Re: Cowan's sum opaque

[Gerald Koenig <jlk@NETCOM.COM>]

djan quotes me, djer:
>la djer. cusku di'e
>
>> John and others seem to agree that all the meaning  in the English "any"
>> can be captured by a universal quantifier or an attitude marker.
>
>Not at all.  Sometimes "any" is existential, not universal.
>
>> I disagree.  Consider this meaning from my Webster's:
>>
>> "1: one or some indiscriminately of whatever kind:
>> 1a: one or another taken at random <ask ~man you meet>."
>>
>> There are two anys here. One taken indiscriminately or some taken
>> indiscriminately. I want to consider the case of one taken
>> indiscriminately. It certainly cannot be expressed as "all". Neither is
>> it an just an attitude.  We're talking about quantification here, namely
>> one something.
>
>Note, however, that the Webster example is an imperative!  You wouldn't say
>in English "I asked any man I met."  You would say "I asked a man that I met"
>(existential) or else "I asked every man that I met" (universal).

djer:

        Trying to preserve the dictionary sense of "any" as: one taken
        randomly out of at least two, in a past tense; I would *not say,
        "I asked a man that I met" because this doesn't express the act
        of random choice which is built in to this definition. The man
        in your sentence could be Adam--well, at least the only man in
        the universe of discourse. So no choice would be possible.
        I would say "I asked someone at random that I met" to preserve
        this meaning of "any" in a past tense.
        Webster:
                someone: some person   [note singular jlk]
                some: 1: one indeterminate quantity, part, or number as
                distinguished from the rest.

        There is no "rest" in your version.

        Although xorxes has said that the meaning of my "any" is close
        to his xe'e I want to use xe'o to clearly express: the random
        selection of exactly one thing from a set of at least two.
        xe'o then is neither an E(x) [there exists at least one] nor an
        All(x) [for all x].  Our troubles begin when we try to make
        "any" mean one or the other of these in order to shoehorn them
        into FOL and thence to lojban.  That is why I advocate coining
        this new word, xe'o. New in that it has no exact equivalent in
        any language because it represents one and only one of the
        m-any faces of the English "any". And "cunpa lo su'o re" does
        the same thing, if user friendliness is no consideration.

>The need for "any"s comes up when we have some kind of opaque context,
>including an imperative; invariably (I claim) this involves a subordinated
>abstraction clause.

djer:
        I think "any" causes opacity by not ever pointing to a specific
        object, concrete or abstract. It's like trying to know the
        winner of the California lottery before the draw. Since the
        referent cannot be known until after the fact of selection, and
        then only if expressed, you are free to claim that it
        invariably involves a subordinate abstraction clause. Or at
        least until the referent is identified later.

        On this interpretion, "I saw someone playing pool" is also
        referentially opaque until the "someone" is explicitly
        identified. It claims only that one person was seen as
        distinguished from the rest. It doesn't claim to identify that
        person. The difference is the one that is so crucial in court.
        You can see I've changed my views 180 deg. concerning pycn's
        example.
        This is how I see it until convinced otherwise.

        djer

>John Cowan              sharing account <lojbab@access.digex.net> for now
>                e'osai ko sarji la lojban.
I did.