A Grammar of Tamashek (Tuareg of Mali)

356 7 Verbal morphology

(359) Derivations for Suffixed PerfN Stems of -vnsu- 'excise'

2MaPl 2FePl comments

/t-ansa-asm/

/t-ansae-aem/

/t-ansae-m/

/[t-anse-m]/

/t-ansa-maet/

/t-ansas-maet/

/[t-anse-m]aet/

note: [...] shows rebracketing of ablaut domain

t-anse-m t-anse-maet surface forms

after basic perfective ablaut

Presuffixal α-Shortening

along with Default

Accentuation

VV-Contraction

PerfN ablaut (e-pclf)

The Shlmpf of this class has a stem that I represent as /-aeCCi-/ (see below
for /-asji-/ from -vju- 'do'). The phonology of /i/, cf. (353), above, is illustrated
by the Shlmpf paradigm of -vksu- 'eat' in (360). The paradigm seems valid
throughout the zone (including at least T-md, T-ka, K-f, R, and Ts).

(360) Shlmpf Paradigm of -vksu- 'eat'

subject category Shlmpf form

a. stem-final V appears as zero with no subject suffix, surface «L»

melody

1P1 n-aeks

3MaSg 0-seks

3FeSg t-aeks

b. /i + ae/ contract as as before suffixal V

lSg asks-aer

c. /i + ae/ contract as ae before suffixal V, triggering Short-V Harmony

2Sg t-aksa-d

2MaPl t-aksa-m

3MaPl aksa-n

d. V appears as a before suffixal C, triggering Short-V Harmony

2FePl t-aksa-maet

3FePl aksa-naet

I recorded e.g. 3MaPl asksa-n for some Kidal-area dialects.
In (360.b-c), the ν accent indicates that accent does not shift to a preverbal
Future particle ad, as in ad t-aks-ad 'you-Sg will eat'. This is because the
underlying stem-final hi and the suffix-initial V both count in Default
Accentuation, which precedes VV-Contraction. In (360.a), even with III the