End vowel for slot?

[TR NS]

Hope it's okay to discuss conlang variations of Lojban on this list. As an armchair conlanger, I can't help but wonder about a few "what-ifs" as I read the CLL. One that keeps coming back to me is the idea that the last vowel of a brivla (is that the right word?) could specify the slot intended, instead of having to to use `se`, `te`, `ve` or `xe`. 

    klama     =>  klama

    se klama  =>  klame

    te klama  =>  klami

    ve klama  =>  klamo

    xe klama  =>  klamu

It just seems a very natural extension of the predicate nature of Lojban's words.

[Jorge Llambías]

That would work for gismu, but not for brivla in general, unless you also drastically reduce the number of available rafsi.

mu'o mi'e xorxes

[TR NS]

Yea, it would effect the rafsi. Not sure how that would work. Don't think it has to be "drastic" though, but it would significant.

[TR NS]

Try that again...

On Monday, August 18, 2014 10:11:49 PM UTC-4, TR NS wrote:

That would work for gismu, but not for brivla in general, unless you also drastically reduce the number of available rafsi.

Yea, it would effect the rafsi. Not sure how that would work. Don't think it has to be "drastic" though, but it would be significant.

 

[MorphemeAddict]

IOW, it might have worked if it had been designed that way originally, but not as things are now. 

stevo

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.
To post to this group, send email to loj...@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Gleki Arxokuna]

Loglan worked this way:

logli = x1 is a Loglan speaker/Loglander.

loglo = x1 is an element/feature/expression of the Loglandic culture.

logle = x1 is an Loglandic place/territory, a region inhabited by Loglanders.

logla = x1 is an element/feature/expression of Loglan.

[.arpis.]

Seems like it would be easy to mishear in a noisy environment.

--
mu'o mi'e .arpis.

[TR NS]

On Tuesday, August 19, 2014 1:56:40 PM UTC-4, .arpis. wrote:

Seems like it would be easy to mishear in a noisy environment.

True, but I've notice that in few other places too. I think context is usually enough to counter that.

[TR NS]

On Tuesday, August 19, 2014 1:50:36 AM UTC-4, la gleki wrote:

Loglan worked this way:

logli = x1 is a Loglan speaker/Loglander.

loglo = x1 is an element/feature/expression of the Loglandic culture.

logle = x1 is an Loglandic place/territory, a region inhabited by Loglanders.

logla = x1 is an element/feature/expression of Loglan.

I did not know. I have the Loglan 1 book, but I haven't read much of it yet compared to the LLC. I'll have to look for that.

Thanks. 

.trans.

[John E Clifford]

Well. this features must be relatively new. i.e., since JCB,  It i not in the Loglan 1 of th 1980s.

[Pierre Abbat]

I know two words that differ in a vowel, not the final vowel, and that are easy
to confuse: In Alice, the Queen is le noltruni'u, and the Duchess is le
noltroni'u.

Pierre
--
When a barnacle settles down, its brain disintegrates.
Já não percebe nada, já não percebe nada.

[MorphemeAddict]

The vowel groupings are in L4 with notations from 1975. They seem to be part of the original language. 

logla, 

log <Loglan, the language>  L-Prim L4 '75 7.4
   (1n) B is an element/feature/expression of Loglan. [B-]  
    (a) Loglan, pert. to the loglan language. 
   lo — (n) Loglan, the mass of such elements. 
   Used in: damlogla;  logcia;  logdia;  logduo;  logmao;  logpli;  logpozfa;  logtaa;  norlogla;  

logle, 

<Loglan, the language>  D-Prim JRK '83 3.6
   (1a) B is an Loglandic place/territory, a region inhabited by Loglanders.  [B-]  

logli, 
<Loglan, the language>  D-Prim L4 '75 1.0
   (1n) B is anLoglan speaker/Loglander.  [B-]  
    (a) Loglan.  
   lo — (n) (the) Loglanders, the mass of such persons.  
   Used in: cnulogli;  

loglo, 
<<Loglan, the language>  D-Prim L4 '75 1.0
   (1n) B is an element/feature/expression of the Loglandic culture.  [B-]  
    (a) Loglan, pert. to the Loglan culture.  
   lo — (n) Loglan, the mass of such elements.  

stevo

[John E Clifford]

