There is no pure English language, and to pursue one is to run the risk of 'ethnolinguistic cleansing', says Duncan Wu.

To many people's disgruntlement, St Catherine's College, Oxford, my place of work, is at present a building site, full of enormous skips with the word Aasvogel on their sides. Thanks to David Crystal's excellent The Stories of English I now know that Aasvogel is an Afrikaaner word meaning "vulture". And the impression you get from this book is that the English language is itself a kind of predatory creature with ever-open maw, hoovering up everything within reach so as constantly to reinvent itself: Singlish (Singalese and English), Tex-Mex (English and Mexican Spanish) and Taglish (Tagalog and English) among others.

A number of books are already on the market that purport to narrate the history of the English language, but none is so well informed or as opinionated as this one. And that is why it is so compelling.

Crystal's central argument is that non-standard Englishes are necessary for the continuing good health of the language: that their proliferation is in fact the norm. Far from there being a single story of the language - that of a standard - there have always been a multiplicity of Englishes, ever since the beginning: "We must regard dialect mixing as a normal part of the Old English linguistic situation," he says in chapter two. Like its peoples, English has always been a mongrel tongue, incorporating the contributions of outsiders with ease.

ÁñÁ«ÊÓƵ

Crystal points out that when Anglian, Saxon and Jutish invaders arrived on our shores in the 5th century, they did not bring with them three "pure" Germanic dialects, but "a wide range of spoken varieties, displaying different kinds of mutual influence". Variation was there from the start.

Incidentally, his account of early English is one of the best things about this book. He is particularly good on Old English kennings , the compound nouns devised as synonyms for such words as "sea" - "seal bath" ( seolbpb ), "fish home" ( fiscesepbel ), "swan road" ( swanrad ) and "whale way" ( hwlweg ), for instance.

ÁñÁ«ÊÓƵ

Central to the more polemical side of Crystal's narrative is the perennial demand for greater linguistic "purity". As early as the 13th century, Ranulph Higden, a monk at St Werburgh's in Chester, moaned in his Polychronicon about how English was being corrupted by foreign influences, "by commyxstion and mellyng, furst wipb Danes and afterward wipb Normans, in menye pbe contray longage ys apeyred, and som usepb strange wlaffyng, chyteryng, harryng, and garryng grisbyttyng." That is, "by intermingling and mixing, first with Danes and afterwards with Normans, in many people the language of the land is harmed, and some use strange inarticulate utterance, chattering, snarling, and harsh teeth-gnashing." Wonderful phrase, that - "garryng grisbyttyng" - I must remember it next time I'm at a governing body meeting.

But the irony is that in expressing his distaste of foreign influence, Higden uses a Latin word, commyxstion , and four French ones, mellyng , contray , longage and apeyred . It was inevitable: by the early Middle Ages, Britain was trilingual, using Latin, French and English. Even so, it did not stop people from being snobbish about certain kinds of usages and regional accents. When in 1364 a skinner from Peebles was summoned as a witness in York, his way of speaking was judged to be inherently untrustworthy. By Chaucer's time there was a widespread feeling that the language was in a parlous state, having been "contaminated" by so many different sources. Even Latin was believed to have been somehow tainted as it crossed Europe, so that Constance in The Man of Law's Tale is said to speak "a maner Latyn corrupt".

Far from subsiding, these fears continued to seethe, fuelled by the xenophobia that seems always to have been characteristic of those who occupy these shores. In the 16th century, it was believed that borrowing had gone too far, and that the Germanic word stock needed to be protected.

As a result, such writers as Spenser attempted to revive obsolete Anglo-Saxon words: "algate", "eld", "sicker", "yblent". This was but a short step from the work of the orthographers, orthoepists and grammarians who in the 18th century attempted to standardise the language. Their efforts, as Crystal shows, were in vain. The grammarian could legislate against the splitting of infinitives and placing of prepositions, but had no control over the evolution of the language at home or abroad.

Crystal is particularly good on how the language put down roots in India, the Caribbean, Africa, Asia and Australasia. "Standard English, conceived as a uniform mode of linguistic behaviour uniting English speakers everywhere, began to fragment almost as soon as it had appeared. While Johnson, Lowth, Walker, Sheridan and the other prescriptivists were busy inserting the remaining bars into a cage that they thought would keep English under proper control in Britain, on the other side of the Atlantic the cage door was about to be opened by Noah Webster, who was proposing a different set of linguistic norms for American-English."

