DubstepKazoo's Dumping Ground o' Random Shit

Shavian

Background

Back in the 1900s, there was a British playwright named George Bernard Shaw, known for penning such works as Pygmalion and Androcles and the Lion. He was also known for absolutely hating English spelling, often ranting about how it needed to be fixed, specifically with a new alphabet altogether, since he considered Latin-script orthography unsalvageable. He was so passionate about this that he set aside over 100,000 British pounds in his will to fund the creation and proliferation of a new alphabet with which to write English, even stipulating a few requirements for it: one letter per sound, one sound per letter, one stroke per letter, and no diacritics (more on this one later), among others.

Unfortunately, Shaw forgot one simple rule of capitalism: when a rich person dies, the law starts applying to them. And the law is whatever living rich people want it to be. Shaw had also named in his will recipients of whatever of his money was left over after his alphabet project, and those organizations weren't content with having just some money; like any good capitalist, they wanted all of the money. So they sued to contest the will, and they won. Suddenly, there was only a tiny portion of money remaining to fund a new alphabet.

With that money, a trust was set up and a competition held to design this new alphabet. The judges narrowed it down to four entries, but found that none of them individually were entirely satisfactory. Of them, the closest was an entry by a man named Kingsley Read, who had been close to Shaw in life, so they tasked him with combining and refining the four winning entries into a final product. What he finally submitted in 1962, twelve years after Shaw's death, was the new alphabet named after its sponsor: Shavian, which would make its debut to the world in a publication of Shaw's Androcles and the Lion, with Shavian on the left-hand pages and the corresponding Latin spelling on the right. Working to Shavian's advantage was the fact that it's quite compact: in order to ensure the two scripts contained the same text on each page, the column of Shavian text had to be two thirds the width of the Latin column.

Design

First of all, Read yeeted the concept of "uppercase" and "lowercase" out the window. Only an insane person would intentionally design an alphabet where each letter had two forms that made all the same sounds but were meant to be used in different contexts. We don't know a sentence is beginning because it starts with a capital letter; we know a sentence is beginning because the sentence is beginning. Likewise, there's no reason the word "I" needs to be uppercase. The closest thing uppercase letters have to a valid use case is names, but even then, we get by just fine in spoken language, and Read didn't see that as a problem you needed an entirely separate script to solve. He determined that names could be easily marked by a simple dot preceding them.

Next, he took inspiration from the design of lowercase Latin script. He noticed that letters could be divided into three categories that he called Tall (e.g. "d"), Deep (e.g. "p"), and Short (e.g. "a"). But their distribution was largely random; he thought he could put these categories to better use. Shavian's Tall letters are (mostly) voiceless consonants, while Deep letters are (mostly) voiced consonants. Finally, Short letters are vowel sounds and a handful of consonants that kinda exist outside the voiced/voiceless dichotomy, those being /l/, /r/, /m/, and /n/.

Another major feature of Shavian is that nearly every letter is either a 180-degree rotation or a mirror of another. In the case of the consonants, this is meant to show their relations to each other. The Shavian letters for /t/ and /d/ are 𐑑 and 𐑛 to reflect that the only difference between the two sounds is their voicedness; otherwise, your mouth is doing the exact same thing to produce both of them. I respect the hustle, and it's an elegant idea on paper, but I think this was a bit of a misstep. People with dyslexia and other reading impairments already struggle with letters like p, d, b, and q; extending that to the entire alphabet would make things even worse.

Spelling

Some Short letters are compounds of other letters, mostly for rhotic vowels. Read recognized that even though speakers of non-rhotic accents of English would pronounce, say, "law" and "lore" identically in isolation, they'd still pronounce them differently if the following word started with a vowel sound. They really are two different phonemes, even in non-rhotic English, so they really do need to be spelled differently to reflect that.

But what about phonemic distinctions that are in some accents but not others? As I mentioned on the landing page of the Spelling Reforms section of this site, my accent lacks the LOT vowel, despite its incredible abundance in plenty of other accents. I also have the MARY-MERRY-MARRY merger and the WINE-WHINE merger, meaning that I pronounce the words in those sets exactly the same as each other, even though other people pronounce them distinctly. This is the biggest problem that faces anyone trying to design a pronunciation-based reform to English spelling.

There's always the option of just letting people spell according to their own accent, but Read didn't like that idea. He felt that each word should be immediately recognizable to the reader, and variant spellings made that harder. So his solution was to come up with a set of standard spelling rules that reflected no single extant accent, but was instead a compromise between the major ones. The WINE-WHINE split is relatively rare, so he merged them, but the FATHER-BOTHER and COT-CAUGHT splits are common outside of General American, so he kept them distinct.

As a result, basically everybody is going to have to learn some kind of distinction in order to spell in accordance with the standard. Speaking as a GenAm speaker, I actually find it pretty easy to tell when to use LOT; I just imagine how a Brit would pronounce any given word. What I struggle with is the distinction between rhotic vowels and vowels followed by /r/. What do you mean, "horror" is 𐑣𐑪𐑮𐑼? I pronounce it as 𐑣𐑹𐑼. Or, you know, you could do what most people in the Shavian Discord server do, and spell everything however the hell you want to.

