Talk:morphology: Rules of formation
I have written the rules for generating lojban words, as opposed to the algorithm for breaking a speech stream into words, here:
Please check that I didn't miss anything.
(I didn't go here into the details of phonotactics: permissible clusters and such. Those can be found in: .)
mu'o mi'e xorxes
> > Re: Morphology: Rules of formation > > I have written the rules for generating lojban words, as opposed to the > algorithm for breaking a speech stream into words, here:
> > Please check that I didn't miss anything. > > (I didn't go here into the details of phonotactics: permissible clusters and > such. Those can be found in:
> .) > > mu'o mi'e xorxes > > > > > >
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Sports - Sign up for Fantasy Baseball. http://baseball.fantasysports.yahoo.com/
>It can't contain la, lai, la'i, doi unless preceded by a >consonant and/or followed by an h.
What about {laus} and {ala'um}? The rule should be "...and/or followed by a non-consonant".
phma -- Without glasses, I can't even distinguish smells... -Les Perles de la médecine
> >It can't contain la, lai, la'i, doi unless preceded by a > >consonant and/or followed by an h. > > What about {laus} and {ala'um}?
They're OK.
> The rule should be "...and/or followed by a > non-consonant".
Right.
(I'm planning to remove the doi/la/lai/la'i stuff from my proposal anyway. But I will keep it commented out so it can eventually be restored if the change is not approved.)
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - Find what you need with new enhanced search. http://info.mail.yahoo.com/mail_250
>Between two cmavo of form CVCy (without an intervening pause) and a brivla or CVV cmavo.
What about two cmavo of form CVCy and a longer cmavo?
phma -- My monthly periods happen once per year. -Les Perles de la médecine
> >Between two cmavo of form CVCy (without an intervening pause) and a brivla > or > CVV cmavo. > > What about two cmavo of form CVCy and a longer cmavo?
CVCyCVVV breaks as {CV Cy CVVV}, no need to pause, is there?
The only reason CVV cmavo are special is that they look like rafsi, but longer (or shorter) cmavo don't.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail
At 12:13 PM 2/28/05 -0800, Jorge wrote:
>--- Pierre Abbat wrote: > > > >Between two cmavo of form CVCy (without an intervening pause) and a brivla > > or > > CVV cmavo. > > > > What about two cmavo of form CVCy and a longer cmavo? > >CVCyCVVV breaks as {CV Cy CVVV}, no need to pause, is there? > >The only reason CVV cmavo are special is that they look >like rafsi, but longer (or shorter) cmavo don't. > >mu'o mi'e xorxes
How about "resytosmAbru"? re + sy + to (a CV, not a CVV) + smabru? Or lujvo resytosmabru (as current language has it)?
mi'e noras [email protected] Nora LeChevalier
On Wednesday 02 March 2005 21:13, Nora LeChevalier wrote: > At 12:13 PM 2/28/05 -0800, Jorge wrote: > >--- Pierre Abbat wrote: > > > >Between two cmavo of form CVCy (without an intervening pause) and a > > > > brivla > > > > > > or > > > CVV cmavo. > > > > > > What about two cmavo of form CVCy and a longer cmavo? > > > >CVCyCVVV breaks as {CV Cy CVVV}, no need to pause, is there? > > > >The only reason CVV cmavo are special is that they look > >like rafsi, but longer (or shorter) cmavo don't. > > > >mu'o mi'e xorxes > > How about "resytosmAbru"? re + sy + to (a CV, not a CVV) + smabru? Or > lujvo resytosmabru (as current language has it)?
Good example. The rule I follow in valfendi is that a pause is required between a cmavo of form Cy and the next word which contains CCV, CVV, or CV'V. So if you wanted to say {re sy to smabru}, you have to pause between {sy} and {to} or between {to} and {smabru}. I normally put the pause right after the Cy unless the next word is {bu}.
phma -- Without glasses, I can't even distinguish smells... -Les Perles de la médecine
> At 12:13 PM 2/28/05 -0800, Jorge wrote: > > > > >Between two cmavo of form CVCy (without an intervening pause) and a > brivla > > > or > > > CVV cmavo. > > How about "resytosmAbru"? re + sy + to (a CV, not a CVV) + smabru? Or > lujvo resytosmabru (as current language has it)?
