English Orthography Reform

[Xing]

As others have pointed out, /ɜ:(r)/ distribution is fairly predictable, deriving from older /ɛr ɪr ʊr/.

I was just thinking, is /ɜːr/ ever contrastive in GA? Don't most Americans rhyme <furry> with <hurry>? So that /ɜːr/, for an American, is basically /ʌ/ followed by /r/? (Or am I way off here??)

[Salmoneus]

Further?

Damnit, I'm an idiot.
For some reason I was reading that 3 as a backward 3 and thinking they were talking about E:...

[Glossaphile]

Perhaps I should clarify that, while I have tried to generally reduce the number of re-associations foreign learners have to make in mastering pronunciation, I realized almost immediately that eliminating them altogether would be impossible, given the limits of English phonology and the sheer variety of native phonologies that non-native students bring to the table.

Using <æ> for /ʌ/ would not only re-introduce what is to most a new letter, it would also conflict with its own original historical purpose. That's two strikes against it, whereas using it for /æ/ instead means only one strike. This seems especially apparent if we continue to presume that <æ> would be added to the alphabet no matter which phoneme it ends up representing. The unfamiliarity of the letter itself will be a factor in any case, so I think it would be more persuasive if we can at least justify the apparent intrusion with some historical legitimacy, which would only really work if it has its original sound value.

The use of <x> for /ʒ/ could be said to be the opposite scenario. The /ʒ/-phoneme needs its own symbol, which it's never had before, so something's bound to end up sounding different than it traditionally does in any case. So why not at least let the symbol itself be a familiar one? The same goes for affricate-<c>, at least under the premise that digraphs are a last resort.

One might think that the imports <ç> for /ʃ/ and <ñ> for /ŋ/ are contradictions, being both foreign and assigned to novel pronunciations, but there are two justifications for that. First, all other options have essentially been exhausted (redundancies in the current 26-glyph repertoire have already been ironed out). Fortunately, the US-International keyboard offers two accessible symbols that also happen to visually resemble conventional letters that have been assigned to sounds which are closely related to the two in question. This seems too fortuitous of a convenience to pass up.

As for other keyboard layouts, I think the US-International would be the most readily available worldwide, given the role of the US in the computer industry. Nevertheless, investigating other options such as the Dvorak or UK-International may not be a bad idea, so thank you for that, Xing.

As to the predictability of /ɜː(ɹ)/, I have two reasons for its treatment in RLS. First, I took my cues from what seems to be the prevailing phonological analysis of broadcast English, which maintains it as a separate phoneme. Secondly, it supports the dialectal flexibility I mixed into the system. In GA, according to the Cambridge Online Dictionary (British edition, which lists both RP and GA pronunciations), the formerly rhotic diphthong /ɛə˞/ tends to collapse into just /ɛɹ/ in American English, thus making words like "fur" and "fair" minimal pairs (RLS <fér> versus <fer> for GA, <fe> versus <fèø> for RP, and <fér> versus <fèør> for the neutral hybrid that I've proposed). There are also a few words in GA, such as "tour," which Cambridge transcribes as /tʊɹ/, where /ʊɹ/ also appears distinctly from /ɜ˞/. With that contrastivity in mind, the free value of <e> seemed like the overall best slot to host /ɜː/, completing the checked/free symmetry of the five conventional vowel letters which I suspect will be mnemonically helpful. Given that it's also not entirely without precedent in traditional spelling, as in "herd" or "jerk," it seemed like a pretty solid choice.

As to the diphthongal status of /eɪ/ and /oʊ/, again, I designed the system primarily with the US and UK national standards in mind, and neither one uses /eː/ or /oː/ in their place. However, as part of my developing dialectal toolkit, I do reserve the digraphs <ee> and <oo> for precisely those forms (I'm much more permissive with digraphs when it comes to notable accents other than GA and RP). As for the diphthongal variants [ij] and [uw], I hadn't considered them before, but they could easily be accommodated with <iy> and <uw>.

I can see how you might also see a contradiction in my diphthong policy with my treatment of affricates, but while I agree that affricates can generally be thought of as the consonantal equivalent to diphthongs, there is one behavioral difference, stemming from the very fact that their components are consonants rather than vowels. Whether or not two consecutive vowels form a diphthong or remain distinct affects the number of syllables present. A stop-fricative combination, on the other hand, doesn't affect how many syllables there are regardless of whether it's parsed as a cluster or as an affricate. One can also see this as minor compromise to already-literate natives, who classify the affricates clearly as wholly contrastive and unitary phonemes (in fact, this might be the stronger argument of the two).

Yes, <férðør> for /fɜ˞ðə˞/ and <feðø> for /fɜːðə/ are the rhotic and non-rhotic forms of "further," respectively.

As for those rhotic words...

IHS = international hybrid (of GA and RP) standard

Mary = <Mèørri> for IHS /mɛə˞(ɹ)iː/, <Mèri> for GA /mɛɹiː/, <Mèøri> for RP /mɛəɹiː/

merry = <mèri> for /mɛɹiː/ (all three accents agree on this one)

marry = <mæri> for IHS & RP /mæɹiː/, <mèri> for GA /mɛɹiː/

furry = <férri> for IHS & GA /fɜ˞(ɹ)iː/, <feri> for RP /fɜːɹiː/

hurry = <hàri> for IHS & RP /hʌɹiː/, <hérri> for GA /hɜ˞(ɹ)iː/

ferry = <fèri> for /fɛɹiː/ (all three accents agree on this one)

sorry = <sòri> for IHS & RP /sɒɹiː/, <sarri> for GA /sɑɹiː/

lorry = <lorri> for IHS & GA /lɔɹiː/, <lòri> for RP /lɒɹiː/

story = <storri> for IHS & GA /stɔɹiː/, <stori> for RP /stɔːɹiː/

The definition of "rhotacization" is extended for spelling purposes to refer to the shortening of /ɑː/ and /ɔː/ before r-codas. This is because the same combinations are variously transcribed as /ɑ˞/, /ɑɹ/, or /ɑə˞/ and /ɔ˞/, /ɔɹ/, or /ɔə˞/, suggesting distributional behavior at least somewhat distinct from purely compositional /ɔː.ɹ/ and /ɑː.ɹ/. This classification also prevents redundant acute accents in words like <start> or <fort>, where the vowels that the acutes would prevent, /ʌ/ and /ɒ/, are phonotactically prohibited in those positions anyway.

This discussion wouldn't be the first time RLS has been criticized for being a bit too detailed, but honestly, I think it better that the system be slightly over-equipped than at all under-equipped. If it's ultimately adopted, people may dispense with a few of the finer distinctions I tend to make, but at least the potential was there in case they unexpectedly did find such distinctions useful.

As a consequence of the fact that any intervocalic <r> must be doubled if the preceding vowel is to be rhotacized (see Lesson 4 at http://www.hsmespanol.com/RestLatSpellSite/Routiks.html), one very hair-splitting distinction RLS can not make is that between examples like "overate" (past tense of "overeat") and "overrate" (to rate too highly), both of which would be rendered as <ouvørreit>. From time to time, I question whether it's really safe to call them homophones, at least in rhotic accents (in non-rhotic, there's no doubt as to their homophony), but it so far hasn't bugged me enough to change the system.

[Dormouse559]

I was just thinking, is /ɜːr/ ever contrastive in GA? Don't most Americans rhyme <furry> with <hurry>? So that /ɜːr/, for an American, is basically /ʌ/ followed by /r/? (Or am I way off here??)

I rhyme them. But I don't really perceive /ɜ˞/ as being any non-rhotic vowel plus /ɹ/, probably because it's written with several different vowel letters. For me, it's just "that rhotic vowel".

[Xonen]

In my opinion, a much better way of keeping unconventional characters to a minimum would be not introducing them in the first place, considering that we already have a bunch of digraphs that do the job just fine. A word like adulthood is so transparently obviously a combination of adult + hood that most people have no trouble at all realizing that the <th> combination is not supposed to be pronounced as a digraph here.

You can hear a phoneme. You can't hear a morpheme boundary.

I never said anything about "morpheme boundaries" - or "phonemes", for that matter. Laypeople tend not to know what those terms mean - but that doesn't mean they often aren't intuitively aware of this stuff. In fact, understanding the morphemes and their combination rules in a language is kind of necessary in order to speak it productively.

Besides, even if native speakers are somehow more conscious of phonemes than they are of morphemes, having the same spelling for /T/ and /th/ would be no problem for them (especially considering the rarity of the latter); they would know the proper pronunciation regardless of spelling. And non-native learners need to learn about morpheme boundaries anyway.

When it comes to teaching children, I think we should keep things as concrete as feasible.

School-age children are quite capable of learning - or even figuring out by themselves - that there's a relationship between words like adulthood and adult. If anything, that would seem to me to be more "concrete" a concept than the difference between /T/ and /th/. And why would grade-school children making a few spelling errors be such a huge problem anyway? They're going to do that no matter how good your orthography is. Which is one reason why I wouldn't go out of my way to try to make it toddler-proof.

People have internalized completely separate pronunciations for <c> and <ch>; removing the <h> from the latter and expecting people to still associate it with the old <ch> rather than the old <c> would probably work about as well as removing the ascender from <d> and expecting people not to see it as <ɑ>.

