A Transcription of English

[R.Rusanov]

This morning I had some free time in between classes so I came up with a transcription of English that is, I guess, semi-phonemic (rather than phonetic or etymological).

It reflects my English most of all, so it doesn't distinguish between Bach's, box, and balks though most AME speakers do. Nor do I r-drop, but I do l-drop (or l-vocalize).

Here's a sample:
Ar fadhyr, Hu art ın hevyn,
Hæloud bi Dhai Neim.
Dhai cıngdym cym,
Dhai wıl bi dyn,
an yrth æz ıt ız ın hevyn.
Gıv ys dhıs dei ar deili bred,
ænd fyrgıv ys ar trespæses,
æz wi fyrgıv dhouz hu trespæs ygenst ys;
ænd lid ys nat ıntu temteissyn,
byt delıvyr ys frym ivıl.
Eimen.

Here are the sound correspondences, if you care about them:

i “lean”
[ɪ] ı “pin”
[eɪ/e] ei “bait”
[ɛ] e “bed”
[æ] æ “bat”
[a/ɑ/ɒ/ɑə] a “cot”
[ɒ/ɔ/oə] a “caught”
[oʊ/o] ou “load”
[ʌ] y “cut”
[ʊ] y “foot”
u “food”
[ju] iu “cute”
[aɪ/ɑɪ] ai “bite”
[æʊ/aʊ/ɑʊ] au “bounce”
[ɔɪ] oi “noise”
[ə] y “about”
[ɪɹ] ir “fear”
[eɹ] eir “fair”
[aɹ/ɑɹ/ɒɹ] ar “far”
[ɔɹ] or “for”
[oɹ] or “four”
[ɝ] yr “fur”
[ʊɹ] ur “pour”
[ɚ] yr “butter”
[ð] dh “that”
[ŋ] n (prevelar) “singing”
[ŋ] ng (alone) “sing”
[ɡ] g “goose”
[ɹ] r “ring”
[ʃ] ss “shoe”
[ʒ] zz “leisure”
[θ] th “thin”
[ʍ] w “when”
b “bag”
[dʒ] gg “juice”
[d] d “dare”
[f] f “fear”
[h] h “ham”
[j] j “young”
[k] c “castle”
[l] l “look”
[m] m “mass”
[n] n “new”
[p] p “part”
[s] s “static”
[tʃ] cc “church”
[t] t “tan”
[v] v “van”
[w] w “wart”
[z] z “zoo"

What do y'all think?

[Travis B.]

Well, aside from concerns about the realism of proposing any new orthography for English...

You should describe the distinctions you are making in terms of phonemes, not in terms of phones.

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American. (Bonus points for also containing all the distinctions made by Standard Scottish English.) Particularly, being able to distinguish PALM, LOT, and THOUGHT, and STRUT and FOOT is a must. Likewise, NEAR should be distinguished from KIT + /r/, SQUARE should be distinguished from DRESS + /r/ and from TRAP + /r/, and NURSE should be distinguished from STRUT + /r/. (How CLOTH is to be treated depends upon whether one is favoring conservatism versus favoring progressive forms found in some given variety. How NORTH and FORCE are to be treated should depend on the level of conservatism aimed at, and how much one wants to accommodate dialects outside any standard variety.)

You also represent reduced vowels in a number of places where many speakers may have unreduced vowels, especially in more careful speech. You might be better off representing vowels as unreduced whenever possible and leaving vowel reduction up to the individual speaker, especially as in many placed reduced and unreduced vowels are actually in free variation.

Aesthetically I have a number of concerns:

Using <ı> anywhere is a very bad idea. Even introducing this glyph was a major mistake made in the design of Turkish orthography.

Marking postalveolars with doubling glyphs, as in the <ss>, <zz>, <cc>, and <gg> seen here is basically without precedent and goes against just about everything doubled consonants have been used anywhere for.

Using <c> for /k/ in general is a bad idea, due to historically in English being /s/ before historical front vowels (and in Old English /tʃ/ when palatalized, typically in association with front vowels). One is better off just using <k> for /k/ and <s> for /s/. English is not a Celtic language.