It's a lujvo, as in the current language. I should have said "a string of one or more rafsi" instead of "a brivla", and that will in fact also cover CVV cmavo. I will fix that in the wording of the rules of formation, thanks. The relevant PEG rule is:
CVCy-lujvo <- CVC-rafsi y h? initial-rafsi* brivla-core / stressed-CVC-rafsi y short-final-rafsi
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web http://birthday.yahoo.com/netrospective/
> The rule I follow in valfendi is that a pause is required > between a cmavo of form Cy and the next word which contains CCV, CVV, or > CV'V.
That requires too much though. For example, you don't allow {.ibymai} without pause, but that can't be anything but {.i by mai}.
The simple rule is to require a pause *always* after Cy. If we relax the rule, then let's relax it as much as possible, not to some arbitrary intermediate point.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web http://birthday.yahoo.com/netrospective/
On Sun, 27 Feb 2005 09:52:47 -0800, [email protected] wrote: > *You must pause in the following cases: snip > **Before and after a closing zoi-delimiter word.
Ah! I didn't remember the "after" bit about the closing zoi-delimiter.
> **If the first rafsi is a CVC-rafsi, and the consonant plus what follows is a lujvo, then the CVC-rafsi must have a y-hyphen.
I presume that "the" consonant is the second one (the capitalised one in "cvC")?
mu'o mi'e .filip. --
Philip Newton
> > *You must pause in the following cases: > snip > > **Before and after a closing zoi-delimiter word. > > Ah! I didn't remember the "after" bit about the closing zoi-delimiter.
The delimiter pauses are not really always required by the morphology. As long as the delimiters are not part of a non-lojban-word the morphology will detect them. For example, {zoizoibrOdazoivAlsi} will be parsed as {zoi zoi broda zoi valsi}. All the morphology requires is that the delimiters don't mix with a truly non-lojban stuff. For example {zoizoibrokdazoi} is a single non-lojban word, because it can't be fully separated into Lojban words. Requiring pauses after the opening delimiter and before and after the closing delimiter is overkill, in particular the one after the closing delimiter is never really needed.
> > **If the first rafsi is a CVC-rafsi, and the consonant plus what follows is > a lujvo, then the CVC-rafsi must have a y-hyphen. > > I presume that "the" consonant is the second one (the capitalised one in > "cvC")?
Yes, fixed.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail
On Sun, Feb 27, 2005 at 07:29:06PM +0100, Philip Newton wrote: > On Sun, 27 Feb 2005 09:52:47 -0800, [email protected] > wrote: > > *You must pause in the following cases: > snip > > **Before and after a closing zoi-delimiter word. > > Ah! I didn't remember the "after" bit about the closing > zoi-delimiter.
I don't either. I think he means after the opening delimiter and before the closing one.
-Robin
> On Sun, Feb 27, 2005 at 07:29:06PM +0100, Philip Newton wrote: > > On Sun, 27 Feb 2005 09:52:47 -0800, [email protected] > > wrote: > > > *You must pause in the following cases: > > snip > > > **Before and after a closing zoi-delimiter word. > > > > Ah! I didn't remember the "after" bit about the closing > > zoi-delimiter. > > I don't either. I think he means after the opening delimiter and > before the closing one.
No, I really meant it as I wrote it. I had the rule for the opening delimiter separately. That is what I thought was the outcome of the discussion that included this post:
but I didn't manage to find the rest of the thread once I found that post, so I can't check if I'm remembering correctly. You said something that you were going to modify the zoi delimiters so that they dealt with spaces differently.
In any case, I have changed the rule now, because zoi delimiters don't really pertain to the morphology. The only rule that matters for the morphology with respect to this is:
"You must pause before and after any non-lojban-word."