In my experience, the notion that <c> is /k/ before a back vowel can be extremely difficult for people to override once they've internalized it (even though - or perhaps because - many of them are probably only subconsciously aware of the whole "before a back vowel" thing in the first place). If I got a nickel for every time I've heard students of eastern European languages (where <c> tends to stand for /ts/ regardless of what follows) pronounce <c> as /k/ despite being way too far in their studies not to have gone over basic pronunciation rules, I'd have... well, perhaps a couple of dollars at this point (the best example might be [lä:tsäkot] for lazacot from one girl after about six months of studying Hungarian; this stuff is hard for some people). So it wouldn't exactly make me rich, but I'd say it's still indicative of a trend.

Interesting. I wonder what's unique about <c> that makes it harder to re-learn for some than <z>, which that girl had apparently mastered with its non-English affricate pronunciation.

Um. Said girl's native language wasn't English - nor is /ts/ even the proper pronunciation for <z> in Hungarian! That would be just plain /z/, in fact. And her native language was (presumably) Finnish, where <z> isn't properly a part of the orthography at all, but its name when reciting the alphabet is pronounced with a /ts/ (and /z/ doesn't exist).

Not that I'd be surprised if the same girl would actually have no trouble pronouncing <z> as /z/ (or at least a better approximation thereof) in English; again, for whatever reason, people seem to be much more mindful of the fact that they have to change their pronunciation when speaking English.

Still, if you want a single-letter representation for /ʃ/, I'd suggest something like <š>, which is also conveniently related to a letter with a related sound, and, again, has the advantage of actually already being in use in several real-world orthographies. (Another possibility would be the aforementioned <x>.)

The problem with that is that it's not available on the US-International keyboard layout, to which I relegate myself as a sort of compromise for skeptics worried about easy typability. This keyboard layout can be activated within minutes on any Windows PC without any new hardware or software. If reform were ever implemented, it probably wouldn't matter that much in the long run, as manufacturers would probably figure out there's money to be made in upgrading their physical keyboards. But in the initial stages, I think such an appeasing concession to conservatives could be crucial in getting the ball rolling.

So essentially, you're making the spelling less recognizable so it'd be more acceptable to conservatives? Somehow I don't see that working.

On the other forum, the point was raised that it's relatively easy to learn new sounds for old letters, so if you accept that, then continental vowels/diphthongs shouldn't seem nearly as disruptive.

I'd say that's conditional: It's relatively easy to learn new sounds for old letters if 1) you're learning a new language and 2) the new sounds are at least somewhat intuitive. Again, I've met numerous people who can't seem to wrap their head around the fact that <c> isn't always /k/ before back vowels (or that <z> isn't /ts/ in every language that isn't English).

lus, English speakers are not complete strangers to these more Latinate correspondences. They may not be the dominant ones, but they do occur in English, and we don't necessarily have to look at really obscure terminology to find them. Consider "father," "herd," "machine," "or," and "rude" for examples where the vowels are pronounced as they would be in RLS (assuming an RP accent). Most other Roman-script languages, on the other hand, have absolutely no words that uses English-like vowels (except for obvious borrowings). So really, the native English speakers are the ones making the lesser adjustment.

You seem to be comparing native speakers already literate in the old orthography of English to non-natives just beginning to learn it... But ignoring non-natives already literate in English and native speakers just beginning to learn to spell.

Of these, the only group this could possibly benefit are the non-native beginners; to native beginners not yet literate in any language it doesn't matter what letters you use, and to people who'd need to relearn literacy in English this whole thing would just be a huge disruption. Probably most huge to the native speakers, though, since they're the ones most likely to need English in their daily lives. That's not such a small adjustment, in my opinion.

Also, as Xing points out, keeping things more distinct can actually help people remember to mind the differences. Despite all the /k/ for <c> and /ts/ for <z> confusion mentioned above, I've never heard anyone pronounce English with Finnish values for the vowels (except as a joke or a quick way to spell out a word).

But by the same token, if they're similar enough in the first place, the perceived accent or effect of any mistakes decreases significantly. If we assume your view is correct, greater differences mean less frequent but also more egregious mistakes. Smaller differences would mean more frequent but also less severe mistakes. Does either approach have any net edge, then?

That doesn't... necessarily follow from what I said, but... I dunno, maybe?

Most importantly, though, a change this big would be a huge disruption to anyone literate in the old system anyway, so I guess what letters you use for the vowels wouldn't really make that much of a difference. So if you prefer "Romance" values, sure, why not? As I said previously, there's nothing wrong with that as a matter of personal esthetics. Trying to sell the change as offering an advantage over the current system in terms of actual usefulness just isn't cutting it, though.

What would you prefer, then, Ossicone? Hopefully not the traditional <ng>, since that would be ambiguous.

It's not, for the same reason that <th> isn't in adulthood. Also, even if it is to some people, the contrast is so marginal that it hardly warrants introducing a whole new letter (and with an unusual value for said letter, to boot).

See above. Also, how is /θ/ an unusual value for <þ>?

/ŋ/ is an unusual value for <ñ>. While most languages don't use <ñ> for anything, quite a lot of people know it due to Spanish being (probably) the second-most widely spoken language on the planet, and that it stands for what most of us would approximate with something like [nj].

[Xing]

As to the predictability of /ɜː(ɹ)/, I have two reasons for its treatment in RLS. First, I took my cues from what seems to be the prevailing phonological analysis of broadcast English, which maintains it as a separate phoneme. Secondly, it supports the dialectal flexibility I mixed into the system. In GA, according to the Cambridge Online Dictionary (British edition, which lists both RP and GA pronunciations), the formerly rhotic diphthong /ɛə˞/ tends to collapse into just /ɛɹ/ in American English, thus making words like "fur" and "fair" minimal pairs (RLS <fér> versus <fer> for GA, <fe> versus <fèø> for RP, and <fér> versus <fèør> for the neutral hybrid that I've proposed). There are also a few words in GA, such as "tour," which Cambridge transcribes as /tʊɹ/, where /ʊɹ/ also appears distinctly from /ɜ˞/. With that contrastivity in mind, the free value of <e> seemed like the overall best slot to host /ɜː/, completing the checked/free symmetry of the five conventional vowel letters which I suspect will be mnemonically helpful. Given that it's also not entirely without precedent in traditional spelling, as in "herd" or jerk," it seemed like a pretty solid choice.

/ɜː(r)/ is a result of historic /ɪr ɛr ʊr/ in two environments: (1) before consonants, and (2) word-finally. /ɪr ɛr ʊr/ is retained before vowels, as in <mirror>, <merry> and <marry>. The only (?) cases where /ɜː/ contrasts with /ɪr ɛr ɜr/ is when a vowel suffix is added after /ɜː(r)/, as in <furry>.

The vowel in <tour> does not derive from historic /ʊr/, but from historic /uːr/. The conservative pronunciation is something like [tʰʊə̯(ɹ)] (fc <poor>) and changes to other pronunciations are later developments.

Mary = <Mèørri> for IHS /mɛə˞(ɹ)iː/, <Mèri> for GA /mɛɹiː/, <Mèøri> for RP /mɛəɹiː/

merry = <mèri> for /mɛɹiː/ (all three accents agree on this one)

marry = <mæri> for IHS & RP /mæɹiː/, <mèri> for GA /mɛɹiː/

furry = <férri> for IHS & GA /fɜ˞(ɹ)iː/, <feri> for RP /fɜːɹiː/

hurry = <hàri> for IHS & RP /hʌɹiː/, <hérri> for GA /hɜ˞(ɹ)iː/

ferry = <fèri> for /fɛɹiː/ (all three accents agree on this one)

sorry = <sòri> for IHS & RP /sɒɹiː/, <sarri> for GA /sɑɹiː/

lorry = <lorri> for IHS & GA /lɔɹiː/, <lòri> RP /lɒɹiː/

story = <storri> for IHS & GA /stɔɹiː/, <stori> RP /stɔːɹiː/

It seems like you are messing up things that are perfectly regular in the current orthography. If we look at the words <mirror merry marry sorry hurry>, we see that they have the regular checked vowels - /ɪ ɛ æ ɒ ʌ/, just that they are pronounced before an intervocalic /r/. The checked nature of the vowels is indicated by the doubling of <r> - a regular way of indicating checked vowels in English orthography. Now you propose that we keep abandon it in <merry marry sorry hurry> (but for some reason keep it on <lorry>...)

Using <æ> for /ʌ/ would not only re-introduce what is to most a new letter, it would also conflict with its own original historical purpose. That's two strikes against it, whereas using it for /æ/ instead means only one strike. This seems especially apparent if we continue to presume that <æ> would be added to the alphabet no matter which phoneme it ends up representing.

So, you are basically saying that we should introduce a new spelling that would constitute a major break with established English spelling, and make English less similar to continental European languages, in order to make it more similar to Old English???

[ol bofosh]

The vowel in <tour> does not derive from historic /ʊr/, but from historic /uːr/. The conservative pronunciation is something like [tʰʊə̯(ɹ)] (fc <poor>) and changes to other pronunciations are later developments.

You mean sure and shore aren't homophones?

[Ithisa]

The vowel in <tour> does not derive from historic /ʊr/, but from historic /uːr/. The conservative pronunciation is something like [tʰʊə̯(ɹ)] (fc <poor>) and changes to other pronunciations are later developments.

You mean sure and shore aren't homophones?

Not in my dialect! I pronounce "sure" as [ʃɝ:], as in "curse" and "mirth", and "shore" as [ʃɔ˞:], as in "more".

[Glossaphile]

You can hear a phoneme. You can't hear a morpheme boundary.

I never said anything about "morpheme boundaries" - or "phonemes", for that matter. Laypeople tend not to know what those terms mean - but that doesn't mean they often aren't intuitively aware of this stuff. In fact, understanding the morphemes and their combination rules in a language is kind of necessary in order to speak it productively.

