Making English Spelling Sensible: A Thought Experiment

This is an idea I’ve been noodling in my head for a long time, years even: how to make English Language more sensible. By “sensible” I mean more consistent, intuitive, and easier for people to learn.

The English Language suffers from a few challenges, not in how it’s spoken, but how it’s written:

• Silent letters: knight, for example, only pronounces 3 letters (n i and t). The “l” in would is not pronounced either, nor is the “h” in ghost. Also, many words have a silent “e” at the end.
• Sight Words: sight words are words that are core to the English language, but are spelled very differently than how they are pronounced: one, caught, could, use, and so on. These are high-frequency words, and crucial for basic English literacy, but as a former volunteer tutor for grade-school kids, they also cause lots of friction and considerable effort for children to learn. It shouldn’t be this hard.
• Overloaded vowels: the letter “a” can be pronounced like “a” in father, “a” in cat, or “a” in rate (with a silent “e”). The letter “o” too: “o” as in hot or gone (the silent “e” is somehow ignored), “o” as in fork, or the “o” in sought.
• Inconsistent consonant sounds: sometimes “c” is pronounced like an “s”, other times like a “k”: center vs. cat. Also “hard” g versus “soft” g.
• Implicit sound combinations: that “ph” sounds like an “f” is , but where did “tion” (as in “shun”) come from? We don’t pronounce the “e” in -ed endings either.

There are great articles on the Internet that help to explain why English spelling has evolved the way it did, but I haven’t really found an ideas online about how to make the spelling more consistent and intuitive. People know that spellings and pronunciation have diverged over centuries, but how to make the spellings match the pronunciation again is a tough nut to crack.

In fact, if you look at Internet slang, some of these changes are already happening, using spellings like “wut”, “nao”, and “gurl”. They may seem frivilous at first glance, but it’s a sign that people are already developing more efficient and effective ways of communicating. A kind of “ground-up” reform.

The purpose of this post is to propose ways to reform the spelling of the English with the following self-imposed guidelines:

1. No new writing systems: The Roman alphabet works fine, and many European languages have tailored it to meet their needs, reforming it over time as things change and evolve. English can do the same. No need to invent a whole new script.
2. Intuitive for native speakers: if a writing system isn’t useful and intuitive for the people who need to use it, it’s not worth learning. So, spelling reforms have to be something that native speakers would naturally pick up.
3. Minimize rules and exceptions: some cannot be avoided, but many others are just relics that can be simply dropped.
   • For example, I don’t believe it’s possible to change the first-person pronoun “I” to anything simpler (why is it even a capital letter?) even though its spelling is a bit non-standard.
4. Make spelling consistent: in other words, make the writing system as WYSIWYG (what you see is what you get) as possible.

To that end, I have been playing around with different solutions for making English more intuitive and straightforward while trying to keep its essence as much as possible. This is a tricky balance, and for me it’s still an ongoing work. Some changes such as dropping silent letters are fairly easy to implement, others such as untangling vowel sounds, are much harder. Readers who find this page are welcome to adopt or reject these proposals as appropriate. Also, since the pronunciations are biased toward North American English, speakers of other English dialects would have to adjust as needed.

Disclaimer

This is just a thought-experiment, and not meant as serious academic endeavor. This is just a language nerd playing around with English. The ideas below probably are flawed, and would need further refinement before they became serious proposals for English spelling reform.

Consonants

Some easy fixes here, some a bit more difficult.

• No silent E – we can use other ways to convey long vowel sounds (see below)
• Replace final Y with other appropriate endings – Y only makes sense at the beginning or middle of a word.
• Break up the letter C:
  • “Soft C” becomes S. This one feels pretty intuitive.
  • “Hard C” becomes K (or Latin C?) – The “K” spelling is more Germanic looking, while hard “C” is more like Latin spelling. Either choice can work. However, either choice will become awkward with some words: Kat versus cat. King versus cing. Killer vs. ciller, or komputer vs. computer. The “K” sound feels a bit more intuitive to me, but I admit I still like the Latin C for sentimental reasons.
• Soft G becomes J – voyaj instead of voyage,
• PH becomes F – Rafael instead of Raphael
• KN becomes N – nife instead of knife
• GHT just becomes T – weit instead of weight, thot instead of thought

More Consonant Changes

As above, some of these are fairly easy and sensible, some are a bit less intuitive.

• TU becomes CHU – virchu instead of virtue
• TURE becomes CHUR – overchur instead overture
• Final S becomes Z – goez instead of goes
• Final CE becomes S – chans instead of chance
• Final LE becomes UL – incredibul instead of incredible
• Final ED becomes D or T – fixt instead of fixed, calld instead of called
• Final CK becomes K, or C – quic or quik instead of quick.
• Final TION becomes SHUN – informashun instead of information

Short Vowels

I decided to keep the short vowels mostly unchanged. They are pretty intuitive already.

• A as in “fat”, except AR as in “far”
• E as in “get”
• I as in “fish”
• O as in “father”, except OR as in “or”
• U as in “fun”, except UR as in “fur”

Long Vowels

As if things weren’t complicated already, trying to find a way to express long vowels without using a silent e or other tricks is one of the harder things to fix. We have five vowels in the English alphabet, but we have a lot more vowel sounds than we can fit onto 5 letters.

There are a few ways to approach this:

• Latin-style diacritics: a is short, ā is long (as in rān for rain). Not all English-language software is equipped to handle diacritics, plus other challenges. This is the most WYSIWYG method, but also requires the most effort to re-learn.
• Diphthongs: using two vowel-letter combinations to make each sound distinct. This doesn’t require substantial changes to software, and we already do this for some words (float, boat, through, avoid, etc). These are pretty distinct and visually look pretty good, but some diphthongs work better than others.
• Vowel-Consonant combinations: instead of two vowels put together, maybe a vowel and a consonant? For example, day uses a + y, so we can apply that consistently to words like rayn, payn, informaytion and so on. Or the o+h combination for boht, goht and so on. Sometimes this work, but in practice I found this to be more awkward than expected.
• A mixed solution: it might be best to simply use diphthongs where it seems intuitive, and vowel + consonant where it seems intuitive, as long as each combination is unique and sensible to native speakers.

I tested out all three options on my own, and I felt that diphthongs were the most intuitive and least effort to re-learn. For example:

• AI as in “night”: nait, fait.
• AE as in “day” or “rain”: dae, raen (alternatively, EI).
• AO as in “now”: nao, paower, flaower
  • alternatively, “OW” might be sufficient here.
• AU as in “law”: lau, rau
  • alternatively, Irish style “EA” (as in Sean) might work, or stick with “AW” instead.
• EE as in “feed”: feed, same as before.
• EI as in “weight” (alternatively, use AE): weit, dei
• EU as in “few”: feu, geus, leus
• OA as in “boat”: boat, soap, soap
• OI as in “avoid”: avoid, boi
• OO as in “good”: good, hood
• OU as in “loose”: lous, foud

Conclusion

Theez ar just sum aideeuz I’v bounsd uraond, soa feel free tou trai them aot, see wut wurks, wut duzn’t, and drop a noat.