Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

anticipating language change

6 views
Skip to first unread message

Matthew Montchalin

unread,
Apr 27, 2003, 5:07:25 PM4/27/03
to
When people concoct their artificial languages, do they take into
account the likelihood of language change? Should an artificial
language be resistant to language change?

Paul O. BARTLETT

unread,
Apr 27, 2003, 5:18:15 PM4/27/03
to

Are you talking about an artificial language constructed for fun,
hobby, art, literature, and the like? Or are you talking about a
language constructed to be an actual international auxiliary language?
The is some debate in the IAL community about how much (and even how)
a conIAL should change.

--
Paul Bartlett
bartlett at smart.net
PGP key info in message headers

Klaus Scholl

unread,
Apr 27, 2003, 5:26:20 PM4/27/03
to
| When people concoct their artificial languages, do they take into
| account the likelihood of language change? Should an artificial
| language be resistant to language change?

With Language 'Change' you mean chaotic, planless, Change of the
Spelling and Pronounciation of Words,
and a chaotic, planless Change of Meanings, like we have it in all
natural Langs?

An Artlanger will tell you probably: changes are the Livelyness of
Languages, its wonderful its natural.
A Auxlanger will tell you probably: no, Naturality is bad, its
illogical, its stupid.

Im an Auxlanger, so you know my Opinion.
I am not a Person that thinks: natural=good & artifical=bad,
i think: logical=good & illogical=bad.
Natural Languages are all bad. They *keep* Mankind stupid with its
Chaoticness.

bye
Klaus


Garth Wallace

unread,
Apr 27, 2003, 6:48:39 PM4/27/03
to
Klaus Scholl wrote:
>
> I am not a Person that thinks: natural=good & artifical=bad,

and then you say...

> Natural Languages are all bad.

Make up your mind.

--
Happy denizen of the Nightstar IRC Network
Webcomics discussion: irc://us.nightstar.net/webcomics
Progressive rock chat: irc://us.nightstar.net/progrock

Automort

unread,
Apr 27, 2003, 6:58:05 PM4/27/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>Should an artificial
>language be resistant to language change?

If it isn't used it won't change. If it's used for any length of time by a
large number of people it will change.
A dead language lasts forever, a living one does not.

>When people concoct their artificial languages, do they take into
>account the likelihood of language change?

How would one go about doing this except to specify how to derive new words?

Garth Wallace

unread,
Apr 27, 2003, 11:53:13 PM4/27/03
to
Automort wrote:
>>From: Matthew Montchalin mmon...@OregonVOS.net
>
>
>>Should an artificial
>>language be resistant to language change?
>
>
> If it isn't used it won't change. If it's used for any length of time by a
> large number of people it will change.
> A dead language lasts forever, a living one does not.

Not sure if that's exactly true. As long as it's taught only as a second
language, using books as a source, it probably won't change. If it
becomes as community's first language, however, you can expect it to
change over time.

Cellus Purfluxius

unread,
Apr 28, 2003, 4:58:27 AM4/28/03
to
Automort wrote:

It is not difficult to show that Esperanto has changed over the years
and even Interlingua, even though it has fewer speakers than has Esperanto.

Cellus P.

Automort

unread,
Apr 28, 2003, 10:30:32 AM4/28/03
to
>From: Garth Wallace gwa...@despammed.com

>As long as it's taught only as a second
>language, using books as a source, it probably won't change. If it
>becomes as community's first language, however, you can expect it to
>change over time.

It will have to change if it's a first language. In fact, if it's a widely used
second language it will change.

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 11:41:44 AM4/28/03
to
On Sun, 27 Apr 2003 23:26:20 +0200, "Klaus Scholl"
<Kl...@sensualium.cjb.net> wrote:

> | When people concoct their artificial languages, do they take into
> | account the likelihood of language change? Should an artificial
> | language be resistant to language change?
>
> With Language 'Change' you mean chaotic, planless, Change of the
> Spelling and Pronounciation of Words,
> and a chaotic, planless Change of Meanings, like we have it in all
> natural Langs?

First we need to realize something about language that often is ingored
by enthusiasts. Language is a *tool* used for the purpose of
communicating idea between humans. All natural languages are capable of
doing that, and when a new linguistic need is required, the language
will adapt to accomodate the needs whether by borrowing words from other
languages or reinventing uses for old words.

> An Artlanger will tell you probably: changes are the Livelyness of
> Languages, its wonderful its natural.
> A Auxlanger will tell you probably: no, Naturality is bad, its
> illogical, its stupid.

Well, they both have different goals in mind. Artlangers are out to
create an atmosphere for literary purposes. Variations and illogic
demonstrate something about the culture.

An auxlanger is creating something practical that's designed to be used
by humans for communicating. As with all language, once it becomes used
it will take on a life of its own. Standardization is important though
to help keep the general direction so the language doesn't fragment into
so many dialects which given enough time will eventually become their
own languages.


> Im an Auxlanger, so you know my Opinion.
> I am not a Person that thinks: natural=good & artifical=bad,
> i think: logical=good & illogical=bad.
> Natural Languages are all bad. They *keep* Mankind stupid with its
> Chaoticness.

Well, illogic will exist in language as long as it exists in people's
minds. I don't see this changing for a long time considering that the
typical mindless citizen has no understanding or interest in reality.
They'd rather rot their lives away watching mindless garbage on
television.

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 2:39:21 PM4/28/03
to

I would expect any artificial language to undergo change as soon as a
significant amount of people start using it beyond control of the
"authority" that created it.

What I'd like to know (which is why I was inquiring about those native
Esperanto speakers in a previous post) is are there any specific changes
that were/are taking place. Or more to the point, are there any
linguistic habits that tend to be universal? I am definitely not a
professional linguist, but have made a few observations of my own and
believe that keeping these habits in mind, it would be possible to
design a language that is highly resistant to change.

Specific things I've been seeing:

Tendency to form CV syllables. A good example of this is modern
French where the spellings still indicate those "lost" consonants at the
end. Italian is leaning in this direction too with vowels at the end of
almost every word. Much different from their common ancestor, Latin.
On the other hand spoken Japanese seems to do just the opposite by
dropping the final vowel.

Consonants clusters (where they exist) seem to have a tendency
to assimlate given enough time.

Tendency towards more isolating grammar, especially as languages
increase in popularity. Examples (English, Mandarin, Romance
languages, etc.)

Tendency to voice consonants between vowels.


Anyone out notice anything different? Any professional studies done in
this area?

Noé Falzon

unread,
Apr 28, 2003, 3:43:05 PM4/28/03
to
Klaus Scholl <Kl...@sensualium.cjb.net> wrote:

> Natural Languages are all bad. They *keep* Mankind stupid with its
> Chaoticness.

Well said !
But I'm not sure it's due to the language...
Noé
--
"Je ne deteste que les bourreaux" -- Albert Camus

Pour m'écrire, veuillez enlever PASDEPUB de mon adresse ;P

Automort

unread,
Apr 28, 2003, 3:46:54 PM4/28/03
to
>From: PASDEPUBn...@tiscali.fr (=?ISO-8859-1?Q?No=E9_Falzon?=)

>> Natural Languages are all bad. They *keep* Mankind stupid with its
>> Chaoticness.
>
>Well said !
>But I'm not sure it's due to the language...
>Noé

No, it's heredity and education.

Paul O. BARTLETT

unread,
Apr 28, 2003, 6:00:43 PM4/28/03
to
On Mon, 28 Apr 2003, Dana Nutter wrote (excerpted):

> I am definitely not a
> professional linguist, but have made a few observations of my own and
> believe that keeping these habits in mind, it would be possible to
> design a language that is highly resistant to change.
>
> Specific things I've been seeing:
>

> Tendency to form CV syllables. [...]


>
> Consonants clusters (where they exist) seem to have a tendency
> to assimlate given enough time.

Hmmm. The English you are writing in seems to go against your
observations. English still abounds in closed syllables and lots
of consonantal clusters, and I see no tendency to eliminate either.

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 6:02:18 PM4/28/03
to
On Mon, 28 Apr 2003 21:43:05 +0200, PASDEPUBn...@tiscali.fr (Noé
Falzon) wrote:

> Klaus Scholl <Kl...@sensualium.cjb.net> wrote:
>
> > Natural Languages are all bad. They *keep* Mankind stupid with its
> > Chaoticness.
>
> Well said !
> But I'm not sure it's due to the language...
> Noé

Exactly! I don't think it's a cause, but a symptom.


Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 6:34:51 PM4/28/03
to

>
> Hmmm. The English you are writing in seems to go against your
> observations. English still abounds in closed syllables and lots
> of consonantal clusters, and I see no tendency to eliminate either.

Written English, no! But in spoken English there are plenty of examples
such as final /t/ or /d/ fading into /?/, "inter-" tends toward
"inner-" when spoken, etc. Words like "can" and "can't" are often (or
is that "offen") indistiguishable unless the final /t/ is stressed.
What about the total disappearance of final "gh": night, bright, light,
thigh, eight, etc? How about the dropping of "r" at the end of
syllables as with most British dialacts, and even some here in North
America? Then there's final /l/ which will often turn into a rounded
vowel like /u/ or /o/. etc...