Your arguments against distinguishing /θ/ from /th/ implicitly invoked morpheme boundaries. The knowledge that "adulthood" is composed of "adult" and "-hood" relies on subconscious awareness of where one morpheme ends and another begins, though as you rightly point out, a laymen wouldn't put it in those terms.

Besides, even if native speakers are somehow more conscious of phonemes than they are of morphemes, having the same spelling for /T/ and /th/ would be no problem for them (especially considering the rarity of the latter); they would know the proper pronunciation regardless of spelling. And non-native learners need to learn about morpheme boundaries anyway.

Non-native students, at least in the beginning, are much less likely to have the same intuitions about the rarity of certain sound combinations as opposed to others. Even native children may not have them to the extent that you seem to presume.

School-age children are quite capable of learning - or even figuring out by themselves - that there's a relationship between words like adulthood and adult. If anything, that would seem to me to be more "concrete" a concept than the difference between /T/ and /th/. And why would grade-school children making a few spelling errors be such a huge problem anyway? They're going to do that no matter how good your orthography is. Which is one reason why I wouldn't go out of my way to try to make it toddler-proof.

While it is true that phonemes are an abstraction, morphology is generally more abstract still. Phonemes can almost always be "heard" in the sense of being easily distinguished by native speakers. The ease with which morphemes can be parsed, however, seems to show much more variation.

Let's take an example that's a little harder to parse into roots and affixes, using the voiced counterpart of /θ/: "adherent." I suspect that one would likely need to have some knowledge of Latin, something that not even most adults enjoy, in order to know that the "ad-" is actually a prefix, though a particularly clever reader might at least have better luck identifying the "-ent" suffix. I don't think the average Joe should have to study any Latin before he can know not to pronounce it /ˈæðə˞(ɹ)ənt/ or something with /ð/.

Even allowing for some latent consciousness of "ad-" as an unproductive bound morpheme, only sufficiently experienced readers will probably have that, but much of the point of spelling reform would be to accelerate mastery for beginners, not make things only slightly easier for those who are already literate anyway.

I wonder what's unique about <c> that makes it harder to re-learn for some than <z>, which that girl had apparently mastered with its non-English affricate pronunciation.

Um. Said girl's native language wasn't English - nor is /ts/ even the proper pronunciation for <z> in Hungarian! That would be just plain /z/, in fact. And her native language was (presumably) Finnish, where <z> isn't properly a part of the orthography at all, but its name when reciting the alphabet is pronounced with a /ts/ (and /z/ doesn't exist).

I was under the impression, since the comment was about students of eastern European languages, that she was a native Anglophone studying Hungarian. I take it the incorrect /ts/ is interference from her native tongue, then? The fact that, as you say Slavic languages tend to pronounce <c> as an affricate more frequently than others could be, in a way, another argument in favor of its RLS assignment. I have the vowels/diphthongs of Romance, a few letters and their corresponding sounds from old Germanic (thorn, eth, and ash), and a touch of Slavic. If such diverse features can help broaden the system's appeal, most of Europe at least just might be a little easier to persuade.

The problem with [<š> for /ʃ/] is that it's not available on the US-International keyboard layout, to which I relegate myself as a sort of compromise for skeptics worried about easy typability. This keyboard layout can be activated within minutes on any Windows PC without any new hardware or software. If reform were ever implemented, it probably wouldn't matter that much in the long run, as manufacturers would probably figure out there's money to be made in upgrading their physical keyboards. But in the initial stages, I think such an appeasing concession to conservatives could be crucial in getting the ball rolling.

So essentially, you're making the spelling less recognizable so it'd be more acceptable to conservatives? Somehow I don't see that working.

It's effectiveness is dependent on the prioritization of easy typability versus familiarity among the general public, and I don't think there are any firm indications yet one way or the other.

On the other forum, the point was raised that it's relatively easy to learn new sounds for old letters, so if you accept that, then continental vowels/diphthongs shouldn't seem nearly as disruptive.

I'd say that's conditional: It's relatively easy to learn new sounds for old letters if 1) you're learning a new language and 2) the new sounds are at least somewhat intuitive. Again, I've met numerous people who can't seem to wrap their head around the fact that <c> isn't always /k/ before back vowels (or that <z> isn't /ts/ in every language that isn't English).

Nevertheless, the consensus seems to be that the coping potential is there. The challenge, albeit still daunting, is not necessarily minimizing the need for it but tapping into it under circumstances different from those that usually bring it out. It's a motivational thing, I think.

Plus, English speakers are not complete strangers to these more Latinate correspondences. They may not be the dominant ones, but they do occur in English, and we don't necessarily have to look at really obscure terminology to find them. Consider "father," "herd," "machine," "or," and "rude" for examples where the vowels are pronounced as they would be in RLS (assuming an RP accent). Most other Roman-script languages, on the other hand, have absolutely no words that uses English-like vowels (except for obvious borrowings). So really, the native English speakers are the ones making the lesser adjustment.

You seem to be comparing native speakers already literate in the old orthography of English to non-natives just beginning to learn it... But ignoring non-natives already literate in English and native speakers just beginning to learn to spell.

It seems to me that advanced non-native speakers would probably approximate the literate natives in their concerns, while what's most relevant to native children would be more akin to the issues faced by beginning non-natives (though the difference in that latter pair may be somewhat greater than the former, depending on age). To me, beginning non-natives and highly literate natives are the two extremes of the spectrum, so if I consider them, I'll have implicitly addressed those somewhere in the middle as well.

Of these, the only group this could possibly benefit are the non-native beginners; to native beginners not yet literate in any language it doesn't matter what letters you use, and to people who'd need to relearn literacy in English this whole thing would just be a huge disruption. Probably most huge to the native speakers, though, since they're the ones most likely to need English in their daily lives. That's not such a small adjustment, in my opinion.

Native Anglophone children would also benefit from it. Nativity alone cannot compensate for the retardation in literacy acquisition caused by the current orthography. It can help, I'm sure, but I doubt it will cancel it out entirely. As for the already literate, given that previous generations have gone to war and died for the sake of future generations, I don't think it's asking too much to learn a simplified orthography, especially if the new regularity is such that it can be learned very quickly (ranging from a few hours of independent study to a week's seminar at most). The burden of researching precisely what the future benefits will be and proving their worth, however, does admittedly rest on the shoulders of the reformers themselves.

I experimented with teaching RLS to a 13-year-old cousin of mine (bright enough and old enough to partially represent both children and literate adults), and while we ran out of time to complete the instruction before she had to return to her home in another state, she seemed to be getting the hang of it after about four non-consecutive hours. She was by no means perfect, but I estimate that another four hours or so would have at least given her the skills to be functional in a hypothetical Anglophone world that used RLS. It was the informal prelude to a more rigorous experiment I hope to carry out on my university campus, so I didn't really document it, but even if you remain skeptical of its validity, perhaps you can at least see where some of my conviction comes from.

/ŋ/ is an unusual value for <ñ>. While most languages don't use <ñ> for anything, quite a lot of people know it due to Spanish being (probably) the second-most widely spoken language on the planet, and that it stands for what most of us would approximate with something like [nj].

True, the velar nasal doesn't exactly match up to the original Spanish usage, but I suspect it's close enough that it's not quite as big of a leap as you seem to think.

/ɜː(r)/ is a result of historic /ɪr ɛr ʊr/ in two environments: (1) before consonants, and (2) word-finally. /ɪr ɛr ʊr/ is retained before vowels, as in <mirror>, <merry> and <marry>. The only (?) cases where /ɜː/ contrasts with /ɪr ɛr ɜr/ is when a vowel suffix is added after /ɜː(r)/, as in <furry>.

For me, even a comparatively rare contrast still counts.

The vowel in <tour> does not derive from historic /ʊr/, but from historic /uːr/. The conservative pronunciation is something like [tʰʊə̯(ɹ)] (fc <poor>) and changes to other pronunciations are later developments.

That's interesting in its own right, but I'm not sure how all this talk of historic derivations is relevant. The ultimate point is what contrasts exist in modern English, with little or no reference to how they got there (or didn't). The changes themselves are not as important as the results that have come down to us.

Incidentally, the suggested international standard pronunciation for "tour" is /tʊə˞/, which would become <tùør> in RLS.

Mary = <Mèørri> for IHS /mɛə˞(ɹ)iː/, <Mèri> for GA /mɛɹiː/, <Mèøri> for RP /mɛəɹiː/

merry = <mèri> for /mɛɹiː/ (all three accents agree on this one)

marry = <mæri> for IHS & RP /mæɹiː/, <mèri> for GA /mɛɹiː/

furry = <férri> for IHS & GA /fɜ˞(ɹ)iː/, <feri> for RP /fɜːɹiː/

hurry = <hàri> for IHS & RP /hʌɹiː/, <hérri> for GA /hɜ˞(ɹ)iː/

ferry = <fèri> for /fɛɹiː/ (all three accents agree on this one)

sorry = <sòri> for IHS & RP /sɒɹiː/, <sarri> for GA /sɑɹiː/

lorry = <lorri> for IHS & GA /lɔɹiː/, <lòri> RP /lɒɹiː/

story = <storri> for IHS & GA /stɔɹiː/, <stori> RP /stɔːɹiː/

It seems like you are messing up things that are perfectly regular in the current orthography. If we look at the words <mirror merry marry sorry hurry>, we see that they have the regular checked vowels - /ɪ ɛ æ ɒ ʌ/, just that they are pronounced before an intervocalic /r/. The checked nature of the vowels is indicated by the doubling of <r> - a regular way of indicating checked vowels in English orthography. Now you propose that we keep abandon it in <merry marry sorry hurry> (but for some reason keep it on <lorry>...)

