LANGUAGE DEATH: A THEORETICAL APPROACH

André Horak

University of Berne

 

 

 

 

    Two years ago, I asked a young schoolboy who was just about to enter high school if he wanted to study Latin later on. He told me that he didn’t, because dead languages weren’t fashionable any more. The boy obviously wanted to say that he did not want to learn a language that was not spoken any more; he was not aware of what a ‘dead language’ actually is.

    Latin and Ancient Greek are most commonly thought of as being ‘dead’ languages and therefore the reason why the term ‘dead language’ is well-known. However, the difficulties one is confronted with when determining ‘language death’, although unknown to most non-linguists, are more far-reaching than they seem to be. Even when trying to determine the terms ‘language death’ or ‘dead language’ from a linguistic point of view, an overall accepted definition is not easy to find. I took the fact that a folk notion is nevertheless difficult to define and explore linguistically as a motivation to make some further investigation into this field of language studies, trying to clarify ‘language death’ as much as possible. As I have limited possibilities to do some field work, I shall attempt to make a mainly theoretical approach, which should then be applicable to any concrete example of a dying language.

    This paper consists of three main sections. The first section deals with the theory necessary to understand the reasons for language death and the process that a dying language undergoes. To begin, I shall try to determine what ‘language death’ is. Further, there are some other issues to clarify, such as the explanation of the metaphor ‘death’, which is clear when referring to human beings, but rather ambiguous when used in a linguistic context. It is likewise interesting if or how it is possible to determine when a language is considered to be dead.

    I will then try to find an answer to the question as to why language death appears, referring not only to linguistic, but also to important social and socioeconomic factors. Finally, I shall come to an investigation of the process of gradual language death, that is, the process how a language dies. Doing this, I will mainly refer to Sasse’s theory (Sasse 1992: 3-20), who suggests three phases of language death. I shall also explain the terms language shift and semi-speaker in more detail, as I consider them to be important factors in language death.

    In the second main section, I attempt to examine some reasons proposed by linguists why people should be concerned about language death, concluding with arguments that verify the necessity of linguistic diversity. As a consequence of the need for languages, the third section of this paper shall propose possible ways of avoiding language death.

    I will now again consider the introductory example of to Latin, in which the set of difficulties becomes apparent if one wants to determine language death. McMahon (1993: 285) claims that ‘what has happened to Greek or Latin is not death, but metamorphosis’. According to her theory, these two languages have been affected by usual linguistic processes, and have therefore not disappeared, but changed into Modern Greek and today’s Romance languages (French, Italian, Spanish, Portuguese, Catalan, Romanian, …). However, this theory seems problematic to me. It is of course undeniable that modern French derives from Latin, and that these languages are therefore genetically related; nevertheless, they remain two separate, different languages. Hence, I claim that it cannot be said that one language ‘survives’ in another only because it appears in a transformed way or because some words of it are used again. If such a theory worked, it would then mean that most of the languages never die because of loan words other languages adopt as borrowings. For instance, one could say that Latin is not dead because of the borrowings ‘Gymnasium’ and ‘Corpus’ existing in modern German, or that any indigenous language that has no speakers any more is not dead only because some few words survive in some colonizing languages. Certainly, what McMahon meant is that Latin is not dead because French is a direct descendant from it. However, the strength of relations between languages is arbitrary. I therefore claim that her theory is rather vague and cannot be sufficiently supported.

    One could even go further and compare the linguistic examples with human ones. If a parent physically dies, this would, according to the previous ideology, mean that he is not dead, but still alive in the body of his children. Such a comparison is of course too far-reaching and the result of problematic metaphors such as death applied in linguistics.

    Nonetheless, the comparison of languages and human beings becomes interesting when trying to determine what a language is and when it stops being a language; a question which leads me back to my original purport: to answer what language death is. Languages wouldn’t exist without speakers, because there would be no need for them. Therefore, if there are no remaining speakers of a language, I see no reason why it should be claimed that the language is not dead. As a result, I come to my first provisional conclusion, namely that a language is dead when all its speakers have disappeared. To prove that I am not the only one supporting this idea, I shall quote Brenzinger, who says that ‘a language is considered to be extinct when there is no longer a speech community using the language’ (Brenzinger 1992: 3).

