Justin Rye, whose post on xenolinguistics I wrote about here, also has one on spelling reform:
For some years now I’ve been amusing myself by planning exactly what I would try in the way of “spelling reform” if I woke up one morning and found that the Revolutionary Stalinist-Linguist Party had mounted a coup and appointed me as World Dictator. Details of my proposal for a Revolting Orthography (modestly entitled Romanised English) are unlikely ever to become available; for now I want to get it clearly established exactly how mad this scheme is. The problems with our current system are sufficiently well-known that I feel no need to rehearse them all here; and people have been protesting about the situation for centuries. So just what is wrong with the idea of switching to something better? Anti-reformists come in thirteen basic flavours, with arguments summarisable as follows.
As I said in the MetaFilter post where I first saw it, “it’s the best thing I’ve ever seen on the subject, comprehensive, knowledgeable, and funny.” That was a couple of years ago; I’m not sure why I didn’t blog it then, but thanks to a comment by David Marjanović in this thread, I’m doing so now. Enjoy!
A very good piece. I really hope some kind of incremental reform project can get going. Our current spelling system is bonkers, but many of the reform proposals go way too fast.
I’ve also read the Xenolinguistics essay you linked and, as a conlanger who focuses on languages for extraterrestrials (and non-humanoids at that, not just the Klingons) I very much agree on the stupidity of a lot of stuff he mentions, including the universal translators.
The primary problem lies with the bureaucrats tasked to implement such reforms. I wouldn’t trust these people to install parking meters rightside up. I don’t want’em touching the language.
Axel Wijk’s Regularized Inglish is the One True Spelling Reform.
Beautiful rhetorical technique to simply take the overwhelmingly compelling nature of your own position for granted, but I’m unpersuaded that making the reform “incremental” (which anyway seems undercut by his premise of “revolution”) would overcome the massive expense and loss of productivity associated with (a) transliterating the entire corpus of English language writings, from historical texts to programming languages, into the modern spelling scheme, or more likely (b) forcing every child who hopes to advance in the world to master both modern and archaic spellings for every word, in school systems that are already struggling. So I trust this is mainly for fun. And I anyway have enough trouble getting people to remember the correct spelling and pronunciation of my 4-letter surname without endorsing a movement that further entrenches American prejudices.
You also linked to his page picking apart esperanto several years ago.
You can’t have incremental spelling reform.
I still say (guestimating out of thin air) that some systematic changes to about 10-15 of the wordstock would aliminate about 80-90 % of the problems.
And this would not make printed materials inaccesible. Anyone who wanted/needed to read them could learn to passively decipher them relatively easily.
Of course non-linguistic considerations mean all this is moot.
… the massive expense and loss of productivity associated with (a) transliterating the entire corpus of English language writings …
Within a decade, this cost is going drop to near-zero: Google Books, print-on-demand.
It will be pretty easy to set up an automated transliteration system to convert from the old spelling to the new. The only difficulty I can see would be the rare heteronyms like bow (down) and bow (arrow), but even there errors could be reduced through grammar-checker logic being applied during the transliteration process.
Regarding the article, I think his response to argument #11 is weak. There’s no way you could possibly implement a phonetic spelling system without forsaking the mutual intelligibility of the different English dialects (or forcing people to learn lots of alien pronunciations, which would largely defeat the entire point).
An objection to spelling reform that the article doesn’t point out is that attempt at introducing major reforms to English spelling would probably just result in people having to learn both the current system and the reformed system. As far as I understand, this is essentially the case in Norway (where they have two systems of spelling) and also in China (where most educated people wind up learning both simplified and traditional characters).
And, what exactly are the problems with the current system? Most educated native speakers have no real problems with English spelling. Even most educated foreigners that I meet have few problems (to the extent that they have mastered the language in general). Most uneducated people can *read* just fine, and to the extent that they can’t I don’t think it’s a matter of the difficulty of the orthography. As far as writing, automated spellcheck has largely eliminated that problem for digital texts, so the issue is really one of handwriting – and it is really worth the hassle of spelling reform so that uneducated people don’t make mistakes when writing by hand?!
Thanks for the plug!
Perhaps I can reduce the temperature by pointing people at the
subtitle, and section 13. I might support “revolutionary” changes
in principle, but in practice they’d require a world dictator with
powers of mind-control, so it’s “revisionist” changes or nothing.
Creeping reform isn’t implausible, it’s already happening to
UK-standard spellings like “anaesthetised”. It’s just a question of
whether it’s systematic.
