[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
Jorge Llamb�as, On 19/09/2012 01:02:
On Mon, Sep 17, 2012 at 11:14 PM, And Rosta<and.rosta@hidden.email> wrote:Jorge Llamb�as, On 18/09/2012 03:01:le fe la q boris qa mrsbnxa li lkni fi la va dtca lo vo vndo ju plnako gsnake As for Boris' death, the likely thing is that the Germans rather than whoever, poison rather than whatever, using it they caused it. Though I'm not sure whether the two v's are evaluated independently or one under the other.le fe la q boris qa mrsbnxa li lkni fi la vi dtca lo vi vndo ju plnako gsnake fi binds vi (ni'uki). I think it means that Germans using poison to kill Boris is more likely than nonGermans using nonpoison to kill Boris.I'm still trying to grok v-/ni'uk- I'm not sure what you mean by "fi binds vi" because f- is not a binder, it doesn't close a formula wrt a variable.
I'd meant there is a long-distance dependency between fi and vi, expressed inflectionally by sameness of vowel. I would treat this as an instance of the syntactic binding relation, but not of the semantic binding relation.
Fi itself would be bound by a higher quantifier.
The way I was thinking vV/niukV works is that in:
Qa [Xa je Y va Z] bcda
the phrase headed by Xa, the binder of va, expresses the pair {"Xa je Y ni Z", "Xa je Y nu Z"}, and predicates bcd of this pair; and X- needn't be f-. But on reflection it might be better to say that it is Qa that (syntactically) binds va, and that
Qa [... va ...] bcda
predicates bcd of {[... ni ...], [... nu ...]}.
In "fi la vi dtca lo vi vndo ju plnako gsnake" the variable "i" is still free. Or maybe fi does bind the inner i, and simultaneoulsy introduces a new free i for li to bind? But in:la je ckfa va mlka prfraka'a I prefer my coffee milky (to not milky)there's no fa to bind va.
Right. The reformulated rule now allows this.
The way I read the vo/va version is: lo vo vndo ju plnako gsnake Poison rather than whatever was used by A in doing E.
I think this requires an extra intepretation convention that when a predicate is, like pln, not the sort that expects a pair of formulas as its argument, the interpretation is this "rather-than" one. But okay, let's provisionally suppose there to be such an extra interpretation convention, unless we discover it causes problems.
la va dtca lo vo vndo ju plnako gsnake Germans rather than whoever used poison rather than whatever in doing E. li lkni fi la va dtca lo vo vndo ju plnako gsnake The likely situation is one in which Germans rather than whoever used poison rather than whatever in doing E.
Okay.
But since you get approximately the same reading with vi/vi instead of va/vo, I'm puzzled (again) about what the variable that goes with v- does.
"le fe la q boris qa mrsbnxa li lkni fi la vi dtca lo vi vndo ju plnako gsnake"
means that lkn is predicated of {[fi la ni dtca lo ni vndo ju plnako gsnake], [fi la nu dtca lo nu vndo ju plnako gsnake]} (not quite the right meaning because it doesn't treat the germans and poison as separate likelihoods).
I suppose v- is a focus marker, and the reason you need the variable is because it's possible there may be more than one thing fighting for something focused, so "lkni" could be competing with something else for the focus and the -i ties it to lkni. So I propose that "ni'u" be the focus operator, and ni'ukV the marked focus that you may need in some extreme cases.
No, v/niuk isn't (in essence) a focus marker (AFAICS); it marks the variant portion of a pair of otherwise alike formulas, and the reason it has to be 'bound' is that the binder marks the start of the formula. It's true that what it marks will indeed tend to be in focus, tho.
For example, for: lo ni'u vndo ju plnako gsnake "Poison rather than whatever was used by A in doing E." we can't really use a variable because the thing that needs it is hidden in the implicit illocutionary operator: [ca lu xsra'aku fu] lo ni'u[ku] vndo ju plnako gsnake "[Hereby my assertion is that] it is poison rather than whatever that was used by A in doing E" la je frmra si xsli je pnsaki ni'u drxaki msta "Farmers that own some donkey and DO beat it are most." (most, that is, with respect to farmers that own some donkey whether they beat it or not). You could use "ni'uka" there for more precision, for "msta" to claim that focus, but I think it's unnecessary because there's nothing else competing for it. Does that sound half right?
I can see how you get from ni'u to the interpretations. I'm not sure that we need a focus marker or that the best way of getting one is a unary operator, but I have no stronger objection to it than that. And yes, for precision the niukV version is needed. --And.