After developing the Latin-based EBEO orthography for English, which uses diacritics and Unicode characters extensively, I started wondering about older electronic devices that do not support Unicode. I asked myself: “How would I be able to type phonetically in the most efficient way on these old things?”
This made me think about ways to write English phonetically and efficiently using just the 26 letters of the Latin alphabet. I revisited an older version of EBEO from a while back, BEO (Better English Orthography), and gave it an extensive makeover to make it more suitable for practical use, taking inspiration from an alternate English orthography called SoundSpel to get the most compact form.
Unlike SoundSpel, which contains some exceptions to mimic the existing English spelling conventions without making it more compact (e.g. spelling words ending with /ɔl/ as ‘all’ instead of ‘ol’), BEO provides a more literal way to write English that is easier to understand.
Easy to type – no need to learn a new keyboard layout
Many common English words retain their original spellings, such as ‘man’, ‘red’, and ‘hop’
Many other words only differ by a few letters and are easily recognisable: ‘can’ becomes ‘kan’, ‘simple’ becomes ‘simpul’, and ‘fire’ becomes ‘faiur’
1 letter or digraph represents 1 phoneme, and only that phoneme
Usually more compact than conventional English orthography
Just over a year ago from today, I discovered Quikscript, a minimalist phonetic alphabet for English that was designed by the British designer Kingsley Read. (Quikscript was a revised edition of an earlier and more famous script also by Read, called Shavian.) Quikscript’s most notable feature is that each letter can be written with a single stroke, and some letters can be written together without lifting the pen up, reducing writing effort. If you want to know more about how Quikscript works, including the associated complex ligature system, click here for the original Quikscript manual by Read.
Here’s a chart showing the letters of Quikscript and their phonetic values.
The problem with existing Quikscript fonts
As I delved further into typing Quikscript on my computer, however, I found that the very few Quikscript fonts out there were just plain terrible. For example, here’s the Universal Declaration of Human Rights in Quikscript, using a Quikscript font called ‘Kingsley’:
Note that most of the glyphs have slightly uneven stroke width, squashed-up kerning (glyph spacing) and inconsistent glyph shapes – put together, the whole thing just looks messy and ugly. Damn, even Comic Sans looks way better in comparison!
Continue reading “Quikscript Sans, a sans-serif font for the Quikscript alphabet”→
Southeast Asia is home to many different writing systems, all of which are pretty much descended from the Pallava script originating from southern Asia. After having a go at adopting Thai and Burmese to write English phonetically a while back, I shall now turn my attention to another elegant Southeast Asian script: Khmer.
Khmer script allows consonants to be placed under or around each other, a useful property for writing a language like English with potentially complex syllables like /stɹɛŋθs/, which can be easily rendered as ស្ត្រេង្ឋ្ស in this Khmer script adaptation. (It looks complicated at first, so I’ll get around to explaining it soon)
Shortly after adapting the Kannada script, I took a look at the closely related Telugu script. Both Kannada and Telugu scripts have a common ancestral origin, and have even been said to be typographical variants of the same script. Hence, this adaptation of Telugu is basically my adaptation for Kannada with each letter replaced with their Telugu equivalents.
Having said that, I now prefer using the Telugu adaptation as its vowel diacritics are more compact than Kannada’s. (For example, /noʊ/ in Kannada is ನೋ while Telugu is నో, much more compact indeed)
The ‘Look of Disapproval’ ಠ_ಠ is a popular emoticon on the Internet that also happens to use a letter from the Kannada script. As someone who enjoys reading and learning new things, I found myself reading about Kannada script. As I read more about it, I noticed that Kannada had a lot of letters, like the Thai and Burmese scripts I had adapted previously, and supported consonant clusters too. These are features that, in my view, made Kannada script especially suitable to adaptation for writing languages like English.
Note: This blog uses Unicode Myanmar script. If your default Burmese font is the non-Unicode ‘Zawgyi’, the Burmese text may not display as intended
After creating the Thai script adaptation, I began checking out other Brahmic scripts to learn how they work, and ended up learning a few to see if I could write English with them. Burmese script had a large inventory of letters and diacritics that are arranged differently depending on context, giving it a complex appearance. And because I like complex challenges, I decided that I had to learn Burmese script next.