The apparent irregularities are really part of the compromise in accent rather than any inconsistencies in sound-symbol correspondence. For example, Americans have pulled a rather funky stunt with original /ɒɹV/ sequences (there may be a predictive rule, but it probably refers to criteria too jargonic for laymen). In some cases, as in "sorry," it went the way of the father-bother merger and became /sɑɹiː/. In others, as in "lorry," it became essentially a rhotacized /ɔ(ː)/, hence /lɔɹiː/ or, if you prefer, /lɔ˞(ɹ)iː/ (at least, that's how it would be pronounced if Americans actually used this word at all). Seeking to be as fair as possible, I tentatively decided to side with the Americans on the "lorry"-group but with the British on the "sorry"-group (this is also consistent with my overall retention of /ɒ/ as separate from /ɑ(ː)/).

I don't deny that standard RLS wouldn't have occasional discrepancies when compared to any one natural dialect, but the goal was to spread them out as evenly and thinly as possible across the broadest proportion of the Anglosphere. In any case, a few regional divergences is still infinitely better than the multitude of universal infidelities that we live with now.

In RLS, except for a comparative few cases induced by prefixation, 'r' is the only consonant that is ever doubled intervocalically, and that's to mark rhotacization of the preceding vowel. Grave accents take the place of traditional consonant doublings, much like acute accents could roughly be said to take the place of "magic E."

Using <æ> for /ʌ/ would not only re-introduce what is to most a new letter, it would also conflict with its own original historical purpose. That's two strikes against it, whereas using it for /æ/ instead means only one strike. This seems especially apparent if we continue to presume that <æ> would be added to the alphabet no matter which phoneme it ends up representing.

So, you are basically saying that we should introduce a new spelling that would constitute a major break with established English spelling, and make English less similar to continental European languages, in order to make it more similar to Old English???

[/quote]

I'm pretty sure that Old English was more similar to continental European languages anyway, both past and present, so I'm not really seeing the conflict there? The two paths go in the same direction, at least approximately.

[Xing]

The apparent irregularities are really part of the compromise in accent rather than any inconsistencies in sound-symbol correspondence. For example, Americans have pulled a rather funky stunt with original /ɒɹV/ sequences (there may be a predictive rule, but it probably refers to criteria too jargonic for laymen). In some cases, as in "sorry," it went the way of the father-bother merger and became /sɑɹiː/. In others, as in "lorry," it became essentially a rhotacized /ɔ(ː)/, hence /lɔɹiː/ or, if you prefer, /lɔ˞(ɹ)iː/ (at least, that's how it would be pronounced if Americans actually used this word at all). Seeking to be as fair as possible, I tentatively decided to side with the Americans on the "lorry"-group but with the British on the "sorry"-group (this is also consistent with my overall retention of /ɒ/ as separate from /ɑ(ː)/).

Your approach when it comes to prerhotic vowels is inconsistent with your general rule for checkedness - according to which a vowel becomes checked when followed by two intervocallic consonants. Why should this rule not apply to vowels when followed by /r/?

I'm pretty sure that Old English was more similar to continental European languages anyway, both past and present, so I'm not really seeing the conflict there? The two paths go in the same direction, at least approximately.

Your reform makes English less similar to Romance and Continental languages first by introducing a letter that does not exist in any of then. Then by respelling lots of words with <a>, which would make both old cognates and recent loanwords containing <a> less recognisable. Further, in many English dialects, there is a tendency to lower /æ/, so that it becomes more similar to the Romance /ä/. Yet further, since your respelling includes words in the bath-set, it would force many English speakers to spell /A:/ with <æ>, which would be rather counterintuitive.

In general, I think your spelling reforms lacks a sense of proportion. You seem to be very concerned with minor discrepancies that hardly pose any real problems for anyone (cf the discussion on ng and adulthood above), while having no problems with drastic breaks with the orthographic tradition (that for lots of speakers don't make things more consistent, and in some cases less consistent).

[Xonen]

Besides, even if native speakers are somehow more conscious of phonemes than they are of morphemes, having the same spelling for /T/ and /th/ would be no problem for them (especially considering the rarity of the latter); they would know the proper pronunciation regardless of spelling. And non-native learners need to learn about morpheme boundaries anyway.

Non-native students, at least in the beginning, are much less likely to have the same intuitions about the rarity of certain sound combinations as opposed to others. Even native children may not have them to the extent that you seem to presume.

I'm not presuming that people have any kind of intuition about the rarity of certain sound combinations - if anything, I'm presuming that people tend not to be that consciously aware of distinctions between the different pronunciations of <th> at all! What I am presuming is that this doesn't much matter: /th/ is such a rare combination of sounds in English that you can pretty much ignore it completely when teaching people how to read. If this results in some clueless foreigner then pronouncing adulthood with a /T/ (which, IME, quite a few foreigners tend to approximate as [tʰ] anyway), then that's a sacrifice I would be willing to make for the sake of not introducing unnecessary extra letters.

Let's take an example that's a little harder to parse into roots and affixes, using the voiced counterpart of /θ/: "adherent." I suspect that one would likely need to have some knowledge of Latin, something that not even most adults enjoy, in order to know that the "ad-" is actually a prefix, though a particularly clever reader might at least have better luck identifying the "-ent" suffix. I don't think the average Joe should have to study any Latin before he can know not to pronounce it /ˈæðə˞(ɹ)ənt/ or something with /ð/.

You're right in that the morphemes are not quite as transparently obvious here. However, why would anyone be tempted to pronounce <dh> as /ð/ in English?

I wonder what's unique about <c> that makes it harder to re-learn for some than <z>, which that girl had apparently mastered with its non-English affricate pronunciation.

Um. Said girl's native language wasn't English - nor is /ts/ even the proper pronunciation for <z> in Hungarian! That would be just plain /z/, in fact. And her native language was (presumably) Finnish, where <z> isn't properly a part of the orthography at all, but its name when reciting the alphabet is pronounced with a /ts/ (and /z/ doesn't exist).

I was under the impression, since the comment was about students of eastern European languages, that she was a native Anglophone studying Hungarian.

That's... quite a peculiar assumption to base on one little preposition (unless the assumption is that non-Anglophones somehow can't be students of said languages...), but okay then. It's usually pretty safe to assume that yours truly's personal experiences come from Finland, unless otherwise specified.

I take it the incorrect /ts/ is interference from her native tongue, then? The fact that, as you say Slavic languages tend to pronounce <c> as an affricate more frequently than others could be, in a way, another argument in favor of its RLS assignment.

Or not. As I was trying to point out, this pronunciation is pretty unintuitive to many of those of us who don't speak "Slavic" languages (like Hungarian? ), which might, you know, be an argument against this kind of reassignment.

I have the vowels/diphthongs of Romance, a few letters and their corresponding sounds from old Germanic (thorn, eth, and ash), and a touch of Slavic. If such diverse features can help broaden the system's appeal, most of Europe at least just might be a little easier to persuade.

Quite frankly, most of Europe probably doesn't give a rat's ass about old Germanic. However, it is familiar with the current spelling of English, and hence that's the orthography you should keep as your starting point if minimizing the difficulty of persuading people is your goal. A complicated mess of different traditions with a bit of everything thrown in and yet no perfect match with anything could easily end up being the kind of compromise that doesn't actually satisfy anyone.

The problem with [<š> for /ʃ/] is that it's not available on the US-International keyboard layout, to which I relegate myself as a sort of compromise for skeptics worried about easy typability. This keyboard layout can be activated within minutes on any Windows PC without any new hardware or software. If reform were ever implemented, it probably wouldn't matter that much in the long run, as manufacturers would probably figure out there's money to be made in upgrading their physical keyboards. But in the initial stages, I think such an appeasing concession to conservatives could be crucial in getting the ball rolling.

So essentially, you're making the spelling less recognizable so it'd be more acceptable to conservatives? Somehow I don't see that working.

It's effectiveness is dependent on the prioritization of easy typability versus familiarity among the general public, and I don't think there are any firm indications yet one way or the other.

I'll admit I haven't done any studies on this, but somehow I suspect most people would be less annoyed by the need to update their keyboard layouts than by having to learn a whole new unfamiliar spelling for everything.

[Glossaphile]

Non-native students, at least in the beginning, are much less likely to have the same intuitions about the rarity of certain sound combinations as opposed to others. Even native children may not have them to the extent that you seem to presume.

I'm not presuming that people have any kind of intuition about the rarity of certain sound combinations - if anything, I'm presuming that people tend not to be that consciously aware of distinctions between the different pronunciations of <th> at all! What I am presuming is that this doesn't much matter: /th/ is such a rare combination of sounds in English that you can pretty much ignore it completely when teaching people how to read. If this results in some clueless foreigner then pronouncing adulthood with a /T/ (which, IME, quite a few foreigners tend to approximate as [tʰ] anyway), then that's a sacrifice I would be willing to make for the sake of not introducing unnecessary extra letters.

I suppose we'll just have to chalk it up to differences in priorities, then. One of my most important design principles from early on was that no single unique spelling should have more than one possible pronunciation (as allowed by the system's stated rules). The converse thereof was almost equally important. No single unique sequence of phonemes should have more than one possible ways to spell it.

I tend to believe that, at its heart, alphabetic writing is almost purely a cipher for pronunciation. Any derived meaning comes from the encoded sounds, not from the encoding itself. Of course, advanced readers learn to skip the phonemic middleman, but I think it significantly aids learning for children and non-natives if that's at least how it all starts out.

