« Is this a sustainable business model? | Main | You first »
Rational Spelling
Here's a great, little video from Ed Rondthaler former president of the American Literacy Council and author of The Dictionary of Simplified American Spelling. Loyal readers will know that simplified spelling or what I call rational spelling holds a special place in my heart.
I worry that the tyranny of spell checkers impedes evolution towards rational spelling.
Hat tip to Boing Boing.
Posted by Alex Tabarrok on September 6, 2008 at 12:52 PM in Science | Permalink
Comments
If spelling becomes truly rational, won't that make much of American/English literature illegible to those taught rational spelling in grade school?
Posted by: Dennis Tuchler at Sep 6, 2008 12:59:05 PM
There is a lot to be said for rational keyboards too; but we are up against real big industry standards.
Posted by: Diversity at Sep 6, 2008 1:04:00 PM
@Dennis: when Maoist China did orthography reform, that was regarded as a feature, not a bug.
Posted by: Cyrus at Sep 6, 2008 1:24:25 PM
I'm not fond of most spelling reforms; English orthography is messy, but has the advantage of being uniform across the United States and very similar across the world. If we adopted phonemic spelling, then we would have to choose a standard dialect, which might not solve much for people who spoke dialects different from the standard. If we adopted the pronunciation of the speakers within the area of the Great Northern Cities Vowel Shift, then we would have to spell 'Pen' and 'pin' the same way as for many of those speakers, those two vowels have merged, just as I pronounce "marry" and "Mary" identically, but "merry" differently from those , but "horse" and "hoarse" or "cot" and "caught" the same way, even though some speakers distinguish Marry-Mary-Merry, horse-hoarse and cot-caught whereas some do not. There might be some limited reform that could be done without running afoul of dialectal differences, but it would be quite marginal and I barely see any reason that it might be useful. I expect that it would create a new standard that is only marginally less difficult than the old whilst making centuries of texts obsolete.
I don't think that English orthography is to blame for what little illiteracy remains in the United States; Japanese is written with an intricate system of two syllabaries and a logography, which I can personally attest to be very difficult to learn, but Japan has an exceptional literacy rate.
If anyone was keen to raise the specter of George Bernard Shaw and his silly claim that, "ghoti," could be pronounced, "fish", I must object to your propagating that wiseacre ignorance. The digraph is pronounced as /f/ only world-finally, perhaps erratically ("though" v. "tough"), but I cannot be sure without more research, there might be a rule. The letter is pronounced as /I/ only in the eccentric word, "women," which is a result of the spelling being altered centuries ago for harmony with the root; in old English "woman" was "wimman" and its plural was "wimmen", but over time the vowel of "wimman" assimilated to the roundness and backness of /w/, then "wimman" -> "woman", but the vowel in the plural was not affected, even though the spelling was altered. Lastly, the sequence is /S/ because of palatalization, alike to why and as sometimes /s/ and /dZ/, which only occurs before historic vowel. Ghoti can only be something sounding rather like, "goatee." (for me: [go4i])
Posted by: Paludicola at Sep 6, 2008 1:32:52 PM
1. Simplified spelling might someday evolve out of texting. Seems unlikely, but any other path is far more unlikely.
2. What about divergent pronunciation (cot/caught, marry/merry/Mary, Korea/career)? What is the benefit of brand new "phonetic" spellings that you will have to relearn and memorize anyway because they don't match your expectations?
3. The word "country" has arguably survived in its current form only because it is not phonetically spelled.
4. Transitioning to a new mandatory simplified spelling system would incur severe economic costs: prolific readers would find their reading speed slashed, probably permanently. Anyone whose job depends on doing lots of reading (financial analysts, policy wonks, etc) would be significantly impaired, as would anyone who depends on being widely read. The blogosphere would grind to a halt. Just like a credit crunch or an oil crunch, the textual parts of the attention economy would face a long and crippling attention crunch, with very real economic consequences.
5. The rise of Amazon Kindle and websites and electronic ink vs. printed paper might suggest a way to transition to a simplified spelling: transform any given text at the touch of a button. But we have long had the ability to transform keyboards to Dvorak at the touch of a button, and few bother. And there will still be road signs, billboards, and anything else that is meant to be read by more than one person at a time, which will have to pick one spelling and stick to it. This will only be workable if we transition to augmented reality, where things like road signs will be virtually imaged in every person's heads-up visor goggles, much like the yellow first-down line in American football appears in every TV set and doesn't actually physically exist on the football field. In this case, the road sign text could be adjusted to each person's preferences: foreign visitors could even read them in their own preferred language.
