The Primitive Elvish long vowel [ę̄], a slacker and lower long [ē], developed into [ai] in Sindarin (PE18/96). A similar sound change occurred in Noldorin, except at that conceptual stage Tolkien represented the sound as [ǣ], with a slightly different pronunciation. Tolkien described this sound change in several places, with somewhat different details each time:

In the earliest recorded Noldorin, [ē, ǣ, ā, ǭ, ō] remained distinct though changed to [ī, ei, ǭ, ou, ū]; later ei, ou products of ǣ, ǭ became ai, au coalescing with original ai, au (first version of Tengwesta Qenderinwa [TQ1], 1930s, PE18/46).

The diphthong ai̯ and archaic ei̯ (from PQ ǣ), which later became ai, were written ]Ö (lÖ) (ai was never written ]`B, but ]~B was sometimes used in the archaic period) (Feanorian Alphabet, 1930s, PE22/27).

In the earliest Noldorin ē, ę̄, ā, ǭ, ō appear as ī, ę̄, ǭ, ǭ, ū; later ę̄, ǭ coalesce with as ai, au (second version of Tengwesta Qenderinwa [TQ2], circa 1950 written shortly before the change of Noldorin to Sindarin, PE18/96).

The sound ę̄ (or ǣ) was fairly rare in Primitive Elvish, arising only from the a-fortification of e. There are no clear examples of this phonetic development in Sindarin, but there are a number of Noldorin examples in The Etymologies:

• ᴹ✶ndǣr > ON. ndair > N. doer “bridegroom” (Ety/NDER) or daer (Ety/DER).
• ᴹ✶kwǣ-nē > ON. paine > N. poen “small gull, petrel” (EtyAC/KWǢ).
• ᴹ✶wǣdē > ON. waide > N. gwaedh “bond” (Ety/WED).

The phonetic development of [ǣ] > [ai] is clear; later developments are obscured because of Tolkien’s vacillation on whether [ai] became [oe] or [ae] in Noldorin. Since [ai] became [ae] consistently in Sindarin, it seems likely that the complete phonetic development at that later conceptual stage would have been [ę̄] > [ai] > [ae].

In the Comparitive Tables for phonetic developments composed in the 1930s, Tolkien shows a more intricate development than the one given above. In particular, both ai and ǣ merged into ę̄, which then becomes ei and finally ae (PE19/25). Perhaps also ei > ai > ae, but this is not shown. There are hints of this fuller development in The Etymologies:

• ᴹ✶gaisrā > gǣsra > ON. gērrha [gę̄rrha] > N. gaer “dreadful” (Ety/GÁYAS).

Assuming this fuller phonetic development was the rule, the development of primitive [ǣ], [ai] into Noldorin [ae] (or [oe]) would more closely parallel the complex sound shifts whereby of [ā], [ǭ], [au] ultimately developed into Noldorin [au]:

• ǣ/ai > ę̄ > ei > ai > ae or oe
• ā/au/ǭ > ǭ > ou > au

This could help explain why sometimes [ae] became [e] in polysyllables in Sindarin and Noldorin; it may have been a shortening of the intermediate form [ę̄], in much the same way that [au] may have become [o] in polysyllables due to the shortening of intermediate form [ǭ]. It could also be a factor in why [ei] sometimes became [e] in unstressed final syllables (Noldorin only), but this is less clear.

With its longer series of sound changes, the Noldorin development of [ǣ] > [ę̄] > [ei] > [ai] was spread out over more time. The evidence in The Etymologies and the sound changes described in the Comparative Tables indicated that most if not all of these changes took place in the Old Noldorin period. Without attested examples, dating the analogous changes of [ę̄] > [ai] in Sindarin is impossible. For convenience of discussion, this lexicon places it immediately after the better attested sound change whereby [ǭ] became [au].

Conceptual Development: There are only a few indications of ǣ in Tolkien’s earliest conceptions of Primitive Elvish from the 1910s and 1920s: there is a passing mention in a very early chart of Gnome Vowels: “æ̂ gave ê” (PE15/13), and ǣ appears as a final vowel in primitive ᴱ✶Ŏrŏmǣ > G. Orma (GL/63). Tolkien properly introduced a-fortified vowels into the Primitive Elvish in the 1930s (PE18/46). The basic development seems to have remained the same thereafter ([ǣ] or [ę̄] > [ai]), but there were some vacillations on intermediate changes and later developments ([ai] > [ae] vs. [oe]).

@@@