When starting out, it's common to think about the meaning of noun phrases and the best noun phrase to translate a concept. Is a phone an "ilo nanpa lili"? An "ilo sona leko"? Something else? This can be a good way to explore the semantic spaces of words, but it often doesn't help much with actually communicating. Why? Because noun phrases in toki pona are inherently vague! I'll explain what this means and then talk about how often, the solution is grammar. # What do I mean by "noun phrases are vague"? What is vague about a noun phrase like "nasin wawa" or "telo sewi" or "jan pi olin suli"? Let's consider what information a noun phrase gives us: > A noun phrase declares that *some* relationship exists between the headnoun and the modifier. Teachers in kama sona will often describe an X Y as an "X related to Y (in some way)". It turns out this is not very much information. Related *how*?! Consider an "ilo toki": - Is the ilo related to toki because **it is used to communicate**, like say, a phone or TTD? Maybe! - Is the ilo related to toki because **the ilo itself communicates**, like, say, C3PO or Mosscap? Potentially! - Or maybe the relationship between "ilo" and "toki" is one of **possession, location, appearance, or description?** Or some other way that two concepts can be related, or that one concept can describe or modify or specify another? To be clear, I don't think that there is something "wrong" with this vagueness or that it's "bad", and I wouldn't exactly call this a "flaw" of toki pona. It's kind of just a consequence of having no morphology to specify these relationships inline, and of the fact that toki pona is a very sentence-oriented language that doesn't like to use noun phrases to convey large amounts of information (unlike, say, English, where I can use phrases like "sentence-oriented language" a a a). # Using sentences instead of vague noun phrases A commonly taught-to-beginners solution to this is "context!" I suppose I agree, but in the following sense: usually, existing context isn't enough to specify the relationship, and you need to create context using grammar. Where noun phrases specify the existence of a relationship but not its nature, toki pona's grammar (particles, prepositions, preverbs, etc.) can concisely specify the nature of the relationship between two words. Here's an example. Let's say I'm at a toki pona meetup and we're going for a really nice hike in the woods. It's windy and the trees are leaning and bending in the breeze. I turn to the group and say: "a! mi pona pilin tan tawa kasi" Do I feel good because of our tawa kasi - our walk in the woods? Or, do I feel good because I'm looking at the beautiful tawa kasi - the movement of the trees as they bend and their leaves flutter in the wind? Unclear! I used a noun phrase and it was too vague to specify. No worries, though - a little grammar can help. I might instead say: "a! mi pona pilin tan ni: mi tawa lon kasi." or... "a! mi pona pilin tan ni: kasi li tawa" Now everyone can easily tell the difference. Why? Because the grammar removes ambiguity by specifying the **nature of the relationship** between tawa and kasi! In the first case, the tawa we are taking is *located in* the kasi via the *prepositional phrase* "lon kasi". In the second case, the kasi are *the subject*, and the *thing they are doing* is tawa (as specified in the predicate). Play around with this and see if it helps you to think and express yourself more easily in toki pona! Here are some other phrases that came to mind while I was brainstorming ideas for this post. For each of them, think about how it's the *nature of the relationship* between the headnoun and the modifier that needs to be clarified. - is this "moku soweli" food *for* animals or *made out of* animals? - is this "ilo kiwen" a tool *made of* metal or a tool *for working* metal? - is this "kalama musi" a sound made *because* someone is amused or a sound made *in order to amuse* others? --- # Appendix Two side notes that I didn't elaborate on here but that seem worth separate discussion, perhaps in another post: 1. It occurs to me that this same phenomenon might help explain why multiple-pi phrases are hard to understand: even though you are adding context (i.e., by being more specific about your modifiers by modifying them), you're also **adding** ambiguity since you still can't specify the relationship between the modifiers and headnoun or between the modifier-modifiers and the modifier-headnouns, so your listener has **more** variables to hold onto at once. 2. This type of vagueness can exist in other languages too. One strategy that non-minimalist languages use to solve this is lexicalization - as an English speaker, I have memorized the fact that birdsong is sung *by* birds, not *about* birds, but a love song is sung *about* love. Another strategy is morphology. I know that a "fire fighter" is a person who fights fires and a "jazz singer" is a person who sings jazz because the suffix "-er" indicates an agent who performs an action. Eschewing lexicalization and lacking any morphology to add such meanings to noun phrases, toki pona tends to rely on sentences and grammar instead to clarify such extra information.)