From: cbmvax!uunet!math.ucla.edu!jimc
Message-Id: <9104221532.AA04604@luna.math.ucla.edu>
To: lojban-list@snark.thyrsus.com
Cc: Arthur Hyun <ash@rpi.edu>
Subject: Re: Your first sentence 
In-Reply-To: Your message of "Fri, 19 Apr 91 10:11:36 EDT."
             <9104191412.AA05304@its.rpi.edu> 
Date: Mon, 22 Apr 91 08:32:12 -0700
Status: RO

> Date:  Fri, 19 Apr 91 10:11:36 -0400
> To:  lojban-list@snark.thyrsus.com
> From:  Arthur Hyun <ash@rpi.edu>
> Subject:  Re: Your first sentence 
   
> * * *
> > dei pamoi le'i mi lojbo jufra
> > This-utterance is-first-in the-set-of my Lojbanic sentences.
> 
> "lojbo jufra" here could mean "sentences IN lojban" as well as
> "sentences ABOUT lojban" could it not?  Furthermore, could not
> "sentences ABOUT lojban" refer to either the language, or the
> "culture"?

That's the "beauty" of tanru; the relation is semi-explicitly
ambiguous. I prefer to avoid the concept of metaphor except when I
really need it, and to use what I call "lawful compounds" instead.  I'm
working on something I can post to the list about that.  In summary, if
you look at the tanru in Old Loglan you find that nearly 90% follow
regular rules for interpreting them, and with minor fiddling with place
structure, even more can be made lawful.  

> This may seem like a nit, but I'd like to ask nevetheless.  Is
> "le'i" necessary here?  I find in my cmavo lists that the place
> structre for "pamoi" includes a place: "among set/list/group x[2]".
> Thus: "dei pamoi le mi lojbo jufra".

pamoi X2 requires a set.  le'i causes the sumti to be interpreted as a
set.  "lemi lojbo jufra" means the (specific) sentence I have in mind,
and the referent of dei would presumably be the first part (word?) of
this sentence considered as an ordered set.  Not what you want.  By the
way, I think lo'i (set of referents actually fitting the bridi (mulbri)
of "my Lojban sentences") better produces the intended sumti referent set. 

Actually in -gua!spi I can recognize the start of a sumti even if it
lacks an article, and so each case of a bridi comes with a default
article (usually the equivalent of le).  In this case it would be (the
equivalent of) le'i.  This is not possible in Lojban because the
article is mandatory to cue the start of the sumti.

>   An argument:  If one should use "le'i" to make explicit the second
>   place as a set, then why define place structures?  Would it not also
>   be best to make *every* place explicit as such?  * * *
>   Again, perhaps it would be better to make the place
>   explicit with a "preposition"?

Think of motion words where the difference between arguments is
semantic but where there is no set/extension contrast.  Here the
article (le, le'i, etc.) can't do double duty as a caselink.  Actually
in Institute Loglan there was a short-lived proposal to make modal
operators (sumti tcita) mandatory on all sumti, replacing both
numerical caselinks and the whole concept of fixed ordered cases.  It's
not unreasonable -- just too verbose (for my taste).  

> Furthermore, I note that "pamoi" has an x3 place: "in property x[3]".
> Can we not say, then, that:
>   "This utterance is first among the set of my sentences such that they 
>    regard lojban", or...

This isn't authoritative, but I believe pamoi X3 specifies the sort
order for the list.  E.g. a restaurant menu ordered by price,
contrasted to ordered by how much I like the dish, or alphabetically. 
This is one way to deliver a "best" or "most extreme" meaning.

James F. Carter        (213) 825-2897
UCLA-Mathnet;  6221 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA  90024-1555
Internet: jimc@math.ucla.edu            BITNET: jimc%math.ucla.edu@INTERBIT
UUCP:...!{ucsd,ames,ncar,gatech,purdue,rutgers,decvax,uunet}!math.ucla.edu!jimc