Authorial versus societal mixing

The appearance of words from other languages in a text dominated by one language need not mean that the text is multilingual.  A critic must first determine if the language of the text merely mirrors the language of readers or of the characters in the narrative.  A writer must not be credited with incorporating other languages into a text if the language used actually already is mixed.  Verisimilitude is quite different from authorial invention.

An easy mistake to make, for instance, involves Chaucer, who is often cited as a genius for cleverly incorporating French into his English text, thereby "creating" Middle English.  On the contrary, Chaucer might have been simply using the language of his time; Middle English might have created Chaucer.

Here are William Rothwell's comments on this issue in his “Aspects of Lexical and Morphosyntactical Mixing in the Languages of Medieval England,” in Multilingualism in Later Medieval Britain (2000), edited by D. A. Trotter:

"In a recent book, The Empire of Words:  The Reign of the OED, John Willinsky quotes without demur a passage from Owen Barfield written seventy years ago on the role of English and French in the late fourteenth century.  Barfield writes of the modern poet envying ‘Chaucer with this enormous store of fresh, unspoiled English words ready to his hand and unlimited treasury across the Channel from which he could pick a brand-new one whenever he wanted it.’ Apparently convinced by this fanciful and utopian picture of Chaucer’s working practices, Willinsky simply puts the second edition of the OED through his computer and declares roundly that: ‘Chaucer coins more new words in English than any other author.’  However, he does not understand the reality of the linguistic situation in late medieval England, so that, whilst the computer might be able to pick out for him all the dictionary entries in which Chaucer’s name appears before the earliest attestation of a word, this facile exercise cannot tell him whether such words were already current in England in either French or Latin documents, or whether they even existed in continental French for Chaucer to borrow.  For the computer to provide any meaningful assessment of the growth of English in the Middle Ages, it would have to analyse the entire contents of the Middle English Dictionary, the Dictionary of Medieval Latin from British Sources and the Anglo-Norman Dictionary as they all advance over a long time-scale, steadily recording the true state of written language in medieval England.  It would then have to do the same with the vast dictionaries of continental medieval French in order to see whether developments in England were paralleled by similar growth in France itself.  Both Barfield and Willinsky have their gaze firmly focused on the narrow literary sphere of Middle English, apparently in blissful ignorance of the fact that both French and Latin had been developing in an English society on English soil for hundreds of years, independently of either Rome or Paris, not only  in the general field of imaginative literature, but also in a  wide spectrum of administrative and technical tales."

The New Critics frowned on literary critics moving out of the "work itself," but New Historicism and various other newer approaches to literature have shown us that literature cannot be divorced from history or society.  We have to give credit where credit is due, and that is to writers who really invent language not really used by men and women (to contradict Wordsworth).