The Mother Tongue

by Bill Bryson

  Some of the most interesting theories about language development in recent years have been put forward by Derek Bickerton, an English-born professor at the University of Hawaii, who noticed that creole languages all over the world bear certain remarkable similarities. First, it is important to understand the difference between pidgins and creoles. Pidgins (the word is thought to be a Chinese rendering of the English word business) are rudimentary languages formed when people from diverse backgrounds are thrown together by circumstance. Historically, they have tended to arise on isolated plantation-based islands which have been ruled by a dominant Western minority but where the laborers come from a mixed linguistic background. Pidgins are almost always very basic and their structure varies considerably from place to place—and indeed from person to person. They are essentially little more than the language you or I would speak if we found ourselves suddenly deposited in some place like Bulgaria or Azerbaijan. They are makeshift tongues and as a result they seldom last long.

  When children are born into a pidgin community, one of two things will happen. Either the children will learn the language of the ruling class, as was almost always the case with African slaves in the American South, or they will develop a creole (from French créole, “native”). Most of the languages that people think of as pidgins are in fact creoles. To the uninitiated they can seem primitive, even comical. In Neo-Melanesian, an English-based creole of Papua New Guinea, the word for beard is gras bilong fes (literally “grass that belongs to the face”) and the word for a vein or artery is rop belong blut (“rope that belongs to the blood”). In African creoles you can find such arresting expressions as bak sit drayva (“back seat driver”), wesmata (“what’s the matter?”), and bottom-bottom wata waka (“submarine”). In Krio, spoken in Sierra Leone, stomach gas is bad briz, while to pass gas is to pul bad briz. Feel free to smile. But it would be a mistake to consider these languages substandard because of their curious vocabularies. They are as formalized, efficient, and expressive as any other language—and often more so. As Bickerton notes, most creoles can express subtleties of action not available in English. For instance, in English we are not very good at distinguishing desire from accomplishment in the past tense. In the sentence “I went to the store to buy a shirt” we cannot tell whether the shirt was bought or not. But in all creoles such ambiguity is impossible. In Hawaiian creole the person who bought a shirt would say, “I bin go store go buy shirt,” while the person who failed to buy a shirt would say, “I bin go store for buy shirt.” The distinction is crucial.

  So creoles are not in any way inferior. In fact, it is worth remembering that many full-fledged languages—the Afrikaans of South Africa, the Chinese of Macao, and the Swahili of east Africa—were originally creoles.

  In studying creoles, Bickerton noticed that they are very similar in structure to the language of children between the ages of two and four. At that age, children are prone to make certain basic errors in their speech, such as using double negatives and experiencing confusion with irregular plurals so that they say “feets” and “sheeps.” At the same time, certain fairly complicated aspects of grammar, which we might reasonably expect to befuddle children, cause them no trouble at all. One is the ability to distinguish between stative and nonstative verbs with a present participle. Without getting too technical about it, this means that with certain types of verbs we use a present participle to create sentences like “I am going for a walk” but with other verbs we dispense with the present participle, which is why we say “I like you” and not “I am liking you.” Very probably you have never thought about this before. The reason you have never thought about it is that it is seemingly instinctive. Most children have mastered the distinction between stative and nonstative verbs by the age of two and are never troubled by it again. Intriguingly, all creole languages make precisely the same distinction.

  All of this would seem to suggest that certain properties of language are innate. Moreover, as we have seen, it appears that the earth’s languages may be more closely related than once thought. The links between languages—between, say, German bruder, English brother, Gaelic bhrathair, Sanskrit bhrata, and Persian biradar—seem self-evident to us today but it hasn’t always been so. The science of historical linguistics, like so much else, owes its beginnings to the work of an amateur enthusiast, in this case an Englishman named Sir William Jones.

  Dispatched to India as a judge in 1783, Jones whiled away his evenings by teaching himself Sanskrit. On the face of it, this was an odd and impractical thing to do since Sanskrit was a dead language and had been for many centuries. That so much of it survived at all was in large part due to the efforts of priests who memorized its sacred hymns, the Vedas, and passed them on from one generation to the next for hundreds of years even though the words had no meaning for them. These texts represent some of the oldest writings in any Indo-European language. Jones noticed many striking similarities between Sanskrit and European languages—the Sanskrit word for birch, for instance, was bhurja. The Sanskrit for king, raja, is close to the Latin rex. The Sanskrit for ten, dasa, is reminiscent of the Latin decem. And so on. All of these clearly suggested a common historical parentage. Jones looked at other languages and discovered further similarities. In a landmark speech to the Asiatick Society in Calcutta he proposed that many of the classical languages—among them Sanskrit, Greek, Latin, Gothic, Celtic, and Persian—must spring from the same source. This was a bold assertion since nothing in recorded history would encourage such a conclusion, and it excited great interest among scholars all over Europe. The next century saw a feverish effort to track down the parent language, Indo-European, as it was soon called. Scores of people became involved, including noted scholars such as the Germans Friedrich von Schlegel and Jacob Grimm (yes, he of the fairy tales, though philology was his first love) and the splendidly named Franz Bopp. But, once again, some of the most important breakthroughs were the work of inspired amateurs, among them Henry Rawlinson, an official with the British East India Company, who deciphered ancient Persian more or less single-handed, and, somewhat later, Michael Ventris, an English architect who deciphered the famously difficult Linear B script of ancient Minoa, which had flummoxed generations of academics.

  These achievements are all the more remarkable when you consider that often they were made using the merest fragments—of ancient Thracian, an important language spoken over a wide area until as recently as the Middle Ages, we have just twenty-five words—and in the face of remarkable indifference on the part of the ancient Greeks and Romans, neither of whom ever bothered to note the details of a single other language. The Romans even allowed Etruscan, a language that had greatly contributed to their own, to be lost, so that today Etruscan writings remain tantalizingly untranslated.

  Nor can we read any Indo-European writings, for the simple reason that not a scrap exists. Everything we know—or, to be more precise, think we know—is based on conjecture, on finding common strands in modern-day languages and tracing these strands to a hypothetical mother tongue, Proto-Indo-European, which may never even have existed. The lack of documentary evidence isn’t too surprising when you bear in mind that we are going back an awfully long time. The early Indo-Europeans were Neolithic—that is, late Stone Age—people who can be dated back to about 7000 B.C. The descended languages of Indo-European almost always show some kind of kinship in their names for primary family relationships, such as mother and father; for parts of the body, such as eye, foot, heart, and ear; for common animals, such as goat and ox; and for natural elements, such as snow, thunder, and fire. We can deduce something about how these people lived from these cognates. They had a common word for snow and cold, so the climate obviously was not tropical, and yet they appear to have had no common word for sea. Those tribes that reached the sea each came up with words of their own, so presumably they began their migration from a point well inland. Among the other words held in common are oak, beech, birch, willow, bear, wolf, deer, rabbit, sheep, goat,
pig, and dog. They had no common word for horse or window. By studying the known range of certain flora and fauna, linguists have placed their original homeland in various places: the Russian steppes, Scandinavia, central Europe, the Danube valley, Asia Minor—indeed, almost everywhere.

  Their common existence is thought to have ended between 3500 and 2500 B.C., when they began to fan out across Europe and Asia. For the most part these were probably not great exoduses but rather gradual encroachments as each new generation sought new pastures and hunting areas. Over the millennia they spread over wide areas—even reaching China. Explorers at the turn of the century were astonished to find a cache of Buddhist documents written in two related but unknown languages in what is now the Chinese province of Sinkiang, along the Old Silk Road. The languages, which they called Tocharian, were clearly Indo-European, as can be seen, for instance, in their words for the number three: tre and trai. As the centuries passed, the original Indo-European language split into a dozen broad groups: Celtic, Germanic, Greek, Indo-Iranian, Slavonic, Thraco-Illyrian, and so on. These further subdivided into literally scores of new languages, ranging from Latin to Faroese to Parthian to Armenian to Hindi to Portuguese. It is remarkable to reflect that people as various as a Gaelic-speaking Scottish Highlander and a Sinhalese-speaking Sri Lankan both use languages that can be traced directly back to the same starting point. With this in mind, it is perhaps little wonder that the Greeks and Romans had no idea that they were speaking languages that were cousins of the barbarian tongues all around them. The notion would have left them dumbfounded. Just within Europe the degree of divergency is so great that only relatively recently have two languages, Albanian and Armenian, been identified as being Indo-European.

  Of all the Indo-European languages, Lithuanian is the one that has changed the least—so much so that it is sometimes said a Lithuanian can understand simple phrases in Sanskrit. At the very least, Lithuanian has preserved many more of the inflectional complexities of the original Indo-European language than others of the family.

  English is part of the Germanic family, which gradually split into three branches. These were North Germanic, consisting of the Scandinavian languages; West Germanic, consisting principally of English, German, and Dutch (but also Frisian, Flemish, and other related dialects); and East Germanic, whose three component languages, Burgundian, Gothic, and Vandalic, died off one by one. Many other European languages disappeared over time, among them Cornish, Manx, Gaulish, Lydian, Oscan, Umbrian, and two that once dominated Europe, Celtic and Latin.

  Celtic, I must hasten to add, is not dead. Far from it. It is still spoken by half a million people in Europe. But they are scattered over a wide area and its influence is negligible. At its height, in about 400 B.C., Celtic was spoken over a vast area of the continent, a fact reflected in scores of place-names from Belgrade to Paris to Dundee, all of which commemorate Celtic tribes. But from that point on, its dominions have been constantly eroded, largely because the Celts were a loose collection of tribes and not a great nation state, so they were easily divided and conquered. Even now the various branches of Celtic are not always mutually comprehensible. Celtic speakers in Scotland, for instance, cannot understand the Celtic speakers of Wales a hundred miles to the south. Today Celtic survives in scattered outposts along the westernmost fringes of Europe—on the bleak Hebridean Islands and coastal areas of Scotland, in shrinking pockets of Galway, Mayo, Kerry, and Donegal in Ireland, in mostly remote areas of Wales, and on the Brittany peninsula of northwest France. Everywhere it is a story of inexorable decline. At the turn of the century Cape Breton Island in Nova Scotia had 100,000 Gaelic speakers—most of them driven there by the forced clearances of the Scottish Highlands—but now Gaelic is extinct there as a means of daily discourse.

  Latin, in distinct contrast, didn’t so much decline as evolve. It became the Romance languages. It is not too much to say that French, Italian, Spanish, Portuguese, and Romanian (as well as a dozen or so minor languages/dialects like Provençal and Catalan) are essentially modern versions of Latin. If we must fix a date for when Latin stopped being Latin and instead became these other languages, the year 813 is a convenient milestone. It was then that Charlemagne ordered that sermons throughout his realm be delivered in the “lingua romana rustica” and not the customary “lingua latina.” But of course you cannot draw a line and say that the language was Latin on this side and Italian or French on that. As late as the thirteenth century, Dante was still regarding his own Florentine tongue as Latin. And indeed it is still possible to construct long passages of modern Italian that are identical to ancient Latin.

  The Romance languages are not the outgrowths of the elegant, measured prose of Cicero, but rather the language of the streets and of the common person, the Latin vulgate. The word for horse in literary Latin was equus, but to the man on the street it was caballus, and it was from this that we get the French cheval, the Spanish caballo, and the Italian cavallo. Similarly, the classical term for head was caput (from which we get capital and per capita), but the street term was testa, a kind of pot, from which comes the French la tête and the Italian la testa (though the Italians also use il capo). Cat in classical Latin was feles (whence feline), but in the vulgate it was cattus. Our word “salary” comes literally from the vulgar Latin salarium, “salt money”—the Roman soldier’s ironic term for what it would buy. By the same process the classical pugna (from which we much later took pugnacious) was replaced by the slangy battualia (from which we get battle), and the classical urbs, meaning “city” (from which we get urban), was superseded by villa (from which the French get their name for a city, ville, and we take the name for a place in the country).

