Response to Anatoly Liberman on simplified spelling

This is a response to Anatoly Liberman's Oxford Etymologist post "Spelling reform: not a 'lafing' matter".

My general principle is that words with consistent spellings should be left alone, and words with inconsistent spellings should be changed. The idea is to make English about as hard as French: given a written word, we can reliably determine its pronunciation in our own accent, but not always vice versa. (A few double pronunciations, as for long and short single vowels and for stop and fricative "c", are in my view too deeply embedded to change.)

In particular, initial kn is always /n/, so it can and should be left alone. Likewise, c before a back vowel should be left alone when it is /k/; there is no more need to change scan to skan or cesspool to sesspool than there is to change car to kar. On the other hand, I would actually go further than you with giraffe and write jiraf, except that this would render it unrecognizable to speakers of other languages.

For syllable-final gh, I think it should remain where it is written today but be absolutely silent, not changing in any way the usual pronunciation of the preceding vowel. Thus since au/aw is a usual way of writing /ɔ/, let us keep caught, change thought to thaught, and change bough to bow. (The weapon, but not the gesture, becomes boe.) When the sound is /f/, write ff finally and f when a consonant follows or in a few international words like clef (musical). Thus for laugh I would write laff (already in jocular use), and likewise laffing for laughing (because lafing would suggest a long vowel pronunciation /eɪ/. The spelling larf, also in jocular use, I would reject because it is very misleading to Americans, Scots, and Irish people.

Masha Bell: I agree with the spellings frend, sed, ruff, blud, munny (not muny, which would rhyme with puny). However, changing any is more problematic. It must be replaced, for as written it appears to rhyme with zany, but to what? England and most of America would be well-served with enny, but the American South says inny due to a regular sound change, and Ireland says anny (and for them any is only slightly irregular). I think the majority must rule here, and I recommend enny.

In short and with a few caveats (I see no need for dh or aa) I would follow Axel Wijk's Regularized English. This is a scanned and OCRed PDF, but with some pages unfortunately out of order. It's rather a big book, but it serves not only as a simplified-spelling proposal but a detailed analysis of every letter, digraph, trigraph, and tetragraph in English and how they are used today. As such, it is invaluable to anyone interested in spelling reform of any sort. (It deals only with RP and General American, but one certenly can't/caan't hav evrything.)


David Marjanović said...

this would render it unrecognizable to speakers of other languages


Unknown said...


As I said in my response to Anatoly, it's too late if one does not want to split English into a number of national and regional systems. When the sun did not set on the British Empire, there was still a chance of imposing then-current RP on the world, but political and social developments have carried us far beyond that. I say PIN/PEN and similar words alike, and I sometimes misunderstand speakers who don't, but I learned to distinguish the spelling in first grade. I had a friend in Texas with the surname KAISER who went to a conference in Boston, and the clerk at the hotel wrote down her name as CARSER. Are you happy with spelling TOT/TAUGHT alike?

The former state director for English in Texas once complained to me that a local TV ad for a DAIRY pronounced it like DERRY, and another was distressed that people were not distinguishing THEIR/THERE. I would resist having Yankee pronunciation imposed on my spelling. Again, modern spelling has already become like Chinese characters, and it is too late to change it. The real future is in Chinese characters, which can be used to write English and other languages (as in Japanese, which is unrelated to Chinese)irrespective of pronunciation, and provide a truly international written medium of communication.

--Rudy Troike
Professor emeritus
English and Linguistics
University of Arizona

John Cowan said...

Thanks for commenting!

As I noted, Wijk's system does not attempt to eradicate difficulties with spelling: that would indeed be impossible with so many different mergers in different accents, like the ones you mention. I certainly do not advocate a merger of pin and pen, or tot and taught, dairy and Derry, or even con and Khan (not a minimal pair for me, though the others are). I would not object to changing vein to vain, as this particular merger has gone to completion in all accents of English, but I don't think it's necessary either.

Fortunately, the number of variable splits is still small enough to make a diaphonemic spelling system possible. If you had, as a precocious first-grade linguist, asked your teacher "But why are pin and pen spelled differently?", the informed answer would be "Because people in other parts of the country don't pronounce them the same, so spelling them differently helps those people." Australians and some people in the Eastern U.S. who have undergone the lad-bad split (a matter of phonemic vowel length Down Under, of tensing hereabouts) will still have to guess which pronunciation goes with which word, but I don't know of any truly minimal pairs.

So with Wijk's reform, you can read either tot or taught with 100% reliability according to your accent, but if you merge the two (I don't, but I suppose you do), you will have to memorize which spelling goes with which meaning. But at least you will not have to stare at choir and wonder which word that is: the spelling quire will tell you everything you need to know.

Finally, you overestimate the ability to write different languages with Han characters. Before the 20C, many languages were indeed written with Chinese characters. People learned to read Classical Chinese texts in their own Sinitic variety, or indeed Japanese or Korean, and wrote their own languages as if they were Classical Chinese with additional vocabulary. But with the increasing movement toward writing Modern Standard Mandarin rather than Classical Chinese, this ceased to be true in China. With the exception of limited amounts of Cantonese writing in Hong Kong and the diaspora, children and illiterate adults must pick up Modern Standard Mandarin as their written language, in the same way that Austrians pick up Standard German as theirs. After World War II, Japanese sharply reduced the number of Chinese characters in general use, while Korean has now abandoned them almost completely in favor of their native syllabary (with a few exceptions like newspaper headlines, where conciseness beats clarity).

If you want to see how English might be actually written logographically, see this lighthearted but serious post on Yingzi.

Unknown said...


Thanks for all the information. Masha's uninformed or misinformed comments were so egregious that I couldn't let them go for the sake of other readers. I am definitely not a mergerer of COT and CAUGHT, though my young sons, growing up here in Arizona instead of Texas, resist my efforts to get them to maintain them (along with icebox), sadly. As understanding as I am of variety, it still bothers me every time I hear a commentator on TV pronouncing LAW as /la/ because the open o has been lost. And I still occasionally have trouble distinguishing someone's PEN from PAN, when there is no clear preceding context. On the other side, I somehow grew up not distinguishing /hw/ and /w/, and not until my mother was in her 90's did I notice that she had been making the distinction all the time.

Re hanzi, a few years ago one of my students at my request converted a whole paragraph to characters just as a demonstration of the feasibility. My sense of the possibility was sparked by an experience at the U of Illinois. A scholar from China gave a talk in Mandarin, along with a handout in hanzi. I was attending the meeting with a student from Korea who had gone through the schools before nationalism expelled the use of characters. At the end of the talk, she asked a very acute question (in English, which had to be translated into Mandarin for the speaker) regarding one of the points in the presentation. I was surprised, and asked her afterward if she had understood him, and she responded that she had not understood a single word of his talk, but she was able to follow the argumentation in the handout because of her recognition of the characters.

I spent a year in Taiwan on a Fulbright, and learned about 300 characters (which through disuse has fallen to some 30), and I've written about a syntactic analysis of locatives. It was interesting to have my name properly rendered into characters so that a chop could be made to use as my legal signature on a lease to a house. You may know that a prominent street in Taipei is named for Theodore Roosevelt, with his name rendered in three characters. The reverse is simply using the characters for /hwoche/ for 'train' in all languages, so that one would not have to learn to read Russian and Hindi, etc. before one could read scientific papers in those languages.

Thanks for the fun lesson on Yingzi -- very good and clever.