i don't think these are the pages i remember talking about derivation but i'm almost home give me like a half hour im busy
Can you do this with transitive verbs? "jan wile"
EDIT: can intransitive verbs become nouns? "wile li esun" or something
Thinking about this again; wouldn't it be simpler to say that there are no nouns, and then derive nouns from intransitive verbs?
this is roughly the vein of the breakthrough i came under after reading this thread, mainly influenced by being a tokiponist who is learning toaq now!
realizing that toaq seems to prefer promoting verbs into nouns by making the noun be the subject of the verb:
e.g. kıa “to be red” > kía “the red one(s)”, kueqtua “to collect” > kúeqtua “the collector(s)”,
i then realized that toki pona seemingly would do the opposite, aside from being able to abstract the verb in the same way an OOP language like Python can treat methods:
palisa “to be like a stick” > “a stick”, moku “to eat” > “food” (something that is eaten)
(in the process of realizing this, it also helped me realize that sitelen is more than something that is seen and kute is more than something that hears ahah)
non-copular verbs, from my experience, seem to applicational in a few ways:
causing an action (these examples are copular, but this seems like it can also allow for non-copular ones with implicit objects)
sina suwi e pan “you will sweeten the dough”, “you cause the dough to be sweet”; soweli li moku e kasi “animals eat plants”, “animals cause plants to be eaten”; ni li pakala e nasin “this broke the bridge”, “this causes the path to be broken”
causing the presence of (i.e. applying) a noun:
ona li sijelo e ilo mi “they bumped into my boat” (credit to lipamanka for the idea), “they applied a body to my tool”; sina moku ala moku e sike ni “did you put peanut butter on this ball?”, “did you apply food to this ball?”; jan li kiwen e akesi “the people are stoning the dragon”, “the people are applying stones to the dragon”
1 Like
I would classify sense 3 (moku1) as simply being sense 4 (moku2) without specifying the object
In addition, sense 2 (moku0?) could take the transitive form "X turns Y into food".
Anyway, I would understand primarily "mi moku e pan" as "I eat the bread", and not the other transitive form "I turn the bread into food" because the bread already is food.
For "mi moku e kasi", I would be inclined to interpret it as the plant already being cooked, but I can see "I make the plant edible (i.e. by cooking it)". However, I would use "seli" for "to cook" instead.
For the variants with two objects, I see them as being the same verb applied to both, e.g. "I eat the bread and the (implied to be cooked) plant". Assuming the kasi isn't already cooked, then none of the English translations fit the Toki Pona text, as "pan e kasi" and "kasi e linja" both would need two different verbs, and "pan e linja" is eating both of them.
1 Like
oh yeah, that might be where I get that impulse from. I think this might be something it inherits from lojban? which is all about shuffling places around until you get the thing you want into the subject place so you can nounify it
isn't the arrow the wrong way around here? if you're doing the opposite of getting a noun from the subject of a verb, shouldn't you be deriving "to be like a stick" from "a stick"?
As for the rest, I'll have to take a look at it tomorrow, it's quite late here. I also want to actually take a look at pu.
I got distracted writing a CFG for toki pona...
1 Like
both jan wile and wile li esun are perfectly sensible. Note that grammatically, li is the particle that disambiguates transitive verbs from adjectives derived from transitive words in the subject. Consider:
jan wile li tawa tomo pi jan olin ona // The wanting-person goes to their lover's house.
jan li wile tawa tomo pi jan olin ona // The person wants to go to their lover's house.
Unsure whether this interpretation actually produces a materially different analysis. I do think that it's a bit nonsensical (at least in tp) to talk about verbs as actually having a transitivity class sinceall verbs can be slotted in both a transitive an an intransitive slot.
Lipamanka mentioned that all verbs could be thought of as transitive and just missing an argument when spoken intransitively. I say that this is also a little bit nonsensical (which, I suppose, also means that tp's underfilling is not so straightforward) -- there are a handful of words that truly work differently in an intransitive capacity. Consider mi lon. Applying a direct object to make lon transitive always changes the meaning of this sentence. Intransitively, it refers to just existing. Transitively, it actually refers to being somewhere or having some type of context. (I suppose you could make the claim that transitive lon is "existing in some context" but I am not sure whether transitive lon actually makes the veridical claim that you do, in fact, exist. Its capability to be used metaphorically suggests to me that it does not make said claim.)