  The grammar of the vulgate also became simplified as Latin spread across the known world and was adopted by people from varying speech backgrounds. In classical Latin word endings were constantly changing to reflect syntax: A speaker could distinguish between, say, “in the house” and “to the house” by varying the ending on house. But gradually people decided that it was simpler to leave house uninflected and put ad in front of it for “to,” in for “in,” and so on through all the prepositions, and by this means the case endings disappeared. An almost identical process happened with English later.

  Romanians often claim to have the language that most closely resembles ancient Latin. But in fact, according to Mario Pei, if you wish to hear what ancient Latin sounded like, you should listen to Lugudorese, an Italic dialect spoken in central Sardinia, which in many respects is unchanged from the Latin of 1,500 years ago.

  Many scholars believe that classical Latin was spoken by almost no one—that it was used exclusively as a literary and scholarly language. Certainly such evidence as we have of everyday writing—graffiti on the walls of Pompeii, for example—suggests that classical Latin was effectively a dead language as far as common discourse was concerned long before Rome fell. And, as we shall see, it was that momentous event—the fall of Rome—that helped to usher in our own tongue.

  3.

  Global Language

  All languages have the same purpose—to communicate thoughts—and yet they achieve this single aim in a multiplicity of ways. It appears there is no feature of grammar or syntax that is indispensable or universal. The ways of dealing with matters of number, tense, case, gender, and the like are wondrously various from one tongue to the next. Many languages manage without quite basic grammatical or lexical features, while others burden themselves with remarkable complexities. A Welsh speaker must choose between five ways of saying than: na, n’, nag, mwy, or yn fwy. Finnish has fifteen case forms, so every noun varies depending on whether it is nominative, accusative, allative, inessive, comitative, or one of ten other grammatical conditions. Imagine learning fifteen ways of spelling cat, dog, house, and so on. English, by contrast, has abandoned case forms, except for possessives, where we generally add ’s, and with personal pronouns which can vary by no more than three ways (e.g., they, their, them), but often by only two
(you, your). Similarly, in English ride has just five forms (ride, rides, rode, riding, ridden); the same verb in German has sixteen. In Russian, nouns can have up to twelve inflections and adjectives as many as sixteen. In English adjectives have just one invariable form with but, I believe, one exception: blond/blonde.

  Sometimes languages fail to acquire what may seem to us quite basic terms. The Romans had no word for gray. To them it was another shade of dark blue or dark green. Irish Gaelic possesses no equivalent of yes or no. They must resort to roundabout expressions such as “I think not” and “This is so.” Italians cannot distinguish between a niece and a granddaughter or between a nephew and a grandson. The Japanese have no definite or indefinite articles corresponding to the English a, an, or the, and they do not distinguish between singular and plural as we do with, say, ball/balls and child/children or as the French do with chateau/chateaux. This may seem strange until you reflect that we don’t make a distinction with a lot of words—sheep, deer, trout, Swiss, scissors—and it scarcely ever causes us trouble. We could probably get by well without it for all words. But it is harder to make a case for the absence in Japanese of a future tense. To them Tokyo e yukimasu means both “I go to Tokyo” and “I will go to Tokyo.” To understand which sense is intended, you need to know the context. This lack of explicitness is a feature of Japanese—even to the point that they seldom use personal pronouns like me, my, and yours. Such words exist, but the Japanese employ them so sparingly that they might as well not have them. Over half of all Japanese sentences have no subject. They dislike giving a straightforward yes or no. It is no wonder that they are so often called inscrutable.

  Not only did various speech communities devise different languages, but also different cultural predispositions to go with them. Speakers from the Mediterranean region, for instance, like to put their faces very close, relatively speaking, to those they are addressing. A common scene when people from southern Europe and northern Europe are conversing, as at a cocktail party, is for the latter to spend the entire conversation stealthily retreating, to try to gain some space, and for the former to keep advancing to close the gap. Neither speaker may even be aware of it. There are more of these speech conventions than you might suppose. English speakers dread silence. We are all familiar with the uncomfortable feeling that overcomes us when a conversation palls. Studies have shown that when a pause reaches four seconds, one or more of the conversationalists will invariably blurt something—a fatuous comment on the weather, a startled cry of “Gosh, is that the time?”—rather than let the silence extend to a fifth second.