A lot of old literature is “misspelled” already, and often it’s published in spelling-modernized editions.
Writing is dominant these days, not speech. We would be better off keeping the spelling and scrapping the current pronunciation(s) in favor of spelling pronunciation.
The post-WW2 reform of Japanese kana spellings is a counterexample to the idea that spelling reform would just result in everybody learning both systems (although inevitably this will be true at the beginning). I’m only a second-language-learner, so I could be mistaken, but as far as I can tell the reformed system is used for all practical purposes. I have a few textbooks from the 70s which describe the historical usage, so it was presumably still likely that a foreigner would encounter it then (25 years or so post-reform), but 60 years on it seems to have almost vanished.
So it is possible to push through sweeping spelling reforms (and I’m certainly glad they did :-))
¿Cómo?
I can’t hear you!
Erm, firstly, it would be phonemic, not phonetic…
Secondly, people already learn lots of alien pronunciations. How many millions of Americans pronounce lord and lard the same and have to learn by heart which is spelled which way? How many of them write things like idealogy which sound patently wrong when read with a British accent, but are indistinguishable from the correct spelling in their own accent? If you want to write all standard pronunciations of English in the same way, you have to make more distinctions than any of these accents makes, and indeed the current orthography does that*. It just goes beyond that and adds distinctions that nobody has made for centuries.
* Only after I had finished school I figured out that there are plenty of Americans who still don’t pronounce w and wh the same.
In Norway, those are not just two systems of spelling, they are two different standards of the language. The traditional Chinese characters are still in use in Taiwan and Hongkong — that’s why people in the PRC who want to read stuff from there have to learn the traditional characters.
After having successfully finished an education that could have been quite a bit shorter (leaving time for other subjects) and could have had a considerably larger success rate, yes, of course.
As I said before: for a large proportion of English words it is easier to learn the spelling first and the pronunciation at the same time or later than to do it the native-speaker way. If an orthography is more difficult for native speakers to learn than for foreign learners, you’re doing something wrong.
Spellcheckers make a lot of annoying mistakes. Have you noticed how the word “allot” has become so common (and seems to be mainly used by people who probably don’t know that rare word in the first place)? That’s because spellcheckers correct alot to allot instead of to a lot.
I wrote:
And it fails to make certain distinctions that probably everyone has made for at least as long. Someone mentioned bow.
Hey, English already has ablaut; you don’t need to make the etymological connection that obvious even if you think etymology should be a prime concern of orthography. German: Bogen — verbeugen/Verbeugung. And biegen “bend” and so on.
Okay, so I stand by pretty much everything I already wrote:
– Any new phonetic/phonemic system to be used across all dialects of English would still be quite alien to most speakers (not as different from everyday speech is from the current system, obviously, but kids are still going to be spending a lot of time taking spelling tests).
– Given that many people will want to read older texts and that it’s unlikely that the entire English-speaking world would want to adopt the reforms (just as the entire Chinese-speaking world didn’t want to accept simplified characters), most educated people will have to learn *both* systems – a major hassle.
– The current system is not a barrier to learning to *read* the language for anybody. Most people, when looking at a strangely-spelled word in context, will be able to figure out what it is.
– Your single example aside, spellcheckers largely fix the problem when we’re talking about composing texts on the computer, which is how most things are written these days (and spellcheckers are constantly getting better, so problems of the kind you mention will become scarcer and scarcer).
So we’d essentially be going through a lengthy, miserable process just to help uneducated people write properly by hand. This is probably why nobody seriously advocates spelling reform.
I could not resist.
The European Union commissioners have announced that agreement has been reached to adopt English as the preferred language for European communications, rather than German, which was the other possibility.
As part of the negotiations, the British government conceded that English spelling had some room for improvement and has accepted a five-year phased plan for what will be known as EuroEnglish (Euro for short).
In the first year, “s” will be used instead of the soft “c”.
Sertainly, sivil servants will resieve this news with joy. Also, the hard “c” will be replaced with “k”. Not only will this klear up konfusion, but typewriters kan have one less letter.
There will be growing publik enthusiasm in the sekond year, when the troublesome “ph” will be replaced by “f”. This will make words like “fotograf” 20 per sent shorter.
In the third year, publik akseptanse of the new spelling kan be expekted to reach the stage where more komplikated changes are possible. Governments will enkorage the removal of double letters, which have always ben a deterent to akurate speling. Also, al wil agre that the horible mes of silent “e”s in the languag is disgrasful, and they would go.
