Why IPA simplification is bad, regarding aspirant–plosive de-aspiration and CUBE is actually wrong
About CUBE
CUBE sought to rectify a bunch of mistakes in the modern rendition of the IPA for English.
It did a reasonable job, but still went to reduce the complexity and thus introduced another set of errors. They did this because their specific education modality is limited to teaching Indo-European language speakers. They are not familiar with the trials and tribulations that members of other language families go through, therefore, they did not recognise the new errors that they introduced.
standard IPA [A]
real pronunciation [B]
CUBE’s idea [C]
what we should write
pit
pɪt
pʰɪt
pɪt
pʰɪt [B]
bit
bɪt
pɪt
bɪt
bɪt [A]
spit
spɪt
spɪt
sbɪt
spɪt [A]
Chain of reasoning:
Bit makes people think that b, /p/ is /b/
Pit makes people think that p, /pʰ/ is /p/
Spit makes people think that sp, /sp/ is /spʰ/
Spit as /spɪt/ is already accurate
Pit as /pit/ is wrong – it should be /pʰɪt/
Bit as /bɪt/ is correct between voiced phonemes, but wrong in isolation. (This is what leads some linguists to believe that English’s voice stop/plosive phonemes are essentially unvoiced. Actually they are voiced and that is why we cannot alter the IPA to show them as their unvoiced counterparts.)
Spit as /spɪt/ could be adjusted to /sbɪt/, which is correct within the erroneous system, but it would cause people to over-voice the p as /b/. b, p and sp are three different sounds. Actually,
P as /p/ is a shortcut for /pʰ/
B as /b/ is a shunted replacement for /p/
SP as /sp/ results in learners pronouncing incorrectly as /spʰ/
SP as /sb/ results in learners over-voicing the p as real /b/
SP as /sp/ is already accurate correct.
The only real correction that can be made is adjusting all plosive consonants in IPA, as such:
letters
was
actually
example
solution #1 (bad)
(CUBE, bad)
(combo)
p
/p/
/pʰ/
pit
/pʰɪt/
/pɪt/
/pʰɪt/
b
/b/
/p/
bit
/pɪt/
/bɪt/
/bɪt/
sp
/sp/
/sp/
spit
/spɪt/
/sbɪt/
/spɪt/
t
/t/
/tʰ/
to
/tʰuw/
/tuw/
/tʰuw/
d
/d/
/t/
do
/tuw/
/duw/
/duw/
ft
/ft/
/ft/
after
/ˈɑːf.tə/
/ˈɑːf.də/
/ˈɑːf.tə/
thèd
/θt/
/θt/
mathed
/maθt/
/maθd/
/maθt/
st
/st/
/st/
stop
/stɔp/
/sdɔp/
/stɔp/
shèd
/ʃt/
/ʃt/
washed
/wɔʃt/
/wɔʃd/
/wɔʃt/
ch
/ʧ/
/ʧʰ/
chin
/ʧʰɪn/
/ʧɪn/
/ʧʰɪn/
j, ge, dge
/ʤ/
/ʧ/
gin
/ʧɪn/
/ʤɪn/
/ʤɪn/
stio
/sʧ/
/sʧ/
question
/ˈkwɛs.ʧən/
/ˈkwɛs.ʤən/
/ˈkwɛs.ʧən/
str
/stɹ/
/sʧɹ/
string
/sʧɹɪŋ/
/sʤɹɪŋ/
/sʧɹɪŋ/
k, c, ch
/k/
/kʰ/
cap
/kʰap/
/kap/
/kʰap/
g
/ɡ/
/k/
gap
/kap/
/ɡap/
/ɡap/
sk, sch
/sk/
/sk/
skip
/skɪp/
/sɡɪp/
/skɪp/
scr
/skɹ/
/skɹ/
scrap
/skɹap/
/sɡɹap/
/skɹap/
Solution 1 Issues:
· Appears more complex.
· ‘Voiced’ coda is not yet satisfactorily resolved.
· Causes erroneous devoicing in onsets and codas during connected speech.
Solution 2 Issues:
· /sd/ results in incorrect over-voicing.
· Also negatively affects word boundaries in connected speech.
Solution 3:
This is the only reasonable solution, because voiced plosive consonants are fully voiced whenever they occur between other voiced phonemes. Additionally, marking de-aspirated consonants as if they were there voiced counterparts is also erroneous because many languages do fully voice their voiced consonants, so teaching students this would result in over-voicing of such phonemes. They would end up pronouncing words like stop as “zdop”, spot as “zbot”, after as “avder”, etc. You see the problem.
