The African « lax » question prosody
Annie Rialland
Laboratoire de phonétique et phonologie,UMR 7018, CNRS/Sorbonne-Nouvelle, Paris
Second TIE Conference: Typology of Tone and Intonation.
Berlin, 7-9 September 2006
• We will show that a « lax » question prosody is an areal feature of the African Sudanic belt.
As first, we will give some background and explain what we call a « lax » question prosody.
Question Prosody in Africa :Background
• Our talk at the TIE1 conference held in Santorini
Question prosody: an African perspective
• which was based on a database of 74 languages belonging to the 4 African language phyla (Niger-Congo, Afro-Asiatic, Nilo-Saharan, Khoisan)
We found a diversity of prosodic yes-no question markers
HIGH-PITCHED MARKERS
• cancellation/reduction of downdrift, register expansion
• raising of last H(s) (not necessarily sentence-final)
• cancellation/reduction of final lowering
• final H tone or rising intonation (final H%)
• final HL melody
• final L tone or falling intonation (final L%)
• final polar tone or M tone
• length (VV or V…)
• breathy termination
• cancellation of penultimate lengthening
• [open] vowel
We introduced the « lax » prosody
It is defined by a set of characteristics :
• a falling intonation
• a lengthening
• a breathy termination
• an [open] vowel
In a given language, this « lax » prosody may be represented by all of these characteristics or by a subset of them.
• In this talk, we will investigate the distribution and the realization of this « lax » prosody in various language families (with sound examples):– in the Niger-Congo phylum
• Gur , Kru, Kwa, Mande, Adamawa-Ubangi, Benue-Congo families
– in the Nilo-Saharan phylum• Central Sudanic, Eastern Sudanic families
– in the Afro-Asiatic phylum• Chadic family
• Currently, our database includes 51 languages with a form of « lax » prosody.
The question « lax » prosody in the Gur family
• Why the Gur family, first?– It is better represented in our database, with
many sound examples– Gur languages are the most «central »
languages in Africa, considered as a linguistic area (Heine and Lewey, in press).
(in the same way as Bulgarian is the most « central language » of the Balkan area, based on the number of «Balkanic» features that it includes)
Heine and Lewey, in press, « Is Africa a linguistic area? »
Gur languages
Number of typical African features or « Africanisms »
« the child »
« the child? »
In Ncam (Togo), the « lax » prosody occurs with its full set of characteristics:
• a falling intonation
• a lengthening
• a « breathy termination » (with progressive opening of the glottis)
• a vowel [a], except after a monosyllabic word ending with a vowel
Ncam (Togo) examples
•Falling intonation• [open] V •lengthening• « breathy termination »
« It is equal. »
« Is it equal? »
« a slave »
« a slave? »
Contraction of a word-final -u and -a
In Wule Dagara (Burkina), the « lax » prosody involves:
• a lengthening
• a « breathy termination »
« He saw a hyena »
« Did he see a hyena? »
The final L is streched over the lengthened ãã.
Zoom on the « breathy termination »
Progressive intensity decrease
Weakening of F5, F4, F3
due to the progressive opening of the glottis
« You said to Kut »
« Did you say to Kut? »
In Wule Dagara, a final flap can be lengthened
The final H is streched over the lengthened r, which becomes a long trill.
