© Charles Chandler

 

Phonetic English old example new a pain, pane path taught, awe aa a aw eɪ æ ɒ b bird b b c cat chicken k c k t͡ʃ d dog d d e eel edit, spotted ee e i ɛ​ f fish f f g goat g g h horse h h i wise sit ii i aɪ ɪ j jaguar j d͡ʒ k kangaroo k k l leopard l l m monkey m m n neanderthal think starling n nk ng n ŋ ŋ o antelope osprey trout hoist oo o ow oy o ɑ aʊ ɔɪ p panther p p q quail kw r rabbit r r s seal sheep pleasure s x jz s ʃ ʒ t tiger that t q,θ t θ u through put about, run porcupine uu u u yuu u ʊ ə,ʌ yu v vampire v v w wolf w w x fox ks x y yak y j z zebra z z

Originally, all spoken languages were represented phonetically, and they all had symbols that corresponded unambiguously with individual phonemes. Some languages (e.g., Spanish) are still spelled phonetically, making them easy to read.

 

English, of course, is a different story. So how did we wind up with so many words that are not pronounced the way they are spelled? Part of the problem is that English has a lot of words that were borrowed from other languages, which were originally spelled and pronounced as foreign words, but which were naturalized into something that rolls easier off the English tongue, though the new sound and the old spelling no longer match. Another part of the problem is that England has been overrun by wave after wave of migrants, each of whom brought distinctive speech patterns with them, shifting conventional pronunciations in arbitrary and meaningless ways.

 

Many systems have been developed to attempt to remedy this situation, for particular languages, or for all languages. The biggest problem at this point is that technology is now getting in the way. We have computer keyboards that work quite nicely for a limited number of characters, but would be unworkable if we had to use different keys for each different phoneme. And typing two-letter sequences to access all of these characters would be a nightmare.

 

Yet technology can fix what it broke. Electronically, it's easy to swap one spelling for another. This means that computer software can offer the option of displaying a language using any spelling scheme that the user chooses. It might be dictionary English, or phonetic English (using the system recommended below, or any other), or it might be English but with the phonetic spelling from some other language (so that non-English speakers can learn the sound of words, without first having to learn the alphabet). In other words, phonetics can be a display/input mode, and we need not address the issue of how we're going to convert everything to a new system. The conversion should be done on the fly, and as a user preference, not a global mandate.

 

The table at right shows the principle phonemes in American English that need to be represented, along with the existing characters that are used, and the IPA characters (most of which are not even in the extended set, much less on the keyboard, so nobody is going to type these). The proposed characters for these sounds are shown, chosen on the basis of being easy to type, not relying on anything that isn't on the standard keyboard. The long vowels are double-stroked; the common short vowels are single-stroked; and short vowel variations are designated with a second character.

 

Most of it is pretty straight-forward, but a few items deserve comment.

• "C" is not used for the "K" or "S" sounds, which already have letters, leaving it free to represent the "Ch" sound, as in Italian.
• "Q" and "X" are not used for their existing sounds, being replaced by "Kw" and "Ks" respectively.
• This leaves two English phonemes that do not already have dedicated characters, and that cannot be reasonably constructed from character combinations, namely "Th" (as in "Think") and "Sh" (as in "Sheep"). These are represented by re-tasking the two obsolete characters (Q and X). Since Q looks more like the Greek theta (Θ) with the same sound, Q is used for the "Th" sound. X is then used for the "Sh" sound, since it already had a bit of "S" in it. It would be less confusing the display these characters as Θ and ʃ respectively.
• Consonants shouldn't need to be doubled to suggest short preceding vowels, especially if doubled vowel characters are used to represent the long sounds.
• There is still a little bit of ambiguity in the short vowel sounds. For example, the same character (U) is used for "put" and "putt". So more vowel modifier keys could be assigned, to unambiguously represent these sounds. This would be especially true for other languages, some of which have a wider variety of vowel sounds.

So here's what American English would look like (with "Θ" in place of "Th", "ʃ" in place of "Sh", and long vowels doubled):

Θu kwik brown fox jumpd oovir θu laazee dawg.

Awl in awl, θis wud taak sum geting yuusd tuu — kiind uv liik kidz first lirning tuu sownd owt wirdz. Θu gud nuuz iz θat kidz wud lirn fastir θan us. Θat's if θaa kan forget teksting langwij for just u minit, wic θaa yuuz raθir eksessivlee, θoo θaa ʃudn't.

The mapping couldn't be accomplished entirely by an algorithm, since English spelling doesn't already obey rules. Thus this could only be done with a dictionary mapping.