A Grammar of Tamashek (Tuareg of Mali)

8.1 ] Instrumental and related nominals 559

(587) Other Instrumental Nominals with <H> Melody

form gloss related verb and gloss

a. Ma ά-s-CaC, Fe t-a-s-CaC-t (from -vCvC-)

ά-s-fal 'roof material' -vfVl- 'be roofed'

ά-s-jan 'animal area' -vjvn- 'kneel'

ά-s-har 'lid' -vhvr- 'shut, block'

a-s-kad 'landmark' [none]

t-a-s-har-t 'door' -vhvr- 'shut, block'

t-a-s-hat-t 'bellows' perhaps -vhvd- 'swear'

t-a-s-jar-t 'flint cloth' -vjvr- 'throw'

t-a-s-ni-t-t 'vehicle' -vnvy- 'mount'

[with /ay/ -»i]

b. Fe t-a-s-aCCi-t-t (from -vCCu-, or -vCCvy-)

t-a-s-anji-t-t 'channel' -vnjvy- '(water) run'

[with /ay/ —> i]

t-a-s-aswi-t-t 'drinking place' -vswu-

In (587.a), the core stem is reduced to -CaC. One can argue that ά-s-CaC
is syncopated from α-s-aCaC. In T-ka and several other dialects, α-s-aCaC is
in fact the surface pronunciation of the corresponding causative VblN, so there
is a clear opposition between causative VblN α-s-aCaC and instrumental
nominal ά-s-CaC. In K-d, on the other hand, both patterns have fallen together
as ά-s-CaC.
The two examples in (587.b) have identical shapes, but 'channel' involves
/ay/ monophthongizing to i, while 'drinking place' is based on a true V-final
stem.
In (588), I present instrumentals that have i or u in the penult. When the
final is a CvC syllable with short V, we get ae in (588.a), and a in (588.b-c). In
(588.d), one cannot distinguish between ae and a because of the BLC.

(588) Instrumental Nominals with {i u}

form gloss related verb and gloss

a. Ma a-s-iCCasC, Fe t-a-s-ICCaeC-t (from -vPQvC- or -uPQvC-)

a-s-ilkasw 'dipper, ladle' -vlkvw- 'draw water'

b. Fe t-a-s-ICaC-t (frozen)

t-a-s-Isak-k 'Bergia herb' [none]