| In a message dated 10/4/2002 10:40:40 PM Central Daylight Time, jjllambias@hotmail.com writes: << If I was forced to accept your saying so, definitely. >> Yet again: they are rarely used and when needed can be perfectly mimiced using the importing sort (useless). If they are taken as basic, it is impossible to get the importing sort in, even though those are the ones we generally need (screwed up). << >But, to the point, once you have >them, it is virtually impossible to get importing expressions, which >presumably -- since they cover 99+% of the cases, you do want. Non-importing expressions cover 100% of the cases, so I don't need the importing one. Also, it is just as easy to get importing from non-importing as the other way: >From importing {ro broda cu brode} -> non importing {ganai da broda gi ro broda cu brode} >From non-importing {ro broda cu brode} -> importing {ge da broda gi ro broda cu brode} >> Where, I assume, contrary to the claim, the free floating {da} is importing. Try again. << >In most cases it makes no difference -- except that inferences that go >through directly with impoarting quantifiers require an extra step with >non-importing ones (not a big problem, but an inelegance surely). The inelegance of importing for me comes from not being able to infer {ro broda cu brode} = {no broda naku brode} = {naku su'o broda naku brode} = {naku me'iro broda cu brode}. >> Explain to me again why these don't go through just fine? << >And, of >course, when it makes a difference, the non-importing one is what you want, >precisely because it does make a difference. If you don't want the >non-importing one, then it doesn't make a difference to be considered. >And, >should that occasion arise -- I'll trust you that it does -- you know where >to get what you want within the importing quantifier framework. I can say the same to you the other way. I just don't see that the inferrence {ro broda cu brode} -> {su'o broda cu brode} is more elegant than the four relationships that hold among the different quantifiers. To me that inferrence seems forced. If that is the "screw up" of the non-importing ro you were referring to, then I can live with it, as I don't see it as a screw up. >> Well, it is generally a correct inference, and thus not forced. As history shows the problem has been to get people to reject it, not to accept it. Since the four relationships work just fine for the importing quantifiers, there is no loss there. And since we can get the non-importing version out when needed, even that is not lost. whereas the non-importing sort loses all the importing inferences -- the ones we generally use. |