Vowel Harmony?

The best topics from Languages & Linguistics, kept on a permanent basis.
Nikolai(lazy)

Post by Nikolai(lazy) »

Eddy the Great wrote:That certainly puts some strict limits on the possible affixes.
It doesn't so much as put a strict limit on possible affixes (as pharazon said), it does do something else. It provides a limit on the variety of meaning that vowel color could provide to a suffix. So that < -ek >'ll always be < -ek > if the stem vowel is front and you'll never be able to pull off that same suffix with a seperate vowels of oposing frontiness* e.g. Attic Greek with its nom. acc. fem. singular in < -e:n > or < -a:n >but it's genitive plural in < -o:n >. Vowel harmony would destroy such a distinction.

* :mrgreen:

Now, there was a discussion long ago where certain things about consonantal harmony and nasal harmony were brought up. Anyone can refresh us on that?

User avatar
Ghost
Lebom
Lebom
Posts: 94
Joined: Mon Mar 31, 2003 10:59 am
Location: Great Britannia
Contact:

Post by Ghost »

OK, I kinda get it, but what's the point? How does it help?

Ghost :?
[url=http://www.emalaith.com/census.html]ZBB Census 2006[/url]

User avatar
pharazon
Lebom
Lebom
Posts: 192
Joined: Thu Sep 04, 2003 1:51 am
Location: Ann Arbor
Contact:

Post by pharazon »

Ghost wrote:OK, I kinda get it, but what's the point? How does it help?

Ghost :?
It doesn't really have a point; it's just a historical development: the vowels assimilate to be more like each other.

User avatar
gach
Avisaru
Avisaru
Posts: 472
Joined: Mon Feb 17, 2003 11:03 am
Location: displaced from Helsinki

Post by gach »

According to http://privatewww.essex.ac.uk/~spena/Ch ... r2.html#vh there's a hig-low harmony in Chukchee. It goes somewhat /E o a/ vs. /i u e/ the first secound and thrird ones being pairs. There's also a shwa that's neutral.

User avatar
Tayanrai
Niš
Niš
Posts: 10
Joined: Sat Apr 12, 2003 5:46 am
Location: Pacific Ring of Fire

Post by Tayanrai »

I don't know how you would consider a vowel harmony system like the one in one of my main conlangs, Kahrhtehno, to be. The 10 vowels of Kahrhtehno are put into a chart - vowels in a word can only be followed by the vowels that are adjacent to that vowel in the chart. For example the vowel 'ah' can be followed by 'a', 'oh', or 'o' because they're adjacent to each other, but it can't be followed by 'i' or 'u'.

Can someone tell me if a system like this exists in any known natlang?

~Tayanrai

User avatar
Drydic
Smeric
Smeric
Posts: 1652
Joined: Tue Oct 08, 2002 12:23 pm
Location: I am a prisoner in my own mind.
Contact:

Post by Drydic »

Tayanrai wrote:I don't know how you would consider a vowel harmony system like the one in one of my main conlangs, Kahrhtehno, to be. The 10 vowels of Kahrhtehno are put into a chart - vowels in a word can only be followed by the vowels that are adjacent to that vowel in the chart. For example the vowel 'ah' can be followed by 'a', 'oh', or 'o' because they're adjacent to each other, but it can't be followed by 'i' or 'u'.

Can someone tell me if a system like this exists in any known natlang?

~Tayanrai
Unless you have a vowel /e/ (I can't be sure-- my goddamn connection won't open RTFs for some STFU reason) It's just height harmony. It all depends on what vowels you mean.

EDIT: GODDAMN BROWSER!!!
Image Image
Common Zein Scratchpad & other Stuffs! OMG AN ACTUAL CONPOST WTFBBQ

Formerly known as Drydic.

dragonprince99
Niš
Niš
Posts: 6
Joined: Tue Dec 30, 2003 6:22 pm
Location: SoCal

Post by dragonprince99 »

I understand the basic ideas of vowel harmony as outlined in this thread, but there's one thing that's been bugging me. What do languages with harmony do with compound words?

If the first part of the compound has vowels that are not "compatible" with those in the second part of the compound, what happens? And which set of affixes (the ones agreeing with the first part or the ones agreeing with the second part) are used?

Thanks.
Unfortunately, I don't like this username anymore, but seeing as how I'm too lazy to start a new account, I'll have to stick with this.

