Forum:Group Project

Start-General-Vocabulary-Organization-Innovations I have split this up to ease things The Emperor Zelos 15:48, May 28, 2010 (UTC)

'''  since this is growing FAST i suggest you QUICKLY split things!  The Emperor Zelos''' 22:56, May 30, 2010 (UTC)

Alphabet and Regularity (proposed)
It would be best to have absolute regularity with all parts of speech. But in order to do that, we need to decide which letters can be used.

Vowels: a e i o u

Consonants: b d f g k l m n p r s t v w y z

Diphthongs: ng, ay, ey, oy

So for regularity: Anyone want to add/subtract anything? Razlem 23:48, May 27, 2010 (UTC)

Maybe not so regular. The -o rule drives me nuts in esperanto. How about nouns end in a consonant, and thenSo for regularity:

Stuff that requires shorter words should be able to end in consonants. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 02:09, May 28, 2010 (UTC)

As long as it's a vowel followed by a consonant. The whole idea is to be able to change meaning without creating a new word/ending (Esperanto: teatro- "theater" teatra-"theatrical"). This is assuming that the roots all end in a consonant. Razlem 02:24, May 28, 2010 (UTC)

I was thinking, say theatre is "teater", then theatrical is "teatera," and theatrically is "teateri." Much like Esperanto. Note the changes I made above to verbs. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 03:24, May 28, 2010 (UTC)

I suppose, but would every word end in R? If it has to be a consonant, L may be easier to say (and to hear). Razlem 03:51, May 28, 2010 (UTC)

Well no. Not every noun has to end in r. There could be a group of consonants that would be natural nouns (say r, p, t, k), artificial nouns (s, f, g, b, d) or something like that. But lol I was thinking a Korean/Japanese system where there's only one liquid which could sound like an l or an r depending on the situation. But check out my syllabic system and that would probably change all this. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 04:02, May 28, 2010 (UTC)

I think Razlem's system is fine, the -o thing in Esperanto is fine and it sounds great. —Preceding signed comment added by TimeMaster (talk • contribs) 17:52, May 28, 2010 (UTC)

That was not biased at all… —Detectivekenny; (Info) Preceding text certified by R. Xun as of 02:00, May 29, 2010 (UTC)

Word Building
I'm not a huge fan of consonant clusters. I was thinking of a system similar to Japanese, with a successive alternation of vowel/consonant. Of course, this would eliminate some diphthongs. Razlem 01:56, May 29, 2010 (UTC)

That actually drives me nuts because I tend to find happiness in scant syllables (like Mandarin or Vietnamese which can be spoken surprisingly quickly and acccurately despite ambiguity in the former and large phonemic inventory in the second). I don't like consonant clusters much either, though. Maybe a system more like Thai/Khmer would be better. Like you can have some consonant clusters like phra, khla, phnom, krung, but nothing to the extent of Dutch. It increases the phonemic inventory, but you don't hear Thai speakers slow down to the hebetude of English. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 02:05, May 29, 2010 (UTC)

We could do a maximum of any two consonants, any more must be separated by a vowel. I'm just trying to keep things regular. Razlem 02:58, May 29, 2010 (UTC)

Okay. And two consonants can't be next to each other at the end of a word. But we still need to decide on the phonology, right? —Detectivekenny; (Info) Preceding text certified by R. Xun as of 03:05, May 29, 2010 (UTC)

Right, but I'm not sure how to go about doing that. Voting is an option, but entrants shouldn't be able to vote. Razlem 03:42, May 29, 2010 (UTC)

Why not just debate on the divergences:


 * 1) Include "ǝ" as a vowel?


 * 1) One liquid (r) or two (l and r)?


 * 1) Voiced consonants?


 * 1) Consonants: nh, h, nq, q, x?


 * 1) Tones?


