POS - first pass

[janKipo]

Speaking of not getting the point of a discussion ... . I thoroughly agree that, as a matter of intellegibilty, the use of a separate, repeated, preposition is preferable to the more compact style. My point is only that the compact style is grammatical and may occasionally have a useful rhetorical effect. I gather that you deny that it is grammatical and apparently base this on the fact that it is classified a a preposition. This fact is of only minor relevance to tp, since words of almost all classes occupy almost all the available slots. 'lon' is preposition because it takes a NP complement in various places other than the PP place at the end of a sentence including in VP.
The comma suggestion has had no takers that I am aware of, but, on the other hand, the temptation to include a PP in the modifier string has also almost vanished (and no one uses 'pi' for it).

[jan-ante]

[janKipo]

Glad to hear someone is using the comma; it does make for clarity in a number of cases (especially Ps as modifiers).
The point of the rest is that, in 'mi lon kasi', 'lon' is in the verb slot and still takes a complement. It is not a case of 0-verb and going directly to the PP slot.(as the complementless case and the transitive one show).

[jan-ante]

[janKipo]

But tp doesn't have a copula verb, like Eng. "be" in "He is nice" or He is a man" or "He is on the swing", does the copulation (if you will) by juxtaposition, or, rather, by making the content word into the head of a VP (if you don't like calling that a verb). No copula is implied, because there is not one to be implied. To be sure, we have to add one in English, but that is a feature of English, not of tp.
Sorry, "make herbally" was your reading, not mine. Mine was "put in tree", 'lon' as a transitive verb, taking its NP complement rather than delaying to a repeated PP at the end (actually, probably not repeated, since the appropriate prep. sees to be 'tawa'). So, maybe your point is that transitively 'lon' loses its ability to take an NP complement. But I don't see any reason to think that, other than your discomfort (which isn't a datum really). I'm even willing to think that 'lon' keeps its complement as head of an NP ("noun"), though I haven't seen a case that I know of. It certainly does as a modifier, since that was what started the problem to which the comma was one part of the proposed solution.

[janMato]

I think this discussion is related to my firm opinion that everything following the verb but before the first e is ill-define outside of a few situations. From jan Sonja's sample sentences, I think she meant this on purpose, that toki pona have weak verbs. Maybe not as weak as , but weak like some papua new guinea or Australonesian languages that have 50, 20 or as few as 3 verbs, and these verbs don't really do much work. (Piraha verbs actually do a lot of work, one of the few places where even Everett admits the language is a remarkably complex and expressive)

S li V [stuff] e DO PP

When the stuff is clearly an adverb, no problems. We see pona, ike, etc in this slot. (somewhere above in this thread I listed all the things I found in the corpus that people had put here when they clearly had an adverb in mind) Well usually no problems, sometimes people gripe that these should be moved to la phrases.
Similarly, when the post verb stuff is "ala", no problems.

When it is anything else, imho, it just isn't defined anywhere in jan Pijes lessons, the jan Sonja lessons.

(A) S li V/P (preposition implying the sorts of things that verbs normally do) noun phrase. This is actually a predicate. mi tawa tomo mi, worried me until I learned about natural languages that extend predication to cover things involving motion and action.

(B) S li V (preposition masquerading as a verb or a verb masquerading as a preposition) (unmarked noun complement) e DO PP. This will continue to worry me until I find a natural language that does this as well.

(C) S li V (verbs of talking, thinking, etc) (unmarked complement, usually what we are talking, thinking about) e DO PP. We rarely see this in practice, the "mi pilin e ni: S" construction is really, really common though, but really verbose.

(D) S li V pi (nouns plus modifiers plus optionally more pi phrases) e DO PP. Who knows what this means? At least in this construction it is clear the post verb stuff is a noun.

I think that these issues mostly have to do with phrases with higher valence (more actors on the stage) and I don't think that has ever been dealt with, and when it has been, the nth actor gets put in a prepositional phrase.

