Re: TECH: only, {me} place structure

[John Cowan <cowan@SNARK.THYRSUS.COM>]

la lojbab. cusku di'e

> Seeing your posting on this to Lojban, I read it a bit more carefully and saw
> that you indeed had called upon "srana".  But the "me" place structure as is
> IS basically the one you propose, but with the shortening that "aspect"
> which should be a "ka/property abstract" should actually give the 2nd place
> of "srana" rather than repeat the srana.

You said this to me on the phone, but looking at the actual place structure
of "srana" it doesn't hold up.  The latest gismu list shows:

        x1 pertains to/is relevant to/concerns/is related/associated with x2

So in fact the x2 place of srana corresponds to the sumti following the "me",
not to any place of "me"; "me la nuIORK." means "srana be la nuIORK.".

So what >is< the x2 place of "me" supposed to be?  I suggest that it is a
property, a "ka" abstraction.  In fact, looking over the related gismu,
"me" is most like "ckini".

> I suspect that "in aspect "leka du"
> would still work with the existing place structure, but I repeat my argument
> based on Zipf in my private message to you:  "le du be [sumti]" is
> shorter and clearer than "me [sumti] be leka du", and likewise for the
> "srana" version of your proposal, whereas the existing place structure for
> "me" does give a slight Zipfean advantage over using "srana"
> "srana be [sumti] bei [leka aspect]" vs "me [sumti] <me'u> [leka aspect],
> though it gives no advantage over "du" (which could be the selbri in the
> main-selbri version of "me").

This sounds to me like you actually agree with my main point:  the x2 place
is a property.  I am not arguing for a change in the place structure,
merely a clarification of what the x2 place is supposed to do.  The result
is that when the x2 place is omitted, as it typically is, "me" is vague
about the relation specified.  This is essentially what pc said back in
TL4:3 pp. 20-21 (language updated to Lojban):

        [A]ny sumti can give rise to a selbri: a sign for the
        property of being specifically related to the thing named
        by the sumti.  The device for this is the cmavo {me}
        attached directly to the front of the first word of the
        sumti (usually, but not necessarily, a descriptor):
        {mela djan.}, {me le ninmu}, {me ko'a}.

        It is not possible to specify, in isolation, the exact
        property to which such predicates refer, or the exact
        relation to the named object.  These must be deciphered
        from the particular context....

By assigning {me} an x2 place, the specification of the desired property
becomes possible: it may be identity, or associated-ness, or whatever.

> Thus I favor the status quo.

So do I, but with a clarification that with the x2 place ellipsized,
the property that belongs there must (as always) be glorked from context.

--
John Cowan      cowan@snark.thyrsus.com         ...!uunet!lock60!snark!cowan
                        e'osai ko sarji la lojban.