Hmm. Since they share the same origin, I think amba could serve as an alternate way to form an intensive adjective, but my intuition is that it would be more emphatic than the normal intensive: “even more” or “very much”, similar to emphatic pronouns.

So nán amba alasse alassea cene tye “I am very much (or even more) happy to see you”

I could see children stacking intensifiers for further emphasis: nán anamba alassea cene etye “I am even even more happy to see you”.

+Paul Strack yes but amba alasse doesn't mean more happy, it's more happiness, since alasse is a noun - so the questions is if we use 'more' with an adjective, should we use the adjectival amba or the adverbial ambe? :) is it nán amba *alassea kene tye or nán ambe alassea kene tye?

Yes, sorry, I meant to type alassea, not alasse. That’s what I get for posting right after I wake up.

Couldn't one just use the intensifier? As in analassea?

+Tamas Ferencz Adverbs modify not only verbs but also adjectives (and other adverbs). So I would go for amba alasse 'more joy' but ambe alassea 'more joyful', cf. English "moderate joy" vs. "moderately joyful".

+Björn Fromén good point on adverbs with adjectives

+Björn Fromén we don’t have much information on how adverbs behave in Quenya, but they may not function the same way as in English. According to PE17/73-74, at the very least adverbs from adjectives were used less often than in English, with a simple adjective used instead. There are other examples of abstract nouns being used where adverbs might be in English (PE17/58-59 and PE17/93-94).

It is my opinion that Quenya would string together adjectives where English might use an adverb, for example saying “a moderate joyful celebration” rather than “moderately joyful”. Based on this, I’d use adjectival amba in front of another adjective in a noun phrase rather than adverbial ambe, and limit the adverb primarily to modifying verbs.

Since we have no examples one way or another, this is just a guess on my part, however.

There is a difference between those two in Polish. We say bardziej radosny 'more joyful' but więcej radości 'more joy' using different adverbs. The first one answers the question jak? 'how' (whereas adjectives answer the question jaki? 'what kind is it?'), the second belongs to the group answering different questions (how much, how often, where, when).
By this analogy I would use adjectives before adjectives, other adverbs and verbs (when they answer the question 'how'), but adverbs before nouns and verbs (when they answer the other questions).
But it's only a speculation of course.

At any rate boundaries between the two are blurred in Q, if we only think of the so well known laurie lantar lassi do those leaves fall goldenly, or do golden leaves fall?
Almost tempted to collect as many attestations of adverbs in sentences/phrases as possible to see if we can get any insight from that.

+Tamas Ferencz that isn’t as big a list as you might think. I skimmed through attested phrases in Eldamo and found relatively few examples of adverbial use, and not one example of an adverb modifying an adjective. Of course, there are less than a thousand attested phrases in total.

My attempt to contribute...

Let's not forget, that amba isn't marked as an adjective, but as a noun and an adjective. I believe this clarifies the matter a bit: it's more about syntactic roles than about parts of speech as such.
I believe amba constitutes noun phrases, either alone = in positions usually occupied by nouns (a direct object: "I want more"; a subject: "there is more"; etc.), or as an attribute to a noun = in positions usually taken by adjectives ("I want more bread"; "there is more bread"; etc.).
Whereas ambe constitutes adverbials, adverbial phrases ("I want it more (than you)"; "he is more reliable (than you)"; etc.).

+Lokyt L. that's an excellent insight - but then you support the idea that before adjectives the adverbial form is required (more reliable)

+Tamas Ferencz Well... Surely ambe is not permitted in substantival and adjectival roles. But whether amba is permitted in adverbial roles, I don't dare say (as I don't generally know enough about the problem of adjectives taking adverbial roles).

... Which means that IMHO either both ambe and amba can go with adjectives, or just ambe - but certainly not just amba.

+Lokyt L. while amba is marked as substantive as well as adjectival, that may not mean all that much, since we know that, like English, Quenya can use adjectives as nouns.

I think if Quenya used adverbs to modify adjectives, it would be hard to determine their exact function. Consider these two English sentences:

“I saw quickly a dying man.”

“I saw a quickly dying man.”

But in Quenya: cennen lintave quelesta Atan.

With its freer word order and lack of indefinite article, I think Quenya would need to be more rigid about the grammatical function of adverbs. In particular, I think they should modify either the verb or the context of the entire sentence, whereas nouns and noun phrases should only be modified by adjectives. I think I would write:

cennen lintave quelesta Atan “I saw quickly a dying man”

cennan linta quelesta Atan “I saw a quick[ly] dying man”

Granted, this introduces a new ambiguity: is the man “quickly dying” or both “quick and dying”. Still, I think this could generally be deduced from context and where it couldn’t, it could be managed by shorter vs. longer pauses between the adjectives: linta-quelesta “quick-dying” (a near compound) vs. linta quelesta or even linta yo quelesta.

