Most of English's inconsistent spelling came about from (sometimes incomplete) pronunciation shifts. Now knight and night are homophones where they once weren't, and while sky still rhymes with by, it no longer rhymes with archery. Fixing this is difficult, not just because people are resistant to change, but also because the variations in accents.
However, we're also introducing a lot of new inconsistencies due to a relatively recent shift to adopting foreign words without changing the spelling or pronunciation as we would have in the past - something no other language does. This forces English readers to learn multiple foreign orthographies and English's to read English.
The British are sometimes better about it. They see the word "jalapeno" or "tortilla" and pronounce it like you'd expect (and get mocked for it) with English's orthography, as opposed to forcing everyone to use Spanish orthography to pronounce them halapenyo and tortiya.
Having spent quite a bit of time among the British dialect speaking population, one of the things that stood out was the systematic and unapologetic anglicization of obviously foreign words. It was very regular and obvious, which has significant benefits, but the practice was different than I am used to with American English.
In the US, words of Spanish origin are commonly pronounced according to Spanish rules, reflecting that a significant percentage of the US population has a passing familiarity with Spanish for historical reasons. This is expected as an American dialect speaker. You see the same, to a lesser extent, for French. Americans in many parts of the US code-switch pronunciation to reflect the origin of the word without thinking about it. People will poke at you a bit if you don’t know the pronunciation of foreign words. But that is far from universal among English dialects. It was a noticeable difference with UK English aside from the dialectical differences that everyone is familiar with.
I’m not making a value judgment, it is just an interesting example of differences in English dialects that go beyond the language. In my experience, North Americans tend to hold truer to the original pronunciation of foreign words even as they butcher English with their regional variations. North Americans, by and large, are oblivious to it. I only noticed by spending a lot of time among native English speakers on other continents.
I have never once heard anyone pronounce the ‘l’s in ‘tortilla’ as if it was a Germanic word. And if anything, I hear people hypercorrect and insert ‘ñ’ where it shouldn’t (It’s not ‘habañero’, dammit) more than I hear them drop it from ‘jalapeño’.
I say we should double down and say ‘English is no longer phonetic. There are 26 symbols which are arranged into words arbitrarily, and you just have to learn every reading by rote.’
Every sentence should be like trying to read: ‘Siobhan Llewellyn-Nguyen ate the jalapeño and dulce de leche gyōza that she kept in her Versace bag alongside an unopened bottle of Moët-flavoured weißbier.’
All those non-English words you used do actually come from languages with phonetic alphabets (albeit only in one direction for French).
And while English spelling is inconsistent an educated native speaker can often pronounce newly-seen words with decent accuracy based on previous encounters with similar words so it’s not completely arbitrary.
English spelling gives you a pretty good idea of a word’s origin, which gives you a hint to its pronunciation. And this phenomenon of importing foreign words unchanged is hardly new so I think it’s something most native speakers would be familiar with.
Personally I really like English’s spelling. I like being able to look at a word and have a good idea of what its origin is, I think it adds something to the language. Reading English is like looking back through time. I understand the arguments for phoneticisation but I really think we would lose something great in the process.
> Fixing this is difficult, not just because people are resistant to change, but also because the variations in accents.
The relevance of accents is greatly overstated. The argument is of the form "we should let the perfect be the enemy of the good, and therefore it's impossible". There are a great many words in English whose pronunciation is irregular: these are the ones we should fix. For these, accent is irrelevant; you can pronounce your r's hard or your a's broad, and it doesn't matter: "bury" is pronounced to rhyme with "merry" in probably every accent of English that's ever been, from Old English (ic byrge vs myrge) on. You could just fix 100 words like "bury" and "could" and "are" whose spellings are either wrong or etymological but don't reflect extant variants, and the spelling would be reformed, children's lives would be improved, and it wouldn't be a problem from any perspective of accent variation or etymology or anything.
> "bury" is pronounced to rhyme with "merry" in probably every accent of English that's ever been
I've definitely heard speakers for whom "bury" rhymes with "furry", and that's without the "Merry–Murray merger" (i.e., the same person would pronounce "berry" to rhyme with "merry" and quite distinctly from "bury".)
> You could just fix 100 words like "bury" and "could" and "are" whose spellings are either wrong or etymological but don't reflect extant variants, and the spelling would be reformed, children's lives would be improved, and it wouldn't be a problem from any perspective of accent variation or etymology or anything.
