From BestATN@xxx.xxx Sun Apr 4 10:40:50 1999 X-Digest-Num: 104 Message-ID: <44114.104.575.959273824@eGroups.com> Date: Sun, 4 Apr 1999 13:40:50 EDT From: BestATN@xxx.xxx Subject: semantics I like the idea of defining gismu in terms of semantic primitives. Such definitions should make it possible for machines to understand them and better interact with people. Someone already mentioned defining common gismu first, then using those gismu in later definitions. Seems like a good idea. How about a Lojban dictionary in Lojban? There are lots of people on this list, and if each one wrote a definition for a single gismu, a lot of definitions would get written, especially if a format were chosen beforehand to adhere to. Has anyone ever written *any* definition in Lojban? How did it turn out?