New CBB

I have grammatical concepts and such for my language already, my problem is I'm not very good at the technical terms or how to not confuse people on here when trying to describe them. Can someone please take a look and help me organize it properly?

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

I'd be glad to help! :3

_________________
♀♥Ø

Thank youuuu! Where would you like me to start? Maybe the phoneme inventory, if that's the correct term. I can at least do IPA.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

Phonology should generally be organized this way:

Inventory

- Allophony

- Syllable weight and stress rules

- Constraints/Neutralizations (phonotactics)


With each section being dependent on the one prior.

_________________
♀♥Ø

I have a vague idea of what those are...

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

Use the Wikipedia, Luke...

_________________
UNUS•ET•UNICUS

Well, I know what they are, I just don't know how to do it for my language.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

The Language Construction Kit says you should start with the sounds.

The phonology should be done in the order Micamo recommended, for the reasons she briefly touched on.

---

Inventory --
What phonemes does your language have?

---

Allophony --
Are there some phonemes that are sometimes pronounced as one phone and sometimes as another, different phone? If so, for each phoneme, show all the phones it can be realized as. (You may also want to show for each phone, which (if any) phoneme(s) it might be the realization of.)
Different phones that may all be realizations of the same phoneme, are called "allophones".
Allophony includes describing the rules by which speakers decide (probably unconsciously) to pronounce a given phoneme as one allophone rather than (any of) the other(s).
Be careful that, if one phone could be an allophone of two different phonemes, the allophony rules of your language are such that, the listener can always be sure which phoneme the speaker actually means.

---

Syllable weight and stress rules --
Every word longer than one syllable has exactly one and only one primarily-stressed syllable.
What constitutes "primary stress" in your language?
How do speakers decide which syllable gets the primary stress?

Every word longer than three syllables, and also three-syllable words unless their middle syllable is primarily stressed, alternates (as nearly as possible) (in most languages) between secondarily-stressed syllables and unstressed syllables.
How do speakers "decide" which syllables get secondary stress and which are unstressed?
(The answer to this is called the language's "rhythm".)
What does "secondary stress" sound like in your language?
How does it sound different from "primary stress"? And from "unstressed"?

Some languages divide syllables into "heavy" and "light", and the pattern of "heavy" and "light" syllables is part of what the stress-assigning and/or rhythm-assigning rules are. Ordinarily heavy syllables are likelier (but often not certain) to have stress than light syllables.
Generally you want to avoid (sometimes with exceptions) having a stressed light syllable next to an unstressed heavy syllable.

You always want to avoid having two stressed syllables right next to each other.
You always (in most languages) want to avoid having three unstressed syllables right in a row.
As nearly as possible you want (in most languages) to avoid having too many pairs of two unstressed syllables in a row right next to each other.

Some languages don't "care" about "syllable weight". If yours doesn't, you don't need to worry about it.

Some languages have three "syllable weights"; light, heavy, and superheavy. As I understand it, and if I understand correctly, any "superheavy" syllable has to receive primary stress, so there can't be more than one in a word; so, in a compound word, if both components have a superheavy syllable, something has to be done to make one of them not superheavy.

The most-common rule is that primary stress goes on the penultimate (next to last) syllable.
The second-most-common rule is that primary stress goes on the first syllable.
The third-most-common rule is that primary stress goes on the last syllable.

See http://www.unileiden.net/stresstyp/ and/or http://wals.info/chapter/14 and/or http://wals.info/chapter/15 and/or http://wals.info/chapter/16 and/or http://wals.info/chapter/17 for more.

---

Constraints and Neutralizations (Phonotactics).
Phonotactics is about which phonemes can come just before which other phonemes; which phonemes can come just after which other phonemes.

"Constraints" are all the prohibited pairs. Usually you start with the most general (say, maybe, a fricative can never be followed by an alveolar), and go on to the most particular (say, a voiceless lateral fricative can never be followed by a voiced pharyngeal).

"Neutralizations" are how your phonology handles things when, as a word is put together, a forbidden pair arises. Somehow the word needs to be changed so that no forbidden pair remains.
This may mean leaving out one sound or the other.
It may mean sticking in a "helping" sound that has no meaning except to keep the forbidden pair separated.
It may mean "mutating" one of the sounds (changing it so that the pair is no longer a forbidden one).
It may mean having the two sounds switch places.

Because which neutralization gets used often depends on which syllables are stressed and which unstressed, "phonotactics" depends on "stress and rhythm rules".

---

---

Well, the inventory part is easy. Just pick some phones from the IPA and say they're the "central phones" of your phonemes, and write those symbols with / / around them instead of [ ].
For vowels, if you don't want to be exceptional, just use /a e i o u/; that's the most common vowel-phoneme inventory. Or, use /a i u/; that seems to be the second-most common, and the minimal one (every language has those three, many also have others).
For consonants, do include / m k j p w b h g N /. For each one of those, most of the world's natlangs have it.
Include also at least one from each of these pairs:
/ k / or / g /
/ s / or / S /
/ g / or / N /
/ n / or / l /
/ n / or / t /
/ n / or / d /
/ s / or / t /
/ v / or / z /
/ l / or / l_d/
/ +tS) / or / S /
/ +tS) / or / +dZ) /
/ +tS) / or / +ts) /
For each of those pairs, at least half of the world's languages have one or the other or both.
And there's more; I'll put it in a "spoiler", since maybe it's kind of long.


