jeff_l
(Jeff_L)
#1
352 7 Verbal morphology
d. nominalization
VblN [see discussion below]
To get the long imperfectives to come out correctly, it is useful to have a
pre-ablaut reconfiguration (cf. §3.4.1.5) of -vPQu- to -PvQu-. The regular
long imperfective melody and local formatives will convert this -PvQu- to e.g.
LoImpfP /-PdQQA-/, with the proviso that the stem-final V appears as
deletable /A/ rather than a. This /-PaQQA-/ is parallel to LoImpfP -PdQQaeC-
for -vPQvC- verbs (§7.3.1.1). As noted elsewhere, this pre-ablaut
reconfiguration has the effect of smuggling in a templatic element into long
imperfective ablaut.
In some eastern dialects, especially A-grm, the LoImpfP of -vPQu- verbs
has a melody similar to that of the short imperfectives. There is some
fluctuation in the A-grm forms, but -PdQQa- and -PaQQu- with
overt final V are attested, and some other eastern dialects especially around
Gao have /-PdQQi-/ with deletable high V, e.g. -PdQQa- rather than -PdQQae-
before C-initial suffix. Examples of the LoImpfP: for -vslu- 'hear' -sdll-
(A-grm I R T-ka T-md) alongside -salla- (Gao K-d) and -sdllu- (A-grm Gao);
for -vjlu- 'go' -jail- or -gdll- (Im Gao R T-ka T-md) alongside -gdlla-
(A-grm), -gdllu- (A-grm), and oddly Kidal-area -gillu- (K K-d K-f). The forms
-sail- and -j/gdll- are variably /-PdQQA-/ and /-PdQQi-/ when we add
suffixes. Thus LoImpfP -t-ajj- 'do' (see below) has 3MaPl t-djjae-n (e.g. T-ka)
but t-djjs-n (some Gao-area dialects).
Even in T-ka we get a stem-final V in the long imperfectives of
syllabically comparable causatives, e.g. LoImpfP -s-άηηα- 'cook' (cf. Imprt
s-seqq), causative of 'be cooked, ripe' (Imprt aeqq, LoImpfP -ηάηη-, cf. A-grm
LoImpfP -ndgrja-).
The big difference between 'vomit' and 'go' in (355) is the
resyllabification of 'go' in the short imperfective system. This
resyllabification, required when the second C is a sonorant, involves Final-CC
Schwa-Insertion (44) and (for T-ka but not other dialects) Epenthetic-Vowel
Accentuation (70), accompanied by Short-V Harmony (§3.2.4, §3.3.2, §3.2.6).
The derivation of ajal could be represented as /aejli/ —> /aejsl/ (Final-CC
Schwa-Insertion and simultaneous Epenthetic-Vowel Accentuation) —» ajal
(Short-V Harmony). In other dialects we get forms like aglu (K Ts) with the
final V retained, and forms like agal (R) where the loss of the final V forces
Final-CC Schwa-Insertion but the accent does not shift.
In (356) I give complete PerfP and Resit paradigms for another verb of
the α/ι type. Note the varying (morpho-)phonological treatments of the stem-
final V.
