Last updated January 27th, 1997, and again September 13, 2002
It is an often illustrated fact that the spelling of the English language is inconsistent, and I won't repeat the examples here. In spite of its inconsistency, personally I find it so beautiful that I am opposed to any proposal to change it. Yet, for the first time in 1978 and again several times after that, I have attempted to design an English spelling system that would work strictly according to rules, while inheriting something of the traditional beauty. And because my ideas on the subject have been nearly stable for some years now, I thought I should present it to whoever wants to read it. It is meant as a serious proposal, although it may also be looked upon as a demonstration of how strictly applying rules leads to perhaps beautiful, often familiar, but sometimes bizarre results.
The rest of the article consists of four major parts:
The Read It chapter lists letters and combinations thereof, and explains to what sounds they correspond in different positions and situations.
A more difficult objective will be to write in the new spelling. The Write It chapter presents phonemes (largely by using examples, so it may be understandable even for those who don't know what a phoneme is), and how to represent them.
Then, in Design Principles there is an explanation of the underlying principles used when designing the system, and of conflicts and difficulties that arise from them.
Finally, in Sample Text some parts of this article are repeated, rewritten in the new spelling.
In order to avoid problems regarding fonts and computer platforms, the phonemic representation does not use the IPA symbols (IPA = International Phonetic Association), but rather their ASCII-representation, following the scheme designed by Evan Kirshenbaum,
The transcription is based on British English, the so-called Received Pronunciation, in the vain hope that in other kinds of English the sounds are different but the phonemes are the same.
The best way to read texts in the new spelling is to read it (aloud, if that helps) as if they were in the traditional spelling, and try to grasp whole sentences. Most of it will be readable at first glance. To test that, you can try the Sample Text below. For any remaining cases of doubt, and to see the design-principles illustrated by examples, you'll find letters and letter-combinations, together with the sounds they represent.
a
|
/@/ |
In unstressed position, if not followed by a written r in the same syllable. Unstressed means that the syllable has no primary, and no secondary stress. In a word like inthuzyazm (enthusiasm), the a does not have this sound, because of the weak secondary stress. Arize (arise), ago (ago), ahed (ahead), apere (appear), akur (occur), akazhen (occasion), dha (the; not before a vowel), idea (idea), kamputer (computer), fual (fuel), diamand (diamond), uzhual (usual). |
|
/&/ |
When followed by a consonant, which is not followed by a vowel. The combinations sh, zh, ch, th, dh, st, nj and pr are counted as one consonant for this rule. To make them count as double, double their first letter, e.g. ssh and ddh. Kat (cat), chat (chat), smash (smash), smasshing (smashing), gaddher (gather), marry (marry), karry (carry), inthuzyazm (enthusiasm), fast (fast; showing American, North-English, Scottish, etc. ways to pronounce it; not suggesting these all sound the same!; the British variant looks like fahst). |
|
/eI/ |
When followed by a single consonant (or sh, zh, ch, th, dh, st, nj, pr), which is followed by a vowel. When followed by a vowel. Make (make), taste (taste), nashen (nation), nacher (nature), perswazhen (persuasion), badhe (bathe), brane (brain), plane (plane, plain), iksplane (explain), Apral (April), apran (apron), chanje (change), chanjiz (changes), chanjed (changed), daze (daze; but not: days, which becomes dayz). |
|
/A:/ |
When followed by one r, which in turn is followed by one or more consonants (not being a second r), or which is final. Ar (are), kar (car), far (far), fardher (farther; cf. father), start (start), star (star), starry (starry). Note that this last word should be pronounced like carry according to the rule just presented, but it is not. The reason is that the relation between star and starry is of a different nature than that between car and carry. |
