Friday, February 5, 2010

Ich bin ein aux-langer. nature of the language design community

We artificial human language designers are not community-minded.
The sheer chutzpah required to think one can make a new frickin' language is staggering.
These aux-langs rarely seem to be collaborative efforts.
We are lone wolfs.

----
I have been corresponding with one of the most talented aux-langers this week. He is known for being aggressive and confrontational about language. He even sought out language-specific forums to say as much.
While he seems to have moderated his views in recent years, he said something at the end of the last e-mail to me.
He pondered which aux-lang designer would 'win'. By implication, the rest would lose.

I once said:
"If I can contribute just a few insights into some
hypothetical eventual world language then I will be
satisfied.
Yes, I'd really like if mine was selected (pending
posting)."

-----
Funny. The rest of the world is either oblivious to our existence, or bewildered and apathetic if not.
And here we are, turning on each other and savaging each other.
Certainly, I have over-reacted defensively.

The question is whether we more closely resemble:
1) The Highlander movie. "There can be only one!" The villain is trying to kill all the other immortals in order to steal their power.
2) The Round Table and Camelot. Yes, there must be a king- whoever can pull the sword out of the stone. BUT the round table is a way to make all equal and prevent bickering about station. And they are united in cause.

At what point do we stop wasting energy trying to put down every other aux-lang design?
At what point do we all agree to unite behind a common banner?
What is the 'sword in the stone' in this contest?

Well first we look at the biggest success story to date, though it have plateaued and remained stagnant for a century.

Finnish linguist Jouko Lindstedt, an expert on native-born Esperanto speakers, presented the following scheme[45] to show the overall proportions of language capabilities within the Esperanto community:
1,000 have Esperanto as their native language.
10,000 speak it fluently.
100,000 can use it actively.
1,000,000 understand a large amount passively.
10,000,000 have studied it to some extent at some time.

D: who will break through this plateau? Who can geometrically increase its base of speakers into the mainstream?
We need not use a "League of Nations" - the UN -for this.
A regional or even national government would work. NGO's could work. To save on translation costs.
The private business community of MNC's have a motivation. To create an in-house interlang. Or to allow a single advertising message to reach more people. In essence, each natural language community is another niche market.
The initial success of each aux-lang to date has been essentially private, by the founder and a band of die-hards.

OK so who gets to be king? Each of us benefits from all efforts that have gone before. We stand like giants only because we stand on the shoulder of generations of aux-lang designers.
I'd argue that we should back the language - all of us SHOULD back -
1) suited to be an aux-lang. Easy and then powerful.
2) the first to get both widespread recognition and a large group of followers.
3) able to promoted by PR and sales successfully. Politically viable. This *may* mean culturally neutral. It may not.
This could even mean a heavily pidgined natural language- who knows?

Well, looking at the above Esperanto stats touches heavily on this list. If easy but powerful, an aux-lang.
"Studied at some time", compared to "speak fluently" is a litmus test. Initial publicity can easy allow brief exposure to a language. But does enthusiasm wane in the face of learning obstacles? Is it unfriendly to too many linguistic backgrounds, rendering it no easier than some nat-lang for them? Do the grammatical elements at the intermediate level stump English speakers?

1,000 have Esperanto as their native language.
10,000 speak it fluently.
100,000 can use it actively.
1,000,000 understand a large amount passively.
10,000,000 have studied it to some extent at some time.
D: if we divide this all by 1,000, we get the following.
1 native speaker
10 fluent speakers
100 actively use
1,000 passively understand
10,000 had contact with it.
Native speakers are only possible after an entire generation. Native speakers are also only possible with fluent parents.
So I focus on fluency.
"To some extent" is a trap, as is "passively understand".
That means someone looked at it and gave up promptly, or can only understand some and hardly speak it.
This level of competence in English means a job barrier to an immigrant, and exclusion from an English community.
Only 1 in 1000 Esp-o-ists who had contact with the language become fluent. Pretty damning.
Only 1 in 100 passes "passively understand" to fluency.
1 in 10 surpass "actively use" to fluency".
"Passively understand" suggests a language which has basic elements too difficult for many learners.

