órerámar wrote:But as shown by i·coimas Eldaron, the waybread made and used by the Eldar alone
Well, this is an incomplete citation. The "Of Lembas" says "...The Eldar did not give it to Men, save only to a few whom they loved, if they were in great need". In other words, they did seldom give it away for the reasons explained, but they did. The bread could be eaten by others. This justifies the genitive apart from the fact that there was a tendency to prefer that form.
The relevant passage in "Quendi and Eldar" tells us that a shift from lambe Eldaiva to lambe Eldaron for 'the language of the Eldar' could be justified only if it meant that the whole language was adopted by another people. Should really a take-over like that be comparable to some occasional gifts for a few others in great need? I'm afraid I can't follow you there.
órerámar wrote:The possessive of the general plural must at most have been a rare form, considering that the only instance on record is Eldaiva/Eldaive in one single text ("Quendi and Eldar") and that there was an increasing "tendency to prefer the [genitives] or use them in place of the [possessives]". The form is also conspicuously missing in the Plotz declensions.
In the same paragraphe WJ:369, where Tolkien states that there was a tendency to prefer the derivative genitives and says that "alkar Oromeo" or "alkar Oroméva" could both be used for "the splendour of Orome", he continues however explaining that those forms do not say exactly the same thing.
What he explains is that they didn't originally. To me it's obvious that the account in "Quendi and Eldar" is essentially etymological. It gives us the historical background of the two case forms but doesn't describe the actual use of them in Middle-earth Quenya. Otherwise there would be too many contradictions with the bulk of the extant corpus. Even a text as canonical as Galadriel's Lament doesn't conform!
órerámar wrote:Well, as the instrumental (ciry)ainen corresponds to (lass)ínen, I think analogy points to *(oront)íva as counterpart of (Eld)aiva.
Sorry, but I do not think that these forms are simply the counterparts of the others. Lassínen (lasse +inen) and orontiva (oron - stem oront + iva) are suffixed in a different way.
As no such plural as *orontiva is attested, we can't very well be certain how it would be suffixed, can we? The universal instrumental marker is -nen (PE 17,p.62). The instrumental plurals are formed by adding this suffix to the archaic accusative, thus ciryai-nen, lassí-nen (acc.pl. ciryai, lassî ) and presumably *orontí-nen (acc.pl. *orontî). The only general plural possessive on record is similarly formed by the suffix -va: Eldai-va (acc.pl. *Eldai). Analogy would yield *lassíva and *orontíva; but of course Tolkien may have thought otherwise, if he ever decided on the matter.
órerámar wrote: The rule seems to be that words with three syllables or more and of which the last syllables are short get a stress and that stress is placed where normally the secondary stress is, (cf. RGO) : Orome - Oroméva.
I agree, something like this seems to be rule when inflexions are added to stems ending in -â, -ê, and -ô, as shown by pairs like ciryali:vanimáli(on) and Valinóreva: Eldaliéva; but it doesn't apply with -î: tárí-va, lassí-nen.
órerámar wrote: Yes, táríva has finally confirmed that two "i" would merge into a long "í"
Two "i"?? The stem târî- has one long i, which is shortened finally (nominative tári) but remains long when the suffix -va is added.
órerámar wrote:But why wrestle with this dubious form, which we can easily do without? I can't think of any situation where neither the genitive *orontion nor the possessive *orontelíva would work equally well.
Because, as Tolkien demonstrates in WJ:369, you may more or less translate into the same thing in English, but it is not the same.
Nevertheless he himself apparently didn't find much use for the general plural possessive.