On April 21st 1995, Bristol University Philosophy Department will host a one-day  seminar to discuss the scientific and cultural arguments for conserving the world's stock of languages. The event is organised by the Centre for Theories of Language and Learning. Here Andrew Woodfield, the Centre's director, sets out some of the issues. According to reliable estimates, half of the world's six thousand languages will become extinct in the next century. Two thousand of the remaining three thousand languages will be threatened during the century after that. In the UK these startling facts have received media attention through the publication in 1994 of the Atlas of the World's Languages. The rapid decline is largely due to economic and political pressures on minority communities which remove the new generation's motivation for sticking to their traditional language.

The languages are not inherently deficient in any way. The situation is a global tragedy in the classical Greek mould, for the demise of native languages is an unintended byproduct of external forces that the participants cannot control. No one actively wants these thousands of languages to die. Yet they seem fated to die. In some ways the extinction of languages is like the extinction of biological species. Of course, the parallel is not complete. For instance, a world with only one species is an impossibility, whereas it is a possibility, at least in principle, that the world will eventually become unilingual. In attempting to conceptualize the dynamics of spread and decline, people have explored a range of biological analogies comparing languages to genes, individuals, characters, populations, and ecosystems.

The relation betweeen a language and its speakers can also be compared to the relation between a homeland and its inhabitants: people exploit their land and their language in the course of making a living, yet at the same time they are stewards of the geographical and linguistic environments. The problem of language-extinction raises fundamental questions about value. If you believe that international agencies and governments should make an effort to conserve lesser used languages, you presumably think such languages are valuable. What exactly is the value of a language? Does the existence of linguistic diversity have any value? Let us distinguish two broad perspectives from which the value of a given language X can be viewed.

On the one hand, linguists, scholars and scientists may treat X as an object of study. On the other hand, there are the people who use X as a means of communication. This population can be subdivided into people for whom X is their only language, people who speak X as their first language and who speak another language as well, and people who speak X only as a non-primary language. Consider the emperilled language Waimiri-Atroari, a language of the Carib family spoken by a precarious community of 500 Indians in Brazilian Amazonia, most of whom are (or were until very recently) monolingual.
 
 
 
 

These people need to share information in order to stay alive, so X has extremely high instrumental value for about 500 people. It has zero instrumental value for everyone else, apart from the traders and officials who wish to communicate with the Waimiri-Atroari. The fact that a population needs somemedium of communication does not provide a knock-down reason in favour of the particular language X, especially in circumstances where X is competing against another language. Compared with a world language, the volume of messages actually carried in Waimiri-Atroari is minuscule because the number of speakers is so small. If the same people spoke Portuguese, they would have potential communicational access to 160 million Brazilians. Many of the 500 individuals, and their future children, will in fact learn Portuguese in the next few years. As the proportion of bilinguals in the tribe goes up, so the instrumental advantage of the native language gets eroded. By the year 2000, many of the message-carrying functions of Waimiri-Atroari within the Waimiri-Atroari community may have been usurped by Portuguese. Adult speakers will stop using X as their default language.

When the children are not even motivated to learn X, the language will have become moribund. Because the comparative advantage of one medium over another is so contingent upon external circumstances, the abiding reasons why a language has value for its users must lie beyond its message-carrying utility. A language is not just a medium, a symbol-system or a code. It is also the repository of a cultural tradition, a way of living and of expressing which helps to confer a sense of identity upon its native-speakers. Some of this intrinsic value can rub off on to those who learn X as a later language - this is a premiss upon which education in the humanities is based. People define themselves partly in relation to their culture. It is not surprising that the Waimiri-Atroari tribe celebrates its own language and teaches the young people to prize it as a thing of value.

A language is an intricate collectively produced artefact inherited from previous generations. Whichever language one absorbs as a child, the truth is that the language is a gift. Let us now turn to the impersonal and scientific reasons for preserving an endangered language based on its value as an object of study. Here we must distinguish between existing and living. Let us say that a language's existence is secure, provided there exists an authoritative dictionary of its words, plus a description of its phonetics and phonology, plus a reasonably complete account of its syntax, plus rules for assigning meanings to its sentences. This information defines X and distinguishes it from all other languages. Additional facts about X that one would wish to find in an archive include X's position in a scheme of classification of the world's languages, its origin, its historical development, the geographical location of its speakers, and so on. Some researches are greatly helped if there exist recordings of speech in X on tape and video. In the case of a written language, a sample of authentic texts will provide extra information.

