There's a welsh proverb I've known for as long as I can remember: "Cenedl heb iaith, cenedl heb galon." It means, "A nation without a language [is] a nation without a heart," and it's become more poignant over the years as more and more families who live around me in North Wales speak in English instead of Welsh across the dinner table.

Welsh, the direct descendant of the Celtic language that was spoken throughout most of Britain when the Anglo-Saxons invaded, has long been under threat from English. England's economic and technological dominance has made English the language of choice, causing a decline in the number of Welsh speakers. And although the decline has steadied in the past 15 years, less than 20 percent of the population of Wales today can speak Welsh in addition to English.

The Welsh language is clearly in trouble. Someday, it may even join the rapidly growing list of extinct languages, which includes Gothic and Hittite, Manx and Cornish, Powhatan and Piscataway. If present trends continue, four of the world's languages will die between the publication of this issue of CIVILIZATION and the next. Eighteen more will be gone by the end of 1997. A century from now, one-half of the world's 6,000 or more languages may be extinct.

The decline is evident the world over. Consider the case of Sene: In 1978 there were fewer than 10 elderly speakers remaining in the Morobe province of Papua New Guinea. Or Ngarla: In 1981 there were just two speakers of the Aboriginal language still alive in northwest Western Australia. And in 1982 there were 10 surviving speakers of Achumawi out of a tribal population of 800 in northeastern California. Does it matter? When the last representatives of these peoples die, they take with them their oral history and culture, though their passing is rarely noticed. Sometimes, years later, we find hints of a culture's existence, in the form of inscriptions or fragments of text, but many of these--the Linear A inscriptions from ancient Crete, for example--remain undeciphered to this day.

There is some controversy over exactly how to count the number of languages in the world. A great deal depends on whether the speech patterns of different communities are viewed as dialects of a single language or as separate languages. The eight main varieties of spoken Chinese, for example, are as mutually unintelligible as, say, French and Spanish--which suggests that they are different languages. On the other hand, they share a writing system, and so perhaps are best described as dialects of the same language. If you opt for the first solution, you will add eight to your tally of the world's languages. If you opt for the second, you will add just one.

Taking a conservative estimate of 6,000 languages worldwide, one fact becomes immediately clear: Languages reveal enormous differences in populations. At one extreme, there is English, spoken by more people globally than any other language in history, probably by a third of the world's population as a first, second or foreign language. At the other extreme is Ngarla (and most of the other languages of the native peoples of Australia, Canada and the United States), whose total population of speakers may amount to just one or two. And then there are closely related groups of languages like the Maric family in Queensland, Australia, which consists of 12 languages. When it was surveyed in 1981, only one of these, Bidyara, had as many as 20 speakers. Most had fewer than five. Five of them had only one speaker each.

The loss of languages may have accelerated recently, but it is hardly a new problem. In the 19th century, there were more than 1,000 Indian languages in Brazil, many spoken in small, isolated villages in the rain forest; today there are a mere 200, most of which have never been written down or recorded. In North America, the 300 or more indigenous languages spoken in the past have been halved.

People sometimes talk of "the beauty of Italian" or of "German's authority," as if such characteristics might make a language more or less influential. But there is no internal mechanism in a language that settles its fate. Languages are not, in themselves, more or less powerful. People don't adopt them because they are more precise. They gain ascendancy when their speakers gain power, and they die out when people die out or disperse. It's as simple as that.

A dramatic illustration of how a language disappears took place in Venezuela in the 1960s. As part of the drive to tap the vast resources of the Amazonian rain forests, a group of Western explorers passed through a small village on the banks of the Coluene River. Unfortunately, they brought with them the influenza virus, and the villagers, who lacked any immunity, were immediately susceptible to the disease. Fewer than 10 people survived. A human tragedy, it was a linguistic tragedy too, for this village contained the only speakers of the Trumai language. And with so few people left to pass it on, the language was doomed.

Other languages -- such as Welsh and Scottish Gaelic--have been threatened when indigenous populations have moved or been split up. Brighter economic prospects tempt young members of the community away from their villages. And even if they choose to stay, it doesn't take much exposure to a dominant culture to motivate ambitious young people to replace their mother tongue with a language that gives them better access to education, jobs and new technology.

A language's fortunes are tied to its culture's. Just as one language holds sway over others when its speakers gain power--politically, economically or technologically--it diminishes, and may even die, when they lose that prominence. Latin, now used almost exclusively in its written form, had its day as a world language because of the power of Rome. English, once promoted by the British Empire, is thriving today chiefly because of the prominence of the U.S.A., but it was once an endangered language, threatened by the Norman invaders of Britain in the 11th century, who brought with them a multitude of French words. In South America, Spanish and Portuguese, the languages of colonialists, have replaced many of the indigenous Indian tongues.

