Date: Mon, 22 Dec 1997 22:34:39 -0500 (EST) Message-Id: <199712230334.WAA11451@locke.ccil.org> Reply-To: "=?iso-8859-1?Q?Jorge_J._Llamb=EDas?=" Sender: Lojban list From: "=?iso-8859-1?Q?Jorge_J._Llamb=EDas?=" Subject: Re: xor questions (was Re: indirect Qs (was Re: On logji lo X-To: lojban To: John Cowan X-Mozilla-Status: 0011 Content-Length: 1629 X-From-Space-Date: Mon Dec 22 22:34:41 1997 X-From-Space-Address: LOJBAN@CUVMB.CC.COLUMBIA.EDU la djan cusku di'e >> >whereas {lu'a ci girzu} is "a member of three groups", not one of the >> >three groups but a common member of the three. > >I think that "lu'a" is vacuous (not invalid) when placed before a non-set, >non-mass sumti. I thought {girzu} was a mass. Ok, change the example to {lu'a ci selcmima} or {lu'a ci gunma}. Those would be "a member of three sets" and "a component of three masses" respectively. >The trouble with the above interpretation is that then >we don't know what to make of "lu'a ci gerku": a common "member" >of three dogs? Right, pretty meaningless unless the context makes it clear what are the members of dogs. Out of context we don't know what {la'e lo gerku} is either. If the only relevant LAhE were the outermost, it wouldn't make sense for it to be grammatical to stack them. >Note that "lu'i" can be (somewhat) usefully iterated: > > lu'i lu'i ci gerku > A set whose sole member is a set whose members are three dogs. Right, and {lu'a lu'i lu'i ci gerku} would be a member of that set, i.e. a set whose members are three dogs. {lu'a lu'a lu'i lu'i ci gerku} would be [at least] one of the dogs. Lojbab: >> That was the other, later, use of lu'a - to allow grouped sumti to >> be labelled with a relative clause. The earliest use, though was defined >> for lu'i alone, and was specifically to allow selection of a number >> of members from a set, such as the "Would you like coffee, tea, milk, or > water?" It would be interesting to see that question in Lojban. There we have the additional problem of the intentionality of djica, of course. co'o mi'e xorxes