I don't think it would be beyond the competence of learners to cope with the fact that there's more than one possible word for daughter. I wouldn't exclude anything from the vocabulary unless it's contradicting or extremely speculative. (I'm not talking about grammar)

It's common for languages to have multiple words for the same thing... Heck, that is why we have thesauruses

Actually I disagree. I think it is not easy for new learners to make good choices between contradictory vocabulary words. I know it took me several years of study before I knew enough to make proper distinctions.

It’s not just a question of synonyms either. There are a large number homonyms with wildly different and even contradictory meaning, which makes reading comprehension very difficult.

I don’t think Tolkien would have tolerated a language in which yelde, yende and selde all simultaneously meant “sister” and I don’t think we should either.

However, making the thousands decisions needed to produce an internally consistent Elvish vocabulary is a stupidly huge amount of work, requiring years of research, so I’m not surprised it hasn’t been done yet.

+Paul Strack I normally work from the roots. The good starting point is to check whether the whole set of words derived from the root makes sense diachronically in-universe. But it will always be a case-to-case scenario. There is no other algorithm than trying to be as inclusive as it still makes sense, while preferring the words with a strong historical connection. Such as the family relations from Fingers.

+Paul Strack I really wouldn't exclude any word from the vocabulary only because it causes ambiguity (and here I am again not talking about mutually exclusive versions of pronominal endings for instance, but cases like you mentioned with several words meaning "sister"). We cope with ambiguity on a daily basis in our real life languages; we even have contronyms and we don't even bat an eyelid at them.
I also think that having choice in the vocabulary would give the langauge a chance to evolve organically - for instance, if the number of speakers picks up and people start using the language, in time the various words having the same meaning could start to diverge and gain added colour (e.g. users may start applying one of the words for "daughter" to daughter-in-laws).

+Tamas Ferencz I get where you are coming from, I just don't agree. I don't think anyone here would suggest using Early Quenya sui or G. suil for “daughter”, so why should we accept all of yelde, yende, selde and anel just because Tolkien invented these forms after 1930? While Tolkien considered each of these forms a valid word for daughter at one point in time or another, at no time did he conceive of them all being simultaneously valid.

Regarding which I think is “best”, I think the question is easy to answer by comparing to Sindarin. In the Etymologies we have archaic †sell “girl”, from which is derived N. iell “daughter” on analogy with N. ionn “son”. But in the King’s Letter we have sellath “daughters” and in the Túrin Fragment have the word sel: unglossed, but from context clearly “daughter”. This indicates that by the 1950s, Tolkien changed N. iell >> S. sel(l). For consistency with its Sindarin cognate, we should probably use Q. selde for “daughter”. The later validity of this form is supported by the diminutive finger name selye from “Eldarin Hands, Fingers and Numbers”.

If we all agree that having a normative grammar for the languages is a good thing, I think it is also worth aspiring to have a normative vocabulary list as well, especially if we want to have a coherent language. Such a normative vocabulary is a larger and more difficult undertaking, but it is, in fact, the ultimate goal of my Eldamo project. At the rate I am going, it will probably be another 5 or 6 years before I get there, though. Something community based might get there sooner, and produce a better result.

And +Александр Запрягаев I agree that basing such a normative vocabulary on proper roots and phonological histories is essential.

I know that many natural languages have dialects where they may have different words even for basic vocabulary. And there is nothing bad in using them. But the same languages are striving for a unified national standard for school and media.
That is why I would vote for a normative vocabulary, too.

In the end only usage can determine whether a specific word becomes part of an active vocabulary or not. If speakers start using a word in a specific setting then it will become part of the vernacular. If Quenya ever gets to that blissful state when it has a lot of speakers...
(and having unified national standards is absolutely fine until people start getting stigmatised for not speaking "the norm", as I have seen so many times in Hungary, for instance, or what went on in the UK at the BBC for a very long time - but that is veering off-topic now)

+Tamas Ferencz that's an eternal battle between descriptivists and prescriptivists. We've got no luxury of having any significant number of Neo-Quenya speakers to let the usage determine. Nevertheless, I want us to remain friends even if we disagree on such crucial issues.

+Ицхак Пензев that's no question:)

+Tamas Ferencz nalvë acca mancë costien!

+Ицхак Пензев namin lá kostalve, rie artaquetilve:)

+Tamas Ferencz To be sure, even though I believe that a normative vocabulary is desirable and the selde is the “best” choice for the Quenya word for daughter, I can’t say that someone who thought otherwise, or used yelde, is “wrong”. The only person who could make those kinds of statements hasn’t been with us for over 45 years. Even if I finish putting together such an Elvish dictionary, it will necessarily have to remain a set of suggestions rather than anything authoritative (since I have no such authority).

Anyone using anel should be publicly ridiculed, however ;)

One more comment: even though I intend at some point to produce a normative dictionary, I also intend to maintain the Eldamo lexicon in its current, primarily descriptive form. I think normative vocabulary would be particularly helpful to new students of the language, but I hope new students eventually graduate to experts and start making their own contributions.

Indeed, the lexicon should include everything attested. And the so called "normative vocabulary" should be only a starting point for a total novice.