The heroine of Ian McEwan's new novel, Sweet Tooth, is Serena Frome. Rhymes with "home"? No: McEwan intends you to pronounce her surname as you do the Somerset town – rhymes with "room".
Nomenclature in Great Britain offers many such traps. Beaulieu, Cholmondeley, Knollys (pronounced like "Knowles"), and the particularly teasing Featherstonehaugh (pronounced like "Fanshaw"): they appear to be designed to mock the ill-bred, ignorant or foreign. Pity the tourist in Oxford who pronounces the "g" when asking for directions to Magdalen College! However, these eccentricities, while symptomatic of the way our spelling has developed, are comparatively trivial examples of the idiosyncratic nature of the entire spelling system – if system may be used of such a haphazard phenomenon – of the English language.
David Crystal, author of this sprightly survey, would challenge the word "haphazard". A prolific author and editor of language books, he can write with authority on trends in the spelling of "rhubarb" ("rubarb" is gaining ground, he reports), and indeed on the history of the spelling of any tricky word you care to mention. For him, the patterns are clear. The rest of us may often remain puzzled.
Crystal begins his story with the arrival in Britain of Christian missionaries, who rapidly discovered that they had at their disposal only about half the number of letters required to denote the sounds – phonemes – of the Germanic tongue of the natives. "That, in a nutshell, is the problem of English spelling," he writes. For example, what were the monks to do about the sound "th"? The eventual answer lies of course in the letters I have written, but the scribes' solution was to create two new symbols, since discontinued. How was one to distinguish between short and long vowels? The scribes of Norman Britain came up with the silent "e", to turn "hop" into "hope" for example, and with letter doubling, to turn "hoping" to "hopping", or "met" into "meet". Later, writers settled on spellings that indicated the etymology of words, so that "debt" – spelled "det", "dett", "dette", and "deytt" in the 13th century – acquired a "b" in acknowledgment of its origin in the Latin "debitum". Sometimes, spellings came about more whimsically: Caxton's Flemish assistants introduced the "h" in "ghost" from their own language, though they failed to get "ghoos" (goose), "ghoot" (goat), or "gherl" (girl) to stick. Objecting to what he saw as the many illogicalities in British English, Noah Webster set about creating a more pleasing, American orthography – color, center, defense, and so on.
All Crystal's explanations make sense, and are highly entertaining to read, but as they accumulate they are likely to leave readers feeling that they are no closer to the key to improving their own spelling. There are so many exceptional cases. Is a long vowel never followed by a double consonant? No: "droll", "all", "small", and, if you're from the south, "class" and "grass". Can a short vowel be followed by a single consonant? Yes: "criminal", "typical". Does spelling always respect etymological origins? No: the spelling of "island" is based on a mistake. And the rules can be hard to grasp: if a Latin prefix ends in "s", and the root word begins with "s" followed by a consonant, there is no doubling: so, "ascribe" rather than "asscribe". Got that?
Nevertheless, Crystal must be right in saying that children will learn spelling more easily if they understand contexts and etymologies. He dislikes lists, and is hard on Dr Seuss too: names such as Wumbus, Wump, and Zlock serve to baffle, he argues. But he is tolerant of – indeed, enthusiastic about – textspeak.
One supposes that the proofreaders tried to be particularly vigilant when working on this book. I spotted only one error: a missing question mark on page 259.