Compiling such an archive is usually a feasible task. For each threatened language X that still has not been documented, a team of linguists and anthropologists might need to spend ten years collecting data and processing. After that work is done, if X stops being spoken, there will be a theoretical possibility of reviving it or creating an approximation to it.

Cornish died out in the 19th century and then got revived by enthusiasts in the 20th. According to the  Mini-Guide To The Lesser Used Languages of the EC , published by the European Bureau for Lesser Used Languages, about two hundred people can now speak fluent Cornish and several thousand more can understand it (though none speaks it as a native language). The paradigm of a revived vernacular is Hebrew. The fact that young Israelis now learn Hebrew at their mothers' knee is due in great measure to the efforts of one Russian immigrant, Eliezer Ben Yehuda, who organised a system of Hebrew schools in Palestine at the turn of the century. Of course, he could not have
achieved this revival had there not existed a set of very special political circumstances. And it must be remembered that Hebrew had been continuously used in liturgical and legal contexts, and had been routinely taught to Jewish males during its 1700 year-long coma, so it was never entirely dead.
 
 

Documented profiles of languages, whether alive or dead, are valuable to cultural historians and anthropologists, to theoretical linguists, to psychologists, philosophers, and many other kinds of scholars. Each language contains clues about the history of a people. The stock of general words (kind names, adjectives, etc.) reveals how the speakers classified things in their environment. Hypotheses about migration patterns can be supported by seeing whether words have been assimilated from the languages of neighbouring populations. The syntax of a language is a guide to its evolutionary descent too.

Linguists glean a lot from comparing the syntactical rules of different languages. For example, Chomsky's well- known hypothesis that all human languages conform to an innately specified general pattern which he called 'Universal Grammar' is supposed to be an empirical hypothesis. Evidence from a wide variety of languages is essential in order to test it properly. For all we know, some language exists which disconfirms Chomsky's hypothesis. But if X died out without being documented, we would never know if X was a counterexample or not.

Cognitive psychologists devise models of speech- production and understanding in which specific assumptions are made about representations, levels of representation, and processing. Actual human languages provide constraints on such model-building. Lesser-known languages might turn out to impose constraints that had never been noticed before.

Since a language is a product of human brains (though not a deliberate artefact) the phonology and the syntax of X can be relevant to hypotheses about neural structures and processes. Each language sets neuroscience the task of identifying the hardware capable of generating the sentences of that language. The acquired structures responsible for competence in X differ in some respects from those that underlie competence in another language. What must a brain that can handle Waimiri-Atroari be like? Also, what must all infant brains have in common, such that any child can acquire any language?

Philosophers sometimes construct artificial languages that are simpler and more precise than natural languages. However, real actual languages provide food for reflection equal to anything that the imagination can devise. The Australian language Guugu Yimithir (and some other Australian languages) employs an 'absolute' system for specifying spatial position rather than a 'speaker-relative' system. One has to say: 'the tree to the east', not 'the tree to my left'; 'the tree to the north of the house', and not 'the tree in front of the house (relative to me)'; and so on. What consequences does this impersonal but memory- taxing method of location have for the speakers' concepts of Space and Self? Perhaps the sceptical relativistic doctrine, so influential in the history of Western and Eastern thought, that 'facts' are partly subjective, is not an option for a person brought up in Guugu Yimithir.

Students of the modern American philosopher W.V.O.Quine ought to be interested in whether any actual language contains nouns that are referentially indeterminate. Quine made up a language in which the word 'Gavagai' refers either to a whole object, or to a time-slice of the object, or to an undetached part of the object, but where nothing in the native speakers' behaviour indicates which of these alternatives they mean. If no term with this semantic property actually exists in any real language, Quine's thesis of indeterminacy starts to look speculative and non- naturalistic.

Because languages contain so much information, it is easy to see reasons why science and the humanities will benefit if we compile archives of all the languages that are now alive. Provided the documents about X are safely kept, the language X continues to exist, at least as an abstract entity. But does it matter to the academic world whether or not a language stays alive, once it has been fully documented?

The fact that a language continues to be alive does yield additional scientific information about processes of change. For a living language is a dynamic entity. Pronunciation and intonation patterns change, new words are coined, even grammatical norms slowly change. Every language has pragmatic rules for performing different kinds of speech acts, manners, gestural accompaniments, styles of delivery, and so on, which are often highly mutable. Video and sound recordings capture only a small sample of utterances, and these immediately become fossilized.

Archives, however rich, provide only a partial description of that part of a language which has been historically actualized at the time they happen to be compiled. Whereas if X stays alive, more can be found out about it by using the methods of experimental science. The testing of hypotheses involves intervening in speech- episodes, putting questions to native speakers, generating data which would not otherwise have existed. So there are scientific advantages in having access to the real thing, and not merely to a record of the thing.

Are these advantages ever important enough to provide an overriding reason to strive to keep X alive, in cases where X appears to be heading for extinction? Massive affirmative action - the establishment of government ministries, radio and TV stations, newspapers and books, cultural events, employment, bribes - might encourage a small tribe in America, Australia or Africa to carry on using its native language. This raises difficult ethical questions not only about the interests of the native people, but also about the allocation of the resources available for carrying out such actions. It might be argued
that some endangered languages (Waimiri-Atroari, for example?) are more valuable than others (like Guugu Yimithir?).

In the next few years, international organizations like UNESCO may be obliged to decide which languages shall be kept alive and which ones shall be left to die. At present the fate of many small languages appears to lie in the hands of unenlightened governments and large corporations. The media mogul Rupert Murdoch is well aware that linguistic diversity has implications for his own business. In the 11th annual John Boynthon lecture delivered in Melbourne on 20th October 1994, Murdoch said,  Indian leaders have long been desperately worried about disunity in their vast, teeming, multilingual country. There has been an effort ever since independence to promote Hindi as the lingua franca,
what in India is called the "link language". But the effort has failed for a number of reasons. Until now. With the coming of the electronic mass media, Hindi is finally spreading because everyone wants to watch the best television programming. And I suspect we will see this story repeated throughout the developing world, not least in China with Mandarin. In which case, it will be not only prosperity that we will catch in our networks, but also order - and ultimately, peace.' ( The Australian, 21/10/94, p11).
 
 

In Bristol on 21st April 1995, a seminar will be held to explore the urgent topic of language extinction. Five distinguished speakers and an audience of around sixty persons will explore the reasons why Academe should worry about the catastrophic loss of languages that is looming. Some of the reasons have been sketched briefly here, and many others could be articulated. I should like to end, however, with the frightening suggestion that human beings do not yet know all the ways in which linguistic diversity is important.

The richness and complexity of Nature outstrips our understanding. Language is a both a cultural product and a natural phenomenon. Its richness and complexity demands our respect in the same way that the biosphere does. The natural human capacity for language - linguisticity - is something that human beings are only just beginning to study in a scientific way. Although grammarians and lexicographers have been around for millenia, linguistic science did not really take off until the late 19th century. One century later, our best linguists admit that we have so far achieved only a little understanding. Even a well-studied language like English is not yet understood completely. For example, there is no adequate descriptive theory of the syntax of English, let alone the semantics.

The fact is, no one knows exactly what riches are hidden inside the less studied languages. We have inductive evidence based on past studies of well-known languages that there will be riches, even though we do not know what they will be. Perhaps Waimiri-Atroari or Guugu Yimithir will give us some really important new information. The argument for conserving unstudied but endangered plants has a similar logic: strange plants may contain medically valuable ingredients, so there ought to be a presumption in favour of their survival.  It seems paradoxical but it's true. By allowing languages to die out, the human race is destroying things it doesn't understand.
 
 

THE CENTRE FOR THEORIES OF LANGUAGE AND LEARNING
UNIVERSITY OF BRISTOL DEPARTMENT OF PHILOSOPHY

REPORT  of seminar held on  April 21st 1995 on
 

THE CONSERVATION OF ENDANGERED LANGUAGES

                         Preamble
 

In 1994 and 1995, the public woke up to the fact that  the number of languages spoken around the world is currently in sharp decline. The cultural, ethical, political and
evolutionary implications are not generally understood. Nor have they been addressed sufficiently by experts. This seminar was intended to help raise the level of
consciousness of the issue amongst academics and laypersons in the UK.

The problem was officially addressed in 1992 by UNESCO, when its affiliate CIPSH (Conseil International de la Philosophie et des Sciences Humaines) announced a project to compile a 'Red Book' on Endangered Languages. A Clearing House and Data Bank Centre was eventually set up under the direction of Prof. S. Tsuchida at the University of Tokyo. An interim progress report is due later this year.

A periodically updated record of the world's languages is published by the Summer Institute of Linguistics. Ethnologue: Guide to the World's Languages  is currently in
its 12th edition (ed. Barbara Grimes). Recent collections on  the topic include:
Endangered Languages (1991) (ed) R.H.Robins and E.M.Uhlenbeck (Berg), and Investigating Obsolescence: Studies in Language Contraction and Obsolescence (1989) (ed) Nancy Dorian (Cambridge University Press).