Using <y> for /ə/ is basically without precedent, not matching the varied uses of it in other languages, present and historical. It just feels like a way to shoehorn in a particular glyph for a purpose for lack of any use elsewhere, and in that actually goes against the tendencies towards merging phonemes glyph-wise. (I would personally just use <e> here, following the precedent set by both historical English and other Germanic languages, especially because /ɛ/ normally does not show up in unstressed syllables in English anyways.)

[R.Rusanov]

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American.

That's pretty impossible. People learning the new orthography (presumably on the basis that it's phonetic) would resent having to learn extra distinctions that they themselves don't make, as that would make the new script not much superior to the eld.

An English orthography that would work worldwide couldn't be entirely phonetic for every speaker, because it would have tons of consonants and vowels that people could hardly keep straight. Our current orthography, while burdensome, at least works for every speaker with a bit of rote memorization.

Anyway I think it's very Eurocentric and borderline-racist of you to not distinguish some of the most spoken dialects of English, such as Indian English, Chinese English, Nigerian English etc. Standard American English has maybe a few million native speakers round Ohio and Southern Florida and RP, well, even less (I guess BBC presenters are the only ones keeping that dialect alive). And even on BBC nowadays you can hear urban British English and guttural t's and dropped or mistakenly replaced r's. Younger speakers, who represent the future of the English languages, definitely stray farther from old-fashioned prestige dialects than the old and I think it's quite ageist to discriminate against them thus.

English dialects are very different one from the other and are still diverging, so trying to gather them back into the fold via a single alphabet is silly & unnecessary. Better for each speaker or dialect group to have an orthography to represent their speech.

Particularly, being able to distinguish PALM, LOT, and THOUGHT ... is a must

This isn't intended to be used for all speakers. If you have extra vowels you can add them in.

STRUT and FOOT

"sccryt" vs "fut", there is a distinction.

NEAR should be distinguished from KIT + /r/

Dunno why you think that, unless you're a Brit.

SQUARE should be distinguished from DRESS + /r/ and from TRAP + /r/

"skwer" - square - vs hypothetical "ccrær" - *trar - is distinguished, though.

and NURSE should be distinguished from STRUT + /r/. (How CLOTH is to be treated depends upon whether one is favoring conservatism versus favoring progressive forms found in some given variety. How NORTH and FORCE are to be treated should depend on the level of conservatism aimed at, and how much one wants to accommodate dialects outside any standard variety.)

Conservatism in language is quite discriminatory against younger and more socially mobile speakers. Whatever happened to the principle of "Write as you speak and read as it is written"?

Using <ı> anywhere is a very bad idea. Even introducing this glyph was a major mistake made in the design of Turkish orthography.

Any reason you think so? It may look striking in computer transcription but it's not hard to do on paper, you just drop the dot. Anyway a lot of dialects merge [ɪ] with [ʌ] or with [ɛ], so they can write their speech without the <ı> and just use the value that seems closest to them.

Marking postalveolars with doubling glyphs, as in the <ss>, <zz>, <cc>, and <gg> seen here is basically without precedent and goes against just about everything doubled consonants have been used anywhere for.

Old English had <cc> = [tʃ] and <gg> likewise for, uh, [dʒ]. Modern English (at least my dialect) doesn't have gemination so the "ss" and "zz" spaces are more or less available; why not double the consonants? It saves the need for special diacritics or, say, "sh" which does conflict with actual English clusters in words like "hogshead".

Using <c> for /k/ in general is a bad idea, due to historically in English being /s/ before historical front vowels (and in Old English /tʃ/ when palatalized, typically in association with front vowels). One is better off just using <k> for /k/ and <s> for /s/. English is not a Celtic language.

That's just your European privilige showing. Many languages have <c> for [k], and anyway [k] + front vowels is very common in English. For example, "key", "keen", etc.

Using <y> for /ə/ is basically without precedent

Welsh does it :/

Anyway I appreciate your comments and am thinking about revising items like doubled consonants and the symbol for [ɪ], as I would like this orthography to have as few "special" characters as possible.

[Travis B.]

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American.

That's pretty impossible. People learning the new orthography (presumably on the basis that it's phonetic) would resent having to learn extra distinctions that they themselves don't make, as that would make the new script not much superior to the eld.

