Date: Thu, 26 Feb 1998 21:34:16 -0500 (EST) Message-Id: <199802270234.VAA02181@locke.ccil.org> Reply-To: Chris Bogart Sender: Lojban list From: Chris Bogart Subject: Re: Meaning of BAI tags X-To: lojban To: John Cowan X-UIDL: 1316788a974bb635ac54b9e15cf56eb6 Status: O X-Mozilla-Status: 8011 X-From-Space-Date: Mon Mar 02 13:32:24 1998 X-From-Space-Address: - >Sentences like "I opened the door with my foot" and "I taught math >in Spanish" should clearly be false if I used my hand or spoke English, Absolutely -- I'm not advocating that {P bai S} be identical to {P}, just that it should imply {P}. {P} definitely should not imply {P bai S}! >but "open" has no tool place, and "teach" no language place, so we >must add then with BAIs. If we don't give the BAI places full status >as arguments to the predication just like the brivla places, there's >no other way to form these new predications except awkward tanru. I do advocate giving the bai's full status as places. In fact I think you should be able to drop *any* place in a true predication and have the result still be true. If you opened the door with your foot, it should follow that you opened the door. If you taught math in Spanish, it should follow that you taught math, and more generally, that you have taught. This useful for software that's looking through Lojban resumes trying to find math teachers to hire. If it sees "Lee taught Math in Spanish" and *can't* conclude that you have taught math, you'll be left out in favor of people who simply wrote "I taught math", without specifying a language or any other optional tag. But on the other hand, if the program is looking for people who have taught in Spanish, and it sees that "la kris pu ctuca" (Chris taught), it should NOT be able to conclude that "la kris pu ctuca bau la spano". (Chris taught in Spanish). I don't propose making BAIs irrelevant! co'o mi'e kris