By the fourth year, peopl wil be reseptiv to steps such as replasing “th” by z” and “w” by ” v”.
During ze fifz year, ze unesesary “o” kan be dropd from vords kontaining “ou”, and similar changes vud of kors be aplid to ozer kombinations of leters.
After zis fifz yer, ve vil hav a reli sensibl riten styl. Zer vil be no mor trubls or difikultis and evrivun vil find it ezi tu understand ech ozer.
Ze drem vil finali kum tru.
I wish you had resisted. I’ve seen that one so many times before. And it’s kind of racist.
Simplified spelling comes and goes in waves. In 1906, Andrew Carnegie gave $250,000 to establish the Simplified Spelling Board. The Board quickly produced a list 300 words with a simplified spelling alternative (but these were comparatively un-radical suggestions like ax rather than axe, and catalog rather than catalogue).
Theodore Roosevelt ordered the Board-approved spellings to be adopted by the Government Printing Office. But the movement petered out when the Simplified Spelling Board started calling for spellings like tuf, def, troble, filosofy and others. There’s certainly a case to be made for these spellings. But they do look goofy (or is that gufy?).
@Reader: spellcheckers largely fix the problem
The problem is the time and brain cells students and people have to waste on arbitrariness. Compared to that, the fact that people need spellcheckers is a small concern.
@bathrobe: it’s kind of racist
Wiktionary tells me “racist” colloquially means “discriminatory”, is that the meaning you intend here? Because surely you’re not talking about the “German race”? Or am I missing something?
@David Marjanović: it would be phonemic
A close-to-phonemic English orthography used across varieties of English could have graphemes for each diaphoneme, i.e. make phonemic distinctions that are made in any variety.
Different varieties have made different mergers, but if the distinction is already made in the current orthography, people would be no worse off. E.g. people pronouncing “father” and “farther” the same would still have to spell them differently, but they’re used to that.
Splits are a bigger problem, though. E.g. Americans would have no idea which a’s are TRAP and which are BATH. Similarly for the LOT/CLOTH split. Of course, the orthography could merge those splits.
Another similar problem is weak vowels. I don’t think the weak /ɪ/ – /ə/ distinction is straightforward from spelling, so if the reformed orthography were to make it, those that merge weak vowels would face a new headache.
Finally, even if these problems were straightened out, there would be some words using different diaphonemes in different varieties, the most prominent one I can think of being “what”, which has STRUT in AmE, but LOT/CLOTH in BrE. “Been” and “again” are others.
Dainichi: Axel Wijk worked out how such a diaphonemic orthography would work in fine detail, making as few changes as possible. His goal is to get English spelling to the state of French spelling: the rules are complex and they run only from orthography to pronunciation, not the other way, but at least there are no (or vanishingly few) exceptions. Another goal is to make no unnecessary changes: thus ph doe not become f, because the spelling rule is uniform. He spells PALM as aa and TRAP as a, and accepts that BATH (and DANCE) words will have different spellings in different countries (as many words already do for basically trivial reasons: -our/-or, -re/-er, etc.) Given how low the functional load actually is, I’m inclined to stick with traditional orthography. As for CLOTH/LOT, very few people still have this split (which implies a CLOTH=THOUGHT merger), either because they have re-merged it or because they have submerged it into the LOT=THOUGHT merger. Again, I’m inclined to stick with the 14C here.
Regularized Inglish, my recension: theory, wordlists, extended example.
Oh no, they’d face a lot fewer headaches than nowadays – instead of distinguishing two spellings for fully reduced vowels, the current orthography distinguishes something like ten! (Or twenty if you’re non-rhotic.) And that’s even without counting one-off exceptions like the silent or in comfortable.
Wijk concluded that it was hopeless to try to do anything about the single reduced vowels, though he does suggest changing reduced ai to e: capten, bargen versus detain, retain. Keeping the etymological vowels is the most important concession to word relationships (nation, nationality, e.g.) My idea is to have a children’s-books-and-poetry mode in which stressed syllables are marked with acute, and unstressed but unreduced (aka secondarily stressed) vowels with grave.
> Oh no, they’d face a lot fewer headaches than nowadays – instead of distinguishing two spellings for fully reduced vowels, the current orthography distinguishes something like ten! (Or twenty if you’re non-rhotic.)
