Home Contact GitHub

Iridium Spelling

Iridium spelling

á rə̀t ðis pə̀st tù əksplén má sistəm fə rátiŋ iŋgliʃ yùziŋ ə fənemik (ofən kōld fənetik) skript. dí igzakt fōm əv ðə sistəm həz dəveləpt ənd cénjd ə̀ve mʌltipəl yīz, bət it həz ōlwéz bín ə komprəmáz bətwín ðə simbəlz yùzd tə rát fənemik ingliʃ in ðí əfiʃəl intənaʃənəl fənetik alfəbet (/IPA/), ən ðə simbəlz yùzd kənvencnəlí in standəd ingliʃ speliŋ. má mʌðə tʌŋ iz nonrə̀tik britiʃ ingliʃ frəm ístən iŋglənd (kémbrijʃī), sə̀ ðə sistəm əz á yùz it wə̄ks best fə ðis dáəlekt əv ingliʃ

Standard English spelling

I wrote this post to explain my system for writing English using a phonemic (often called phonetic) script. The exact form of the system has developed and changed over multiple years, but it has always been a compromise between the symbols used to write phonemic English in the official International Phonetic Alphabet (IPA), and the symbols used conventionally in standard English spelling. My mother tongue is non-rhotic British English from eastern England (Cambridgeshire), so the system as I use it works best for this dialect of English.

The alphabet

I have two different sort orders:

By consonant/vowel, then by placement:

pbmwfvθðtdnszʃʒcjrlykgŋhaāáàiīíuūùeēéoōóòʌəə̄ə̀

Extended alphabetical

aāáàbcdðeēéfghiīíjklmnŋoōóòprsʃtθuūùvwyzʒəə̄ə̀ʌ

Why

English spelling is known for being difficult and taking a long time to learn both for native speaker children and non-native adult learners. This is mostly because there is a many-to-many mapping between symbols (letters) and sounds (phomemes). In English, if you come across a new word you don’t reliably know how it should be pronounced. For example should the made-up word “tead” be pronounced to rhyme with “mead” or with “head”? In the other direction, if you hear a new word, you can’t necessarily tell how it is spelled. For example should the made-up word pronounced /dait/ (rhymes with “write”) be spelled ‘dite’, ‘dyte’, or ‘dight’?

If you grew up only speaking English, it’s not obvious that spelling doesn’t have to be so complicated. English spelling is actually uncommonly complex, and has inspired a long tradition of people trying to ‘fix’ the problem – it’s even got its own wikipedia page: https://en.wikipedia.org/wiki/English-language_spelling_reform. Many languages have a much more straightforward mapping between letters and phonemes. Spanish and Polish are examples of languages with high degrees of correspondence between letters and phonemes. See https://en.wikipedia.org/wiki/Phonemic_orthography for a longer list of languages and more detail.

I use this system for almost all of my handwritten text (mostly to-do lists and journaling), and increasingly for my typed text. The reasons I like using this system:

it’s fun to write
it’s not obvious to most other people what is written, especially at a glance
it increases my awareness of the actual sounds in English words and their relative frequencies

Other reasons why you might like this system:

it’s not that hard to learn to read or write
using this script, if you hear a word you can write it down, and if you see a word written down, you know how to say it

My main criteria for this spelling system were:

a one to one phoneme (sound) to symbol (letter) mapping – each letter represents one phoneme, each phoneme is represented by one symbol only, this ensures that if you hear something, you know how it’s written, and if you see something written down, you know how it’s pronounced
avoid digraphs as much as possible, but potentially allow them for diphthongs (TODO: explaind diphthong) and triphthongs because otherwise there are just too many unique symbols
- a digraph is when two symbols together are used to indicate one sound, e.g in standard English spelling “sh” is a digraph which represents the single phoneme /ʃ/
- the reason to avoid digraphs is because they can cause ambiguity when reading a word, for example, it’s not always possible to tell if “th” represents the “th” sound, or a “t” sound followed by an “h” sound
if there is a pretty consistent or intuitive single character used to represent a phoneme in English spelling already, keep that in the new system, this makes it more familiar and intuitive
if there isn’t an obvious or intuitive single character already used in English spelling to represent a given phoneme, use the standard IPA symbols (e.g. when replacing the digraph consonants of English “sh”, “ch”, etc)
minimise use of characters from outside the standard English alphabet. Unfortunately this isn’t entirely feasible as English has ~37 phonemes even if you discount diphthongs, so I can’t fit all those phonemes into the 26 letters of the standard English alphabet
allow for an ASCII version which can be automatically converted to the proper symbols so it can be typed without a special keyboard layout
easy to write by hand

Letter choices:

Here I’ll go through the phonemes of English group by group to slowly build up from more familiar to less familiar symbols.