Oops, my comments were about the general principle (final vowels for converting place structure), not about the Loglan "example".  Note that the Loglan is, in fact, four primitive predicates whichshare a common central concept and a four-letter chunk, not a single predicate with variations on the final vowel.  There is no common predicate "x1 is a piece of the Loglan language as spoken in area x2 by speakers x3 reflecting cultural traits x4" or so. And, even if there were, this is a unique case and the principle was not generalized beyond this (and a fewe cultural predicates which also occurred in clusters).  These cases are from before the "Great Morphological Revolution" (1980) which tended to disparage such similarity in primitives in the interest of a more controlled derivation process.  And the usual suggestion about final vowels i Loglan was to indicate the number of arguments, a significant fact that was not visible at that stage.

[ianek]

If Lojban was intended to be usable in a noisy environment, it wouldn't include over hundred one-syllable words.

mu'o mi'e ianek

[John E Clifford]

Like most conlangs, Lojban gave virtually no consideration to the noisy room problem -- and has resisted any attempt to deal with the issue since (but then, so has English  "talk louder, repeat yourself, ...")

[TR NS]

On Wednesday, August 20, 2014 5:53:20 PM UTC-4, clifford wrote:

Like most conlangs, Lojban gave virtually no consideration to the noisy room problem -- and has resisted any attempt to deal with the issue since (but then, so has English  "talk louder, repeat yourself, ...")

Well considering lojban has just five pure vowel sounds, there is certainly room for playing with vowel harmonies (https://en.wikipedia.org/wiki/Vowel_harmony).

[John E Clifford]

Not much.  Vowel harmony would wreak havoc with the etymology of primitives, since vowels count as much as consonants in the word-building algorithm.  But, once that issue were solved, the harmony would be lost in derive predicates, since the vowels carry over to the components and changing them to fit some preestablished pattern would interfere with their identification.  Finally, and perhaps most damning, vowel harmony does not affect function words (mostly one syllable) where the problem of noisy channels is the gravest, because of the packed word space and the relatively little use context has in that area (many hard to distinguish words belong to categories that occur in the same frames).  So, vowel harmony does not help nearly enough to be worth pursuing.

[And Rosta]

TR NS, On 21/08/2014 04:08:

> Well considering lojban has just five pure vowel sounds, there is certainly room for playing with vowel harmonies (https://en.wikipedia.org/wiki/Vowel_harmony).

Lojban has seven vowels, whose allophonic ranges exhaust the entirety of vowel space.

--And.

[Gleki Arxokuna]

aeiouy = 6

What is the seventh one? 

[Jorge Llambías]

It's not represented in writing. See:  http://lojban.github.io/cll/3/8/

[Gleki Arxokuna]

o'anai

[Alex Richmond]

This makes me wonder, do the few people that speak Lojban ever tend to skip the last vowel of popular selbri? For instance, do they ever say {pend} instead of {pendo}, or {pram} instead of {prami}?

I would imagine that people who have been speaking/listening awhile might elide them sometimes. {ma cmen do} is one example that I could imagine hearing

[Gleki Arxokuna]

2014-08-24 4:00 GMT+04:00 Alex Richmond <greens...@gmail.com>:

This makes me wonder, do the few people that speak Lojban ever tend to skip the last vowel of popular selbri? For instance, do they ever say {pend} instead of {pendo}, or {pram} instead of {prami}?

I would imagine that people who have been speaking/listening awhile might elide them sometimes. {ma cmen do} is one example that I could imagine hearing

http://mw.lojban.org/index.php?title=gism
 

[Alex Richmond]

On Saturday, August 23, 2014 9:40:19 PM UTC-7, la gleki wrote:

2014-08-24 4:00 GMT+04:00 Alex Richmond <greens...@gmail.com>:

This makes me wonder, do the few people that speak Lojban ever tend to skip the last vowel of popular selbri? For instance, do they ever say {pend} instead of {pendo}, or {pram} instead of {prami}?

I would imagine that people who have been speaking/listening awhile might elide them sometimes. {ma cmen do} is one example that I could imagine hearing

http://mw.lojban.org/index.php?title=gism

I do like having the final letter, don't get me wrong. It's certainly helped me with memorization. But I could see people eliding it while speaking. Is that frowned upon by the community or do people not really care? 

