaát- (tr.)
The Omotic Language Family 829
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
The nominal demonstratives may be used adnominally in which case they
often occur in the basic form shown in Table 26.3, e.g. ha naʔa
´y‘this boy’, he
´y‘that boy’. They can also be nominalized and used pronominally, e.g.
´‘this is a boy/child’. In the latter usage the basic demonstratives are
affixed with the nominalizers -ge (masculine) or -nn (feminine) plus gender-,
number- and case-marking morphemes that are used with other nominals
as well. The derived forms are given in Table 26.4.
Similar derivations and functions are attested, among others, in Dawro,
Dizi, Haro and Sheko.
Functionally, both basic and derived nominal demonstratives can be
used exophorically and anaphorically respectively pointing to entities in
the physical world in the speech-context and those identified in discourse.
The following examples are from Wolaitta:
(15) a. b-a
go-sg:imp dstl:n:dem-m:
´faa ʔekk-a
book:def:acc take-
‘Go, open that, that door over there and bring the book.’
b. he ʔuu
´ra g-ı
distl:n:dem ensete:def:m:acc ensete:sp say-ipfv:rel-
‘If one feeds that ensete type [mentioned earlier] which is known
as giʃı
´ra ...
The demonstratives ya
´a‘there’ and ha
´a‘here’ need the locative suffix -n if
only a locative meaning is intended: yaa
´n wotta
´‘put it there!’ Otherwise the
demonstratives may express direction or location depending on context,
e.g. ha
´a eha
´‘bring here/or towards speaker/!’ Example (16) is another
illustration of the directional meaning.
(16) ʔa
´-ppe ha
3msg-abl prox:adv:dem take-1sg:pres:neg:q
Table 26.4 Nominalized demonstratives in Wolaitta
Proximal masc. Distal masc. Proximal fem. Distal fem.
Sbj. hagéthis (one)hegéthat (one)hannáthis (one)hinnáthat (one)
Obj. hagáathis (one)Hegáathat (one)hannóthis (one)hinnóthat (one)
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
´-ppe ya
dist:adv:dem take-3msg:pres:aff:q
‘Shouldn’t I take (money) from him? Instead of he taking away
from me?’
(Lit. ‘Don’t I take here from him? Does he take there from me?’)
Interestingly, the distal local-adverbial demonstrative ya
´a‘there’ is used as
input in the verbal demonstratives yaa
´n- ‘be thus’ and yaa
´t- ‘do thus’ (see
Table 26.3). Compare (17a) and (17b) on their difference with regard to
(17) a. ha
´a yaa
‘Sweeping here (into the bowl in speaker’s hand), doing like
b. ʔagg-a
give_up-sg:imp say-ds:
‘[The brothers repeatedly] tell him “stop” he does not stop; (he)
who lived being like that ...
Dixon (2003:77) states that ‘in a number of languages, interrogatives fall
into the same paradigm as demonstratives’. This holds for the Wolaitta
demonstrative verbs discussed above since these have structurally parallel
interrogative verbs: waa
´n- ‘be.what’ and waa
´t- ‘do.what’
Some languages have deictic forms that give information about the
altitude of the location that is being pointed out relative to where the
speaker is found (cf. Hirut Woldemariam 2001 and 2007 on Dawuro and
Haro, Mulugeta Seyoum (2008) on Dime, among others).
26.3.6 Numerals
Most Omotic languages have a decimal system synchronically. In Wolaitta,
for example, the basic cardinal numerals 1–10 are ʔista
´, naaʔʔa
´, heezza
´puna and ta
The decimal system gives indications that it was historically developed
from a quintisimal system. See, for example, the highlighted parts of the
Wolaitta numerals for ‘six’, ‘seven’, ‘eight’ and ‘nine’ above. These respec-
tively contain segments that are found in the numerals ‘one’, ‘two’, ‘three’
and ‘four’. Based on such correspondences, Zaborski (1983) identifies three
historical typological systems in Omotic: (1) the numerals for ‘eight’ and
The Omotic Language Family 831
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
‘nine’ in Hamar, Karo, Ari, Chara, Zayse, Doko and Yemsa suggest that
these have a system going back to ‘two subtracted from ten’ and ‘one
subtracted from ten’. (2) Koorete and Central Ometo (including Gamo,
Gofa, Wolaitta) have a system based on addition to ‘five’, i.e. ‘six’ goes
back to ‘one plus five’, ‘seven’ to ‘five plus two’, etc. (3) Zayse, Yemsa and
Sheko use both systems: ‘six’ and ‘seven’ are formed through additions to
the base for ‘five’, while ‘eight’ and ‘nine’ are formed through subtraction
from ‘ten’. (Cf. Hirut Woldemariam 2005 for historical analysis of Ometo
numeral systems.)
Ordinal numerals are derived from the corresponding cardinal ones by
adding a special morpheme. For example, in Wolaitta -nta
´is suffixed and
tone-accent shift takes place as in naaʔʔanta
´‘second’ from naaʔʔa
In higher numerals some segments are deleted in the derivation e.g.
´‘eighth’ from ho
´na ‘eight’, not *hospunanta
´as might be
expected. A similar derivation is attested in languages from all branches
of Omotic.
26.4 Syntax
26.4.1 Phrasal Structure: On Modifier- or Complement-Head
Omotic languages are verb-final. Nominal modifiers precede the head
nouns; object nouns and adverbial-locative complements occur before the
verb. The position of the verb is generally stable, and constituent order
within a verb phrase does not show much variation across the languages.
For this reason, we do not discuss it further in the present section.
In contrast, the order in noun phrases is not strictly head-final in all lan-
guages. In some languages different positional restrictions apply to different
kinds of modifiers. For example, in Benchnon basic demonstratives occur
only after the head, whereas possessive nouns and pronouns only precede
the head noun. Relative clauses, adjectives and numerals/quantifiers, on the
other hand, can occur either preceding or following the noun they modify
(cf. Rapold 2006: 552–9). For Sheko, Hellenthal (2010: 251–6) reports that
possessives and adjectives have a fixed position relative to the head noun:
possessors always precede the head, while adjectives must follow the noun
they modify. Numerals and relative clauses, however, may precede or follow
the head noun. The preferred place for demonstratives in Sheko is following
the head noun, but they may also occur preceding it.
As Bender (1991: 88–90) and Mulugeta Seyoum (2008: 172–3) show, Aari
and Dime allow both head-modifier and modifier-head orders, as in the Aari
phrases: edin lak’amta ‘good people’ and lak’amta noqa ‘good water’. In both
languages relative clauses may occur before or after the head noun. In Ometo
languages such as Basketo, Haro Maale and Wolaitta, the modifier-head order
is most frequent in noun phrases; in this sequence the modifier occurs in its
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
lexical form, not marked for gender and number. However, if the word
order is changed to head-modifier, the modifier is marked for number,
gender and case. Thus in Basketo, either the adjective mints ‘strong’ or the
noun gabara ‘farmer’ may be affixed with the definiteness/number/accu-
sative case marker -(a)ntsi in the alternative expressions: mints gabara-ntsi
‘the strong farmers’ and gabara mints-antsi ‘the strong farmers’ (cf. Azeb
Amha 1995: 3).
Constituent order can be strict and grammatically important in
possessive phrases in which the basic sequence is possessor-possessed.