EDIT: see Polysemy testing - #36 by _at
1 Like
Cool, yes, that's what I wanted to confirm! That the transitive→intransitive (by filling the object with "ijo") works in other grammatical contexts than where it's the main verb.
I think they were only arguing that there are no intransitive verbs? Only nouns, adjectives, and transitive verbs.
But I disagree with that, there's at least "pu".
Not sure about the with-object part, mostly because I don't recall the grammar for prepositions-as-main-verbs, but at least you can argue that "mi lon" has "lon" as an adjective rather than an intransitive verb.
jesus so much discussion damn uh I will try to get back to this by tomorrow!
3 Likes
Pedantic note: pu is an adjective. lipu pu is the book describing Toki Pona.
Ehh, I think this one's pretty cut and dry in that mi lon ijo has lon as the main verb. The dropping of the li particle can happen only when the subject is a bare mi or sina and therefore is a pretty good indication that the main verb is next. If this was a noun phrase and not a full sentence, you would be right, it would be ambiguous whether lon is a verb or adjective.
I note that there would be no issue with translating that as "wanting-person", the "wanted-person", or the "person of want". You can kind of put the ke'a wherever you want, lol.
a, nimi.li has it as "interacting with the book Toki Pona: The Language of Good by Sonja Lang", which I took as an intransitive verb.
Looking at the semantics spaces dictionary definition, it sounds more like an adverb?? I didn't know toki pona had those...
Yes, but:
(don't wanna do emphasis, because I hope this way Discourse can identify the context, but imagine I bolded adjective)
Also,
my point was only about "mi lon" used as a full sentence, where I argue that it could have the adjective as the "main verb" (think "mi laso") which means that "mi lon" is not necessarily evidence for the intransitive verb "lon" existing (unless you want to instead say that adjectives don't exist and "laso" and "lon" are both intransitive verbs, which, like, sure I'd be fine with that)
1 Like
remember that there are no lexical intransitive verbs in toki pona. all lexical "verbs" are transitive. This goes for lape too, which is an adjective, not a verb. (which is hard for us anglophones to wrap our heads around because every word in english for being asleep is a verb or some form of a verb)
1 Like
Point taken, mi lon is not proof of the existence of intransitive verbs. Going off of the exact verbage from lipu pu,
it seems it is sensible to interpret what looks like intransitivity as another word class entirely, filling the slot of the main verb. I think we're on the same page now.
1 Like
On what basis do you much such a claim as there not being intransitive verbs in Toki Pona?
it's almost 03:30 here, and despite my sleep schedule being severely off from my ideal (22 bedtime, awake around 07; unfortunately I do need that much sleep), it is definitely past my bedtime even in my internal time zone, so I will have to return to this thread tomorrow. I skipped parts of some posts for "too sleepy to think about this rn" reasons where I did respond to other parts, but I have set bookmarks for those so I can return to them :D
Not lipamanka but after reading the rest of this thread I think I find this analysis sensible:
-
lipu pu p25:
A verb is a word that expresses an action done to a noun. To form a sentence with a verb, follow this model: NOUN + li + VERB + e + NOUN
-
Pigeonhole it -- there are only about 20 verbs given lexical definitions in lipu pu. All of these are transitive by definition.
-
Intransitive presentations of true lexical verbs are similar to filling the objects with ijo, as per Polysemy testing - #16 by _at
As per Polysemy testing - #31 by magnap it may also be sensible to interpret what looks like an intransitive verb as an adjective filling the verb slot.
1 Like
lexically, in lipu pu, all words are either a verb (with a transitive meaning), an adjective, or a noun. There are also preverbs and prepositions. This matches up with usage. No word is lexically a verb while also having an intransitive base meaning.
the word "lexically" here is important though. You could analyze the word after li in a sentence that lacks e as an intransitive verb. I personally don't like this analysis but it works fine. This would be positional, not lexical. When I say that toki pona lacks lexical intransitive verbs, it means that there is no word with a base meaning that is an "intransitive verb."