Paul O. BARTLETT

unread,
Apr 28, 2003, 6:43:43 PM4/28/03
to
On Mon, 28 Apr 2003, Dana Nutter {post responses only to the list} wrote:

> > Hmmm. The English you are writing in seems to go against your
> > observations. English still abounds in closed syllables and lots
> > of consonantal clusters, and I see no tendency to eliminate either.
>
> Written English, no! But in spoken English there are plenty of examples

> [...]

About all I can say is that the (dialect of American) English *I*
speak still abounds in closed syllables and consonant clusters. (And
post-vocalic /r/ is alive and well.)

Automort

unread,
Apr 28, 2003, 6:51:31 PM4/28/03
to
>rom: Dana Nutter {post responses only to the list} ab...@nutter.net

>English still abounds in closed syllables and lots
>> of consonantal clusters, and I see no tendency to eliminate either.
>
>Written English, no! But in spoken English there are plenty of examples
>

Maybe a person could design a language with instructions on forming new words
and instructions that every 100 years the spelling be revised to keep up with
actual pronounciation, with the sounds assigned the letters never to change.

Matthew Montchalin

unread,
Apr 28, 2003, 7:47:18 PM4/28/03
to
|>Should an artificial language be resistant to language change?

On 27 Apr 2003, Automort wrote:
|If it isn't used it won't change. If it's used for any length of time
|by a large number of people it will change.

But will it change in a predictable manner?

|A dead language lasts forever, a living one does not.
|
|>When people concoct their artificial languages, do they take into
|>account the likelihood of language change?
|
|How would one go about doing this except to specify how to derive
|new words?

Since no human is capable of absolutely perfect hearing, let alone
perfect reproduction of the sounds that others speak, there is
going to be imperfect transmission. Noise will creep in. It will
creep in somehow, somewhere, but it is always going to be compensated
for, either by reliance on some other sound, or on some particular
grammatical feature that will Do The Job.

Matthew Montchalin

unread,
Apr 28, 2003, 7:50:16 PM4/28/03
to
On Mon, 28 Apr 2003, Cellus Purfluxius wrote:
|>>When people concoct their artificial languages, do they take into
|>>account the likelihood of language change?
|>>
|>>
|>
|>How would one go about doing this except to specify how to derive new
|>words?
|>
|>
|It is not difficult to show that Esperanto has changed over the years
|and even Interlingua, even though it has fewer speakers than has Esperanto.

How about Volapuek, as a case in point? Most Americans can't pronounce
the umlaut sounds. Sure, they can figure out how to emphasize the last
'important' syllable of a word, but there is going to be a fair amount
of 'drift' simply for the set of sounds that they haven't had any
exposure to. As languages go, the grammatical structure behind Volapuek
is going to be around longer than the overall root vocabulary is.

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 8:37:25 PM4/28/03
to
On Mon, 28 Apr 2003 18:43:43 -0400, "Paul O. BARTLETT"
<bart...@smart.net> wrote:

> On Mon, 28 Apr 2003, Dana Nutter {post responses only to the list} wrote:
>
> > > Hmmm. The English you are writing in seems to go against your
> > > observations. English still abounds in closed syllables and lots
> > > of consonantal clusters, and I see no tendency to eliminate either.
> >
> > Written English, no! But in spoken English there are plenty of examples
> > [...]
>
> About all I can say is that the (dialect of American) English *I*
> speak still abounds in closed syllables and consonant clusters. (And
> post-vocalic /r/ is alive and well.)

They still exist in English, but will they in 100 years. How about 500
years? Which direction is it headed? Especially in a world where
English is still expanding to different areas and cultures where it's
destined to undergo some evolutionary changes influenced by the habits
of its new speakers.

I too speak English with my final /r/'s but I have lived in some areas
that don't. The habit of dropping that /r/ is easy to pick up. I don't
see any cases of people from NYC or Boston moving elsewhere and
starting to pronounce them unless it's at a very young age.

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 9:21:21 PM4/28/03
to
On 28 Apr 2003 22:51:31 GMT, auto...@aol.com (Automort) wrote:
> Maybe a person could design a language with instructions on forming new words
> and instructions that every 100 years the spelling be revised to keep up with
> actual pronounciation, with the sounds assigned the letters never to change.


Well, that's something like what I'm doing. There will be a standard
set for the creation of new words. I'd rather just create a simple
enough phonology to avoid any need for spelling reform. Of course the
language will tend to go its own direction once in use.

I'm using a simple phonology: / p t k f s x m n N j w i e a o u @ /
which will have a generous inventory of allophones to help accomodate
those who may have trouble with any of these. The phonemes are written
in the Roman alphabet respectively as ( P T K F S H M N Y J W I E A
O U X ) I will have a website up for this soon. It's mostly complete
but I still have a couple of design decisions to make first.

Spelling reform is nothing new. English seems to be the language most
resistant to it despite the fact it was one of the first to have a
standardized orthography. Many languages undergo such reforms, and some
even have some authority, or "language police", attepting to control
the direction of the language. The French are notorious for attempting
to keep the English influence from their lexicon. Iceland has worked to
preserve its language for several hundred years now. So even "natural"
languages have some "artificial" influence. German has seen some minor
reforms in modern times. Russian was reformed in 1917. And finally
even Chinese saw a reform of their written language to what is now known
as "simplified" writing, although it's still logographic.

Of course I'm sure most here have seen, and have opinions on,
orthographic reforms for Esperanto. Personally, I prefer: cx, gx, hx,
jx, sx, & w since it causes the least confusion. These are not reforms
needed because of the language evolving though but rather the result of
a poor design from the start. I doubt there were many typewriters back
thes with H^ on them!

Dana Nutter {post responses only to the list}

unread,
Apr 28, 2003, 9:28:51 PM4/28/03
to

Working on my current (still unnamed) project, I'm toying with some
different ideas on how to handle verb modals and verbal aspects. The
major questions are:

What are the most common modes in natural language?

I'd like to see examples of different methods of handling these:
affixes, adverbs, etc.

The same for aspects. Any good example from natural languages?


Right now, I'm leaning toward using adverbs for both.

Garth Wallace

unread,
Apr 28, 2003, 9:34:17 PM4/28/03
to
Dana Nutter {post responses only to the list} wrote:
> On Sun, 27 Apr 2003 14:07:25 -0700, Matthew Montchalin
> <mmon...@OregonVOS.net> wrote:
>
>
>>When people concoct their artificial languages, do they take into
>>account the likelihood of language change? Should an artificial
>>language be resistant to language change?
>
>
> I would expect any artificial language to undergo change as soon as a
> significant amount of people start using it beyond control of the
> "authority" that created it.
>
> What I'd like to know (which is why I was inquiring about those native
> Esperanto speakers in a previous post) is are there any specific changes
> that were/are taking place. Or more to the point, are there any
> linguistic habits that tend to be universal? I am definitely not a
> professional linguist, but have made a few observations of my own and
> believe that keeping these habits in mind, it would be possible to
> design a language that is highly resistant to change.

Nasalization tends to spread to adjacent phonemes, IIRC.

Automort

unread,
Apr 28, 2003, 9:43:40 PM4/28/03
to
>From: Dana Nutter {post responses only to the list} ab...@nutter.net

>I'd rather just create a simple
>enough phonology to avoid any need for spelling reform.

Sooner or later, pronounciation will change.

Automort

unread,
Apr 28, 2003, 9:44:46 PM4/28/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>But will it change in a predictable manner?

Try to predict the changes, and not allow your predictions to be read for, say,
300 years. You won't know if you predicted correctly, but someone might.

Crandadk

unread,
Apr 29, 2003, 3:22:50 AM4/29/03
to
>When people concoct their artificial languages, do they take into
>account the likelihood of language change? Should an artificial
>language be resistant to language change?

Some changes in pronunciation happen when people use easier pronunciations of
words they find difficult to pronounce. This type of change may be reduced if
the language is designed to be easy to pronounce from the start, avoiding for
example, the "scii" (to know) of Esperanto, and avoiding various distinctions
that many people find difficult to make, e.g. "as" vs. "like".

Matthew Montchalin

unread,
Apr 29, 2003, 4:38:23 AM4/29/03
to

Okay. And is there a rule of thumb that can be counted on to determine
whether the grammar will change before the vocabulary will?

Matthew Montchalin

unread,
Apr 29, 2003, 4:39:37 AM4/29/03
to
On 29 Apr 2003, Automort wrote:

Okay, that is a practical, empirical way of determining the success
behind any predictions made. Still, making a prediction is better
than making no predictions at all.

Matthew Montchalin

unread,
Apr 29, 2003, 4:43:45 AM4/29/03
to
On 29 Apr 2003, Crandadk wrote:
|>When people concoct their artificial languages, do they take into
|>account the likelihood of language change? Should an artificial
|>language be resistant to language change?
|
|Some changes in pronunciation happen when people use easier pronunciations
|of words they find difficult to pronounce.