The death of languages is most noticeable in parts of the world where large numbers of languages are concentrated in a few small geographical regions. Travel to the tropical forests of the Morobe province in Papua New Guinea and you'll find five isolated villages in a mountain valley where fewer than 1,000 people speak the Kapin language. They support themselves by agriculture and have little contact with outsiders. Other tiny communities, speaking completely different languages, live in neighboring valleys. Linguists estimate that in the country as a whole there is approximately one language for every 200 people. Indeed, three countries, which together amount to less than 2 percent of the earth's land area, support 1,700--or a quarter--of the world's living languages: Papua New Guinea has 862; Indonesia, 701; and Malaysia, 140. These countries' isolation and physical geography account in large part for the existence of such concentrations, and it is hardly surprising to find that, as remote areas of the globe have opened up for trade or tourism, there has been a dramatic increase in the rate of language death. Valuable reserves of gold, silver and timber in Papua New Guinea, for example, are bringing speculators to the islands--and with them their languages.

There has been little research into exactly what happens when a language begins to die. The process depends on how long there has been contact between the users of the minority language and their more powerful neighbors. If the contact has been minimal, as in the case of Trumai in the Amazon, the minority language might remain almost unchanged until the last of its speakers dies. But if two languages have been in contact for generations, the dominant language will slowly erode the pronunciation, vocabulary and grammar of the minority language. Take the Celtic languages of northwest Europe. Following the death of the last mother-tongue speakers of Cornish (spoken in Cornwall until the 19th century) and Manx (spoken in the Isle of Man until the 1940s), the only remaining Celtic languages are Breton (in northwest France), Irish and Scottish Gaelic, and Welsh. All have been in steady decline during the 20th century. Equally, all have been the focus of strenuous efforts to revive their fortunes (or, in the case of Cornish and Manx, to resurrect a new first-language base). But the effects of four centuries of domination by English are evident everywhere.

Walk into the stores in the strongly Welsh-speaking areas of North Wales, as I regularly do, and you will hear the Welsh language widely used--and apparently in good health. But there is also a great deal of recognizable English vocabulary scattered throughout the speech. Of course, all languages have what linguists refer to as "loan words"--words taken from other languages to supplement the vocabulary. English itself has tens of thousands of words borrowed from French, Spanish, Latin and other languages. But there is an important difference between traditional vocabulary borrowing and what takes place in an endangered language. When arsenic, lettuce and attorney came into English in the Middle Ages, it was because these items did not already exist in the English-speaking community: The nouns were introduced to describe new objects, and so to supplement the existing vocabulary. But in the case of an endangered language, the loan words tend to replace words that already exist. And as the decline continues, even quite basic words in the language are replaced.

I meet this phenomenon every day on the Welsh island of Anglesey, where I live. It's become quite unusual to hear locals referring to large sums of money in anything other than English. In a Holyhead butcher's shop recently, I overheard someone say "Mae'n twelve fifty" (It's twelve fifty), where the first part of the sentence is colloquial Welsh and the second part is colloquial English. As I waited for a train at the station the same day, I heard a porter calling out to disgruntled passengers "Mae'n late" (It's late). And I later overheard a group of people using the English word injection as they stood in a street describing in Welsh someone's visit to a doctor's clinic. In all these cases, perfectly good Welsh words already exist, but the speakers did not use them. Why they chose not to is not at all clear. Maybe they did not know the Welsh words, or maybe it is a sign of status or education to use the English equivalents. But when something as basic as its number system is affected, a language is clearly in danger.

Mixed languages are an inevitable result of language contact, and they exist all over the world, often given a dismissive label by more educated speakers: Wenglish, Franglais, Spanglish. Such mixed varieties often become complex systems of communication in their own right--and may even result in brand-new languages, or pidgins such as Tok Pisin, which is now spoken by more than 1 million people in Papua New Guinea. But when one of the languages in question has no independent existence elsewhere in the world, as in the case of Welsh, mixed languages are a symptom of linguistic decline.

In the West, when a population fears that its language is threatened, speakers often react defensively, establishing a committee or board to oversee and coordinate political policy and to plan dictionaries, grammars and local broadcasting. The best-known example is France, home of the Academie Francaise, where there is now a law banning the use of English words--such as le week-end and le computeur--in official publications if a native French term already exists (in these cases, la fin de semaine and l'ordinateur). Often two levels of language ability emerge as a consequence. There is an educated standard, used as a norm in education and the media. And there is a colloquial standard, used by the majority of the population (including many educated users, who thereby become bilingual -- more technically, bidialectal--in their own language). It is the usage of the elite minority that is called by the majority the "proper" or "correct" language, even though it often represents a far more artificial style of speech than the language of the streets.

