LM and LS research is now in its seventh decade since its establishment as a dedicated field of study in the 1960s. Like other fields of specialist study, the field of LM and LS has changed over time, reflecting both internal and external developments in the discipline. The preceding chapters of this book have documented the main developments that took place between the 1960s and the late twentieth century. We have shown how a variety of disciplines not only became increasingly interested in this topic of study but also contributed to it through different methodological approaches and theoretical insights: they included but were not limited to (linguistic) anthropology, social psychology of language, sociology of language and sociolinguistics. Indeed, other disciplines such as demography, political science and even economics have made contributions (e.g., Grin and Vaillancourt 1997; Ozolins 1993; Veltman 1983; amongst others). As we mentioned in Chapter 2, scholars interested in the language situation of migrant communities in the ‘new world’, primarily North America, spearheaded the development of the field. Later, its geographical range expanded to other parts of the world – Europe, South America, parts of Asia and Southern Africa – where new waves of immigrants had settled. A particularly interesting development concerned the expansion of LM studies to migrant communities in Europe. Until the 1960s Europe had in fact been the major ‘source’ region for migration to the new world. Its citizens, communities and their language practices constituted the focus of the pioneering early studies in North America and Australia, e.g., studies of Norwegians, Swedes, Dutch, Germans, Greek, Italians and Poles. In the second half of the twentieth century, Europe changed its role from being the source for to the recipient of migration, as we described in Chapter 2. The other main setting for LM and LS research is that of territorial linguistic minorities around the world. However, this setting has become increasingly the subject of a separate branch of language contact studies: language endangerment. Unfortunately, an already large and still growing number of the languages of the world's linguistic minorities find themselves in this state. In these cases urgent linguistic documentation is the primary objective to be achieved before any revitalisation or revival can be undertaken.

CHALLENGES OF USING QUESTIONNAIRES IN MULTILINGUAL RESEARCH

Questionnaires are widely used to examine various forms of social behaviour and practices. They are the preferred data-gathering tool for macro-scale projects and surveys. Census surveys, discussed in the previous chapter, are a prime example. Questionnaires are also widely used in macro-scale research focusing primarily on language issues: usage patterns and linguistic practices. Typically, scholars investigating questions of LM or LS try to gather data from a representative sample of the group or community in question to make assessments on the state of play. However, questionnaires are also a useful tool in small(er)-scale research, for example, when you want to identify what languages primary-aged school children use at home and with their friends. In such types of research, the questionnaire is seldom the only data-collection tool but is complemented by other modes of data collection, including case studies, ethnography, participant observation, testing and experiments. The use of questionnaires in research on linguistic diversity shares many elements with its use in other types of social research: sampling issues, the distribution and administration of the questionnaire, the formulation of questions. Some of these, however, are more challenging because of the linguistic diversity of the respondents. For example, the standard sampling techniques may not be very efficient for locating participants when the focus is on their bi- or multilingualism. Similarly, a decision needs to be made in which language the questionnaire should be presented. If the questionnaire is to be administered by an interviewer, should this interviewer be bilingual, should he or she be a member of the same ethnolinguistic group or not? These are some issues and questions facing researchers who rely on the use of questionnaires in exploring language use and language attitudes in a multilingual context. In the next sections we will discuss some of the challenges particularly pertinent to research into LM or LS. They include the sampling of participants and the language(s) used in the questionnaire. The rest of the chapter will be devoted to a discussion of the types of questions used to probe both (reported) language use and language attitudes. Finally, we will mention some aspects of data processing.

BEYOND SURVEYS AND QUESTIONNAIRES

Talking with people about their language use (interviews) and observing how people use languages in a range of settings (participant observation) are probably the two most frequently used methods, besides questionnaires, in the exploration of LM and LS. Experiments and tests are another tool, although these are used primarily for investigating attitudes and linguistic proficiency.