Then it helps to construct a language employing the sounds that a given
people is already familiar with. (Just because Americans can tell the
difference between liquids or retroflexes like l and r does not mean
that people in Japan can. And that is just an example.) Deciding
a priori what is 'easy' to pronounce and what is not, is problematical.

|This type of change may be reduced if the language is designed to be
|easy to pronounce from the start, avoiding for example, the "scii"
|(to know) of Esperanto, and avoiding various distinctions that many
|people find difficult to make, e.g. "as" vs. "like".

Are you worried merely about sequences of phonemes, rather than the
phonemes themselves? Shouldn't you also worry about sequences of
tones? What about sequences of quantities, if your language has
those characteristics for the vowels?

Klaus Scholl

unread,
Apr 29, 2003, 8:47:34 AM4/29/03
to

"Dana Nutter {post responses only to the list}" <ab...@nutter.net>
schrieb im Newsbeitrag
news:g99rav8dkcfqq81kh...@4ax.com...

It's not Cause, not Symptom its a mutual effect, its both.
The natural growing of Language is that of a Tumor:
new concepts are build on a old primitive Basis,
the Fundaments stagnate since thousands of Years.
It's similar to the Development of CPU's for Personal Computers:
to be forced to create downwards-compatible CPU's,
new CPU's still base on x86 Technology, which drains the Speed and
Efficiency
in comparision to what were possible with new Technologies.
--
To get a Glimpse of what would be possible in Language:
Since Natlangs scale Properties in the vast Majority by only 2
Grading-Adjectives
(like good-bad, thats why some call it the Good-bad-Scheme) this Scheme
is
dominating their thoughts, their Brain, their Thinking, and causes
primitive Opinions
(Things are either good or bad), resulting in primitive Actions (bad
People must be made punished).
Of course People can Images more refined Grades, but they can hardly
express them, hardly communicate them,
and what is hardly communicated does not exist in the Public-Mind, in
the Public-Thinking,
in the Public-Person. Though there are Thoughts which thousands of
People think over and over again (e.g. certain Feelings): if they are
not expressable they are lost, they are Ghosts, they are Souls without
Bodys, they never appear in public, they do not exist de-facto.
The good-bad-Scheme is a primitive Scheme which embosses itself into the
Mind of People,
if they want it or not, and which does NOT exist anymore because People
cant think further than in good-bad Steps, but because its a Relict of
ancients Times, which People cant get rid of:
its been traded over and over again and is not replaced by more
temporary Schemes,
which rather accord to nowadays average Intelligence, and nowadays
Mentalities.
Orwell painted the Devil at the Wall in his Book 1984:
he accused Conlangs to be capable of controlling ones Mind.
The Natlangs ARE the Devil, who controls the Mind (though no Authority
pulls the Puppets Strings) and the Auxlangs (or Modlangs, but i dont
explain here what a Modlang is)
are the only Escape from this Devil. So i dont like Orwell for his Book,
he accused the Wrong Kind of Languages,
he accused the Conlangs instead of the Natlangs.
He saw the Devil on the wrong Side.


Greet from Klaus.


Automort

unread,
Apr 29, 2003, 10:40:12 AM4/29/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>is there a rule of thumb that can be counted on to determine
>whether the grammar will change before the vocabulary will?

It likely won't.

Automort

unread,
Apr 29, 2003, 10:44:45 AM4/29/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>that is a practical, empirical way of determining the success
>behind any predictions made. Still, making a prediction is better
>than making no predictions at all.

I agree. Perhaps computer models of possible, likely changes can be made and
reserved for a later date. They HAVE to be hidden till the right time to avoid
becoming an influence on the language.
A very long term experiment.

I once suggested to an archaeology publication that if artifacts of an unknown
culture are found, the rest of the culture should be predicted from those items
before further investigations are carried out. I didn't mean language and
particular historical events, but the level of complexity and nature of
government and religion, etc.
Essentially their response (in more complicated language) was that nobody had
the balls to do that because it could cost them jobs and grants. At least we'd
be dead before our predictions were tested.

Automort

unread,
Apr 29, 2003, 10:46:29 AM4/29/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>Then it helps to construct a language employing the sounds that a given
>people is already familiar with.

Of course, if Americans do that they are "ethnocentric" and "prejudiced."
Not in my opinion, since I'm less interested in sounds than in meaning and
grammar, but it's the prevailing prejudice.

Matthew Montchalin

unread,
Apr 29, 2003, 11:45:36 PM4/29/03
to
On 29 Apr 2003, Automort wrote:

What do you think of Volapuek? In this language, the last syllable of
a word is 'stressed.' I seem to have missed the information describing
location of secondary stresses, though. Would it be as readily received
by a people whose natural language is more concerned with tones than
syllabic stresses?

hs

unread,
Apr 30, 2003, 5:05:43 AM4/30/03
to
Klaus Scholl wrote:

> Orwell painted the Devil at the Wall in his Book 1984:
> he accused Conlangs to be capable of controlling ones Mind.
> The Natlangs ARE the Devil, who controls the Mind (though no Authority
> pulls the Puppets Strings) and the Auxlangs (or Modlangs, but i dont
> explain here what a Modlang is)
> are the only Escape from this Devil. So i dont like Orwell for his Book,
> he accused the Wrong Kind of Languages,
> he accused the Conlangs instead of the Natlangs.
> He saw the Devil on the wrong Side.
>

Not necessarily. Conlangs can be used for either, the good and the bad - and
after all, purposefully making people stupid is even worse than being stupid
by nature.
BTW, it seems to me that for what you have in mind, Lojban is an answer.
What is your opinion of Lojban?
--
Hans Straub

Klaus Scholl

unread,
Apr 30, 2003, 7:45:57 AM4/30/03
to

|
| Not necessarily. Conlangs can be used for either, the good and the
bad - and
| after all, purposefully making people stupid is even worse than being
stupid
| by nature.
| BTW, it seems to me that for what you have in mind, Lojban is an
answer.
| What is your opinion of Lojban?
| --
| Hans Straub

Lojban is unattractive for me (for the same Reasion as it is
unattractive for others):
because of its randomized Word Generation, the Language may have a
logical Fundament,
but its Words are lifeless and hard to store in Memory.
Also it lacks many Features i invented like
'Universal Wordclass Derivation', 'Scheme of Scale-Terms',
'Preposition-Affix-Hybrids', 'Simulated Word-Staunching',
'the extended Gressive-Model', ...
(to explain them would take here to much Space)

Im an Auxlanger, who seeks for the golden Language,
the perfect Language, but i see that an Auxlang cannot
get wide Acceptance in the Folk,
thats why i am mainly a Modlanger than an Auxlanger.
A Modlanger is someone who alters Natlangs into an Language which is
a Mixture of a Natlang and a articifial-Lang (called Modlang),
by integrating artificial Modules (called MOD's).
This is a new kind of Philosophy: the NatLang is altered not replaced!
the Disadvantatages/Advantages between an Auxlang and an Modlang are:
1. Auxlang must be *completely* contructed before it can be used, which
is enormous Time-consuming,
a small Mod can be constructed in 5 Minutes.
2. an Auxlang must be *completely* learned by the Speaker to be used,
which is Time-consuming,
a small MOD can be learned in 5 Minutes at used at once.
3. an Modlang can be altered Step-by-Step, an Auxlang hardly so.
4. MOD's can be quickly replaced, and Auxlang has no MODs, and Parts of
it can hardly be replaced.
5...

the Advantages of a Modlang are overwhelming, and MOD's are the only
realistic Way to the golden Language,
which will evolve out of the Natlang, when People apply the MOD's.
E.g. you and me could agree on a MOD and apply them at once,
if the MOD is good others will probably imitate it,
that could be Rules ('a' is written 'an' no matter if it precedes a Word
beginning with a Vowel or not)
that could also includes new Words ('Tomorrow i unbecome a Prisoner',
'Beleft the Tree grows a Mushroom', 'I won Money, whichwhy im rich
now'),
that could include almost everything which alters Language into a more
golden one.

Greet from Klaus.


Automort

unread,
Apr 30, 2003, 9:44:08 AM4/30/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>What do you think of Volapuek?

> Would it be as readily received


>by a people whose natural language is more concerned with tones than
>syllabic stresses?
>

I don't know. Where is it most used?

Automort

unread,
Apr 30, 2003, 9:49:21 AM4/30/03
to
>From: "Klaus Scholl" Kl...@sensualium.cjb.net

>Lojban is unattractive
.........


>because of its randomized Word Generation,

Basically, all words are arbitrary sounds assigned to meanings. I suppose you
mean that related concepts in that language don't necessarily have similar
sounding words?

>Im an Auxlanger, who seeks for the golden Language,
>the perfect Language,

I'd like for a dozen supermodels to fall passionately in love with me,
submitting to my every whim and handing over their paychecks unquestioningly.

>This is a new kind of Philosophy: the NatLang is altered not replaced!

You mean like the basic English someone designed for non-speakers? Indeed, you
mean like a pigin, where a couple of languages are simplified and combined?


Klaus Scholl

unread,
Apr 30, 2003, 10:49:06 AM4/30/03
to
|
| Basically, all words are arbitrary sounds assigned to meanings. I
suppose you
| mean that related concepts in that language don't necessarily have
similar
| sounding words?

