12.2 Focalization 643(744) waedder assaral [ά i-t-ajj],
not work [Foe 3MaSgS-LoImpf-do.LoImpfP]
edss [a i-t-ajj]
sleep.VblN [Foe 3MaSgS-LoImpf-do.LoImpfP]
'Work [focus] is not what he does, (rather) sleep [focus] is what he
does.' (lit. "It's not [the case that] work [focus] is what he does,...")
Since Focus ά precedes any preverbal particles, if the focalized clause
contains a Future morpheme, the latter will be the non-initial Future
morpheme (mar in T-ka, e in several other dialects) rather than ad.
(745) a-sakka [a-\d mar 0-aemma2t] [T-ka]
a-sakka [a-\d e 0-aemmaet] [R]
tomorrow [FocAComit Fut 3MaSgS-die.ShImpf]
'It's tomorrow [focus] that he will die.'
There is normally only one focalized constituent. However, as in other
local languages, it is possible to embed a focalized clause in a 'why?'
interrogative. This is arguably a layered double-focus construction, but the
only Focus morpheme is associated with the internal focus (746).
(746) ma-\fael [i-s-αη [a i-rhd]]
what?\on [Pl-meat-MaPl [Foe 3MaSgS-want.Reslt]]
'Why (is it) meat [focus] that he likes?' [Gao]
The ά Focus morpheme is regularly omitted in WH-interrogatives (§12.3).
In my T-ka data, with this exception the morpheme is normally audible in
focalized constructions. In my Kidal-area data, omission is fairly common in
texts even in non-interrogative focalizations, though informants are always
willing to repeat the examples more carefully with the ά audibly present, so I
am inclined to interpret omission as due to a low-level contraction. My K-d
informant, in elicitation, allowed omission in subject but not other
focalizations. However, in my Κ texts (from other speakers), I did observe
omission of ά in object focalization as well as subject focalization.
12.2.1 Subject focalization
In the case of subject focalization, the verb is expressed in indefinite participial
form, but takes invariant 3MaSg subject prefix (i- or 0- depending on the
initial segment of the verb, and zero for those perfective verbs that cannot take
subject prefixes). The participle likewise has MaSg suffix -aen, regardless of
the actual pronominal category of the subject (747). Examples (747.h-i) do not
differ in meaning (to my knowledge) and illustrate the optionality of ά in K-d
dialect.