Welcome to the Alternate Script Bureau! This is a blog dedicated to the various writing systems and alternate orthographies for English and other languages that I have devised, including a phonetic English orthography, and English-language adaptations of non-Latin scripts such as Hangul and Khmer.
What can you do with alternate scripts?
Understand how other writing systems work
Develop an appreciation for different cultures
Write secret notes for yourself
Here’s a video showcasing some of the many writing systems I have adapted:
Arabic and Hebrew are the most well-known right-to-left scripts, but there are actually a few more RTL scripts in widespread use today. One of them is the Syriac script, a distant cousin of the Hebrew and Arabic scripts which is used to write the Neo-Aramaic languages spoken by Christian communities in Iraq and Syria. Like Arabic, Syriac is a cursive script with letters that flow into one another and which uses dots on some letters so that they can be told apart, although to a much lesser degree compared to the oversimplified Arabic letterforms.
Adaptation process
Overall, the consonants were straightforward. However, like Hebrew, some of these letters can have 2 pronunciations: a ‘hard’ plosive sound or a ‘soft’ fricative sound.
/p/~/f/: ܦ
/b/~/v/: ܒ
/t/~/θ/: ܬ
/d/~/ð/: ܕ
/k/~/x/: ܟ
/g/~/ɣ/: ܓ
In English, the plosive sounds are statistically more frequent on the whole than their fricative counterparts, so most of these letters shall get the ‘hard’ plosive pronunciations by default.
/p/ = ܦ (as in ‘port’)
/b/ = ܒ (as in ‘born’)
/d/ = ܕ (as in ‘done’)
/g/ = ܓ (as in ‘get’)
The only exceptions shall be ܟ and ܬ, which will now solely represent the ‘soft’ fricative pronunciation…
/θ/ = ܬ (as in ‘thank’)
As there is another letter we can use for plain /t/! This one, ܛ, represents the incredibly similar /tˤ/ sound and is cognate to Hebrew ט, which has also changed it to a regular /t/: it would be a shame to leave it unused, so let’s repurpose this for /t/ in this adaptation. Likewise for ܩ /q/, let’s make it /k/.
Africa is home to many unique writing systems. One of them is the Ge’ez fidel syllabary/abugida used to write the languages of Ethiopia and Eritrea, which originally developed from a South Arabian offshoot of the highly influential Phoenician consonantal script (abjad) which is ancestral to both Arabic, Hebrew and the Western alphabets. Compared to its ancestral Semitic script, the Ge’ez fidel includes vowels as a core part of its design, theoretically making it much more capable of representing other languages as well.
And now, let’s find out a way to use the Ethiopic script to write English phonetically!
Adaptation process
First though, let’s look at what makes the Ge’ez script an interesting one to work with.
Like other syllabaries such as Japanese kana and Cherokee, a Ge’ez syllable consists of a consonant-vowel (CV) pair. There are 7 vowels, meaning that there are 7 possible syllables for each of the consonants. What makes it unique, however, is that the vowels in syllables are indicated by modifying parts of the original consonantal letter in a largely systematic way, similar in mechanism to the vowel markings in Indic scripts such as Devanagari and Telugu, which is why Ge’ez is nowadays classified as an abugida.
The first syllable with the vowel /-ä/ is represented with the base consonantal letter, while the 6 other syllables for /-u/, /-i/ /-a/, /-e/, /-ɨ/ and /-o/ are indicated by modifying the base consonant in a largely consistent way. Consider the letter for /b/ as an example:
በ /bä/ – the base consonant.
ቡ /bu/ – has a flag diacritic on the right side at the centreline.
ቢ /bi/ – has a flag diacritic on the bottom right corner.
ባ /ba/ – the left side of the consonant is shortened.
ቤ /be/ – has a loop diacritic on the bottom right corner.
ብ /bɨ/ – has a flag diacritic on the left side at the centreline.
ቦ /bo/ – the right side of the consonant is shortened.