Yes, and there is too less Derivation,
and its Concept of 'Gismu' and 'Selbri' is too wicked for a Person with
standard Intelligence
to understand.
Recitation:
>>A selbri is not a type of word (like a 'verb' in English); it is
something that some types of words can do. >>Various types of word can
act as selbri, but cmene, as we've seen, cannot.
Eh? Who shall understand this? Its awfull.
Also the Language is Ugly (Not that french or english would be less
ugly, but i expect from an Auxlang that
it aims for Aesthetics in Sounding, like Quenye does.)

| >Im an Auxlanger, who seeks for the golden Language,
| >the perfect Language,
|
| I'd like for a dozen supermodels to fall passionately in love with me,
| submitting to my every whim and handing over their paychecks
unquestioningly.
|

If you dont seek out for your Dreams you will hardly make them come
true.
And please argument abit less polemically.

| >This is a new kind of Philosophy: the NatLang is altered not
replaced!
|
| You mean like the basic English someone designed for non-speakers?
Indeed, you
| mean like a pigin, where a couple of languages are simplified and
combined?
|

No. I explain it.
What is a perfect Car?
A perfect Car is a perfect Motor + perfect Tires + perfect seats and
thousands of other Parts.
So the innovative modular Approach of my Idea of a Modlang does nothing
else than
take the old primitive Car (Natlang) and replaces Step by Step each Part
of the Car by artificial more Advanced Parts.
Mods can be simple or more complex:
a simple (nevertheless important) MOD of the English Language would be
to flect all
strong Verbs weak, e.g. Present Tense 'put' -> Past Tense 'puted'.
a more complex MOD would be to replace all personal pronouns by an
unambigous,
derivative Set of new ones, e.g.

Singular:
Kasus| 1.Pers.| 2.Pers.| 3.Pers.| 4.Pers.| 5. Pers.|
-----|--------|--------|--------|--------|---------|
Nom. | i | tu | he | tui | se |
Acc. | im | tum | hem | tuim | sem |
Gen. | ime | tume | heme | tuime | seme |
Plural:
...

or anything similar to that, i showed this Table just to give a View
what i mean with 'new Set';
So MODs i call Alterations thourgh new Words or Rules or Bundles of
Rules
which are apriori, that means they do not comply with existing
Wordmaterial or existing Meaning-material,
and they are artificially and planfully constructed.
Their Goals are Unambiguity, Logic, Uniformity through Derivation,
Simpicity.


Greet from Klaus.


Automort

unread,
Apr 30, 2003, 4:18:37 PM4/30/03
to
>From: "Klaus Scholl" Kl...@sensualium.cjb.net

>Who shall understand this? Its awfull.
>Also the Language is Ugly

I plan to study the language again when I have learned about the natural
languages I have decided to study. What I have looked at seems to show it quite
useful.
How many people use any "artificial" language? Like the others, it will
probably maintain a group of users for some time.
I'm interested in its development, though I won't see more than a small part of
it.


>i expect from an Auxlang that
>it aims for Aesthetics in Sounding

lojban wasn't designed specifically for aesthetics.

>please argument abit less polemically.

My way of stating that you can't expect perfection. Otherwise Claudia Schieffer
would be doing my laundry right now, Vendela would be cooking, and Rachel
Hunter would be coming down the hall to ask what she can do for me.

>What is a perfect Car?

One I can afford to buy and keep running.
Like the Mitsubishi in my driveway.


>take the old primitive Car (Natlang) and replaces Step by Step each Part
>of the Car by artificial more Advanced Parts.

But what defines the more advanced parts? Do you need fine discriminations in
time, or do you need to carefully describe social relationships, or do you need
to emphasize geomettical shapes?
The matter of a constructed language, it seems to me, is whatever you would
prefer to be able to discuss in minute detail or the particular way you view
the world. It isn't a patchwork like a natural language, but the perfection is
how well it does what you want it to do, not some abstract and universal
perfection.

Matthew Montchalin

unread,
Apr 30, 2003, 7:59:19 PM4/30/03
to

Cryptography? ;)

Automort

unread,
Apr 30, 2003, 9:41:46 PM4/30/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>|I don't know. Where is it most used?
>
>Cryptography? ;)
>

And stress on the last syllable matters there?

Andrew McIntosh

unread,
May 1, 2003, 5:06:51 AM5/1/03
to
> almost every word. Much different from their common ancestor, Latin.
> On the other hand spoken Japanese seems to do just the opposite by
> dropping the final vowel.


Actually, Japanese doesn't drop final vowels. The vowels u, i. and
sometimes even o can be unvoiced in Japanese (usually between unvoiced
consonants, ie "suki", "sita"), but also quite often at the end of a
word. Since unvoiced vowels aren't common in other languages, most
foreign speakers hear it as the vowels being "dropped". They're not.
They're just pronounced in a manner you're not accustomed to. Japanese
is very strictly CV.

Matthew Montchalin

unread,
May 1, 2003, 8:21:56 AM5/1/03
to

Volapuek being a language whose grammar expects a great deal of
significance in the last syllable, the grammar is probably more
resistant to linguistic drift than those which are prepositionally
based.

Cryptography that involves more than just character swapping, can
get quite intense when a language whose parts can be represented
very reliably by mathematical shorthand, and yet are altered or
changed according to an algorithm.

nyra

unread,
May 1, 2003, 6:05:34 PM5/1/03
to
Klaus Scholl schrieb:
[...]

> Also the Language is Ugly (Not that french or english would be less
> ugly, but i expect from an Auxlang that
> it aims for Aesthetics in Sounding, like Quenye does.)

Now _that_ is really subjective. I find many german dialects
aesthetically pleasing that most folks i know find just plain
'orrible. And i found the quenya (or was it sindarin?) spoken in the
Lord of the Rings films rather bland and boring. For most
aesthetically pleasing natural language my current vote would go to
spanish, with russian coming in a close second.

> | >This is a new kind of Philosophy: the NatLang is altered not
> replaced!
> |
> | You mean like the basic English someone designed for non-speakers?
> Indeed, you
> | mean like a pigin, where a couple of languages are simplified and
> combined?
> |
>
> No. I explain it.
> What is a perfect Car?

The one which best fits the individual user's needs and desires. A
"perfect" formula one racing car is one of the worst cars you can
possibly use for a family trip.

Of course, people already "tune" their spoken language to some degree,
by using it with more or less skill, having a larger or smaller active
vocabulary etc.

> So the innovative modular Approach of my Idea of a Modlang does nothing
> else than
> take the old primitive Car (Natlang) and replaces Step by Step each Part
> of the Car by artificial more Advanced Parts.

Sounds quite interesting, as long as enough attention is paid to how
it possibly interferes with already-existing words (don't want to
produce reams of new homophones) and general "speakability" (avoid
generating hard-to-pronounce phoneme clusters and the like), and as
long as the modifiers are sufficiently critical of their own work. I
don't think you can make the One Perfect Language, you can only hope
to make one which does the things you want it to do particularly well.

(big modifications)


> Their Goals are Unambiguity, Logic, Uniformity through Derivation,
> Simpicity.

I don't consider any of these universal values a language ought to
strive for. But then i'm rather strictly interested in pseudo-natlangs
which rather strive to imitate the behaviour of natural languages, so
i'm at least as biased as you are.

--
Bqsfttfusftptusbjutbmbvupsjufeftspijtfueftfnqfsfv-stbqsftbxp
jsqspdmbnfuspjtgpjttbmjcfsufmbgsbodftftutpvnjtbeftdp
nqbhojftgjobodjfttesfsftrvjejtqptfoueftsjdifttfevqbztfuqbsmf
npzfo evofqsfttbdifuffeijsjhfoumqjojpo. - Voufnpjoxfsjejrvf.

Matthew Montchalin

unread,
May 1, 2003, 7:28:17 PM5/1/03
to
On Thu, 1 May 2003, Matthew Montchalin wrote:
|On 1 May 2003, Automort wrote:
||>From: Matthew Montchalin mmon...@OregonVOS.net
||
||>|I don't know. Where is it most used?
||>
||>Cryptography? ;)
||
||And stress on the last syllable matters there?
|
|Volapuek being a language whose grammar expects a great deal of
|significance in the last syllable, the grammar is probably more
|resistant to linguistic drift than those which are prepositionally
|based.

But if the syllabic stress were always on the first most meaningful
syllable of the word, then a language that was a 'prefix' based
language would be more resistant to linguistic drift.

Abrigon Gusiq

unread,
May 2, 2003, 1:13:07 AM5/2/03
to
Also do you mean planned change, or chaotic change, or maybe even
expected change.

Sort of like how the French Academy allows new words in, but they often
have to be based on previous french roots.

Then you have natural change, it just happens, no real planning to it,
popular lingos are much like this, with slang and new meanings of old
words.

Then there is change you expect to happen, like how Old German became
New German, or basically Grimms Law, where as some forms become
different forms due to just how people are, speak and like. I don't know
to much of it, but the basic concept, so please forgive.

