It is not too difficult to make a success of an Illustrated Encyclopedia of Furry Animals or to create a similarly beguiling Profusely Illustrated Encyclopedia of Technical Wizardry. To use the same format for something as verbal and (text and graphology apart) non-visual as language is to court disaster. Add the fact that we all consider ourselves equipped to pontificate on the subject, then The Cambridge Encyclopedia of the English Language (hereafter CEEL, pronounced “seal” according to its creators) might be a sitting duck. It’s not.

David Crystal begins by recounting the various language stories that show how “English” has evolved from the dialect of Hengist and Horsa to its present status as a multi-form world language. Subsequent sections are devoted to vocabulary (the lexicon, its etymology, structure and “loaded” elements such as taboo language, jargon, slang and political correctness); grammar; the natures of spoken and written English; the variety and usage of the language; and, finally, a section on the ways we acquire and can study English.

Most controversial is CEEL’s own layout. Crystal has enthusiastically embraced a format damned by many critics of children’s information books: the tyranny of the double page spread. “Sentences never cross turn-over pages,” he boasts. “I have tried to ensure that it will be possible for readers to dip into this book at any point.” So each spread offers a self-contained article, associated “panels” of related information, minor articles - and illustrations.

Of the latter, many are fascinating. The reader is quickly absorbed in a newspaper cutting, trying to deduce in which English-speaking country it originated. Cartoons, adverts and medieval manuscripts all show distinctive usages - but do we need a colour picture of a man putting down a carpet to show that “relaid” does not mean the same as “relayed”? And does a photograph of replica Anglo-Saxon huts really help us to decide whether we should refer to “Anglo-Saxon” or “Old English”? Without them, of course, the spreads might look dull.

Many of the panels, however, offer valuable digressions. What was it that started us using abbreviations 150 years ago? Why are males more likely to have monosyllabic first names than females and why do names ending in a vowel tend to be female? How do we explain a language in which the word “colour” in the world of (caucasian) health means red (as in glowing cheeks) and anything but red in snooker? And how come one simple, near redundant letter can mean a kiss, wrong, right (as on a ballot paper), Christian, adult, unknown, multiply and ten?

The author is particularly good on the question of grammar and whether it can be prescriptive. In summary, he suggests (with approval) that we are moving into an age when once again we see the need for a metalanguage (language with which to talk about language) but not for one more suited to the study of Latin.

So, for example, we all need to know something about word classes (not, please, parts of speech) - even if we cannot classify, say, “garden” in “garden party”. (It is no longer a noun because it cannot be pluralized; it cannot be an adjective because it cannot be compared, as in “gardenest”.) This, then, is not a book that can be seized upon as an ally by those who persist in believing that following some arbitrary rules will result in “correct” English. As the author points out, there are over 100 exceptions to the “i before e except after c” mnemonic (ancient, beige, conscience, efficient, eight, etc).

Nor is CEEL concerned only with wayward, standard English. It embraces regional variants from Middle Scots (which developed when Gallic first gave way to English) to the many varieties of American, African, Caribbean and Antipodean English.

There are fascinating introductions to legal and religious English; the language of broadcasting, computing, weather forecasting, sports commentating, jokes and the “delayed single-exchange pseudo-dialogue” of the answerphone. Crystal is especially witty on accent, recalling the judge who apologised for leaving a document in his country cottage. “Fax it up,” said a helpful barrister. “Does rather,” replied the judge.

There are drawbacks. The double-page spread leads to occasional padding and unnecessary lists (do we really need a catalogue of all the dictionaries published in the 1840s?) and the often populist approach results in one or two woolly statements (“Vocabulary is the Everest of a language”). More seriously, despite the useful glossaries and reading lists, the actual index is frustratingly cryptic.

So the result is a book for browsing and study as much as reference. The language of language (hyponymic hierarchies et al) is always explained but is still daunting - so the average school student may need direction in its use but every English teacher will want it for the library, department or personal use as soon as possible. It is a veritable treasure trove of information, examples and asides that will enrich the teaching of language at all levels.