devanagari orthography: Difference between revisions

From Lojban
Jump to navigation Jump to search
No edit summary
 
mNo edit summary
Line 1: Line 1:
''If you have more than ten sublists, the wiki breaks.''


''On another note, if people ever stop editing this page, it should be reorganized.''
''I created a very, very experimental Devanagiri orthography. Until I get an image up, I will describe it here:''


The page hasn't been edited (before now) since August, and it does seem to have some interesting stuff that could be dug out. Any volunteers to reorganize it?
The syllabic script is abandoned. So are the letters 'ei', 'ai', 'ou' and 'au'. Vowel marks now form dipthongs. The sound for retroflex 'sh' is used for Lojban 'j'. The aspirated stops are dropped. Double line for capitalization, 'aha' for full stop and other symbols unchanged. (I think writing an apostrophe would be pretty hard, though...)


''OK, sorry, d00d. I agree, but I think it best it not be me that do so. Also, um, sorry about the [[jbocre: ce'u 1|ce'u 1]] and [[jbocre: ce'u 2|ce'u 1]] I panickedly created; could whoever's last to turn out the lights refactor any changes from them back here? Again, apologies for losing my composure on this. --- Nick Nicholas''
I'm currently making a graphic that will map Lojban to Devanagari, one-to-one, and I'll add a link here soon.


''As a further note, I accept that I was wrong about this. -- [[User:Nick Nicholas|nitcion]]''
''A Devanagari orthography has been discussed on the list. See [http://groups.yahoo.com/group/lojban/message/1323,] http://groups.yahoo.com/group/lojban/message/1324, [http://groups.yahoo.com/group/lojban/message/1341''


----
Since] the topic is pretty old there, I think I'll continue the discussion here.
 
When absent from a ''ka'' phrase, ''ce'u'' is presumed to occupy the x1 of that phrase's ''selbri'' --- whether it is already occupied by another place or not. Thus: ''le ka mi xendo'' = ''le ka mi no'u ce'u xendo'', not ''le ka mi xendo ce'u''. In this, ''ce'u'' differs from ''[[jbocre: ke'a|ke'a]]''.
 
'''WHAT?''' First time I see this. Is ''le ka mi no'u ce'u xendo'' in any way different from ''le du'u mi xendo'' or ''le nu mi xendo''?
 
*Yes, I, [[User:Nick Nicholas|Nick Nicholas]], am responsible for (a) formulating this, and (b) putting this in the lessons; and there is no way it can be anything but this. I am comparing not ''ka'' to ''nu'', but ''ce'u'' to ''ke'a''. It is clearly untenable to suggest that ''ce'u'' occupy the first empty place, like ''ke'a'' does. I honestly don't see what's controversial about this.
**What is so untenable about ''la djan zmadu la pol le ka mi xendo''?
 
***I sure had to do a double-take on it. What's not so much less difficult to parse about ''la djan. zmadu la pol. le ka se xendo mi''?
****Nothing. I find them both equally acceptable.
 
**In what context would you use ''le ka mi no'u ce'u xendo''?
***How else would you say "He appreciated '''my kindness'''"? ''ko'a nelci le ka mi xendo''.


****There's no ''ce'u'' there, it might as well be ''nu'' instead of ''ka''.
Yes, indeed, the "unwritten vowel" is a schwa, just like Lojban y. However, it would be much more useful in to have it represent Lojban "a" in this orthography. Now, to write a conjuct consonant starting with r that doesn't have an equivalent in an Indic language, the problem is easily solved by "r" sign (like a hook on top) Others aren't so simple.
*****That's specious: there is a ''ce'u'' there, it's just elided. Or are you saying one cannot like properties?


******It sure should be ''nu'' instead of ''ka''. It is a mistake to always translate -ness as ''ka''. Just because a word is an adjective in English does not mean that it should be always abstracted with ''ka'' in Lojban. One can like properties in the same way one can like numbers, sets or propositions, they are not the kind of things people tend to like.
The suggestion for using ph for Lojban "f" and bh for Lojban "v" was a good one.
*******Still can't agree. I do not just like the abstraction; I like that the abstraction holds for a particular entity.