If you only quote lojban words with zoi, there need not be any pause for the parser to divide the words properly.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - Easier than ever with enhanced search. Learn more. http://info.mail.yahoo.com/mail_250
On Sun, Feb 27, 2005 at 05:48:00PM -0800, Jorge Llamb?as wrote: > No, I really meant it as I wrote it. I had the rule for the > opening delimiter separately. That is what I thought was the > outcome of the discussion that included this post: > > > > but I didn't manage to find the rest of the thread once I found > that post, so I can't check if I'm remembering correctly.
> You said something that you were going to modify the zoi > delimiters so that they dealt with spaces differently.
Already done. They require a pause/space after the opening delimiter and before the closing delimiter.
The forums appear to be a bit upset, but see:
(Threaded mode isn't working; I'll go fix this soon)
> In any case, I have changed the rule now, because zoi delimiters > don't really pertain to the morphology.
Agreed.
-Robin
On Sun, Feb 27, 2005 at 06:10:12PM -0800, Robin Lee Powell wrote: > On Sun, Feb 27, 2005 at 05:48:00PM -0800, Jorge Llamb?as wrote: > > No, I really meant it as I wrote it. I had the rule for the > > opening delimiter separately. That is what I thought was the > > outcome of the discussion that included this post: > > > > > > > > but I didn't manage to find the rest of the thread once I found > > that post, so I can't check if I'm remembering correctly. > > > You said something that you were going to modify the zoi > > delimiters so that they dealt with spaces differently. > > Already done. They require a pause/space after the opening > delimiter and before the closing delimiter. > > The forums appear to be a bit upset, but see: > > http://www.lojban.org/tiki/tiki-view_forum_thread.php?forumId=1&comments_parentId=3938&comments_maxComments=1&comments_style=commentStyle_plain > > (Threaded mode isn't working; I'll go fix this soon)
- DAMN* that was a pain in the ass. :-(
Fixed.
Full thread is at
It's three pages long. For xorxes' agreeing with me on the zoi delimiter issue, search for "Yes, I suppose it's not that hard,".
-Robin
> On Sun, Feb 27, 2005 at 06:10:12PM -0800, Robin Lee Powell wrote: > > On Sun, Feb 27, 2005 at 05:48:00PM -0800, Jorge Llamb?as wrote: > > > You said something that you were going to modify the zoi > > > delimiters so that they dealt with spaces differently. > > > > Already done. They require a pause/space after the opening > > delimiter and before the closing delimiter.
Are you sure? I don't see any changes.
I think we should go either with the morphological break (as I think it is set up now) so that neither {to} nor {go} can be used to quote "Togo", or require pauses both sides of the closing delimiter, so that both {to} and {go} could be used. The middle-of-the-road rule that requires a pause only before the final delimiter but not after it means you can use {go} but not {to} as delimiters to quote "Togo". That doesn't seem very reasonable.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - now with 250MB free storage. Learn more. http://info.mail.yahoo.com/mail_250
> On Sun, Feb 27, 2005 at 06:10:12PM -0800, Robin Lee Powell wrote: > > On Sun, Feb 27, 2005 at 05:48:00PM -0800, Jorge Llamb?as wrote: > > > You said something that you were going to modify the zoi > > > delimiters so that they dealt with spaces differently. > > > > Already done. They require a pause/space after the opening > > delimiter and before the closing delimiter.
Are you sure? I don't see any changes.
I think we should go either with the morphological break (as I think it is set up now) so that neither {to} nor {go} can be used to quote "Togo", or require pauses both sides of the closing delimiter, so that both {to} and {go} could be used. The middle-of-the-road rule that requires a pause only before the final delimiter but not after it means you can use {go} but not {to} as delimiters to quote "Togo". That doesn't seem very reasonable.
mu'o mi'e xorxes
'''''''''''''''''''''''''''''''''''''''''''__ Do you Yahoo!? Yahoo! Mail - now with 250MB free storage. Learn more. http://info.mail.yahoo.com/mail_250