The plight of the indigenous languages of America was made vivid by James Fenimore Cooper as long ago as 1826, when the Indian chief Tamenund lamented that "before the night has come, have I lived to see the last warrior of the wise race of the Mohicans." There are 200 North American Indian languages, but only about 50 have more than 1,000 speakers, and only a handful have more than 50,000. Just over a year ago,Red Thunder Cloud, the last known fluent speaker of the Siouan language Catawba, died. The only surviving fluent speaker of Quileute is 80-year-old Lillian Pullen, of La Push, Washington. But at least the decline of American Indian languages has begun to attract widespread attention from politicians and the media--sources of support that are unlikely to help such equally threatened but less well known cases as Usku in Irian Jaya or Pipil in El Salvador.

In Europe, public attention is regularly focused on language rights by the European Bureau of Lesser Used Languages, headquartered in Brussels. A recent book, A Week in Europe, edited by the Welsh magazine editor Dylan Iorwerth, offers a glimpse of Western European life by journalists writing in minority languages. Some of these are minority uses of major languages, such as German in Denmark, Swedish in Finland, and Croatian in Italy; but in most cases the entire language-using community is found in a single region, such as Scottish Gaelic, Galician, Alsatian, Welsh, Catalan, Asturian, Breton, Friulian, Basque, Sorbian, Occitan, Provencal, Frisian and Irish. Political concern over the status of minority languages is regularly voiced by the European Parliament, and occasionally words are backed up with financial commitment-- to local newspapers and broadcasting, literary festivals and teaching programs.

When an endangered language (such as Gaelic) is spoken in a culture whose historical significance is widely appreciated -- perhaps because it is associated with prowess in arts and crafts, or because it is known for its literary achievements--it may provoke widespread concern. And sometimes endangered languages that have suffered as a result of colonial expansion win support from speakers of the dominant language, who wish to distance themselves from the aggression of their ancestors. But in most cases, anxiety, like charity, begins at home. In the 1970s, Gwyn for Evans held a hunger strike as part of his (successful) campaign for a Welsh-language TV channel. And in 1952 in Madras, India, Potti Sriramulu died following a hunger strike in support of the Telugu language. Language, as that Welsh proverb reminds us, is truly at the heart of a culture. It is a matter of identity, of nationhood.

With enough personal effort, time and money, and a sympathetic political climate, it is possible to reverse the fortunes of an endangered language. Catalan, spoken in northeast Spain, was allocated the status of an official regional language, and it now has more native speakers than it did 30 years ago. And the Hocak, or Winnebago, tribe in Wisconsin is hoping to develop a full Hocak-speaking school system. In an effort funded entirely by profits from the tribe's casinos, schoolchildren use interactive multimedia computer programs to gain familiarity with a language that was traditionally passed down orally from parent to child. Such advances generally depend upon collaboration between minority groups, such as those who united to form the European Bureau for Lesser Used Languages. Together they have a realistic chance of influencing international policies, without overlooking the vast differences between the political and cultural situations of minority languages: Welsh, Gaelic, Maori, Quechua and Navajo demand very different solutions.

Welsh, strongly supported by Welsh-language broadcasting and Welsh-medium schools, is alone among the Celtic languages in stopping its decline. The census figures for the last 20 years show a leveling out, and even some increase in usage among certain age groups, especially young children. A similar vigorous concern seems to be stimulating Navajo and several other American languages, as well as some minority languages in continental Europe. But it is quite clear that most of the endangered languages of the world are beyond practical help, in the face of economic colonialism, the growth of urbanization and the development of global communication systems. And, given the difficulty there has been in achieving language rights for such well-known communities as the Navajo or the Welsh, the likelihood of attracting world interest in the hundreds of languages of Papua New Guinea, each of which has only a few speakers left, is remote. Clearly, with some 3,000 languages at risk, the cost of supporting them on a worldwide scale would be immense. Can, or should, anything be done?

On an intellectual level, the implications are clear enough: To lose a language is to lose a unique insight into the human condition. Each language presents a view of the world that is shared by no other. Each has its own figures of speech, its own narrative style, its own proverbs, its own oral or written literatures. Preserving a language may also be instructive; we can learn from the way in which different languages structure reality, as has been demonstrated countless times in the study of comparative literature. And there is no reason to believe that the differing accounts of the human condition presented by the peoples of, say, Irian Jaya will be any less insightful than those presented by writers in French, English, Russian and Sanskrit. Moreover, the loss of a language means a loss of inherited knowledge that extends over hundreds or thousands of years. As human beings have spread around the globe, adapting to different environments, the distilled experiences of generations have been retained chiefly through the medium of language. At least when a dying language has been written down, as in the case of Latin or Classical Greek, we can usually still read its messages. But when a language without a writing system disappears, its speakers' experience is lost forever. The Bithynian, Cappadocian and Cataonian cultures are known today only from passing references in Greek literature. Language loss is knowledge loss, and it is irretrievable.

