Ilk. [ŋg] vanished before [w] lengthening the preceding vowel; [-Vŋgw-] > [-V̄w-]
In Ilkorin, it appears that [ng] vanished before [w] with compensatory lengthening of the preceding vowel, as suggested by Helge Fauskanger (AL-Doriathrin/líw). The best evidence for this change is ᴹ✶liñwi > Dor. líw “fish” (Ety/LIW). Since medial [ŋw] became [ŋgw] in Primitive Elvish (PE18/54), the medial development of this word was [-iŋw-] > [-iŋgw-] > [-īw-].
A second possible example is ᴹ✶neñwi > Dor. nîw (Ety/NEÑ-WI), also suggested by Helge Fauskanger (AL-Doriathrin/nîw). The development of this example is less obvious, but if the rule that [e] became [i] before [ŋg] also applied, then its medial development could have been [-eŋw-] > [-eŋgw-] > [-iŋgw-] > [-īw-].
A possible counter-example to this rule is Dor. ring from the root ᴹ√RINGI (Ety/RINGI). Its cognate ᴹQ. ringwe implies a primitive form [riŋgwi], which should have become **rîw if its development was consistent with the rule above. However, it is possible that this Ilkorin word was not a direct cognate the Quenya, developing instead for the simple root form [riŋgi].
On the other hand, in earlier (deleted) entries for the root ᴹ√LIW, the development was ᴹ✶lingwi > Dor. ling “fish” (EtyAC/LIW) instead of líw. This matches the development described in the Comparative Tables: ᴹ✶-ŋw- > [-ŋgw- >] Ilk. -ŋg- (PE19/23). This indicates that Tolkien was uncertain of the development of [-ŋgw-] in Ilkorin, vacillating between having it become [-ŋg-] or having it follow the rule above.
|After||01700||[e], [o] became [i], [u] before [nn], [nd], [ŋg]||ᴹ✶NEÑ-WI > Dor. nîw||Ety/NEÑ-WI|
|Before||04600||[w] vanished after medial velars||ᴹ✶liñwi > Dor. líw||Ety/LIW|
Phonetic Rule Elements
Phonetic Rule Examples
|liŋgwe > līwe||-Vŋgw- > -V̄w-||ᴹ✶liñwi > Dor. líw||✧ Ety/LIW|
|niŋgwe > nīwe||-Vŋgw- > -V̄w-||ᴹ✶NEÑ-WI > Dor. nîw||✧ Ety/NEÑ-WI|