We now move on to look at SDA in educational contexts. This differs from SDA in naturalistic contexts in several ways. Most significantly, the learners are normally students enrolled in a course or an educational institution who want to, or are required to, learn a new dialect for a specific purpose – usually for progressing in the formal education system or for working in a particular profession.

Chapter 1 mentioned that there are two different types of educational contexts for SDA. One normally involves individual or small-group tutoring in dialect coaching or accent modification training. In both kinds of tutoring, the learners are generally adults who have chosen to be involved in SDA, and the emphasis is on phonological features of the D2. Teaching approaches for dialect coaching and accent modification training are discussed in Chapter 8.

The other, and much more common, type of educational context for SDA is in the classroom. In most cases, the learners are initially children who come to school speaking a regional or ethnic dialect, a colloquial variety, or a pidgin or creole as their D1 – what we refer to here as unstandardised dialects or vernaculars. These students need to acquire as a D2 the standard dialect that is used in the formal education system. Adolescents and young adults who speak unstandardised dialects continue with SDA in secondary and tertiary formal education. In this context the lexical and morphosyntactic features of the D2 needed for reading and formal writing are more important than phonological features.

As mentioned in Chapter 1, the popular consensus appears to be that SDA is easier or “less traumatic” than SLA (Escure 1997: 7). This is because of the small linguistic distance between the D1 and the D2 compared to that between the L1 and L2. However, some scholars have the opposite point of view – that SDA is more difficult than SLA – and for the same reason – i.e. the small distance between the D1 and D2. For example, Haugen (1964: 125) wrote: “Bidialectalism may actually be harder to acquire than bilingualism. All scholars have agreed that it is harder to keep two similar languages apart than two very different ones.” And Wolfram and Schilling-Estes (1998: 287) noted: “In some ways, it may be easier to work with language systems that are drastically different, since the temptation to merge overlapping structures and ignore relatively minor differences is not as great.”

Rather than get into an argument about whether SDA is more difficult than SLA overall, this chapter presents some of the particular aspects of SDA that are especially difficult compared to SLA. Of course, because of the age effects described in Chapter 4, these difficulties do not pertain to children 7 years and younger, and are relevant mainly to older children, adolescents and adults. While I still restrict the discussion to naturalistic contexts, some of the observations are also relevant to educational contexts, as described in later chapters.

This chapter is concerned with the educational approaches and programs that have been used specifically for the purpose of assisting the acquisition or use of another dialect. The characteristics of each of these are described, as well as any evaluative research that is available.

The chapter starts off with depicting two types of dialect teaching that normally do not occur in the classroom: dialect coaching and accent modification. As mentioned in Chapter 7, students in such programs are usually adults who are interested in acquiring the phonological features of another dialect for professional purposes. The remaining sections deal with classroom approaches for younger students who are required to learn the standard dialect in the formal education system. Some historical background is first presented about approaches and methodologies used in the 1960s and 1970s. Then, three types of current approaches are described: instrumental, accommodation and awareness.

Dialect coaching and accent modification

Dialect coaching

In Chapter 3, I talked a bit about actors' imitations of dialects, and mentioned that most actors who perform in dialects other than their own have used a dialect coach. There are dozens of websites on the internet for dialect coaches (or voice coaches) – often with lists of films, plays and other productions that they have worked on. Two well-known examples are Victoria Mielewska in Australia and Paul Meier in the USA. There is also an association of dialect coaches and related professionals: the Voice and Speech Trainers Association (VASTA), which has a newsletter, accessible over the internet (www.vasta.org).

This chapter looks at some of the phenomena involved in acquiring a new dialect and ways of examining them, based on the studies described inChapter 2 and other research as well. First, I give an explanation for the ways that dialect acquirers are perceived by both D2 and D1 speakers. Then, after talking about imitation, I describe two types of SDA, replacive and additive. In the section that follows I discuss various approaches used to examine SDA.

Perceptions of dialect acquirers

In Chapter 1, I mentioned my own experience of always being recognised as an American in Australia but being told I have a British accent when I'm in America. Shockey (1984: 87) also commented that her American subjects in England “sounded like Americans to British ears and like British people to Americans”. Chambers (1992: 695n) made similar observations:

The first part of this chapter explains some of the reasons for this “double foreignness”.

Factors contributing to non-nativeness in the D2

First, let's look at why dialect acquirers continue to be perceived as “non-native”.

In addition to examining the extent of D2 learning, most studies of SDA also aim to determine some of the factors that lead to learners differing in their overall attainment, and to some D2 features being acquired more successfully than others. In fact, determining these factors is often the main focus of some SDA studies.

This chapter introduces some of the factors relating to both learners and particular linguistic features, and then concentrates on age of acquisition and two related linguistic factors, linguistic level and complexity of rules. These factors intersect to affect the degree of ultimate attainment. The chapter concludes with a discussion of some possible reasons for these effects.