Such intellectual arguments may persuade the dispassionate observer, but most arguments in favor of language preservation are quite the opposite: They are particular, political and extremely passionate. Language is more than a shared code of symbols for communication. People do not fight and die, as they have done in India, to preserve a set of symbols. They do so because they feel that their identity is at stake--that language preservation is a question of human rights, community status and nationhood. This profoundly emotional reaction is often expressed in metaphors. Language nationalists see their language as a treasure house, as a repository of memories, as a gift to their children, as a birthright. And it is this conviction that has generated manifestoes and marches in Melbourne in support of Aboriginal languages; referendums, rioting and the defacing of public signs in Montreal on behalf of French; civil disobedience in India and Pakistan, in Belgium and in Spain.

Such demonstrations stand in stark contrast to places where cultural and linguistic pluralism works successfully, as in Switzerland and Sweden, where the dominant culture respects the identities and rights of its linguistic minorities, and provides educational opportunities for speakers. Successful multilingual communities such as Sweden's serve as examples for the United Nations, UNESCO, the Council of Europe, and the European Parliament as they act to preserve minority language use.

Conversely, several countries have actively repressed minority languages, such as Basque by the Spanish fascists, or Sorbian (a Slavic language spoken in southern Germany) by the Nazis. And laws forbidding the use of minority languages have been commonplace; children have been punished for using a minority language in school; street signs in a minority language have been outlawed; the publication of books in the language has been banned; people's names have been forcibly changed to their equivalents in the language of the dominant power. Whole communities, such as several in the Basque-speaking parts of northern Spain, have had their linguistic identity deliberately eliminated.

Political arguments for and against preservation have been expressed with such vehemence that they tend to dominate any discussion of minority languages. Does the loss of linguistic diversity present civilization with a problem analogous to the loss of species in biology? Not entirely. A world containing only one species is impossible. But a world containing only one language is by no means impossible, and may not be so very far away. Indeed, some argue strongly in favor of it. The possibility of creating a unilingual world has motivated artificial-language movements (such as Esperanto) since the 16th century, and there are many who currently see the remarkable progress of English as a promising step toward global communication. They argue that mutual intelligibility is desirable and should be encouraged: Misunderstandings will decrease; individuals and countries will negotiate more easily; and the world will be more peaceful.

This kind of idealism wins little sympathy from language nationalists, who point out that the use of a single language by a community is no guarantee of civil peace--as is currently evident in the states of the former Yugoslavia or in Northern Ireland. But language nationalists are faced with major practical concerns: How can one possibly evaluate the competing claims of thousands of endangered languages? Is it sensible to try to preserve a language (or culture) when its recent history suggests that it is heading for extinction? In the next few years, international organizations may have to decide, on chiefly economic grounds, which languages should be kept alive and which allowed to die.

The publication in the early 1990s of major surveys of the world's languages has brought some of these issues before the public. UNESCO's Endangered Languages Project, the Foundation for Endangered Languages (established in the U.K. in 1995) and the Linguistic Society of America's Committee on Endangered Languages and their Preservation are fostering research into the status of minority languages. Information is gradually becoming available on the Internet--such as through the World Wide Web site of the Summer Institute of Linguistics. And a clearinghouse for the world's endangered languages was established in 1995, by request of UNESCO, at the University of Tokyo.

But after the fact-finding, the really hard work consists of tape-recording and transcribing the endangered languages before they die. The fieldwork procedures are well established among a small number of dedicated linguists, who assess the urgency of the need, document what is already known about the languages, extend that knowledge as much as possible, and thus help preserve languages, if only in archive form.

The concept of a language as a "national treasure" still takes many people by surprise--and even English has no international conservation archive. It is hard to imagine the long hours and energy needed to document something as complex as a language--and it's often a race against time. Thirty years ago, when anthropologist J.V. Powell began working with the Quileute Indians in Washington state, 70 members of the tribe were fluent speakers. Around that time the tribal elders decided to try to revitalize the language, writing dictionaries and grammars, and imagining a day when their children would sit around chatting in Quileute. "But," says Powell, "their prayers haven't been answered." Now they've scaled back to a more modest goal: basic familiarity rather than fluency. Powell recognizes that they will not save Quileute, but it will be preserved in recordings for future scholars--and will serve as a symbol of the tribe's group identity. That may seem like a small success, but it's a far better fate than the one facing most endangered languages.

PHOTO (COLOR): In Wales, less than 20 percent of the population speaks Welsh

PHOTO (COLOR): In Papua New Guinea (left), there is one language for every 200 people; many of the Aboriginal tongues of Western Australia (below) are at risk

PHOTO (COLOR): Lillian Pullen (above), the last fluent speaker of Quileute; a Hocak language instructor (right) helps five-year-old Alexandria Lange gain familiarity with her tribe's native tongue

~~~~~~~~

By DAVID CRYSTAL