Revised species (eliminating species) Posted by pycyn on Mon 21 of Jun, 2004 22:44 GMT posts: 2388 Use this thread to discuss the Revised species (eliminating species) page.
Posted by pycyn on Mon 21 of Jun, 2004 22:44 GMT posts: 2388 I have posted notes for a first revision of the species paper as a comment to that paper. xorxes was, of course, right; it can be done without species, using only proeprties (and what goes with them in Lojban).
Posted by xorxes on Tue 22 of Jun, 2004 15:26 GMT posts: 1912 I'm not sure I have digested everything yet, but here are some comments. pc: > {lo broda cu brode} iff *P overlaps Q and S(P) > intersects S(Q) > Remember, “intersects” is a more stringent requirement > than mere mingling. Is it really necessary to separate intensional broda from the rest? With XS-lo, the interpretation would be P overlaps Q whatever brode is. (And any {naku} would say that P does not overlap Q, not that it does overlap ~Q.) > For every property P there is a set S(P) = {x : Px} > For every set s, there is a property P(s) such that s = {x :P(s)x} Does this mean that for each P there is one and only one set S(P), and for each s there is one and only one property P(s)? The second one doesn't sound right. > So, s = M(P(s)) and P = P(M(P)) If there can be many properties corresponding to the same set, then the first one might mean that for any P(s), s=M(P(s)), the second one that there is some P(M(P)) such that P=P(M(P)) > i is an individual concept iff i is a property that pervades > no other property. (i may overlap and number of other property > and – with at most one exception – does.) The concepts "... is 2" and "... is 1+1" seem to pervade each other, so neither is an individual concept. (Or are they both the same concept?) What would be an example of an individual concept? What is the possible exception, "nothingness"? mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Posted by pycyn on Tue 22 of Jun, 2004 15:26 GMT posts: 2388 A> Yes. Not doing so is the temptation that kills so many attempts to do things with properties. P overlaps Q in no wise gurantees that S(P) mingles with S(Q), though the converse does seem to hold. thus the compromise given. And, of course, in the intensional contexts, the mingling is largely irrelevant. I wish I could think of a way around this since the dual system is irritating. but I can't see it — except to reverse the procedure and make the intensional the norm and add a special clause for all the intensional ones. And that is merely a stylistic difference, without substance. If your system really works the way you claim and your understanding of overlap is the same as mine, then your system is seriously flawed in a more concrete way than usual. B> I am not sure you can pull this off and still get a meaningful transfer of negations. To be sure, negation doesn't transfer as nicely as one would like anyhow, but it does doe so in enough cases to make it useful to be able to mark it. And, of course, we need in any case to be able to say that P does overlap -Q. C> Not one and only one, but at least one. A given set may be the extension of several intensions, but each intension determines a unique set (which may be equally determined by some ther intension). I see that is not clear and that I use it ambiguously later. I'll clean it up. D> Yes; see above. E> If they are ic and pervade one another, they are identical — is the way it goes. I'm not sure that these are cases of that sort, but, since the connection seems analytic, I suppose they are. The exception is the null property, the "nexus" of no semantic threads. I suppose that "nothingness" comes about as close as any natural language can come to expressing it. Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote: I'm not sure I have digested everything yet, but here are some comments. pc: > {lo broda cu brode} iff *P overlaps Q and S(P) > intersects S(Q) > Remember, “intersects” is a more stringent requirement > than mere mingling. A>Is it really necessary to separate intensional broda from the rest? With XS-lo, the interpretation would be P overlaps Q whatever brode is. (B>And any {naku} would say that P does not overlap Q, not that it does overlap ~Q.) > For every property P there is a set S(P) = {x : Px} > For every set s, there is a property P(s) such that s = {x :P(s)x} C>Does this mean that for each P there is one and only one set S(P), and for each s there is one and only one property P(s)? The second one doesn't sound right. > So, s = M(P(s)) and P = P(M(P)) D>If there can be many properties corresponding to the same set, then the first one might mean that for any P(s), s=M(P(s)), the second one that there is some P(M(P)) such that P=P(M(P)) > i