Glenn
Avisaru
Avisaru
Posts: 316
Joined: Mon Sep 16, 2002 3:43 am
Location: Virginia, USA/Tiolu, Kiarlon

Post by Glenn »

dragonprince99 wrote:I understand the basic ideas of vowel harmony as outlined in this thread, but there's one thing that's been bugging me. What do languages with harmony do with compound words?

If the first part of the compound has vowels that are not "compatible" with those in the second part of the compound, what happens? And which set of affixes (the ones agreeing with the first part or the ones agreeing with the second part) are used?

Thanks.
To answer the first question: a couple of things. :) In the languages with vowel harmony that I know (chiefly the Turkic languages), compound words or foreign borrowings with "incompatible" vowels frequently remain unchanged, at least in their "official" forms. However, I have noticed that in Kazakh, for example, "harmonized" versions, where the vowels in one part of the word are changed to match those in another, do crop up, perhaps from colloquial origins, and I could easily see such forms becoming generally accepted. Languages with stricter vowel or consonant harmony probably do this more regularly (and more quickly) than those with "looser" tendencies.

To answer the second question: the Turkic languages are almost exclusively suffixing, with stress on the final syllable of most words (or before the personal endings on verbs), so it is the vowel in the final syllable of the word root that determines the vowels in the suffixes that follow. A language that was primarily prefixing could well do the opposite, with the prefix vowels determined by the first syllable of the root. I think such a language was brought up in an earlier thread--I can't remember whether it was a conlang or a natlag, however.

In other words, the general principle seems to be that the vowel harmony (in whatever form) of an affix or set of affixes is determined by the vowel in the adjacent syllable of the root, which makes sense. This is not to say, however, that you couldn't do things differently. :wink:

p@,
Glenn

dragonprince99
Niš
Niš
Posts: 6
Joined: Tue Dec 30, 2003 6:22 pm
Location: SoCal

Post by dragonprince99 »

Thanks very much for the answer...I have another question, then. :wink:

This might be sort of an artificial situation, but say that two words in the same language are differentiated only by the vowels that they have; for example, /kile/ and /kulo/ (in a made-up system where /i/ and /e/ are in one category while /u/ and /o/ are in another category). If one of these forms goes onto a compound word, then does it make the compound ambiguous? (Because it could be with the meaning of /kile/ or the meaning of /kulo/).

Or is this all a bit too contrived and these types of /kile/ vs. /kulo/ words never occur in reality?
Unfortunately, I don't like this username anymore, but seeing as how I'm too lazy to start a new account, I'll have to stick with this.

Glenn
Avisaru
Avisaru
Posts: 316
Joined: Mon Sep 16, 2002 3:43 am
Location: Virginia, USA/Tiolu, Kiarlon

Post by Glenn »

dragonprince99 wrote:Thanks very much for the answer...I have another question, then. :wink:

This might be sort of an artificial situation, but say that two words in the same language are differentiated only by the vowels that they have; for example, /kile/ and /kulo/ (in a made-up system where /i/ and /e/ are in one category while /u/ and /o/ are in another category). If one of these forms goes onto a compound word, then does it make the compound ambiguous? (Because it could be with the meaning of /kile/ or the meaning of /kulo/).

Or is this all a bit too contrived and these types of /kile/ vs. /kulo/ words never occur in reality?
Hmm...this is indeed an issue that might occur (if rarely). Most languages that I've seen are "tolerant" enough of combined compound words that the two elements (/kile/ and /kulo/) could retain their original forms without much trouble. If the vowel harmony is extremely strict, however, or the compounded forms of the two words "blur" together over time, then either the two might still be distinguished from context, or an additional element (such as a new word added to the compound, or a separate adjective or noun, say) might be added to make the meaning clear. I hope that helps...

p@,
Glenn

User avatar
pharazon
Lebom
Lebom
Posts: 192
Joined: Thu Sep 04, 2003 1:51 am
Location: Ann Arbor
Contact:

Post by pharazon »

dragonprince99 wrote:Thanks very much for the answer...I have another question, then. :wink:

This might be sort of an artificial situation, but say that two words in the same language are differentiated only by the vowels that they have; for example, /kile/ and /kulo/ (in a made-up system where /i/ and /e/ are in one category while /u/ and /o/ are in another category). If one of these forms goes onto a compound word, then does it make the compound ambiguous? (Because it could be with the meaning of /kile/ or the meaning of /kulo/).