Even when the research is more quantitatively focused, such as in large-scale language surveys, it is sometimes complemented with a small sample of interviews in which a more detailed discussion about language use and language choice takes place. Where the focus of the research is on the language behaviour and language choice of a specific community or group, such as Maori adults in Wellington, New Zealand; Bangladeshi children in London; Puerto Rican youngsters in New York or Turkish immigrant women in Hamburg, Germany, interviewing members of these groups and communities is the standard way of exploring relevant language issues. Participant observation is also a common way of investigating language behaviour as it allows researchers to observe actual linguistic behaviour and practices in various settings or locations rather than having to rely on self-reports.

In the following sections we will discuss these main ‘other’ tools for researching language practices in the context of LM and LS. Section 5.2 deals with the interview, followed by Section 5.3 on participant observation. In Section 5.4, we discuss the use of experiments in exploring attitudes towards language use. In Section 5.5, we touch upon some other ‘methods’ that are used in exploring questions of LM and LS. These could be collectively described as biographic and/or autobiographic narratives.

INTERVIEWS

We have repeatedly noted that the question ‘Who speaks what language to whom, when, where and to what end/for which purpose?’ has become the guiding beacon for the exploration of language-use patterns in multilingual settings, through either written questionnaires or interviews. In fact, if an interview is used to pursue this question in the context of macro-oriented research, then such an interview is best described as an interviewer-administered questionnaire. In such an interview, the content of questions will be very similar to that discussed in relation to written questionnaires. However, more attention may need to be paid to the formulation of questions. For example, using more open-ended questions may be more suitable in such situations.

FOCUSING ON INTERGROUP DIFFERENCES IN LM AND LS

The existence of intergroup or interethnic differences in maintenance or shift rates is probably the most consistent finding emanating from the multitude of studies. Some linguistic minorities or ethnolinguistic groups maintain their language better than other groups, despite being exposed to the same environment. For example, Polish Australians maintain their language more and longer than Dutch Australians in the same Australian environment. Another observation regarding interethnic or intergroup differences in LM or LS rates relates to a specific group managing to maintain its language in very diverse settings. The Greek diaspora around the world is an example of this. Conversely, there are ethnolinguistic groups whose LM or LS rates are quite susceptible to the environment in which they find themselves. This could be said, for example, of Turkish immigrants in Australia in contrast to those living in the Netherlands or Germany. They seem to maintain their language better in the latter countries. It also applies to Basque people living in Basque territories of France and Spain. In the previous chapter we delved into some trends that seem to be common across heritage and territorial linguistic minorities. These related mainly to personal characteristics, i.e., sociodemographic factors, and to the contexts (domains) that promote LM or LS. Here, the focus is mainly on the group level in an attempt to shed light on these intergroup differences. In particular, we review a number of approaches that have attempted to identify factors influencing language behaviour at the group level. In most cases, the models developed are in fact taxonomies or typologies focusing on group factors. Some of these are described as ‘theories’ because their proposers contend that they are able not only to enhance our understanding of specific language-use patterns but also to predict whether a group is likely to maintain its language.

TAXONOMIES OF GROUP FACTORS INFLUENCING LM AND LS

Kloss’ clear-cut and ambivalent factors

One of the first approaches to understanding differences in the language behaviour and language-use patterns of ethnolinguistic groups was Kloss’ (1966) study of German-American language maintenance efforts during the late nineteenth and first half of the twentieth century. In this paper Kloss created a list of two types of factors that contribute to shaping the dynamics of LM or LS. He referred to them as clear-cut and ambivalent factors.

In this part, we move beyond description and discussion of the dynamics of LM and LS to focus on the efforts, actions and initiatives of individuals, families, groups and the communities themselves to maintain their minority or heritage language and pass it on to future generations. Part III has two chapters. Chapter 8 is mainly occupied with documenting the kinds of activities and initiatives that minority and heritage language communities themselves are known to undertake in pursuit of LM. Sections 8.1 to 8.5 describe these efforts around the main domains of the home or family (Section 8.1), the community-based school (Section 8.2), religion (Section 8.3), secular community groups (Section 8.4) and the media (Section 8.5). In Section 8.6 we present some examples of how the majority or dominant community assists in LM in relation to schooling and the media. Inevitably, there is some overlap between Chapter 8 and Chapter 6 (to some extent, also Chapter 7) as they all deal with investigating the central question – Who speaks what language to whom, where and to what end? – and describing and analysing patterns of language use in relation to the key domains mentioned above.