is an individual concept iff i is a property that pervades > no other property. (i may overlap any number of other property > and – with at most one exception – does.) E>The concepts "... is 2" and "... is 1+1" seem to pervade each other, so neither is an individual concept. (Or are they both the same concept?) What would be an example of an individual concept? What is the possible exception, "nothingness"? mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Posted by xorxes on Tue 22 of Jun, 2004 15:27 GMT posts: 1912 pc: > Yes. Not doing so is the temptation that kills so many attempts to do > things with properties. P overlaps Q in no wise gurantees that S(P) mingles > with S(Q), though the converse does seem to hold. thus the compromise given. Why "thus"? The assumption seems to be that we must guarantee mingling whenever mingling could be relevant, but I don't see why that is such an absolute requirement. We have other ways to guarantee mingling: {su'o broda cu brode}. > And, of course, in the intensional contexts, the mingling is largely > irrelevant. I wish I could think of a way around this since the dual system > is irritating. but I can't see it — except to reverse the procedure and > make the intensional the norm and add a special clause for all the > intensional ones. And that is merely a stylistic difference, without > substance. If your system really works the way you claim and your > understanding of overlap is the same as mine, then your system is seriously > flawed in a more concrete way than usual. What is the flaw? > I am not sure you can pull this off and still get a meaningful transfer > of negations. To be sure, negation doesn't transfer as nicely as one would > like anyhow, but it does doe so in enough cases to make it useful to be able > to mark it. And, of course, we need in any case to be able to say that P > does overlap -Q. We have ways of saying that: {lo broda cu me lo na brode} is the most obvious one. Or, if you still don't know what {me} means, use whatever relationship expresses overlap between the two arguments {lo broda} and {lo na brode}. Using naku, which denies that a relationship holds, to affirm instead that some other relationship holds is, in my opinion, an unnecessary complication. (And you will have to deal with negation in the multiple argument cases too, which don't convert so easily to overlap talk.) > If they are ic and pervade one another, they are identical — is the way > it goes. I'm not sure that these are cases of that sort, but, since the > connection seems analytic, I suppose they are. OK. Not sure what you will be using "individual concepts" for, then. "...is 2" and "...is 1+1" are intensionally different, i.e. different properties, even though by your definition they are the same individual concept. mu'o mi'e xorxes __ Do you Yahoo!? Yahoo! Mail - Helps protect you from nasty viruses. http://promotions.yahoo.com/new_mail
Posted by pycyn on Tue 22 of Jun, 2004 15:27 GMT posts: 2388 A>Well, I think we were talking about thinks like either {su'o broda cu brode} or {lo broda cu brode}, in both of which mingling is relevant. Sure, if it weren't we could ignore it, but it almost always is — and certainly is in the cases where I have called for it. B> That it can't distinguish between overlap and mingling when the distinction is crucial — as it usually is. C> {me} can't be the magic bullet for solving all problems (though I have to admit that JCB used it thus for a while. But even if it were, why would we want this periphrasis in place of a straightforward distinction that was nicely built into the language 50 years ago and hasn't been challenged up to now (though occasionally misused)? D> ISuppose they are intesnionally distinct; what follows from that is that they do not pervade one another. To be sure, it is hard to imagine a world in which they had different extensions, but, if they are intensionally distinct, there must be at; least one such, which proves they do not pervade one another. In a word, the two concepts, intensionally distinct and pervade one another, do not hang together. An individual concept is a property unique to a — real or imaginary — individual. Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote: pc: > Yes. Not doing so is the temptation that kills so many attempts to do > things with properties. P overlaps Q in no wise gurantees that S(P) mingles > with S(Q), though the converse does seem to hold. thus the compromise given. A>Why "thus"? The assumption seems to be that we must guarantee mingling whenever mingling could be relevant, but I don't see why that is such an absolute requirement. We have other ways to guarantee mingling: {su'o broda cu brode}. > And, of course, in the intensional contexts, the mingling is largely > irrelevant. I wish I could think of a way around this since the dual system > is irritating. but I can't see it — except to reverse the procedure and > make the intensional the norm and add a special clause for all the > intensional ones. And that is merely a stylistic difference, without > substance. If your system really works the way you claim and your > understanding of overlap is the same as mine, then your system is seriously > flawed in a more concrete way than usual. B>What is the flaw? > I am not sure you can pull this off and still get a meaningful transfer > of negations. To be sure, negation doesn't transfer as nicely as one would > like anyhow, but it does doe so in enough cases to make it useful to be able > to mark it. And, of course, we need in any case to be able to say that P > does overlap -Q. C>We have ways of saying that: {lo broda cu me lo na brode} is the most obvious one. Or, if you still don't know what {me} means, use whatever relationship expresses overlap between the two arguments {lo broda} and {lo na brode}. Using naku, which denies that a relationship holds, to affirm instead that some other relationship holds is, in my opinion, an unnecessary complication. (And you will have to deal with negation in the multiple argument cases too, which don't convert so easily to overlap talk.) > If they are ic and pervade one another, they are identical — is the way > it goes. I'm not sure that these are cases of that sort, but, since the > connection seems analytic, I suppose they are. D>OK. Not sure what you will be using "individual concepts" for, then. "...is 2" and "...is 1+1" are intensionally different, i.e. different properties, even though by your definition they are the same individual concept. mu'o mi'e xorxes __ Do you Yahoo!? Yahoo! Mail - Helps protect you from nasty viruses. http://promotions.yahoo.com/new_mail
Posted by xorxes on Tue 22 of Jun, 2004 15:27 GMT posts: 1912 pc: > That it can't distinguish between overlap and mingling when the > distinction is crucial — as it usually is. But it can: lo broda cu brode - overlap su'o broda cu brode - mingling mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Posted by pycyn on Tue 22 of Jun, 2004 20:31 GMT posts: 2388 Well, then rather to my surprise it turns out that both {lo cinfo cu fetsi} and {lo cinfo cu nakni} are true. Indeed, so is {lo tirxu cu blabi}. But, if that is the case, then — outside of intensional contexts — there is no difference between {su'o broda} and {lo broda}. This is not at all the imprssion you have given over the last few years. I should add that, as far as I can see, there is no difference in intensional contexts either, unless the context specifies that in it there are to be no broda. Notice that the difference between {lo} and {su'o} in extensional contexts does not rest on the distinction between overlap and mingling but on the distinction between mingling and intersection. Both reduce to overlap in the intensional case, where differences in actual distribution do not apply. Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote: pc: > That it can't distinguish between overlap and mingling when the > distinction is crucial — as it usually is. But it can: lo broda cu brode - overlap su'o broda cu brode - mingling mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Posted by xorxes on Tue 22 of Jun, 2004 20:31 GMT posts: 1912 pc: > Notice that the > difference between {lo} and {su'o} in extensional contexts does not rest on > the distinction between overlap and mingling but on the distinction between > mingling and intersection. Both reduce to overlap in the intensional case, > where differences in actual distribution do not apply. XS-lo does not distinguish between mingling and intersection, or indeed any other distributions. The instance distribution is simply not a part of the claim. If you want to infer something about distribution you have to deduce it from context or ask the speaker for more specific information. mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail
Posted by pycyn on Tue 22 of Jun, 2004 20:48 GMT posts: 2388 Well, then {lo broda cu brode} means exactly the same thing as {su'o broda cu brode} — contrary to what you have often said — and I don't see what all the fuss was about. Intersection is an attempt to make sense of your repeated claim that {lo broda cu brode} amd a general claim that {su'o broda cu brode} did not. If you no longer hold that view (which has beena peculiar one from the get-go), then we pretty much agree on everything and I can drop all of this except the renewed suggetion that , since you don't need it, you get rid of Mr. Broda. jllambias2000@yahoo.com.ar> wrote:pc: > Notice that the > difference between {lo} and {su'o} in extensional contexts does not rest on > the distinction between overlap and mingling but on the distinction between > mingling and intersection. Both reduce to overlap in the intensional case, > where differences in actual distribution do not apply. XS-lo does not distinguish between mingling and intersection, or indeed any other distributions. The instance distribution is simply not a part of the claim. If you want to infer something about distribution you have to deduce it from context or ask the speaker for more specific information. mu'o mi'e xorxes __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail
Posted by xorxes on Tue 22 of Jun, 2004 21:33 GMT posts: 1912 pc: > Well, then {lo broda cu brode} means exactly the same thing as {su'o broda cu > brode} — contrary to what you have often said — and I don't see what all > the fuss was about. We must be speaking different languages. I say that {lo broda cu brode} makes no claim about instances, and from that you conclude that it's exactly the same thing as making the claim about at least one instance. {su'o da poi broda} is just not interreplaceable with {zo'e noi broda}. One claim is much more precise than the other. Sometimes we want such extra precision, sometimes we don't. > Intersection is an attempt to make sense of your > repeated claim that {lo broda cu brode} amd a general claim that {su'o broda > cu brode} did not. It can be used to make a more general claim, yes, in the right context. {zo'e} is very context-sensitive. It does not always make a general claim. If you present a sentence out of the blue with no context, the general claim tends to be the first that comes to mind, that's all. > If you no longer hold that view (which has beena peculiar > one from the get-go), then we pretty much agree on everything and I can drop > all of this except the renewed suggetion that , since you don't need it, you > get rid of Mr. Broda. I have no particular attachment to Mr Broda. He is not mentioned in the proposed definitions. mu'o mi'e xorxes __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail
Posted by pycyn on Tue 22 of Jun, 2004 22:22 GMT posts: 2388 So, we're back to {lo} is general — or just more vague --than {su'o}. Not it is strictly not possible to be more vague than {su'o} unless you allow that the number might be zero. On the other hand, I present you with a notion for {lo} that is as vague as {su'o} but neither implies {su'o} nor is implied by it. Whereas, your {lo} is both, if it is to do its job. Your second attempt here (less a failure than the first), is that {lo} is gnerally used to make general claims, but not always/ That is the meaning of {lo} that I have advocated for some time — even at the beginning of the species paper. It also does away with Mr. Broda codswallop and leaves {lo} and {su'o} materially equivalent still. I doubt that {zo'e} is more context sensitive than intersection, by the way — as I noted in the first revision. I wonmdered what happened to Mr. Broda (I still do), but at least he has vanished after all this time. I hope that the lunacy that went with him has gone too, but vestiges (mild psychoses?) still seem to be around. Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote: pc: > Well, then {lo broda cu brode} means exactly the same thing as {su'o broda cu > brode} — contrary to what you have often said — and I don't see what all > the fuss was about. We must be speaking different languages. I say that {lo broda cu brode} makes no claim about instances, and from that you conclude that it's exactly the same thing as making the claim about at least one instance. {su'o da poi broda} is just not interreplaceable with {zo'e noi broda}. One claim is much more precise than the other. Sometimes we want such extra precision, sometimes we don't. > Intersection is an attempt to make sense of your > repeated claim that {lo broda cu brode} amd a general claim that {su'o broda > cu brode} did not. It can be used to make a more general claim, yes, in the right context. {zo'e} is very context-sensitive. It does not always make a general claim. If you present a sentence out of the blue with no context, the general claim tends to be the first that comes to mind, that's all. > If you no longer hold that view (which has beena peculiar > one from the get-go), then we pretty much agree on everything and I can drop > all of this except the renewed suggetion that , since you don't need it, you > get rid of Mr. Broda. I have no particular attachment to Mr Broda. He is not mentioned in the proposed definitions. mu'o mi'e xorxes __ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail
