English Orthography Reform

[ABC]

I'd be more happy with proto-Indo-European than Esperanto. After all, this was the original language of most Europeans' ancestors.

[Iron]

Everyone should just speak Esperanto.

No.

Only Europeans, and Non-Native North Americans and Oceanians should learn Esperanto. Then every other continent/sprachbund should learn one language, after that all aforementioned minor auxlangs should merge into one large auxlang. Along with the speaker's cultural language, of course.

We should all learn Georgian and eject our ways through life.

[ol bofosh]

Ai käm from Inglênd. Ai liv in Catêleunìê. Beuth hav Seint Jòj az thè peitrên seint. It meiks sens dhat ai shud lën Jòjên.

I come from England. I live in Catalonia. Both have St. George as their patron saint. It makes sense that I should learn Georgian.

[Ànradh]

Ai käm from Inglênd. Ai liv in Catêleunìê. Beuth hav Seint Jòj az thè peitrên seint. It meiks sens dhat ai shud lën Jòjên.

I come from England. I live in Catalonia. Both have St. George as their patron saint. It makes sense that I should learn Georgian.

George is my middle name, can that count too? (As if I'm not failing at enough languages...)

[ol bofosh]

If yù wë neimd fò Seint Jòj, dhat'l bì fain.

If you were named for St. George, that'll be fine.

[C.J.]

This is my best effort so far at reforming English spelling. It may not be pretty

It's not too bad, IMO, at least not the first version there. I personally don't really like using digraphs for commonly occurring short vowel sounds, though, so I'd replace at least <ea> with just <e>. Of course, I realize this is a problem if you consider it important to keep /E/ and /@/ distinguished in writing. However, my opinion would be that /@/ wouldn't really necessarily need its own letter even in an otherwise phonemic writing system, since it only occurs as a result of mostly predictable vowel reduction in unstressed syllables. (Some would even analyze it as fully predictable, thus reducing the status of [@] to that of an allophone and not an independent phoneme at all... But it's pretty complicated.)

Also, the need for <æ>/<ae> could be eliminated by sticking with <a> for /æ/ and <o> for /A/; /oU/ could then be spelled <ou> (which would also make things more systematic, considering the usage of <ei> for /eI/). EDIT: Except, of course, that you're already using <ou> for /u/. But maybe that, in turn, could be spelled with <ue> or something?

but I think it's functional, with one letter or digraph representing one sound consistently.

Well, seems to work in that respect. Although I'd point out that an orthography can be considered "functional" even if it's not fully phonemic (in fact, my opinion is that in some situations a not-fully-phonemic one can work better - but that's a whole other discussion).

It is based on General American, the kind spoken in the Pacific Northwest if that matters.

It does kind of matter, in that the system at least stops being phonemic (and possibly also making sense) if you try to read it with another dialect's pronunciation.

I've always hated <o> for /A/. I did take aesthetics into consideration with this, but there's only so much that can be done using 26 letters to represent about 33 sounds. Native English words tend not to look to bad, but longer Latin borrowings can look pretty ugly (institoushenelaizd). I was thinking of using a simple acute accent on the vowel for stress, like Spanish does.

The reason I put <ea> for /E/ is because I think the schwa is the most commonly occurring sound in the English language, whether actually phonemic or not, so it has the shorter spelling. <ou> for /u/ is an aesthetic consideration inspired by French, because I had to solve the dilemma of how to represent /u/ /V/ and /U/ without making the orthography disgusting.

I was trying also to get the vowel letters closer to what they originally represented. In no other language that I can think of that is transcribed with the Latin alphabet is <o> /A/. <ue> actually isn't a bad idea. But I also couldn't think of another language with phonemic /V/ in it, so I used what was already there.

The dialect barriers are already here, and people can read the written language. This system would have to be standardized (preferably with each predominantly-English speaking country with its own dialect having its own variant). People within one country would know the writing system, and that would leave communication mostly unimpeded, but it could potentially cause problems abroad and on the internet.

