spelling bee

I assume that English is quite an easy language to learn to speak, given the number of people who speak it in one form or another, but it is probably no easier than any other language to speak well. And when it comes to writing the language down, the problems are exacerbated. There may not be a problem remembering the gender of nouns, or the myriad inflexions of verbs that plague other languages, but English is probably the least phonetic in its spelling of any major alphabetic language (an ideographic language such as Chinese carries no phonetic information in its script). This means that English is so riddled with irregularities and disparities between pronunciation and spelling that even native speakers can easily make mistakes with the latter.

There are twenty-four consonants in English as it is usually spoken but only eighteen usable consonants in the alphabet (because C can be rendered by k or s, depending on the context, Q by kw, and X by ks). However, four of the six remaining consonants can be represented by two-letter combinations (ch, ng, sh, th), and th can easily represent the unvoiced consonant in thin or the voiced version in this without confusion, while the voiced version of sh heard in words like leisure and treasure is sufficiently uncommon that the absence of a specific signifier shouldn’t cause any difficulties.

This doesn’t mean that a given written consonant will always be pronounced in the way indicated by the letter or letters. Most people will now be familiar with George Bernard Shaw’s fatuous attempt to ridicule English spelling with his invented word ghoti, which if you’ve never encountered it is pronounced ‘fish’ (gh as in cough, o as in women, and ti as in nation). However, this misses the point. The real problem is with the vowels.

Take a look at the following grid, in which words in the same column rhyme, but the way the rhyming part is spelled is different in each case. Words in the same horizontal row have the same spelling, but the pronunciations differ. It is possible to construct many such pronunciation grids for words of one syllable, some much bigger than the one shown here, although this example includes what is probably the largest number of different letter combinations to represent the same pronunciation.

I do not wish to imply that there are no problems with consonants. Someone whose only exposure to English is in its spoken form would think the following spellings perfectly reasonable: casl, resl, brisl, josl, rusl. Note that in each case there is a ‘silent’ t in the orthodox spelling, while resl also includes a redundant w. All five words also include a superfluous final e. There is a trend towards eliminating such final e’s, although care would need to be taken if this tendency is to be encouraged. For example, definite, doctrine and hypocrite could lose their final e’s without creating ambiguity, just as the final e has been dropped from deposit and fossil, and, more recently, from proletariat and secretariat. But dropping the final e from hearse would obviously be a mistake.

Initial w’s are a bit of a nuisance in a range of situations: silent in wreck, wring, writhe, write and wrong (among others); silent in who and whom, but pronounced in what, which and why; and distorting the following vowel, as in warm and worm, which if pronounced as they read should rhyme with harm and form, but don’t.

There are other letters that don’t appear to be playing a logical part in the accepted pronunciation: the s in island; the c in muscle; and the n in column and solemn. However, in the last two of these examples, in the words derived from these nouns (muscular, columnar, solemnity), the ‘silent’ letter is pronounced, so rectification of these spellings would be illogical.

Perhaps the most egregious of the redundant letters that crop up in English words are the b in debt and doubt, and the p in receipt. We have the great eighteenth-century lexicographer Samuel Johnson, in his Dictionary of the English Language of 1755, to thank for the last of these. At a time when spellings were not standardized—Shakespeare spelled his name in several different ways—Elizabethan scholars had added a p to conceit, deceit and receit, and although the three words are cognate, Johnson removed the offending p from only the first two. Another spelling anomaly introduced by Johnson is seen with deign and disdain, which derive from the same root.

The English language is very good at appropriating words from other languages when the need arises, but in the case of French in particular, such words have often been imported with their original pronunciations intact. Why, for example, is ballet not pronounced to rhyme with mallet; debut to rhyme with rebut; and debris to rhyme with hubris? The spelling would then be obvious, but the effect of the current arrangement is to separate pronunciation and spelling into different compartments, with the potential for yet more confusion.

Systematic reform of English spelling would seem to be called for, but there is a major caveat. It should be more important for a printed word to convey its meaning than its pronunciation for someone seeing the word for the first time, and if we were to spell nation as nayshon, for example, then its meaning would not be grasped automatically by speakers of other languages that contain a recognizable variant. The US spelling fotograf conceals its origin in ancient Greek, and for anyone with some knowledge of this language also its meaning.

Although I am not recommending any such reform, this doesn’t mean that there won’t be occasional ad hoc changes in pronunciation driven by what are essentially illogical spellings: how long, for example, can a word like clerk resist the spelling clark? Or might a new pronunciation emerge in which clerk rhymes with jerk? Given that most people learn to speak a language before they can read or write it, the pressure is likely to be on the spelling to change.