Our Magnificent Bastard Tongue

by John Mcwhorter

  But one Germanic language doesn’t care so much about dotting i’s and crossing t’s. And it used to. Old English had a good old-fashioned trio: hider, þider, and hwider. These were passed down into Middle English as hither, thither, and whither. But they eventually blew away like autumn leaves. Today they are gone.

  Quite a few European languages have a word that refers to people in a generic sense. Spanish’s Se habla español is the most familiar example: se here means “you” in the sense of “one.” In French this is on. In German it is man: Hier (not her!) spricht man Deutsch (“One speaks German here”). As it is in most of the other Germanic languages (an exception is that in Icelandic the word for men, maður, subs for man). This means that in Germanic languages there is almost always a nice, filled-out array of pronouns making lots of distinctions, like in Swedish:

  In comparison, English settles for making poor you do an awful lot of work:

  Notice that while Swedish has its “se habla” pronoun man, in English we drag you in to do that job: You have to be careful with these big corporations. In Old English, though, there was a man pronoun, too. But in Middle English documents, over three hundred years it blows away like autumn leaves. Today it is gone.

  Learn a European language, including any Germanic language but Swedish, and note that quite often, while most verbs form their past perfect with the verb have—Ich habe gesprochen (“I have spoken”)—a good little bunch do it with the verb be, too—Ich bin gekommen (“I ‘am come’ ”). Just like in Old English: Learning had fallen away was “Learning was fallen away”: Lār āfeallen wæs.

  Marking some verbs with be instead of have is a matter of being explicit about a certain nuance: in the perfect, the verbs marked with be refer, technically, to a state rather than an action; i.e., something that bes. When you say you have arrived, you mean that you have now achieved the state of being there: “I’m here, so let’s get started.” On the other hand, when you talk about how you raked leaves this afternoon, you usually are getting across that you performed the action of raking leaves, not that you have achieved the state of having raked the leaves and are now ready to have your picture taken.

  We English speakers think, “Well, yeah . . .” but hardly feel it necessary to split that hair. The other Germanic languages do split it—and Old English did.

  But something strange started happening in Middle English, as usual; now it was the be-perfect that was falling away (like autumn leaves). By Shakespeare, be is used with only a few verbs (“And didst thou not, when she was gone downstairs, desire me to be no more so familiarity with such poor people?” Henry IV, Part II, II, i, 96) and today, it lingers on only in a frozen form such as The autumn leaves now are gone. Even there, you may well have thought of gone as an adjective (The leaves are red, The leaves are gone), and in any case you can also say The autumn leaves have gone, which, in this case of the grand old Old English be-perfect, they have, as always in English.

  In any self-respecting member of the Germanic family, one (man?) puts the verb in the second slot in the sentence, hell or high water. So for I saw a film, German has Ich sah einen Film—nothing odd there. But if you want to say Yesterday I saw a movie, “saw” has to stay in that second slot, and so “I” has to come after it: Gestern sah ich einen Film (“Yesterday saw I a film”). The verb sits tight in that second slot and everything else has to manage. In all the Germanic languages it has to be “Yesterday saw I . . .” to keep the verb in second place. Swedish for Today she’s driving the car has to be:

  which gets kører (“drives”) into that number two slot after I dag (“today”).

  I dag kører hun bilen

  today drives she the car

  This quirk of word order, which linguists call “verb-second” or “V2” for short, is by no means common in the world, and to my knowledge is only a family trait today in Germanic, in which it is as normal as Apfel pie. The details differ from language to language, but all Germanic languages have it—except one. Its absence in that one (guess which one!!) is odd. Although, given that the one it is absent from also shucked off so much else, maybe it’s not odd. Maybe there is a reason behind all of this.

  English’s autumnal leaf-dropping quality involves even more cases,6 but I need not list them all: you get the point. No Germanic language has shed as much of what Proto-Germanic passed down to it as English, by a long shot. Of course some drop a stitch here and there more than others. Afrikaans has no gender, because it is what happened when Dutch was learned by so many Africans that, unlike any Germanic language on the Continent, it went as far as English did and lost gender. However, in terms of the many other features that make a language a descendant of Proto-Germanic, Afrikaans is very much a card-carrying member: in Afrikaans, you “remember yourself,” you come hither, there is a nice man pronoun, a be-perfect, and the V2 tic. Swedish, as noted, has lost its be-perfect (although it holds on in Norwegian and Danish). However, Swedish is otherwise as Germanic as, well, German.

  English’s grammar, then, is “easier” than the other Germanic languages’. The Grand Old History of English describes these “difficult” features as just mysteriously melting away. But none of these authors have had occasion to consider how very many such features just melted away, and that nothing similar was happening in other Germanic languages. The question beckons: why has English been so strangely prone to just letting it all go?

  Back in the twenties, pioneering linguist Edward Sapir groped at the question in an elegantly put discussion of the whither/hither/thither case:

  They could not persist in live usage because they impinged too solidly upon the circles of meaning represented by the words where, here and there. That we add to where an important nuance of direction irritates rather than satisfies.

  Sapir’s writing, as always, satisfies—but it does leave a question as to why, oh why, speakers of this and only this Germanic language found nuance so irritating. Scholarship on English has proceeded with about as little interest in that question as Sapir evidently had. Yet the question has an answer. It’s as much a part of the story of the English language as Chaucer and Caxton.

  Whodunit?

  When I was about eight, I remember letting a neighborhood friend take a spin on my bike. He was a more highly spirited fellow than me and gave it a good zip up and down hills, bumping it down some curbs, doing “pop-a-wheelies” and so on. Finally he skidded to a stop in front of me and some of our pals. We heard some screw or washer from somewhere in the bike clink to the ground. Then, a pedal fell off, followed by the handlebars. The seat screws went loose and the seat tipped limply forward, and finally the back wheel fell out. My friend ended up squatting on a mangled heap of bars and gears, and we and our friends watching howled with laughter for the next fifteen minutes.

  I swear I remember this, and yet upon reflection I can’t help suspecting that the memory has been distorted in my mind over time. After all, bicycles do not just fall to pieces like that—it would be like something out of a Looney Tune. Maybe something happened just to the seat, or just a wheel slipped—but with all due acknowledgment of entropy, the bike cannot have completely disintegrated right under him. Why, after all, would a bike do that?

  Languages are no more likely to toss off massive amounts of grammatical features than bikes are to fall to dust. For example, in language groups other than Germanic, there is never one language that just miraculously becomes a stripper. Supposedly, what happened to English is so unremarkable. But the 250 languages of Australian Aborigines are known for having lots of suffixes, and even though the languages have been spoken there for several tens of thousands of years, not a single one has drifted into a state like English’s.

  What this means is that something happened to English. Someone did something to it. If a bike does collapse under its rider, then we know that earlier that day, somebody loosened all of its screws so that it would fall apart after being ridden hard for a while. Somebody unscrewed English. Attention must be paid.


  Let’s pay some, and line up the suspects. It has, actually, been bruited about that English was turned into a simpler language by the Norman French. The idea is tempting, but impossible. There were, for one, never all that many Normans on the ground in England—one estimate is about ten thousand amid a British population of one or two million. The Normans were an elite living amid masses of ordinary people speaking English as they always had. Thus, even if Normans tended to speak English in an inaccurate way, there is no reason that English-speaking folk would imitate them—if they ever even met them.

  This even includes people as influential as the kings of England who, for a while, were men of Norman birth who likely did not even speak English. Think about it—let’s imagine that the king does speak English, but as a second language, like, say, Maurice Chevalier. You, on the other hand, talk like Eric Idle. If you ever actually heard the king speak—and that’s a big “if,” especially since there’s no radio or TV—no matter what esteem you hold the king in, why would you start walking around talking like him, so consistently that your children hear you talking only that way? If by chance you really were so odd a person, what would the chances be that whole villages would take to doing what you were doing, 24/7, for a century?

  Besides, evidence suggests that the Normans didn’t speak funny English for long anyway. By a hundred years and change after the Battle of Hastings in 1066, there are reports of Normans needing to have French taught formally to their children, and of people of Norman ancestry speaking good English like anyone else. By the early 1300s, William of Nassyngton famously had it that:

  And somme understonde wel Englysch

  that can nother Latyn nor Frankys.

  Bothe lered and lewed, olde and gonge,

  Alle understonden english tonge.

  Lewed, by the way, meant “unlearned”—neat how the word has evolved into its modern Hustler connotation. And gonge, by the way, was young.

  We can assume, then, that the Norman impact on English was in terms of words, and lots of them. That’s old news. Who beat up English’s grammar?

  The Viking Impact

  Here is where our Vikings come in. Grown men raised on Old Norse were suddenly faced with having to do their raggedy best speaking Englisc on a regular basis whenever they spoke with anyone besides the guys they came over with. The simple fact is that adults have a harder time learning languages than children and teenagers—and this was an era when there was no Berlitz, no language instruction beyond someone on the fly telling you, “Here’s the word for . . . ,” and for the most part, not even any writing.

  They came in one wave after another over a century—for generations there were ever new hordes of men from across the sea not speaking the language right. Crucially, whereas French came to England as an elite language spoken by rulers living remotely from the common folk, the Vikings took root on the ground, often marrying English-speaking women, such that their children actually heard quite a bit of their “off” English. All of this had an effect on the English language.

  The waves in question started in 787, Danes on the eastern side and Norwegians round the western one. For the next hundred years England coped with increasing numbers of these invaders, culminating in an agreement in 886 that the Vikings would confine their dominion to the northern and eastern half of England, thence termed the Danelaw.

  The power that the Vikings wielded is clear in traditionally noted things such as the proliferation of Scandinavian-derived place names in the Danelaw area ending in -by and -thorp, and names ending in -son (like Orm Gamalson, he of the sundial), as well as transformations of bureaucratic procedure. These things alone, however, cannot, in the strict sense, tell us much about whether these people were passing their rendition of English down to new generations of people of both Scandinavian and Anglo-Saxon (and Celtic) descent. Power can be wielded by almost counterintuitively small numbers of people, and thus have no effect on how everyday language is spoken.