All in all, I think Read struck a pretty good compromise in his spelling rules, landing on a system that's basically Received Pronunciation with rhotic vowels. He even had the great idea to stipulate that the words "to," "the," "and," "for," and "of" never be spelled fully, but instead with just the first consonant in them, saving significant time and space in writing. He also specified that "a" and "an" are always spelled with a schwa instead of the vowels those words would have when stressed, but all other "little words" are spelled as if they were stressed.

...Which brings me to one of Shavian's biggest flaws, though it's not one that Read could've done anything about. The blame for this one lies squarely on Shaw. Stress in English is phonemic; it's an important distinction between words. In fact, it's the only distinction between "insight" and "incite." Therefore, in a phonemic spelling system like Shavian's, it's important to clearly mark stress... But that would require a diacritic, which Shaw explicitly proscribed. The man didn't even like apostrophes, for crying out loud, which is why Shavian doesn't use those either.

Read's solution was the best I think he could've come up with under his restrictions, though it tends to bounce people off at first, especially if they were lured by promises of "one letter, one sound." It comes in two parts: the schwa-STRUT split and the HAPPY-KIT merger.

Back in Read's time, lots of people tended to pronounce the schwa, which was never stressed, and the STRUT vowel, which was always stressed, as two different sounds, so it made sense for Shavian to have a distinct letter for each (and their rhotic versions). General American merges those two sounds into just the schwa, and it did even in Read's time, too. But the accents that split them back in the day are starting to merge them more and more now, so in effect, most speakers can see the Shavian letters 𐑳 and 𐑩 as denoting the difference between a stressed and unstressed schwa, respectively.

However, this other thing, despite being just as simple, seems harder for people to get used to. The unstressed vowel at the end of "happy" used to be pronounced in many different ways, ranging from the FLEECE vowel to the KIT vowel to even the DRESS vowel. A reasonable person would recognize it as its own phoneme and just come up with a new letter for it, but Read instead decided that it should be spelled with the same letter as the KIT vowel. This had a ripple effect on Shavian spelling. Just like with the schwa and STRUT, a lot of these varying pronunciations of the HAPPY vowel are dying out, and most people nowadays pronounce it as an unstressed FLEECE. This results in us having to spell what sounds like a FLEECE vowel as KIT, so we look like we're saying "happih." In effect, for the vast majority of people, a stressed FLEECE is spelled as 𐑰, while an unstressed FLEECE is spelled as 𐑦, just like all KITs regardless of stress.

This really grated on me at first, but I got used to it. After all, KIT never ends a word otherwise, so there's no ambiguity there... until you start adding suffixes, resulting in stuff like "candied" and "candid" both being spelled 𐑒𐑨𐑯𐑛𐑦𐑛. On the flip side, this rule does create a distinction between "trusty" and "trustee," so all in all, it's a wash. Still a relic of an older time that would never make it into a system designed in the modern day, though. And yeah, I think 𐑦 at the end of a word is more aesthetically pleasing and easier to write than 𐑰. It just looks unstressed, you know?

Some people really don't like it. Sure, it has its flaws, and it certainly isn't how I would've designed the system, but it at least lets you mark stress for the FLEECE vowel, right? Personally, I find it harder to determine whether a reduced vowel (such as the first one in "believe") is a schwa or KIT.

However, all this still means that only the schwa and FLEECE distinguish their stress in writing. Other vowels can still get fucked. Case in point, the aforementioned "insight" and "incite" are both spelled 𐑦𐑯𐑕𐑲𐑑. Kinda stumbling at the finish line there.

Actually using Shavian

Despite its flaws, I think Shavian is quite good, and I actually journal in it semi-daily. As of late, my Shavian handwriting has actually gotten more compact than my Latin handwriting, where I already tend to squish letters together pretty hard.

And in case you couldn't tell from how much Shavian I included on this page, it's actually part of Unicode! You can get keyboard software that lets you type in Shavian, too. Though if it's showing up as boxes for you, you might need to install the Noto Sans Shavian font.

And then there's Quikscript

I don't have enough to say about this one to justify giving it its own page, but a few years after completing Shavian, Read tried to improve upon it, releasing the new version under the name Quikscript. It got very little attention, and nobody bothered much with it.

And I can see why, to be honest. For every improvement it makes to Shavian, it introduces another drawback. It merges the schwa and STRUT, but unmerges WINE and WHINE and still doesn't mark stress for some reason. It eliminates a lot of lookalike letters, but the ones that remain are even harder to tell apart now. It makes better use of Tall and Deep letters, but they're mostly distinguished by the parts that jut out from the writing line now, making Quikscript hard to read if the Talls of one line overlap the Deeps of the previous. It introduces more abbreviations and ligatures, making writing even faster and more compact, but rules for when you're supposed to write the schwa in affixes are largely arbitrary. And besides, the abbreviations could just be imported into Shavian as-is if its users agreed upon them as a convention.

Resources