This is the case both in constructions in which the relation is morphologi-
cally marked by the genitive case, and in constructions in which the
relation is designated by word order only, as in the Dime examples zo
´ttu or zo
´t‘a lion’s leg’, which can be used alternatively (cf. Mulugeta
Seyoum 2008: 50). See also the Wolaitta examples in section
According to Rottland (1990: 195), Shinasha allows both possessor-
possessed and possessed-possessor orders. However, the possessed-
possessor order is not mentioned in Gebre B. Guadie’s description of
Shinasha (2013: 146–51).
26.4.2 Sentential structure
Typical constituent order in transitive clauses is AOV. Languages differ in
the relative flexibility of alternating this order. Some languages, e.g.
Northern Mao are more consistent in keeping this order. Likewise, in non-
verbal clauses the predicate is always the final constituent; subject and any
other constituent (e.g. possessive, dative complement) precede the predi-
cate. Agent/subject and object are marked by nominative and accusative
cases respectively. See sections and on case marking and
on verbal inflection respectively.
26.4.3 Argument Structure and Verbal Extension
In Omotic, morphological processes deriving verbs from other lexical
categories are rare. However, non-category-changing ‘derivations’ or
stem extensions of verbs are common. These include causative, passive/
reciprocal, intensive/pluractional, and inchoative. The following examples
are from Wolaitta:
(18) a.
´tta-a k’ant’-iı
wood-m:acc cut-3msg:past:aff:decl
‘He cut wood.’
´tta-a k’ant’-iss-iı
´si (causative)
wood-m:acc cut-caus-3msg:past:aff:decl
‘He made someone cut wood.’
The Omotic Language Family 833
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
´tta-y k’ant’-e
´si (passive)
wood-m:nom cut-pass-3msg:past:aff:decl
‘The wood is cut.’
´tta-a k’ant’-erett-iı
´si (intensive/pluractional)
wood-m:acc cut-ints-3msg:past:aff:decl
‘He cut many trees (or he cut one tree into several pieces).’
In Wolaitta and many other languages, passive, reciprocal and reflexive or
medial are not formally distinguished on the verb stem.
Where the meaning allows, combination of up to two derivational
morphemes can be made. Examples:
(19) ʔe
´-ʔadda ʔe
he-nom 3refl-father
‘He made his father be thanked.’ (Haro, Hirut Woldemariam
2003: 126)
And from Wolaitta:
(20) ʔas
´wa k’ant’-erett-iss-ı
meat-m:acc cut-freq-caus-ss:a:cnv
‘They got the meat cut into several pieces and gave it to the people.’
26.4.4 Subject-Agreement Marking on the Verb
None of the Omotic languages marks the object on the verb. Some Omotic
languages, including Haro, Maale and Malo, do not mark subject agree-
ment except in the imperative where second person singular and plural
are distinguished. In others, just a two-way distinction is made. For exam-
ple, affirmative declarative in Dime only distinguishes first person subject
from all other subjects, respectively marked by –tand -n:
(21) ʔate
´d-i-t ‘I came’
´d-i-t ‘we came’
´d-i-n ‘he came’
´d-i-n ‘she came’
yaye ʔa
´d-i-n ‘you (sg) came’
´d-i-n ‘you (pl) came’
´d-i-n ‘they came’
(Mulugeta Seyoum 2008: 128)
A parallel construction in Basketo also has just two morphemes -i- and -a-,
used depending on the number or gender of the subject: -i- is used when
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
the subject is plural or third person masculine, whereas -a- indicates that
the subject is singular, including third person feminine:
(22) 1sg tani lukk-a-de ‘I went’
2sg neni lukk-a-de ‘you went’
3fsg iza lukk-a-de ‘she went’
3msg ii lukk-i-de ‘he went’
1pl nuni lukk-i-de ‘we went’
2pl yinti lukk-i-de ‘you went’
3pl inti lukk-i-de ‘they went’
(Azeb Amha 1995: 5)
Benchnon as well as the East Ometo languages Koorete, Zayse and Zargulla
have elaborate subject-agreement inflection, using distinct verbal mor-
phemes corresponding to the eight distinct subject pronouns including
the first person plural inclusive and exclusive ones, as illustrated in (23).
Note that the consonantal segments in the verbal subject-agreement mar-
kers correspond to segments in the independent pronouns.
(23) (ta
´) yee
´nne ‘I came’
´) yee
´nne ‘you (sg) came’
´) yee
´nne ‘he came’
´) yee
´nne ‘she came’
´) yee
´nne ‘we (excl) came’
´) yee
´nne ‘we (incl) came’
´ni) yee
´nne ‘you (pl) came’
´ni) yee
´nne ‘they came’
In contrast, in the North Ometo languages Gamo, Gofa, Dawro (also known
in the literature as Kullo) different sets of subject-agreement morphemes
are used depending on tense-aspect, mood and polarity values. This typo-
logically rare system of distinguishing various sentence types through
inflectional means is discussed in section 26.5.2.
26.4.5 DisplacedSubject-Agreement Markers
In a few Omotic languages it has been observed that verbal subject-
agreement suffixes may be attached to a non-verbal category in
a sentence. The phenomenon is observed in some languages from different
branches of Omotic that are not geographically contiguous: Zayse and
Zargulla from East Ometo, Sheko from the Maji/Dizoid group of North
Omotic and Aari from south Omotic. In the East Ometo languages Zayse
and Zargulla, for example, verbal pronominal elements move out of the
verb and are affixed to a focused constituent in the clause (cf. Hayward
1990a, Azeb Amha 2007a and 2007b). In example (24), the subject-
agreement marker -s is part of the verb, as this category is in focus (see
The Omotic Language Family 835
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
also the paradigm in (23) above). In (25), however, the same suffix is
attached to the focused form guta
(24) ʔe
3msg:nom tomorrow arrive-ints-foc-3msg-fut
‘He will arrive tomorrow.’
(25) ʔe
3msg:nom tomorrow-foc-3msg arrive-ints-fut
‘He will arrive tomorrow
With content question words which are inherently in focus, the same
obligatory displacement is observed:
(26) a. ʔaa
´nde-n yee
when-2sg come:past
When did you (sg) come?’
b. ʔaa
´nde-it yee
when-2pl come:past
When did you (pl) come?’
c. ʔas’o-y ʔa
man-nom where-3msg exist:pres
Where is the man?’
´-y ʔa
woman-nom where-3fsg exist:pres
Where is the woman?’
In examples (27a) and (27b) the object content question words:
´nnesa‘which (m:obj)’ and ʔa
´nniʃa‘which (f:obj)’ are respectively
marked by the morphemes -nand -us, coindexing the second person sin-
gular and third person plural subjects of the two sentences.
The independent pronouns in parentheses are optional:
(27) a. (ne
(2sg:nom) which-3msg-acc-2sg see-past
‘Which one (m) did you (sg) see?’
b. (ʔu
´nn-iʃ-a-us s’ee
(3pl:nom) which-3fsg-acc-3pl:sbj see-past
‘Which one (f) did they see?’
A similar phenomenon is observed in Ari, where in the interrogative, the
adverbial complement rather than the verb is assigned verbal subject
agreement. Thus, in (28), seni ‘tomorrow’ is marked for person and number
of the subject; the morpheme -d- on the verb its(i) ‘eat’ marks the imperfec-
tive and the final morpheme -oindicates that the sentence is interrogative.
