[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
One of the early examples (from xorxes) that brought the lo'e debate to life again was "This depicts a snake". In subsequent discussion we have rightly distinguished (i) the unique snake, and (ii) various sorts of representative imaginary snakes (archetypal, modal average, etc. etc.). Originally, in discussing "This depicts a snake", I had argued that "a snake" is a Unique. I want to point out in this message that that analysis still seems right to me, that analysing "a snake" as a Representative Imaginary leads to some problems, and that therefore we all need Uniques (i.e. we need them not just because me & xorxes & xod like the idea of them). The problems that analysing "a snake" as a Representative Imaginary leads to are these: There is an ambiguity between the following readings: "Being depicted by this is a Typical property of the Representative Imaginary snake." "In this picture we see what the Representative Imaginary snake would look like if it were visible." Obviously the second reading is more like what is wanted than the first, but the ambiguity is one that has to be got rid of, perhaps by treating it as a matter of scope (as I think Jordan wanted), so that the first reading belongs to 'lo'e' since ti se pixra and the second to ti pixra lo'e since Anyway, assuming we can get the second reading, it is still not quite right. On a reading (of "this depicts a snake") where no particular snake is depicted, it is still not necessarily the case that the snake is *representative* of snakedom. For instance, if the representative imaginary snake is green, the snake in the picture can be red. So there is a difference between plain "this depicts a snake" and "this depicts the archetypal snake". To summarize: * We *do* need Unique. * Representative Imaginaries involve as-yet-unsolved scope-like problems. --And.