An Attempt to Fix English Spelling (AKA Mission: Impossible)

[KathTheDragon]

In regards to gh, you can leave it in after i and ei at least, since it is always silent, and lengthens a preceeding <i>. However, there are a few situations where <eigh> is pronounced as a long <i>. My solution is to respell those as <egh> (which apparently doesn't even exist yet.)

[ol bofosh]

I've been hanging round on a forum, and most of them quote the difficulty of learning English. But once you've learnt it there's little problem, so most people won't question it until there's a serious problem in it's use. As Adjective Recoil sed "English spelling is not going to be reformed until British, Australian, and American English are mutually unintelligible both in speech and in writing". And now, through present media, the Anglosphere may be converging agen, and until the Internet and Hollywood crash down, differences might actually close for a while (pronuncation patterns etc.).

I'm using this chart, which I was surprised doesn't reduplicate anything anywhere.

That's a huge obstacle to a universal spelling reform. To put it mildly. My favourite idea so far is to create a system that compromises between RP and GA, both being well known in most English speaking countries, and also the basic pronounciation model for learning English in non-English countries. I saw a spelling reform that did this: trap-bath, palm, lot-cloth, thought, which I thought was a good compromise between GA and RP's
RP: trap, bath-palm, lot-cloth, thought
GA: trap-bath, palm-lot, cloth-thought
Words like garage, route and thorough, that don't share lexical sets, could have their "national variation", but by and large a lot of words would share lexical sets and could be organised along those lines.

If each accent were to create it's own variant, it might work, but that could but a further wedge between each dialect and English-as-lingua-franca. Austrailia, for example, could create a as-close-as-humanly-possible phonemic orthography, but they'd distance themselves from the Anglosphere, and even more since no one else would adopt it (I might be tempted, since there are some similarities with my accent).

We're also lacking organisations to lead reforms, and it doesn't look like we'll be getting any soon (I dont mind this... at all). It's left to spelling conventions to make changes. Textspeak has some influence here, but is mostly based on the writer knowing how to break conventional spelling rules rather than making a simple phonemic system.

The most practical solution for me is to make a few changes to the present orthography, and popularise common (yet informal) spelling alternatives: nite, thru, tho, aswell, alot, etc.. SR1 is a favourite of mine (including the Spelling Society's further suggestions), and I've alredy employed it in this post. Cut Spelling goes a tiny bit too far IMO. Visually it's too different and also suggestions like "remove letters irrelevant to pronunciation" is fine until you drastically change the form of a word (eye y/ey/ye?). It may be irrelevant to pronounciation, but it might be relevant to recognition.

I think I'll adopt SR1 to make a start. The Austrailian Government had it for a while: Australian Ministry of Helth.

[Melteor]

If you really want to have uniform spelling and uniform language that's a different story. It probably is worth noting that English has mostly succeeded in cannibalizing its closest sibling Scots, perhaps with the help of the writing system (but I am no expert.) On the other hand, certain dialect features are supposedly getting stronger. From a learner's perspective you still have to get used to hearing all sorts of accents and dialects. I'm not sure having to face the same thing in writing would be better or worse than what we have now. We have a mild version already in British-American spelling differences anyway. Some of the proposals are not new - through -> thru is from Webster. Possibly publishers could pull it off as a means of generating money if people would buy a book spelled according to their local dialect. The closest real example I can think of is Nynorsk in Norwegian, where the orthography can show all the regionalisms but everyone just writes in a compromise anyway. Of course, the biggest difficulty comes from the fact that the people who it works for aren't the ones who think it needs fixing. Even if people mess up the spelling there's still a great level of recognition - English writing is weird like that. In another thread it was suggested that diacritics could be used in concert with digraphs to disambiguate them in context. Then you just don't write the diacritics once you've learned the words. I kind of like that idea.

[rr]

I don't actually believe that there will be a reform of English spelling. Well, I don't know, at least that's what I think you're saying; I could be (and most likely are) wrong. This was mostly just for fun.

[Melteor]

Well yeah it's still fun right?

[KathTheDragon]

I reckon there'll be one in at least the next hundred years.

[rr]

I reckon there'll be one in at least the next hundred years.

It's very possible, but it would be subtle.

Well yeah it's still fun right?

Yup

[Melteor]

Well good.

I played around with the spelling-diacritics-for-the-exceptions thing and I am done. "The süites åre open tó yóu sirs." I just added diacritics for the minor exceptions which are also regular. There can be others for the weirder patternless surprises and some for dialectical nuances if anyone wants that.

[KathTheDragon]

Something needs to be done with <ough>. Every instance of <gh> is either regular, or can be made regular without breaking anything. Not <ough>.