********That's ''nu'' then.
Dipthongs, on second thought, would be better written, (also as suggested on the list) using the available dipthongs. The Devanagari diphthong "ai",  although beginning with a schwa, can be used for Lojban "ai". Indeed, all diphtongs in Lojban are taken care of save "ii" and "uu". If we were using an orthogaphy like Kannada or Telugu, the solution for this would be quite simple: use the existing letters for long "i" and long "u". Unfortunately, Devanagari has the fomer, but not the latter, so to be consistent, we must use the "initial form" of each -with- the diacritic on the vowel.


''' -> [[jbocre: I like you being beaten by me|I like you being beaten by me]]'''
''Maybe I missed this in the discussion, but when I studied Sanskrit, there certainly was a letter (and a diacritic) for long-u, as well as long-i.  What I don't see, though, is a diphthong for '''oi'''.''(Because /o/ is already a diphthong in Sanskrit, the short form of /au/. pc)''  As to the other falling diphthongs, would you write them with semivowels w/y instead of vowels?  And are we using visarga for apostrophe, as I would think should be obvious? --mi'e mark. Eep, it looks like some of what's said there was by '''me'''.  I should read it first.''


*******Whether you call it ''nu'', ''ka'', or ''su'u kei lenu lunbe dansu'' is irrelevant --- the point is that you can need ''ce'u'' in this context, and it would be nice to have a default interpretation for elided ''ce'u'', just as with ''ke'a''.
::Bap!:: I was thehink about 'ei'. Sorry. Yes, you can write the other falling dipthongs with semivowels. Hm... and I never noticed '''oi''' either. What about '''eu'''? Same problem. ''(Exactly; /e/ is short /ai/. pc)'' And the visarga is logical for the apostrophe. Wouldn't the danda be logical for the full stop? And the puurNaviraama for the comma might work too.
********I like the same default interpretation as for ''ke'a'', I don't see why we need to complicate it.


********* *sigh* Because that would mean that ''le ka mi xlura'' does '''not''' mean 'my influence', but 'the susceptibility of others to me'. Which breaks usage and common sense. I will not agree on this, so we might as well take it to arbitration...
Make c into Lojban c and ch into Lojban tc, and j into Lojban j and jh into Lojban dj, Devanagari "h" into Lojban x, and all you have to worry about is punctuation. Now a full stop could be indicated using "aha", but as I said earlier, the "'" becomes a problem. So does capitalisation and the comma, because none of the existing Devanagari fonts I've seen so far have anything equivalent to those symbols. Any suggestions?
**********A better rendering of ''le ka mi xlura'' might be ''my influencing [[jbocre: them|them]]'', a property that someone else might have. It certainly does not braek common sense. As for usage, usage of ka was pretty chaotic before we came to understand it better.


*********Arbitration is '''not''' acceptable. The traditional arbitrators are biased.
(Was this in the earlier messages?  Transcribing English and Arabic and whatever, modern Hindi uses dots creatively: I remember dotted /j/ for /z/ -- and monolinguals saying things like /dijaster/ and /sijors/-- a brand of ciggyboos. pc)
 
BTW, another use of ''ce'u'' is with ''li'i'', which I think should also always have an implicit one like ''ka''. As in ''mi nelci le li'i ce'u jinga fi do''.
 
''I don't think so. An experience is pretty individual specific. I'm not sure of what an "experience of beating you" might be, divorced from an experiencer. (Then again, even better is to just replace "li'i" with "nu" :-)''
 
''ro da nelci le li'i ce'u jinga fi do''. It is precisely because it cannot be divorced from the experiencer that the place needs to be a ce'u, otherwise you wouldn't be able to tell which place corresponds to the experiencer.
 
What's the difference between that and ''ro da nelci le li'i da jinga fi do''?
 
I don't know whether le li'i da jinga fi do is supposed to be da's or do's experience.
 
How can you like someone else's experience? If you like their experiencing it, then say that: ''ro da nelci le nu do lifri le li'i da jinga fi do''.
 
''He means it like this: in "li'i klama", is this is experience of being a le klama, a le selkla, a le terkla, etc? I agree, ce'u can work with li'i and is a good thing. An experiencer takes a sumti slot, not a whole bridi. --xod''
 
Okay, whatever. I really think that we should just drop ''li'i'' and ''si'o'' anyway; you won't find me using them.