In Chapter 9 we tackle two questions pertinent to LM: Can LS be reversed? Here, we will draw upon Fishman's Graded Intergeneration Disruption Scale to seek a response. The other question – Should LS be reversed? – deals with a more controversial topic, at least for languages that will disappear or be lost, if they are not maintained. In the context of this book, this affects territorial minority languages rather than heritage languages in migrant settings.

KEEPING THE HERITAGE OR MINORITY LANGUAGE GOING IN THE FAMILY

Throughout this book and in line with the dominant opinions of LM scholars we have stressed the key role that the family plays in efforts to maintain and pass on the heritage or minority language to future generations. So what can or do families do to keep the heritage or minority language alive? In the next sections we discuss various strategies, initiatives and practices found in families living in situations where there are pressures to shift to the dominant language(s) of the nation.

In Part II we described various data collection methods and tools that have been used or developed to investigate language practices and patterns associated with our two main settings: migrant communities and territorial linguistic minorities. In Part III we move to a discussion of findings emanating from the wealth of studies that made use of these various methods to examine questions of LM and LS. Indeed there have been hundreds, if not thousands of studies around the world examining LM and LS in our two main settings. Given this wealth, it is not surprising that a multitude of factors and forces influencing the process of LM or LS has been unearthed. Here we will focus on those factors that have been studied most widely and on discussing common trends or patterns that have been observed across this wide variety of studies. Some factors influencing LM and LS concern individual characteristics (variables) such as age, gender, educational background, social class, race/ethnicity, religious affiliation or marital status. Other factors apply mainly to groups. These include but are not limited to, the numerical strength of a group, its settlement patterns and its linguistic or cultural similarity to the dominant or majority group. Further factors that influence LM and LS relate to the majority group and/or the polity within which the minority finds itself. They include, for example, the majority's attitude towards the minority, its language and culture and the existence of laws and policies that support or oppose linguistic diversity, bi- and multilingualism. In this chapter, our discussion of common trends and patterns centres around two focal points: in Section 6.1 our focus is on variables and characteristics of the speaker or language user; i.e., this concerns the ‘who’ in the central question of ‘Who speaks what language to whom and when?’ The other focal point relates to other elements of the key question – to whom and when – that refer to contexts of use or domains. These will be taken up in Section 6.2. In both sections we will try to establish whether a specific speaker characteristic or context can be clear-cut, i.e., clearly enhancing LM or clearly leading to LS.

THE STUDY OF LM AND LS AS A SUBFIELD OF LANGUAGE CONTACT STUDIES

In the introduction of this book we already hinted that the study of LM and LS is heavily intertwined with that of related phenomena such as multilingualism, language loss, language death, language endangerment and language revival and revitalisation. All these phenomena are, in one way or another, the outcome or the consequence of a plethora of language contact situations. Contact between languages and language varieties is not only omnipresent but also the norm rather than the exception in linguistic constellations around the globe.

Mobility of people – of individuals, families, groups and entire communities – is at the source of most language contact. Reasons for mobility are multiple, ranging from individual adventurous exploring and a desire or need for trade, education or employment beyond one's own territory to the results of expulsion or invasion. All these forms of mobility, voluntary or forced, short or long term, are most likely to involve contact between speakers of different languages or language varieties. Short-term language contact between individuals or groups is not likely to induce (major) changes in linguistic constellations or the language practices of the speakers. For example, many people now holiday in places where other languages are spoken. While some holidaymakers may be motivated to learn the local language, this is not likely to reshape or affect their linguistic practices drastically, once they return from their holiday. Conversely, long-term contact between people is much more likely to have a linguistic effect. Migration – voluntary as well as forced – colonisation and invasion are the types of mobility that have had the most impact on changes in linguistic constellations and language practices. In most cases, speakers of one language move (long term or permanently) into the territory of another speech community. Where colonisation or other types of ‘invasion’ are the trigger for the contact, the language or languages of the ‘intruders’ often not only upset the existing linguistic hierarchies and dynamics but may also lead to the extinction of some of the indigenous languages.