I would argue that we see here, in 2 aspects, a test for the future king's "sword in the stone".
1) a community of speakers of a certain size. Critical mass.
2) the ratio between "passively understand" and fluency. For esp-o, that ratio is 1 in 100.
ONE PERCENT of those which begin down the path to learn Esperanto are able to fully function in it.
That is NOT a great batting average. Not exactly Babe Ruth's hitting scores.

The best way to break through the plateau of speakers is to provide utility. Yes, that means a link to some ideology. In the case of Esp-o, that patron group would be internationalists.
These would be the critical "early adopters". They are necessary - though not sufficient.
But then what? Here is where ease of learning and powerfulness become the issue instead.
Those outside of the demographic niche without any particular interest in internationalism want utility.
A good portion of utility is energy and time spent to acquire it. If the learning requirements are too high, then the reward for learning the aux-lang is not likely to make learning it worth the while of the non-ideologically motivated learner.

Here is where the "cultural neutrality" can be of use. Incorporated properly, this principle ought to make the structure of the language easier to learn. Freed from the fetters of single linguistic tradition, pure form and function are emphasized, leading to a sleek and non-dogmatic phonology, grammar and syntax.
I'd argue that the counterpart, the 4th implicit consideration, is how marketable the language is to the powers that be.
Realpolitik. That means the USA. Eventually China, then India. In a general sense, anybody already with a UN Security Council veto vote. For France and Russia also.
France. I have a love-hate relationship with them. They obstructed the best effort at an IAL to date.
But if they had not done so, then I'd be wasting my breath now, and would not have even started down this path I've chosen.

----
Let's take a brief look at both French and Russian.

French.

The velar nasal /ŋ/ is not a native phoneme of French, but occurs in loan words in final position such as parking or camping.[1] People who have difficulty with this sound replace it with a prenasalized [ŋɡ] sequence instead of a single consonant [ŋ].[citation needed] This sequence also appears almost systematically where there is a possible liaison with the initial vowel of a word pronounced just after it

D: Hmm. No getting around that for Decimese, not with a Mandarin layout for the words.
Sounds like they manage to cope.
inné [in(n)e] ('innate')
D: some consonant gemination.
D: also on an initial stressed syllable's vowel.
formidable [ffɔrmidabl] ('terrible')

D: their even-timing, even on most stressed syllable, is irrelevant to Decimese.
French is SVO , but it incorporates or cliticizes objective pronouns before the verb. ...
Bon!

Mandarin is SVO but has many SOV characteristics.
D: so SVO is really about appeasing English speakers, who are least inclined to pick up a second language, let alone an aux-lang.
D: I see a lot of diphthongs. Often nasalized prior to a word-final nasal consonant.
I never could say onion- oignon - properly...

----
Russian.

Diphthongs
Russian diphthongs all end in a non-syllabic [i̯], which can be considered an allophone of /j/, the only semivowel in Russian. In all contexts other than after a vowel, /j/ is considered an approximant consonant. Phonological descriptions of /j/ may also classify it as a consonant even in the coda. In such descriptions, Russian has no diphthongs.
D: diphthongs are plain out!

They have lots consonants. Huh. I don't see NG. Gah. Well, no way to avoid it with a Mandarin basis.
It seems I will make some enemies along with friends...

They devoice word-final consonats. I don't have any. I wonder what they do to word-mid consonants?
Doesn't much matter. It's allophonic in Decimese, a mere convention for Anglos, really.

Consonant Clusters
Russian is notable for having fewer phonotactic restrictions than many other languages,[45] producing word-initial clusters that would be difficult for English speakers. Some, such as in встретить [ˈfstrʲetʲɪtʲ] ('to encounter'), can have as many as four segments.
D: Interesting. They can handle very complex consonant clusters -but ONLY on the FIRST syllable.
OK, I'll keep that in mind as I develop later, advanced optional rules for "chunking" word particles into the main related word.
E.g. If LA HO PIBUM, then PL- versus -BL- will be desirable.
Even more into the future, I am trying to plan for the option of new consonant clusters of triple form, as well as the possibility of the introduction of the TH pair. I use the English word STRENGTHS as a guide to the potential of this approach.
But that is another story.

Some languages are more complicated: Russian allows all possible combinations SVO, OVS, SOV, OSV, VSO, VOS.
D: they should manage to the SVO word order without too much difficulty.
----
D: in summary, not too bad, though with a few bumps along the way.

-----
Decimese. Coupla thoughts.