  For example, China was ruled by foreigners for much of its history, including the famous Mongolian regime of Genghis Khan, as well as Manchus from 1644 to as recently as 1911. However, the languages of the rulers had no effect on Chinese. The foreigners ruled from their compounds, using interpreters to communicate with the outside. Actual Chinese people largely encountered the foreigners in occasional interactions with soldiers—if at all. Chinese as spoken by millions across a vast land was unaffected.7 In Africa, colonial languages, like English, French, and Portuguese, have certainly poured words into small local languages—but they have had almost no effect on these languages’ grammars. How one uses the grammar in, say, Chichewa in Malawi has nothing to do with the English Malawians learn in school and see in the movies.

  As such, even the fact that the very king of England was Danish for a spell (Cnut, from 1016 to 1035) can tell us nothing as to whether his native Old Norse had any significant effect upon the grammar of English as spoken by everyday people. What we need is evidence that Scandinavians speaking incomplete English would have been so common that children would have heard this faulty rendition as much as, if not more than, regular English—to the extent that “foreigner” English affected what they grew up using as everyday speech.

  Imagining this requires putting on our “antique” glasses: children in this era did not go to school, did not read, and there was no “standard English” that they encountered in the media, because in ordinary daily life there was no media to speak of. To children in Anglo-Saxon Britain, language was something you heard people around you—and no one else—speaking. You didn’t see it on the page. In getting a sense of how in such a setting the Vikings would have passed on their rendition of English to the ages (and, eventually, to the page you are reading), three things are useful.

  First, in many places they were quite densely concentrated: in some parts of the Danelaw most people were of Danish ancestry. This means that “Scandi”-sounding English would have been a matter of not just the occasional Dane or Norwegian here and there (“Mommie, hwy spæketh he like thæt?”),8 but a critical mass of people.

  Second, in documents, we clearly see that English gets simpler first in the north—where the Scandinavians were densely settled. Old English came in at least four dialects. The one usually written in was West Saxon, which is to us today “normal” Old English. But one of the dialects spoken in the Danelaw region was Northumbrian. In Northumbrian toward the end of Old English, as the Battle of Hastings was looming, the conjugational endings were already wearing out, as if someone were having trouble keeping them apart. Sometimes, all the endings in the present tense except the first person singular were the same, -as. This is Old English? And by Middle English, in the north, this erosion continued—in the plural the final consonant flaked away, leaving a mere -e:

  But take a look at what Southern Middle English was still like, where there had been no Vikings—normal Old English:

  So it’s not that the endings just fell apart all over England out of some kind of guaranteed obsolescence. They fell apart in a particular place—where legions of foreigners were mangling the tongue!

  It was the same with gender: it starts flaking away in Northumbrian Old English, while down south all three genders held on possibly into Middle English. Even as recently as the late nineteenth century, rural folk in the extreme southwest in Dorset were still dividing things between a “personal” and an “impersonal” gender. “Personal” things were not only people but all living things and, for some reason, tools. So, of a tree, He’s a-cut down. But of water, It’s a-dried up. Even demonstratives still came in two genders: this water was impersonal, but in the personal one, said theäse tree. Also, the V2 rule started unwinding, predictably, in the north; it held on in the south, including in Kent in the southeast, for much longer.

  Finally, our friend Orm Gamalson even left us a crucial window into English as rendered by speakers of Old Norse. Gamalson was writing in the Northumbrian dialect, in which noun suffixes were as much a mes
s as the verb ones were. In other dialects of Old English, when one wrote of something happening on a ship, ship was in the dative and took an -e ending: scipe. But in Northumbrian, one just said in scip: the -e was gone. In the same way, in Gamalson’s inscription, as it goes on from where we left off with it earlier, he places the rebuilding of the minster “in King Edward’s days,” which he wrote as in Eadward dagum. There are two interesting things about those three words.

  First, Orm left off the possessive ending. Just as today, in other Old English sources Eadward would take an -s suffix to indicate the possessive: in Eadwardes dagum. In Orm’s part of England people were leaving off endings—but not elsewhere. Orm’s part of England was, also, where, well, Orm was. It was Scandi-land, where people not raised in English were speaking it as part of the everyday routine, leaving the niceties off.

  But second, the -um suffix on dagum is revealing. It is a dative plural, and in Northumbrian, this solid suffix hangs on even while the others are wearing away in cases like scip for scipe and Eadward for Eadwardes. Even as the Old English era is winding down and even in other Old English dialects, this -um suffix is starting to undergo natural wear and tear and morph into the likes of -en, in Northumbrian it’s always right there as -um, shining like a star.

  And there’s a reason. You can see it in this table. We come back to Old English’s stān for stone, and armr in Old Norse meant (get ready . . . !) “arm”: