3.4 Ablaut 99sections on nominal and verbal morphology. Here we describe some basics of
Tamashek ablaut.
3.4.1 Stem shapes and templates
In this section I am concerned only with the representation of stems that serve
as the input to ablaut processes. Some nouns, and all invariant forms (particles,
minor adverbials, etc.), cannot serve as ablaut inputs and are not considered
here.
3.4.1.1 Nouns
The only ablaut that nouns are subject to is plural. Some nouns do not have an
ablaut plural and are disregarded here.
For nouns that are subject to ablaut, the fully-spelled out Sg stem, along
with the attached vocalic prefix (if any), is the input to the PI ablaut process.
Nothing is gained by decomposing the Sg stem into, say, a CV-skeleton and a
vocalic melody.
It should be noted, however, that a marked (as opposed to default) accent
must be part of the representation. The accent may be on the core noun stem
itself, or (if the core stem is monosyllabic) on the vocalic prefix. The accent is,
in most cases, carried over from Sg to (ablaut) PI. Thus as-mdtuj 'work gear',
PI i-mutaj, and ά-mnss 'camel', PI 1-mnas.
There is only case where phonological information not easily discernible
from the Sg stem is needed to get PI ablaut to come out properly. This is when
a feminine noun, with FeSg suffix -t, has word-penultimate accent in the Sg,
e.g. t-a-jarjis-t 'shoulder blade'. This is because -t is one of the suffixes that
disallows antepenultimate word accent. In the unsuffixed ablaut PI, there is (by
stipulation) no suffix. This effectively forces the hand of the stem, which must
reveal whether it has a lexical accent on the penult that is overridden by
Default Accentuation in the Sg. In the case of 'shoulder blade', it does turn out
that the stem has a lexical accent on the penult, so we get t-i-jsrjas instead of
#t-l-j3rjas. As a result, the basic lexical representation of the Sg stem plus
vocalic prefix is -a-jarjis-.
3.4.1.2 Verbs (specific stem-shapes)
The issue of basic lexical representation is much more difficult for verbs.
This is because one can argue that every actual stem (PerfP, Imprt, LoImpfP,
etc.) is formed by combining a basic stem shape with a vocalic melody, with or
without additional ablaut formatives. If this is the case, then the audible V's in
each stem acquire at least their high (H) or L (low) feature values from the