There are three Adûnaic verb forms appearing in the first draft of the Lament of Akallabêth that share two common characteristics: (1) their last vowel is -i and (2) they have an augment: a repeated stem vowel prefixed to the verb. The examples are:

• udūbanim “fell” (plural) ← ?dub-
• akallabi “fell in ruin” (singular) ← *kalab-
• akhaini “lay” (singular) ← ?khay-

These two traits resemble how the Quenya perfect tense is formed: with a vowel augment and with a final -ië suffix (e.g. Q. tuv- → utúvië). I believe all three of these examples belong to the similar verb tense, which I label the “draft-perfect”. I believe that the past and draft-perfect tenses serve the same role in the draft of the Lament that the aorist and past tenses do in the final version of the Lament.

As noted elsewhere (SD/439), in the Adûnaic grammar of the later versions of the Lament, the aorist tense can function as a past tense in historical narratives, so that the past tense functions as the pluperfect. This usage is largely consistent with the past and draft-perfect use in the first draft of the Lament. Many of the first draft’s past forms changed to aorist forms in the second draft. Similarly, many of the first draft’s “perfect” forms changed to past tense:

	First draft	Second draft	Verb Development
S1	Kadō zigūrun zabathān unekkū	Kadō zigūrun zabathān unakkha	past >> past
S2	ēruhīn udūbanim dalad ugrus	ēruhīnim dubbudam ugru-dalad	perfect >> past
S4	bārun-adūnō rakkhatū kamāt sōbēthumā eruvō	bārun an-adūn {urahhata >>} urahta dāira sāibēth-mā ēruvō	past >> past >> aorist
S5	azrē nai phurrusim akhās-ada	azrīya du-phursā akhāsada	past >> aorist
S6	anadūni akallabi	anadūnē zīrān {hikallaba >>} hikalba	perfect >> past >> aorist
S11	adūn batān akhaini ezendi	adūn batān ēluk izindi yadda	perfect >> past

This pattern of changes is not entirely consistent. The first sentence remained in the past tense. The fourth sentence also initially remained in the past in the second draft before changing to the aorist. The sixth sentence changed all the way from the perfect to the past to the aorist (and Tolkien continued to vacillate between the past and aorist tenses in later versions). There is enough of a pattern, though, to indicate that all three draft “perfect” forms served a similar grammatical function as the later use of the past tense.

Other Theories

I haven’t seen anyone present a theory for these verb forms quite like mine, but Andreas Moehn did note (LGtAG) that the initial a- values in akhaini and akallabi might mark the completion of a process. Carl Hostetter and Patrick Wynne suggested something similar but only for akallabi (VSH/37). All these authors generally analyse the vowel prefixes in these first draft verbs as singular pronouns: neuter in the case of a- and masculine in the case of u-.

It is clear that draft Adûnaic used pronominal prefixes: the initial u- in the verb unekkū “he-came” appearing in the first sentence is certainly the pronoun “he”. I think, however, that it is the only instance of a pronoun in the first draft of the Lament, as it is the only verb with an explicit pronominal gloss. The a- in akhaini could possibly be a neuter singular pronoun “it” matching its subject “road”, but the u- in udūbanim should be plural yu- or ya- “they” to match “Númenóreans” and the a- in akallabi should be feminine hi- “she” to match its subject anadūni “Númenor” (the names of Númenor were elsewhere consistently feminine).

For this reason, I don’t thinks the theory that “all verbal prefixes were pronouns” fits the first draft of the Lament very well, even though such prefixes consistently function as pronouns in later versions. Carl Hostetter and Patrick Wynne do make some interesting suggestions (VSH/34) about the possible past-tense function of the -n- appearing in akhaini, udūbanim and azgaranādu, but these ideas are not incompatible with my theory that “verbal prefixes can be augments of a perfect tense”.

All these arguments are quite pedantic and extremely speculative (including my own), since they apply only to the rejected draft form of the Adûnaic language, for which very little information is available. I only discussed them at such length because they have interesting correlations to the functions of the aorist and past tenses in later versions of Adûnaic grammar.