Parallel to this, in the ‘imperfect continuous interrogative’ in (29) we find
similar subject-agreement markers (bolded) between the two reduplicants
of the verb its(i) ‘eat’. The grammatical or pragmatic motivation for the
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
shift of subject-agreement markers in Aari is not stated; it is probable that
focus plays a role here, too. The data are from Bender (1991: 99–100):
(28) 1sg ita sen-iitsi-d-o ‘Will I eat tomorrow?’
2sg ana sen-aa itsi-d-o ‘Will you eat tomorrow?’
3sg na/no sen-uk itsi-d-o ‘Will s/he eat tomorrow?’
1pl weta sen-oitsi-d-o ‘Will we eat tomorrow?’
2pl yeta sen-eitsi-d-o ‘Will you (pl) eat tomorrow?’
3pl keta sen-ak itsi-d-o ‘Will they eat tomorrow?’
(29) 1sg its-i-itsi-d-o ‘Am I eating?’
2sg its-a-itsi-d-o ‘Are you eating?’
3sg its-uk-itsi-d-o ‘Is s/he eating?’
1pl its-o-itsi-d-o ‘Are we eating?’
2pl its-e-itsi-d-o ‘Are you (pl) eating?’
3pl its-ek-itsi-d-o ‘Are they eating?’
In Sheko too, verbal subject-agreement markers are displaced and
affixed to a focused constituent (Hellenthal 2010: 448).
26.5 Sentence-Type Distinctions
One of the interesting features of Omotic is that many of its member
languages overtly mark the function of a sentence as a statement (declara-
tive), a request/question (interrogative) or command or wish (imperative/
optative). The strategy used in making these distinctions differs from
language to language. These are briefly discussed below.
26.5.1 Declarative and Interrogative Clauses: Edge Marking
In some languages, interrogatives are distinguished from declarative
clauses by adding an invariable affix to the verb, which may or may not
be accompanied by special question intonation.
Different affixes may be used to mark the same sentence type depending
on polarity or tense-aspect values. This is the case in Maale (cf. Azeb Amha
2001a: 147–58). Compare the question-and-answer pairs in (30a–d). Verb-
final -ı
´ya in (30a) and -a
´in (30c) indicate that the verbs they are affixed to
respectively head perfective and imperfective interrogative sentences.
These two contrast with the corresponding declarative sentences in (30b)
and (30d) in which aspect is distinctly marked from sentence type: the -e
and -a
´alternation marks perfective and imperfective aspect respectively,
whereas -ne designates the two sentences are affirmative declarative.
(30) a. ʔatsı
man:m:nom come-pfv:q
‘Did the man come?’
The Omotic Language Family 837
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
b. ʔatsı
man:m:nom come-pfv-aff:decl
‘The man came.’
c. ʔatsı
man:m:nom come-ipfv:
‘Is the man coming?’
d. ʔatsı
man:m:nom come-ipfv-aff:decl
‘The man is coming.’
The affirmative declarative marker -ne (30b and 30d) contrasts with the
negative declarative marker -se in (30e–f):
(30) e. ʔatsı
man:m:nom come-pfv:neg-neg:decl
‘The man did not come’.
f. ʔatsı
man:m:nom come-ipfv:neg-neg:decl
‘The man is not coming.’
Northern Mao morphologically distinguishes two subtypes of declara-
tive sentence-type markers on the verb: simple declarative and hearsay
declarative. The interrogative, imperative and jussive-optative sentences
each have their own characteristic endings. In (31) we illustrate the
declarative and interrogative clauses:
(31) a. ra
Ram-sbj aff-go_out-decl
‘Rama left.’
b. ra
Ram-sbj aff-go_out-hrsy-decl
‘Rama left (they say).’
(Ahland 2012: 472)
c. ı
`n ha-ho
def woman-sbj 2sg-com aff-go-intr
‘Did the woman go with you?’
(Ahland 2012: 474)
In interrogatives with content question words, the affirmative prefix ha-
must be dropped:
(31) d. kı
who-sbj 1sg-gen porridge-obj eat-intr
Who ate my porridge?’
(Ahland 2012: 476)
In Koorete affirmative declarative and interrogative sentences are marked
with -so and -a, respectively (cf. Binyam S. Mendisu 2008: 124–6).
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
Besides the sentence-final (edge-marking) discussed in the present sec-
tion, two other interesting morphological strategies are attested in inter-
rogative-declarative distinction in some Omotic languages, as discussed
next in sections 26.5.2 and 26.5.3.
26.5.2 Special Interrogative Verb Inflection
One unique feature of Omotic languages, especially languages from the
Ometo branch e.g. Wolaitta, Benchnon and Aari, is having different verb
inflection for the declarative and interrogative (cf. Hayward 1995, Bender
1991, Breeze 1990, Rapold 1996, Azeb Amha 2012). The data from Wolaitta
in (32) illustrates this. For the same person, number and gender, we find
different exponents in different sentence types (declarative vs. interroga-
tive), polarity (affirmative vs. negative), tense (present vs. past; the future
declarative has an invariable form ʔekkana
´‘I, you, etc. will take’ but future
interrogative inflects for different person, number and gender):
(32) a. Affirmative declarative
Present Past
´si ‘I take’ ʔekk-aa
´si ‘I took’
´sa ‘You take’ ʔekk-a
´dasa ‘You took’
´si ‘He takes’ ʔekk-iı
´si ‘He took’
´su ‘She takes’ ʔekk-aa
´su ‘She took’
´si ‘We take’ ʔekk-ı
´da ‘We took’
´ta ‘You (pl) take’ ʔekk-ı
´deta ‘You (pl) took’
´sona ‘They take’ ʔekk-ı
´dosona ‘They took’
b. Affirmative interrogative
Present Past Future
‘Do I take?’ ʔekk-
‘Did I take’ ʔekk-
I take?’
´y‘Do you
‘Did you
‘Will you
´‘Does he
‘Did he
‘Will he
´y‘Does she
‘Did she
‘Will she
‘Do we
‘Did we
‘Will we
‘Do you (pl) take?’
‘Did you
‘Will you
‘Do they
‘Did they
‘Will they
The negative declarative in Wolaitta merges the present and future
tenses. Thus, in the negative the distinction is between past and non-
past, as shown in (32c) for the same verb ʔekk- ‘take’:
The Omotic Language Family 839
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
(32) c. Negative declarative
Present/Future Past
´kke ‘I don’t/won’t take’ ʔekk-a
´kke ‘I did not take’
´‘You don’t/won’t take’ ʔekk-
‘You did not take’
´‘He doesn’t/won’t
‘He did not take’
´‘She doesn’t/won’t
‘She did not take’
´kko ‘We don’t/won’t take’ ʔekk-ı
´kko ‘We did not take’
‘You (PL) don’t/won’t
‘You (PL) did not
‘They don’t/won’t
‘They did not
In the negative-interrogative of Wolaitta verbs, the same past and non-
past distinction is made as in the negative-declarative:
(32) d. Present/Future Past
‘Don’t/won’t I take?’ ʔekk-
‘Didn’t I take?’
´‘Don’t/won’t you
‘Didn’t you
‘Doesn’t/won’t he
‘Didn’t he take?’
´‘Doesn’t/won’t she
‘Didn’t she take?’
‘Don’t/won’t we
‘Didn’t we take?’