This 7-vowel system also happens to be ideal for writing English with only minimal modifications. However, English has a lot of consonant clusters and it would be ideal to indicate the absence of a vowel somehow. For this reason, in this adaptation, the first syllable series shall now solely represent the consonant on its own, without a vowel. Also, while the other vowels can easily be mapped to their English equivalents, /ɨ/ has no equivalent, so I’ll adjust it and make it represent the very common /æ/ sound in ‘hand’ and ‘tap’ instead. Here’s our example with /b/ as a comparison:
Hangul is a truly unique writing system in that it was created by humans almost from a blank slate, instead of evolving from an older pre-existing script. The script was created by King Sejong and his ministers of the Joseon dynasty of Korea in 1446 to improve literacy, being designed to represent the Korean language in a far more intuitive way than the pre-existing hanja (Chinese characters)-based system ever could. Letters for related sounds are derived from base letterforms in a highly systematic manner, making Hangul well-known for being incredibly easy to learn. In theory, this would also make the script capable of representing every possible sound in every human language, including English.
But what would it take to write English with Hangul in the most concise way possible?
To find out, I had a look at the archaic Hangul letters (yethangul, 옛한글), which have a much wider selection of letters and letter combinations for representing Korean and Chinese as spoken in the 1400s. These include consonant clusters such as ᄹ /st/ and letters for some fricatives like ᅗ (/ɸ/) and ᄫ (/β/), which will come in handy for writing English sounds that do not exist in Korean. Unicode offers full support for these archaic letters via the ‘Hangul Jamo’, ‘Hangul Jamo Extended-A’ and ‘Hangul Jamo Extended-B’ character blocks. The only catch? Korean language input methods generally do not support these characters, but we’ll get to that in a moment.
Also, this isn’t the only time that someone’s tried writing English with Hangul – plenty of other individuals have had a go at adapting Hangul for English, but this adaptation is the only one out there which is based on how Hangul is actually used in the real world and not as a substitution cipher for the alphabet or something else entirely, which is why I’ll refer to my adaptation as ‘Hangul for English: The Real Deal’, or HETRD for short!
Note: This article contains archaic Hangul letters which will not display properly without fonts that support it well, such as Noto Sans CJK KR or Noto Serif CJK KR.
Adaptation process
Adapting Hangul for English was a notable challenge, since it was originally not designed to write languages with lots of fricatives in them, such as English. However, with a bit of work, it’s more than capable of representing English in a phonetically consistent manner.
A Hangul syllable is written in a single square block in a similar manner to Chinese characters, which can be divided into 3 parts: the initial consonants (I) on the top left side, medial vowels (M) on either the right or centre, and the final consonants (F) at the bottom. Here’s a picture depicting the ways in which letters can be arranged:
In Korean Hangul orthography, there are 19 initial consonants, 21 medial vowels and 28 final consonants. That seems like a lot of consonants to play around with, until one realises that:
Korean Hangul does not have letters for the following sounds: /f/ (as in ‘fun’), /v/ (as in ‘van’), /θ/ (as in ‘thing’), /ð/ (as in ‘the’), /z/ (as in ‘zoo’), /ʃ/ (as in ‘share’) and /ʒ/ (as in ‘vision’).
Korean orthography does not distinguish between the /ɹ/ and /l/ sounds: ㄹ is pronounced with an /l/ sound only at the end of a syllable, and /ɹ/ elsewhere.
Some of the consonants cannot be final consonants.
Consonants
As usual, the majority of Korean Hangul consonants can be assigned rather easily to their English counterparts: ᄏ /k/, ᄃ /d/, ᄉ /s/, ᄀ /g/, ᄒ /h/ and so on. Likewise, if there’s no initial consonant, the ieung null consonant letter is used as a filler: ᄋ = /∅/. Hence, the syllable /a/ on its own is 아, which is the vowel for /a/ ᅡ with the ieung ᄋ before it.