Let's take an example that's a little harder to parse into roots and affixes, using the voiced counterpart of /θ/: "adherent." I suspect that one would likely need to have some knowledge of Latin, something that not even most adults enjoy, in order to know that the "ad-" is actually a prefix, though a particularly clever reader might at least have better luck identifying the "-ent" suffix. I don't think the average Joe should have to study any Latin before he can know not to pronounce it /ˈæðə˞(ɹ)ənt/ or something with /ð/.

You're right in that the morphemes are not quite as transparently obvious here. However, why would anyone be tempted to pronounce <dh> as /ð/ in English?

If we go the more digraph-heavy route, which you seem to prefer, the <dh> would be the most logical choice for /ð/ opposite <th> for /θ/ (much like <zh> would be for /ʒ/).

I was under the impression, since the comment was about students of eastern European languages, that she was a native Anglophone studying Hungarian.

That's... quite a peculiar assumption to base on one little preposition (unless the assumption is that non-Anglophones somehow can't be students of said languages...), but okay then. It's usually pretty safe to assume that yours truly's personal experiences come from Finland, unless otherwise specified.

Well, the original message didn't specify a native language for the student, and since the topic is overwhelmingly English, I didn't think there'd be any harm in making that assumption, though it did briefly occur to me that it could be something else. Perhaps I should have asked first. My apologies if I seemed presumptuous at all. Now that I know you deal alot with Finnish, I'll be better advised.

I take it the incorrect /ts/ is interference from her native tongue, then? The fact that, as you say Slavic languages tend to pronounce <c> as an affricate more frequently than others could be, in a way, another argument in favor of its RLS assignment.

Or not. As I was trying to point out, this pronunciation is pretty unintuitive to many of those of us who don't speak "Slavic" languages (like Hungarian? ), which might, you know, be an argument against this kind of reassignment.

Fair enough.

I have the vowels/diphthongs of Romance, a few letters and their corresponding sounds from old Germanic (thorn, eth, and ash), and a touch of Slavic. If such diverse features can help broaden the system's appeal, most of Europe at least just might be a little easier to persuade.

Quite frankly, most of Europe probably doesn't give a rat's ass about old Germanic. However, it is familiar with the current spelling of English, and hence that's the orthography you should keep as your starting point if minimizing the difficulty of persuading people is your goal. A complicated mess of different traditions with a bit of everything thrown in and yet no perfect match with anything could easily end up being the kind of compromise that doesn't actually satisfy anyone.

True, diersity can lead to a breakdown in cohesiveness (and therefore appeal) if not applied carefully. This is no small part of what went wrong with traditional orthography. All those borrowings from all those sources over the centuries, Anglicized in pronunciation but never in spelling. Combine that with the more general archaism of it, and no wonder it's a messǃ

So essentially, you're making the spelling less recognizable so it'd be more acceptable to conservatives? Somehow I don't see that working.

It's effectiveness is dependent on the prioritization of easy typability versus familiarity among the general public, and I don't think there are any firm indications yet one way or the other.

I'll admit I haven't done any studies on this, but somehow I suspect most people would be less annoyed by the need to update their keyboard layouts than by having to learn a whole new unfamiliar spelling for everything.

All the more reason why more research is needed. I've made my own humble first attempts with this poll: http://spellingreform.net/forums/viewto ... f=15&t=223. Sadly, far too few people have responded for it to be at all scientific, though the current meager results do mildly support your suspicions.

The apparent irregularities are really part of the compromise in accent rather than any inconsistencies in sound-symbol correspondence. For example, Americans have pulled a rather funky stunt with original /ɒɹV/ sequences (there may be a predictive rule, but it probably refers to criteria too jargonic for laymen). In some cases, as in "sorry," it went the way of the father-bother merger and became /sɑɹiː/. In others, as in "lorry," it became essentially a rhotacized /ɔ(ː)/, hence /lɔɹiː/ or, if you prefer, /lɔ˞(ɹ)iː/ (at least, that's how it would be pronounced if Americans actually used this word at all). Seeking to be as fair as possible, I tentatively decided to side with the Americans on the "lorry"-group but with the British on the "sorry"-group (this is also consistent with my overall retention of /ɒ/ as separate from /ɑ(ː)/).

Your approach when it comes to prerhotic vowels is inconsistent with your general rule for checkedness - according to which a vowel becomes checked when followed by two intervocallic consonants. Why should this rule not apply to vowels when followed by /r/?

The rhotic combinations /ɑɹ/ or /ɑ˞/ and /ɔɹ/ or /ɔ˞/ are treated mostly as holistic units of their own, just like the other two, /ɜ˞/ and /ə˞/. This is due, at least in part, to a broad distributional rule that allows the reduction of diacritic usage. If they were subject to the normal checked/free rules, words like "start" and "fort" would require an acute accent, since the vowels would otherwise be pronounced /ʌ/ and /ɒ/. But non-pre-vocalic /ʌɹ/ and /ɒɹ/ never occur in English, so the acutes would be redundant. It's similar reasoning to that which explains why the tilde disappears from <ñ> before /k/ or /g/.

Furthermore, these apparent exceptions are redeemed by forming a broad pattern of their own. Double-r is also used intervocalically after diphthongs to signal the epenthetic rhotic schwa that's automatic in all other environments. This insertion is included in the definition of "rhotacization" for the purposes of explaining the system, and a sequence like /aɪə˞/ is referred to as a "rhotic" or "rhotacized" diphthong. It's virtually impossible for a rhotic speaker to genuinely say something like /faɪɹ/, even though that's exactly what the spelling "fire" implies. It inevitably comes out as /faɪə˞/, with such predictability that it becomes redundant to write <fayør> when <fair> would do just as well. This predictability, however, lapses when there's another vowel immediately after, which is why we have two <r>s in <fairri> (/faɪə˞(ɹ)iː/; "fiery") but only one in <pairøt> (/paɪɹət/; "pirate"). R-doubling then consistently signals rhotacization of whatever precedes it.

I'm pretty sure that Old English was more similar to continental European languages anyway, both past and present, so I'm not really seeing the conflict there? The two paths go in the same direction, at least approximately.

Your reform makes English less similar to Romance and Continental languages first by introducing a letter that does not exist in any of then. Then by respelling lots of words with <a>, which would make both old cognates and recent loanwords containing <a> less recognisable. Further, in many English dialects, there is a tendency to lower /æ/, so that it becomes more similar to the Romance /ä/. Yet further, since your respelling includes words in the bath-set, it would force many English speakers to spell /A:/ with <æ>, which would be rather counterintuitive.[/quote]

When I advocate for greater affinity with continental European languages, I mean affinity with how the sounds map onto the symbols, not to how words ultimately look on the page. For example, <neiçøn> may not look much like Spanish <nación> or Italian <nazione>, but at least the pronunciation is less surprising from a Romanophone beginner's standpoint than it would be if it were spelled <nation> or even something like <naishun>. The tonic diphthong is very intuitively spelled, as is the /n/, and the two letters that probably will take some getting used to (<-çø->are at least perfectly consistent in how they sound. In fact, that just might be a good summary of RLS as a system from a continental European perspective. Almost everything is either already intuitive or so unvarying that the learning effort is likely to be minimal.

In general, I think your spelling reforms lacks a sense of proportion. You seem to be very concerned with minor discrepancies that hardly pose any real problems for anyone (cf the discussion on ng and adulthood above), while having no problems with drastic breaks with the orthographic tradition (that for lots of speakers don't make things more consistent, and in some cases less consistent).

The way I see it, because the code from which we would be divorcing ourselves is so irregular to begin with, any replacement which was regular enough to reap worthwhile gains would inevitably be a "drastic break," so I saw little point in placing proximity to the traditional system very high at all on the priority list. If you try too hard to maintain visual similarity to the current spelling, you're still going to end up with so many rules, sub-rules, and corollaries that I think many would ask if anything palpable was even accomplished.

[Xing]

The rhotic combinations /ɑɹ/ or /ɑ˞/ and /ɔɹ/ or /ɔ˞/ are treated mostly as holistic units of their own, just like the other two, /ɜ˞/ and /ə˞/. This is due, at least in part, to a broad distributional rule that allows the reduction of diacritic usage. If they were subject to the normal checked/free rules, words like "start" and "fort" would require an acute accent, since the vowels would otherwise be pronounced /ʌ/ and /ɒ/. But non-pre-vocalic /ʌɹ/ and /ɒɹ/ never occur in English, so the acutes would be redundant.

The vowels in start and fort - as well as bird are basically the "broad" versions of the traditional short or "checked" vowels. The former are almost in complete complementary distribution with the latter. Basically, the "broad" versions occur whenever a checked vowel plus an <r> is followed by another consonant, or appears word-finally, while the "traditional" realisation of the checked vowel occur when the <r> is doubled. The only exceptions that I can think of arise when a vowel-initial suffix is added to a word-final broad vowel - as is starry and furry. And would be much simple to mark the few exceptions - if that was ever deemed necessary - than to redo the whole regular system in order to accommodate for the few exceptions.

When I advocate for greater affinity with continental European languages, I mean affinity with how the sounds map onto the symbols, not to how words ultimately look on the page. For example, <neiçøn> may not look much like Spanish <nación> or Italian <nazione>, but at least the pronunciation is less surprising from a Romanophone beginner's standpoint than it would be if it were spelled <nation> or even something like <naishun>. The tonic diphthong is very intuitively spelled, as is the /n/, and the two letters that probably will take some getting used to (<-çø->are at least perfectly consistent in how they sound. In fact, that just might be a good summary of RLS as a system from a continental European perspective. Almost everything is either already intuitive or so unvarying that the learning effort is likely to be minimal.