In Moba (Togo), the « lax » prosody involveslengthening and « breathy termination », only:
Statement
•Lengthening•Prolongation of F0•Intensity decrease
Question
« Stones. »
« Stones? »
Statement
Question
airflow decreases
airflow increases
opening of the glottis
From Rialland A. , 1984, "Le fini/l'infini ou l'affirmation/l'interrogation en moba (langue voltaïque parlée au Nord-Togo)" , Studies in African Linguistics, supp. 9
The melodic contour stretches out the tone realization
In Moba, question prosody :•no specific tone or melodic contour.•lengthening•breathy termination+ base form of words (without truncation or metathesis occurring elsewhere)
« Beans. »
« Beans? »
The « lax » prosody has informational and expressive variants in some languages:
• Tem examples (Tchagbalaye 1976):
bòòbó « They went » (statement)
bòòbóóò « Did they go? »
bòòbóò « They went, didn't they? »
(asking for confirmation)
bòòbóòóòóò « Did they go? (exclamatory) »
Patterns of variation of the « lax » prosody in Gur languages
lengthening lengthening L%
lengthening L% [open] vowel
Moba, Nateni, Wule Dagara
Dagaare, Gulmancema, Kasem, Kabiye, Konni, Kusaal, Lobiri, Moore, Moyobe, Nawdem, Tem
Akaselem Ncam
In blue, languages with breathy termination. No data on this point for the other languages.
All of the largest Gur languages have a lax question prosody.
2 out of the 17 Gur languages in our database do not have this prosody (Kulango, Farefare)
The « lax » prosody in Kwa languages
L L -a or e
Akan (Ashanti) Baoulé Porto-Novo Gun
Adjoukrou Avikam Ewe Fon Gun
Patterns of variation of the « lax » prosody in Kwa languages
No sound examples, no data on breathy termination.
The « lax » prosody occurs in all Kwa language groups and in all largest languages of the family.
The « lax » prosody in Kru languages
Patterns of variation of the « lax » prosody in Kru languages
lengthening lengthening
L%
lengthening
[open] vowel
lengthening
L%
[open] vowel
Wobé Bassa Vata Godié
Neyo
No sound examples, no data on breathy termination.
The « lax » prosody occurs in all Kru language groups.
1 out of the 6 Kru languages in our database do not have this prosody (Klao in Liberia)
The « lax » prosody in Mande languages
Southeastern Mande languages
The « lax » prosody in Southeastern Mande
languages – Southeastern Mande languages spoken in Côte
d’Ivoire share areal features with Kru and Kwa languages: a tendency to monosyllabicity and a large number of tones (3, 4 or even, 5 tones)
– The « lax » prosody is widespread among them• in Toura (lengthening + L% or -è)
• in Gouro (lengthening + L%)
• in Wan (lengthening with L% in some contexts).
Sporadic « lax » prosody in Western Mande languages
• Western Mande languages generally have high-pitched markers (Soninké, Bambara, Mende)
• The « lax » prosody occurs sporadically :– in Bambara, there is a -wà question marker,
beside a H% and other segmental morphemes.
Sound file: from An Ka Bamanankan Kalan: Intermediate Bambara, C. Bird and M. Kante
The « lax » prosody in Benue-Congo(non-Bantoid) languages
L%, L or "stepdown" tone
L%, L or "stepdown" tone + a
Degema Gwari Isoko
Bekwarra Engenni Fyem Yala
Patterns of variation of the « lax » prosody in Benue-Congo (non-Bantoid) languages
« Lax » prosody is widespread, occurring in many groups of this family: Edoid, Cross-river, Plateau, Nupoid, Idomoid.
Languages with « lax » prosody are interspersed with languages having high-pitched markers (Efik, Igbo, Yoruba).
The « lax » prosody in the bantoid subfamily of the Benue-Congo family excluding Bantu.
L or L% -a
Slight fall -wa
Downstepped H or M + a
Ejagham Mambila Limbum Tikar
Patterns of variation of the « lax » prosody in Benue-Congo Bantoid languages (non-Bantu)
In Mambila, the final falling contour differs from the realization of a L tone or any of the 4 tones (Connell 2004).
There is a variety of question markers in this family, some languages having the « lax » prosody, others having high-pitched markers (Bafut).
« lax » prosody in Adamawa-Ubangi languages
Length
L or L%
L or L% (y)+a
M (default tone)
-wa
Mbum Munzombo Zande
Banda-Linda -à or - Mono Mumuye
Ngbaka
Patterns of variation of the « lax prosody »
in Adamawa-Ubangi languages
The largest languages of this family (Zande, Banda, Gbaya) have a form of « lax » prosody.