Mike

Matthew Montchalin

unread,
May 2, 2003, 4:21:49 AM5/2/03
to
On Thu, 1 May 2003, Abrigon Gusiq wrote:
|Also do you mean planned change, or chaotic change, or maybe even
|expected change.
|
|Sort of like how the French Academy allows new words in, but they often
|have to be based on previous french roots.
|
|Then you have natural change, it just happens, no real planning to it,
|popular lingos are much like this, with slang and new meanings of old
|words.
|
|Then there is change you expect to happen, like how Old German became
|New German, or basically Grimms Law, where as some forms become
|different forms due to just how people are, speak and like. I don't
|know to much of it, but the basic concept, so please forgive.

Half of the problem is trying to predict the change before it happens.
If the prediction is accurate, then the change was foreseeable and some
kind of planning should have been useful. If the prediction fails,
then the change could just as well be chaotic, and there is a matter
of pure luck as to whether the change could have been dealt with.

I expect that most changes are phonetic in nature, and not grammatical;
the phonetic change happens first, and then grammatical change happens
second.

Garth Wallace

unread,
May 2, 2003, 5:06:46 AM5/2/03
to
Dana Nutter {post responses only to the list} wrote:
> Working on my current (still unnamed) project, I'm toying with some
> different ideas on how to handle verb modals and verbal aspects. The
> major questions are:
>
> What are the most common modes in natural language?

According to Palmer's "Mood and Modality", all mood comes down to Realis
(asserting an actual event) versus Irrealis (no assertion, or an unreal
event). If you want a modal system, you'll need to handle epistemic or
evidential possibility and necessity (how likely it is that the event
happened/is happening/will happen), permission and obligation (whether
the subject is allowed or required to perform the action), commission
(promises and threats), and ability and willingness. The problem is that
different languages divide these up among modal lexemes differently. For
example, English uses "may" for epistemic possibilty and for deontic
permission. Also, epistemic modals get complicated when you throw in
negation. Modal systems in natlangs are a big mess.

I really recommend reading "Mood and Modality" by F. R. Palmer. It
covers the subject in depth, and it's actually very readable.

> I'd like to see examples of different methods of handling these:
> affixes, adverbs, etc.

Not sure about adverbs (I can't remember seeing any examples in Palmer,
but that doesn't necessarily mean it's impossible). However, beyond
modal verbs like in English, affixes and particles both seem common.

> The same for aspects. Any good example from natural languages?

The two most common aspects in natlangs are probably progressive and
perfective. At least, that's the most basic aspectual distinction I ca
think of: action is still going on vs. action thought of as a single event.

Just guessing, I think affixes may be the most common way of marking
aspect. I have no evidence of this, so take it with a grain of salt.

> Right now, I'm leaning toward using adverbs for both.

I think I remember reading somewhere that aspect is a more fundamental
concept than tense: that languages tend to develop aspect markings
before they develop tense (and the tense markings may be derived from
aspects, for instance the perfect aspect frequently becomes a simple past).

--
Happy denizen of the Nightstar IRC Network
Webcomics discussion: irc://us.nightstar.net/webcomics
Progressive rock chat: irc://us.nightstar.net/progrock

Garth Wallace

unread,
May 2, 2003, 5:10:39 AM5/2/03
to
Dana Nutter {post responses only to the list} wrote:
> On 28 Apr 2003 22:51:31 GMT, auto...@aol.com (Automort) wrote:
>
>>Maybe a person could design a language with instructions on forming new words
>>and instructions that every 100 years the spelling be revised to keep up with
>>actual pronounciation, with the sounds assigned the letters never to change.
>
>
>
> Well, that's something like what I'm doing. There will be a standard
> set for the creation of new words. I'd rather just create a simple
> enough phonology to avoid any need for spelling reform. Of course the
> language will tend to go its own direction once in use.
>
> I'm using a simple phonology: / p t k f s x m n N j w i e a o u @ /
> which will have a generous inventory of allophones to help accomodate
> those who may have trouble with any of these. The phonemes are written
> in the Roman alphabet respectively as ( P T K F S H M N Y J W I E A
> O U X ) I will have a website up for this soon. It's mostly complete
> but I still have a couple of design decisions to make first.

That's an interesting orthography. Why <Y> for /N/ and <X> for /@/? It
seems to me that <Y> is more often used as a vowel than <X> (<Y> is even
/@/ is Welsh, AIUI), and <X> is almost always a consonant. Neither is
commonly used for /N/ though (why not use an n-tilde, or an eng?). Also,
why the capitals?

Dana Nutter {post responses only to the list}

unread,
May 2, 2003, 6:25:32 PM5/2/03
to
On Fri, 02 May 2003 09:10:39 GMT, Garth Wallace <gwa...@despammed.com>
wrote:

> Dana Nutter {post responses only to the list} wrote:
> > On 28 Apr 2003 22:51:31 GMT, auto...@aol.com (Automort) wrote:
> >
> >>Maybe a person could design a language with instructions on forming new words
> >>and instructions that every 100 years the spelling be revised to keep up with
> >>actual pronounciation, with the sounds assigned the letters never to change.
> >
> >
> >
> > Well, that's something like what I'm doing. There will be a standard
> > set for the creation of new words. I'd rather just create a simple
> > enough phonology to avoid any need for spelling reform. Of course the
> > language will tend to go its own direction once in use.
> >
> > I'm using a simple phonology: / p t k f s x m n N j w i e a o u @ /
> > which will have a generous inventory of allophones to help accomodate
> > those who may have trouble with any of these. The phonemes are written
> > in the Roman alphabet respectively as ( P T K F S H M N Y J W I E A
> > O U X ) I will have a website up for this soon. It's mostly complete
> > but I still have a couple of design decisions to make first.
>
> That's an interesting orthography. Why <Y> for /N/ and <X> for /@/? It
> seems to me that <Y> is more often used as a vowel than <X> (<Y> is even
> /@/ is Welsh, AIUI), and <X> is almost always a consonant. Neither is
> commonly used for /N/ though (why not use an n-tilde, or an eng?). Also,
> why the capitals?

I don't care for diacritics or digraphs. I'm trying to keep it clean
and simple. Actually /N/ is actually spelled with an eng, but
establishing "Y" as an alternative for those who are typographically
challenged. I mainly chose "Y" as an alternate because the lower case
has a tail similar to eng. "X" for /@/? Well just because it's there.
I needed another letter to reppresent that vowel. "X" will only be
used for one purpose anyway, as a connector between consonants when
forming compound words. Ex: HONXTOM (library) = HON (book) + TOM
(home). Lojban & loglan also use "y" for /@/. I don't want this being
confused with their usage.

It doesn't have to be all capitals. It can also be written in lower
case as long as it is written in all one or the other. No capitalizing
the first word of a sentence, and no capitalizing proper nouns. For
example you could write headlines and titles in all upper case, and the
entire body of text in all lower case. I believe this is similar to how
Georgian and Armenian scripts are used, somebody else would probably be
able to answer that better.

I actually did create a simple "native" script, complete with TTF font,
for this language, but have mostly scrapped that idea since it would be
highly impractical for an auxlang.

Matthew Montchalin

unread,
May 2, 2003, 8:53:34 PM5/2/03
to

Does your language possess tones, tone colors (rising/falling tones), or
syllabic stress? Should they be represented with their own unique
charactes?

Dana Nutter {post responses only to the list}

unread,
May 3, 2003, 12:22:03 AM5/3/03
to

> Does your language possess tones, tone colors (rising/falling tones), or
> syllabic stress? Should they be represented with their own unique
> charactes?

Syllabic stress is indicated by vowel lengthening, however that stress
is only to provide special emphasis when necessary. Word stress will
most likely be indicated by a particle. Questions are formed by
interrogative words or an interrogative particle with no special
intonation. The "official" intonation pattern will be basically a
"flat" tone, but any tonal pattern will be acceptable since it serves no
semantic or grammatical purpose.

Cellus Purfluxius

unread,
May 3, 2003, 12:27:25 PM5/3/03
to
Kvanto al kambio in le moderne linguas: an le fakto ke nos possede films
e son-rekorderos non va konservar le lingva plus facilmente kvam in le
tempores kvando solo un elite habeva le kapabilitate skriber? E kval
rolo joka le skriptura ipse?

English version:
Don't you think that in our days and in the future when we have sound
and video recorders and films this will mean that the change in
languages will be slower? And what role does writing play?

Cellus P.

Cellus Purfluxius

unread,
May 3, 2003, 12:39:25 PM5/3/03
to
Dana Nutter {post responses only to the list} wrote:

>(...)
>
> Iceland has worked to
>preserve its language for several hundred years now. So even "natural"
>languages have some "artificial" influence.
>
(Interlingva kon un ortografia fonetik, just for the fun of it):
Le islandese lingva era multo mixate kon danese in le passate sekulo.
Per un konscie labor on purifikava le lingva de omne non-islandik
formas. On rekreava le lingva per le auxilio del lingva in le islandese
sagas.

In nostre tempore on krea islandese parolas pro omne nove parolas. pro
laser imprimitor, komputator etc.

Cellus P.