|
/E@/ |
When followed by one r, which in turn is followed by a vowel. Are (air), kare (care), stare (stare), skarez (scares), skares (scarce). |
ah
|
/A:/ |
This digraph is written for this sound, if it is not due to written r. Fahdher (father), Shekahgo (Chicago), and, for British pronunciation rahdher (rather), fahst (fast), bahth (bath). |
au
|
/O:/ |
Written for long open o-sounds not due to written r. Thaut (thought), sau (saw), wauter (water), aul (all), kaul (call). |
ay
|
/eI/ |
Written at the end of a word, and in derivations and compounds of such words. Day (day), dayz (days; but not daze), daylite (daylight), say (say), saying (saying), dhay (they), nay (neigh, nay), trane (train). |
e
|
// |
The letter e is not pronounced at all, if it serves only to influence the quality of the vowel that precedes the preceding single consonant (or a group which is counted as one consonant). For examples refer to: phonetic symbols a pronounced as /eI/ e pronounced as /i:/ i pronounced as /aI/ o pronounced as /OU/ u pronounced as /ju:/ |
|
/I/ |
In words with a stressed second syllable, in which the first syllable is be-, de-, pre-, re-, se-, te-, or e-, followed by a single consonant, or a combination of such prefixes. In words which seem to have such a prefix, but where it is not really a prefix. In monosyllabic words, that have a long i-sound when pronounced in isolation, but a short i-sound in rapid speech. In other cases where it helps avoid ugly doubling of consonants. Examples: enuf (enough), befor (before), dezert (desert, as a verb; the noun is spelt dezzert), deni (deny), pretend (pretend), remember (remember), redevellap (redevelop), devellap (develop), selekt (select), tenashas (tenacious), elevvan (eleven), eluzhan (elusion, illusion), annemal (animal), be, me, she, he, we, dhe (the, before a vowel), idea (idea), ideal (ideal), kureas (curious), akremoneas (acrimonious), kreate (create), kreashen (creation), theater (theater, theatre), theatrikl (theatrical), realistik (realistic), efishnt (efficient), rekkwezit (requisite), peanno (piano). |
|
/@/ |
When unstressed, and followed by r. Better (better), Peter (Peter). |
|
/e/ |
When followed by a consonant, which is not followed by a vowel. The combinations sh, zh, ch, th, dh, kw and st are counted as one consonant for this rule. To make them count as double, double their first letter, e.g. ssh and ddh. Bed (bed), bet (bet), get (get), getting (getting), sez (says), breth (breath), bredhran (brethren), ekkwitty (equity), mezzher (measure), enny (any), menny (many), ennytthing (anything), preffas (preface). |
|
/i:/ |
When followed by a single consonant (sh, zh, ch, th, dh, kw, st) which is followed by a vowel. When followed by a vowel. Fele (feel), bredhe (breathe), beste (beast), ekwal (equal), sezen (season), seze (seize), sezed (seized), seset (ceased), leset (least, leased), leshet (leashed), feste (feast). |
|
/@:/ |
When followed by one r, which in turn is followed by one or more consonants (not being a second r), or which is final. Her (her), herd (herd, heard), therd (third), hert (hurt), werd (word), defer (defer), deferring (deferring; this should have a sound as in error according to the rules, but it has not. This is because of the analogy with defer). |
|
/I@/ |
When followed by one r, which in turn is followed by a vowel. Here (here, hear), bere (beer), hering (hearing). |
ee
|
/i:/ |
Double e is written for this sound at the end of a word, and in derivations and compounds of such words. See (see, sea), seez (sees, seas; compare seize), tree (tree), treez (trees). |
i
|
/I/ |
When followed by a consonant, which is not followed by a vowel. The combinations sh, zh, ch, th, dh and nd are counted as one consonant for this rule. To make them count as double, double their first letter, e.g. ssh and ddh. Begin (begin), bitter (bitter), wimmin (women), in (in, inn), bizzy (busy), biznis (business), kist (kissed), list (list), mist (missed, mist), forrid (forehead), invalid ('invalid), invallid (inv'alid), villij (village), villijjerz (villagers), shortij (shortage). |
|
/aI/ |