It is possible, with the qualification that in practice most representations would make BATH and CLOTH favor one or the other, rather than representing them separately unto themselves. (I personally would favor aligning the representation of BATH with GA and of CLOTH with RP, as that is more conservative*, and I favor more conservative representations when in doubt.)

* properly speaking, more conservative RP actually has CLOTH like GA, but diachronically that derives from a conditional vowel lengthening that applied inconsistently, so the historical pronunciation is closer to that of more modern RP.

An English orthography that would work worldwide couldn't be entirely phonetic for every speaker, because it would have tons of consonants and vowels that people could hardly keep straight. Our current orthography, while burdensome, at least works for every speaker with a bit of rote memorization.

Anyway I think it's very Eurocentric and borderline-racist of you to not distinguish some of the most spoken dialects of English, such as Indian English, Chinese English, Nigerian English etc. Standard American English has maybe a few million native speakers round Ohio and Southern Florida and RP, well, even less (I guess BBC presenters are the only ones keeping that dialect alive). And even on BBC nowadays you can hear urban British English and guttural t's and dropped or mistakenly replaced r's. Younger speakers, who represent the future of the English languages, definitely stray farther from old-fashioned prestige dialects than the old and I think it's quite ageist to discriminate against them thus.

English dialects are very different one from the other and are still diverging, so trying to gather them back into the fold via a single alphabet is silly & unnecessary. Better for each speaker or dialect group to have an orthography to represent their speech.

Since when is being "phonetic" a plus? In practice we need an orthography where everyone spells words either the same way or very similarly (considering that at the present there are more than one standard spelling for certain words). Without that, reading would be much more difficult, as one would have to think about the pronunciation of each word one read and mentally translate it into one's own dialect, and simple things like searching online for text simply would not work.

And if we are going to have an orthography that is crossdialectically phonemic (read: not phonetic), it only makes most sense to base it on either the standard varieties or idealized historical varieties corresponding to the areas where English is most spoken, i.e. the US, the UK, Canada, and Australia. And here, most Canadian English outside the Maritime provinces is phonemically close to GA (with only a limited set of words that differ phonemically), and likewise General Australian phonemically differs little from southern English English (except that it splits /æ/ into /æ/ and /æː/, unlike RP). This is not racist, this is simply taking into account the reality of just who actually speaks English.

And of course we must base an orthography on either standard varieties or idealized historical varieties because we cannot simply digest a wide smattering of different dialects with their own individual variations into a single orthography; even if we are to avoid using standard varieties we must still beforehand digest those dialects down into idealized historical varieties from which they would be treated as having been descended from or closely allied to.

Particularly, being able to distinguish PALM, LOT, and THOUGHT ... is a must

This isn't intended to be used for all speakers. If you have extra vowels you can add them in.

(This goes back to where I pointed out that we need to have everyone spelling the same way.)

STRUT and FOOT

"sccryt" vs "fut", there is a distinction.

Note the following:

[ʌ] y “cut”
[ʊ] y “foot”

You merged the two here.

NEAR should be distinguished from KIT + /r/

Dunno why you think that, unless you're a Brit.

Well, it is important that a new orthography represent the distinctions that most English (and for that matter, most Scottish) people have, since the UK is only the country with the second-largest natively English-speaking population.

and NURSE should be distinguished from STRUT + /r/. (How CLOTH is to be treated depends upon whether one is favoring conservatism versus favoring progressive forms found in some given variety. How NORTH and FORCE are to be treated should depend on the level of conservatism aimed at, and how much one wants to accommodate dialects outside any standard variety.)

Conservatism in language is quite discriminatory against younger and more socially mobile speakers. Whatever happened to the principle of "Write as you speak and read as it is written"?

As stated before, we need people to write the same way; things will not practically work otherwise. And the reason to favor conservatism is that in some places we just have to choose one form over another, and it is ultimately more fair to consistently choose conservative forms, which can be objectively chosen on the basis of consistent principles, than progressive forms, which have to be cherry-picked in a manner that ultimately favors one dialect over another without any objective basis.

