I recently read John McWhorter's book Our Magnificent Bastard Tongue: The Untold Story of English, which is an excessively self-conscious revision of the history of the English language. Excessively self-conscious because McWhorter's jabs at the "history of English experts" quickly grow tiresome, and frankly make him sound like something of a crank - the sort of tone that creationists take towards scientists. But an interesting revision because he also explains why a different set of skills allows a different kind of linguist to see a different history to our language.
The primary difference is that McWhorter doesn't specialize in a single language: he specializes in creoles, those curious half-and-half languages that arise when competing native languages occupy the same space. And he argues that English grammar displays the tell-tale signs of miscegenation, a salad bowl of features that tell the history of Angles and Saxons landing in Celt-occupied England in the 5th Century CE, the Viking invasion of same 300 years later, and the Norman conquest of 1066.
These incidents are well-known, of course, but McWhorter believes their effect on the people's speech is not. Traditionally, the Celtic language is believed to have gone extinct and been more or less completely replaced by the speech we know as Old English. The evidence is the extreme shortage of words in English that can be traced to Celtic origins - the subjugated Celts seem to have contributed almost nothing to the invaders' language. In my own linguistics class years ago, my instructor presented this as a mystery: the absence of Celtic words would suggest that the Celts were quickly killed off; however, the archaeological evidence didn't show an unusual amount of warfare and suggested a long period of coexistence. How to resolve the paradox?
The key to the puzzle may be to approach language with less attention to vocabulary and more attention to grammar. This is what McWhorter does. Throughout the book he displays a familiarity with a vast number of different grammars, especially the funky features that are rare and even unique to one language or another. And through grammar, he sees strong evidence that the Celts lived on for many years, kept their own language, and occasionally spoke English. When they did use English, they added a couple twists that seemed natural to themselves, but probably sounded quite strange to the Angles. Eventually, those oddities began to sound familiar and became the common way to speak.
One of these is the "meaningless do," the semantically pointless addition of "to do" in a sentence when we ask a question or state a negative. It's the reason why all other speakers of a Germanic language can say, "Read you the book?" and get the reply, "I read not the book," but if an English speaker tries that, he sounds like he's channeling Yoda (who, to be fair, did learn his English a long time ago, in a galaxy far, far away1). The English speaker says, "Did you read the book?" and hears, "No, I didn't read the book." "Do" adds no meaning to the sentence, but it is necessary in order to be grammatically correct.
It's a weird construction because none of English's close relatives construct their questions and negatives this way. In fact, this sort of thing is rather rare across the globe. But there are a couple dead languages that are known to have used the meaningless do: Celtic and Welsh. The people who were ruled by invading Angles and Saxons, learned the invaders' language, but continued to speak their own at home. What are the odds that English picked up do from their subjects' language rather than inventing in de novo?2
The Vikings had a different effect on English, an effect that McWhorter describes as "battering" the grammar. If you've put some study into German, you know that its nouns and adjectives rarely appear in sentences as they do in the dictionary. Instead, they take all these damned endings: one of three different endings, depending on whether its gender is masculine, feminine, or neuter, and another for the plural. And each of these will change depending on whether the word is a subject, direct object, indirect object, or possessive. Germans don't seem to have trouble with it, but what a pain for the rest of us!
The Vikings found those endings to be a total pain, too. In fact, it appears that they couldn't get the hang of them at all and just ... dumped them. It must have been a horrid-sounding English to the natives, as horrible to them as it would be to me if someone relentlessly used -ed for every past tense (gived, throwed, hitted, etc.). But the Vikings came in waves of all-male groups, married local women, and raised children who heard this nasty tongue from their fathers, if not their mothers. The version without endings spread, across the country and through the generations. English differs from German today because, to some extent, we speak German like lazy, dumb foreigners.
These developments have been obscured because historical linguists tend to look at written texts for their evidence of past languages. And the written record doesn't seem to show these developments. But McWhorter's explanation is plausible enough: writing was uncommon, reserved for formal occasions, and deliberately preserved the ancient modes. The vernacular changed, but the professional scribes did not deign to record it.3
Then came the Normans. For 150 years, French was the lingua franca4 and the style of formal, written Old English was forgotten. When the English cast off French and returned to their native tongue, they began writing it as they spoke. Suddenly, Middle English enters the written record.
And the Norman influence on English? Although they wrote in French, there is documentary testimony that the Norman overlords were speaking more English than French. However, there were only few of them and they didn't mix with their subjects - their influence on English grammar was negligible. The language picked up an enormous number of French words -- it was written, not spoken, French that left its mark -- but this doesn't seem to interest McWhorter very much. He's much more interest in grammar than vocabulary.
At every turn, McWhorter's analysis of English's evolution is grounded in historical circumstance. The Celts added only a few subtle grammatic changes, a consequence of continuing their own speech apart from the English speakers, but over enough centuries to add a few distinctive constructs. The Vikings mixed with the local people, intermarrying and adopting the local language immediately - but badly, in a grammatically stripped-down form. And the Normans arrived in very small numbers, and remained a tiny elite, apart from the rest of the people; their written documents enriched English vocabulary, but left no mark on its grammar.
Finally, recall that McWhorter studies creoles, specializing in the mixing of languages. He has bones to pick with grammatical nitpickers, the sort who fret over split infinitives, prepositions at the end of sentences, who v. whom, or "He done read the book" (my example, not his). I don't know what he thinks of "refudiate," but you can almost hear him laughing at those prissy grammar Nazis5 who think they're speaking some sort of "pure" English, when in fact their speech is descended through a long line of folks who just didn't know how to talk good.
[Addendum]
As a followup to Heidi's comment, here's a map showing the maximum spread of the Celts. As you can see, they had a pretty wide influence in Europe, although not so much the Scandinavian countries:
_________________________________
1. Although Yoda, as I recall, sometimes uses "do" in the modern manner. He's had a long time to listen to those young Jedi.
2. The progressive "He is reading" is another form that existed in Celtic and today is peculiar to English.
3. It should be no surprise to academics. After all, they're the ones who have to teach their students "You don't write like you talk, dude, or your grade'll suck!"
4. Yes! Pun intended! Would you doubt me?
5. Which, yes, often describes me.
Final Demolition of Kresge Buildings, University of Michigan, July 2010.