‘Don’t/won’t you (pl)
‘Didn’t you (pl)
‘Don’t/won’t they
‘Didn’t they
26.5.3 Reductive Morphology
In a few Omotic languages, declarative and polar interrogative sentences
contrast by the absence of certain grammatical morphemes in the inter-
rogative sentence. The missing element is a person, modality or focus
marker that may not be expected to have direct bearing on sentence-type
distinction. That is why we use the notion of ‘reduction’ instead of ‘zero
marking’. Interestingly, the languages that exhibit the phenomenon are
not geographically contiguous and belong to different branches of the
family: Zayse and Zargulla from East Ometo, Dime from Eastern Omotic
and Sheko from the Maji-Dizoid group. This suggests that probably the
strategy is/was broadly used in Omotic.
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
In Dime the verb is obligatorily marked with subject-agreement mor-
phemes in perfective and imperfective affirmative-declarative clauses.
In the corresponding interrogative clauses, however, first and third person
subject-agreement markers must be dropped (cf. Mulugeta Seyoum 2008).
The author states that the perfective interrogative aspect marker -ibears
high tone, but next to this prosodic indication, the contrast in the marking
of verbal-grammatical categories is important. The paradigms in (33) illus-
trate this:
(33) Perfective affirmative declarative Perfective interrogative
´d-i-t ‘I came.’ ʔate
´‘Did I come?’
wɔtu ʔa
´d-i-t ‘we came.’ wɔtu ʔa
´‘Did we come?’
´d-i-n ‘he came.’ nu
´‘Did he come?’
´d-i-n ‘she came.’ na
´‘Did she come?’
´d-i-n ‘they came.’ ke
´‘Did they come?’
(Mulugeta Seyoum 2008: 166)
The second person interrogative in Dime behaves differently, since in this
case a specific interrogative marker -a
´is used:
(34) ya
´d-i-n ‘you (sg) came.’ yaye ʔa
´‘Did you (sg) come?’
´d-i-n ‘you (pl) came.’ yese
´‘Did you (pl) come?’
(Mulugeta Seyoum 2008: 166)
In Sheko declarative sentences, ‘non-future’ and ‘future’ affirmative
declarative verbs are marked by the morphemes -ke and -me respectively.
Corresponding non-future and future polar interrogative sentences do not
take these affixes. Otherwise the two sentences have exactly the same
structure. The question-and-answer pairs in (35a) and (35b) from
Hellenthal (2010) illustrate the contrast.
(35) a. sook’-a
´ahee sook’-a
sleep-3msg:q yes sleep-3msg-nf:decl
‘Did/does he sleep?’ ‘Yes, he slept/is sleeping.’
b. gabak’a a-tag-a
market 2sg-go-fut:q 1(sg)-go-fut-f:decl
‘Will you go to the market?’ ‘I will go.’
In Zargulla, polar interrogative clauses differ from their corresponding
declarative clauses in that the verb of the declarative clause must have the
focus marker -tte- (36b), while this must be dropped in the interrogative
counterpart (36a). The same phenomenon is observed in Zayse (cf.
Hayward 1990a).
(36) a. ʔe
3msg:nom tomorrow arrive-ints-3msg-fut
‘Will he arrive tomorrow?’
The Omotic Language Family 841
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
b. hoo,(ʔe
´) guta
yes, 3msg:nom tomorrow arrive-ints-foc-3msg-fut
‘Yes, he will arrive tomorrow.’
26.5.4 Imperative
In most languages the imperative is morphologically marked.
Benchnon and Dime, however, use the basic lexical form of the verb
for imperative singular without additional marking. In the plural
imperative an additional morpheme is added to the lexical form as in
the Benchnon examples: so
¯ʔkȉt‘draw water (sg)!’ vs. so
water (pl)!’
Wolaitta, along with Koorete, Haro and Maale, is an example of the
languages that mark both the singular and the plural imperative.
Examples from Wolaitta:
(37) a. kiı
message-m:nom tell-2sg:imp
‘Tell the message!’
b. kiı
message-m:nom tell-2pl:imp
‘Tell the message!’
A number of Omotic languages have a special polite imperative form.
Such languages include among others Maale, Wolaitta and Dizi.
The following examples are from Dizi (Beachy 2005: 109):
(38) a. wu-ŋ
b. ʔn
-zur-a-s-e e koj-n-dɛj
-voc yes say-imp-pol:imp
‘My relative, say “yes”.’
c. jin-ɨ-kojs jɛ-n-ti
I-v-All come-imp-pl:imp
’Come to me (pl)!’
Note that in the examples in (38), the plural and polite imperatives are
built on the singular one, by adding additional morphemes -ti and -dɛjto
the already imperative form.
Negative imperative and negative optative are marked by a special nega-
tive morpheme in many Omotic languages. This is the case in Aari, Dime,
Dizi, Haro, Maale, Koorete, Wolaitta, Zayse and Zargulla. For example, Dizi
generally uses a verbal negative proclitic nan- and the suffix -
neously (see the bolded morphemes in example (39a)). In the negative
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
imperative and optative, however, only the suffix -is is added to the
verb (39b).
(39) a. c
ˇ’abt-bab-a nan-a
illness-poss:m-prox neg:vb-3msg-get_well-pfv-neg
‘The sick man has not gotten well.’
b. ʔŋ
‘Let’s not steal!’
(Beachy 2005: 105)
In Hamar, negative imperatives can only be formed by using the verb
¨‘leave, stop’ after the negated lexical verb, as in kuma
¨‘eat!’ vs. kuman
¨‘don’t eat!’; the bound negative morphemes -ma
¨and -ma cannot be
used for this purpose (Lydall 1976: 427). Similarly, Benchnon uses one of
the converb forms of a lexical verb and a positively inflected form of the
verb s
´d‘remain’ to express negative imperative, whereas in other contexts
the bound affixes -arg
or -ar
mark negation (cf. Breeze 1990 and Rapold
(40) a. ham-a
´Benchnon (Rapold 2006: 259)
‘I did not go.’
b. ha
‘Go, please!’
c. ha
Benchnon (Breeze 1990: 37)
go-cnv:2sg remain-2sg
‘Don’t go (2sg)!’
Imperative and optative form a single inflectional pattern in many
Omotic languages, except that the former concerns second person
whereas the latter involves first and third persons. For example, in
languages that use different verb stem classes, the two take the same
stem. Or they have a special negative form that is not used in other
verb forms. The following example is from Zargulla (Azeb Amha
2009b: 212).
(41) a. Negative Imp/Optative: b. Affirmative Imp/Optative:
‘Let me not go’ ʔepp-ana
´‘Let me take’
‘Don’t go!’ ʔepp-a
´‘take! (2sg)’
‘Let her
not go’
‘Let her take’
The Omotic Language Family 843
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
‘Let him
not go’
´sse /
‘Let him take’
‘Let us not
go (inc)’
´nno ‘Let’s take (inc)’
‘Let us not
go (exc)’
´nno ‘Let’s take (exc)’
‘Don’t go!’ ʔepp-a
´-ite ‘take! (2pl)’
‘Let them
not go’
´sso /
‘Let them take’
´‘Let him/her/
them take!’
26.5.5 Non-Verbal Equative Clauses
In a number of languages, attributive or equational non-verbal clauses in
the present tense are morphologically unmarked. For example, Wolaitta
has no present or past tense copula marker. In the future tense form, it
uses the lexical verb gid- ‘be enough’, as in gos
ˇa gid-ana ‘(he/she) will be
a farmer’ (cf. Azeb Amha 2007c). In contrast, Dizi distinguishes three
copulas, which are obligatorily used: the attributive copula ti-, the existen-
tial copula -ki- and the negative copula nan- (identified as ‘auxiliary nega-
tive verb’ in Beachy 2005). The three copulas can be inflected for the
number or gender of the subject of the nominal clause (42a–d).
(42) a. yaaba jes
man good be-3msg
‘The man is good.’
b. bac
ˇ-a aŋgat ki-go
clothes-prox here ext:be-3msg
‘The clothes are here.’
c. nan-ı
neg-3fsg it-loc ext:be-neg
‘There is no one (f); There are none.’
d. kek tɛ-j,ee kek ti-go
correct be-fut:q yes, correct be-3msg
‘Is it correct? Yes, it is correct.’
There are languages that have a present tense copula but use it
optionally or in a pragmatically restricted context. The latter case
reflects the situation in Haro, as described by Hirut Woldemariam
(2003): in Haro the addition of the copula -kko emphasizes the nominal
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
predicate; pragmatically neutral attributive clauses in this language
occur without this morpheme. Compare: ha
´ti ‘This is a chief’ vs.
´ti-kko ‘This is achief’. Aari, Dime and Kullo (or Dawro) have
a copula, which is optional for present tense attributive and equational
clauses. From available literature, it seems that the absence or presence
of the copula in these languages is not associated with a special func-
Past and future tense attributive/equational clauses often use an obliga-
tory copula verb, e.g. Dime (cf. Mulugeta Seyoum, 2008). Moreover, unlike
attributive/equational clauses, existential non-verbal clauses tend to be
morphologically marked by a special copula or are headed by lexical
verbs corresponding to ‘sit’, ‘exist’ or ‘live’ in English.
In the negative, an obligatory copula or a verbal form heads attributive
or equational non-verbal clauses and existential clauses. In Aari, for exam-
ple, there is a general non-verbal negative copula, daki, used as in laqami
daki ‘it is not good’, dofen daki ‘it is not true’ and etsina maana daki ‘the man is
not a smith’ (cf. Bender 1991). Others mark the negative on equivalents of
the English verbs ‘become’ or ‘happen’, e.g., in Wolaitta the expression
´kka ‘You (sg) are not a farmer’ contains the second person
singular negative verb gid-a
´kka ‘you will not become, you will not be
26.6 Complex Sentences
Dependent clauses such as the conditional, concessive, adversative and
converb are morphologically distinguished from main clauses by special
endings attached to the verb of the dependent clause (43). Generally,
dependent clauses occur before the matrix verb, but cases of the reverse
order are reported for the conditional (cf. Azeb Amha 2001a).
(43) a. The adversative in Maale (Azeb Amha 2001a: 188)
go-f:ipfv- rain-
‘They would go to Jinka, but it rained hard.’
b. The conditional in Wolaitta
´-w ko
1sg give_up-
‘If you want to stay longer (here), I will leave you.’
The Omotic Language Family 845
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
26.6.1 The Converb
The converb is another dependent clause type that is widely attested in
Omotic. The following examples from Wolaitta illustrate the construction:
(44) mas’aa
book:def:acc take-ss:a:cnv
‘He sent (her) saying “bring the book”’
The highlighted affixes -a
´indicate number and gender distinc-
tion. The former is used with all singular subjects except the third person
masculine. In (44) it indexes the second person singular addressee.
The morpheme
´, on the other hand, marks third person masculine
singular and all plural subjects. As such the inflection of the converb is
much less elaborate compared to Wolaitta main clauses (cf. §26.5.2).
On the other hand, the converb is the only inflecting dependent clause
type in the language. Some languages, e.g. Zayse, Zargulla, have converbs
that do not inflect at all. Others, e.g. Aari and some converb types in Yemsa,
have converbs that inflect for each person, number and gender. Wolaitta is
thus in between these extremes.
The most common function of the converb is to indicate temporal
information such as anteriority or simultaneity of the event expressed in
the dependent clause in relation to the matrix clause. But the converb
form itself is not marked for tense-aspect. The converb is also used for
clause chaining, in which switch reference is marked. This is discussed in
the next section. (For a survey of the converb in Omotic, see Azeb Amha
and Dimmendaal 2006b.)
26.6.2 The Converb and Switch Reference
The converb is one of the few dependent clause types in Omotic in which
switch-reference is morphologically marked. Different-subject converb
and the same-subject converb are morphologically distinguished in most
Omotic languages. The former indicates that the agent/subject of the state
of affairs expressed by a converb is different from that of the immediately
following converb (or series of converbs) or main-clause verb. Consider
example (45), where there are three converb clauses
and four subject-
switches. Each verb is marked by a different-subject marker to indicate this
(45) b-iı
ˇin b-iiiis
ˇin saa
´-y k’amm-ı
earth-nom be_dark-ds:
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
´tta giddo
´-n asa
´-y ak’-ı
´n kap-ı
tree:acc inside-
‘While (the people) were going (a very long distance), it became dark
and the people having spent the night in the forest, the bird ....’
The same-subject converb, on the other hand, indicates that the agent/
subject of a converb is the same as the subject of the subsequent converb-
(s), other dependent clauses and/or the main-clause verb. In contrast to
example (45), in (46), the subject of the four consecutive converb clauses is
the same (3msg subject), and this lack of switch is marked on each verb by
the short (-ı
´)and long (-ı
´) forms of the anterior converb.
(46) he pars-u
´wa Ɂúy-í ka
that beverage-
Ɂoiddú Ɂuraa gid-ı
four man:acc become-
‘Having drunk that beverage, having been satiated, having become
as fat as four men and risen ...’ (Adams 1983: 163)
Switch-reference system is reported among others for Benchnon, Dizi,
Maale, Sheko, Zargulla and Yemsa.
26.6.3 The Converb and Verb Compounding or Complex Predicates
The same-subject anterior converb is used in many Omotic languages to
form complex predicates or verbal compounds. Such languages include
Benchnon, Dime, Maale, Wolaitta, Yemsa and Zargulla. Various authors
have examined Wolaitta V
combinations of which V
must be
a shortened form of the same-subject anterior converbs -ı
´and -a
The construction is used to express various meanings including manner,
immediacy, completion, etc. See Azeb Amha and Dimmendaal (2006a) and
Azeb Amha (2010) for details. In (47) and (48) we present a few examples,
showing the gender contrast in the (a) and (b) forms:
(47) a. mokk-ı
‘He disappeared after having been seen briefly.’
b. mokk-a
‘She disappeared after having been seen briefly.’
(48) a.
come_out- ss:Sh.a:cnv
‘He arrived unexpectedly.’
The Omotic Language Family 847
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
‘She arrived unexpectedly.’
26.7 Concluding Remarks
In this chapter we presented an overview of several aspects of the phonol-
ogy, morphology and syntax of Omotic languages which would be of inter-
est in typological comparison. Some features that are widespread in Omotic
are not that common in other Ethiopian languages. These include robust
tone, terminal vowels, sibilant harmony, demonstrative and interrogative
verbs, switch reference, and special interrogative and negative paradigms,
all of which are found in every branch of Omotic. Onthe other hand, there is
significant diversity within the language family. This is evident from the
frequently attested cases of languages that are exceptions in having or not
having particular morphosyntactic properties. For example, with regard to
gender marking, while a two-way (masculine–feminine) gender distinction
is common throughout the family, Northern Mao is unique in not marking
gender at all. Similarly, almost all languages distinguish pronominal and
verb-agreement gender in the third person singular. Yemsa is exceptional in
making feminine–masculine distinction in the pluralas well. Some ofthese
could probably be explained by language contact, but others may be indica-
tions of the historical depth of the language family.
