The picture of Nubia and Nubian history was originally drawn by historians and archaeologists; in recent times, anthropologists and modern linguists have contributed their works. Actually, the first linguists were travellers in the seventh century who discovered and recorded the Nubian dialects of the Nile Valley. But it was not until the nineteenth century that specific studies dealt with these dialects, studies such as that of Reinisch in 1879, or Lepsius in 1840–53.

In the twentieth century starting as early as 1910, there was a discovery of ancient text written in Nubian. Scholars became interested in finding out the affinity between the Nubian language and other languages, and between the different dialects of Nubian.

Nubian has historically been classified as:

1. Mixed negro by Lepsius (1880);

2. Proto-Hamitic by Reinisch (1911) (“Proto” postulates Hamitic in its earliest underdeveloped stage);

3. Sudanic by Westermann (1911) and Meinhof (1912);

4. Eastern Sudanic by Greenberg (1948), MacGaffey (1961) and Trigger (1965).

Today there is a preponderance of opinion that Nubian is an Eastern Sudanic language, a branch of Nilo-Saharan. The Eastern Sudanic people settled in the Nile Valley and brought their language, which became the Nubian of the present day.

Linguistic conservatism and language maintenance

An intriguing question that echoes through many of the studies in this volume is whether there is a relationship between linguistic conservatism or corruption and language maintenance or death. The relationship between persistence of form and persistence of use can be phrased and questioned in two directions. The functional linguist asks whether, when we know that a language has undergone contraction in number of speakers or domains of use, we can predict certain types of changes in linguistic structure as a consequence. The editor of this volume has led in formulating and investigating that question (e.g. Dorian 1981, 1986c), and articles in the second section of this volume in particular bring important evidence to bear on the issue.

A different ordering of the variables gives rise to a different question, one of considerable interest to those readers concerned with the social, political, and social-psychological processes of language maintenance and shift, those of us still struggling to understand how the battery of factors that seem to overwhelm minority languages combine to produce their effects, or occasionally are turned back. The question is raised and confronted most directly in Eric Hamp's linguistically erudite review of phonological change in dying languages, but it is a question that is fundamentally about the social processes underlying language change and language loss. That question is whether, when we find interference or convergence phenomena in the structure of a language in contact with others, we should necessarily see these as signs that shift is in progress and loss impending.

Introduction

The first point I would like to stress in this chapter is that Warlpiri is not yet a dying language. The positive attitudes of the Warlpiri people towards their language and way of life encourage language maintenance. However, the young people are attracted by aspects of the non-Aboriginal way of life. The adoption of new ways, together with exposure to the English language, have led to changes in the language. There are many changes taking place quite rapidly, particularly in the language of the young people. Given statements that are frequently made in the Australian context, statements to the effect that all Aboriginal languages will soon have died out, it is worthwhile to point out some of the changes in progress that we have observed in the speech of the people in the community, particularly in that of the children. The data I will use in the paper have been collected as part of a longterm project on the children's acquisition of Warlpiri as a first language.

In this chapter I will discuss two major aspects of the changes in progress. One is the borrowing of lexical items from English; these borrowings are for both new and existing concepts, and what is of particular interest is the replacement of traditional words with English borrowings as well as the assimilation of the loanwords into the Warlpiri system. It is not unusual to find lexical borrowings when two languages are in contact. Dorian (1986c:259), for example, states that English words have been readily taken into East Sutherland Gaelic and adapted to the Gaelic grammar and morphophonology.

This volume came about in response to a need which had become time-consumingly obvious to the editor: people investigating a wide variety of precariously placed speech forms, in geographically diverse locations, were working in ignorance of each other's efforts. A kind of clearinghouse role had fallen to me, simply because I had published a book in a subfield which had no journal or other regular publication outlet, the subfield which had come to be known (for better or worse) as “language death”. I was spending increasing amounts of time putting researchers in touch with colleagues whose work was likely to be of interest to them, arduously passing along information on a case-by-case basis. Very late one night I found myself considering alternative solutions to the problem; this volume represents one of those alternatives.

Because of the circumstances in which the volume was conceived, a firm, ongoing part of the enterprise has been intercommunication: dissemination of methodologies, data, ideas, analyses, implications, hypotheses. A set of focus questions (see Introduction) was circulated to all potential contributors to help stimulate thinking along various but shared lines, and a month or so after the target date for first drafts from contributors, each manuscript then in hand was circulated to those who were clearly working on matters of relevance to one another's subject. Contributors were encouraged to correspond with each other and to incorporate crossreferences to each other's papers in their own chapters. All of the internal crossreferencing in this volume arises from authors' perception of what in other papers has special relevance to their own work, not from the editor's perception.

Introduction

The Breton language has been and remains today a powerful symbol of identity for Bretons. Both positive and negative symbolic values attached to Breton are critical variables in this language's persistence and in its impending death. The symbolic opposition of Breton with French is the most important variable in identifying why today Bretons speak most often of language death.

Language use has never been a part of the census of France, but various estimations of Breton speakers have been made by researchers in the nineteenth and twentieth centuries. In 1808 Cocquebert de Monbref estimated 967,000 native speakers, and in 1886 Paul-Yves Sébillot estimated the number to be 1,300,000. The Breton literary group Gwalarn estimated in 1927 that there were one million everyday users of Breton, and Francis Gourvil, a scholar of the Breton language and literature, estimated a figure of 1,200,000 for people who knew Breton. Figures for the 1970s and 1980s vary. Gwennole Le Menn (1975:72), a researcher on Celtic languages for the Centre National de la Recherche Scientifique (CNRS), estimated that 12 years ago there were approximately 1,200,000 people who knew Breton and 300,000 capable of using it as their everyday language, out of a total population in Brittany of approximately 3.7 million. In a 1987 survey of 999 people in Lower Brittany (of a total population of 1.7 million), Fanch Broudic, a Breton-language broadcaster for Radio-France/Bretagne Ouest (FR3), calculated that 61.1 percent understand Breton and 51.7 percent speak it ‘very often’, ‘fairly often’, or ‘sometimes’. The survey showed that an estimated 240,000 Bretons in Lower Brittany use Breton every day (Kuter 1987:4).