It's impossible to make an average-sounding consonant-phoneme-inventory without including some sounds that fewer than half the world's natlangs have.
But you might make a good one (if by "good" you mean "more-or-less typical") from just the consonant-phonemes that are used by one-third or more, or by 25% or more, of the world's natlangs:


See http://wals.info/chapter/1, http://wals.info/chapter/2, and http://wals.info/chapter/3.

---

Now, for phonotactics.

Go through your consonant inventory and decide which consonants can occur in syllable onsets, which can occur in syllable codas, and which can occur in both.
If you allow clusters in your onsets, decide which consonants can occur as the first consonant in an onset-cluster, and which can occur as the last consonant in an onset-cluster.
If you allow clusters in your codas, decide which consonants can occur as the first consonant in a coda-cluster, and which can occur as the last consonant in a coda-cluster.

Now go through all the ordered pairs of consonants that the above decisions might allow to come up in that order in an onset-cluster, and decide which ones your language would actually allow to occur consecutively in that order in an onset-cluster, and which ones it woudn't.
Also, go through all the ordered pairs of consonants that the above decisions might allow to come up in that order in a coda-cluster, and decide which ones your language would actually allow to occur consecutively in that order in a coda-cluster, and which ones it woudn't.

Decide then what your maximal syllable-structure can be.
For instance, I have a conlang whose maximal syllable-template is (C)(C)V(V)(C)(C).
VV in my 'lang would be a diphthong; in other languages it could be a long vowel.
I do not allow homorganic consonants to occur consecutively. Alveolars can't follow alveolars, retroflexes can't follow retroflexes, palatals can't follow palatals, velars can't follow velars, etc.
My conlang follows a "sonority rule" of the following kind.

If two consonants in the same syllable are on the same side of the nucleus (i.e., both in the onset, or both in the coda), then the consonant closer to the nucleus must be more sonorous than the one further away from the nucleus.
If two consonants in the same syllable are equally distant from the nucleus, then the consonant in the coda must be more sonorous than the consonant in the onset.

Such a rule is common in natlangs.

My "sonority scale" is; semivowels > liquids (including rhotics (some may be trills) as well as lateral liquids and also all approximants that aren't semivowels) > nasals > fricatives > stops.

However, you may wish your conlang's syllable structure to be (C)(R)V(C) at maximum, where "R" means "a sonorant", in this case a semivowel or a liquid (nasals are also sonorants, but you may not want them in that position).

Anyway;
Look at all the sounds that can be the last sound in a syllable, and all the sounds that can be the first sound in a syllable.
Say /x/ can be last and /z/ can be first.
If /xz/ can't occur in a syllable, do you want it to be allowed in a word?
This is mostly going to be either two consonants, last-in-a-coda followed by first-in-an-onset; or else two vowels, one in an open (codaless) syllable followed by one in an onset-less syllable.

Now, look at all the forbidden pairs, and decide which ones might be generated during word-building unless you do something about it. And decide, What are you going to do about it?

That will settle your phonotactics and your syllable structure.

---

So, that's the first two steps.

Does that help?

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

I'd like to make a few corrections to Eldin here.

Firstly, I meant "constraint" within the context of optimality theory. In OT "constraint" doesn't solely refer to forbidden pairs, but also to instances of productive elision, umlaut, epenthesis, mutation, metathesis, etc. The order in which the constraints apply is critically important, as the application of one constraint can stop another constraint from applying that otherwise would: This is called "bleeding."

Secondly, while Eldin's "neutralizations" are important (though I consider them a part of "constraints"), I meant something different: In many languages, not all phonemes are distinct in all situations. Navajo, for instance, doesn't distinguish /j w/ before back vowels. So it distinguishes ye/we but not ya/wa. We thus say that /j w/ are "neutralized" in this environment. Neutralizations should be explained alongside constraints because they're often the cause of constraints. For example, an epenthesis constraint that turns /np/ into /nep/ in some domains, because /n m/ are neutralized before another consonant.

_________________
♀♥Ø

Good point.
I'd explain "constraints" this way;
A "constraint" is something you'd rather not do, unless you have to in order to satisfy some more important "constraint".
So, make a list of the things you don't want to happen, then order them in order of how much you don't want them to happen.
If the only way to not break a less-important constraint is to break a more-important one (if the only way to abide by a more-important constraint is to break a less-important one), go ahead and break the less-important constraint.
"Forbidden pairs" may not be all there is to your phonological constraints.
And sometimes you may find it necessary to violate some of your constraints.



I completely misunderstood what you meant by "neutralizations".

I totally disagree on thinking that "neutralizations" (correctly so-called) are important, especially for a newby conlanger or a first conlang.