When followed by a single consonant (or sh, zh, ch, th, dh, nd), which is followed by a vowel. When followed by a vowel. At the end of a word. I (I, eye), likes (likes), kinde (kind), behinde (behind), linez (lines), gidez (guides), sidez (sides), sites (sites), bi (by), diagram (diagram), deni (deny), denial (denial), deniing (denying), diaper (diaper), moddefi (modify), hi (high), hier (higher; compare hire below), li (lie), lize (lies), lier (liar). |
|
/aI@/ |
When followed by one r, which in turn is followed by a vowel. Fire (fire), hire (hire), tire (tire, tyre). |
o
|
/O/ |
When followed by a consonant, which is not followed by a vowel. The combinations sh, zh, ch, th,dh, st and ld are counted as one consonant for this rule. To make them count as double, double their first letter, e.g. ssh and ddh. Woz (was), wosh (wash), wosshing (washing), lost (lost), twmorro (tomorrow), worranty (warranty), worrantee (warrantee), progres (progress; American), not (not), orthografy (orthography), fatografy (photograpy), ov (of), of (off). |
|
/OU/ |
When followed by a single consonant (or sh, zh, ch, th, dh, st, ld), which is followed by a vowel. When followed by a vowel. At the end of a word. So (so, sew), do (doe, do (the note); compare the verb do), going (going), no (no, know), noing (knowing), o (O, oh, owe), oing (owing), bote (boat), note (note), hoste (host), toste (toast), sole (soul, sole), solde (sold), kolde (cold), progeram (program, programme), progeres (progress; British), aproche (approach), aprochet (approached). |
|
/O:/ |
When followed by one r, which may be followed by some other consonant. Mor (more), for (for, four), yor (your), hors (horse). |
oo
|
/u:/ |
Fool (fool), soop (soup), hoo (who), too (too, two), doo (do), choo (chew). |
|
/U@/ |
When followed by r. Poor (poor), toor (tour). |
ou
|
/aU/ |
Toun (town), hou (how), out (out). |
|
/aU@/ |
When followed by r. Tour (tower), our (our, hour). |
oi
|
/OI/ |
Boy (boy), poynt (point), anoy (annoy), avoyd (avoid). (Note: oy was chosen instead of oi in order to avoid ambiguity with words like "going"). |
u
|
/V/ |
When followed by a consonant, which is not followed by a vowel. The combinations sh, zh, ch, th and dh are counted as one consonant for this rule. To make them count as double, double their first letter, e.g. ssh and ddh. Just (just), kup (cup), butter (butter), hurry (hurry), udder (udder), uddher (other), wun (one), dun (done), nun (none, nun), sun (sun), sum (some, sum), sumwun (someone), kuller (colour, color), frunt (front), kupl (couple). |
|
/ju:/ /jU/ |
When followed by a single consonant (or sh, zh, ch, th, dh, st), which is followed by a vowel. When followed by a vowel. At the end of a word. Nu (new, knew), nuer (newer), fu (few), ku (queue, cue), kuer (queueer), tune (tune), uze (use, as a verb), use (use, as a noun), uziz (uses, as a verb), usiz (uses, as a noun), uzed (used), uset tw (used to), kure (cure), kurabl (curable). |
|
/u:/ /U/ |
Where the /ju/ turned an s into sh or a z into zh. Shuger (sugar), uzhual (usual). |
|
/jU@/ |
When followed by one r, which in turn is followed by a vowel. Pure (pure), kure (cure), kureas (curious). |
|
/U@/ |
Where the /jU@/ turned a s into sh or z into zh. Sure (sure). |
w
|
/U/ |
Gwd (good), pwt (put), fwl (full), lwk (look), wwd (wood, would), swd (should). |
|
/w/ |
See the section on consonants. |
y
|
/I/ |
Before vowels, and in final position. Also in derivations of words in which the i-sound was final. Pritty (pretty), lately (lately), hastyly (hastily), happy (happy), happyer (happier), berry (bury, berry), berryd (buried), dinnasty, dinasty (dynasty), privvassy, privassy (privacy). |
|
/j/ |
See the section on consonants. |
b
|
/b/ |
Be (be), bee (bee), bere (beer), baby (baby), eb (ebb), throbbing (throbbing), basen (basin), base (base, bass), basik (basic), basis (basis), baseze (bases), biollajy (biology), bialojjikl (biological), bothe (both), boddher (bother). |
c
|
// |
Only used in proper names, and for foreign languages. |
ch
|
/tS/ |
Chane (chain), kach (catch), kaccher (catching), kacching (catching), nacher (nature), cherch (church). |
|
/x/ |
Loch (loch). |
d
|
/d/ |
Ded (dead), Dad (Dad), hard (hard), meddo (meadow). |
dh
|
/D/ |
Dhat (that), dha (the; before a consonant), dhe (the; before a vowel), dhis (this), dheze (these), feddher (feather), bredhe (breathe), bredhran (brethren), widh, with (with), fahdher (father), rahdher / raddher (rather), muddher (mother) |
