robin's gadri proposal: Difference between revisions

From Lojban
Jump to navigation Jump to search
mNo edit summary
 
mNo edit summary
Line 1: Line 1:


-=na tanxe=-
I'm insane for doing this.


The following is a translation of Anotinette Portis' [http://www.amazon.com/Not-Box-Antoinette-Portis/dp/0061123226/ref=pd_bbs_sr_1/104-8101861-8524715?ie=UTF8&s=books&qid=1181085233&sr=8-1 ot a Box], which demonstrates nicely different type of lojban negationSince this cute children's book loses a lot without the illustrations, I have appended to each line a description of the drawing on that page. The protagonist is a little child bunny:
My goals are maximum expressive power across a wide range of axes, and minimum change of CLL LojbanThis document is horrifically incomplete; at this point, I just want to see if it has enough '''potential''' support to justify writing up a better version.


mu'i ma do zutse  ne'i lo tanxe (bunny sitting in box)
The basic idea is to identify all of the axes (plural of axis) along which we want to be able to express things in a gadri-like fashion, assign each a UI, and specify a transformation that occurs when said UI is used after a gadri.  Since we have the power to define UI usage, this seems perfectly acceptable to me, and many of the UI are under-utilized and under-defined.


na'i go'i .i ti na tanxe (bunny sitting in racing car)
Note that I am '''''not''''' attached to this particular selection of UI cmavo; we can find other ones, or make some up, as necessary.


tezu'e ma do cpana le va tanxe  (bunny standing atop box)
If LX was a '''completely''' generic article (i.e. xorlo), then in this system lo broda == LX sa'e cu'i je'u .o'e nai da'i nai broda, and le broda == LX sa'e je'u cu'i .o'e nai da'i cu'i broda


na'i go'i .i ti ba'e na tanxe (bunny atop mountain)
* Personal Specificity -- sa'e
** The specific thing(s) I have in mind (i.e. "le").


mu'i ma do cecla renro lo djacu lo tanxe (bunny squirting box with hose)
*** le broda ~= ro lo sa'e broda == ro da poi mi pensi ke'a zi'e poi ke'a broda
** A specific thing(s), but I have no particular one in mind (i.e. "su'o da").


na'i go'i .i mi pu ba'e cusku le sedu'u ti na tanxe (bunny as firefighter)
*** lo broda ~= le sa'e cu'i broda == su'o da poi ke'a broda
** The idea of a thing / Mr. Thing / intensionality. Note that this is '''not''' si'o; that relates to a whole relationship, not merely the x1 thing.


xu caku do dasni lo tanxe (bunny wearing box)
*** lo sa'e nai broda ~= le sa'e nai broda == zo'e noi ke'a broda (which, obviously, I stole from xorlo).  Note that in this one case, default quantifiers are dropped.


na'i go'i .i ti na tanxe (bunny as robot)
* Existence -- da'i
** su'o da, or something like it: insisting that the thing really does exist.


xu do ca'o sanli ne'i leva tanxe (bunny standing in box)
*** lo broda == le da'i nai broda
** Imaginary, or otherwise fundamentally non-existant (with respect to the current semantic space).


na'i go'i .i ti ba'e ba'e ba'e ba'e na tanxe (bunny in crow's nest of pirate ship, basket of hot air balloon, howdah on elephant, and as tugboat captain)
*** lo da'i broda == le da'i broda
** Agnosticism about existence is da'i cu'i, and is the default for le.  Please note that this is arguable from CLL, but I think it's the best choice.


.o'o ri mo (bunny standing inside box with cut-out window)
** Used to exist (which I think is just "pu", but this needs to be addressed so that we can so "a doctor" and make it clear that agnosticism about the tense means that all the doctors having been killed isn't much relevant).


(bunny sitting atop box, thinking hard)
* Distributivity -- .o'e, or mass cmavo.
** (See http://philosophy.syr.edu/mckay.html for an explication.  This relates to what Lojban has historically called "masses".)


ri me lemi ba'e na'e tanxe (bunny in rocket ship, blasting off)
** The short version: Non-distributive just means that the elements are only considered as a whole in this predication.  Distributive is the default; each individual is considered by themselves.  "The students surround the building" cannot be represented in standard predicate logic without non-distributives / plurals / masses, because it implies that '''each''' student surrounds the building.
** Lojban has historically mixed up what I'm calling "Distributivity" and "Property Transference".  I'm not at all certain that seperating the two out is terribly useful, but I'm providing .o'e to allow it.
 
** It has been noted that .o'e, as a pure emotion cmavo, kind of sucks for this purpose, but I can't think of anything better.  If people would prefer to go back to conflating the two issues, or grab an unassigned cmavo, that's fine.  Consider this the weakest part of the proposal.
** Distributive
 
*** LE .o'e broda == LEi broda == su'o da poi gunma zi'e broda
** Non-distributive
 
*** LEi .o'e nai broda == LE broda
 
* Property Transference
** Handled solely by choice of gadri.
 