I think the things I mistakenly called "neutralizations" are much more important, and easier for a newby to see the point of, than the things to which the term "neutralization" correctly applies.

Not every language, probably not even the majority, has any (correctly so-called) neutralizations.

---

Also;

While a lot of what you've said is in fact important, and it's important in the order you said it, still, a lot of it is said in a way that a first-time conlanger or a newby won't understand.

Usually I'm the one who has to be told that.

But I know you can explain things I can't explain, and even explain things so I understand them when I didn't before your explanation.

So can you do that for the rest of this thread? (Not on other threads!)

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

That's a wonderful way to explain constraint ranking!



I'm not so sure about this one: (Some types of) harmony can be explained as a sort of neutralization; The type found in Chumashan languages, for example. I'm not so sure if we can write off harmony as being something beyond what a beginning conlanger should worry about. You can get away with lacking neutralization, I was just trying to explain why they should be explained alongside constraints/phonological processes.



I'm sorry. It can be difficult to remember what it's like to be a newbie at times. I should have explained the whole "constraints and neutralizations" thing at the beginning.

So, wake, do you have any further questions about how to organize phonology?

_________________
♀♥Ø

You're right that if his/her conlang has neutralizations, s/he needs to explain them.

I was just trying to say that things like harmony, and whatever other neutralizatons there might be, would be better known to a beginner one feature at a time, rather than all lumped together in the label "neutralizations", especially with that label unexplained.
(BTW http://www.sil.org/linguistics/GlossaryOflinguisticTerms/ doesn't have anything about "archiphonemes" nor "underspecification" nor "neutralization" in it.)

Even for harmony, the newby or first-timer or beginner (as the case may be) should probably go and look up what kinds of harmony there are, or, at least, what kinds are common and what else is typical of languages that have harmony. (For instance, they tend to be synthetic (agglutinating or fusional) rather than isolating/analytic, and tend to have mostly suffixes and not very many prefixes, and grammatical information is likelier contained in affixes than in clitics or independent words/particles or stem-mutations or lexical suppletion.)

Probably s/he should look up assimilation and dissimilation as well.

I guess it's more the term "neutralizations" that IMO probably isn't a good beginners' term, than the neutralizations themselves.

Similarly, the whole thing about "syllable-weight" may be a complication one may not have to consider in his/her first conlang; not all languages have weight-sensitive stress nor weight-sensitive rhythm. (Indeed, possibly not all languages have both stress and rhythm, though IIRC all have at least one or the other.)

---

(BTW: what kinds of neutralizations are there?)

---

Well-said and well-written. Thanks!


____________________________________________________________________________


@wakeagainstthefall: Have you made any progress yet on your Phoneme Inventory?
If so, how much if any progress have you made on your Allophony? your Stress and Rhythm? or your Phonotactics?

My Mon 09 Jan 2012, 19:51 post was edited at Mon 09 Jan 2012, 20:39 ; it now contains some suggestions for Vowel Inventory and Consonant Inventory, and something about Phonotactics.

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

Yeah, I tried posting the allophones last night, but it timed out and didn't save my post. It was wonderful.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

A a, (À à only used at the end of words) /ɑ/, /a/ at the end of words
B b /b/
C c /ʃ/
D d /d/
E e /ɛ/
F f /f/
G g /g/, gg /ʀ/
H h /h/
I i /ɪ/
J j /ʒ/
K k /kʰ/
L l /l/, at the end of words… I couldn’t find this sound.
M m /m/
N n /n/
O o /o/, /ʊ/ at the end of words
P p /p/
Q q /k/ at the end of words
R r /ɾ/
S s /s/
T t /t/
U u /ʊ/ (does not go at end of words)
V v /v/
W w /w/
X x /ks/
Y y /j/
Z z /z/

Ā ā, ai /aɪ/
Ç ç /tʃ/
Ē ē, ei /ɛɪ/
Ī ī /i/
Ō ō /ɔx/ used at the end of words or to separate from another vowel in what would be a diphthong (A diaresis I think it's called?)
Ó ó /o/ at the end of words
Ū ū /ʏ/
Ŕ ŕ /r/
th /θ/
Ţh /ð/
ou, (òu used to distinguish a masculine noun) /u/
oi /ɔɪ/
ao /ɑʊ/
ia /ja/ only used at end of words, everywhere else is "ya"

These are all the sounds in Ruic. It has a conscript too, but it needs 1 more majuscule and minuscule form of a letter and a majuscule form of another.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

Please let me know if it's too English-y or anything.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

_________________
L1:
Fluent (on a good day):
Written:
Beginner:
Working on: ~ Eil

Why am I not getting a response after I asked for help? This happened the last time I wanted help for my conlang on here. I'm a noob, I'd like help. I've got basic grammar and phonology. I don't know why I stop getting responses after awhile. Seriously, is someone going to continue helping me or what?

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

Where have you asked for help? I'd want to answer to your questions if I find them.

They wanted an allophony, I posted it, then asked if it was too Englishy, and no one responded.

_________________
Magnae clunes mihi placent, nec possum de hac re mentiri. -Multomixtor

Where? Post the link.