D: I once theorized about an aux-lang that used consonants for vocabulary, but used vowels for grammar.
Essentially what English does in part in SING and RUN.
SING - verb. SONG - noun. Tenses - SANG, SUNG.
RUN - verb and noun. RAN- tense.
I think that is called ablaut.
Well, if we gut the vowels out of Decimese, and use the taxonomic vocabulary approach, we could explore this.
Essentially instead of Ygyde's V - C- V... format, we get a C (gap) C (gap) format for lexical meaning.
THEN we insert -V- -V- for grammar.
In English, prefixes tend to provide nuanced meaning (pseudo-, re-). Suffixes tend to be more grammatical (-ment, -ness).
If we resort to a 2-syllable base for most definitions, much like Lojban, then we'd have the format C*C*n. N for nasal.
It is fun to explore. Vowels, then diphthongs for nuance- subsets of meaning. Conversely, consonant clusters (STR-.BL-, -RN) for more lexical nuance.
----
D: I was beginning preliminary planning into standard vocabulary items. That means less focus on HLRWY syllables, and more on PB TD et al. I looked to the future, to a distant hypothetical future where the basic Decimese is widely used.
At that stage, and in certain linguistic backgrounds by competent speakers, consonant clusters become an option to improve brevity. A language which is superficially Mandarin sans tones simply cannot be that terse!
Anyway, the a look at the use of HLRWY in consonant clusters in English yielded a coupla results.
1) I knew H could not be used.
2) LRW can all be used by 3-4 of the basic 5 voiced/voiceless consonant pairs.
(Shameful admission here - I don't use the 2 Mandarinesque pairs since I am less comfortable with them as an Anglo...!!!)
3) Y cannot be used either. This suggests that Y within a word should likely denote diphthongs.
Back to 2) The consonant pairs that the THREE special consonants can form vary, and need to be planned for.
IF I plan enough at this stage for what L, R, and W (as well as H and Y) are to denote, then I can match those 3 with the appropriate pairs to allow likely, common, and sensible consonant cluster truncations.
For example, body parts, due to our bilateral symmetrical layout, often need to be expressed as pairs- a duality.
Arms. 2. Legs 2. Related items: pants (not pant-leg). Socks. Shoes. Sometimes we have 2 sets of 2. <:
That reminds me of those old sci-fi horror/action movies. Alien. Aliens! Aliens.... ez?
Then there are collective noun concepts to consider. Grains comes to mind.
Also, are we referring to stones as in rocks, or the unit stones?
Obviously, LRW must be consciously match to PTFSK et al during the design stage of Decimese.
There is no flying by the seat of my pants. I must understand the overview of the entire language before I select even one vocabulary item!
H and Y are special. H can be dropped to form a diphthong, or in theory a geminated vowel.
However, Y cannot. If Y as a diphthong is simply the I (ee- machine) sound, then it overlaps with I in forming diphthongs.
This suggests that Y in the form of word particle Y + vowel (YV) should be relegated to a certain particular class of modifier words.

My head hurts.

D: in retrospect, I have no idea if using the word-final nasal consonant for grammatical indication yields better results that the word-final Esperanto vowel, or the word-initial Ygyde vowel to denote grammar.
I do think it is appealing to Mandarin speakers. I also think that is allows very easy parsing of word boundaries.
Some indication was necessary to off-load the need for overt word particle grammatical indicators, eve with rigid word order.
Lord knows that tying up M, N, and NG for non-lexical functions severely reduces the number of syllables available for vocabulary!
In the end, this will be a good thing. I will be forced to make a very dense and concise system to generate vocabulary. But right now it is kicking my butt!

2 comments:

Dino Snider said...

A language that has motivated students reaching the fluency of Esperanto at 10x the rate - so 1 in 100 versus 1 in 1000 - would reach an important benchmark. This would involve a staggering magnitude shift compared the old 'Model T Ford' of the aux-lang world. Doable, though - just like a high-tech hybrid could accomplish that in performance compared to a model T Ford. So there is the goal. I believe it is doable.

dino snider said...

If a child could be taught to be fluent at Lojban, then we could see what the language can accomplish. For example, does it really encourage clear thought? The quandary is that no adult is able to be become fluent.
So who would teach the child?
And would this constitute child abuse? <: