Rabin, Chaim, “The Genesis of the Semitic Tense System”, in John Bynon (ed), Current progress in Afro-Asiatic linguistics, Amsterdam studies in the theory and history of linguistic science, Series IV, Current issues in linguistic theory v. 28 (Amsterdam, John Benjamins Publishing Company, 1984), 391-397.
Rabin breathes new life into the discussion of the development of tense/mood/aspect (TMA) systems of the older Semitic languages by dropping two long-held assumptions: 1) That proto-Semitic was an absolute beginning and must represent a simple original stage of verbal inflection, and 2) That Classical Arabic was more or less identical in its structure and forms with proto-Semitic. Instead, Rabin suggests that Classical Arabic represents the final stages of a process that began in a situation exemplified by Akkadian.
At first both “perfect” (Akkadian preterite) and “imperfect” (Akkadian present/future) were expressed by prefixed forms while the suffixed form served a separate function outside of the TMA system (Akkadian stative). All later languages built the suffix form into the TMA system, but in different ways, and they also differed in how/if they maintained the old prefix “perfect” form. Wherever two prefix forms are maintained, the “imperfect” is distinguished by being longer either by internal additions (Akkadian, Ethiopic) or by adding a final vowel (West-Semitic).
This may also help explain the vexing use of the suffixed form as an apparent “imperfect” in Biblical Hebrew weqaṭal constructions. If the qaṭal “perfect” form indeed developed from a form similar to the Akkadian stative which did not originally mark time or aspect, then it makes more since that an “imperfect” weqaṭal form could have emerged contemporary (or even prior) to this rather than see weqaṭal as a later development from the qaṭal form as some type of conversion in analogy to wayyiqṭol.Rabin, Chaim, Semitic Verbal System