----
----


Whoa! John Cowan very recently [http://groups.yahoo.com/group/lojban/message/9481 made it clear on the mailing list], that there is NO implicit assumption that ce'u goes in the first place, or the first empty place. He said the proper location should be [[jbocre: glork|guessed]] from context. If that bothers you, swear an oath to always write your ce'us, and too bad about the text that's already been written. --xod
Whoever was working on this, if you could describe this orthography in terms of the Unicode Devanagari code chart at: [http://www.unicode.org/charts/PDF/U0900.pdf] then a [[jbocre: yudit|yudit]] keymap can be made very easily, and then pictures can be made even easier. --[[jbocre: Jay Kominek|Jay Kominek]]
 
''That's no different from the status of ''ke'a''; it too is properly meant to be [[jbocre: glork|glork]]ed from context, and that it should go to the first empty x1 is merely a convention, and a defeasible one at that:''
 
;: Like any sumti, ``ke'a'' can be omitted. The usual presumption in that case is that it then falls into the x1 place ...
 
;: However, ``ke'a'' can be omitted if it is clear to the listener that it belongs in some place other than x1 (http://www.lojban.org/files/reference-grammar/chap8.html)
 
''I'm just saying we should have a similarly non-binding convention for ''ce'u'', and x1 makes a lot more sense for that than first empty. Nothing forbids you using ''ke'a'' for second empty, and likewise nothing forbids you saying ''la djan zmadu la pol le ka mi xendo''. I think that's bucking the trend of what most people have been using for both, and is thus a smidgeon uncooperative, but that's why this is being proposed as a convention, not a rule.'' -- mi'e [[User:Nick Nicholas|nitcion]]
 
The Book gives explicit examples of ce'u being assumed to be in non-x1 and filled-x1 places. Does it give such examples of ke'a being elided and assumed to be elsewhere than x1? In usage I have never seen anyone pull either stunt. They elide ce'u and ke'a when in x1, and add it when not in x1, and really don't stick the sumti in a place that's already filled, something I don't understand but would use "poi ce'u" if I did.--xod
 
;:'''-> [[jbocre: ke'a|ke'a]]'''
 
''Course, that matches English, and I'm surprised you say you haven't seen ke'a=x2 (I assume you meant "elsewhere than x1".) I'm sure I pulled that stunt plenty in my day. And as for sticking ce'u in a place already filled, people have already been doing that, for the simple reason that almost noone knows ce'u exists. Are saying people are '''not''' saying ''le ka mi xendo'' for 'my kindness'? And again, ke'a and ce'u are not comparable: ke'a is anaphoric, so there's no point filling its slot with another sumti --- ce'u is not anaphoric, and you can fill in its value whenever you want without making it any less of a property. Indeed, you '''have''' to be able to fill in its value, or it's ultimately useless.''
 
You can fill in its value with the selbri, i.e. ''ko'a mutce le ka ce'u xendo'' and ''ko'a ckaji le ka jinga fi ce'u''. A property by itself is pretty meaningless, I mean, what's ''le ka ce'u te jinga'' by itself. It doesn't have a truth value, and it doesn't occur. It just sits in people's minds, waiting to become part of a bridi, where the ''ce'u'' will get filled in according to the meaning of the selbri. ''Sorry? ''ce'u'' is a sumti, so when it does become part of a bridi, it gets filled with a sumti value: ko'a mutce le ka ce'u no'u ko'a xendo. At least, that's how lambda variables work...'' Well, I'm sure you can walk all over me as far a lambda variables go, but it's contrary to my understanding of properties. Still, I don't think it makes much of a difference in this case.
 
''Now I'm wondering when this gets taken to the list... :-)''
 
''Don't take it to the list! There's too much traffic there as it is.''
 
''Here we have another irreconcilable difference between us. I think it is irresponsible '''not''' to maintain traffic between the Wiki and the list; the inevitable outcome of that is that the Wiki and the list evolve distinct linguistic norms. In any case, too late :-) , and I will now shut up about this topic; I've said more than enough'' -- [[User:Nick Nicholas|nitcion]].
 
''I was just commenting on the amount of messages on the list, not on whether there should be cross-traffic.'' OK, sorry about misunderstanding. Did not forward to list after all, anyway...
 
OK, I understand le ka mi xendo. With ce'u, is this le ka mi ce'u xendo? What is the grammar of ce'u, can this work? --xod
 
<nowiki>[Not adding content, only clarification, not adding content, only clarification...]</nowiki> No: ce'u is a sumti like ke'a. -- [[User:Nick Nicholas|nitcion]]
 
''I seem to remember seeing somewhere in the old archives that ''ce'u'' was originally a [[jbocre: PA|PA]]. How did that work?''
 