In many cases it would take existing homophones and turn them into additional meanings of the same spelling, which would actually reduce clarity and comprehensibility of written text.
A Plan for the Improvement of English Spelling, sometimes attributed to Mark Twain:
For example, in Year 1 that useless letter "c" would be dropped to be replased either by "k" or "s", and likewise "x" would no longer be part of the alphabet. The only kase in which "c" would be retained would be the "ch" formation, which will be dealt with later. Year 2 might reform "w" spelling, so that "which" and "one" would take the same konsonant, wile Year 3 might well abolish "y" replasing it with "i" and Iear 4 might fiks the "g/j" anomali wonse and for all.
Jenerally, then, the improvement would kontinue iear bai iear with Iear 5 doing awai with useless double konsonants, and Iears 6-12 or so modifaiing vowlz and the rimeining voist and unvoist konsonants. Bai Iear 15 or sou, it wud fainali bi posibl tu meik ius ov thi ridandant letez "c", "y" and "x" -- bai now jast a memori in the maindz ov ould doderez -- tu riplais "ch", "sh", and "th" rispektivli.
Fainali, xen, aafte sam 20 iers ov orxogrefkl riform, wi wud hev a lojikl, kohirnt speling in ius xrewawt xe Ingliy-spiking werld.
The European Commission has just announced an agreement whereby English will be the official language of the European Union.
As part of the negotiations, the British Government conceded that English spelling had some room for improvement and has accepted a 5- year phase-in plan that would become known as "Euro-English".
In the first year, "s" will replace the soft "c". Sertainly, this will make the sivil servants jump with joy. The hard "c" will be dropped in favour of "k". This should klear up konfusion, and keyboards kan have one less letter.
There will be growing publik enthusiasm in the sekond year when the troublesome "ph" will be replaced with "f". This will make words like fotograf 20% shorter.
In the 3rd year, publik akseptanse of the new spelling kan be expekted to reach the stage where more komplikated changes are possible.
Governments will enkourage the removal of double letters which have always ben a deterent to akurate speling.
Also, al wil agre that the horibl mes of the silent "e" in the languag is disgrasful and it should go away.
By the 4th yer people wil be reseptiv to steps such as replasing "th" with "z" and "w" with "v".
During ze fifz yer, ze unesesary "o" kan be dropd from vords kontaining "ou" and after ziz fifz yer, ve vil hav a reil sensibl riten styl.
Zer vil be no mor trubl or difikultis and evrivun vil find it ezi tu understand ech oza. Ze drem of a united urop vil finali kum tru.
Und efter ze fifz yer, ve vil al be speking German like zey vunted in ze forst plas.
English spelling irregularity serves a functional role similar to Japanese kanji. It distinguishes homophones (e.g., night and knight) and evinces the shared meaning of words which have the same root with different pronunciations (e.g., sign and signature). Regularizing the spelling would make the language harder to read for those fluent in it, even for a hypothetical population who learned the reformed spelling from early childhood.
Perhaps interestingly, China did two rounds of simplifying Chinese characters. The first one was well received, but the second one[0] fell out of favor and was rescinded, because it was making up a bunch of new characters (in English that would be equivalent to making up new words). People don't want to remember new words.
I think perhaps more interestingly, they presumably had the opportunity to create a phonetic system like Hangul or hiragana, and certainly the institutional power to force any change through if they really wanted to, but elected not to.
Ah, yes. The romanization movement of Chinese characters[0]. We do have a pinyin system that bacame extremely important when computers came into China. If you didn't know pinyin, you would have to use other complicated methods (e.g. Wubi) to type in Chinese, which requires memorizing a whole new system.
about a time traveling author who goes to the future (via Satanic dealmaking) and finds out nobody remembers him, there is this text copied from that text of the future
From page 234 of "Inglish Littracher 1890-1900" bi T. K. Nupton, publishd bi th Stait, 1992.
Fr egzarmpl, a riter ov th time, naimed Max Beerbohm, hoo woz stil alive in th twentith senchri, rote a stauri in wich e pautraid an immajnari karrakter kauld "Enoch Soames"—a thurd-rait poit hoo beleevz imself a grate jeneus an maix a bargin with th Devvl in auder ter no wot posterriti thinx ov im! It iz a sumwot labud sattire, but not without vallu az showing hou seriusli the yung men ov th aiteen-ninetiz took themselvz. Nou that th littreri profeshn haz bin auganized az a departmnt of publik servis, our riters hav found their levvl an hav lernt ter doo their duti without thort ov th morro. "Th laibrer iz werthi ov hiz hire" an that iz aul. Thank hevvn we hav no Enoch Soameses amung us to-dai!