    However, even between the linguists supporting the theory that language depends on human beings and disappears with its last speakers, there are controversial ideas. For instance, David Crystal considers language as a ‘tool of communication’ (Crystal 2000: 2), and therefore claims that it is only alive when there is more than one speaker left, because only one speaker makes communication impossible. While understanding his idea, I only partly share his opinion. It is undeniable that there cannot be communication with one remaining speaker of a language. One last speaker is however enough to pass on his linguistic knowledge to a child, who easily acquires the capacity of a native speaker. As a result, it can be said that with one speaker communication can easily be established and Crystal’s understanding of what the aim of a language is, that is, to serve as a ‘tool of communication’, is therefore satisfied. Therefore, I claim that with one speaker, a language is only restricted in its function, but nevertheless existing as a living language.

    To find a definition of (gradual) language death, I again refer to McMahon, who claims that language death ‘involves a transfer of allegiance of part of a population from a language which has been native in the area, to a more recently introduced language in which the indigenous population has become bilingual’ (McMahon 1993: 285). This ‘transfer of allegiance’ goes that far, that the less prestigious language completely disappears in favor of the dominant one. I shall come to the reasons of language death further below.

    We saw that the question as to what language death is may result in a long discussion leaving more questions open than originally raised. I now come to a personal conclusion saying that language death is a phase of linguistic evolution, in which no speakers of a language are left any more. Subsequently, I attempt to find the more explicit reasons why a language dies.

    Language death is a common phenomenon in language studies. It does not necessarily appear under a particular environmental condition; a language can die in a civilized society, e.g. Latin, or on a remote Pacific island. Thus, it can be assumed that there must be more than one reason why a language dies.

    As mentioned before, language exists through and depends on its speaker. If these speakers are physically threatened, their language is automatically in danger as well. Crystal illustrates some examples of physical threatening, which are mainly in form of catastrophic natural causes. Especially ‘small communities in isolated areas can easily be decimated or wiped out by earthquakes, hurricanes, tsunamis, floods [or] volcanic eruptions’ (Crystal 2000: 71). Such an ‘elimination’ of all speakers of a language would have as a consequence the sudden and immediate death of the language affected. In the Roman period, Latin disappeared gradually, resulting first in Pre-Romance varieties, before finally transforming itself into today’s modern Romance languages. Now, we have also seen sudden death, which is the most rapid way of language extermination.

    Economic situation is another factor, which may become a threat to a speech community and therefore the reason for migration. One can find numerous examples where people are forced to abandon their homes due to a bad economic condition in their country. For instance, natural disasters, such as droughts or floods, may lead to famines in financially unstable Third World countries and thus to unbearable living conditions. A similar case, although caused by humans themselves, is war, which results in a reduction of the population not only because of the victims killed in combat, which I call ‘primary victims’, but also because of ‘secondary victims’. These ‘secondary victims’ die due to diseases, such as tuberculosis, especially present in Third World countries which have been devastated by civil wars for many years and where bad economic situations do not allow an elaborated health care (Cf. also Crystal 2000). It shall be mentioned at this point that large-scale migrations do not necessarily lead to language death. Nevertheless, if the number of emigrants is significant enough, migration may be the first stage in the process which finally results in language death.

    Civil war implies a problematic political situation, which is another reason for migration. However, political problems do not always result in wars. A good example to illustrate other forms and consequences of political oppression is delivered when considering the colonialist period in the past as well as the neocolonialism these days. When a dominating nation conquered new lands, the indigenous population was suppressed by both physical power and the ideology that the new language brought in by the conquerors was ‘better’ and superior to their own. Moreover, the official language and the language of the education system was mostly the colonists’ one, which was therefore regarded as more prestigious. Due to its frequency, I claim that prestige is one of the most important factors in gradual language death.