However, Solution 3 will not be followed, because all unvoiced plosives are always aspirated except for this environment, therefore, nobody will follow the aspiration marking with ʰ for long – they may start out with best intentions, but after a while, people will default back to the shorthand, “oh there’s no need to mark the aspiration, it’s always like that”.
However… it would be really beneficial if language teachers were taught about this, because that would make them actually aware that English has aspirated consonants while many languages have unaspirated consonants.
The solution seems to be simply that language teachers need to be better at their job. They should learn about the intricacies of pronunciationof the language that they’re teaching, and they should learn how best to teach these concepts and train their students into the new pronunciation model of the target language. These things aren’t excessively complicated, these things are just things that every student can and must learn in order to pronounce the new language correctly. Anything after this is what we can consider as an “accent”; any deviations prior to developing their skills aligned with model, we can say are “mistakes”. Mistakes inasmuch as the deviations cause reasonable misunderstandings, through increased ambiguity, causing issues with tenses, word boundaries, and other similar things.
more details:
Solution 1 is unsuitable, because, in connected speech, between other voiced sounds, these unaspirated plosives are voiced.
So, do we teach that, too?
“Between voiced sounds, the ‘voiced’ letters B D J G are fully voiced; at the start and end of strings of speed (first word, last word, before a comma), they are unvoiced, but the vowel extends right up to their edges. Create these sounds correctly by activating your voice up to their edges. Ensure that you release your mouth position after the consonant.”
This may seem complicated but actually this occurs in many languages, and is taught in those languages; e.g. Korean features this effect but even more strongly than in English, and students studying the Korean language learn about this. To not learn it results in them severely mispronouncing every single word.
Languages are complex and complicated.
That’s ok.
English is a language. It’s not a problem for some things to be complicated about it. Just teach those things carefully, systematically, and rigorously. Train the students in those things from day 1, get the concepts & systems into their brains, and help them understand and stabilise their skill.
Don’t create a reductionist pattern, a false system of over-simplification. You are not doing anyone a favour: not the students, and not all the people who they will speak to in their future lives. We owe it to the students to teach them English properly so that a) they can actually communicate, b) they feel confident knowing that their skills are based on reality and not simplified fantasy, and c) they can talk to non-Vietnamese people and know that they are speaking well enough for anyone to understand them.
As it is now, every language centre and school is using the same, frankly terrible, method of teaching pronunciation. Nobody understands the system of sounds in the language, nobody is learning how to control their mouths carefully, and nobody is learning how to actually create the ending sounds. You create simplified methods of teaching them, but all of your methods are just re-inforcing the Vietnamese way of pronouncing sounds.
There is a difference between a mistake and an accent. And, unfortunately, most vietnamese people are not speaking with simply a “vietnamese accent”. They are speaking with a whole array of mistakes that honestly make it nearly impossible for foreigners to understand anything more than trying to order food and drinks. Moreover, with the combination of poor pronunciation and lack of understanding of grammar, our students are unequipped to understand foreigners clearly when they speak English. This is a shame, because every single usage of English around them is an opportunity to learn and import the sound & grammar. But if their pronunciation skills are weak, then their listening skills are inherently weak too; and therefore so is their ability to hear the words and thus the grammar of the sentences, so they cannot be learning from their exposure in the real world. This is such a shame.
Solution #2 – replacing all the de-aspirated consonants with their voiced IPA counterparts – introduces another problem: connected speech. When linked to the next word, starting with a vowel, final consonants that undergo de-aspiration will render completely incorrect connected speech if the speaker uses the fully-voiced consonant. A de-aspirated T, /tʰ/ → /t/, is not the same as a voiced consonant between voiced phones.
E.g. “washed up”:
/wɔʃt ʌp/
/wɔʃd ʌp/
Obviously, a /t/ in coda position is not actively aspirated as /tʰ/, but a de-aspirated coda-/t/ is notably different from a regular coda-/t/ – compare “at a place” and “washed a plate”.
Now try “bored a hole”. The coda /d/ is NOT like the coda /t/ of “washed”.
These are THREE DIFFERENT SOUNDS. And, as most students unsurprisingly desire, to speak clearly and correctly, these three different sounds should be pronounced differently.
Therefore, the solution is not to find which is the best way to simplify it, but rather, to create a sensible way of teaching it.
Last updated