Using <ı> anywhere is a very bad idea. Even introducing this glyph was a major mistake made in the design of Turkish orthography.

Any reason you think so? It may look striking in computer transcription but it's not hard to do on paper, you just drop the dot. Anyway a lot of dialects merge [ɪ] with [ʌ] or with [ɛ], so they can write their speech without the <ı> and just use the value that seems closest to them.

It does not work typographically, the reason being that when it is used, in Turkic latin scripts, it breaks the capitalization of <i> as well, by taking the capital <I> while forcing the capital of <i> to be dotted.

Marking postalveolars with doubling glyphs, as in the <ss>, <zz>, <cc>, and <gg> seen here is basically without precedent and goes against just about everything doubled consonants have been used anywhere for.

Old English had <cc> = [tʃ] and <gg> likewise for, uh, [dʒ]. Modern English (at least my dialect) doesn't have gemination so the "ss" and "zz" spaces are more or less available; why not double the consonants? It saves the need for special diacritics or, say, "sh" which does conflict with actual English clusters in words like "hogshead".

For starters, English can have geminates at morpheme boundaries, and this breaks the representation of those. And Old English for <cc> was both /kː/ and /ttʃ/ and for <gg> was both /ɡː/ and /jː/ (realized [ddʒ]) (but this was normally actually written <cg>), with neither of these pairs being originally distinguished in writing but being distinguished with dots on the latter two in some modern scholarly standardizations.

As for the problems of using things like <sh>, in the rare case where that is ambiguous, one could always just introduce using an apostrophe or a hyphen to separate the two. And these cases are probably rarer than the cases of morphological geminates in English.

Using <c> for /k/ in general is a bad idea, due to historically in English being /s/ before historical front vowels (and in Old English /tʃ/ when palatalized, typically in association with front vowels). One is better off just using <k> for /k/ and <s> for /s/. English is not a Celtic language.

That's just your European privilige showing. Many languages have <c> for [k], and anyway [k] + front vowels is very common in English. For example, "key", "keen", etc.

The main thing is that the main languages where <c> is used for /k/ everywhere are Celtic languages only; <c> before a front vowel is something else in Romance languages, and was too in other Germanic languages (except that these cases have been weeded out in more recent times, e.g. <c> + front vowel being changed to <z> in modern Standard German).

Using <y> for /ə/ is basically without precedent

Welsh does it :/

I forgot that Welsh had that as another value for <y>; I thought <y> only represented high vowels in it. Still, this use in a Germanic language would be idiosyncratic, as in Germanic languages <y> has almost always represented some high front vowel (whether rounded or not) or front diphthong.

[Imralu]

You pronounce "father" as /ˈfɑθər/?

[Rory]

Travis,

Rusanov has come up with a transcription of English. Not a spelling reform, not a proposed "fix" to how we all should write, nothing else. Just a transcription of how (s)he speaks. So as far as I can see, it doesn't matter that it doesn't apply pandialectically.

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American.

That's pretty impossible. People learning the new orthography (presumably on the basis that it's phonetic) would resent having to learn extra distinctions that they themselves don't make, as that would make the new script not much superior to the eld.

It is possible, with the qualification that in practice most representations would make BATH and CLOTH favor one or the other, rather than representing them separately unto themselves. (I personally would favor aligning the representation of BATH with GA and of CLOTH with RP, as that is more conservative*, and I favor more conservative representations when in doubt.)

* properly speaking, more conservative RP actually has CLOTH like GA, but diachronically that derives from a conditional vowel lengthening that applied inconsistently, so the historical pronunciation is closer to that of more modern RP.

Actually, I agree with Rusanov that it is pretty impossible to do well. The transcription simply can't be consistent for every speaker. In American Englishes, there are three possible phonemically distinct vowels productions attested for the simple word dog - so, do we transcribe it with the FATHER vowel, the LOT vowel, or the CAUGHT vowel? How about merry - is that TRAP, GET, or FACE? Same question for blanket and egg. Vowels aside, there are also issues of words that people just say differently - for instance, tourniquet, which most Americans pronounce with final /Et/ and most Brits with final /eI/. How do we transcribe that?

[Melteor]

None of these is too bad IMO. I personally like Rusanov's because there are fewer letters, more diacritics, but both diacritics and longer words is undesirable. I favor more graphs as opposed to diacritics as well. Digraphs are okay only occasionally. I think these sort of topics, proposals belong in the orthography thread tho.

[R.Rusanov]

Note the following:

[ʌ] y “cut”
[ʊ] y “foot”

You merged the two here.

You're absolutely right. The "oo" of "foot" threw me off.

You pronounce "father" as /ˈfɑθər/?

The old orthography, uh, cross-dialectically contaminated my transcription, lol. Thanks for pointing it out.

Here is a longer sample:

Goldilacs ænd dhy Thri Berz

Wyns upan y taim, dher wyz y lıtyl gyrl neimd Goldilacs. Ssi went for y wac ın dhy forest. ıt wyz y lyvli dei. Prıti sun, ssi ceim ypon y haus. Ssi nact, ænd wen nou wyn ænsyrd, ssi wact raıt ın.
Æt dhy teibyl ın dhy cıccen dher wyr thri bols yv porıgg. Goldilacs wyz hyngri. Ssi tyc y fiu baıts frym dhy fyrst bol. Nau dher wyr fiur baıts dhæn dher hæd bın!
“Dhıs porıgg ız tu hat!” ssi ecscleimd.
Sou ssi teisted dhy porıgg frym dhy secynd bol.
“Dhıs porıgg ız tu cold,” ssi sed.
Sou ssi teisted dhy læst bol yv porıgg.
“Aaaa, dhıs porıgg ız ggyst raıt,” ssi sed hæpıli, ænd ssi eıt ıt all up. Ssi thyryli enggoid ıt!
Æftyr dhıs ssi desaided ssi wyz filıng y lıtyl tard. Sou ssi wact ıntu dhy lıvıngrum wer ssi sa thri ccerz. Goldilacs sæt ın dhy fyrst ccer tu rest hyr fit.
“Dhıs chair ız too big!” ssi ecscleimd.
Sou ssi sæt ın dhy secynd ccer.
“Dhıs ccer ız tu bıg, tu!” ssi waind.
Sou ssi traid dhy læst ænd smalest ccer.
“Aaaa, dhıs ccer ız ggyst rait,” ssi said. Byt ggyst æs ssi setyld daun ıntu dhy ccer tu rest, ıt brouc ıntu pisez!
Goldilacs wyz veri tard bai dhıs taim, sou ssi went ypsterz tu dhy bedrum. Ssi cyd hardli cæri hyrzelf yp dhy sterz! “Ai thinc ai’l ggyst lai daun,” ssi sed. (Dher wyz y mıryr ın dhy rym, byt ssi dıdynt noutıs ıt.) Ssi lei daun ın dhy fyrst bed, byt ıt wyz tu hard. Dhen ssi lei ın dhy secynd bed, byt ıt wyz tu saft. Dhen ssi lei daun ın dhy thyrd bed, ænd ıt wyz ggyst rait. Goldilacs fel yslip.
Æs ssi wyz slipıng, dhy thri berz ceim houm.
“Symwyn’z bın itıng mai porıgg,” grald Papy ber.
“Symwyn’z bın itıng mai porıgg,” ed Mamy ber.
“Symwyn’z bın itıng mai porıgg, ænd dhei eit ıt al yp!” craid Beibi ber.
“Symwyn’z bın sıtıng ın mai ccer,” grald Papy ber.
“Symwyn’z bın sıtıng ın mai cer,” sed Mamy ber.
“Symwyn’z bın sıtıng ın mai ccer, ænd dhei’v broucen ıt al tu pisez,” craid Beibi ber.
Dhıs wyz cwait y prablem! Dhei desaided tu lyc yraund sym mor, ænd wen dhei gat ypsterz to dhy bedrym, Papy ber grald, “Symwyn’z bın slipıng ın mai bed,”
“Symwyn’z bın slipıng ın mai bed, tu,” sed Mamy ber.
“Symwyn’z bın slipıng ın mai bed, ænd ssi’s stıl dher!” ecscleimd Beibi ber.
Ggyst dhen, Goldilacs wouc yp ænd sa dhy thri berz. Ssi ggympt yp ænd ræn aut yv dhy rym. Goldilacs ræn daun dhy sterz, oupend dhy dor, ænd ræn ywei ıntu dhy forest. Ænd æftyr dhıs ssi nevyr rityrnd tu dhy houm yv dhy thri berz.

[zompist]

Spelling reforms belong in the C&C forum.

[Travis B.]

Travis,

Rusanov has come up with a transcription of English. Not a spelling reform, not a proposed "fix" to how we all should write, nothing else. Just a transcription of how (s)he speaks. So as far as I can see, it doesn't matter that it doesn't apply pandialectically.

It should have been marked as a transcription of a particular dialect of English then.

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American.

That's pretty impossible. People learning the new orthography (presumably on the basis that it's phonetic) would resent having to learn extra distinctions that they themselves don't make, as that would make the new script not much superior to the eld.

It is possible, with the qualification that in practice most representations would make BATH and CLOTH favor one or the other, rather than representing them separately unto themselves. (I personally would favor aligning the representation of BATH with GA and of CLOTH with RP, as that is more conservative*, and I favor more conservative representations when in doubt.)

* properly speaking, more conservative RP actually has CLOTH like GA, but diachronically that derives from a conditional vowel lengthening that applied inconsistently, so the historical pronunciation is closer to that of more modern RP.

Actually, I agree with Rusanov that it is pretty impossible to do well. The transcription simply can't be consistent for every speaker. In American Englishes, there are three possible phonemically distinct vowels productions attested for the simple word dog - so, do we transcribe it with the FATHER vowel, the LOT vowel, or the CAUGHT vowel? How about merry - is that TRAP, GET, or FACE? Same question for blanket and egg.

In these kinds of cases I would always favor a diachronically conservative pronunciation from which most of the extant pronunciations can be derived. Hence for dog I would choose LOT and for merry I would choose DRESS.

Vowels aside, there are also issues of words that people just say differently - for instance, tourniquet, which most Americans pronounce with final /Et/ and most Brits with final /eI/. How do we transcribe that?

These kinds of cases which are simply irreconcilable even by resorting to favoring diachronically conservative pronunciations would probably be best handled by just having multiple spellings corresponding to the different pronunciations and considering them to all be acceptable, as long as they are consistent with the rest of one's writing.

[Salmoneus]

Words like 'tourniquet' are no problem - Americans can just pronounce the word their way, by the 'rules', and British people can just say it the way it looks.

[Imralu]

"She was feeling a little tard"??
She'll get locked up for that!

[maıráí]

It should have been marked as a transcription of a particular dialect of English then.

It reflects my English most of all

Not exactly what you're demanding, but still.

[R.Rusanov]

"She was feeling a little tard"??
She'll get locked up for that!

"r" has an offglide in stressed syllables after a front vowel, for example <bead> [bid] vs <beard> [biəɹd]. This makes <chair> [t͡ʃ ͡ɛjəɹ] and <Cher> [t͡ʃ ͡ɛəɹ] very near homophones. Whence [tajəɹd] ~ [taəɹd] ~ <tard>.

I've decided that "x" (heretofore unused) will substitute for "cs", and "q" for now-deprecated "cw".

For example:
✗cwin
✓ qin

✗ecscleim
✓ excleim

In ūsō:
"Dhy qin yv ınglınd excleimed dhæt hyr pawyrs deraivd frym dıvain rait, nat frym dhy veigyriz yv pyblıc opıniyn, cyrent paliticyl that natwıthstændıng."

A resolution has also been reached on the transcription of the <tr> and <str> (phonetically [tʃɹ] and [ʃtʃɹ]) clusters:
<tr> will be written "tr". Use of "ccr" is now frowned upon
<str> will be written "str". Use of "ssccr" is disparaged

[Rory]

Travis,

Rusanov has come up with a transcription of English. Not a spelling reform, not a proposed "fix" to how we all should write, nothing else. Just a transcription of how (s)he speaks. So as far as I can see, it doesn't matter that it doesn't apply pandialectically.

It should have been marked as a transcription of a particular dialect of English then.