Or is this all a bit too contrived and these types of /kile/ vs. /kulo/ words never occur in reality?
It does seem a bit unlikely, but there are few things that I think could happen:

-The normal rules would apply and people would just depend on context to resolve the ambiguity (I'm assuming the normal thing is to assimilate the vowels, if it wasn't obvious)

-Another word would replace one of the compounds

-The normal process would be resisted to avoid the homophony (not very likely, but it happens)

Cor
Niš
Niš
Posts: 5
Joined: Sat Aug 23, 2003 9:35 pm
Location: Wisconsin
Contact:

Post by Cor »

Is voiceless/voiced harmony feasible?

bi's sock puppet
Sanci
Sanci
Posts: 19
Joined: Thu Aug 28, 2003 12:46 am
Contact:

Post by bi's sock puppet »

Cor: That'll be consonant harmony, not vowel harmony. Reportedly, very few languages have consonant harmony.

Siride
Niš
Niš
Posts: 13
Joined: Mon Mar 17, 2003 9:53 pm
Location: UNC-Chapel Hill, North Carolina, USA

Post by Siride »

Consider examples in English like "reseed" and "recede". I would imagine context is more than good enough to handle such a problem.

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Post by Soap »

Oh, yeah ... don't worry about homophones, people in China and Polynesia get by just fine with dozens of homophones per word in some cases. Just as long as the two words that become homophones are words for totally different things, like "apple" and "telephone cord", you should be OK, because in ordinary situations nobody would confuse the two. Unless you want to be deliberately mischievous and give your language lots of words like "inflammable", which is ambiguous in English because there are two homophonous <B>in-</B> prefixes which in some contexts can mean opposite things.

Cor
Niš
Niš
Posts: 5
Joined: Sat Aug 23, 2003 9:35 pm
Location: Wisconsin
Contact:

Post by Cor »

bicoherent wrote:Cor: That'll be consonant harmony, not vowel harmony. Reportedly, very few languages have consonant harmony.
That'd only be consonant harmony if the only voiced/voiceless distinction were in consonants. In many of my conlangs, however, the vowels are also distinguished by voicing.

User avatar
con quesa
Lebom
Lebom
Posts: 159
Joined: Sun Apr 27, 2003 1:34 pm
Location: Fnuhpolis- The City of Fnuh

Post by con quesa »

I have some vowel harmony questions.

In my conlang, <i> <?> <?> are front vowels, <?> <?> <o> <u> are back, and <?> <a> <e> are neutral. There is front/back harmony, although a neutral vowel can be in either group. Normally, the vowels in a root are in the same group as the final vowel in the root: k?nasu is OK, but not *kinasu. I want suffixes' vowels to also harmonize with the last root vowel, even though if there is a suffix, the last root vowel isn't the last vowel in the word. Can there be such a distinction between a root and a word? Or is this impossible, in the same way as making different sound changes for nouns and verbs?

Also, if the last root vowel is a neutral vowel, is it possible for vowel harmony to simply not happen, allowing words like ticuz'a ? Or would the vowels harmonize in a different way, creating tucuz'a or ticiz'a ?
con quesa- firm believer in the right of Spanish cheese to be female if she so chooses

"There's nothing inherently different between knowing who Venusaur is and knowing who Lady Macbeth is" -Xephyr

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Post by Soap »

I vote for letting the last vowel in the root decide it all. That's close to what you find in Finnish and other Uralic languages that have vowel harmony. A real Uralic speaker would know more than me.
Sunàqʷa the Sea Lamprey says:
Image

gsandi
Lebom
Lebom
Posts: 110
Joined: Fri Apr 25, 2003 10:13 am
Location: Vancouver, BC, Canada
Contact:

Post by gsandi »

con quesa wrote:I have some vowel harmony questions.

In my conlang, < i > <?> <?> are front vowels, <?> <?> <o> < u > are back, and <?> <a> <e> are neutral. There is front/back harmony, although a neutral vowel can be in either group. Normally, the vowels in a root are in the same group as the final vowel in the root: k?nasu is OK, but not *kinasu.

Also, if the last root vowel is a neutral vowel, is it possible for vowel harmony to simply not happen, allowing words like ticuz'a ? Or would the vowels harmonize in a different way, creating tucuz'a or ticiz'a ?
It's your language, you can make your vowel-harmony rules work any way you like. :)

If you want to make these rules lifelike, you may wish to check out how vowel harmony works in real languages - the most accessible European languages with vowel harmony being Turkish, Finnish and Hungarian. I am no expert on the first two, but can provide a reasonable description of Hungarian.