** Set-wise; result has only properties such as cardinality, number, and so on
** Mass-wise; result has all properites borne by any of the entities contained
 
* Indentification
** Handled by choice of LA vs LE cmavo, not UI.
 
** By Name
** By Descriptive selbri
 
* Accuracy / Veridicality -- je'u
** Contained identifier is intended to objectively describe the thing; IOW, the description is not specific to the speaker at all.
 
*** lo broda ~= le je'u broda
** Contained identifier is innaccurate / untruthful / nonce in some fashion known to the speaker; IOW, the description is specific to the speaker.
 
*** lo je'u nai broda ~= le je'u nai broda
** Cointained identifier may or may not be objectively accurate.
 
*** lo je'u cu'i broda ~= le broda

Revision as of 17:11, 4 November 2013

I'm insane for doing this.

My goals are maximum expressive power across a wide range of axes, and minimum change of CLL Lojban. This document is horrifically incomplete; at this point, I just want to see if it has enough potential support to justify writing up a better version.

The basic idea is to identify all of the axes (plural of axis) along which we want to be able to express things in a gadri-like fashion, assign each a UI, and specify a transformation that occurs when said UI is used after a gadri. Since we have the power to define UI usage, this seems perfectly acceptable to me, and many of the UI are under-utilized and under-defined.

Note that I am not attached to this particular selection of UI cmavo; we can find other ones, or make some up, as necessary.

If LX was a completely generic article (i.e. xorlo), then in this system lo broda == LX sa'e cu'i je'u .o'e nai da'i nai broda, and le broda == LX sa'e je'u cu'i .o'e nai da'i cu'i broda

  • Personal Specificity -- sa'e
    • The specific thing(s) I have in mind (i.e. "le").
      • le broda ~= ro lo sa'e broda == ro da poi mi pensi ke'a zi'e poi ke'a broda
    • A specific thing(s), but I have no particular one in mind (i.e. "su'o da").
      • lo broda ~= le sa'e cu'i broda == su'o da poi ke'a broda
    • The idea of a thing / Mr. Thing / intensionality. Note that this is not si'o; that relates to a whole relationship, not merely the x1 thing.
      • lo sa'e nai broda ~= le sa'e nai broda == zo'e noi ke'a broda (which, obviously, I stole from xorlo). Note that in this one case, default quantifiers are dropped.
  • Existence -- da'i
    • su'o da, or something like it: insisting that the thing really does exist.
      • lo broda == le da'i nai broda
    • Imaginary, or otherwise fundamentally non-existant (with respect to the current semantic space).
      • lo da'i broda == le da'i broda
    • Agnosticism about existence is da'i cu'i, and is the default for le. Please note that this is arguable from CLL, but I think it's the best choice.
    • Used to exist (which I think is just "pu", but this needs to be addressed so that we can so "a doctor" and make it clear that agnosticism about the tense means that all the doctors having been killed isn't much relevant).
    • The short version: Non-distributive just means that the elements are only considered as a whole in this predication. Distributive is the default; each individual is considered by themselves. "The students surround the building" cannot be represented in standard predicate logic without non-distributives / plurals / masses, because it implies that each student surrounds the building.
    • Lojban has historically mixed up what I'm calling "Distributivity" and "Property Transference". I'm not at all certain that seperating the two out is terribly useful, but I'm providing .o'e to allow it.
    • It has been noted that .o'e, as a pure emotion cmavo, kind of sucks for this purpose, but I can't think of anything better. If people would prefer to go back to conflating the two issues, or grab an unassigned cmavo, that's fine. Consider this the weakest part of the proposal.
    • Distributive
      • LE .o'e broda == LEi broda == su'o da poi gunma zi'e broda
    • Non-distributive
      • LEi .o'e nai broda == LE broda
  • Property Transference
    • Handled solely by choice of gadri.
    • Set-wise; result has only properties such as cardinality, number, and so on
    • Mass-wise; result has all properites borne by any of the entities contained
  • Indentification
    • Handled by choice of LA vs LE cmavo, not UI.
    • By Name
    • By Descriptive selbri
  • Accuracy / Veridicality -- je'u
    • Contained identifier is intended to objectively describe the thing; IOW, the description is not specific to the speaker at all.
      • lo broda ~= le je'u broda
    • Contained identifier is innaccurate / untruthful / nonce in some fashion known to the speaker; IOW, the description is specific to the speaker.
      • lo je'u nai broda ~= le je'u nai broda
    • Cointained identifier may or may not be objectively accurate.
      • lo je'u cu'i broda ~= le broda