2 I am following the tradition of most work on Bantu grammar (Doke 1943, 1954, Meeussen 1967, etc.) in describing /a/ as the least marked IFS. As this earlier work points out, /a/ is arguably the default IFS, as it occurs most frequently, being compatible with a wider range of tenses and aspects than the other IFS. See Mutaka (1994) for discussion of other phonological processes in Kinande which take the I-Stem as their domain. And see work like Myers (1987, 1998), Mchombo (1993) for arguments motivating the I-Stem in other Bantu languages. 3 As Hyman et al. (1999) argue, an alternative approach to denning the RED as a verb stem would be to define Kinande reduplication as a form of stem self-compounding at the I-Stem level. It is beyond the scope of this paper to compare these approaches in detail. The important point is they both agree in defining the RED as a verb stem whose segmentism and morphological parse are required to correspond to its Base stem. ' sort of tower' ' toy rat' ' many women' ' sort of chest' 'small stones' ' sweets' ' monthly' ' person'