Sure, I completely understand the point, but people would rather keep their current headaches than exchanging them for new ones, even if they’re much lighter. They’ve invested in the pills that help with the current type. My point is that for any reform to be realistic (by which I mean less unrealistic), people would have to be able to know how to spell things, either from their own pronunciation or the old spelling.
As for CLOTH/LOT, very few people still have this split (which implies a CLOTH=THOUGHT merger),
I’m confident that you could round up well over 100 million Americans with it; I wouldn’t call that “very few”. (Vaux found 61% of respondents distinguishing LOT/THOUGHT in 2003, and this distinction nearly always co-occurs with the LOT/CLOTH split in North America.) That said, it’s capriciously variable in its scope (no two American dictionaries can ever agree on the same word set), and like other “local splits” such as TRAP/BATH, I’d definitely exclude it from a reformed orthography. (On the other hand, I’d be a lot more sympathetic to endangered “original” distinctions like w/wh.)
the silent or in comfortable.
Or better yet, metathesized: it’s commonly “comfterble” in North America.
Portuguese makes do without: nação, nacionalidade…
Now that I think about it, I think I’ve noticed!
David M: “comfterble” – I think I’ve noticed!
And I don’t think I have! But of course, you and I don’t hear the same speakers.
I wonder if such people (consciously or not) associate comfortable with comforter (a kind of blanket or bedspread) (main stress on the first syllable, slight stress on the third, none on the second) rather than comfort. So “comforterable” with loss of the totally unstressed “or”.
No, it’s an extremely common pronunciation — I use it myself — and it’s purely a simplification of the excessively long word. It certainly has nothing to do with “comforter,” a word that is barely part of my vocabulary. I’m sure you’ve heard it and just haven’t noticed; now that you’re paying attention, I’ll bet you hear it frequently.
a word that is barely part of my vocabulary
Presumbaly [sic], comforter is like chifforobe: you use it if you have one, you don’t use it if you don’t. We have lots of comforters in our house and no chifforobes.
@John Cowan: Keeping the etymological vowels
… is a bad idea, if you ask me. Orthography designers are likely to overestimate the understanding of etymological relationships in the general population. Also, I doubt this can be done consistently. Nation – nationality, but maintain – maintainance? Pronounce – pronounciation?
> As for CLOTH/LOT, very few people still have this split (which implies a CLOTH=THOUGHT merger), either because they have re-merged it or because they have submerged it into the LOT=THOUGHT merger.
By re-merged, you mean (at least) RP, but not GenAm, right? As Lazar said, my understanding is that practically all non-cot-caught merged AmE speakers maintain the split, although details vary.
@Lazar: and like other “local splits” such as TRAP/BATH, I’d definitely exclude it from a reformed orthography.
Probably just stating the obvious here, but the problem is that often one or both of the split branches merge with something else, such as BATH with PALM or LOT-PALM and CLOTH-THOUGHT. So leaving the split out from a reformed orthography not only means more phonemes per grapheme, but also more graphemes per phoneme.
JC: comforter is like chifforobe: you use it if you have one, you don’t use it if you don’t
I think most people have at least one comforter at home, which they use to add extra warmth to their bedding, and even the ones who don’t personally own one are familiar with the item from seeing it in other people’s homes or in stores selling bedroom supplies. I have sometimes offered the use of a comforter to overnight guests, and no one has ever asked what the word meant. On the other hand, I think the word chifforobe was described here quite some time ago and appeared to be unknown to most readers, including myself. It refers to a hybrid piece of furniture that was popular at one time but is no longer so, and few people even remembered seeing one.
As Lazar said, my understanding is that practically all non-cot-caught merged AmE speakers maintain the split, although details vary.
Yep, that’s me.
Yes, I was using “racist” very loosely in the discriminatory sense. I probably wouldn’t use that now because racism has strongly coalesced around “black vs white vs yellow (vs a few others maybe)”, despite the fact that geneticists tell us these are not very meaningfully categories genetically.
At any rate, I wasn’t referring to the German or French “races” but attitudes among English speakers making fun of people who don’t speak English natively.
In point of fact, the joke was only mildly discriminatory. I’m certainly not averse to people bunging on a foreign accent for fun. But I found the low-level kind of humour in that joke silly and in poor taste.
If I encountered maintainance, I’d stress it on the unreduced-looking second syllable…
If I encountered maintainance, I’d stress it on the unreduced-looking second syllable…
I think I used to hear that pronunciation, though I haven’t noticed it for a long time.
