sona pi pali moku

[janMato]

jan ni li pana sona e ni: jan ante li ken pali e moku kepeken ilo lili.

http://thestonesoup.com/blog/
"stone soup, minimalist home cooking"
ni li telo moku kepeken kiwen. ni li nasin pali moku tawa tomo mute sina kepeken ilo lili.

[janKipo]

I would have said 'nasin pi pali moku' but I'm not sure what the difference would be.

[janMato]

nasin pali moku tawa vs nasin pi pali moku

They're different, imho, there is a subtle difference between pali (working-- descriptive of an activity, a verb-y sense) and pi pali (work-- a noun-y sense) and with pi being just a universal relationship and modification itself being something of a universal relationship, there isn't any mechanism to pin down the meaning further.

[janKipo]

I have more or less the opposite sense, 'nasin pi pali moku' is "a way of making food" 'nasin pali' is, to me, a procedure, a way to do something. Thus 'nasin pali moku' is a foody way to do something, but not necessarily make food (I imagine a tax accountant working on a cluttered desk with burger wrappers and mo shu pork boxes all over everything). Of course there are dozens of other readings for both, these are just (to me again) the most likely in a contextless case.

[janMato]

I see your point. But universal relationships are so broad that they're impractical in the long run. I think that is why pidgins evolve to either deal with the problem lexically (an idiom requiring memorization) or by introducing a new grammatical device to nail down the meaning.

nasin moku
road (that one can buy) food (at) -- the outdoor farmer's market
road (made of) food -- a cake decorated to look like asphalt.
technique (for making) food- cooking

pali moku
work (to prepare) food
work (accomplished using) food
work (that happens in the presence of) food
work (enabled by) food

[janKipo]

Well, the tp motto is "Context will decide" and, while I agree that all the things you list are possible readings, only a couple seem at all likely context-free. I (and you) could fadge any number of contexts that would make some of the others look likely or even obviously correct, And we could each surely come up with several more reading, each into an ever more odd context. Trying to decide what an expression means without context is at best a guessing game involving assuming some context, usually the "nothing special" one. And, of course, insisting on one same reading of an expression in every context is equally useless (why common expressions are not quite idioms). I suppose the point here ought always be, does it communicate the right message. If so, then it was good tp, whatever someone may want say it "really " means; if not, it is bad tp, even if its "real" meaning is just right. But, of course, there are covertly statistical reasons for thinking that one expression is more likely than another to convey the right message in a given context.