[ol bofosh]

And wot hppns if there is no re4m:

"It appears, therefore, that we are stuck with our established spelling, with all its drawbacks. Moreover, as our pronunciation continues to change, and it is certainly continuing to change - it must inevitably move even further from the spelling. In fact, English appears to be drifting towards a system like that of Chinese, in which the written form of a word has nothing to do with its pronunciation: instead, each word is represented on paper by a set of arbitrary but conventional marks, and becoming literate means learning to recognize and produce the arbitrary marks attached to each spoken word. Well, it works for the Chinese, so no doubt it will work for us, too." R.L.Trask, Mind the Gaffe

In wich case txtspk wud b mor eficient, IMO.

[Melteor]

I don't know why anyone hasn't made Tweets audio sound bites. We've got vine. When the audio goes over a time limit it gets sped up to fit in.

If those kids are smart they have spellchecker working for all their abbreviations.

[KathTheDragon]

I have created a spelling system for English that besides working chiefly on a morpheme level and often leaving vowel length unmarked (vowel length is sometimes explicit), it works extremely well.

[jmcd]

@KathAveara: Show us then? It does seem a reasonable idea for a reform.

It probably is worth noting that English has mostly succeeded in cannibalizing its closest sibling Scots, perhaps with the help of the writing system (but I am no expert.)

It's more despite the differences in orthography: spellings such as <quh> for /ʍ/ and <ou> for /u/ are less common nowadays, often being replaced by their English equivalents and other Scots spellings are still well-known like <ae> for /e/. It's just your normal scenario of big dominant languages beating up the little ones.

In another thread it was suggested that diacritics could be used in concert with digraphs to disambiguate them in context. Then you just don't write the diacritics once you've learned the words. I kind of like that idea.

But then you'd need to remember the diacritics as well as the pronunciation and the spelling so it actually makes there be yet another thing to remember unnecessarily.

[KathTheDragon]

(The below is all written in the new orthography which attempts to preserve as many spellings as possible, whilst still removing deviant spellings)
Most wurds ar stil the same, though sum hav changed dramaticaly. Pronunciations are derived as folows (and ar geared towords RP, unfortunatly):
Wurd-final VCe secwences drop the <e> and lengthen the vowl.
Vowls ar long befor anuther vowl and generally wurd-finally.
When <y> is a vowl, it has the same value as /i/, eccept wurd-finally, wher it is long /e/.
Certain vowl secwences have fixed values, such as <igh egh> being long /i/, <ugh> being 'uf', etc (I'l provide a ful list if recwested)
Cle and Cre drop the <e> and hav sylabic l and r respectivly.
Wurd-final <s> is /z/ after a vowl or a voiced consonant.
<si ti> are /sh/, <zi> is /zh/. <ce ci> are /s/ (wurd-final <ce> also lengthens a preveous vowl), <ge gi> are /j/. If vowel lengthening needs to be blocked, use <sce dge>.
<s> is /s/, /sh/ or /sy/ and <t> is /t/, /ch/ or /ty/ before long /u/ (depends on whether yu have you-dropping, -coalescense, or neither)

The abuv all apli to morphemes. However, ther ar stil ecceptions, mostly in bound morphemes as far as I can tel.

Ther ar also a number of optional orthographic changes tu change the aesthetic:
qu > cw/kw.
When not before <i e>, or when twu wurds are distingwished by the diference, k > c
In some cases, gh > h.


Ther may be mor rules I hav either forgoten or not worked out yet. This is bi no means complete or perfect.


Edit: Aded pronunciation rule.

[Chargone]

Simply marking stress makes English spelling a lot more consistent.
After that, the biggest thing is getting the vowels to line up. Very tricky, as it's the main difference in pronounciation between dialects, though less so if stress is marked.

Tossing the 'words can't end in v' rule to get ov and of instead of of and off wouldn't hurt, and would eliminate a number of confusing CVCe words where the e is both silent and doesn't modify the vowel. (Love, constructive, have), and there's a bunch of other weird fiddles we could probably do without, but the two at the start of this post are probably the most useful.

On a personal note: its vs it's. By the standard rules they should both be it's, to the best of my knowledge there is no valid overlap in their placement, and it's an incredibly common error to get them mixed up (no matter how many times I am told which is which, it doesn't stick, and there is no obvious cue to turn into a mnemonic to help. By the rules they should both be it's, and that's the bit that sticks in my head. I've given up on getting the right one for any given situation and just use it's in all cases these days. "Surprisingly", while some notice the error, no one is ever confused by it.)

[awer]

why don't people just learn not to be lazy and pronounce words as they're written instead? you'd prolly sound like you're from the middle ages.