I worked on the present contribution during my fellowship at the LCRC in
the James Cook University in Cairns, Australia in 2013 and 2014. I am
grateful to the directors of the Centre, Professors Alexandra Aikhenvald
and R. M. W. Dixon for their helpful comments, suggestions and for their
hospitality. I would like to thank Professors Ton Dietz, Harry Wels and Jan
Abbink at the African Studies Centre in Leiden for their support.
Colleagues at the LCRC enriched my work through helpful discussions
and through their friendship: thanks to Amanda Parsonage, Angeliki
Alvanoudi, Brigitte Flick, Diana Foker, Valerie Guerin, Elena Mihas,
Simon Overall, Nick Piper and Hannah Sarvasy.
1. In Benchnon the bilabial syllabic nasal m
˙is attested only in two related
words and its phonemic status is questionable (cf. Rapold 2006: 57–62).
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
2. Some languages have a restricted number of nouns that end in
a consonant. The latter tend to be sibilant. Comparison across lan-
guages suggests that tvs following sibilant consonants get deleted.
3. The vocative is another case type that has different forms depending on
the gender and number of the noun, e.g. ʔa
´‘young boy!’ vs. ʔa
‘young girl’ in Maale. This case form is not further discussed in the
present work (see Azeb Amha 2012 for an overview of this case in
various Omotic languages).
4. Glossing in the original work is simplified and adapted to make it
closest to the convention used in this work, e.g. 3msg = ‘third person
masculine singular’ rather than pm = ‘person marker’.
5. Benchnon has several subject forms. This is the ‘long strong subject
form’. For other forms see Rapold (2006: 472).
6. In this example, we count the first two verbs as part of a single clause,
because they are repetitions to express duration. Note that the speaker
further indicates the very long duration of the motion event by the
extra-lengthening of the first vowel [i] in the duplicated verb b-iiiis
Adams, B. A. 1983. A Tagmemic analysis of the Wolaitta language. PhD thesis,
School of Oriental and African Studies, University of London.
Ahland, Michael Bryan. 2012. A grammar of Northern Mao (MA
PhD thesis, University of Oregon.
Aklilu Yilma. 1994. A sketch of the Nayi grammar. SLLE linguistic reports
16: 1–20.
2003. Comparative phonology of the Maji languages. Journal of Ethiopian
Studies 36(2): 59–88.
Allan, Edward J. 1976b. Dizi. In M. L. Bender (ed.), The non-Semitic languages of
Ethiopia, pp. 377–92. (Monograph No. 5, Occasional Papers Series,
Committee on Ethiopian Studies 1). East Lansing: African Studies
Center, Michigan State University.
Azeb Amha. 1995. Case in Basketo. African Languages and Cultures 28
(1): 1–17.
1996. Tone-accent and prosodic domains in Wolaitta. Studies in African
Linguistics 25(2):111–38.
2001a. The Maale language. CNWS Publications, Vol. 99. Leiden University.
2001b. Ideophones and compound verbs in Wolaitta. In Christa Killian-
Hatz and F. K. E. Voeltz (eds.), Ideophones, pp. 49–62. Amsterdam and
Philadelphia: John Benjamin.
2006. Gender and animacy in Maale in comparative perspective. In S. Uhlig
(ed.), Proceedings of the 15th International Conference on Ethiopian Studies,
pp. 704–14. Wiesbaden: Harrassowitz.
The Omotic Language Family 849
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
2007a. Are -a- and -o- in the indicative verb paradigms of Zargulla nomi-
nalizers? In Ahma Azeb, Sava
´Graziano and Maarten Mous (eds.),
Omotic and Cushitic languages studies: Papers from the Fourth Cushitic-
Omotic Conference, Leiden, 10–12 April 2003, pp. 1–22. Cologne: Ru¨ diger
Ko¨ ppe Verlag.
2007b. Questioning forms in Zargulla. In Rainer Voigt (ed.), ‘From Beyond
the Mediterranean’: Akten des 7. internationalen Semitohamitistenkongresses,
Berlin 13–15, September 2004, pp. 197–210. Aachen: Shaker Verlag.
2007c. Non-verbal predication in Wolaitta. In Joachim Crass and
Ronny Meyer (eds.), Deictics, copula and focus in the Ethiopian Convergence
Area, pp. 99–117. Cologne: Ru¨ diger Ko¨ ppe Verlag.
2009a. Participant marking and valence in Wolaitta.
In Gerrit Dimmendaal (ed.), Coding participant marking: Construction
types in twelve African languages, pp. 355–84. Amsterdam and
Philadelphia: John Benjamins.
2009. The morpho-syntax of negation in Zargulla. In Leo Wetzels (ed.),
The linguistics of endangered languages: Contributions to morphology and
morpho-syntax, pp. 199–220. Utrecht: LOT.
2010. Compound verbs and ideophones in Wolaitta revisited.
In Mengistu Amberber, Brett Baker and Mark Harvey (eds.), Complex
predicates: Cross-linguistic perspectives on event structure, pp. 259–90.
Cambridge University Press.
2012. Omotic. In Zygmunt Frajzyngier and Erin Shay (eds.), The Afroasiatic
languages, pp. 423–504. Cambridge University Press.
2013. Demonstratives and directionals in Wolaitta. Paper presented at
the LCRC weekly meeting. The Cairns Institute, JCU Cairns, Australia.
Forthcoming. Male drums and female drums: Gender and inanimate
nouns in Oyda. Gender, culture and cognition. Special issue of the
International Journal of Language and Culture (ed. Angeliki Alvanoudi).
Azeb Amha and Gerrit J. Dimmendaal. 2006a. Verbal compounding in
Wolaitta. In Alexandra Y. Aikhenvald and R. M. W. Dixon (eds.), Serial
verb constructions: A cross-linguistic typology, pp. 319–37. Oxford and
New York: Oxford University Press.
2006b. Converbs in an African perspective. In F. K. Ameka, Alan Dench
and Nicholas Evans (eds.), Catching languages: Issues in grammar writing,
pp. 393–40. Berlin: Mouton de Gruyter.
Beachy, Marvin Dean. 2005. An overview of central Dizin phonology and
morphology. MA thesis, University of Texas at Arlington.
Bender, M. Lionel. 1991. Comparative Aroid (South Omotic) syntax and
morphosyntax. Afrika und U
¨bersee 74: 87–110.
Bender, M. L. 2000. Comparative morphology of the Omotic languages. (Lincom
Studies in African Linguistics 19). Munich: Lincom Europa.
2003. Omotic lexicon and phonology. Southern Illinois University: SIU
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
Biniyam S. Mendisu. 2008. Koorete verb Morphology. PhD thesis, Oslo
Breeze, Mary J. 1986. Personal pronouns in Gimira. In Ursula Wiesemann
(ed.), Pronominal systems, pp. 47–69. (Continuum 5). Tu¨ bingen: Gunter
1990. A sketch of the phonology and grammar of Gimira (Benchnon).
In R. J. Hayward (ed.), Omotic studies, pp. 1–67. London: School of
Oriental and African Studies.
