[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
xorxes:
> la nitcion cusku di'e
>
> >When you say {lo merko}, you
> >claim some objective knowledge of the membership of {American}. The
> >same should hold when you say {lo'e merko}: one should not be
> >intrinsically more subjective than the other
>
> To me, {lo'e merko} is not at all about the membership, i.e. the
> extension, of {lo'i merko}. It is about its intension
Its intension conceived of as an archetypal member -- a prototype-
theoretic intension. As opposed to an intension conceived of as a set
of membership criteria, which is expressed by ka.
To me -- and I hesitate to claim this as true of Lojban rather than
merely my intuitions about it -- these two sorts of intension are
rather different. The ka-type is defined axiomatically, while the
lo'e-type is defined inductively from its extension, at least
potentially, which is the source of its typicality effects.
The two major theories of categorization are thus catered for, the
one by ka and the other by lo'e.
> >There's an out here: you can say stuff about {lo merko} without needing
> >to know every single American, because you use the sense of the word
> >rather than its denotation --- the definition, rather than the
> >enumeration ("An American is defined as..."). But I currently don't
> >think this extends to {lo'e}. I think lo'e is meaningless without a
> >notion of surveying: a reasonable sense of what the actual population
> >does, not just a definition
>
> I think that's the key misunderstanding of {lo'e}. It is NOT about
> the enumeration. No averages or modes. It is the sense and only
> the sense that matters, much more than with {lo}. You can even
> squint {lo'e broda} into existence when {lo'i broda} is empty
It is because the lo'e-type intension is derived inductively from
the extension that the other guys are getting this idea of it
being like a statistical mode. And I think that prototype theory
would accept that the statistical mode is an important factor
in defining prototypes, though it is not the only factor nor an
overriding one.
In other words, Nick could respond to you by asking how we know
what the sense is like. And I think you would be bound to answer
that partly our mental representation of the sense is based
on the statistical mode of the category members from our own
experience. But of course this does not mean that a claim about
a prototype is a claim about a statistical mode.
--And.