Step 1: (non-nasal stops and easy short vowels)

p, t, k, b, d, g, a, i, u, e, o

five short vowels:

a: kat, bat (cat, bat)
i: big, gig (big, gig)
u: gud, kuk (good, cook)
e: get, bed (get, bed)
o: dog, got (dog, got)

stop consonants:

p: pat, ap, pet (pat, app, pet)
t: tap, tik, bat (tap, tick, bat)
k: kat, akt, bak (cat, act, back)
b: bad, bet, dab (bad, bet, dab)
d: dig, dek, gud (dig, deck, good)
g: get, eg, dog (get, egg, dog)

Step 2: (nasals and difficult short vowel)

ʌ, m, n, ŋ

short vowel: ʌ, as in:

kʌt, bʌg (cut, bug)

nasals:

m: met, dam (met, damn)
n: not, tʌn, tent (not, tonne, tent)
ŋ: soŋ, nitiŋ, baŋ, baŋk, piŋk (song, knitting, bang, bank, pink)

Step 3: (liquids)

r, l, y, w

liquids:

r: rant, rip, rat, rʌn, roŋ (rant, rip, rat, run, wrong)
l: log, lent, bul, mil (log, lent, bull, mill)
y: yet, yel, yap (yet, yell, yap)
w: wot, wen, wud (what, when, would)

Step 4: (schwa)

The schwa: ə, this vowel is found all over English in unstressed syllables, it’s sortof a nothing placeholder vowel.

indefinite article:

get ə kat (get a cat)
got ən ap (got an app)

syllabic nasals and liquids:

lemən, bʌtən, kitən (lemon, button, kitten)
botəl, padəl, pikəl (bottle, paddle, pickle)

unstressed syllable nucleus:

wintə, anəməl, kamrə (winter, animal, camera)

Step 5: (long vowels)

ā, ī, ū, ē, ō, ə̄

These are the lengthened versions of the 5 ‘cardinal’ short vowels we encountered in Step 1, plus a lengthened version of the schwa vowel.

ā: dān, māk, ām, fā (darn, mark, arm, far)
ī: ī, bīd, nī (ear, beard, near)
ū: kyū, lū (cure, lure)
ē: lē, ē, kē (lair, air, care)
ō: bōd, dō, wōp (board, door, warp)
ə̄: wə̄d, tə̄n, bə̄d, gə̄l (word, turn, bird, girl)

Step 6: (easy fricatives)

f, v, s, z, h

fricatives:

f: fām, grafik, lāf, hāf, sofən (farm, graphic, laugh, half, soften)
v: vent, hav, sə̄v (vent, have, serve)
s: sit, send, dāns, mes, hōs (sit, send, dance, mess, horse)
z: zip, fʌz, haz (zip, fuzz, has)
h: hed, hāt, hī (head, heart, hear)

Step 7: (diphthongs ending with i)

í, é, á, ó

there are four diphthongs ending with the i sound, see ‘English after RP’ for explanation of why í is here if confused

í: líf, fíd, sití, síliŋ (leaf, feed, city, ceiling)
é: tébəl, vén, sé, dézí (table, vein, say, daisy)
á: dáv, dá, táp, bá (dive, die, type, buy)
ó: tó, kón (toy, coin)

Step 8: (affricates)

c, j

affricates:

c: cop, cíz, bac (chop, cheese, batch)
j: jʌmp, majik, brij, əjʌst (jump, magic, bridge, adjust)

Step 9: (diphthongs ending with u)

ù, ə̀, à, ò

there are four diphthongs ending with the u sound, see ‘English after RP’ for explanation of why ù is here if confused