Comrie, Bernard. 1989. Language universals and linguistic typology. University
of Chicago Press.
Debela Goshu and Girma A. Demeke. 2005. A grammatical sketch and
some socio-linguistic aspects of a dying language, Anfillo. Ethiopian
Languages Research Centre Working Papers 1(1): 53–86.
Dimmendaal, Gerrit. 2007. Africa’s verb-final languages. In Bernd Heine
and Derek Nurse (eds.), A linguistic geography of Africa, pp. 271–307.
Cambridge University Press.
Dixon, R. M. W. 1982. Where have all the adjectives gone? And other essays in
semantics and syntax. Berlin and New York: Mouton.
2003. Demonstratives: A cross-linguistic typology. Studies in Language 27
(1): 61–112
2004. Adjective classes in typological perspective. In R. M. W. Dixon and
Alexandra Y. Aikhenvald (eds.), Adjective classes, pp. 1–49. Oxford
University Press.
Fleming, Harold. 1976. Omotic overview. In M. L. Bender (ed.), The non-
Semitic languages of Ethiopia, pp. 298–323. (Monograph No. 5, Occasional
Papers Series, Committee on Ethiopian Studies 1). East Lansing:
African Studies Center, Michigan State University.
Gebre B. Guadie. 2013. Shinasha noun morphology. PhD thesis, Oslo
Hayward, R. J. 1982. Notes of the Koyra language. Afrika und U
¨bersee 65(2):
1987. Terminal vowels in Ometo nominals. In H. Jungraithmayr and
W. W. Mu¨ ller (eds.), Proceedings of the fourth International Hamito-Semitic
Congress, Marburg, September 20–22, 1983, pp. 215–31. Amsterdam and
Philadelphia: John Benjamins,
1988. Remarks on Omotic sibilants. In M. Bechhaus-Gerst and F. Serzisko
(eds.), Cushitic and Omotic: Papers from the International Symposium on
Cushitic and Omotic Languages, Cologne, January 6–9, 1986, pp. 263–99.
Hamburg: Helmut Buske Verlag.
1990a. Notes on the Zayse Language. In R. J. Hayward (ed.), Omotic lan-
guage studies, pp. 210–355. London: School of Oriental and African
1990b. Notes on the Aari language. In R. J. Hayward (ed.), Omotic
language studies, pp. 425–93. London: School of Oriental and
African Studies.
The Omotic Language Family 851
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
1995. The Challenge of Omotic (an inaugural lecture). School of Oriental
and African Studies, University of London.
2003. Omotic. The ‘empty quarter’ of Afroasiatic Linguistics.
In Jacqueline Lecarme (ed.), Research in Afroasiatic grammar II: Selected
papers from the fifth Conference on Afroasiatic Languages, Paris, 2000, pp.
241–62. Amsterdam: John Benjamins.
Hellenthal, Anne-Christie. 2010. Sheko grammar. Utrecht: LOT.
Hirut, Woldemariam. 2001. Demonstratives in Dawuro. Afrikanistische
Arbeitspapiere. (Special issue on North-East Africa linguistic and literary
studies) 65: 157–67.
2003. The grammar of Haro with comparative notes on the Ometo linguistic
group. PhD thesis, Addis Ababa University.
2005. Historical notes on numerals in Ometo: The obsolete quinary
system. Journal of Ethiopian Studies 38(1/2): 131–50.
Hirut Woldemariam. 2007. Deictics in Gamo. In Joachim Crass and
Ronny Meyer (eds.), Deictics, copula and focus in the Ethiopian Convergence
Area, pp. 129–38. Cologne: Ru¨ diger Ko¨ ppe Verlag.
´va. 1990. Grammatical relations in Gamo: A pilot sketch.
In R. J. Hayward (ed.), Omotic language studies, pp. 356–405. London:
School of Oriental and African Studies.
Hudson, Grover. 2005. 75 Ethiopian languages: 19 Cushitic, 20
Nilosaharan, 23 Omotic, 12 Semitic and 1 unclassified. Unpublished
paper. Available online at:
Ko¨ nig, Christa. 2006.Marked nominative in Africa. Studies in Language 30(4):
Lydall, Jean. 1976. Hamar. In M. L. Bender (ed.), The non-Semitic languages of
Ethiopia, pp. 393–438. (Monograph No. 5, Occasional Papers Series,
Committee on Ethiopian Studies 1). East Lansing: African Studies
Center, Michigan State University.
Mahder Tesfu. 2001. The morphology of Malo. MA thesis, Addis Ababa
Moges Yigezu. 2007. The vowel system of Kara from a historical-
comparative perspective. In Rainer Voigt (ed.), ‘From beyond the medi-
terranean’. Akten des 7. Internationalen Semitohamitistenkongresses, pp.
245–51. Aachen: Shaker.
Mulugeta Seyoum. 2008. A grammar of Dime. Utrecht: LOT Publications.
Rapold, Christian J. 2006. Towards a Grammar of Benchnon. PhD thesis,
Leiden University.
Rottland, Franz. 1990. A sketch of Shinasha morphology. In R. J. Hayward
(ed.), Omotic Language Studies, pp. 185–209. London: School of Oriental
and African Studies.
Taddese Ado. 2001. Tone in Kafinoono. MA thesis, Trondheim University.
Theil, Rolf. 2007. Kafa phonology. Journal of African Languages and Linguistics
28(2): 193–216.
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
2013: Koorete tonology. In M.-C. Simeone-Senelle and Martine Vanhove
(eds.), Proceedings of the Fifth International Conference on Cushitic and Omotic
Languages, Paris, April 16–18, 2008, pp. 167–74. Cologne: Ru¨ diger Ko¨ ppe
Unseth, Peter. 1990. Linguistic bibliography of the non-Semitic languages of
Ethiopia. East Lansing: African Studies Center, Michigan State
Wedekind, K. 1983. A six-tone language in Ethiopia: Tonal analysis of
Benchnon (Gimira). Journal of Ethiopian Studies 16: 129–56.
1985. Why Bench (Ethiopia) has five level tones today. In Ursula Pieper
and Gerhard Stickel (eds.), Studia linguistica diachronica et synchronica, pp.
881–901. Berlin and New York: Mouton de Gruyter.
1990. Gimo-Jan or Ben-Yem-Om: Beng-Yemsa Phonemes, Tones and
Words, In R. J. Hayward (ed.), Omotic language studies, pp. 68–184.
London: School of Oriental and African Studies.
Zaborski, Andrzej. 1983. Basic numeral in Omotic languages. In S. Segert
and A. J. E. Bodrogligeti (eds.), Ethiopian studies, pp. 123–42. Wiesbaden:
Zaugg-Coretti, Silvia. 2013. Yemsa verb morphology. PhD thesis, University
of Zurich.
The Omotic Language Family 853
at https:/
Downloaded from https:/ Universiteit Leiden / LUMC, on 03 Jul 2017 at 12:47:47, subject to the Cambridge Core terms of use, available
... These are Semitic, Cushitic, Omotic and Nilo-Saharan. Omotic is one of the six language families within the Afro-asiatic phylum predominantly spoken between the lakes of southern rift valley and Omo River [7]. ...
... On the other hand, Ometo language which have around four million speakers are facing a number of problem due to unavailability of resource [7].This hinders the communication between the people who speak different language specifically English and Ometo languages. To facilitate the communication between the speaker of Wolaita, Gamo, Gofa and Dawro languages and also to use the documents and information produced in technologically favored languages like English, the documents need to be translated. ...