f
|
/f/ |
Fase (face), faze (phase), differ (differ), defer (defer), enuf (enough), fite (fight), faut (fought). |
g
|
/g/ |
Giv (give), givving (giving), go (go), eg (egg), egz (eggs), begger (beggar). |
h
|
/h/ |
Hi (high, hi), hier (higher), hire (hire), hoo (who), hooz (whose), gras-hopper (grasshopper). |
j
|
/dZ/ |
Juj (judge), jujjiz (judges), ej (edge), ejjiz (edges), brij (bridge), unnabrijd (unabridged), chanje (change), chanjed (changed), langwij (language), anjal (angel), bannjo (banjo), igzajjerate (exaggerate), sajest (suggest; British), sagjest (suggest; American), sajeschn, sagjeschn (suggestion), jale (jail, gaol). |
k
|
/k/ |
Kepe (keep), kis (kis), kissing (kissing), kar (car), kool (cool), nok (knock, nock), kwk (cook), nokking (knocking), rekwire (require), rekkwezit (requisite), kahky, kakky (khaki), teknollajy (technology), teknalojjikl (technological). |
l
|
/l/ |
Luv (love), luvver (lover), luvving (loving), lwk (look), mil (mill), miller (miller), milk (milk), kerl (curl), loos (loose), looz (lose). |
m
|
/m/ |
Moov (move), mene (mean), room (room), kreme (cream), annemal (animal), parammaterz (parameters). |
n
|
/n/ |
In (in, inn), nevver (never), no (no, know), noing (knowing), nu (new, knew), nollij (knowledge), nollijjabl (knowledgeable). |
|
/N/ |
Link (link), sinking (sinking), munk (monk), zinkt (zincked). |
ng
|
/N/ |
Sing (sing), singer (singer), long (long), longger (longer), fingger (finger), lungz (lungs), Ingglish (English) |
p
|
/p/ |
Paper (paper), pipe (pipe), pool (pool), pwl (pull), pahst, past (passed, past), paste (paste), spase (space), spashas (spacious), plad (plaid), plat, plait (plait), playd (played), spelling reform (spelling reform), projjekt (project), prajekt (project). |
q
|
// |
Only used in proper names, and for foreign languages. |
r
|
/r/ |
Ranje (range), morover (moreover), berry (berry, bury), rede (reed, read), red (read, red). |
|
// |
When following a vowel, and not followed by a vowel, it is not pronounced at all (in one style of pronunciation), but only influences the quality and length of the preceding vowel. Far (far), ferst (first), hors (horse), pure (pure). |
s
|
/s/ |
Spekes (speaks), spekerz (speakers), spasiz (spaces), skarz (scars), skares (scarce), skarez (scares), sezen (season), sitty (city), loos (loose), lisnz (listens), goos (goose), gese (geese), say (say), sed (said), sez (says), sad (sad), sno (snow), salillakwy (soliloquy). |
sh
|
/S/ |
Fish (fish), nashen (nation), nashnal (national), nasshanallitty, ship (ship), kash (cash, cache), rashnal (rational), rashn (ration), rasho (ratio). |
t
|
/t/ |
Take (take), letter (letter), ahskt, askt (asked), taste (taste), kwkt (cooked), tucht (touched). |
th
|
/T/ |
Thingz (things), breth (breath), lengthan (lengthen), nutthing (nothing). |
v
|
/v/ |
Voual (vowel), delivver (deliver), liv (live), hav (have), havving (having), hevvy (heavy), veikl (vehicle), veamans (vehemence), veamantly (vehemently). |
w
|
/w/ |
Only at the beginning of a word or syllable, and after k or t. Wile (while, wile), wich (which, witch), west (west), kweschn (question), wor (war, wore), worranty (warranty), worrantee (warrantee), intertwine (intertwine), wwd (wood, would), wwman (woman), wwlf (wolf). |
|
/U/ |
See the section on vowels. |
x
|
// |
Only used in proper names, and for foreign languages. |
y
|
/j/ |
Only at the beginning of a word or syllable. Yesterday (yesterday), yoo (you), kortyard (courtyard). |
|
|
See the section on vowels. |
z
|
/z/ |
Zele (zeal), zone (zone), iz (is), dhoze (those), dheze (these), woz (was), faziz (phases), fasiz (faces), baseze (bases), looz (lose), zilafone (xylophone), dayz (days), nuze (news), pazes (possess), pazesiz (possesses), pazest (possessed), pazeshn (possession), sizzerz (scissors). |
zh
|
/Z/ |
Lezzher (British) lezher (American) (leisure), mezzher (measure), trezzhery (treaury) kahezhen (cohesion). |
Here we see phonemes (based on the RP style of pronunciation; if that isn't helpful, ignore them and just look at the examples), with the letters used to write them, and some examples. This table is much leaner than the previous one, no rules, no explanations, fewer examples.