Cellus Purfluxius

unread,
May 3, 2003, 12:48:42 PM5/3/03
to
Dana Nutter {post responses only to the list} wrote:

>Working on my current (still unnamed) project, I'm toying with some
>different ideas on how to handle verb modals and verbal aspects. The
>major questions are:
>
>What are the most common modes in natural language?
>

>I'd like to see examples of different methods of handling these:
>affixes, adverbs, etc.
>

>The same for aspects. Any good example from natural languages?
>
>

>Right now, I'm leaning toward using adverbs for both.
>

Varii grammatikas usa varii termines. Il existe species de aktion e
aspektos. To no es necessarimen le sami kosa. Ma noi lassa ti problema e
usa varii prefixi pro varii sorts de aktion. Assi:

Itterative e distributive aktion:
Harry Potter, Herminone e sui fratres viagiadava al station. Harry
Potter vidava kome le gente kurradeva per le porta al perons un, du e
kvadr.

Inkoativi e momentani aktion:
Subite il era li turn de Harry. Ille ekkurreva. Perkurreva kon sukcess
li mur!

On krea li varii types de aktion per li auxilie de varii prefixes,
infixes e suffixes.

Cellus P

Cellus Purfluxius

unread,
May 3, 2003, 12:58:22 PM5/3/03
to
hs wrote:

In mi opinion le humanos pote inventar le parolas ke illes besonia. Si
isto non esseva le fakto nos ankora parlarea komo durante le tempore
kvando nos chassava le mammuts in antike tempores!

Per konsekvente in le kvartiros del "proles" on certo haberea nove
parolas le kval le partito non poteva kontrollar!

Cellus P.

>
>

Automort

unread,
May 3, 2003, 3:31:31 PM5/3/03
to
>From: Cellus Purfluxius Ja...@nospam.com

>Don't you think that in our days and in the future when we have sound
>and video recorders and films this will mean that the change in
>languages will be slower?

Possibly. Or there may develop "movie" languages and "used" languages much as
people from England, etc. sing as though they were from Tennessee but speak
their regular gibberish.

>And what role does writing play?

It can standardize spelling till written words no longer reflect the spoken.

Matthew Montchalin

unread,
May 3, 2003, 7:21:33 PM5/3/03
to
On Sat, 3 May 2003, Dana Nutter {post responses only to the list} wrote:
|> Does your language possess tones, tone colors (rising/falling tones), or
|> syllabic stress? Should they be represented with their own unique
|> charactes?
|
|Syllabic stress is indicated by vowel lengthening, however that stress
|is only to provide special emphasis when necessary.

In Classical Latin, short syllables could be stressed. Stress was not
predictable on the basis of syllable length. (Hence the modern practice
of supplying vowels with macrons or 'horizontal strokes' above the vowels,
but syllable accent was something you had to memorize.)

Cellus Purfluxius

unread,
May 4, 2003, 10:59:39 AM5/4/03
to
Automort wrote:

At least Swedish spelling has had a substantial influence on
pronunciation. Nowadays nearly everyone pronounces the preposition
corresponding to English "of" as "av" (as it is spellt, and not as "åv"
as elderly people said in my childhood). So by the written language we
have gone back some 4 centuries!

Cellus P.

Automort

unread,
May 4, 2003, 5:13:19 PM5/4/03
to
>From: Cellus Purfluxius Ja...@nospam.com

>At least Swedish spelling has had a substantial influence on=20
>pronunciation.

Well, I've heard people pronounce the English word "often" as "off-ten" instead
of the correct way, "off-in."

Ja Mess

unread,
May 5, 2003, 10:58:50 AM5/5/03
to


What language is that you use? For it seems to me a mix of itaolian and
spanish with some unsuccesfull wish to use latin words.
It's ugly like hell!
What neo-latin languages have you studied so far?
And why do you use imperfect - conditional? That's wrong according to
the most ot the neo-latin languages' grammar.
Where did you find the form "esseva"? As all the languages use "era"?
I would like to have a chat on this because I'm so curious :-)
Jamess

(I speak very well italian, castellan, catalan, I am romanian, and I
understand pretty well french. Maybe I could help you with something
--
Direct access to this group with http://web2news.com
http://web2news.com/?alt.language.artificial
To contact in private, remove no-3spp0a5mm

Cellus Purfluxius

unread,
May 5, 2003, 11:25:15 AM5/5/03
to
Ja Mess wrote:

>>(...)

>>
>>>
>>>
>>>
>>In mi opinion le humanos pote inventar le parolas ke illes
>>besonia. Si
>>isto non esseva le fakto nos ankora parlarea komo durante le tempore
>>kvando nos chassava le mammuts in antike tempores!
>>
>>Per konsekvente in le kvartiros del "proles" on certo haberea nove
>>parolas le kval le partito non poteva kontrollar!
>>
>>Cellus P.
>>
>>
>
>
>What language is that you use? For it seems to me a mix of itaolian and
>spanish with some unsuccesfull wish to use latin words.
>It's ugly like hell!
>

Isto era un joco. Ego scribeva Interlingua ma usava un ortografia multo
fonetic. Le variante es totalmente mi proprie. Iste texto in contrario
es normal interlingua. Le sol differentia in iste texto incomparation
con interlingua "normal" es que ego non scribe "orthographia" que es le
forma traditional in interlingua.

Lege plus a http://www.interlingua.com

Amicalmente

Cellus P.

Ja Mess

unread,
May 5, 2003, 12:50:24 PM5/5/03
to


Gracies, ara vaig a veure la web de que m'has dit tu. M'interessa molt
la cosa, pero perque estic curios. Aixo que et poso en aqui es catalan.
El entens? M'agrada molt perque sembla molt al romanés que es la meva
llengua.

:-)
Ok. Where are you from? And what languages (natural ones ) have you
studied?

Thanks again,
James

Paul O. BARTLETT

unread,
May 5, 2003, 2:35:52 PM5/5/03
to
On Mon, 5 May 2003, Ja Mess wrote:

> [cut]

> What language is that you use? For it seems to me a mix of itaolian and
> spanish with some unsuccesfull wish to use latin words.
> It's ugly like hell!
> What neo-latin languages have you studied so far?

Kjell is a fluent user of (IALA) Interlingua, but he likes to play
around with his own idiosyncratic conlang which he calls Communicando.
Personally I think that that is counterproductive in respect to the
conIAL movement, as it dissipates energy in a movement that has enough
difficulty as it is.

--
Paul Bartlett
bartlett at smart.net
PGP key info in message headers

Ja Mess

unread,
May 5, 2003, 3:43:29 PM5/5/03
to
> On Mon, 5 May 2003, Ja Mess wrote:
>
>> [cut]
>
>> What language is that you use? For it seems to me a mix
>> of itaolian and
>> spanish with some unsuccesfull wish to use latin words.
>> It's ugly like hell!
>> What neo-latin languages have you studied so far?
>
> Kjell is a fluent user of (IALA) Interlingua, but he
> likes to play
> around with his own idiosyncratic conlang which he calls
> Communicando.
> Personally I think that that is counterproductive in respect to the
> conIAL movement, as it dissipates energy in a movement
> that has enough
> difficulty as it is.


I have just had an exam at the university and didn't have time tu study
this Interlingua, but 2nite I'll work on it. It's cute and strange. I
understant everything but as I am a grammar maniac and in love with the
natural languages it is difficult to find a compromise regarding the
artificial languages for humans.
I study computer engineering and programming, and have some basic
knowledge of IA, so I am pretty willing to finda compromise. :-)
Ja Mess (Bogdan-Alexandru Constantin)

Cellus Purfluxius

unread,
May 6, 2003, 11:08:32 AM5/6/03
to
Ja Mess wrote:

Ego poteva ben comprender tu texto in catalano post un poco de pensar in
alcun locos. Ante alcun annos ego studiava le catalano multo
superficialmente.

Ego es svedese. In le schola ego apprendeva le anglese, germano,
francese (un poco), latino, al universitate ego ha studiate russo e
polonese. In le polonese ego me senti plus confidente. Le romaniano ego
cognosce solo sur le nivello touristic. Ego sape leger un jornal si il
non es un texto troppo specialisate.

Amicalmente

Cellus Purfluxius (proprie latinisation de mi nomine :-)


Ja Mess

unread,
May 6, 2003, 1:04:10 PM5/6/03
to
> Ja Mess wrote:
>
>>>Ja Mess wrote:
>>>
>>> =20

>>>
>>>>>(...)
>>>>>
>>>>>In mi opinion le humanos pote inventar le parolas ke illes
>>>>>besonia. Si
>>>>>isto non esseva le fakto nos ankora parlarea komo
>>>>> durante le tempore
>>>>>kvando nos chassava le mammuts in antike tempores!
>>>>>
>>>>>Per konsekvente in le kvartiros del "proles" on certo haberea nove
> etc..

>>>>What language is that you use? For it seems to me a mix of
>>>>itaolian and
>>>>spanish with some unsuccesfull wish to use latin words.
>>>>It's ugly like hell!
>>>>
>>>> =20

>>>>
>>>Isto era un joco. Ego scribeva Interlingua ma usava un
>>>ortografia multo
>>>fonetic. Le variante es totalmente mi proprie. Iste texto
>>>in contrario
>>>es normal interlingua. Le sol differentia in iste texto
>>>incomparation
>>>con interlingua "normal" es que ego non scribe
>>>"orthographia" que es le
> etc..