Individual and linguistic factors

Why do some subjects in the studies described inChapter 2 attain native-like usage of D2 variants while others do not use any D2 variants at all? Various non-linguistic factors have been examined to account for such differences in individuals' performance. Some of these factors are connected with intrinsic characteristics of the learners, such as their age. Others have to do with the learning context – age of arrival in the new location or when D2 acquisition began, often called age of acquisition, and length of residence in that location. Another set of factors is associated with the complex notions of motivation to acquire the D2 and attitudes towards it and its speakers.

This book is about learning a new dialect, and how it is different from learning a new language. In this introductory chapter, I start by describing the contexts where this kind of learning occurs and some of the questions the book aims to answer. As I have tried to make the content accessible not just to linguists, I have also included some basic information for readers without a strong background in linguistics. This is about differentiating dialects, describing speech sounds and studying variation in language. The final section presents a brief outline of the book.

Second language acquisition and second dialect acquisition

The study of second language acquisition (often abbreviated as SLA) examines how people who already speak a first language (L1) subsequently acquire a second or additional language (L2). This book focuses on a special type of SLA – when the relationship between the L1 and the L2 is close enough for them to be considered by their speakers to be varieties of the same language, or different dialects, rather than different languages. In this situation, the term “second dialect acquisition” (SDA) can be used. The study of SDA examines how people who already speak one dialect (D1) acquire a different dialect (D2) of what they or their community perceive to be the same language.

Of course it is notoriously difficult to distinguish language and dialect.

Chapter 7 described some of the negative outcomes of both mainstream and ESL approaches in the formal education of speakers of unstandardised dialects. Chapter 8 showed little evidence of success in actual D2 acquisition in accent modification programs and the failure of SESD classroom approaches in the 1970s. However, much more positive outcomes were described for dialect coaching for performance and for instrumental, accommodation and awareness approaches in the classroom. This chapter explores some of the reasons for these results, both negative and positive. In doing so, it refers back to some of the factors that affect naturalistic SDA, as presented in Chapters 3 to 6. The chapter moves on to a discussion of further steps that can be taken to promote SDA in classroom contexts, followed by a short conclusion with some suggestions for further research.

Unsuccessful outcomes

Accent modification

As we saw in Sections 3.2 and 7.1 above, while actors can learn to produce other dialects for performance, it is much more difficult to acquire another dialect to the extent that it can be maintained in everyday use. The difference is that production for performance requires only explicit knowledge or conscious awareness of some of the features of the D2, and the ability to imitate them accurately. Actual acquisition of the D2, however, requires not only this explicit knowledge but also implicit competence, or the set of computational procedures needed to produce the D2 features automatically and consistently, as well as accurately.

This chapter presents summaries of the methodology and results of studies relevant to SDA in naturalistic contexts. The primary aim is to describe how much of the D2 was acquired by individuals and groups of subjects. As in studies of SLA, the particular degree of acquisition is referred to here as the level of attainment. The first section of the chapter deals with the question of how attainment in SDA can be measured. The following two sections outline seventeen studies of SDA, eight done on dialects of English and nine on dialects of other languages. After a summary of the overall results, there are brief descriptions of three other studies that focus on shift from the D1 rather than attainment in the D2. The chapter concludes with a discussion of some methodological issues.

Measuring attainment in SDA

How can we measure attainment in SDA? Because of the linguistic similarities that usually exist between the D1 and D2, many of the methods used in SLA studies are usually not suitable – for example, grammaticality judgements and measures of fluency and communicative success. This is because D1 speakers normally know the basic grammar of the D2, since it is virtually the same as that of the D1. Furthermore, because of mutual intelligibility, they can usually communicate fluently by using the D1 with D2 speakers. That leaves us with examining observable behaviour – i.e. performance data. More specifically, we need to concentrate on linguistic features that distinguish the D2 from the D1.

As mentioned in the preceding chapter, in order to account for differential results in D2 attainment, studies of SDA have examined both individual factors relating to learners, and linguistic factors relating to variables. This chapter looks at additional factors in both of these categories.

Additional individual factors

In addition to age of acquisition (AoA), individual factors studied in SDA include length of residence, social identity, gender, degree of social interaction with D2 speakers, motivation and attitudes, and occupation. These factors and their relationships are described here.

Length of residence (LoR)

One of the most commonly examined factors in studies of naturalistic SDA is length of residence (LoR) – i.e. how long the acquirer has lived in the D2 region. The LoRs, or ranges of LoR among subjects, for each of the seventeen studies examined inChapter 2 are given in Table 4.1 in the preceding chapter. A quick glance at the first three studies seems to indicate that LoR is not a consistently significant factor, since the children in Tagliamonte and Molfenter's (2007) study reached native-like usage after six years but those in Trudgill's (1981) did so in less than one year, and some in Berthele's (2002) did so in two years whereas others did not. The rest of the figures for LoR are not very helpful because of the wide ranges. But we do find many examples of subjects with long LoRs yet very low percentages of use of D2 variants.