ù: bùt, trù, jùs, flùt, lùz, sùp (boot, true, juice, flute, lose, soup)
ə̀: gə̀, fə̀, bə̀t, yelə̀, mə̀v, sə̀ (go, foe, boat, yellow, mauve, sew)
à: nà, àl, əlàd (now, owl, aloud)
ò: còl, òld, sòl, bòl (coal, old, soul, bowl)

Step 10: (difficult fricatives)

θ, ð, ʃ, ʒ

θ and ð are mostly ‘th’ in Standard English spelling, but they’re actually two separate phonemes ʃ is mostly ‘sh’ in Standard English spelling, ʒ has no particular consistent spelling

fricatives:

θ: θiŋ, fōθ, sinθ (thing, fourth, synth)
ð: ðə̀, síð (though, seethe)
ʃ: ʃin, ə̀ʃən, məʃín, speʃəl, ʃugə (shin, ocean, machine, special, sugar)
ʒ: ʒonrə, béʒ, leʒə (genre, beige, leisure)

Extras:

Step 11: (“triphthongs”)

éə, áə, óə, ùə, àə, ə̀ə

I write the “triphthongs” using the relevant diphthong symbol plus the schwa.

éə: léə, pléə (layer, player)
áə: láə, fáə, táə, háə (lyre, fire, tire, higher)
óə: lóə, implóə (lawyer, employer)
ùə: fyùə, brùə (fewer, brewer)
àə: àə, flàə, pàə, (hour, flour, power)
ə̀ə: lə̀ə, mə̀ə (lower, mower)

Step 7: more schwa

TODO

Notes on the vowels

Vowels are a bit harder because in standard English spelling there is an extremely poor mapping between vowel phonemes and the letters used to spell them (see https://aeiouy.onlyskin.dev), there are so many different options to represent a given vowel phoneme, which often using digraphs, patterns split out into different parts of the word, patterns which combine with consonant letters, etc. There are also only really five “vowel” symbols in English spelling, while phoneme-wise, there are ~13 pure vowels and ~8 diphthongs.
Diphthongs I chose to write using diacritics. I really like how this is a featural aspect of my writing system: you can see whether a diphthong ends in ‘i’ or ‘u’ from the direction of the diacritic. In combination with using a macron diacritic for all long pure vowels, you can also see the feature of syllable nucleus weight from the presence of the diacritic.

Vowel table:

short	long	opening	closing
a (had)	ā (lard)	á (buy)	à (cow)
i (hid)	ī (leer)	í (bee)
u (hood)	ū (lure)		ù (coo)
e (head)	ē (lair)	é (bay)
o (odd)	ō (lore)	ó (boy)	ò (coal)
ə (uhuh)	ə̄ (lurk)		ə̀ (cone)
ʌ (bud)

Writing in ASCII

This system uses quite a few symbols which aren’t on any standard keyboard layouts, so I have a system for typing it using ASCII characters. The ASCII version can then be run through a program which simply find/replace’s the characters. It works pretty well. I’ve also created a custom keyboard layout on MacOS, where I have mapped mostly the same keys/key combinations as the ASCII version to output the correct characters.
- ʃ as S
- ʒ as Z
- θ as T
- ð as D
- ŋ as N
- indicate the long vowel by adding a colon after the vowel: a:, i:, u:, e:, o:
- indicate opening diphthong with a following i and closing diphthong with a following u, e.g. sei, huu
- ɜ type as numeral 3
- ʌ type as ^ (caret)
- ə type as ; (the reason for this is because you hardly ever use a semicolon, it looks unintrusive in the pre-conversion typed text, and it’s on the keyboard’s home row which is important because the schwa is so common)