For both bilabial fricatives /f/ and /v/, let’s co-opt the archaic letters for /ɸ/ ᅗ and /β/ ᄫ respectively, since these sounds are incredibly similar to each other, almost indistinguishable to an untrained ear. They also pair nicely with the corresponding stop sounds /p/ ᄑ and /b/ ᄇ, being basically these letters plus the bottom circle of the /h/ letter ᄒ.
The Hebrew script has a blocky and angular aesthetic that makes it easy to recognise, being the only other widely used right-to-left script in the world besides Arabic. In fact, this relationship is not just co-incidental, as both the Hebrew and Arabic scripts share the same origins: both were once variants of the Aramaic script, which was once widely used across the Middle East a few thousand years ago. Since then, Hebrew has been adapted for writing the wide variety of languages spoken by the Jewish diaspora – but what would it take to write English with it in the most optimal manner? Let’s find out.
Since many others have also had a go at using Hebrew to write English, I’m going to call this ‘Yet Another Hebrew English Alphabet’, or YAHEA for short.
Adaptation process
Most of the Hebrew consonant letters have clear English equivalents: /s/ ס, /m/ מ/ם, /ɹ/ ר, etc.
However, some of these letters can have 2 pronunciations depending on whether a vowel sound occurs before them: a ‘hard’ plosive sound or a ‘soft’ fricative sound.
/p/~/f/: פ/ף
/b/~/v/: ב
/t/~/θ/: ת
/d/~/ð/: ד
/k/~/x/: כ/ך
/g/~/ɣ/: ג
In English, the plosive sounds are statistically more frequent on the whole than their fricative counterparts, so the letters פ – ב – ד – ג shall get the 1st pronunciations:
The Arabic script, one of the most widely used writing systems in the world, is easy to recognise due to its very distinctive aesthetic: connected letterforms of swishes, lines, curves and loops, many of which are interconnected to each other in a stream of sorts, with dots above or below to distinguish each distinct letterform. Unlike the Western alphabets, Arabic script runs from the right to the left and only tends to represent consonantal sounds, leaving the vowels largely unwritten, just like the Aramaic script which was once widely used across the Middle East a few thousand years ago and from which both Arabic and Hebrew scripts have descended from. Although it had originated in Arab-speaking regions, it has since then been adapted to write a wide variety of languages such as Persian (Farsi), Urdu and Malay.
Having said that, what would it take to make a version of Arabic for English? Let’s find out.
Since many others have also had a go at using Arabic to write English, I’m going to call this ‘Yet Another Arabic English Alphabet’, or YAAEA for short.
Adaptation process
Most consonant letters were fairly straightforward: ف = /f/, س = /s/, ر = /ɹ/, etc. Unlike many of the other writing systems out there, it also has distinct letters for /θ/ ث and /ð/ ذ.
Indonesian is the standardised variety of the Malay language that is widely spoken across the Indonesian archipelago. Given that Malay was once written in the Indic Kawi script widely used across Maritime Southeast Asia from the 8th to the 16th century, it’s actually straightforward to write it in the Javanese script, which is essentially a highly cursified and stylised form of Kawi, since Malay shares the same set of key phonemes and phonotactics as the Javanese language and many of the letters therefore can be used as-is with only some minimal adjustments.
Almost all consonants in Indonesian can be easily assigned to their corresponding Javanese script letter: /r/ ꦫ, /p/ ꦥ, /t/ ꦠ, /dʒ/ ꦗ, etc.
This leaves just /f/, /v/, /x/ and /z/. These consonants were not originally part of the indigenous Malay phoneme inventory and hence are found only in loanwords from other languages that do have these sounds. Traditional Javanese just slaps a 3-dot cecak telu diacritic ꦳ on ꦥ, ꦮ, ꦏ and ꦗ respectively, but I think that this is not needed because there are actually some surplus murda letters which once represented distinct sounds in Sanskrit, some of which sound really close to the 4 aforementioned consonants and can be re-assigned as follows (IPA indicating their original Indic pronunciation):
ꦦ /pʰ/ (murda /p/) -> f /f/
ꦨ /bʱ/ (murda /b/) -> v /v/
ꦑ /kʰ/ (murda /k/) -> kh /x/
ꦣ /ɖʱ/ (murda /d/) -> z /z/
There also is a variant of /k/ ꦏ with a mark below that is used to transcribe the Latin letter ‘q’ – in normal pronunciation, it is pronounced identically to ꦏ.
ꦐ -> q /k/
When the /ŋ/, /r/ and /h/ consonants end a word, they are written with their own syllable-final consonant diacritics: cecak ꦁ, layar ꦂ and wignyan ꦃ respectively.
After having used my other alternate English orthography YAAEO for quite a while, I wondered if there was a way to have the vowel letters represent English phonology accurately without re-adjusting the pronunciation of ⟨A⟩ and ⟨U⟩ as I had done in YAAEO, which was one of my biggest gripes with that scheme. This was the end result, and this of course means giving it a funky new name to distinguish it from the old! I’m just gonna call this the Random Dotless I Orthography, or RDIO for short, since its most prominent feature is the ‘dotless I’ and I’m aware that English spelling isn’t going to be reformed anytime soon, so this is just a fun and random take on how I feel it should be done.
As with YAAEO, the key design principle was efficiency. It uses just 1 additional diacritic-capped vowel letter (À), 1 modded vowel letter (the ‘dotless I’, I/ı) and 5 additional consonant letters (Ț, Ð/þ, Ŋ, Ș and Ż) to complement the existing 26 letters of the alphabet and allow them to be able to cover all of the English phonemes with minimal ambiguity. And a whole bunch of highly opinionated stuff that not everyone might agree with – but we’ll get to that shortly.
Since this shares a lot of principles with YAAEO, most of the text here has been copied over from my YAAEO article.
Devanagari is the most well-known script of the Indian subcontinent, tracing its roots to a northern form of the ancient Brahmi script. Originally developed to write manuscripts in the liturgical Sanskrit language, it has since been adapted for many widely spoken South Asian languages such as Hindi and Nepali.
But what would it take to get it to write English phonetically in the most efficient manner? Let’s find out.
Adaptation process
Making Devanagari work for the English language was a straightforward process, as there are many consonants which have clear English equivalents: /n/ न, /ŋ/ ङ, /s/ स, /ɹ/ र, /h/ ह, and so on.
There’s some differences, of course. Devanagari has separate letters for aspirated, palatal retroflex and breathy voiced consonants, whereas English doesn’t have retroflex or breathy voiced consonants, but quite a bit of fricatives like /f/, /v/, /θ/, /ð/, /z/, /ʃ/ and /ʒ/. In this adaptation, I used some of these extra consonant letters to represent the English fricatives and affricates, using them in such a way that they are mapped to the most closely related sound where possible:
फ /pʰ/ -> /f/
भ /bʱ/ -> /v/
ट /ʈ/ -> /θ/
ड /ɖ/ -> /ð/
ध /dʱ/ -> /z/
च /c/ -> /tʃ/
ज /ɟ/ -> /dʒ/
ढ /ɖʱ/ -> /ʒ/
Also, while व has a sound more closely corresponding to /v/ in Hindi today, in this adaptation it shall represent /w/, which is how it was originally pronounced. भ /bʱ/ shall be used to represent /v/ instead, as it is the voiced counterpart of फ /pʰ/ which is used for /f/ in this adaptation, allowing them to ‘pair up’ nicely.
Vowels were a little more complicated situation. English has more consonants than the traditional Devanagari script as used for Sanskrit, but since Devanagari has been adapted for other languages as well, there are some extra vowel marks and letterforms we can use.
There’s the ‘chandra E’ ॅ for /æ/ as in words like ‘can’ and ‘hand’, and the ‘OE’ diacritic ऺ for the schwa /ə/ in words like ‘comma’ – let’s use both. For the /ɔ/ vowel in ‘hot’ and ‘tall’, I’ll use ‘chandra O’ ॉ whereas the plain ‘O’ ो shall be used for /oʊ/~/əʊ/ as in ‘code’ and ‘bold’, since its original Devanagari pronunciation /oː/ is closer to the /oʊ/ sound.