/&/
a kat (cat)
/e/
e bed (bed)
/I/
e befor (before)
i begin (begin)
y pritty (pretty)
/i:/
e seze (seize)
ee see (sea, see)
/I@/
er here (here, hear)
ea idea (idea), sereas (serious)
/eI/
a make (make)
ay day (day)
/E@/
are kare (care)
/A:/
ar kar (car)
ah fahdher (father)
/O/
o not (not, knot)
/O:/
au thaut (thought)
or mor (more)
/OU/
o so (so), bote (boat)
/U/
w fwl (full)
u shuger (sugar)
/U@/
oor poor (poor)
/u:/
oo fool (fool), sooper (super)
/V/
u kup (cup)
/@/
a ago (ago)
er better (better);
/@:/
er refer (refer)
/aI/
i hi (high), mine (mine), reli (rely)
/aI@/
ire fire (fire)
/aU/
ou out (out)
/aU@/
our our (our, hour)
/OI/
oi boy (boy)
/ju:/
u uze (use)
/p/
p pipe (pipe)
/b/
b bi (by, bye)
/f/
f fite (fight)
/v/
v verry (very), vary (vary)
/m/
m mi (my), me (me)
/w/
w we (we)
/t/
t tee (tea, T)
/d/
d doun (down)
/T/
th thinks (thinks), thingz (things)
/D/
dh dheze (these), dhat (dhat)
/s/
s say (say)
/z/
z maze (maze, maize)
/n/
n noun (noun)
/l/
l lo (low)
/r/
r ro (row)
/S/
sh ships (ships), shelz (shells)
/Z/
zh espeanahzh (espionage)
/tS/
ch chor (chore)
/dZ/
j jam (jam)
/j/
y yoke (yoke)
/k/
k kis (kiss), tikl (tickle)
/g/
g gote (goat)
/N/
n stinks (stinks)
ng yung (young), yungger (younger)
/h/
h haste (haste)
/x/
ch loch (loch)
Phonemic consistency
A phoneme may be defined as a group of related sounds, the common characteristics of which can distinguish words from each other. For example the l-sound in look, like, leave, full are all different, due to influence from adjacent sounds, or due to the position of the sound in the word. Yet we consider /l/ to be one phoneme. It can distinguish otherwise identical words, for example the pair light/right.
The ideal spelling system consistently uses one letter for each phoneme, but most languages, including English, have more phonemes than letters. This can be overcome by using diacritics, which is not done in the traditional English spelling system, and is not proposed here either, because in despite standardisation, it often leads to heated debates when texts are used on computers and exchanged via Internet or otherwise. One other approach is to use letter combinations, digraphs, to stand for phonemes. The disadvantage is that the phonemes represented by a combination take more space that those written as one letter. A compromise is to use combinations sparingly, but also make some single letters stand for more than one phoneme, but then dictated by unambiguous, or nearly unambiguous, rules.
