[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Jordan: > On Thu, Dec 26, 2002 at 06:57:06PM -0000, And Rosta wrote: > > Jordan: > > > On Tue, Dec 24, 2002 at 08:01:13PM -0000, And Rosta wrote: > > > > Jordan: > > > > > I'll retract my statement about "lo'ei" being > > > > > poorly defined---it's certainly better defined than And's Unique > > > > > stuff at least, > > > > > > > > Which definition did you find poorly done, and which particular > > > > elements of the definition did you have trouble with? > > > > > > The definition for Unique makes sense, but it's not a very precise > > > way of looking at things. Under the way you define it, the exact > > > properties of the Unique-thing are completely up to the > speaker---basically > > > anything can be predicated of it > > > > > > So the problem wouldn't be the definition, so much as the meaning > > > > Surely not everything can be predicated of uniques -- of Jordan, > > of oxygen, of London, of homo sapiens, of magenta, of Tuesday, etc > > It seems to me that the properties of these are not completely up > > to the speaker, and that we don't have a problem with their meaning > > But my point is that the definition you've given of Unique doesn't > specify how it is determined what can be predicated. So, while > that maybe doesn't constitute a "bad" definition, it is not a precise > one either This is true. Sometimes definitions can be overspecified and are better left vague or given a prototype definition (so that membership of the category is determined by degree of resemblance to the prototype case), since vague or prototype definitions can be more useful than highly specified ones. In Lojban there are no grammatically-defined rules about what can be predicated of individuals, so this aspect of Unique is simply an instance of the way Lojban treats individuals in general. That is, Lojban doesn't specify how it is determined what can be predicated of individuals, and because Unique is an individual, it naturally follows that Lojban doesn't specify how it is determined what can be predicated of Unique. --And.