[Xing]

I was trying also to get the vowel letters closer to what they originally represented. In no other language that I can think of that is transcribed with the Latin alphabet is <o> /A/. <ue> actually isn't a bad idea. But I also couldn't think of another language with phonemic /V/ in it, so I used what was already there.

I don't know what languages you have considered. In many languages (including many dialects of English) <o> represents a back, rounded vowel of varying height, and <a> represents an open, unrounded vowel (though in some languages it can represent a more tense vowel, [æ~ɛ]).

It seems to me like one could have either <o> or <a> for [ɑ] - the two would be roughly equidistant from their "average" values.

A problem/issue in most American forms of English is, of course, the father-bother merger - that /ɑ/ is found both in palm-set word (which are usually spelt with <a>) and lot-set words (which are usually spelt with <o>)

The dialect barriers are already here, and people can read the written language. This system would have to be standardized (preferably with each predominantly-English speaking country with its own dialect having its own variant). People within one country would know the writing system, and that would leave communication mostly unimpeded, but it could potentially cause problems abroad and on the internet.

I think we should keep as much as possible of dialect-neutrality in any spelling reform. It would make things more complicated, if people had to learn two or more orthographies.

We could divide proposals of English spelling reform in three categories:

(1) Reform proposals that attempt to eliminate spellings that are irregular or weird in all or most dialects. IMO, this is what spelling reforms in the first hand should focus on.

(2) Reform proposals that are more dialect-specific - that may make spelling more consistent in some dialects, but not in others. For example, by incorporating the father-bother merger or the trap-bath split in the orthography. Such reforms could make spelling easier to learn for speakers of some dialects. But that has to be weighed against the complications when many people would have to learn two or more orthographies.

(3) Reform proposals that attempt to changes spellings that are regular/consistent in most English dialects, in order for English spelling to become more similar to spellings found in some other languages. This is IMO the least urgent area to reform. Written English is so widespread, and there's no reason to regard - for example - Spanish or Italian pronunciations of the letters as more "correct". It could be a problem, though, when foreign loanwords enter the language. Some older loanwords have been adapted to English spelling (thuggee, chutney), but many newer ones haven't.

A problem/issue is that some loanwords are nativised differently in different dialects of English. For instance, <a> in many recent loanwords is often nativised as /ɑ/ in American English, but as /æ/ in many other varieties.

*******************

Btw: I may have said it before, but I can say it again. I think /ɑ(ː) is a particularly tricky phoneme in English. Partly because it did not exist originally in English, but have evolved form various sources, often different in different dialects. Any proposed orthography that designates a specific letter for /ɑ(ː) is therefore bound to be highly dialect specific. But it's of course not a problem if its one'e goal to have a dialect-specific orthography...

[Xonen]

I've always hated <o> for /A/.

Yes, well. Me, I've always hated <ae> for /æ/, but that may be partially due to having seen Finnish surnames like Väätäinen being written as "Vaeaetaeinen" a few too many times.

The reason I put <ea> for /E/ is because I think the schwa is the most commonly occurring sound in the English language, whether actually phonemic or not, so it has the shorter spelling.

Certainly, I wouldn't suggest using <ea> (or any other digraph) for /@/. The point is that you could just use <e> for both, since the possibility for confusion is well-nigh zero.

But I also couldn't think of another language with phonemic /V/ in it, so I used what was already there.

I suppose one possibility would be using <o> for /V/, which would free up <u> for /U/.

The dialect barriers are already here, and people can read the written language.

Yes, but then, the current orthography isn't dialect-specific. Not that a dialect-specific orthography would be impossible to read for speakers of other dialects, but would kind of treat speakers unequally, and learning to write would pretty much require learning the written dialect as well.

I was trying also to get the vowel letters closer to what they originally represented. In no other language that I can think of that is transcribed with the Latin alphabet is <o> /A/. <ue> actually isn't a bad idea. But I also couldn't think of another language with phonemic /V/ in it, so I used what was already there.