I concede that there are a lot more speakers with unmerged LOT/THOUGHT but merged LOT=PALM in the U.S. than I thought (and tried to palm off on the rest of you). Labov’s survey back in 1996 showed 40% merged, 60% unmerged. The trend is against us, though: younger people are more likely to be merged in any given locale. WP says the main areas of resistance are the Northeast Corridor (where THOUGHT is raised), the Inland North (where LOT is fronted), and the Inland South (where THOUGHT is broken to a diphthong). Of course, NE New England, Canada, and Scotland are fully merged, but the populations of these places are small compared to (most of) the U.S.
Wijk’s reform is meant to be a reform, not a revolution like most spelling “reforms”. It clears out existing irregularities like “one” that come under no intelligible rule, rather than redesigning the written language from scratch. Wijk spent about forty years analyzing the existing spelling system from the viewpoint of how it actually works, rather than just saying “Let me enumerate the phonemes (or at best diaphonemes) and devise a spelling for each.” Thus consonants and clusters as well as vowel digraphs end up with one pronunciation each, single vowels with at most three (“long”, “short”, and schwa).
I think Wijk went a little too far in distinguishing aa (PALM) from a, and in writing dh for eth; in these cases, he was catering to the second-language rather than the native speaker. But the functional load of both distinctions is very low.
Surprisingly unmentioned yet is Mark Rosenfelder’s computer-assisted version of (what appears to be) basically the same thing Wijk did (or, at least, something very similar).
…It doesn’t seem to have /ð/ at all that I could see, except in the initial phonology listing – presumably because it’s so scarce in vocabulary terms. Apparently the non-initial version is normally an intervocalic allophone? I’m not sure of the details, but as far as I could tell from my attempts at googling, it seems fairly regular.
Initial /ð/ is limited to a few (mostly common) words, and can probably be safely spelled dh in there, because initial dh is otherwise exceedingly rare* – though this isn’t really the way Mark Rosenfelder’s system works.
That leaves with, which is indeed an exception, but apparently doesn’t have /ð/ at all in some dialects, ether, which looks like it has /ð/ but doesn’t, and probably a few other exceptions I couldn’t find quickly.
(I’m not sure what it has for PALM either; best I can tell, this lexical set mostly consists of a bunch of exceptions to other rules, and examples of it end up variously lumped with LOT, THOUGHT or COMMA.)
*) dhow and its plural, dharma and family, a bunch of exotic place names (Dhofar, Dhekelia, Dhaulagiri…), and everything else is even rarer
If spelling were to be reformed, I think it would be important (from the perspective of getting people to adapt it, not that I’m personally in favor of it) to maintain all existing distinctions that are currently shown in spelling, regardless of whether they are inherited or come from splits or whatever. When I was a kid, I was very into music (I still am, but what’s changed is that I’ve learned some linguistics in the meantime), and there are two spellings in particular that I remember really bothering me back then.
In the late 80s, I listened to a lot of hip hop, & thru it learned the word “wack” (always (in my experience) thus spelled in lyric sheets). I have the w-/wh- distinction (phonetically the difference in my production is subtle, I think, but it’s definitely important phonologically), and it really bothered me that when newspapers wrote about rap music, they invariably (again, in my experience) spelled this word “whack” (which I guess is probably correct etymologically, but I haven’t checked & am not sure). This was very clearly wrong to me (tho I don’t think I even knew I made that distinction & had no idea why it was wrong) — wack and whack were completely different words!
A little later, in the early 90s, I listened to a certain amount of indie rock. Many bands spelled “thought” (I distinguish lot/thought & lot/cloth) “thot” on their lyric sheets, which just seemed incredibly wrong to me (again, I had no idea why it seemed so wrong, or why it would ever make sense to anyone to write “thought” as “thot”), to the extent that, when I was in a band with a friend of mine who used the “thot” spelling in his fanzine, I was worried about making sure that when we put out our tape, its liner notes nowhere used the “thot” spelling (the issue ended up never coming up, which I was relieved about).
And I had no issues with alternate spellings, as, e.g., “fuk shit up” was perfectly cromulent to me. But there was some psychic damage in distinctions that I made which were generally preserved in writing being eliminated, tho I had no idea what was going on until I took a linguistics class a few years later. And I also had no idea that anyone distinguished between pin & pen, for example (I don’t). Plenty of distinctions (e.g., trap/bath, which of course I don’t have, or ð/θ, which of course I do) already don’t show up orthographically, which is fine, but taking ones away seems like trouble to me.
(For what it’s worth, I was born in 1977 in North Carolina, & I pronounce “comfortable” as “comfterble”.)