[Pierre Abbat]

On Saturday, August 23, 2014 17:00:46 Alex Richmond wrote:
> This makes me wonder, do the few people that speak Lojban ever tend to skip
> the last vowel of popular selbri? For instance, do they ever say {pend}
> instead of {pendo}, or {pram} instead of {prami}?

The only time I do that is in the song "There was an old woman who swallowed a
fly", and even then you can tell that ".i mutce se xrani" is not ".i mutco se
xranu".

> I would imagine that people who have been speaking/listening awhile might
> elide them sometimes. {ma cmen do} is one example that I could imagine
> hearing

I'd answer "macmendo ki'a".

Pierre
--
Don't buy a French car in Holland. It may be a citroen.

[TR NS]

On Sunday, August 24, 2014 12:40:19 AM UTC-4, la gleki wrote:

http://mw.lojban.org/index.php?title=gism

"However, they are there from the gismu creation process to provide mnemonic assistance. How quickly would you learn banguwithout the angu of the English (pronounced languedj)? Would a Chinese speaker learn vanju as easily without the chinese "jiu" (pronounced [dziou]) being the entire second syllable?"

I can't really take this explanation seriously. I personally think this one place Logan/Lojban failed. Creating words by statistically mixing the letters of words from other languages just ends up creating words that on the whole no one recognizes. 

[Gleki Arxokuna]

.ie

However, what are the alternatives? 

[MorphemeAddict]

I've said this for years. 

stevo

--

[MorphemeAddict]

A priori creation is one. It could be much better than what was actually done. 

stevo

[Gleki Arxokuna]

2014-08-24 20:21 GMT+04:00 MorphemeAddict <lyt...@gmail.com>:

A priori creation is one. It could be much better than what was actually done. 

Then show one.

[John E Clifford]

Well, "better" is tricky, since we never tested  how "good" the Loglan/Lojban system was.  But a priori creations would help with issues that the present system ignores, like better word separation (the noisy channel problem or just finding more rafsi) or more equitable phoneme distribution (for some aesthetic reason).  But it is a real pain to do, whereas the present system gave a good starting point that got the creators most of the way through the list.  Of course, a program could be devised to generate a field of  words of predetermined form, satisfying the distribution criteria and then assign them willy-nilly (or even under some criteria) to a list (from somewhere -- maybe Loglan, after all) of meanings. I'm not sure that would be significantly harder to learn the the gismu list (and, indeed, it might well provide a real test for the claim that the present list is easy to learn -- against everyone's experiences).

[And Rosta]

On 24 Aug 2014 17:32, "Gleki Arxokuna" <gleki.is...@gmail.com> wrote:
>
>
>
>
> 2014-08-24 20:21 GMT+04:00 MorphemeAddict <lyt...@gmail.com>:
>
>> A priori creation is one. It could be much better than what was actually done. 
>
> Then show one.

I hope you don't mean to imply that unless Stevo shows one his claim is empty. (I think that would be foolish.)

Perhaps a discussion about better designs than Lojban's would be better on Conlang or Engelang lists, and I invite you to pose your questions there about better designs.

--And.

[TR NS]

On Sunday, August 24, 2014 12:22:11 PM UTC-4, stevo wrote:

A priori creation is one. It could be much better than what was actually done. 

stevo

A priori can be difficult though. Yet it's not too hard to do some word derivations using transformation rules, e.g. deriving `library` from the word for `book`.

Personally, I would have just borrowed words whole-clothe from different languages without trying to mix them. I also would not have given Chinese the full weight of its population in selecting words simply b/c it is a rather geographically locked language and also b/c it is tonal.

[MorphemeAddict]

I have a relex of Lojban with completely a priori gismu. It has the advantage that every gismu has just one short rafsi and one long rafsi. 

How do you want me to show it? I'm not going to post them all here, and I don't have a web site to put them on. 

stevo

[Gleki Arxokuna]

This is not even related to loglangs. I wish there was an active a priori conlang so that we can see that it can be successful.

do klingon and quenya meet your requirement of a priori languages?

can a language distant to Chinese and SAE languages be successful? How many people from China and SAE countries (except Latin America itself) are learning Aymara/Quechua these days?

[TR NS]

On Monday, August 25, 2014 11:17:44 AM UTC-4, stevo wrote:

I have a relex of Lojban with completely a priori gismu. It has the advantage that every gismu has just one short rafsi and one long rafsi. 