I don't know what languages you have considered. In many languages (including many dialects of English) <o> represents a back, rounded vowel of varying height, and <a> represents an open, unrounded vowel (though in some languages it can represent a more tense vowel, [æ~ɛ]).

It seems to me like one could have either <o> or <a> for [ɑ] - the two would be roughly equidistant from their "average" values.

Well, to be fair, there are plenty of languages that use <a> for [ɑ], while <o> for that value is rather rare. So I can sort of see where this is coming from. However, I personally don't find anything wrong with <o> for [ɑ], especially in a language where plenty of dialects still have a rounded vowel instead of said [ɑ].

[Xing]

Well, to be fair, there are plenty of languages that use <a> for [ɑ], while <o> for that value is rather rare. So I can sort of see where this is coming from. However, I personally don't find anything wrong with <o> for [ɑ], especially in a language where plenty of dialects still have a rounded vowel instead of said [ɑ].

You're probably right here. It may have been a bit premature to say that <a> and <o> are equally plausible candidates for [A], when many more languages would have <a> for [A]. Especially if a language only has a single open vowel, it's very likely that it's represented by <a>.

The situation may be a little different if we are dealing with languages with more unusual vowel systems, especially with two open or near-open vowels - a front one and a back one. In such a language, it'd be roughly as plausible - IMO - to use <a> for either the front or the back one, and find another letter or letter combination for the other.

[C.J.]

The reason I put <ea> for /E/ is because I think the schwa is the most commonly occurring sound in the English language, whether actually phonemic or not, so it has the shorter spelling.

Certainly, I wouldn't suggest using <ea> (or any other digraph) for /@/. The point is that you could just use <e> for both, since the possibility for confusion is well-nigh zero.

I would never use <ea> for /@/. You do have a good point, but I think that making the two orthographically distinct would better help to indicate where the stress falls in a word.

The dialect barriers are already here, and people can read the written language.

Yes, but then, the current orthography <i>isn't</i> dialect-specific. Not that a dialect-specific orthography would be impossible to read for speakers of other dialects, but would kind of treat speakers unequally, and learning to write would pretty much require learning the written dialect as well.

I should have better phrased that. My point was more that people already learn spellings which make no sense to them. A standardized orthography for each English-speaking country would be based on the majority dialect in said countries and would need to be learned by everyone else, which wouldn't make it any harder than it already is to learn to write, seeing as our current orthography is based on the way English was spoken at least 400 years ago. It would actually be easier because, while they still have to learn rules that don't apply to their dialect, the rules would be very consistent.

A better idea, I suppose, would be to base a standard English orthography on the dialect with the most phonemes, probably RP, but with stress differences reflected (elouminem vs. aeluiminyem for aluminum/aluminium) while rhoticity is retained (er for /3/ or the syllabic /r\/ which result in schwas at the end of words in nonrhotic dialects) because such things would be confusing. /A/, /O/, and /Q/ are all /A/ in GA, which would be an easy rule to remember. However it would result in problems in words in which the vowels /}/ and /A/ differentiate. EDIT: But that could be learned.

[Xing]

Some of the worst offenders in English orthography, though, are the result neither of regional differences, nor by the fact that English orthography was standardised a few hundred years ago - but by the fact that the standardisation process was not 100% regular. Spellings like <ie>, <ei>, <ou> and <ow> could be used for different sounds, even before the sound changes and dialectal splits that have led to the present situation occurred.

A better idea, I suppose, would be to base a standard English orthography on the dialect with the most phonemes, probably RP, but with stress differences reflected (elouminem vs. aeluiminyem for aluminum/aluminium) while rhoticity is retained (er for /3/ or the syllabic /r\/ which result in schwas at the end of words in nonrhotic dialects) because such things would be confusing. /A/, /O/, and /Q/ are all /A/ in GA, which would be an easy rule to remember. However it would result in problems in words in which the vowels /}/ and /A/ differentiate. EDIT: But that could be learned.

