Message-Id: <199309290705.AA12078@eli.CS.YALE.EDU>
Date:         Wed, 29 Sep 1993 03:03:11 -0400
Reply-To: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Sender: Lojban list <LOJBAN%CUVMB.bitnet@YaleVM.YCC.YALE.EDU>
From: Logical Language Group <lojbab@ACCESS.DIGEX.NET>
Subject:      Re: TECH: input on gismu place structures wanted - sumti raising?
To: Erik Rauch <rauch>
Status: RO

Nick writes:
NS> To Logical Language Group respond I thus:
NS>
NS> #sarcu - JCB's 1975 equivalent appears to have a du'u (neither nu nor
NS> #object) defined for x1 (though he had no real way to express a du'u).
NS> #It appears in any case that x1 is a sumti raising, but should the
NS> type #of abstraction be limited to a du'u?  (I am coming to think of a
NS> du'u #abstract as a second order abstract:  ledu'u broda = lenu lenu
NS> broda cu #fatci - with the implication that any other abstract is a
NS> sumti raising #from a du'u place)
NS>
NS> I doin't see why x1 should be just a {du'u}; I also think your
NS> thinking on the second order abstract is correct.

If du'u abstractions are 2nd order, then it seems to me that having them
alternate with 1st order abstractions in the same place is roughly the
same as allowing abstraction and concrete in the same place.  If you say
that "food is necessary for life"  you could be saying the less-raised
"the event of consuming food" is necessary for life', or the 2nd order
'the fact that "... consumes food" is true' is necessary for "the state of
... is alive".  Since the x2 of sarcu is an abstract, any attempt to pull
"..." is pretty tricky and either one or two levels of sumti raising, but then
the food is also a 2nd level sumti raising to put IT in x1.

I don't feel strongly about this, but multi-levels of sumti raising makes
me have multi-levels of semantic queasiness about the langauge and our
analysis of it %^)

lojbab