if the rule about no la or lai or doi in cmene didn't exist: Difference between revisions

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


'''mingi'u'''
From another page:


''Preceding these with ''za'e'' will endear you to those of us who are not committed [[jbocre: hardliners|naturalists]]. (Not using them at all will endear you even more.)''
(I personally believe we should have abolished the nonsense about no la/lai/doi in cmene and just required a mandatory pause before as well as after in all settings.) --mi'e [[User:Mark Shoulson|.mark.]]


* [[jbocre: loglo fu'ivla|loglo fu'ivla]]
----
* [[jbocre: Resurrected gismu|Resurrected gismu]]


* [[jbocre: experimental cultural gismu|Experimental cultural gismu]]
I agree. Could we add a rule that says that names that begin in a glottal stop CAN contain la/lai/doi? --[[User:And Rosta|And Rosta]]
* [[jbocre: Slang gismu|Slang gismu]]


* [[jbocre: Free Gismu Space|Free Gismu Space]]
*That wouldn't help. To take an example from the list, {la .dalais. .lamas.} could still be interpreted as {la da lai s. la mas.} because "s" and "mas" would not be required to begin with a glottal stop. --[[jbocre: rab.spir|rab.spir]]
* [[jbocre: Rant: Against Experimental Gismu|Rant: Against Experimental gismu]]
*Yeah.  That wouldn't be enough for the (putative) word-parser.  It has to know where cmene begin.  It knows they end at '''C.''', and searches back to find the beginning, which is either a pause or a la/lai/doi, whichever comes first.  If we allow this, we'd force it always to search back to the pause, which, in the case of an "old-style" cmene, could be back at the start of the jufra.  That is, ''.i mi'o tavla la lojban.'' could just as easily be one biiig cmene, ''.imi'otavlalalojban.''  The only way is to require that '''all''' cmene be preceded by a pause, '''everywhere'''.  Me, I don't think that's such a hardship.  I'd just train myself that the word ''la'' was pronounced with a glottal stop at the end (like the Klingon word '''la''''/commander), and so it would always have a pause afterwards.  Likewise ''lai'' and ''doi''.  [[jbocre: hardliners ardliner|hardliners ardliner]] though I be, I get the feeling that usage will probably bring this to pass.  I've been involved in Lojban for a long time now, and I '''still''' see Lojbanists as experienced as I accidentally sticking illegal syllables into their cmene.  The la/lai/doi rule probably will simply not survive.  Hardliners might try to keep the "pause-before" rule to compensate, but that may be hard to instill. ''--mi'e mark.''


* [[jbocre: Experimental gismu proposal|Experimental gismu proposal]]. Let's settle this the real, macho (or as macho as lojban gets) way: Tinker at each other!
**I don't think it matters if it's hard to instil. No human brain will actually speak and hear according to the morphophonological rules, and the sounds people actually make don't matter all that much. Getting from actual speech sounds to a phonological representation is a pattern-matching exercise more suited to a neural net sort of thingo than to your ordinary linear algorithm. [[User:And Rosta|And Rosta]]
* [[jbocre: new color gismu|new color gismu]]. Proposal for 4 new gismu for colors
**Probably true. In writing, even those who write their dots often forget to write them after ''mi'e'' for example. --mi'e [[User:xorxes|xorxes]]


=== Staple food ===
***True and guilty as charged, but I don't think that's very damning.  Just because a dot isn't written doesn't mean it isn't pronounced.  I usually remember that a pause is required after ''mi'e'' even if I don't write it.  Dots are optional reminders, not mandatory parts of the orthography, IMO. ''--mi'e mark''
**.i mi'o tavla la lojban doesn't make a whole lot of sense - did you mean .i mi'o tavla fo la lojban or .i mi'o tavla fi la lojban, by any chance?


'''plumu''' plum:x1 is a peach/plum/cherry/almond of species x2
***''fi la lojban.'' I misremembered the structure of ''tavla.'' I think I was thinking of ''casnu.''
* I also agree. I've been experimentally putting pauses after ''la'', etc., and I find that it's easier to remember to pause after the word when I do. -- Adam


* (''prunu'' conflicts with ''pruni'' "elastic")We have fu'ivla for some of the species, but no concise way to distinguish between a plum and any member of the whole genus without using fu'ivla rafsi (''ziryrutrprunu''). With this gismu, it could be zirplumu - not that all plums are the same color. --[[jbocre: Pierre Abbat|phma]]
----
** This might be a good temporary gismu, but plant and animal gismu should be done by the algorithm. -- [[jbocre: Adam|Adam]]


'''smela''' plum (instead of ''plumu'')
We've made some progress in that la/lai/doi are now allowed after a consonant. How about allowing them when the syllable is accented? If the la/lai/doi was really supposed to be a cmavo, it would either be unaccented or followed by a stop. --[[jbocre: rab.spir|rab.spir]] ''No it wouldn't; cmene aren't brivla, so that rule doesn't apply.''
 
