« In praise of little people | Main | Groupthink in companies »

July 11, 2007


Mark Wadsworth

English spelling is a hilarious in-joke for those who can spell, a nightmare for those who can't (or haven't been taught).

Other countries took their spelling by the scruff of the neck and simplified it. But English language is far too widespread, it's not like Malaysia, Turkey, France or Germany where the language is/was largely confined to one country and so one body can appoint itself the keeper and guardian.

And besides, I'm quite good at spelling.


That's "AN hilarious in-joke", Mark.


Some Dvorak stories from the real world:


People should switch to Dvorak. Less mistakes, less movement, less RSI, move productivity, move happiness.

Ken Houghton

The Fable of the Keys appears to depend upon the conceit that, since the Dvorak keyboard was either not selected or not available as typewriters were standardizing, it is either inferior (in the first case) or counterintuitive (in the second).

This is another of those cases where Equilibrium Theory leads to "this is the best of all possible worlds."

Mark Wadsworth

Sanbikinorian, "AN" before "hilarious" is a bit pretentious, like saying "myself" instead of "I" or "me".

"AN honest" on the other hand is correct. As indeed is "an 'ilarious" if you are slovenly.


Depends, I suppose, on whether (a) you count the English language as inefficient entirely because its spelling is inefficient and (b) you think efficiency is something for a language to aspire to.

Matt Munro

An efficient languge wouldn't use words at all, it would use symbols, as the brain "thinks" in symbols.
The perceived requirement for words to have a high correspondence between spelling and sound is completely nebulous, and something that only school children, bad teachers and dyslexics complain about.


Spelling reform is a topic for madmen. On the other hand, Max, that should be "Fewer mistakes....".

Tim Worstall

But the language is simplyfying itself: as the above little tiff over an and a shows. 50 years ago everyone would have said an was correct, a wrong. Now we think insisting upon an is pendantic. In 50 years time it will be wrong and a right (maybe).
That I think is where Chris' analogy fails (mildly, to be sure): we're not locked into our language in the same way that we are keyboards: language evolves, piece by piece.


The problem from the point of view of reinventing the language is that it wasn't invented as such in the first place. It evolved from Latin, French (which itself has Latin roots) and the Germanic languages along with a pot pourri of other influences over the centuries. Changing the spellings would disguise the roots, making learning the language more difficult. When learning new languages, I look for the roots that I can recognise and can therefore start a rudimentary translation. The SSS would undermine that.

The language is what it is; let it evolve naturally. If it ain't broke (and it ain't) don't fix it.


Spelling reform is great, as long as you can stop pronunciation changing.


Tim Worstall, there are constantly new pieces of pedantry being evolved, invented or rediscovered so that there will always be new complex 'rules' to replace the ones that we no longer care about.


The outcomes are only suboptimal if you ignore the cost of transition. To claim that English is a suboptimal language one must demonstrate that their was a point that there was a moment where some sort of linguistic central planner could have switched us to another language with improved grammar, vocabulary, and spelling and had that investment be a Pareto improvement with transfers. But since people in the past are poorer than we are now, that would seem to be a very tough case to make.

Marcin Tustin

The question is not does decentralisation lead to suboptimal results, but whether centralised decisions tend to be better. I would suggest that constructed languages with chosen normative versions tend to be grammatically complex, whilst those that grow (mostly) organically, even if not folk languages, like English, Norwegian (yes, I'm aware of the diversity that word covers), or Mandarin, are marked by very simple grammars. This probably indicates that what people find most beneficial is a simplified or flexible grammar. The grammar of those languages may be complex in computational terms (many things are allowable), but in human terms are easy to apply.

Roger Thornhill

I find Chris' points very interesting. I think it is more about inertia. The QUERTY was all about preventing an original typewriter from jamming by the twofold process of isolating common letter pairs and slowing down typing (fat chance!).

AS to words: Have you thought that the proper way to pronounce "though" is to actually voice the final "h"? As to the differnt treatment of the 'ou', this could be resolved by the use of accents as other European languages do.

If the Spelling-SS had their way, we would be forced to spell and so speak like some estuarine hag and our language would be doomed to a flat, nasal, glottaled sludge.

An alternative could be to go towards an ideographic language. Japanese combines a phonetic and ideographic systems (the latter borrowed from the Chinese) with a neat way to isolate and sandbox foreign words using a parallel phonetic script (basically an uglier, spikey version of the parochial set).

Maybe we begin to write spoken English in Hiragana, write French in Katakana and put down our thoughts in Kanji, though personally I like Traditional Chinese which has a richer set of ideographs stretching to over 65,000.

I once mastered Hiragana. It was an amusing skill in Japan to be able to answer the question "what does that say?" by replying in Japanese. The follow up question "no, what does that MEAN?" was beyond me, as all I could do was pronounce the phonetics! :-)


In any case, whatever the answer to the question about language, of course decentralised evolution can lead to suboptimal results (this is trivially true in biology, too). The question is whether state regulation would lead to more optimal results, on average; and the answer is a resounding no. People just aren't clever enough, and in particular, their ability to forecast almost anything thing is approaching zero. (And the lock-in problem itself would probably be worse because there would be people already invested in one solution).

That's why democracy and the regulated free market are the least bad ways of organising ourselves.

Actually, the analogy with biological evolution is quite interesting in respect of languages. The spelling of English may be silly, but misspelled words don't usually detract that much from understanding; also apart from spelling, English is fairly easy to learn, being relatively uninflected. If, through an accident of history, America spoke Russian (very inflected) or Chinese (tonal, and difficult to write), would those languages be becoming the lingua franca the way English is? My guess is not.

Mark Wadsworth

Dearime 'On the other hand, Max, that should be "Fewer mistakes...."'

Dearime, you are a star.

tom s.

I thought you were going to quote Krugman on English food, from about 10 years ago. It's a fine essay on evolution leading to inefficient results. http://web.mit.edu/krugman/www/mushy.html

Philip Hunt

Potentilla: """If, through an accident of history, America spoke Russian (very inflected) or Chinese (tonal, and difficult to write), would those languages be becoming the lingua franca the way English is? My guess is not."""

Well, Latin (a highly inflected langauge) used to be the lingua franca of Europe. And Spanish and Portuguese (also inflected) are very widely spoken around the world. In being inflected doesn't seem to stop a language from spreading.

I very much doubt if the linguistic characteristics of a language are at all important in how many speakers it has; English has spread, like Latin, Spanish and Portuguese, primarily through military conquest -- indeed the countries where English is the main language spoken are all to my knowledge places that used to be ruled by Britain.

Hilary Wade

There's an essay somewhere (?Dennett) in which the writer suggests that the human brain has an upper limit beyond which it can't process information any faster. Which would explain why translations of a text into different languages tend to be more or less the same length. Which, in turn, suggests that written language has already evolved to as "efficient" a point as it can get - more or less.


Spelling in the UK is moving towards the American simplifications - this is partly a European influence, as the oldies there tended to learn American as a result of WWII and its aftermath. I hate "organization".

john b

Dreamingspire - the 'ize' extension is thoroughly British in origin. Indeed, it's Oxford University's house style, and has been since anyone started caring about consistency of spelling...

I'm sceptical that UK spelling is moving towards American simplification - I've not seen a British publication or a Brit in email writing 'color', 'center' [*] or 'favorite', for example (or 'skeptical', come to that...)

[*] except for those awful DIY shops called Plumb Center, etc, which I assume are pretending to be American.


The other big problem with "simplified" and/or "phonetic" spelling is accents. Who decides (and on what basis) which accent is the true phonetic version of a language?


The spelling of words is important not just for communication but also for cultural reasons. Reforming spelling for efficiency is a bit like suggesting that a Tudor manor house be knocked down because a block of flats is more efficient.

In any event there has been a similiar experiment to this idea, where the prevailing wisdom was that the established system was only present due to a market failure and a new, more efficient system was imposed by the government. Of course this was the metric system units, which replaced the old organic English units. What is interesting about this is that still, many decades after the transition, most people are still using the English units (even young people who never learned them at school, like me), even though in many cases this means extra cost in translating from where the new units have been officially imposed. So perhaps there was no market failure and the old units are somehow more efficient than the metric system afterall.

tom s.

dreamingspire - Many people thought a few years ago that Canadian spelling was becoming inevitably American, but the differences have stayed. In fact some newspapers (Toronto Star in particular) have changed from the American spelling of color etc - which they used on the grounds that it was coming soon anyway, so get used to it - to a more anglo version.

For anyone who cares, Canadian spelling uses our (colour) re (centre) but ize (nationalize).

The comments to this entry are closed.

blogs I like

Why S&M?

Blog powered by Typepad