I could repeat the criteria I suggested a couple of pages ago. A hypothetical "reference accent" for a spelling reform would:

-be rhotic
-be yod-retaining, but would not have undergone yod-coalescence (tune and toon would pronounced differently; as as would dune and June)
-not have undergone the trap-bath split (glass would rhyme with gas, and dancer would rhyme with cancer)
-not have undergone the lot-cloth split (cross would not rhyme with sauce)
-not have merged /ɑː/ and /ɒ/ (father would not rhyme with bother)
-not have merged /ɒ/ and /ɔː/ (cot and caught would be pronounced differently)

Those splits and mergers are the causes of many of the present phonological differences between English dialects.

Some phonological differences would be little more problematic. An "etymological" approach to forming a reference dialect would ignore the foot-strut split (so that cut would rhyme put). But since the vast majority of dialects have the foot-strut split, one could argue that a hypothetical reference dialect should incorporate it.

Another problematic issue is how to handle recent loanwords with <a> (pasta, mafia). GA usually have /ɑ(ː)/ in those, while many other accents have /æ/. Since many of those words has been borrowed quite recently, it does not make sense to ask which English dialect is most conservative here (the words have so to speak entered the English language in different forms in different accents).

[zee]

Á think dhis nyú òthogrùfí shùd bí yúzd wôldwád.
I think this orthography should be used worldwide

Dhen pípùl wil bû dûn tù mí ùnd á wil bí súpírù tù yú òl...
Then people will bow down to me and I will be superior to you all...

Á...á...mín it wil bí fun? :)
I...I.. mean it will be fun? :)

[Xing]

Here is my suggestion for a moderate spelling reform:

●The <ea> digraph ⇒ <ee> when /iː/, <e> when /ɛ/, and /ai/ when /ɛɪ̯/
It quite pointlessly reduplicates other spellings; the resulting homographs are quite few and should not cause any problems, and it can also remove certain problematic homographs ("read", "lead"; etc.
●<oo> for /ʌ/ ⇒ <u>
This change might seem needless for accents without the foot-strut split; but it would be a quite small change.
●Get rid of apostrophes in words like can't and don't.
These have functioned as separate words long enough in English, that it's superfluous to have to indicate that they "really" are contractions.

Here is my suggestion for a more drastic spelling reform (it only concerns vowels so far):

(Historical) long vowels and diphthongs

/iː/ <ee>
/ɛɪ̯/ <ai>, or <ay> word-finally
/ɑɪ̯/ <ie>
/əʊ̯/ <oe>
/aʊ̯/ <ou>, or <ow> word-finally
/ʉː/ <oo>
/jʉ/ <ew>

(Historical) short vowels

/ɪ/ <i>
/ɛ/ <e>
/æ~a/ <a>
/ɒ/ <o>
/ɵ/ and /ʌ/ <u>

"Broad" vowels

/ɑː/ <a> or <á> (when it's historically the result of a retracted /æ-a/, today written <a>)
/ɔː/ <o> or <ó>

Vowels before (historical) /r/:

/ɛː(ɹ)/ <air>
/ɑː(ɹ)/ <ár>
/ɜː(ɹ)/ <ir>
/ɔː(ɹ)/ <ór>
/ʊə(ɹ)/ <ur>

Note that the above are those vowels that historically have been altered by a following /r/; there are still regular short vowels that can occur before /r/: /a/ in <marry>, /ɛ/ in <merry>, /ɒ/ in <lorry> and /ɪ/ in <mirror>.

Unstressed vowels:

/ə/ <a>
/əɹ/ <er>
/i/ (final) <y>

Some comments: The accent to mark the "broad" vowels may be considered optional. The reason for using accents, is that these two "broad" vowels are the cause of some important phonological divisions among English dialects: the trap-bath split and the cot-caught merger. People could write either <ásk> or <ask>, depending on whether or not they have the trap-bath split (and want to indicate it). An accent more or less does not cause to much visual divergence in the written language. The same for the /ɔː/ - those who do distinguish it from /ɒ/ would have the option to indicate it with an accent.