The Atlas of the World's Languages edited by C. Moseley  and R.E.Asher (1994) incorporates 'time of contact' maps for the Americas and Australia showing the indigeneous languages  that were spoken when European colonisers arrived. These are
juxtaposed with inset maps charting the current distribution of languages. By comparing the two sets of maps the reader can see the degree to which native languages have been
usurped.

Various learned societies have conducted symposia on language extinction. In February 1995 the meeting of the American Association of the Advancement of Science attracted world-wide attention when Prof. Michael Krauss predicted  that up to 95% of the world's 6,000 languages would be extinct or moribund by the end of the next century, and Prof. Leanne Hinton reported that Northern Pomo, one of the 50 native languages of California, had become extinct with the death of its last speaker three weeks prior to the
meeting.

The Internet has helped to increase dramatically the flow of information about who is doing what in this field. The email forum based at the Australian National University
known as Endangered-Languages-L, which was inaugurated in  September 1994, has played an important role. The Bristol seminar would have been much diminished had this list not existed. Endangered-Languages-L was a major source of  factual information and contacts.
 
 
 

                        Summary of Proceedings

There were five presentations of around 50 minutes each  followed by a 2 hour discussion.

Mr. Christopher Moseley (co-editor of Atlas of the World's  Languages 1994)

 Mapping the Future of the World's Languages

The speaker gave an account of the overall design of  the Atlas with emphasis upon methodological and cartographic issues. He used the example of Canada to illustrate the need to separate linguistic identity from ethnicity. Moseley then  described in detail the history and social circumstances of  Livonian, a highly endangered Finno-Ungaric language spoken  in Latvia. It is crucial to realize that the drive for preservation is not based upon a romantic illusion, but upon the objective fact that each language is a unique repository of the accumulated thoughts and experiences of a community.

Moseley made a broad distinction between 'internal' and 'external' factors that affect the fortunes of a language over time. He provided a handout (appended) listing some of
the more important factors. These included location, distribution, status, the presence or absence of norms, public services and official contacts, education, media, and
cultural use. Another sheet (also appended) gave a breakdown by continent of the main endangered languages.

He identified two major tasks for the future: to establish a trans-national mechanism for monitoring changes in the distribution of languages, and to assemble speedily a
corpus of material on each language. He mentioned a project started by Routledge publishing company to compile The World's Languages Corpus, which will be made available on CD-ROM .
 
 
 
 

Dr. Mark Pagel (Lecturer, Department of Zoology, Oxford University)
 
 

        Should Linguistic Diversity be Preserved?
 

The talk was divided into two parts. Part 1 discussed scientific reasons for wishing to have a full record of languages in documented form, abstracting from any
considerations about the survival of speakers of those languages. These reasons are based upon the wealth of information about human evolution that can be extracted from
languages by the use of statistical techniques. Hypotheses about human migration patterns and the evolution of languages can be used to support and illuminate findings and hypotheses in genetics and population biology. Among the many pieces of research cited was a recent study by Ruth Mace and Mark Pagel demonstrating that there is a
latitudinal gradient in linguistic diversity on the North American continent analogous to the latitudinal gradient with respect to diversity of biological species. This suggests that both kinds of diversity are partly ecologically determined.

Part 2 discussed the role of language-learning as a cause of cognitive and neural changes in the individual. The speaker argued that when the child acquires dispositions to
categorize objects through early word-learning, some neuronal connections in the brain are strengthened while others are weakened or eliminated. He cited a neural net
simulation to show how previous learning affects a system's way of categorizing new stimuli, and concluded that, although it may be true that all human beings 'think in the
same way', one's native language influences one's perceptions. Loss of linguistic diversity implies a consequent loss in the range of potential ways of experiencing the world.
 
 

Dick Hayward (Professor, School of Oriental and African

Studies, London University)
 
 

 Submerging Islands: Language Loss in the Horn of Africa
 

The extremely complex linguistic profile of present-day Ethiopia is the outcome of geographical, social and political pressures that need to be viewed historically.
Amharic and Oromo are the major languages, but there exist nine groups of minority languages. Altogether the country possesses around 90 languages belonging to 4 families:
Semitic, Cushitic, Omotic, Nilo-Saharan. It is quite common for a person to have a working knowledge of four tongues, and many Ethiopians speak six or more. Linguistic
differences mirror cultural and tribal differences between mountain people and plains people, between hunter-gatherers, pastoralists and agriculturalists. During most of the 20th century the Amhara people wielded power, and Amharic was imposed as the language of government and commerce. The present Tigrinha-speaking regime has adopted an enlightened policy of tolerance towards minority languages. Nowadays one
can buy things in the street in languages other than Amharic.

