The first one made me wonder, can we tell how to form an instrument-name (noun) from an A-stem verb, should one exist for "sweeten"? (If not, then a possible way around that would be to take noun meaning "sweetness" and add an element like *tyarya [cf. matya] or antë).

N.B: Qenya Lexicon has the distinctive synonyms for "powder" mar (mard-) (#2) (ixt. marma, Qe "sand") and mul (muld-).

Etymology of the European word: http://www.etymonline.com/index.php?term=sugar&allowed_in_frame=0
[The known verbs for "to grind" are Qe mul- and marda-.]

+ܤܡܝ ܦܠܕܢܝܘܤ an instrument name - or an agent? For are agents necessarily animate? Could we have perhaps *lissetar for 'sweetener'?

Tamas: I mused on whether it was possible to make a noun that would mean "sth. which is used to achieve/perform X-ing". (In Finnish -in[1] is used to form these; I don't know if Hungarian has a corresponding affix.)

[1] http://en.wiktionary.org/wiki/-in#Suffix_6

+ܤܡܝ ܦܠܕܢܝܘܤ we use the ó/ő present participle (and adjectival) ending for that purpose:
szór 'sprinkle' szóró 'sprinkling, sprinkler'
édes 'sweet' édesít 'sweeten' édesítő 'sweetener'

Somehow any kind of descriptive compound sounds very artificial to me.. Sugar belongs to a group of commodities which spread along trade routes and whose names were loaned across languages rather than invented. If we have no evidence of sugar being present in Middle-earth (it doesn't appear in the LotR, that much I can say), perhaps a loan into Neo-Quenya might actually be best - *su(k)kar or something like that.

I find *tyavinqua better than *alatyave because of the confusion with the negative prefix (see alasaila). One might also use amya-, mai- *'excellent, admirable' from PE17:163, *maityave/*maityávea, *amyatyave/*amyatyávea. Or actually simply *tyávea.

I cannot find QL OLOK-..

We have N. cevnas 'earthenware' (VT45:19) -> Q. *kemnasse. It looks like a mass noun, but might be usable for a single jar as well.

+Roman Rausch mea culpa, not OLOK, LOKO-!

On FB Helge cited Qenya matsa 'delicious' as a possibility.

A good discussion. I'm still thinking over all the proposals.
And yes, I find prefix mai- less ambiguous than ala-.

+Ицхак Пензев
indeed it is less ambiguous. Probably the best recourse is to use the attested words that use ala- 'well' but consider the prefix as improductive.

Note that PE21:12,13 gives Qenya tambe "pot, jar"

Yes! Yes! Now I remember that I saw this word, and even put it into my notes, and totally forgotten where I wrote about it, and couldn't find it.