In some important ways, this system fails to be phonemic for many dialects:
It doesn't indicate the father-bother merger (which is prevalent in most American accents) - speakers of such dialects would have to learn that /ɑ(ː)/ is sometimes spelt <á>, and sometimes <o>. The system is also underliyngly rhotic, in that it indicates non-prevocalic /r/, which is silent for many speakers. Again, speakers of certain dialects would just have to learn that some words have silent <r>'s.

Possibly, the system will also presuppose a yod-retaining accent (or maybe I could allow a choice between <oo> and <ew> in many words).

[Xonen]

<oo> for /ʌ/ ⇒ /u/

My first reaction to this was , but then I realized you probably meant to type <u>.

●Get rid of apostrophes in words like can't and don't.
These have functioned as separate words long enough in English, that it's superfluous to have to indicate that they "really" are contractions.

Maybe... Although I'd say that whether or not they count as fully separate words is still debatable. The relationship between them and the uncontracted forms cannot and do not continues to be fully transparent - indeed, the latter are still acceptable variants, and even the preferred ones in more formal registers. Anyway, I wouldn't consider this to be among the biggest problems in English spelling.

[Prinsessa]

Also bear in mind that they're still inflected. Doesn't and don't.

[wakinstube]

I always find dhaet altrnit orthagrufees look lame so I just use a conscript (though in my conscript there are two different spelling systems for GA & RP.)

[Xing]

I always find dhaet altrnit orthagrufees look lame so I just use a conscript (though in my conscript there are two different spelling systems for GA & RP.)

Could you show your conscript?

What phonemic systems do you consider to be "GA" and "RP", respectively?

[wakinstube]

How would I go about showing it to you? I'm not very tech-savvy and I don't have a font for it. GA I view as the usual accent you will find, not southern, not yankee, sort of normal sounding and nondescript. the sort of voice that online dictionaries with audio use. As to RP I'm not as sure, as I don't actually come from england but rather the same I would say, sort of generally english sounding, not pronouning r's in coda etc.

[Xing]

As long as we are concerned with the purely phonemic level, I would say that the (main) differences are:

-RP is non-rhotic, GA is rhotic.
-RP distinguishes between /ɑː/ and /ɒ/ (so that father does not rhyme with bother), GA does not.
-RP pronounces marry, merry and Mary differently, GA pronounces them the same.
-GA is yod-dropping after /t/, /d/ and /n/; RP retains /j/ after these consonants.
-RP (except for some conservative varieties) does not have the lot-cloth split; those varieties of GA that haven't gone through the cot-caught merger have the lot-cloth split.
-RP has the trap-bath split, and pronounces words like bath, pass, dance, command etc. with /ɑː/; GA does not have said split, and has /æ/ in such words.
-RP pronounces some recent loan-words and foreign names (such as pasta, mafia, kebab, Sri Lanka, Vietnam, etc.) with /æ/, GA with /ɑ(ː).

Apart from those general accentual differences (that concerns general differences, applicable to large groups of words), there are many differences in the pronunciation of single words (garage, laboratory etc.)

(Even though I have spoken in terms of "RP" and "GA" - the more "standard", "posh", or sometimes "conservative" accents on either side of the pond - many other British and American accents would retain the same phonemic systems, though the exact phonetic realisations may be different.)

---

Just wanted to check if you had the same differences between GA and RP in mind when you designed your scripts

[ol bofosh]

many other British and American accents would retain the same phonemic systems, though the exact phonetic realisations may be different.

Yeah, I'm from SE England and share a lot in common phonemically with RP. A lot of the mergers and splits in my accent are along similar lines, but I'm phonetically distinct in most instances.