 * 1) Diphthongs with w? —Detectivekenny; (Info) Preceding text certified by R. Xun as of 14:02, May 29, 2010 (UTC)

What we must not forget is that it is an auxiliary language. So, we must keep it simple. Not to the point of minimalism, but we should avoid very complex clusters or very uncommon sounds. But there is one thing we can do so that it is at the same time simple and phoneme-rich: allophones. See e.g. how in Classical Arabic there are only three vowels, 'a', 'i' and 'u', but they are actually realized as at least five different sounds, 'a', 'e', 'i', 'o', 'u', because some consonants affect the pronunciation of vowels (mostly of 'a'). So we can establish limits of variation for some sounds. This would have several consequences. Imagine for instance we decide not to make a difference between 'r' and 'l'. "International" words would keep the original sound ("telephone" would be something like 'telefon'; "formula" would continue like it is), while original words would have only one letter or the other. Of course, this free variation would be applied wisely, only to a defined set of phonemes, for it should be a positive characteristic of the language, not something to create a complete mess. Panglossa | Talk 14:26, May 29, 2010 (UTC)

I don't get the big deal with "international" words. Chinese doesn't do that stuff and even then, it doesn't let "international" words disrupt its phonology. For example, there are some words that are made using compounding, i.e. diànhuà, lit electricity-speech, meaning telephone, or gōnggòngqìchē, meaning "bus" which translates literally as public-vapor-vehicle. And then there are some which are edited to fit the phonology, like hànbǎobāo, which translates to Chinese-castle-bag (???) but it doesn't add extra letters to the alphabet. I think we should take an approach like the Chinese because it reduces irregularities. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 15:16, May 29, 2010 (UTC)
 * The reason I proposed that is to allow words continue to be pronounced the way they already are pronounced throughout the world. Imagine e.g. the language doesn't include the letter 'v', but we decide to keep words like "television", the sound [v] would be allowed in this word. That is more of a tolerance level, considering it is an international language, rather than a complex dual pronunciation system. More or less the same thing that happens in Esperanto: although the letter 'r' has a standard pronunciation, Americans pronounce it like English 'r', Italians like Italian and so on, and this doesn't interfere with communication.Panglossa | Talk 18:07, May 29, 2010 (UTC)


 * 1) Include "ǝ" as a vowel?

Perhaps as an accented letter or diphthong

Two, Alveolar lateral approximant for L, alveolar trill for R.
 * 1) One liquid (r) or two (l and r)?

L, M, N, R, V, Z
 * 1) Voiced consonants?

nh, h, nq , q , x (as kh?)
 * 1) Consonants: nh, h, nq, q, x?

Not for an auxlang
 * 1) Tones?

Ok
 * 1) Diphthongs with w?

Razlem 15:36, May 29, 2010 (UTC)

1. I would rather not include diacritics. They are sort of redundant. I think many people can type ǝ easily on their keyboard, and keep in mind it would be easier to write than ǒ or something. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 15:48, May 29, 2010 (UTC)

2. Cierto.

3. Let's add NG and take out V and Z because neither exists in Chinese or Korean, and voiced obstruents might be difficult for Chinese and Korean people anyway. :P

4. X as in kh.

5. At least keep long and short consonants. They seem intimidating at first but they come naturally after a short while.

6. Cierto. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 15:48, May 29, 2010 (UTC)

ǝ can't be easily typed on an English Mac. There aren't any keyboard shortcuts for it. Razlem 16:19, May 29, 2010 (UTC)

Nevermind, I figured out a way :)

For future reference (Mac Users)

System Preferences > Language > Keyboard Input > Azeri

Set up the [Command+Space] shortcut to switch keyboards

ə is the apostrophe/quotation key

Razlem 17:56, May 29, 2010 (UTC)


