The "particles conlang"

[Zaromachian]

Good morning! (or evening, it depends)

Allow me to introduce you to my current conlanging project, one I like to call "the particles conlang" (obviously a working title).

I came up with the idea while reflecting on japanese particles. I said to myself, "they come after the word they mark; if you start paying attention to what the speaker is saying only when you hear a particle like は or が, you might be missing what he's talking about."
So I started thinking about a language with particles placed before the nouns or phrases they refer to, so as to put the listener in "subject" or "object" mode before they hear what the subject, or the object, etc. is. It is an a priori language, and I think some features I want it to have may be absent from natlangs altogether.

I'm not planning on giving this language a distinct culture, but I have something in store should I change my mind.

So, without further ado, here comes the...

*** (not terribly original, I know) PHONOLOGY ***

= Consonants =
Nasals: /m n ɲ/
Plosives: /p b t d k g/
Fricatives: /ɸ β s z ʃ ʒ h/
Affricates: /ts tʃ/
Approximants: /j w/
Lateral approximant: /l/

= Vowels =
Front: /i e ɛ a/
Center: /ə/
Back: /u o ɔ/

*** PHONOTACTICS ***

The language has a strictly syllabic structure; every syllable is of the form CV. Once again, I'm not running for complexity here.

The phoneme /e/ is realized as [ɛ] in stressed syllables, [e] everywhere else.
Likewise, the phoneme /o/ is realized as [ɔ] in stressed syllables, [o] everywhere else.

The stress almost always falls on the first syllable of each lexeme. Several accents may therefore be present in compound words.

The schwa is mostly used to represent sounds from other languages, such as isolated consonants.

*** ROMANIZATION ***

/m/ > m
/n/ > n
/ɲ/ > ñ
/p/ > p
/b/ > b
/t/ > t
/d/ > d
/k/ > k
/g/ > g
/ɸ/ > f
/β/ > v
/s/ > s
/z/ > z
/ʃ/ > x
/ʒ/ > j
/h/ > h
/ts/ > c
/tʃ/ > ç
/j/ > y
/w/ > w
/l/ > l
/i/ > i
/e,ɛ/ > e
/a/ > a
/ə/ > ë
/u/ > u
/o,ɔ/ > o

*** PARTS OF SPEECH ***

This is where the language starts becoming a little bit "alien", I'm afraid.
There are three parts of speech in this conlang: nouns, particles and specifiers.

Nouns belong to one of several categories that largely determine what particles and specifiers may be applied to them.
- Material nouns represent physical objects (such as kuço "apple" or paji "table").
- Abstract nouns refer to concepts, immaterial things, places and time periods (such as kumu "thought, idea").
- Action nouns obviously refer to actions, processes and sometimes the results of actions (such as koyobe "drinking" or lexe "walking").
- State nouns represent qualities that may be possessed by other nouns, or states other nouns may find themselves in (such as tahi "hot").

Specifiers, of course, specify what thing we are referring to when using a noun. This is a part of speech whose workings I can picture, but am unable to explain for now because I haven't come up with many specifiers yet. Basically, the idea is, if you say "kuço", you're talking about apples in general; but you may be wanting to talk about one apple in particular, one that has already been mentioned in the conversation, or you may be talking about one kind of apple, etc. The way you clarify what you are talking about exactly is with specifiers. Numbers are also included in this category.
Some specifiers are suffixes, most notably the ones that express possession or accumulation.

Particles mark nouns and related specifiers by function.

MA: this is the particle of existence; it indicates that something exists or is present.
ex.: Ma di kuço -> There is an apple.

KO: particle of location.
ex.: Ko je paji ma kuço -> There are apples on this table [over there].

***

There's still more to come.
I'm sorry if this language looks messy for now.

[mèþru]

What you posted is phonotactics, not phonetics. European languages and Hebrew generally have particles before the noun, among many others. It is part of a concept known as head-directionality, where the most important word of a phrase is at one side of the phrase.

[Travis B.]

Both /e/ and /ɛ/ are transcribed as [e]; this vowel is pronounced /ɛ/ in stressed syllables, /e/ everywhere else.
Likewise, the character [o] is pronounced /ɔ/ in stressed syllables, /o/ everywhere else.

This would be the opposite, that you have two phonemes /e o/ which are conditionally realized as [e] versus [ɛ] and as [o] versus [ɔ] respectively depending on stress.

[Zaromachian]

Thanks for the correction, Travis.

I got one question, though: is it realistic / plausible / normal / etc. to have only two vowels that are pronounced differently depending on stress?

At first, I wanted to have /i e̞ a/ as front vowels and /u o̞ ɒ/ as back vowels, because I liked the symetry; so then I decided to spice up the vowel system, by dividing both mid vowels into mid-open and mid-close vowels to keep this symetry.

(Then I dropped the /ɒ/ because I don't like that sound outside of english, but that's another story)

[Travis B.]