>>Gracies, ara vaig a veure la web de que m'has dit tu.
>> M'interessa molt
>>la cosa, pero perque estic curios. Aixo que et poso en
>> aqui es catalan.
>>El entens? M'agrada molt perque sembla molt al roman=E9s

>> que es la meva
>>llengua.
>>
>>:-)
>>Ok. Where are you from? And what languages (natural ones ) have you
>>studied?
>> =20

>>
> Ego poteva ben comprender tu texto in catalano post un
> poco de pensar in =
>
> alcun locos. Ante alcun annos ego studiava le catalano multo=20
> superficialmente.
>
> Ego es svedese. In le schola ego apprendeva le anglese, germano,=20

> francese (un poco), latino, al universitate ego ha
> studiate russo e=20

> polonese. In le polonese ego me senti plus confidente. Le
> romaniano ego=20

> cognosce solo sur le nivello touristic. Ego sape leger un
> jornal si il=20

> non es un texto troppo specialisate.
>
> Amicalmente
>
> Cellus Purfluxius (proprie latinisation de mi nomine :-)

svedese? Congratulations. As a romanian it wasn't difficult for me to
study the languages I know now... how difficult is it for a Sweedish
person?

My name is Bogdan-Alexandru. Nice to meet you.

Matthew Montchalin

unread,
May 7, 2003, 5:38:24 AM5/7/03
to
On 29 Apr 2003, Automort wrote:
|>From: Matthew Montchalin mmon...@OregonVOS.net
|
|>that is a practical, empirical way of determining the success
|>behind any predictions made. Still, making a prediction is better
|>than making no predictions at all.
|
|I agree. Perhaps computer models of possible, likely changes can be made and
|reserved for a later date. They HAVE to be hidden till the right time to avoid
|becoming an influence on the language.
|A very long term experiment.

Yes

|I once suggested to an archaeology publication that if artifacts of an
|unknown culture are found, the rest of the culture should be predicted
|from those items before further investigations are carried out. I
|didn't mean language and particular historical events, but the level
|of complexity and nature of government and religion, etc.

Interesting...

|Essentially their response (in more complicated language) was that
|nobody had the balls to do that because it could cost them jobs and
|grants. At least we'd be dead before our predictions were tested.

There are nevertheless some sites that need to be gone through, before
they are lost. Not all sites are 'more' secure for being underground.

Matthew Montchalin

unread,
May 7, 2003, 5:39:45 AM5/7/03
to

I prefer 'off-in' but there is no way of convincing certain other
speakers of my correctness. :)

Matthew Montchalin

unread,
May 7, 2003, 5:43:02 AM5/7/03
to
||Word stress will most likely be indicated by a particle. Questions are
||formed by interrogative words or an interrogative particle with no
||special intonation. The "official" intonation pattern will be basically
||a "flat" tone, but any tonal pattern will be acceptable since it serves
||no semantic or grammatical purpose.

One of the things that I was getting at, is that not all sounds are the
same, and some sounds are inherently more resistant to linguistic drift
than others are. This resistance becomes even more pronounced when the
sounds are found to be adjacent to each other.


Automort

unread,
May 7, 2003, 11:03:28 AM5/7/03
to
>From: Matthew Montchalin mmon...@OregonVOS.net

>There are nevertheless some sites that need to be gone through, before
>they are lost. Not all sites are 'more' secure for being underground.

Yes, especially in the US and in developing countries where construction is
constant.
Apparently now in Egypt and other countries that are trying to develop their
natural resources, which they need to do.
China has a good historical record, respect for their own history, at least.

Ja Mess

unread,
May 7, 2003, 3:37:21 PM5/7/03
to

Yes... it's off-in.
and it's addréss not àddress
and neether, not nayther
and I don't remember :-) I talked to an english teacher because I had
many arguments on these issues

Ja Mess

Automort

unread,
May 7, 2003, 6:38:44 PM5/7/03
to
>From: "Ja Mess" kmchatk.ne...@web2news.net

>it's addréss not àddress

Well, the first is a verb, the second a noun.

Cellus Purfluxius

unread,
May 8, 2003, 9:13:27 AM5/8/03
to
Ja Mess wrote:

Multsumesc! Le amusamento es integremente a mi latere!

Naturalmente il es difficile graduar difficultate de un lingua.

Claro que le linguas romance differe del germanic linguas, ma mi
cognoscentia de illos es grandemente passive, ego recognosce parolas.
Activemente ego los maestra solo in parlar e leger, ma non in scriber.
Pro isto ego usa interlingua.

Amicalmente

Cellus P.


Illo se lege Kjell Rehnström in svedese :-)

John Stone

unread,
May 8, 2003, 11:27:32 AM5/8/03
to
"Ja Mess" <kmchatk.ne...@web2news.net> writes:

> > On 4 May 2003, Automort wrote:
> >>>From: Cellus Purfluxius Ja...@nospam.com
> >>
> >>>At least Swedish spelling has had a substantial influence on=20
> >>>pronunciation.
> >>
> >>Well, I've heard people pronounce the English word "often"
> >> as "off-ten"
> >>instead of the correct way, "off-in."
> >
> > I prefer 'off-in' but there is no way of convincing certain other
> > speakers of my correctness. :)
>
> Yes... it's off-in.
> and it's addréss not àddress

Which syllable of "convict" gets stress?

The *convict* will now address the crowd. He is giving them the
address of the United Nations General Assembly building, which will
*convict* them stupid and evil because they are ignorant of Nature's
Harmonic Time Cube Creation Principle.

> and neether, not nayther

Your transcription is confusing. If "ee" is /iy/, I would expect "ay"
to be /ey/, but no one says /neydh'r/, only /niydh'r/ and /naydh'r/.

> and I don't remember :-) I talked to an english teacher because I
> had many arguments on these issues

And the teacher sang "Let's Call the Whole Thing Off".

--
John Stone| The dinosaurs didn't | Find out about Christian
----------+ become extinct. It's | Communism, Bible prophecy,
1972, and Richard Nixon is President | Noah's Ark, U.S. Holocaust and
of the United States. -- Dan Goodman | evil dalai lama. -- Tom Goodman

Ja Mess

unread,
May 8, 2003, 11:42:04 AM5/8/03
to
John Stone wrote:
> "Ja Mess" <kmchatk.ne...@web2news.net> writes:
>
>
>>>On 4 May 2003, Automort wrote:
>>>
>>>>>From: Cellus Purfluxius Ja...@nospam.com
>>>>
>>>>>At least Swedish spelling has had a substantial influence on=20
>>>>>pronunciation.
>>>>
>>>>Well, I've heard people pronounce the English word "often"
>>>>as "off-ten"
>>>>instead of the correct way, "off-in."
>>>
>>>I prefer 'off-in' but there is no way of convincing certain other
>>>speakers of my correctness. :)
>>
>>Yes... it's off-in.
>>and it's addréss not àddress
>
>
> Which syllable of "convict" gets stress?
>
> The *convict* will now address the crowd. He is giving them the
> address of the United Nations General Assembly building, which will
> *convict* them stupid and evil because they are ignorant of Nature's
> Harmonic Time Cube Creation Principle.
>
>
>>and neether, not nayther
>
>
> Your transcription is confusing. If "ee" is /iy/, I would expect "ay"
> to be /ey/, but no one says /neydh'r/, only /niydh'r/ and /naydh'r/.
>
>
>>and I don't remember :-) I talked to an english teacher because I
>>had many arguments on these issues
>
>
> And the teacher sang "Let's Call the Whole Thing Off".
>


nayther: same as why
neether: same as fee

sorry for the "barbarisme" :-)

Automort

unread,
May 8, 2003, 2:50:31 PM5/8/03
to
>From: John Stone jd...@softhome.net

>
>Which syllable of "convict" gets stress?

If it's a noun, as in a prisoner, it's the first syllable: CONvict. If it's a
verb, as i to find guilty, it's the second: conVICT.

Anton Sherwood

unread,
May 9, 2003, 7:55:24 PM5/9/03
to
>> From: "Ja Mess" kmchatk.ne...@web2news.net
>> it's addréss not ŕddress

Automort wrote:
> Well, the first is a verb, the second a noun.

In my dialect (I'm from a Midwestern university town),
the verb <address> is always iambic, but for the noun, it depends:
a "street address" is always iambic; in a data-processing context
("address registers and data registers") the noun is always accented
on the first syllable.

--
Anton Sherwood, http://www.ogre.nu/

Anton Sherwood

unread,
May 9, 2003, 7:59:14 PM5/9/03
to
Ja Mess wrote:
> and neether, not nayther

i do say /ni:Dr/ but i suspect i'd say /najDr/ in response
to a question; it's more emphatic.