The visarga diacritic ः, which represents final /h/, has been repurposed to represent some diphthongs without their own diacritics such as /eɪ/ and /ɪə/.
The anusvara dot diacritic ं is used to represent final /ŋ/ at the end of a syllable, e.g. /kɪŋ/ = किं.
As with my previous abugida adaptations on this blog, I won’t be using an inherent vowel since it only adds an extra layer of complexity due to the fact that English has a lot of words ending in consonants/clusters.
Note: this is an updated version of the previous Thai script adaptation I did a few years ago. This new edition aims to more fully utilise all the available vowel diacritics and remaining Thai consonants that were left unused initially.
The Thai script was originally derived from an earlier version of the Khmer script to better fit the properties of the Thai language while retaining its ability to represent Sanskrit and Pali, in which Buddhist texts were written. Today, Thai script is well-supported on just about every modern computing device, so I wondered if I could use Thai script to write English efficiently. The end result repurposes ‘surplus’ Thai consonants to represent English consonant clusters and is reasonably compact in length while also being hard to read for those who do not understand this complex entanglement of consonants, consonant clusters and vowel diacritics. Hence, this would probably be the ideal script for keeping your secrets safe from prying eyes – that is, as long as the ‘prying eyes’ are not literate in Thai.
Adaptation process
In the Thai orthography, a single consonantal phoneme can be represented by up to 6 letters, all of which are assigned a particular class that is used to indicate a Thai tone. Many of these letters represent Indic consonants which do not exist in the Thai language (such as palatal, retroflex and breathy voiced consonants), as well as distinct Thai sounds which had since merged with other consonants. For example, the following Thai letters are all pronounced (albeit with different tones) /tʰ/: ฑ, ฒ, ฐ, ถ, ท and ธ. Likewise for /kʰ/: ข, ฃ, ค, ฅ, and ฆ. And for /pʰ/: ผ, พ, and ภ. There are a few more examples of such instances throughout the Thai script.
I utilised the extra letters to represent English sounds not present in Thai, with a particular focus on assigning the more simple letters to core English consonants. If the target sound doesn’t exist in Thai (e.g. /ð/), I’ll pinch one of the simpler letters from any large collection of letters with a similar pronunciation. For example, the letter for /θ/ (the ‘th’ sound in ‘teeth’), ท, was taken from the above group of /tʰ/ letters. (In this adaptation, ถ is used to represent the /t/ sound.) Likewise for /ð/ (the ‘th’ in ‘this’), which shall be ธ, also from the /tʰ/ letter group.
The vowels were more straightforward as the common ones could be easily mapped to their Thai equivalents. Some things to note:
I elected not to use แ for /æ/, since it takes up more space than using an existing tone mark.
The ‘maitaikhu’ diacritic อ็ is used to represent /ɔ/ (the ‘o’ in ‘pot’).
The vowel sign for Thai /uː/ อู is used to represent the very common English diphthong /oʊ/ (the ‘ough’ ‘in ‘dough’). This was based on the equivalent Khmer script vowel diacritic អូ, which is also pronounced as /ou/ for certain consonants.
Instead, English /uː/ shall be written as /ʊ/ อุ with the ‘mai ek’ tone marker อ่ like this: อุ่.
Thai script has 2 diacritics for /aj/: ไอ and ใอ. In this adaptation, ไอ is used for /aɪ/ while ใอ is used for /eɪ/.
Recently, I created a personal blog called ‘Glory to Randomness‘. It might not sound like much, especially with the pretentious name, but I’ve decided to write all articles in my RDIO alternate English spelling scheme just for fun – I’ll be posting articles about topics I find interesting every now and then. If you’ll like to see how a more consistently spelt English would look like, feel free to check it out.
Alternate Script Bureau 