I'm editing a flash card deck of lojban words for my own use, and one of the things that's struck me is how often gismu constrain the kinds of things in a position.
Most of the time the constraint makes sense (at least in a "sure, I can see it" kind of way) but as I find more gismu where a position is defined as being a person or an event (and all the ones that are {ka}) when the other way might make sense, I begin to wonder more and more why (or if?) the suggested kinds for a place aren't merely a guidance.
For example, I'm just looking at
jenca: x1 (event) shocks/stuns x2
What would be the problem with "do jenca mi"? I suppose there's some elided abstraction that I could use in the x1 place to mean "some event related to you" - but why would that be neccesary?
Or do I misunderstand - is the point that jenca describes a shock related to an event, and *therefore* "do jenca" implies that you did something shocking on the face of it?
If it's the latter, are the other abstraction requirements (and other implied types) likewise looser than I imagine?
Judson