Cayuga, an Iroquoian language, was originally spoken in what is now New York State. When first encountered by Europeans, the Cayuga lived in villages around Cayuga Lake, surrounded by the other related Five Nations Iroquois. To the west were the Seneca, to the east the Onondaga, beyond them the Oneida, and finally, the Mohawk. Due to a series of unfortunate events, Cayuga speakers now live in two distantly separated communities, one in Ontario and the other in Oklahoma. While the language is still very much alive in Ontario, it is receding in Oklahoma, as fewer and fewer speakers use it on rarer and rarer occasions. Not surprisingly, the Cayuga spoken in Oklahoma has begun to differ in subtle ways from that spoken in Ontario.

Historical background

The Cayuga had occupied the same land for centuries before the American Revolution. When war erupted, they were drawn into battle on the side of the British, and in 1779 an American expedition destroyed their villages. This led to their dispersal. Some Cayugas remained with their land, but others went to live with Senecas at Buffalo Creek, and still others went with other Iroquoians to Ontario to form the Six Nations Reserve. By the end of the Revolution, only about 130 Cayugas remained at Cayuga Lake, 350 were with the Seneca at Buffalo Creek, and 382 were at Six Nations. (For a detailed discussion of the history of the Cayuga, see Sturtevant 1978; White, Englebrecht and Tooker 1978.)

The primary objective of second language acquisition (SLA) research is to uncover the principles which govern the process by which someone moves from no knowledge of a second language through various intermediate “stages” towards near-native control of the second language. This book provides an opportunity to view some of these principles from a different vantage point as well as an opportunity possibly to provide some new insights into the process of language obsolescence.

The standard view of second language acquisition (by immigrants and their children) makes it seem as if the chronologically first-acquired language will always be totally native and “successful” and any subsequent languages will be nonnative and often unsuccessfully acquired. The cases discussed in this volume show how one-sided the standard view of SLA is: SLA often proceeds the way it does because of inadequate access to and interaction with the people who speak and use the “second” language as their primary, fully-developed native language, not just because it is not the native (“first”) language of the learner. Immigrants typically have restricted access to the situations where they can successfully acquire the language of the country to which they have immigrated. Second- and third-generation speakers of a contracting or dying language are similarly limited in their access to the type of linguistic input and interaction that is necessary for them to become or remain competent speakers of their ancestral language. From this perspective, languages being partially acquired and dying languages seem very similar. My purpose in this commentary is to suggest why they are so similar.

Discussion from the perspectives of child language and aphasia; of historical linguistics; of social process; of pidgins, Creoles and immigrant languages; and of second language acquisition

As noted in Chapter 4, not all texts from the London–Lund and LOB corpora were included in the study, because of the time involved in editing the tagged texts. All genres included in the corpora, however, are represented in the study.

In addition, many of the text samples in the London–Lund corpus were divided. Texts were divided for one of two reasons. The first is that many of these texts, which are 5,000 words long, actually comprise two or more shorter texts. For example, a typical telephone ‘text’ consists of several conversations which are juxtaposed so that the total number of words in the text sample exceeds 5,000. In these cases, each conversation (or speech, broadcast, etc.) was separated and treated as a distinct text. If a text thus separated was shorter than 400 words, it was excluded from the analysis. (For this same reason, several of the letters that had been collected were excluded.)

Text samples that did not consist of several different texts were divided to obtain two samples of approximately 2,500 (continuous) words each. For this reason, these are not ‘texts’ in the sense that they are not bounded and do not contain all of the structural (textual) properties of a text. Many of the 2,000-word samples in the LOB corpus are of this type also; the text samples do not represent entire books, articles, or even chapters, and so do not represent entire ‘texts’.

Overall linguistic generalizations

There is a long history of research on the linguistic characterization of speech and writing. Although a variety of approaches has been adopted, the shared goal of most previous studies has been to identify specific linguistic features that distinguish between the two modes. Many studies also offer overall linguistic characterizations of speech and writing. In general, writing is claimed to be:

1. more structurally complex and elaborate than speech, indicated by features such as longer sentences or T-units and a greater use of subordination (O'Donnell et al. 1967; O'Donnell 1974; Kroll 1977; Chafe 1982; Akinnaso 1982; Tannen 1982a, 1985; Gumperz et al. 1984);

2. more explicit than speech, in that it has complete idea units with all assumptions and logical relations encoded in the text (DeVito 1966; 1967; Olson 1977);

3. more decontextualized, or autonomous, than speech, so that it is less dependent on shared situation or background knowledge (Kay 1977; Olson 1977);

4. less personally involved than speech and more detached and abstract than speech (Blankenship 1974; Chafe 1982; Chafe and Danielewicz 1986);

5. characterized by a higher concentration of new information than speech (Stubbs 1980; Kroch and Hindle 1982; Brown and Yule 1983); and

6. more deliberately organized and planned than speech (Ochs 1979; Rubin 1980; Akinnaso 1982; Brown and Yule 1983; Gumperz et al. 1984).