Despite the illusory sense so many of today's users have of a golden age of Fowlerian good usage, the language has never been static: pronunciation, grammar and vocabulary all continue to evolve, in defiance of efforts to govern it. As Crystal wryly observes: "There is perpetual instability within the notion of a standard."

Against the suspect impulse to "purify", Crystal praises Dr Johnson, who "was linguistically omnivorous, interested in all varieties of language and indeed in all languages". Johnson revelled in the profusion of dialects, and though his Dictionary did not set out to record them all, he did record a number of Scottish and Irish words, perhaps because his amanuenses used them ( scelerat, stuckle, thrapple ).

ÁñÁ«ÊÓƵ

Crystal is a sort of latter-day Johnson, welcoming of new coinages, and many will find the most engaging part of this book to be its final chapters, which account for the adoption of English as a global tongue and its use on the internet. He argues that we are now in a transitional period in which non-standard and standard Englishes are mutually informing. The correct response is not to go on complaining monkishly, but to adapt the culture so that we can "maintain the literary momentum of the 19th and 20th centuries, and develop a more accepting frame of reference for handling non-standard English". As he makes clear, this means getting away from the prescriptiveness of the grammarians, orthographers and pronunciation experts, instead ensuring that we continue to communicate with accuracy, precision and elegance.

ÁñÁ«ÊÓƵ

This thesis is embedded in what is in essence a history book. It could so easily have been a dry, unrewarding exercise, instead of which it is an exhilarating read, unstuffy and well informed even when discussing female agent nouns, orthographical variation and dialect morphology. It is fascinating, for instance, to hear that changes in pronunciation are led by young, middle-class women, as for instance in the increasing use of the glottal stop: "This in turn leads to the intriguing conclusion drawn by some sociolinguists: it is the pronunciations that women use that become the prestige forms in a language. Men may be the dominant voice in society but their accent has been given a female sanction."

Unlike many other historians, Crystal is attentive to developments in the vernacular and early on notes that writers seldom "exercise much of an influence on the emergence of a linguistic standard for general use".

Indeed, he elsewhere observes that "literature is but the icing on a huge linguistic cake". He thus writes well on such down-to-earth matters as "well"; the survival of "stane" north of the border; and the influence of dialect.

He wears a lifetime's erudition lightly, with no detachment from contemporary culture - noting, among other things, that Yoda in Star Wars regularly inverts his word order, placing the object initially ("If a Jedi knight you will becomeI"), like the Anglo-Saxons. There is an excellent section on Middle Earth, which analyses the standard English used by the Hobbit heroes (Bilbo and Frodo Baggins, Pippin and Merry), alongside the non-standard English of Sam Gamgee, Tom Bombadil and other rustic characters. And in a fascinating chapter on vocabulary, Crystal acknowledges the range of words used in The Sun , pointing out that its use of underprivileged, non-standard forms ("sarnies", "wanna", "yep" and "dammit") has led commentators to accuse it unjustly of "expressive inadequacy".

I cannot help but wonder whether he is a trifle hard on Wordsworth's claims to use the language of the middling or lower classes of working people.

Perhaps those claims were not entirely accurate, but Wordsworth does occasionally use dialect words that he must have heard used by peasant-folk, such as "canty" in Goody Blake and Harry Gill , while the "chatters"/ "waters" rhyme in the first stanza of Resolution and Independence is true to the accent used by ordinary Cumbrians.

But that is a minor point, with no bearing on the effectiveness of the book as a whole. It is impossible to come away from it unimpressed by the vigour with which Crystal affirms English to be "the most etymologically multilingual language on earth". If you doubt that, it is worth musing on these words, all apparently English though in fact absorbed from other lands: caravan (Persia); yoghurt (Turkey); bungalow (India); skunk (Amerindian); assassin (Arabia via Italy); muscle (France); design (Italy); lettuce (Latin); dunk (German); pariah (India again). Just as these have been embraced by the language, so (Crystal argues) we should beware of ghettoising new usages, split infinitives and the like, a reaction that too often betokens "a harmful narrowness of vision". "In its worst excesses, when it focuses on the usage of particular minority groups, it amounts to ethnolinguistic cleansing. It is intolerance masquerading as vigilance."

Duncan Wu is professor of English language and literature, St Catherine's College, Oxford.

ÁñÁ«ÊÓƵ