*** fixed version of scoreGismu gives the answer. "plum" in all langs:
*** Chinese  Méi > mei
 
*** Hindi    Bēra > bera
*** English  plum > plam
 
*** Spanish  ciruela > siruela
*** Russian sliva > sliva
 
*** Arabic  Brqwq > byrkuk
*** mei 0.347 bera 0.196 plam 0.160 siruela 0.123 sliva 0.089  byrkuk 0.085
 
*** The highest score of 0.53 has  "smela". And I suggest {zirsmela} for plum, {xunsmela} for cherry, {najysmela} for peach, {pelsmela} for apricot, {ri'onsmela} for almond, {blasmela} for sloe. --mi'e [[jbocre: najrut|najrut]]
 
'''kokso''' for coconut. Coconut is just as important as bread in some countries. So we have
 
Chinese / Hindi / English / Spanish / Russian / Arabic
 
keke 0.347 / koko 0.196 / koko 0.160 / koko 0.123 / kokos 0.089 / kuku 0.085
 
And the highest score of 0.7662 have "koklo kokpo skoko kokro kokco kokso ckoko kokno kokfo kokmo kokto". So I suggest {kokso} (as the sound "s" is present in one of the six langs)  --mi'e [[jbocre: najrut|najrut]]
 
'''pamga''' for papaya that is prevalent in America.
 
We have "mugua 0.347 papita 0.196 papaia 0.160 papaia 0.123 papaia 0.089 babaia 0.085" and scoreGismu outputs "pimga pamga" with the score of 0.4922
 
=== Gismu for living beings ===
 
'''korvo''' crow: x1 is a crow/raven/magpie/jay of species x2
 
* Derived from Latin ''corvus'', this word matches English ''crow'', Spanish ''cuervo'', and Hebrew ''`orev'' (''what's the Arabic word?'').
**Since Latin "corvus" first of all means "raven", I'd put it first: x1 is a raven/crow/... "`orev" is both raven and crow, so it is in Malti (very close to Arabic) where it's "cawlun" (tcaulun); "cock-crow" also is "il-ghajta tasserduk" (il 'aita...).
 
a gismu for '''crustacean'''
 
*There should be for crustacean, arachnid, and insect a gismu. Most people AFAIK have no problem for most arthropods telling which they are, with a few exceptions like the horseshoe crab, which despite its name is a chelicerate, like the arachnids. With three gismu we could easily tell the crab month from the scorpion month. It makes no sense to me that the same gismu is used for both. -[[jbocre: Pierre Abbat|phma]]
**I'm not so sure. Hardly anyone will be able to tell you that a pill bug is a crustacean.
 
**An improved version of scoreGismu outputs:
**Crustacean: lonxia 0.347 krefica 0.196 lopstr 0.160 langosta 0.123 rak 0.089  srtan.albxr 0.085    {lokra klora  0.44}
 
**Insect: cinki
**Arachnid: jiju 0.347 makari 0.196 spaidr 0.160 arana 0.123 pauk 0.089  xnkbut 0.085    {jijru jidju 0.40}
 
**We have suggested gismu and their score in braces. I'm not sure whether I've done the proper transliteration. So please put me right if necessary. I think lokra and jidju would be nice for new gismu and jukni should become obsolete in favor of them -[[jbocre: najrut|najrut]]
*I think there are proportionally a few too many gismu for mammals as opposed to other animals. For example, there is "cmacu" (mouse) as well as "ratce" (rat) but no gismu for "turtle," "frog," or any aquatic invertebrates - [[User:PlasticRaven|Plastic Raven]]
 
=== Other concepts that might have gismu ===
 
See [http://arj.nvg.org/lojban/unofficial-gismu.html]
 
Here are some concepts that should have gismu, but don't.
 
'''frili''' and '''sampu''' might be merged into one gismu.
 
*frili:x1 is easy for x2 under conditions x3
*sampu:x1 is simple in property x2
 
**For now, we may use temporary lujvo ''filsampu'' - s1=f1 is simple/easy in property s2 for f2 under conditions f3 --najrut
 
a gismu for '''intension''' (''This should be 'intention'. See also [[jbocre: intensional|intensional]]'')
 
*Something along the lines of ''x1 is intended to be/supposed to be x2 (ka) as intended by x3''. Unless someone can tell me how to say this with existing gismu, which would be great. ''jinzi'', ''bilga'' don't seem quite right. -- [[jbocre: Adam|Adam]]
** Something with djica dunli.
 
*** The idea is to translate those pesky things like "wine bottle". According to the most common interpretation (as far as I can tell), "botpi lo vanju" must actually contain wine. A wine bottle could be something like "broda le ka vanju botpi" if "broda" has the place structure above.
**** How many people would have a problem with "vanju botpi"? Do you think somebody would ask how a bottle could be made out of wine?
 