The effect of theoretical ambiguity will be limited by practical language facts; for example the new spelling pazes for "possess") is ambiguous in theory, but the incorrect reading (taking a to mean /eI/ and e to be mute and only indicate the length of the preceding vowel) leads to the pronunciation /peIzs/, which is so unlikely that it isn't a reasonable candidate.
Retain existing consistency
Although somewhat chaotic, the existing English spelling does have rules, which make it possible to guess the right pronunciation of an unknown word at a glance, with at least some chance of success. The well-known example ghoti, which is said to read as fish (as in enough, women, nation), illustrates this, for no one would spontaneously read it as fish, because the letters can mean the sounds, but not in this situation.
The new system proposed here uses as much as possible of the existing rules, which makes it possible that, although it is drastically different, the new spelling is readable to a certain extent for anyone familiar with the traditional spelling. (Try it yourself by reading the Sample Text ). This also means the new system is not made unnecessarily "un-English": That in most other languages the letter a sounds something like the a in father, and the sound in make is often written with an e or a digraph containing it, doesn't mean this has to be the case in English too. Thus, make and take stay the same in the newly proposed system.
The historic difference between long and short sounds, like in licking and liking, tacking and taking, etc., shown in spelling by following them with a single or double consonant, is retained in the new system, together with the sound quality to which they developed. All five letters have such originally short and long forms. Some letters also have other meaning independent of this scheme, and they do not need doubling, but they are often governed by stress (like the a in ago), or they are implicit, meaning the new system is not completely unambiguous, though much more so than the old.
To avoid some ugly new spellings, some letter combinations are supposed to be single for the purpose of deciding whether the vowel before them is "short" or "long" (short and long in the historic sense, not corresponding to present-day sounds). Thus in taste (new and old spelling alike), the st is counted as one letter, which the a "long". Now, if a word tass-ting existed (luckily it does not), in the new system it would have to look like tassting, to distinguish it from the gerund form of taste: tasting (again, old and new spelling the same). So to double a single letter, it is doubled (k becomes kk, never ck), but to double a double letter which is counted as single (for example sh in smash), double its first letter, so it becomes ssh: smasshing, so it doesn't sound like smay-shing.
This doubling of the first letter is admittedly very un-English, and is borrowed from Hungarian (where it denotes the letter itself, and does not influence the preceding sound, by the way).
One other very un-English trick is to use the letter w (which can be a consonant too), as a vowel. This a taken from Welsh.
At the end of a word, a single letter is enough to get the "short" sound, but if another vowel follows, doubling is required. We follow the Dutch (and to a large extent, English) system here, not the German system, which uses doubling even at the end of a word (sometimes English does that too, as in off and inn and sick). Examples of this are let and letting, get and getting (new and old system alike). A typically English feature in addition to this is, that a mute e (as in mute, where the e makes the u is "long", but is not itself heard) lengthens the preceding vowel. This is retained in the new system too. Thus, the new spelling kahezhen (cohesion) suggests that the last syllable does not have a vowel, but that the e is mute and serves only to make the previous e "long", and the last syllable has a syllabic n. The new spelling kahezhan would mean that the last syllable has a schwa vowel (as a in ago), followed by an n.
Consistency for all speakers
English is spoken in many different countries, and even within one country in many different ways. This wouldn't have to complicate a spelling system, if only all variants had the same phoneme system, but realised the same phonemes with different sounds. Unfortunately, although the differences are not too many, this is not the case. Different phonemes in one style of pronunciation fall together in another. A phoneme in one word may defect to another. What is conveniently seen as one phoneme in one style is better considered two in another.
All this makes compromises inevitable. In many cases, my basis has
been British English (RP - Received Pronunciation), while trying to
arrive at solutions also reasonably acceptable in North America.
(Perhaps that's why the dollar becomes doller, but the pound stays pound; the euro becomes uro, though).
One reason is that RP in many cases coincides (not regarding actual
sounds, but rather the phoneme structure) with usage in Australia,
New Zealand, South Africa, Ireland and Scotland. Another is that
showing distinctions that some do not hear may make writing more
difficult, but reading easy. It also keeps the new spelling closer
to the old, which again eases reading.
And finally a practical but utterly invalid reason: I am more familiar with British than with American pronunciation.