Hayward then presented case studies of some small languages (the 'islands' of the title), showing the situation and prospects of each.. Koyra, a language spoken by a mountain community surrounded by Oromo-speaking plains-dwellers, is not particularly threatened by Oromo. The greatest threat is that it will be usurped by standardisation, that is, by levelling down the fifteen languages in its group to a single common denominator. Harar
is a language spoken only in one particular city. Daseneche is a Cushitic language spoken by people living near the north shore of Lake Turkana. Its presence there is due to
the migration of distinct groups having various prior linguistic affiliations, who were forced to become cultivators through loss of lands and cattle caused by the disease rindepest. Bayso, a language like Somali but a long way from the main Somali-speaking area, is used by a cattle-grazing people who escaped the Oromo expansion because they
settled on a island. Their cattle-based lifestyle is now precarious owing to overgrazing. Yam is another mountain language which has remained isolated from Oromo pastorality.
It is governed by strict rules linked to the social hierarchy; regal discourse is different from normal discourse. Although only five hundred persons speak Yam, half a million claim Yam ethnicity despite not knowing the language.

These micro-profiles illustrate that each language has a unique story to tell. Which language happens to be spoken in a given area can depend upon unpredictable contingencies but may also reflect deliberate choices regarding commerce and lifestyle.
 

Allan Wynne-Jones (European Bureau for Lesser Used Languages) Orchestrating Language Revival

The speaker applied concepts derived from theories of management and marketing to the task of nurturing a minority language. He argued that the 'product' to be promoted should
be regarded as the active using of the language within the life of the community, which is not accurately measured in terms of the number of speakers competent in the language.
Although the number of children who know Welsh has been steadily increasing, further language planning is needed to achieve full revival of the Welsh language.

Wynne-Jones divided the process from survival to revival into the following stages: idealism shown by enthusiasts, then protest, followed by legitimacy and institutionalisation, then parallelism, and ultimately normalisation. The last stage is characterized by a generally shared aspiration to a fully bilingual society.

The marketing approach emphasizes the need to adjust programmes to the wishes of consumers. Since the population of Wales contains a mixture of people having different
attitudes and preferences, crude hard-selling of Welsh would be counter-productive. Attention must be paid to background factors that might alter perceptions of Welsh, for example by dispelling the common prejudice that it is a language better adapted to poetry and to political rhetoric than to business.

 Language planning must be integrated with spatial planning and economic development. The man-made environment determines whether or not a language that can be used is in
fact used. J. Fishman claimed rightly that intergenerational transmission of a language is the key to its survival, but face-to-face relationships of all sorts are important.

The speaker briefly explained the activities of the organization Menter a Busnes, which aims to develop enterprise and business as intrinsic and creative elements in Welsh language culture. One of its latest activities, the Economic Development Forum,  is a consortium encompassing fifteen minority language communities in Europe. The
initiative is supported by the European Bureau for Lesser Used Languages. This is a unit within DG XXII, having at itsdisposal a budget of 3.75mecu in 1994-5 to finance projects
and run the Bureau.

Sibn Wyn Siencyn (Language Consultant, author of The Sound of Europe)

Thinking Twice: issues in Welsh as a second language in children under 5

Through bilingual education there has been a halt in the decline in number of Welsh speakers in all counties of Wales. If present trends continue, 56% of children will be
Welsh speakers in 2021, which means a trebling of the numbers. The key date will be 2011, as that year's census will show if the newly bilingual children turn into parents
who teach Welsh to their children.
 

The speaker described the highly successful 'submersion' technique for teaching Welsh to English-speaking preschoolers through play-activity in which all adults always speak Welsh - with plenty of repetition, short sentences, rephrasing and pointing - and where the child's responses in English are understood and acknowledged. The idea that bilinguality is desirable has caught on amongst non-Welsh-speaking parents. These groups have produced cohorts of children suitable for Welsh language education at
school.

Wyn Siencyn emphasised the excitement of bringing out the potential bilingualism of the child. Linguists and psychologists find it rewarding to study incipiently bilingual children for the light these children shed upon the processes of cognitive development. One robust
phenomenon is the simultaneous acquisition of parallel constructions, e,g Welsh pronouns and English pronouns are mastered at the same time.

The Organisers: Dr. Andrew Woodfield (director of CTLL)

Mr. Daniel Brickley (CTLL research assistant)
 © Copyright 1997
Garifuna World