This isn't directly quoted from lipu pu, but I do have the derivational patterns in toki pona memorized:
N -> V: applying the noun to something
N -> A: general "genetive" relation ("of"); having the qualities of N
V -> A: quality of relating to the activity, quality of performing the activity
V -> N: the transitive object of V, the act of V
A -> N: something with quality A
A -> V: making something have quality A
Sometimes you can stack these, but some of these higher level derivations don't work. For example, you can't do N -> V -> A -> N (example: kiwen becomes "turn something into a rock" becomes "related to the act of turning something into a rock" becomes "that which turns things into rocks." This one doesn't work because if one used kiwen to mean "someone who operates a freezer," people would misunderstand). Some do work though, like N -> A -> V -> N ("rock" -> "hard" -> "make something hard" -> "that which is made hard." This works because if you used "kiwen" to describe an ice cube, something that has been made hard, it is understandable).
Once I tried to go through and catalogue every single one of these that worked and every single one that didn't. Eventually it got really tiering. and I gave up, and I never published anything on it. But really with just the basics I've provided you can derive whatever the meaning "should" be and every time you'll get something that speakers use, because usage still matches up with the definitions in lipu pu.
Remember that some words have more than one lexical part of speech. The most common example is lukin, which is that was sees, not that which is seen. This isn't an exception, it just has a lexical noun definition that means "eye," so you can't derive "that which is seen" as you could normally do.
4 Likes
Replying to everything that @magnap has said to me
What? This is like never true in formal linguistics. You see "this has not been attested in usage" all the time in grammars. Descriptivists don't speculate about what could happen without saying "this is speculation and is not attested." At least they sure shouldn't! Generalizations cut corners that linguists often have the luxury to leave uncut and flesh out.
Yep! This is fair, but it's worth mentioning that we speakers have an intuition on what is right and wrong in the language we speak, and the more experienced we are the more likely we are to have a sense for what is right and wrong for the majority of speakers. lipu pu was @jan_Sonja's self-descriptive document, but it is also pretty close to how most people speak. Deviations from nasin pu are more likely to be small.
You're correct! It is bad analysis to speculate and call that speculation descriptivism. That doesn't mean it's prescriptive though; speculation is not inherently either. It just means it isn't descriptivist.
In my opinion the construction ended a while ago. I can't pinpoint when, but toki pona is definitely not a group project at the moment in the slightest. If you try to suggest changes, they will rarely if ever catch on, probably as often as changes in English catch on for people outside of your friend group.
Yeah sure! The "I am food" meaning is using "moku" as a derived noun in the predicate position. Because toki pona lacks a copula, the predicate can act as either a noun, and adjective, or a verb. In this case it's a noun.
For "I eat," moku is acting as a verb. Semantically, it has a patent (that which is eaten), but positionally in the sentence it does not have a transitive object. Derivationally, it is still in its base meaning, which is as a transitive verb.
Yes, this is ambiguous. People often use e to disambiguate these meanings: "mi moku e ijo" is unambiguously a transitive verb because if a lexical transitive verb is followed by an e, it must semantically be one as well. If it isn't, it could semantically be a transitive verb or a noun in the predicate position.
"e" is the particle that marks a transitive object. It also derives the predicate of the sentence into a transitive verb meaning if it didn't yet have one, but if it was already semantically a transitive verb, it just specifies what the transitive object was.
Nope! This is explicitly NOT how derivation works in toki pona. A few words are lexically both a verb and a noun, like lukin. If lukin were just lexically a transitive verb, its nominalized meaning would be "image" or "a vision," not "eye."
yep! "the desire is to sell" or something for "wile li esun," and "desired person" OR "person who desires" for "jan wile."
Huh! interesting! maybe! I don't actually know! The problem here is that what people call "intransitive verbs" are also often adjectives, which derive differently into verbs and can be derived into nouns. This would require more thinking though; I don't dislike this analysis!
1 Like
Also @tbodt let me know that the derivation system is the same in Esperanto! Go figure. Makes so much sense!