You will of course note that the future is now our past, and the inevitable future did not come to pass, even though some people are still obviously working on it, and this is very troubling because it means the devil lied and poor Enoch Soames lost his soul and went to hell for nothing, because of a "a swindle, a common swindle" as Max Beerbohm puts it in the story.
At a certain point I think we need to just acknowledge that we actually all learn two quite different dialects—Written English and Spoken English—and we should not expect our Written English to match 1:1 with our Spoken English.
It's a well-known problem in public speaking that if you write out a speech the way that most people want to it will sound terrible and stilted because the rules of speech are dramatically different than the rules of writing. The reverse is also true: if you've ever tried to transcribe speech (as in a talk or an interview) you'll quickly find that to make it at all legible you're actually not transcribing it verbatim, you're translating from the Spoken English dialect to the Written English dialect.
These differences aren't random or accidental either, they developed naturally to solve the dramatically different problems that the two media are meant to solve. Retaining spelling that no longer matches Spoken English is a great example of this in that it serves a number of very important purposes for readers and writers:
* It allows easy communication across large (100+ year) time gaps, which is one of writing's most valuable contributions.
* It distinguishes words that are now homophones but weren't always, providing greater clarity when reading.
* It provides a wider variety of word shapes than a consistent orthography would, increasing the visual differences between words on the page.
Every spelling reform effort gets stopped when the very literate (the ones who you really do have to get on board with a spelling reform effort) realize that the outcome will be that they will have a much harder time using formal writing for its intended purpose.
Where spelling reform actually does happen and works is where language change always happens—new dialects of Written English are evolving all the time, at a faster and faster rate as more people use text for a larger chunk of their communication. But a concerted, centralized effort to change the formal writing system is never going to work.
I’ll have to give this a read! It reminds me of this quote often attributed to Mark Twain (although not verified):
“For example, in year 1 that useless letter ‘c’ would be dropped to be replased either by ‘k’ or ‘s’, and likewise ‘x’ would no longer be part of the alphabet. The only kase in which ‘c’ would be retained would be the ‘ch’ formation, which will be dealt with later. Year 2 might reform ‘w’ spelling, so that ‘which’ and ‘one’ would take the same konsonant, wile year 3 might well abolish ‘y’ replasing it with ‘i’ and Iear 4 might fiks the ‘g/j’ anomali wonse and for all. … After this bakward progression, there would be no more trubl with kommas, semikolons, etc.”
Respelling using only the 26 letters works pretty well for the General American dialect. The most unintuitive specific respellings (otherwise pretty easy) are:
uu for 'oo' in "book". A relatively rare vowel but some very common words use it.
dh for 'th' in "the". The two sounds of 'th' are not fully separated in English.
igh for 'i' in "price". Common but lacks any unambiguous shorter spellings.
The second issue is trying to prevent the default rules from applying. [aeiou]_e is usually easy to fix by doubling the consonant, but that fails with multi-letter consonant sounds (splitting into syllables helps this one). The s->z change at end of words is very difficult to stop (we can easily respell "fleas"/"flees" to "fleez", but respelling "fleece" to "flees" is problematic). The only reason respelling usually works is because most words don't change into existing words.
The third issue, of course, is that there are many dialects with extra sounds (mostly original, but some later inventions), or different pronunciations for only a subset of words that are indistinguishable in other dialects. I have no idea what sound is used to distinguish "for" from "four" in dialects that distinguish them.
But the issue that almost never gets addressed is stress, which is not really a word-level phenomenon but significantly affects pronunciation. For example, "to" has maybe 5 different pronunciations just based on exactly how stressed it is compared to adjacent words (t, ta, to, too, tooo), and it also has variants based on what particular sounds are around it. I'd guess that stress variations exist for every semi-common word that isn't a noun or lexical verb (and some that are).
https://archive.ph/AtTLx
Most of English's inconsistent spelling came about from (sometimes incomplete) pronunciation shifts. Now knight and night are homophones where they once weren't, and while sky still rhymes with by, it no longer rhymes with archery. Fixing this is difficult, not just because people are resistant to change, but also because the variations in accents.