Some English letters are pronounced very differently from their Continental counterparts. But not all - especially not the checked/lax/short vowels. In the case of /æ/, to change that spelling would at most benefit some accents at the expense of others. Say, if you have a strong northern inland accent, your /æ/ might be closer to [ɪə] or [eə] or something like that - in which case it make little sense to identify /æ/ with the Continental /ä/. But in many other dialects, /æ/ is a much more open vowel - in which case such identification would be quite natural.

The way I see it, because the code from which we would be divorcing ourselves is so irregular to begin with, any replacement which was regular enough to reap worthwhile gains would inevitably be a "drastic break," so I saw little point in placing proximity to the traditional system very high at all on the priority list. If you try too hard to maintain visual similarity to the current spelling, you're still going to end up with so many rules, sub-rules, and corollaries that I think many would ask if anything palpable was even accomplished.

There are in fact many regularities in English spelling. And it would be much easier to change the exceptions and perhaps the more narrow sub-rules, that to change the big patterns of the orthography. The reasons for you drastic changes seems to be:

(1) A wish to make English vowel letters more similar to the Continental ones.
(2) A wish to distinguish monophthongs from diphthongs.
(3) A wish to avoid potential ambiguities in pronunciation - however marginal they may be.

Each of the above would have - at most - an marginal benefit. (1) Has been discussed quite a lot both here and on the ZBB. Regarding (2), one could wonder why you not favour a completely featural writing system? Regarding (3), it is important to bear in mind that few writing systems are perfect phonetic or phonological ciphers of the spoken langue. It's not uncommon for writing systems to omit things like word- or morpheme-boundaries, as well as pitch, intonation, and sometimes also stress and vowel length. (Many such things are often intuitively clear to people who are competent in the language, so that obligatory marking might appear as excessive pedantry...)

[Xing]

Just for fun, I'll try out a spelling reform adopting (more or less) Romance vowel values.

The basic thought is that many vowels can be thought of as short, long or broad. Long vowels are doubled. (or alternatively, vowels are long when they are followed by a single, prevocalic consonant.) Broad vowels are optionally marked with an accent (which is often left out, since broadening typically happens in certain semi-predictable environments). Word final vowels may take either the long or the broad value, depending on the vowel.

Short <a>: æ
Broad <a>: ɑː

Short <e>: ɛ
Long <e>: eɪ̯

Short <i>: ɪ
Long <i>: iː

Short <o>: ɒ
Broad <o>: ɔː
Long <o>: oʊ̯

Short <u>: ʊ, ʌ
Broad <u>: ɜː
Long <u>: uː

I make a few simplifications/regularisations when it comes to the consonants. The unstressed vowels are spelt <er>, <a> and <i>, respectively.

kit - kit
ship - ship
rip - rip
dip - dip
spirit - spirit

(Basically, no change)

dress - dres
step - step
ebb - eb
hem - hem
terror - terrer

trap - trap
bad - bad
cab - cab
ham - ham
arrow - arro

lot - lot
stop - lot
rob - rob
swan - swon

("Swon" has been regularised. It is quite regular with today's spelling - but this spelling requires one subrule fewer...)

strut - strut
rub - rub
cub - cub
hum - hum

foot - fut
full - ful
look - luk
could - cud

(The reference dialect has not undergone the foot-strut split!)

bath - bath, báth
staff - staf, stáf
clasp - clasp, clásp
dance - dans, dáns

(The reference dialect has not undergone the trap-bath split!)

cloth - cloth, clóth
cough - cof, cóf
long - long, lóng
laurel - lorrel, lórrel
origin - origin, órigin

(The reference dialect has not undergone the cloth-lot split!)

nurse - nurs, núrs
hurt - hurt, húrt
term - turm, túrm
work - wurk, wúrk

(/ɜː/ may be regarded as the broad version of either /ɪ/, /ɛ/ or /ʌ/.)

fleece - fliis
seed - siid
key - ki
seize - siiz

face - feec
weight - weet
rein - reen
steak - steek

palm - pam, pám
calm - cam, cám
bra - bra, brá
father - father, fáther

(/ɑ:/ is basically the "broad" version of /æ/.)

thought - thót
taut - tót
hawk - hók
broad - bród

goat - goot
soap - soop
soul - sool
home - hoom

goose - guus
who - whu
group - gruup
few - fyu

price - prais
ripe - raip
tribe - traib
aisle - ail
choir - kwair

choice - chois
boy - boy
void - void
coin - coin

mouth - mauth
pouch - pauch
noun - naun
crowd - craud
flower - flawer

near - niir
beer - biir
pier - piir
fierce - fiirs
serious - siirias

(Basically, long <i> followed by <r>)

square - squeer
care - keer
air - eer
wear - weer
Mary - Meri

(Basically, long <e> followed by <r>)

start - start, stárt
far - far, fár
sharp - sharp, shárp
farm - farm, fárm
safari - safari, safári

north - north, nórth
war - wor, wór
storm - storm, stórm
for - for, fór
aural - oral, óral

force - foors
floor - floor
coarse - coors
ore - oor
oral - ooral

(Basically, long <o> followed by <r>)

cure - kyuur
poor - puur
tour - tuur
fury - fyuuri

(Basically, long <u> followed by <r>)

(Some cure word have more or less "jumped" to the north/force set, and might at least optionally be spelt accordingly.)

happy - happi
silly - silli
Tony - Toni
merry - merri

letter - letter
beggar - begger
martyr - marter
visor - vaizer

comma - komma
China - China
sofa - sofa


Tha north wind and tha sun wur dispyuuting which wos tha stronger, when a travaller keem along, rapped in a worm clook. The agried that tha won whu furst succiided in meeking tha travaller teek hiz klook of, shud bi considdered stronger than tha uther. Then tha North Wind blu az hard az hi kud, but tha moor hi blu tha moor cloosly did the travaller foold hiz klook araund him. Then tha Sun shained aut wormli, and immiidieetly tha travaller tuuk of hiz clook. And tha North Wind woz oblaiged tu confes that tha Sun woz the stronger ov tha tu.

(This is something I made on the fly; so there might some errors.)

[Glossaphile]

The rhotic combinations /ɑɹ/ or /ɑ˞/ and /ɔɹ/ or /ɔ˞/ are treated mostly as holistic units of their own, just like the other two, /ɜ˞/ and /ə˞/. This is due, at least in part, to a broad distributional rule that allows the reduction of diacritic usage. If they were subject to the normal checked/free rules, words like "start" and "fort" would require an acute accent, since the vowels would otherwise be pronounced /ʌ/ and /ɒ/. But non-pre-vocalic /ʌɹ/ and /ɒɹ/ never occur in English, so the acutes would be redundant.

The vowels in start and fort - as well as bird are basically the "broad" versions of the traditional short or "checked" vowels. The former are almost in complete complementary distribution with the latter. Basically, the "broad" versions occur whenever a checked vowel plus an <r> is followed by another consonant, or appears word-finally, while the "traditional" realisation of the checked vowel occur when the <r> is doubled. The only exceptions that I can think of arise when a vowel-initial suffix is added to a word-final broad vowel - as is starry and furry. And would be much simple to mark the few exceptions - if that was ever deemed necessary - than to redo the whole regular system in order to accommodate for the few exceptions.

This is exactly my point. The only thing missing is that the long/free vowels never occur in "broad" positions either. So the checked/free contrast is, in effect, neutralized before any /ɹ/ which is not itself pre-vocalic. The merged archiphoneme could be said to take on the quality of the free vowel and the length of its checked counterpart.

/ɔ˞(ɹ)#/
/ɔ˞C/
/ɔ˞(ɹ)V/
/ɒɹV/

*/ɔːɹ./
*/ɒɹ#/
*/ɒɹC/

We seem to agree on the phonological analysis. What we disagree on is on how cases of intervocalic /ɹ/, where the checked/free contrast remains, should be handled. While you prefer to let the default sound be broad and double the <r> to indicate a checked pronunciation, I prefer to let the default sound be free, double the <r> to mark it as broad (RLS would refer to broadening as a manifestation of "rhotacization"), and add a grave accent to mark it as checked.

So for example,...

<so> = saw = /sɔː/
<sor> = sore/soar = /sɔ˞/ or /sɔɹ/
<sort> = sort = /sɔ˞t/ or /sɔɹt/
<sortiñ> = sorting = /sɔ˞tɪŋ/ or /sɔɹtɪŋ/
<sorriñ> = soaring(GA) = /sɔ˞(ɹ)ɪŋ/
<soriñ> = soaring(RP) = /sɔːɹɪŋ/
<sòri> = sorry = /sɒɹiː/
<sórt>= ???? = */sɔːɹt/

Granted, the contrast between an example like /sɔ˞(ɹ)ɪŋ/ and /sɔːɹɪŋ/ within the same dialect is questionable, but the fact that a hypothetical compound like "sawring" is phonotactically permissible and could potentially contrast with the actual word "soaring" in rhotic speech is the main reason behind my better-safe-than-sorry policy when it comes to that distinction. At the very least, it contributes to RLS' general capacity to represent both rhotic and non-rhotic accents, which is further exemplified with the contested use of <e> for /ɜː/.

