A foot-based reanalysis of edge-in tonal phenomena in Bambara
Zoll (2003) demonstrates that "directionality" in tone association need not be stipulated, but
emerges from the interaction of constraints on tone sequencing, namely CLASH (no high tone sequence on
adjacent TBUs) and LAPSE (no high tone sequence on non-adjacent TBUs). However, Bambara "edge-in"
forms (1), where tone appears to be distributed from the edges of the word inward, are problematic for her
theory (Zoll 2003:264), as they violate both CLASH and LAPSE equally. In this paper we show that
Bambara tone does not require edge-in association if tones are associated within optimally bisyllabic
"tonal feet" (2) (Leben 1997, 2002, 2003, Zec 1999, deLacy 2002) and regulated by constraints on foot
construction (FT-BIN, RH-TYPE=TROCHAIC). Moreover, we demonstrate that headed tonal feet offer a
better characterization of three puzzling properties of Bambara tonal melodies: i) alternate tonal patterns
of trisyllabic nouns, ii) association of the LHL tonal pattern, and iii) distribution of lexical tone caused by
the `liaison high tone' appearing in definite phrasal contexts. An edge-in analysis of the data (Rialland &
Badjimé 1989) is forced to propose additional, conflicting, directional association rules.
The inventory of tonal melodies and noun shapes in Bambara is given in (3). Tones associate within
binary feet in bisyllabic and quadrisyllabic nouns. But, full binary footing is not possible for mono- or
trisyllabic forms. Exhaustive parsing of syllables into feet is assumed, and a degenerate foot is located at
the left edge of trisyllabic nouns: ( )( ). For LH and HL melodies, the second tone spreads within the
binary foot rather than crossing foot boundaries, the result of a constraint ALIGN-T-HD, which requires
association of lexical tones to foot heads: (bàn)(fúlá) and (mán)(gòrò). Under an edge-in analysis, the
trisyllabic patterns of LHH and HLL require an extra rule of spreading from the right edge. Trisyllabic
nouns are the only ones to have an alternate tonal melody (4). We attribute this to different positioning of
the degenerate foot: (bànfù)(lá); edge-in adopts a different spreading rule from the left edge.
The LHL melody maps to a quadrisyllabic noun as LLHL (kòròkárà). We argue that this is due to
constraints on tonal heads. Heads of feet prefer H tones (*HD-L in deLacy 2002). Therefore, (LL)(HL) is
preferable to (LH)(LL). The edge-in analysis deals with the LLHL pattern by associating the H at the
right edge, and spreading low tone from the left edge, the opposite of the standard pattern assumed for
trisyllables. Furthermore, the ability to spread from either edge predicts an alternate LHLL form, with low
tone spreading from the right. This is not attested. Under the tonal foot analysis, no alternate
quadrisyllabic forms are predicted, as alternate tonal patterns follow from non-binary footing.
In definite phrasal contexts, a `liaison' H tone associates to the final syllable of the noun. Rialland
& Badjimé (1989) show that this affects nouns that end in L tone (5), and in particular, with LHL
melodies, the H tone pushes lexical tones leftward instead of creating a contour tone word-finally. They
offer no explanation as to why the contour tone appears in initial position. We show that it follows from
the two previously mentioned constraints, ALIGN-T-HD and *HD-L (6). Furthermore, even though
alternate footings are considered, only one is optimal; there is no alternate tonal pattern for this word.
Finally, we compare our analysis to another tonal foot analysis of trisyllabic nouns (Leben 2002,
2003). Leben's analysis differs in three fundamental ways: i) the liaison H tone is treated as part of the
lexical tone melody, ii) there are two tonal foot melodies (H and LH), and iii) no foot heads. We concur
with the concept of a tonal foot, but argue that the analysis has several drawbacks. It incorporates a tone
lowering rule to change final high tones to low in indefinite phrasal contexts and to adjust internal
sequences of high tones. It predicts that LHL trisyllables should alternate in indefinite/definite contexts as
sàkénè / sàkén , but in fact s kèné is the definite form. Moreover, the analysis cannot extend to
quadrisyllabic forms: forms such as LLHL kòròkárà are predicted to surface as *LHLL in indefinite
contexts, as the final H of (LH)(LH) is changed to L. Low toned quadrisyllabic forms such as bùgùnìnkà
cannot be derived, as the analysis can only generate a maximum of two adjacent low toned syllables.
In conclusion, constraints on tonal feet, incorporating the notion of a tonal head, offer a superior
account of Bambara tonal patterns than an edge-in directional analysis. Zoll (2003)'s theory of Optimal
Tone Mapping is no longer undermined by the case of Bambara. It remains to be seen whether tonal feet
can be extended to deal with other instances of tonal melodies which also appear `binary' in nature.
kulukutu
(1) `a ball' (2) (ku.lu)(ku.tu)
H L H L
(3) H bá `river' L bà `goat'
bálá `balafon' bàlà `porcupine'
súngúrún `young girl' gàlàmà `ladle'
jánkárúbú `rogue' bùgùnìnkà `whip'
HL mángòrò `mango' LH bànfúlá `hat'
kúlúkùtù `ball' gàrìjégé `chance'
LHL sàkénè `lizard'
kòròkárà `tortoise'
(4) HL mángòrò `mango' LH bànfúlá `hat'
mángórò bànfùlá
definite phrasal
(5) L bà b `goat'
bàlà bàlá `porcupine'
gàlàmà gàlàmá `ladle'
bùgùnìnkà bùgùnìnká `whip'
HL mángòrò mángòró `mango'
kúlúkùtù kúlúkùtú `ball'
LHL sàkénè s k èn é `lizard'
kòròkárà kòrókàrá `tortoise'
sakene `the lizard' ALIGN(T, HD) *HD-L
(6)
LHLH
**! *
a. (sa)(ke.ne)
* **!
b. (sa.ke)(ne)
**! *
c. (sa.ke)(ne)
* **!
d. (sa)(ke.ne)
* **!
e. (sa)(ke.ne)
* *
f. (sa.ke)(ne)
References
deLacy, Paul. (2002). The interaction of tone and stress in Optimality Theory. Phonology 19:1-32.
Leben, William R. (1997). Tonal feet and the adaptation of English borrowings into Hausa. Studies in
African Linguistics 25:139-154.
Leben, William R. (2002). Tonal feet. In Ulrike Gut & Dafydd Gibbon, eds. Proceedings, Typology of
African Prosodic Systems. Bielefeld Occasional Papers in Typology 1: 27-40.
Leben, William. (2003). Tonal Feet as Tonal Domains. In J. Mugane (ed.) Trends in African Linguistics
5: Linguistic Typology and Representation of African Languages. Africa World Press, 129-138.
Rialland, Annie and Mamadou Badjimé. (1987). Reanalyse des tons du Bambara: Des tons du nom a
l'organisation generale du système. Studies in African Linguistics 20: 1-28.
Zec, Draga. (1999). Footed tones and tonal feet: rhythmic constituency in a pitch-accent language.
Phonology 16: 225-264.
Zoll, Cheryl. (2003). Optimal Tone Mapping. Linguistic Inquiry 34: 225-268