However, we're also introducing a lot of new inconsistencies due to a relatively recent shift to adopting foreign words without changing the spelling or pronunciation as we would have in the past - something no other language does. This forces English readers to learn multiple foreign orthographies and English's to read English.
The British are sometimes better about it. They see the word "jalapeno" or "tortilla" and pronounce it like you'd expect (and get mocked for it) with English's orthography, as opposed to forcing everyone to use Spanish orthography to pronounce them halapenyo and tortiya.
Having spent quite a bit of time among the British dialect speaking population, one of the things that stood out was the systematic and unapologetic anglicization of obviously foreign words. It was very regular and obvious, which has significant benefits, but the practice was different than I am used to with American English.
In the US, words of Spanish origin are commonly pronounced according to Spanish rules, reflecting that a significant percentage of the US population has a passing familiarity with Spanish for historical reasons. This is expected as an American dialect speaker. You see the same, to a lesser extent, for French. Americans in many parts of the US code-switch pronunciation to reflect the origin of the word without thinking about it. People will poke at you a bit if you don’t know the pronunciation of foreign words. But that is far from universal among English dialects. It was a noticeable difference with UK English aside from the dialectical differences that everyone is familiar with.
I’m not making a value judgment, it is just an interesting example of differences in English dialects that go beyond the language. In my experience, North Americans tend to hold truer to the original pronunciation of foreign words even as they butcher English with their regional variations. North Americans, by and large, are oblivious to it. I only noticed by spending a lot of time among native English speakers on other continents.
English is such an interesting mess.
I have never once heard anyone pronounce the ‘l’s in ‘tortilla’ as if it was a Germanic word. And if anything, I hear people hypercorrect and insert ‘ñ’ where it shouldn’t (It’s not ‘habañero’, dammit) more than I hear them drop it from ‘jalapeño’.
I say we should double down and say ‘English is no longer phonetic. There are 26 symbols which are arranged into words arbitrarily, and you just have to learn every reading by rote.’
Every sentence should be like trying to read: ‘Siobhan Llewellyn-Nguyen ate the jalapeño and dulce de leche gyōza that she kept in her Versace bag alongside an unopened bottle of Moët-flavoured weißbier.’
All those non-English words you used do actually come from languages with phonetic alphabets (albeit only in one direction for French).
And while English spelling is inconsistent an educated native speaker can often pronounce newly-seen words with decent accuracy based on previous encounters with similar words so it’s not completely arbitrary.
English spelling gives you a pretty good idea of a word’s origin, which gives you a hint to its pronunciation. And this phenomenon of importing foreign words unchanged is hardly new so I think it’s something most native speakers would be familiar with.
Personally I really like English’s spelling. I like being able to look at a word and have a good idea of what its origin is, I think it adds something to the language. Reading English is like looking back through time. I understand the arguments for phoneticisation but I really think we would lose something great in the process.
> Fixing this is difficult, not just because people are resistant to change, but also because the variations in accents.
The relevance of accents is greatly overstated. The argument is of the form "we should let the perfect be the enemy of the good, and therefore it's impossible". There are a great many words in English whose pronunciation is irregular: these are the ones we should fix. For these, accent is irrelevant; you can pronounce your r's hard or your a's broad, and it doesn't matter: "bury" is pronounced to rhyme with "merry" in probably every accent of English that's ever been, from Old English (ic byrge vs myrge) on. You could just fix 100 words like "bury" and "could" and "are" whose spellings are either wrong or etymological but don't reflect extant variants, and the spelling would be reformed, children's lives would be improved, and it wouldn't be a problem from any perspective of accent variation or etymology or anything.
> "bury" is pronounced to rhyme with "merry" in probably every accent of English that's ever been
I've definitely heard speakers for whom "bury" rhymes with "furry", and that's without the "Merry–Murray merger" (i.e., the same person would pronounce "berry" to rhyme with "merry" and quite distinctly from "bury".)
> You could just fix 100 words like "bury" and "could" and "are" whose spellings are either wrong or etymological but don't reflect extant variants, and the spelling would be reformed, children's lives would be improved, and it wouldn't be a problem from any perspective of accent variation or etymology or anything.
In many cases it would take existing homophones and turn them into additional meanings of the same spelling, which would actually reduce clarity and comprehensibility of written text.
> "bury" is pronounced to rhyme with "merry" in probably every accent of English that's ever been
Bury rhymes with hurry around Philadelphia (NJ, Maryland, some parts of NY).
