4 1 Introductionrectified in future editions are the absence of scientific identifications of flora-
fauna species, and the omission of accents on stems other than verbs (verbs
generally have grammatically predictable accents).
For Burkina Faso we now have David Sudlow's handy work, The
Tamcisheq of north-east Burkina Faso (2001), which includes some lexical
lists organized by domains as well as grammatical information. The migrations
into Burkina began in earnest around 150 years ago and have been continuous,
as desertification has pushed Tuaregs south. The dialects spoken in Burkina
reflect the diverse origins of the migrants, but Tawellemett is well represented.
Prasse's three-volume work, Manuel de grammaire touaregue (1972-74),
is an elaborate systematization of data from Foucauld from the Hoggar. In this
work, Prasse analyses Tuareg phonology and morphology (the fourth volume
on syntax did not appear), in historical linguistic terms. The Manuel is an
important work for Berberists, but the author's perspective is somewhat
idiosyncratic, since he connects attested modern Tuareg forms directly with
Pre-Proto-Berber reconstructions. Much of his interest was in presenting a Pre-
Proto-Berber reconstruction with three short vowels in addition to the less
controversial long (full) vowels. Synchronically, Tuareg has two short vowels
ae and a, which are themselves partially merged in some dialects depending on
following consonants, and the non-Tuareg Berber languages generally have
just one short vowel phoneme. Unfortunately, the Manuel was written before
the author's Niger-based Lexique, and was therefore based largely on
Foucauld, whose transcriptions of short vowels for Tamahak were highly
unreliable. In addition, the Manuel (like the Lexique) was written in the
absence of good synchronic analysis of accent, which turns out to interact in
significant ways with the ablaut system. In general, the Manuel is of greatest
interest for its analysis of vocalic ablaut, and of a series of other, specific
historical linguistic issues.
Another important recent work on Proto-Berber is Maarten Kossmann's
Essai sur la phonologie du proto-berbere (1999). This is essentially a
collection of topical studies, addressing a number of cruxes of Berber
historical phonology.
1.3 Historical background
Historians and archeologists are in broad agreement that there has been a long-
term southward shift of Saharan populations over at least 1500 years. Around
the year 500 AD, significant portions of the Sahara were sufficiently moist and
temperate to accommodate permanent human settlement, while the Middle
Niger River area was swampy and too infested with disease-bearing flies to
permit year-round habitation. Over the centuries, climatic changes have made
some of the Sahara almost completely uninhabitable (and impassible),
including a significant expanse west of Araouan (north of Timbuktu). Other
areas could be used for grazing part of the year (assuming normal wet-season