A language is a tool fashioned by a man; it is the only gift that identifies human beings from the rest of life. Language is a means of communication in our day-to-day activities to do various things, like giving commands, asking questions, and expressing feelings, but we use it especially to communicate information about the world. There are a lot of languages spoken by human beings. It is difficult to learn and speak all languages spoken in this world, for this reason not all peoples are communicated with each other. Usually, this communication gap is solved by using a human interpreter. However, the use of human interpreters is expensive and inconvenient. Many types of research are being done to resolve this problem using machine translation techniques. Machine translation is an automatic translation of a source language to a target language. This can be spoken to speech or text to text translation. In this work, a bi-directional text-based multilingual machine translation for English and Ometo languages pair using recurrent neural networks is proposed. We started our study with the objective of designing and developing multilingual machine translation for English to Ometo by making the translation bi-directional by applying a recurrent neural network on translations between these language pairs. In order to achieve our objective, we collected parallel corpus data from different sources and divided it into training, validation, and testing sets. We trained our model using four experiments; the first experiment was done by combining four language datasets to Ometo, the second experiment was done by combining Wolaita, Gamo, and Dawuro datasets for training and Gofa dataset for testing and tuning. The third experiment was done by combining Wolaita, Gamo, and Gofa datasets for training and Dawuro dataset for testing and tuning. The fourth experiment was done by combining Wolaita, Gofa, and Dawuro datasets for training and Gamo dataset for testing and tuning. After training and testing these systems on corresponding training and testing datasets, the first experiment shows a BLEU score of 5.7 and 11.7 for English – Ometo, and Ometo-English respectively. For experiments II, III, and IV the experiment shows a BLEU score of 4.5, 2.4, 3.4 for experiments II, III, and IV respectively.
... Ethiopian languages which are under resourced and technologically disadvantaged have limited NLP application due to unavailability of NLP resources [7]. Among these Ethiopian languages, Wolaita, Gamo, Gofa and Dawuro languages which belongs to Omotic language family highly suffer from the lack of language resource to take the advantage of the technological supported language [8]. In addition to this, manual translation is expensive, time consuming, needs professionals, and it is complex to provide the translated material in short period of time [9]. ...
Conference Paper
Full-text available
In this paper, we described an effort towards the development of parallel corpora for English and Ethiopian Languages, such as Wolaita, Gamo, Gofa, and Dawuro machine translation. The corpus is collected from the religious domain and to check the usability of the collected parallel corpora a bi-directional Neural Machine Translation experiments were conducted. The neural machine translation shows good results as a baseline experiment of BLEU score of 13.8 in Wolaita-English and 8.2 English-Wolaita machine translation. The Wolaita-English translation shows a better performance than the other pairs of local language and the result of neural machine translation performs well when the amount of dataset increases, thus the amount of dataset has a great impact on the performance. Besides these, the morphological richness of the Ethiopian language contributed to the low performance of neural machine translation when the Ethiopian language is used as the target language. Further, we are working on minimizing the effect of morphological richness through different morphological processing techniques in the translation of Ethiopian languages.
Full-text available
The verb-final type in a crosslinguistic perspective The position of the verb relative to other constituents within a clause has been claimed by a number of authors to be a predictor of an additional set of syntactic features. Thus, according to Greenberg (1966), verb-initial languages tend to be prepositional rather than postpositional, putting inflected auxiliary verbs before rather than after the main verb; with more than chance frequency, verb-final languages tend to use postpositions, with auxiliary verbs following the main verb. Such inductively based generalizations about the nature of language of course require further analyses and explanations, e.g. in terms of preferred parsing or processing structures for the human mind. In more recent correlative studies of this type, e.g. by Dryer (1992), a distinction is drawn between phrasal and non-phrasal elements. Whereas phrasal elements, such as subject and object phrases or adpositions appear to follow a more consistent right-branching or left-branching pattern crosslinguistically, the position of non-phrasal categories such as adjectives, demonstratives, negative particles, or tense–aspect markers does not seem to correlate with constituent order type, as argued by Dryer. Obviously, constituent order is but one of various factors determining the typological portrait of a language, morphological techniques used in expressing syntactic and semantic relations being another important parameter. This latter observation of course is not new; Sapir (1921:120–46) already pointed out that languages may differ considerably in the techniques used for the expression of syntactic relations.
Full-text available
Full-text available
Two analyses are presented, Kafa I (with foreign sounds) and Kafa II (without foreign sounds). Kafa I is interpreted as the contemporary phonology, while Kafa II is the historical stage preceding the time when Amharic loanwords started to enter the language in larger quantities. Not all phonemes of Kafa I are equally integrated in the phonological system; phonemes with a low frequency are least integrated, and tend to be replaced by phonologically similar phonemes with a higher frequency.| Both analyses differ from earlier ones, primarily because some distinctive consonant oppositions postulated earlier turn out to be non-distinctive when quantity is taken care of in a more exact way. Kafa II consonant clusters are shown to constitute strictly defined sets, and most foreign clusters in Kafa I do not differ radically from the native ones.
Full-text available
Thesis (M.A.)--University of Texas at Arlington, 2005. Includes bibliographical references (leaves 173-178). Photocopy.
Africa is a continent where grammaticalized case systems are a rare phenomenon. But there is one exception: East Africa is a region where there is a relatively high occurrence of case languages (that is, languages with a grammaticalized case system). With regard to the type of case systems which occur in Africa, again, the picture is crosslinguistically unbalanced as there are hardly any ergative languages. In other words, of the two most common case types worldwide, accusative and ergative(/absolutive), essentially only one is represented in Africa, namely the accusative type. From a worldwide perspective, Africa seems to be a continent where case has nothing special to offer. However, in East Africa there are so called marked-nominative languages which seem to be quite unique worldwide. They are somehow a mixture: On the one hand they share features with prototypical accusative languages, on the other hand they share features with prototypical ergative languages.  In the present paper I will, first, define the typical features of a marked-nominative language. Second, I will give an overview of the languages which have a marked-nominative system. Third, I will deal with the question of whether the distribution of marked-nominative languages is genetically or areally motivated. And fourth, I will speculate on how such unusual systems could have developed.
This book presents the first comprehensive study of Dime, an endangered Omotic language spoken by about 5400 speakers in south-west Ethiopia. The study presents analysis of the phonology, morphology and syntax of the language as well as a sample of ten texts and an extensive word list. The author identifies a number of interesting comparative and typological phenomena. These include a series of uvular and velar fricatives which have not been reported in related languages. Dime has a two-way grammatical gender distinction and a special plural-agreement, both manifested on modifying categories. Rather than inflecting the same base pronoun-forms for various cases, as is common in other Omotic languages, Dime uses distinct subject pronoun sets that are formally different from object, dative and other pronoun types. Phrasal word-order is flexible; there is also a degree of flexibility in marking grammatical morphemes such as number, definiteness and case which may be marked either on the head noun or on the modifier or on both. Sentence-type distinction between interrogative and declarative clauses is partly expressed through morpheme reduction on the verb. That is, in the declarative, person-agreement morphemes are obligatory whereas these must be dropped in the interrogative. These and a number of other issues discussed in the study make the work interesting for specialists on Omotic and Afroasiatic studies as well as to general linguists interested in language typology.