Granted, this is all pretty speculative, but I like that it makes Quenya grammar more distinct from English, and it seems be in the spirit of what Tolkien wrote about the limited use of adverbs in Quenya.

-sta is a verbal noun ending, IMO what you need here is an active participle (quelila); we also have lintie/lintiénen attested as the adverbial form of linta.
This ambiguity arises in this particular case because Q does not have an indefinite article which would separate the noun phrase ("I saw quickly a dying man" "I saw a quickly dying man"); but as soon as we're talking about a specific man the ambiguity disappears (kennen lintie i quelila Atan/kennen i lintie quelila Atan). Word order can resolve the ambiguity: lintie kennen quelila A vs. kennen lintie quelila A, even with Acc+Infin kennen Atan quele lintie.

+Paul Strack I must admit I'm not sure what you mean by "either the verb or the context of the entire sentence"... As far as I understand, modifying the main verbal phrase (the predicate) pretty much means modifying the whole clause (as all other constituents are by definition modifiers of the predicate as well), doesn't it?

+Paul Strack But anyway, the important thing is that we agree on "[adverbs] should modify ... the verb ..., whereas nouns and noun phrases should only be modified by adjectives" :)
And as for what are adjectives modified by... As I said previously, It's a complex question that I can't attempt to answer with the knowledge I have.
(This being said, I however somehow arrived at a conclusion that the nature of the "not-much-use-for-adverbs" property of Quenya is rather that adjectives can be employed as adverbs in any position, no matter what they modify, if the context doesn't require an explicit adverb in order to make the construction comprehensible/unambiguous. Which indeed might be the case of "a quick, dying man" vs. "a quickly dying man".)

And by the way, if the default modifiers of adjectives would be other adjectives, than "a dying man" might also bring a question whether these participles / verbal adjectives do act like adjectives or like verbs in regard to the nature of their modifiers.

+Lokyt L. I agree that we are in broad alignment, and that the main open question is whether Quenya adverbs can modify adjectives. Since we have no clear attested examples, any “answer” must be speculative and subject to our personal tastes. I also think there's an interesting discussion to be had on the behavior of verbal participles as adjectives, but I would rather defer that to a separate thread.

+Tamas Ferencz I agree that my choice of quelesta “dying” was a poor one. It is an attested adjective, but it is “Middle Quenya” from the period when the verbal noun prefix was -ste. This conflicts with later form of the verbal noun suffix as you pointed out, so I agree that quelila is a better choice.

My choice of lintave over lintie is, however, more deliberate. In particular, I don’t think that lintie is actually an adverb, but is rather an abstract noun (“swiftness, speed”) which may be used as an adverb to modify a verb.

I think the phrases on PE17/58-59 make this clear:

norne lintie “he ran quickly, (lit.) he ran [with] swiftness”

This is even clearer with:

norne lintieryanen “he ran as swiftly as he could, (lit.) he ran with his speed”

Since lintieryanen is in the instrumental case, it is definitely being used as a noun.

The fact that so many Quenya adverbs (lintie, márie, laurie) also function as abstract noun is one of the reasons why I think Quenya would not use “adverbs” to modify adjectives, only verbs. Allowing nouns to modify adjectives would be extremely counter-intuitive.

I chose the more “adverbial” lintave in my examples above because I think it is conceivable that it could be used to modify an adjective, even though I personally would prefer to use linta. However, I think the chances that lintie “swiftness, speed” could be used to directly modify an adjective is extremely low.

+Paul Strack you marked qelesta in Eldamo as a noun.

+Tamas Ferencz it is a noun. Then my use of it was even more wrong :)

+Paul Strack ^^^ I see what you did there. Lol

+James Coish that was ... entirely accidental. I wish I could claim to be that clever.

+Paul Strack Lintieryanen is an instrumental of a noun? That's interesting, 'cause some natural languages (my native included) really can create adverbs from nouns by simply switching them to the instrumental case. And what is of relevance, these desubstantival adverbs CAN (at least in some Slavic languages) modify adjectives, BUT only deverbal ones, i.e. participles; not the true, primary adjectives.

+Lokyt L. Tolkien classified certain noun cases as “adverbial”. These were originally used to form adverbs from nouns, but by the Third Age were “normal” noun cases. I suspect that they were no longer true adverbs.

+Paul Strack Oh. May I ask which noun cases are these?

It depends on the time period. In later writings it’s the allative, ablative and locative cases (and maybe also instrumental). Earlier there were additional comitative and “adverbial” cases (which I call the similative), but Tolkien seems to have abandoned these.

Ah, now I understand - "nouns in adverbial cases" actually means "noun forms capable of constituting adverbial phrases". Thanks.
(And yes, we have these in Slavic too :) )

+Paul Strack But back to the original topic:
You don't think amba is a declinable noun, or do you?

+Lokyt L. Maybe? It’s hard to guess. Clearly plural and possessive inflections make little sense. But I can imagine other inflections might work. For example:

lauyuvalme ambanen “We will flourish with more”.