I would say that there is a subtle distinction between vespermanĝi and manĝi vespere. The latter could refer to any snack or act of eating in the evening, whereas the former is more of a regular occasion. Vespermanĝi as a compound word for manĝi vespere could be a bit misleading because of the established meaning of the regular, even ritual vespermanĝo.

Therefore, in my view, whereas the root manĝ' is verbal, the compound root vespermanĝ' should be regarded as substantival. Thus vespermanĝi can take an indirect object, e.g. Mi vespermanĝis per kokidaĵo, but if we are talking only about the act of eating, not the occasion of a meal, it might be better to say Mi manĝis kokidaĵon vespere instead of Mi vespermanĝis kokidaĵon.

BTW PIV 2005 marks vespermanĝi as intransitive!

Miland:
Therefore, in my view, whereas the root manĝ' is verbal, the compound root vespermanĝ' should be regarded as substantival.

Is there such a thing as a compound root in Esperanto? I'd have thought that was an oxymoron.

"...whereas the root manĝ' is verbal, the compound root vespermanĝ' should be regarded as substantival"

Meh, part-of-speech/semantics confusion. I don't want to say I told you so, but...

"With some roots the meaning of them in a different grammatical guise does not attach itself easily to something in the real world." - Sudanglo

But shouldn't they? Shouldn't all forms be applicable? What I did in Angos (sorry to keep bringing it up, but it's how I can explain things), was make every root a substantive, and assign no specific meaning to the verb form. The equivalent of "pani" would be whatever you do with bread using the context given. It could be "give bread" or "bake bread" or "eat bread"- whatever the context allows. I think it allows for more freedom in meaning while cutting down on vocabulary. In context, many verbs are understood (as demonstrated by 3rdblade).

I'm no expert in Esperanto, but going off of what erinja said, it would probably be "be bread" or to "bread" something (to coat in bread crumbs).

I agree with you Miland - and Todd there's a reason why you have never heard vespermanĝi used transitively. People don't use it this way.

As regards whether manĝ has a different status to vesper so that one can be called a root and the other not, I see no reason for this. They are both lexical radikoj. And together they become almost like a new root.

There is no need for a convoluted theoretical superstucture here. The building blocks of Esperanto are just strung together to in a necessary and sufficent way to produce words which convey the meaning. Compounds that can't readily be attached to something in the real world are not used.

Because we humans have regular meals, one of which is in the evening the 'o' that is vesper and manĝ is dinner. And the 'i' that is vesper and manĝ is the act of having dinner.

Short answer Razlem. There is no need for all possible compounds to be applicable - why should there be.

To coat in bread crumbs would be paneri, or panerumi, or panumi. NPIV documents panumi in this sense. The reason why it isn't pani is because it is felt that it is not clear what this should mean.

razlem:"...whereas the root manĝ' is verbal, the compound root vespermanĝ' should be regarded as substantival"

Meh, part-of-speech/semantics confusion. I don't want to say I told you so, but...

It's okay. This particular issue has been kicked around for a very long time.

"With some roots the meaning of them in a different grammatical guise does not attach itself easily to something in the real world." - Sudanglo

But shouldn't they? Shouldn't all forms be applicable? What I did in Angos (sorry to keep bringing it up, but it's how I can explain things), was make every root a substantive, and assign no specific meaning to the verb form. The equivalent of "pani" would be whatever you do with bread using the context given. It could be "give bread" or "bake bread" or "eat bread"- whatever the context allows. I think it allows for more freedom in meaning while cutting down on vocabulary. In context, many verbs are understood (as demonstrated by 3rdblade).

I'm sure that would work, and in cases where ambiguity would result, additional language could be used.

Would you like to come to my house later and bread?

It could mean eat bread, bake bread, or apply breading to something. It could be easily disambiguated simply by saying more specifically what the plan is.

The point isn't that this wouldn't work. It's just not Esperanto. And there's really nothing to be done about that.

sudanglo:I agree with you Miland - and Todd there's a reason why you have never heard vespermanĝi used transitively. People don't use it this way.

That, however, isn't normative. There's no apparent reason why one couldn't use it this way. At least, I haven't seen one yet.

As regards whether manĝ has a different status to vesper so that one can be called a root and the other not, I see no reason for this. They are both lexical radikoj. And together they become almost like a new root.

There is no need for a convoluted theoretical superstucture here.

Indeed? They why invent a novel, and undefined, entity called a compound root? Or "almost like a new root"? What exactly is the point of adding this bit of superstructure? In this case, I'm the one applying the principles of Esperanto as they are, without inventing new ones.

Compounds that can't readily be attached to something in the real world are not used.

That's fine, but vespermanĝi, used transitively can readily be attached to something in the real world!

Because we humans have regular meals, one of which is in the evening the 'o' that is vesper and manĝ is dinner. And the 'i' that is vesper and manĝ is the act of having dinner.

Indeed, and when we dine, we don't just sit there and breathe. We dine on something. Every time! Vespermanĝi is a verb, with two roots, of which the primary one is a transitive verb. Why should the addition of vesper- change it to an intransitive one? The fact that we get manĝo=meal from manĝi=to eat has no implication for the transitivity of manĝi; to make it otherwise for compounds with manĝi is precisely to add convoluted rules that accomplish nothing.

Shall we say, then:

Ni vespermanĝis je bovaĵo? What possible misunderstanding could we be preventing by saying that instead of Ni vespermanĝis bovaĵon? Indeed, using the preposition replacement property of the -N ending, we get Ni vespermanĝis bovaĵon anyway.

I think that the following Ekzercaro passage should make it clear that use of vespermanĝi with -n does not violate any rule.

El la dirita regulo sekvas, ke se ni pri ia verbo ne scias, ĉu
ĝi postulas post si la akuzativon (t. e. ĉu ĝi
estas aktiva) aŭ ne, ni povas ĉiam uzi la
akuzativon. Ekzemple, ni povas diri “obei al la patro” kaj
“obei la patron” (anstataŭ “obei je la
patro”). Sed ni ne uzas la akuzativon tiam, kiam la klareco de la
senco tion ĉi malpermesas; ekzemple: ni povas diri
“pardoni al la malamiko” kaj “pardoni la
malamikon”, sed ni devas diri ĉiam “pardoni al la
malamiko lian kulpon”.

Miland:Mi vespermanĝis per kokidaĵo

I just did a search in Tekstaro and found a phrase in La ŝtona urbo that puts it in the way you suggest: Mi kaptis la restaĵon de la pano per kiu ni vespermanĝis, kaj ŝovis gin en la kuirpoton. Notably this was the only example where the object of the "suppering" was even mentioned at all; the others were all of the normal intransitive usage.

I would agree with the intransitive classification of vespermanĝi as that reflects its common usage, but I don't think I would disagree with the use of the accusative. To take a well-worn example, iri is basically intransitive but the preposition-dropped form "iri vojon" is considered acceptable, despite it's rarity, because the meaning is clear in terms of the relationship between the action and the associated object. For me the same clarity exists with "vespermanĝi bovaĵon" and so I think it should not be deemed wrong. Though on stylistic grounds I can well imagine that using a preposition in the way Miland suggests, which helps to preserve the intransitive idea inherent to the verb, could be the better option.

sudanglo:Short answer Razlem. There is no need for all possible compounds to be applicable - why should there be.

Maybe I misunderstood you. I was talking about vowel classifiers, not compounds.