[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
| In a message dated 10/11/2002 7:36:23 AM Central Daylight Time, araizen@hidden.email writes: << . But that does not make a property and its degree of >> I wonder if this is true. And, worse, I wonder if the move is coherent: can {ka ce'u broda} for example refer to both a property and a measuring rod -- unless the property just is the measuring rod, which is possible, I suppose, as variant on Zadehsing measures rather than truth values. But, in that case, I would keep {ni} for this notion of a property and use {du'u} for the more familiar one. << The vast majority of actual uses of properties are 'bound' in this sense, but there has been some use & experimentation of 'unbound' properties (e.g. 'mi nelci le ka ce'u prami', 'I like Love/Loving', since 'nelci' certainly doesn't require any abstraction in x2, this is an 'unbound' use.) >> I have trouble with the example, but once I get over that I have no trouble with {mi nelci le ni ce'u prami}, meaning something quite different, though equally strange (or maybe not: "I like this scale for evaluation the degree of loving"). So, the {ka}-{ni} distinction seems to be needed after all. << I use klani as 'x1 is something quantifiable, and its quantity is x2'. I haven't though about how to use the x3 of 'klani'. >> One of the presuppositions of {ni} is that everything is quantifiable and so the first part is superfluous. x3 is just ni ce'u [bridi] with the [bridi] supplied by x1 somehow ("the obvious way"). on xod: << > [0, 1] contains an infinite number of reals. Inside that interval can be > mapped the unique vrude-ness of every atom in the universe. I don't see > why restriction to [0, 1] should give one a sense of limitation. Because if you ever assign the truth value/quantification of 1 to a given bridi, there can never be something higher (more true/more), but I can certainly conceive of quantities which are open-ended and can always be increased no matter how large they are. >> xod's solution, I think, is to never assign 1 (nor 0, probably, though bottom limits seem easier to come by:0 Kelvin for cold, 0.00 inches for height, and so on). The mapping is not proportional, of course, but all values get in, even if second order comparisons don't always work: " Kareem is more taller than Abdul than Abdul is Haroon." |