Vowel Model v2
Last updated
Last updated
Vowel Model v1 is the model that I am currently using to teach English pronunciation.
This is my second Vowel Model hypothesis, which has problems.
It is a nice theory, it works pretty well, and I will keep developing it until it breaks or is resolved.
View the others via the buttons above.
Vowel Model v3 is currently the most stable and symmetrical.
ɑj
ɑɹ
ʌ
ɛj
ɛɹ
ɛ
ɪj
ɪɹ
ɪ
oj
oɹ
ɔ
ɵʉ
ɵɹ
ɵ
aʉ
ɜɹ
a
The rhotic pattern is used primarily. This can be parsed into your own accent easily:
/ɹ/ mutates to /ɚ/, /ə/ or an elongated base vowel, e.g. /ɛɹ/ → /ɛː/.
Reanalysis: ɑj – ɑɹ – ʌ
ɑj – tense, long, glide-bound
ɑɹ – rhotic, long, R-bound
ʌ – relaxed, short, checked
Rhotic acç
ɑj
ɑɹ
ʌ
Non-rho acç
ɑj
ɑː
ʌ
Standard IPA
aɪ
ɑː
æ
kite /kɑjt/
cart – /kɑːt/
cut /kʌt/
Standard
aɪ
aː, aɹ
ʌ
CUBE
ɑj
ɑː, aɹ
ʌ
Rhotic
ɑj
ɑɹ
ʌ
difference
glide
tense, longer
relaxed, shorter
example
kite
/kʰɑjt/
cart
/kʰɑːt/
cut
/kʰʌt/
Reanalysis: ɛj – ɛɹ – ɛ
ɛj – tense, long, glide-bound
ɛɹ – rhotic, long, R-bound
ɛ – relaxed, short, checked
Rhotic acç
ɛj
ɛɹ, ɛɚ
ɛ
Non-rho acç
ɛj
ɛə, ɛː
ɔ
Standard IPA
eɪ
eɚ
ɒ
bait /bɛjt/
bear /bɛə/
bed /bɛd/
Note that there is a characteristic difference here between the British Englishes and the American Englishes: the diphthong /ɛə/ is a closing diphthong in British English, while it’s an opening diphthong /eə/ in American English.
Many accents simplify the /ɛə/ diphthong into a simple long vowel, /ɛː/ or /eː/. In a Rhotic accent, this may manifest as /ɛ˞ ː/ or /e˞ ː/. This is the same effect as with /ɪə/ becoming long, /ɪː/ (see Diphthongs).
bait /bɛjt/ – bear /bɛː/ – bed /bɛd/
standard
ɛɪ
ɛə/eə, ɛɚ/eɚ
ɛ
modern
stressed
ɛj
ɛə, ɛɚ
ɛ
strong
ɛj
ɛː, ɛɚ
ɛ
weak
ej
eː, eɚ
e
difference
tense, longer, glide
tense, longer, rhotic
relaxed, shorter, checked
example
they
/ðɛj/
there
/ðɛː/
then
/ðɛn/
The "strong" form is used in important words and stressed words, such as when the point of the word needs to be emphasized in a word that is usually unstressed.
The "weak" form is what emerges in unimportant, common and casual words.