So, on your questions:
I want suffixes' vowels to also harmonize with the last root vowel, even though if there is a suffix, the last root vowel isn't the last vowel in the word. Can there be such a distinction between a root and a word? Or is this impossible, in the same way as making different sound changes for nouns and verbs?
By suffix, do you mean a derivational suffix (like English -ness, -dom, -ize) or an inflexional suffix (like plural, genitive and 3rd person -(e)s in English)?

In Hungarian, both types may have vowel harmony, and adjust their internal vowel according to the last vowel of the root. It may be a 2-way distinction:

-ban, -ben (in): gy?r-ban (in the factory) / kert-ben (in the garden). [NB: there is no hyphen in standard orthography, I insert it for clarity's sake]

It can also be 3-way:

-en, -on, -?n (on): kert-en (on the garden), h?z-on (on the house), gy?m?lcs-?n (on the fruit)

The principle is the same (although may be different in detail) for verbal endings and derivational suffixes:

-ek, -ok, -?k (first person singular, present indicative, indefinite): besz?l-ek (I speak), ugat-ok (I bark [why not, can't dogs talk? :evil: ]), ?r?l-?k (I am happy).

-s?g, -s?g (-ness, -ing): k?v?n-s?g (wish [noun], from kiv?n-ni [to wish]), v?n-s?g (extreme old age, from v?n [very old]).
Also, if the last root vowel is a neutral vowel, is it possible for vowel harmony to simply not happen, allowing words like ticuz'a ?
Normally, when vowel harmony operates, there is no such thing as it not happening. Every possible vowel in the root has its well-defined "forced" correspondent in the endings. However, exceptions may occur, for any number of reasons:

1. There may be a vowel that is historically the coalescence of two (or more) older vowels belonging to more than one vowel class. In Hungarian /i/ is such a vowel - it may be derived from the former front high unrounded vowel /i/ or from the former back high unrounded vowel /?/. The endings may reflect the distinction to this day:

h?d-ban (in the bridge), but sz?v-ben (in the heart).

2. Loanwords and compounds may have mixed vowel types, in which case the last vowel is the decisive factor:

beton (cement) - beton-ban (in the cement)
borj?-b?r (calf-skin) - borj?-b?r-ben (in the calf-skin)

3. Sometimes root vowels sporadically (i.e. not as part of a regular sound change) change into a vowel in a different class, while the suffix may keep the vowel corresponding to the former class of the root vowel. E.g. in Hungarian:

Root -i- may change to -?- in a labial environment, as f?l (ear), from earlier *fil. The first person possessive for words with root vowel /?/ is normally /-?m/, as in k?rt-?m (my horn [musical instrument]), yet for f?l it is f?l-em (my ear), where the ending reflects the original root vowel.

---------------------------------------

I hope this answers your question.

Sometimes I've been tempted to create a Uralic conlang, on analogy with my creation of a Romance conlang (Tundrian). It would be a fun way to help me learn Finnish!

Good conlanging, and Happy New Year.

Glenn
Avisaru
Avisaru
Posts: 316
Joined: Mon Sep 16, 2002 3:43 am
Location: Virginia, USA/Tiolu, Kiarlon

Post by Glenn »

bicoherent wrote:Cor: That'll be consonant harmony, not vowel harmony. Reportedly, very few languages have consonant harmony.
Well...I haven't necessary encountered consonant harmony throughout a word, but a kind of front/back "consonant harmony", with certain consonants pronounced further forward before front vowels and back before back vowels, is quite common--a good example is a distinction between palatal/velar or velar/uvular /k/ and /g/. In English, this form of harmony exists, but is not phonemic; in Kazakh, on the other hand, it is considered phonemic, and different letters are used to represent the front and back sounds in Kazakh orthography. In other languages, of course, these might be considered different phonemes altogether, and may be used regardless of the following vowel.

I see this as a consonant/vowel harmony of sorts, although it may be closer to, for example, the kinds of consonant assimilation seen in English and many other languages (for an example of progressive voicing assimilation in English, check out the voiced <s> ([z]) in begs, or the unvoiced <d> ([t]) in danced.)

p@,
Glenn

User avatar
abeygail
Sanci
Sanci
Posts: 17
Joined: Sat Jul 25, 2009 8:21 pm
Location: Sacramento, California

Post by abeygail »

Alright, I understand vowel harmony... how does tone harmony operate or develope?
http://sites.google.com/site/sahuho/home

Post Reply