<fe> = fur(RP) = /fɜː/
<fér> = fur(GA) = /fɜ˞/
<feri> = furry(RP) = /fɜːɹiː/
<férri> = furry(GA) = /fɜ˞(ɹ)iː/
<fèri> = ferry(neutral)/fairy(GA) = /fɛɹiː/
<fest> = fest = /fɛst/
<fést> = first(RP) = /fɜːst/
<férst> = first(GA) = /fɜ˞st/

Some English letters are pronounced very differently from their Continental counterparts. But not all - especially not the checked/lax/short vowels. In the case of /æ/, to change that spelling would at most benefit some accents at the expense of others. Say, if you have a strong northern inland accent, your /æ/ might be closer to [ɪə] or [eə] or something like that - in which case it make little sense to identify /æ/ with the Continental /ä/. But in many other dialects, /æ/ is a much more open vowel - in which case such identification would be quite natural.

You seem to have confused my goal of mild-to-moderate dialectal flexibility with that of always accommodating every single accent in the world. No practical orthography, no matter what the specifications, can possibly do the latter, at least not with a language as diverse and widely used as English. If someone in the northern inlands ever finds cause to explicitly write in his/her own accent, RLS offers them the choice of spelling the standard <pæt> as <pèøt> or <pìøt>. In most situations, however, if he/she hopes to communicate as broadly as possible, then he/she will spell according to the agreed-upon international standard, even if it does diverge from their own pronunciation. I do this all the time in my sample RLS writings, routinely distinguishing /ɒ/ from /ɑː/ even though my native US dialect merges them.

As I've said, some discrepancies with respect to local vernaculars are inevitable no matter what the principles and/or parameters of a proposed spelling system are. All I've endeavored to do is spread out those discrepancies as evenly and fairly as possible, with some focus on major players like RP and GA, so that those divergences are sufficiently infrequent within any single dialect and equally infrequent across a broad range of dialects.

Plus, part of the goal of reform (at least for me) is to provide a more reliable guide for non-natives to practice pronunciation, and it seems virtually unheard of for any foreigner to actively seek to emulate anything besides either GA or RP.

There are in fact many regularities in English spelling. And it would be much easier to change the exceptions and perhaps the more narrow sub-rules, that to change the big patterns of the orthography. The reasons for you drastic changes seems to be:

(1) A wish to make English vowel letters more similar to the Continental ones.
(2) A wish to distinguish monophthongs from diphthongs.
(3) A wish to avoid potential ambiguities in pronunciation - however marginal they may be.

Each of the above would have - at most - an marginal benefit. (1) Has been discussed quite a lot both here and on the ZBB. Regarding (2), one could wonder why you not favour a completely featural writing system? Regarding (3), it is important to bear in mind that few writing systems are perfect phonetic or phonological ciphers of the spoken langue. It's not uncommon for writing systems to omit things like word- or morpheme-boundaries, as well as pitch, intonation, and sometimes also stress and vowel length. (Many such things are often intuitively clear to people who are competent in the language, so that obligatory marking might appear as excessive pedantry...)

The benefits to be gleaned from an orthography based on the above principles are debatable, but not necessarily insignificant. I think the possibility of some measurable advantage is at least plausible enough to merit exploration until such time as research provides a more definitive conclusion. I'm not even convinced I'm really changing that much. Four of the five checked vowels are pronounced in ways that would be wholly unsurprising to native literates, and the assignments given to the free vowels, though clearly not as frequent, have some precedent in the traditional code as well. Some are even pervasive enough to influence the pronunciation of relatively unknown words. If I were to present a random Anglophone with my cat's name "Tiki," I strongly suspect he/she would spontaneously say /tiːkiː/ rather than /taɪkaɪ/. In essence, many English speakers do seem to have a certain latent sense of more universal vowel correspondences that we simply have to promote back to mental prominence. Not a cakewalk by any means, but nor is it the same as asking them to make associations which are genuinely and completely new. To find instances of that, you can only refer to the uses of <ç>, <ñ>, <x>, <q>, and checked <a>. Proportionally speaking, I'm not so sure the change is as great as it may superficially appear to be.

Also, who said anything about pitch, intonation, etc? The only intonation-marking in RLS would be incidental on the sentence level due to optional choices between strong and weak forms of function words such as <tu> versus <tø> for "to." Nothing explicit. Stress-marking exists in RLS but is completely optional, and pitch is completely ignored. I seek only to be strictly phonemic, not at all phonetic. Actually, a significant thrust in designing the system was to avoid the explicit marking of easily predictable alternations. Part of my treatment of rhotacized (or "broad") vowels, for instance, is to prevent the use of acute accents where phonotactics render them redundant.

My main critique of your experimental system would be that it locks itself in to foot-strut-merged dialects, which would be of very limited use outside regions that host such accents. You also seem to use a stress-based rule to deduce the pronunciation of final <a>, but that only works if you mark stress, which you don't. I'm also not particularly fond of the digraphs for reasons already discussed. Overall, though, it's far better than some other schemes I've seen.

I actually toyed around myself with classifying centering diphthongs like /ɪə/ and /ʊə/ as pre-codal-liquid realizations of /iː/ and /uː/, but then I found a few near-minimal pairs in Cambridge (such as "kneel" and "real," listed as /niːl/ and /ɹɪəl/) that pretty much disproved their predictability, at least according to criteria simple enough for the average reader/writer.

[Xonen]

Non-native students, at least in the beginning, are much less likely to have the same intuitions about the rarity of certain sound combinations as opposed to others. Even native children may not have them to the extent that you seem to presume.

I'm not presuming that people have any kind of intuition about the rarity of certain sound combinations - if anything, I'm presuming that people tend not to be that consciously aware of distinctions between the different pronunciations of <th> at all! What I am presuming is that this doesn't much matter: /th/ is such a rare combination of sounds in English that you can pretty much ignore it completely when teaching people how to read. If this results in some clueless foreigner then pronouncing adulthood with a /T/ (which, IME, quite a few foreigners tend to approximate as [tʰ] anyway), then that's a sacrifice I would be willing to make for the sake of not introducing unnecessary extra letters.

I suppose we'll just have to chalk it up to differences in priorities, then. One of my most important design principles from early on was that no single unique spelling should have more than one possible pronunciation (as allowed by the system's stated rules). The converse thereof was almost equally important. No single unique sequence of phonemes should have more than one possible ways to spell it.

That's fine in and of itself, but you must note that this clashes quite heavily with some of the other things you've claimed are also priorities, namely avoiding unconventional characters and, most importantly, trying to make the changes easy to sell to the general public. And my opinion is that the problem you're trying to fix here is ridiculously small compared to the problems the solution itself would cause in these other respects.

I tend to believe that, at its heart, alphabetic writing is almost purely a cipher for pronunciation. Any derived meaning comes from the encoded sounds, not from the encoding itself. Of course, advanced readers learn to skip the phonemic middleman, but I think it significantly aids learning for children and non-natives if that's at least how it all starts out.

Well, quite obviously, the encoding can and does convey information; that's why an English-speaker can immediately tell the difference between words like rain, rein and reign in writing even though they're identical in pronunciation. Of course, it could be argued that since we don't need those words to be distinguished in speech, we shouldn't need them to be distinguished in writing, either - but counterarguments to that could also be provided (and have been, even on these boards; this isn't the first time we're having this discussion). Writing and speech are ultimately quite different media of communication, so it can't just be taken for granted that the same principles apply to both.

But yes, for native-speaking children first learning to write it probably would make things easier, or at least quicker, if the spelling of a word could reliably be told just from knowing its pronunciation. In my mind, though, this doesn't override all other concerns in designing an orthography, and again, children are going to make spelling errors no matter how good the orthography is. It's not that dangerous.

Let's take an example that's a little harder to parse into roots and affixes, using the voiced counterpart of /θ/: "adherent." I suspect that one would likely need to have some knowledge of Latin, something that not even most adults enjoy, in order to know that the "ad-" is actually a prefix, though a particularly clever reader might at least have better luck identifying the "-ent" suffix. I don't think the average Joe should have to study any Latin before he can know not to pronounce it /ˈæðə˞(ɹ)ənt/ or something with /ð/.

You're right in that the morphemes are not quite as transparently obvious here. However, why would anyone be tempted to pronounce <dh> as /ð/ in English?

If we go the more digraph-heavy route, which you seem to prefer, the <dh> would be the most logical choice for /ð/ opposite <th> for /θ/ (much like <zh> would be for /ʒ/).

I don't prefer digraphs, I prefer the already existing digraphs over new unfamiliar spellings. There's a difference.

Now, in my own spelling reform idea (some examples of which you can find, say, here), I have, in fact, introduced a new symbol for /ð/, namely <ħ> (which has the advantage of looking like a ligature of <t> + <h> and hence doesn't change the way words look like too much). But even that's actually probably going a bit too far, IMO; it might be better to just accept that <th> has two main pronunciations. Native speakers know the proper pronunciation anyway, and those non-natives who actually 1) manage to produce interdental fricatives at all, 2) are perfectionist enough to care about the distinction and 3) yet somehow fail to have enough exposure to spoken English to instinctively learn when to use which could perhaps be expected to learn this much by heart. Especially considering that both sounds are rare and minimal pairs between them almost non-existent.

I was under the impression, since the comment was about students of eastern European languages, that she was a native Anglophone studying Hungarian.

That's... quite a peculiar assumption to base on one little preposition (unless the assumption is that non-Anglophones somehow can't be students of said languages...), but okay then. It's usually pretty safe to assume that yours truly's personal experiences come from Finland, unless otherwise specified.

Well, the original message didn't specify a native language for the student, and since the topic is overwhelmingly English, I didn't think there'd be any harm in making that assumption, though it did briefly occur to me that it could be something else. Perhaps I should have asked first. My apologies if I seemed presumptuous at all. Now that I know you deal alot with Finnish, I'll be better advised.

