THE MOTHER TONGUE
by William F. Allman, et al.

Linguists are working back from modern speech to re-create the first language of the human race

     In 1786, Sir William Jones, an Englishman serving the Crown as a judge in India, turned a series of seeming coincidences into an extraordinary discovery about human nature. A scholar of the Orient by training, Jones had embarked on an effort to learn Sanskrit, the language in which many ancient Indian religious and literary texts are written. To his amazement, Jones found that Sanskrit's grammatical forms and vocabulary bore a striking resemblance to those of Greek and Latin, so much so that "no philologer could examine them all three without believing them to have sprung from some common source." As Charles Darwin was to assert almost a century later about the human body, Jones suggested that a fundamental part of the human psyche--language--had a hidden ancestry of its own.

     Today, scientists are leading a new revolution in understanding the roots of language. While linguistic pioneer Noam Chomsky and his followers have focused on language as a psychological phenomenon, a small band of renegade scholars is revealing how languages are a product of cultural evolution. Sifting through modern tongues for linguistic "fossils" in the form of common words and grammatical structures, these "linguistic paleoanthropologists," many of whom have worked in obscurity in the Soviet Union, are reconstructing the pathways by which the world's roughly 5,000 languages arose from a handful of ancient "mother" tongues. A few radical linguists have gone even further, claiming they have reconstructed pieces of the mother of them all: The original language spoken at the dawn of the human species.

     These linguistic findings are a windfall for archaeologists, anthropologists and other social scientists who are trying to piece together the story of the peopling of the earth. "We've come to realize," says Alexis Manaster Ramer, a researcher at Wayne State University in Detroit, "that a lot of the answers to the big questions lie in something you might call anthro-psycho-socio-linguistics." Language is an integral part of the cultural glue that binds people together and signals their presence. Tracing the evolution of language can reveal how ancient peoples migrated into new lands, for instance, just as reconstructing the vocabularies of lost languages can give researchers clues to what ancient people saw, ate and thought, or how one culture coexisted--or collided--with another. The new linguistic findings also neatly dovetail with conclusions drawn from a very different area of evolutionary research. Comparisons of human genes worldwide have produced a "family tree" of the human race whose branches closely mirror the branching of languages proposed by linguists, leading to the startling suggestion that all people--and perhaps all languages--are descended from a tiny population that lived in Africa some 200,000 years ago.

     English pedigree. The idea that languages are constantly evolving is obvious from looking at English over time: Consider Shakespeare's Elizabethan "Shall I compare thee to a summer's day?"; Chaucer's 14th-century, Middle English "Whan that Aprille with his shourse sote," and the opening line from the eighth-century Old English epic BEOWULF: "Hwaet! We Gar-Dena, in geardagum."

     These dramatic sound changes within a single language are possible only because, with the exception of onomatopoeic words like sizzle, the sound of a word has no direct connection to its meaning, says Merritt Ruhlen, a scholar in Palo Alto, Calif., who is tracing the relationships among the world's languages. Like coins, words get their value from the community at large, which must agree on what they represent. The word dog may mean a furry creature with four legs and a wagging tail, for instance, but hippopotamus or ziglot would serve just as well, as long as both speaker and listener agreed on its meaning.

     Cream in your coffee. These arbitrary associations between sounds and meanings provide the key to reconstructing the linguistic past, says Ruhlen. Because any number of sounds could be associated with a particular meaning, the presence of similar- sounding words with similar meanings in two different languages suggests that both languages had a common ancestor. For instance, diners might order their coffee au lait, con leche or latte, depending on whether they are in a French, Spanish or Italian restaurant. Using these similar-sounding "daughter" words for milk and a knowledge of how the sounds of words change as languages evolve, linguists could come close to reconstructing the Latin form, lacte, even if this mother tongue of Romance languages were unknown.

     Similar comparisons among words are what led Jones to suspect that Latin, along with Greek and Sanskrit, had descended from an even more ancient mother tongue. The word for the number three, for instance, is tres in Latin, treis in Greek and tryas in Sanskrit. Over the years, scholars following up on Jones's suggestion have demonstrated that dozens of languages, including English, Swedish, German, Russian, Polish, Hindi, Persian, Welsh and Lithuanian, are all descendants of this same ancient "proto-language." Called Indo-European by linguists, this mother tongue was spoken some 8,000 years ago before the invention of writing, and is known only by the traces left behind in the vocabularies of its daughter languages.

     From these remnants, however, linguists have reconstructed a vast lexicon of proto-Indo-European words, providing clues to the origins of the ancient people who spoke the language when they populated nearly all of Europe. According to recent work by two Soviet linguists, Thomas Gamkrelidze and Vyacheslav Ivanov, words for domesticated animals such as cows, sheep and dogs as well as plants such as barley, flax and wheat suggest that the people who spoke proto-Indo-European were farmers. Likewise, the prevalence of words evoking mountains and rapidly flowing rivers suggests the Indo-European people originally lived in a hilly terrain.

     Using these and other linguistic clues, Soviet researchers have offered new evidence that Indo-European originated in an area known as Anatolia, which is now part of Turkey, and from there spread throughout Europe and the sub-Continent. Linguists had long thought that the Indo-European proto-language had originated in southern Russia and had been spread throughout Europe by hordes of conquering warriors. But Gamkrelidze, Ivanov and other Soviet scholars cite words in proto-Indo-European that appear to have been borrowed from the languages of Mesopotamia and the Near East, suggesting that the speakers of proto-Indo-European lived in close geographical proximity to these cultures. The proto-Indo-European word for wine, for instance, appears to have its ancient roots in the non-Indo-European Semitic word wanju and the Egyptian wns.

     Farming in Europe. The Soviets' linguistic work has found unexpected support in new research by British archaeologist Colin Renfrew, who, unaware of the linguistic studies, independently determined that the Indo-European homeland was in Anatolia, based on a reassessment of the archaeological evidence. Renfrew suggests that it was farmers, not warriors, who were responsible for the spread of the Indo-European language into Europe. He notes that even if a farmer's offspring had moved only 10 miles from the family farm to set up farms of their own, the resulting wave of agriculture could have swept throughout Europe from Anatolia in about 1,500 years, carrying the Indo-European language with it. Because farming can support a larger number of people than hunting and gathering, the existing inhabitants of Europe were probably pushed out or adapted to farming on their own, says Renfrew.

     While the existence of proto-Indo-European has been accepted among scholars for years, linguists have now begun to trace the lineage of languages back even further. Linguists studying languages from other areas of the world have identified ancestral mother tongues such as Altic, which gave rise to east Asian languages including Japanese and Korean, and Afro-Asiatic, the ancestor of Semitic. Working backward from reconstructions of Indo-European, Altic, Afro-Asiatic and several others, Soviet scholars have found that these ancestral tongues derived from an even more ancient language. Called "Nostratic," meaning "our language," this ancestral tongue was reconstructed independently by Soviet linguists Vladislav Illich-Svitych and Aharon Dolgopolsky during the 1960s, though their work was not translated into English until recently.

     To re-create this ancient mother tongue, the Soviet scholars examined words considered by linguists to be the most stable parts of a vocabulary, such as names for body parts, personal pronouns and natural objects such as the sun and moon. Analyzing how the sounds for these words changed among Nostratic's various daughter languages, they were able to reconstruct hundreds of words. The Nostratic word for young man, for instance, is majra, which evolved into merio in Indo-European and thousands of years later became mari in French, meaning husband, and marry in English.

     The reconstructed words of Nostratic vocabulary offer a glimpse of how the people who spoke the language lived, and they suggest a date when the language thrived. The absence of words for domesticated plants suggests the Nostratic speakers were probably hunter-gatherers, says Vitaly Sheveroskin, a former student of Dolgopolsky's who is now at the University of Michigan. Even more intriguing is the word kuyna, which can mean either dog or wolf; the "k" evolved into an "h" in Germanic languages, leading to hound in English. The ambiguity of meaning in the word suggests that wolves were in the process of becoming domesticated, says Sheveroskin, who notes that the oldest bones of dogs date to 14,000 years ago, giving a time frame when Nostratic was spoken. The speakers of Nostratic were well-traveled: Not only is the lexicon peppered with words that refer to "long journey" but over the next several thousand years, Nostratic split into several major language families as its speakers migrated from the Near East, their suspected homeland, into Europe, Persia and India.