    In the past, there have also been examples of colonist behavior with the extreme high point of total prohibition of indigenous languages and systematic persecution of its speakers. In this context, the term language murder has been created to designate the procedure which tolerates the dominant language, now called killer language, as the only possible way of communication (Cf. also Crystal 2000 and Fishman 1991). In linguistics, there are several other metaphors in use referring to language death, such as healthy language or language suicide. The use of these metaphors is not inappropriate; however, one has to bear in mind that they are metaphors. The danger in the use of such terms is that one could easily misunderstand the meaning of the metaphor and consider language as a living organism, which it is not. Therefore, I shall at this point insist on the idea that language is only a tool of its speakers and cannot exist on its own.

    Another economic reason for emigration is the need for jobs which are often not available in the countryside. The result is an increasing urbanization accompanied with the adoption of the language or linguistic norms necessary to get a job in town. There, the dominant language or language variety has mostly become the only possible way of communication, that is, a lingua franca.

    Undoubtedly, people from the countryside mostly speak the same language as those from urban areas, although with different dialectal features. In the case of dialect speakers living in cities where the supposed ‘standard’ variety of the same language is spoken, the rural dialect rather than the language itself would be affected and endangered. It would thus be an example of ‘dialect death’. However, as I shall argue subsequently in this paper, the boundary between dialect and language is not clearly defined; therefore, ‘dialect death’ and ‘language death’ are to be treated as equal.

    The term lingua franca has created controversial opinions among linguists. Crystal (2000: 29) argues that ‘if one language does […] become the world’s lingua franca, […] it does not follow that this must be at the expense of other languages.’ David Crystal is right, if one considers international communication as the only role of a lingua franca, without bearing in mind other consequences that the propagation of one single language might bring. I argue that a lingua franca is another factor suppressing other languages, which then are more likely to disappear.

    The reason for my assumption is connected with the natural, human need for communication, which people try to satisfy by learning foreign languages. If the inhabitants of a country, where two languages of globally minor importance are being spoken, try to facilitate their national communication, they most probably learn the other national language which is not their native one. If then a third, foreign language is ‘imposed’ on such a system of communication, because it has become the common language of commerce, education, etc., e.g. English, the two minor languages may lose their importance, and thus their learners. Of course, the immediate consequence is not language death. What occurs in such a case is most probably diglossia, that is, a functional distribution of the different coexisting languages. However, as I have personally observed, the step between diglossia and obsolescence of one language is very small. It can sometimes be remarked in the case of a family, where the parents are immigrants and therefore speak a foreign language, whereas the child’s education is in a different language. In some cases, the child will continuously refuse the communication in its actual native language, before it may finally become monolingual or semi-speaker of the native language due to its lack of practice.

Such a dominant language can be called lingua franca, which due to its frequency hinders the propagation of other languages. As a consequence, the superseded languages may lose their speakers and gradually disappear (as we have seen before, some jobs may require applicants to be speakers of the lingua franca, which might result in the fact that some people stop speaking their native language). In such a case, a lingua franca becomes a killer language and the reason why another language dies.

    These days, the propagation of the dominant language has become easier than ever through the media. I consider the youth as the most affected social group, which today feels attracted by English as a modern dominant language that may later on develop into a killer language. The first affected languages would then be those with only few speakers left, whereas well-established languages would adopt amounts of new borrowings rather than disappear. Examples of English borrowings, already in a considerable number, can be found in everyday life (especially in science). Surely, we cannot speak of language death these days, but if the intensity of English borrowings continues at the same level, other languages may become endangered in a few generations’ time. 

    To sum up the part in which I tried to give an answer to the question why language death appears, we have seen that there may be more than one single reason for a language to die. A radical and rare way is the immediate death of all speakers of a language in natural catastrophes. The most frequent possibility is gradual language death due to the ‘domination effect’ by another language, for social or economic reasons. In consequence of the processes cited above, bilingualism, which is typically established in former colonies, becomes less frequent or impossible (due to prohibition) between the indigenous populations, until the last speaker of the native language disappears. What remains is a monolingualism of the new dominant language. Although it attempts to facilitate communication, a lingua franca may become such a dominant language and thus be another reason for language death.

    After now being aware of the term language death and the reasons why it arises, I shall have a closer look at the process of how a healthy language dies. To do this, I will refer to Hans-Jürgen Sasse, who proposes three phases of language death (Sasse 1992: 20-21).