Some cases in detail:
Morphology
In some cases the principle of morphology is used, like in daylite instead of dalite, hier (higher) instead of hire, dayz (plural of day), not daze (as in "dazed and kanfuzed"), star and starry.
Where compounds could lead to ambiguity, a hyphen may be used: gras-hopper (grashopper would sound like gray-shopper).
Many frequent words have strong forms (when said in isolation) and weak forms (in connected speech). In some cases there even more forms than two. Although the weak form occur more often, the spelling for the sake of clarity shows strong forms. So for is always written like that, even though is more often sounds like fa or fer, as is written as as, at as at (the strong and weak forms have the same spelling here, and this is a feature, not a bug), and is always and, has, had and have are always that. That stays dhat and than stays dhan (again the weak forms would be spelled the same).
The principle of writing strong forms applies even to compounds: twmorro, not tamorro. In "He came to" and "he came to his senses", to is always spelt tw (he kame tw; he kame tw hiz sensiz), although is the first case, it may sounds rather like too. The spelling too is reserved for the meanings "also" and "a number one higher than one).
The special meaning of the letter e, as in be, we, she etc. is a special case of the same principle. Thus be and bee are identical in new and old, and distinct in both.
Etymology
An etymological principle is hardly applied at all except perhaps when writing you as yoo, not u.
I found that Mark Rosenfelder also thinks that the existing English spelling does have more underlying consistency than it seems at first sight.
He even wrote a
Sound Change Applier
program
that is able to transform Latin into Portuguese by applying sound transitions,
but also manages to predict the correct pronunciaton of English words with
a remarkably high rate of success.
This leads me to believe that his program, if fed with the appropriate
transition definitions, could also convert existing English into my
Ingglish spelling with a reasonable degree of accuracy. I didn't try
it yet, and probably never will, but I expect it can be done.
Intradukshn
It iz an offan illustratid fakt dhat dha spelling ov dhe Ingglish langwij iz inkansistant, and i wonet repete dhe igzahmplz here. In spite ov its inkansistansy, persanaly i finde it so butyfwl dhat i am apozed to enny prapozl tw chanje it. Yet, for dha ferst time in 1978 and agen sevral timez ahfter dhat, i hav atemptid tw dezine an Ingglish spelling sistim dhat wwd werk striktly akording tw rulez, while inherritting sumthing ov dha tradishnal buty. And bekoz mi ideaz on dha subjekt hav bene nerely stabel for sum yerez nou, i thaut i shwd prezent it tw whooevver wonts tw rede it. It iz ment az a sereas prapozl, auldho it may aulso be lwkt apon az a demmanstrashen ov hou striktly apliing rulez ledez tw perhaps butyfwl, offan familyer, but sumtimez bezar rezults.
Dha rest ov dha artikl kansists ov for majer parts:
Dhe Rede It chapter lists letters and kambinnashenz dharov, and iksplanez tw wot soundz dhay korrispond in diffrant pazishnz and sittuashenz.
A mor diffikult abjektiv wil be tw rite in dha nu spelling. Dha Rite It chapter prezents fonemez (larjly bu uzing igzahmplz, so it may be understandabl even for dhoze hoo donet no wot a foneme iz), and how tw reprezent dhem.
Dhen, in Dezine Prinsiplz dhare iz an iksplanashen ov dhe underliing prinsiplz uzed wen dezining dhe sistim, and ov konflikts and diffikultiz dhat arize from dhem.
Finaly, in Sampl Tekst sum parts ov dhis artikle ar repetid, reritn in dha nu spelling.
Ingglish iz spoken in menny difrant kuntriz, and even widhin wun kuntry in menny difrant wayz. Dhis wwdnt hav tw kamplikate a spelling sistim, if onely aul vareants had dha same foneme sistim, but realized dha same fonemez widh difrant soundz. Unforchanatly, auldho dha difransiz ar not too menny, dhis iz not dha kase. Difrant fonemez in wun stile ov pranunseashen faul twgeddher in anuddher. A foneme in wun werd may defekt to anuddher. Wot iz kanveneantly seen az wun foneme in wun stile iz better kansiderd too in anuddher.
Copyright © 1996, 1997, 2002 by R.Harmsen, all rights reserved.