***** The problem isn't "vanju botpi". If I have an bottle filled with water which is a "wine bottle" in English by shape and probably former contents, I can't call it a "ca'a botpi lo vanju" in lojban, and thus "vanju botpi" may be misleading.
***** I am afraid you are being too literal. A bottle that is has only held water, but was made to hold wine, is still a vanju botpi.
 
***** So you would say ''ti vanju botpi lo djacu'' in the case above?
***** Yes. And if anybody complains, let them first read Chap 12 section 2 (p 275 in the printed version). The English word "for" in the definition of botpi ("...bottle for x2") may create some confusion. It may be interpreted to mean that is what the bottle is ''intended'' to contain, but not what it actually contains! Insisting upon that usage would drive us to ''ti botpi lo vanju fi'o vasru lo djacu''. Of course, none of this is anything like a general solution to your original issue! How would you like a ''fi'o djitai'' modal?--[[User:xod|xod]]
 
*''ti tutci lo'e nu botpi lo vanju''This is used for bottling wine.''ti vanbo'itci'' --[[User:xorxes|xorxes]]
** That sounds like a device for filling wine bottles to me.
 
*** The English translation may be misleading. Maybe it should be 'This is used to contain wine'. ''botpi'' of course does not mean 'to bottle' in the sense of putting into a bottle, there is no such action implied in the Lojban version.
 
'''[[jbocre: brivo|brivo]]''' predicate-word: x1 [[jbocre: zo|zo]] is a predicate-word expressing relationship x2 among arguments x3
 
'''pitsa''' pizza: x1 is a pizza [[jbocre: of type x2?]] ''with ingredients/toppings x2? mi'e .filip.''
 
'''salsa''' salsa: x1 is a quantity of salsa containing ingredients x2
 
* But '''sanso''' already means that! --mi'e [[User:xorxes|xorxes]]
** this is just more specific.
 
*** ''No it isn't, salsa is just the Spanish word for sauce. This can be done with sanso if you put chips in the x2 place. - [[jbocre: .kreig.daniyl.|.kreig.daniyl.]]''
**** ''salsa is also the Italian word for sauce. When an Italian says "salsa", what he means more likely has tarragon and oregano in it than chili peppers. -[[jbocre: Pierre Abbat|phma]]''
 
* ''mexsanso'' also works. I hardly think this concept is so important that a lujvo isn't good enough.
** Any particular reason this is ''mexno'' and not ''xispo'' or whatever?
 
***The Spanish word "salsa" means ''sanso'': '''any''' kind of sauce. The English word "salsa", borrowed from Spanish, means ''mexsanso'', some typically Mexican sauce, I suppose. To an Argentinian, "salsa" does not have any strong association to that kind of sauce. Using ''salsa'' for ''mexsanso'' is malmerko. --[[User:xorxes|xorxes]]
*''Salsa'' is sort of like ''Barbeque''. To some people, it is a specific thing, to others it is not. And to those for whom it is specific, the answer may differ. (Ask me sometime about the confusion this can cause when you want to buy salsa or offer tips on barbequeing). - mi'e. [[jbocre: .kreig.daniyl.|.kreig.daniyl.]]
 
'''taksi''' taxi: x1 is a taxi/cab
 
* [[User:xorxes|xorxes]] pointed out that this should have a place for passenger/cargo, probably in x2. (Compare karce, carce, and marce, which all have "for carrying x2".) I suggest that '''taksi''' (should it exist) also have a place for usual area of operation, and possibly one for the driver. Perhaps something like '''''taksi''' taxi: x1 is a taxi/cab for carrying x2, operating in area x3, with driver x4''? - mi'e .filip.
**Generally, within reason, the more places (up to five), the better. In this case, .filip.'s proposed places sound reasonable to me. --[[User:PlasticRaven|Plastic Raven]]
 
'''tango''' tango: members of set x1 consisting of lead x2 and follower x3 dance a tango
 
* '''This clashes with ''tanko'''''
* x1 (couple) tangoes to accompaniment/music x2 (''the place structure of '''dansu''' seems much more appropriate, besides the tango is a kind of music before being a dance'')
 
** '''tangygi'a''' for the lead and '''tangyselgi'a''' for the follower.
* ''.i lo'e remei cu sarcu le nu tango''
 
**In addition to the clash with "tanko", "tango" encourages pronunciation of the outside-Lojban-ubiquitous "ng" sound, which does not exist in Lojban. --[[User:PlasticRaven|Plastic Raven]]
 
'''zbiga''' Asbergerish: x1 is a manifestation of Asberger's Syndrome in x2 as detected by x3
 
* ''do you mean "Asperger's Syndrome"? -b- seems a common mispeling on the Web, however. Maybe '''spiga''', then? --pne''
**This seems like it would be used much more often by outsiders in reference to the Lojban community than the other way around. Since Asperger's syndrome isn't normally a common topic of discussion, this is almost like asking for trouble, but not necessarily in a bad way. --[[User:PlasticRaven|Plastic Raven]]
 
'''zombi''' zombie: x1 is a zombie in event/activity x2
 
* As amusing as that would be, I think gismu for various types of undead is probably going just waaay too far. (Should be a feasible lujvo, though.)
** Take a look at what happened to the innocent word "vampire" in [[jbocre: round two|round two]] of Broken Phone.
 