    The first stage Sasse suggests is the phase of a primary language shift. The reasons for language shift to arise may be the same as illustrated above in connection with language death, that is, prohibition, persecution, physical threat, natural catastrophes or wars. However, one of the most frequent origins is a new, recently introduced dominant language, which continuously suppresses (or kills, to use a metaphor) another dominated one, until this last becomes restricted in use or entirely disappears.

    What remains to be explained now the term shift used at this point. In a closed linguistic area, such as a small island, the indigenous language is the only possible way of communication. It is used in all registers (as long as there are some) and social classes. Then, usually a dominant nation opens the closed circle of the indigenous speech community, introducing and imposing their language as the one used in administration and politics. The natives, who want to start careers in such a new political situation, are then forced to learn the dominant language, which has gained in prestige and become official. What has happened is that the new language has become prestigious due to its use by the dominant nation. Of course, the natives who have not been forced to work in town will keep their language alive. If the native population is, however, small enough, or if the dominant language has reached all indigenous areas, the entire speech community of a native language can easily be affected. Examples of linguistic territories occupied by dominant languages are numerous. For instance, Basque is dominated and highly endangered by Spanish, in spite of the local attempts to keep it alive (through literature, separatist movies, etc.).

    Although easily understandable, language shift is difficult to define (which is confirmed by the fact that linguists often try to avoid short definitions). Theoretically, I would explain language shift as a change in linguistic attitude in the indigenous people’s minds towards their and the new language. ‘Once a new language becomes dominant in a certain speech community the old one is potentially endangered […]’ (Sasse 1992: 21). Language shift is ‘often a [gradual], slow and cumulative process’; it is therefore difficult to be identified (Fishman 1991: 40). The original language loses its power and prestige in favor of the newly introduced one; it becomes restricted to only a few situations, such as prayers, among friends and in informal speech.

    Once language shift has begun, it is very difficult to stop this process. In particular cases, it is nevertheless possible to resist the oppression by the dominant language. For instance, a minor cultural-linguistic area that is politically embedded in a country, such as Chechnya or Catalonia, often distinguishes itself by using its own language and thus at least keeping its linguistic independence. Another case is the delimitation of youth groups, who want to separate themselves from the adults using their own slang and resisting the pressure of the formal language taught in school. What again occurs in such a situation is diglossia, although with two dialects or language varieties, and not languages. However, the process remains the same: the youth slang may become dominant and supersede the formal language in the youngsters’ speech. Youth jargons can be found in every language, and as mentioned before, (English) borrowing is not an infrequent method to obtain a special youth register.

    After the initial phase of language shift, the second stage in Sasse’s language death model is the period of language decay, which ‘is defined as the serious linguistic disintegration which is typical for the speech of so-called semi-speakers’ (Sasse 1992: 15). Semi-speakers’ speech appears in a reduced form compared to the speech of native speakers. Their imperfect knowledge of a language can be noticed considering their faulty morphology or the loss of grammatical categories in their speech. Further, they mostly use phonemes which only exist in the dominant language; that is, they apply the phonological system of the dominant language to the dominated one, even if the latter requires different phonological rules. I shall give further examples of semi-speakers’ linguistic imperfections subsequently. What I will examine now is why this type of speaker appears.

    One theory is that semi-speakers have as their models native speakers who have not maintained regular communication in their language and therefore lost their capacities of full speakers. The semi-speakers would then have acquired a ‘perfect’ knowledge of the variety as spoken by speakers who have become imperfect due to lack of practice; thus, semi-speakers do not have the fluency and correctness of natives. I have personally made such an experience, although not to the extent of a semi-speaker. The parents of a friend emigrated from Czech Republic more than 15 years before the communist regime broke down. Therefore, they did not have the possibility to maintain regular contact with their former home, which is the reason why linguistic innovations could not get through to them. As the span of time in which the contact was interrupted was not large enough, these innovations, mainly in form of (English) borrowings, are not of major importance. However, reading a Czech newspaper, they realize that somehow the flow of information has been interrupted. As my friend has never studied Czech at school, she has almost completely relied on her parents’ and their friends’ speech (besides the graphic medium). Therefore, as their language, unlike standard Czech, has not remarkably changed for the last 30 years, I would qualify my friend as a native (bilingual), but nevertheless imperfect speaker of Czech. Of course, her competence in Czech is on a much higher level than a semi-speaker’s, and probably higher than the language competence that can ever be reached by someone who is learning Czech as a second (foreign) language; however, if she passed her language as it is today (that is, without studying Czech as spoken in the mother country) to her children and so on, semi-speakers would probably occur in two or three generations’ time. I shall now come back to some language deficiencies of semi-speakers, which I partly remark in my friend’s speech (although very weekly).

    First, semi-speakers tend to lose grammatical forms, such as the gerund. Complicated sentence constructions are often reduced to main clauses only, whereas subordinate clauses are often lost. Tenses are mostly saved, but mixed up, which results in their inappropriate use, e.g. in reported speech or conditional phrases. Semi-speakers use tenses and apply grammar unconsciously as it is done by ‘full’ speakers, but their grammatical feeling is defective and thus often misleading. As a result, their speech is characterized by agrammatism. As already mentioned, semi-speakers tend to use the phonology of the dominating language also when speaking the dominated variety. Once they realize this phonological inadequacy, they try to improve on it, which however results in hypercorrection; for instance, speakers pronounce phonemes where they should not, as in the case of the ‘r-intrusion’ in English. To give an example, I shall consider the phrase ‘my idea is’. In fluent speech, there may arise an r-intrusion between ‘idea’ and ‘is’, resulting in /ai’diər/. A ‘full’ speaker will surely be aware of the correct word ‘idea’ due to his natural mental grammar. However, a semi-speaker whose mental grammar is most probably defective may think that /ai’diər/ is correct in the form as he commonly pronounces the word, that is, with an ‘-r’. When he then notices that he was wrong, he will in many cases correct all words with the same ending, such as ‘clear’, which results in /kliə/ even in fluent speech; this phenomenon is called hypercorrection (cf. Sasse 1992: 61ff.).

    To reformulate the last section, the phenomenon appearing in language decay is reduction. The dying language loses some of its important qualities which facilitate communication and becomes thus reduced in function; in other words, a part of a language is lost. A reduced language is mostly affected in the vocabulary, in its grammar and tenses. Trudgill claims that ‘reduction is found only in the case of languages which have no native speakers and are of restricted use’ (e.g. in the use of a lingua-franca, a pre-pidgin or a pidgin). Saying this, he most probably assumes that the mental grammar of a pidgin speaker is defective, whereas the one of a creole speaker is not. However, I do not agree with this claim. Being a native speaker of Czech, I feel that the mental grammar of my already mentioned friend, who is a bilingual speaker of Czech (which is a language, not a creole), is not as ‘healthy’ as the one of a Czech living in the mother country, which leads me to raise that (week) reduction may very well occur in her speech (and thus also in the speech of semi-speakers).

    Reduction is typical of obsolescent languages and must not be mixed up with simplification, whose aim is to facilitate the use of a language (through the increase in invariable word forms, loss of inflexions and affixes or regularization, e.g. in the formation of the plural, in orthography, etc.). The speaker who is affected by reduction is called semi-speaker. To avoid confusion, a semi-speaker is not bilingual. A bilingual speaker has similar knowledge in both languages, whereas semi-speakers’ speech is defective (reduced) and therefore a phenomenon of language decay. Claiming this, I am aware of the problem of determining whether someone is a semi-speaker or rather bilingual. Such a determination is difficult and will most probably lead to the conclusion that the boundary between both types of speakers is not clearly defined and thus arbitrary. In other words, the difference between semi-speaker and bilingual is gradual, and therefore leads to a proficiency continuum. What I consider as a bilingual speaker in this paper is someone whose speech remains intact in the main functions (e.g. tenses, syntax) and thus approaches to a ‘full’ speaker’s speech, whereas a semi-speaker is clearly identifiable due to the incorrect use of his language.

    As the semi-speaker is a central notion pertaining to the process of language death, I shall have a look at another related problem, that is, the problem of how semi-speakers can be identified. The easiest case is the identification of these imperfect speakers among full speakers. If the members of a speech community show different linguistic capacities, it then only remains to assume the most competent speech to be the most reliable one in terms of the fluent speaker. Once this is done, semi-speakers can now be easily recognized comparing the different varieties within the same speech community and assuming that the most imperfect one pertains to the semi-speaker.