After having used my adaptation of Hangul for a while, I wondered if other writing systems could be used to write English phonetically. I soon focused my attention on Thai script, with its large inventory of letters and diacritics.
In the original Thai orthography, up to 6 letters were mapped to the same phoneme, but used to represent Thai tones.
Hangul is an alphabetic script widely used to write the Korean language, and was originally developed in 1446 by Joseon Dynasty ruler Sejong the Great and his ministers to improve literacy. As the only widely used script in the world to be based on linguistic features, I thought it would be cool to have a go at adapting it to write English.
English has a very complicated orthography that is inconsistent and troublesome, with its many silent letters such as ‘gh’ in the word ‘though’ (đò) and the ‘b’ in ‘debt’ (dét). And shockingly, some letters can be pronounced differently – ‘gh’ is also pronounced ‘f’ in words like ‘laugh’ (laf) and ‘cough’ (kof)! There’s surely a better way to write it phonetically, isn’t it?
Introducing EBEO (Even Better English Orthography), a revised Latin script orthography that utilises 15 additional diacritic-capped letters and 3 additional letters to expand the traditional 26 letters to cover all 43 English phonemes. Unlike traditional English orthography, each letter represents exactly 1 single English phoneme, and only that phoneme.
It’s not the prettiest orthography out there, but it makes a good starting point for describing any alternate scripts I choose to adapt. It can also be used for transcribing texts written phonetically in other alternate scripts for English, such as Shavian, Deseret and Quikscript.
It uses simple diacritics and simple letter modifications (so you don’t have to learn too many weird letters like Þ)
Supported by just about every professionally designed Latin script font
Each letter represents a single English phoneme, avoiding the ambiguity that can be caused by using 2 or more letters (digraphs) to write a phoneme
Common English diphthongs are also written with 1 letter
More compact than conventional English orthography
Some common English words retain their spelling
It uses diacritics (need to use a special keyboard layout)
Many other common English words will have different spellings (kof kof)
Vowels (11 letters total)
5 vowel letters are not enough to describe the full range of English vowels. To represent 5 of the monophthongal vowel sounds, the acute ´ is placed over a closely related vowel, and the ‘a’ sound in ‘father’ is written with a dot ̇ over the letter ‘A’.
Some of the letters are assigned 2 phonetic values, which sound close enough to be considered variations of the same sound.
Guide: /IPA phoneme/ Letter (example word with letter’s indicated phoneme in bold)
/ʌ/ a (sun)
/æ/ á (can)
/aː/ ȧ (father)
/ə/~/ɜ/ e (sure)
/ɛ/ é (red)
/ɪ/ i (bid)
/iː/ í (bead)
/ɔ/ o (pot)
/ɔː/ ó (call)
/ʊ/ u (pull)
/uː/ ú (pool)
Diphthongs (8 letters total)
Diphthongs are indicated by one out of 2 diacritics, which are used depending on the property of the diphthong:
Rising to an ‘ɪ’ or ‘ə’ sound = ¯ (macron): ai = ā, ue = ū
Falling to an ‘ʊ’, ‘u’ or ‘ə’ sound = ˋ (grave): au = à, iu = ù
/aɪ/ ā (side)
/aʊ/ à (now)
/ɔɪ/ ō (toy)
/oʊ/ ò (dough)
/eɪ/ ē (say)
/ɪə/ è (hear)
/ʊə/ ū (tour)
/ju/ ù (news)
For any other diphthong not defined here, just put the 2 letters of the diphthong in order (example: /ʊɔ/ = uo).
These are the most commonly found 3-vowel sequences in English. The first 2 vowels in the triphthong are indicated by one of the diphthong letters described above.
/aɪə/ āe (flyer)
/aʊə/ àe (tower)
/jʊə/ ùe (cure)
Rhotic vowel sequences
IPA charts for English often include these phonetic vowel sequences as well. This shows how they are represented with EBEO:
/aː(ɹ)/ ar (far)
/ɔː(ɹ)/ or (north)
/ɛə(ɹ)/ ér (chair)
/ɜː(ɹ)/ er (nurse)
All other rhotic vowel sequences (e.g. /oʊɹ/) are represented by appending the letter ‘r’ at the end of the vowel, e.g. /oʊɹ/ = òr.
Consonants (24 letters in use, 26 total)
In addition to 19 existing Latin consonants, there are 7 additional letters to indicate the remaining consonants.
The letters for the consonant sounds in ‘Thank’ and ‘The’ were derived from ‘T’ and ‘D’ respectively as the 2 sounds are closely related, while the ones for ‘Share’ and ‘Azure’ were derived from ‘S’ and ‘Z’ likewise.
The letter ‘C’ is currently used in English orthography to represent 2 other phonemes that are already represented by the letters ‘S’ and ‘K’ depending on context. This is why in EBEO, ‘C’ is used only to represent the ‘Ch’ sound’ in ‘Change’, similar to its pronunciation in Indonesian and Malay.
The letters ĥ and ĝ are not used in English, but can be used for transcribing other languages that do have their sounds (e.g. German).
/p/ p (port)
/b/ b (born)
/f/ f (free)
/v/ v (van)
/m/ m (muse)
/t/ t (tree)
/d/ d (drive)
/θ/ ŧ (thank)
/ð/ đ (the)
/n/ n (new)
/k/ k (call)
/g/ g (get)
/x/ ĥ (loch)
/ŋ/ ŋ (sing)
/s/ s (soon)
/z/ z (zoo)
/ʃ/ ś (share)
/ʒ/ ź (azure)
/tʃ/ c (change)
/dʒ/ j (joke)
/ɹ/ r (run)
/l/ l (laugh)
/h/ h (house)
/w/ w (way)
/j/ y (yell)
Special contractions (1 letter total)
EBEO uses ‘x’ to represent both /ks/ and /gz/, as /ks/ and /gz/ do not contrast with each other and are both very common sequences in English words, particularly those derived from Latin.
/ks/, /gz/ x (box)
It should be noted that ‘x’ is only to be used if the /s/ or /z/ sound does not indicate a plural. Hence, /æksɛs/ (access) is spelt ‘áxés’ and /bɔks/ (box) is ‘box’, but /lɔks/ (locks, plural of lock) is spelt ‘loks’.
The following letters can be used for transcribing particular accents of English.
/ʍ/ ŵ (what)
ŵ is to be used if the 1st consonant in ‘what’ and ‘way’ needs to be distinguished.
Universal Declaration of Human Rights (in my idiolect)
Ùniversel Diklárēśen of Hùmen Rāts
Ol hùmen bíŋs ar born frí ánd ikūl in digniti ánd rāts. Đē ar éndàd wiŧ rízen ánd konśens ánd śud ákt tòerds wan enađer in e spirit of brađerhud.
(Artikel 1 of đe Ùniversal Diklárēśen of Hùmen Rāts)
Excerpt from a short story I wrote a while ago
For comparison, you can view the original one here.
Ā hád e strēnj drím đát nāt.
In đát drím, Ā fànd māsélf ewēkeniŋ, lāiŋ on soft grín gras, in e fántesi 8-bit ūrld serànded bā kompùters. Đe luminens of bliŋkiŋ mòdéms ánd wóm, cèrful ciptùn mùzik fild đi ér. Olđò évriŧiŋ lukd bloki ánd skwér, it brót mi bák tu đòz dēs. Of ol đe kompùters Ā só, 1 of đém ūs plēiŋ mā fēveret soŋ! Ā jamp ánd líp in jō òver đe sāt. Ā đén só mā hàs, ánd Ā séd “Hā” tu mā bést mēts, hu ūr wētiŋ àtsād. Wi wókd tugéđer, háviŋ e cèri cát ebàt đe kompùter gēm Ā ūs ūrkiŋ on erlèr.
“Sò wats đát kul gēm gona bi ebàt, ē?” wan of đém askd. “If ù levd Mariò, ùl lev đis!” Ā séd. “Ósem!!! Kant wēt tu sí it!” Insād mi đe fāer tu kíp mi gòiŋ bikēm stroŋger.
Wi wókd intu e vivid sansét. Ā réminisd đe mémoris of past samers, plēiŋ rétrò vidiò gēms in đe kul śēd, iven đò đe san àtsād píkd át 42 digrís ánd mélted évriŧiŋ éls.
EBEO Keyboard layout
To type EBEO efficiently, I created a keyboard layout that augments the traditional QWERTY layout with the additional accented and stroked letters. The additional letters are accessed by pressing the Right Alt key on your keyboard.
Want to try EBEO on your computer? Here you can download the Windows, Linux and Android (Multiling O) versions I have made.