Anton Sherwood

unread,
May 9, 2003, 9:51:21 PM5/9/03
to
Cellus Purfluxius wrote:
> In mi opinion le humanos pote inventar le parolas ke illes besonia.
> Si isto non esseva le fakto nos ankora parlarea komo durante le
> tempore kvando nos chassava le mammuts in antike tempores!

This implies that all language change is a response to need,
which I think absurd.

Anton Sherwood

unread,
May 9, 2003, 9:55:53 PM5/9/03
to
Dana Nutter wrote:
> Working on my current (still unnamed) project, I'm toying with some
> different ideas on how to handle verb modals and verbal aspects.
> The major questions are:
>
> What are the most common modes in natural language?

Indicative, imperative, subjunctive, conditional, optative
are the ones familiar to me - but this shows an IE bias.

...
> Right now, I'm leaning toward using adverbs for both.

Why not tense as well?

Automort

unread,
May 10, 2003, 10:53:13 AM5/10/03
to
>From: Anton Sherwood bro...@pobox.com

>This implies that all language change is a response to need,
>which I think absurd.

Like mutation, a certain amount of it is random with respect to communication
needs.
Does it matter whether one says "bawth" as in England or "bath" as in the US?
Not really. With respect to communication in itself the difference is random.
The acquisition or invention of new words is more a matter of necessity, though
there, too, a certain randomness exists. Did the American colonists have to
adopt the Indian word for "skunk" but invent the word "rattlesnake?" I doubt
it.

Automort

unread,
May 10, 2003, 11:03:19 AM5/10/03
to
>From: Anton Sherwood bro...@pobox.com

>In my dialect (I'm from a Midwestern university town),
>the verb <address> is always iambic, but for the noun, it depends:
>a "street address" is always iambic; in a data-processing context
>("address registers and data registers") the noun is always accented
>on the first syllable.

The English spoken in Texas is the correct, benchmark form of the language.

John Stone

unread,
May 10, 2003, 11:39:55 AM5/10/03
to
auto...@aol.com (Automort) writes:

> >From: Anton Sherwood bro...@pobox.com
>
> >This implies that all language change is a response to need, which I
> >think absurd.
>
> Like mutation, a certain amount of it is random with respect to
> communication needs. Does it matter whether one says "bawth" as in
> England or "bath" as in the US?

To normal British people, "bawth" is pronounced the same as "borth"
and "bath" is pronounced the same as "barth". These are people who
rhyme "morning" with "yawning".

> Not really. With respect to communication in itself the difference
> is random. The acquisition or invention of new words is more a
> matter of necessity, though there, too, a certain randomness
> exists. Did the American colonists have to adopt the Indian word for
> "skunk" but invent the word "rattlesnake?" I doubt it.

We could randomly call an animal the "bawth", because that syllable
isn't being used by anyone.

Cellus Purfluxius

unread,
May 10, 2003, 3:12:35 PM5/10/03
to
Anton Sherwood wrote:

> Cellus Purfluxius wrote:
> > In mi opinion le humanos pote inventar le parolas ke illes besonia.
> > Si isto non esseva le fakto nos ankora parlarea komo durante le
> > tempore kvando nos chassava le mammuts in antike tempores!
>
> This implies that all language change is a response to need,
> which I think absurd.

Il es un resultato de necessitate e que le vocabulos cambia con le
tempore, justo proque on los usa in un poco alterate signification.


Cellus P.

Cellus Purfluxius

unread,
May 10, 2003, 3:18:40 PM5/10/03
to
Anton Sherwood wrote:

> Dana Nutter wrote:
> > Working on my current (still unnamed) project, I'm toying with some
> > different ideas on how to handle verb modals and verbal aspects.
> > The major questions are:
> >
> > What are the most common modes in natural language?
>
> Indicative, imperative, subjunctive, conditional, optative
> are the ones familiar to me - but this shows an IE bias.

narrativo, potential...

Cellus P.

Automort

unread,
May 10, 2003, 5:17:49 PM5/10/03
to
>From: John Stone jd...@softhome.net

>We could randomly call an animal the "bawth", because that syllable
>isn't being used by anyone.
>

Which one, the skunk or the snake?

only to the list}abyss@nutter.net Dana Nutter

unread,
May 10, 2003, 6:20:12 PM5/10/03
to
On Fri, 09 May 2003 16:59:14 -0700, Anton Sherwood <bro...@pobox.com>
wrote:

> Ja Mess wrote:
> > and neether, not nayther
>
> i do say /ni:Dr/ but i suspect i'd say /najDr/ in response
> to a question; it's more emphatic.

Only /ni:Dr/ here. /najDr/ tends to be snobby.


only to the list}abyss@nutter.net Dana Nutter

unread,
May 10, 2003, 6:58:49 PM5/10/03
to
On Fri, 09 May 2003 18:55:53 -0700, Anton Sherwood <bro...@pobox.com>
wrote:

> Dana Nutter wrote:


> > Working on my current (still unnamed) project, I'm toying with some
> > different ideas on how to handle verb modals and verbal aspects.
> > The major questions are:
> >
> > What are the most common modes in natural language?
>
> Indicative, imperative, subjunctive, conditional, optative
> are the ones familiar to me - but this shows an IE bias.
>

What I'm trying to do is look beyond the IE bias and find out if there
are any constants across different languages/families. Or, just get
some ideas how different languages handle modals.


> > Right now, I'm leaning toward using adverbs for both.
>
> Why not tense as well?

Still toying with that idea. The way it's set up right now, tense is
marked by suffixes, but only when needed for emphasis. They are not
marked at all if the verb if modified by an adverb denoting time such as
"last week", "now","tomorrow", etc. I may just give both suffixes and
adverbs as possible choices to allow a little more freedom with usage.


Silvercat

unread,
May 10, 2003, 7:25:53 PM5/10/03
to

"Automort" <auto...@aol.com> wrote in message
news:20030510110319...@mb-m28.aol.com...

What are you talking about? California is correct of course!

only to the list}abyss@nutter.net Dana Nutter

unread,
May 10, 2003, 8:11:25 PM5/10/03
to
On Sat, 10 May 2003 16:25:53 -0700, "Silvercat" <silv...@qnet.com>
wrote:

Yes! Well Southern California anyway.

Automort

unread,
May 10, 2003, 9:16:30 PM5/10/03
to
>From: Dana Nutter {respondonly to the list}ab...@nutter.net

>What I'm trying to do is look beyond the IE bias and find out if there
>are any constants across different languages/families. Or, just get
>some ideas how different languages handle modals.

A good idea.
But I don't consider it a bias. More material is available to Westerners, the
majority of persons here probably, on IE than on other families.

Automort

unread,
May 10, 2003, 9:19:12 PM5/10/03
to
>From: "Silvercat" silv...@qnet.com

>
>What are you talking about? California is correct of course!

They are too confused and believe themselves in the vanguard.
A place that has blended languages for a couple of centuries (Spanish, Correct
English, and German in this case) in relative isolation for most of it produces
a stable, mature language.

Automort

unread,
May 10, 2003, 9:20:06 PM5/10/03
to
>From: Dana Nutter {respondonly to the list}ab...@nutter.net

>


>Yes! Well Southern California anyway.

Well, it isn't the gibberish one hears in English movies.

Anton Sherwood

unread,
May 10, 2003, 3:54:11 PM5/10/03
to
Automort wrote:
> The English spoken in Texas is the correct,
> benchmark form of the language.

My long search is over!

Anton Sherwood

unread,
May 10, 2003, 8:46:37 PM5/10/03
to
>> Dana Nutter wrote:
>> > What are the most common modes in natural language?

> Anton Sherwood wrote:
>> Indicative, imperative, subjunctive, conditional, optative
>> are the ones familiar to me - but this shows an IE bias.

Cellus Purfluxius wrote:
> narrativo, potential...

How is narrative different from indicative? Is it used to report
matters not known personally but heard from others?

David Thomas

unread,
May 11, 2003, 2:21:16 AM5/11/03
to
In article <3EB19A2E...@gmx.net>, nyra <ny...@gmx.net> writes:

>Klaus Scholl schrieb:
>[...]
>> Also the Language is Ugly (Not that french or english would be less
>> ugly, but i expect from an Auxlang that
>> it aims for Aesthetics in Sounding, like Quenye does.)
>
>Now _that_ is really subjective.

Klaus is somewhat of a Disgruntled Philosopher, a stereotype I often use for
description that includes a quality of not understanding that some
things--indeed, many things--are just frickin' relative!

- Vae
Sleep, Fate, Death, and I sat one sunday down at tea.
Fate offered up his Ziggy mug before I poured,
Sleep yawned in his PJs, seeming mildly bored,
And Death politely asked, "Another pirouline?"
- Vae

Cellus Purfluxius

unread,
May 11, 2003, 6:02:43 AM5/11/03
to
Anton Sherwood wrote:

I think it is used to tell what another person has said, but what you
have not witnessed yourself.

Cellus P


David Thomas

unread,
May 11, 2003, 6:25:41 AM5/11/03
to
In article <20030510110319...@mb-m28.aol.com>, auto...@aol.com
(Automort) writes:

>The English spoken in Texas is the correct, benchmark form of the language.

Does that have anything to do with the president?

It is loading more messages.
0 new messages