Date: Thu, 25 Sep 1997 01:45:00 -0500 (EST) Thu, 25 Sep 1997 11:03:41 +1000 (EST) Date: Thu, 25 Sep 1997 11:03:41 +1000 (EST) From: HACKER G N To: John Cowan Cc: Lojban List Subject: Re: na`e In-Reply-To: <0EH000EL2YVEKQ@newcastle.edu.au> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mozilla-Status: 0015 Content-Length: 1298 X-From-Space-Date: Thu Sep 25 01:45:01 1997 X-From-Space-Address: drv.cbc.com!c9709244@cbgate.cbc.com On Wed, 24 Sep 1997, John Cowan wrote: > > "cei" doubles up the functions of "goi" and "poi". When applied > to an assignable pro-bridi, it assigns it; when applied to an > existential pro-bridi, it restricts it. This isn't explicitly > stated in the book because I was leery of saying too much about > second-order quantification when my understanding of it is > quite shaky. > > But anyway, relative clauses can only be applied to sumti, and > while "su'o bu'a" is technically a sumti, in the prenex > (by special exception) it is functioning as a quantifier + > pro-bridi. So the true grouping is > > su'o (bu'a cei (na vreta)) zo'u ... > ` For-some (relationships which are (not reclining)) ... > > rather than > > (su'o bu'a) (poi na vreta) zo'u ... > Speaking-of-(some-things which-satisfy "bu'a") > (which do not recline) Oh! That's a real twist on what's written in the grammar, I think. I would perhaps have been a bit more comfortable with "su'o nu bu'a", "su'o su'u bu'a" or some other such abstraction to express a predicate relation in a prenex, because it seems more consistent with the grammar elsewhere, but I'm not overly fussed about it, because I don't consider myself likely to use these kinds of constructions. Geoff