07 December 2011


by Ben Zimmer

[I’ve taught writing and composition in colleges, often to ESL and remedial students, so English spelling was frequently an issue. Spelling of American English is very hard, even for native-born speakers. It’s full of irregularities and sounds that are not spelled the way they’re pronounced. In addition, English has borrowed many words from other languages and these words sometimes (but not always) keep their original foreign spelling and pronunciation. I used to keep a list of words that give me trouble—before the advent of spell-checkers and on-line dictionaries. Most writers learn to recognize their common spelling problems (mine are i-e and -eed/-ede words) and double-check those words. (We have to be careful about relying on spell-checkers: some misspellings are real words, but with different meanings—like discrete and discreet—and the spell-checker won’t catch them.) In many cases, it’s sad to say, the best way to learn English spelling is simply to memorize it.

[To help make my point on the first day I broach the subject of spelling, I used to write the letters
g-h-o-t-i on the board and then ask how the word’s pronounced. The answer’s supposed to be ‘fish.’ The following column on the subject appeared in “On Language,” the New York Times Magazine feature by Ben Zimmer, on 25 June 2010. I thought it would be fun to reintroduce it to ROT readers. ~Rick]

* * * *
The spelling of English is a bizarre mishmash, no doubt about it. Why do we spell acclimation with an “i” in the middle but acclamation with an “a”? Why do we distinguish between carat, caret, carrot and karat? For those who feel strongly that something needs to be done, there’s no better place to vent some orthographic rage than the Scripps National Spelling Bee. The 2010 bee, held earlier this month, was no exception, as a handful of protesters from the American Literacy Council and the British-based Spelling Society picketed the Grand Hyatt in Washington, while inside young spellers braved such obscurities as paravane (an underwater mine remover) and ochidore (a shore crab).

When talk turns to the irrationality of English spelling conventions, a five-letter emblem of our language’s foolishness inevitably surfaces: ghoti. The Christian Science Monitor, reporting on the spelling-bee protesters, laid out the familiar story (while casting some doubt on its veracity): “The Irish playwright George Bernard Shaw is said to have joked that the word ‘fish’ could legitimately be spelled ‘ghoti,’ by using the ‘gh’ sound from ‘enough,’ the ‘o’ sound from ‘women’ and the ‘ti’ sound from ‘action.’ ”

Just one problem with the well-worn anecdote: there’s not a shred of evidence that Shaw, though a noted advocate for spelling reform, ever brought up ghoti. Scholars have searched high and low through Shaw’s writings and have never found him suggesting ghoti as a comical respelling of fish.

The true origins of ghoti go back to 1855, before Shaw was even born. In December of that year, the publisher Charles Ollier sent a letter to his good friend Leigh Hunt, a noted poet and literary critic. “My son William has hit upon a new method of spelling ‘Fish,’ ” Ollier wrote. You guessed it: good old ghoti. Little is known about William Ollier, who was 31 at the time his father wrote the letter. According to Charles E. Robinson, a professor of English at the University of Delaware who came across the ghoti letter during research on the Ollier family about 30 years ago, William was a journalist whose correspondence reveals a fascination with English etymology.

As a language fancier in mid-19th-century England, William Ollier would surely have come into contact with the strong current of spelling reform—championed by the likes of Isaac Pitman, now remembered for inventing a popular system of phonetic shorthand: what Pitman called “phonography.” In 1845, Pitman’s Phonographic Institution published “A Plea for Phonotypy and Phonography,” by Alexander J. Ellis, a call to arms that laid the groundwork for ghoti and other mockeries of English spelling. To make the case for reform, Ellis presented a number of absurd respellings, like turning scissors into schiesourrhce by combining parts of SCHism, sIEve, aS, honOUr, myRRH and sacrifiCE. (If you’re wondering about the last part, the word sacrifice has historically had a variant pronunciation ending in the “z” sound.)

Ellis thought scissors was a downright preposterous spelling of sizerz, and he went about calculating how many other ways the word could be rendered. At first he worked out 1,745,226 spellings for scissors, then adjusted the number upward to 58,366,440, before finally settling on a whopping 81,997,920 possibilities. Isaac Pitman and his brothers liked to use the scissors example when proselytizing for phonetic spelling, and the 58 million number even worked its way into “Ripley’s Believe It or Not!”

Don’t believe it. Ellis admitted that “the real number would not be quite so large,” since English spelling does not actually work by stitching together parts of words in Frankensteinian fashion. Ghoti falls down for the same reason, if you stop to think about it. Do we ever represent the “f” sound as gh at the beginning of a word or the “sh” sound as ti at the end of a word? And for that matter, is the vowel of fish ever spelled with an “o” in any word other than women? English spelling might be messy, but it does follow some rules.

Robinson suggested to me that William Ollier could have come up with ghoti in a parlor game of Ellis-inspired silly spellings. Victorians often amused themselves with genteel language games, so why not one involving the rejiggering of common words? Into the 20th century, other jokey respellings made the rounds, such as ghoughphtheightteeau for potato (that’s gh as in hiccough, ough as in though, phth as in phthisis, eigh as in neigh, tte as in gazette and eau as in beau).

Ghoti was elevated above these other spelling gags when it became attached to the illustrious name of Shaw—who, like Churchill and Twain, seems to attract free-floating anecdotes. If Shaw never said it, who was responsible for the attribution? I blame the philologist Mario Pei, who spread the tale in The Los Angeles Times in 1946 and then again in his widely read 1949 book, “The Story of Language.” Pei could have been confusing Shaw with another prominent British spelling reformer, the phonetician Daniel Jones (said to be one of the models for Shaw’s Henry Higgins in “Pygmalion”), since Jones really did make use of the ghoti joke in a 1943 speech.

With Shaw’s supposed imprimatur, ghoti lingers with us. Jack Bovill, chairman of the Spelling Society, told me that despite its jocularity, ghoti is nonetheless “useful as an example of how illogical English spelling can be.” I beg to differ: if presented with ghoti, most people would simply pronounce it as goaty. You don’t have to be a spelling-bee champ to know that written English isn’t entirely a free-for-all.

[Believe it or not, there are some English spelling rules that hold true more often than not. These "rules," however, will not keep you from making mistakes, even if you memorize them all. English has too many words that just do not follow any rules.]


  1. Why do we distinguish between carat, caret, carrot and karat?

    Though I admit that the difference is subtle and might not be apparent to everyone, I happen to pronounce them differently.

  2. Ian--

    Dunno, actually. I'm not an expert. (I usually pronounce 'caret' as 'carr-AY,' but maybe that's a pretention.)

    Some words that are pronounced alike but spelled differently (homophones) are written that way expressly to make them look different, like 'Sunday' and 'sundae.' I think 'carat' and 'karat' may be related that way, too.

    Some words sound alike because they came into English from different routes (say Latin/French vs. German). After going through vowel shifts and other changes as they are absorbed into English, the pronunciations coincidentally end up similar.