2 out of the 7 Adamawa-Ubangi languages in our database do not have this prosody.
In Zande, the melodic marker differs from a lexical tone as it is not associated with a tone-bearing unit (Boyd 1980).
The « lax » prosody in the Niger-Congo Phylum
We have found it in all the Niger-Congo families except Atlantic, Bantu and Kordofanian.
(Map: web resources for African languages)
« lax » prosody area
areas without« lax » prosody
The « lax » prosody in the Nilo-Saharan phylum
(Map: web resources for African languages)
The lax question marker in Ngambay,
a Nilo-Saharan language (Central Sudanic, Sara group)
DOUMPA MIAN-ASMBAYE
Université de N’Djaména-Tchad
In Ngambay, the question prosody is breathy. We can
hear an « h » at the end of the question.
« -wàh » is the question marker
Declarative utterance : ngokon Dè
Interrogative utterance : ngokon lè Dè wàh
h
Zoom of the « breathy termination »
Half of the vowel -a is voiceless
Patterns of « lax » prosody in Central Sudanic languages
• -wà
in Kabba, Ngambay-Mundu, Sara-Ngambay
(data on breathy termination only in
Ngambay-Mundu)
• -à or -wà in Mbay
• slight fall in Bagiro
Question prosody in Songhay,Western Sudanic and Eastern
Sudanic languages
• Songhay : only high-pitched markers
• Western Sudanic languages– -wá (Kanuri), a « hybrid form »
• Eastern Sudanic languages– à in some languages (Zaghawa, Turkana)
otherwise, languages of this family have high-pitched markers (Anywa, Arusa, Dholuo, Nandi)
The « lax » prosody in the Afro-Asiatic phylum
(Map: web resources for African languages)
The « lax » prosody in the Afro-Asiatic phylum
• We have found it only in the Chadic family
• In the Chadic family, some languages have a form of « lax » prosody associated with downdrift reduction– àa (Angas, Sayanci)– à (Pero)
otherwise, Chadic languages have high-pitched question markers (Hausa, Tera).
The « lax prosody » : areal distribution
Sudanic belt
breathiness
The area in which the « lax » prosody has been found coincides roughly with the Sudanic belt.
from Clements & Rialland, « Africa as a phonological area », in press, in The Linguistic geography of Africa, Heine and Nurse (eds.), Cambridge University Press (with new additions).
Another prosodic characteristic commonly found in the Sudanic belt : 3 or more level tone systems
From Clements & Rialland, op. cit.
2 areas
Other phonological characteristics commonly found in the Sudanic belt
• labial velar stops• implosive consonants• nasal vowels• two series of high vowels (±ATR)• ATR vowel harmony
(Clements and Rialland, op. cit.)
• We suggest that the « lax » prosody originated in the Niger-Congo phylum, and was later borrowed by neighboring languages (Chadic and Nilo-Saharan).
• Its pattern of extension recalls the spreading of other Niger-Congo features such as labial velar stops in the Sudanic belt.
Alternatives to the « lax » prosody
• When, in a given language, the lax prosody is not present, generally high-pitched markers occur (downdrift reduction, raising of the last H tones, rising intonation or an HL melody).
We have found very few languages with segmental markers only: Ngiti (-tí), Samba Leko (-gú, -ì), Rugciriku (-ndí), Ngangela (-ndí), Shi (kà), out of 110 languages.
The « lax » prosody as a typical feature of the Sudanic belt
• High-pitched markers are extremely common outside of Africa
• High-pitched markers are also overwhelmingly common in Africa outside the Sudanic belt
• The « lax » prosody is a special feature of the Sudanic belt
• It is currently unknown whether such question prosodies occur elsewhere in the world
Many thanks
• To my colleagues, my students, my informants for providing me data included in this presentation or helping me to find them.
in particular, A. Some, N. Podi, D. Mian-Asmbaye, R. Boyd, G. Dimmendaal, T. Schadeberg and N. Clements