How do you want me to show it? I'm not going to post them all here, and I don't have a web site to put them on. 

stevo

You do? That would be cool to see. I can put it up on a Github project if you'd like.

[Jacob Errington]

.i la stevo cu cusku di'e

I have a relex of Lojban with completely a priori gismu. It has the advantage that every gismu has just one short rafsi and one long rafsi.

If such a relex exists, then with it we can eliminate the rafsi/gismu distinction.

Let's suppose we drop all the gismu, and keep only their associated rafsi. For CVC and CCV rafsi, we suffix -no or -mo (for those CVC that end in 'n'), and we can make the gismu forms, and for the CVV, we can't make true gismu forms, but we can construct a CV magic word that will bind the previous cmavo and produce something with the grammar of a selbri. Of course, that would just be in the "standalone gismu" cases. For lujvo, we would proceed as normal.

.i mi'e la tsani mu'o

[Jorge Llambías]

Or you could still produce a true gismu form by infixing for example CVV -> CV-CC-V  (you'd need one CC for monosyllabic CVV and a different one for disyllabic ones, and you couldn't use "Cn" or "nm" as the infix so as not to conflict with the CVC and CCV forms).

[MorphemeAddict]

All my gismu (noglo) are CVCCV, where the initial CV is a semantic classifier and the rest is the unique rafsi. 

Two examples to illustrate how I form lujvo (nonla).

1) From "brajmadegycalku" (barda jamfu degji calku), 'big toe nail', 

I get su-njo bi-bla bi-gdo bu-dza => budza-gdo-bla-njo. 

2) From "rilpemsa'a" (rilti pemci sanga), 'perform rap music', 

I get no-bre no-xma pe-jma => pejma-xma-bre. 

stevo

--

[Jacob Errington]

.i la stevo di'e cusku

All my gismu (noglo) are CVCCV, where the initial CV is a semantic classifier and the rest is the unique rafsi. 

Two examples to illustrate how I form lujvo (nonla).

1) From "brajmadegycalku" (barda jamfu degji calku), 'big toe nail', 

I get su-njo bi-bla bi-gdo bu-dza => budza-gdo-bla-njo. 

2) From "rilpemsa'a" (rilti pemci sanga), 'perform rap music', 

I get no-bre no-xma pe-jma => pejma-xma-bre. 

stevo

How could I have expected anything less elegant than that? .u'i

For the first lujvo, wouldn't it be possible to write "dza-gdo-bla-njo", since dz is initial-permissible? And for the latter, wouldn't it be "pejmy-xma-bre"?
Unless you have changed the lujvo rules or the morphology, in which case this is more than just a relex.

[MorphemeAddict]

It's a relex in the broad sense of a one-to-one correspondence between the parts of words. 

The first syllable of a word is its semantic classifier, which for lujvo is usually the same as for the main word of the tanru it's based on. 

I changed the lujvo rules and the morphology (phonology too, but that didn't show in the example). Whether it's still self-segregating or not, I don't know (and don't much care). 

It helped alleviate my frustration with Lojban gismu. 

stevo

[.arpis.]

On Mon, Aug 25, 2014 at 11:17 AM, MorphemeAddict <lyt...@gmail.com> wrote:

I have a relex of Lojban with completely a priori gismu. It has the advantage that every gismu has just one short rafsi and one long rafsi. 

How do you want me to show it? I'm not going to post them all here, and I don't have a web site to put them on. 

stevo

I'd love to see it too. Pastebin or Google Docs come to mind as minimal solutions; neocities.org or github come to mind as hosting options too, and I'm sure that in a pinch, you can throw them on to a page in one of the wikis.

--
mu'o mi'e .arpis.

[Pierre Abbat]

On Monday, August 25, 2014 22:43:14 Jacob Errington wrote:
> For the first lujvo, wouldn't it be possible to write "dza-gdo-bla-njo",
> since dz is initial-permissible? And for the latter, wouldn't it be
> "pejmy-xma-bre"?

"gd", "nj", and "xm" are not permissible initials, so they cannot appear at
the beginning of rafsi.

Pierre
--
When a barnacle settles down, its brain disintegrates.
Já não percebe nada, já não percebe nada.

[MorphemeAddict]

There is always a semantic classifier (CV-) at the beginning, so yes, they can appear at the beginning of my rafsi. 

stevo