Reanalysis: ɪj – ɪɹ – ɪ
ɪj – tense, long, glide-bound
ɪɹ – rhotic, long, R-bound
ɪ – relaxed, short, checked
Rhotic acç
ɪj
ɪɹ, ɪɚ
ɪ
Non-rho acç
ɪj
ɪə, ɪː
ɪ
Standard IPA
iː
ɪɚ
ɪ
seat /sɪjt/
seer /sɪə/
sit /sɪt/
standard
iː
ɪə, ɪɚ
ɪ
modern
stressed
ɪj
ɪjə, ɪjɚ
ɪ
strong
ɪj
ɪə, ɪɚ
ɪ
weak
ɪj
ɪː, ɪ˞ː
ɪ
difference
tense, longer, glide
tense, longer, rhotic
relaxed, shorter, checked
example
heat
/hɪjt/
here
/hɪə/
him
/hɪm/
Reanalysis: oj – oɹ – ɔ
oj – tense, long, glide-bound
oɹ – rhotic, long, R-bound
ɔ – relaxed, short, checked
Rhotic acç
oj
oɹ
ɔ
Non-rho acç
oj
oː
ɔ
Standard IPA
ɔɪ
ɔː
ɒ
boy /boj/
bought /boːt/
bottle /ˈbɔt.ɫ̣/
toy /toj/
taught /toːt/
top /tɔp/
standard
ɔɪ
ɔː
ɒ
modern
oj
oː
ɔ
difference
tense, longer, glide
tense, longer, rhotic
relaxed, shorter, checked
example
boy
bought
bottle
Reanalysis: ɵʉ – ɵɹ – ɵ
ɵʉ – tense, long, glide-bound
ɵɹ – rhotic, long, R-bound
ɵ – relaxed, short, checked
Rhotic acç
ɵʉ
ɵɹ, ɵɚ
ɵ
Non-rho acç
ɵʉ
ɵə, ɵː
ɵ
Standard IPA
uː
ʊɹ, ʊɚ
ɒ
food /fɵʉd/
cure /kjɵː/
good /gɵd/
-ure (etc.) produces /ɵɹ/ in Rhotic accents, and long/tense /ɵː/ in non-Rhotic accents:
pure, cure, injure, sure, adventure, posture, manicure
jury, fury, furious, curious
standard
uː
ʊɹ
ʊ
modern
ɵʉ
ɵɹ, ɵɚ
ɵə, ɵː
ɵ
difference
tense, longer, glide
tense, longer, rhotic
relaxed, shorter, checked
example
food
cure
good
Reanalysis: aw – a – ɜː
Here I’ve merged the orphaned sets of a–aw and ɜː–ə.
I’m not sure if it’s valid but it does work spatially on the graph. They could be related also in that stressed ER is /ɜː/, and unstressed ER and unstressed A are both /ə/.
ɜː is a tense, elongated schwa.
a is a tense but short sound, the default value of ⟨a⟩. ⟨a⟩ is the first to become the schwa /ə/ sound.
In Rhotic accents, stressed ER is /ɜ˞ ː/ or /ɚː/, unstressed ER is /ɚ/ and unstressed A is /ə/.
There is some asymmetry compared to the other vowel sets, in that the vowel at the mid-point along the path is long instead of the vowel at the end point… But this may be to distinguish it from the schwa position, as it’s really similar, and is even identical in many accents.
In summary these vowels may not really bear a true relationship. However, all the other vowels are connected via this relationship, and the two sounds that I've put into this set are both orphaned. So it works even just as a way to store them apart from the other vowels.
aw – tense, long, glide-bound
a – relaxed, short, checked
ɜɹ – rhotic, long, R-bound
Rhotic acç
aʉ
a
ɜɹ, ɜ˞
ə
ɚ
Non-rho acç
aʉ
a
ɜː
ə
ə
Standard IPA
aʊ
æ
ɜː, əː
ə
ɚ, ə
house /haʉs/
sat /sat/
certain /ˈsɜː.tʰən/
a: about /əˈbaʉt/
er: hunger /ˈhʌŋ.ɡə/
standard
aʊ
æ
ɜː
ə
modern
aw
a
ɜː
ə
difference
tense, longer, glide
relaxed, shorter, checked
tense, longer, rhotic
it's a schwa, relax
example
house
/hsaʉs/
sat
/sat/
certain
/ˈsɜː.tʰən/
a: about /əˈbaʉt/
er: hunger /ˈhʌŋ.ɡə/
if /ɛə/ is a diphthong, then it and its simplification /ɛː/ DON’T belong in the E series.
That leaves the E series with only 1 member!
Ok, visit another page to read my third model hypothesis: Vowel Model v3
The above vowel series theory emerged accidentally during an attempt to develop a model of vowel declension, intended to help English learners understand & predict how vowel sounds change in words depending on word stress in a sentence, syllable stress in a word, syllable position in a word, and base pronunciation value of a word part.
This declension series isn't part of the my Tension Path theories, but it is the origin of them, so I'll link to it here when I've actually come up with the model. As of now (May 2025), there is no model, because these vowel series theories have taken more time & focus than I expected!
I expect these vowel set series to provide the foundation for mapping out the declension series.
Vowel Model v1
Vowel Model v2
Vowel Model v3 (latest)