 * Lol yes it can; I have a mac and I can type almost anything. Set the input to "US extended." There are several random ways to create symbols. Fot ǝ, try shift+alt+; and then press e. You can easily type ŋ, ɲ, ƥ, ǒ, ɔ̂̀, Ƣ, ő, ḿ, ũ, ŭ, ŷ, ì with a little practice. —Detectivekenny; (Info) Preceding text certified by R. Xun as of 18:08, May 29, 2010 (UTC)
 * Please don't forget us siple mortal people who don't have Macs... :D Panglossa | Talk 18:14, May 29, 2010 (UTC)
 * Lol there's a good keyboard shortcut for you. Double click: ǝ press ctrl+c, then the useful keyboard shortcut is control+v. :) —Detectivekenny; (Info) Preceding text certified by R. Xun as of 18:18, May 29, 2010 (UTC)

Panglossa | Talk 18:07, May 29, 2010 (UTC)
 * 1) Include "ǝ" as a vowel? -> yes, provided we find a simple way to represent it (that is, no accents or unusual characters)
 * 2) One liquid (r) or two (l and r)? -> A single one, maybe varying, like in Japanese or Korean.
 * 3) Voiced consonants? -> OK
 * 4) Consonants: nh, h, nq, q, x? -> What sounds do they represent? We should rather avoid digraphs.
 * 5) Tones? -> I like tones, but for an auxlang, definitely no!!!
 * 6) Diphthongs with w? -> yes

1. Well if "ǝ" is the only special character, you could just find it on wikipedia and use copy paste. Or if you have a mac you could just use shift-alt-;+e.

2. Okay. I thought that would be better, but a trill is fine too.

4. Yeah, keep x (x in IPA) and h (h in IPA). The others are too foreign for most.

5. By tones I didn't mean those Chinese things that have American's begging pitifully to Chinese people to understand them. I mean the relatively common lengthenings that exist in almost all languages (except languages with "real" tones lol) which could easily be demonstrated with an accent mark as in Hungarian or Irish or by dooubling leetteerrs as in Finnish. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 18:18, May 29, 2010 (UTC)


 * Do you mean Vowel Length? Or simply Stress change? Because those are very different from tones. Also, what about Pitch Accent? It seems to be the easiest form of tonal usage (only having two tones; up down, or down up). That, I think, could be easily represented by the use of graves and acutes. Also, I think the easiest thing for people though would be Stress, like in Spanish or even English (compare produce to produce, or ésta versus está). Much more compact! :) LctrGzmn 00:38, June 1, 2010 (UTC)
 * Double letters are fine (and would be way easier to type). Also, the shift-alt-e combo won't work for me, even with U.S. Extended. Razlem 18:57, May 29, 2010 (UTC)
 * I didn'tsay that. I said shift+alt+; and then e. (Detectivekenny)


 * №́<-- That's what I get from shift, alt, ;, e.
 * XDD you have to let go of shift and alt before you press e. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 01:31, May 31, 2010 (UTC)
 * Ah, I see. But for me it's easier to just do command+space+;

Final Decision: Phonology
To make a final decision, we need the IPA for the vowels, consonants, and diphthongs:

Vowels: a e i o u ə

Consonants: b d f g h k l m n p r s t w x y

Dipthongs: ng ay ey oy əy aw iw ow əw

Tones: aa ee ii oo uu əə


 * What do the letters x and y stand for in the Consonants list? If x stands for its IPA value, what about merging it with h, as they are veri similar and are in free variation in a great number of languages? As for y, I guess it stands for [j].
 * Should we add palatal sibilants [ʃ] and [ʒ]?
 * What does Diphthong ng stand for?

<span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 17:52, May 31, 2010 (UTC)
 * What does the Tones line mean, exactly?
 * If we clarify these questions, this proposed set of phonemes has my green flag.

<span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 17:54, May 31, 2010 (UTC)

From my understanding, y= [j]; a merge of h and x is agreeable. I vote X should take the place of both

NG =[ŋ] I guess it's technically a digraph.

What I'm not sure of is the sound for ə. In Azeri, it's ae, but in English, it's a mid-central reduced vowel (if it's the former, the combination ae would suffice)

The tones represent longer sounds.

Razlem 19:04, May 31, 2010 (UTC)


 * So, they are not "tones", they are long vowels. Vowel length as a distinctive feature may be interesting, but is it desirable for an auxlang?
 * For the schwa ([ə]), we could do like in Welsh and use the letter 'y' for it. So we keep using only plain ascii characters and use one letter that would otherwise be left out (we then use 'j' for the [j] sound, sticking to IPA).

<span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 02:34, June 1, 2010 (UTC)

Long vowels do exist in almost every language, English "law", Spanish "manzana", so I think it would be good to have for a Terran auxlang.

That would work for the schwa and the y, but that was just my interpretation of the letter. I'm not exactly sure what Detectivekenny had in mind when he suggested it.

Razlem 02:58, June 1, 2010 (UTC)


 * What I mean about long vowels is that it may present some problem if we use it as a distinctive feature. Even languages which have long vowels (like English and Spanish as you said) do not rely on this distinction to tell one word from another.
 * But I've just thought of something else: if we settle for a monosyllabic language (or something next to that), we could use vowel length as a morphological element. We could use it, i.e., to make plural of nouns or to indicate past tense on verbs, something like 'ka' "person", 'kaa' "people" or 'ne' "say", 'nee' "said". Or also some kind of derivation, like 'ne' "word", 'nee' "speak". <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 03:28, June 1, 2010 (UTC)
 * Maybe not from different words, but longer vowels tend to be where the stress is. Relating it to the topic below, we could use long vowels to denote stress.
 * Razlem 03:57, June 1, 2010 (UTC)
 * Use h for x/h. Chinese uses the letter h which makes the x sound except it makes the h sound before a. Also it may be a good idea to take out "l." Seeing how much trouble my three-year-old cousin has with it, it may be difficult for people not used to it. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 14:32, June 1, 2010 (UTC)
 * Alright, H is acceptable.


 * Many major languages use the L sound (Chinese, English, Arabic, Spanish...), it would be unusual not to include it in a Terran auxlang. If it's the distinction between L and R, we can allow a retroflex approximant. I'm curious, what is your cousin's parent language? Razlem 15:23, June 1, 2010 (UTC)
 * English :) Idk, it's okay to include it. I'm just suggesting.  <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 18:37, June 1, 2010 (UTC)

Stress
Fixed stress (like in Esperanto, Polish, Hungarian...) or variable stress position? In the case we decide for fixed stress, where would it be? (a good average would be next-to-last like Esperanto or Polish) In case we decide for variable stress, will it be predictable (like in Arabic or Portuguese) or not (like in English or Italian)? If not predictable, how shall we decide where it falls? <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 03:29, June 1, 2010 (UTC)

Penultimate is fine, especially if we're going to have regular endings. If we use the long vowel system, it could be variable. Razlem 03:52, June 1, 2010 (UTC)

Hungarian has first-syllable-stress and long/short vowels. That system works, even for long words like goocbye (viszontlaataasra). <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 14:32, June 1, 2010 (UTC)


 * I like the idea of stress following long vowels. A system like that of Classical Arabic seems very appropriate, as it is both predictable (does not require diacritics) and yields a good variety in pronunciation, contributing to the beauty of the language. Basically, stress falls on the last long syllable of a word. A syllable is considered long if it has a long vowel or if it ends in more than one consonant. One possible argument against this system is that it produces words in which stress falls long before the last syllable, what may not be desirable for an auxlang. But this is solved if we do like in Latin, which uses the same system but establishes a limit for stress placement (it can not go beyond the antepenultimate syllable). <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 15:30, June 1, 2010 (UTC)
 * Just use the Hungarian system. It works for words like szaamiigooteep with all stressed and also for words like magyar with none. It would be a better auxlang. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 18:37, June 1, 2010 (UTC)
 * Stressing all first syllables is rather clumsy to most people not used to it. Seems to me a bad idea for a conlang. Anyway, if our language tends to have mostly short words (one, two or tops three syllables), it does not present a problem though.<span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 19:13, June 1, 2010 (UTC)
 * I believe we decided on particles rather than agglutination, so stressing the first syllable shouldn't be a problem. Razlem 19:38, June 1, 2010 (UTC)
 * So the proposal, based on this discussion, is a fixed stress placed on the first syllable. Monosyllabic words do not have double letters (yes? no?). Stress in words with two or three syllables is denoted by an elongated vowel (words can not exceed 3 syllables). Razlem 00:26, June 2, 2010 (UTC)
 * Whoah, not what I was thinking at all. I was thinking just like Hungarian, long vowels are a whole new set of vowels so you have words with any number of long or short vowels.  And if we want compound words then better not limit our opportunities with compound words.  And fyi first-syllable stress is a little clumsy but you get used to it.  If you'd rather have final stress, that's less clumsy.  <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 04:47, June 2, 2010 (UTC)

Future Plans
I'm showing this page to a couple of my friends; one offered to create a Rosetta-Stone-like program when we're finished. I thought that was an excellent idea (but it might be while before we get to that point). Opinions? e easily on 02:08, May 28, 2010 (UTC)

Isn't this just sort of practice? <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 02:09, May 28, 2010 (UTC)

Wow, great plans for our dear offspring :D <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 02:14, May 28, 2010 (UTC)

Well, if we like how it turns out at the end, then we should try it. Razlem 02:18, May 28, 2010 (UTC)

Syllable Structure
(C = consonant; V = vowel; S = semivowel)
 * V
 * VC
 * VS
 * CV
 * SV
 * CVS
 * SVC
 * CVC
 * SVS

I guess we won't include syllables more complex than this (CSVSCC for instance...), I just listed all possibilities, please share your opinions. <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 18:21, May 29, 2010 (UTC)

My set of choice:
 * CV
 * SV
 * CVS
 * SVC
 * CVC

Syllables starting with a vowel (V, VC, VS) may lead to ambiguity in connected speech, mainly if we are going to stress on monosyllabic words/roots. We can establish a glottal stop to be used whenever a syllable begins with a vowel (just like in German), or at least when some ambiguity is possible.<span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 18:29, May 29, 2010 (UTC)

That set is fine with me Razlem 18:52, May 29, 2010 (UTC)

Add:

CCV

CSV

CCVS

CSVS

CCVC

CSVC. If we are going to do a Thai system, then these are crucial. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 12:47, May 30, 2010 (UTC)

I guess a syllable of the CCVC type would be something like 'trap' or 'klat', is it right? So we should establish types of consonants (as it was already mentioned by someone) and limit those consonant clusters. E.g., if we use L for "liquid" (meaning 'r' and/or 'l'), we would have: But what happens if we have a compound word with two syllables like CLVCCLVC (imagine something like 'klaktrep'), what seems to me a bit too complex for an auxlang. Limiting syllables to CVC tops gives space for some nice clusters, CVCCVC 'kakrep' for example, which may be understood either as CV-CCVC or as CVC-CVC. <span style="font-family:calibri,lucida sans,arial,tahoma,verdana,sans-serif;font-weight:bold;font-variant:small-caps;">Panglossa | Talk 22:50, May 30, 2010 (UTC)
 * CLV
 * CSV
 * CLVS
 * CSVS
 * CLVC
 * CSVC
 * CLVL
 * CSVL

Listen to any youtube video in Thai (i.e. search "Thai News Broadcast"). You can see it works fine for them and you would never guess just by listening that Thai roots are mostly monosyllabic. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 01:32, May 31, 2010 (UTC)

I agree with Panglossa, and his initial choice of consonants is perfect to create a vocabulary. We shouldn't add more than what's required for an auxlang. Razlem 04:41, May 31, 2010 (UTC)

Large numbers of phonemes are necessary in a language for it to be efficient. If you do not want the language to be efficient, I should abdicate this project and work on my own auxlang because I work towards efficiency. <small style="color:#7F7F6F">—Detectivekenny; (Info) Preceding text certified by R. Xun as of 04:54, June 2, 2010 (UTC)