Thanks for the correction, Travis.

I got one question, though: is it realistic / plausible / normal / etc. to have only two vowels that are pronounced differently depending on stress?

If you mean have one vowel phoneme that is always stressed and another that is always unstressed, that is possible, but I would only do that if there is not a one-to-one relationship between stressed and unstressed phonemes (e.g. a greater number of stressed "full" phonemes and a smaller number of unstressed "reduced" phonemes); if there is a one-to-one relationship between phonetically similar phones where one is stressed and one is unstressed, I would just analyze them as one phoneme.

At first, I wanted to have /i e̞ a/ as front vowels and /u o̞ ɒ/ as back vowels, because I liked the symetry; so then I decided to spice up the vowel system, by dividing both mid vowels into mid-open and mid-close vowels to keep this symetry.

(Then I dropped the /ɒ/ because I don't like that sound outside of english, but that's another story)

You could always have /i e a/ and /u o ɑ/ - but then I would go for /i e æ/ and /u o ɑ/ to make the low vowels more distinct.

[Nortaneous]

Qiangic languages are conventionally transcribed as contrasting /a A/, but I haven't seen any vowel diagrams, so I'd bet that just comes from some linguist somewhere in China not picking up on the unwritten knowledge that nobody uses { and a the way IPA intends.

[KathTheDragon]

nobody uses { and a the way IPA intends.

Given that that's SAMPA, I'm sure nobody does use { the way IPA intends... >.>

[Zaromachian]

Thanks for the correction, Travis.

I got one question, though: is it realistic / plausible / normal / etc. to have only two vowels that are pronounced differently depending on stress?

If you mean have one vowel phoneme that is always stressed and another that is always unstressed, that is possible, but I would only do that if there is not a one-to-one relationship between stressed and unstressed phonemes (e.g. a greater number of stressed "full" phonemes and a smaller number of unstressed "reduced" phonemes); if there is a one-to-one relationship between phonetically similar phones where one is stressed and one is unstressed, I would just analyze them as one phoneme.

I'm sorry, I don't know if you're talking about the frequency of accentuated syllables in the language as a whole, or the number of stressed vs unstressed phonemes in the inventory.

I'll give a couple of examples to try and clarify what I mean:
- "koyo" (a drink, a beverage) is pronounced ['kɔjo] because the first syllable is stressed.
- "lexe" (to walk, the act of walking) is pronounced ['lɛʃe], for the same reason.
- But "kumu" (a thought, an idea) is pronounced ['kumu]; even though the first syllable is stressed, its vowel is still realized as .

Given that most words will be 2-3 syllables long, plus the (always unstressed) particles, I'd say there should be 1 stressed syllable for 3-5 unstressed syllables.

[Yng]

Unless the distribution of the phonetic realisations (allophones) is predictable, then I would say analyse them as separate phonemes. But if /o/ when stressed is always realised as [O], that makes perfect sense as an allophone.

[Travis B.]

Thanks for the correction, Travis.

I got one question, though: is it realistic / plausible / normal / etc. to have only two vowels that are pronounced differently depending on stress?

If you mean have one vowel phoneme that is always stressed and another that is always unstressed, that is possible, but I would only do that if there is not a one-to-one relationship between stressed and unstressed phonemes (e.g. a greater number of stressed "full" phonemes and a smaller number of unstressed "reduced" phonemes); if there is a one-to-one relationship between phonetically similar phones where one is stressed and one is unstressed, I would just analyze them as one phoneme.

I'm sorry, I don't know if you're talking about the frequency of accentuated syllables in the language as a whole, or the number of stressed vs unstressed phonemes in the inventory.

I'll give a couple of examples to try and clarify what I mean:
- "koyo" (a drink, a beverage) is pronounced /'kɔjo/ because the first syllable is stressed.
- "lexe" (to walk, the act of walking) is pronounced /'lɛʃe/, for the same reason.
- But "kumu" (a thought, an idea) is pronounced /'kumu/; even though the first syllable is stressed, its vowel is still realized as .

Given that most words will be 2-3 syllables long, plus the (always unstressed) particles, I'd say there should be 1 stressed syllable for 3-5 unstressed syllables.

Remember, slashes are for phonemes, brackets are for phones; as you are speaking of how something is pronounced, you should use brackets.

[Hallow XIII]

Qiangic languages are conventionally transcribed as contrasting /a A/, but I haven't seen any vowel diagrams, so I'd bet that just comes from some linguist somewhere in China not picking up on the unwritten knowledge that nobody uses { and a the way IPA intends.

Consider that technically, <a> can be a front low vowel in IPA. Khmer is transcribed the same way, and its /a A/ really are front and back low unrounded vowels, at least for my teacher most of the time -- A sometimes becomes Q though.

[Zaromachian]

I've added romanization and the first part of the grammar, parts of speech.
The way I'm presenting the language may not be optimal, but hopefully you can figure out where I'm going.