***You could just use compounds involving "crida" and any other morbid gismu. --[[User:PlasticRaven|Plastic Raven]]
 
'''nisku''' fraik: x1 is a fraik/marrot/eligug of species x2
 
* This is a back-formation from ''sfenisku'' "[[jbocre: P|Penguin]]", which is a kind of ''nisku'' that stays on or goes under the surface, as opposed to other kinds of ''nisku'' which can fly. The great auk was a ''sefta nisku'', but not a ''sfenisku''.
* {xatci} for 'yawn'. - Lindar
 
** Why {xatci} ?
*** Chinese Dǎ hāqian daxakian
 
*** Hindi Jambhā'ī jambai
*** English         yawn ion
 
*** Spanish bostez-o bostez
*** Russian zev zev
 
*** Arabic ttha'b  tytyxaxb
**Then scoreGismu suggests "ximna/kibni/kimna/xibni/dibni". But isn't fu'ivla much better for 'yawn' ? -- mi'e najrut
 
'''Clockwise''' and '''counter-clockwise''' probably deserve gismu (complete with axis/frame of reference place).
 
Ctino proposes to '''[http://www.lojban.org/tiki/new%20color%20gismu dd four new color gismu to the list]'''.

Revision as of 16:52, 4 November 2013

From another page:

(I personally believe we should have abolished the nonsense about no la/lai/doi in cmene and just required a mandatory pause before as well as after in all settings.) --mi'e .mark.


I agree. Could we add a rule that says that names that begin in a glottal stop CAN contain la/lai/doi? --And Rosta

  • That wouldn't help. To take an example from the list, {la .dalais. .lamas.} could still be interpreted as {la da lai s. la mas.} because "s" and "mas" would not be required to begin with a glottal stop. --rab.spir
  • Yeah. That wouldn't be enough for the (putative) word-parser. It has to know where cmene begin. It knows they end at C., and searches back to find the beginning, which is either a pause or a la/lai/doi, whichever comes first. If we allow this, we'd force it always to search back to the pause, which, in the case of an "old-style" cmene, could be back at the start of the jufra. That is, .i mi'o tavla la lojban. could just as easily be one biiig cmene, .imi'otavlalalojban. The only way is to require that all cmene be preceded by a pause, everywhere. Me, I don't think that's such a hardship. I'd just train myself that the word la was pronounced with a glottal stop at the end (like the Klingon word la'/commander), and so it would always have a pause afterwards. Likewise lai and doi. hardliners ardliner though I be, I get the feeling that usage will probably bring this to pass. I've been involved in Lojban for a long time now, and I still see Lojbanists as experienced as I accidentally sticking illegal syllables into their cmene. The la/lai/doi rule probably will simply not survive. Hardliners might try to keep the "pause-before" rule to compensate, but that may be hard to instill. --mi'e mark.
    • I don't think it matters if it's hard to instil. No human brain will actually speak and hear according to the morphophonological rules, and the sounds people actually make don't matter all that much. Getting from actual speech sounds to a phonological representation is a pattern-matching exercise more suited to a neural net sort of thingo than to your ordinary linear algorithm. And Rosta
    • Probably true. In writing, even those who write their dots often forget to write them after mi'e for example. --mi'e xorxes
      • True and guilty as charged, but I don't think that's very damning. Just because a dot isn't written doesn't mean it isn't pronounced. I usually remember that a pause is required after mi'e even if I don't write it. Dots are optional reminders, not mandatory parts of the orthography, IMO. --mi'e mark
    • .i mi'o tavla la lojban doesn't make a whole lot of sense - did you mean .i mi'o tavla fo la lojban or .i mi'o tavla fi la lojban, by any chance?
      • fi la lojban. I misremembered the structure of tavla. I think I was thinking of casnu.
  • I also agree. I've been experimentally putting pauses after la, etc., and I find that it's easier to remember to pause after the word when I do. -- Adam

We've made some progress in that la/lai/doi are now allowed after a consonant. How about allowing them when the syllable is accented? If the la/lai/doi was really supposed to be a cmavo, it would either be unaccented or followed by a stop. --rab.spir No it wouldn't; cmene aren't brivla, so that rule doesn't apply.