     The reconstruction of another such "macrofamily" of languages has given new clues to another mass migration, the original settling of the Americas. Joseph Greenberg, a linguist at Stanford University, recently proposed a controversial theory that all the languages spoken by Native Americans can be grouped into three families that correspond to three waves of migration from Asia into the New World thousands of years ago. The largest, oldest and most controversial group proposed by Greenberg is a macrofamily that he calls Amerind, which is made up of all the languages in South and Central America as well as many in North America. The other two language groups arc Na-Dene, which includes tongues spoken by Native Americans in the Northwest as well as Navajo and Apache, and Eskimo-Aleut, which contains languages spoken mostly in the Arctic: this group was the last to arrive in the New World.

     Scholarly furor. Greenberg's theory has created a furor among linguists, even some of those who champion the deep reconstruction of languages. At a recent conference in Boulder, Colo., linguists attacked Greenberg's admittedly unconventional methodology, in which he compares common-sounding words across many languages rather than attempting to reconstruct the sound shifts that occurred when one tongue diverged from another. Yet Greenberg's defenders cite his track record: A classification of African languages he made 20 years ago created a similar furor among linguists and is now widely accepted.

     Greenberg's theory is being given new weight by an upheaval in archaeological thinking about the peopling of the Americas. Archaeologists have long believed that the first migration to the New World occurred some 12,000 years ago--too short a time, some linguists argue, for the hundreds of Indian languages to arise from Greenberg's proposed Amerind mother tongue. But recently, several archaeological sites in the Americas have been shown to be far older than 12,000 years, suggesting that the first migration to the New World may have occurred much earlier--thus allowing more time for languages to diverge. One site, a rock shelter in Pennsylvania has been dated at 16,000 years and another site in Chile may date back as far as 33,000 years. New studies that compare changes in the genes of Native Americans suggest that the date of the first migration might stretch back as far as 60,000 years.

     Research on American languages is also throwing light on a longstanding linguistic mystery in Europe--as well as testifying to the remarkable wanderlust of ancient humans. Linguists have long wondered about the origins of Basque, a language spoken in the north of Spain that is one of the few non-Indo-European languages on the Continent. Soviet linguists have uncovered evidence that Basque is related to Na-Dene, and that both languages are part of yet another language macrofamily that includes tongues ranging from Chinese to the ancient Mediterranean tongue Etruscan. Called Dene-Caucasian, this ancient language was reconstructed in large part by Soviet linguist Sergei Starostin, another student of Dolgopolsky's. This wide-ranging tongue, spoken on both sides of the Bering Strait and at both ends of the Eurasian land mass, reflects the vast movements of ancient peoples who took their language with them and in some cases, such as Basque, kept it alive despite their being surrounded by other tongues.

     The survival of an exotic linguistic island like Basque suggests that language, like genes, can sometimes serve as a marker for a distinct group of people. Historical linguistic studies are generating widespread interest among scientists involved in one of the most exciting new developments in science: Tracing the evolutionary history of human genes. "It's quite clear why you have a correlation between genes and language," says Stanford University geneticist Luigi Cavalli-Sforza, a pioneer in the new genetic techniques. "When the human expansion around the earth took place some 50,000 years ago, it caused a number of separations between groups that didn't communicate again, genetically or linguistically. As the genes become different, the languages become different too.

     Cavalli-Sforza and his colleagues recently examined genetic markers from 42 different indigenous peoples from around the world and used the divergence among the genes to construct a family tree for the entire human race. The tree shows the human diaspora over tens of thousands of years, as a single population split into several large groups and then into the smaller tribes that exist today.

     African split. More important, Cavalli-Sforza found that the groupings of the human family based on genetic evidence closely mirrored the language groupings laid out independently by historical linguists. The oldest split occurred between Africans and other world populations, reflecting the migration of Homo sapiens out of Africa. This split is reflected in languages as well: Africa's Khoisan languages, such as that of the !Kung San, which uses a clicking noise denoted by an exclamation point, is only distantly related to other languages in the world. A recent paleoanthropological discovery in the Near East of the oldest fossils of Homo sapiens gives a rough date for when this first split might have occurred: The fossils date back 92,000 years.

     Another study that similarly traced the human genetic lineage suggests that all languages may have their roots in a small population that lived some 200,000 years ago. In their famed "Eve" hypothesis, Allan Wilson, Mark Stoneking and Rebecca Cann of the University of California at Berkeley traced genetic material from women around the world and concluded that all humans alive today are descendants of a tiny population of Homo sapiens that lived in Africa.

     If the human race did arise from this small group of people, then it is likely they all spoke the same language, contends Sheveroskin. What's more, he says, the same techniques that gave rise to the reconstructions of ancient macrofamilies can also be used to dredge up bits of the original mother tongue of the human race. For example, the Nostratic word for leaf, lapa, is similar to tlapa in Dene-Caucasian and dap in Amerind. And the Amerind word for woman, kuni, closely resembles the Nostratic word for woman, kuni. It is from this word that the English word queen is derived.

     Fleas and in-laws. Sheveroskin and other linguists have reconstructed dozens of words in this original mother tongue, which has been dubbed simply "proto-World." The word for I, for instance, is ngai; nas means nose. The linguists have also reconstructed proto-World words referring to body parts, fleas, in-laws and a category of words that referred to pairs of objects--reflecting, perhaps, a culture that before the invention of mathematics counted the world in ones, twos and many. Niwha and hwina refer simultaneously to life, breath and blood, but, strangely, notes Sheveroskin, there appear to be no words in proto-World that refer to human emotions.

     In the end, discovering the roots of language is inexorably tied to the still unresolved task of defining what language is, and this is where the ultimate impact of the new linguistic research may lie. The deep connections between languages demonstrate that far from a mere communication device, language is the palette from which people color their lives and culture. Intimately connected to the human experience, language oils the gears of social interactions and solidifies the ephemera of the mind into literature, history and collective knowledge. It is the calling card of the human race, announcing the presence not only of those alive today but, with its deep roots into the past, the ancient ancestors who came before us.

     Spreading the word

     New research suggests that the ancestral tongue from which most modern European languages are descended was spread throughout the Continent by farmers. According to British archaeologist Colin Renfrew, this language, called Indo-European, arose in Anatolia--part of modern-day Turkey--some 8,000 years ago, and over the following millennia gave rise to the precursors of English, French and dozens of other modern tongues.

     Soviet linguists have found that Indo-European is itself a descendant of a more ancient tongue. Dubbed Nostratic, this proto-language arose some 14,000 years ago, spawning several language families that spread into Africa, India and Europe.

     Basic data: ARCHAEOLOGY AND LANGUAGE by Colin Renfrew, Cambridge University Press, 1988.

     All in the family

     Comparing genes from people around the world, biologists have created a family tree for the human species. The human race can be divided into seven major groups, each the result of ancient migrations. Strikingly, these genetic groups roughly mirror the major language families reconstructed by historical linguists.

     Basic data: "Reconstruction of Human Evolution: Bringing Together Genetic, Archaeological, and Linguistic Data," L.L. Cavalli-Sforza; "Evolution of Language: A Global Perspective," Merritt Ruhlen.

     First words: Original utterance: Chimps, dolphins and early humans

     As historical linguists trace the world's languages back to their earliest sources of diction and syntax, other researchers are taking on the deep mystery of how and when humans started to talk in the first place. In the forthcoming book UNIQUELY HUMAN: THE EVOLUTION OF SPEECH, THOUGHT AND SELFLESS BEHAVIOR (Harvard University Press, $27.95), linguist Philip Lieberman of Brown University argues that ancient hominids lacked the modern human's elongated vocal tract, which he contends is essential for the wide range of sounds characteristic of language. Fossil evidence suggests that fully developed language may have arisen only with the appearance of anatomically modern humans some 200,000 years ago, and that their sophisticated powers of communication may have been a key to the species' migration around the globe.

     But some crude form of language may have existed long before the evolution of language as we know it. Studies of fossil skulls indicate that a part of the brain crucial to language production in modern humans existed in hominids millions of years ago, suggesting that our ancient ancestors may have had at least rudimentary linguistic abilities. In his new book, LANGUAGE AND SPECIES (University of Chicago Press, $24.95), linguist Derek Bickerton of the University of Hawaii contends that primitive language may have arisen not so much from the need to communicate as from the mind's effort to grasp the world around it. Vestiges of this language remain with us, he says, in the form of the simple, one and two-word utterances of toddlers and adults who speak to each other in pidgin.