I don't just deal with Finnish, I am Finnish. My signature does sort of list Finnish as my first language, and I believe I did compare my experience with [unspecified people] learning eastern European languages to Finns learning English... Not that I think those hints should necessarily have been obvious enough to make my nationality clear - but in the absence of clear hints in any direction, I do think you shouldn't assume anything about someone's nationality on the net.

Why would this topic be overwhelmingly English, anyway? We're all linguists or at least interested in the subject here, and obviously we all know English, so why shouldn't English spelling be a topic of interest even (and perhaps especially) to those of us for whom it's not a native language?

[Xing]

As for /Vr/

The reason we should treat "broad" (pre-rhotic) vowels as special cases of short/checked vowels, or as special cases of long/free vowels, is that they historically evolved from long vowels, and that this is reflected in the current orthography. It's much better to keep the current system, as long as it's regular - and if it needs to be change, we should change those instances where it goes wrong - such as <starry> and <furry>.

The centring diphthongs /ɪə̯/ and /ʊə̯/ are basically the "long" vowels /iː/ and /uː/ as they occur before rhotics. Why not keep them so? /ɪə̯/ - contrasts (or at least, contrasted) with both /ɪr/ and with sequences of /iː/ followed by /e(r) - cf serious vs Sirius. The straightforward solution would be to spell /ɪə̯/ and /ʊə̯/ like regular /iːs/'s and /uː/'s - just that they are followed by <r>'s.

Although some dictionaries might list a word as <near> as pronounced [nɪə̯(ɹ)] (at least in the UK) I suppose that's a rather old-fashioned pronunciation. Most Britons, I suppose, would pronounce it either with a monophthong [nɪː(ɹ], or with two syllables: [niːə(ɹ)]. In the latter case, it would effectively merge with sequences of /iː/ followed by /ə(r)/ - so that idea would make a perfect rhyme with dear.


As for /æ/

I see no reason why we should not continue to spell /æ/ with <a>. One goal of your spelling reform is that the vowel letters should approach their "continental" value. The problem is that many English dialects have no straight equivalent to the continental /ä/. In some dialects, /ʌ/ is pronounced quite similar to continental /ä/. In other dialects, /ɑː/ is pronounced quite similar to continental /ä/. In yet others, /æ/ is pronounced quite similar to it. (In my variety of English, I'd suppose /æ/, /ɑː/ and /ʌ/ are roughly equidistant to the Continental /ä/). To assign some other value for (checked) <a>, and spell /æ/ in some other way, would at most benefit some dialects, at the expense of others, if the goal is to make the spelling of English vowels more "Romance".

There is yet a reason why we should keep the same letter for /æ/ and /ɑː/, rather than for /ʌ/ and /ɑː/. There are lots of words that vary between /æ/ and /ɑː/ between English dialects. Apart from the famous bath-set, there are also a large group of recent loanwords that take either /æ/ or /ɑː/ depending on dialect. (Though the distribution is different than for the bath-set: US accents tend to favour /ɑː/.) There are, on the other hand, few words that vary between /ɑː/ and /ʌ/.

[Glossaphile]

As for /Vr/

The reason we should treat "broad" (pre-rhotic) vowels as special cases of short/checked vowels, or as special cases of long/free vowels, is that they historically evolved from long vowels, and that this is reflected in the current orthography. It's much better to keep the current system, as long as it's regular - and if it needs to be change, we should change those instances where it goes wrong - such as <starry> and <furry>.

The centring diphthongs /ɪə̯/ and /ʊə̯/ are basically the "long" vowels /iː/ and /uː/ as they occur before rhotics. Why not keep them so? /ɪə̯/ - contrasts (or at least, contrasted) with both /ɪr/ and with sequences of /iː/ followed by /e(r) - cf serious vs Sirius. The straightforward solution would be to spell /ɪə̯/ and /ʊə̯/ like regular /iːs/'s and /uː/'s - just that they are followed by <r>'s.

Although some dictionaries might list a word as <near> as pronounced [nɪə̯(ɹ)] (at least in the UK) I suppose that's a rather old-fashioned pronunciation. Most Britons, I suppose, would pronounce it either with a monophthong [nɪː(ɹ], or with two syllables: [niːə(ɹ)]. In the latter case, it would effectively merge with sequences of /iː/ followed by /ə(r)/ - so that idea would make a perfect rhyme with dear.

There is definitely some predictability there, but I just don't think it's quite predictable enough to justify what you're suggesting, as shown by the near-minimal pair I gave above: /niːl/ ("kneel") versus /ɹɪəl/ ("real"), or even better, /niːl/ ("kneel") versus /mɪəl/ ("meal"). I almost always assume that lexicographers have good reason for the distinctions they make.

While dictionaries may sometimes be a bit conservative, I tend to have faith that the modern ones are decent at avoiding prescriptivism. Moreover, I need to to stay pretty faithful to at least one of what I call "the Big Three" (Oxford, Cambridge, and Webster) in order to normalize and legitimize my spellings. Otherwise, differences in the native accents of readers can distort the reform system's mapping of sound to symbol. This way, I can point to a cohesive and respected reference.

As for /æ/

I see no reason why we should not continue to spell /æ/ with <a>. One goal of your spelling reform is that the vowel letters should approach their "continental" value. The problem is that many English dialects have no straight equivalent to the continental /ä/. In some dialects, /ʌ/ is pronounced quite similar to continental /ä/. In other dialects, /ɑː/ is pronounced quite similar to continental /ä/. In yet others, /æ/ is pronounced quite similar to it. (In my variety of English, I'd suppose /æ/, /ɑː/ and /ʌ/ are roughly equidistant to the Continental /ä/). To assign some other value for (checked) <a>, and spell /æ/ in some other way, would at most benefit some dialects, at the expense of others, if the goal is to make the spelling of English vowels more "Romance".

There is yet a reason why we should keep the same letter for /æ/ and /ɑː/, rather than for /ʌ/ and /ɑː/. There are lots of words that vary between /æ/ and /ɑː/ between English dialects. Apart from the famous bath-set, there are also a large group of recent loanwords that take either /æ/ or /ɑː/ depending on dialect. (Though the distribution is different than for the bath-set: US accents tend to favour /ɑː/.) There are, on the other hand, few words that vary between /ɑː/ and /ʌ/.

Other dialects seem to be coming up alot lately, but the heart of RLS takes its cues from RP and GA, since they tend to be the most widely understood, respected, and therefore useful accents. As I said before...

You seem to have confused my goal of mild-to-moderate dialectal flexibility with that of always accommodating every single accent in the world. No practical orthography, no matter what the specifications, can possibly do the latter, at least not with a language as diverse and widely used as English. If someone in the northern inlands ever finds cause to explicitly write in his/her own accent, RLS offers them the choice of spelling the standard <pæt> as <pèøt> or <pìøt>. In most situations, however, if he/she hopes to communicate as broadly as possible, then he/she will spell according to the agreed-upon international standard, even if it does diverge from their own pronunciation. I do this all the time in my sample RLS writings, routinely distinguishing /ɒ/ from /ɑː/ even though my native US dialect merges them.

As I've said, some discrepancies with respect to local vernaculars are inevitable no matter what the principles and/or parameters of a proposed spelling system are. All I've endeavored to do is spread out those discrepancies as evenly and fairly as possible, with some focus on major players like RP and GA, so that those divergences are sufficiently infrequent within any single dialect and equally infrequent across a broad range of dialects.

[Xing]

There is definitely some predictability there, but I just don't think it's quite predictable enough to justify what you're suggesting, as shown by the near-minimal pair I gave above: /niːl/ ("kneel") versus /ɹɪəl/ ("real"), or even better, /niːl/ ("kneel") versus /mɪəl/ ("meal"). I almost always assume that lexicographers have good reason for the distinctions they make.

There is no rhotic involved in real - it has, or at least had, a regular /iː/ followed by a /ə/. If you look at Cambridge Dictionaries Online, you will see that /ɪə/ is the regular way in which they transcribe sequences of /iː/ + /ə/ - see for instance their transcription of idea.

Other dialects seem to be coming up alot lately, but the heart of RLS tales its cues from RP and GA, since they tend to be the most widely understood, respected, and therefore useful accents. As I said before...

If a reform proposal benefits some (major) dialect at the expense of other (major) dialects, it is, IMO, highly preferable not to reform. If you have, for example, a strong northern inland accent, your /æ/ is probably a quite close vowel, that's very different from the Romance /ä/. (This is probably one reason why Americans tend to pronounce recent loanwords with /ɑː/ rather than /æ/) For speakers of such an accent, letting short <a> represent some other vowel than /æ/ might make spelling more Romance. But in many other dialects, /æ/ is much more open. For speakers of such dialects - not using <a> for /æ/ would make spelling less Romance.

Even if it's impossible to please all dialects, certain changes would be more problematic than others. Not to write /æ/ and /ɑː/ with the same letter would lead to lots of mismatches, since the distribution of /æ/ and /ɑː/ differs greatly among English dialects. (Or alternatively, dialectal spelling standards would have to proliferate.) As far as I know, there are no such differences in the distribution of /ɑː/ and /ʌ/. (Or maybe there are a few; I just can't think of any right now...) Spelling /ɑː/ and /ʌ/ with different letters would therefore not pose any great problems from a dialectal point of view.

[threecat]

The whole concept of spelling reform should actually be avoided because a spelling-reformed version of English would greatly hamper international trade, science etc. If you designed a phonetic orthography, then English would orthographically splinter into many different and unconnected dialects, making international cooperation impossible. If you pronounced English just a little differently, then you'd need a completely different way to spell words. Ridiculous!