6. Even with technological means to assist any transition, you'd probably end up with a Bokmål/Nynorsk fiasco, two eternally competing standards.
Posted by: at Sep 6, 2008 2:04:41 PM
Paludicola: thanks for the very sensible argument against "rational" spelling.
Posted by: samson at Sep 6, 2008 2:06:04 PM
Shouldn't that be "rashanull" spelling?
Posted by: Yancey Ward at Sep 6, 2008 2:14:24 PM
Fixed spell-checkers, as in word-processors impede spelling reform; but google-as-spellchecker is good for it.
Posted by: Douglas Knight at Sep 6, 2008 2:31:44 PM
Isn't spelling reform a bit like linguistic inflation? It steals from those who have a lot saved up, to give to those with without. In the former category go not only those of us who can already write, but all our libraries, archives, literature. In the latter go children who already speak but don't write.
I'm not in favour, because I don't think the burden of learning is unreasonable high, and the benefit of being able to easily read books from a century ago is huge.
The burden of learning to read classical Chinese was, I believe, much bigger than English. And making old texts unreadable was one of the goals of the reformers, which I'm told they mostly achieved.
As Paludicola points out, phonetic spelling requires you to choose a standard pronunciation. If my spelling is going to store extra information beyond my choice of words, I'd prefer it to remember which languages the words were borrowed from than to encode how some 20th century American who can't tell a writing instrument from a sewing tool spoke.
Posted by: improbable at Sep 6, 2008 2:58:55 PM
"@Dennis: when Maoist China did orthography reform, that was regarded as a feature, not a bug."
Except that the great percentage of "simplified" characters had been in use for hundreds and even thousands of years already. Most educated Chinese in the mainland can still read "traditional" characters, even if they can't always write them. The switch from the classical to vernacular written languages is what cut many Chinese off from the classical culture, and that was definitely not attributable to the Maoists alone.
Posted by: wugong at Sep 6, 2008 3:00:54 PM
That's interesting, wugong. I didn't know there were two changes, 'spelling' and vernacular, to chinese in the 20th century.
Do you have any good articles or books on this which you could recommend?
Posted by: improbable at Sep 6, 2008 3:15:59 PM
1. Good sources on evolution of modern Chinese orthography:
Binyong Yin and John Rohsenow (1994). Modern Chinese Characters.Sinolingua
Insup Tayler and Martin Taylor (1995). Writing and literacy in Chinese, Korean and Japanese. Philadelphia: John Benjamins.
2. The arguments for "rationale" spelling assume that the goal of writing is pronunciation. It probably isn't, entirely, and it probably should not be, either. For example, the words "South" and "Southern" are pronounced quite differently. Would you rather have a writing system that preserved the morphemic relation at the expense of pronunciation, or the reverse? Given that writing systems are arguably written for the convenience of native speakers of that language, I'd argue that it's better to keep the morphemic relation than the phonemic one.
Sure, it's a nuisance for learners, but it's manifestly learnable (were it not, either the written or the pronunciation discrepancies would disappear), and note that educated adults tend not even to be aware of these oddities.
Learning to read in an alphabetic writing system will still be hard, no matter how "rational" the system is. (See Gough's (1980) classic paper, Learning to read: An unnatural act). Compared to the inherent difficulty involved in reading and writing, the additional costs due to spelling oddities is pretty much a rounding error.
Posted by: Kevin Miller at Sep 6, 2008 3:58:40 PM
Improbable-- An older book is Nationalism and Language Reform in China by John DeFrancis. It should give a good overview. Any history of modern China (Spence is a popular one), will have a section or more on the vernacular movement and other aspects of language reform in the 20th century.
Posted by: wugong at Sep 6, 2008 4:10:12 PM
Pauludicola: You argument against Shaw is entirely unconvincing. The whole point of a straightforward conception of spelling is that one letter = one sound. Yes, we all realize that if you invoke all sorts of contextual rules and historical contingencies, you can explain away the anomalies where the straightforward conception fails. But Shaw's point remains, that under a straightforward conception of spelling, there are English words where gh+o+t+i gets pronounced as f+i+sh.
Posted by: David Wright at Sep 6, 2008 4:35:08 PM
There are many so problems with English spelling that act as unnecessary hurdles for early learners. These hurdles account for disillusioning many of the 25% who fail to achieve their literacy milestones by age 11. The links between high illiteracy and difficult-to-teach spelling systems is incontrovertible. Many other writing systems don't have these early obstacles to learning to read.