     Apes and other intelligent animals may also share this rudimentary form of language. Studies of chimpanzees have shown that animals can communicate with signs or symbols only in a very limited fashion. But just last month, researchers from the University of California at Los Angeles and Emory University's Yerkes Regional Primate Research Center reported that a 5-1/2- year-old pygmy chimp named Kanzi spontaneously learned to use grammatical rules similar to those of a 2-year-old child. UCLA psychologist Patricia Marks Greenfield and Yerkes biologist Sue Savage-Rumbaugh found that the order in which Kanzi used symbols was an integral part of their meaning. Recent studies with dolphins and sea lions demonstrate that these creatures can learn to understand word order as well.

     Whether the utterances of animals represent true humanlike language abilities remains a matter of intense scientific debate, however. Critics of animal-language studies argue that the animals are only using rote behavior to get food or are responding to unconscious cues from their trainers. "All the evidence suggests that the animals are merely using sophisticated ways to request things," says Columbia University psychologist Herbert Terrace, whose ape research convinced him that animals could not learn to communicate in a humanlike language. Evidence that language may be a uniquely human trait, say some researchers, ultimately lies not in animals' language abilities but in how they use that ability. In animals, language appears to be just one more way to fulfill physical desires; in humans, language reflects not only our earthly needs but our heavenly desires, thoughts and emotions.

     Early writing: From tax audits to poetry

     Since the time of the first cave drawings in 35,000 B.C., humans have revealed their nature by scratching, painting and otherwise representing the world around them. But not until 3200 B.C., with the rise of the first city-states in Mesopotamia, did rulers show a need for intricate records of taxation, tributes and inventories--and for an elaborate system of symbols for communicating. Called cuneiform, the first writing was made with sharpened reeds that cut wedge-shaped symbols into soft clay. With a vocabulary of more than 2,000 signs, it was used not only for bookkeeping but also for medical prescriptions, epic tales and religious writings.

     Although cuneiform is the earliest known writing, several other systems sprouted up independently throughout the world, looking very different from one another in many ways but sharing common features that suggest why and how man invented writing. The birth of most writing systems coincided with the rise of a state in which a dominating elite used written records to establish and conserve power. In Egypt, for example, early writings depicted a leader receiving tributes from the masses, and in China some of the earliest documents are construction records for huge public monuments. In Central America, Mayan writings commemorated seasonal ceremonies and royal conquests.

     While early writing systems were developed by and for the elite, however, most evolved over time to become instruments of the common man. Writing systems tended to start out as "pictographic"; that is, drawings of the sun, oxen or rain signified those objects. Although this method seems simplistic, in fact a new sign had to be added for each object, rendering vocabularies burdensome and accessible only to scribes who made writing their life's work. Gradually, symbols expanded in meaning to become "ideograms," representing ideas associated with objects. The sign for sun came also to mean heat or light, for example. Eventually, symbols were used phonetically; the signs for bee and leaf, for instance, were combined to mean belief. This early type of linguistic pun simplified both writing and reading, and at the same time made communication of more abstract ideas possible.

     It was not until 1500 B.C. and the invention of the first true alphabet, Old Canaanite, that written communication became readily available to the masses. For the first time, a writing system was based purely on phonetics, affording a simple means to write anything that could be spoken. Moreover, one could master the 22 to 32 symbols of the early alphabets in a matter of days or weeks, in contrast to the years it took to master earlier systems. But because the earliest alphabets lacked vowels, they still posed a problem for the average reader, who found words made only of consonants difficult to decipher. For instance, the word dbt could mean debt, or debit, or doubt, or debut. The Greeks introduced vowels in 740 B.C., clarifying written text and allowing anyone to record debates, speeches, personal thoughts and ideas. The alphabet transformed writing into a tool that gave individuals of any social class the opportunity to express themselves, and most scholars conclude its invention was inseparable from the first inklings of democracy and the birth of poetry.--by Joannie M. Schrof

     Lineage of a word: Haku (Proto-World); Hakw (Amerind); Kwa (Dene-Caucasian); Haku (Nostratic); Hakw (Indo-European); Aqua (Latin);  Wazzar (Old German);  Water (English)

     Lineage of a word: Hita (Proto-World); Hit- (Amerind); -ita (Nostratic);  Hed (Indo-European); Edmenai (Greek); Edere (Latin); Ezzan (Old German); Eat (English)

     Lineage of a word: Kuni (Proto-World); Kuni (Nostratic); Gwen (Indo-European); Queen (English)

     Lineage of a word: Kujan (Proto-World); Kuyna (Nostratic); Kuon (Indo-European); Hound (English)