Other notes

I put any standard English Spelling I want to embed (e.g. for names, etc.) between <> or //, as in “á spə̀k wið má frend /Chloe/ tədé”
If there is ever any ambiguity about which syllable a sound belongs to, I distinguish a syllable boundary with a “.”, for example in the compound noun “pricetag”, without a “.” would be “praistag”, but with is “prais.tag”, which makes it clear that the s is part of the first syllable coda
I reduce schwas to an apostrophe in certain places, but it’s mostly just stylistic
- for a syllabic nasal/liquid consonant (e.g. “button” bʌtən -> bʌt’n or “scrabble” skrabəl -> skrab’l)
- shortened forms of grammatical words/prepositions, which is similar to in normal English (he’d’ve gone = hídəv gon -> híd’v gon, mightn’t we = mátənt wí -> mát’nt wí), just note that the apostrophe in my spelling simply represents the schwa vowel, whereas in standard English it represents an omission or contraction, so I don’t use them for “I’m” = ám, “you’re” yō etc.
allomorphs – in standard English spelling, the plural morpheme is always written with an “s”, but it is actually pronounced in several different ways depending on the final consonant of the word to which it is attached, this is called allomorphy. In my system, I write allomorphs as they sound, which is how most languages with phonemic writing systems do it.
- the noun plural can be with ‘s’, ‘z’, or ‘iz’, so for cats, dogs, and horses -> kats, dogz, and hōsiz
- same for the genitive ‘s’ as in “my friend’s house”
- the past tense can be with ‘t’, ‘d’, ‘id’, so for walk, wave, and wade -> wōkt, wévd, and wédid
- the article “the” is pronounced differently depending on the following word: the apple, the pear are ðí apəl, ðə pē
- the article “a” already reflects its allomorphs in standard English spelling – “a” vs “an”: an apple, a pear: ən apəl, ə pē
- it’s also useful to be aware of when using this system, that the citation/isolation form of many function words (e.g. prepositions, auxiliary verbs) is the strong or stressed form, so when you write them it will often look different, usually the vowel will reduce to schwa e.g.:
  - the word “can” -> ðə wɜd “kan”
  - I can speak English. -> á kən spík iŋgliʃ
  - full list: https://en.wikipedia.org/wiki/Stress_and_vowel_reduction_in_English#Weak_and_strong_forms_of_words
    - always reduced: a, an, and, be, been, but, he, her, him, his, just, me, or, she, than, that (as conjunction), the, them, us, we, who, you, your
    - reduced except at end of sentence: as, at, for, from, of, to, some, there (to also doesn’t reduce before vowel)
    - reduced, except at end of sentence or contracted with not: am, are, can, could, do, does, had, has, have, must, shall, should, was, were, will, would.
web font: Noto Sans Display or Noto Serif

Variations

hyphens to join clitic grammatical words to other words e.g. “the apple” -> ðí-apəl, “the pear” -> ðə-pē, “an apple” -> ən-apəl, “a pear” -> ə-pē, “to him” -> tə-him, “for me” -> fə-mí, “to the shops” -> tə-ðə-ʃops
diphthongs second later as y/w: sé => sey, sà => saw

100 most common English words in Iridium spelling (this will be formatted nicer)

come – kʌm
get – get
give – giv
go – gə̀
keep – kíp
let – let
make – mék
put – put
seem – sím
take – ték
be – bí
do – dù
have – hav
say – sé
see – sí
send – send
may – mé
will – wil
about – əbàt
across – əkros
after – āftə
against – əgenst
among – əmʌŋ
at – at
before – bifō
between – bətwín
by – bá
down – dàn
from – from
in – in
off – of
on – on
over – ə̀və
through – θrù
to – tù
under – ʌndə
up – ʌp
with – wið
as – az
for – fō
of – ov
till – til
than – ðan
a – ə
the – ðə
all – ōl
any – ení
every – evrí
no – nə̀
other – ʌðə
some – sʌm
such – sʌc
that – ðat
this – ðis
I – á
he – hí
you – yù
who – hù
and – ən
because – bikəz
but – bʌt
or – ō
if – if
though – ðə̀
while – wál
how – hà
when – wen
where – wē
why – wá
again – əgen
ever – evə
far – fā
forward – fōwəd
here – hī
near – nī
now – nà
out – àt
still – stil
then – ðen
there – ðē
together – təgeðə
well – wel
almost – ōlmə̀st
enough – inʌf
even – ívən
little – litəl
much – mʌc
not – not
only – ə̀nlí
quite – kwát
so – sə̀
very – verí
tomorrow – təmorə̀
yesterday – yestədé
north – nōθ
south – sàθ
east – íst
west – west
please – plíz
yes – yes

Questions and tricky ones:

He’d (híd or híəd or hīd)
feel (fíəl or fíl), fail, field, etc
we’re (wī or wíə)

Notes for other dialects

this system would need a lot of adaptation for someone who speaks a rhotic dialect
the class vowel