The full forms of biconsontal and triconsontal roots could be modified by suppressing the characteristic vowel from its normal position between the first and second consonant (SD/422), for example: ✶-z’ri < √ZIR or ✶-k’lab < √KALAB. In such cases (a) the final vowel would necessarily be the characteristic-vowel (SD/434) and (b) some sort of suffix was need to create a viable word (SD/422).

The most common suffix was another instance of the chararactistic vowel: ✶izri, ✶aklab. More rarely, some other formative prefix could be substituted, as ✶da-zri, ✶da-klab (SD/422). This second (rare) case was the only situation in which the first vowel of a primitive word could be something other than the characteristic vowel of the root.