    The investigation begins to be problematic with only semi-speakers left. If all members of a speech community, which is sometimes reduced to only one person, are semi-speakers, it is difficult to identify them as such. With only imperfect speakers, it becomes difficult to judge the reliability of their speech, e.g. it is problematic to determine whether or not their grammar is correct, and thus, if they are fluent speakers or not (Cf. also Dorian 1977: 23-32). The reason why I touch on the identification problem of semi-speakers is that their appearance is a clear indication of a dying language. The reconstruction of the original language from the speech of semi-speakers exclusively is a process which I shall explain in the last section of this paper.

    After having seen the reasons for language death and the processes involved in it, I shall now come to the second main section of my paper. Crystal (2000: 27-67) has collected a number of theories by various linguists, why we should and why we should not be concerned about language death. Reformulating some of those ideas, I shall also argue that some of them are not legitimate ways of thinking. Nevertheless, the aim of the following section is to demonstrate that language death is not to be taken lightly.

    It has been said that we should not be concerned about language death, justifying this claim with the ‘Biblical story of Babel, [which says] that the proliferation of languages in the world was a penalty imposed on humanity, the reversal of which would restore some of its original perfectibility’ (Crystal 2000: 27). This ideology is based on the Bible, as there are others based on different myths. Such an argument does of course not work, as one cannot involve mythology, religion or ideology into scientific examination. Science is based on facts; it can therefore not involve theories based on simple beliefs or assumptions.

    A more serious reason why language death is desirable rather than to be prevented can be found in the propagation of one single, world-wide language (lingua franca), which would enable a fast international communication without any problems of linguistic code (one single existing language implies the death of all other languages). A unique language would also be an economic advantage, as there would be no more need for translations. Once the linguistic barriers are removed, one single language should contribute to the peace on earth. However, as civil wars in monolingual countries prove, it is utopian to think that quarrels can be avoided by only standardizing the language. Such a standardization would further impose the problem to decide which language should be the one to serve as an international, generally-accepted code.

    Of course, one might argue at this point that we are far away from aiming at one single language, and that a reduction of 200-300 languages would nonetheless be a good solution. However, it is too easy to consider language studies as an isolated field of investigation; language is moreover interwoven into other fields, such as history, as we shall see below. The consequences of a reduction of the number of languages would therefore be much more drastic than they seem to bee at first sight.

    It is undeniable that the continuous death of languages affects the linguistic diversity in the world. As mentioned before, language is strongly related to other fields, such as culture. However, in some cases, such as Ireland or Scotland, language and the maintenance of tradition can be separated the one from the other. For instance, this occurs when the original language (e.g. Irish) does not have many speakers left any more; in such a case, language necessarily has to be separated from culture (otherwise, any maintenance of culture would be impossible). Nevertheless, tradition and culture are often transmitted through language; a diminution of languages would therefore have as its consequence a diminution of cultures. As an example, most indigenous populations in Middle and South America keep celebrating their tribal customs and traditions in their native languages (e.g. Guaraní, Nahuatl, Quechua, etc.). Again, this is possible because the Native Americans in those regions have kept their languages alive and consider them as important factors of identification. If those languages and hence the relation of the indigenous to their origins were lost in favor of Spanish, there would at least be a decrease in traditional rituals. As a consequence, the loss of a language or language variety may affect the financial situation of a country, whose economy relies on tourism. Hence, the conservation of culture and thereby also the maintenance of language variety is not only desirable, but may be important to keep the living standards. If language contributes to the well-being of a nation in such a significant way, its role is not only to be a ‘tool of communication’ (Crystal 2000: 2), but also to ensure a nation’s future.

    The importance of language diversity can also be remarked from a historical point of view. Today’s languages consist in huge parts of remnants of old, dead languages, such as Latin. Those dead languages survive in modern languages in form of borrowings, or leave us some structural or morphological features. This fact not only contributes to a lexical variety, but also allows us to investigate the exact processes a language has undergone from its beginnings until today. The history of a language is closely linked to the history of its speakers. The knowledge of when a certain feature first appeared in a language and from which foreign language it was taken makes it possible to reestablish the genealogy of a nation.