[dead]
A Plan for the Improvement of English Spelling, sometimes attributed to Mark Twain:
For example, in Year 1 that useless letter "c" would be dropped to be replased either by "k" or "s", and likewise "x" would no longer be part of the alphabet. The only kase in which "c" would be retained would be the "ch" formation, which will be dealt with later. Year 2 might reform "w" spelling, so that "which" and "one" would take the same konsonant, wile Year 3 might well abolish "y" replasing it with "i" and Iear 4 might fiks the "g/j" anomali wonse and for all.
Jenerally, then, the improvement would kontinue iear bai iear with Iear 5 doing awai with useless double konsonants, and Iears 6-12 or so modifaiing vowlz and the rimeining voist and unvoist konsonants. Bai Iear 15 or sou, it wud fainali bi posibl tu meik ius ov thi ridandant letez "c", "y" and "x" -- bai now jast a memori in the maindz ov ould doderez -- tu riplais "ch", "sh", and "th" rispektivli.
Fainali, xen, aafte sam 20 iers ov orxogrefkl riform, wi wud hev a lojikl, kohirnt speling in ius xrewawt xe Ingliy-spiking werld.
The European Commission has just announced an agreement whereby English will be the official language of the European Union.
As part of the negotiations, the British Government conceded that English spelling had some room for improvement and has accepted a 5- year phase-in plan that would become known as "Euro-English".
In the first year, "s" will replace the soft "c". Sertainly, this will make the sivil servants jump with joy. The hard "c" will be dropped in favour of "k". This should klear up konfusion, and keyboards kan have one less letter.
There will be growing publik enthusiasm in the sekond year when the troublesome "ph" will be replaced with "f". This will make words like fotograf 20% shorter.
In the 3rd year, publik akseptanse of the new spelling kan be expekted to reach the stage where more komplikated changes are possible.
Governments will enkourage the removal of double letters which have always ben a deterent to akurate speling.
Also, al wil agre that the horibl mes of the silent "e" in the languag is disgrasful and it should go away.
By the 4th yer people wil be reseptiv to steps such as replasing "th" with "z" and "w" with "v".
During ze fifz yer, ze unesesary "o" kan be dropd from vords kontaining "ou" and after ziz fifz yer, ve vil hav a reil sensibl riten styl.
Zer vil be no mor trubl or difikultis and evrivun vil find it ezi tu understand ech oza. Ze drem of a united urop vil finali kum tru.
Und efter ze fifz yer, ve vil al be speking German like zey vunted in ze forst plas.
I aprov. Dis is god.
English spelling irregularity serves a functional role similar to Japanese kanji. It distinguishes homophones (e.g., night and knight) and evinces the shared meaning of words which have the same root with different pronunciations (e.g., sign and signature). Regularizing the spelling would make the language harder to read for those fluent in it, even for a hypothetical population who learned the reformed spelling from early childhood.
Perhaps interestingly, China did two rounds of simplifying Chinese characters. The first one was well received, but the second one[0] fell out of favor and was rescinded, because it was making up a bunch of new characters (in English that would be equivalent to making up new words). People don't want to remember new words.
[0]: https://en.wikipedia.org/wiki/Simplified_Chinese_characters#...
I think perhaps more interestingly, they presumably had the opportunity to create a phonetic system like Hangul or hiragana, and certainly the institutional power to force any change through if they really wanted to, but elected not to.
Ah, yes. The romanization movement of Chinese characters[0]. We do have a pinyin system that bacame extremely important when computers came into China. If you didn't know pinyin, you would have to use other complicated methods (e.g. Wubi) to type in Chinese, which requires memorizing a whole new system.
[0]: only Chinese version available. https://zh.wikipedia.org/wiki/%E6%B1%89%E5%AD%97%E6%8B%89%E4...
To Quote from Enoch Soames - https://www.gutenberg.org/ebooks/760
about a time traveling author who goes to the future (via Satanic dealmaking) and finds out nobody remembers him, there is this text copied from that text of the future
From page 234 of "Inglish Littracher 1890-1900" bi T. K. Nupton, publishd bi th Stait, 1992.
Fr egzarmpl, a riter ov th time, naimed Max Beerbohm, hoo woz stil alive in th twentith senchri, rote a stauri in wich e pautraid an immajnari karrakter kauld "Enoch Soames"—a thurd-rait poit hoo beleevz imself a grate jeneus an maix a bargin with th Devvl in auder ter no wot posterriti thinx ov im! It iz a sumwot labud sattire, but not without vallu az showing hou seriusli the yung men ov th aiteen-ninetiz took themselvz. Nou that th littreri profeshn haz bin auganized az a departmnt of publik servis, our riters hav found their levvl an hav lernt ter doo their duti without thort ov th morro. "Th laibrer iz werthi ov hiz hire" an that iz aul. Thank hevvn we hav no Enoch Soameses amung us to-dai!
You will of course note that the future is now our past, and the inevitable future did not come to pass, even though some people are still obviously working on it, and this is very troubling because it means the devil lied and poor Enoch Soames lost his soul and went to hell for nothing, because of a "a swindle, a common swindle" as Max Beerbohm puts it in the story.
At a certain point I think we need to just acknowledge that we actually all learn two quite different dialects—Written English and Spoken English—and we should not expect our Written English to match 1:1 with our Spoken English.
It's a well-known problem in public speaking that if you write out a speech the way that most people want to it will sound terrible and stilted because the rules of speech are dramatically different than the rules of writing. The reverse is also true: if you've ever tried to transcribe speech (as in a talk or an interview) you'll quickly find that to make it at all legible you're actually not transcribing it verbatim, you're translating from the Spoken English dialect to the Written English dialect.
These differences aren't random or accidental either, they developed naturally to solve the dramatically different problems that the two media are meant to solve. Retaining spelling that no longer matches Spoken English is a great example of this in that it serves a number of very important purposes for readers and writers:
* It allows easy communication across large (100+ year) time gaps, which is one of writing's most valuable contributions.
* It distinguishes words that are now homophones but weren't always, providing greater clarity when reading.
* It provides a wider variety of word shapes than a consistent orthography would, increasing the visual differences between words on the page.
Every spelling reform effort gets stopped when the very literate (the ones who you really do have to get on board with a spelling reform effort) realize that the outcome will be that they will have a much harder time using formal writing for its intended purpose.
Where spelling reform actually does happen and works is where language change always happens—new dialects of Written English are evolving all the time, at a faster and faster rate as more people use text for a larger chunk of their communication. But a concerted, centralized effort to change the formal writing system is never going to work.
I’ll have to give this a read! It reminds me of this quote often attributed to Mark Twain (although not verified):
“For example, in year 1 that useless letter ‘c’ would be dropped to be replased either by ‘k’ or ‘s’, and likewise ‘x’ would no longer be part of the alphabet. The only kase in which ‘c’ would be retained would be the ‘ch’ formation, which will be dealt with later. Year 2 might reform ‘w’ spelling, so that ‘which’ and ‘one’ would take the same konsonant, wile year 3 might well abolish ‘y’ replasing it with ‘i’ and Iear 4 might fiks the ‘g/j’ anomali wonse and for all. … After this bakward progression, there would be no more trubl with kommas, semikolons, etc.”
<https://www.angelfire.com/va3/timshenk/codes/meihem.html>
Kinda odd that they dont seem to mention: https://www.math.rug.nl/~ernst/linguistics/mark_twain_spelli...
NPR did a story about this book a couple days ago.
https://www.wbur.org/hereandnow/2025/04/15/enough-is-enuf-bo...
Count your blessings: it's way, way worse in French..
Respelling using only the 26 letters works pretty well for the General American dialect. The most unintuitive specific respellings (otherwise pretty easy) are:
The second issue is trying to prevent the default rules from applying. [aeiou]_e is usually easy to fix by doubling the consonant, but that fails with multi-letter consonant sounds (splitting into syllables helps this one). The s->z change at end of words is very difficult to stop (we can easily respell "fleas"/"flees" to "fleez", but respelling "fleece" to "flees" is problematic). The only reason respelling usually works is because most words don't change into existing words.The third issue, of course, is that there are many dialects with extra sounds (mostly original, but some later inventions), or different pronunciations for only a subset of words that are indistinguishable in other dialects. I have no idea what sound is used to distinguish "for" from "four" in dialects that distinguish them.
But the issue that almost never gets addressed is stress, which is not really a word-level phenomenon but significantly affects pronunciation. For example, "to" has maybe 5 different pronunciations just based on exactly how stressed it is compared to adjacent words (t, ta, to, too, tooo), and it also has variants based on what particular sounds are around it. I'd guess that stress variations exist for every semi-common word that isn't a noun or lexical verb (and some that are).
[dead]