I'm not sure how you interpreted A Transcription of English as An English spelling reform proposal for everyone, but to each their own, I suppose.

Underlyingly any new English orthography should at least be able to contain all the distinctions made by Received Pronunciation and General American.

That's pretty impossible. People learning the new orthography (presumably on the basis that it's phonetic) would resent having to learn extra distinctions that they themselves don't make, as that would make the new script not much superior to the eld.

It is possible, with the qualification that in practice most representations would make BATH and CLOTH favor one or the other, rather than representing them separately unto themselves. (I personally would favor aligning the representation of BATH with GA and of CLOTH with RP, as that is more conservative*, and I favor more conservative representations when in doubt.)

* properly speaking, more conservative RP actually has CLOTH like GA, but diachronically that derives from a conditional vowel lengthening that applied inconsistently, so the historical pronunciation is closer to that of more modern RP.

Actually, I agree with Rusanov that it is pretty impossible to do well. The transcription simply can't be consistent for every speaker. In American Englishes, there are three possible phonemically distinct vowels productions attested for the simple word dog - so, do we transcribe it with the FATHER vowel, the LOT vowel, or the CAUGHT vowel? How about merry - is that TRAP, GET, or FACE? Same question for blanket and egg.

In these kinds of cases I would always favor a diachronically conservative pronunciation from which most of the extant pronunciations can be derived. Hence for dog I would choose LOT and for merry I would choose DRESS.

I agree that's a sensible approach for a spelling reform proposal. Nevertheless, it doesn't "contain all the distinctions made by Received Pronunciation and General American", at least not in a consistent way. (I'm assuming that consistency is a desideratum of a spelling reform.) That's why I think that it's impossible.

(On the subject of choosing diachronically conservative yet extant pronunciations, would such a reform distinguish between meat, meet, and mate?)

[R.Rusanov]

Two have /i/ from /e:/ and /e:a/ < /au/ respectively, the other has /ɛi̯/ < /a:/. Presumably at least two of these would be distinguished.

meat : meat
meet : meet
mate : maat

Something like that perhaps.

Our fadhir dhaet art in heven,
Halowed bee dhii naam

But that's silly. It's hard to teach people an etymological pronunciation when they've got, well, any other in their heads. Either etymological or modernized.

[Hallow XIII]

That's just your European privilige showing. Many languages have <c> for [k], and anyway [k] + front vowels is very common in English. For example, "key", "keen", etc.

That is possibly one of the most mentally deficient statements I have ever seen. Also, you misspelt "privilege", you ridiculous twat.

And what exactly is the point of this? If it's not a suggested new spelling for Engloish, and just you transcribing your dialect... Why not just use phonemic IPA for English and be done with it? And then why engage in debates with Travis over

"Write as you speak and read as it is written"?

Oh, wait. You had to make a point about us all being racists and xenophobes.

it's very Eurocentric and borderline-racist

quite discriminatory against younger and more socially mobile speakers

[clawgrip]

<Cher> [t͡ʃ ͡ɛəɹ]

If you're talking about the singer/actress here, it's /ʃ/, not /tʃ/.

[Travis B.]

Two have /i/ from /e:/ and /e:a/ < /au/ respectively, the other has /ɛi̯/ < /a:/. Presumably at least two of these would be distinguished.

meat : meat
meet : meet
mate : maat

Something like that perhaps.

Our fadhir dhaet art in heven,
Halowed bee dhii naam

But that's silly. It's hard to teach people an etymological pronunciation when they've got, well, any other in their heads. Either etymological or modernized.

It is probably not necessary to distinguish Late Middle English /eː/ (as in meet) and /ɛː/ (as in meat) unless one's own dialect specifically distinguishes the two and one is aiming at specifically representing one's own dialect, as this is a distinction that has by this point been lost in most English varieties (even though apparently there are relicts in some rural parts of England that distinguish the two, even though I am not sure how well they have survived to date). The problem that representing this distinction is that then one will have to decide that each instance of standard /iː/ belongs to one or the other (or to LME /ɪ/ as in HAPPY), when there are almost certainly instances of such, e.g. in loans and neologisms, that cannot be traced back to Middle English, and will hence have to be simply assigned to one or the other arbitrarily.