Scenario (B), imho, reminds me of the constructions you see in really old languages, the ones that the children don't really learn to speak well until they're 11, the ones where saying something simple like "She owns a house" is some surprising construction like "It is a house that surrounds her." (and it's all said in one word)

Moving into the land of "jan nasa li wile ante e toki pona", I'd say that constructions B and C should have a marker. It feels like there should be one there. Nothing in toki pona that smacks of a phrase is allowed to be juxtaposed to another phrase without a separating particle. In natural languages, these particles are usually drawn from prepositions, so my if I was feeling avant garde, I'd use one of the free prepositions for the purpose. And once one went down that rabbit hole, I suspect a post verb-pre e-phrase prepositional phrase would come to indicate manner and the post e-phrase would come to represent place. Which would have a interesting correspondence to how Germanic languages seems to care about PMT (place manner time) ordering of prepositional phrases.

[janKipo]

Well, I don't know what languages these might be that a native speaker doesn't learn until he's 11 or so (unless you mean like English which many students don't quite grasp even at 21); so far as I cab tell, the native speakers of x are about as good at x as those of y are at y at a given age. to be sure, there may be niceties that many people never have occasion to use, but, by and large, if the occasion arises, they come up with (or understand) the right form (cf the French shepherd and the pluperfect subjunctive). There are, of course, languages which are harder for the speaker of one language to learn as a second language (Basque notoriously, Finno-Ugrics for English speakers), but that is just habit conflicts.

From what I remember of the talk a few years ago, the claim that kelen is verbless is a matter of definitions. It is certainly less verbless that the LoCCans but a lot more so that Menominee or English even. Claims about New Guinean languages and especially about Piranha, are all to be held at arms length for a few decades until some further studies (by people without one sort of an agenda or the other -- or with opposing agendas) are done. The history of languages that don't ... or have only ... or are otherwise totally unique is too cluttered with bad analysis, inadequate data, and just plain lying to encourage confidence.

I'm not sure that regular adverbs (pona, ike, mute. musi, etc.) should ever go in the 'la' phrase; the place and time cases are another matter -- and have pretty much sorted out to time in 'la' and place in the terminal PP.

For the rest of the "stuff," mainly what looks like NPs (and VPs for modals, which no one seems to find problematic for some reason). Using a preposition (class P) as a verb with the following NP complement is just definitional of class P and not at all surprising (it is harder to explain if you treat the members of P as verbs, though I suppose something could be worked out). Using this with a DO gets messy however, and I suppose good taste recommends deferring the complement to the terminal PP, with a repeated (or appropriately modified) preposition. There are a few verbs which also take NP complements: definitely 'nimi' (N) takes the name involved when transitive: 'jan li nimi "Jowan" e jan lili sama' (deferred using 'kepeken' apparently). There has been a suggestion that the verbs (all actually V, I think) that take quotations as deferred direct objects mark the subject of the thought/talk/writing as a modifier of the verb (so, using 'pi' if more than one word is involved) and I suppose that making it a NP complement instead would serve as well (though I am coming around to thinking that the habitual solution, just putting it in as the DO, makes as much sense as -- or more than -- other proposals). (btw, how is 'toki e ni:S' significantly more verbose than "say that S", which it mirrors exactly?)

No POS masquerades as another POS, but some words may function in the home function of a POS not its own. Indeed, in tp, most words can and a goodly number do.

[jan-ante]

[janKipo]

OK. 'lon' gets its complement as vt from having it as vi and prep. It also seems to have it as modifier, as witness the several natural uses of it in this way. I suspect it can have it as noun as well, though I know of no natural cases, but this (constructed) one seems OK to me: 'lon telo suli li pona lon" "(Life) on the sea is a good life". So, I have written the possibility for all roles, including vt, since it makes description easier. I may find that some of these cases are forbidden (the modifier case, though used naturally, is suspect). Your failure to see the meaning of the phrase 'lon kasi e ilo suno' is not a serious datum for grammar (I misread things all the time as do most folks), but it is for style: this construction is too complex and needs to be replaced in most uses by another (deferred preposition). It may also turn out that there is a better way to describe the parts of speech of tp, such that the complement is no longer part of the defining notion for 'lon' and other explanations will need to be found for their occurrence. But I don't see one now or on the horizon. Appeal to Sonja's sketches and Pije's introductory lessons (which do talk about this a little) does not help when dealing with this, a complex case, clearly from the more advanced course (never written).

[jan-ante]