But obviously, if my spelling in my previous comment is anything to go on, I should be in favor of merging “adapt” and “adopt”.
“whack” (which I guess is probably correct etymologically, but I haven’t checked & am not sure).
It appears to come from whack ultimately, but mediated through wacky – so I think wack is preferable.
Many bands spelled “thought” (I distinguish lot/thought & lot/cloth) “thot” on their lyric sheets,
Time has marched on with this one: thot is now a popular slang term of unrelated meaning and origin (see Urban Dictionary).
I’ve only ever seen that “thot” written before and had wondered about how to pronounce it, but had forgotten about it when I commented. Is that pronounced /θɑt/ (assuming no caught/cot merger)? That was my guess, but I really have no idea.
Yeah, it’s been in the LOT class when I’ve heard it from speakers who make the distinction.
And θ, not ð, right?
After typing that, I looked it up on wiktionary (I didn’t expect to find it there!), and they say it comes from an acronym starting with “that”, which would suggest ð, but that seems unlikely. They also give the pronunciation as “(UK) IPA: θɒt/”; “(US) IPA: /θɔt/”; “(cot–caught merger) IPA: /θɑt/”, but that suggests to me that whoever wrote that has got it backwards in terms of US merged/unmerged usage, or, really, that they don’t understand the difference between phonemic and phonetic (that is to say, I think Lazar’s right about the unmerged vowel, & people with the merger could say either of those vowels, depending on their circumstances, and that there’s no reason for separate merged/unmerged phonemic transcriptions). But whatever.
I was also surprised to learn from that page, thot [sic], that “thot” is a Scots way of spelling not just “thought” (dating from at least the 16th century) but also “though” (dating from at least the 15th c). Could this be right?
What I mean is, did 15th century (or later) Scots speakers really end “though” with a -t?
It used to be the intervocalic allophone of /θ/, and not much has happened to change that:
1) the treatment of unstressed words as not phonological words, so their initial (the, this, that, there, though; thou, thy) or final (for some but not all speakers: with) */θ/ came to be treated as intervocalic;
2) apocope stranding intervocalic */θ/ *[ð] in final position, creating a word-final contrast with originally final */θ/ *[θ] (mouth is a minimal pair even graphically: /θ/ for the noun, /ð/ for the verb);
3) Greek words with th never get /ð/, regardless of position; the Biblical name Ethan is lumped in with these.
I don’t know if the fourth process, the shortening of the very rare Old English phoneme /θː/, has any effects that remain today. Moth is one such word, and people who extend the intervocalic voicing of */θ/ to the position before the plural ending |z| pronounce moths with /ðz/.
…where the dh in Dhekelia and Abu Dhabi is already meant to represent foreign [ð].
As Rosenfelder says, he’s using his own pronunciation, which evidently has PALM = LOT (but quite possibly the THOUGHT vowel for the word palm itself); further, as in the US generally, half doesn’t belong to PALM in the first place, but to BATH.
And, of course, since his system worked on spelling that didn’t indicate stress, he ended up treating all instances of final -a as unstressed, which of course most of them are (comma, panda), and thus (as far as I can tell) didn’t account for the relatively few words (mostly monosyllabic, and mostly abbreviations/truncations) that did have a stressed final -a, which are traditionally counted in the PALM lexical set.
could have graphemes for each diaphoneme, i.e. make phonemic distinctions that are made in any variety
I’ve heard somewhere of a proposed 19th century Bulgarian spelling system that tried to do the same thing for Bulgarian dialects, and ended up using 16 different vowel letters; the Cyrillic system of spelling vowels meant it wasn’t quite as ridiculous as it sounded like – 19th century Russian already had 13 vowel letters, counting the Ё, so IIRC they just used that, Ъ, and the yuses – but it still was a bit tricky.
[I checked and apparently this isn’t quite the vowel set that system used, but the number 16 is correct.]
With the sheer amount and variety of English vowel sounds (even for actual phonemes within a single dialect, never mind cross-dialect diaphonemes), and the relatively few vowel letters, we might have to resort to trigraphs, and I suspect that even that might not be enough.
Of course a diaphonemic French-style spelling (i.e. with the spelling-to-pronunciation mapping being [almost] entirely rule-based for every individual dialect) is what Irish has; and I would not exactly call Irish spelling particularly sane.
But of course Irish dialects had diverged a lot more than the English ones did (at least, if we don’t count Scots among English dialects).