Changes in nation-states are also possible triggers for language contact. Some of these are in fact outcomes of invasions, so they are ultimately linked to mobility.

LINGUISTIC DEMOGRAPHY AND THE STUDY OF LM

In this first chapter dealing with data collection, we focus on linguistic demography. Linguistic demography is concerned with the study of numerical or statistical aspects of languages and their distribution across the world. Major sources of information include Census and large-scale surveys. A number of countries collect information on language use and language proficiency through their periodic Census surveys, usually done every five to ten years. The language data collected in this way have been of assistance to language scholars interested in the linguistic make-up of a country, state or region and in the dynamics of these linguistic landscapes, including major and minor shifts in the usage of particular languages and language varieties. In line with a purpose of Census surveys, they have also been used in the shaping or changing of language policies and planning initiatives. In this chapter we will look at some issues that surround the use of language data drawn from Census surveys in the study of LM and LS.

LANGUAGE DATA FROM CENSUS SURVEYS

Census surveys are periodical population surveys that collect information from individuals (typically over the age of five years) about a large set of variables, factors and behaviours. These surveys facilitate planning in a variety of fields such as transport, education, social services, health, housing, trade, business, the economy and even political representation. Countries with a linguistically diverse population increasingly include questions around language use and language proficiency. Reasons for such inclusion can range from a desire to document and monitor trends in the linguistic diversity of the national population to a need for the planning of language services (interpreting, translation and language learning) to formulating policies on language use and language rights. These questions, albeit that their formulation is often problematic, are regularly used by a variety of scholars, including those working on LM and LS, to assess the state and vitality of territorial minority, immigrant or heritage languages within a country or large polity. In some cases where the Census does not include language questions researchers may draw upon other variables (e.g., ethnic origin, birthplace) to gain some insight into language matters. Of course the latter way of gaining information on language matters is very fraught as neither ethnicity nor birthplace is an appropriate substitute or reliable indicator of linguistic background.

THE ESTABLISHMENT OF LM AND LS AS A SEPARATE FIELD OF ENQUIRY

Although tracing the exact origins of this field of enquiry is no doubt a fascinating enterprise, it is best left to forensic enthusiasts more knowledgeable than me in discovering the relevant historical data. The task of identifying the origins and pioneers is further complicated by the fact that this area of study is intricately intertwined with other areas such as the study of language contact, bi- and multilingualism, language death and language change. Trying to disentangle the origins of each of these is probably futile. However, these complexities and complications should not be an excuse for ignoring the contexts and phenomena that encouraged scholars to establish LM and LS as a significant field of enquiry linked to the overarching fields of language contact and multilingualism.

There is a considerable degree of consensus that LM and LS emerged as a field of enquiry in the mid-twentieth century and that it was initially linked to contexts of language contact arising from migration, especially from the ‘old’ world, Europe, to the ‘new’ world, North America and, later, predominantly Australia and New Zealand.

A key text in its establishment is undoubtedly the paper entitled ‘Language maintenance and language shift as a field of enquiry. A definition of the field and suggestions for its future development’, written by Joshua Fishman in 1964. In this article he states:

SPRACHINSELFORSCHUNG – STUDY OF LINGUISTIC ENCLAVES – AS A FORERUNNER OF THE FIELD

In Fishman's (1964) seminal text he identifies scholars such as Uriel Weinreich, Einar Haugen, Heinz Kloss and Charles Ferguson as key contributors to the establishment of the field, at least in the Anglophone world.