From Pycyn@aol.com Sat Aug 28 07:43:17 1999 X-Digest-Num: 224 Message-ID: <44114.224.1238.959273825@eGroups.com> Date: Sat, 28 Aug 1999 10:43:17 EDT From: Pycyn@aol.com Subject: Anselm summary Thank you for all your cards and letters,... Let's see. Since Anselm's point in introducing the description (of God, by the way) is to prove that a being satisfying that description exists, the lo and da poi versions have to be excluded immediately, since they contain implicitly the conclusion of the argument in the premise (circular reasoning of a rather obvious sort -- and Anselm ain't obvious). That seems to leave le, which has the added advantage of showing up one of the common attacks on Anselm's argument, that "the F is not an F" is not a contradiction, as Anselm would have it (and, indeed, is a proof that there is no F). On the other hand, it is still not possible to do the complicated buried referent with the gadro. So, that leads to da voi, which makes a lot of sense historically. Anselm is in fact proposing this description as his way of talking about God (inspired by the Holy Spirit, he says). And he wrote decent Latin, so he does not actually use articles, though he does switch between "something" and "it" in a way that pretty closely parallels the English "a/the", introduction/continuation, pattern. Of course, he is convinced that his description is veridical, but voi does not deny that -- it merely leaves it up to the situation to determine. And Anselm clearly thinks that Guanilo's almost identically formulated description, of a most perfect island, is not veridical (the situation is somewhat more complicated than that sounds like, but that is close enough for present purposes). Speaking of Guanilo, Anselm's best response to him (not the one he is usually credited with making) is that while an island -- or any other specific kind of thing -- than which none greater can be conceived is not coherent, for we can always thing of ways of improving a particular sort of thing, an unqualified that than which ... does not allow this sort of improvement beyond a limit. Thus, while Xorxes' first suggestion helps a lot, the possibility that even broda counts as a specific sort of thing may be enough to sink it in Anselm's eyes. And, of course, the poi would have to be voi to avoid immediate circularity. Anselm wants his none-greater to be unique of course, but another attack on his argument is to note that he has not eliminated the case of two (or twenty) equally great being at the apex, none conceivably surpassing any other. So, the pa and the uniqueness of logical "the" are not entirely to the point either, though good to think about. I do not see that ko'a voi offers any advantages over da voi at this point, though it clearly is permitted without a previous (or later) goi. Before somebody else mentions it, I note that the da format does commit me (and Anselm) to there being something called that than which ..., even if it is not .... But the crucial point is whether anything is ... and if that so-called thing is not, it does not matter what it is otherwise. So, in particular, we have not smuggled God in in the variable, Anselm's idea of God (which clearly does exist -- even in the mind of the Fool who says that God does not) will do fine for the purpose. The whole discussion here has shown, however, that Aquinas had good reasons for suspecting that Anselm's proof failed from well-concealed circularity, presupposing the existence of what he claimed to prove. Thanks for helping clarify this for me; I'll let you know how it finally comes out (well, I'll let you read a trat of Proslogion 2, which will not -- this list being what it is -- ever be final). By the way, I am now persuaded that Lojban etc. does not conflate specificity and veridicality. The distinction is merely specific/non-specific. Non-specific must be veridical else there is no hook for identifying an appropriate object at all (Note &'s list does not contain non-veridical non-specific cases). Specific, on the other hand, works on the principle that, since the object is identified, it does not matter what you call it, though pragmatic (again -- that is what forces the issue for non-specific) recommends something that will help the ignorant to find the right thing, thus either veridical or pretty transparent in context. Note, le and voi do not REQUIRE non-veridicality and probably would not usually use it, Mike Hammer cases aside. But, as Lojbab notes, I have argued so many sides of this over the years that I am not sure where the world is any more on it. pc