[Jipí]

why don't people just learn not to be lazy and pronounce words as they're written instead? you'd prolly sound like you're from the middle ages.

Because a) it's nothing to do with lazyness and b) this is not how it works.

[rr]

why don't people just learn not to be lazy and pronounce words as they're written instead?

I seriously doubt that the Great Vowel Shift is a result of laziness. See this Wikipedia article for more information on the Great Vowel Shift: http://en.wikipedia.org/wiki/Great_Vowel_Shift

you'd prolly sound like you're from the middle ages.

The Great Vowel Shift did occur during the late Middle Ages.

[awer]

Because a) it's nothing to do with lazyness and b) this is not how it works.

well, apart from the gvs, most pronunciation changes are based on the "because it's easier to pronounce" rule. think t-flapping, cluster reductions (kn-, ps-), taught-taut,vowel reduction etc. *you're* making all the mess

[Jipí]

Thanks, finally someone who explains to me how sound change works Let's stop this mess of changing pronunciation then, before laziness makes us degenerate fully into just grunting like animals. Instead, let's keep pronouncing things properly, i.e. the way they're spelled.

[Tropylium]

[starting to speak like this is one of the fastest ways to stop complaints about mangled pronunciation]
(Those are IPA phonemic brackets.)

anyway, contrary to popular belief about the impossibility of reform, English spelling reforms actually happen all the time; they just only affect certain registers of the language in the usage of certain speakers (or, maybe rather, writers).

take e.g. the 2nd person singular pronoun having been fairly commonly reformed to "u" in Txt English

I suppose Reformers are expecting some kind of officiality and credit, but in the absense of an Academy of English, who exactly is supposed to be instituting this reform, the ghost of Noah Webster by means of levitating bedsheets to intimidate anyone who doesn't comply?

(and hence my approach to spelling reform is not to create Proposals, it is to just use forms like "tho", "offen", "comprize", "adjectiv" in casual and ambiguously formal contexts, and to wait & see if anything catches on.)

[KathTheDragon]

"offen"

I presume that's "often"? Not everyone pronounces it like that, you know.

[ol bofosh]

(and hence my approach to spelling reform is not to create Proposals, it is to just use forms like "tho", "offen", "comprize", "adjectiv" in casual and ambiguously formal contexts, and to wait & see if anything catches on.)

Like. Since English is overwhelmingly governed by convention, then we can only make our own suggestions and encourage ones alredy in use (informally). I'm learning there's a lot of competition in the spelling reform community (partly why nothing takes on) and not enough collaboration.

I'm all for <e> for /E/ as a start!

And choose hiccup over hiccough!

Tropylium wrote:
"offen"

I presume that's "often"? Not everyone pronounces it like that, you know.

I'm all for offen! It could just be a style choice. Like color and colour...

[Salmoneus]

The real problem with spelling reform is that it's largely pointless. It's unlikely much will happen along those lines, other than a few of the really weird spellings being forgotten (eg hiccough), and maybe increasing neutralisation of letters for schwa to <e> or <u>, particularly in -@nt endings where there's a fair random alternation between <a> and <e>. But by and large, English spelling is far too useful as it is to be degraded by regularisation.


Having said that (and yes, I've worked out my own schemes in the past), one interesting idea that didn't occur to me until this thread is to have a system with one-way sight-to-sound predictability, but NOT sound-to-sight predictability. That is, if you can read the word you know how it is said, but if you hear the word you don't know how it's written.

What's the point of that? Well, it makes the job of reform a lot easier for a start - first because it makes it easier to create a reform system if you only want it to be one way, and second because it lets you leave a lot more existing spellings as they are, making it easier to learn and less alien-looking. But more than that, such a scheme would have a potential purpose - because on the one hand it would make reading easier (so, eg, making it easier for non-natives to learn), while on the other hand it would allow homophones to be disambiguated. And this wouldn't necessarily be limited to retaining current multiple spellings, but could be extended by introducing new doublets. For instance, "sanction" meaning 'to approve or support' could be <sanction>, while "sanction" meaning "to condemn or prevent support for" could be <sanxion> or <sankshun>. Likewise cleeve vs cleave, bite vs byte (oh wait, we already have that one), and so on.

Of course, the new alternative spellings would be more or less arbitrary, so they couldn't really be introduced without a centralised body. But it's an interesting idea.

[KathTheDragon]

one interesting idea that didn't occur to me until this thread is to have a system with one-way sight-to-sound predictability, but NOT sound-to-sight predictability. That is, if you can read the word you know how it is said, but if you hear the word you don't know how it's written.

My scheme is almost there, but it doesn't fully distinguish long and short vowels in all positions.