All Primitive Adûnaic biconsonantal and triconsontal roots had “full-form” with two instances of the characteristic-vowel, one in its normal position between the first and second consonant and an extra vowel after the second consonant (SD/422), for example: ✶ziri < √ZIR or ✶kalab < √KALAB. This full-form can be considered the simplest form of words derived from the root.

The extra vowel could appear in places other than its full-form position, however, so that it was prefixed or suffixed instead. Furthemore, the characteristic vowel could be suppressed from its normal position between the first and second consonant, appearing elsewhere instead (SD/422).

It is rare for such variations to include both a prefix and suffix, and the characteristic vowel rarely appears more than twice (SD/422). In addition, since Primitive Adûnaic words are not allowed to end in consonant clusters, a prefixed forms like **akalb is not possible without a final vowel. Such a form requires a suffixed subordinate vowel different from the characteristic vowel, such as ✶akalbu or ✶akalbi (SD/422). This is the only basic primitive word pattern in which three vowels appear.

Because of subordinate-vowel-variation, even full-form words could show some variation by replacing the second characteristic vowel with a different vowel, for example: ✶zira, ✶ziru, ✶kalib, ✶kalub. Once the positioning of the vowels was determined, the word could undergo further root-modifications as part of its derivational process.