However, one might think that the areas that are addressable are, firstly the double consonant rule (ferry berry merry but very; common but comic; muddy but study etc). This rule is only followed 50% of the time. And then there are the many silent letters that pepper the language (dumb-rum, receipt-deceit, debt-met).
There are other niggling inconsistencies: a promise/ to promise, a service/to service but a practice/to practise - comes to mind.
See: http://www.englishspellingproblems.co.uk/index.html
None of this has anything much to do with pronunciation. Also the etymology argument is something of a red herring once you accept that words change and there is no "absolutely proper" convention of spelling a word: Chaucer wrote "lern" "iland" "hadde": Victorians wrote "shew" (show) "musick" "comick" etc.
If we are talking about an evolutionary & generation change it is unlikely that fixing this would slow down reading speeds. The grammar system underwent a simplification in mediaeval times; we also simplified our number system from Roman to Arabic and our currency system from LSD to decimal. We saw the back of these "dumb" systems and introduced things that are fit for purpose. As good spelling & literacy are immensely important it is time we thought about upgrading the spelling system for the benefit of learners.
Posted by: NJH at Sep 6, 2008 5:06:53 PM
The analogy to simplified Chinese is terrible since characters and orthography are not directly comparable and characters have always evolved. Those above who thought that Chinese simplification was designed by "Mao" to eliminate the reading of older texts are merely showing their absolute ignorance of China, since simplification predated the Cultural Revolution and even Communism, almost all educated Chinese can read traditional characters, traditional characters have never been banned, and all the classical texts are available in simplified print.
A better analogy would be to the spelling reforms in other Latin-based languages, such as French (1990), German (1996), Dutch (1947), etc.
And let's not forget that Shakespeare did not use the spelling that we use today. For example, he would write sentences like "Houer through the fogge and filthie ayre," yet English readers today are still able to read Shakespeare. That's because texts get reprinted in the new spelling system.
Posted by: LZ at Sep 6, 2008 6:27:48 PM
Chaucer wrote "lern" "iland" "hadde"
Yes, but the spellings evolved organically, not by fiat. Imagine if everything you read is as difficult to read as, say, the Canterbury Tales. That will be what it will be like for the old generation reading new texts and the new generation reading old texts. Are schools going to have to stop using any old texts before the change?
And it's not just spelling that evolves, pronunciations evolve too. We can change the spellings of all words to some standard, but in 50 years people are going to pronounce them in a different way. Do we then have to change the spellings again?
To me it all seems like a solution in search of a problem.
Posted by: Hei Lun Chan at Sep 6, 2008 7:13:28 PM
LZ, it's true that classical Chinese texts are now available in simplified characters, but the reform did make old books harder to read for the younger generation (of course, then the Chinese destroyed the old books, so that problem became less relevant). And it's not precisely true that "all educated Chinese can read traditional characters"; they can do pretty well, sure, but even my graduate student friends at top Chinese universities don't know their traditional characters perfectly. It's not an insuperable barrier, but simplified character education has made it marginally more difficult for PRC students to partake in the classical traditions of literature and calligraphy.
Posted by: Nick at Sep 6, 2008 9:18:00 PM
Nick-- I still maintain that the switch from classical to vernacular as the standard for the written language had a FAR great effect than the switch to "simplified" versions for some characters. You may think that the vernacular language movement was a bad idea for this reason, but there is no doubt that it contributed to massive increases in literacy in the Chinese speaking (and reading and writing) world. In fact, since the literacy rate is so monumentally higher than it was in, say, 1870, one could argue that far more people in the PRC can read "traditional" characters now than before the language reforms. Would you object to the use of the written Romance languages because it's led to people being less able to read Latin?
As for "then the Chinese destroyed the old books, so that problem became less relevant," what a completely absurd statement! That's like saying "The Europeans destroyed the books written by Jews" because the Nazis burned some books. Sure, some books were destroyed during the Cultural Revolution by a tiny minority of Chinese people, but even the PRC continued to put out massive typeset editions of essentially all the classical works throughout most of the Maoist period. And I'm pretty sure almost none of the books in any of main national libraries suffered any harm.
As for calligraphy, many of the simplified characters are derived from cursive calligraphic forms and any students truly interested in calligraphy would have no trouble learning traditional forms. Hell, I did, and I'm not a native speaker!
Posted by: wugong at Sep 6, 2008 10:55:57 PM
'Improbable' and others wunder how new spellings would be possible with our varius dialects. Traditional spelling (TS) mor or less copes with them now; so why cant a renewed spelling system?
My guess is that it would be based on 'standard English'. as represented by BBC and NBC newsreeders pronunciation. Yes, there ar two 'standaard Englishes', and we in the rest of the world cope with them, both in heering them and reeding them.
So, why couldnt an international group of professional peeple (linguists, publishers, broadcasters, busness groups, parents, educationists) meet and hammer out a new 'standard English' spelling sistem that was logical, predictable, and did the job that alfabets wer desined for – making reeding and riting available to all?
Posted by: Allan at Sep 6, 2008 11:42:45 PM
English speakers, like people in any literary culture, do not sound out a written word's individual phonetic components. Each word is read very much like a Chinese character, as a whole symbolic piece. Words retain historical baggage by virtue of their spelling, but that spelling should not be discarded lightly. When encountering a new word, native English speakers pick up on a dozen different signals, which, in context, teach them the meaning of that word.
Posted by: J. Cuttance at Sep 7, 2008 6:38:25 AM
I'm skeptical of simplified spelling. I can't claim to be knowledgeable, but it seems that Pinker's _The Language Instinct_ (scared to do italics on this site!) did a good job of rebutting the concept. Unless you want to go to phonetic pronunciation like Spanish, the simpler you make the spelling rules, the more complex pronunciation rules become. So you're left with a similar level of complexity overall. While an evolutionary move to simpler spelling (thru, lite, etc.) is fine, I have a hard time believing any organized change will work well.
I bet the publishing industry would be highly in favor of a drastic spelling change. Imagine how many new editions they could sell (at least to the people who care about these things).
Posted by: Greg at Sep 7, 2008 6:24:30 PM
I'll just point out that the United States already went though a period of rational spelling, thank goodness. Written English would be much more of a mess if it hadn't. Unfortunately English speaking nations outside of North America have been slow to adopt the US rationalizations, but there is hope they are slowly coming around. Another round of rationalisation seems called for but there is no need to do it quickly. We can just pick one of the worst problems, fix it and if problems somehow arise we can hold off on further improvements. Of course simplifying spelling is only likely to work if major media organizations get on board. But simplified spelling will help children learn and make English more attractive as a second language, and so could have significant economic benefits in the long run.
Posted by: Ronald Brak at Sep 7, 2008 8:18:26 PM
Ronald Brak,
Last time I checked, Canada was still IN North America. Most of us here still spell English properly (eg, colour, harbour, labour, centre, metre, theatre, etc), and have no intention of "coming around", as you put it. Anyway, any true "rationalisation" of English spelling would also entail the instertion of the letter "f" into the word lieutenant (the rest of the world pronounces it "leftenant"). I'm sure you'd be glad to go along with that.
Meanwhile, I'm still waiting for you guys to switch to metric and stop measuring things in hogsheads or furlongs or miles or whatever it is you use. Maybe then you will be in a position to start lecturing the rest of us on what's rational and what's not.
Posted by: Jib Halyard at Sep 7, 2008 9:32:13 PM
I quite agree with NJH. There is no reason why Inglish spelling coodn't be at least no wurse than French. The rules of French spelling are certenly complicated, but thare are aulmost no wurds hoose spelling is completely irregular and unpredictable. Wunce you have lerned the system, you can pronounce eny wurd at sight.
French has a bunch of silent letters at the ends of wurds, but in Inglish the silent letters (I doan't mean silent final 'e') doan't serve eny purpos at aul, offen not even historical: thare never wos a "b" in "dett" (in Old French it wos "dette"). In the cases whare different accents pronounce wurds differently (like pronouncing "bath" like "bahth" in RP, or pronouncing "cloth" like "clawth" in most American accents), traditional orthography usually is sutable for breaking ties.
In eny case, it shood be obvious that the spelling I am using here is not such a radical break with the past that it wood make reading eether/ither old or new texts very difficult at aul, but *wood* make lerning to read much easier (lerning to spell wood still be hard, but not *as* hard by eny means). Lerners wood not hav to suffer thrue lerning to read the 15% or so of wurds hoose spelling is plain silly.
As for Japanese literacy, it's very overrated: aulthough it's true that most Japanese people lern to read and rite the aulmost 2000 characters needed, reserch shows that they forget meny of them before very long. Fortunately, Japanese computers let you type in Roman letters and automatically convert to Japanese writing.
Posted by: John Cowan at Sep 7, 2008 9:52:04 PM