    Without language diversity, historical investigation would be more difficult to make. As an example, we know that a large number of French borrowings have come into English during a certain period, that is, at the time of the Norman Conquest, when the two languages were in close contact. Language (in this case English) contains in its etymology the best form of historical testimony. Crystal (2000: 35) explains that ‘[e]ach etymology demonstrates through its presence a point of contact, an index of influence. Words become part of the evidence of social history.’ Without such words (borrowings), historical research becomes more difficult. Thus, we see that besides the linguistic factors, language can also deliver important testimonies of populations from the past. Linguistic diversity is a way to make language more attractive. Moreover, it can hardly be imagined to make some investigation in historical linguistics without different language situations available through language diversity. However, the point that linguistic diversity facilitates historical research is a doubtful reason to maintain different languages (cf. Crystal 2000: 32-36).

    Language equals identity, which is another reason why the maintenance of language diversity is important. When speaking of languages, we should not only focus on so-called official (standard) languages. Moreover, dialects can also have the function of identification and are therefore to be treated the same way as languages. ‘The boundary between dialect and language is arbitrary, dependent on sociopolitical considerations […]. Dialect death is language death […]’ (Crystal 2000: 38). Although the argument that language helps to keep one’s identity is evident, the consequence of dialect death is remarkable and can be noticed in the fact that people have always tried to collect and compile old words and regional tales (often in dialect) containing rural expressions.

    Finally, I shall mention a purely linguistic reason why languages should be kept alive. From a linguistic point of view, each dead language is equally a loss of a resource, which could be used for investigation. Especially isolated, indigenous languages might be of major importance, as they tend to develop independently from other languages and have therefore not been influenced by borrowings or by any superstrate language. Non-linguists often consider such indigenous languages to be primitive, and therefore think that they are not as important as ‘civilized’ languages. Thus, not enough attention is paid to the threat of indigenous languages due to the wrong belief that they are not valuable enough. However, this is an error, as such languages may even contain more complex phonological, grammatical or lexical features than English, Spanish or German. In the end, the fact that each language is a part of the science that linguists are dealing with should be significant enough to care about the existence of every language.

    The final section of this paper deals with the question of how language death can be avoided, and, if it has already begun, how a language can be fully revitalized. In order to be able to establish some general rules applicable to any case of language death, most linguists have first considered some concrete examples, which should serve as a basis for their investigation. The main idea is to find out how languages could survive in the past. Based on this knowledge, it should be possible to apply the same ‘rescue’ method to endangered languages.

    So far, we have only little knowledge about the reasons why minority languages survive and why speakers of such languages are motivated to keep them alive. Some cases have shown that a language is likely to survive in geographically isolated regions, where any form of language contact is interrupted. Isolation is thereby often linked with either no education at all, or monolingual education only. Such an example of geographical delimitation can however only be significant in areas where the media do not play a significant role, e.g. in some regions in Africa and South-East Asia. An artificially created isolation of civilized regions (e.g. in Europe) is of course not possible, a fact leading to the conclusion that this example cannot be a starting point to find some general rules how language death can be avoided.

    In other cases, languages have been kept alive through the media, that is, through written and spoken forms of communication. Moreover, an established political independence has often led to a new national pride, which required the maintenance of an own national language. However, this method is not applicable to regions where there are no media yet, or where the speakers of a language do not have such a patriotic feeling (cf. Crystal 2000: 128-129).

    These two examples are sufficient to illustrate that it is difficult to find some general rules for language reconstitution. Nevertheless, Crystal has established six factors ‘which turn up so frequently that they could be recognized as postulates for a theory of language revitalization’ (Crystal 2000: 130). What he basically suggests is that a language is more likely to survive if it is written down. In order to realize a written language, a sufficient educational system has to be established. The next step is that a speech community which is able to record its oral conversation in form of written documents should gain prestige, because only a prestigious nation can be wealthy, and – connected with wealth – powerful. Based on power, technological innovations and their propagation in the originally endangered language should finally lead to a language which is no longer threatened. Some of these points have been applied to regions with endangered languages, such as Welsh, Basque and Gaelic in Europe, or Quechua, Guaraní and Navajo in America. The application of all six factors together is, of course, an idealization of language reconstruction (cf. Crystal 2000: 130-144).

