[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
[An expansion of what I've just written in another message, something
I have been thinking for the last day or two.]
The discussion of lo'e and genericity and also the discussion
about gadri choice when dealing with singleton categories leads
me to feel that we are converging on two series of gadri.
I'm not trying to corral people into a decision they don't
want; I'm just trying to shift the focus of debate from discussing
meanings to making decisions about actual cmavo.
* One series, 'AVERAGE', yields a *representative* and *imaginary*
member of lV'i, derived by averaging and squinting.
* The other series, 'UNIQUE', simply presupposes that lV'i is
singleton.
The rationale for Unique is:
* It lets the speaker decide where to draw the boundary between
individuals and multiple instantiations of categories, when imposing
the distinction upon the world. (I've discussed this in lots of
other messages, & xod too, so I won't say more here.)
* It solves the 'problem' of gadri choice when dealing with
singletons. (I recognize that not everybody feels it to be a
problem.)
The rationale for Average is:
* It *encodes* the notion of representativeness.
* It approximates to the majority understanding.
There has been disagreement about different sorts of Average or
representativeness -- mean, median, mode, prototypicality. I don't
think there should be different gadri for different sorts of
Average, nor {pe} phrases to specify. Rather, the gadri should
simply presuppose that the referent is arrived at by some process
of averaging and squinting such that the outcome is representative
of the set in some relevant way. If one wants to be more precise
about the exact kind of averaging involved, then one can use
a lujvo whose x2 is lo'i or le'i; the x1 gives the product of
the averaging.
I would like to make two further proposals (I want them to be
taken jointly, not separately):
* cmavo for Average are lo'e & le'e
* cmavo for Unique are loi'e & lei'e and are made official.
If the consensus was against making loi'e and lei'e official,
then I would argue that lo'e/le'e should then be assigned to
Unique, because the ground covered by Average is contained
within the ground covered by Unique.
--And.