The original idea was that as su'o da means E(x) and roda means A(x),
 
so ce'u da would mean lambda(X).  Lambda is not really a quantifier,
 
of course, but it does bind a variable.  Under that old usage,
 
self-love (see mailing list) would be "ka ce'u da prami da".
 
This was rejected, on just what grounds I don't recall. --[[jbocre: John Cowan|John Cowan]]
 
So then, always using ce'u, how do I express le ka mi xendo?
 
Well, I've already said how I express it, so you'd best wait for someone else to answer.... -- [[User:Nick Nicholas|nitcion]]
 
If you have to use ''ce'u'' and ''ka'', ''le ka mi zo'u ce'u xendo'' is probably the lesser of evils. nitcion would have you say ''le ka mi no'u ce'u xendo''. Using the recently proposed x2 of ''ka'', it could be ''le ka ce'u xendo kei be mi''. Personally, I doubt that it's really a property that you want here. -- Adam
 
You express it as le nu mi xendo. --[[jbocre: John Cowan|John Cowan]]

Revision as of 16:46, 4 November 2013

I created a very, very experimental Devanagiri orthography. Until I get an image up, I will describe it here:

The syllabic script is abandoned. So are the letters 'ei', 'ai', 'ou' and 'au'. Vowel marks now form dipthongs. The sound for retroflex 'sh' is used for Lojban 'j'. The aspirated stops are dropped. Double line for capitalization, 'aha' for full stop and other symbols unchanged. (I think writing an apostrophe would be pretty hard, though...)

I'm currently making a graphic that will map Lojban to Devanagari, one-to-one, and I'll add a link here soon.

A Devanagari orthography has been discussed on the list. See [1] http://groups.yahoo.com/group/lojban/message/1324, [http://groups.yahoo.com/group/lojban/message/1341

Since] the topic is pretty old there, I think I'll continue the discussion here.

Yes, indeed, the "unwritten vowel" is a schwa, just like Lojban y. However, it would be much more useful in to have it represent Lojban "a" in this orthography. Now, to write a conjuct consonant starting with r that doesn't have an equivalent in an Indic language, the problem is easily solved by "r" sign (like a hook on top) Others aren't so simple.

The suggestion for using ph for Lojban "f" and bh for Lojban "v" was a good one.

Dipthongs, on second thought, would be better written, (also as suggested on the list) using the available dipthongs. The Devanagari diphthong "ai", although beginning with a schwa, can be used for Lojban "ai". Indeed, all diphtongs in Lojban are taken care of save "ii" and "uu". If we were using an orthogaphy like Kannada or Telugu, the solution for this would be quite simple: use the existing letters for long "i" and long "u". Unfortunately, Devanagari has the fomer, but not the latter, so to be consistent, we must use the "initial form" of each -with- the diacritic on the vowel.

Maybe I missed this in the discussion, but when I studied Sanskrit, there certainly was a letter (and a diacritic) for long-u, as well as long-i. What I don't see, though, is a diphthong for oi.(Because /o/ is already a diphthong in Sanskrit, the short form of /au/. pc) As to the other falling diphthongs, would you write them with semivowels w/y instead of vowels? And are we using visarga for apostrophe, as I would think should be obvious? --mi'e mark. Eep, it looks like some of what's said there was by me. I should read it first.

Bap!:: I was thehink about 'ei'. Sorry. Yes, you can write the other falling dipthongs with semivowels. Hm... and I never noticed oi either. What about eu? Same problem. (Exactly; /e/ is short /ai/. pc) And the visarga is logical for the apostrophe. Wouldn't the danda be logical for the full stop? And the puurNaviraama for the comma might work too.

Make c into Lojban c and ch into Lojban tc, and j into Lojban j and jh into Lojban dj, Devanagari "h" into Lojban x, and all you have to worry about is punctuation. Now a full stop could be indicated using "aha", but as I said earlier, the "'" becomes a problem. So does capitalisation and the comma, because none of the existing Devanagari fonts I've seen so far have anything equivalent to those symbols. Any suggestions?

(Was this in the earlier messages? Transcribing English and Arabic and whatever, modern Hindi uses dots creatively: I remember dotted /j/ for /z/ -- and monolinguals saying things like /dijaster/ and /sijors/-- a brand of ciggyboos. pc)


Whoever was working on this, if you could describe this orthography in terms of the Unicode Devanagari code chart at: [2] then a yudit keymap can be made very easily, and then pictures can be made even easier. --Jay Kominek