    Crystal’s ideas are right; however, I doubt if they are effective enough to prevent a language from dying. The reason for my concern is that a language may be identified as an endangered one when there are only semi-speakers left. Thus, a ‘rescue’ has to be fast. The procedure cited above, that is, the establishment of a literate society, education, power and so on, may be too slow, as it would be too time-wasting to reconstitute a fully fledged language from a semi-speaker’s speech. Time is something we do not have in such situations, as we often realize too late that a language is about to die. Therefore, I claim that Crystal’s factors are mainly useful in situations where the process of language death has not yet started, but where it is probable that such a process may start.

    Now, I shall briefly mention Hans-Jürgen Sasse’s theory, which suggests different methods of revitalization according to the three phases of language death illustrated above, that is, language shift, language decay and language death. During the first phase, there has to be a motivation to keep a language, such as ‘political situation, […] separatist movements […], migration and gain of new prestige’ (Sasse 1992: 21). In the phase of language decay, a number of remaining full speakers have to start teaching their children in the endangered language in order to keep it alive. If there are not enough full speakers left, the language has to be reconstituted with semi-speakers. However, it is doubtful if the resulting language is the same as the original one.

    Finally, there is also the more complex situation of language reconstruction with no speakers of a language left at all, that is, in the case of language death. An example of such a reconstruction or revitalization is ‘Modern Hebrew, which was created on the basis of the codified holy texts after more than 2000 years of interruption of regular language transmission’ (Sasse 1992: 22).

    Language reconstruction is a task which becomes more difficult the more advanced the process of language decay and thus language death is. The different theories of revitalization underline that each situation requires its own procedure. Therefore, an establishment of general rules how a dying language can be saved seems to be problematic.

    To sum up this paper, we have seen that a language death situation occurs when there are no speakers left any more. According to a different point of view, which considers language as a tool used for communication, a language is already dead when there is still one speaker left. With the illustration of such different opinions, I attempted to show that the metaphor ‘death’ is in linguistic terms ambiguous and allows various interpretations. We have further seen that not only linguistic factors, but also social and socioeconomic factors are involved in language death. The process which a dying language undergoes can be divided into three phases. At the beginning, prestige and pressure of the dominant language are reasons for language shift to occur. Once the first stage has begun, the emergence of semi-speakers, the resulting language decay and the final language death are only a question of time.

    The second section discussed if and why people should be concerned about language death. As we have seen, an argumentation based on mythology is, together with the demand for a world-wide lingua franca, utopian and moreover not reasonable. Due to the strong link of languages to fields such as culture, history, and to the notion of identity, and only for purely linguistic reasons, a prevention of language death is desirable.

The final section of this paper attempted to find some general rules that could be used to avoid language death. However, some examples have illustrated that the different situations and regions in which languages die make it difficult to establish such generally-applicable rules, or at least rules that are practically realizable in a short span of time.

    In the end, I cannot and do not want to judge whether language death is a positive or negative phenomenon. I shall only point out that language is mostly adapted to the natural need of human beings. Therefore, if for some reason a language does not contribute to a ‘positive’ quality of life, it should not be kept alive artificially.

 

 

 

 

 

 

 

 

 

 

References

 

 

Brenzinger, M. (1992),   Language Death. New York: Mouton de Gruyter.

 

Crystal, D. (2000),     Language Death. Cambridge: Cambridge University Press.

 

Dorian, Nancy C. (1977),  «The problem of the semi-speaker in language death», In Fishman, J. A. (ed.), International Journal of the Sociology of Language. New York: Mouton Publishers, 23-32.

 

Fishman, J. A. (1991),  Reversing Language Shift. Clevedon: Multilingual Matters LTD.

 

McMahon, April M.S. (1993),  Understanding Language Change. New York: Cambridge University Press.

 

Sasse, H.-J. (1992), «Theory of language death». In Brenzinger, Matthias (ed.). Language Death. New York: Mouton de Gruyter, 7-30

 

Trask, R.L. (1996),  Historical Linguistics. New York: Oxford University Press.